New Airport Report

Report of the Commission of Inquiry on
the New Airport

CONTENTS

Pages
Abbreviations i -
viii
Brief Summary A - C
Detailed Summary I - XI
CHAPTER 1 INTRODUCTION AND
ACKNOWLEDGEMENT
1 - 5

CHAPTER 2 TERMS OF REFERENCE 6 - 7

CHAPTER 3 METHODOLOGY,
CRITERIA AND
TREATMENT OF
EVIDENCE
8 - 30
Section 1 : Methodology
Section 2 : Concerned
Parties' Role,
Responsibility and
Liability
Section 3 : Resolution of
Issues
Section 4 : Criteria
Section 5 : Standard of
Proof and Treatment of
Evidence

CHAPTER 4 THE HEARING AND
CONFIDENTIALITY OF
DOCUMENTS
31 -
41
Section 1 : The Hearing
Section 2 : Confidentiality
of Documents
Section 3 : Hearing of
Witnesses by Group

CHAPTER 5 THE ROLES AND
DUTIES OF KEY
PARTIES AND
COORDINATION
42 -
79
Section 1 : Introduction
Section 2 : The Roles and
Duties of the Key Parties
(a) AA

(b) ADSCOM
(c) NAPCO
Section 3 :
Communication Channels
Section 4 : Adequacy of
Communication and
Coordination

CHAPTER 6 AOD - DECISION,
PREPARATION AND
RESPONSIBILITY
80 -
118
Section 1 : Planning for
AOR
Section 2 : Decision on
AOD
Section 3 : Preparation for
AOD
Section 4 : Responsibility

CHAPTER 7 AIRPORT OPENING -
THE PROBLEMS AND
THEIR DEBILITATING
EFFECT
119 -
127

CHAPTER 8 THE STANDARDS OF
CLASSIFICATION --
FROM TEETHING TO
MAJ OR PROBLEMS
128 -
151

CHAPTER 9 TEETHING AND
MINOR PROBLEMS
AND REMEDIAL
MEASURES
152 -
201

CHAPTER 10 MAJ OR PROBLEM -
THE OPERATION OF
FIDS
202 -
228
Section 1 :Importance of
FIDS in AOR
Section 2 :FIDS Operation
on AOD
Section 3 :Remedial
Measures and the Present
Status

CHAPTER 11 MAJ OR PROBLEM -
CARGO HANDLING
229 -
253
Section 1 : Importance of
Cargo Handling in AOR
Section 2 : Cargo
Handling on AOD
Section 3 : Remedial
Measures and the Present

Status

CHAPTER 12 THE OTHER MAJ OR
PROBLEM AND
MODERATE
PROBLEMS
254 -
313
Section 1 : The Other
Major Problem: Baggage
Handling
Section 2 : Moderate
Problems

CHAPTER 13 RESPONSIBILITY -
FIDS
314 -
367

Section 1 : History of
Development, Installation,
Testing and
Commissioning of FIDS -
Delays and Problems
Section 2 : What was
Wrong with FIDS?
Section 3 : Repairs after
AOD
Section 4 : Causes and
Responsibility

CARGO HANDLING
368 -
424
Section 1 : The
Development of the Cargo
Terminal Operators at the
New Airport
Section 2 : Causes for the
Problems on AOD - AAT
Section 3 : Causes for the
Problems on AOD -
HACTL
(a)The Alleged Causes
(b)The Opinions of
HACTL's Experts
(c)Dust
(d)The Main Causes
Section 5 : HACTL's Best
Endeavours Basis
Section 6 : HACTL's
Attitude in the Inquiry

CHAPTER 15 RESPONSIBILITY - THE
OTHER MAJ OR
PROBLEM AND
MODERATE
425 -
472
PROBLEMS
Section 1 : Major
Problem: Baggage
Handling
Section 2 : Moderate
Problems

TEETHING AND
MINOR PROBLEMS
473 -
500
Section 1 : Teething and
Minor Problems

CHAPTER 17 RESPONSIBILITY OF
THE AIRPORT
AUTHORITY
501 -
551
Section 1 : AA's
Obligations under the
Airport Authority
Ordinance
Section 2 : Coordination
and Communication
Section 3 : Overview of
What Went Wrong
Section 4 : Misstatements
and Responsibility for
Them
(a)FIDS
(b)ACS

(a)W3 Townsend
(b)W48 Lam
(c)W43 Oakervee
(d)W44 Heed
(e)W45 Chatterjee
(f)The AA Board

CHAPTER 18 CONCLUSIONS 552 -
655
Section 1 : The Decision
to Open the Airport
Section 2 : Extent of
Readiness and the
Problems
Section 3 : Causes of the
Problems and
Responsibility
Section 4 : Adequacy of

****************
Appendices
Communication and
Coordination
of AA
Section 6 : The Present
Situation
Section 7 : Could the
Chaos and Confusion have
been Avoided?
Section 8 : Lessons
Learned
Chapter(s) in
Which referred
Pages

Appendix I Parties in the Inquiry 1 656 - 658
Appendix II Legal representatives of
parties in the Inquiry
1 659 - 663
Appendix III List of experts in the
Inquiry
1 and 3 664 - 669
Appendix IV List of witnesses in the
Inquiry
4 670 - 684
Appendix V Members of the Board
of the Airport Authority
(J une 1998)
5 685 - 686
Appendix VI Channel of documentary
communication to
ADSCOM
5 687
Appendix VII Organization chart of
the Airport Authority
(31 J uly 1998)
5 688
Appendix VIII Diagrammatic
presentation of air-
conditioning problems
on AOD
8 and 12 689
Appendix IX Pictorial diagram
showing the three kinds
of flight times
10 690
Appendix X Diagram showing use of
FIDS
10 and 13 691
Appendix XI Flight information
experiences of various
FDDS users and MTRC
10 692 -695
Appendix XII Diagram showing CHS
of HACTL with its five
levels
11 696
Appendix XIII Pictures showing the
problem bag area in the
Baggage Hall
12 697 - 699
Appendix XIV Diagram showing the
inter-link between FIDS
and other systems
13 700
Appendix XV Diagram showing the
allocation of FIDS
related components of
airport operations into
problem areas
13 701
Appendix XVI Brief daily summary of
some airport operational
statistics
18 702

Abbreviations

A

A buses Airbuses
AA Airport Authority
AAT Asia Airfreight Terminal Company Limited
ACC Apron Control Centre
ACCS Air Cargo Clearance System
ACP Airport Core Programme
ACS Access Control System
ADSCOM Airport Development Steering Committee
AE Airport Express
AEC Airport Emergency Centre
AEH AEH J oint Venture
AFC Airport Fire Contingent
AFFC Airport Freight Forwarding Centre
AIDB Aeronautical Information Database
AMD Airport Management Division, AA
AMFSRC Airport Main Fire Station Rescue Control
Ansett Ansett Australia O/B Ansett International Limited
AOCC Airport Operations Control Centre
AOD airport opening day
AODB Airport Operational Database
AOR airport operational readiness
APA Aircraft Parking Aid
APM Automated People Mover
APV Apron Passenger Vehicle
AR Airport Railway
ARA Airport Restricted Area
ATA actual time of arrival
ATC Air Traffic Control
ATCC Air Traffic Control Centre
ATD actual time of departure
Atlas Air Atlas Air, Inc.
ATV automatic transfer vehicle
AVAS Audio and Visual Advisory System

i
AVSECO Aviation Security Company Limited

B

BAA British Airport Authority
BCJ British-Chinese-J apanese J oint Venture
BCR Baggage Control Room
Bechtel International Bechtel Company Ltd.
BGR boarding gate reader
BHO baggage handling operator
BHS Baggage Handling System
BMS Building Management System
BP boarding pass
BRH Baggage Reclaim Hall
BSI Building Systems Integration
BSM Baggage Source Message
BSS Box Storage System
Bukaka Ramp PT. Bukaka Teknik Utama-RAMP J oint Venture

C

C & ED Customs and Excise Department
CAD Civil Aviation Department
CAL China Airlines Ltd.
Canadian Airlines Canadian Airlines International Limited
Carrier Carrier Hong Kong Limited
CAS Common Antenna System
Cathay Pacific Cathay Pacific Airways Limited
CCTV closed circuit television
CDG Control Systems Development Group
CEM Controlled Electronic Management Systems Limited
CEO Chief Executive Officer
Cevasa Imagen Cevasa Imagen S.A.
the Chief
Secretary
W36 Mrs Anson CHAN, the Chief Secretary for
Administration and Chairman of ADSCOM
CHO cargo handling operator
CHS Cargo Handling System
City U City University

ii
CLK Chek Lap Kok
CLP China Light & Power Company Limited
CMT China Motion United Telecom Limited
CNIM Constructions Industrielles De La Mediterranee SA
Commission Commission of Inquiry on the New Airport
COSAC Community System for Air Cargo
CPCS Cathay Pacific Catering Services
CPM Consultant Project Manager
CPU Central Processing Unit
Crisplant Crisplant Limited
CSE CSE International Ltd.
CSS Container Storage System
CTO cargo terminal operator
CUTE Common User Terminal Equipment

D

DAC Distributed Access Controller
DAN Distributed Antenna Network
DCA Director of Civil Aviation
DCS Departure Control System
Demag Mannesmann Dematic AG Systeme
DHL DHL International (Hong Kong) Ltd.
Dragon Air Hong Kong Dragon Airlines Limited

E

E buses external buses
EDS Electronic Data Systems Limited
EEV EEV Limited
EMC Equipment Motion Control
ENG Engineering Department, HACTL
ESRA Enhanced Security Restricted Area
ETA estimated time of arrival
ETD estimated time of departure
ETV elevating transfer vehicle
ExCo Executive Council

iii
F

FAT factory acceptance test
FDDFS Flight Display Data Feed Services
FDDS Flight Data Display System
Ferranti Ferranti Air Systems Limited
FIDS Flight Information Display System
FIMI FIMI-Philips S.r.l.
FS Financial Secretary
FSA Support Agreement Relating to the Financing,
Construction and Operation of the Airport
FSCC Fire Services Communication Centre
FSD Fire Services Department

G

GBMS General Building Management System
GCPA General Coverage Public Address
GEC G.E.C. (Hong Kong) Ltd.
GPY Gammon Paul Y J oint Venture
Grant Grant Ameristone Limited
GSM Global System for Mobile Telecommunications
GTC Ground Transportation Centre
Guardforce Guardforce Limited

H

HACTL Hong Kong Air Cargo Terminals Limited
HAECO Hong Kong Aircraft Engineering Company Limited
HAS Hong Kong Airport Services Ltd.
HATS Hong Kong Air Terminal Services Ltd.
Hepburn Hepburn Systems Limited
HKASP Hong Kong Aviation Security Programme
HKIA-ASP Hong Kong International Airport-Airport Security
Programme
HKSAR Hong Kong Special Admnistrative Region
HKT Hong Kong Telecom CSL Limited

iv
hrs hours
Hughes Hughes Asia Pacific (Hong Kong) Limited
Hutchison Hutchison Telecommunications (Hong Kong) Ltd.

I

IATA International Air Transport Association
ICL International Computers Limited
ID Identification
IS Interchange Server
ISD Information Services Department, HACTL
IT Information Technology

J

J AL J apan Airlines Company Limited
J ATS J ardine Air Terminal Services Ltd.

K

KLM KLM Royal Dutch Airlines

L

LCD liquid crystal display
LCS Logistic Control System
Level(s) Level(s) of the passenger terminal building
Level(s) Level(s) of the baggage security screening
level(s) levels of the Cargo Handling System in
SuperTerminal 1
Los Los Airport Cleaning Services Limited

M

MAZ manual all zone
MCC motor control centre
MFT Multi-Functional Terminals

v
MHI Mitsubishi Heavy Industries, Ltd.
MHS Material Handling System
min minute(s)
MMI Man Machine Interface
Mott The Mott Consortium
MTR Mass Transit Railway
MTRC Mass Transit Railway Corporation
Murata Murata Machinery (HK) Ltd.

N

NAPCO New Airport Projects Co-ordination Office
Nishimatsu Nishimatsu Construction Co., Ltd.
NWT New World Telephone Limited

O

OCPM Operations Computer Project Manager
OFTA Office of Telecommunications Authority
Ogden Ogden Aviation (Hong Kong) Limited
OOG out-of-gauge
OP occupation permit
OPS Operations Department, HACTL
OPT Operations Project Team
Oracle Systems Oracle Systems Hong Kong Ltd.

P

PA Public Address System
PAA Provisional Airport Authority
PABX Private Automatic Branch Exchange
para(s) paragraph(s)
PCHC Perishable Cargo Handling Centre
PD Project Division, AA
Pearl Pearl Delta WMI Limited
PIN personal identification number
PLC Programmable Logic Controller
Preston The Preston Group Pty Ltd.

vi
PRs problem reports
PTB Passenger Terminal Building
PTS particular technical specifications

R

RAM Random Access Memory
RASTI Rapid Assessment of Speech Transmission Index
Reliance Reliance Airport Cleaning Services Limited
RFI radio frequency interference
RHOs ramp handling operators
RMS Resources Management System
Rotary Rotary (International) Limited

S

S buses shuttle buses
SAC Sort Allocation Computer
Safegate Safegate International AB
SAR Special Administrative Region
SAS Stand Allocation System
SAT site acceptance test
SC Senior Counsel
SCC Scheduling Committee Computer
Securair Securair Limited
SES Secretary for Economic Services
SESL Swire Engineering Services Ltd.
Siemens Siemens AG
SigNET SigNET (AC) Limited
SITA Societe Internationale de Telecommunications
Aeronautiques
SmarTone SmarTone Mobile Communications Limited
SRF Shipment Release Form
SSS system segment specification
ST1 SuperTerminal 1
STA scheduled time of arrival
STD scheduled time of departure

vii
T

TCO Trucks Control Office
Thai Airways Thai Airways International Public Company Limited
TMR Trunk Mobile Radio
TMS Terminal Management System
TODC Time of Day Clock
TOP temporary occupation permit
TU Timed Updates
TUSC Twenty-Foot ULD Storage Centre

U

UK United Kingdom
ULD unit load device
UPS uninterrupted power supply
US United States
USA United States of America

V

Vanderlande Vanderlande Industries Hong Kong Ltd.
VHF Very High Frequency
Virgin Virgin Atlantic Airways Limited
VIS Vehicle Information System
VRS Voice Routing System

W

name of witness
followed by
(with)
Witness who gave evidence together with one or more
witnesses
W (followed by a
number)
Witness number as called at the public hearing
WB Works Bureau
WDUM WUF Data Update Manager a background computer
process which defines the flight information sent to
the man machine interface of the Flight Information

viii
Display System
WHC Western Harbour Crossing

Y

YDS YDS Engineering Ltd.
Youngs Young Engineering Company Limited

ix
Brief Summary

1. The decision to open the new airport on 6 J uly 1998 for operation
was made by the Airport Development Steering Committee
(ADSCOM) in J anuary 1998, after carefully taking into
consideration the state of the construction and the development works
for the systems and facilities in the Passenger Terminal Building
(PTB) and in the major franchisees premises and the assurances of
Airport Authority (AA) since about mid-1997 that the new airport
would be ready for operation in April 1998. The main determinant
for a date later than AAs suggested target of April 1998 was
ADSCOMs wish that the new airport would have the full
complementary ground transportation provided by the Airport
Railway which would only be ready in late J une 1998. The decision
was approved by the Chief Executive in Council, accepting
ADSCOMs views without comment. No political or ulterior
consideration was involved in the decision making.

2. The two main culprits for the chaos on airport opening day (AOD)
were the deficiency of the Flight Information Display System
(FIDS) and the breakdown of the Cargo Handling System (CHS)
of the Hong Kong Air Cargo Terminals Ltd (HACTL) which built
SuperTerminal 1 (ST1).

3. Due to FIDS difficulties, incorrect and incomplete flight information
was provided to all airport users, passengers and airport operators
alike. All had difficulty in knowing when a flight would arrive and
where it was going to park. This affected and delayed the
operations of the ramp handling operators (RHOs) in serving
aircraft and passengers and in unloading baggage and cargo.
Passengers and airlines did not know which departure gates were
assigned to the flights, especially when such gates were subjected to
many changes. Planes were late in both arrival and departure.
Passengers were delayed by the flight movements and also by the late
arrival of baggage. The means of communication such as the trunk
mobile radios and mobile phones which could be and were relied on
A
by airport operators to obtain the necessary flight information were
overloaded. Only about one third of the public telephones were
operational. The airbridges linking aircraft and PTB were not
always working, and the doors that were supposed to be operated
through the access control system from the airbridges to PTB
occasionally malfunctioned, causing further delay to both aircraft and
passenger movements. A full apron resulted as early as about noon
and lasted till about 5 pm, and another one was experienced between
8 and 11 pm. Aircraft had to wait to be provided with a parking
stand. Passengers were greatly inconvenienced and anxious while
service providers were sweating to cope. All kinds of operators of
the new airport were generally unfamiliar with the environment and
experienced difficulties in the operation of FIDS, and despite their
tremendous efforts, the chaos could not be avoided.

4. HACTLs CHS was operating slowly and inefficiently. The
slowness compelled operators to use the manual mode, which could
not cope with the workload that the automatic mode would.
Operators were not familiar with the manual mode either. The
unpreparedness of the operation of CHS was precipitated by the long
delays experienced in the construction of ST1, which also cascaded
down to delaying the installation, testing and commissioning of CHS
equipment and systems as well as the training and familiarisation of
HACTLs staff. There was a breakdown of the hand-over procedure
that had been agreed between HACTL and RHOs, resulting in a large
backlog of inbound cargo being left on the ramp interfacing with
HACTLs premises. Not only the customers of HACTL suffered,
but Hong Kongs airfreight-forwarding trade also sustained severe
losses.

5. A number of other problems which can be considered to be teething
and minor would by themselves have only caused minor
inconvenience to passengers, freight forwarders and other airport
users had FIDS and ST1 been running smoothly. However, the
effects of each of the small problems were enhanced, snowballed and
spiralled when they interacted with each other and the trouble with
FIDS and CHS.

B
6. AA failed in its duty imposed by the Airport Authority Ordinance that
in operating the new airport, it shall have regard to the efficient
movement of air passengers, air cargo and aircraft. For the chaos in
PTB, AA must therefore be primarily responsible. HACTL is
primarily responsible for not being able to provide a cargo handling
terminal ready with 75% of its full capacity that it had assured AA
and Government. There were progressive delays in the construction
and commissioning works in both PTB and ST1, and yet AAs top
management did not pay sufficient heed to the risks that various
systems were barely ready for operation on AOD. There was no
overall risk assessment but only sketchy contingency plans. AAs
top management and HACTL were over confident with what they
could achieve by AOD, for which not only they but also Hong Kong
paid a dear price.

7. Had a deferment been suggested by either AA or HACTL, the chaos
could have been avoided by a postponement of AOD by about two
months. Unfortunately, no one ever made any such suggestion, and
everyone was working diligently but blindly towards the common
goal of AOD. FIDS was operating reasonably efficiently about a
week after AOD, and the other less serious problems also subsided
within a short time. However, it had taken HACTL about six weeks
to recover. Although there may still be glitches and hitches, the new
airport is now running as a pride of Hong Kong.
C
Detailed Summary

DECISION TO OPEN THE NEW AIRPORT

1. The decision to open the new airport on 6 J uly 1998 for operation
was made by the Airport Development Steering Committee
(ADSCOM) in J anuary 1998 after carefully taking into
consideration the state of the construction and the development
works for the systems and facilities in the Passenger Terminal
Building (PTB) and in the major franchisees premises and the
assurances of Airport Authority (AA) since about mid-1997 that
the new airport would be ready for operation in April 1998.

2. AA was established as a statutory corporation by the Airport
Authority Ordinance (the Ordinance) that came into force on 1
December 1995. Under the Ordinance, AA has the functions and
duties to provide, operate, develop and maintain an airport for civil
aviation. While AA shall conduct its business according to prudent
commercial principles, the Ordinance provides that it shall have
regard to safety, security, economy and operational efficiency and
the safe and efficient movement of aircraft, air passengers and air
cargo. AA had to examine and evaluate when the new airport and
all the structures, facilities and systems that were required for
airport operation would be ready for opening the new airport for
operation, in other words, it had to examine and evaluate airport
operational readiness (AOR).

3. Prior to J anuary 1998, the target date for the opening of the new
airport had always been scheduled for April 1998, and the works
regarding the construction of buildings, the installation of facilities
and the commissioning of various systems for the operation of the
new airport had been awarded by AA or its predecessor, the
Provisional Airport Authority, to contractors with completion dates
corresponding to or compatible with April 1998. Similarly, those
works regarding the premises of AAs franchisees and business
partners whose services were required for the operation of the new
I
airport would also be completed at such time as to meet the April
1998 target.

4. Under AAs franchise agreements with its franchisees, AA was
obliged to give a three-month advance notice to the franchisees of
the date of opening of the new airport. ADSCOM was mindful of
the importance of fixing an airport opening date well in advance so
that the public as well as all concerned parties would know this date
for their own planning purposes. It was therefore necessary for
ADSCOM to take a decision on a firm airport opening date at least
three months ahead of April 1998.

5. As early as October 1995, ADSCOM considered that since the
smooth opening of the new airport is essential to Hong Kong,
ADSCOM was best placed to be the overall monitor of AOR.
ADSCOMs executive arm is the New Airport Projects
Co-ordination Office (NAPCO). NAPCOs primary duty was to
co-ordinate between Government departments which were
responsible for the development of seven of the 10 Airport Core
Programme (ACP) projects and the three bodies that were
responsible for the development of the remaining three ACP
projects, ie, AA in respect of the new airport, Mass Transit Railway
Corporation (MTRC) in respect of the Airport Railway (AR)
and Western Harbour Tunnel Co Ltd in respect of the West Harbour
Crossing. ADSCOM also directed NAPCO as its executive arm to
monitor the progress of AOR.

6. Towards the end of 1997, ADSCOM asked AA to recommend a date
for airport opening. NAPCO assisted ADSCOM by critically
examining the progress of AOR critical items, and ADSCOM put
details of matters of concern to AA to obtain its comments.
ADSCOM was particularly concerned with three matters: (a) when
AR would be ready for operation to provide ground transportation
to complement the operation of the new airport; (b) the delays in the
completion of the construction and system works in PTB; and (c)
the slippages in the construction of SuperTerminal 1 (ST1) to be
operated by Hong Kong Air Cargo Terminals Ltd (HACTL), the
major cargo handling operator at the new airport. AA advised that
II
PTB would be ready to open for operation in early April 1998.
HACTL had given an assurance that it would be ready with 50% of
its throughput capacity (which was the required amount in April
1998) by early April 1998. However, considering the delay in
ST1s construction works, AA was confident that the new airport
would be ready at the end of April 1998, which date it
recommended to ADSCOM. On the other hand, MTRC was
adamant that AR would only be ready by 21 J une 1998, the
contractual date of completion, with little hope to advance the date.
ADSCOM considered that it was essential for the new airport, as a
world-class airport, to be complemented by an efficient ground
mass transportation system, as opposed to ground transportation
consisting of makeshift arrangements. It decided that 1 J uly 1998
to be the ceremonial opening day and 6 J uly 1998 to be the
operational opening date (AOD). The ceremonial opening day
was subsequently altered to 2 J uly 1998. 6 J uly 1998 was chosen
because it was a Monday, when air traffic would be lighter than
other days of the week and road traffic would be lighter the night
before. The lighter air traffic would hopefully reduce the duties of
operators of the new airport and the light ground traffic. This
would facilitate the execution of the enormous relocation exercise to
move the equipment, facilities and staff from the Kai Tak airport to
the new airport on the Sunday night before AOD. The two
months between the original target of April 1998 and AOD would
also provide a comfortable float for the completion of necessary
AOR works in PTB, and HACTL in particular. HACTL was
happy with the added time, and subsequently gave assurances to AA
and Government that it would be ready on AOD with 75% of its
cargo handling throughput capacity, instead of the 50% throughput
by April 1998. On the evidence, the Commissioners conclude that
the decision on AOD was proper and wise.

7. The decision was approved by the Chief Executive in Council,
accepting ADSCOMs views without comment. There is no
evidence before the Commission to indicate that the decision was
made with any political or ulterior consideration.

III
PROBLEMS ON AND SINCE AOD

8. On AOD, numerous problems were encountered, and confusion and
chaos ensued. The problems affected man, cargo and machine:
there was no efficient movement of air passengers, air cargo or
aircraft. Even after AOD, problems continued to occur. The
Commission has investigated all the problems that it has been able
to identify and made findings as to their causes and where
responsibility lies. The problems, which were classified as
teething or minor, moderate and major, in accordance with their
seriousness and scope, are set out below. The findings of the
Commission as to the causes of the problems and the responsibility
for them can be found in the appropriate chapters of this report,
while Chapter 18 contains a summarised account of all the
problems.

Teething or Minor Problems:

[1] Mobile phone service not satisfactory
[2] Trunk Mobile Radio (TMR) service not satisfactory
[3] Public telephones not working
[4] Escalators breaking down repeatedly
[5] Insufficient or ineffective signage
[6] Slippery and reflective floor
[7] Problems with cleanliness and refuse collection
[8] Automated People Mover (APM) stoppages
[9] Airport Express ticketing machine malfunctioning
[10] Airport Express delays
[11] Late arrival of tarmac buses
[12] Aircraft parking confusion
[13] Insufficient ramp handling services
[14] Airbridges malfunctioning
[15] No tap water in toilet rooms and tenant areas
[16] No flushing water in toilets
[17] Urinal flushing problems
[18] Toilets too small
[19] Insufficient water, electricity and staff at restaurants
[20] Rats found in the new airport
IV
[21] Emergency services failing to attend to a worker nearly
falling into a manhole while working in PTB on 12 August
1998
[22] Traffic accident on 28 August 1998 involving a fire engine,
resulting in five firemen being injured
[23] A maintenance worker of Hong Kong Aircraft Engineering
Company Limited slipped on the stairs inside the cabin of a
Cathay Pacific Airways Limited (Cathay Pacific) aircraft
on 3 September 1998
[24] A power cut occurring on 8 September 1998, trapping
passengers in lifts and on the APM as well as delaying two
flights
[25] Missed approach by China Eastern Airlines flight MU503
on 1 October 1998

Moderate Problems:

[26] Delay in flight arrival and departure
[27] Malfunctioning of the Access Control System (ACS)
[28] Airside security risks
[29] Congestion of vehicular traffic and passenger traffic
[30] Insufficient air-conditioning in PTB
[31] Public Address System malfunctioning
[32] Insufficient staff canteens
[33] Radio frequency interference on air traffic control
frequency
[34] Aircraft Parking Aid malfunctioning: a Cathay Pacific
aircraft was damaged when hitting a passenger jetway
during parking on 15 J uly 1998
[35] An arriving passenger suffering from heart attack not being
sent to hospital expeditiously on 11 August 1998
[36] Fire engines driving on the tarmac crossed the path of an
arriving aircraft on 25 August 1998
[37] A Hong Kong Airport Services Ltd. tractor crashed into a
light goods vehicle, injuring five persons on 6 September
1998
[38] Tyre burst of United Arab Emirates cargo flight EK9881
and runway closures on 12 October 1998
V
[39] Power outage of ST1 due to the collapse of ceiling
suspended bus-bars on 15 October 1998

Major Problems:

[40] Flight Information Display System (FIDS)
malfunctioning
[41] Cargo Handling System (CHS) malfunctioning
[42] Baggage handling chaos

9. The two main culprits for the chaos on AOD were the deficiency of
FIDS and the breakdown of CHS of HACTL.

10. Due to FIDS difficulties, incorrect and incomplete flight
information was provided to all airport users, passengers and airport
operators alike. All had difficulty in knowing when a flight would
arrive and where it was going to park. This affected and delayed
the operations of the ramp handling operators (RHOs) in serving
aircraft and passengers and in unloading baggage and cargo.
Passengers and airlines did not know which departure gates were
assigned to the flights, especially when such gates were subjected to
many changes. Planes were late in both arrival and departure.
Passengers were delayed by the flight movements and also by the
late arrival of baggage. The means of communication such as
TMR and mobile phones which could be and were relied on by
airport operators as alternative means of obtaining the necessary
flight information were overloaded. Only about one third of the
public telephones were operational. The airbridges linking aircraft
and PTB were not always working, and the doors that were
supposed to be operated through ACS from the airbridges to PTB
occasionally malfunctioned, causing further delay to both aircraft
and passenger movements. A full apron resulted as early as about
noon and lasted till about 5 pm, and another one was experienced
between 8 and 11 pm. Aircraft had to wait to be provided with a
parking stand. Passengers were greatly inconvenienced and
anxious while service providers were sweating to cope. All kinds
of operators of the new airport were generally unfamiliar with the
environment and experienced difficulties in the operation of FIDS,
VI
and despite their tremendous efforts, the chaos could not be
avoided.

11. HACTLs CHS was operating slowly and inefficiently. The
slowness compelled operators to use the manual mode, which could
not cope with the workload that the automatic mode would.
Operators were not familiar with the manual mode either. The
unpreparedness of the operation of CHS was precipitated by the
long delays experienced in the construction of ST1, which also
cascaded down to delaying the installation, testing and
commissioning of CHS equipment and systems as well as the
training and familiarisation of HACTLs staff. There was a
breakdown of the hand-over procedure that had been agreed
between HACTL and RHOs, resulting in a large backlog of inbound
cargo being left on the ramp interfacing with HACTLs premises.
Not only the customers of HACTL suffered, but Hong Kongs
airfreight-forwarding trade also sustained severe losses.

12. The baggage handling problem is also a major one, because of its
impact on numerous passengers on AOD and the days following.
However, had FIDS been operating properly, RHOs resources and
energy could have been focussed on alleviating the baggage
problem. The other 39 problems are classified as teething or minor
and moderate. Some of them were isolated incidents, and did not
contribute to the chaos on AOD. The others would by themselves
have only caused minor inconvenience to passengers, freight
forwarders and other airport users had FIDS and ST1 been running
smoothly. However, the effects of each of the small problems were
enhanced, snowballed and spiralled when they interacted with each
other and the trouble caused by FIDS and CHS, to the extent that
nobody could have reasonably anticipated.

THE KEY RESPONSIBLE PARTIES

13. The main cause of the inefficient operation of FIDS was the
problems with the FIDS software, giving rise to slow response time
and causing great difficulty to the operators of FIDS on AOD. The
VII
lack of training on the part of the operators was also a major
contributing factor. The root cause of all these was the lack of
time. Due to delays in the development of FIDS, software
development time, testing time as well as training time had all been
compressed. There is insufficient evidence for the Commission to
decide whether it was AA or its contractor and subcontractors for
FIDS that should be responsible for the delays. For the software
problems, Electronic Data Systems Ltd., the subcontractor
providing the software, should be responsible. AA should mainly
be responsible for the failure to have its operators properly trained.

14. The delays in development of the software at the early stage was
mainly caused by the lack of co-ordination between AAs Project
Division (PD) led by W43 Mr Douglas Edwin Oakervee and
Airport Management Division (AMD) of which W44 Mr Chern
Heed is the director. There were also other instances of lack of
co-ordination, the most important of all was AMD and the
Information Technology (IT) Department headed by W45 Mr
Kironmoy Chatterjee failing to effectively arrange the staff of the
subcontractors for the software of FIDS (inclusive of the Terminal
Management System) and IT Department to timely attend the Apron
Control Centre to assist the operators there when help was most
needed, although such staff were present in other parts of PTB.
For this lack of co-ordination, W44 Heed and W45 Chatterjee
should be responsible. W45 Chatterjee also failed to advise AMD
of the risk of deferring the stress and load tests of FIDS.

15. Though there were risks involved in using FIDS on AOD, W44
Heed did not make any global contingency plan or have an overall
risk assessment. W44 Heed should be responsible on this account.

16. For the chaos in PTB on AOD and the days after, W44 Heed as
AMD Director must be primarily responsible. W3 Dr Henry
Duane Townsend, as the Chief Executive Officer (CEO), should
also be responsible. He is further found to be responsible for not
co-ordinating AMD and PD as the CEO.

17. The Commission also finds that the lack of co-ordination between
VIII
AMD and PD was probably caused by the personalities and
characters of those occupying the posts as directors of these two
divisions as well as the CEO. W3 Townsend did not give
sufficient priority to the operational requirements of AMD, and did
not give adequate support to W44 Heed.

18. AA as a whole failed in the duty imposed on it by the Ordinance to
have sufficiently regard to the efficient movement of air passengers,
air cargo and aircraft in operating the new airport. As to the chaos
in PTB, despite the responsibility of W44 Heed and W3 Townsend
for the same matter, the AA Board must bear the ultimate
responsibility, because the Ordinance has imposed the duty on it,
although it has power to and did delegate that duty to the CEO and
the AA management.

19. The evidence shows that W3 Townsend had made two
misstatements to ADSCOM, one on paper and the other orally.
The written misrepresentation was that FIDS was, as a whole,
98.7% reliable, and the oral one was that ACS had been successfully
tested. The Commissioners do not have sufficient evidence to
conclude that the misstatements were uttered with intent to mislead,
but ADSCOM was in fact misled. For these misstatements, W3
Townsend must be responsible. W45 Chatterjee is found to be
grossly negligent in not disabusing ADSCOM of the misstatement
on FIDS, but is not found responsible regarding ACS. W44
Heeds attitude that he would not bother if ADSCOM was misled
betrayed the trust that ADSCOM reposed in him, and exposed a
weakness in his integrity.

20. The top management of AA was over-confident in what they could
achieve, and were too busy to step aside to look at the risks involved.
As a result they assured ADSCOM that the new airport would be
ready.

21. The root cause of ST1s paralysis, similar to the cause of FIDS
deficiency, is also the lack of time. There were progressive delays
in the construction and commissioning works in ST1, compressing
the time for testing and for training of operators. There is
IX
insufficient evidence as to whether HACTL or its main contractor
and subcontractors were responsible for the delays.

22. The Commission finds that probably the main cause of the
breakdown of HACTLs CHS was the problems with the integration
between the Logistic Control System and the mechatronics of the
Container Storage System and the Box Storage System. The
operators were also not trained well enough to operate CHS in
manual mode.

23. HACTL is primarily responsible for not being able to provide a
cargo handling terminal ready with 75% of its full throughput
capacity that it had assured AA and Government. HACTL was
also over-confident in CHS that they had developed.

24. AA should also have monitored the readiness of HACTL. While
AA had professionals to check on the physical construction side of
the works carried on in ST1, it did not have any expertise to
effectively monitor CHS. AA therefore did not have sufficient
regard to the efficient movement of air cargo in preparing the new
airport for operation and should be responsible.

25. NAPCO was the overall monitor of AOR. It should have critically
examined and evaluated AOR critical issues, including the readiness
of PTB and ST1 in effecting the efficient movement of air
passengers, aircraft and air cargo. In discharging these functions,
NAPCO committed two errors: (a) assuming that AA had the
necessary expertise to monitor HACTLs CHS, without even asking
AA if it actually had the expertise; and (b) failing to critically
examine the contingency plans of AA and to query if it had made an
overall risk assessment.

26. NAPCO therefore failed to discharge its duties as the overall
monitor of AOR in its position as the executive arm of ADSCOM
and as directed by ADSCOM. However, as ADSCOM itself was
the overall monitor of AOR, it is ultimately responsible for the
duties of such an overall monitor not having been satisfactorily
discharged by NAPCO.
X

27. Both AA and HACTL were too confident to appreciate the risks
involved in the compression of their testing and training time.
They never sought any postponement of AOD. Had a deferment
been suggested by either AA or HACTL, the chaos could have been
avoided by a postponement of AOD by about two months.
Unfortunately, no one ever made any such suggestion, and everyone
was working diligently but blindly towards the common goal of
AOD.

28. FIDS was operating reasonably efficiently about a week after AOD,
and the other less serious problems also subsided within a short time.
On the other hand, it had taken HACTL about six weeks to recover.
Although there may still be glitches and hitches, the new airport is
now running as a pride of Hong Kong.
XI

CHAPTER 1

INTRODUCTION AND ACKNOWLEDGEMENT

1.1 On 2 J uly 1998, the day following the first anniversary of the
reunification of Hong Kong with the Peoples Republic of China, the new
Hong Kong International Airport (the new airport) at Chek Lap Kok
(CLK) was opened by our President J iang Zemin. The publicity of
the new airport was further enhanced by the departure of Air Force One
with the President of the United States of America on board on the same
day, although the Hong Kong International Airport at Kai Tak (the Kai
Tak airport) was still operating. The Kai Tak airport was to be replaced
by the new airport which was due to open for operation four days after
the ceremonial opening.

1.2 Since Chinas resumption of exercise of sovereignty over
Hong Kong a year ago, people in the Hong Kong Special Administrative
Region have become generally more interested in this place in which they
live and work, as evidenced by the record turnout for the election of the
Legislative Council that took place on 24 May 1998. The building of
the new airport as well as its ground transportation support systems and
the infrastructure items connected with them were complete. These
projects had taken many years of planning and preparation and involved
the largest sum that was ever expended in Hong Kongs history. While
many members of the public went to the Kai Tak airport to have a last
glimpse of it and used the camera to retain their memory before it would
be closed, many more paid visits to the new airport to obtain a personal
feeling of it. Most of those who roamed around the new airport and
those who followed the media coverage on it were impressed with its size
and the spaciousness of the Passenger Terminal Building (PTB) and
rightly so, because PTB is the largest single air terminal building in the
whole world. They entertained little doubt that the services the new
airport offered would be better than those available at the Kai Tak airport.
It is with this kind of expectation that the public and members of the
media were looking forward to the operational opening of the new airport

1

on 6 J uly 1998. The dark cloud of economic downturn that stemmed
from some South East Asian countries and J apan had spread over Hong
Kong; the opening of the new airport was a silver lining that everyone
was anticipating, at least as a booster of confidence that Hong Kong
would have an early recovery.

1.3 There was a huge relocation exercise in the night between 5
and 6 J uly for moving personnel and equipment from Kai Tak to CLK.
This was a tremendous task, involving an enormous amount of planning,
preparation and organisation, counting on the weather being not too
difficult, relying heavily on the coordination and cooperation amongst
members of the airport community, and hoping that the public would not
participate in such a manner as to cause disruption. The electronic
media reflected public interest in the new airport by televising many parts
of the process, which contributed towards keeping interested persons at
home instead of going out into the way of the move. Everything went
on smoothly and nothing appeared to give rise to any worry to the public.
The expectation that the new airport would be a great success was
elevated.

1.4 In the morning news on 6 J uly 1998, a couple on the first
arriving flight were shown on the television, being welcomed to Hong
Kong and given souvenirs to commemorate their being the first arrivals at
the new airport. Later in the day, however, there was news that
passengers had to wait for a long time to get their baggage, that baggage
and air cargo processing was delayed, and that there was congestion in
PTB and the areas around. Everything did not seem that well after all.
On the following days, the media extensively reported the problems
experienced by passengers, visitors as well as cargo consignors and
consignees. Hong Kong peoples great expectation with the linked
auspice of an early economic recovery was dashed.

1.5 For many days, media coverage identified various problems
and reported incessant criticisms, culminating in an outcry that there must
be an investigation of the fiasco that blemished Hong Kongs reputation
as the Asian hub of civil aviation and damaged the business of the air
import and export trades. As a result, the Government, the Legislature
as well as the Ombudsman each announced that an inquiry would be held.

2

The Commission of Inquiry on the New Airport (the Commission) was
thus established on 21 J uly 1998.

1.6 The Commission was given six months within which to
report its findings and conclusions to the Chief Executive. Since the
date of their appointment, the two Commissioners and the Secretary to
the Commission, Mrs Marion LAI CHAN Chi Kuen, started work
without any delay. Mr Benjamin YU SC, Mr J AT Sew Tong and Ms
Yvonne CHENG were appointed as counsel for the Commission, and
Messrs Baker & McKenzie as solicitors. Many meetings were held to
consider various matters necessary to initiate and proceed with the inquiry,
including the selection of experts to assist the Commission, the approach
to identifying the problems with the new airport and obtaining documents
and evidence from various persons or organisations that might be
involved. Media reporting was screened to help in identifying the
problems. Apart from sending out letters to various persons or
organisations to seek information, a hotline was set up to receive
evidence from the public and interested parties. Eventually four experts,
Professor Vincent Yun SHEN, Mr J ason G YUEN, Professor Xiren CAO
and Dr Ulrich Kipper were appointed and they duly participated in the
inquiry whenever necessary.

1.7 Starting from 14 August 1998, three preliminary hearings of the
inquiry were held for various procedural purposes such as arranging for
the persons or organisations who so wished to be made parties to the
proceedings, dealing with legal representation of the parties, and giving
directions on witness statements, order of examination of witnesses,
documents to be used at the hearing, confidentiality of such documents,
and recording of the proceedings, etc. The hearing of evidence started
on 7 September 1998 and concluded on 11 December 1998 and counsels
replies were heard until the last day of the year. Altogether 61 days
were spent for the hearing, both on preliminary and substantive matters.
The parties with a short description of their interest or involvement are
listed in Appendix I to this report, whereas their legal representatives can
be found in Appendix II. The experts appointed by the Commission
and the parties are set out in Appendix III.

1.8 In response to the request of the Commission, around 800

3

box files, containing about 500 pages each, of documents were delivered
to the Commission. The Commission also received 245 witness
statements including supplemental statements, some of which were from
persons not based in Hong Kong. The witness statements from the
parties who were legally represented were all prepared by or with the
assistance of their lawyers. Although there were occasions when parties
or witnesses provided their statements slightly beyond the Commissions
prescribed time, which was tight on all accounts, the documentary
evidence was overall supplied expeditiously. Counsel and solicitors for
the parties were restrained in asking questions of the witnesses called, so
that the number and length of such questions were kept to the minimum
required in the circumstances. All these were very important in view of
the time given to the Commission to complete its inquiry. The
Commission is most grateful for the assistance and cooperation of the
parties and their legal representatives, as well as the organisations and
persons who supplied witness statements and information from overseas
in such limited time, without which this report would not have been ready
so soon.

1.9 Through the hotline, e-mail and post, about 100 persons supplied
information and lodged complaints about the problems encountered at the
opening of the new airport. The Commission is thankful to their
contribution. Before and during the hearing, the Commission made
several visits to various parts of the new airport. Each time, the Airport
Authority and some of the parties made all necessary arrangements to
facilitate access and direct the attention of the Commission to matters in
issue or of interest, for which the Commission is obliged to them.

1.10 The Commission also acknowledges its debt to members of the
media. The media gave wide coverage to the problems faced by the new
airport and dealt with the hearing of the inquiry in no lighter manner.
These, though not easily noticeable, not only helped the Commission in
identifying the problems and occasionally the parties, but also enabled the
public to be made aware of the evidence received by the Commission on
each day of the hearing. The latter aspect is very important, for the
process and progress of the Commissions work was given a certain
degree of transparency.

4

1.11 The Commission is very pleased with the Legislative Council
Select Committee and the Ombudsman, who were separately conducting
their own inquiries into the new airport, for having arranged the order of
receiving evidence from the witnesses who were required to attend the
Commissions inquiry in such a manner as not to cause any disruption to
the Commissions hearing, let alone not to make life difficult for the
witnesses.

1.12 During the past six months, Mrs Marion LAI and her staff in the
Secretariat, all legal representatives for the Commission and all the
experts were working very hard, often staying in the office late in the
evening during Mondays to Saturdays, and frequently on Sundays.
Without their dedication and exemplary diligence, the work of the
Commission would have been impossible. To each and everyone of
them, the Commission expresses admiration and gratitude.

5

CHAPTER 2

TERMS OF REFERENCE

2.1 The two Commissioners were appointed on 21 J uly 1998 by
the Chief Executive in Council under the Commissions of Inquiry
Ordinance, Chapter 86 of the Laws of Hong Kong, to inquire into the
operation of the new airport and the problems encountered since it
opened with the following terms of reference:

(1) To examine the planning and preparation for the opening of
the new airport including the adequacy of communication
and coordination between all interested parties.

(2) To examine the decision to open the new airport on 6 J uly
1998 and the extent to which it was ready to begin
operation on that date.

(3) To examine the operation of the new airport since it opened
on 6 J uly 1998 (including but not limited to flight
information display system, franchised air cargo services,
ramp handling and baggage handling and airside security)
and to identify the roles of the various parties involved.

(4) To identify problems encountered in the operation of the
new airport and to establish their causes and where the
responsibility for each of them lies.

(5) To report to the Chief Executive with findings and
conclusions within 6 months of the date of appointment or
such time as the Chief Executive in Council may allow.

2.2 The Chief Executive in Council directed that the civil
liability of any party for any loss or damage and its quantification should
be outside the terms of reference of the Commission.

6

2.3 The Commission was empowered to appoint experts to
provide reports on any matters covered by the inquiry.

7
CHAPTER 3

METHODOLOGY, CRITERIA AND
TREATMENT OF EVIDENCE


Section 2 : Concerned Parties Role, Responsibility and Liability

Section 3 : Resolution of Issues


Section 5 : Standard of Proof and Treatment of Evidence


3.1 The Commission appreciated from the terms of reference
that it had to inquire into and examine a number of main issues, namely,

(a) whether the decision to open the new airport on 6 J uly was
made correctly or properly when it was made;

(b) what planning and preparation was made to open the new
airport, including the adequacy of communication and
coordination between all interested parties;

(c) whether the new airport was ready for operation on 6 J uly
1998, the airport opening day (AOD);

(d) what problems were encountered in the operation of the new
airport on AOD and thereafter;

8
(e) the cause or causes for such problems;

(f) the roles played by various involved parties; and

(g) the identity of the persons or bodies who were responsible
for the problems.

3.2 Though the decision on AOD can be treated as a separate
topic, whether it was correctly and properly made is connected with the
readiness of the new airport being operational on that day and the
problems encountered then and thereafter. The remaining subject
matters of the inquiry involve a very wide scope, not only because the
new airport was a mammoth project involving the construction of many
buildings, structures and facilities, but the readiness of its opening would
also necessitate the examination of the efficiency of the operation of a
number of services. The Commissioners decided at the outset that the
first step to be undertaken was to identify the problems encountered at the
opening. Once the problems were identified, the Commission would
investigate the causes for them and the roles played by various interested
parties, whereby the persons responsible could be found. The problems
themselves and their causes would enable the Commission to come to a
conclusion whether the new airport was ready to open for operation on
AOD, and if not, whether the communication and coordination of the
persons making the planning and preparation were adequate, and whether
the decision on AOD was proper or correct in the circumstances.

3.3 For identifying the problems and the possible persons that
might have information on the areas in which the problems occurred, the
Commission would first consult media reports as from AOD. Those
persons so identified would be asked about the causes for the problems,
first through correspondence, and when allegations as to the causes had
been received, the Commission would seek witness statements and
relevant documents from those from whom the allegations stemmed.
The Commission would then set up hearings for oral testimony to be
received so as to enable any person or organisation who might be
implicated by the allegations to put forward his or its case and refute any
such allegations. The hearing was to be conducted in public so that all
the evidence that needed to be dealt with would be thrashed out in public,

9
and the whole process of how the Commission reached its findings and
conclusions on all issues would be transparent.

3.4 The areas to be covered by the inquiry were quite wide, if
not for any other reason, mainly because the problems in the operation of
the new airport on AOD and thereafter as reported continuously by the
media were quite numerous.

3.5 With the above approach in mind, the Commission set in
train the following steps:

(a) Gathering as much information as possible about the issues
within its terms of reference from media reports;

(b) Writing letters to various persons or organisations mentioned
in the media reports or who might be involved regarding
each of the problems, seeking information and documents
about the existence of the problems, the causes of the
problems and the roles played by them regarding the
problems;

(c) Appointing counsel and solicitors to deal with matters in (b),
and also to prepare for the hearing of evidence to be held by
the Commission;

(d) Trying to identify from all the information received the areas
on which expert assistance would be needed by the
Commission, for understanding the issues and for expert
opinions on technical and scientific issues to be provided;
and

(e) Getting and setting up a venue for the hearing of evidence to
be conducted by the Commission.

3.6 It is necessary to elaborate on each of the steps mentioned in
the preceding paragraph, except perhaps the venue. The venue was
acquired by the Secretariat in accordance with the size and set-up as
required by the Chairman of the Commission who had the experience of

10
having conducted the Garley Building Fire inquiry in 1997. Although
the large number of parties who were interested in participating in the
hearing was not anticipated, the size of the venue was fortunately just
sufficient to house the teams of legal representatives retained by the
parties and members of the public including the media.

3.7 The Commissioners were appointed on 21 J uly 1998,
about two weeks after AOD. In this intervening period, there was
extensive coverage by the media of the problems encountered in the
operation of the new airport and that was a good starting point in the
Commissions operation. The problems identified with the help of the
media reports would enable the Commission to commence a train of
paper inquiries with the persons or organisations who might also have
been identified by the media reports or who the Commission, with the
assistance of the Secretariat, counsel and solicitors, thought might be
involved. The responses and documents supplied by the various persons
explained the roles played by themselves or provided information to the
Commission as to other persons who were or might be involved. All
such persons or organisations would then be required to provide witness
statements of people who had personal or indirect evidence on the issues.

3.8 Although the process of sending inquiry letters was used
initially for preparation of the hearing, it was also employed for apprising
the parties and non-parties to the hearing of the allegations or possible
allegations against them. The Inquiry started with only some problems
as identified by media reports and the first batch of inquiry letters were
sent to organisations that were thought to be able to explain the problems
and provide information about the causes and the persons responsible
therefor. When answers with allegations of the causes and the persons
responsible were received, other inquiry letters were sent to these alleged
persons, so that these persons were given a chance to respond and more
information could be extracted. The process of paper inquiry, seeking
information and witness statements, and putting forward allegations and
possible allegations for the addressees to respond, went on almost up to
the conclusion of the entire hearing.

3.9 The importance of having a hearing of evidence cannot be
over-emphasised. The person against whom allegations have been made

11
must, to be fair to him, be given an opportunity to answer such
allegations and put forward his own case as to what exactly happened and
how and why he is not or should not be responsible. The hearing has to
be conducted in public, so that all the evidence relating to any person or
any issue will be disclosed for public scrutiny, ensuring that justice must
not only be done but must be perceived by the public to be done. The
hearing is to be conducted in similar manner as a court trial in which
witnesses are to give evidence, either by way of oral testimony or by
producing documents to help establish what they have to say. The
witness will be cross-examined by any party who takes any issue with
him or seeks to establish something favourable to that party, and by
counsel for the Commission in order for the Commission to get at the
truth and raise matters of concern. Although the Commission is
appointed by the Chief Executive in Council, it is important to appreciate
that the Commission was in fact conducting an inquiry on a matter of
public concern, and the public interest was what the Commissioners as
well as counsel for the Commission had to bear in mind, and always bore
in mind.

3.10 The duties of counsel appointed by the Commission were
onerous. Not only did they have to prepare the inquiry letters and make
further inquiries arising from the responses to such letters, they had to
prepare for the examination of the witnesses at the open hearing. It has
to be clearly stated, for avoidance of any misunderstanding, that counsel
were not involved in any decision making of the Commission, in that the
findings and conclusions of the Commission were reached independently
of counsels views, and for that matter, independently of any other
persons. Nonetheless, the views of counsel for the Commission, those
of counsel for the parties who had addressed the Commission, those of
the experts appointed by the Commission and the parties, as well as those
of the witnesses and the representations of the parties had all been
considered before the Commission arrived at its findings and conclusions.

3.11 During the course of receiving information from various
persons and organisations, the Commission was able to get some idea as
to the areas that would involve technical and scientific knowledge, on
which experts assistance would be needed. As a result, four experts
were appointed for the Commission. Their brief curricula vitae can be

12
found in Appendix III. Professor Xiren CAO of the Hong Kong
University of Science and Technology is an expert on mechatronics,
mainly to look into problems alleged by Hong Kong Air Cargo Terminals
Limited (HACTL), one of the two cargo operators franchised to operate
as such at the new airport, to be related to the mechanical, electrical and
electronic equipment of HACTL. Professor Vincent Yun SHEN from
the same university is an information technology (IT) expert who was
to examine problems that were encountered by the Flight Information
Display System installed at the new airport, and also the problems that
might have occurred with HACTLs computer system. Mr J ason G
YUEN, an airport expert from San Francisco Airport, USA, scrutinised
the planning, preparation, communication and coordination necessary for
the opening of the new airport, obviously from the American perspective.
Dr Ulrich Kipper is an IT expert with the added advantage of having been
applying his knowledge of IT in airport operations. He is from the
Frankfurt Airport in which he is employed and from which he normally
operates. He also had the assistance of Dr Markus Leins, a fellow
colleague specialised in IT at the Frankfurt Airport. The participation of
Dr Kipper with Dr Leins as an expert in the inquiry is to straddle between
the fields covered by Professor Shen and Mr Yuen, as well as to provide
input from a European angle. While Professor Shen and Professor Cao
helped the Commission in understanding technical and scientific issues,
the contributions from Mr Yuen and Dr Kipper enabled the
Commissioners to widen their horizons in looking at the issues to be
determined by them in a more cosmopolitan and international manner.

3.12 Shortly after its appointment, when the Commission
started to gather information about the problems encountered on AOD, it
realised from media reports that the problems facing the operation of the
new airport were quite numerous, and many of them were still surfacing.
The Commission was concerned about the sufficiency of the time within
which they should complete their work in inquiring into all the problems
and the depth of their examination of the causes for the problems. Due
to the terms of reference that the Commission had to inquire into the
problems encountered since (the new airport) opened, it appeared that
the inquiry to be conducted was to be an on-going and never-ending
exercise if some problems crept up every now and then during the course
of the inquiry till 20 J anuary 1999 when the Commissions report is due.

13
This must be the case on a strict interpretation of the terms of reference,
but the deadline for the submission of the report would in no
circumstances allow that course being taken. The Commissioners felt,
therefore, that as they were given a fixed period within which to report
their findings and conclusions to the Chief Executive, there must be a
self-imposed end to their inquiry. The Commission would investigate a
problem that occurred after their appointment, as opposed to those that
were known before, if the problem was considered to be of significance,
but on the other hand, it would not deal with any other post-appointment
problem in any great detail.

3.13 Item (4) of the terms of reference requires the Commission
to identify the problems encountered in the operation of the new airport
and to establish their causes and where the responsibility for each of them
lies. Thus, it is clear that the Commission has to inquire into the causes
for the problems and the responsibility for such causes. Upon
examination of some of the documents obtained by the Commission from
various concerned parties, one vexing question immediately surfaced.
The question is: what is the extent to which the Commission should go in
inquiring into such causes and responsibility? In the normal
circumstances of a court trial, the court entrusted with such a task will
certainly have to get to the root of the problem, thereby finding the causes
and attribute the responsibility to one or the other of the parties to the suit.
In most court cases relating to contractual liabilities, the court will make
findings as to which of the contracting parties, who are invariably parties
to the proceedings, is liable to the other party or parties. In most cases
relating to claims of tortious liability, the court will conclude on whom
the liability lies, and if more than one person is liable, the court will
apportion the blame. However, the Chief Executive in Council
expressly directs that civil liability of any party for any loss or damage
and its quantification shall be outside the terms of reference of the
Commission. In view of this express direction, the tasks of the
Commission seem to be lighter than that of the courts in their resolution
of disputes between parties to civil litigation. Moreover, the
Commission is required by item (5) of the terms of reference to report to
the Chief Executive with findings and conclusions within 6 months of its
appointment or such time as the Chief Executive in Council may allow.
Although there is always a possibility that the inquiry entrusted to the

14
Commission cannot be completed with findings and conclusions within
six months of its appointment, it is appreciated that unless there are very
cogent reasons, the given time limit should be adhered to. This is
obvious for at least two reasons. The request for extension of time
should be properly seen as only providing a safety measure to cater for
any circumstances unforeseen by the Chief Executive in Council, but
otherwise the Commission is required to complete its tasks within six
months. The time limit must have taken into consideration the publics
concern in the matter under inquiry, so that the Commission has to
finalise its tasks with dispatch for the public to be apprised of what was
going on with the new airport at its opening within a reasonable time
frame.

3.14 Many problems encountered at the opening of the airport
require examination of voluminous documentation, detailed
understanding of the problems and discernment of where the truth lies
through accounts by various parties who were involved. Such accounts
comprise answers provided to questions posed by the Commissions
counsel in letters or in the examination of witnesses. The parties were
invariably linked by contracts for their basic relationship, and which party
was responsible for the performance of certain obligations in the contract
could normally be determined by the true interpretation of the contract
itself. However, there are at least two obstacles to make the
determination of the cause and responsibility difficult, namely,

(a) A party to a contract alleges that the other party to the
contract is responsible for the problem inquired into by the
Commission because the other party had breached a
contractual obligation owed by the other party, whereby
causing delay or difficulty to it resulting in its failure to
ensure that the problem would not arise. The other party
claims that its breach of obligation was in turn caused by
the first party in failing to perform another obligation under
the same contract. This kind of allegation of breach of a
prior obligation can go on several times between the two
parties, going round a circle of cause and consequence.
The determination of such disputes is very time-consuming.
Although this kind of circles of obligations frequently

15
appears in normal civil litigation before the courts, the
courts have little time constraint in resolving the dispute to
arrive at a final decision. As an example, a building
contract dispute will need a number of months or even
years of hearing by the court before a final decision can be
reached. The Commissions position differs in that the
Commission had to operate in a limited time scope.

(b) One of the two contracting parties makes various
allegations against the other contracting party (the second
party), and the second party blames the non-performance of
an obligation of a third party in another contract between
the second party and the third party. The allegations may
go on for several layers to link a fourth and even more
subsequent parties. Again, this dragging in of parties as to
be the culprit for the event that caused the damage often
happens in normal civil litigation, and the courts are able to
reach a conclusion after a lengthy trial. However, the
Commission did not have the luxury of time.

3.15 The Commission therefore had to decide on the extent to
which its inquiry should attempt, or else there would be no hope for the
inquiry to be completed within the time allowed. Based on the answers
to queries raised by the Commission addressed to various parties, the
Commission had a general appreciation and understanding of the
problems and the allegations of the concerned parties. These allegations
related closely to the causes of the problems and the responsible persons
or parties. Bearing in mind the time required for receiving oral
testimony to allow the parties to have a fair hearing, the Commission was
constrained to impose a stop to the length of its inquiry by setting targets
on the extent of the inquiry. The limitation of this approach is that the
findings and conclusions on the causes for the problems and
responsibility for such causes might not be too definite in that

(a) although the causes for a problem encountered could be
identified, the exact root of the problem might not be found;
and

16
(b) the responsibility for the causes could not be definitely
determined as to be attributable to a party, but rather two and
more parties might be identified as the culprits with no
apportionment of blame, or the responsibility might lie on
one or two or more parties and there is no conclusion as to
which particular one.

3.16 The Commission has realised all along that its findings and
conclusions with the above-mentioned limitation or disadvantages are not
too satisfactory either for the Chief Executive to whom the Commission
is to report or for the public if the Commissions report is released
generally. However, due to the time requirement and the express
direction that they are not to investigate the civil liability of the concerned
parties, the Commissioners feel that the self-imposed extent of the inquiry
is the proper and appropriate approach and it is the best they can do in the
circumstances. Further, the Commission is not entrusted with the task of
finding solutions for the problems. Fortunately, most if not all of the
problems have been rectified and those that still remain are subject to
urgent and earnest remedies, the Commissions inability to get at the root
of the problems would have little adverse consequence. The definite
identification of the culprit, which would be very relevant for the
attachment of civil liability, could be left to the courts or arbitration which
will be resorted to by the concerned parties. After all, section 7 of the
Commissions of Inquiry Ordinance clearly provides that

Evidence given by any person before the Commission shall
not be admissible against him in any civil or criminal
proceedings by or against him, except where he is charged
with any offence under Part V (Perjury) of the Crimes
Ordinance (Cap. 200) or is proceeded against under section 8
or 9 [contempt].

3.17 The inability of the Commission in making definitive
findings of the responsibility of a party in no way hinders any civil or
arbitration proceedings amongst the concerned parties, even though
admittedly it would create a feeling of dissatisfaction on the reader of the
Commissions report that he cannot see a perfect, instead of a partial,
ending of a narrated story.

17

3.18 The Commissioners find consonance of parts of their above
views in the Victorian Communism Commission (1949) Report (Australia)
page 7, where Commissioner Lowe said:

I should not treat the matters investigated before merely as a
piece of litigation between parties in which findings should be
made on the evidence in favour of one party or the other, but as
matters in which the Executive desires to know, not merely
what I find proved by the evidence, but also what the evidence
does not satisfactorily determine and which I think may
nevertheless be possibly true. In what follows there are some
matters in which I am able to say on the evidence are in
accordance with the allegation, and some others which I am
able to say are not in accordance with the allegation, but there
are a number of matters which all I can say is that I am not
satisfied on the evidence that the allegation is true. Such a
finding is not intended to be, and must not be taken to be,
equivalent of finding not guilty. It indicates only that I think
I have not been able to discover what the truth is, and that
further evidence may show the allegation to be true or untrue.

3.19 Many points and arguments were raised by the parties and
their counsel and counsel for the Commission, obviously to look after the
parties interests and to assist the Commission in reaching fair and
reasonable conclusions. While the Commission has dealt with many of
these propositions and arguments in the report, numerous such points
have not been expressly mentioned. This approach of the Commission
must not be taken as its failing to pay attention to or consider all such
ideas. The reasons for not stating them are many. This report is unlike
a court judgment where all arguments of counsel are often expressly
considered, for otherwise the report would give the reader, who is not
necessarily a person trained in the law, a view of too many trees but not a
forest. It would also be too burdensome on the Commissioners who
should bear firmly in mind the necessity of stating their findings in an
expeditious manner. Many points though examined may not lead to any
definite conclusions, because the evidence obtained by the Commission is
not sufficient to enable it to reach a firm view. Some arguments

18
presented are not rational and if stated would simply need to be dismissed.
Other unaccepted or rejected arguments may, on the other hand, require
lengthy analysis and recital of a number of items of the evidence to show
why they are specious or unsound, and their relative unimportance does
not warrant the increase in the complexity and volume of the report. A
report consisting of all such matters would certainly be confusing to the
reader and clouding the main and important issues that Commissioners
ought to decide and have determined. It must, however, be stressed that
an argument or point or evidence which has not been stated in this report
should not be taken as it having not been considered.

Section 2 : Concerned Parties Role, Responsibility and Liability

3.20 The role that a person or organisation plays in an activity is
always a ready and important guide to his or its involvement in that
activity. The involvement will point to the area of duty or obligation.
The obligation may arise out of contract, or it may not. An obligation
under contract is defined by law and a breach of the obligation will give
rise to civil liability. Another way that may attract civil liability is the
commission of an act or omission proscribed by the law of torts. As the
Commission is tasked by its terms of reference not to make any finding as
to civil liability, the Commissions findings and conclusions are on the
roles, acts and omissions of the concerned parties to find out the party
responsible, with or without reference to legal positions under the laws of
contract or torts. Despite that limitation of the Commissions purview, it
is sometimes necessary to look into the legal position of a party. For
example, if there is a statute governing the status and activities of a party,
such as the Airport Authority Ordinance (the Ordinance) establishing
the Airport Authority (AA), or if there is a contract whereby the partys
obligations are defined, the legal position of the party in accordance with
the Ordinance and the contract can thus be ascertained. However, the
Commission has also examined other matters not necessarily relating to
the laws of contract and torts in order to base their findings of
responsibility, such as whether there was sufficient coordination or
communication, or whether a certain work should have been accepted
under the particular circumstances. These matters do not have any
implications on contractual or tortious liability, but they are relevant to

19
the question of responsibility that the Commission has to determine.

3.21 Most of the roles of the parties arose out of contract. In
constructing the new airport and providing it with various facilities, AA
had to employ numerous contractors. Many contractors appointed
subcontractors, splitting the responsibility for performing the works under
the main contract or franchise that was granted to them by AA. There
were even sub-subcontractors appointed by subcontractors. On the other
hand, the roles of the Airport Development Steering Committee
(ADSCOM) and the New Airport Projects Coordination Office
(NAPCO) are mainly not contractually based. For the Government
entrusted works that AA and its franchisees were to perform, the
relationship may be contractual, but in respect of ADSCOMs decision on
AOD and NAPCOs monitoring role over AA, that was purely a matter of
administration of Government. These various roles and relationships
were carefully examined by the Commission. However, responsibility is
not only related to the roles of the concerned parties, but it also hinges on
the causes for the problems.

Section 3 : Resolution of Issues

3.22 For ascertaining the causes of the problems encountered at
the new airport, the Commission needs to determine the issues raised by
various parties to the hearing and non-parties. Such issues were raised
by way of representations presented to the Commission, in the oral
testimonies received during the hearing, or in the examination of such
testimonies. Similarly, the issues on responsibility were raised by
written representations or through oral evidence.

3.23 All parties to the Commissions hearing and non-parties who
were implicated by any allegations were provided with opportunities to
answer such allegations and present their own case. Although many
issues were identified at the early stage of the inquiry, not a small number
of issues only became apparent during the oral testimonies of witnesses at
the hearing or when answers to inquiry letters sent by the Commission
were received. In order to ensure that the length of the hearing was kept
to a manageable extent, numerous inquiry letters were sent, pointing out

20
allegations raised by parties or areas of concern of the Commission, so
that the concerned person or body could respond. For the purpose of
further ensuring fairness to the persons against whom criticisms might be
made, on 14 December 1998, three days after the conclusion of oral
evidence, the Commission issued a broad outline of possible allegations
against parties and non-parties for their consideration. This broad
outline was separate and independent from the final written submissions
of counsel for the Commission and for the parties, and those who wished
to make any representations and submissions to respond to the broad
outline were allowed to do so.

3.24 A copy each of the final written submissions of the parties
were provided to the other parties in the evening of 21 December 1998,
and thenceforth they were also made available to members of the media.
This was to ensure transparency since the submissions were not read out
openly at a hearing. The non-parties were also allowed to inspect these
submissions and respond thereto.

3.25 As is said above, the roles of the parties involved can be
more readily ascertained. However the issues and allegations raised by
the parties and non-parties on cause and responsibility are numerous, and
sometimes extremely involved. For example, the air-conditioning
system, which did not operate efficiently or without fault, is a
conglomeration of the work of a number of parties, from providing the
design to supplying the various equipment and systems.
Correspondence making representations by various parties and
non-parties on the issues raised is voluminous, which makes
determination of cause and responsibility difficult. Even in a simpler
matter that is covered by a single contract with only two parties and
where allegations are conflicting, determination is not rendered any easier.
The reason is the time within which the Commission has to finalise its
inquiry. In view of the importance of enabling the public to be apprised
of the Commissions findings on cause and responsibility in a relatively
short period, the Commission at the early stage of the inquiry decided on
its approach not to get to the root of the causes or the ultimate
responsibility of interested parties wherever time did not permit. A
more detailed discussion of this can be found in paragraphs 3.12 to 3.16
above. The situation of the inquiry is thus very different from that of a

21
normal civil suit, where issues are crystallised by pleadings or through
interlocutory proceedings, and parties are allowed to call their own
witnesses at the trial. The witnesses are examined thoroughly in court
and conflicting evidence can be determined after hearing the witnesses.
Even where evidence is conflicting and there is no material supporting
one version or the other, and other rational bases for determining the issue
being equal, the court can as a last resort decide the issue upon
observation of the witnesses demeanours. Due to the time constraint,
not all persons who had made conflicting allegations were called, and the
Commission was deprived of the opportunity to observe the demeanour
of witnesses. Besides, the Commission has been cautious to uphold an
important principle which is that no person should be condemned until he
has a chance to be heard. The best way to test the allegations is to put
them to a witness called by one of the involved parties, so that he may
answer on behalf of the party. By this method, the witness is given a
chance to proffer whatever explanations he deems necessary and his
demeanour in the witness box will be examined by the adjudicator. This
avenue, however, was not always open to the Commission wherever the
seriousness or otherwise of the problem to which the evidence would
relate did not warrant a considerable amount of hearing time being spent.
In such circumstances, it is difficult, if not impossible, to find out the
precise cause for the problem on which the concerned parties had made
conflicting allegations or which of the parties who made such allegations
should be responsible for the fault. Notwithstanding the limitation, the
Commission has considered all the relevant materials and analysed the
situations very carefully to reach its views and findings.

3.26 There is an instance where an allegation or its seriousness
was not realised during the hearing, and therefore it was not put to the
relevant witnesses when they gave evidence. That is in relation to
ADSCOMs overall responsibility in monitoring airport operational
readiness (AOR), dealt with in the concluding part of Chapter 5.
Although the allegation was not put to relevant witnesses for them to
answer, the Commissioners think that as their opinion and finding are
based on their understanding of the circumstances surrounding the issue
and the law, rather than dependent on any answers that might have been
given by the witnesses, it is proper to include their views in the report.

22


3.27 Some of the matters the Commissioners are tasked to
examine are the adequacy of communication and coordination of persons
responsible for the opening of the airport, and the readiness of the airport
for being opened on 6 J uly 1998. Readiness involves whether the new
airport was safe, secure and efficient for users, including aeroplanes,
passengers, and those working in it. In performing their functions, the
Commissioners will have to determine the criteria against which the
involved persons should be judged in relation to the various issues within
their remit, in particular readiness, efficiency and adequacy in their
respective context.

3.28 In the Commissioners opinion, readiness, efficiency and
adequacy are all matters of degree and they have to be examined in the
surrounding circumstances. The Commissioners are Hong Kong people,
and the Hong Kong perspective will be taken into account. However,
for judging these issues in relation to the operation of the new airport,
which is an international airport, the Commissioners decided that it is
proper also to take into consideration the international viewpoint and
experience. That was the reason why the four experts with different
backgrounds were appointed. While Professor Shen obtained his first
degree in Taiwan, Professor Cao got his in China, and both of them were
conferred a doctorate by a US university. Mr J ason Yuen is an airport
expert based at the San Francisco Airport, although he has experience
with airports outside the USA. On the other hand, Dr Kipper operates
from his airport management company at Frankfurt Airport and has
worked in airports in Europe and South East Asia. The expertise from
all of them would enable the Commission to view the subject matter of
the inquiry in an international perspective.

3.29 A few examples on readiness may help to explain why the
Commission considers that the issue has to be examined in all the
surrounding circumstances. If a person were to say that he is ready to
leave home for going to the market to buy things, he would only be ready
when he has brought with him some money for the purpose. If he were
to go out to a dinner party, he would only be ready when he is properly

23
dressed for the occasion. If he were to go to catch a plane, he would not
be ready unless he brings with him his travelling documents. In the
context of AOR, especially the huge airport at Chek Lap Kok (CLK),
the steps to be taken for readiness must be magnified, to say the least,
thousands of times. An example which is very much smaller by scale is
organising a picnic for a school with 1,000 students. Each of the
students will have to be notified of the programme, where to meet and
what to bring and the food arrangements, etc. The teachers who are
responsible will have to decide what happens if one or more students are
late or sick to attend the meeting place, or what to do if they feel sick in
the middle of the trip, etc. The organisation and works of the airport are
many more times larger, and there are at least the systems integration
issues that are not normally required to be handled in a school situation.
In a school organisation, there has been an established class system, say
each consisting of 40 students, headed by a class teacher. Each class
teacher is under a head teacher or supervisor responsible for a number of
classes or forms, and the supervisors are under the direction of the
headmaster. The small degree of integration, if need be, is to be made
through the line of control, by the class teacher over the class, then by the
supervisors over the class teachers, and by the headmaster having charge
of all the supervisors, teachers and students. This system of control and
integration is simple and can be appreciated by most people who have
gone through school.

3.30 However, building an airport and making it ready for
operation is a very different matter. There were at one time over 20,000
labourers of various disciplines involved in the building of the airport,
employed by over 80 main contractors. Contractors were employed by
AA and its 28 business partners and franchisees, and many contractors
shared their work with many sub-contractors and sub-subcontractors.
Building works themselves require very careful and close coordination.
It may be easier for the general public to appreciate the situation by using
an example when a person wishes to decorate or refurbish a flat involving,
say simply four types of work: painting the walls and ceilings for the
whole flat, building some shelves, laying carpets and changing the floor
tiles of a bathroom. Instead of entrusting all the four types of works to
a single contractor, the flat owner asks four different contractors to do
each type of the works required, for the sake of saving some expenses.

24
Which of the four kinds of works should be done first, and which last?
It seems that the changing of the bathroom floor tiles has nothing to do
with the other three and so it can be done first. But if the bathroom
walls have to be painted, should the floor tiles be laid first or the walls be
painted first? Should the shelves be put up first and then the walls
painted, or vice versa, or should the carpets be laid after everything else is
done? Or rather, should each piece of the works be done partially to
await other pieces of the works to be done partially, and various stages of
the works would have to be programmed very carefully in order to avoid
delay and workers of different contractors crossing each others paths and
doubling the tasks. The existence of four different contractors will
certainly complicate matters and the flat owner will have to do all the
necessary coordination all by himself. That is why that sometimes in
order to save the trouble, the flat owner will entrust all the works to one
contractor and let the latter do all the coordination. But that will in most
cases involve greater expenditure, for the contractor will certainly include
in his charges a sum for covering the time and effort he has to spend in
coordinating the various pieces of works.

3.31 Doubtless, there are many officers and staff of AA who are
experienced people in various well-established professions, such as
engineering, building, IT and management. As funding was provided
piecemeal at the initial stages of the construction of the new airport, AA
did not contract out all the works required to achieve AOR to one single
contractor. In fact, over 80 contractors were employed to construct
various buildings, to provide various building services and facilities and
to supply and commission various technological, IT and computer
systems. AA has had to perform the coordination of all these various
works and systems. Moreover, various buildings and facilities were to
be provided by business partners and franchisees, such as catering,
provision of aeroplane fuels, baggage and cargo handling, to name but a
few. While these business partners and franchisees were bound by
contract to complete their works and therefore make their buildings,
services and facilities ready for airport users, if they slip one way or
another, AOR will be affected though AA may charge them penalties
according to the contractual terms or may claim damages for breach of
contract against them in the courts. AA oversaw their works so as to
ensure completion in time, but AA could hardly do anything else except

25
to impress upon them and the sub-contractors employed by them the
importance of completing promptly. There could not be direct
interference, and even if there could be, in the short time available since
J anuary 1998, nothing could possibly be done to improve even if AA
were to take over their works.

3.32 All that said, the Commissioners do not forget that AA had
teams of highly professional officers and staff. They are trained and
experienced in the works and non-works activities and in coordinating
them. It is on that plane that their performance is to be judged.
Another context in which the whole matter must be viewed is that they
were not only putting up a house or a building, but they were building a
huge airport involving air traffic and thousands of users each day.
Readiness in this sense must necessarily mean that the buildings, services
and facilities are not only available, but that they have to be safe, secure
and efficient in performing all their proper and expected functions.

3.33 On the other hand, the involved parties did not envisage that
on AOD, the new airport would be fully ready as if it were that everything
that was available in Kai Tak would be there in CLK, because for instance,
phase 5 of the relocation exercise, to take place between 06:30 hours on 6
J uly to 5 August 1998 had still to be performed. The readiness required
to be judged must be viewed in this light.

3.34 People in Hong Kong are always proud of their efficiency.
The social welfare status of Hong Kong has been such that except for the
very needy who may be taken care of by the Government, every one
looks after his own welfare. It is a densely populated community where
survival and flourishing depends on the personal endeavours made by the
individual, in constant competition with others. When one can afford it,
there are numerous and multifarious entertainment avenues available. It
is a vibrant and hectic place where the flame of life burns vehemently.
Remarkable efficiency amongst the people has developed throughout the
years because of keen competition and a feeling of allowing nothing to be
missed, either in work or enjoyment, within a given time span. For
example, people do not have to wait for long inside a bank to transact
deposit or withdrawal transactions. Bank tellers, while courteous, will
not spend time to chit-chat with customers, so that the next customers in

26
line will not need to wait any longer than necessary. Throughout the
territory which is less than 400 square miles in area, there are thousands
of automatic teller machines provided by various banks for people to
withdraw money, 24 hours a day. Apart from fast food stalls, small
restaurants can be found in many places where one can have a meal
served within a couple of minutes, etc. This efficiency applies to all
sorts of things and activities, while improvements are constantly
attempted and peoples expectation of efficiency keeps on growing.

3.35 Nevertheless, the Commissioners, as Hong Kongers
themselves, feel that efficiency and expectation of it must be put in their
proper perspective. Efficiency is in most cases relative, and
expectations are generally based on former experience. If an outgoing
passenger gets into the airport and is immediately served by the check-in
counter, the immigration counter, the customs check, then these services
can be considered to be absolutely efficient. But if he is the second
person in a queue to be dealt with by the airline staff, the immigration and
customs officers, then he may feel that the services are not efficient.
The worse if he ranks tenth or later in such a queue. It is difficult to set
a proper standard for efficiency, especially a situation may change with
numerous permutations depending on the length of the lead time before
the estimated time of departure of the flight that the passenger is catching
and the time when he arrives at the airport as well as other events such as
when other passengers boarding the same flight arrive and many other
circumstances which do not necessarily depend on the number of staff
deployed at each of the counters that he has to go through before boarding
the plane. It is therefore necessary to compare with the situations at Kai
Tak and in airports throughout the world or at least in developed countries
in order to gauge efficiency.

3.36 In this respect, the survey on user friendliness of the new
airport carried out by the Tourist Association in a period of five days on
10 to 14 J uly 1998 is instructive. It is pointed out, as evidenced by the
responses to the questionnaire used by the Association, that overseas
visitors were more easily satisfied than local residents, regarding almost
every aspect of the airport services and facilities.

3.37 Adequacy is another issue that has to be judged in all the

27
surrounding circumstances and in the context to which it relates. In
Hong Kong, a flat of about 100 square metres is said to be a middle range
residence, whereas in many parts of the USA, a place of that size is
considered to be very small. When translated in terms of adequacy, a
residence of such a size will sometimes be considered to be adequate for a
family of four, but sometimes inadequate. In the context of
communication and coordination in the planning and preparation for the
opening of the new airport, adequacy therefore has to be gauged against
many factors such as how much the airport was ready for operation,
whether the services and facilities rendered were safe, secure and efficient,
the importance of the task to be performed, the qualifications and
experience of the persons involved, the positions of such persons, etc.
All these will have to be viewed against the Hong Kong standard and in
the international perspective.

3.38 Having considered all evidence and matters, the
Commissioners are firmly of the opinion that the minor problems
encountered on AOD and shortly thereafter should not fairly be used as
the basis for treating that the new airport was not ready for operation, that
minor human errors should be excusable even according to the high
standard of efficiency of Hong Kong people and their expectation and
that slight oversight with insignificant consequence should not be
considered as the matter not having been adequately attended to or
considered.

3.39 The Commissioners have also borne in mind and at heart the
danger of using hindsight as the basis of criticism. Pollock MR said in
City Equitable Fire Insurance Co Ltd [1925] 1 Ch 407, CA, at 509:

As I have already said it is quite easy to charge a person after
the event and say: How stupid you were not to have discovered
something which, if you had discovered it, would have saved us
and many others from many sorrows.

3.40 Hindsight is important for discovering what lessons to be
learned from a past event, and perhaps even more important for the
purpose of the inquiry in finding out the truth, but it is not a proper
yardstick against which blame should be evaluated. This approach is

28
also shared by W51 Mr J ason G YUEN when in cross-examination he
agreed to the following proposition: Hindsight is necessary for
ascertaining what is the cause of things, but hindsight is not a good
measure for responsibility and blame. In the process of reaching their
findings and conclusions, the Commissioners have judged the
responsibility of each of the persons subject to inquiry by what he knew
or should reasonably have known at the time of his conduct or activity
and in the light of all the surrounding circumstances.

Section 5 : Standard of Proof and Treatment of Evidence

3.41 Under the terms of reference, while the Commission is
tasked to inquire into the decision on AOD, problems affecting the
operation of the new airport and the causes for such problems, it has to
make findings and draw conclusions as to where the responsibility lies.
On the other hand, it is expressly proscribed from deciding on civil
liabilities amongst involved parties. In civil cases, normally the onus of
proof lies on the party who makes the assertion, and the standard of proof
is on the balance of probabilities, meaning more likely than not. The
Commission is of the view that for matters to be decided by the
Commission, although there is generally no onus of proof on any party, it
is always safe and proper to adopt the same civil evidence rule that he
who alleges must prove. The standard of proof adopted by the
Commission is also generally on the balance of probabilities, but the
more serious the nature of the allegation or criticism, the weightier the
evidence there must be for the Commissioners to be satisfied. A finding
on an issue must be supported by a standard of proof commensurate with
the seriousness of the issue. Where it is stated in this report that the
Commission reaches any finding or conclusion, the standard in support is
that on the balance of probabilities. When the finding or view is based
on more cogent evidence, the Commission will state the higher standard
that has been reached, by using terms such as beyond all reasonable
doubt, sure, undoubtedly, doubtless or absolutely, etc.

3.42 In the course of the inquiry, voluminous documents have
been supplied by parties and non-parties to the Commission. Witness
statements of over 200 witnesses have also been obtained. The length of

29
hearing by the Commission was kept to the minimum, so as to save time
and enable the Commission to reach its conclusions within the time
allowed by the terms of reference and to prevent expending any amount
of public funds more than absolutely necessary. Only 56 witnesses were
called and examined on oath or affirmation at the hearing, and some of
them were called as a group so as to optimise effect and minimise time.
Greater detail of this unusual procedure can be found under Section 3 of
Chapter 4. The witness statements of the witnesses who were not called
are considered although the persons had not been subject to oral
examination by the Commission or the parties. Over 1,200 inquiry
letters were sent from time to time to parties and non-parties to seek as
much information on various issues as possible to avoid having to call
persons or organisations who dispute the issues, a measure also to keep
the length of the hearing to the minimum. The witness statements, the
answers to the Commissions inquiry letters and a large amount of the
documents have been examined and many of the Commissions findings
are based on them. The Commissioners have been conscious of the fact
that sometimes it would not be very satisfactory to rely on documents and
witness statements when there was no opportunity for the person or
organisation affected by them to cross-examine the makers to test their
evidence. However, the Commissioners feel that in the circumstances
and for public interest, that has to be done, or else the inquiry could only
be concluded within years, and their findings would only be made when
the publics memory and interest in the subject would have long
evaporated. The Commissioners have exercised great care when
documentary evidence is preferred to witnesses oral testimony.
Moreover, oral testimony based on a witnesss memory of events may not
be of better evidential value than contemporaneous documents. The
Commissioners also observed witnesses demeanours in evaluating their
evidence and sometimes rely on the inherent probabilities of matters to
help determine where the truth lies and whether a witness is truthful.

30
CHAPTER 4

THE HEARING AND CONFIDENTIALITY OF DOCUMENTS


Section 2 : Confidentiality of Documents

Section 3 : Hearing of Witnesses by Group


4.1 As alluded to in Chapter 2, the hearing conducted by the
Commission began on 14 August 1998, to deal with preliminary
procedural matters before the substantive hearing of testimony. The
preliminary hearings were to ensure that the substantive hearing was to be
conducted smoothly and with as little interruption as possible.
Altogether there were three sittings on preliminary matters and 58 days of
substantive hearings. The hearing dates and witnesses appearing at each
are set out in Appendix IV to this report.

4.2 All the hearings were conducted in public, like any court
hearing of civil or criminal litigation. Everything done by the
Commission was transparent, and the evidence that the Commission
would or might rely on in the consideration of its findings and
conclusions were all mentioned at the public hearings. Those who
might be implicated or concerned in the subject matter of the inquiry
were at their request duly made parties to the proceedings. Almost all of
them were represented by counsel or solicitors, whose proper questioning
of witnesses was invariably allowed. They were all informed of the
experts appointed by the Commission, their respective expertise and the
issues to be dealt with by them. The parties were free to appoint their
own experts and have the difference in expert opinions resolved either by

31
agreement between various camps including the Commission or by orally
examining the expert evidence at the hearing. Timetables of witnesses
attendance, which were revised from time to time as circumstances
demanded, were supplied to the parties to the proceedings as well as
members of the media who had expressed an interest of getting them.

4.3 Due to time constraint and the voluminous documents
involved, no hearing bundles of documents could be prepared for the use
of the parties to the inquiry. However, to enable the parties to prepare
for the witnesses to be called, counsel for the Commission were directed
to provide to the parties with an index of files or documents with page
reference to be referred to in respect of a particular witness within three
days before he/she was to give evidence or within two days after his/her
witness statement was submitted, whichever was later. Parties who
wished to refer to further documents relating to the witness were required
to give notice of the additional documents with relevant page reference
within 48 hours of receipt of the index. Such parties were also required
to prepare 27 copies of the files or documents covered by their notices for
the Commission and the parties on the day when the witness was called.

4.4 All the above was done for ensuring fairness to the parties,
especially to the persons who might be implicated by the Commissions
findings and conclusions. Members of the public had full liberty to
attend any of the hearings as they pleased. The presence of many
members of the media at the hearings enabled the proceedings to be
reported and made known to the public who did not attend. The
Commissioners are confident that justice has been done and has been seen
to be done.

Section 2 : Confidentiality of Documents

4.5 The only thing that was not disclosed to the public or the
media is the contents of the documents on which a successful claim for
confidentiality was made by the parties. During the preliminary hearing
to hear interested persons as to their participation in the substantive
hearing of evidence, their legal representation, and various other
procedural matters, the question of confidentiality of documents was

32
raised. The parties had supplied or would supply numerous documents
to the Commission. They expressed their willingness and readiness in
providing documents to the Commission to assist in its inquiry but were
concerned that the disclosure of some of these documents to the public or
other parties would be detrimental to their interest. While the
Commission had power to order production of documents, and the parties
were cooperative in that respect, the Commission was mindful that any
person supplying the documents should not have his interest, commercial
or otherwise, unnecessarily affected. A balance must be properly drawn
in order to enable the public to know the evidence adduced to the
Commission and the interests of the parties that ought in all fairness to be
protected.

4.6 At the preliminary hearing on 21 August 1998, the
Commission made rulings on the documents that were irrelevant and
those that would be regarded as confidential, in the sense that they should
not be released to the public or other persons during the public hearing of
evidence, and if such documents had to be referred to in the hearing,
protection would be provided to ensure the least impact on the party
supplying the documents. The Commissions rulings and directions are
summarised below, with brief reasons:

(1) Regarding materials that are irrelevant, they should not be
used for the inquiry at all, and no party to the inquiry except
the one who has supplied them to the Commission should
have access to them. The Commissions Secretariat will
check the documents claimed by all parties to be irrelevant
and exclude them from being accessible to anyone other than
the supplier. Any disagreement between the party making
the claim and the Commissions counsel on irrelevance will
be determined by the Commission at a later hearing.

(2) Regarding the relevant materials, the following grounds in
support of the claims of confidentiality of documents supplied
to the Commission are allowed:
(a) Security materials relating to the security of the new
airport and related operational procedures. These
materials should generally be excluded on the ground of

33
public interest for the protection of the security and
safety of the airport and its users.
(b) Intellectual Property Rights materials showing the
design and specifications of devices and systems that are
subject to intellectual property rights. These materials
should generally be excluded on the ground of trade
secrets and for the protection of intellectual property
rights and technical know-how.
(c) Commercial Sensitivity materials concerning pricing
and costing. These materials should generally be
excluded on the ground of trade secrets for protecting the
marketing techniques and competitiveness of business
concerns.
(d) Potential Litigation and Claim materials that are
subjects of potential litigation and claim, including
materials that are subjects of legal professional privilege
and discussions on how claims or potential claims are to
be dealt with. The materials subject to professional
privilege should be excluded absolutely (whether they
are relevant or irrelevant to the inquiry) on the
well-established basis of safeguarding confidence, trust
and candidness between client and lawyer in the context
of fair administration of justice. The discussions on
how claims or potential claims are to be dealt with
should be generally excluded on the ground that they are
secrets which the others concerned with such claims
should not be given any opportunity to get to know, or
otherwise the party involved in the discussions would be
unfairly prejudiced.

(3) Regarding the materials that are subject to the general
exclusion under the four allowed grounds, however, there
may be certain materials that are germane to various issues
of the inquiry. These materials may be used at the hearing.
The Commission can exclude the public and parties who are
not concerned with the particular topic from the hearing, and
that part of the transcript relating to such closed door hearing
can be excised before the transcript is made available to

34
parties other than those participating in the closed door
hearing.

(4) The following directions arise from and are ancillary to the
above rulings:
(a) The materials, either in the form of a document in its
entirety or in the form of identified parts of a document,
should first be excluded or blocked out from the Copying
Bundles kept by the Commissions Secretariat. They
will therefore not be disclosed to the other parties
allowed to have access to the Copying Bundles. The
non-disclosure will be maintained regarding the
documents that may be compiled for the use of the
hearing.
(b) The parties claiming the exclusion must identify, by
reference to the page numbering used by the
Commission, all the documents in their entirety or the
specific parts of each of the documents sought to be
excluded pursuant to the four allowed grounds, and
notify the Commissions Secretariat accordingly. They
shall use their best endeavours thereafter to render
assistance to the Secretariat in effecting the exclusions
and shall be prepared to attend the Secretariat for that
purpose. The Secretariat will make arrangements with
them for their attendance. Parties who do not submit
their identification expeditiously will be deemed to have
waived their claim of confidentiality on the documents
and materials supplied by them to the Commission.
(c) Immediately before the generally excluded materials are
to be referred to at the hearing, the party wishing to rely
on the Commissions general ruling on exclusion should
be on the alert to make an application to the Commission
for exclusion of the public and unconcerned parties from
the hearing and the consequent excision of the transcript.
This kind of application will be dealt with on an ad hoc
basis.

(5) To further protect the parties who have supplied documents

35
and materials to the Commission for the purpose of the
inquiry, each party to these proceedings and their legal
representatives must each give a written undertaking to the
Commission that no document, material or information
obtained from the Commission or the inquiry, save the
Commissions report to the Chief Executive or any part
thereof which has been made public, shall be used for any
purpose other than for the inquiry. The form of the
undertaking will be settled by the Commission.

4.7 Based on the above rulings and directions, many parties who
had supplied documents to the Commission started their claim of
confidentiality and irrelevance. The documents provided to the
Commission were voluminous and when sorted by the Commission
Secretariat, they comprised not less than 800 box files each consisting of
about 500 pages. Messrs Baker and McKenzie, solicitors for the
Commission who worked under the direction of the Commissions team
of counsel, had to deal with all the claims in a preliminary manner, going
through the claims made by the parties, agreeing to them and overseeing
the parties in taking steps to expunge the documents or to obliterate parts
of the documents from the files before they were allowed to be copied by
other parties. There were some initial disagreements of confidentiality,
which were fortunately resolved in the spirit of goodwill and cooperation
between all concerned. The Commissioners task in determining on
particular documents or portions of them within the ambit of their rulings
was greatly relieved. As a result of this onerous exercise, about 500
files remained to form the centre of the attention of the parties and the
Commission and its team of lawyers.

Section 3 : Hearing of Witnesses by Group

4.8 The inquiry hearing followed the same procedure as a court
trial where witnesses are called one by one. Each witness gives
evidence in chief, led by the party calling him. He will then be
cross-examined by the other party or parties, and thereafter re-examined
by the calling party. About a week after the commencement of the
hearing of evidence, the Commissioners were concerned about the slow

36
speed at which the hearing was proceeding and the fact that quite a
number of answers given by witnesses were based merely on their
understanding of the situation from information supplied to them by other
person or persons. The witnesses alleged that because of their position
in the organisation in which they worked, they were not responsible for
the area with which an issue was related, and someone else in his
organisation had the responsibility for that area. That meant, either they
said they could not answer the questions, or when they answered the
questions, their answers were based on second-hand knowledge. The
consequence was that the person or persons who were alleged to have
direct knowledge would have to be called. The Commissioners found
this unsatisfactory because not only that issues raised could not be
resolved immediately, but also that calling other witnesses on the same
issues would inevitably prolong the proceedings. Moreover, the former
witnesses might have been allowed to shirk the responsibility to answer a
question. Although they might be recalled, that would result in time
being wasted.

4.9 At a meeting to discuss the progress of the hearing, the
Commissioners learned that their concern was shared by counsel and
solicitors for the Commission. Everyone was trying to see a way to
alleviate the situation, and Dr Edgar Cheng raised a novel and interesting
suggestion. He proposed that witnesses from an organisation should be
called as a group so that all those who were or might be responsible for
areas relating to particular issues of the inquiry would all be brought
before the hearing to answer questions put at the same time. While the
proposal was attractive as being able to solve the problem facing the
Commission, all the lawyers at the meeting were feeling uneasy about it
as being anomalous to their training and conventional practice. In an
ordinary court case, the normal practice is for witnesses to be called one
after another, each individually giving evidence in chief, cross-examined
and re-examined. There has not been a case in Hong Kong known to
those present at the meeting where a group of witnesses gave evidence
together in the witness box. After considering the proposal for a day
and seeing that there was nothing against the principles of fairness and
justice, the Chairman agreed to put the proposal to the parties at the
hearing.

37
4.10 All counsel for the parties who addressed the Commission
on the proposal were unanimously against the idea. Having considered
counsels submissions very carefully, the Commission was of the view
that the proposal should be put into practice, and gave the following
ruling on 21 September 1998:

For the purpose of saving time and concentrating on
particular issues, and to help establish the identity of the person who
has personal knowledge of matters relating to those issues, at the
hearing last Friday, the Commission proposed that a group of witnesses
from a party should be called to give evidence en masse. We heard
counsels views on our proposal and all those for the parties who spoke
were against it. Their views can be summarised as follows:
(a) the proposed course is probably unworkable because if a group
of witnesses are asked a question, none of them would know
who is going to answer what;
(b) if one of the group is criticised, it would be unfair to him because
the questions put were answered by someone else in the group
and not necessarily himself;
(c) there will be a risk of treating the evidence from the group as the
evidence of the individual, especially if that individual is to be
criticised;
(d) a witness in the group may not only be relevant to the issues
examined, and he will have to be recalled individually when
other issues relevant to him are examined;
(e) it may be difficult or unmanageable if all the witnesses in the
group are to give evidence in chief simultaneously, or
cross-examined simultaneously;
(f) it may make counsels tasks in preparation of the examination
more difficult; and
(g) it may create practical difficulty because if the witnesses to be in
the group are important for the running of the business of the
party, and the requirement of all of them to attend the hearing
together will debilitate the partys operation.

We have considered the matter very carefully, especially
in view of the objection by counsel, almost in unison. However, in
view of the fact that we have all to work towards a time target, ie, to

38
complete the hearing by sometime in December, and in order to enable
better focus on certain issues in the inquiry, we feel that it may be
profitable to make an attempt to have several witnesses called together
on a particular issue or issues. This direction of course will not apply
to general matters or evidence required of a witness who will cover a
number of matters that may or may not be related to a single issue.

Our direction will no doubt enable the parties concerned
to concentrate on a particular issue, by having all the witnesses (from a
party) who are or may be responsible for that issue to be called together.
Anyhow in preparation of the examination of each of the witnesses in a
group, counsel would certainly have to take into account the contents
of the witness statements of the others in the group. It will be
beneficial to clarify amongst all the witnesses concerned with the issue
from a party to be asked a global question who in fact was responsible
for what, and then the questions can be directed at the person who
claims or admits to be so responsible. We do not think that counsels
task will be rendered harder. But even if there is some risk of it being
so, we think that the benefit of having all matters relating to one issue
and the responsibility for it clarified at the same time with all those
who have personal knowledge or may be responsible would outweigh
the little disadvantage that might be encountered by counsel.

We propose the following guidelines for the examination
of a group of witnesses so that most, if not all, of the problems
postulated by counsel would unlikely occur:
(a) Where a number of witnesses are called at the same time, each of
them will be sworn or affirmed individually.
(b) For the LiveNote record, each will be assigned an alphabet, from
B onwards, because A is normally reserved for meaning
answer.
(c) All the witnesses will have made a witness statement or
statements, and their witness statements are to be treated as
evidence in chief under oath.
(d) The party who leads the evidence of a group of witnesses will be
allowed to ask a few questions of each of the witnesses
individually, to clarify or add to what is stated in his own
witness statement(s).

39
(e) The cross-examination of the witnesses will, depending on
circumstances, be either directed at the group or at each
individual witness. For example, Who is responsible for
testing a particular area of a system? will be a global question
to all, whereas the witness whose responsibility is thus
identified can be asked questions directed to him alone. At the
end of each question or at the close of the partys
cross-examination, a wrap-up question may be asked of the
group if any member would like to say anything further on
what the other or others have said. When a global question is
asked, the witness who wishes to answer will be asked to raise
his hand.
(f) Re-examination can also be done either by asking a question of
all of the witnesses or directed at one particular witness.

Our guidelines seem to be able to answer the first six of
the points made by counsel. Hereunder, we examine each of the
points made, following the same sub-paragraph numbering under the
first paragraph of this ruling:
(a) The argument that none of the witnesses in a group would know
who is going to answer what will not arise as the decision on
who is to answer what rests with the counsel asking the
question.
(b) If one of a group of witnesses is to be criticised, the criticism
should be directed by counsel to him, and he will not take
responsibility for an answer not given by him.
(c) Any criticism that may be made by counsel or the Commission
of an individual witness will be based on all the evidence
received; and any counsel who wishes to put an allegation or
accusation against the witness personally will give him an
opportunity to answer.
(d) A witness who is in a group of witnesses and whose evidence is
required for other issues apart from the issues examined during
his participation in the group will have to be recalled
individually; but this will not increase his burden, as anyhow he
will need to be examined on both sets of issues.
(e) The group of witnesses will be in the witness box together, but
our guidelines do not have the consequence of their giving

40
evidence simultaneously.
(f) As we said earlier, counsels tasks in preparation of the
examination may or may not be more difficult, but the benefit
of enabling everyone, counsel and witnesses alike, to focus on a
particular issue should outweigh any small disadvantage that
may be experienced.
(g) Regarding the practical difficulty that might be caused to the
operation of the business of the party, we feel that if the
witnesses, whether individually or in a group, will be required
to attend the hearing, their absence from work, either staggered
or globally, will have similar effect to the party. However, we
will keep an open mind on this, and when the timetable, which
will list the witnesses to be called in a group by enclosing their
names in a pair of parenthesis with an indication of the issue or
issues to be dealt with, the party concerned who sees the
difficulty can address us, although of course, we do not wish to
spend too much time on this sort of application.

4.11 Pursuant to the ruling, several groups of witnesses from the
following parties were called dealing with the same or related issues: the
Airport Authority, Hong Kong Air Cargo Terminals Limited (HACTL),
Murata Machinery (HK) Ltd, New Airport Projects Co-ordination Office,
Cathay Pacific Airways Limited, the two experts appointed by HACTL
and two of the information technology experts appointed by the
Commission. The witnesses who gave evidence as a group can be
identified in the Hearing Dates and Witnesses at Appendix IV by the
word (with) appearing after the name of those who gave evidence with
one or more witnesses. Nothing unfair or unjust or untoward happened,
and no counsel for any party criticised the procedure or addressed the
Commission further on it. The procedure operated smoothly and
effectively. In the Commissioners view, the procedure worked
satisfactorily and contributed to saving time, effort and costs of all
concerned.

41
CHAPTER 5

THE ROLES AND DUTIES OF KEY PARTIES AND
COORDINATION


Section 2 : The Roles and Duties of the Key Parties
(a) AA
(b) ADSCOM
(c) NAPCO

Section 3 : Communication Channels

Section 4 : Adequacy of Communication and Coordination


5.1 Under the September 1991 Memorandum of Understanding
Concerning the Construction of the New Airport in Hong Kong and
Related Questions (Memorandum of Understanding) signed by the
Governments of the Peoples Republic of China and the United Kingdom,
a new airport was proposed and to be completed to the maximum extent
possible by 30 J une 1997.

5.2 The new airport was a portion of the Airport Core
Programme (ACP) including altogether 10 major infrastructure
construction projects with the new airport at Chek Lap Kok (CLK)
being the ultimate focus of attention.

5.3 In J anuary 1994, a 45-month programme was established by
the then Provisional Airport Authority (PAA) and endorsed by the
Airport Development Steering Committee (ADSCOM), based on

42
step-by-step funding considerations and forecast award dates for the
foundations of the Passenger Terminal Building (PTB), superstructure
and other major contracts. This would lead to an airport opening target
of 30 September 1997.

5.4 After the Support Agreement Relating to the Financing,
Construction and Operation of the Airport was signed on 1 December
1995, it was announced that the new airport would be opened in April
1998, followed by the opening of the Airport Railway (AR), which was
later known as Airport Express, in J une 1998. Since then, April 1998
had been adopted as the target date for opening the new airport for the
purposes of planning, programming and preparation.

5.5 On 1 December 1995, the Airport Authority Ordinance (the
Ordinance), Chapter 483 of the Laws of Hong Kong, came into force,
whereby PAA was reconstituted to become the Airport Authority (AA).
The aim of the Ordinance is, inter alia, to enable AA to provide, operate,
develop and maintain an airport for civil aviation in the vicinity of CLK
and makes provision for the safe, secure and efficient operation of such
airport and for connected purposes.

5.6 While no specific day in April 1998 was mentioned, AA
targeted 1 April 1998 in their programmes, milestone lists and reports.

5.7 Towards the end of 1997, AA revised the target to the latter
part of the month, ie, on or about 29 April 1998.

5.8 On 13 J anuary 1998, the Government announced that the
new airport was going to open for operation on Monday 6 J uly 1998.
The decision to open the new airport on airport opening day (AOD)
was made by ADSCOM with the approval of the Chief Executive in
Council.

Section 2 : The Roles and Duties of the Key Parties

5.9 In order to understand how the decision to open the new
airport was made and the duties of each of the parties regarding the

43
problems that occurred on AOD, it is necessary to consider and
understand the roles played by each of the parties. The duties of most of
the parties are mainly based on contract and their roles are apparent from
the contractual relationship. There are only a few exceptions, which are
AA, ADSCOM and the New Airport Projects Co-ordination Office
(NAPCO), who were the key parties in the setting up of the new airport.
The roles and duties of these key parties and their relationship amongst
each other are examined here.

(a) AA

5.10 PAA was established on 4 April 1990 for putting through the
airport project, while other organisations were to be responsible for
various parts of the ACP. PAA was succeeded by AA in December 1995,
by virtue of the Ordinance.

5.11 AAs creation was heralded in the Memorandum of
Understanding of September 1991, which provides that an Airport
Authority should be established by legislation to be modelled as far as
possible on the Mass Transit Railway Corporation Ordinance. Under
the Ordinance, AA is a statutory corporation. It has the status like a
private company developing and running the new airport. The rationale
behind, which can be found in the Consultation Paper published by
Government in J anuary 1994 together with the Airport Corporation Bill,
is to enable AA to act in a more commercial manner than would be
possible for a Government department, with two benefits, namely,

(a) AA would be able to work more quickly and be better placed
to fast-track development to meet urgent development
timetables; and

(b) the level of direct funding support required from the
taxpayer that would otherwise be required if the airport was
to be run by Government or any of its departments would be
reduced, because AA as a commercial concern would be
able to borrow substantial amounts successfully without
requiring full Government guarantees of its debt, thus
reducing the direct equity funding required from

44
Government and reducing or eliminating the liabilities
involved for Government in guaranteeing debts.

5.12 When W36 Mrs Anson CHAN, the Chief Secretary for
Administration and Chairman of ADSCOM (the Chief Secretary), gave
evidence in the inquiry, she told the Commission the background of the
decision to establish a statutory corporation to build and run the new
airport. At the material time, Government considered very carefully
several options as to the best way of providing for the planning,
construction and operation of the airport and commissioned consultants to
advise on the institutional arrangements. In the light of the consultants
recommendations, having examined various options, and taking account
of the efficient and effective way that the Mass Transit Railway
Corporation, an independent statutory corporation, had operated the Mass
Transit Railway for some time, Government decided that the best way
was to go for an independent statutory corporation that would have
statutory responsibility for the planning, construction and operation of the
airport. A corporation that would be required to operate on sound
commercial principles, free of Government bureaucratic interference and
the need to adhere to Government regulations, would be in the best
position to deliver an airport within a shorter timeframe than a
Government department, in a more cost-effective manner, and would be
able to raise funds in the open commercial market, and thus keeping
public expenditure down to a minimum. The aim was not only to
minimise public expenditure, but more for providing Hong Kong with an
efficient, safe and secure airport that the community could be proud of.

5.13 In order to fully understand AAs role and responsibilities,
reference should first be made to the relevant provisions of the Ordinance.
Section 5(1) of the Ordinance stipulates the purposes of AA, as follows:

maintaining Hong Kongs status as a centre of international and
regional aviation, provide, operate , develop and maintain, at
and in the vicinity of Chek Lap Kok, an airport for civil aviation.

Under the same subsection, AA may provide at or in relation to the airport
such facilities, amenities or services as are, in its opinion, requisite or
expedient.

45

5.14 Section 6 of the Ordinance is also important. Section 6(1)
provides that AA shall conduct its business according to prudent
commercial principles and shall, as far as practicable, ensure that, taking
one year with another, its revenue is at least sufficient to meet its
expenditure. Section 6(2) further provides that AA shall, in
conducting its business or in otherwise performing its functions, have
regard to safety, security, economy and operational efficiency and the safe
and efficient movement of aircraft, air passengers and air cargo. The
only limitations and restrictions of AAs powers are provided for in
various sections of the Ordinance:

(a) AA is not to establish or operate meteorological service or air
traffic control service, or make any air service agreement or air
service arrangement with the government of any country or
territory outside Hong Kong, etc [s 8];

(b) The Chief Executive in Council may make regulations for
various purposes, the most important of which is for securing
the safe or secure operation, or the proper maintenance, of the
airport or for securing the safety of persons or a specified class
or description of persons who are within the airport area [s 18];

(c) The Chief Executive in Council may, if he considers the public
interest so requires, give to AA such directions (in writing) as
regards the performance of any of its functions as he considers
appropriate [s 20];

(d) DCA may, in consultation with AA, give a direction to AA in
order to discharge or facilitate the discharge of an international
obligation regarding civil aviation, etc [s 21];

(e) The Chief Secretary may require AA to execute works or
measures to ensure a risk of injury to persons within the airport
area due to the defective condition of the airport or any vehicle,
vessel, machinery or other plant or equipment in any place in
the airport area to be eliminated or significantly reduced [s 39];
and

46

(f) The Chief Executive in Council may make regulations
providing for the appointment by the Chief Executive of
persons to be inspectors for enforcement the regulations under s
18 or s 39 [s 38].

5.15 AAs affairs shall be under the care and management of a
board whose functions shall comprise such care and management [s 4]
but the board is at liberty, subject to certain exceptions, to delegate any of
its functions to any member or employee, including the Chief Executive
Officer [s 9].

5.16 According to the above statutory provisions, AA is a
statutory corporation having the purpose of providing, developing,
operating and maintaining a new airport with the objective of maintaining
Hong Kongs status as a centre of international and regional civil aviation.
It must carry out the purpose and objective in accordance with prudent
commercial principles but in conducting its business or performing its
functions, it must have regard to safety, security, economy and
operational efficiency and the safe and efficient movement of aircraft, air
passengers and air cargo. Save for the limitations and restrictions of its
powers by possible Government intervention, under the provisions of the
Ordinance set out in paragraph 5.14 above, the AA Board enjoys full
autonomy and can delegate all its functions to its employees and the
Chief Executive Officer (CEO), who was for all purposes of this report
W3 Dr Henry Duane Townsend. It is in this context that the relationship
between the key parties relating to the opening and operation of the new
airport should be examined.

5.17 Apart from AA, there is the Airport Consultative Committee
which is mainly responsible for gathering public opinion on matters
relating to the new airport and passing it onto Government. The
Commissioners have not found any of the subject matters of the inquiry
relates to the Airport Consultative Committee, and therefore its role and
participation are excluded for further consideration in this report.

5.18 According to the above-mentioned declared policy of
Government and the statutory provisions, AA would and should be run as

47
a commercial concern, with minimal bureaucratic interference or
intervention of the Government. Indeed, apart from matters that relate
to public expenditure and that may affect the public interest, in which
Government intervention may be possible by way of calling in an audit
by the Director of Audit or appointing an inspector to investigate AAs
affairs, the Ordinance confers full autonomy and independence on AA.

5.19 AA is structured with a Board and a management. The
Board, chaired by W50 Mr WONG Po Yan, has 15 members, consisting
of six ex-officio members and nine non-official members. The
ex-officio members are the Secretary for Economic Services, the
Secretary for the Treasury, the Secretary for Works, the Director of Civil
Aviation, the Chief Executive of the Hong Kong Monetary Authority and
the Director of NAPCO. The names of the 15 members in 1998 are set
out in Appendix V, and W3 Townsend, the Chief Executive Officer of
AA is among them. The AA management was headed by W3 Townsend,
and under him there were over 1,800 staff, notably for the purposes of the
inquiry W43 Mr Douglas Edwin Oakervee (Director of Project Division
(PD)), W44 Mr Chern Heed (Director of Airport Management Division
(AMD)) and W45 Mr Kironmoy Chatterjee (Head of Information
Technology Department). It is to be noted that W43 Oakervee was an
experienced engineer, W44 Heed is a seasoned airport manager and W45
Chatterjee is a well-established IT professional, and each of them is
assisted and was at all material times assisted by a number of experienced
professionals in the same field.

5.20 While AAs counsel accepted that AA had to closely
scrutinise the development of Flight Information Display System (FIDS)
and other systems and works in PTB which were to be provided, installed
and built by its own contractors and subcontractors, they submitted at
length that AA did not have any statutory duty but only administrative
function to monitor the activities of franchisees such as Hong Kong Air
Cargo Terminals Limited (HACTL). It was further submitted that AA
had no standing to scrutinise the work done by HACTLs contractors,
namely Murata Machinery (HK) Ltd. and Mannesmann Dematic AG
Systeme. Under the franchise agreement whereby AA granted a
franchise to HACTL to operate as a cargo terminal operator in the new
airport, HACTL was to report and to permit specified inspections by AAs

48
designated representatives. Where proactive monitoring was called for,
AA did not shrink from pouring all necessary resources into the task.
Counsel also argued that the provisions of section 6(2) of the Ordinance
did not impose duties on AA to ensure that there would be safe and
efficient movement of air passengers, aircraft and air cargo. The term
shall have regard to these subjects in section 6(2) only made them AAs
primary targets in operating the new airport. It was further argued that
given the respectability of HACTL and the apparently satisfactory nature
of its progress reports, verified so far as possible by observation of
activities on site, AA did enough by way of monitoring and should not be
criticised for failing to penetrate what might represent a misleading
picture presented by HACTL.

5.21 The Commission accepts that AAs duty under the
Ordinance may not give rise to a civil liability. Nonetheless it is a duty.
Section 2 of the Ordinance expressly defines functions as including
powers and duties, and as AAs function under section 5(1) of the
Ordinance is to maintain Hong Kongs status as a centre of international
and regional aviation, by providing, operating, developing and
maintaining an airport for civil aviation, there is a duty for it to have
regard to those primary objectives as set out in section 6(2). For the
purpose of this inquiry, it does not matter whether the duty is called duty
or whether it is called target purpose or objective. The duty is to have
regard to those objectives. Having regard means having appropriate and
sufficient regard, not merely having thought about an objective and then
forgetting it or not making much effort in having it carried out. These
objectives, the safe and efficient movement of passengers, aircraft and air
cargo, are fundamental elements of an efficient airport to which sufficient
weight has to be given. As far as air cargo movement is concerned,
whether it is sufficient to award a franchise to a reputable franchisee is a
matter of degree. Whether monitoring is required and the extent of the
monitoring required is also a matter of degree. In the present case,
where the franchise agreement makes provisions for the possible
termination of the franchise in case of unsatisfactory performance on the
part of the franchisee, AA must to a certain extent maintain its position
and power in having regard to the safe and efficient movement of cargo
in the new airport. If the provision of the services entrusted to the
franchisee fails, is AA entirely without responsibility? How about its

49
assurances given to Government that the new airport was ready for
operation on AOD? What about the publics expectation that the new
airport provided by AA for Hong Kong would be a world-class one, and
the new airports featuring in the reputation of Hong Kong as a whole?
While certain monitoring work had been performed over the construction
works of HACTLs SuperTerminal 1 (ST1) with the necessary
expertise, why should there be no expertise regarding the systems
development, installation and testing? What exactly happened was that
the monitoring work was only done partially, mainly over the
construction works but little over the systems. If, for example, a person
with expert knowledge of cargo handling systems had monitored the
testing of the Cargo Handling System (CHS), at least HACTL and AA
could have been warned of the absence of a sufficient throughput test for
the Box Storage System, part of CHS. Another aspect was that AA had
given assurances to ADSCOM that the new airport (which must be
inclusive of HACTLs services) would be ready on AOD. Since AA did
not monitor the systems of HACTL effectively or at all, AA should have
warned ADSCOM that there was no monitoring over the systems or that
AA possessed no expertise for such monitoring, for otherwise the
assurances would be defective and tend to be misleading. For this
partial monitoring and for the failure to warn ADSCOM when the
assurances were given, AA was in breach of its duties.

(b) ADSCOM

5.22 Government established ADSCOM chaired by the Chief
Secretary to make strategic and policy decisions regarding the ACP,
which included the project to build the new airport. Government, as the
initiator of the ACP which involved a colossal investment of public funds,
needed to take steps to ensure that these infrastructure works would be
carried out within budget and on time. Apart from the Chief Secretary,
the other members of ADSCOM are as follows:

Financial Secretary
Secretary for Economic Services
Secretary for Planning, Environment & Lands
Secretary for the Treasury
Secretary for Transport

50
Secretary for Works

5.23 ADSCOMs terms of reference were not stated in any
legislation, but can be found stated in slightly different terms in various
documents. In ADSCOM Paper 1/90 of February 1990, the terms of
reference of ADSCOM were stated as follows:

(1) to review the general progress of the new airport project and
associated works, including the transport infrastructure; and

(2) to resolve problems referred to it by policy secretaries.

The role of ADSCOM was also described in subsequent ADSCOM
papers:

(3) ADSCOM has the overall responsibility for establishing
policy, guiding the implementation of the ACP projects and
coordinating action taken by the Hong Kong Government
with regard to the Memorandum of Understanding
Concerning the Construction of the New Airport in Hong
Kong and Related Questions [para 3 of ADSCOM Paper
29/91 of August 1991];

(4) Since the drawing up of the AOR will run into different
policy areas, and since the smooth opening of the new
airport is essential to Hong Kong, ADSCOM is best placed
to be the overall monitor. The PAA/AA should submit
regular progress and funding reports through NAPCO to
ADSCOM. Should there be issues which, for reasons
beyond PAA/AAs control, are threatening to hold up the
CLK AOR, or matters which cannot be resolved at the
working level, NAPCO would in the first instance, refer
them to the relevant policy secretary for resolution at an
existing forum. If that fails, NAPCO would then escalate
the matter to ADSCOM for resolution. [para 20 of
ADSCOM Paper 45/95 of October 1995]; and

(5) Significant policy issues and matters affecting more than

51
one Policy Branch are subject to collective decisions at
ADSCOM which has the overall responsibility of overseeing
the smooth implementation of the ACP and coordinating
actions taken by the Hong Kong Government with regard to
the MOU. [para 2 of ADSCOM Paper 49/91 of December
1991]. (MOU means the Memorandum of Understanding)

In paragraph 6 of her statement to the Commission, the Chief Secretary
described the role of ADSCOM as follows:

As a policy group and co-ordinating body, ADSCOM has a
wide remit for the 10 mega infra-structural projects
constituting the ACP For these 10 projects, ADSCOM
provides an overall steer on issues with significant policy or
resource implications. ADSCOM oversees progress and
cost control but does not concern itself with matters of an
operational nature.

5.24 ADSCOM had the overall steering responsibility for the
planning and implementation of the 10 ACP projects, including the new
airport, on critical issues on policy and resources. It would also
intervene to resolve any impasse, if any, between Government
departments amongst themselves and between them and AA.

5.25 As AA is to be the builder and operator of the new airport,
all the duties and functions of the Civil Aviation Department (CAD) of
the Government regarding airport operations and management, save air
traffic control, which the department was administering at Kai Tak, were
to be transferred to AA upon the opening of the new airport. There are
yet many services required for the operation of the new airport that will
be performed by Government departments. For instance, CAD has to
deal with air traffic control, the Immigration Department has to perform
passenger immigration, entry and exit functions, the Customs and Excise
Department has to ensure duty clearance of imported and exported goods
and passenger belongings, the Police has to maintain public order and the
Fire Services Department has to ensure fire safety and protection. The
involvement of Government departments in the operation of the new
airport meant that their presence in the new airport would have to be

52
catered for, with premises to house their officers and facilities to enable
them to perform their functions. While direct communication with the
various concerned Government departments would be engaged in by AA,
Government set up NAPCO to ensure that there was full coordination
between the departments and AA, and to resolve any difficulty that might
be encountered in such coordination. NAPCO was also to monitor the
performance and progress of the new airport project and act as the
executive arm of ADSCOM.

(c) NAPCO

5.26 Apart from para 20 of ADSCOM Paper 45/95 (paragraph
5.23(4) above) in which reference was made to NAPCO, NAPCOs role
and responsibilities can also be found in other ADSCOM papers, as
follows:

(1) NAPCO serves as the executive arm of ADSCOM and is
responsible for the overall management of project
implementation and co-ordination. NAPCO gives advice
and guidance to departments in respect of the resolution of
interface issues and, ensures the timely completion of the
projects and that approved ACP policies and procedures are
followed. In conjunction with Finance Branch, NAPCO
will also exercise overall project cost control and
contingency fund management across the ACP projects
(except the airport, AR and WHC) [para 2 of ADSCOM
Paper 49/91 of December 1991]. (AR means the Airport
Railway and WHC means the Western Harbour Crossing)

(2) NAPCO would have 2 general areas of responsibilities
related to the implementation of the ACP projects. They
are the overall management of project implementation and
coordination and, the Governments public information and
community involvement programmes In addition, there
would be other areas of NAPCO responsibility related to the
overall coordination of the ACP projects such as a clearing
house for contract administration, project insurance,
mediation services and importation of labour. [para 5 of

53
ADSCOM Paper 29/91 of August 1991];

(3) NAPCO would provide overall programme management
services in coordinating and guiding the implementing
Departments and Agencies as required to resolve interface
issues, control overall ACP costs, assure timely completion
of the projects and carrying out approved ACP policies and
procedures. [para 9 of ADSCOM Paper 29/91];

(4) NAPCOs Programme Management responsibilities would
include:

(a) Establishment of ACP project procedures,

(b) Coordination of interface issues and resolution of
conflicting requirements between Departments,

(c) Oversee detailed coordination between Government
and other interfacing non-Government ACP projects,

(d) Review of project scopes and budgets to assist
Finance Branch in maintaining budget control,

(e) Development and up-dating of a Baseline
Implementation Plan, trend programmes and
expenditure forecasts,

(f) Monitoring and control of scope, cost and
programme,

(g) Recommending corrective actions and expediting
critical decisions,

(h) Recommendation on the allocation of financial and
staff resources,

(i) General review and coordination of contract
document formulation, contract administration and

54
construction management,

(j) Provision of technical specialists and administrative
support on an as needed basis, and

(k) Other duties as directed by the Chief
Secretary/ADSCOM.
[para 11 of ADSCOM Paper 29/91];

(5) On the government side, at least 16 departments have been
identified as likely to play some role in preparation for AOR.
Most of the departments have no in-house programming
capability. NAPCO will therefore assist departments in the
preparation of their programmes while they deal directly
with the PAA/AA over the planning for AOR. NAPCO
will however assist in conflict resolution between the
PAA/AA and Government departments. [para 21 of
ADSCOM Paper 45/95 of October 1995];

(6) At present, no single budget has been assembled for the
AOR although the PAA has identified a small sum for this
purpose which is part of their Head Office budget for
1995/96. Government departments will presumably rely
on their departmental budgets for this purpose, which are
outside the scope of the $158.2 billion ACP works budget.
NAPCO will continue to monitor PAAs overall airport
development budget which includes provision for the AOR,
and will assist departments to identify AOR requirements.
[para 22 of ADSCOM Paper 45/95];

(7) As executive arm of ADSCOM, NAPCO will monitor the
progress and funding position of the AOR, liaise with
PAA/AA on problem areas (whether on programme, cost or
interface issues) and refer matters to policy secretaries and
ADSCOM as appropriate for speedy resolution. [para 23 of
ADSCOM Paper 45/95];

5.27 From all the documents referred to above, it can be seen that

55
NAPCO had a number of roles to play with various responsibilities
relating to ACP. Putting them in the proper perspective within the scope
of the Commissions inquiry, merely relating to the new airport,
NAPCOs main role and responsibilities were to

(a) act as the executive arm of ADSCOM;

(b) monitor the progress of the new airport project; and

(c) coordinate the interface between Government departments
and AA.

5.28 In her letter dated 20 November 1995 to the concerned
Government departments and policy bureaux, the Chief Secretary stated:
The Provisional Airport Authority (PAA)/future Airport Authority (AA)
will be responsible for drawing up and implementing an Airport
Operational Readiness (AOR) programme. It will co-ordinate with all
agencies involved to consolidate and agree on a comprehensive list of
tasks ahead, and manage the implementation of the entire AOR
programme to ensure that the airport is ready, in all respects, to open on
schedule.

5.29 About 16 Government departments were identified as having
something within their jurisdiction and ambit that related to the new
airport, and NAPCOs task under (c) in paragraph 5.27 above is to
coordinate between these departments and AA. In case NAPCO could
not resolve any interface problems, it would resort to ADSCOM for
assistance.

5.30 NAPCOs roles and responsibilities under (a) and (b) in
paragraph 5.27 above are closely connected. In ADSCOM Paper 45/95
of October 1995, it was stated since the smooth opening of the new
airport is essential to Hong Kong, ADSCOM is best placed to be the
overall monitor and that As an executive arm of ADSCOM, NAPCO
will monitor the progress and funding position of the AOR (see
paragraphs 5.23(4) and 5.26(7) above.) It was because NAPCO was the
executive arm of ADSCOM that NAPCO was responsible to monitor the
progress of the works for making the new airport ready for opening.

56
However, NAPCOs role was that of an overall monitor, as it was
performing duties on behalf of ADSCOM as the latters executive arm.
W33 Mr KWOK Ka Keung, the Director of NAPCO since J anuary 1998,
told the Commission that NAPCOs monitoring over AA was one on a
high level because AA was responsible to operate the airport under statute.
NAPCOs monitoring was proactive to the extent that was practicable but
not being offensive, so as not to spoil the good relationship at the working
level between NAPCOs staff and AAs personnel. NAPCO should not
operate above the law and interfere with AAs work. Further, AA had a
team of 550 staff while NAPCO only had 38 professional officers many
of whom were deployed for duties other than the new airport. NAPCO,
according to W33 Kwok, would not look for detailed operational
procedures and contingency plans that AA might have, for otherwise that
would be beyond NAPCOs mandate.

5.31 W36 the Chief Secretary testified that the underlying policy
of deciding on AOR was to have the new airport operating safely,
securely, efficiently and smoothly. ADSCOM had never resiled from
those criteria through the course of its examination of the readiness of the
new airport to open on AOD, from the time before the decision was made
right up to AOD. She described the role of NAPCO as one of a critical
observer over AA regarding the development of the new airport and AOR.
A critical observer is one that observes and critically assesses what one
has observed. Notwithstanding the statutory functions of AA to plan,
develop, operate and manage the new airport, Government was the body
that decided on AOD, as that involved the interest of the Hong Kong
community as a whole. Before the decision on AOD was made,
ADSCOM required NAPCOs assessment of AOR, and based on that
assessment also asked AA to provide answers to various matters of
concern that might affect AOR. As referred to above, AOR means that
the new airport would have to operate safely, securely, efficiently and
smoothly. After the decision was made in early J anuary 1998,
Government still maintained NAPCO as a critical observer of the
progress in order to ensure the four criteria would be met.

5.32 The Commissioners find that NAPCO was rather in the
position of an interested and critical observer instead of a supervisor,
controller or auditor, in that it would observe the progress of the works,

57
assess the progress critically, give comments on the progress, sometimes
provide advice when sought, while not giving nor able to give orders,
directions or instructions that may affect AAs autonomy as conferred by
the Ordinance. It was because of NAPCOs position as ADSCOMs
executive arm under (a) in paragraph 5.27 that NAPCO reported the
progress of the works, and in particular issues critical to AOR to
ADSCOM and kept ADSCOM apprised of matters related to those issues.
Yet, while NAPCO might advise ADSCOM of such issues and NAPCOs
own views on them, neither it nor ADSCOM would act in any manner to
interfere with AAs autonomy, which was to establish and run the new
airport in accordance with prudent commercial principles.

5.33 ADSCOMs overseeing the progress of the new airport,
through its executive arm NAPCO, must be seen in the light of AAs
independence and autonomy and also that ADSCOM as well as NAPCO
were not involved nor were empowered to get involved in the day-to-day
running of the new airport project, which was entirely a matter for AA.
Apart from matters of public interest and international obligations, in
respect of which power is conferred by various sections of the Ordinance
referred to in paragraph 5.14 above for Government to act or intervene,
ADSCOM and NAPCO had no authority to deal with the development,
maintenance and management of the new airport which were left entirely
with AA. The evidence received by the Commission also bears out this
situation. Where NAPCO identified a delay or a problem with the
progress of the new airport project affecting AOR, it could and did draw
it to the attention of AA since no one other than AA could either
accelerate the progress or resolve the problem. NAPCO could not direct
AA to do either, let alone seek to accelerate the programme or resolve the
problem on its own accord and with its own resources. Where the delay
or problem persisted and thus had a bearing on AOD or AOR, NAPCO
would raise it with ADSCOM so that the matter could be resolved at that
level. Where the problem as raised at the level of ADSCOM was
addressed by AA and an assurance given, unless the assurance was
blatantly incorrect or untrue, it would be unrealistic and difficult for
ADSCOM or NAPCO to challenge it granted the limitation of ADSCOM
and NAPCO in not being closely and directly involved with the project.
Challenging such an assurance would also require employment of
resources. Had either of these two bodies attempted to do so, it would

58
involve at least two undesirable or illegitimate consequences, namely,
having to lay out expenses from public funds for the purpose of falsifying
AAs assurance, which might or might not succeed, and interfering with
AAs independence. Either of these consequences would certainly have
generated serious criticisms and could hardly be justified in view of the
fact that trusting reliance should be placed on AA and its numerous
professional officers and specialist contractors in handling the project
ably and that the autonomous status of AA as entrenched by the
Ordinance must be respected.
5.34 Counsel for the Commission propositioned that ADSCOM
was in the position of a de facto AA Board in respect of matters which
were AOR critical as it had exercised overriding control over the AA
management. From the evidence, the only true intervention by
ADSCOM was its decision on AOD. Even on the significance of having
a standby FIDS as a fall-back in case of the failure of the main FIDS, it
only recommended to AA to have it commissioned, in view of the
instability of the main FIDS throughout the tests that had been gone
through up to March 1998, although in the notes of the ADSCOM
Meeting on 21 March 1998, this recommendation was termed a final
decision to be made by ADSCOM. W46 Mrs Elizabeth Margaret
Bosher said that the duty of AA towards ADSCOM was primarily to
report on progress, and ADSCOM could insist on the quality of the
reporting. This view is consistent with section 19(1) of the Ordinance,
which reads:
The Authority shall supply the Governor with such information
relating to any of the Authoritys affairs as he may from time
to time require.
5.35 Nonetheless, from the evidence, it is clear that sometimes the
directive word instructed was recorded as used when ADSCOM wished
the AA management to do something, and W45 Chatterjee, W43
Oakervee, and W44 Heed thought that ADSCOM was an overriding
body or one whose wishes should be seriously taken into account. That
was basically due to the fact that ADSCOM represented Government and
in turn represented the public in seeing that Hong Kong should have a
world-class international airport. In the opinion of the Commissioners,

59
however, this does not make ADSCOM a de facto Board of AA, as
counsel for the Commission put it.
5.36 AA is a statutory corporation empowered to provide and
operate the new airport. All the necessary functions for such purposes
are vested by the AA Ordinance in the AA Board which is allowed to
delegate to the AA management. There is one thing that is not
mentioned in the Ordinance, which is who is to make the decision on the
opening of the new airport for operation. That was apparently reserved
for Government and the decision fell to be made by ADSCOM. This
was perhaps foreshadowed by ADSCOM Paper 29/91 which reads:
A number of ACP projects are, or are intended to be, assigned
to a private sector agency for implementation. They include
the PAA (Provisional AA) for CLK Airport, MTRC for the
Airport Railway and, a Franchisee for the Western Harbour
Crossing. In each of these cases the particular Agency will be
independently responsible for the financing, as well as the
development and implementation of the projects. Their
responsibilities relative to the ACP projects will be detailed in
the enabling legislation and agreements with Government. In
overall (macro) programming terms, however, there is a need
for the Government at ADSCOM level to exercise on-going
monitoring and, as and when necessary, in the overall interests
of Hong Kong, to take decisions affecting those agencies. In
that connection, NAPCO will need to maintain close liaison
with them.
The decision on AOD was in fact made by ADSCOM after very careful
and close consultation with AA. The purpose of the consultation was to
ensure that the new airport would be ready on the opening date to be
nominated by ADSCOM. As a responsible Government which
ADSCOM represented in this matter, the opening date must be based on a
cautious and reasonable assessment that the new airport would be ready
when it was to open for operation. For that, ADSCOM continued to
meet to satisfy itself that the objective of a safe, secure, efficient and
smooth airport would be achieved on AOD, and NAPCO carried on with
its monitoring exercise over AOR issues. All these activities and the

60
respect shown by the AA Board and management towards ADSCOMs
wishes do not establish a reasonable basis and should not be properly
used as such for attributing to ADSCOM the position of a de facto AA
Board, nor to increase the duties of Government or ADSCOM so that it
guaranteed that the new airport would be operational ready on AOD. To
consider otherwise would ignore the Ordinance which entrusts and
empowers AA and nobody else with the functions of providing and
operating the new airport.
5.37 Despite the presence of a number of Government
representatives on the AA Board, there is no evidence that Government
attempted to control AA or that it used the official members on the Board
to monitor the operational readiness of the new airport. Indeed, on the
proper date for airport opening, official members of the Board abstained
from voting for the Board to make a recommendation to ADSCOM.
5.38 Counsel for the Commission relied on Article 128 of the
Basic Law to argue that Government had the duty to ensure the provision
of an operational and efficient airport, including an efficient air cargo
handling service. Article 128 provides:
the Government of the Hong Kong Special Administrative
Region shall provide conditions and take measures for the
maintenance of the status of Hong Kong as a centre of
international and regional aviation.
5.39 The Commissioners do not accept that Article 128 imposes a
duty on Government as argued by counsel. On the other hand, the
enactment of the Ordinance, that established AA as a statutory
corporation with the functions of providing and operating the new airport
with the objectives of maintaining Hong Kong as an international and
regional hub of civil aviation, can and should properly be considered as
part and parcel of the conditions and measures required by Article 128
to be taken by Government. In establishing AA through the enactment
of the Ordinance by the Legislative Council, Government had also taken
into account that as a statutory corporation, AA would eliminate or reduce
bureaucratic constraints and delays in developing and running the airport
and enhance the opportunities and efficacy of raising loans for those

61
purposes. These two objectives could hardly be achieved if Government
itself were to build and operate the new airport.
5.40 Nonetheless, counsel for the Commission were correct in
pointing out that Government must see to it that AA would discharge its
duties. This is because that Government had decided on AOD, and by
this decision alone, Government should not only have assessed that AOD
was a date that the new airport would be ready for efficient operation, but
should also be satisfied that AA would discharge its duties of providing
an efficient airport for Hong Kong on AOD after the decision on AOD
was taken. In the opinion of the Commission, however, this role of
Government through ADSCOM was subject to two limitations: the
importance of observing the law and the proper use of public funds, both
of which are vital to the public interests. As AA is the statutory
authority in the form of a corporation in charge of the affairs of the new
airport, any undue interference by Government would be an unjustified
and even unlawful usurpation of AAs statutory functions and obstruction
of AAs autonomy. The importance of Government and everybody else
respecting the law does not need any explanation. It is also important
that public interest demands that public funds should not be expended
except for good reasons. ADSCOMs overall monitoring of AAs work
and the progress of the airport project was performed through NAPCO,
the executive arm of ADSCOM. The overall monitoring was to enable
Government to keep an eye on AA in discharging its functions under the
Ordinance for the purpose of satisfying Government that there would be
an efficient new airport on AOD. If the monitoring role of NAPCO
were to take a full audit of the development of the new airport or to
supervise AAs performance from time to time, NAPCO would have to
employ a large number of professionals and experts to examine every
step taken by AA and its contractors. That would undoubtedly duplicate
the efforts and expenses that AA was incurring, and would therefore be
unjustifiable and unnecessary in the interest of preserving public funds.
Thus, it is reasonable for NAPCO not to employ resources to examine the
operational and procedural details that AA planned for the running of the
new airport, insofar as AAs plans produced for NAPCOs scrutiny were
satisfactory under critical examination.
5.41 As W36 the Chief Secretary said in evidence, NAPCOs

62
monitoring role was that of a critical observer which, the Commissioners
consider, struck a fair balance between the conflicting considerations.
On the one hand, Government had to be satisfied in the interest of the
public that on AOD there would be an efficient new airport. On the
other hand, Government had to respect the law by not unduly interfering
with the statutory autonomy of AA and also had to protect public funds
from being unnecessarily spent for fully auditing the work of AA and
delve into its procedural and operational details when AA had its own
large professional teams and consultants.
5.42 This role of a critical observer of NAPCO is, however, not
limited to evaluating the progress of the development of the new airport
through AAs reports on various AOR critical issues, as counsel for
Government argued, but should cover the examination of the progress by
NAPCOs own professional staff. This role of NAPCO is not merely
the Commissioners opinion, but is borne out by the evidence. NAPCO
did have its professional staff from Government departments and from
International Bechtel Company Ltd. (Bechtel) which was a company of
airport consultants that it employed, very often on site, to monitor the
progress of the construction and system development works in PTB and
ST1, not merely relying on AAs reports on the progress.

Section 3 : Communication and Coordination Channels

5.43 The new airport was only one of 10 major infrastructure
projects comprising the ACP under NAPCOs monitoring responsibility.
There was a distinction in NAPCOs role in relation to the Government
versus non-Government ACP projects, the latter including the new airport.
In regard to the former, Government had direct involvement. The
Government works agents were subject to procedures and administrative
controls imposed by NAPCO, as well as ACP conditions of contract
authored by NAPCO. NAPCO was in a strong position to recommend
remedial measures when delays or problems were identified. For the
new airport, AA being an independent statutory corporation managed its
contracts and works directly under its own professional management.
NAPCOs role was limited to reviewing AAs plans and programmes,
monitoring overall progress on site, and flagging up problems or potential

63
problems to be followed up by the Project Manager of NAPCO, Director
of NAPCO and ADSCOM in consultation with the AA management.
NAPCO was not a party to AA contracts and franchises, and had no direct
say in the performance of the AA contracts. Although AA provided
information to NAPCO for overall monitoring purposes, NAPCO did not
have unrestrained access to all AA contract documentation. Further,
NAPCO only had a limited coordination and field staff to carry out
monitoring assignments, including only five field staff who were
primarily responsible for monitoring Government works at the new
airport.

5.44 For the purposes of the inquiry, the Commission
concentrated on the roles and responsibilities of AA, ADSCOM and
NAPCO in respect of FIDS and HACTL, the two major problems that
have been identified as having the greatest impact on the operation of the
new airport on AOD. FIDS is the computer system installed in PTB for
the provision to users of information about flights coming in and going
out of the new airport, and HACTL is the main cargo handling franchisee
that is responsible for about 80% of all the cargo that would be moved
through the new airport. After Government through ADSCOM had
decided and announced 6 J uly 1998 to be AOD, NAPCO continued to
play a role in relation to AOR. NAPCO acted as a focal point for all
Government departments involved in the AOR process, and monitored
AA in its planning and implementation of the overall AOR programme.

5.45 Regarding HACTL, NAPCOs role was predominantly
limited to monitoring the monthly reports produced by HACTL for AA in
accordance with their Franchise Agreement; for other franchisees,
NAPCO relied on AAs own monthly reports. In the nine months or so
prior to opening, however, NAPCO became involved in detailed
exchanges with AA and with HACTL in respect of HACTL, on account
of the serious delays being experienced with the Government facilities
and systems in ST1.

5.46 To discharge its functions, NAPCOs Airport team was
divided into working groups, each headed by an Area Manager, who
reported to the Chief Coordinator, Mr J ohn Lloyd Smith. The areas
covered by the working groups included:

64

(a) AA Building Works, including Government entrustments;

(b) directly funded Government works, including stand-alone
buildings and entrustments to Franchisees;

(c) Airport Systems; and

(d) AOR.

5.47 The entirety of the working groups was made up of up to 16
staff, nine of whom were from Bechtel. The responsibilities of Bechtel
were contained in the Consultancy Agreement No. CE 85/95 between
Bechtel and Government, dated 1 February 1996. In accordance with
this Agreement, Mr Tudor Walters, who became the Consultant Project
Manager of NAPCO, was directly responsible to the Director of NAPCO.
On technical matters, Mr Walters generally advised ADSCOM directly at
its regular meetings and became directly involved in the discussion of
such matters as they arose with other parties at ADSCOM. He was in
attendance at almost all ADSCOM meetings between 1 February 1996
and 6 J uly 1998, providing his assessments of the situation and
recommendations from time to time. The assessments and
recommendations were based on his judgment, his perusal of relevant
documentation and discussions held with many of the directly involved
parties. During the nine months prior to AOD, the AOR working group
was augmented by the addition of W32 Mr J han Schmitz of Bechtel as
the Deputy Consultant Project Manager who led the AOR group up to
AOD. As from J anuary 1997, three of the Bechtel staff were based in
CLK to specifically monitor the progress of (a) fit out works in PTB, (b)
Government Entrusted Works in franchisee buildings and (c) systems
installation, integration and testing.

5.48 The following were routinely produced as part of NAPCOs
airport project management process:

(a) Weekly Situation Reports on Key Issues and Critical Items
prepared by NAPCO for ADSCOM circulation (Weekly
Situation Reports).

65

(b) Weekly Site Reports produced by Senior Engineers based at
CLK to the Chief Co-ordinator of NAPCO(Weekly Site
Reports).

(c) Detailed Fortnightly Reports (produced roughly over the last 12
months before AOD) on Airport Key Issues and Critical Items
for discussion at the NAPCO Directorate Meeting
(Bi-weekly CLK Reports).

(d) ACP Monthly Progress Reports, prepared on the basis of routine
reporting from work agents and NAPCOs overall schedule,
budget and interface management perspective (ACP Monthly
Progress Reports).

(e) Detailed Reports and Presentations on Airport Critical issues and
AOR prepared by AA with NAPCOs comments for
ADSCOMs consideration.

(f) Minutes of technical meetings with AA (and HACTL).

(g) Key correspondence with the AA, including letters from Mr
Walters to AAs Project Director, W43 Oakervee.

5.49 NAPCO was ADSCOMs executive arm. The Director of
NAPCO was a member of the AA Board. The Secretary for Works, a
member of ADSCOM, was also a member of the AA Board. While
NAPCO was performing overall monitoring of the airport project,
amongst the ACP projects, the Works Bureau (WB)(then the Works
Branch), sometimes also monitored those AAs works which gave rise for
concern. For the purposes of such monitoring, reports were prepared by
AA, WB, NAPCO and its officers. The reports or a gist of them were
submitted to ADSCOM. After receiving these reports, and other papers
for ADSCOM prepared by NAPCO, ADSCOM had meetings on the new
airport about once every two weeks. These reports were the main
channel of communication between ADSCOM with the other bodies.
There were the following reports:

66
(a) NAPCOs senior engineers on site submitted Weekly Site
Reports to the Chief Coordinator of NAPCO.

(b) Bechtels professional staff seconded to NAPCO submitted
Bi-weekly CLK Reports to NAPCO.

(c) AA submitted the following reports to NAPCO, namely, draft
ACP Monthly Progress Reports, monthly progress reports,
monthly construction reports and HACTLs ST1 monthly
progress reports.

(d) From (a), (b) and (c) above, NAPCO submitted to ADSCOM
Weekly Situation Reports, ACP Monthly Progress Reports,
ADSCOM Papers, Chairmans Briefs and other documents
prepared by NAPCO.

(e) AA submitted ADSCOM Papers and other documents
prepared by AA to ADSCOM.

(f) WB submitted to ADSCOM a Situation Report on AOR.

The channel of the documentary communication can be seen in
Appendix VI.

5.50 There was no direct communication between Government
bodies and HACTL, save on matters connected with the Government
entrusted works to be carried out in ST1. However, apart from
submitting the reports and papers to NAPCO and ADSCOM referred to in
sub-paragraphs (c) and (e) of the preceding paragraph, AA also allowed
NAPCO staff to participate in technical meetings between AA and
HACTL.

5.51 Within AA, there are various divisions under W3 Townsend
as the CEO, each responsible for a certain area of work. An
organisation chart of AA is at Appendix VII. There does not appear to
be anything wrong with the structure of the organisation within AA.
Communication within AA should have little problem as the staff of all
the relevant divisions, notably PD and AMD, for the building and running

67
of the new airport were at CLK. Their coordination was, however,
problematic, and will be briefly alluded to in the next section and more
fully dealt with in Chapter 17.


5.52 Communication and coordination between various divisions
in AA were not satisfactory. That was noted by NAPCO staff who were,
for the purposes of monitoring, at the site in CLK.

5.53 The deficiency of coordination within AA was noted by
NAPCO in various records. In ADSCOM Paper 34/97 by NAPCO
dated 19 September 1997, it was stated that NAPCO found that
coordination within the AA itself, particularly between AMD and PD and
the Commercial Division, as well as coordination and cooperation
between AA, its business partners, Government and all others required
intensified attention and immediate improvement. The coordination and
cooperation between AMD and PD was particularly important from about
this time, as the new airport was transitioning from the construction stage
to the operation stage, the responsibility of PD in relation to the
construction and system works was in the course of being handed over to
AMD. AMD was eventually to use the works and systems developed
under the auspices of PD, and AMD had to operate the services and
facilities so provided for the purpose of running the new airport.

5.54 The point was made in the 170
th
ADSCOM meeting on 20
September 1997: AMD should be in the driving seat of the airport
project at this point in time, but because of the personalities involved, it
was being pushed round parameters set by PD and had yet to gear itself
up. The Deputy Director of NAPCO advised that W3 Townsend
should, but did not, quickly and firmly resolve this problem.

5.55 The notes of the ADSCOM special meeting on 7 November
1997 also recorded DCA as saying that he had no faith in the top
management of AA. The project was driven by the Project Director
W43 Oakervee who always tried to bulldoze his way through. W3
Townsend was not in control and the organisation was not functioning as

68
it should.

5.56 NAPCO also recorded the lack of cooperation from AA in its
Weekly Site Report of 7 March 1998. NAPCOs attempts to find out
what was going on regarding systems integration during the period were
continually thwarted because AA staff were warned not to say anything.
It was not surprising that NAPCO started to distrust AA. NAPCO
further reported that AA would claim that, all the scheduled tests were
completed; however, the reality was that the system could not yet
display flight information at a number of locations.

5.57 In its Weekly Situation Report of 1 May 1998, NAPCO
reported that it had still not received the AAs quantification of additional
requirements for the contingency plan in case of FIDS failure, as
promised.

5.58 Another week passed by, NAPCO again reported that AA
claimed to have corrected many of the FIDS critical software issues and
resolved the Societe Internationale de Telecommunications Aeronautiques
(Common User Terminal Equipment) /FIDS interface problems with
implementation at site continuing. However, a number of software
issues, which AMD stated as critical, were still outstanding and this
raised concerns on AAs ability to establish Day One operating scenario.
AA was developing the contingency FIDS with General Electric
Company (Hong Kong) Ltd and Hong Kong Telecom CSL
Limited(HKT) but the time available for development was short.
Work to interface FIDS with other systems such as Airport Operational
Database (AODB), Baggage Handling System, etc continued and
updates to AODB software was due in mid May. NAPCO had been
chasing AA but had still not received its quantification of additional data
transfer requirements [NAPCO Situation Report, 8/5/98].

5.59 In the ADSCOM Chairmans brief prepared by NAPCO and
at the 183
rd
meeting of ADSCOM, both of 22 may 1998, NAPCO pointed
out that by opening, the airport systems would largely operate on
standalone mode. It was clear from the AA report that lots of integration
were still underway and programmed for completion by the end of May
1998. ADSCOM had been assured that systems existed for manual data

69
transfer. However, as most systems had to be operated on a standalone
basis, more staff, procedures, etc, had to be organised. The
quantification of what this involved in terms of equipment, staff, changed
procedures, training, etc which NAPCO had been after for months had
yet to be forthcoming from AA. In the Summary of Critical CLK Issues,
dated 19/6/98, NAPCO continued to state that the demonstration of the
viability of workarounds, schedule and procedures of installing
enhancements, system status etc were all expected in a detailed report
which was still not yet received. NAPCO had yet to receive from AA
the quantification of additional data transfer requirements under the
contingency scenario.

5.60 As late as May 1998, the coordination between AMD and
PD still caused concern. In the Weekly Report by NAPCOs Mr David
Thompson for the week ending 23/5/98, he reported that in order to
accommodate the new back up FIDS, AMD needed to have some more
workstations, without which there would be problems for system
development and training functions. In answer to a NAPCO question,
concerning the reason why five additional workstations had not simply
been purchased, it appeared that PD was not willing to spend money and
AMD did not have access to funds.

5.61 There was also a coordination problem regarding the testing
of Government entrusted works. In a memorandum dated 28/5/98 from
W33 Kwok it was noted that the continuing delays in testing and
commissioning of CAD systems were the result of ongoing AA
installation, testing and commissioning problems with the AA primary
systems. Thus, until the primary AA systems were fully functional and
operational, CAD systems which were dependent upon the AA master
system could not be adequately tested or commissioned.

5.62 W31 Mr J ames WONG Hung Kin, the Project Manager of
NAPCO, gave evidence before the Commission about coordination
between AA and NAPCO. He said that the very detailed internal project
reports prepared by AA were originally only supposed to be available to
the AA Board members. That practice was changed in middle of 1996.
After that, AA was much more open to Government and shared with
NAPCO its internal reports. From those working level reports, NAPCO

70
staff on the site knew a lot more about the true picture in addition to
having day-to-day contact with AAs working level staff. The
relationship gradually improved a lot, particularly towards the end of the
project. In the half year before AOD, AA was quite open towards
NAPCO by allowing NAPCO staff to take part in the site acceptance tests
and to visit Interface House which was previously quite closed to
outsiders, including NAPCO. W31 Wong said that towards the end of
the project, NAPCO generally had quite a good grasp about the progress
of a wide spectrum of the AA works.

5.63 Regarding the structural arrangement of AAs organisation in
the development of the new airport, W51 Mr J ason G YUEN, the airport
expert appointed by the Commission, was of the view that consultants
should have been employed within AA to oversee the works that were
performed by contractors, so that they would monitor the works closely
and report to the AA Board direct, apprising it of their views as a source
independent of PD to assist the Board in assessing the quality and
progress of the works. The situation in fact was that NAPCO as a
separate and independent monitor was monitoring the works and
reporting to ADSCOM, an outside and higher authority, with the result
that the AA Board could not efficiently carry out its statutory task of
developing the new airport. However, the lack of outside consultant to
advise the AA Board, as opposed to the AA management, albeit desirable,
is not established by the evidence as a cause to the problems on AOD and
the Commission therefore opines that this subject should not be a matter
for serious criticism. The Commissioners will return to this matter in
the subsection dealing with the AA Board in Chapter 17. What
contributed to causing the problems was that AA did not possess any
expertise or employ any in examining the progress and effectiveness of
HACTLs systems which formed part of the AOR programme. Had the
AA Board or management retained consultants to monitor the
development or at least the testing and commissioning of HACTLs
systems, its assessment of the systems readiness would have been on
sound basis.

5.64 Both W36 the Chief Secretary and Chairman of ADSCOM
and W33 Kwok maintained that NAPCO had sufficient resources to
discharge its monitoring functions regarding AOR. In relation to the

71
identified critical issues of FIDS and HACTL, apart from the
Government staff that were assigned to it, NAPCO had the professional
assistance from the Bechtel personnel who were experienced in airport
matters. NAPCO also had professional staff on site to observe the
progress of the works, which basically means observing the progress of
construction works of both PTB and ST1 and the Express Centre. On
the other hand, NAPCO had Bechtel to monitor FIDS and help it
understand what was going on with FIDS. As far as HACTL was
concerned, NAPCO did not have contractual relationship with HACTL
which was putting up ST1 and the Express Centre and installing a CHS
which it was developing in these buildings. CHS includes the
equipment and computer systems. NAPCOs personnel could visually
look at the completeness of the equipment installation but the
development and readiness of the computer systems was not visible, at
least not visible to the untrained eye. If NAPCO were to inquire into
HACTLs systems, as opposed to merely observing the state of the
construction works of HACTLs premises and the installation of the
equipment, it would be without any contractual basis and might be
affecting HACTLs proprietary interest in the systems. NAPCO
therefore only relied on AAs monitor of HACTLs systems since it was
AA that had to ensure AOR. NAPCO did not after all have the expertise
in HACTLs CHS and it as well as ADSCOM assumed that AA must have
such expertise in order to discharge its functions over HACTL relating to
AOR. As HACTL and AA had never raised any concern or doubt
about the operational readiness of HACTLs systems, NAPCO examined
the progress of the construction works of HACTLs buildings and the
installation of the equipment to assess the degree of HACTLs operational
readiness, based on the fact that CHS was modular in design, in that each
module could work independently and the cargo handling capacity
depended on how many of the modules would be ready for operation on
AOD. NAPCO assumed that the degree of completion of the buildings
and equipment would be tantamount to the proportion of cargo handling
capacity being ready.

5.65 The Commissioners accept that NAPCOs relationship with
HACTL was different from that between AA and HACTL which affected
NAPCOs role of monitoring over HACTL relating to AOR. HACTL
was a franchisee of AA and had contractual obligations owed to AA,

72
which enabled AA to oversee whether HACTLs CHS would be
operationally ready for AOD. On the other hand, NAPCO could only
monitor HACTL indirectly through AA, although when NAPCO was
overseeing the progress of the Government works at the HACTL
premises, it would also take the opportunity of observing the progress of
the physical construction and installation.
5.66 The Commission opines, however, that in performing its
functions under the role as found by the Commission, NAPCO failed in
two aspects. First, it should have inquired with AA whether it had the
necessary expertise in monitoring HACTLs progress relating to the
installation, testing and commissioning of ST1s 5-level CHS equipment
and systems, but it did not do so. Secondly, it should have checked
whether AA had plans and contingency measures and should have had an
overall assessment whether such plans and measures were adequate in
view of the then prevailing circumstances. As a corollary, NAPCO
should also examine if AA had an overall risk assessment.

5.67 These duties on the part of NAPCO could have been easily
discharged without draining too many resources. Regarding the first
duty, NAPCO should simply ask a question of AA. There was a missing
link in NAPCOs monitoring of HACTLs AOR through AA, in that AA
did not have sufficient expertise in assessing HACTLs systems in its
CHS. NAPCO never inquired or ascertained whether AA had the
necessary expertise in monitoring HACTLs systems, but merely relied on
AAs assessment and HACTLs report to AA that the system was ready.
Merely assuming that AA had the necessary expertise may, in the opinion
of the Commissioners, be acceptable as far as ADSCOM was concerned,
for it was relying on NAPCO to do the monitoring work and advise it, but
the same is not acceptable vis--vis NAPCO. NAPCO had the
monitoring and advising functions to discharge towards ADSCOM.
NAPCOs assumption that AA was qualified to monitor HACTLs
systems, on which NAPCO did not possess any expertise, and its failure
to inquire if in fact AA was so qualified rendered risky NAPCOs reliance
on AAs assessment and reflected a lack of attention to HACTLs systems
as compared with the attention that NAPCO paid to FIDS. In this
respect the Commission finds that NAPCO should be responsible for
failing to discharge its functions fully to ADSCOM. Had NAPCO

73
coordinated in a better manner with AA, NAPCO should also have asked
this very pertinent question, and for this failure in communication,
NAPCO is responsible. Having said that, it would not be reasonable to
put too much blame on NAPCO relating to the failure of ST1, for that
would primarily be the responsibility of HACTL who should provide the
necessary cargo handling facilities on AOD as it had assured from time to
time. For HACTLs responsibilities, please see Section 4 of Chapter 14.
AA was also responsible for failing to ensure that HACTL was ready to
provide the necessary cargo handling facilities on AOD. AA should
have retained experts in the field to monitor if CHS would be
operationally ready, and they are responsible for failing to do so, instead
of merely relying on HACTLs words. Had NAPCO asked AA the
question and found that AA did not have the expertise in understanding
CHS being installed in ST1, NAPCO should and could warn ADSCOM
of the deficiency enabling ADSCOM to decide whether it should urge AA
to employ the necessary expertise.
5.68 For the second duty, the documentary evidence shows that
NAPCO did request and press AA for contingency plans and measures,
contrary to W33 Kwoks oral testimony. However, what NAPCO failed
to do was to examine such plans and measures critically and warn
ADSCOM (which warning would undoubtedly be passed onto AA) that
there was no global contingency plan commensurate with the prevailing
situation and possible risks identified. This approach would not have
involved NAPCO in delving into AAs operational details, save where a
critical examination of AAs plans and measures unearthed evidence that
AOD was at risk. For instance, the risk of a possible failure of FIDS on
AOD was exposed by the trials in J anuary and February 1998 that had
taken place showing its instability. When this was reported to
ADSCOM by NAPCO, ADSCOM urged AA to commission a standby
FIDS as a contingency. When NAPCO and ADSCOM knew that the
standby FIDS, which was the contingency plan for the possible failure of
the main FIDS, had been successfully tested, then it was satisfied that
AOD could go ahead. As to when the standby was to be used and the
procedure of how it would be invoked, those related to the operational
details with which NAPCO and ADSCOM were not concerned, and the
Commissioners opine rightly so. There was also a standby Stand
Allocation System as a contingency measure for the possible failure of an

74
integral component of FIDS, ie, the Terminal Management System, that
would perform the stand allocation job. ADSCOM was also satisfied
with this arrangement. The contingency of using Public Address System
to make announcements and whiteboards to show flight information was
also a measure to disseminate updated information, when necessary,
during the interval when there was to be a switchover to the standby
FIDS. However, no thought was given by AA to the situation in a global
and overall manner, for instance, when the main FIDS failed and resort
had to be made to the standby FIDS and whiteboards, whether the means
of communications available were adequate in obtaining the necessary
flight information and directing and disseminating the information to the
parties requiring the same. The Commissioners are therefore of the
view that NAPCO also failed in examining the contingency plans made
by AA in a critical manner. There should also have been an overall
assessment of the risks involved in opening the new airport for operation
on AOD. This assessment would involve evaluating the risk of failure
of the systems critical for AOR, what would be needed in the event of a
failure of one system or another, and whether proper contingency
measures had been put in place to cover the problems which might arise.
AA should have made such an assessment, but admittedly did not do so.
NAPCO could have asked AA if it had made such an assessment. Such
an inquiry could have alerted AA. Had AA produced a report on risk
assessment, NAPCO should also have examined it critically and advise
ADSCOM accordingly. If, as the fact turned out to be, AA did not make
such an assessment, NAPCO should have done it independently
according to the knowledge of the progress that it had obtained from its
own monitoring of critical AOR issues and AAs reports. This, however,
NAPCO failed to do. This failure can be said to be an instance of lack
of coordination with AA.
5.69 The Commissioners turn to examine the position of
ADSCOM: whether it should be responsible for NAPCOs said failures.
ADSCOM had assigned to NAPCO the duty to monitor the progress of
the new airport, and the failure of duty on NAPCOs part in the two
aspects mentioned above should not justly be a basis for attaching blame
to ADSCOM because it is only fair and reasonable that ADSCOM
expected NAPCO to fully discharge that duty. ADSCOM made an
overall assessment of the readiness of the new airport at the end of 1997

75
when a decision on AOD was needed. The Chief Secretary as chairman
of ADSCOM then wrote to W50 Wong, the Chairman of the AA Board,
impressing upon him the importance of the new airport being ready with
safe, smooth and efficient operation when it opened. In that letter of 15
November 1997, detailed questions were set out covering AOR critical
subjects for AA to answer. The request was for an honest assessment
by AA. After careful consideration of all matters raised, AA assured
ADSCOM that the new airport would satisfy the requirement of safe,
smooth and efficient operation at the end of April 1998. Even with that
assurance, ADSCOM decided that AOD should be in J uly 1998, having
placed great importance on the availability of AR which would only be
ready by the end of J une 1998. Thus, a comfortable float was provided
to AA from its recommended date of end of April to the beginning of J uly
1998. There is no evidence to suggest that AAs assurance of the new
airport being ready at the end of April 1998 was not carefully or honestly
made. In such circumstances, it would be unreasonable to regard that
ADSCOM was rash or acting carelessly or improperly in deciding that
AOD should be in J uly 1998.
5.70 Counsel for the Commission criticised ADSCOM for failing
to re-assess the readiness of the new airport after it had reviewed the
situation very carefully in J anuary 1998 to make the decision on AOD.
Since the decision was made and announced, ADSCOM continued to
examine critical AOR issues. It continued to meet to examine the
progress of these issues reported by AA and NAPCO. From time to time,
it was given reassurances by both AA and HACTL that PTB and ST1
respectively would be ready on AOD. It would therefore be
unreasonable to suggest that ADSCOM should repeat the assessment
exercise every month or at any time after J anuary 1998 unless there was
clear evidence available to it that the J uly date was at risk. Yet, no such
evidence was presented to it. On the contrary, there were the
reassurances from AA and HACTL, which albeit were based on
over-confidence on their part. The over-confidence is referred to in
paragraphs 17.33 and 17.62 of Chapter 17. Nor did NAPCO advise
ADSCOM that AOD was at risk or a deferment should be considered.
NAPCO did not examine the contingency plans of AA critically or query
AA as to whether an overall risk assessment had been made. However,
there is no evidence that ADSCOM should have any reason to doubt that

76
NAPCO was discharging its functions of a critical observer dutifully. In
J une and early J uly 1998, AA also represented to ADSCOM that the main
FIDS as a whole achieved 98.7% reliability, that the Access Control
System and the Public Address System had been successfully tested, and
that the standby FIDS had also been successfully tested. All these
provided confidence to ADSCOM that AOD was safe. The problems
surfacing on AOD were multifarious, compounding and feeding each
other in a downward spiral, culminating in the chaos that no one,
including all the experts and professionals involved in the works and in
the assessment of the works, could have reasonably foreseen in all the
surrounding circumstances. Although the Commission is able to find
most of the causes for the problems facing AOD, that is done with the
benefit of hindsight which was unavailable to the involved parties at the
material times. It would be unfair to castigate anyone purely through
the lens of hindsight, and the Commissioners have refrained from doing
so.

5.71 One question, however, remains. As has been emphasised
in paragraph 5.30 above, ADSCOM was the overall monitor of the
progress of the development of the new airport and NAPCO as
ADSCOMs executive arm was to monitor the progress of AOR. The
relevant passages are repeated here for easy reference:

(a) Since the drawing up of the AOR will run into different
policy areas, and since the smooth opening of the new airport
is essential to Hong Kong, ADSCOM is best placed to be the
overall monitor. The PAA/AA should submit regular
progress and funding reports through NAPCO to ADSCOM.
[para 20 of ADSCOM Paper 45/95 of October 1995];

(b) NAPCO serves as the executive arm of ADSCOM and is
responsible for the overall management of project
implementation and co-ordination. [para 2 of ADSCOM
Paper 49/91 of December 1991]

(c) NAPCOs Programme Management responsibilities would
include: (k) Other duties as directed by the Chief
Secretary/ADSCOM. [para 11 of ADSCOM Paper 29/91 of

77
August 1991]; and

(d) As executive arm of ADSCOM, NAPCO will monitor the
progress and funding position of the AOR [para 23 of
ADSCOM Paper 45/95 of October 1995].

5.72 It therefore appears clear that the duty of an overall monitor
was ADSCOMs and not NAPCOs. However, as it had the power to do
so, ADSCOM delegated the duty of the overall monitor of the progress of
AOR to its executive arm, NAPCO, and directed it to discharge the duty.

5.73 The question on responsibility of ADSCOM can be framed
in various ways:

(a) Does the delegation of the duty of ADSCOM as an overall
monitor to NAPCO fully discharge that duty on the part of
ADSCOM?

(b) Does the delegation absolve ADSCOM from any
responsibility as an overall monitor?

(c) Is ADSCOM responsible for NAPCOs failures in discharging
the overall monitoring duties as found by the Commission?

5.74 As said in paragraph 5.70 above, there is no evidence that
ADSCOM should have any reason to doubt that NAPCO was not
discharging its functions of a critical observer dutifully. W36 the Chief
Secretary testified that she considered NAPCO had discharged its
functions conscientiously. The Commissioners are satisfied that
ADSCOM did not know that NAPCO had committed the two errors,
namely, failing to ask AA the pertinent question whether AA had the
necessary expertise to monitor the readiness of HACTLs systems, and
failing to critically examine AAs contingency plans and query AA as to
the existence of an overall risk assessment. If ADSCOM were
duty-bound to inquire with NAPCO on these two areas, ADSCOM would
be doing NAPCOs job or advising NAPCO as to how to do its job.
That does not seem to be right, for an executive arm is to execute the
decisions of the policy maker. The policy maker will decide on policies

78
which will be carried out by the executive arm but is not supposed to
advise the executive arm as to how to do its job. The executive arm is
responsible for carrying out such decisions and the functions entrusted to
it by the policy maker, and it is unreasonable to hold the policy maker
responsible for the faults committed by the executive arm in the
execution of the functions. On this analysis, it does not appear that after
the delegation of the overall monitors duties to NAPCO, ADSCOM
should be responsible for NAPCOs failures in the discharge of such
duties. Another perspective is to compare ADSCOMs position with
that of the AA Board. The Ordinance imposes on the AA Board the duty
to have regard to the efficient movement of air passengers and air cargo
in operating the new airport. The delegation of such duty by the AA
Board to the AA management does not appear to absolve it from being
responsible if such duty has not been discharged, for the Ordinance looks
upon the AA Board to discharge that duty, although at the same time
allowing it to delegate. The duty is a primary duty that cannot and, in
the opinion of the Commissioners, should not be discharged by mere
delegation. Taking into account the importance of the readiness of the
new airport on AOD to the Hong Kong public, in ADSCOMs own words
the smooth opening of the new airport is essential to Hong Kong, and
bearing in mind that it was ADSCOM who made the decision on AOD
and also that it was ADSCOM and ADSCOM alone in the circumstances
that could make a decision to defer AOD, if necessary, the
Commissioners come to the view that the responsibility as an overall
monitor should not be allowed to be discharged by delegation. It should
be considered as a primary and crucial duty, the ultimate discharge of
which should rest with ADSCOM. The public also looks upon
ADSCOM to have that duty discharged. In the premises, the
Commissioners hold that ADSCOM should be responsible for NAPCOs
failures, though not without some hesitation. The holding, however,
does not mean that ADSCOM was at fault, for it did not commit any error
in entrusting the overall monitoring job to NAPCO, its executive arm, nor
did it commit any error in not advising how NAPCO was to do that job.
Nonetheless, towards the public, its responsibility as the overall monitor
of AOR was not discharged satisfactorily.

79

CHAPTER 6

AOD DECISION, PREPARATION AND RESPONSIBILITY

Section 1 : Planning for AOR

Section 2 : Decision on AOD

Section 3 : Preparation for AOD


Section 1 : Planning for AOR

6.1 Up till the decision was announced by Government in
J anuary 1998 that airport opening day (AOD) for operation was to be
on 6 J uly 1998, Airport Authority (AA) had targeted April 1998 as the
opening day, and everything in the development of the new airport was
geared to this goal. AA originally planned for a clear trials period, that
is, a period dedicated to training, trials and other operational transition
activities following substantial completion of airport facilities and
systems, and their hand-over to the operators. As stated in a letter dated
14 September 1994 from W43 Mr Douglas Edwin Oakervee, the Director
of AAs Project Division (PD), to the then Director of New Airport
Projects Coordination Office (NAPCO), W48 Mr Billy LAM Chung
Lun, AA at the time planned a 7-month airport operational readiness
(AOR) phase, including a 4-month period for airfield trial operation
upon completion of the runway and taxiways, and a further 3-month
period for airport-wide trials following completion of airport
commissioning and system integration.

6.2 Consistent with AAs planning, the Airport Trial Operations
Strategy Document prepared in draft form by Mr Daniel Ough of AA in

80

1996 included a 4-month clear trials period, and stated that Airport
construction and testing must be complete before Airport Trial Operations
commence to avoid trialing incomplete facilities.

6.3 By August 1997, a period during which acceleration
measures were being planned and instructed by AA for the various
systems contracts, AA continued to maintain a clear trials period. The
period was then subdivided into a 1-month Transfer to Operations
period and a 3-month Airport Trials Period. While a number of works
activities were extending into the 3-month clear trials period, all major
works activities including systems installation, testing and
commissioning were planned to be completed well in advance of the
trials period.

6.4 In response to a request by Airport Development Steering
Committee (ADSCOM) at its meeting on 9 August 1997, NAPCO
prepared an assessment of the AOR process for the new airport, with the
cooperation of AA. NAPCO presented to ADSCOM its preliminary
observations on 23 August 1997 and its findings and recommendations on
20 September 1997. The definition of AOR used as a basis of the
assessment was as follows:

The AOR process encompasses all steps necessary to transition
the new airport from construction, testing and commissioning
through familiarisation, training, trials and relocation to ensure safe,
smooth and efficient operations from the first day of airport
opening, at a demonstrated and acceptable level of operational
standard. This can only be achieved when the operator and all
involved parties are fully familiar with the airport facilities,
systems and procedures, and only after systems and procedures are
reliable, practised and proven. The success of the process
depends on a comprehensive level of integrated planning,
coordination and management, and the unconditional commitment,
participation and sharing of information by all parties.

6.5 This definition of AOR was also agreed to or accepted by
AA, and was substantially identical to what ADSCOM saw as the
requisite for airport opening.

81

6.6 In NAPCOs assessment report, NAPCO advised that
continuing delays in the physical works and systems had had the effect of
compressing the AOR period in relation to the opening date target, which
at the time was April 1998. Given this programme compression, the
AOR activities, including training and trials, would have to be conducted
in parallel with works completion activities. The report pointed out that
such overlapping of the construction works phase and the AOR phase was
perhaps unprecedented when comparing AOR programmes for other new
major airports. While such an approach was not necessarily
unacceptable, a potential risk to AOR existed. To mitigate the risk to
AOR and a smooth opening, NAPCO recommended that AA undertake a
well-defined and programmed iterative process of AOR access, training
and trials activities in parallel with works completion activities, starting
six months from opening.

6.7 While AA planned to initiate its training programmes in
Interface House and other off-site locations to get a necessary early start
on training, NAPCO pointed out in the report that there was no substitute
for on-site, hands-on training. To accommodate this, NAPCO
recommended that the Airport Management Division (AMD) be
granted access to the Airport Operations Control Centre (AOCC)
facility and systems, as well as facilities and systems designated for use
in the first trial, at least 18 weeks prior to opening in order to start
hands-on operator training. NAPCO also recommended that business
partners and Government staff be granted unimpeded access to facilities
and systems designated for use in the first trial at least two weeks prior to
the first trial date (at least 12 weeks prior to opening). NAPCO further
recommended that all systems deemed essential for Day One operations
be fully tested, commissioned and available for final operational training
and trials not less than six weeks prior to opening. The
recommendations were based on discussions between the NAPCO AOR
Coordination Team and members of AMD, including W44 Mr Chern
Heed, AMD Director, and reflected AMDs views at that time.

6.8 Other findings from NAPCOs AOR assessment report
included:

82

(a) While it was found that AAs training plans appeared to be
well developed, NAPCO was concerned that AA planned to
proceed with training on systems that had not been fully
developed, tested and commissioned. NAPCO considered
it likely that AA staff had to be retrained on the actual Day
One systems following their testing and commissioning, and
that AA needed to take this into account in its training plans.

(b) Lack of integrated AOR plans and programmes, and no
demonstrated ability to track the sequence of required AOR
steps within AAs master programme. NAPCO found it
difficult to determine the true status of constructions works
as well as AOR activities in AAs master programme leading
to Day One operations. There were a number of schedule
and interface mis-matches in AAs programmes. Many of
these mis-matches related to the use of mandatory dates that
were not linked to other construction and AOR activities.
NAPCO pointed out that without such linkage, progress
measurement and the effects of delay could not be fully
assessed, and that the ability to chart a critical path and
develop plans to recover from potential delays was limited.

(c) Lack of coordination within AA itself, and between AA and
its business partners and Government departments on AOR
issues. NAPCO found that the matrix organisational split
of AOR responsibilities between the various AA Divisions
was not functioning efficiently, that information bottlenecks
existed, and that there should be a shift in organizational
focus and decision making from the construction side, ie PD,
to the operations side of AA, AMD. NAPCO
recommended that AA consider appointing a single-point
responsible senior-level executive to direct the AOR process
and coordinate action inclusive of all participants, including
the various AA Divisions.

6.9 AA agreed with the major findings of NAPCO and
addressed the concerns of NAPCO. On training, AA gave assurance that
re-training would be undertaken as required. This was subsequently

83

reconfirmed by AA in its reply of 10 December 1997 to a series of
questions posed by W36 Mrs Anson CHAN, the Chief Secretary for
Administration and Chairman of ADSCOM (the Chief Secretary). AA
stated that necessary refresher training would be provided in a 10-week
period from the final handover of Day One systems (then targeted at 15
February 1998) to the airport opening date (which was then late April
1998). On integration of the construction works with the AOR
programme, during the course of NAPCOs assessment, AA had already
given assurance that a programme integrating works and AOR activities
was already in place, and would be further developed as necessary. As
noted in ADSCOM Paper 34/97 dated 19 September 1997, AA agreed to
produce an AOR master programme linked to its construction programme,
called the Integrated Accelerated Programme. On coordination, AA
gave assurance that steps would be taken to improve coordination and
communication between PD and AMD. W43 Oakervee stated at the
ADSCOM meeting on 20 September 1997 that a task force had been set
up on the interface between PD and AMD. At the ADSCOM meeting
on 13 October 1997, AAs Divisional Manager of Planning and
Scheduling, Mr J J esudason, stated that the relationship between the two
Divisions was getting better every day. At that meeting, W3 Dr Henry
Duane Townsend, the Chief Executive Officer (CEO) of AA, was
requested by the Chief Secretary to keep a close watch on the situation
and to promptly sort out any difficulties between the Divisions. AA also
assured ADSCOM that the AOR milestones tabled at the ADSCOM
meeting on 20 September 1997, based on ADSCOM paper 34/97, would
be met. At the ADSCOM meeting on 20 September 1997, W3
Townsend confirmed that AA agreed with NAPCO on what constituted
critical dates in the training and trials programme, and he was positive
that AA could achieve the milestones, including having integrated
systems fully tested and commissioned not less than six weeks prior to
opening.

Section 2 : Decision on AOD

6.10 The decision to open the airport on 6 J uly 1998 was taken by
ADSCOM in J anuary 1998. In her witness statement, the Chief
Secretary as Chairman of ADSCOM recapitulated major events that

84

illustrated how this decision was arrived at. The target date for the
opening of the new airport was originally scheduled for April 1998 in the
announcement made by the then Financial Secretary in J une 1995
following the Sino-British agreement on the financing arrangements for
the new airport and the Airport Railway (AR), which was later known
as Airport Express. It was based on the Provisional Airport Authoritys
(PAA) programme of works and the time required for commissioning,
trial operations and planning the move from Kai Tak. Since then AA,
Government departments and many other parties which were involved
with the Airport Core Programme (ACP) had been using the date as the
target completion date of their own programmes. However, it was
always understood that as a target date, it would require confirmation by
a formal announcement to be made closer to the time by Government in
conjunction with AA, in the light of the overall airport readiness achieved
and the prospects of AR being ready ahead of time. AR had a planned
completion date of 21 J une 1998 but there was the expectation that, given
the past record of Mass Transit Railway Corporation (MTRC), progress
on AR could probably be accelerated to support airport opening in April
1998.

6.11 In the franchise agreements of Asia Airfreight Terminal
Company Limited (AAT) and Hong Kong Air Cargo Terminals Limited
(HACTL), as well as in others, AA was obliged to give a three-month
advance notice to the franchisees of the date of the opening of the new
airport. While this obligation encumbered AA to have a date fixed for
the opening long beforehand, it had to examine and monitor the progress
of all kinds of works relating to the building, building services and
facilities in order to keep to the expected opening in April 1998. On the
other hand, ADSCOM was mindful of the importance to fix an airport
opening date well in advance so that the public as well as all concerned
parties would know this date for their own purposes and planning. It
was therefore necessary for Government to take a decision on a firm
airport opening date to be made at least three months ahead of April 1998.

6.12 With the above in mind, Government started to examine
critically the state of overall AOR from February 1997 onwards. On
supporting transport facilities, ADSCOM re-examined with MTRC
prospects of AR being ready ahead of time and in parallel considered the

85

feasibility and acceptability of bringing in contingency transport
arrangements in the case of AR acceleration not being possible.

6.13 From May 1997, ADSCOM had been requesting AA to
advise on the overall readiness of the airport to open in April 1998. In
May and J une 1997, W3 Townsend made the assessment that the physical
structure of the Passenger Terminal Building (PTB) should be
completed in November 1997 and other elements such as the fit-outs, the
computer systems and the retail and commercial operations should be in
place by J anuary or February 1998. He also advised that an AA Master
Programme was in place to keep track of AOR by integrating franchisees
programmes, systems contracts and training.

6.14 ADSCOM looked seriously into the possibility of advancing
the completion of AR. At the ADSCOM meeting of 6 September 1997,
MTRC made a presentation to ADSCOM on progress of the AR when
MTRC expressed reservations in its ability to advance AR completion
date to April 1998. ADSCOM asked the Chairman of MTRC to put
together a proposal setting out all relevant considerations such as
practical, operational and financial implications.

6.15 At the ADSCOM meeting on 20 September 1997 referred to
in paragraph 6.9 above, ADSCOM asked AA for an analysis of the critical
issues and an unequivocal statement whether April was a realistic target
date to enable ADSCOM to arrive at a definitive view on airport opening.

6.16 At the ADSCOM meeting on 13 October 1997, AA provided
a more comprehensive report with a revised work programme and briefed
ADSCOM on the works progress, training and trial preparations, and
contingency plans. W3 Townsend also reported that he believed
HACTL should be able to achieve 50% of its designed capacity by end
April 1998. However, in spite of the general optimism expressed by AA,
ADSCOM asked for further reports on training and trial, as well as on
systems integration.

6.17 At the ADSCOM meeting on 24 October 1997, MTRC made
a detailed presentation indicating fundamental problems with advancing
AR completion date to April 1998. The Chief Secretary assured the

86

Chairman of MTRC that in no case would Government compromise the
principle of safety and reliability. At the same meeting, ADSCOM
considered a paper prepared by the Transport Bureau on the impact of a
mismatch between the commissioning dates of the new airport and AR.
While it was agreed that alternative contingency transport arrangements
for April opening should be further explored, the prospects of a later
opening date in late J une or early J uly were also discussed and 1 J uly
1998 was raised as a possibility. The Chief Secretary asked for a careful
analysis of all essential factors before a final decision was taken.

6.18 At the ADSCOM meeting on 3 November 1997, AA
provided an updated assessment on AOR, addressing the key questions
raised by ADSCOM, including construction programme, status of
systems integration and scenario for Day One operation. AA was
positive that an April 1998 opening date could be achieved. W3
Townsend reported that the AA Board was following the subject closely
and that the consensus of the Board was that they would be able to
operate the airport on 1 April 1998, but taking account of HACTLs
progress, the end of April would be a more suitable date. AAs
optimism at that time, however, was not entirely shared by ADSCOM
members. From the NAPCO reports, it was noted that the works
programmes had slipped, and the plan for systems training was tight.
The Chief Secretarys view was that as the AA Board and its management
were closest to the actual construction of the airport and its operational
readiness, the AA Board must provide Government with a categorical
confirmation that everything essential for an efficient airport on Day One
would be available if they recommended an April 1998 date. At the
same meeting, Transport Bureau presented a discussion paper on
contingent transport arrangements. ADSCOMs general view was that
the contingent arrangement, though technically feasible, might not be as
efficient as AR and would not be commensurate with the image of a
modern airport.

6.19 At a special meeting on 7 November 1997, ADSCOM
remained concerned that AA had continued to qualify its statements with
provisos. ADSCOM noted AAs apparent difficulty in keeping to
milestones in its own programmes. The Chief Secretary decided to
write to the Chairman of AA asking a series of specific questions based

87

upon information provided by NAPCO. At her request, W1 Mr Richard
Siegel, Director of Civil Aviation ( DCA), attended this and subsequent
ADSCOM meetings to provide additional input from the civil aviation
perspective.

6.20 Accordingly, the Chief Secretary wrote to the Chairman of
AA, W50 Mr WONG Po Yan, on 15 November 1997 expressing
ADSCOMs serious concern as to whether April 1998 was a realistic
opening date. In his reply dated 10 December 1997, W50 Wong
responded to each and every question and assured her that

the Board has undertaken a very thorough review of progress in
all areas, with particular reference to areas of concern identified in
your letter and the attached questions. Following this careful
scrutiny, we are satisfied that the airport can be ready to open for
safe, smooth and efficient operation on an appropriate date in the
last week of April.

6.21 Notwithstanding these reassurances from AA, the Chief
Secretary again wrote to W50 Wong on 17 December 1997 saying that
Government members of the AA Board still had concern in various areas
and that they had suggested that as a number of key milestones would be
coming up in the next few weeks, both AA and Government would be in
a better position to assess whether an April opening date would be
achievable if these key milestones were indeed achieved according to the
latest programme presented by the AA management. She urged the
Board to continue to monitor developments closely with a view to
reaching a firm conclusion on the airport opening date in early J anuary
1998.

6.22 The crucial ADSCOM meeting took place on 2 J anuary 1998.
There were different opinions amongst members and regular attendees of
ADSCOM. The implications of an April or J une 1998 opening were
carefully examined. After a thorough discussion, the Chief Secretary
eventually decided, with the endorsement of the entire ADSCOM, that
given the doubts on the adequacy of contingency transport arrangements
and the state of readiness of airport systems and HACTL, airport opening
should be deferred, with the aim of producing on Day One a world class

88

airport supported by efficient transport facilities.

6.23 Bearing in mind the repeated positive assurances from AA
that the new airport would be fully operational in April 1998, it was
thought by ADSCOM members that an extra three months would provide
added comfort to both AA and its franchisees to strive for a safe, secure
and efficient airport upon its opening.

6.24 Having ruled out the April date, the Chief Secretary asked
those present in the meeting if 21 J une 1998 (which was the scheduled
date for AR commissioning) or 1 J uly 1998 (which was the first
anniversary of the Hong Kong Special Administrative Region) should be
the new opening date. The general inclination of ADSCOM members
was for 1 J uly 1998, to allow more time for AR to get ready and for
public relations reasons.

6.25 Recognising that there should be a time gap between the
ceremonial opening of the airport and its actual coming into service, the
Chief Secretary requested NAPCO to find out what the shortest gap
should be, taking into account the logistical requirement for the airport
relocation exercise and the airport opening ceremony.

6.26 At a special meeting on 8 J anuary 1998, NAPCO suggested
and ADSCOM accepted Monday 6 J uly 1998, because a few days would
be needed between the airport ceremonial and operational opening for the
critical phase of the airport relocation exercise. Opening the new airport
on a Monday would offer the advantage of the night move taking place
when road traffic was light and when a big spectator turnout would be
unlikely. Air traffic was also lighter on a Monday.

6.27 The Chief Secretary then met the Chief Executive in Council
(the Chief Executive) and explained to him the reasons why ADSCOM
had decided to defer the opening date to J uly. He endorsed the decision
and agreed that the Executive Council (ExCo) should be informed of
ADSCOMs recommendation. Airport opening was discussed in ExCo
on 13 J anuary 1998, and ADSCOMs recommendation was noted.

6.28 The 1 J uly date was eventually altered to 2 J uly 1998 for the

89

ceremonial opening of the new airport, while the day for opening the new
airport for operation was unaltered to remain as 6 J uly 1998.

Section 3 : Preparation for AOD

6.29 Prior to the announcement of AOD, both AA and
Government appreciated that there were slippages regarding the
construction, fit-out and system works in PTB and in HACTLs premises.

6.30 HACTLs premises consist of SuperTerminal 1 (ST1) and
an Express Centre, but the major part of the cargo handling operation
would be performed within ST1. Construction of ST1 suffered from a
series of late completion dates, with building works delays affecting
equipment installation, commissioning, testing and the installation and
testing of Government department support systems. However, NAPCO
was not aware of any reports of problems with HACTLs own computer
system development. Within ST1, HACTL was to provide a designated
number of rooms for Government departments, the services of which
were essential to the air cargo business. These departments, particularly
the Customs and Excise Department (C&ED), were in turn dependent
on their own operating systems, such the Air Cargo Clearance System,
Trunk Mobile Radio (TMR) system, Information Network, etc. These
systems were installed by direct Government contracts, entrustments to
HACTL contractors and entrustments to AA contractors. For the last
few months before the J uly opening, NAPCOs primary concerns merely
related to the installation of Governments systems and HACTLs ability
to satisfy the requirements for obtaining an occupation permit. By AOD,
the ability of the Government departments to support the air cargo
operation at HACTL had been achieved, and no problems arose therefrom.
In the circumstances, the rest of this report will not deal with the
Government entrusted works in ST1.

6.31 NAPCO did not directly monitor HACTLs development
apart from the Government entrusted works. The reasons for the
positions of Government and AA relating to HACTL can be found in
paragraphs 5.64 and 5.65 of Chapter 5.

90

6.32 The routine monitoring carried out by NAPCO in relation to
HACTL was initially focused on the basis of HACTLs Monthly Reports
to AA, copies of which AA passed to NAPCO. Later on, AA also
provided NAPCO with copies of AAs own internal project reports, which
contained very detailed information on both PTB and HACTL projects.
NAPCO staff were also allowed to participate in meetings between AA
and HACTL.

6.33 ST1s piling and building contracts started in September and
November 1995 respectively and by August 1996 delays of around two
months in the building works started to impact the erection of the cargo
handling equipment. By November 1996, these equipment delays had
increased to three months and HACTL acknowledged that it would have
to control progress very closely.

6.34 It was stated in NAPCOs Weekly Situation Report dated 2
May 1997 that broad agreement had been reached in principle with its
main building contractor, Gammon Paul Y J oint Venture (GPY) on a
revised programme which would address delays of around 20 weeks and
achieve 50% of its full capacity throughput in April 1998. Yet no
agreement incorporating any accelerated programme was signed.

6.35 The delays suffered in HACTLs construction works caused
great concern to NAPCO and ADSCOM, and the subject was discussed
continuously since the middle of 1997. Mr Tudor Walters, Consultant
Project Manager (CPM) of NAPCO, provided an account in his
statement as follows:

(a) By the ADSCOM meeting on 12 J uly 1997, HACTLs senior
management had made public statements to the effect that
they doubted their ability to support an April airport opening.

(b) By the ADSCOM meeting on 20 September 1997, W3
Townsend confirmed that HACTL was striving to meet an end
April 1998 opening.

(c) At the ADSCOM meeting on 8 December 1997, lack of
confidence in HACTLs ability to meet an April 1998 opening

91

without a major acceleration to its existing contracts was
noted.

(d) At the ADSCOM meeting on 14 February 1998, even though
the airport opening date had by then been postponed to J uly,
HACTL remained high on the agenda of outstanding concerns.
Acceleration measures had not yet been contractually
concluded with GPY.

(e) By 21 March 1998, a supplemental agreement for
acceleration had been negotiated between HACTL and GPY,
but it was not until a month later that the parties formally
signed it.

(f) By the ADSCOM meeting on 22 May 1998, the main
concern was with the time when an Occupation Permit for
ST1 could be obtained. NAPCO had written directly to W7
Mr Anthony Crowley Charter, the Managing Director of
HACTL, on 21 May expressing concerns over progress, but
had received a response at the end of the month expressing
confidence in ST1 being ready in time. The detailed
Monthly Reports from HACTL at this time were conveying
improvements which would support a 75 percent throughput
by airport opening.

(g) By 6 J une 1998, HACTLs target for the Occupation Permit
had slipped to 20 J une 1998.

(h) HACTL was part of the itinerary for the ADSCOM visit to
the new airport on 14 J une 1998. The lack of readiness of
offices within the building for Government departments was
noted, but W7 Charter expressed confidence and offered
unqualified assurance that 75% throughput capacity would be
achieved by AOD.

6.36 The problems with the PTB may be generalised as
predominantly caused by lateness of first the phased completion of the
foundations, then late completion of the building structure, finishes,

92

systems installation, integration, testing and commissioning, with training
and familiarisation being squeezed into a continually decreasing time
frame.

6.37 Efforts were made by the AA management to address the
initial lateness in the building works by effecting two supplemental
agreements in September 1996 with the main building and building
services contractors. The purpose of these supplemental agreements
was to address accumulated delays and introduce accelerations while at
the same time extending the completion dates for these contracts from the
end of J une 1997 to the end of October and November 1997 respectively.
This would have preserved four clear months for training trials and
familiarisation ahead of a 1 April 1998 opening.

6.38 The contractors for the AA systems were also subsequently
approached for proposals to recover delays and accelerate the completion
of their works by December 1997. Most systems contractors refused (or
were unable) to guarantee completion by the requested date, but offered
phased completion on a prioritised basis to December 1997, with certain
items of system functionality deferred to and after airport opening. At
the time, this was conveyed by AA to be acceptable as all system
functions necessary for airport opening would be in place, although
training and familiarisation would have to be phased to match systems
availability. Specific shortcomings persisted in a number of the airport
operating systems, however, particularly the Flight Information Display
System (FIDS) and its integration with the Airport Operational
Database (AODB) on which many other airport operating systems
relied for their basic information.

6.39 At the ADSCOM meeting on 8 December 1997, it was noted
that the AOCC facility, agreed by all at the 20 September meeting to be
so crucial in its hand-over to AMD on 15 November 1997, was
anticipated for hand-over on 5 J anuary 1998. Similar slippage had taken
place on Government areas within PTB. None of the Government areas
had been completed by 1 December 1997, and the best estimate by AA
was for completion in March 1998.

6.40 Following the announcement of 6 J uly 1998 as the AOD, a

93

lot of preparations in various areas were made in earnest by AA for AOR.
AA took steps to ensure that all AMD staff who would be working at the
new airport received adequate operational training for the roles which
they would perform at Chek Lap Kok (CLK). A large number of the
staff worked at Kai Tak, and would continue to work there until AOD.
There were further staff who were due to work at CLK but did not work
at Kai Tak. It was necessary to ensure that all staff received experience
of the operational environment at CLK before AOD. The new staff all
undertook training at Kai Tak in order to gain real life operational
experience. This facilitated the release of existing Kai Tak staff to
undergo training and familarise themselves with CLK operations. The
new CLK staff also attended operational training courses at CLK, which
included both classroom sessions and hands on sessions in which they
were able to gain experience of the systems and facilities at CLK.

6.41 In addition, the AOCC staff also took part in the testing and
commissioning of systems organised by PD and the Information
Technology (IT) Department and thus acquired familiarity with
operational procedures and the fall-back and workaround systems which
would be required if the primary systems or facilities did not function as
designed. To reinforce what had been learned from the various training
courses, all the terminal operations staff were scheduled to participate in
at least one of the five terminal operation trials and other relevant tests so
that they could benefit from the experience of working in a simulated
terminal operations environment.

6.42 Airport operational trials were an essential element of AOR
Programmes. The organisation, consultation and management of these
trials required a long lead-time for planning which was started in late 1996.
AA had programmes for the airport trials to be conducted between J anuary
and March 1998, before the airport became operational at the then targeted
April 1998 opening. The objectives were to identify deficiencies in the
facilities, and to test the effectiveness of the operational training received by
staff. Another purpose of the trials was to identify potential problems
before AOD so as to make improvements to minimise the problems which
might arise when the airport became operational. These objectives were
reflected in the initial paper on airport trials which was circulated on 8

November 1996 to, among others, all general managers in AMD.

94

6.43 The scenarios for the trials were carefully planned in
consultation with all parties concerned. This included the airlines, ramp
handling operators, the baggage handling operator, relevant Government
agencies, and HACTL and AAT in case of airside trials. AA began
consultation with business partners and Government departments with a
view to establishing their requirements for the trials in early 1997. After
this initial consultation process a strategy paper was prepared incorporating
the comments of the business partners and Government agencies. A
working group was also formed to discuss and agree on trial scenarios and
the way to carry them out in the trials. The working group, including
business partners and Government agencies, held its first meeting on 11
April 1997 and continued to meet up to the time of the final airport trial.

6.44 The trials were designed to simulate real operation as much
as possible. However, as AA was not able to get many aircraft from the
airlines to participate in the trials, it was impossible to simulate the stress
of real operation.

6.45 AA originally planned only three trials of landside and PTB
facilities and systems, for mid-J anuary, mid-February and mid-March
1998. As AOD was moved from the expected April to J uly 1998, AA
reviewed the strategy for the trials and decided to add two additional
trials. The organisation of the operational trials was a major logistical
undertaking. The scale of the AA trials increased each time. The first
airport trial on 18 J anuary 1998 involved only approximately 500
participants; the second on 15 February 1998 had approximately 1,000
participants; the third on 28 March 1998 comprised approximately 1,200
participants; the fourth on 2 May 1998 had approximately 2,000
participants; and the fifth trial on 14 J une 1998 involved 12,000
participants. However, because the construction project was operating
to a very tight timetable, the possible scenarios had to be matched up with
the facilities that would be available at the date of the trials.

6.46 Although it was always planned that AA would organise and
manage a series of operational trials, with progressive integration of the
landside, airside and PTB operations, these were not the only trials which
would be undertaken by airport users before AOD. Other trials would

95

be undertaken by contractors, which were monitored by AA, in relation to
particular systems and facilities. Further, it was always the case that
other airport users, such as the business partners and Government
agencies would organise their own trials and training for their staff.
Indeed AA actively encouraged them to do so. These trials did not have
to involve AA but it would always be prepared to help if required.

6.47 After each trial there were review meetings, attended by a
number of AA staff and representatives from NAPCO, Government
departments such as the Fire Services Department, Immigration
Department, C&ED and Hong Kong Police, as well as business partners
like Cathay Pacific Airways Limited, Hong Kong Airport Services Ltd.
and J ardine Air Terminal Services Ltd., HACTL, etc.

6.48 After these meetings, participants in the trials would prepare
reports focusing on areas for improvement in accordance with the
objectives of the trials. These would then be collated in a summary. In
addition, AA undertook surveys with the volunteers who played the role
of passengers in the trials in which it sought views on matters such as
signage, cleanliness, lighting and temperature within the PTB, and also
asked them to make comments on any other matters on which they
wished to express a view. The summaries prepared following the review
of meetings of the trials were sent to W44 Heed. He would discuss the
matters arising from the reports with the Deputy Chief Executive Officer
of AA, W48 Lam, and W43 Oakervee. Their collective views were then
incorporated in reports to the AA Board which discussed the results of the
trials and issued recommendations for future action with a view to
remedying defects discovered in the trials.

6.49 As a result of these actions, a number of problems were
identified and changes and improvements accordingly made by AA to
PTB facilities or procedures. For instance, after the first trial, the
feedback showed that the queuing area for Immigration was too short.
AA therefore moved the screens back to allow a longer queuing area.
There were also comments that the toilet facilities were insufficient in the
Meeters and Greeters Hall and as a result two additional bigger toilets
were built inside the hall. In addition, other improvements such as
modifications to the toilets, revision and addition of direction signs were

96

suggested. However, AA was not able to have all of these changes
implemented before AOD.

6.50 NAPCO was closely monitoring the airport trials as
observers. NAPCO reported its observations with critical comments to
ADSCOM. In the Weekly Situation Report for 24 J anuary 1998,
NAPCO commented that the first trial served to demonstrate that
incomplete systems remained the greatest risk to AOR. Although the
trial was a useful familiarisation exercise for the participants, the original
trial objectives had not been achieved. FIDS crashed and both FIDS and
the interface between FIDS and Common User Terminal Equipment
(CUTE) were inoperable; the terminal-wide Public Address System
(PA) did not function; signage was minimal and inadequate; close
circuit television was not yet set up for Government departments; the
telephone and TMR systems were only available on a limited and
restricted basis. The required operational and support facilities for both
the Government and business partners were only partially available.
Five of the seven originally planned incident trials, involving the airport
systems such as communicating a gate change via FIDS, had to be
cancelled. It was subsequently discovered by AA that the crash of FIDS
was caused by Societe Internationale de Telecommunications
Aeronautiques incorrectly loading the software for CUTE and that the
problem was eventually resolved.

6.51 At the ADSCOM meeting on 14 February 1998, AA reported
further slippage in FIDS, with the site acceptance test (SAT) again
delayed from 15 February to 25 March 1998. That latter date essentially
became the date for finally deciding whether FIDS should be an
integrated system or should be in stand-alone mode.

6.52 In the Weekly Situation Report for 21 February 1998,
NAPCO emphasised that the second trial showed that the greatest risk to
airport operational readiness was systems development, integration,
testing, commissioning and training. The airport systems had yet to
achieve a state of pre-operational functionality and significant problems
continued with the stability of various systems, and in particular FIDS.
The Day One version of FIDS, scheduled to be in place by the second
trial on 15 February 1998, some 20 weeks ahead of opening in order to

97

provide a platform for system integration as well as training and trials,
had slipped to late March 1998. At the second trial, FIDS crashed again
during the initial check-in process, and both FIDS and the FIDS/CUTE
interface were only available with limited functionality on an intermittent
basis at the trial. As with the first trial, given that required operational
facilities and systems were only minimally and partially available, the
second trial was more an extension of the familiarisation programme
rather than a meaningful real life operational trial.

6.53 During late February, as noted in NAPCOs Weekly
Situation Report for 28 February 1998, the follow-on trial dates were
confirmed for 2 May and 14 J une 1998, the latter to be a fully integrated
dress rehearsal prior to AOD. The airside trial originally scheduled for
24 February was cancelled, and airside trial elements were combined with
the landside/PTB trial scheduled for 2 May 1998. In the Weekly
Situation Report for 7 March 1998, NAPCO noted that the Day One
version of FIDS, inclusive of basic functionality as well as required
systems interfaces, had slipped further, from 15 February 1998 to 1 May
1998.

6.54 In order not to interfere with FIDS development and testing
activities, the landside/PTB trial originally scheduled for mid-March was
re-scheduled for 28 March 1998. However, in the Weekly Situation
Report for 14 March 1998, NAPCO noted that given continuing FIDS
problems, further training by airline staff on FIDS/CUTE stations and
check-in procedures had been postponed to mid-April 1998, which did
not support the objective of having a functional system and sufficiently
trained staff for a meaningful trial on 28 March 1998. To help alleviate
this situation somewhat, and provide a more real-life simulated setting
for FIDS testing, AA scheduled and conducted a number of pre-trial
exercises with staff from 13 airlines on 25 March 1998. In this pre-trial
exercise, the FIDS/CUTE interface appeared to be functional and stable,
although a number of airlines experienced problems logging on and off of
the system, and a number of blank FIDS monitors was noted.

6.55 At the ADSCOM meeting on 21 March 1998, AA reported
that the SAT of FIDS functionality on a limited number of monitors (on a
localised network) had achieved a 90 percent pass rate. However,

98

interfacing problems with AODB, the Baggage Handling System (BHS)
and CUTE persisted. At ADSCOMs insistence, the AA Board had
instructed the development of a standby FIDS. There was concern that
the Day One systems configuration including any standby and
contingency arrangements should be in place by the first week of May, so
as to leave two clear months for final training and familiarisation.

6.56 The NAPCO observation team also attended the third trial
on 28 March 1998. In its Weekly Situation Report for 4 April, NAPCO
stated that this trial was generally successful for those systems functions
that were made available, although some facilities and most systems
required by the Government departments remained unavailable. The
report also noted that FIDS testing was on-going although some problems
remained. The report expressed the urgent need for AA to identify the
system functionalities critical to support airport opening in order to
achieve a state of pre-operational functionality.

6.57 On 30 March 1998, NAPCO staff with the personnel from
International Bechtel Company Ltd (Bechtel), a US firm of airport
consultants, held an internal review of AA systems status, including
contingency measures, based on drafts of ADSCOM Papers 14/98 and
16/98 that AA was to present to ADSCOM on 1 April 1998 on the status
of FIDS. Following the concern expressed by ADSCOM members over
FIDS and their urges to have a standby FIDS developed, the AA Board
approved the development of a standby FIDS on 23 March 1998. A
standby system is a fallback and will be used when the main system fails
to function. In the final version of AAs ADSCOM Paper 14/98, dated 1
April 1998, AA stated its intention to proceed with the permanent FIDS,
which it reported as having now been sufficiently developed to provide
an operational system which could be satisfactorily operated by AAs
AMD staff and the airlines. However, AA reported that it was also
proceeding with the development and implementation of a standby FIDS
that would be available in case the permanent system failed. While
NAPCO questioned whether there was sufficient time to successfully
develop and commission such a standby FIDS prior to opening, as stated
in its ADSCOM Paper 14/98, AA showed that they planned to complete
development and testing of standby FIDS functions by 15 J une 1998 and
to test the switch-over or cut-over from the permanent FIDS and to train

99

operators from that point to 30 J une 1998. NAPCO was also concerned
about the additional equipment, revised operational procedures and staff
training that would be required for implementing such a standby system.
In a letter faxed by Mr Tudor Walters to W3 Townsend, Mr Walters
expressed the need, in regard to the standby FIDS, for AA to develop a
quantification of essential data transfer requirements, workstation and
other equipment needs, software modification requirements, staff
requirements, procedure modifications and a programme to bring together
all these elements.

6.58 In the Weekly Situation Report for 18 April 1998, NAPCO
reported that many critical FIDS software problems remained, and they
needed to be resolved before the end of April in order to establish the Day
One operational software. The report also noted that AA was
proceeding to develop its contingency strategy for FIDS, including
manual data transfer that would be required if resort had to be made to
use the standby FIDS. On the other hand, at the ADSCOM meeting on
18 April, the level of confidence being reported by AA in relation to
systems readiness and integration was markedly increased.

6.59 In the Weekly Situation Report for 25 April 1998, NAPCO
noted that AA had identified the system functionalities critical to support
airport opening, which were reported to ADSCOM on 18 April 1998.
AA proposed that Day One systems operation would be to a large degree
on stand-alone mode. A stand-alone mode means that the system in that
mode operates independently and not in a mode integrated with other
systems. When a stand-alone mode is used, its integration with other
systems will be required to be done manually, by operators inputting the
required information and data derived from the stand-alone into the other
systems. NAPCO again expressed the urgent requirement that AA
provide a quantification of the manual data transfer across system
interfaces so that the amount of resources and procedural changes could
be identified and implemented. It noted in the Weekly Situation Reports
for both 25 April and 2 May 1998 that the airport systems would not
achieve a state of pre-operational functionality before mid-May 1998, and
that this remained a risk to AOR.

6.60 On 2 May 1998, the NAPCO observation team attended the

100

fourth trial. In the Weekly Situation Report for 9 May, NAPCO noted
that the trial was generally successful for those systems functions that
were made available, although some facilities and most systems required
by the Government departments still remained unavailable. This trial
encompassed expanded landside/PTB elements, and included a number of
airside trial scenarios using a B-747 aircraft. In the PTB portion of the
trial, a number of continuing system faults and failures were reported.

6.61 In its Weekly Situation Reports for 9 and 16 May 1998,
NAPCO reported that a number of FIDS software issues which AAs
AMD considered critical remained outstanding. There was an internal
NAPCO review of AAs Systems Availability Checklist, which identified
those systems that were crucial for AOR and identified the integration
status of each system. NAPCO also wrote to W45 Mr Kironmoy
Chatterjee, the Head of IT Department of AA, to point out its
understanding that problems continued with FIDS, that parts of PA were
still not available and that 30 percent of the telephone circuits were not
functioning.

6.62 In the Weekly Situation Report for 23 May, NAPCO
reported that little overall progress had been achieved by AA in
rectification of the remaining FIDS software problems. NAPCO noted
that the airport systems would not achieve a state of Day One operational
stability before early J une 1998, and that this represented a risk to a
smooth opening. However, AA continued to develop contingency
measures and had conducted a successful demonstration of the standby
FIDS on 22 May 1998, which was reported in NAPCOs Weekly
Situation Report for 29 May 1998. At the ADSCOM meeting on 22
May 1998, in response to a direct question from the Chief Secretary, W43
Oakervee gave specific assurance that the airport would be ready to open
on 14 J une 1998, the date of the fifth and final trial, also termed the dress
rehearsal. AA continued to give assurance that all critical software
problems had been resolved on the primary FIDS. According to W32
Mr J han Schmitz, the then Deputy CPM of NAPCO, on 29 May 1998 AA
reported that FIDS had been tested at 120 percent of its design capacity
with no major problems.

6.63 At the ADSCOM meeting on 6 J une 1998, AA reported that

101

the main airport operating systems were finally commencing reliability
testing in integrated mode. W3 Townsend also reported that the Access
Control System (ACS) had been successfully tested.

6.64 The fifth and final trial took place on 14 J une 1998, and
NAPCOs observation team again attended. In the weekly Situation
Report for 20 J une, NAPCO stated that a number of systems problems
had been observed. NAPCO noted that the Day One configuration of
airport systems, including manual modes, were now to be in place by the
week of 22 J une 1998, and that there were continued systems delays and
operational problems. NAPCO also observed that the results of the trial
were somewhat inconclusive, that many FIDS displays were not working,
that baggage from the check-in areas seemed not to be arriving at the
lower parts of the baggage system, that departure gates in the South
Concourse could not be operated properly, possibly due to problems with
ACS, and that TMR was suffering from interference. These and other
observations served as background for a note NAPCO prepared for
tabling at the ADSCOM meeting on 24 J une 1998. Although the five
trials were conducted on the basis of what the facilities would permit at
the time, slippage in construction and installation of equipment and
systems restricted the scope and value of the trials. Even though the
fifth trial, the final one, involved 12,000 participants, there was no true
airport-wide trial to test how the various systems, including cargo
handling, would interact and function together. As W51 Mr J ason G
YUEN, one of the experts appointed by the Commission pointed out in
his report, AA did not plan properly for any major failures that might
show up at the final airport trial. The last trial was only three weeks
from AOD which hardly allowed sufficient time for recovery, retrial,
training and practice should any major system fail.

6.65 On 14 J une 1998, ADSCOM members visited the new
Airport by way of AR. Assurances were given by the senior AA
management to the Financial Secretary who was acting as Chairman of
ADSCOM that within PTB all outstanding electrical and mechanical
systems works would be complete by 17 J une in Government rooms and
that in AOCC, the standby FIDS and gate allocation systems were ready,
and all cabling would be complete by 17 J une.

102

6.66 Between 22 and 24 J une 1998, NAPCO and Bechtel staff
conducted an internal review of ADSCOM Paper 34/98 entitled Airport
Operational Readiness Status Report prepared by AA for presentation to
ADSCOM on 24 J une 1998. In that report, AA stated that reliability
tests on FIDS version 2.01C commenced on 14 J une and were completed
on 20 J une using live data from Kai Tak through AODB, that the
reliability of the system as a whole has been 98.7% available, and that
the reasons for unavailability of some monitors and liquid crystal display
(LCD) boards at the 14 J une trial had been identified and the problems
were being rectified. In ADSCOM Chairmans brief of 24/6/98
prepared by NAPCO, NAPCO stated We suspect that the FIDS
problems have not been fully resolved. The 98.7% reliability of the
system is not satisfactory. In a note prepared for ADSCOM, which was
tabled at the ADSCOM meeting on 24 J une 1998, NAPCO pointed out
that while the permanent FIDS had completed a 5-day test with
satisfactory results, the system remained unstable, and that there were
outstanding cable problems and connections affecting FIDS monitors and
LCD boards at remote locations. NAPCO also noted that the standby
FIDS was reported as being ready, and that the cut-over time from the
permanent FIDS to the standby system was in the range of 45 minutes.
NAPCO further reported that ACS was still unreliable. However, in the
said ADSCOM Paper 34/98, AA gave categorical assurance that airport
systems, including FIDS, would be operationally ready in time for airport
opening on 6 J uly 1998. AA also indicated that testing had confirmed
the stability of the permanent FIDS. It recommended that the permanent
FIDS rather than the standby FIDS be used on AOD, and stated that in the
event of the failure of some FIDS functions, planned workarounds could
be depended on. AA also gave firm assurance that ACS software
problems had been resolved and the remote control of door locks was
being progressively implemented.

6.67 ADSCOM members revisited the airport on 24 J une 1998.
Again, firm assurances were delivered by the AA management that all
essential outstanding works would be completed before airport opening.

6.68 In the Weekly Situation Report for 27 J une 1998, NAPCO
noted that the outstanding number of FIDS software problems had
increased, including high priority items. Reliability tests continued, but

103

FIDS performance still showed problems of instability, with significant
outages and downtime. It also expressed concern that a real risk to a
smooth opening existed. However, AA continued to focus on
contingency measures, and the cabling for the standby FIDS, which had
been delayed, had now been completed and a reliability test had been
conducted on 24 J une 1998. Testing of the standby FIDS was to have
re-commenced by 15 J une 1998, following the system demonstration held
on 22 May 1998 and initial testing, but priority work had been focused on
the permanent FIDS. A final trial of the standby FIDS was subsequently
conducted on 30 J une 1998, as originally scheduled by AA in its 1 April
1998 presentation to ADSCOM, with representatives of 35 airlines in
attendance. The success of this trial was confirmed by AA in its
ADSCOM Paper 36/98 dated 4 J uly 1998.

6.69 The occupation permit for PTB was issued on 29 J une 1998.
Also on the same day, the Aerodrome License, effective from 1 J uly 1998,
was approved by DCA, W1 Siegel, certifying from an aviation
perspective that the new airport was in a sufficient state to operate safely
and securely. NAPCO prepared an updated assessment of the current
progress of critical airport works items having a potential impact on AOR
as at 30 J une 1998 which was subsequently tabled at the ADSCOM
meeting on 4 J uly 1998. In the assessment NAPCO reported that its
information on FIDS was that the system was down for 9% of the time
during the continuous test run between 14 and 27 J une 1998, and that was
not satisfactory, noting that the overall FIDS remained unstable, and that
ACS in PTB was unreliable. NAPCO further noted that AA was not
giving priority to the standby FIDS. However, as late as 4 J uly 1998,
AA continued to give firm assurance that any residual systems problems
had been or would be resolved, and that the systems would be in a fully
operational state, with standby contingency measures in place as required,
prior to AOD. In its ADSCOM Paper 36/98, AA confirmed that the
permanent FIDS was sufficiently stable, and that it would be used to
provide the primary flight display function in PTB. It further stated that
when some displays or functions failed, available workarounds could be
depended on, and that immediate cut-over to the standby system, which
had been satisfactorily trialed on 30 J une 1998, could be made if the
permanent FIDS failed. AA also stated its plan for using the standby
Stand Allocation System (SAS) for gate allocation, a function that

104

could, if FIDS was absolutely reliable, be performed by Terminal
Management System (TMS) as part of the permanent FIDS . AA
planned to use the standby SAS with parallel input into TMS to ensure
that terminal operations were up to date. Insofar as the Building
Systems Integration (BSI) package system which included critical
systems such as PA had not yet been fully commissioned, AA planned to
use those systems in a standalone mode. As to ACS, AA stated that
system stability had improved and that the system was now on line, with
work continuing on improving the reliability of card readers.

6.70 From NAPCOs observations on site, (conveyed to
ADSCOM via the NAPCO update as of 30 J une 1998) NAPCO was
aware that the permanent FIDS servers (two of them, one backing up the
other) had experienced outages during the extended trial, which if
repeated in operation could result in a down time of around 10 percent for
FIDS. This was likely to occur intermittently resulting in a freezing of
displayed information for the period when both servers were down. In
NAPCOs view, this might create a nuisance. At the ADSCOM meeting
on 4 J uly 1998, W45 Chatterjee, the Head of IT Department of AA, gave
the following assurances. He reported that the permanent FIDS had
been running continuously since 22 J une and was stable. There would
be workarounds when a function of FIDS went down and the
workarounds had been tested and found to work well. The switch over
from permanent system to the standby FIDS had been tested with 35
airlines. Within 30 minutes, most displays were switched on. He
confirmed that that was acceptable from the operational point of view.
During the switchover, the information displayed on the LCD boards and
monitors would become outdated, but PA could be used to disseminate
up-to-date information as a remedy. Whiteboards and extra hands
would also be available to help with directing the passengers in the
problem area when necessary.

6.71 Despite AAs assurances, W32 Schmitz, in his witness
statement to the Commission, stated that he was concerned that a smooth
opening of the new airport might be at risk. He anticipated that there
would be some initial operations problems, disruptions and inefficiencies
at the new airport, particularly in regard to airport systems and
coordination between the various operators. However, given the

105

continued assurances of AA management and staff that potential
operational problems had been catered for and that contingency measures,
procedures and resources were in place to recover from systems failures,
he was led to believe that most of the problems would be manageable and
would not unduly impact passenger or aircraft processing. He did not
anticipate the severity or compounding of problems as they actually
occurred.

6.72 W32 Schmitz also stated that his concern was not shared by
AA. He knew of no member of AAs management or operations staff
who expressed doubts as to whether the airport opening would in the end
be operationally successful. AA had made known that the airport would
not be 100 percent physically complete at opening, that construction,
rectification and fit-out works would be on-going in PTB and the Ground
Transportation Centre after opening, and that a number of systems
functions and systems integration steps not considered AOR critical had
been deferred. Yet, while AA anticipated that a number of teething
problems could be expected, it maintained a consensus of confidence.
On the basis of the assurances provided by the AA management who were
closer to the situation and had the full picture, that at least the basic
facilities, systems contingency measures were in place for the
commencement of Day One passenger and aircraft processing, W32
Schmitz did not feel that any suggestion of postponement of the airport
opening date could be justified. He maintained the same view when he
testified before the Commission.

6.73 Mr Tudor Walters wrote to the Commission regarding the
discrepancy between the 98.7% reliability and the 9% downtime of
FIDS referred to in paragraphs 6.66 and 6.69 respectively above. He
said that he did not consider the discrepancy as important, for he was of
the view that the downtime of FIDS would only cause a nuisance since
there were two host servers, one acting as a fallback for the other. He
pointed out that W45 Chatterjee stated at the ADSCOM meeting on 4
J uly 1998 that the permanent FIDS had been running continuously since
22 J une and was stable, and that switch-over from permanent to standby
FIDS had been tested with 35 airlines and was successful. He relied on
AAs assurance that FIDS was stable and that the standby FIDS and
contingency measures were fully available in case of FIDS failing. He

106

felt therefore that any intermittent outage of the permanent FIDS would
have a nuisance value, but would not be catastrophic, and he did not
believe it was necessary to specifically clarify the discrepancy. He was
in fact rather optimistic towards the end and did not raise any real
problems in the last ADSCOM meeting before AOD.

6.74 All the ADSCOM members with whom the Commission
inquired about AAs statement that the reliability of the system as a
whole had been 98.7% available, namely, the Chief Secretary, the
Financial Secretary, the Secretary for Economic Services and the
Secretary for Works, answered that they understood that to mean that the
whole of FIDS was 98.7% reliable. Had they known that the figure only
related to the availability of the hardware of FIDS and not the reliability
of the system as a whole, they would have asked the AA management at
the ADSCOM meeting of 4 J uly 1998 to explain the significance of the
figure regarding the overall reliability and stability of FIDS. Some of
them further stated that they would also have inquired as to the reliability
and stability of the standby FIDS. When the Chief Secretary testified
before the Commission, she said that she was not too concerned with the
discrepancy between the FIDS figures reported by AA and NAPCO, since
she was confident that in case of FIDS failure, resort could be made to the
standby FIDS which AA had reported to have been successfully tested on
30 J une 1998. She knew that FIDS was not quite reliable but she had
AAs confirmation that FIDS had been running continuously since 22
J une. While she expected some teething problems, she was led to
believe that there would not be any significant problems on AOD.


6.75 Having examined all the evidence very carefully, the
Commissioners find it clear beyond peradventure that the Chief
Executive was not involved in any way in the decision making of the
opening of the airport, although he approved that decision. The decision
was taken by ADSCOM which was then reported to him by the Chief
Secretary and also reported to ExCo at its meeting on 13 J anuary 1998.
In the documents disclosed by ExCo to the Commission, there was a note
of the meeting dealing with airport opening and it was to the effect that

107

the airport opening date was noted by ExCo. Apart from that, there was
a discussion on the ceremonial opening of the new airport on 1 J uly 1998.
The date was eventually altered to 2 J uly 1998 in order to prevent any
clash of the AOD ceremony with the activities anticipated for
commemorating the first anniversary of Hong Kongs reunification with
the Mainland. The Commissioners therefore conclude that the role of
the Chief Executive was merely approving the decision and is not
responsible in any way for it.

6.76 The position of ADSCOM is now examined. The Chief
Secretary testified that the policy of deciding on AOR was to have the
new airport operating safely, securely, efficiently and smoothly.
ADSCOM had never resiled from those criteria throughout the course of
its examination of the readiness of the new airport to open on AOD, from
the time before the decision was made right up to AOD.

6.77 AA awarded a great number of contracts for the construction
of various buildings and provision of various services and facilities to the
new airport. Most of the contracts had their completion dates by the end
of 1997, so that the aim of having the new airport operational in April
1998 could be achieved. In regard to HACTLs franchise for cargo
handling, however, the date of operational readiness was agreed to be
beyond April 1998, ie, 18 August 1998.

6.78 As stated before, in the franchise agreements of AAT and
HACTL, as well as in others, AA was obliged to give a three-month
advance notice to the franchisees of the date of the opening of the new
airport. While this obligation encumbered AA to have a date fixed for
the opening long beforehand, it had to examine and monitor the progress
of all kinds of works relating to the building, building services and
facilities in order to keep up to the expected opening in April 1998. On
the other hand, ADSCOM was mindful of the importance to fix an airport
opening date well in advance so that the public as well as all concerned
parties would know this date for their own purposes and planning.

6.79 As from the middle of 1997, whereas AA established a
Steering Committee on AOR to deal with issues relating to AOR,
ADSCOM also started discussions in earnest on AOR issues. In the

108

papers for discussion at the meeting of ADSCOM on 3 November 1997
(ADSCOM Papers 44/97 of 31/10/97 and 45/97 of 31/10/97), NAPCO
reported to ADSCOM the progress of various works, facilities and
services essential for the operation of the new airport. A number of
matters of concern were raised and considered at that meeting and a
subsequent ADSCOM meeting on 7 November 1997. Members of AAs
senior management were invited to attend the meeting on 3 November
1997. On the training courses planned to be provided to AAs AMD
staff, Government and business partners, W44 Heed assured ADSCOM
that all staff would be trained in time, ie, by April 1998, and he was
satisfied that the required equipment would be made available for training
purposes. W46 Mrs Elizabeth Margaret Bosher confirmed that the
priority items essential for airport operation would be available for the
airport trials. In respect of the concern expressed by Mr Tudor Walters
about the readiness of the systems at PTB, while W43 Oakervee
acknowledged that there were some problems with the BSI (which was
considered not critical for AOD), FIDS and BHS were on target. W3
Townsend also informed the meeting that the consensus of the AA Board
was that AA was able to operate the new airport in early April 1998
insofar as AA works were concerned, but given HACTLs progress, the
end of April would be a more suitable opening date.

6.80 The ADSCOM meetings on 3 and 7 November 1997
culminated in the Chief Secretary, as the Chairman of ADSCOM, sending
a letter dated 15 November 1997 to the Chairman of the AA Board, W50
Wong. W50 Wong by letter of 10 December 1997 replied, dealing in
length with the items raised in the Chief Secretarys letter.

6.81 In order to simplify matters, it is only necessary to mention
the three issues which gave rise to the utmost concern of ADSCOM at the
time, namely, the availability of FIDS to the new airport, the progress of
the works relating to HACTLs premises, (that is, ST1 and the Express
Centre) and the readiness of AR when the airport was expected to open
for operation. FIDS was at that time just installed in the new airport,
and was undergoing testing and commissioning. However, many
significant problems in the systems operation were found, and the worry
was whether it would be operating reliably when the airport opened.
HACTLs premises were still under construction, and there were reports

109

that the progress of the construction works had slipped with many weeks
of delay, and it was feared that HACTLs services would not be available
in April 1998. ADSCOM was also at about that time investigating with
MTRC regarding whether AR would be ready for use in April 1998.

6.82 Regarding the first two of these three significant matters,
W50 Wong on behalf of the AA Board assured ADSCOM that the
problems with FIDS and HACTL would be duly resolved and both of the
services provided on these scores would be ready in April 1998. AAs
assurance to ADSCOM regarding the readiness of HACTL was based on
assurances given by HACTL to AA that HACTL, with the supplemental
agreement with its main contractor, GPY, would be ready for operation in
April 1998 with 50% of its throughput capacity, which percentage
together with the share of cargo handling that was to be done by AAT was
assessed at that stage as to be sufficient to cater for the cargo facilities
required of the new airport. On the other hand, as MTRC maintained
that the project regarding AR could not be properly accelerated to be
ready in April 1998, AA was making contingency transportation measures
with the assistance of related Government departments that the new
airport would be ready to open in April 1998.

6.83 On 12 J anuary 1998, the AA Board had a meeting in which
the date that the AA Board members considered being desirable for
opening the new airport was discussed in detail. When the resolution to
open the new airport was put to a vote, the ex-officio members of the
Board abstained because they thought it only proper that the non-official
members should make the decision without their participation as
Government officials. The decision reached was that the new airport
was ready to open in the last week of April 1998.

6.84 On the part of ADSCOM, in addition to the two meetings on
3 and 7 November 1997 referred to above, it held no less than four more
meetings, namely those on 15 November, 8 and 20 December 1997 and 2
J anuary 1998, before it finally decided on AOD. The decision was that
the new airport should open on 1 J uly 1998, and that day was to be for the
ceremonial opening. At a special meeting of ADSCOM on 8 J anuary
1998, NAPCO suggested and ADSCOM accepted that 6 J uly 1998 was to
be the operational commencement date of the new airport.

110

6.85 Whilst there was a certain amount of scepticism harboured
by ADSCOMs members as to the airport being able to be ready in April
1998, mainly caused by the problems still encountered with FIDS and the
delay regarding the completion of the HACTL buildings, the main reason
for the decision to defer the expected date of April 1998 to 1 J uly 1998
was the inability of having AR ready earlier than 21 J une 1998. This
was despite the fact that there were contingency transportation plans in
hand to provide adequate transportation facilities to link the new airport
with other parts of Hong Kong. ADSCOM was of the view that if the
new airport was to open with AR ready for operation, rather than relying
on a scheme of makeshift transportation facilities, the operation of the
new airport would be more smooth and efficient. The added period
from April to 1 J uly 1998 would undoubtedly give more time to AA and
HACTL to make FIDS and the cargo handling services better equipped,
both of which were expected to be ready by the end of April. Based on
the considerations of operating the new airport in a safe, secure, efficient
and smooth manner, ADSCOM made the decision.

6.86 ADSCOM further considered that it would be necessary or
at least more advisable for the airport to open operation on a Monday and
reached the conclusion that 1 J uly 1998 should be the ceremonial opening
date whereas Monday, 6 J uly 1998 should be the operation
commencement date. There was an enormous relocation plan which
was divided into five phases, the first one starting on 8 May 1998 and the
penultimate one on the eve of the operation opening of the new airport.
This relocation exercise was very important because it was only on the
eve of the operational opening of the new airport that all the facilities and
personnel from the Kai Tak airport that were required for CLKs
operations, which or who were to remain in service until that eve, would
have to be completely moved to CLK. The fifth phase would only take
place after AOD. The phase 4 relocation would require participation of
over a thousand heavy-duty vehicles, barges and aeroplanes, and
ADSCOM accepted the advice that traffic on the roads between Kai Tai
and CLK would be the lightest in the evening of Sunday night. The
added advantage of opening the airport for operation on a Monday was
that air traffic was the lightest on Monday and that would give a little
more time and practice to the personnel and facilities at the new airport to

111

gear up to the heavier air traffic expected of the days following.

6.87 Immediately after the decision was made by ADSCOM, the
Chief Secretary informed the Chief Executive of the decision and the
major reasons behind it. A paper was prepared for the ExCo meeting to
be held on 13 J anuary 1998. During the ExCo meeting, its members
were informed of the decision and the reasons which they noted. The
Commission finds that the Chief Executive and the ExCo members noted
the decision and the reasons, and none of them expressed any view on the
subject. The Chief Executive and the ExCo members should properly
be treated as having approved the decision but not being involved in any
process of the decision making.

6.88 Further, there is no evidence whatsoever that the decision to
open the airport in J uly 1998 was a result of any political consideration or
ulterior motive. There was only one single occasion during the
ADSCOM meetings in November 1997 through to J anuary 1998 where
the officiating guest for the ceremonial opening was mentioned, and it
was on 15 November 1997, before the decision on AOD was reached.
The Chairman of ADSCOM stated that members should aim for the
President of the Central Peoples Government as the officiating guest. In
ADSCOM Paper 48/97 of 6/12/97 for discussion at the meeting on 8
December 1997, NAPCO proposed that the Chief Executive should be
invited to sound out the President on the role of the Principal Officiating
Guest as soon as possible. Those were the only references in all the
voluminous documentation on the decision on AOD that could possibly
be imagined as having any political flavour. Eventually, obviously in
order to avoid clashing with other activities to celebrate the first
anniversary of Hong Kongs reunification with the Mainland, the
ceremonial opening of the new airport was rescheduled to 2 J uly 1998.

6.89 Since the decision on AOD was announced through the
media on 13 J anuary 1998, subsequent reports, either those orally made
during meetings or in writing through papers to ADSCOM, from AA and
NAPCO never hinted to ADSCOM that there were such risks on AOR
that AOD should be altered. No one ever mentioned to the Chief
Secretary or ADSCOM that there should be a consideration of deferring
AOD. From all the voluminous contemporaneous documentary

112

evidence that has been supplied to the Commission, there is no record
that such an idea had been raised by anyone. ADSCOM was still
overseeing the progress of the facilities and services essential for the
operation of the new airport on AOD, but no one within or outside
ADSCOM had ever intimated, let alone warned, that there was any risk
of difficulty regarding the airports safe, secure, efficient and smooth
operation. On the contrary, AA was confident that PTB was ready for
operation. It assured ADSCOM as late as 4 J uly 1998 that the
permanent FIDS was stable, reporting that the standby FIDS had been
successfully tested and confirming that various workarounds were well
prepared. Apart from the assurances on a best endeavour basis given by
the HACTL management that 75% throughput capacity could be
achieved on AOD, W7 Charter, HACTLs Managing Director, also
assured the Financial Secretary during the latters visit to ST1 on 14 J une
1998 that HACTL would be ready for AOD.

6.90 When the Chief Secretary gave evidence before the
Commission, she categorically told the Commission that no one had ever
informed her of any risk of allowing the new airport to be open for
operation on 6 J uly 1998 or raised with ADSCOM a deferment of that
date, and from what she was informed throughout her chairmanship of
ADSCOM she did not anticipate that the new airport would open with so
many and so serious problems.

6.91 In her witness statement to the Commission, the Chief
Secretary made the following concluding remarks:

(a) The ACP was the most ambitious and complex
infrastructural project ever undertaken by the Hong Kong
Government. ADSCOM was established to provide the
necessary oversight and perform essentially a trouble shooting
role, ensuring that all parties involved worked towards the
ultimate goal of delivering the entire programme within
budget and on schedule. ADSCOM performed its functions
in a proactive and critical manner, working together with AA,
MTRC and the relevant Government departments in
maintaining momentum, keeping costs under control and
resolving the many interface problems that arose.

113

(b) From the start, ADSCOMs intention was to build a world
class international airport that the community could be proud
of. ADSCOM was aware of the serious operational
problems that had confronted some new airports elsewhere on
opening, with the consequent adverse publicity which took
time to overcome. It was anxious to avoid this happening
with the new airport. It wanted a safe, secure and efficient
airport of international standard on Day One and these
considerations were borne uppermost in mind when deciding
on the opening date. No political considerations were
involved.

(c) In all the circumstances, and having regard to AAs
assurance that end April 1998 was a realistic opening date,
ADSCOM felt that a J uly opening would give greater
confidence. ADSCOM accepted that on first opening there
would inevitably be some teething problems but it believed
that the airport could cope reasonably well with both
passengers and cargo on Day One. Unfortunately this did
not prove to be the case. Any decision to postpone the
opening of the airport should not be taken lightly. However,
had there been the slightest suggestion from AA or HACTL
that the new airport could not cope with either the passenger
or the cargo flow on Day One for any reason at all, ADSCOM
or Government would not have hesitated to reconsider the
airport opening date.

6.92 The Chief Secretary was cross-examined as to her statement
that ADSCOM would not have hesitated to reconsider AOD had there
been any suggestion that it should do so. Her attention was drawn to the
word irreversible in her letter to W50 Wong when discussions about the
proper date for airport opening were on going at the end of 1997, as well
as her utterance as noted in an ADSCOM meeting during the same period
that once AOD was decided, she would expect everyone to stick to it.
The Chief Secretary explained that these were confidential letter and
notes and were to convey the meaning to AA that once decided, all
persons concerned should exert their utmost to work towards meeting the

114

target. She said that there would be consideration to defer AOD, for
example, when a typhoon was expected, which would undoubtedly have
an adverse effect on the relocation exercise from Kai Tai to CLK. She
also said that if the Airport Control Tower was on fire, the opening must
be deferred. W32 Schmitz gave another example : failure of AA to
obtain an aerodrome licence. The Commissioners find that AOD by its
nature simply could not be irreversible but, as pointed out by the Chief
Secretary, it all depended on the cogency of the reasons in support of a
deferment. In his testimony, W43 Oakervee, one of those who were
privy to the confidential correspondence and notes of meetings between
Government and AA, also denied that the date fixed by Government was
irreversible in the sense that it could not be altered. He recognised
that AOD, when decided, was the common target towards which
everyone should work. While the AA Board was adamant that late April
1998 should be the opening date, ADSCOM rejected the idea and instead
chose 6 J uly 1998 so as to allow sufficient time to have AR ready to
provide a smooth and efficient transportation service to users of the new
airport. Of course, once decided, AOD should not be lightly changed,
for it was a decision creating the certainty on which many people such as
airport operators relied. Nonetheless, if sufficiently weighty material
was proffered, the Commission has no doubt that ADSCOM would
certainly consider whether a deferment was necessary. The unfortunate
thing was that no one ever suggested a deferment or put situations before
ADSCOM that would, at the time, justify a revisit of the decision.

6.93 There was a suggestion by HACTL to various witnesses that
there should be a soft opening of the new airport in that Kai Tak should
be retained for HACTL to partially handle cargo for a short period of time
after AOD before ST1 was put into full operation. That was also raised
in the testimonies of W7 Charter and W2 Mr YEUNG Kwok Keung, the
Managing Director and Deputy Managing Director of HACTL. In fact,
the idea was discussed in correspondence between HACTL on the one
hand and PAA and Government on the other in August 1995. At the
time Government requested HACTL to be ready to provide 1.2 million
tonnes of cargo handling capacity by April 1998, but if HACTL required
operating partially at Kai Tak, Government would consider such
requirement sympathetically. However, HACTL did not vigorously
pursue the idea with ADSCOM or Government eversince that time.

115

Instead, HACTLs management gave assurances to AA and ADSCOM
that ST1 would be ready to have 75% throughput capacity on AOD.
When the announcement of AOD was made on 13 J anuary 1998, W2
Yeung was reported by the media as saying that HACTL welcomed the
added time from the original target of April to J uly 1998 and felt relieved.
HACTL never hinted that it was not able to be ready for AOD. It was
imbued with confidence about the systems the development of which it
was responsible and proud of.

6.94 Having considered all the evidence and counsels
submissions very carefully, the Commissioners find that ADSCOM did
not make any mistake in deciding that 6 J uly 1998 should be the date for
the operational opening of the new airport. Indeed, ADSCOM members
had exercised great care and diligence in reaching that decision. The
main reason for ADSCOMs selecting J uly 1998 for opening the new
airport was to await the completion of AR, and that was despite AAs
insistence that all critical AOR items would be ready by late April 1998.
MTRC maintained that the scheduled completion date of AR being 21
J une 1998 could not be abridged, and ADSCOM decided to have the
transportation service of AR available to airport users when the new
airport opened instead of leaving them to resort to makeshift
transportation facilities before AR was ready. Moreover, the added time
of over two months between late April and early J uly would surely
provide a comfortable float to the PTB and HACTL projects. The only
reasonable conclusion that the Commissioners can reach is that it was a
proper and wise decision.

6.95 During the period between J anuary 1998 after the decision
was made up till AOD, ADSCOM exerted no less efforts and care
regarding the progress of the AOR issues. The continuous assurances
given by AA and HACTL had lulled ADSCOM members into a false
sense of confidence and security, resulting in their not revisiting the
opening date. Indeed, no one had ever even uttered a word that they
should do so. While everybody within AA, HACTL and NAPCO who
provided witness statements to the Commission or testified before it was
of the view that teething or settling problems might be expected, none
ever anticipated the chaotic conditions that occurred on AOD. All
concerned were taken by surprise. The Commissioners feel that it

116

would be improper and unreasonable to hold ADSCOM or any of its
members responsible for not appreciating the risks of keeping to AOD in
the then prevailing circumstances.

6.96 The Commission has also examined the position of
ADSCOM vis--vis the other bodies involved in order to decide on the
responsibility of ADSCOM regarding the problems encountered on AOD
and thereafter. ADSCOM is a high level committee for the purpose of
making policy and strategic decisions in respect of ACP including the
opening of the new airport. It has NAPCO as its executive arm.
NAPCO is responsible for monitoring the work of AA, and contributing
towards the coordination between AA and Government departments.
NAPCO had site engineers to observe and critically assess the
performance and progress of the works carried out by various contractors
employed by AA, and it also had a CPM from outside Government. The
CPM used to be Mr Tudor Walters, and on his retirement in J uly 1998,
W32 Schmitz replaced him. Both Mr Walters and W32 Schmitz were
senior officers of Bechtel, an international firm of airport consultants.
As ADSCOM is basically a policy decision-maker, the every day
monitoring of the performance and progress of the airport works was
reposed in NAPCO. It would be unreasonable to expect members of
ADSCOM to visit the site at CLK daily or even once in a while to
attempt to exercise physical monitoring over the works. On the other
hand, NAPCO with its site engineers and professionals should be able to
observe the physical condition, assess the performance and progress of
the works, and report to ADSCOM anything which should be
ADSCOMs concern.
6.97 Indeed, ADSCOMs serious concern about the readiness of
FIDS and HACTL before the decision on AOD was made in J anuary
1998 was prompted by NAPCOs reports and critical comments through
papers and oral discussions at the ADSCOM meetings. After J anuary
1998, NAPCO had not slackened in its efforts and continued to report on
the progress of the works, and critically assessed the progress of the
essential AOR issues. The Commissions deliberation on NAPCOs role
and duties is contained in Section 2 of Chapter 5. Having examined all
the evidence and submissions very carefully, the Commission finds that
there was generally sufficient communication and coordination between

117

ADSCOM and NAPCO. W36, the Chief Secretary told the Commission
that she was satisfied with NAPCOs work and she was of the view that
NAPCO had discharged its functions effectively and conscientiously.
The Commission opines, however, that in performing its functions
towards ADSCOM, NAPCO failed in two aspects. First, it should have
inquired with AA whether it had the necessary know-how in monitoring
HACTLs progress relating to the installation, testing and commissioning
of the CHS equipment and systems, instead of merely assuming that AA
was so qualified, but it failed to do so. Secondly, it should have checked
whether AA had plans and contingency measures and should have had an
overall assessment whether such plans and measures were adequate in
view of the then prevailing circumstances. As a corollary, NAPCO
should also examine whether AA had an overall risk assessment.
NAPCO should be responsible for these omissions. The detailed
analysis can be found in paragraphs 5.66 to 5.68 of Chapter 5.
ADSCOMs responsibility has been dealt with in paragraphs 5.70 to 5.74
of Chapter 5. The evidence shows that ADSCOM had the duty of an
overall monitor and it had delegated the duty of the overall monitor of the
progress of AOR to its executive arm, NAPCO and directed it to
discharge the duty. The Commissioners find that towards the public,
ADSCOM should be responsible for NAPCOs failure.

6.98 The responsibilities of the AA Board and management
regarding AOD will be dealt with in later chapters when their
involvement on the problems encountered at AOD will be addressed in
detail.

118

CHAPTER 7

AIRPORT OPENING THE PROBLEMS AND THEIR
DEBILITATING EFFECT

7.1 The decision of the Commissioners on whether the new
airport was ready to open on airport opening day (AOD), 6 J uly 1998,
depends on their examination of the problems encountered on that day
and shortly thereafter. If only a few minor problems had surfaced, the
airport would obviously be viewed as ready to open for operation on
AOD. Despite the efforts of all involved or interested parties, this was
not to be, unfortunately.

7.2 Numerous problems occurred. Many of the problems were
inter-related and intertwined, and it tends to confuse the observer as to
what was the cause and what was the effect or consequence. In this
chapter, the Commission identifies the problems encountered, describes
their debilitating effect, individually and collectively, and narrates the
situation on all fronts.

7.3 The ceremonial opening officiated by President J iang was on
2 J uly 1998, immediately following the celebration and commemorative
activities for the anniversary of Hong Kongs hand-over that had taken
place the day before. Phase 4 of the relocation exercise in which all the
required facilities and equipment were moved from the Kai Tak airport to
the new airport at Chek Lap Kok (CLK) was performed in the night of
5 and early hours of 6 J uly 1998, in order to make the new airport fully
operational in the morning of 6 J uly 1998. There were occasional
drizzles that night, but as anticipated and planned, traffic on the roads
between Kai Tak and CLK was light, and the relocation was smooth with
little occurrence worthy of concern to anyone interested.

7.4 The first flight was to arrive at the new airport at about 6:30
am on 6 J uly 1998, AOD. The Chief Secretary for Administration of the
Hong Kong Special Administrative Region, who is also the Chairman of

119

Airport Development Steering Committee and some of her fellow
members, the Chairman of the Board of the Airport Authority (AA) and
many of his colleagues, as well as the Chief Executive Officer of AA and
his staff were all there waiting for the first arriving passengers to present
them with souvenirs to mark the occasion. Most of the receiving group
had little sleep the night before, because of their involvement one way or
another with the relocation process. Nothing untoward happened when
a flight from New York landed, and the ceremonial reception of the first
couple to arrive was none other than a pleasing event.

7.5 Shortly after 8 am, after clearance at the immigration counter,
a number of arriving passengers had difficulty finding out where to
retrieve their luggage. The Flight Information Display System (FIDS)
monitors and the liquid crystal display (LCD) boards that were situated
at the entrance to and inside the Baggage Reclaim Hall (BRH) were
either blank or displaying the number of an incorrect reclaim belt on
which baggage was to appear. Those who noticed the few whiteboards
on which reclaim belt numbers were written had to crowd around these
whiteboards to obtain the information, and after that, had to wait for a
long time for their baggage. At the same time, people who came to the
new airport to meet their relatives, friends or customers were not able to
get correct flight information on the monitors and the LCD boards that
were supposed to show all relevant flight information, including flight
status (ie, the plane has arrived or not), the estimated time of arrival, the
actual time of arrival and the meeting gate (ie, either Gate A or Gate B
from which the passengers will exit to the meeting and greeting area).
They were not able to estimate when those whom they expected would
come out. Obviously, all these caused inconvenience and anxiety to the
arriving passengers and their meeters.

7.6 In the meantime, departure passengers were not provided
with boarding gate numbers that should normally appear on boarding
passes. The airline staff could not help, for they did not have the
required information either. All were not able to get any or accurate
flight information regarding the outgoing flights on the monitors and
LCD boards, such as the estimated time of departure, the actual time of
departure and the boarding gate number. Sometimes, information was
displayed on some monitors and LCD boards, but was inconsistent with

120

that shown on other monitors. The situation was the same both on the
landside and the airside (where access by the general public is denied) of
the huge Passenger Terminal Building (PTB). After the departing
passengers had gone through immigration and security clearance, they
reached the airside where AA provided two whiteboards situated at the
Information Counter in the Departures Hall. On these whiteboards,
current flight information that AA staff could get would be shown.
However, the assistance of the whiteboards was rendered only to those
departing passengers who happened to inquire with AAs attendants or
chanced to see the whiteboards. Many people were jostling around the
whiteboards seeking information while AA staff were busy putting on the
most current information. Those passengers who did not know of the
existence of the whiteboards were anxious and confused. There were
also many changes to the gates allocated. Many passengers were
running up and down the large Departures Hall looking for the right
boarding gate for their flight. They had to make great haste to find the
proper gate which they would otherwise normally do leisurely. Some of
them even missed the plane. The same lack of flight information and
status affected airline staff who were either manning the check-in desks
or working elsewhere in the airport, causing confusion and inconvenience
to them.

7.7 Arriving aircraft experienced delay in having a gate or a
remote stand allocated to them for parking. The delay was aggravated
for aircraft that arrived later, because of domino effect. From around
noon to 5 pm and from 8 pm to 11 pm, the apron was full and arriving
aircraft had to wait on the taxiway for a stand to be allocated because a
stand would only be available after another flight had departed. The
delay was so bad that some planes had to stay on the tarmac for over an
hour, when the passengers were not allowed to alight and the air in the
plane became stuffy so as to cause nausea or dizziness to some occupants.
When the plane was able to park at a remote stand, it would have to wait
for the passenger steps and buses. With some of the planes that parked
at the terminal stands, their passengers were exasperated when the
airbridge linking the plane door to the arriving gate was unable to operate.
The late parking of the planes and the stalled disembarkation of arriving
passengers compounded the delay suffered by departing passengers.

121

7.8 Even when arriving passengers were successful in getting off
the plane, their plight was to wait for two to three hours for their baggage,
and some even had to leave the airport without their baggage, which
could not be found. As from about 10 am, the baggage handling
operator experienced difficulty in assigning reclaim belts by the use of
the Flight Information Display System (FIDS) workstation, and the
reclaim belts were not assigned to flights in time or at all. The arriving
passengers had no information from FIDS or from airlines on the proper
baggage reclaim belt, resulting in anguish and frustration. Although the
several whiteboards at the BRH alleviated the situation, the passengers
were exasperated because their baggage would arrive at the belt late.
Three ramp handling operators (RHOs) were to handle departure,
arrival and transfer bags. They had to go to the aircraft to unload the
baggage and bring it to the Baggage Hall on level 2, a sort of basement
area, of PTB. Not that RHOs were not trying to serve the passengers,
but they simply could not cope. Due to the inefficiency of FIDS, they
did not know where the aircraft they had to serve were, and finding them
took effort and time. The stand changes and serious flight delays also
compounded their difficulties and strained their resources. They did not
know which baggage lateral was assigned for them to bring arrival
baggage to. There were thousands of departure or transfer bags
swamping around the problem bag area in the Baggage Hall. Early in
the morning, about 420 bags arrived at the new airport from Kai Tak, 220
of which were fed into the Baggage Handling System (BHS) without
baggage labels. These bags could not be read by BHS, and they became
problem bags that had been outcast from the normal departure laterals
and ended up at the problem bag area. Various problems with baggage
handling on AOD created about 6,000 problem bags lying around in
Baggage Hall, and the place was an eyesore.

7.9 RHOs had to provide services for the disembarkation of
passengers, by operating the airbridges that linked the arrival gates to the
planes or to provide steps for the aircraft parked at remote stands. They
did not know what stands were allocated to the aircraft they served, and
they had to drive around on the tarmac or send runners to the airfield to
locate the planes. Even after stand allocation and other necessary flight
information was available at the Airport Emergency Centre after 4 pm,
RHOs had to send their staff there to obtain such information. These

122

staff attempted to relay the information to their operation staff through
Trunk Mobile Radios (TMR) or mobile telephones, but these facilities
frequently failed them. Delayed by difficulties obtaining flight
information from FIDS and other means of communication, RHOs were
late in providing steps and buses and passengers had to wait to get
disembarked and taken to the PTB.

7.10 When passengers, those accompanying them and those
meeting them were put in a situation without reliable flight information
and were stranded for hours in PTB, they turned to the telephone. There
were also many sightseeing visitors to PTB on the new airports opening
day. To their dismay, many of the public telephones in PTB, both at
landside and airside, were not working. Mobile phones were also
employed, but with unsatisfactory results, because the networks were
overloaded. Overloading resulted from competing demands of users and
operators of the new airport at the same time. The passengers and
visitors who were stuck in PTB did not feel too comfortable because the
temperature was insufficiently low. The air-conditioning was not
functioning properly. Many who used the toilets found that they had to
wait for quite sometime and the facility was over-crowded. Some of the
toilets were filthy, as the flushing system did not always function, tap and
flushing water was occasionally cut off, and there were too many users.
Some had to walk up or down the steep steps of certain escalators that did
not work, and the signage showing directions did not fully satisfy users
needs. Restaurants and refreshment facilities were overcrowded and
refuse bins spilled over. All these added insult to injury.

7.11 Those who attempted to get to the new airport by public
transport found themselves in a jam, for the road works surrounding PTB
were not fully completed. There were long lines of buses congesting at
and close to the bus stop at Cheong Tat Road outside PTB, and there were
long queues of people, arriving passengers and visitors alike, waiting to
board buses. There was congestion in passengers flow as alighting
passengers did not know how to get into PTB since the passenger lifts
and escalators had not yet become operational.

7.12 On the cargo front, the situation was not at all better. There
were two cargo terminal operators (CTOs), namely Hong Kong Air

123

Cargo Terminals Ltd (HACTL) and Asia Airfreight Terminal Company
Ltd (AAT). Apart from serving aeroplanes and passengers, RHOs
also transported cargo to either of these CTOs. They brought the cargo
they unloaded from aircraft on dollies (ie, a kind of wheeled platform
dragged by a motor tractor) to AATs or HACTLs premises to hand over
the cargo. Due to the smaller size of AATs operations, handling about
20% of all cargo, the problems were not too serious. It was very
different with HACTL. RHOs were often not able to find HACTLs
staff at the interface area for cargo transfer. They originally left the
cargo on the dollies at the interface area, and returned to their other duties.
However, as time went on, they found that the dollies that they had left
before at the interface area were still occupied by cargo and they could
not retrieve them for further use. As a result, they put the cargo on the
dollies onto the ground and retrieved the dollies. Hundreds, if not
thousands, of pallets and containers of cargo occupied the interface area
between the ramp and the two CTOs premises, which in turn caused
difficulties and delay to AATs and HACTLs personnel to identify, locate
and remove.

7.13 It was not an exaggeration to describe the new airport as a
pandemonium for men, cargo and aeroplanes resulting from failure of
computer systems. Aeroplanes had to queue to park and be served by
RHOs, RHOs tractors, vehicles, dollies and equipment were working
furiously to try to cope, FIDS including its monitors and LCD boards
functioned hopelessly, the air-conditioning went slow, some escalators
ground to a halt, and the Cargo Handling System (CHS) at HACTL
failed to respond properly. Baggage and cargo did not often succeed in
finding their way to their owners or custodians and when they did they
were long overdue. For many hours, there was a full apron of
aeroplanes, and outside and inside PTB, there were crowds of people,
either running around to look for the proper gate to board planes or
waiting for baggage, friends and relatives, and transport. Although the
staff of AA, airlines, RHOs, and cargo handling operators were working
hard, such work was to little avail, for passengers and visitors remained
greatly inconvenienced, annoyed and alarmed at the services provided.

7.14 The chaos went on for a few days, although some of the
problems might not be noticed by particular individuals. Many

124

problems were identified through the complaints of users of the new
airport and through media reports. Each of these problems is dealt with
in later chapters. They are summarised below:

(1) The unreliable working of FIDS, with malfunctioning
of monitors and LCD boards that were supposed to
show flight information
(2) The breakdown of cargo handling by HACTL which
eventually imposed an embargo on most import cargo
and some export cargo
(3) Delay in flight departure and arrival
(4) Malfunctioning of the Access Control System
(5) Confusion over parking of planes
(6) Malfunctioning of Aircraft Parking Aids
(7) Late arrival of tarmac buses causing delay to
disembarkation of passengers
(8) Insufficient passenger steps and miscommunication
amongst staff handling the same
(9) Airbridges malfunctioning
(10) Chaos in baggage handling including malfunctioning of
monitors and LCD boards in the baggage reclaim area
(11) Public Address System malfunctioning
(12) Airside security risks
(13) Congestion of vehicular traffic and passenger traffic
(14) No tap water in toilet rooms and tenant areas
(15) No flushing water in toilets
(16) Toilets too small
(17) Urinal problems with water flow, infrared sensors and
cleanliness
(18) Insufficient air-conditioning in the PTB
(19) Large number of public telephones not working
(20) Mobile phone service not satisfactory
(21) TMR service not satisfactory
(22) Escalators breaking down repeatedly
(23) Insufficient or ineffective signage
(24) Slippery and reflective floor
(25) Problems with cleanliness and refuse collection
(26) No sufficient water, electricity and staff at restaurants

125

(27) Insufficient staff canteens
(28) Automated People Mover stoppages
(29) The Airport Express (AE) ticketing machines
malfunctioning
(30) The AE train delay.

7.15 AAT had difficulty to deal with all the cargo it was
supposed to handle on AOD. On 18 J uly 1998, AAT successfully made
arrangements with the nearby Airport Freight Forwarding Centre to use
the latter for breakdown, storage and collection of the backlog cargo, and
thenceforth, the severe congestion at AATs terminal started to abate.
The backlog that had been built up from Day One was cleared by 13
August 1998. It took AAT over six weeks to return to normality.

7.16 The CHS of HACTL was operating slowly and inefficiently
throughout the day on 6 J uly 1998. A huge backlog of cargo was built
up along its interface on the airside to the north of SuperTerminal 1
(ST1), HACTLs main premises. In the early morning of 7 J uly 1998,
the inventory of cargo kept by the computer disappeared. The whole
CHS had to be shut down for checking and repairs. All the cargo that
had been moved there from Kai Tak and from the ferry flights arriving at
CLK on 6 and 7 J uly had to be transferred to HACTLs premises and
facilities at Kai Tak for processing. On 8 J uly 1998, HACTL, which
handled about 80% of cargo exported and imported at the new airport,
announced an embargo on all imported cargo save for a very small
number of items. Export cargo had to be sent by shippers to Kai Tak for
processing before they would be transferred to the new airport to board
the flights. The embargo was extended to 9 J uly and then to 18 J uly but
it was further extended with its four-phase recovery programmes. Even
though the amount of export cargo processed by HACTL at CLK grew
gradually, it was not until 23 August 1998 that all import and export cargo
could be normally handled by ST1.

7.17 The embargo and the delay in handling cargo gave a hard
time to cargo owners, importers, exporters, shippers, freight forwarders,
and all those who made a living out of air cargo. While many perishable
goods simply became rotten and had to be discarded, some cargo was
missing. The businesses of the people in these fields were seriously

126

disrupted, and many suffered substantial financial loss.

7.18 The furore that was brought by the opening of the new
airport was replaced with initial dismay and eventual disappointment.
The inconvenience experienced by passengers and visitors and the
hardship suffered by the air cargo trades received intensive and extensive
coverage of the media. Many were complaining about the unreadiness
of the new airport to operate on AOD, others were crying shame on the
lack of efficiency, and the rest were estimating the size of the financial
losses. All of them converged to profess that Hong Kong had not only
lost substantially in money, but suffered severely in its position and
reputation as the most efficient city in the world and as South East Asias
hub of international civil aviation for both cargo and passenger.

127
CHAPTER 8

THE STANDARDS OF CLASSIFICATION FROM TEETHING
TO MAJOR PROBLEMS

8.1 In paragraph 7.14 of the preceding chapter, all the problems
facing the new airport on airport opening day (AOD) and shortly
thereafter have been identified. However, after the commencement of
the inquiry, there were allegations of further problems, namely:

(a) rats found in the new airport;
(b) an arriving passenger suffering from heart attack not
being sent to hospital expeditiously on 11 August 1998;
(c) emergency services failing to attend to a worker nearly
falling into a manhole while working in Passenger
Terminal Building (PTB) on 12 August 1998;
(d) fire engines driving on the tarmac crossed the path of
an arriving aircraft on 25 August 1998;
(e) traffic accident on 28 August 1998 involving a fire
engine, resulting in five firemen being injured;
(f) on 3 September 1998, a maintenance worker of Hong
Kong Aircraft Engineering Company Limited
(HAECO) slipped on the stairs inside the cabin of a
Cathay Pacific Airways Limited (Cathay Pacific)
aircraft;
(g) on 6 September 1998, a Hong Kong Airport Services
Ltd. (HAS) tractor crashed into a light goods vehicle,
injuring five persons;
(h) on 8 September 1998, a power cut occurred, trapping
passengers in lifts and on the Automated People Mover
(APM) as well as delaying two flights;
(i) missed approach by China Eastern Airlines flight
MU503 on 1 October 1998;
(j) tyre burst of United Arab Emirates cargo flight EK9881
leading to runway closures on 12 October 1998;

128
(k) power outage of SuperTerminal 1 (ST1) due to the
collapse of ceiling suspended bus-bars on 15 October
1998; and
(l) radio frequency interference on Air Traffic Control
frequency.

8.2 According to the methodology of the Commission, while
minor problems that occurred after its appointment would generally not
be dealt with in any detail, problems that might be serious were to be
looked into unless they occurred too close to the conclusion of the inquiry
as to make investigation impracticable. For the purpose of
concentrating on serious or major problems, the Commission thought it
proper and reasonable only to spend time and energy on minor or
teething problems to the extent that they are due. Not only did the
Commission have a deadline of six months to meet, and that manpower
and money should not be deployed to deal with insignificant or
inconsequential matters, minor problems should reasonably be viewed as
not having any significant effect on the users of the new airport or the
reputation of Hong Kong.

8.3 There should, therefore, be an initial evaluation or
classification of the problems identified so as to rank their seriousness
whereby the priority in which time and manpower to be spent in inquiring
into them should be properly placed. The classification is also necessary
for better appreciation of the degree of impact that each problem had on
the operations of the airport and on passengers as well as airport operators.
For this purpose, W51 Mr J ason G YUEN and W55 Dr Ulrich Kipper
were most helpful in contributing their respective perspectives on the
standards of quality regarding airport services and facilities, whereas the
Commissioners would determine the appropriate yardstick using their
knowledge of circumstances in Hong Kong and as visitors to various
airports around the world.

8.4 Having considered the opinions of W51 Yuen and W55
Kipper, and bearing in mind the facts as found from the evidence that the
Commission accepts, the Commissioners are of the view that the
problems encountered on AOD and shortly thereafter can be divided into
three classes: minor problems (which include teething problems),

129
moderate problems and major problems. While the categorisation must
be a matter of degree, certain criteria are adopted to judge the proper slot
into which a problem should be put. This criteria is set out below. The
Commissioners freely quote the experts views as adopted by them.

8.5 Despite the scientific and technological advancement in
recent years, human foresight is still very limited, so that not everything
one handles will turn out the way that is foreseen. Good planning is
therefore important in the development of a project, but good planning
does not necessarily mean perfect planning, and problems that cannot be
reasonably foreseen or contemplated by the human brain will often occur.
Further, sometimes it is necessary to run new machines and systems for a
period of time for them to operate smoothly. That is why a new motor
car has to be run-in for several thousands of kilometres after which its
various parts have to be checked, corrected, repaired or replaced. An
airport is thousands of times more complicated, with extremely complex
facilities. W51 Yuen opines that the startup of a new major airport will
inevitably encounter various glitches, malfunctions, mishaps and
technical problems. These abnormalities are sometimes prevalent, and
usually accepted by airport operators and users as minor inconveniences
at startup as teething problems, similar to the temporary pain a child
must go through during the eruption or shedding of teeth. They are not
desirable but are generally viewed as facts of life. W55 Kipper
describes teething problems as those that will inevitably occur during the
first few days of operation, irrespective of the conscientiousness of the
pre-operational testing and commissioning activities. He states that
even if all problems exposed by testing and airport trials before airport
opening have been solved, new problems will occur after airport opening
because of the complexity of all linked airport processes and the
unavoidable discrepancies between simulated and real live load.

8.6 To what extent the abnormalities are considered teething or
minor problems depends on the service standards and the tolerance level
of the airport operators and users. For a premier international airport
such as that at Chek Lap Kok (CLK), the Commissioners adopted the
opinion of W51 Yuen that the startup abnormalities should be no greater
than those experienced at other world-class airports of similar size.
W55 Kipper contributed his views from a European angle. The

130
following are the criteria taken as proper and reasonable for assessing the
seriousness of each of the problems encountered on AOD. For a
problem to be determined as a teething problem, it must meet each and
every one of the following criteria:

(a) Criterion 1 The nature of the problem must not
involve safety. Any malfunctioning that endangers the
personal safety of people should not be considered as a
teething problem, regardless of whether it has occurred
at any other airport, or how commonly it has occurred.
A distinction must be drawn between problems that
cause inconvenience or minor financial loss or loss of
business opportunities against those that put the lives of
people at risk.

(b) Criterion 2 The magnitude of the problem must be
limited. The problem must not be pervasive in relation
to the size of the field. For example, if the terminal
has 50 escalators and three malfunctioned on opening
day, that may be considered a teething problem.
However, if the terminal has only six escalators and
three malfunctioned, there may be something
intrinsically wrong with the escalators.

(c) Criterion 3 The inherent cause of the problem must
not be a fundamental failure. Problems caused by
poor workmanship, misalignment, defective parts,
accidental damage, operator error and the like can be
considered as teething problems. However, problems
caused by design errors, improper planning,
mismanagement, and gross negligence are not teething
problems.

(d) Criterion 4 The recovery period for overcoming the
problem must be short, with rare or no subsequent
recurrence. As a rule of thumb, a teething problem
should not last more than a few days. In cases where
permanent corrective work requires more time,

131
temporary measures must be able to alleviate the
problem within a few days. Moreover, there should
only be rare or no recurrence of the problem after the
recovery period, other than the normal rate of failure
occurring universally.

(e) Criterion 5 The problem must be beyond the
operational norms of an airport. Certain problems will
occur in an airport under normal circumstances. For
example, on the average, one can expect two bags out
of 10,000 to be misdirected or lost. Misdirecting two
bags out of 10,000 should be considered as a fact of life
at the airport, and not a startup problem, teething or
otherwise.

8.7 Although a problem may not be able to satisfy all the five
criteria as to be a teething problem, it can still be considered a minor
problem insofar as it satisfies Criteria 1 and 2 above, as its impact on
users is not widespread or reasonably perceived to be serious. They
normally do not attract widespread public attention and are soon forgotten
when the situation is rectified. If the problem is pervasive, affecting
almost all airport users or a certain large class of airport users, or it
seriously affects an area of airport operation that is considered to be
important, then it is a major problem. In between, there are moderate
problems, where the impact of the problem is widespread, affecting a
large number of users, or the problem is in respect of an area reasonably
considered to be an important area of airport operation, but if the degree
in each of these qualifying factors is less than that for a major problem,
then the Commissioners will classify it to be moderate. It should also be
mentioned that some problems are classified as moderate although
normally they may be classified as minor except for the fact that the
extent and length of inconvenience caused to airport users are much
larger and longer.

8.8 Applying the above criteria, the Commissioners are of the
view that the abnormalities set out hereunder as experienced at the new
airport are teething or minor problems. These problems are grouped
together because there should be little difference in their treatment. The

132
reasons for their conclusion regarding each can be found in paragraphs
8.9 to 8.24:-

[8] APM stoppages
[9] Airport Express (AE) ticketing machine
malfunctioning
[10] AE delays
1998
[22] Traffic accident on 28 August 1998 involving a fire
engine, resulting in five firemen being injured
[23] A maintenance worker of HAECO slipped on the stairs
inside the cabin of a Cathay Pacific aircraft on 3 September
1998
flights
on 1 October 1998

133
8.9 [1] Mobile phone, [2] TMR, and [3] Public telephones not
working. There was no extraordinary or abnormal problem with the
TMR network used by Airport Authority (AA) inside PTB. Among
the other networks, TMR and normal mobile phone services provided and
operated by Hutchison Telecommunications (Hong Kong) Ltd
(Hutchison) experienced problems with system overload. China
Motion United Telecom Limited also experienced congestion in its TMR
network, while SmarTone Mobile Communications Limited and Hong
Kong Telecom CSL Limited also experienced system overload.
However, the huge number of users of the two types of services were not
reasonably anticipated when their capacity was planned, because there
were a large number of visitors, stranded passengers, ramp handling
operators (RHOs) staff and airline personnel using the systems on
AOD due to the inefficiency of Flight Information Display System
(FIDS). Had the necessary flight information been available through
FIDS, RHOs and airlines would not have needed to use TMR and mobile
phones so frequently, passengers and those greeting and meeting them
would not have had to stay in PTB for hours, and the demand on the
services of the two types of networks would not have exceeded their
capacity. The problem was exacerbated because only about one-third of
the public telephones planned to be made ready on AOD were operational.
There were 322 payphones installed in PTB on AOD, but only 118 of
them were operational. Out of the 43 courtesy phones installed on AOD,
only 32 were functioning. Although there was minor malfunctioning of
phone equipment reported by AA, the main cause for the out of service
phones was incomplete cable connections. Insufficient telephones
would normally only cause a little inconvenience and some waiting time
for users. The unexpected demand on AOD, due to the large number of
delayed passengers and visitors, however, made the unavailability of a
large number of public telephones more serious. Subsequently, the
following numbers of telephones became operational: from 150 on AOD
to 329 on 20 J uly, 377 on 25 J uly, and a total of 382 on 3 August 1998.
Hutchison, as with the other network operators, completed the
enlargement of the capacity of their respective mobile phones and/or
TMR services within a few days after AOD, and the problems have not
since recurred. There were problems, on the other hand, with poor
signal strength in some airline offices and portions of the ramp. They
are fundamental inadequacies and therefore cannot be treated as teething

134
problems. However, these problems were rectified by relocating
antennae and installing new ones very soon after AOD, and as such they
were minor. Most of the planned number of public telephones were
ready for operation by 25 J uly 1998.

8.10 [4] Escalators. Two out of 61 escalators were not operating
on AOD. In respect of the 59 escalators in operation, there were 20
incidents of stoppages on that day and 19 such incidents on the following
day. The stoppages were due to one of three causes: (a) improper setting
of safety device, (b) foreign articles jamming the units, and (c) people
pushing the emergency stop button not in a case of emergency. Causes
(b) and (c) are normal occurrences at airports or in public buildings.
The safety device under (a) was set at too sensitive a level, so that slightly
heavier load would trigger a stop. The sensitivity level was adjusted on
8 J uly 1998 having taken into account the actual working condition in the
new airport and no further stoppage was caused by this reason. This is a
classic teething problem.

8.11 [5] Insufficient or ineffective signage. During the first
three airport trials that took place in J anuary, February and March 1998,
comments from participants were collected with the consequence that a
number of signs were redesigned and ordered, albeit some were only
ready after AOD and then had to install without delay. There were
1,500 signs in PTB on AOD and they all worked well, except one with an
arrow pointing the wrong direction. The design of the statutory signs
required by Government departments such as the Fire Services
Department and Buildings Department had all been approved by the
departments and their installation was complete before AOD. The
design of the directional signs was based on the logic of the usage of
various parts of PTB for which the signs were installed. The principal
confusion relating to signage was caused by visitors or departing
passengers wishing to go to the Departures Hall from the Arrivals Hall, as
their wish was not in accordance with the logic of the usage of the
Arrivals Hall where only few signs would show them the way to the
Departures Hall. On AOD, however, most of the confusion among the
passengers and visitors was caused by the unfamiliarity with the new
facilities, an operational change that diverted departing passengers into
the Arrivals Hall, and the FIDS failure, and not by signage. Furthermore,

135
the misdirected sign as mentioned above was corrected in one day.
Temporary signs were put up to alleviate the confusion on AOD and
permanent signs when available were subsequently installed. As opined
by W51 Yuen, the signage philosophy adopted by the architect, as
discussed below, may have also contributed to the publics complaints.
Signage additions, revisions, and refinement is quite common among
major airports after the terminal is put to actual use. There are two
schools of thoughts in airport terminal signage: (1) use the minimum
number of signs necessary to direct people in order to reduce clutter for a
more aesthetically pleasing terminal environment. Most architects
prefer this philosophy unless the airport operator has a strong voice in the
design; and (2) provide signs for directions as above, but in addition to
these necessary signs, install signs in between as backups in case the first
sign is missed. These additional signs also serve as confirmation of
direction to assure people that they are proceeding in the right direction.
Most airport operators and airport users prefer this second school of
thought, even though there may be some sacrifice on aesthetics due to
cluttering. However, the philosophy of signage in the new airport was
based on the first school instead of the second one.

8.12 [6] Slippery and reflective floor. Like the signage situation,
the first few airport trials brought about many criticisms on the floor of
PTB. It was pointed out that the polished Zimbabwe Black granite
insets and borders of the flooring were slippery and too reflective, and
they pose a physical risk and a source of embarrassment for ladies
wearing skirts. As a result, instructions were given by AA to contractors
to hone the reflective and slippery granite. After some honing was done,
the work was considered to be too time-consuming. It was subsequently
decided that a proprietary material should be applied to the granite to
reduce its lustrous and slippery nature instead. Only five incidents of
persons slipping were reported and two of these were associated with
water on the floor and only one man was required to be and was duly sent
to hospital as he received a small wound in his arm. In view of
approximately six million people using PTB during this period, the rate of
these incidents did not appear out of the ordinary. However the
reflective nature of the black granite can still be a cause of complaint,
especially from ladies wearing skirts.

136
8.13 [7] Problems with cleanliness and refuse collection. There
was substantial build-up of rubbish on AOD due to AA and tenant
contractors construction activities. This build-up of rubbish during the
final days of construction is typical of major projects. However, due to
the scale of the work at CLK, even 400 day shift and 150 night shift
refuse cleaners working two weeks prior to AOD could not keep up with
the build-up. All contractors cleaning labour were deployed starting 7
J uly. On AOD, there were a large number of visitors and delayed
passengers using the catering facilities on AOD extensively. The
resulting cleaners, such as used food boxes and unfinished food, found
their way around refuse bins that were full. The amount of rubbish was
unexpected because the number of users of PTB for a lengthy period was
not foreseen. The problem continued on 8 J uly due to access problems
for staff and vehicles removing rubbish. Most rubbish was cleared by
10 J uly.

8.14 [8] APM stoppages. There was one incident at around
11:30 pm on 20 J uly 1998 where one passenger and four airport staff
were trapped in the APM, which was the means of fixed track
transportation situated in the basement of PTB for moving passengers
along its east and west parts. The line is about 800 metres long, with
two parallel tracks joined at the ends by a loop. Train No. 3 arrived at
the West Hall departures station where all passengers were supposed to
alight. One passenger and four airport community staff got on board
when all passengers had alighted. Train No. 3 proceeded as
programmed to the West Turnback, but stopped when it detected that
Train No. 2 had stalled at the West Hall arrivals station. It appears that
the people trapped in Train No. 3 tried to force open the doors which set
off an alarm to the Airport Operations Control Centre. Before the
maintenance staff arrived to restore Train No. 3, a trapped passenger tried
to pry open the door by turning the emergency door release valve and
eventually got onto the emergency walkway. For safety reasons, the
APM operator immediately shut down traction power in the tunnel and
the passengers were escorted to the West Hall departures station. The
problem was caused by the passenger and the staff getting on board
despite announcements not to do so. This occurrence could have been
prevented if security staff had been stationed at the platform to ensure
that no one boarded the train there. This precaution was subsequently

137
adopted by AA soon after the incident. There has been no recurrence.
There were also reports of minor stoppages of the APM due to door
failures resulting from doors being forced open by passengers, or trains
undershooting or overshooting their designated stops. To reduce the
sensitivities of doors to passenger interference, the contractor increased
the time between the door circuit receiving a signal of interference to
door failure. By the end of August 1998, the contractor has not received
any further reports of door failures. Station attendants are also present
to ensure that APM passengers do not force open doors.

8.15 [9] AE ticketing machine malfunctioning and [10] AE delays.
When AE went into operation, 41 out of 52 ticketing machines were
operational. A number of them could accept bank notes but could not
handle coins or give change. Since service counters were available for
purchasing tickets, and staff were assigned at the machines to provide
change, inconvenience to the public was kept to a minimum. Moreover,
since the number of machines installed was based on future growth, the
number of working units was probably sufficient to meet the demand.
All machines were working properly by 24 J uly 1998. There were
minor and singular train service disruptions on 9, 11, 14, 23 and 27 J uly,
when passengers were transferred to another train. The most serious
disruption to the AE service occurred on 23 J uly when the Tsing Yi
station declared a red alert resulting in the temporary suspension of the
temporary suspension of the Tung Chung Line and a 20-minute frequency
for AE. On 27 J uly, AE was delayed for 19 minutes due to a signal error.
These incidents appeared to be isolated cases of equipment failures at the
startup of a new and major rail system, causing relatively minor
inconvenience to passengers.

8.16 [11] Late arrival of tarmac buses. There were delays of
tarmac buses in meeting the demand for transporting passengers between
the terminal and remotely parked aircraft. However, the delays were
generated by: (a) lack of accurate and prompt flight information due to
the deficiency of FIDS; (b) reallocation of aircraft stands and increased
usage of remote stands due to delayed flights and malfunctioning of the
airbridges; (c) communication difficulties due to overloading of TMR and
mobile phones; (d) poor co-ordination between boarding gate assignment
and location of aircraft; (e) usage of tarmac buses as boarding lounges in

138
lieu of holding rooms in terminal; and (f) failure of the Access Control
System (ACS). Without these factors, tarmac buses might not have
been delayed in meeting passengers or at least the delay would not have
been inordinate.

8.17 [12] Aircraft parking confusion. On 6 and 7 J uly 1998,
aircraft stand allocation had to be performed manually due to the
problems with SAS and TMS. Stand allocation was made with delay,
mainly due to the inefficiency of FIDS and TMS. The extended stay of
aircraft due to delays in flight departures eroded parking capacity and
made the allocation task more difficult. Since the vacancy of parking
stands was uncertain, it was nearly impossible to plan parking
assignments effectively. Further compounding the problem was
communication problems among operational staff. In addition,
problems such as the malfunctioning of some airbridges and ACS doors,
insufficient towing tractors due to the amount of aircraft repositioning
required, non-familiarity of push-back procedures by some tractor drivers,
pilots not fully familiar with the apron, taxiways and remote stands and
so farther also exacerbated the problem.

8.18 [13] Insufficient ramp handling services. Two passenger
disembarkation methods were employed at the new airport, via airbridges
when the aircraft was parked at frontal stands at PTB, and via mobile
steps and tarmac buses when the aircraft was at a remote stand. The
problems with airbridges are discussed under item [14] below, and the
late arrival of tarmac buses are discussed under item [11] above. Since
the tarmac buses and the mobile steps work together in the
disembarkation of passengers, most of the discussion on the problem with
tarmac buses also apply to mobile steps.

8.19 [14] Airbridges malfunctioning. There were serious delays
in disembarking passengers via airbridges. Operators could not arrive at
the airbridges in a timely fashion, and the delay was mainly caused by the
lack of flight information due to the slow response of FIDS. Swipe
cards for operating the airbridges malfunctioned and were replaced by
keys two days prior to AOD. However, insufficient keys were issued,
inconveniencing the airbridge operators and resulting in delay in
disembarking passengers. There was also a programming error in the

139
software for controlling the airbridges. The misplacement of one line of
programming code caused incorrect sequencing of the bridge components,
which caused intermittent alarms and airbridge malfunctioning for some
B-747 aircraft. The malfunctioning of doors on airbridges, which were
controlled by ACS, also contributed to the operators inability to access
the airbridges. HAS, one of the RHOs, which operated the airbridges,
reported that emergency glass had to be broken to release a door to allow
passengers to get through.

8.20 [15] No tap water in toilet rooms and tenant areas, [16] no
flushing water in toilets, and [17] urinal flushing problems. Prior to
AOD, Tank Rooms 3 and 8 were manually operated by the contractor due
to defective valves. The valves are used to regulate water flow. This
would not have caused any problems with the supply of potable water.
However, on AOD, airport security measures were implemented and the
contractors personnel were denied access to the tank rooms. The tanks
became empty and no potable water was supplied to the northern part of
the East Hall, the North Concourse, West Hall, North West Concourse,
and South West Concourse. The low water level alarm did not activate
because the Building Management System, which was considered not to
be an airport operational readiness required function, was still not
operational. On 7 and 8 J uly, there was no potable and flushing water in
the toilets and some tenant areas in the southeast side of PTB. These
toilets were served by Tank Room 2, which was flooded. For safety
reasons, the electrical control panel which operated the pumps was
switched off causing interruption to the water supply. Water supply was
restored by 8 J uly. Even before AOD, the plumbing contractor reported
problems with urinals due to (a) flow of flushing water, (b) problem with
sensors, (c) blockage of urinals, and (d) cleanliness of toilets. In
particular, the urinal problems included accumulation of sediment in the
valves, improper setting of sensors, accumulation of rubbish due to public
misuse and the low level of cleanliness of toilets. Sediment in water
supply is an airport maintenance problem. Malfunctioning due to
improper setting of sensors is a workmanship issue. Rubbish in urinals
due to public misuse is common in a busy airport. Cleanliness of toilets
is a janitorial operation issue. By October 1998, rectification to the
flushing system for the toilets were made and generally, very substantial
improvements have been achieved.

140

8.21 [18] Toilets too small. On AOD, people needed to queue
up for toilets. Apart from the many visitors whose number was beyond
the expectation of AA, there were many passengers who were stranded by
the late flight departures, and meeters and greeters who were affected by
the late arrivals. The washroom services were therefore very much
stretched. The philosophy of the toilet design can be explained due to
large size of PTB. Smaller toilet blocks were built and scattered at
reasonable distance so that those who wish to use them would be able to
find one in the near distance. If large toilet blocks had been built, they
would have been a long distance apart and therefore difficult to find as
space is a valuable commercial asset in PTB. The area of the toilets was
also not designed to be large enough to accommodate baggage carts.
This was also not an issue in the airside departures area and the
pre-immigration area on arrival as no baggage cart is provided in these
places. For other areas, it was assessed that only passengers travelling
alone would have no one to look after their baggage on carts while using
the toilet. But such passengers could use the toilets after they checked
in their luggage, or before they retrieved their luggage from the reclaim
belt. If they really needed to bring in a baggage cart, then they could
use the toilets for the disabled which were large enough. Although
small toilets do cause some inconvenience to this particular kind of
passengers, the Commissioners accept the design as being reasonable in
the circumstances, especially taking into account the many toilet blocks
that scattered not too far away from each other which would convenience
most passengers. There were some complaints about the size of the
toilet cubicles, especially those for ladies. W43 Mr Douglas Edwin
Oakervee explained that the 150-mm wide ledge over the cistern at the
back of the toilet could be used to keep ones hand luggage. The small
inconvenience caused by the small sizes of the toilets and the cubicles is
considered to be minor.

8.22 [19] Insufficient water, electricity and staff at restaurants.
Water supply: the discussion for lack of potable water in toilets under
items [15] and [16] above are also applicable for restaurants in the new
airport. Electricity supply: there were electrical outages for short
periods of time as tenant contractors switched off power to perform their
work, and by tripping of circuit breakers due to faulty tenant work. AA

141
also had to disrupt power supply to increase capacity of power
distribution system due to unanticipated electrical power load from
tenants. Staffing problems: services in restaurants were unsatisfactory
due to shortage of staff to handle the volume of business. For landside
restaurants, the surge of visitors during the first week of airport opening
taxed the facilities beyond expectations. For airside restaurants, some
employees did not obtain security passes in time to allow them to report
to work. The problem was further compounded by the inexperience and
the unfamiliarity of the new workers with the new environment. These
problems occasionally caused inconvenience to users for about a week
since AOD.

8.23 [20] Rats. In the middle of August 1998, during the course
of the Commissions inquiry, it was reported by the media that there were
rats pestering the new airport, sometimes even causing damage to electric
cables by their nipping, and some newspapers accentuated the problem by
using out the vernacular name of Rat Island to describe CLK. AA had
in fact commissioned pest control in early May 1998, and rats should not
therefore be a problem of great concern.

8.24 [21] Emergency services failing to attend to a worker nearly
falling into a manhole while working in PTB on 12 August 1998, [22]
Traffic accident on 28 August 1998 involving a fire engine, resulting in
five firemen being injured; [23] A maintenance worker of HAECO
slipped on the stairs inside the cabin of a Cathay Pacific aircraft on 3
September 1998; [24] A power cut occurring on 8 September 1998,
trapping passengers in lifts and on the APM as well as delaying two
flights; and [25] Missed approach by China Eastern Airlines flight
MU503 on 1 October 1998. These problems are treated as minor
because each of them was an isolated incident or accident involving one
person or only a few people, with effect neither extensive nor widespread,
and the evidence does not point to any operational problem of the new
airport. The details about them can be found in Chapter 9.

8.25 Out of all the problems identified, only the above 25 items
are considered to be teething or minor. More details of their causes and
the party responsible are provided in Chapters 9 and 16. The most
serious problems that were pervasive in relation to the operation of the

142
new airport were the inefficiency or unreliable working of FIDS, together
with the malfunctioning of monitors and liquid crystal display (LCD)
boards which were to display flight information, the breakdown of cargo
handling by Hong Kong Air Cargo Terminal Limited, which eventually
imposed an embargo on most import and export cargo, and the chaos in
baggage handling including the malfunctioning of monitors and LCD
boards in the baggage reclaim area. These three items were closely
connected with the delay in flight departure and arrival, confusion over
parking of planes, late arrival of tarmac buses causing delay in
disembarking passengers, insufficient ramp handling services and
miscommunication amongst staff handling the same. The three items
were therefore treated as deserving the greatest attention of the
Commission and its counsel, as each of them caused great financial loss
to various sectors of the Hong Kong business community, or created
serious inconvenience to passengers using the new airport, or severely
affected the reputation of Hong Kong. They are therefore classified as
major problems.

8.26 All other problems that are not major or minor are classified
as moderate in accordance with the criteria set in paragraph 8.7 above.
Some of these moderate problems are so categorised because they could
have serious consequences in risking personal safety or safety of aircraft,
such as the malfunctioning of Aircraft Parking Aid, fire engines driving
on the tarmac crossing the path of an arriving aircraft and an arriving
passenger suffering from heart attack not being sent to hospital
expeditiously. Others relate to the important aspect of security of the
new airport. Fortunately, these problems only occurred once, and no life
or security was jeopardised. All these problems, save a few of them, did
not last long as remedial measures were taken to resolve or rectify them,
otherwise the consequences would have been serious. Some problems
are included in this category because they affected an extensive class of
airport users or they lasted for a considerable time. Moderate
problems are set out below:

[27] Malfunctioning of ACS

143
[31] Public Address System (PA) malfunctioning
[33] Radio frequency interference (RFI) on air traffic
control frequency
[34] Aircraft Parking Aid malfunctioning: including a Cathay
Pacific aircraft was damaged when hitting a passenger
jetway during parking on 15 J uly 1998
[35] An arriving passenger suffering from heart attack not
being sent to hospital expeditiously on 11 August 1998
[37] HAS tractor crashed into a light goods vehicle, injuring
five persons on 6 September 1998

8.27 [26] Flight delays. There were significant delays in both
incoming and outgoing flights on AOD and the next few days. Their
pervasiveness and magnitude disqualify them from being teething
problems. On AOD, there were 213 incoming flights. 51% of them
arrived early or on time when not taking into account of the holding time
of aircraft on the taxiway. Of all the incoming flights, 7% was delayed
within 15 minutes, 23% within 30 minutes, 36% within 60 minutes and
13% more than 60 minutes. The average delay was 24 minutes. There
were 207 outgoing flights, and all were delayed. 3% of them were
delayed within 30 minutes, 13% within 60 minutes and 87% more than
60 minutes. The average delay for all departure flights was 2.63 hours.
On 7 J uly 1998, 62% of outgoing flights was delayed more than 60
minutes. However, these delays were not problems in themselves.
Rather, they were the results and consequences of other problems such as
the inefficiency of FIDS, difficulties in baggage handling, and
malfunctioning of the airbridges and ACS etc. Such lengthy delays are
normally beyond the expectation of passengers, and obviously affected a
large number of passengers, causing them inconvenience and anxiety.

144
8.28 [27] Malfunctioning of ACS. The most serious impact to
the public were incidents of arriving passengers not being able to enter or
exit to and from terminals. There were altogether five incidents of
arriving passengers not being able to exit the airbridges leading from the
plane door to PTB, apparently all due to the malfunctioning of the ACS.
The ACS was to control access to the airbridge so that only authorised
personnel with a swipe card can open the door between the airbridge and
the Arrivals or Departures Hall in PTB. The malfunction rendered the
swipe card useless in opening the doors. Other malfunctioning of doors
caused significant inconvenience to AA and tenant employees. Total
incidents reported: 44 in the first week from AOD, 29 the second week,
48 the third week, and 57 the fourth week. Although security guards
and airport personnel were posted to operate doors manually so that no
public safety was compromised, problems remained more than a month
after AOD, and the incident trend line showed no sign of diminishing
occurrences from AOD to end of J uly. The efficiency and productivity
of the airport staff and tenant employees were impacted when doors
malfunctioned, thus exacerbating other problems by prolonging the
response time. Due to the fact that security guards were posted
immediately to manually operate the doors affected until defective parts
were replaced and the doors tested, this problem is bordering on minor
and even teething, but for the fact that ACS was still not fully completed
and tested at the time of the hearing. Moreover, the ACS problem was
linked with other airside security problems, such as 90 transit passengers
of China Airlines Ltd. being allowed to enter the Departures Hall without
security check, and 55 cases of unauthorised access to the restricted area
between 6 J uly and 17 October 1998. Had ACS been properly
functioning, these other security problems would probably not have
occurred, albeit they did not pose real security risks. This is the main
reason why ACS is considered to be a moderate problem.

8.29 [28] Airside security risks. On 10 J uly, police motorcycles
sought entry into a security restricted area in response to a traffic accident
with two workers slightly injured. Two ambulances were allowed entry
into the restricted area but not the police motorcycles. The security staff
followed established procedures in denying access for a non-emergency
vehicle which had no permit. The police motorcycles had no permit and
could not be considered as emergency vehicles since they had no siren

145
nor flashing lights. Whilst it can be said that security procedures were
followed, there may be issues of ambiguity in these procedures and
miscommunication between the relevant authorities. On 25 J uly, airline
staff took some 90 transit passengers from the aircraft to Departures Hall
directly, without security screening. This was a violation of procedures
for screening transit passengers. W51 Yuen states that in the imperfect
world of airport security, this type of human error does happen
occasionally without fanfare. However, if the ACS had been
functioning properly, the door between the aircraft apron and the
Departures Hall would be locked, preventing access to the transit
passengers. Between 6 J uly and 17 October 1998, there were 55
reported cases of unauthorised access to the Airport Restricted Areas.
The vast majority of these cases involved staff failing to bring permits,
failing to display them, or using colleagues permits for convenience.
On 8 J uly, a KLM Royal Dutch Airlines (KLM) flight departed with the
baggage of two passengers who were not on board. During the boarding
process, it was discovered that the equipment used to scan boarding
passes was not working properly and manual collection and checking had
to be carried out. It became apparent later that some passengers were
missing and the cabin crew conducted a passenger head count, which
turned out to be equivalent to the final number of passengers checked-in.
At this time, the flight was already an hour behind schedule.
Subsequently two passengers showed up at the boarding gate when all
doors had already been closed. It was not until then that the boarding
staff realised that the headcount was inaccurate. This incident was
investigated by Civil Aviation Department which found that KLM had
breached aviation security requirement by not ensuring that the relevant
baggage was removed.

8.30 [29] Congestion of vehicular traffic and passenger traffic.
The traffic congestion on the roadways around PTB and passenger
congestion in PTB were caused mainly by the unexpectedly large number
of visitors, the non-completion of paving construction works and
confusion among passengers inside PTB on AOD. W51 Yuen states that
extraordinary increase in traffic on opening of major airport facilities is a
common occurrence due to drivers circulating the roadways to find their
destination. The problems were resolved when visitors subsided,
additional signs installed, paving works completed and better traffic

146
management implemented. However better traffic planning would have
had kept the problems at bay. Although the period affected was not too
long, the problem caused inconvenience to a large number of people and
it is therefore considered to be a moderate problem.

8.31 [30] Insufficient air-conditioning in PTB. There were a
number of incidents of air-conditioning failure inside PTB, with the
duration and cause, according to AA, set out below:

Duration Cause

4 hrs, 6 J uly Shutdown of one to three operating
chillers due to reasons such as
defective flow switch and low pressure
switch fault (see Appendix VIII for a
diagrammatic presentation of events)

30 min, 10 J uly one of three chillers shut down due to
insufficient water flow caused by
operator error

2.5 hrs, 12 J uly two of four chillers shut down due to
operator error

7-9 hrs, 13 J uly all four chillers shut down due to
lightning strike

45 minutes to all chillers tripped off due to lightning strike
2.5 hrs, 28 Aug affecting the power supply; first chiller
resumed within 45 minutes and
remaining chillers resumed within 2.5
hours

1.3 to 3.3 hrs, all chillers tripped off due to loss of sea water
29 Aug supply; first chiller resumed within 1
hour 20 minutes and remaining chillers
resumed within 3 hours 20 minutes

147
45 minutes to all chillers tripped off due to lightning strike;
2.75 hrs, 30 Aug first chiller resumed within 45 minutes
and remaining chillers resumed within
2 hours 45 minutes

1 hr, 8 Sept all chillers tripped off due to power
failure caused by tripping of circuit
breakers; all chillers resumed in an
hour

4 hrs, 14 Sept all chillers tripped off due to human
error while another contractor carried
out testing on another system; all
chillers resumed within 4 hours

2 hrs, 12 Oct all chillers tripped off due to China
Light & Power Company Limited
system disturbance causing the tripping
of the air handling units and the chillers;
all chillers resumed in over 2 hours

1 hour 10 mins, 28 Nov all chillers tripped off due to loss of sea water
supply; all chillers resumed progressively over
1 hour 10 mins

The not infrequent tripping of the chillers in PTB and the importance of
air-conditioning in Hong Kong especially during the summer months
render the problem more than minor. W51 Yuen opines that aside from
the chiller outages, the publics perception that air conditioning was
inadequate could have been a result of the systems design temperature of
24
o
C instead of a more acceptable 22
o
C. Passengers in airport terminals
generally prefer a slightly cooler temperature than normal due to walking
and luggage toting activities. There were also incidents of no or
insufficient air-conditioning in some tenants premises. The inadequate
air-conditioning in tenant areas was caused by miscalculation and poor
co-ordination of tenant construction works. Some tenant contractors fit
out work was behind schedule for AOD. This delay, in turn, caused late
applications for chilled water connections and compressed the workload

148
of AAs air-conditioning contractor just prior to AOD. There were also
faulty works by the tenant contractors which caused delays in supplying
the chilled water. The problem lasted from AOD to around 13 J uly.

8.32 [31] PA malfunctioning. PA consists of Central PA and
Local PA, the definitions of which can be found in paragraph 12.135 of
Chapter 12. On AOD, central PA was down twice. On 7 J uly 1998, it
was down for six times including one which lasted 2 hours and 5 minutes.
On 8 J uly 1998, five occurrences of downtime were experienced. Local
PA was more unstable than Central PA. During the first week of
operations, 26 gate rooms experienced problems with Local PA. There
were 21 reports of Local PA problems in various zones in the second
week, 25 in the third week, and 122 in the fourth week. Intermittent
problems of local zone PA consoles continued beyond the last week
September. Problem logs from 4 August to 20 September show
numerous local PA problems occurring virtually daily in a random
fashion. When Local PA fails, Central PA can take up its function.
However, since AOD even Central PA had failed from time to time.
Although some reported problems were caused by operators errors, most
were due to the malfunctioning of PA (equipment damage or system fault).
A meeting was held ten days after AOD to develop programme for
completion of outstanding work. Site acceptance tests of PA were not
completed until the end of October 1998. There was also an inherent
problem with the acoustics of PA. Even when the system was
functioning, the barrel vault ceiling structure caused echoes which
reduced the clarity of the announcement. W51 Yuen opines that
boarding instructions at gate rooms are basic requirements for airport
operations. From an airport and airlines operational point of view, any
problem with such system should not be allowed to continue for over a
month. This would normally be a minor problem but for the fact that
FIDS was not working properly on AOD, and there were a number of
gate changes. The malfunctioning of PA aggravated the already chaotic
situation.

8.33 [32] Insufficient staff canteens. The new airport has a
working population of about 44,629, with about 14,600 people working
daily in PTB. Some employees had to wait more than 40 minutes for
food and table. For the first two weeks of airport opening, only one

149
canteen with a seating capacity of 250 was in operation. A second
canteen opened on 14 J uly, and the third on 29 J uly, with the total
capacity being about 800. Another canteen opened on 15 October,
increasing the total capacity to 954 people. W51 Yuen opines that the
ratio of the number of workers to canteen seats vary from airport to
airport depending on the eating habits of the workers (purchasing food
versus bringing their own). However, the ratio of about 15 to 1
(assuming 14,600 in the main shift to 954 seats) appears very low. If,
for example, half the workers do not bring their meals, there would be
over two workers per seat at the same time, even if it is assumed that they
stagger their meal breaks in three periods. Since food prices in airport
public restaurants are higher than normal, workers are not expected to use
them every day. Facilities planning for airports should take into account
the needs of the passengers, visitors, as well as the employees serving
these passengers and visitors. As a large number of people working at
the new airport are affected, this problem is classified as moderate. AA
has given evidence that improvements are being planned to build more
meal facilities for staff.

8.34 [33] RFI on air traffic control (ATC) frequency.
Problems with RFI on the Very High Frequency radio communication
channels of the ATC were reported as far back as late 1994. The sources
of RFI were in the form of spurious signals originated from some
unknown paging stations along the coastal areas in the Guangdong
Province. Hong Kong Government has raised this issue with the
relevant Mainland authorities since December 1994. Since then,
remedial measures have been taken by Hong Kong and the relevant
Mainland authorities. To address the problem, affected frequencies were
replaced and six additional frequencies were used by ATC as extra
backup to further safeguard flight safety since 1996. With the spare
frequencies available, air traffic operations at the new airport have not
been affected. This is an important matter for air traffic safety and it
explains why the problems is classified as moderate.

8.35 Problems [34], [35], [36], [37], [38] and [39] are mainly
isolated occurrences. While item [39] will be dealt with in paragraph
11.15 of Chapter 11, the remaining items are dealt with in Chapters 12
and 15. All these problems are not considered to be minor because they

150
either involved injuries to several people or related to areas of importance
such as the safe operation and security of the new airport.

151
CHAPTER 9

TEETHING AND MINOR PROBLEMS
AND REMEDIAL MEASURES

9.l Chapter 8 sets out briefly the problems which the
Commissioners regard as minor or teething. This chapter deals with
each of these problems in detail, outlining their causes and remedial
measures. The responsibility for each of these teething and minor
problems is reviewed in Chapter 16.

[1] Mobile Phone Service Not Satisfactory

9.2 Mobile phone services play an important role in the
communications system for the new airport. Unfortunately, services
were plagued by network problems on the airport opening day (AOD),
causing difficulties to airport operators and inconvenience to passengers
and the public using the airport.

9.3 There are 11 mobile phone networks sharing the use of a
Common Antenna System (CAS) inside the Passenger Terminal
Building (PTB). SmarTone Mobile Communications Limited
(SmarTone) installed the CAS, while individual mobile phone
operators designed and installed their own equipment to deliver their
respective service. According to SmarTone, the CAS was designed to
cover the public area and the VIP Suite of PTB. The CAS was
commissioned for about two weeks and accepted by participating mobile
phone operators before AOD. Each operator was responsible for using
and monitoring capacity of its own equipment. The responsibility of the
Airport Authority (AA) was to ensure completion of the physical
installation of the antenna system.

9.4 According to AA, some mobile phone users in PTB
152
continually received busy signals because a number of networks were
overloaded on AOD. The peak period of the problem occurred on AOD
when flights were delayed, airlines, ramp handling operators (RHOs)
and other operators were having difficulties getting flight information and
there was chaos in baggage handling. As the situation improved after
AOD and more channels were added to the network, mobile phones
services quickly returned to normal.

9.5 The extent of overloading of a mobile phone network is
measured by the blocking rate, which indicates the unsuccessful rate of
incoming and outgoing mobile phone calls for each sector cell of the
antenna network during a certain period. SmarTone admitted that its
own mobile phone network had a high blocking rate on AOD. While a
blocking rate of 5% is acceptable by industry standards, the design
guideline of SmarTones system provides for a blocking rate of less than
2%. The blocking rate on AOD was as high as 79.9%, falling to 4.7%
on 7 J uly and to 0.35% on 8 J uly 1998. SmarTone argued that the rapid
fall in the blocking rate showed that the SmarTone system was designed
to cater only for a reasonable volume of mobile communication traffic in
PTB. This argument is not entirely correct. According to the
statement of Mr Alan MOK Kai Chau, Senior Manager, Radio Network
Engineering Department of SmarTone, SmarTone increased the number
of channels in the CAS in a bid to ease the problem on 7 J uly 1998. The
overloading of communication channels were unexpected, resulting
directly from the high number of mobile phone call attempts made on the
day. SmarTone pointed out that the high usage of the mobile phone
system was due to the large number of stranded passengers in PTB and
the fact that many public telephones in PTB were not in service at that
time. This accords with the facts found by the Commission.

9.6 Hutchison Telecommunications (Hong Kong) Ltd
(Hutchison) operates three mobile phone networks at the new airport
and experienced similar problems on AOD. Hutchisons system was
launched on 22 J une 1998. According to Mr Edmund SIN Wai Man of
Hutchison, overloading problems were experienced on the Global System
for Mobile Telecommunications (GSM) network. Records from
153
Hutchisons traffic statistics database show that the relevant cells of the
GSM network had blocking rates of between 5% and 16%. Hutchison
also attributed the problem to the sharp upsurge in call traffic which
increased from 38% on 5 J uly to 136% on 6 J uly and 148% on 7 J uly
1998. To address the problem, Hutchison added a total of 83 channels
to its base stations for the GSM network on 7 J uly 1998. An additional
base station was also installed at the Chek Lap Kok (CLK) telephone
exchange. These measures successfully alleviated the overloading
and no problem with Hutchisons networks was reported after 7 J uly
1998.

9.7 Hong Kong Telecom CSL Limited (HKT) experienced
similar overloading problems on AOD with its GSM network, one of the
two networks it operates at the new airport. The average blocking rate
for Cell Code NAA2 was recorded as 6.67% on AOD. According to
HKT, the initial capacity of its network within PTB was 244 voice
circuits, more than three times the service capacity at Kai Tak. To cater
for expansion, HKT also installed hardware to provide 24 additional
voice circuits for operation by year 2001. On AOD, HKT
commissioned 89% of its allocated capacity on the CAS. Around
midnight on AOD, in view of the unprecedented levels of use, HKT
reconfigured its network, utilizing the remaining reserve voice circuits to
increase network capacity. HKT stressed that the problem on AOD was
one-off event and no problems were experienced with its networks since
then.

9.8 Information has also been sought from the three other
mobile phone operators, New World Telephone Limited (NWT),
Peoples Telephone Company Limited and Mandarin Communications
Limited (trading as Sunday). In their responses, all the companies
replied that they had not encountered any overloading on AOD. To
cater for potential growth in traffic, NWT requested SmarTone on 13 J uly
1998 for system expansion.

[2] Trunk Mobile Radio (TMR) Service Not Satisfactory

154
9.9 The TMR system is an important and essential
communication means for many airport operators at the new airport.
The evidence shows that users experienced problems with the system on
AOD and for a while afterwards. These problems exacerbated the
chaotic situation at the airport during its initial period of operation.

9.10 AA has its own TMR system for use at the new airport.
There are two other operators providing TMR services for airport
operators at PTB, China Motion United Telecom Limited (CMT) and
Hutchison. CMT was the contractor for installing and maintaining the
TMR Distributed Antenna Network inside PTB. The network is used by
CMT and Hutchison to provide TMR services to their respective users.
In order to provide services to its users, each operator was required to:

(a) provide indoor radio coverage in PTB by connecting its own
TMR base station(s) to the TMR Distributed Antenna Network ;

(b) provide outdoor radio coverage on the air field; and

(c) provide a switching or linking feature in its system to
facilitate communication between users served by the indoor
antennae and users served by the outdoor antennae.

9.11 Each TMR operator is also responsible for the capacity and
maintenance of its own equipment. According to CMT, the TMR
Distributed Antenna Network was installed and commissioned before
AOD. Coverage, channel efficiency and reliability of the system were
tested and found to conform to standards specified in the contract.
Airport users and operators also tested the TMR network during the
Airport Trials and the test outcome was satisfactory.

9.12 On AOD, problems were reported with the use of both
Hutchison and CMT TMR systems, but not with the AA system. AA
submitted that airlines and RHOs using Hutchisons system had
difficulties receiving signals while working inside airline offices in PTB
and on the ramp on and shortly after AOD. The problem peaked in the
155
first few days of airport opening when a large number of people were
using TMR at the same time. No such problem was reported on the
CMT system. According to CMT, the only problem with its TMR
system on AOD was congestion caused by unfamiliarity of users with the
operation of radios. By adding more channel capacity to its repeater
network and re-tuning radios to different repeaters, CMT solved the
congestion problem within four hours on AOD. CMT also arranged for
its Customer Service Team to provide its users with further training on
radio operation. It stressed that the entire problem was resolved within
one day and no official complaint had been received since then.

9.13 Hutchison acknowledged that some of its users experienced
delay in obtaining a channel for communication on its TMR system on
AOD. The problem was due to an upsurge in the usage of TMR
communication resulting from the abnormal situation at the new airport
on AOD. According to Mr Edmund SIN Wai Man, Director of
Engineering of Hutchison, the system had not been designed to cope with
the huge volume of traffic encountered on AOD. Despite severe
overloading, the system operated in accordance with its specification and
did not break down. Complaints received by Hutchison related mainly
to shortage of channels, coverage problems, patchy signal and poor
reception in certain parts of the new airport. To overcome these
problems, Hutchison installed additional channels at the new airport and
overloading was substantially reduced by 9 J uly 1998.

9.14 There were also reported problems of coverage and patchy
signals such as in the Baggage Hall, on the apron outside SuperTerminal
(ST1) and in landside offices at Levels 5 and 6. According to Mr Sin,
AA planned to build one or two antenna farms within the airport
perimeter to house all external antennae of TMR operators. However,
the proposed antenna farms were not available on AOD and Hutchison
had to locate its main base station at Fu Tung Estate, Tung Chung, about
3 kilometres away from the perimeter of the airport. Hutchison had
explored the possibility of locating the station on the roof of the Cathay
Pacific Catering Services (CPCS) Building but AA did not accept the
proposal. As pointed out by Hong Kong Airport Services Ltd. (HAS),
156
the signals transmitted from the Tung Chung base station were slightly
weak and their strength was further dissipated by buildings around the
perimeter of the new airport. While Mr Sin stressed in his statement that
coverage from the Tung Chung base station over various outdoor
locations of the new airport including the apron was tested by Hutchison
on 7 May 1998 with satisfactory results, he attributed the problem of
patchy signal on parts of the apron outside ST1 to the location of the base
station and commented that if Hutchison had been able to locate the base
station within the perimeter of the new airport, the problem would not
have existed.

9.15 The Baggage Hall was supposed to be covered by a common antenna
system in PTB, which is an internal antenna system, can enhance coverage in large
buildings and is linked with Hutchisons Tung Chung base station. However, on
AOD, the common antenna system for Hutchisons TMR system was out of action
because of a problem with the link between the common antenna system and the Tung
Chung base station. Accordingly, Hutchisons TMR users had to rely on the Tung
Chung base station which produced a slightly weaker signal. The link between the
Tung Chung base station and the common antenna system was put into operation on
29 J uly 1998.

9.16 In the light of the problems on AOD, Hutchison installed a
new base station with seven more channels in PTB on AOD, four
channels on Day Two and three more on 14 J uly 1998 which improved
TMR coverage at the Baggage Hall. Hutchison also put in place another
temporary base station on the roof of ST1 to enhance signal transmission
and added seven repeaters and 30 antennae to improve the coverage on
Levels 5 and 6 landside offices. RHOs and other operators of the airport
community were able to use the system about a week after AOD.

[3] Public Telephones Not Working

9.17 There are about 400 public phones in PTB supplied by NWT.
International Computers Limited (ICL) is AAs contractor for the
cabling work in PTB, including that for public telephones.

157
9.18 As at AOD, NWT installed 365 public phones consisting of
291 powerphones, 31 conventional phones and 43 courtesy phones. A
total of 150 public phones were operational on AOD. Of powerphones
and conventional phones, only 118 (about 30%), and of courtesy phones,
only 32 were operational on AOD:

Type Number in operation on AOD

Powerphones 111 (about 26 of which could only
make local calls)

Conventional phones 7

Courtesy phones 32

Total 150

9.19 Of the public phones that were functioning, some had
operational problems. AA and NWT presented different versions of the
extent of the problems, which are summarised as follows:

Problem

Number
(According to AA)

Number
(According to NWT)

Coins not being accepted

8 powerphones 4 powerphones
No IDD services available

27 powerphones 26 powerphones
Long wait before second
call could be made

Approx. 5
conventional
phones

Common in
conventional phones
Poor quality of reception

Some powerphones None

Hardware problems

Some powerphones None
158
Phones drawing too much
current and tripped power
circuits
Approx. 5
powerphones
16 power phones

9.20 The main reason for the low number of public phones in use
on AOD was the incomplete and poor cabling and jumpering work in the
communications rooms. In normal circumstances, the lack of
telephones in PTB would have caused some inconvenience and waiting
time to users. However, on AOD, there was an unexpectedly high
demand for public phones due to flight delays, baggage reclaim delays,
overloading of mobile phone networks, and the influx of a large number
of sightseers to the airport. The increased demand coupled with the
unavailability of a large number of public phones, despite having been
installed, caused serious inconvenience to the public, particularly in the
light of the chaos caused by the lack of flight information on AOD.

9.21 Before AOD, AA had instructed other contractors to carry
out the cabling and jumpering work that ICL was supposed to do.
Although the bulk of the work was completed before AOD, some work
was outstanding. Due to the delay in the completion of cabling and
jumpering, there was insufficient time to test the cabling circuits and
NWTs payphone network.

9.22 Signs were put up on the payphones indicating that they did
not work. Despite this, members of the public tried to use them which
meant that the inoperative phones had to be covered completely. The
number of phones in operation increased rapidly from 150 on AOD to
329 on 20 J uly, 377 on 25 J uly and 382 on 3 August 1998. As flight
information display improved and more channels were added to the
mobile phone networks, the inconvenience caused by the unavailability of
public phones was significantly reduced by mid-J uly.

[4] Escalators Breaking Down Repeatedly

9.23 Some users were annoyed by the escalators in PTB breaking
down repeatedly. The escalators were designed and installed by
159
Constructions Industrielles De La Mediterranee SA (CNIM).
According to AA, the escalators are of public service grade which is the
same standard as those used by Mass Transit Railway Corporation
(MTRC) in numerous stations throughout Hong Kong.

9.24 Of the 61 escalators installed within PTB and the Ground
Transportation Centre (GTC), 59 were in service on AOD. There
were 20 and 19 escalator stoppages recorded respectively on that day and
the day after. The stoppages were mainly caused by the escalators
protective devices being set at a high sensitivity level.

9.25 The New Airport Projects Co-ordination Office (NAPCO)
identified the cause of the malfunction as being the tripping of overload
switches. The threshold of the overload switches were set too low for
full passenger loads on the escalators. When these were geared to a
higher load setting, the problem was resolved. While testing and
commissioning of the escalators had taken place, and most of the
escalators had been used in the various airport trials, the low overload
settings had not been picked up for a fully functioning airport load
situation.

9.26 Mr Robert J ohn Fluhr, General Manager of the Maintenance
Services Department of AA, said in his witness statement that on AOD, as
he walked around with one of his managers, a number of the escalators
and travellators were not working. The two managers discovered that
the switches had tripped, and further investigation established that this
was due to the over sensitivity of the overload safety mechanism. After
CNIM had adjusted the setting of this overload mechanism, the problem
disappeared.

9.27 Mr WONG Yiu Fai, Manager of Building and Systems
Maintenance of AA, stated in his witness statement that because the exact
loading of the facility was unknown before the actual opening of the
airport, a certain amount of commissioning was necessary for a few days
before the optimum level was found which both protected the facility
from damage and did not result in unnecessary stoppages. It was
160
therefore difficult for the installation contractor to set the sensitivity level
correctly for AOD. CNIM submitted that it was normal to adjust the
sensitivity of protection devices to a level which allowed normal
protection of the customer and that this would be done after the escalator
had been put in use. This was done on 8 J uly 1998.

9.28 Other incidents were caused by people stopping some
escalators not in emergency situations and foreign objects jamming the
steps. The large number of visitors on AOD had increased the
possibility of loose parts being dragged into the combs causing the
stoppages. The emergency stop button was pressed, sometimes by
accident and sometimes by passengers wishing to walk the other way up
or down an escalator.

9.29 On AOD, systems such as the General Building
Management System, the Building Services Integration, the Supervisory
Control and Data Acquisition, and the Mechanical Building Management
System had not been completed. These systems are important to
maintenance services for automatic control, performing remote switching
and as a monitoring and warning system. Had these systems been
available, the breakdown of an escalator would be reported automatically
and a maintenance team could be despatched immediately to fix the
problem. Due to the unavailability of these systems, maintenance staff
relied on regular patrolling to monitor the operation and condition of
plant equipment; hence they were not always in a position to take timely
corrective actions to prevent the disruption of service, but were busy
reacting to complaints and other feedback.

9.30 After the first week of airport opening, the operation of the escalators
stabilised. The few stoppages occurring later were mainly due to loose screws from
luggage and other foreign objects jamming the steps. This is considered to be part of
everyday operation.

[5] Insufficient or Ineffective Signage

9.31 There were complaints that users of PTB were inconvenienced because
161
signage did not provide sufficient information or direction to passengers to ensure
smooth and efficient flow through the various facilities. However, AA maintained
that the signs in place on AOD were adequate to enable all passengers and other users
of PTB to use the building in accordance with the design. According to AA, the
concept of a minimal approach to signage was fundamental to the design of the
directional signage. Proliferation of commercial signage could also have affected
adversely the clarity of signage. The likelihood that signage would be required to
undergo change in future also led to the adoption of a system that could economically
accommodate modifications. However, W51 Mr J ason G YUEN opined that there is
another school of thought that prefers to install signs in between necessary signs as
back-up in case the first sign is missed. Most airport operators and users prefer this
philosophy.

9.32 According to the witness statement of Mr Mark A. Siladi, the
Vice-Chairman of the Board of Airline Representatives, the issue of signage was not
dealt with adequately before airport opening. He understood that since AOD, 2,000
to 3,000 extra signs had been provided by AA. He referred specifically to the
inadequacies of signage directing users to airline offices on the landside of PTB. It
was not until after AOD that AA announced the approved signage for airline offices.
W43 Mr Douglas Edwin Oakervee explained at the 67
th
meeting of the Project
Committee that the original design of signage was driven by the more aggressive
commercial philosophy in that commercial activities, not necessarily airlines, were
key to the airport. The signs had therefore been designed to direct passengers to the
commercial areas. This resulted in some areas where the signage for commercial
and airport operations were in conflict. W40 Mr Peter LEE, Manager Product
Development of Cathay Pacific Airways Limited (Cathay Pacific), said that signage
to indicate their southern baggage inquiry desk and transfer desk was not adequate.
AA refuted allegations of inadequate signage at Cathay Pacifics southern baggage
inquiry desk and transfer desk. It stated that transfer signage was put up shortly
before AOD and temporary signs were erected before AOD for airline baggage
inquiry desks.

9.33 According to NAPCO, they had observed and made specific
comments on the signage problem to AA following the various airport
trials. In his written statement, Mr Nicholas Trevor Reynolds, Chief
Architect of AA, stated that it was difficult to assess the objectivity of
162
responses from participants of the trials on signage. This was because
temporary signs used at the trials were different from the proposed
permanent signage. AA said that they had directed changes to be made
to take account of comments at airport trials. This was confirmed by
The Mott Consortium (Mott), the contractor for the detailed design of
signage in PTB. In his witness statement, Mr Robin Doughty,
Commercial Manager for Cevasa Imagen S.A. (Cevasa Imagen), AAs
contractor for manufacturing the signs, said that AA put in substantial
orders for signs as late as May 1998 and kept changing instructions as
AOD approached. Some of the signs which AA requested to be installed
before AOD were not put up in time. The additional signs it ordered in
May 1998 from Cevasa Imagen were not essential for the proper
functioning of PTB.

9.34 An unanticipated number of passengers and visitors used the
external buses instead of Airbuses and crowded at Cheong Tat Road
which led them to Level 3 (ground level) of PTB at the new airport.
Due to the unanticipated use of PTB, passengers starting at Level 3
without luggage and visitors were diverted to the Departures Hall through
the Arrivals Hall. Unfortunately, the signage was designed to start at the
Departures Hall. AA explained that the basic design of the building
assumed a one-way flow system with all departing passengers entering
the building at the Level 7 Check-in Hall from the Level 8 Departures
kerb or the MTRC platform and carparks via the terminal access structure.
All arriving passengers were assumed to leave from Level 5 Meeters and
Greeters Hall to the trains, buses, taxis, hotel limousines and carparks.
This caused some confusion among users of the building under the
one-way flow signage system because these passengers presumably saw
signs intended for arrival passengers rather than for departure passengers.
Mott suggested that as the signs were designed with a logical process
related sequence in mind, confusion would arise if people flows were not
managed to the intended operational criteria. Apart from unanticipated
use for this reason, AA insisted that the large number of people using
PTB had done so satisfactorily.

9.35 AA however acknowledged that among the more than 1,500 directional
163
signs within PTB, a single arrow within the Meeters and Greeters Hall pointed in the
wrong direction. It alleged this was a mistake on the part of the contractor but it was
corrected within one day.

[6] Slippery and Reflective Floor

9.36 The Commissioners have heard complaints about the
slipperiness and reflective quality of the polished granite flooring used in
PTB. Criticisms center around the Zimbabwe Black granite floors
which are allegedly both slippery and very reflective, the latter causing
potential embarrassment to female users of the new airport.

9.37 The interior architectural design of PTB was carried out by
Mott in its capacity as AAs consultant. The supply and laying of the
hard flooring for PTB was completed by Grant Ameristone Limited
(Grant), nominated sub-contractor selected by AA, the
British-Chinese-J apanese J oint Venture (BCJ ) being the main
contractor. The materials and the types of surface finish were specified
and approved by AA before the sub-contract was awarded. According to
AA, it took into consideration factors like durability under heavy traffic
load, suitability for pedestrian and trolley use, ease of maintenance as
well as aesthetics in the selection of flooring materials. To cater for the
heavy pedestrian and wheeled traffic at PTB, natural granite was used and
such was consistent with its use in prestigious buildings, both throughout
Hong Kong and internationally. Also, to enable floor patterns and
borders to break down visually the large expanse of the floors in PTB, a
selection of five different types of granite were used and, except for
Zimbabwe Black (black) and Rustenberg (dark grey) which had polished
surfaces, all the other granite surfaces were honed.

9.38 According to AAs submissions, a total of five incidents of
people slipping on floors in the public areas of PTB were recorded
between AOD and 31 August 1998. From the records available, a wet
floor was identified as a contributing factor in two of these incidents.
Furthermore, none of the five reported incidents of slipping occurred on
the black granite floors.
164

9.39 The problem of slippery and reflective floors came up during
the first airport trial held on 18 J anuary 1998 when airline staff
complained of the slipperiness and the reflective surface of the polished
black granite floors. As a result of the feedback, AA carried out a series
of tests to measure the slipperiness of all the granite surfaces, both the
polished and honed ones and concluded that the polished stone was
marginally more slippery. In their witness statements, both Mr David
J ohn Corby, Senior Project Manager for PTB, and Mr Nicholas Trevor
Reynolds, Chief Architect of AA, claimed that the presence of dust which
came from continuing construction and cleaning activities was a factor
contributing to the slippery floors. Mr Timothy Graham Stelfox, Head
of Contracts of BCJ , also stated in his witness statement that certain parts
of the granite floors in the public areas had a highly polished surface
finish and, under certain conditions, could possibly be slippery for
footwear with particular characteristics. To address this problem,
remedial actions were subsequently taken to raise the slip resistance of
the polished Zimbabwe Black and Rustenberg surfaces. As a first step,
Grant honed the border areas adjacent to check-in desks where passengers
would queue in order to reduce the polish effect. Honing which
involved grinding the floor surface to make it rougher, however, proved
to be extremely time consuming. In an attempt to achieve quick results,
AA undertook research into the possibility of applying a non-slip surface
coating to the polished floors and eventually decided to carry out the
treatment to all the Rustenberg and Zimbabwe Black surfaces. The
actual process of surface treatment, which required a dust free
environment with no traffic for a period following treatment, began on 1
J uly 1998 as soon as AA had found a suitable product and the areas
concerned had become clear of temporary construction works. The
whole operation was completed after AOD and all the floor surfaces so
treated meet the standard of the American Society of Testing and
Materials for use by disabled persons.

[7] Problems with Cleanliness and Refuse Collection

9.40 Rubbish build-up in some parts of PTB immediately before
165
AOD and shortly thereafter has been cited as a problem.

9.41 Before looking at the problem, it is useful to first have an
understanding of the basic design of the refuse collection system at PTB.
The design of the refuse collection system in PTB as part of the terminal
building design was carried out by Mott. The design provides for a total
of eleven refuse collection rooms at various points around PTB. All
refuse rooms are situated adjacent to the ground level lobbies of the
goods lifts serving all levels of the building, with access to the adjacent
roads. There are also two refuse compactor rooms, one in the western
apron area to handle apron waste and the other located at the ground level
in the southern concourse. All PTB tenants and their janitorial
contractors are required, under the terms of their agreements, to package
waste in a clean and hygienic manner and place it in designated
containers in the refuse rooms. Pearl Delta WMI Limited (Pearl),
which has been awarded the contract to provide waste management
services at the new airport, is responsible for the transportation of waste
from the refuse rooms, both airside and landside, to the two compactor
stations and its subsequent removal to the North Lantau Refuse Transfer
Station for disposal. In addition, a temporary collection point was
established on the landside on AOD to deal with relocation and fitting out
waste. AA has engaged two contractors to provide general janitorial
services at PTB and their distribution of work is described below:

Contractor Scope of service

Los Airport Cleaning
Services Limited (Los)

Provision of janitorial services at certain
parts of PTB and GTC.

Reliance Airport Cleaning
Services Limited
(Reliance)
Provision of janitorial services at some
restricted airside areas of the new airport,
including the Baggage Hall, some AA offices
and some areas at the apron level for ramp staff.

9.42 On AOD and some time thereafter, there was accumulation
of construction debris and other rubbish in PTB although, according to
166
Los, the problem did not manifest itself in PTB and GTC apart from the
airline offices and the common areas on the 3/F, 5/F and 6/F of the
building. As pointed out by Mr Eric WONG Wai Lun, General Manager,
Operations Support of AA, the removal of waste, particularly industrial
waste generated by the construction and fitting out work and removal
operation, was a serious problem in PTB at that time. Tonnes of waste
materials were produced each day, all of which needed removal. To deal
with the problem, both AA and BCJ , the PTB main contractor, hired
additional labour to remove the rubbish. From the submission of AA, it
was noted that BCJ s labour force was increased to about 400 day-shift
and 150 night-shift workers during the two weeks prior to AOD and,
during the same period, an average of approximately 800 cubic meters of
rubbish was removed from PTB every day. However, despite the efforts
made, it was not possible to remove the considerable volume of industrial
waste being constantly generated by the contractors and tenants around
that time. According to Mr Wong, construction, removal and stocking
activities kept going on from the few days before and the week after AOD
and it was almost impossible to prevent the illegal dumping of refuse in
public areas. Although both AA and BCJ did mobilise extra cleaners to
remove the waste, its build up was too fast to be cleared. As to what
caused the problem, the Commissioners find from evidence available that
all the following are contributing factors:

(a) failure of PTB tenants to comply with the proper refuse disposal
procedures;

(b) deficiency in the design of the refuse collection system in PTB and
non-functioning of facilities and equipment;

(c) delay in the issue of access permits or passes to both the workers and
vehicles of cleaning contractors;

(d) insufficient co-ordination between AA and its cleaning contractors; and

(e) the presence of stranded passengers and a large number of sightseers.

167
(a) Failure of PTB tenants to comply with the proper refuse disposal
procedures

9.43 The failure of many PTB tenants to comply with the
requirements of tenant design guidelines in the disposal of rubbish is
perhaps the major cause of the problem. The tenants were late in
completing their fitting-out works and, as a result, their relocation
exercises started later than anticipated. This eventually led to large
overlap of activities that were originally planned to be consecutive and in
turn created significant volumes of construction refuse to be removed
within a short time. Worse still, these tenants or their contractors did not
observe the proper rubbish disposal procedures and simply dumped their
rubbish away from their premises instead of removing it to the designated
refuse collection points. The same story was reflected in the evidence of
W42 Mr NG Ki Sing who confirmed that the reluctance of tenants to take
up their premises until the very last minute before AOD had resulted in a
large build up of waste all around PTB.

(b) Deficiency in the design of the refuse collection system in PTB and
non-functioning of facilities and equipment

9.44 There are a number of design or equipment related factors
contributing to the build-up of refuse. These factors are summarised as
follows:

(i) The design of the refuse room was not adequate to handle the refuse
volume in some areas.

(ii) Refuse chutes between Level 5 and Level 3 are not continuous and
waste collected from Level 5 has to be containerised on Level 4 and
pushed along a walkway to the chute on Level 4 for unloading.
According to Pearl, AA has not contracted out the required
transportation service.

(iii) Several restaurants are located at the area on top of the chutes on
Level 5 where no refuse room is provided. As a result, the
168
restaurant operators cannot make use of the chutes due to its
particular design described in (ii) and simply dump their refuse in
the common areas nearby.

(iv) The two compactor stations were not ready for use on AOD and
there was no access or power supply to them.

(v) The use of electric tugs to tow waste containers to the compactor
stations on the airside was originally proposed by Pearl and
accepted. However, permits for the use of the tugs were denied
shortly before commencement of operation and, consequently, new
equipment and alternative arrangements were put in place.

(vi) The refuse rooms were not ready for use and two temporary areas
had to be made available.

(vii) For some unknown reasons, some refuse rooms that were not ready
for use were however accessible to the tenants but not Pearl.

(viii) The size of the standard litter bins was too small to cope with the
situation on AOD.

9.45 In its submission, Mott refuted the allegation at (i). It
stressed that the design of the refuse room was compliant with
appropriate standards and had been approved by both AA and the
Buildings Department. It further explained that the routing of waste
through Level 4 as described in (ii) was determined by the location of
concessionaries and the absolute desire of AA to keep the movement of
waste totally out of public sight.

(c) Delay in the issue of access permits or passes to both the labour
and vehicles of cleaning contractors

9.46 AA acknowledged that there were problems on 8 J uly 1998
in getting access permits for staff and passes for both workers and
vehicles of cleaning contractors to remove rubbish. This eventually led
169
to some build-up of rubbish. As stated in the submission of Los, the
normal processing time for a permanent permit to restricted areas is three
full working days in accordance with the Permit System Manual of the
Aviation Security Company Limited (AVSECO). Despite this
prescribed time frame, it usually took 10 days prior to AOD for an
applicant to obtain a permit and the situation was even worse from late
J une up to end of J uly 1998 when it took three weeks for the issue of
permits. In its written submission, Pearl referred to an incident where
some filled waste containers could not be removed because the licence
application for vehicular airside access had been denied due to some
problems with insurance certificate. The delay in the availability of
permits or passes for cleaning labour and vehicles unduly affected not
only the number of staff who could be deployed to work but also the
planning of cleaning work within the restricted areas. AVSECO,
however, submitted that prior to AOD, Los had been issued with 309
permits (vis--vis 660 applications) for them to fulfil their contractual
obligations in the Airport Restricted Area including toilet cleaning.
Against the background, it should be noted that there are only 33 public
toilets located airside within Airport Restricted Area, or the Departures
and Arrivals areas of PTB. The reason that only about half of the
permits were issued was due to the failure of the staff of Los in turning
up for photo-taking and collection of the permits themselves. On AOD,
the Permit Office had issued 61 permits to Reliance from a total of 63
permit applications received from it.

(d) Insufficient co-ordination between AA and its cleaning contractors

9.47 From the evidence adduced, it is noted that the co-ordination
between AA and its cleaning contractors was insufficient resulting in the
failure to provide adequate cleaning service. From the daily log kept by
AA, there was an incident in which a job order made to Los to clean up
the goods lifts at the East Hall and the West Hall as well as the lift lobby
areas was not undertaken by the contractor due to difficulties in
communication. In a separate incident, Reliance was requested by AA
on the night of 5 J uly 1998 to clean up its contract area before AOD but
was unable to complete the job simply because of lack of time.
170
Moreover, the amount of waste left over was too much for its cleaners to
cope with in one nights time. AA also suggested that its refuse removal
contractors including Los, Reliance and Pearl stuck rigidly to their
respective boundaries of work and this added to the problem.

(e) The presence of stranded passengers and a large number of
sightseers

9.48 There is also evidence to show that the presence of a large
number of airport sightseers shortly after AOD undoubtedly aggravated
the problem of rubbish build up. As evident from the submission of
Los, the malfunctioning of Flight Information Display System (FIDS)
had resulted in an increased number of stranded tourists who consumed
food and drink on the spot. This, coupled with the large number of
sightseers who used the catering facilities in PTB, created additional
pressure on the provision of janitorial services within the building after
airport opening.

9.49 The problem of rubbish build up lasted only a few days and
by 10 J uly 1998, all rubbish was substantially cleared. Most of the retail
shop tenants have become more considerate in disposing of their own
rubbish. Also, as confirmed by Los, the permit processing time by
AVSECO has returned to normal and there is now sufficient manpower
inside the restricted areas to carry out cleaning services.

[8] Automated People Mover (APM) Stoppages

9.50 Automated People Mover (APM) is automated shuttle
train without a driver which runs along the central concourse of PTB at
the basement level. It is designed to carry passengers and staff from the
East Hall to the West Hall of PTB providing them with easy access to
distant aircraft gates. APM operates in pinched loop modes routing
through four stations and with two turnbacks at the extreme ends of the
track. Each APM train can carry up to 200 passengers and each single
journey takes approximately 90 seconds.

171
9.51 The APM system was designed, built and installed by the
New Hong Kong Airport People Mover System J oint Venture under AA
contract C350. The joint venture consists of Sumitomo Corporation and
the Mitsubishi Heavy Industries, Ltd. (MHI). Except for certain
contractual arrangements which fell within the formers responsibility, all
actual execution of the contract was the responsibility of MHI. On the
operational side, MHI was the contractor under AA contract M008 to
operate and maintain APM for a period of three years. The daily
performance of the APM contractor was overseen by the Airport
Management Division (AMD) of AA.

9.52 The problems with the operation of the APM during its
initial period of airport opening concerned interference to automatic door
movement and train stoppages. In one incident, passengers were
trapped and unable to leave the train for about 50 minutes.

(a) Train stoppages

9.53 On AOD and for two months after, train stoppages were
caused by vehicle door failures, platform door failures or overshooting of
trains. The number of occurrences during this period were as follows:

Occurrence Frequency: number of times
recorded

Failure of vehicle door

34
Failure of platform door

Several
Train overshooting 2 (1 and 25 August 1998)

MHI admitted that these occurrences were caused by initial failure of
equipment, and they decreased as fine-tuning of the system progressed.
In projects of a scale similar to that of the APM system, fine-tuning after
the start of operations was required to improve operational efficiency and
to accommodate actual operating conditions which might not have been
172
exactly simulated or foreseen during the design and testing stage.

9.54 Investigation into the causes of train stoppages revealed that
passengers sometimes forced a door open in order to help other people to get on board
when the door was closing. This disrupted the closing movement of train doors
triggering an alarm from the door control circuit and causing the train to stop.
Platform doors failed because of friction of door equipment with surrounding
mechanical parts and the failure of local door control circuit. The investigation also
concluded that trains overshot their designated stops because of improper contact
between the trains power rails and the power collectors.

9.55 Although MHI stressed that vehicle door failures were
mainly caused by passengers forcing doors open, it did put into effect
some technical remedial measures. After modifications were made to
reduce the sensitivity of the door control circuit, the rate of vehicle door
failures was reduced from 0.7 times to zero per day in early September
1998. Adjustment was also made to platform doors to reduce the
friction of door equipment and the local door control unit replaced, after
which there were no more train stoppages caused by platform door
failures.

9.56 To tackle the problem of trains overshooting, MHI has
replaced all power collector shoes of trains. The Commissioners,
however, note from the submissions of MHI that the problem of stopping
was not rectified completely soon after AOD. Since August 1998, there
have been further incidents of trains overshooting and, on some occasions,
undershooting. The following counter-measures were therefore
necessary:

(i) Renewal of the dip switch counters for tyre diameter settings.

(ii) Recording of data on stopping positions, tyre diameters and dip
switch counter settings to update counter settings if required.

(iii) Adjustment of the range of dip switch counter setting to allow for
bigger tyre diameter.
173

(iv) Modification of the system software.

Implementation of the above counter-measures is expected to enhance
vehicle stopping accuracy.

(b) Trapping of passengers

9.57 On 20 J uly 1998, one passenger and four airline staff
members were trapped inside an APM train and were unable to leave the
train for about 50 minutes. Eventually, the passenger missed his flight.
According to MHIs submissions, the incident began at around 11:30 pm
that day when the group of persons boarded No. 3 train at the West Hall
departures platform. The following chronology of events may be useful
in understanding the nature of the incident and the remedial actions taken:

Time Event

11:30 pm
(20 J uly 1998)
One Cathay Pacific passenger and four airline staff
members boarded, or remained on board, No. 3 train at
the West Hall departures station.
(Normally, all passengers heading for departure gates
should alight at the platform and trains should be cleared
before heading for the West turnback. A recorded
announcement was made continuously at the station to
request passengers to alight. However, for unknown
reasons, the group did not take heed of the
announcement.)

11:36 pm No. 3 train was stopped by its Automatic Train Control
system while proceeding on the West turnback because
the preceding train (No. 2) was stalled at the West Hall
arrivals station due to a vehicle door malfunction.

174
11:53 pm The APM operator at the Airport Operations Control
Centre (AOCC) received an alarm from No. 3 train
indicating that someone had tried to force open the
vehicle door.

The APM operator at AOCC realised that some
passengers were on board No. 3 train and, through the
inter-com, advised them to be patient and wait for the
assistance of maintenance staff. The APM operator then
notified the maintenance staff to go to the West turnback
to restore No. 3 train. However, it took longer time than
expected for the maintenance staff to reach the scene
because the APM operator could not communicate
effectively with them.
At the same time, MHI was trying to restore the operation
of No. 2 train.

11:59 pm The vehicle door problem of No. 2 train was fixed by
MHI. Meanwhile, No. 3 train remained at the West
turnback.

12:17 am
(21 J uly 1998)
Despite the APM operators advice, the passengers on
board No. 3 train tried to open the vehicle door by turning
the emergency door release valve which set off the
Manual Door Open alarm.

12:20 am The APM observed through the closed circuit television
(CCTV) five persons getting out of No. 3 train onto the
emergency walkway. At the instruction of the Airport
Terminal Deputy Manager, the operator shut down the
traction power in the tunnel for the safety of these people.

12:35 am The five persons were safely escorted to the West Hall
departures station through the emergency walkway.

9.58 The five persons on board No. 3 train were not supposed to
175
be there. Measures to prevent people from boarding trains at the West
Hall departures platform were not effective in the event. According to
MHI, there should have been security staff at the station to ensure that
arriving trains were actually cleared and that no one would attempt to get
on board. Apparently, this was not the case in the incident. MHI
attributed the cause of the delay in maintenance staff arriving to restore
the train to the lack of effective TMR equipment provided by AA for
MHIs maintenance staff.

9.59 Following the series of disruption of train service, AA has
taken measures to provide station attendants at each of the four train
platforms to ensure that passengers do not attempt to interfere with the
door operation or to board when they should not. These attendants are
trained to perform evacuation procedures. Emergency procedures have
also been adjusted.

[9] Airport Express (AE) Ticketing Machine Malfunctioning

[10] AE Delays

9.60 Unlike the Kai Tak airport which was situated in the urban
area and was well served by a convenient network of public transport, the
new airport rests on an island far away from the urban centres of Hong
Kong. To enable the new airport to operate smoothly, safely and
efficiently, it is very important to put in place an efficient public transport
system to cater for the daily needs of airport users. As a key component
of that transport system, the Airport Railway, later known as AE, run by
the MTRC was designed to handle 40% of the airport passenger ground
traffic. Its efficient operation is vital to the smooth functioning of the
airport. This accounted for the Governments reluctance to open the
new airport without AE being ready.

9.61 Some problems with the service of AE occurred during the
initial period after the new airport was opened. These problems
pertained to the breakdown of ticketing machines and disruption of train
service.
176

(a) AE Ticketing Machines

9.62 The coin management system of the AE ticketing machines includes
the requirement to accept coins and to give coins as change for tickets purchased with
notes or coins. During the bulk loading tests carried out just before AOD, it came to
light that the machines had difficulty in accepting high volumes of coins and, after
repeated purchases, the machines would go out of service. The problem, however,
did not surface in the acceptance tests carried out at the manufacturers site in the
United Kingdom (UK) or in the test facility established by the responsible
contractor of MTRC at its site office in Hong Kong. As soon as the problem was
discovered, the contractor briefed the specialists in UK and initiated the necessary
investigative work. In parallel, MTRC developed a series of contingency measures
to cope with the operations on AOD. The measures included:

(i) The coin management system on all machines was disabled
so that they would accept notes only.

(ii) A stock of pre-coded AE single journey and return journey
tickets was established for purchase by passengers at the
Customer Service Centres at all AE stations.

(iii) Temporary signs were put up to advise passengers of the
temporary arrangements.

(iv) Staff were posted to assist passengers in the use of the
machines and to provide them with the correct denomination
of notes to purchase tickets.

(v) Technical support from the contractor was put in place to
ensure the acceptable operation of the machines.

9.63 On AOD when AE was first brought into service, 41 out of
the 52 machines were operational. This did not pose any problem since
the total number of machines was designed to cater for future growth and
their full capacity was not necessary for operational needs on AOD. In
177
fact, in order to maximise the effectiveness of customer support and
technical support, some machines were not used due to their location.
Although passengers did experience some degree of difficulty with the
operation of the machines, this was not entirely a result of the lack of coin
acceptance and change giving functions. There were also problems
associated with the functionality of the machines and passengers
unfamiliarity with their operation. According to MTRC, the experience
was comparable to the introduction of new ticket machines in Mass
Transit Railway urban lines stations two years ago.

9.64 The contractors investigation revealed that the problem was
related to the coin identification and validation sub-system and was
brought about by the range of parameters of the coins being much greater
in actual operation than that used during the development and testing
phase. By 8 J uly 1998, a new software was introduced to all machines
to enable them to give coins as change. A further software revision was
finalised five days later which allowed the full functionality of the
machines to be brought into use progressively throughout the system.
By 14 J uly 1998, the software problems were completely solved and all
ticket machines have been working properly since 24 J uly 1998.

(b) Disruption of train service

9.65 Before AE came into operation on 6 J uly 1998, the system
had undergone a 12-week period of integrated system testing and trial
operations. As a result of inspection of the operations towards the end
of that period, it was agreed that AE should open for passenger operations
on 6 J uly 1998 at a service interval less than the design capacity for full
operation and with the journey time longer than the scheduled time of 23
minutes. This was because of the highly complex nature of integration
of the many systems involved and the need to regulate both the Tung
Chung Line service and the AE service which operated on the same pair
of tracks for the most part of the length of the railway. This effectively
meant that the AE service would be run at 12-minute frequencies.
Through a press release dated 30 J une 1998 and a subsequent one on 4
J uly 1998, passengers were advised of the possibility of extended journey
178
times and the need to board trains at least two hours before the scheduled
flight time. MTRC believed that, with the widespread publicity given,
prospective train travellers would be made aware of the initial limitations
of services and the impact on them should be minimal. In terms of
system readiness, the Commissioners note that during the briefing by
MTRC on 16 May 1998, Airport Development Steering Committee was
assured that all systems had been substantially tested and there were no
major technical issues. MTRC, however, expected that there would be
the usual startup problems but they should not impinge on passenger
safety. In the first three days of operation, the average journey time to
and from the airport was 29 minutes and 90% of all scheduled trips were
completed. The Daily Management Report of AAs Landside
Operations Department also recorded that in a couple of incidents
involving train delays, passengers had to be transferred from one train to
another and passenger baggage was late in arriving at the airport.
According to the explanation of MTRC, the extended journey time was
mainly due to difficulties with the train supervision system and the time
had been reduced following fine-tuning of the system in the first week of
operation. There was also a problem with train door operations but this
had been rectified progressively throughout the train fleet.

9.66 There were minor train service disruptions on 9, 11, 14 and
27 J uly 1998. The most serious disruption to the AE service occurred at
9:50 am on 23 J uly 1998 when a train damaged a rail crossing on the
track towards the airport due to an error on the part of the train operator.
The accident resulted in temporary suspension of the Tung Chung Line
and a 20-minute service frequency for AE. The number of passengers
affected in the incident was estimated to be about 4,000. Contingency
measures such as provision of replacement bussing services were
immediately available to deal with the disruption of service. AE
resumed full service at 12-minute frequencies at 12:30 pm on the same
day.

9.67 In its submissions to the Commission, MTRC accepted the
initial failure of AE to meet performance specifications. They argued
that the disruption was caused by human error and the problems
179
encountered were minor in terms of either delay or inconvenience and
were in the nature of teething problems. Throughout the initial period
from AOD, AE was able to operate in a safe, effective and efficient
manner. The incident on 23 J uly 1998 was an isolated event and did not
relate to any system-wide or training problem. Both MTRC and its
contractor were quick in rectifying the problem and introducing effective
and adequate contingency measures to cope with the situation.

9.68 The problem of train delays posed greater inconvenience to
passengers to and from the airport. Train delays which were caused
mainly by signalling and communication problems could possibly be part
of the usual startup problems which can generally be expected for such a
large and complex railway system. On 27 J uly, AE was delayed for 19
minutes due to a signal error. While most of the incidents recorded are
minor in nature, the major disruption of service on 23 J uly 1998 which
resulted in damage to the rail was more serious. There might be some
truth in MTRCs claim that the accident is only an isolated incident since
it has so far not recurred. Taken together, the problem of train delays is
a minor one. The problem was largely resolved by the end of the first
week after AOD and full functioning of the ticketing machines was back
to normal progressively thereafter. MTRC has been able to reduce the
problem speedily and professionally and improve the AE service on an
incremental basis. During the months of August and September 1998,
AE was able to achieve an average service frequency of 10-minute
intervals with a 25-minute journey time. 75% of all journeys were
actually completed in less than 25 minutes. Starting from October 1998,
AE has operated in accordance with the original performance
specifications at 8-minute service intervals with a 23-minute journey time.
No major incidents of service problems have been reported since early
August 1998. The Commissioners find particular comfort in that
passenger safety does not seem at any time to have been compromised.

[11] Late Arrival of Tarmac Buses

9.69 At CLK, HAS is the sole franchisee for the provision of
airside bus service, commonly known as tarmac buses, for the
180
transportation of passengers and airside staff between PTB and remote
stands where the aircraft are located.

9.70 On AOD and Day Two, there was significant delay in the
disembarkation of arriving passengers, both at the frontal stands at PTB
and at remote stands, some delays lasting for up to 2 hours.

9.71 The delay in disembarking arrival passengers at the frontal
stands at PTB docking bays were primarily caused by problems related to
the airbridges, which are discussed under item [14] Airbridges
Malfunctioning.

9.72 The delay in disembarking arriving passengers at the remote
stands was largely due to a combination of factors. Problems with FIDS
resulted in inaccurate flight information on the location and status of
arriving aircraft being provided which resulted in service providers,
including RHOs, having to spend time searching for the aircraft on the
apron. The problems relating to the failure of FIDS are discussed in
Chapter 10. The problems associated with the TMR used by HAS
impeded information flow for the despatch of buses and drivers. The
overloading of the mobile phone network made the situation worse. The
problems relating to the mobile phones and TMR are detailed under items
[1] and [2] above. There was a greater utilisation of remote stands for
parking of aircraft due to serious flight delays, particularly for departure
flights. This put heavier demand on tarmac buses than would normally
be expected. The flight delays and a full apron on occasions created
difficulties in co-ordinating boarding gate assignment and the location of
aircraft which in turn resulted in increased travelling time for buses due to
the longer distance between some Apron Passenger Vehicle lounges in
PTB and certain remote stands. On some occasions buses were forced
to collect passengers from the south apron bus dock and drive for 25
minutes to the north apron parking bay when it would have only taken a
few minutes for north apron passengers to board through the north apron
bus dock. The use of buses as boarding lounges reduced the time
available for buses to carry out its transporting duties. At the Kai Tak
airport passengers were admitted to a holding lounge and were required
181
to wait until there were sufficient passengers to fill the bus. However,
the boarding practice at CLK was such that passengers were allowed to
board the bus immediately after check-in, resulting in buses having to
wait at the bus dock until there were sufficient passengers to fill it.
Furthermore, in some cases on AOD and Day Two, passengers arriving at
a frontal stand where the airbridge did not work were required to be
transported by tarmac buses to PTB.

9.73 AA alleged that there were insufficient bus drivers and
tarmac buses for AOD and subsequent days. This was denied by HAS
who maintained that they had sufficient resources to service
approximately three times the number of scheduled flights. Due to the
insufficient number of security cards made available by AA, sometimes
arriving passengers and airline staff could not gain admittance to PTB
which meant that drivers, who did have security cards, left their buses to
open security doors to admit passengers and airline staff to PTB.

9.74 Since AOD, there have been a number of remedial measures
taken to improve the efficiency of tarmac bus service. The measures
taken in respect of FIDS are described in Chapter 10. HAS
subsequently changed its TMR provider. A more reliable flow of flight
information has been provided by AA to HAS. AA now faxes HAS
allocation charts for remote bay flights at 2:00 am each day which allows
HAS to plan its manpower allocation at beginning of the day. This is
then updated throughout the day. HAS has recruited additional
supervisors and bus drivers and more supervisors have been assigned to
monitor passenger volumes and to patrol ramps during peaks. Once the
allocation of stands became more orderly, the tarmac bus service
significantly improved. By 13 August 1998, bussing operation was able
to meet prescribed targets in over 90% of the assignments.

[12] Aircraft Parking Confusion

9.75 Apron Control Centre (ACC) is responsible for allocating
parking stands for aircraft. Through FIDS, flight information including
parking stands allocated to aircraft was disseminated to operators of the
182
airport community, including airlines and RHOs. Due to the problem
relating to FIDS on AOD and Day Two (see Chapters 10 and 13), ACCs
ability to perform timely allocation of parking locations for departing and
arriving flights was hampered. Extended stay of aircraft due to delays in
flight departures eroded parking capacity and made the allocation task
more difficult. According to W23 Alan LAM Tai Chis evidence, delays
of aircrafts departures and arrivals built up quite quickly on AOD and by
about 1:00 pm, the apron was full. W28 Anders YUEN Hon Sings
recollection was that by about noon on AOD the apron was full, meaning
that all the parking stands were occupied. W29 Mr CHAN Kin Sing,
however, testified that there were two periods of full apron, between 12
noon and 5 pm and between 8 pm and 11 pm. Thus incoming planes
had to queue along the taxiway and would be directed to go to the first
available stand, wherever it might be. Planning of stand allocation was
therefore impossible.

9.76 Lack of flight information caused great difficulties in
communication among the operators in the airport, such as amongst ACC,
AOCC, the airlines and RHOs, etc. This put a strain on their resources.
In addition, problems such as the malfunctioning of some airbridges and
Access Control System doors, insufficient towing tractors due to the
amount of aircraft repositioning required, non-familiarity of push-back
procedures by some tractor drivers, pilots not fully familiar with the
apron, taxiways and remote stands and so forth also exacerbated the
problem.

9.77 AA on the other hand alleged that there was no evidence of
any confusion on the part of pilots or aircraft caused by the failure of
FIDS, though there was short-lived confusion amongst RHOs during the
first few days of the airport opening. It appears from RHOs evidence
that they were able to get back to normal operation by about Day Four.

9.78 On Day Two, a Task Force chaired by W48 Mr Billy LAM
Chung Lun, the Deputy Chief Executive Officer of AA, was formed to
consider immediate actions to be taken in remedying the situation. The
Task Force consisted of senior representatives from AA, the Secretary for
183
Economic Services, the Director of Civil Aviation and the Director of
NAPCO. With the adoption of manual backup procedures to FIDS and
stand allocation and improvement measures implemented in passenger,
baggage and ramp handling services, significant improvements have been
achieved.

[13] Insufficient Ramp Handling Services

9.79 The delay in providing mobile steps for passengers to
disembark from aircraft parked at remote stands was similar to that in the
provision of tarmac bus service discussed under item [11] above,
although all three RHOs, instead of HAS alone, were involved in serving
passengers of the airlines with which they had respectively entered into
contracts. Problems in disembarking passengers at frontal stands due to
malfunctioning airbridges are discussed under item [14] below.

[14] Airbridges Malfunctioning

9.80 An airbridge connects the fixed link bridge, which is part of
PTB, and the aircraft parked at the frontal stand bordering PTB. The
new airbridges at the new airport are quite different from those previously
operated at Kai Tak, and have different operational procedures. The
airbridges are operated by RHOs. The airbridges were supplied,
installed and commissioned by PT. Bukaka Teknik Utama-RAMP J oint
Venture, a nominated subcontractor of BCJ .

9.81 A number of faults were reported on AOD and the days
thereafter. On AOD, four out of 74 airbridges were out of service for
one to two and a half hours. There were 19, 30, 30, 30 and 34 fault calls
from AOD to Day Five respectively. There were a total of 576 fault
calls up to end of J uly. Many of the faults related to auto-leveller failure
alarms. There were also problems in extending or retracting the
airbridges to and from the aircraft. Other than on AOD, these did not
cause significant flight delays. To deal with the operational problems,
two airbridge teams were formed on Day Three by AA and the contractor
to restore service promptly. Service was restored quickly, usually in no
184
more than five minutes.

9.82 The auto-leveller adjusts the height of an airbridge so that it
follows an aircrafts movements during loading and unloading. The
airbridge is raised and lowered by a vertical drive control circuit. The
unusually high number of auto-leveller failure alarms (resulting in
stoppage of the airbridge) was caused by the incorrect sequence of a
timing element in the canopy deployment safety circuit which had been
mistakenly included in the vertical drive control circuit. In servicing a
wide bodied aircraft such as a B747, the canopy deployment safety circuit
timer did not have enough time to satisfy programme requirements and
the electrical power to the vertical drive automatic control was interrupted.
The auto-leveller recognised this as a runaway and for safety reason,
sounded an alarm.

9.83 The incorrect sequencing was caused by a programming
error in one out of approximately 25,000 lines of programming codes.
The software error was identified on 11 J uly 1998 and solved on the
following day. Refresher training was also provided to RHO staff.

9.84 AA required all airbridge operators to be certified and stipulated that
an operator could only operate an airbridge by himself after he had operated under
supervision for more than 50 flights. It has been suggested that some of the delays
to the disembarkation of passengers might have been caused by RHO staff not being
experienced or well trained to operate the airbridges. HAS alleged that its staff were
not given sufficient access to the airbridges at the new airport for training and had to
practice on crude simulations, such as the use of an iron bar as an aircraft.
Notwithstanding the allegation, when the RHOs representatives gave evidence before
the Commission, they all denied that their staff were not experienced or well trained.
It was also said that operating an airbridge was not rocket science but rather a very
simple process. Operator error might also have been due to the unfamiliarity of staff
with the operation of new airbridges. Irrespective of the cause, it did not cause
serious operational problems and the situation improved very quickly.

[15] No Tap Water in Toilet Rooms and Tenant Areas

185
[16] No Flushing Water in Toilets

9.85 On AOD and the few days thereafter, the following problems
with flushing and tap water were reported:

Date Period of
Interruption
Flushing
or Potable
Water

Areas Affected Tank
Room
Involved
Eve-ni
ng of
6 J uly
About 2
hours
Potable Certain toilets and
tenants in the North
Concourse and Northern
part of the East Hall in
PTB

Tank
Room 3
7 and
8 J uly
3 to 4 other
interruptions,
each lasting
less than 2
hours

Potable Same as above Tank
Room 3
6 J uly Less than 4
hours
Potable Certain toilets and
tenants in the West Hall,
North West Concourse
and South West
Concourse.

Tank
Room 8
8 J uly About one
hour
Potable Toilet Block 6-17, 6-18
and 6-29

Tank
Room 8
7 to
8 J uly
From 15:00
on 7 J uly to
07:45 on 8
J uly

Potable and
Flushing
Some toilet blocks and
some shops on the
south-east side of PTB
Tank
Room 2

186
9.86 AEH J oint Venture (AEH) is the contractor employed by
AA for the installation of the systems which provide flushing and potable
water to the toilets in the public areas, and valved connections to the
boundary of the tenant areas. AEH subcontracted the supply,
installation, testing and commissioning of related electrical and hydraulic
works to Rotary (International) Limited (Rotary).

(a) Problems relating to Tank Rooms 3 and 8

9.87 Immediately prior to AOD, the valves which regulated water
flow into the water tanks were not functioning properly. The valves
therefore had to be operated manually by Rotary on a 24 hour basis to
ensure that an adequate level of water was being maintained in these
tanks. On AOD, staff from Rotary were unable to obtain security passes
to enter Tank Rooms 3 and 8 which were in a restricted area. As no one
was operating the tank rooms, the water in the tanks ran dry. It was
alleged that had the Building Management System been operational, the
low water level alarm would have warned AMD of the problem and
immediate remedial action could have been taken to prevent the tank
from running dry.

9.88 Water supply was restored in the morning of 7 J uly 1998
when Rotarys staff were allowed access to the tank rooms. The tank
rooms were under manual operation as late as mid-September 1998 and
there has not been any further interruption to the water supply.

9.89 The other interruptions to water supply from Tank Rooms 3
and 8 were caused by repairs on that day or by the fine tuning to the
pressure settings necessitated by the significant increase in demand for
water after AOD.

(b) Problems relating to Tank Room 2

9.90 The drainage pipes in Tank Room 2 and the Fire Services
Tank Room 1 (which is adjacent to Tank Room 2) are connected to the
foul water drainage system outside PTB via a manhole. Flooding in
187
tank rooms was not an infrequent event as there had been many occasions
of foul water backing up from the external foul water drains into the floor
gully in Tank Room 2 and manhole in Fire Services Tank Room 1.

9.91 In the morning of 7 J uly 1998, flooding in Tank Room 2 was
reported. Flooding was caused by a blocked external foul drain which
resulted in foul water over-spilling into the tank room. The floodwater
had reached some 10 inches high and for safety reason, the electrical
control panel that operated the pumps was switched off, causing flushing
and potable water supply from Tank Room 2 to stop.

9.92 Rotary and AEH staff attended to the problem and temporary
pumps were brought to Tank Room 2 to pump dry the area.
Dehumidifying and special drying equipment were also set up to dry out
the electrical control panel in the tank room. Water pumping from Tank
Room 2 was resumed at about 7:45 am on 8 J uly 1998.

9.93 As a temporary measure, Rotary was instructed by AA on 11
August to install a temporary pump in Fire Services Tank Room 1
standing by and/or operating 24 hours, seven days a week to ensure that
flooding was controlled. The temporary pump diverted the foul water
backing up in the manhole to another drainage network in the Baggage
Hall on Level 2 which diverted the foul water to the northeast end of
PTB.

9.94 The flooding was caused by blockage in the pipe work for
which Nishimatsu Construction Co., Ltd. was responsible. On 18 J uly
1998, through CCTV cameras, it was discovered that a section of the
pipeline was broken. Remedial work was carried out and the pipe was
reinstated on 15 August 1998.

9.95 The Commission has not received any evidence of
recurrence of any of the problems.

[17] Urinal Flushing Problems

188
9.96 Four problems were identified in relation to urinals in the
new airport: (a) difficulties in controlling the flushing water flow; (b)
operational problems of the infrared sensors which activate the flushing
valves; (c) blockage of urinals caused by rubbish; and (d) cleanliness of
the toilets.

9.97 The urinals in PTB were installed by Rotary, a subcontractor
of AEH. Los is the main contractor for providing cleaning services in
the public toilets.

(a) Flow of flushing water

9.98 Difficulties were experienced in controlling the flow of flushing water
through the flushing valves. The desired flow rate should be sufficiently high to self
clean the valve of seawater sediment whilst at same time not causing splashing. The
poor quality of seawater and a low flow rate caused the build up of sediment in the
flushing valves of the urinals. The problem with the flow of flushing water and the
poor quality of seawater had been identified in early 1998. In addressing the
problem of seawater quality, AA issued instructions to AEH to clean the affected
water tanks and also to install stainless steel weirs at each outlet in each tank in front
of the outlet pipe to the pumps to prevent sand and dirt flowing into the pipe. Whilst
the water tanks had been cleaned accordingly, AEH did not install the weirs as
instructed.

9.99 Prior to AOD, Rotary suggested to AA to increase the water
flow to improve the self-cleaning of valves. Rotary also proposed the
installation of hoods to prevent splashing caused by the increase in water
flow. However, this was rejected by AA in March 1998, partly for
aesthetic reason and partly because AA did not consider this to be a
complete solution to the problem.

9.100 In mid-J uly 1998, AA eventually accepted Rotarys
recommendation to install an amended piston within the valves and the
installation of hoods. The new piston was installed by AEH at its own
expense in August 1998. The installation of the hoods was completed
by 11 August 1998.
189

9.101 According to AA, there are still two outstanding problems
with the flushing system. First, corrosion of the solenoids which operate
the valves, as they have been exposed to seawater. Remedial work is
being carried out by AA. Secondly, there is a residual problem of not
having a correct balance of pressure setting for the valves. Rotary,
however, denied that such problem exists. It is not clear to what extent
the operation of the urinals were affected by these outstanding problems.

9.102 AA also alleged that the problem with the flow of flushing
water was attributable to the outstanding need for testing and
commissioning of the hydraulic system, which was substantially
completed only by the end of October 1998. All rectification work by
contractors was completed by 16 October 1998 and generally substantial
improvement has been achieved since then.

(b) Problems with sensors

9.103 In the few days immediately after AOD, it was discovered that not all
sensors had been correctly set to detect a person standing at normal usage distance
from urinal. According to AA, the sensor distance was pre-set by the manufacturer.
AA alleged that the problem was caused by the incorrect measurement of sensor
distance by AEH, but this AEH denied.

9.104 Another problem was caused by users mistakenly pressing
the sensor cover plates, believing this to be a flushing button. This
either damaged the sensors or affected its setting. To avoid any
misconception, a label reading Do Not Push in both English and
Chinese was affixed to each sensor cover plate.

9.105 Replacement of damaged sensors was effected by the end of
August 1998 and they were fitted with more substantial fixtures to
prevent interference and damage.

(c) Blockage of urinals

190
9.106 Blockages in drains had been caused by users disposing of rubbish into
the urinals. The problem was noticeable when the airport first opened, particular on
sightseeing days, when large number of people visited the new airport. Rotary
alleged that the plastic waste strainers in urinals were not fixed and were sometimes
removed, allowing rubbish to get into the system, thereby causing blockages.
Regular attendance by cleaners was thus required to prevent blockages.

9.107 The clearing of blockages in urinals was performed by the
cleaning contractor, Los. This is related to the cleanliness of toilets
which is dealt with below.

(d) Cleanliness of toilets

9.108 There were complaints that toilets in PTB were dirty,
particularly during the first few days of AOD. Even shortly before AOD,
the Financial Secretary had raised the issue of cleanliness in toilets and
AA was asked to put an attendant in each toilet to ensure its cleanliness.
W44 Mr Chern Heed said in his evidence that AA had been let down by
the contractor responsible for janitorial services in toilets (i.e. Los).
W44 Heed also added that the problem lay with staff training and
supervision.

9.109 The toilets were crowded with sightseers and stranded
passengers during the first few days of AOD. The heavy usage of the
toilets made the cleaning task more difficult. According to Los,
cleaners were sometimes crowded out by the huge number of people
queuing up to use the toilets. Other factors affecting the cleaning
service included the interruption of the flushing systems, disruption to
potable and flushing water supply and the urinal blockages described
above. Los has also pointed out that there is no ventilation facility in
toilets and that has caused smell lingering during heavy usage. This
affected the users impression of the cleanliness of toilets. Los alleged
that like some other contractors, it also encountered difficulty in
obtaining permits to restricted areas which prevented its staff from
working in some 33 public toilets within the airside.

191
9.110 After AOD, Los deployed extra staff to clean the toilets. On top of
regular cleaning, special task forces were also sent to check and clean the toilets every
two hours. These measures alleviated the cleanliness problem and there was
improvement to the situation.

[18] Toilets Too Small

9.111 There have been criticisms that the size of the toilets in PTB
of the new airport is too small, causing inconvenience to airport users.
In particular, the inability of air passengers to get their baggage trolleys
into toilets has been highlighted as a problem.

9.112 In his evidence before the Commission, W43 Oakervee attributed the
criticism on the size of the toilet cubicles to the fact that the doors of the toilets went
from floor to ceiling. He said this may create a claustrophobic feeling on the part of
the users. Another reason provided by him was that the Commercial Division of AA
had been trying to expand and keep as much space available for commercial areas as
possible. This resulted in AAs decision not to go for large and elaborate toilets but,
rather, something that was of a finish that met world-class standards and of a size that
was functional and met the general specifications. Mr Barry Ball, Senior Architect
Interiors of AA, said in his witness statement that the fact that larger toilet blocks
would have reduced the availability of revenue generating space was an important
factor which influenced AAs decision not to allow trolleys into the toilets. Other
factors included their desire to avoid having a large dead area in the middle of the
block which would inevitably become clogged by unattended baggage trolleys.

9.113 In fact, the reason for the size of toilets can be found in the design
rationale. Under AA contract C101, the detailed design of PTB including the design
of the toilets was prepared by Mott. In his witness statement, Mr Winston SHU, the
Director in charge of the architectural team at Mott, explained that sizing of toilets in
PTB, as in the case of other facilities there, was based on dwell population number
and the busy hourly flow rate of passengers at 5,500 passengers per hour. This flow
rate was approved by AA as the basic design parameter. AA also approved the
adoption of the planning guidelines of the British Airports Authority (BAA).
These guidelines were formulated on the extensive experience of BAA in operating
large international airports.
192

(a) Distribution

9.114 In accordance with the final design, public toilets in the new airport are
provided conveniently along the main passenger traffic flow, having regard to the
passenger dwell time in different zones of PTB. Generally, toilet facilities for an
airport can either be centralized (ie, with a single large facility at one or two locations)
or dispersed (ie, with a number of smaller facilities strategically provided). Given
the large physical size of PTB, AA adopted the latter approach which would enable
passengers to locate toilets easily and conveniently by shortening the walking
distances between these facilities and the passenger processing points such as
check-in desks and baggage reclaim areas.

(b) Provision

9.115 Following the BAA guidelines, Mott established a baseline
of public toilet provisions in terms of the number of water closets, urinals
and hand basins. The actual provisions, which were approved by the
Buildings Department of Hong Kong Special Administrative Region
Government as adequate for PTB, exceeds the requirements stipulated in
the BAA guidelines by 17% for male urinals and 48% for female closets.
The urinal stall and hand basin separations, and the water closet cubicle
dimensions are based on established architectural standards (eg, AJ
Metric Handbook). The size of toilet varies depending on their locations
within PTB as well as the number of facilities and the space allowed for
circulation in each of them. Large blocks are provided at places such as
passenger processing areas and catering outlets where passenger flow and
dwell time are expected to be more. In departures and arrivals
concourses where the expected dwell time is predictably less, smaller
blocks are located at regular intervals. Disabled toilets and nursing
mothers rooms are located adjacent to toilet blocks at specific locations.
W43 Oakervee testified that the standards adopted by AA conformed to
the legal requirements of Hong Kong and were also the norm in public
buildings.

193
(c) Baggage Trolleys

9.116 There are three main areas in PTB where passengers have access to the
use of trolleys:-

(i) baggage check-in area at the departures level;

(ii) Baggage Reclaim Hall at the arrivals level; and

(iii) the meeters and greeters area at the arrivals level.

9.117 Mr Shu explained that in deciding whether space should be allowed for
trolleys to gain access into toilets in these areas, passenger travelling habits and their
convenience were taken into consideration. Most of the passengers who have to
bring baggage trolleys into toilets travel alone and hence have no one to entrust their
baggage with. As far as toilets at departures level after check-in are concerned, the
majority of passengers travelling on their own would check in their baggage in the
first instance when they arrive at the airport. The number of passengers who have to
carry heavy baggage on trolleys when visiting the toilets at departures level after
check-in is therefore considered very small.

9.118 The majority of passengers who use toilets at baggage reclaim would
do so while waiting for their baggage to arrive at the reclaim carousel. They do not
tend to collect their heavy baggage and place it on trolleys before going to the toilets.
If they do, they can leave their baggage unattended in this area because the reclaim
hall is usually the safest part of the Terminal from theft. It is a restricted area and all
the baggage has to go through Customs Control and is subject to possible scrutiny.

9.119 As for toilets in the arrival meeters and greeters hall, once the
passengers clear Customs, passengers travelling on their own who are not meeting
someone at the airport usually head for transport. Toilet provisions in this area cater
mainly for meeters and greeters who do not have baggage and need not use trolleys.

9.120 After considering the low number of passengers likely to take trolleys
into toilets, AA decided that trolleys would not be allowed into these toilets but
suitable circulation space around the hand basins and urinal stalls would be made
194
available so that baggage trolley could be left in the central area of the toilet even if it
was brought into it. W43 Oakervee confirmed that toilets had been designed
deliberately not to allow people to get a trolley in.

9.121 In his witness statement, Mr Shu also pointed out that an open design
was adopted for toilet entrances because it would facilitate passengers carrying bags
and that there was ample storage and shelving space inside toilets for hand carried
baggage. Space was also available at the toilet entrances for use of parking of
baggage trolleys. Furthermore, for rare occasions where heavy baggage on trolleys
must be taken into toilets, passengers could make use of the toilets for the disabled
which were located next to the regular toilet blocks and were spacious enough to
accommodate trolleys.

9.122 As a result of the comments received from the first terminal operations
trial, AA instructed the contractor to carry out a series of enhancement works to the
toilets. These included provision of additional lighting, installation of hand dryers
and increasing the depth of toilet cubicles and the width of the dry shelves. In
particular, the height of the cubicle doors, which stretched from floor to ceiling
according to the original design in order to prevent theft, was reduced. The width of
some doors was also altered. In addition, new toilets were put in the meeters and
greeters area which are quite large. All these enhancements suggest that AA had
taken on board the comments received and exercised a good degree of flexibility in
modifying and improving the design.

[19] Insufficient Water, Electricity and Staff at Restaurants

9.123 There were problems with water and electricity supply to
restaurants in the first few days after AOD, causing inconvenience to
passengers and staff. AA alleged that for some tenants, insufficient staff
had led to unsatisfactory service, long queues, lack of food variety and
some restaurants had to close early. Customers also complained of
inadequate service staff.

(a) Water supply

9.124 Disruption to the potable water supply to the tenant areas during the
195
first few days of airport operation has been dealt with under item [15] above. There
were complaints that most of the tenants left to the last minute to apply for the
connection of the water supply, resulting in its late availability. AA also alleged that
some tenants often carried out their work not to the required standards, causing delay
to the connection of water until the work had complied with the requirements.
Despite the problem, water supplies were connected in order to make PTB habitable
as early as possible. This resulted in some leakage or flooding though the problems
were not significant.

9.125 The tenants, however, raised their concern over delays by
AA to provide water, gas and power to tenant areas. There were also
complaints by tenants that they were not receiving security permits
promptly to enable contractors to carry out work in restricted areas.

(b) Electricity supply

9.126 Similar allegations were made by AA and the tenants in respect of
electricity supply. AA complained that the tenants took possession of the premises
only at the last possible moments, resulting in a flood of last minute requests for
service connections which could not all be handled in the time available. AA also
alleged that the late submission of applications and supporting materials from the
tenants had been a significant contributing factor to the problem. Most of them
submitted their design in March and April 1998.

9.127 The restaurant tenants preference to use electricity instead
of gas was out of AAs expectation. More power was also requested by
airline tenants at a late stage. This increased the overall demand for
electrical power which resulted in the overall power system to be
upgraded. This necessitated the redesign, specification and procurement
of new equipment which took time to install and commission.

9.128 There were a number of occasions of short electricity outage
on and shortly after AOD. These, according to the AA, were primarily
caused by fitting out contractors of the tenants who switched off power
without the AAs permission so that they could complete outstanding
work.
196

9.129 AA further alleged that some outages on or after AOD were
caused by tripping of circuit breakers caused by faults in electrical
installation put in by tenants. Only limited areas were affected and there
was no major disruption to terminal operations.

9.130 As at 10 August 1998, two relatively serious outages
affecting tenants were reported. The incident on 7 J uly 98 was caused
by improper overload settings between a tenants installation and AAs
installation. This caused a power failure to landside retail shops at
Level 7 for 2 hours and 40 minutes. The long period of outage was
resulted because the maintenance staff of both the contractor and AA
were refused access to the switch room by a security guard.

9.131 Three circuit breakers burnt out in a power outage to Levels
5 and 6 south offices on 17 J uly which lasted about 4 hours. AA
suspected the cause to be related to a contractor staff of Cathay Pacific
working on the CX lounge who left a fire hose reel running. Water ran
through the joint of the floor, ran along the cables and finally fell onto the
live terminal causing a short circuit across the terminals. Staff from the
contractor and AAs maintenance attended to the problem and repair work
was carried out that night.

(c) Staffing problems

9.132 There were complaints that service at restaurants was
unsatisfactory with long queues, a lack of food variety and an inability to
keep the shop open for long hours. These problems were in turn partly
caused by the water and electricity supply problems. In some cases, the
problems were due to insufficient or inexperienced staff at the restaurants.

9.133 Restaurants on airside experienced problems with their staff
not receiving security passes by AOD thereby preventing them from
attending to duties. On the landside, excessive demand for catering
facilities were generated by the influx of an unexpectedly large amount of
curiosity visitors in excess of 60,000 per day in the first week of AOD.
197

9.134 Restaurant operators were experiencing low efficiency
problem and significant staff turnover. A number of caterers were
believed by AA to have employed staff with little or no previous relevant
experience.

9.135 To address the issue, AA reminded all catering licensees to
comply with the service standards incorporated in the licence agreements.
Action has taken on 7 J uly 1998 to ensure 24-hour operation of
restaurants, where necessary, and reliability of stock. Improvement was
made to the process of issuing permits.

9.136 The staffing problem was short-lived since the number of
sightseers significantly decreased after the first week of AOD as curiosity
over the new airport waned with time. Generally, the water, electricity
and staffing problems only caused inconvenience to users for a week or
so and occasionally thereafter.

[20] Rats Found in the New Airport

9.137 Towards the end of August 1998, during the course of the
Commissions inquiry, it was reported by the media that thousands of rats
were pestering the new airport. It was alleged that parts of PTB and the
aircraft maintenance facilities were affected. Some newspapers
accentuated the seriousness of the problem by digging out the vernacular
name of CLK, being Rat Island.

9.138 In response to the Commissions inquiry letter, AA said that
rats were a problem throughout Hong Kong, especially at construction
sites. It had planned a strategic pest management programme, including
rodent control. In October 1997, AA arranged for the employment of a
full time professional pest control contractor to provide pest control for,
inter alia, the common areas of PTB and the ground transportation system.
This involved the implementation of an intensive rodent eradication
programme and the provision of regular maintenance services for rodent
control. An intensive 120-day rodent eradication programme was
198
implemented with effect from 1 May 1998. AA had also employed an
in-house pest control team to carry out rodent control work for the areas
occupied by AA and the common areas at CLK. The work area covered
the airfields, aprons, runways and small airport ancillary buildings when
the airport commenced operation in J uly 1998.

9.139 Airport tenants such as Hong Kong Aircraft Engineering
Company Limited (HAECO), Hong Kong Air Cargo Terminals Limited,
and CPCS and PTB tenants such as restaurants, retail stores, government
and airline offices were required to implement their own pest control
programmes as part of their tenancy agreements with AA. A continuous
monitoring programme was maintained by the Environmental Group
within AAs AMD. Periodic environmental audits, including vermin
control audits, are performed in tenants areas to ensure that adequate pest
control programmes have been implemented.

[21] Emergency Services Failing to Attend to a Worker Nearly Falling
into a Manhole While Working in PTB on 12 August 1998

9.140 On 12 August 1998, a worker nearly fell into a manhole in a
cable tunnel L3 near Gate 61 in PTB. He sustained minor injuries. It
took 17 minutes for ambulance service to reach the cable tunnel and
locate the injured. It was discovered after the arrival of the ambulance
that special service operational crew was required to save the injured
worker who lay below ramp level. Another call was then made to Fire
Services Communication Centre of the Fire Services Department (FSD)
through AOCC 21 minutes after the first report of the incident. AA
confirmed that it was not normal AOCC procedure to request both
ambulance and fire service assistance when a medical emergency was
reported.

[22] Traffic Accident on 28 August 1998 Involving a Fire Engine,
Resulting in Five Firemen being Injured

9.141 On 28 August 1998, a Fire Services Vehicle was travelling
along the slip road of the Airport Road towards Tung Chung. Upon
199
approaching the merging point with East Coast Road, the driver alleged
that he had to turn right to avoid collision with another vehicle.
However, he lost control and hit the kerb embankment. Upon impact,
the vehicle ran across the road surface and down a slope and eventually
came to a stop at another slip road. Five FSD personnel were injured.

[23] A Maintenance Worker of HAECO Slipped on the Stairs inside the
Cabin of a Cathay Pacific Aircraft on 3 September 1998

9.142 On 3 September 1998, a maintenance worker of HAECO fell
from a flight of staircase inside the cabin of a Cathay Pacific aircraft
while at work. He accidentally slipped on the stairs and sustained minor
injuries.

[24] A Power Cut Occurring on 8 September 1998, Trapping Passengers
in Lifts and on the APM as well as Delaying Two Flights

9.143 The press reported that on 8 September 1998, passengers and
airport staff were trapped in lifts and APM for several minutes in parts of
PTB, delaying two flights. The incident is being investigated by Rotary
and no firm conclusion of the exact cause of power failure can be drawn
by the Commission.

[25] Missed Approach by China Eastern Airlines Flight MU503 on 1
October 1998

9.144 On 1 October 1998, a China Eastern Airlines flight MU503
was instructed to carry out missed approach when it was obvious to the
Air Traffic Control that a Cathay Pacific aircraft was unable to vacate the
runway in time. MU503 was about 12 km from the airport when the
Cathay Pacific Airbus on the runway was permitted to depart. The pilot
of the airbus reported that the aircraft could not take off because of a
passenger problem in the cabin. The ATC tower controller judged that
the runway could not be vacated in time for the landing of MU503 and
the pilot of the latter was instructed to carry out missed approach.
Missed approach procedures are safe and standard manoeuvres published
200
in the Aeronautical Information Publication for the pilots to follow.

201
CHAPTER 10

MAJOR PROBLEM -- THE OPERATION OF FIDS

Section 1 : Importance of FIDS in AOR

Section 2 : FIDS Operation on AOD

Section 3 : Remedial Measures and the Present Status

Section 1 : Importance of FIDS in AOR

10.1 FIDS is the acronym for flight information display system.
FIDS receives flight information from various systems interfaced with it
and processes the same for dissemination through data-feed or display for
various users of the new airport. There are at least the following users
of the flight information provided by or through FIDS:

(a) Airport Authority (AA);
(b) Air Traffic Control (ATC) of Civil Aviation Department
(CAD);
(c) passengers;
(d) Baggage handling operator (BHO), ie, Swire Engineering
Services Ltd (SESL);
(e) Ramp handling operators (RHOs), ie, J ardine Air Terminal
Services Ltd (J ATS), Hong Kong Airport Services Ltd and
Ogden Aviation (Hong Kong) Limited;
(f) Cargo terminal operators (CTOs), ie, Hong Kong Air
Cargo Terminals Limited and Asia Airfreight Terminal
Limited; and
(g) airlines and other members of the airport community.

10.2 Flight information consists of a large number of items of

202
information that are required for operation of an airport, such as the time
of arrival and departure of flights, the check-in counter or desk number,
the departure gate number, the flight status such as gate open,
boarding, gate closed or airborne, arriving and arrived, the
arrival gate number, the baggage reclaim belt number, exit gate number,
etc. It is apparent that all these items are necessary for passengers and
the airlines of which they are customers. In addition, other items such
as baggage lateral (spur) allocation, aircraft stand or bay allocation,
aircraft type and registration number are important for the purposes of
airlines, BHO, RHOs, CTOs and other service providers operating in an
airport.

10.3 One of the most important elements of flight information is
time, ie, the time of arrival and departure of flights coming in and going
out of the new airport. There are three kinds of time: scheduled,
estimated and actual, meaning the time scheduled by airlines, the time
estimated by airlines and ATC by the use of the radar tracking processor,
and the actual time witnessed by ATC or AA or airlines. These times are
respectively called scheduled time of arrival (STA), estimated time of
arrival (ETA) and actual time of arrival (ATA) for arrival, and
scheduled time of departure (STD), estimated time of departure
(ETD) and actual time of departure (ATD) for departure, and they are
relied on by passengers, airlines, BHO, RHOs, CTOs and other users of
the new airport. The time information is particularly important to the
airlines, RHO and CTOs and other service providers for their provision of
services in their respective fields and on which their planning and
efficiency are depended. A pictorial diagram prepared by W55 Dr
Ulrich Kipper, an expert appointed by the Commission, showing these
kinds of time can be found at Appendix IX.

10.4 Apart from time, the number of the gate allocated to a
particular flight of an airline is also important as this information is
essential for the airline as well as for the passengers. The check-in
desks assigned to the airline will enable the airline to know where to send
their staff to serve the passengers who likewise need to know where to
check in.

10.5 Baggage lateral allocation by BHO is necessary for RHOs to

203
know from which lateral to pick up departing baggage after the bags have
gone through the check-in counter, the conveyor belts and the security
check, so that the RHO serving the airline can deliver it to the relevant
departing aircraft. RHO will also need to know the stand allocation of
the aircraft to which baggage will be delivered. For arriving aircraft, the
RHO and other service providers need to know the stand allocation, so
that they can deploy their vehicles and manpower to await the arrival of
the aircraft, promptly providing services such as catering, water, fueling,
cleaning and unloading the baggage and cargo from it. RHOs will also
need to know the stand allocation in order to serve the passengers. For
aircraft to be parked at frontal stands bordering Passenger Terminal
Building (PTB), they need to send operators to operate airbridges
promptly to disembark passengers. For aircraft to be parked at remote
stands in the airfield, RHOs should know the stand number well before
arrival of the aircraft for mobile passenger steps and tarmac buses to be
despatched to meet the passengers on arrival. While the baggage will be
delivered to the lateral leading up to the reclaim belt allocated for the
flight for retrieval by passengers, the cargo will be delivered to one of the
two CTOs as consigned. The registration and type of the aircraft are
sometimes also necessary to confirm identification of the aircraft and for
deployment of resources. A diagram showing the items of information
required by main airport users is at Appendix X.

10.6 ATC would also need to know the flight information and
stand allocations so to effect air traffic control. Time is absolutely
necessary for planning of air traffic, and stand allocation is essential for
giving directions to aeroplane pilots on the runway and apron.

10.7 SESL, as BHO, is a provider of information for FIDS as well
as user in baggage handling. SESL assigns reclaim laterals for arriving
bags through the FIDS workstation in the Baggage Control Room
(BCR). The allocations are displayed to RHOs on LCD boards in the
Baggage Hall on Level 2 and to passengers on monitors and LCD boards
in the Baggage Reclaim Hall (BRH) on Level 5. Reclaim laterals are
usually assigned based on a daily template prepared by SESL and
provided to RHOs the night before. However, SESL may make changes
to the pre-arranged allocation based on changes in flight times and stand
allocations provided by FIDS. The new lateral will then be displayed on

204
FIDS for the information of RHOs and passengers. For departure
baggage to be sorted by the Baggage Handling System to departure
laterals, SESL is required to enter the daily flight schedule into the Sort
Allocation Computer which may be updated in accordance with
information received from AOCC, normally via FIDS.

10.8 Other members of the airport community include hotels,
freight forwarders, handling agents and providers of transportation for
passengers, baggage and cargo, etc and they all need flight information.
AA had FIDS installed in the new airport and provided all such
information through monitors and liquid crystal display (LCD) boards
situated at various strategic locations throughout PTB. On the other
hand, airline offices, RHOs and CTOs and other members of the airport
community obtain flight information essential to their operations from a
flight information distribution system known as Flight Data Display
System (FDDS). FDDS is a service provided to the airport
community by Hong Kong Telecom CSL Limited (HKT) which entered
into a contract with AA to obtain the information through a database
installed by AA called Airport Operational Database (AODB)with
which FIDS is interfaced. There were hundreds of users of FDDS in the
airport community, and representatives of a few sectors are set out in
Appendix XI, which also contains a summary of their evidence.
Although RHOs are not included in the appendix, they have always been
FDDS customers. HKT also provides a service connected with FDDS,
which is to provide data feed to computer systems of customers. This
service is known as Flight Display Data Feed Services (FDDFS).

10.9 Timely, accurate and complete flight information and status
is therefore considered by all concerned as critical for the operation of the
new airport. In all the correspondence between the Commission and all
interested parties and throughout the oral testimony provided to the
Commission, there has been no gainsay of the importance of FIDS.
Everyone recognises FIDS as critical to the operation of the new airport.
AA entered into a contract, C381, with G.E.C. (Hong Kong) Ltd (GEC)
for the latter to provide FIDS to the new airport as early as 16 J une 1995,
and never retracted from its position that FIDS, with the necessary
functionality of providing the flight information and status, should be
ready for operation on airport opening day (AOD). The Airport

205
Development Steering Committee (ADSCOM) and New Airport
Projects Co-ordination Office (NAPCO) consistently raised concern
whenever the progress in the installation, commissioning or testing of
FIDS or the training of operators slipped. The importance placed on
FIDS by AA, ADSCOM and NAPCO is evident from the minutes of the
numerous meetings of these organisations and the voluminous reports,
monthly, weekly and even daily, relating to FIDS. Airlines and all
service providers participated in training on FIDS and joined various
airport and airline trials before AOD for purposes including
familiarisation with the use of FIDS. FIDS is indisputably critical to
airport operational readiness.

Section 2 : FIDS Operation on AOD

10.10 Many problems were encountered on AOD with FIDS.
From the allegations made by various parties, these problems arose in
many areas of the operation at the new airport. It is obvious, and no
party has ever challenged, that the flight information necessary for
passengers and other airport users was not available, inaccurate or
incomplete. Although the impact was felt all over the new airport, the
events happening at the following places are most telling:

(a) PTB areas where passengers expected to find flight
information, and on the ramp where RHOs were working.

(b) Apron Control Centre (ACC) and AOCC, both of which
were operated by staff of AA, and ATC, operated by CAD.

(c) BCR operated by the BHO, SESL.

10.11 On AOD and a couple of days thereafter, both arriving and
departing passengers found that there was no, incorrect or inconsistent
flight information displayed on the monitors and LCD boards that were
supposed to show it. In his witness statement, Mr Raymond HO Wai Fu,
the Chief Assistant Secretary for Works (Information Technology) of the
Works Bureau, described the situation. He arrived at the new airport at
1:30 pm on AOD. He observed that the large LCD boards just before

206
check-in counters in the Departures Hall and at the Arrivals Hall were not
showing updated information. He also found that other FIDS display
devices, including the Band-3 monitors at the Immigration counters and
the LCD boards at the baggage reclaim belts were not showing correct
information. Most people waited for a long time to collect their baggage.
Whiteboards were used to show passengers which baggage reclaim belt
should be used. He also noted that no baggage trolleys were there and
the courtesy telephones at the BRH were not working. When he
returned to PTB on Day Two, there were still a lot of people waiting for
baggage at BRH. The LCD boards and monitors were still showing
outdated or missing information. Whiteboards were still used to display
reclaim belt information, departure gates, etc to passengers. He was told
by a member of SESL staff that baggage reclaim status was not being
updated because ATAs for most flights were not available on FIDS on
AOD. In the afternoon of Day Three, Mr Ho observed that the LCD
boards at BRH were showing incorrect and outdated information. There
were occasions when a display suddenly disappeared and went up again,
with some monitors blacked out. Mr Ho also noticed that changes of
gates assigned to a departure flight were displayed in the remark
column. However, fewer people were waiting at BRH for their baggage.
On Day Four, during Mr Hos visit to PTB, he found that the display
monitors at the Arrivals Hall and Departures Hall were showing some
useful information, such as updates on flights arriving about 15 minutes
earlier. However, some monitors were still displaying outdated
information. Except for the passengers from one incoming flight who
were waiting for over an hour at BRH because the baggage was not
transferred from the Baggage Hall at Level 2 to the reclaim belt, Mr Ho
observed that baggage handling seemed to be working effectively.

10.12 Monitors and LCD boards were initially the only devices
supposed to display flight information for the use of passengers, but what
passengers were provided with on AOD were whiteboards showing flight
information. According to EEV Limited (EEV), there were altogether
1,952 monitors and 150 LCD boards. The monitors were screens of
three different sizes, 32, 28 and 15 and located at various places
throughout PTB. There were seven types of LCD boards supplied by
EEV which were of sizes much larger than the monitors, namely,

207

Quantity Type Nature and Location of Board

8 4302
Check in Summary Boards
82 4308 Gate Display Boards
6 4314 Meeters and Greeters Boards
48 4321 Baggage Reclaim Boards
2 4319 Baggage Reclaim Summary Boards
4 4340 External Baggage Reclaim Summary Boards
16 4350 Automated People Mover (APM) Signs
166

4302 Located in four strategic areas above the check in aisles.
4308 Located at departure gates, normally two per gate.
4314 At meeters and greeters area, located above the passenger exit
doors on Level 5.
4321 Located on each baggage reclaim belt: four boards per belt.
4319 Located at place prior to airport immigration for arriving
passengers, advising them which hall to enter in order to
reclaim their baggage.
4340 Located at each ramp entrance to the Level 2 Baggage Hall,
advising BHO where to load arriving bags.
4350 A dynamic sign located above each door for APM and not part
of the flight information system.

10.13 The so-called flight information displayed on the monitors
and LCD boards caused great anxiety and confusion to the passengers on
AOD. Whiteboards that were used instead were limited in number.
For instance, sometime after 7 am on AOD, three whiteboards were set up
at Level 5 Baggage Reclaim Hall, two at Level 6 Departures Airside, and
another two at the Meeters and Greeters Hall. Temporary signage for

208
check-in counter allocations were posted at the Departures Hall at Level 7,
and all gate assignments including gate changes were announced through
the Public Address System (PA). However, these limited number of
whiteboards and devices were no comparable substitute for the monitors
and LCD boards had they worked. W37 Mr Dominic Alexander
Chartres Purvis, the Manager of Customer Services of Cathay Pacific
Airways Limited (Cathay Pacific), had this to say:

(a) The two whiteboards at the Level 6 Departures Airside might
need to display 100 flights leaving over a two-hour period
(35 scheduled movements per hour plus about 30 flights
delayed);

(b) Flight information on the whiteboards in (a) was written up
in red ink and was messy and difficult to read. It was in
practice impossible for AA staff to keep these whiteboards
up to date in such a way that they provided timely
information on departures when passengers were crowding
around them and asking questions; and

(c) The whiteboards in BRH were also messy and difficult to
read. AA staff had similar problems updating these boards.

10.14 The customers of the FDDS service provided by HKT did
not receive correct or complete flight information either. Their
complaints can be found in Appendix XI. Inside the new airport, the
three RHOs did not receive accurate, complete and reliable flight
information from either FIDS or FDDS. For the purposes of RHOs, the
most important items of information were the ETA and stand allocations
of aircraft, which were necessary for them to send vehicles and staff to
parking stands to service aircraft, such as loading and unloading baggage
and cargo, prior to departure or arrival. For aircraft parked at remote
stands, passenger steps and buses would have to be provided for
disembarkation of passengers and bringing them to PTB. Lack of such
information compelled RHOs to use the telephone to seek it from ACC or
AOCC, but the telephone lines were too busily engaged. RHOs sent
staff to attend ACC or AOCC to get the information, or even deployed
persons to go around the apron to chase landing aeroplanes to see where

209
they parked. When the stand location was identified, there was
difficulty in relaying the information to the RHO offices because the
Trunk Mobile Radio (TMR) System that they used was congested
beyond capacity by the unexpectedly large number of calls. These
actions taken to alleviate the situation drained RHOs of their resources
and aggravated the delay in making baggage available at the reclaim belts
for arriving passengers and in loading baggage onto departing aircraft.

10.15 At about 10 am, RHOs had a meeting with AA personnel at
AOCC, and it was decided that whiteboards should be set up at Airport
Emergency Centre (AEC), to which RHOs could send their staff to
check ETA and stand allocations. However, according to W5 Mr Allan
KWONG Kwok Hung, Assistant General Manager, Operations of J ATS,
whiteboards were not set up until about 4 pm, as AA needed time to
source more whiteboards and finalise logistics and procedure.
According to W26 Mrs Vivian CHEUNG Kar Fay, Terminal Systems
Manager of AA, whiteboards were only set up at the AEC, which is
situated next to AOCC, as late as 7 pm on AOD. Most AA witnesses
agreed with this evidence.

10.16 CADs operations at ATC on AOD also experienced
difficulty. AA obtained flight STA and STD from the seasonal schedules
provided by airlines and obtained ETA and ATA from CAD which was
responsible for air traffic control, giving guidance to aircraft for their
landing and departure. ETD and ATD would be available to AA either
from airlines or from CAD. The flight information and flight status so
obtained by AA would be stored in AODB, which was linked to and
interfaced with FIDS. As can be seen from the diagram at Appendix X,
the interface between AODB and FIDS would enable the free flow of
information between the two systems. FIDS would draw the
information from AODB, process, disseminate and display the same
through itself or other systems for users, for instance, AAs ACC and
AOCC, the airport community and passengers.

10.17 Stand allocation was to be done by ACC of AA. Within
FIDS, there was a Terminal Management System (TMS) which would
deal with allocations of stands, gates and check-in and transfer desks.
For the purposes of administering air traffic control, CAD would need to

210
have information as to the parking stand to be allocated by AA to arriving
aircraft and the parking stand of departing aircraft.

10.18 It is not disputed by any of the parties that one of the ways of
AA to obtain ETA and ATA was through CAD. CADs information on
ETA and ATA is derived initially from the flight plans provided by
airlines and subsequently when an aircraft comes close enough to Hong
Kong and can be detected by CADs radar tracker, the information would
be provided by the radar tracker. There was also another way whereby
ETA and ATA could be provided by CAD to AA, namely through the
Aeronautical Information Database (AIDB) of CAD. It was agreed
between AA and CAD in meetings held several months before AOD that

(a) there was to be a link between AAs AODB and CADs
AIDB and radar tracker;

(b) CADs AIDB and radar tracker were to send ETAs to AAs
AODB;

(c) AAs AODB was to pass information on stand allocation to
CADs AIDB;

(d) as the ETA and ATA from CADs AIDB and radar tracker
would not be 100% accurate, AA should manually authorise
those data before they were fed into AODB as updates for
distribution to the other AA systems such as FIDS;

(e) if the AODB/AIDB and radar links should for any reason fail
or become unreliable, then contingency measures should be
taken, so that ETA and ATA would be supplied by CAD to
AA by telephone or fax, and stand allocation would be
supplied by AA to CAD also by telephone or fax; and

(f) CAD would also provide landing sequence displays which
also contained ETA and ATA to AA.

10.19 It should be noted that the ETA and ATA supplied by CAD
through AIDB and radar tracker would, pursuant to the agreement, go

211
through authorisation before being fed into AODB and such information
would then be available to FIDS for dissemination mainly through
display devices to users.

10.20 On AOD, the AODB/AIDB link was not used for fear that
there would be too much drain on manpower to screen the information on
over-flying flights to be transmitted. Only the radar link was connected.
Until 8:30 am, the ETAs provided by CADs radar tracker had in fact not
gone through the manual authorisation process agreed between CAD and
AA. The reason, as described by W24 Ms Rita LEE Fung King, the
FIDS Project Manager of AAs Information Technology (IT)
Department, was that the AODB and AIDB and radar tracker interface
had been tested with flight information for a couple of weeks before AOD,
and it was found that the information was reliable. She and her
colleagues in the IT Department and Operations Department therefore
decided that it was not necessary to screen the flight information from
CAD before feeding it into AODB. This decision was, however, not
communicated to CAD prior to AOD. On that day, before 8:30 am, the
flight information provided through CADs radar tracker was missing or
inaccurate. One situation was that a few ETAs of aircraft were much
earlier than their STAs. This caused TMS Gantt chart boxes to overlap
and be in conflict, with resultant green bars. This confounded ACC
operators and affected their operations on TMS. A stand just outside
PTB, known as a frontal stand, and as distinguished from a remote stand,
is linked to a gate leading into the terminal, and a problem with the
allocation of frontal stands also affects the gate allocation. As a result,
at 8:30 am, the link between AODB and the radar tracker was
disconnected. Moreover, the landing sequence display that would also
provide ETAs and ATAs did not work inside AAs ACC or AOCC,
although the same landing sequence display was available at ATC Tower
without any report of failure. The computers and cables for receiving
the landing sequence display at ACC and AOCC were supplied by AA,
whereas the hardware and dedicated data lines for ATC Tower were
provided by CAD. The contingency plan of providing the ETAs and
ATAs was thereafter put in place, CAD supplying ETAs and ATAs to
ACC at first on the telephone and later that day by fax. The stand
allocation information was similarly provided by ACC to CAD also
through telephone and then fax.

212

10.21 People who were in ACC told the Commission the events
that occurred there. When W28 Mr Anders YUEN Hon Sing, an
Assistant Airfield Duty Manager stationed at ACC, came on duty on 5
J uly, TMS was stable. The seasonal schedule was already loaded in
both TMS and Stand Allocation System (SAS). The daily flight
schedule for 6 J uly was loaded into SAS, and the schedule for 6 J uly was
also automatically rolled out on TMS. ACC staff used SAS for
optimisation in the allocation of parking stands, following a decision
made about three weeks before AOD to use SAS as the primary stand
allocation tool when TMS was at that time found not to be too stable.
At about 4:30 pm on 5 J uly 1998, an ACC operator performed
optimisation of all flights from the scratch area to the allocation area in
TMS without problem. This was for the purpose of preparing TMS for
input and confirmation of stands allocated by the SAS optimisation
function. It was intended that after the stands were set out on the Gantt
chart on the TMS screen through the use of its own optimisation process,
the operators would then adjust the allocations on TMS to make them
consistent with SAS, and then confirm them on TMS. The aim was to
use TMS, an integral component of FIDS, to disseminate and display the
stand allocations (and other flight information) through FIDS, because
SAS is a stand-alone system which does not link with FIDS and cannot
be used for dissemination of flight information. At around 9 pm, both
TMS and SAS were stable.

10.22 At about 9:15 pm, the first of the relocation flights from Kai
Tak arrived. There were 29 such flights which were flying in from Kai
Tak to the new airport as Kai Tak was to close for operation the next day.
The last relocation flight arrived at 1:29 am. For all these flights, ACC
staff entered chocks-on time and the registration number in the FIDS Man
Machine Interface (MMI) which is a workstation for manually
operating FIDS. FIDS then displayed a prompt linking the registration
number to the relevant departure flight. The ACC operator invoked the
function, thinking that this would avoid the need to manually enter the
registration number for the subsequent departure, but not realising that
this would later inhibit flight linking operations by manual linking
procedures in TMS.

213
10.23 The daily flight movement sheets received at about 1 am in
the morning of 6 J uly 1998 from Cathay Pacific and Hong Kong Dragon
Airlines Limited contained a number of changes from the airlines
schedules. Mr C K CHAN, Senior Airfield Supervisor of AA, attempted
to perform the necessary flight swaps in TMS. A flight swap is
necessary where there is a change of aircraft for an arrival or departure
flight from that originally planned. The same aircraft that arrives
bearing an arrival flight number will normally be used for a departure
flight with a different flight number, although the aircraft is the same.
When an aircraft other than the arriving aircraft is used for the departure
flight as originally planned or when a different aircraft is used for a flight
other than that previously designated, a flight swap is required. At about
2:10 am, Mr C K Chan reported to W29 Mr CHAN Kin Sing, another
Assistant Airfield Duty Manager at ACC, that he was unable to link the
relocated Cathay Pacific aircraft, that TMS was operating slowly, and that
it was taking a number of minutes for the system to respond to the link
select command. W29 Chan told Mr C K Chan to prepare the stand
allocations by using SAS.

10.24 W28 Yuen then attempted to do the swapping in TMS
himself manually but was unable to complete any of the swaps. Shortly
afterwards, Mr C K Chan told W28 Yuen that SAS had crashed in the
course of flight swapping. W28 Yuen explained to the Commission how
SAS crashed. When an attempt was made to do a flight swap, the Gantt
chart on the monitor remained unaltered with the input. When further
attempts were made, the SAS screen turned blank and the Gantt chart
disappeared.

10.25 At about 2:30 am, W28 Yuen telephoned City University
(City U), the contractor employed by AA to supply SAS, and reported
the flight swapping problem of SAS. City U agreed to attempt the
swaps themselves from their offices in Kowloon. In accordance with
City Us suggestion, W28 Yuen asked Mr C K Chan to send City U the
flight swapping details by fax.

10.26 At about 3 am, in view of the problems with the two systems,
ACC staff started to prepare a manual Gantt chart and manual allocation
board in the event that these problems could not be resolved. Whilst

214
others began this process, W28 Yuen again tried, without success, to
overcome the flight swapping problem in TMS. By about 4 am, the
manual stand allocation procedures had been put in place and allocations
made for flight movements up till about 10 am.

10.27 At about 5:30 am, W28 Yuen noticed that one or two flights
were displaying green bars next to their entry on the Gantt chart in TMS.
The number of green bars steadily increased, covering, in some cases, the
adjacent entry on the Gantt chart, and eventually they appeared in relation
to about 30 out of 64 of the boxes for flights on the Gantt chart. The
green bar problem rendered the affected flight numbers on the chart
illegible to operators, giving them the impression that there was
something wrong with TMS. The green bars, according to W28 Yuen,
were caused by ETA being earlier than STA by more than 15 minutes.

10.28 W28 Yuen also assisted other ACC operators with the entry
of chocks-on and chocks-off time (plus registration number where
applicable) on a FIDS MMI workstation. The system response time was
very slow, sometimes taking 10 minutes or more to respond. In addition,
in the course of trying to scroll the display, the FIDS MMI frequently
produced a number of query boxes, asking the operator to click to
confirm the command. The processing to clear each box often took
several minutes and accordingly delayed the entry of data. During the
day, W28 Yuen also assisted, on a number of occasions, in the re-booting
of a FIDS MMI workstation, which was necessary when the FIDS MMI
hanged.

10.29 W29 Chan told the Commission that he called the IT
Department of AA when the flight swapping was first encountered at
around 2 am. He spoke to a female person who said that she would go
to check the server, but nothing was heard from her afterwards. On the
other hand, W28 Yuen testified that he knew that W29 Chan had
contacted the IT Department, but when there was no response, W28 Yuen
called W24 Lee at about 3 am and asked for help. He could not recall
the details of the conversation. After that telephone conversation, he
was not able to contact W24 Lee again over the telephone, and she only
arrived at ACC at about 6:30 am. She then dealt with the flight
swapping and was able to do some of them, but she also experienced slow

215
response from TMS. W34 Mr Peter Lindsay Derrick of The Preston
Group Pty Ltd (Preston) arrived at ACC at about 12:30 pm and assisted
in the flight swapping. However, the slow response remained
throughout the rest of the day.

10.30 During the course of the day, starting from early morning,
everyone inside ACC was kept very busy, not only by trying to enter data
on TMS FIDS MMI workstations, but also by having to answer numerous
telephone calls from airlines and airport operators requesting stand
allocation information. The information was supplied by referring to the
manual Gantt chart and the manual allocation board. Information had to
be relayed in this manner because flight information, in particular stand
allocations, was not being disseminated and displayed on FIDS accurately
and completely. This caused delay in the departure of flights and
affected stand allocation to arriving flights, with snowball effect. Full
apron was experienced, according to W28 Yuen and W29 Chan, between
12 midday and 5 pm and between 8 and 11 pm. During those periods,
arriving aircraft had either to land later or to wait on the apron for a next
available stand to be assigned to it.

10.31 During the night of AOD, W28 Yuen continued to assist with
the input of data (chocks-on, chocks-off time and registration numbers).
Much of the data was now outdated but needed to be entered into the
FIDS MMI to bring the system up to date. The system response time
was still slow and by the early morning of 7 J uly the ACC staff had still
not caught up with real-time operation. They continued with manual
stand allocation. It was at about 2 pm on 7 J uly that W24 Lee
demonstrated to W28 Yuen that it was necessary to clear the registration
number and chocks-on time in order to perform flight swapping.

10.32 The backlog of information having been cleared, ACC staff
were able to use TMS for optimisation of stands on 8 J uly, with the
continued assistance of AAs IT Department.

10.33 The situation in AOCC was no better. The events on AOD
were recorded in a contemporaneous FIDS log kept in AOCC, which was
produced as attachment 16 to the statement of W26 Cheung, AAs
Terminal Systems Manager stationed at AOCC on AOD. The following

216
major events were recorded in that log:

(a) 06:00 hrs AIDB/AODB interface was down. The landing
sequence monitor was unstable. FIDS/AOCC was unable
to receive ETA, ATA, ATD and other flight movement
updates. W26 Cheung clarified in her testimony that the
down time of AIDB/AODB should be at about 8:30 am.

(b) 06:30 hrs The confirmed gate/stand allocation entered by
ACC/AOCC operators into FIDS/TMS was automatically
moved/removed by the system to a wrong gate/stand.
Some confirmed gate allocation information in TMS was not
passing through to the FIDS/MMI and displays. Some
confirmed desk allocations in FIDS/TMS were lost in the
system. FIDS/TMS Gantt chart started to automatically
shut down frequently. Restart of the Gantt chart was
required but took about 30 minutes. This caused great
delay in updating the system with information.

(c) 07:00 hrs Duty staff reported that inconsistent flight
information was being displayed on the monitors at different
locations. Some of FIDS displays were not reflecting what
had been updated in the system by the operator. About
80% of displays of boarding gates were either unable to
display correct flight information or to display anything.

(d) 08:00 hrs FIDS workstations performed very slowly, not able
to handle all the input by ACC, AOCC and BCR. This
especially caused great problems in updating TMS with
stand/gate allocations and confirming the allocations.

(e) 10:00 hrs FIDS workstations at BCR hanged and BHO was
not able to input reclaim belt assignment. FIDS had not
been able to show any information at the Baggage Hall on
Level 2 allocation of reclaim belts.

(f) 10:30 hrs AOCC was notified that monitors and LCD boards
were not showing the updates entered into the FIDS

217
workstations. Electronic Data Systems Limited (EDS)
checked the host server which was down and needed to
reboot the system. Displays were refreshed at around 11
am.

(g) 11:00 hrs AOCC FIDS workstations performed even slower.
It took 20 to 25 minutes to allocate a reclaim belt.

(h) 12:00 to 15:10 hrs Failure to update information at various
gates and failure to login at check-in counters and gates were
experienced.

(i) 20:00 hrs FIDS workstations continued to perform slowly.
It took 20 to 25 minutes to allocate reclaim belt. W26
Cheung told the Commission that the reclaim belt function
was especially slow, while the other functions were also slow,
but not to that extent.

10.34 Turning to the events in the BCR, Mr Guy Gerard
Summergood of EDS stated that on AOD, at approximately 8:15 am, an
operator in BCR was progressing the baggage status, cross referencing
with a paper for allocation of flights to reclaims, and marking off flights
as they were allocated and cleared from the MMI. There had not been
many pre-allocations made prior to Mr Summergoods arrival in the BCR,
meaning that additional work had to be done by the already busy operator.
The operator often changed the status of flights from FIRST BAGS (the
time when the first bag is put onto the baggage reclaim carousel) through
to DONE (the time when all the bags have been reclaimed) without a
pause for LAST BAGS (the time when the last bag is put on the baggage
reclaim carousel), with an interval of approximately 30 minutes in
between. If an ATA was supplied but LAST BAGS was set too early, the
Timed Updates (TU) would cause the flight to be cleared from the
display. Mr Summergood noticed that TU was set too short and advised
AA that TU be increased to avoid flights being cleared from display too
early. During the course of the day, Mr Summergood also noticed that
the status progression to INTERNAL (the time when the baggage is
inside PTB external baggage hall) was sometimes delayed and, on one
occasion, the status was progressed to DONE too quickly. The net result

218
of these failures was that on AOD passengers had to wait for a substantial
period before any information regarding the location of their baggage was
displayed.

10.35 Mr Summergood thought that the problems were due to lack
of familiarity with applying the process in a new working environment
and the pressures involved with backlogged flights. While this view
will be examined in Chapter 13, it should be noted that he also stated that
on AOD, baggage reclaim displays were sometimes blank or displaying
outdated information. He further said that the speed of the system was
initially slow and degraded to taking 8-12 minutes to perform any
operation on the MMI, although data was still being processed in the
meantime and the system was not idle. He contacted EDS staff in the
FIDS room at the AOCC and they were aware of the speed problems.
Following reboot of the system at approximately 10:45 am, the
performance improved significantly. However, the AA staff who gave
evidence before the Commission did not agree that there was any
significant improvement in performance after 10:45 am.

10.36 Mr Rupert J ohn Edward Wainwright of EDS was also in the
new airport on AOD. He provided a witness statement to the
Commission. His role was to deal with database related problems
during installation of builds and to diagnose and solve performance
problems. One of the issues that he monitored was database locking.

10.37 Mr Wainwright arrived at the new airport at 10:30 pm on 5
J uly. According to him, from his arrival until 6 am on 6 J uly the
database and FIDS system showed no signs of performance problems.
From then onwards, however, he noticed the Central Processing Unit
(CPU) usage increasing although the system was still functioning.
From about 8:00 am onwards the Oracle database processes were not able
to get all the CPU usage that they required owing to conflicting demands
from other FIDS and TMS processes, which would have caused variable
response times in the system. Soon after 8 am he began to get reports
from the AOCC staff of MMI users encountering the Oracle error
associated with Oracle shared memory allocation being too small, called
ORA-04031. These errors were infrequent, but from 8 am onwards the
disruption caused by the ORA-04031 problem increased slowly to peak

219
just before 10 am. He wanted to deal with the problem by changing the
Oracle shared memory allocation. With the agreement of W21 Mr
Michael Todd Korkwoski of EDS and W25 Mr TSUI King Cheong (the
Project Manager of AA), Mr Wainwright shut down FIDS in order to
effect the change. He was then advised by Mr Michael Hobden, the
hardware manager for EDS, that a Unix operating system parameter
would only allow an insignificant increase in Oracles memory allocation.
He therefore brought FIDS up again without making the proposed change
to increase the shared memory allocation, at about 10:45 am.
Performance temporarily improved at first and then began to slow down
as more MMI and Common User Terminal Equipment (CUTE) users
logged back in. Between 11 am and 1:30 pm, Mr Wainwright also
noticed a small number of standard, transient Oracle locks that occurred
and cleared. These locks were related to either CUTE or MMI
workstations but resolved themselves before they caused sufficient delay
necessitating termination of the workstation sessions.

10.38 When Mr Wainwright returned to PTB at midnight on 6 J uly,
he was informed by a colleague, Mr Stefan Paul Bennett, that there had
been more locks which Mr Bennett tracked down to MMI software errors.
EDS then implemented the Unix and Oracle configuration changes.
These changes affected only the host server and were not related to
workstations. According to Mr Wainwright, the ORA-04031 error did
not occur again. FIDS, however, slowed down again as the airport
became busier in the morning of 7 J uly but the curing of ORA-04031
made the system stable throughout the day. Mr Wainwright remained at
PTB until 11 am on 7 J uly but returned to PTB at 10:40 pm that evening.
Thereafter until 11 J uly, he was regularly at PTB doing more work. On
10 J uly, there was a most significant locking problem between 3 and 4 pm,
and changes were made to the system under System Change 109 in the
early hours of 11 J uly 1998, which led to a significant breakthrough in
resolving performance issues.


10.39 In his witness statement, W21 Korkowski of EDS wrote that
there were two main problems relating to C381 FIDS host server

220
operations in the first week after AOD: performance and database issues.
To address the performance issues, EDS undertook actions to increase the
amount of server resources installed that were allocated to the database
and to eliminate unnecessary queries generated internally by the system.
The performance issues were greatly reduced the night after AOD and
eliminated by 11 J uly 1998. The database issues exhibited themselves
intermittently due to several internal problems with the Oracle database
that the FIDS system utilised causing database locking. The locking
conditions can be eliminated temporarily by restarting the database.
This problem was subsequently solved through changes to FIDS and
TMS to bypass the Oracle features that were causing the problems.

10.40 There were four workstations in the ACC, two for TMS, one
used as FIDS MMI and one a standby. According to W28 Yuen, on Day
Three (8 J uly 1998) and Day Four, additional Random Access Memory
(RAM) was put into the workstations in ACC. The response time was
improved but not significantly. However, TMS became more stable.
The slow response was only improved significantly at the end of the first
week after AOD. The green bar problem has not arisen since Day Three.
On AOD, however, the appearance of green bars was not the major
problem, because it annoyed operators but did not prevent them from
carrying out their tasks. After W34 Derrick assisted in removing the
green bars at around 2 pm, the real problem hindering input of data into
TMS was the slow system response. TMS was too slow for the ACC
operators to catch up with updating stand allocation on a real time basis.

10.41 W26 Cheung told the Commission her understanding was
that on the night of AOD, EDS reconfigured FIDS and reset some
parameters, to make the system go faster. On about Day Three, EDS
also reduced the input refreshing rate of the screen. On about Day Five,
EDS worked on the servers, and before that installed some RAM to the
workstations. W27 Ms Yvonne MA Yee Fong, a Project Manager of
AAs IT Department, gave evidence with W26 Cheung in a group before
the Commission. W27 Ma clarified that the refreshing rate was reduced
from 6 seconds to 45 seconds.

10.42 W27 Ma also told the Commission that since February 1998,
AA had negotiated with Oracle Systems Hong Kong Ltd (Oracle

221
Systems) to be AAs consultant to look into the health of FIDS and
AODB, both of which used Oracle for their database. The Oracle
consultancy was only agreed in late J une 1998 and the consultants came
into the new airport to start working on 29 J une 1998. They then
identified some problems with the Oracle database, some of which were
rectified before AOD, but most of them were only fixed after AOD.
W27 Ma told the Commission that the problems still not rectified by
AOD caused some of the slow response and locking problems.
However, as at 5 November 1998, all but one of the fixes had been done.

10.43 The witness statement of Ms Susan WONG of AA described
the problems experienced with the Oracle database in greater detail. On
the morning of AOD, certain Oracle errors were reported several times on
FIDS workstations, and by about 10 am, certain MMI functions failed to
proceed. The errors were generated by the inability of the Oracle
database to find sufficient memory in the Database Shared Pool. The
database was restarted at 10:39 am to clear up the Shared Pool. This
specific problem was not encountered further that day. As a permanent
fix EDS, on the recommendation of the Oracle consultants, reset the size
of the Shared Pool and the related system parameters to a higher value
and restarted the server and database to take account of the changed
parameters on the night between 6 and 7 J uly 1998. No similar error has
recurred since the change.

10.44 Mr Wainwright of EDS stated that one of the issues that he
monitored was database locking. He briefly explained the three
concepts central to the understanding of this issue, namely,

Alerts: A mechanism to let one computer programme wake
another up by signalling to it. Alerts make use of
locking and therefore a programme which signals an
alert and then gets stuck could cause the whole system
to lock up.

Locking: A standard database feature to preserve the integrity of
records by which a process updating a record locks that
record from updates by any other process until the
update is complete.

222

Deadlock: A deadlock occurs when two different programmes try
to lock the same resource in a different order and end up
both waiting for each other indefinitely. Oracle will
always cancel the last lock request of one programme to
free the other programme. All Oracle locks acquired
before the last lock request will remain until something
is done to terminate the programme.

10.45 Ms Susan Wong also dealt with the locking problem. She
said that locking is a normal feature of almost all database systems which
allow multi-user access. It prevents other users gaining access to the
particular database object whilst it is being accessed. When one user
has finished processing a particular database object, it should be released
to allow other users to work on it. When two or more processes allocate
database objects required by each other, a deadlock occurs. The Oracle
database has the capability to detect and resolve such deadlocks
automatically. However, as this automatic process takes several minutes
to operate, users requiring the deadlocked resource will have to wait
which in turn will result in longer response time. There were the
following significant deadlocking problems occurring on AOD:

(a) Table Storage Parameters. Certain deadlocks occurred as a
result of insufficient settings on various table storage
parameters from AOD. A table in the database stores a
particular type of data record. If, as occurred in this
situation, the table parameters were set too low, users would
have to wait before being allowed access to the relevant
table. On the recommendation of Oracle Systems, EDS
carried out the necessary remedial actions on 17 J uly and 18
August 1998.

(b) WDUM and MMI processes. WDUM is a background
process which defines the flight information sent to the FIDS
MMIs to update the displays. A problem with this process
occurred on AOD which seriously affected performance.
Suggestions on application changes to prevent WDUM and
MMI deadlocks were sent to EDS and were implemented on

223
10 J uly 1998. Similar deadlocking has not recurred since.

(c) SESSION_ALERTS Table. When an MMI is opened
(known as a session creation), the SESSION_ALERTS table
will be searched for obsolete records for deletion. Database
errors occurred when the system tried to delete records from
the table which prevented the operation of the session
creation process. This was one of the causes that W27 Ma
told the Commission, but the cause was identified only on 10
J uly 1998. Records within the table were truncated as a
temporary measure, and MMI sessions resumed normal
functioning. Permanent fixes were implemented by EDS
towards the end of August 1998.

10.46 Ms Susan Wongs statement tallies substantially with what
W27 Ma told the Commission. W27 Ma said that there was a shared
pool problem relating to configuration of the Oracle database. That
problem was fixed in the night between 6 and 7 J uly by EDS enlarging
the shared pool and some parameters on the operating system level. In
the night between 10 and 11 J uly, EDS caused the Oracle consultant
employed by them to do three further things to improve the stability and
speed of FIDS, as follows:

(a) Disabling the deletion of the SESSION_ALERTS table;

(b) Truncating the SESSION_ALERTS table, so that it would
clean up the table every night, and before a permanent fix
was implemented, AAs IT staff truncated the table every
night; and

(c) Modifying the triggers in the TMS table to reduce the
locking of the database of TMS/FIDS.

10.47 Ms Susan Wong also stated that one problem with the Oracle
database was identified on 10 J uly which was not shown to have caused
the performance problem on AOD, and that was eventually fixed by EDS
on 23 August 1998.

224
10.48 Mr Ian CHENG, employed by Oracle Systems, was the
supervisor responsible for monitoring the Database Administration
Consulting Services delivered by Oracle Support Services. He arrived
at the new airport at 7:30 am on AOD and was taken to AOCC by Ms
Susan Wong at around 10:45 am after he had obtained his permit. He
was told that the FIDS database was restarted with an increase on shared
pool sizing at 10:39 am to resolve a shared pool problem since the shared
pool allocation had been smaller than needed. He discussed shared pool
sizing with EDS and recommended several Oracle parameters and
features in order to avoid the recurrence of the shared pool problem.
EDS informed him that they would raise certain Oracle and Unix
parameters, restart the database to take into account the effect of the
changes to be made on the night of AOD in order to implement a
long-term solution.

10.49 Mr Cheng also stated the following:

(a) An Oracle error was encountered on 7 J uly 1998. The error
occurred while a FIDS table was updated by a housekeeping
job. The problematic Structured Query Language
statement was examined and identified to be a problem on
consistent read on Oracle dynamic performance view.
The explanation and workaround were delivered to EDS and
AA on the same day.

(b) From 5 to 10 J uly 1998, related deadlock trace files were
discovered. From 6 to 9 J uly 1998, with the help of EDS,
he studied the design and source code of TMS and MMI, and
the trace files were collected. He concluded that the
application had a potential deadlock, and at peak hours was
causing three to four deadlocks per hour. Two
workarounds were suggested by EDS to which Mr Cheng
agreed. The first workaround was to prolong the frequency
of MMI refresh rate from 6 seconds to 45 seconds, which
would reduce the chance of encountering deadlocks between
WDUM and MMI. He believed that this workaround was
implemented around 8 to 9 J uly. The second workaround
was to separate the transaction of MMI into two so that the

225
FIDS_ALERT_INFO and SESSION_ALERTS tables would
not be locked or requested within one transaction to
eliminate the factor of deadlock situation. This workaround
was implemented on the night of 10 J uly 1998.

(c) An Oracle table has a transaction slot on each data block
which allows the concurrent locking on the records within
the data block by several processes. If the related
parameter is set too low and there is no room for this
transaction slot to grow, the subsequent transactions which
require locks on this particular data block will have to wait
until the occupied transaction slots are released by other
processes. Some of the trace files indicated several tables
encountered this problem, causing up to one deadlock per
day, and recommendation was delivered on 13 J uly 1998 to
EDS to solve this issue.

(d) From 10 J uly 1998, deadlocks on User Locks were found
occasionally, perhaps one or two per day. This might have
been due to a misbehaviour on releasing User Locks while
using a particular package called DBMS_ALERT. The
explanation and workaround were delivered on 13 J uly.
EDS informed him that a new module would be
implemented to avoid using DBMS_ALERT.

(e) Several Oracle internal errors were reported on 10 J uly while
the SESSION_ALERTS table was being deleted. A known
Oracle bug was identified and a patch was delivered on 16
J uly. After the application of the patch on 9 August, the
symptom of the bug did not recur, but the errors were still
reported. A new Oracle bug was discovered and another
patch was delivered on 12 August 1998.

10.50 Mr Wainwright stated that during the night between 6 and 7
J uly 1998, EDS implemented configuration changes to Oracle and the
Unix parameters and System Change 109 effected in the early hours of 11
J uly 1998 made a significant breakthrough in resolving performance
issues. In its response dated 9 November 1998 to the Commission, EDS

226
stated that the problem relating to the shared memory pool allocation was
fixed in the early hours of 7 J uly, that the problem relating to WDUM
Central Processing Unit, ie System Change 109, was fixed in the early
hours of 11 J uly, and the memory of the workstations was increased
between 11 and 21 J uly. It mentioned that there were three factors
causing database deadlocking. Use of Oracle Alerts package and
corruption of the FIDS SESSION_ ALERTS table were significant causes
of locking not processing correctly, while deadlock on Exit of MMI was a
minimal cause of deadlocking. In the early hours of 11 J uly, Preston
made changes to TMS to eliminate the use of the Oracle Alerts package
and a new release of TMS was tested and installed on 23 August 1998.
EDS installed Oracle patches and made changes to the FIDS design to
eliminate the reliance on the FIDS SESSION_ ALERTS table, and the
work was performed in August ending with the design change being
installed on 28 August. The deadlock on Exit of MMI was fixed on 11
J uly, under System Change 109. Once System Change 109 was carried
out, performance of FIDS MMIs improved dramatically and has ceased to
be an issue.

10.51 From AOD to early September 1998, many fixes or changes
to FIDS were also implemented by EDS. The causes for the problems
mentioned by W27 Ma and Ms Susan Wong were all rectified.
Truncating of the SESSION_ALERTS table started on or shortly after 11
J uly. In the early morning of 11 J uly, an additional statement was
inserted into TMS triggers to reduce the incidence of database locks
caused by the users of the Oracle ALERTS package. Other changes
were implemented for better operational usage, providing workarounds
and enhancing diagnostic facilities, all for improving the stability and
performance of the system.

10.52 W22 Mr Edward George Hobhouse described FIDS as
workable on AOD and that by Day Three TMS was almost there,
running as a planning and allocation tool, and that by Day Six, operation
was comparable to the situation of the Kai Tak Airport. W21 Korkowski
described FIDS as operating efficiently after the first week. W28 Yuen
described FIDS functions and speed as acceptable about a week after
AOD. While W26 Cheung agreed generally with W28 Yuen, she told
the Commission that there was a downtime of FIDS on 19 September

227
1998 for almost 2 hours. She thought that FIDS, as a complicated and
new system, ought to have a period of months for tuning.

10.53 Although such a complicated and sophisticated software
system as FIDS must be treated conservatively, it appears from the
evidence of all the parties concerned, namely, AA, GEC, EDS and Oracle
Systems that as from late September 1998 most, if not all, of the
problems that might affect the smooth functioning of FIDS had been
resolved. W26 Cheung said that as FIDS was healthy albeit still having
some problems, her confidence in FIDS was much higher when she gave
evidence before the Commission than shortly before AOD. The
Commissioners see little reason to doubt what W26 Cheung told them in
this regard, as she appeared to them to be a very conservative person
relating to IT matters. The evidence of all tends to suggest that there
was nothing fundamentally wrong with FIDS, and that it has worked
efficiently and smoothly since late September 1998.

228
CHAPTER 11

MAJOR PROBLEM CARGO HANDLING

Section 1 : Importance of Cargo Handling in AOR

Section 2 : Cargo Handling on AOD


Section 1 : Importance of Cargo Handling in AOR

11.1 Right from the commencement of the new airport project,
Airport Authority (AA) recognised that efficient movement of cargo
was an important aspect in the readiness of the new airport for operation.
Section 6(2) of the Airport Authority Ordinance expressly provides that
AA shall have regard to the safe and efficient movement of air cargo.
The airport at Kai Tak was ranked amongst the busiest international cargo
airports in the world and its efficient and speedy handling of air cargo had
played a vital role in maintaining the vibrant economic growth of Hong
Kong. The new airport was expected to achieve no less. Hong Kong
Air Cargo Terminals Limited (HACTL) was the only franchisee
allowed to operate as a cargo terminal operator (CTO) at Kai Tak, and
through more than two decades of operation, it had established a
reputation as a standard setter for efficiency and productivity amongst the
airport communities worldwide. One can realise the significance of air
cargo transportation to the new airport by merely looking at the fact that
US$1 billion was invested by HACTL for the development of
SuperTerminal 1 (ST1). For Chek Lap Kok (CLK), the monopoly
was broken and two franchises were granted, one to HACTL and the
other to Asia Airfreight Terminal Company Limited (AAT), and each
would operate a separate cargo terminal. They were respectively
assessed to cater for about 80% and 20% of the cargo capacity expected

229
of the new airport. The readiness of the two cargo handling facilities
was considered by AA and Government as a critical airport operational
readiness (AOR) issue right from the early stage of the development of
the new airport.

11.2 The delay in the construction of HACTLs ST1 and AATs
premises gave rise to grave concern to AA and Government even before
the announcement of airport opening day (AOD) in J anuary 1998.
While AATs franchise provided a capacity of processing 1,100 tonnes of
air cargo per day by March 1998, HACTLs contractual completion date
to provide 75% throughput capacity, ie, about 1,800,000 tonnes per
annum out of the full capacity of 2,400,000 tonnes (excluding the annual
capacity of 200,000 tonnes for the Express Centre), was 18 August 1998.
HACTL was experiencing progressive and serious delay in the
construction of ST1 and the subject was always on the agenda of the
meetings of the AA Board and Airport Development Steering Committee
(ADSCOM). The construction works of ST1 had slipped so much
that throughout 1997 HACTL itself was worried about meeting the
requirement of readiness in April 1998, the then target date. When AOD
being 6 J uly 1998 was announced in J anuary 1998, W7 Mr Anthony
Crowley Charter, the Managing Director of HACTL and W2 Mr YEUNG
Kwok Keung, the Deputy Managing Director of HACTL both expressed
great relief for having three more months to get ST1 ready.

11.3 Since the contractual date of providing 75% capacity was 18
August 1998, HACTL had been operating on a best endeavours basis, ie,
it promised to use its best endeavours or efforts to get ST1 and the cargo
handling system (CHS) ready for operation to handle a certain
percentage of its yearly throughput capacity. At first when the target
opening date was April 1998, HACTL stated openly and to AA and
Government that it would be able to provide 50% of its full capacity in
April 1998, and later that readiness date was pushed to the end of April
1998. Its full capacity being 2.4 million tonnes a year, 50% meant 1.2
million tonnes a year. In early 1998, HACTLs assurance given to AA
and Government was that it would be able to provide 75% of its
throughput capacity on AOD which was then known to be 6 J uly 1998.
As far as New Airport Projects Co-ordination Office (NAPCO) is
concerned, although there had been constant concern over the readiness

230
of the building of ST1, there was no suspicion that CHS had any
problems. NAPCO knew all along that construction of ST1 suffered a
series of slippages, with building works delays affecting equipment
installation, commissioning and testing and impacting the installation and
testing of Government department support systems housed in ST1.
HACTL never reported any problem with CHS. For the last few months
before AOD, NAPCOs primary concerns related to the installation of
Governments systems and facilities at ST1 for Government departments,
such as the Air Cargo Clearance System (ACCS) to be used by the
Customs and Excise Department (C & ED) relating to cargo customs
clearance. When AAT obtained its occupation permit (OP) on 9 J une
1998 and HACTL eventually obtained its temporary occupation permit
(TOP) on 3 J uly 1998, everyone was relieved. The main worry had
gone. No one suspected that CHS would break down.

Section 2 : Cargo Handling on AOD

11.4 The condition of cargo handling on AOD can simply be
described as chaos. An outsider can see that the ramp at the northern
boundary of both ST1 and AATs building was full of cargo, scattered all
over a very large area. The most significant problem affecting AAT was
that a large backlog of cargo was allowed to be built up which heavily
congested its terminal and the interface area with the ramp. This led to
delay in AATs cargo handling and certain cargo being located. The
problem might have been caused by many more cargo arriving at AAT
than it had anticipated. AATs staff, who began to work in a new
environment and with a new system, simply could not cope. A lack of
co-ordination between AAT and ramp handling operators (RHOs)
exacerbated the problem. AAT made arrangements with the nearby
Airport Freight Forwarding Centre (AFFC) to use the latter for
breakdown, storage and collection of the backlog cargo, and from 18 J uly
1998 onwards, the severe congestion at AATs terminal started to abate.
This allowed AAT to process daily inbound and outbound cargo normally
and without further difficulties. The backlog was cleared by 13 August
1998. The difficulties experienced by AAT and the impact thus caused
were relatively small and manageable than HACTLs, not only because of
the smaller size of its operation as compared with HACTLs, but also

231
because its cargo handling system goes nowhere near the degree of
sophistication as HACTLs. The rest of this chapter therefore focuses on
HACTLs problems.

11.5 For the purposes of the inquiry, one needs to go inside ST1
to see what exactly was experienced by HACTL. For understanding the
difficulties that were encountered at ST1 on AOD, it is necessary to have
some appreciation of HACTLs CHS. CHS in ST1 is a very
sophisticated system consisting of five levels, as follows:

(a) Level 5 the Community System for Air Cargo (COSAC
2) which is the main computer system connected with
outside systems like Flight Display Data Feed Services
(FDDFS) and Societe Internationale de
Telecommunications Aeronautiques (SITA), etc and
accessible by customers, airlines and other users while it is
also linked to the lower levels of CHS;

(b) Level 4 the Resources Management System (RMS)
which manages resources such as availability and
deployment of manpower, shifts of personnel, scheduled
load requirements, etc, after analysing information from
level 5 such as customers instructions and flight information
for the purpose of managing the resources at ST1, resulting
in its giving commands to the lower levels of CHS in an
optimum and best prioritised manner;

(c) Level 3 the Logistic Control System (LCS) which
takes orders from upper levels and possesses the intelligence
of giving orders to the lower levels of CHS in accordance
with the command of RMS or COSAC or it can give orders
to the lower levels of CHS on its own independent from
RMS or COSAC;

(d) Level 2 the Programmable Logic Controller (PLC)
which takes orders from the LCS and then operates the
mechatronics; and

232
(e) Level 1 the mechatronics of the CHS which are the
mechanical, electrical and electronics equipment which
perform the work of handling cargo, including transferring
cargo on conveyor belts and automated transfer vehicles
(ATVs), putting cargo into the storage compartment and
retrieving cargo therefrom.

A diagram showing the CHS with its five level is at
Appendix XII.

11.6 The mechatronics of the CHS consist of two main
components, namely, the Container Storage System (CSS) and the Box
Storage System (BSS). Both CSS and BSS have stacker-cranes which
pick up cargo from ATVs or conveyor belts and lift it to the assigned
compartment for storage and retrieve it from the compartment whenever
needed. W2 Yeung emphasised that the mechatronics, being the lowest
arm of the 5-level CHS, was the most important element in the handling
of cargo. Without them, the whole CHS could not work, while they
could work alone even if the higher levels of CHS all failed. Although
LCS is a single computer system, it is linked separately to CSS and BSS,
giving orders through PLC for the two mechatronics systems to perform
work independently or collectively.

11.7 CHS has five basic operation modes utilising different
combinations of the sub-systems, namely,

(a) optimised mode (automatic mode): all levels are operating
together with RMS operating at full capacity planning for
LCS;

(b) inventory mode (automatic mode): COSAC and LCS
operating together but without RMS; messages defining cargo
locations are exchanged directly between COSAC and LCS;

(c) online mode (automatic mode): LCS, Equipment Motion
Control (EMC) and PLC are working together, without
RMS and COSAC; LCS continues to update COSAC and
RMS with inventory information as to the locations of Unit

233
Load Devices (ULDs) and bulk cargo;

(d) offline mode: the machinery is under the control of the
Interchange Server (IS) system and containers are
transported based on the default routings embedded in the
PLC software;

(e) manual mode: EMC and PLC are used as stand-alone with no
information to LCS, COSAC and RMS; the different units of
the machinery (sensors, drives, switches, etc) are still linked
and interacting; and

(f) maintenance mode: this is a sub-mode of manual mode where
each unit is controlled individually, eg, in maintenance mode
it is possible to control drives (such as to move a ULD)
without interference of messages sent from the sensors.

11.8 It is to be noted that when manual mode is operated, there
will be no automatic update of inventory information contained in
COSAC and RMS as to the locations of ULDs and bulk cargo. In such
circumstances, the operator will have to input data into the
Multi-Functional Terminals (MFT) to update the inventory.

11.9 CHS is a modular design in that each part of it can be
operated as an individual and independent module without having to rely
on another module. Not the entirety of CHS was required to operate in
order to process the projected amount of cargo on AOD, and indeed, not
all the equipment of CHS was fully commissioned. CSS was built on
the east and west sides of ST1, and on AOD, the whole of the west side,
namely, W1, W2 and W3 were to be used together with a part of the east
side, E1. BSS is, on the other hand, divided into north and south, and
both the North BSS and South BSS would be used on AOD. Set out
hereunder is a sketch showing the positions of the various parts of CHS.

234
J
W 1
H
E 1
L
M
G
NBSS
W 2
F
E 2
N
P
E
W 3
D
E 3
Q
R
Airside
Express
Centre
20-foot ULD
Storage Centre
(TUSC)
SBSS
Landside
ST1

11.10 The problems with ST1 and CHS on AOD and the period
thereafter can best be presented in the form of a chronology. The
chronology shown below is adopted from that prepared by Dr Ulrich
Kipper, one of the experts appointed by the Commission, and is mainly
based on a substantially contemporaneous chronology prepared by
HACTL which had been furnished to the Commission and supplemented
with other evidence received by the Commission. The chronology
summarises the events related to ST1 and CHS on AOD and shortly
before and after. For easy reference regarding the contents of the
chronology in this report, a reference number prefixed by AODH is
assigned to each of the events listed.

Date Time Event
2 J uly [AODH 1] Twenty-foot ULD Storage Centre (TUSC)
was flooded during the Fire Services drencher test, and the
Elevating Transfer Vehicle (ETV) was damaged and
therefore, not available on AOD.
3 J uly
[AODH 2] HACTL started to accept cargo in ST1;
minimal cargo were accepted. 291 empty ULDs were
relocated from Kai Tak to ST1.

235
Date Time Event
4 J uly
[AODH 3] Few cargo acceptance: bulk 7 consignments,
118 pieces, 2,130 kg; prepacked 33 ULDs; 109 empty
ULDs were relocated from Kai Tak to ST1.

5 J uly
[AODH 4] Few cargo acceptance: bulk 139 consignments,
2,186 pieces, 46,216 kg; prepacked 519 ULDs; 1,023
empty ULDs were relocated from Kai Tak to CLK of
which 655 were moved during the overnight period.

6 J uly Whole
day,
general
[AODH 5] Most areas of ST1 were available on AOD, and
it was expected that this would deal with anticipated load.
[AODH 6] Zones functional: Zones L and M, entire
Perishable Cargo Handling Centre (PCHC), E1 (Zones L
and M), W1, W2, W3.
[AODH 7] On AOD, all levels of computer systems were
operational.
[AODH 8] Manual issue of truck dock tickets due to
unavailability of VIS (ie, Vehicle Information System) and
Trucks Control Office (TCO) facilities.
[AODH 9] In the field (of an FDDFS message)
estimated_date_atc HACTL received only a few
messages, insufficient to be able to update COSAC with
the necessary information for RMS to effectively plan
operations and send instructions via LCS. (As a
comparison, HACTL received approximately 10,000
messages from FDDFS per day in September 1998.)
[AODH 10] Many units were unable to be located because
of faults in the system. Many flights left without carrying
their designated cargo for the whole day.
[AODH 11] Problems with both BSS and CSS noticed on
AOD.

[AODH 12] Many of the empty ULDs were put into CSS
in the morning of 6 J uly 1998 because they had arrived
with the last flight. There were reports of slow CSS
response. Both placing cargo into the empty ULDs and

236
Date Time Event
the preparation of cargo for despatch were slow.
00:00 [AODH 13] Continued from the late evening, prepacked
ULDs were received at W1, ie, Zones H & J . There were
occasional ATV faults which interrupted the reception
process. However, LCS and CSS were generally
operational.
[AODH 14] Due to long order processing time needed for
a storage order, all export loaded stacker boxes accepted
on 1/F were transferred to 3/F and 4/F workstation floor
for temporary staging via the cargo lift.
[AODH 15] After 00:00 hours, when Flight Data Display
System (FDDS) became operational there appeared to be
no information of real practical use.
[AODH 16] The Private Automatic Branch Exchange
(PABX) system was not functioning properly. Phone
numbers were wrongly directed. Staff started using their
own mobile phones for communication.
00:40 [AODH 17] Information Services Department (ISD)
informed Operations Project Team (OPT) that FDDS
would not be available, and AA would start faxing flights
estimated time of departure (ETD)/actual time of
departure (ATD) and estimated time of arrival
(ETA)/actual time of arrival (ATA) to ISDs fax
machine later in the morning.
02:00 [AODH 18] A CSS stacker crane in Zone J had stopped.
[AODH 19] Shift Manager, W19 Mr TSUI Shek Chui, felt
that it was necessary to override the automatic system and
operate in manual mode at Zone J only.
02:20 [AODH 20] Stacker crane SC0J 8 was switched to manual
mode of operation.
[AODH 21] About 30 CSS orders designated for CSS9J
were found queuing for processing. The processing then
took about three hours to complete instead of the normal
one hour.
03:00 [AODH 22] Control Systems Development Group
(CDG) identified that stacker crane had made incorrect

237
Date Time Event
reservation, causing some orders to be unprocessed. A
malfunction was reported at stacker crane no. SC0J 8 due to
an incorrect storage compartment reservation data within
LCS. The presence of a software bug was the
suspected culprit. CDG immediately carried out recovery
procedures to try to re-activate SC0J 8 so that appropriate
orders could be handled.
04:00 [AODH 23] Build-up activities in progress, but the
build-up staff needed to search for the cargo in loaded
stacker boxes at 3/F and 4/F. ULDs for export were
reported as urgent by OPT users. Decision by OPT to
operate SC0J 8 manually, creating inventory inaccuracies
through bypassing automatic equipment control.
Inaccuracies were either due to inputting incorrect
information or omitting to input certain information.
Such inaccuracies were also caused by delays in the keying
in of data.
06:00 [AODH 24] A lot of completed ULDs were waiting for
ATV pick up after LCS-CSS order had been raised.
System Support Team was called for assistance.
07:40 [AODH 25] The backlog of cargo had increased.
08:00 [AODH 26] A lot of units were still waiting at
workstations for ATV pickup.
09:00 [AODH 27] Most of the stacker cranes (other than SC0J 8)
were being operated in manual mode resulting in further
inventory inaccuracies.
[AODH 28] From 09:00 to 22:00 on 7 J uly maintenance
staff observed substantial number of faults in the
operations of CHS. They tried to keep CHS running
continuously, to reset and to restart the affected equipment,
whenever possible.
10:00 [AODH 29] More and more outbound ULDs accumulated
at workstations.
[AODH 30] ATVs of Level 3 and Level 4 workstation
floors were found unresponsive to serving incoming ULDs
that had arrived at the power conveyers. Engineering
Department (ENG) staff were requested to operate the

238
Date Time Event
ATVs manually.
[AODH 31] Inbound ULDs were spread over the whole of
ST1s northern interface with the ramp.
[AODH 32] Operators tried to use MFT to initiate transfer
orders. Some operators appeared unfamiliar with MFT
user screens, making the situation worse. They were
handling live operations in a new working environment
at ST1 for the very first time. Orders initiated by MFTs
were therefore either stopped or rejected as a result of
LCSs routing and reachability check functions.
Despite efforts being made by the CDG to rectify the
inventory data inaccuracies, more inaccuracies were
simultaneously caused by manual operators.
12:00 [AODH 33] A few inbound units received on 3/F and
cargo breakdown commenced.
[AODH 34] Due to the poor performance of LCS/BSS,
large amount of stacker boxes with export cargo could not
be stored back in system. Therefore, loaded boxes were
temporarily stowed at the eastern side truck docks.
14:00 [AODH 35] A lot of stacker boxes or consignments could
not be located upon cargo build up.
14:30 [AODH 36] All LCS supervisory functions were
unavailable for 1 hour and 15 minutes.
15:00 [AODH 37] Large amount of perishable cargo subject to
immediate release could not be located at the airside.
[AODH 38] Meeting called by W12 J ohnnie WONG Tai
Wah, General Manager-Operations, with W14 Ms Violet
CHAN Man Har, System Manager, W20 Mr Tony KWAN
To Wah, General Manager-Engineering and W10 Mr HO
Yiu Wing, Project Manager-Control. Everyone reported
problems. Determined to enforce shop floor staff to
actively report problems to the support teams. The
Meeting also decided to conduct physical ULD inventory
check starting midnight.
15:30
[AODH 39] The prepacked units acceptance point at W1
(ie, Zones H & J ) experienced frequent faults due to weight
discrepancy at transfer vehicle TVOJ 2.

239
Date Time Event

16:00
[AODH 40] LCS-CSS associated equipment were all
operated under manual mode.

17:00 [AODH 41] A meeting was held between CDG, ENG and
OPT. The following were agreed:
[AODH 42] (a) ULD inventory check would begin at
00:01 of 7 J uly and end at 02:00 7 J uly. During that
period, storage of ULD into CSS should be prohibited;
[AODH 43] (b) After the inventory check all CSS
equipment (in west wing) would be put back to automatic
mode. ATVs and building cargo hoists in W1 would be
partially put back to automatic mode, but ATVs and
building cargo hoists of W2 and W3 would remain
manually operated by ENG;
[AODH 44] (c) Operations Department (OPS) would
adopt fixed path for transferring ULD between levels 3 and
4 and ground level of airside building, in which cargo
hoists would be used.
18:00 [AODH 45] CDG and OPT agreed to suspend optimised
mode of operations. That was to detach RMS and
LCS-CSS linkage. No retrieval orders could be initiated
by RMS automatically for export ULDs. Operations staff
were required to initiate retrieval orders themselves.
19:00

[AODH 46] Operations Computer Project Manager
(OCPM) had sought for permission from General
Manager of ISD and General Manager of OPT to revert the
system to inventory mode.
20:00 [AODH 47] Large crowds of consignees who were holding
Shipment Release Forms (SRF) with dummy storage
locations were waiting for their cargo at the first floor
truck docks.
22:30 [AODH 48] ULD inventory check began at W2 and W3.
7 J uly Whole
day,
general
[AODH 49] On airside there was a backlog of inbound
cargo which had been dumped at the northern side of ST1.
Operations were being conducted manually as the

240
Date Time Event
automatic system of CSS had apparently failed.
[AODH 50] There were various recorded faults in the
computer terminal. Staff were therefore required to
physically look for the required cargo in all compartments,
thus slowing down the process significantly.
[AODH 51] On landside, boxes could not be entered into
BSS system because of the breakdown and were placed
outside the truck dock area.
[AODH 52] During the course of 7 J uly, CDG performed
software updates to enhance LCS (enhanced logging and
control software functions).
03:30 [AODH 53] ULD inventory check at W2 and W3
completed. Update of LCS-BSS inventory records to
correspond with LCS-CSS records. Inventory check at
W1 Zone J began with Zone H continued to operate under
automatic mode.
04:00 [AODH 54] A lot of inbound units from 6 J uly inbound
flight still outstanding and needed to be broken down.
Build-up outbound units waiting at workstations were not
picked up by ATVs.
[AODH 55] CDG found that a batch of ULD records in
LCS-CSS was deleted by a hidden system event.
04:30 [AODH 56] Inventory check at W1 Zone J completed;
check at W1 Zone H began.
05:00 [AODH 57] Acute shortage of dollies. The interface area
was jammed with dollies with inbound and outbound
ULDs.
05:45 [AODH 58] The compartment inventory database was
found corrupted. All CSS zones were manually operated.
06:00 [AODH 59] LCS-CSS compartment inventory as of the
image 6 J uly 23:00 restored by CDG. From this point
onward, all LCS-CSS and associated operations were run
under manual mode.
[AODH 60] A great number of boxes and bins were
dispersed around the warehouse floors as a result of

241
Date Time Event
LCS-BSS slow response to pick up boxes and bins.
06:15 [AODH 61] Due to the continued disruption in CHS
operations, more than 50 trucks were waiting for
pre-packed cargo delivery.
07:00 [AODH 62] Large number of trucks delivering prepacked
ULDs were waiting to be served. The trucks were
instructed to divert to Kai Tak.
08:00 [AODH 63] Due to delay in perishable cargo release, large
amount of perishable cargo from inbound flights of 6 J uly
was not collected by the consignees at PCHC truck docks.
10:00 [AODH 64] AA officers approved the marking of the north
interface area with staging zones and lanes.
12:00 [AODH 65] Meeting with airline representatives and C &
ED to discuss the situation.
15:00
[AODH 66] HACTL announced a 24-hour embargo on
export bulk cargo and import cargo on passenger flights
with the exception of perishables, strong room cargo,
newspapers, livestock and life saving materials (urgent
items).

18:00
[AODH 67] Prepacked cargo for export freighters and
inbound cargo from freighters other than urgent items was
to be processed at Kai Tak.

8 J uly
[AODH 68] HACTL imposed a 48-hour embargo except
urgent items.

9 J uly [AODH 69] TUSC recovered and back in operation.
[AODH 70] HACTL imposed a 9-day moratorium on all
cargo on all aircraft except inbound urgent items.

[AODH 71] HACTL started to clear out ST1 as (1) cargo
release would be quicker at Kai Tak; (2) clearance of CSS
and systems would allow HACTL to rectify the problems
with CHS and clean equipment, machinery and the

242
Date Time Event
building site of ST1.

10
J uly
[AODH 72] BSS suspended operation.
[AODH 73] HACTL did not know how long recovery
would take; might take months. By then, HACTL
decided to concentrate on CSS, as it was needed for export
processing, and BSS less important as BSS was functional
at Kai Tak.
14
J uly
[AODH 74] In a meeting, Mannesmann Dematic AG
Systeme (Demag), the supplier of CSS, put forward a
proposal to HACTL to develop an offline mode based on
the operations if PLC and the mechatronics to enable CSS
to operate only levels 1 and 2 of CHS.
13
Aug
[AODH 75] BSS resumed full operation.
24
Aug
[AODH 76] ST1 was back to full import and export
operation.

11.11 Shortly after midnight on 5 J uly 1998, there were a large
number of ULDs, empty or otherwise, that had been transferred from Kai
Tak to ST1. HACTL was trying to store these ULDs in CSS but from
W7 Charters evidence, the task was not completed even as late as
midday on AOD. It is clear from the chronology that in the small hours
of AOD, stacker crane SC0J 8, one of the three stacker cranes operating
that day for CSS, stopped [AODH 18] and Zone J had to be operated in
manual mode [AODH 20]. The 30 orders designated for CSS9J had to
be processed manually which took about three hours that could have been
completed in less than an hour under automatic mode operation in normal
circumstances [AODH 21]. The manual mode of operation created
inventory inaccuracies for upper levels of CHS, namely, LCS, RMS and
COSAC. The inaccuracies were caused by the operators keying in
inaccurate information of the location of the ULDs, or their delay or
omission in inputting the data [AODH 23,32]. More and more areas of
CHS responded slowly to orders and had to be operated in manual mode.
As a result, more and more inventory inaccuracies were created [AODH

243
27]. The manual mode of operation worked much more slowly than the
automatic mode, and backlogs of unprocessed cargo were increasing
[AODH 29]. The import cargo on the ramp delivered by ramp handling
operators (RHOs) were also building up [AODH 31]. A decision was
taken to manually check the inventory [AODH 38,42], which was
performed in the small hours of 7 J uly 1998 [AODH 48,53,56], but by an
inadvertent application of a programme used for testing, the inventory
was deleted [AODH 55,58]. This gave rise to serious suspicion that
there was something gravely wrong with the systems, and an embargo
was declared at 3 pm on 7 J uly 1998 [AODH 66], while HACTL was
considering ways to recover. The details and the analysis of the causes
for the problems are contained in Chapter 14.


11.12 As can be seen from the chronology above, the embargo was
prolonged from 8 J uly to 18 J uly 1998. In fact, apart from the TUSC,
the Express Centre, which was part of ST1, did not experience any
difficulty on AOD or thereafter. The Express Centre contains strong
room facilities, the HACTL Express module to process onboard courier
material and HACTLs three express operator tenants, UPS Parcel
Delivery Services Limited, DHL International (Hong Kong) Ltd. and
TNT Express Worldwide (Hong Kong) Limited. The facilities in the
Express Centre never stopped operation. The Express Centres
200-position container handling system was fully operational, enabling
HACTL to handle in a programmed manner approximately eight
outbound freighter loads per day. On 16 J uly 1998, HACTL announced
a four-phase recovery programme, as follows:

Phase One: From 23:59 on 18 J uly 1998, HACTL would begin
to process 50% of the projected daily tonnage of both imports and
exports. Cargo to be managed would be restricted to prepacked
cargo on freighters only. During this phase, imports would be
processed at Kai Tak except import urgent items would continue
to be handled by ST1.

Phase Two: By the end of J uly, operations would be extended to

244
cover prepacked export cargo on both freighters and passenger
aircraft and all import cargo on freighters and passenger aircraft,
representing not less than 75% of the projected total daily
tonnage.

Phase Three: From mid-August, the service would be fully
operational using both ST1 and Terminal 2 at Kai Tak. HACTL
would process 100% of projected tonnage of both imports and
exports, with operations extended to cover prepacked and bulk
export cargo.

Phase Four: All export and import cargo operations would be
handled by ST1 at the end of August 1998.

11.13 The four phases of recovery were carried through quite
efficiently and indeed, ST1 apparently recovered fully on 24 August 1998,
ahead of the planned time, when it started to handle all cargo, imports and
exports. The details of the history of ST1s recovery are evident from
the press releases and the statements made by HACTL from time to time.
The periods of the moratoria imposed by HACTL on cargo that it would
handle and the details of the recovery programme are summarised below:

Date of
Announce
-ment
Date or
Period
Covered

Type and Quantity of Cargo

Place of
Processing
6 J uly ST1 opened

ST1
7 J uly 24 hours

From 7
J uly

Embargo on all export bulk cargo and
imports on all passenger freighters
except urgent items

Prepacked export cargo

Freighter inbound cargo

Export and import urgent items

Terminal 2

Terminal 2

ST1

245
Date of
Announce
-ment
Date or
Period
Covered


Place of
Processing
8 J uly 48 hours Arrangements put in place on 7 J uly
extended

9 J uly 9 18
J uly
Moratorium on all cargo on all aircraft,
except inbound and outbound urgent
items

Inbound and outbound urgent items
(approximately 10% of all cargo)

The cargo currently at ST1 would be
moved to Terminal 2 for storage and
distribution

ST1

Terminal 2
15 J uly From 15
J uly
HACTL eased restrictions to accept
outbound cargo on narrow bodied
aircraft

The Express Centre with strong room
facilities and with 3 express operator
tenants, UPS Parcel Delivery Services
Limited, DHL International (Hong
Kong) Ltd. and TNT Express
Worldwide (Hong Kong) Limited was
fully operational

Import perishables accounted for about
319 tonnes of the total cargo handled
in ST1

HACTL able to process approximately
1,400 tonnes of cargo per day out of an
expected daily load of 4,000 tonnes

Terminal 2

ST1

ST1

ST1 and
Terminal 2

246
Date of
Announce
-ment
Date or
Period
Covered


Place of
Processing
All of the over 2,000 containers
previously held in ST1 CSS had been
transported to Kai Tak, by using barges
and trucks

Terminal 2
16 J uly From 16
J uly
HACTL able to process over 1,900
tonnes of cargo per day

HACTL announced four-phase
recovery programme for air cargo
services using both ST1 and Terminal
2 in Kai Tak

ST1 and
Terminal 2
24 J uly From 18
J uly

From 24
J uly
HACTL had been processing an
average of 2,520 tonnes of cargo per
day

HACTL started to accept one pallet of
general cargo per inbound passenger
aircraft, ie, an addition of about 300
tonnes of cargo per day. With the
perishable cargo, HACTL would be
handling over 30% of the projected
daily import tonnage from passenger
aircraft

HACTL also started to accept some
prepacked pallets for a limited number
of outbound passenger flights

ST1 and
Terminal 2

ST1 and
Terminal 2

ST1 and
Terminal 2
28 J uly From 21
J uly
HACTL was able to handle an average
of 2,708 tonnes of cargo per day using
ST1 and
Terminal 2

247
Date of
Announce
-ment
Date or
Period
Covered


Place of
Processing

From 29
J uly
both ST1 and Terminal 2

HACTL began to process two pallets
of general cargo per inbound passenger
aircraft. HACTL brought in 33 more
trucks with roller beds stripped out of
Terminal 1 at Kai Tak welded onto the
trucks to make up a fleet of 200 to
handle the increased inbound workload

Terminal 2

29 J uly From the
week
before

From 30
J uly
Accepting 3 export pallets per flight
for a limited number of passenger
aircraft

HACTL launched full resumption of
all inbound cargo handling services,
accelerating the recovery plan for all
inbound cargo by about 24 hours. All
imports handled at Terminal 2 with the
exception of urgent items
Import and export urgent items

Terminal 2

Terminal 2

ST1
6 August From the
week
before

From 9
August

From 11
August

The tonnage handled by HACTL had
been at 80% of normal levels

HACTL would handle on all outbound
passenger flights 2 pallets, either
prepacked or built up by HACTL, plus
500 kg of loose cargo

Cargo on all Cathay Pacific Airways
Limited (Cathay Pacific) inbound
flights, both passenger and freighter,
ST1 and
Terminal 2

ST1 and
Terminal 2

ST1

248
Date of
Announce
-ment
Date or
Period
Covered


Place of
Processing

From 15
August
would be handled at ST1 (Cathay
Pacific accounted for more than 80%
of Hong Kongs total transshipments
last year)

HACTL would accept all export cargo
and processing would be at either ST1
or Terminal 2

ST1 or
Terminal 2
13 August From 12
August

From 14
August

HACTL had handled all Cathay Pacific
inbound cargo

HACTL uplifted the remaining partial
restrictions on outbound cargo for both
passenger and freighter aircraft
ST1

ST1

HACTL would handle all outbound air
cargo

All inbound cargo from airlines other
than Cathay Pacific would be handled
and processed at Terminal 2

Phase Three of recovery programme
completed
ST1

Terminal 2
18 August From 14
August

HACTL had been handling all
outbound cargo for all flights at ST1

Inbound cargo from aircraft, both
passenger and freighter of China
Airlines Ltd., Thai Airways
International Public Company Limited,
Air Hong Kong Limited, J apan
Airlines Company Limited, Korean Air
Company Limited and Eva Airways
ST1

ST1

249
Date of
Announce
-ment
Date or
Period
Covered


Place of
Processing

On 18
August

From 18
August
Corporation was processed at ST1

Certificate of Operational Readiness
issued

ST1 handled 90% of all HACTLs
cargo, and the remainder, processed at
Terminal 2, would be transferred to
ST1 by the end of August 1998

ST1

ST1 and
Terminal 2
20 August From 24
August
HACTL would handle all cargo at ST1

ST1
24 August From 24
August
HACTL was handling all cargo at ST1

The four-phase recovery programme
which began on 18 J uly 1998 was
completed, some 8 days ahead of time
ST1

11.14 Nothing of note happened between 24 August 1998 and 15
October 1998 although as late as mid-September 1998, Hongkong
Association of Freight Forwarding Agents Ltd and a few individual
freight forwarders still wrote to the Commission to state that their cargo
handled by HACTL were missing. However, the Commissioners
believe that because there were only a few of such complaints, HACTLs
performance as CTO was generally satisfactory as from 24 August 1998,
and the complaints were a legacy of the confusion and deficiency that
plagued ST1 on AOD.

11.15 On 15 October 1998, there was a structural failure of a
section of ST1s power distribution system linking ST1 to a local
substation causing disruption to ST1s operation. A large section of the
ceiling suspended bus-bars and cables used in the distribution of

250
commercial power to certain parts of the building collapsed around 6 am
on 15 October 1998, cutting power to mainly the eastern half of the
terminal building. This resulted in a number of airline offices losing
power, limiting their ability to communicate with counter-parties required
to be informed concerning cargo movement, both in and out of ST1.
The power failure affected ST1s operational efficiency and slowed down
the processing time for some types of cargo. The Express Centre and
the PCHC were operating normally. Temporary power was restored to
the airline offices for operating computers, telephones and faxes some 12
hours after the failure, and other temporary power measures were used to
restore power to other affected areas. The water pumps servicing the
fire hydrants and the drencher systems were not affected. Power
supplies to the pumps servicing the sprinkler systems were restored by 10
am on 17 October 1998. Whilst the stacker cranes in zone E1 of the
terminal continued to be energised in spite of the bus-bar problem, the
airside interface of E1 was cut off from power until temporary power was
made available by 2:30 pm on 17 October 1998. The sub-systems
peripheral to the E1 CSS remained without power until noon on 21
October 1998. This, whilst enabling HACTL to deal with the airside
load without any reduction in capacity, rendered it inefficient for HACTL
to process the inbound load as the overall build-up and breakdown
facility within the terminal was reduced to a little over 60 % instead of
75%. As a result, some amount of outbound cargo shutout from flights
had to be dispatched later in the day or on the next flight on the following
day. There was also some delay to the breakdown of inbound load
which led to a slower service on the landside. Close dialogue was
maintained with customers throughout the period and operations
remained under reasonable control. Permanent power for
air-conditioning and full lighting in the offices was restored on 20
October 1998 and all other affected areas of ST1 were connected with
permanent power on 22 October 1998. HACTLs engineers and
engineering consultants were satisfied that collapse of the remaining parts
of the network was unlikely as its support was different from that of the
collapsed portion. The exact cause of the collapse of the ceiling
suspended busbars is not yet known. Binnie Consultants Limited have
been appointed by HACTL to investigate into the incident.

11.16 The reasons given by HACTL as to how to rectify the

251
problems it encountered can be gleaned from its early press releases after
AOD.

(a) 7 J uly: The impact of the high volume of ULDs moved from
Kai Tak for entry into CSS at ST1 has resulted in inaccuracy
in our ULD inventory. We have also encountered
computer system difficulties. We now have to buy time to
rectify these system problems.
(b) 8 J uly: at the same time (of taking advantage of our
facilities at Kai Tak) allowing our engineers and contractors
adequate time to rectify current hardware and software
problems with our Box Cargo Storage Systems.
(c) 9 J uly: a moratorium will assist the company in rectifying
software and mechanical problems which have impacted upon
the efficiency of the buildings operation.
(d) 10 J uly: It ( the moratorium) will enable us to address and
deal with software and minor electrical and mechanical
equipment problems which have not enabled the ST1 to
operate at the levels of efficiency needed to deal with inbound
and outbound cargo demands. When we restart operations
we will build up gradually and therefore will have to limit and
control the inbound and outbound flow of freight. The use
of Terminal 2 at Kai Tak for import handling and distribution
is likely to continue for a few months.
(e) At a meeting on 15 J uly 1998 on cargo handling operations
between HACTL and Government at Chief Secretarys Office:
FS (Financial Secretary) asked about the progress with sorting
out the bugs. Both W7 Charter and W2 Yeung suggested
that computer software was not the main problem. Rather it
was the electrical and mechanical faults caused by the
environment of a building site. W2 Yeung recounted that
the computer system never really ceased to work since Day
One but because of failures in the mechanical system, workers
were forced to go manual on the CHS and in the process

252
mucked up the database and wiped out memories
inadvertently.
(f) HACTL ST1 Operational Recovery Strategy dated 15 J uly
1998 reads: Operations resumption as from 2359 hours 18
th

J uly will commence on the basis of the use of off-line mode
using PLC controls. This is to minimise the effect of a
possible high rate of equipment faults on the performance of
higher modes of controls, which whilst delivering more
operational efficiency, will carry higher risks. To further
minimise our risk, we will concentrate our operations restart
initially based on CSS resources. We will then introduce
operations requiring automation of BSS. The reason for
this decision (to commence with pre-packed cargo handling
using CSS only) is to avoid the use and therefore the risk and
burden of managing BSS, during the restart operation so that
management and technical attention can focus solely on CSS
initially. This (the reception and processing of export bulk
using BSS) will be dependent on the availability of BSS
which by then will have been checked out and be in a position
to support a stable operation.

11.17 From all the press releases and public statements made by
HACTL, the Commissioners can identify the progress of ST1s recovery.
Yet it can be noticed that as from 15 J uly 1998, HACTL had altered from
their openly stated positions of computer system difficulties, current
hardware and software problems, software and minor electrical and
mechanical equipment problems to computer software was not the
main problem, it was the electrical and mechanical faults caused by the
environment of a building site and the computer system never really
ceased to work since Day One. The course of the inquiry conducted by
the Commissioners was hindered by the fact that HACTL was not too
forthcoming with the actual reasons that caused the difficulties at ST1.
The Commissioners therefore feel hesitant about what had actually been
done by HACTL to rectify the problems with CHS. They can only base
their conclusions as to the causes of the breakdown of ST1 from the
evidence of Murata (the contractor for BSS equipment) and Demag (the
contractor for CSS equipment) and HACTLs press releases. The
contents of the press releases are reliable because they would have been

253
acted upon by air cargo owners, freight forwarders and airlines in their
use of ST1s services. The causes for the problems and the
responsibility for them can be found in Chapter 14.

254
CHAPTER 12

THE OTHER MAJOR PROBLEM AND
MODERATE PROBLEMS

Section 1 : The Other Major Problem : Baggage Handling

Section 2 : Moderate Problems

Section 1 : The Other Major Problem : Baggage Handling

12.1 The Baggage Handling System (BHS) at the new airport
is one of the most advanced systems in the world. It is a centralised,
highly automated system controlled by computer and incorporating a
high level of security. It has three functional areas departures, arrivals
and transfers.

12.2 Departing passengers at the new airport will check-in their
baggage at check-in desks in the Departures Hall on Level 7 of the
Passenger Terminal Building (PTB). At the check-in desks, baggage
will be labelled by airline staff and put on the conveyors to go to the
Baggage Hall. Baggage that cannot be safely conveyed, such as soft
bags and bags with straps, will be placed in plastic tubs before being put
on the conveyor. Oversize or out-of-gauge (OOG) bags are taken by
airline staff down to the Baggage Hall at Level 2 via the OOG lifts.
The baggage then goes through sortation and security screening. Upon
check-in, conveyors will take the baggage down into the Baggage Hall
where one finds the BHS equipment and machinery. Check-in can also
be done at the In-Town Check-In desks at the two major Airport Railway
(AR) stations. These baggage will then be transferred to the Baggage
Hall and be injected into the system. In the Baggage Hall, the
automated sortation system will direct the baggage to the appropriate
flight laterals. At these laterals, which are effectively collection points
254
for flight-sorted baggage, ramp handling operators (RHOs) will
transfer the baggage to containers which are then taken to the aircraft by
road vehicles for loading onto the aircraft.

12.3 Arrival baggage is unloaded from aircraft by RHOs and
brought by road vehicle to the conveyor loading stations, located in the
Baggage Hall. RHOs will then transfer the baggage onto conveyors
that take it to the reclaim units or carousels in the Arrivals Hall on Level
5. Arrival passengers will be directed to the particular carousel,
through information displayed on liquid crystal display (LCD) boards,
to collect their baggage. Transfer baggage is collected from the aircraft
in a way similar to arrivals baggage but is injected into the Departures
sortation system. From there, it is treated as departure baggage and
automatically directed to the correct flight laterals.

12.4 The departure/transfer machinery is independent of the
arrival machinery. The former consists of a large and complex systems
of conveyors, scanners and laterals. The latter consists of feeder
conveyors in the Baggage Hall which link up with the carousels in the
Baggage Reclaim Hall (BRH).

12.5 BHS sorts departure and transfer baggage automatically and
routes them to the correct departure lateral for collection by RHOs who
will then dispatch the baggage to the appropriate aircraft on the apron.
The sorting is done through the reading by BHS of the 10-digit barcoded
licence plate number on the baggage label printed by airlines and by
looking up the corresponding Baggage Source Message (BSM) in the
BHS Sort Allocation Computer (SAC). BSM has been produced and
transmitted to BHS by the airlines via the Common User Terminal
Equipment (CUTE) during check-in. If the baggage label is not read
at the automatic coding station, it will be diverted into a no-read loop
where the baggage will be read by a staff at the manual coding station.
If the baggage is late or it cannot be sorted by automatic or manual
coding, it will be sent to the problem bag area where it will be removed
and dealt with by Swire Engineering Services Ltd (SESL) staff and
RHOs who will take the baggage to the appropriate aircraft.

12.6 Where departure bags miss their flights (which have
255
departed or closed), or have labels that cannot be read by either
automatic or manual coding, or cannot for some other reason be sorted
by BHS onto the correct flight lateral, they will be sorted into the
late/problem bag area. The bags then need to be manually removed
from the system by BHS operators, to await collection by RHOs.

12.7 The Airport Authority (AA) contracted the design and
build contract of BHS to a consortium consisting of SESL, Vanderlande
Industries Hong Kong Ltd. (Vanderlande), Crisplant Limited
(Crisplant) and Siemens AG (Siemens). While BHS is operated
and maintained by SESL, baggage handling is performed by the three
RHOs, namely, Hong Kong Airport Services Limited (HAS), J ardine
Air Terminal Services Limited (J ATS) and Ogden Aviation (Hong
Kong) Limited (Ogden).

12.8 BHS is an important system at the new airport. It affects
flight departures and the time in which arriving passengers can collect
their baggage. As the baggage handling chaos on airport opening day
(AOD) and the few days afterwards show, problems with BHS can
have a huge ramification on the efficient operation of the new airport.
The baggage handling problem will have a direct and significant impact
on passengers, arriving or departing, causing delays and inconvenience
to them. The problems with BHS and the actions taken to remedy the
situation are dealt with in the following paragraphs.

12.9 There was a serious problem in the handling of baggage on
AOD. According to AA statistics, some 10,000 of 20,000 departure
and transfer bags missed their flights on AOD. W30 Mr Ben Reijers,
Senior Design Engineer for BHS, testified that there might be around
6,000 instead of 10,000 problem bags. Some departure bags were
loaded onto flights late, adding to delays in flights departing. Departure
baggage handling started getting unmanageable by about 9 am on AOD.

12.10 On the first week of AOD, arrival passengers experienced
significant delays in reclaiming their baggage. From Days Three to
Seven, arrival passengers had to wait an average of one hour 41 minutes
to collect their bags. There was also some confusion as to where bags
were to be picked up.
256

12.11 The problems relating to baggage handling were serious in
the first few days of airport opening. Passengers were inconvenienced
and the standards previously achieved at Kai Tak were not met at the
new airport until about the second week. The effect of the baggage
handling problem was compounded by the other problems happening on
that day, in particular, the problem with the Flight Information Display
System (FIDS). Flights were delayed, there was confusion over stand
and gate allocation and parking of planes. There were also problems in
the allocation of reclaim carousels at BRH and in the display of the
carousel numbers.

12.12 It is clear that the problems were caused by a number of
separate and discrete matters, including human error. Some problems
were the effect of other problems encountered in airport operations, eg,
with the FIDS and the Trunk Mobile Radio (TMR). Each baggage
handling problem had a significant impact, if not by itself, certainly
when combined with the other problems encountered. The problems
are classified as follows:

(a) accumulation of problem bags;

(b) system stoppages;

(c) delay and confusion in handling arrival baggage;

(d) stretching resources of RHOs; and

(e) inexperience or unfamiliarity of RHO, airline and SESL
staff.

(a) Accumulation of problem bags

12.13 The main cause of the chaos for departure bags was the
accumulation of a very large number of problem bags in the Baggage
Hall which led to system die-back and stoppages. Many of the problem
bags were not sorted and eventually missed their flights. On AOD
approximately 30% of all bags went into the problem bag area as
257
compared to 3% per day in normal circumstances at Kai Tak. About
5,000 problem bags remained at the end of AOD. Pictures showing the
problem bag area with bags piling up are exhibited at Appendix XIII to
this Report.

12.14 Bags are sorted into the problem bag area when departure
bags miss their flights because the plane has departed or the laterals have
been closed, when baggage labels cannot be read by either automatic or
manual coding, or when for other reason they cannot be sorted by BHS
into the correct flight lateral. The bags have to be removed from the
system by hand and collected by RHOs.

12.15 On AOD, the problem bags had to be sorted manually
between the three RHOs and further by each RHO according to their
respective flights. RHOs resources were thus stretched, and delays
were experienced in transporting baggage to departing aircraft and in
delivering baggage to arriving passengers waiting at the reclaim
carousels at BRH on Level 5.

12.16 BHS was designed to deliver 1,400 problem bags per hour.
This, to certain extent, depended on the capacity of RHOs in sorting the
bags. According to W30 Reijers, it was expected that the staff would
handle one problem bag per minute. However, on AOD, bags arrived at
the problem bag area at a rate of around 10 to 15 per minute. This
created difficulties to RHOs who had to remove the problem bags
manually from the system. Bags not removed in time caused system
die backs in the problem bag area and these die-backs together with other
stoppages of the system led to more bags becoming late and problem
bags. This vicious cycle led to the extreme inefficiency of operations in
the Baggage Hall.

12.17 To relieve the problem bag area, at about 3 pm on AOD,
SESL reset BHS parameters to divert problem bags to laterals instead of
to the problem bag area. In order to achieve this, all infeeds had to be
closed for about two hours and bags manually sorted by SESL staff.

12.18 The large number of problem bags was not caused by one
single factor. Rather, a number of incidents happened on AOD led to
258
this result. Some of the incidents are set out below. It should be noted
that the incidents relating to system stoppages described later also
contributed to the large number of problem bags. The particular
problems are numbered to facilitate further reference in this Report.

12.19 [BHS 1] Cathay Pacific Airways Limited (Cathay Pacific)
and Securair Limited (Securair) staff fed about 220 bags from Kai Tak
with no baggage labels into the conveyor system at the new airport.
There were about 815 pieces of interline baggage at Kai Tak, of which
about 420 were brought to the new airport by Securair on AOD. These
bags were not scheduled to depart on any flight from the new airport on
AOD. Despite instructions from Mr Victor WONG Chu King, System
Manager Airport of Cathay Pacific to Securairs staff to use fallback
tags or the OOG lift to deliver the bags to the Baggage Hall, Securair
staff with the assistance of Cathay Pacific staff, put some 220 bags on the
conveyor belt without tags. As these bags had no baggage labels, BHS
identified them as problem bags and rightly diverted them to the problem
bag area.

12.20 In the afternoon, Mr. Wong noticed his instructions had not
been followed and stopped the use of the conveyor belt. The remaining
200 bags were then sent to the Baggage Hall by the OOG lift. The lift
was also used on 8 J uly 1998 when a further 335 bags were delivered to
the new airport. The remaining 40 pieces were taken to Securairs
central tracing office.

12.21 [BHS 2] Airlines checked in bags with incorrect labels or
invalid or no BSMs. Some departure and a large percentage of transfer
bags bore labels with bar codes that were not recognisable by BHS, or
were given BSMs of an incorrect format. W30 Reijers thought that
about half of the problem transfer bags were the result of invalid labels.
J apan Airlines Company Limited (J AL) accepted that it introduced
perhaps 600 bags with unrecognisable BSMs on AOD, because an old
version of its computer programme had been mistakenly loaded in Tokyo.
Thai Airways International Public Company Limited (Thai Airways)
admitted that seven of its transfer bags had labels that could not be read
by BHS.

259
12.22 AA encouraged airlines to use labels that met International
Air Transport Association (IATA ) recommended specifications, which
would satisfy baggage reconciliation security requirements as well as
baggage identification requirements. However, adopting the
specifications was a management issue for airlines who were under no
obligation to do so. AA was aware that not all airlines would provide
labels that met IATA standards. AA also foresaw that , some labels that
met the standards might not be read by BHS. Accordingly, AA
developed problem bag procedures to ensure a bag with a label that
could not be read by the system would be routed to its proper destination.
The drawback of this solution was that it put pressure on manual
resources when the problem bag area was overloaded. This
unfortunately materialised on the first day of operation of the system.

12.23 The inability of BHS to recognise BSMs was not always
caused by airlines. In the case of J apan Asia Airways Company
Limited, the wrong prefix (J L instead of the correct EG) was
programmed for recognition by SAC in BHS for its bags. On AOD, all
bags for this airline were diverted to the problem area as BHS was
expecting BSMs labels for J L206 and bags with EG206 were unknown
to the system. This problem was rectified within a few days after AOD.
It is not clear from the evidence whether it was SESL or AAs Airport
Operations Control Centre (AOCC) who programmed the prefixes for
recognition by SAC.

12.24 [BHS 3] Airlines checked in about 2,000 bags with invalid
flight numbers. Some airlines entered flight numbers for baggage
labels and BSMs that were different from those listed in the flight
schedule, and were thus not recognisable by BHS. These bags were
sent to the problem bag area. In one case, Flight CP8 of Canadian
Airlines International Limited (Canadian Airlines) was destined for
Vancouver and Toronto. On the same flight, there were nine passengers
who travelled from Hong Kong to Montreal via Vancouver with 21
pieces of baggage. Their baggage was tagged through to Montreal
under a funnel flight number CP1088. BHS was unable to identify
these bags which was sent to the problem bag area. Canadian Airlines
admitted that it was responsible for the incident. It claimed that it was
not aware that they should inform AA about the extra flight numbers on
260
BSM as AA did not consult the airlines about the use of extra flight
numbers on BSM. No inconvenience was caused to passengers as the
bags were picked up from the problem bag area and loaded on the same
flight. Another case involved Ansett tagging bags with the originating
Ansett flight instead of the connecting Virgin Atlantic Airways Limited
(Virgin) flight from Hong Kong.

12.25 [BHS 4] Aviation Security Company Limited (AVSECO)
staff rejected a large number of bags at Level 2 security screening,
putting pressure on Level 3 screening, lengthening baggage handling
time and causing more problem bags. It was alleged that AVSECO
operators had rejected more bags at Level 2 security screening than in
normal circumstances, probably being more cautious on the first day of
operation, or did not decide within the set period of time whether to clear
the bags or not. Many bags were thus automatically diverted to Level 3.
Some of the non-conveyable bags had not been placed in tubs and were
lost in tracking. These were also discharged to security screening on
Level 3. AVSECO staff had difficulties in processing such a large
number of bags, resulting in more problem bags. According to
AVSECO, of 6,705 bags screened at [Level 3], 860 were bags rejected
from Level 2, 1,713 were mis-tracked and 4,132 were diverted to Level 3
as a result of tubs not being used and the high number of emergency
stops. The problem with emergency stops will be dealt with below.

12.26 [BHS 5] RHOs delivered transfer bags from inbound flights
into BHS after connecting flight laterals had been closed. This was an
example of the difficulties faced by RHOs as a result of the other
problems faced by them on the apron. The lack of flight related
information from FIDS and the inefficiency of means of communication
meant that RHOs were delayed in meeting inbound flights and thus in
delivering transfer bags to BHS. The frequent stoppages of the system,
including intermittent stoppages of three out of four induction belts at the
Central Transfer System, also contributed to the problem. Since the
connecting flight laterals had been closed when the transfer bags were
fed into BHS, those bags were diverted to the problem bag area. This
problem would have occurred before 3 pm on AOD, as after that time
departure flight laterals were kept open to allow circulation of problem
bags.
261

12.27 [BHS 6] RHOs did not clear bags from departure laterals in
time, resulting in full lateral alarms, which caused subsequent bags to go
to the problem bag area. W30 Reijers thought that this created around
800 problem bags.

12.28 [BHS 7] One of RHOs, Ogden, put about 230 arrival bags
from a KLM flight No. 887 onto transfer laterals. There were
altogether about 260 bags from this flight, consisting of 30 transfer bags
and 230 arrival bags. They were put onto the transfer laterals by Ogden.
Whilst Ogden had rightly put the 30 transfer bags onto the transfer lateral,
the 230 arrival bags should have been put onto the arrival laterals.
These arrival bags were thus sorted by BHS into the problem bag area.
Delay and inconvenience were caused to the arriving passengers on that
flight. The bags were retrieved by Ogden and were placed on the
appropriate reclaim belt. Most of the arriving passengers received their
bags on the same day. W30 Reijers alleged that he saw Thai Airways
and Aeroflot Russian International Airlines incorrectly put arrival bags
onto transfer laterals although this was denied by the airlines.

(b) System stoppages

12.29 There were some 500 stoppages of the system on AOD.
One even lasted from late morning to mid-afternoon on AOD. Airline
staff had to transfer bags from one conveyor belt to another. Stoppages
in turn led to the accumulation of more late and problem baggage. The
sorter system produced problem bags faster than it could discharge and
the whole BHS started to die-back up to the infeed points Hence
system stoppages and problem baggage caused a vicious cycle which
eventually led to extreme delays in baggage handling.

12.30 Stoppages were caused or exacerbated by the actions or
omissions set out below.

12.31 [BHS 8] Bags that could not be safely conveyed were not
put in tubs and OOG bags were fed into the conveyor system instead of
being sent down to the Baggage Hall via the OOG lift. At check-in
counters, airline staff will label departure bags and place them on the
262
conveyor belts to be sent to the Baggage Hall. Bags that cannot be
safely conveyed, such as round or soft bags that will roll along the
conveyors and rucksacks with straps that cannot be secured and so may
become stuck in the system, should first be placed in plastic tubs to be
conveyed on the system. Airline staff should take large bags to the
OOG lifts to be transferred to the Baggage Hall, otherwise the conveyor
belt would be jammed. These procedures were not new and airlines
had been trained and reminded of them. There were about 200 to 250
bag jams on AOD caused by bags not being put in tubs and oversized
bags being put on the conveyor. These jams caused system stoppages.

12.32 According to Mr Klaus Sterzel, Project Manager of Siemens
and Mr Christopher J ames Bleasdale, Contract Director of SESL, at
about 9 am on AOD, BHS started to die back because the secondary
sorter infeeds were stopped too frequently due to incorrectly presented
baggage. The offending bags were usually oversized or not placed in
tubs when they should have been.

12.33 [BHS 9] Too many erroneous emergency stops led to
numerous disruption and system down time. The emergency stop
buttons were pressed some 99 times on AOD. Many of the stoppages
may have been deliberate, as when bags had to be manually removed
from the system by the staff of SESL, RHOs or AVSECO. In one
instance, a SESL staff restarted the system without removing a
non-conveyable bag that went underneath a tilt tray sorter because it was
not put in a tub. This resulted in damage to part of the system and some
system down time.

12.34 RHOs explained that the emergency stop buttons might
have been pressed accidentally, due to the protruding design of the
buttons. AA and SESL, in consultation with the Labour Department
safety officers, subsequently installed a cover to prevent accidental
activation of the button. Although the protruding design could have
resulted in easy accidental activation, there were competing safety
considerations for making buttons easily accessible in an emergency.

12.35 [BHS 10] Communication difficulties between operators in
the Baggage Hall due to TMR overload and unavailability of other
263
means of communication resulted in longer times for the system to be
reset each time it was stopped. W30 Reijers claimed that this
exacerbated the problems caused by system stoppages because operators
had difficulties communicating with each other and resets of the system
which could have taken one to two minutes had taken 10 minutes
instead.

(c) Delays and confusion in handling arrival baggage

12.36 On AOD, delays and inconvenience were caused to arriving
passengers who had to wait longer than usual before they could retrieve
their bags. Arrival baggage is first brought by road vehicle from the
aircraft to the conveyor loading stations located in the Baggage Hall.
There, RHOs transfer the bags onto conveyors that take the bags to the
reclaim carousels in the Arrivals Hall on Level 5. Passengers would
proceed to the reclaim carousel assigned by the baggage handling
operator (BHO) from SESL in the Baggage Control Room (BCR)
and displayed on the FIDS monitors located after immigration clearance
and in the Arrivals Hall, to pick up their bags. A number of problems
arose in the above process, details of which are set out below.

12.37 [BHS 11] RHOs had no reliable flight information from
FIDS and had communication difficulties due to TMR and mobile phones
overload and unavailability of other fixed lines of communication.
There were delays in collecting bags from aircraft and transferring them
to the Baggage Hall. This was the result of the snowball effect of
delays on the apron caused by a number of factors. Stand allocation by
Apron Control Centre (ACC) was delayed due to problems
encountered with flight swapping early in the morning of AOD and with
Terminal Management System (TMS). Stand allocation input into
FIDS was delayed also by the slow response time of FIDS. At about
noon, where the new airport experienced full apron, aircraft that had
landed had to wait on the taxiway for the next available stand. There
were problems disseminating flight information to RHOs, which
increased ground time for handling arriving passengers and baggage.
Flight information was not displayed via Flight Data Display System
(FDDS). TMR was overloaded, creating difficulties in users
obtaining a channel of communication and there were insufficient fixed
264
lines of communications to enable RHOs to get in touch with ACC or
AOCC to obtain stand information and arrival times. No joy could be
found in the use of mobile phones whose networks also experienced
overloading problem. Hence RHOs were hampered in their operations
because they had difficulties knowing the time and at which stand the
aircraft would park.

12.38 At about 4 pm on AOD, a meeting was held in the Airport
Emergency Centre (AEC) between AA, AOC, RHOs and BHO to
discuss communication difficulties and an information centre was set up
in the AEC after that meeting. After that, information on stand
allocation was passed from ACC by phone, fax and TMR to AEC.
Airline staff, RHOs and baggage handling staff had to go to the
whiteboard at the AEC to look for the stand allocation and relay the
information by telephone to their colleagues. With this arrangement in
place but only at 7 pm, it was possible for operators to know the stand
and times of arriving aircraft, but it increased the ground handling time
of RHOs and was a drain on their resources.

12.39 [BHS 12] RHOs did not use both feedlines of carousels.
An allegation was made against RHOs that they did not maximise the
use of the feedlines of carousels as each arrival carousel could be fed by
two conveyors, which increased despatch times. This led to the
slowing down of the baggage handling process.

12.40 [BHS 13] RHOs did not know the assigned lateral for
arrival bags. Reclaim laterals are usually assigned by SESL according
to a pre-arranged allocation, which is distributed to RHOs and BHO on a
template the preceding night. This is also the practice today. However,
on AOD the template could not be relied on because of changes brought
about by flight delays and changes to flight schedules. SESL could not
properly rely on the schedule for actual allocation of laterals on AOD,
because flight delays made it necessary to estimate the time after the
actual time of arrival (ATA) of a flight when its passengers would
arrive at the reclaim carousels. Display parameters were extended to
leave displays on for a longer time to wait for arriving passengers who
might have been delayed in disembarkation.

265
12.41 From the point of view of RHOs, the pre-arranged lateral
allocation schedules for arrival baggage were rendered useless at around
8 am on AOD. This was when SESL reallocated laterals in order to
optimise their use. This was an attempt to reduce delays in despatching
arrival baggage by reallocating laterals on a real time basis. The new
lateral allocations were displayed for passengers inBRH. Unfortunately,
RHOs did not receive the information, as the FIDS LCD boards in the
Baggage Hall were not working. Thus baggage did not arrive at the
announced reclaim belt for passengers. Furthermore, given the limited
means of communication, it took longer for RHOs to receive the
necessary information. In some cases, RHOs also had to run around to
find out the location of the allocated lateral. This increased handling
time for arrival bags.

12.42 One incident of bags not arriving at the announced reclaim
lateral was reported to AOCC at around 8 am on AOD. The problem
was apparently resolved when SESL was told by AA to revert to the
original fixed schedule and stop real time reallocations at about 8 am.

12.43 [BHS 14] RHOs abandoned unit load devices (ULDs)
around arrival baggage feedlines, causing congestion and confusion in
the Baggage Hall. Congestion and some confusion resulted in the
Baggage Hall because RHOs left both full and empty ULDs around
arrival baggage conveyors.

12.44 [BHS 15] FIDS workstation in BCR performed slowly and
hung frequently. It had been suggested that there were serious
response problems with FIDS allocating reclaim laterals on AOD. In
AAs FIDS log, the FIDS workstation in BCR was recorded to have
hung up at 10 am and frequently at other times or took a long time to
execute functions. According to W26 Mrs Vivian CHEUNG Kar Fay,
Terminal System Manager of AA, it took 20 to 25 minutes to make one
reclaim belt allocation at times on Days One and Two. Therefore on
AOD there was either no or delayed displays of reclaim belts to RHOs
and to passengers. W35 Mr Gordon J ames Cumming, Sub-contract
Manager of Electronic Data Systems Limited (EDS), the contractor for
FIDS, and W26 Cheung gave evidence that at about 10 am the
performance of the FIDS workstation in BCR was so slow that AA/EDS
266
decided to reconfigure the parameters. At about the same time, the
whole operation of reclaim belt allocation was taken over from SESL by
AOCC. The workstations in AOCC had the same response problems,
but operators in AOCC could switch from one workstation to another
when the first one hung and was being rebooted.

12.45 [BHS 16] There was no reliable flight information displayed
on the LCD in the BRH. LCD boards in the reclaim area did not display
correct or any information as to where passengers should pick up their
bags. This problem could have been caused by missing components in
the board, cabling problems, slow performance of FIDS inhibiting data
entry or failure of BHO to allocate flights to reclaims or to progress
flights correctly. The result was that displays were cleared off too
quickly or the provision of the relevant information was delayed. To
fill in or supplement missing information, AA put whiteboards with
necessary information written on them at BRH on Level 5 early in the
morning on AOD.

(d) Stretching of RHOs resources

12.46 [BHS17] While RHOs had considered problem scenarios in
their preparation for AOD, for instance, losing sorters, losing power and
being faced with a large number of problem bags, none of them were
prepared for such a large number of problems bags that arose on AOD.

12.47 Additional drain on RHOs resources was caused by the lack
of essential flight information via FDDS or other means of
communications. On AOD, because of the problem with lack of
accurate flight information, runners had to go between AOCCs
whiteboard and staff on passenger and cargo ramps to pass on
information that should have been available from FDDS. The lack of
flight related information from FIDS and FDDS was exacerbated by the
failure of TMR and mobile phones due to overloading of the systems
which was, in turn, caused by sharp increase in usage because of FIDS
failure. This caused difficulties to RHOs as to where to send staff to
pick up or to load baggage.

12.48 There was allegation against RHOs that on AOD they had
267
inadequate manpower deployed at the problem baggage area to remove
the large number of bags going there. The build up of problem bags
meant that RHOs manpower was severely stretched with manually
sorting these problem bags. A consequence of the stretched resources
of RHOs was that the Remote or Hot Transfer System, although
available, was not used to handle transfer baggage. As a result, all
transfer baggage was handled only by the Central Transfer System in the
Baggage Hall, which slowed down operations. [BHS 18] For instance,
J ATS decided not to use the Hot Transfer System in order to conserve its
baggage handling resources for the main baggage handling area to assist
with the clearing of problem bags.

12.49 To cope with the worse than anticipated situation on AOD
and especially the number of problem bags, HAS engaged additional
staff in the days following AOD to clear the backlog. For Odgen, 60
trained staff from associated companies in overseas operations were sent
to the new airport to support operations in the first month of AOD.
J ATS also deployed extra staff and continued working extended hours to
process the backlog. By about Day Three, the situation had improved
significantly and baggage operation began to normalise.

(e) Inexperience or unfamiliarity of airline, RHO and SESL Staff

12.50 [BHS 19] Many of the actions of airline, RHO and SESL
staff demonstrate their inexperience or unfamiliarity with operations in a
new environment and with a larger scale of operation, for example the
airlines incorrect method of introducing unconveyable bags into the
system. Staff of airlines and Securair showed an inability to deal
correctly with new situations such as when they sent Kai Tak bags with
no labels into BHS. Inexperience and unfamiliarity may also have
caused operator and staff to be overwhelmed by the delays and confusion
caused by a lack of flight information vital to their operations, as when a
RHO put arrival bags into the transfer system, although the arrival and
transfer belts were some 25 metres apart.

12.51 On the system itself, HAS argued that on AOD, only one of
the four transfer belts was working properly. This created difficulties
for HAS in handling transfer baggage. It is not clear from the evidence
268
what caused the intermittent stoppage of the three laterals. It might
have caused by the system stoppages and die backs referred to above.
The intermittent nature of the stoppages of the system indicated that the
system could function again quickly after a restart and did not require
remedial measures to be taken on the system itself. Other than HAS
allegation, it does not appear from the evidence that there was any design
fault or error with BHS. AA also maintained that at no stage that BHS
ceased to operate otherwise than as designed. Other than some
parameter settings, no change was made to the functionality of BHS
since AOD. W55 Dr Ulrich Kipper, the Commissions expert,
commented that BHS worked as designed without error. On the
evidence, the Commissioners are of the view that the problems with
baggage handling were probably not related to the system itself, but
rather were the result of causes external to it.

Remedial Measures and the Present Status

12.52 During the days after AOD, there was improvement to the
performance of FIDS, and the direct and consequential problems it
created gradually subsided. On Day Two, the number of bags left over
was 6,000 out of a total of 24,000 bags processed. This was reduced to
2,000 (out of 26,000 bags), 1,400 (out of 27,000 bags) and 220 (out of
27,000 bags) on Day Three, Day Four and Day Five respectively.
RHOs were able to return to normal operation by about Day Three to
Day Four. RHOs, passenger handling entities and airlines had worked
with AA to put more logic into the assignment of gates to minimise the
amount of RHOs travelling time around ramps. Further, as staff and
operators became more experienced and familiar with the system and
operation, baggage handling at the new airport improved significantly.

12.53 AAs statistics showed that by Week 2 of AOD, the average
figures for first and last bag delivery times were similar to figures for
Kai Tak, and were improving. The latest statistics published by AA
show that during 1 December 1998 to 3 J anuary 1999, 90% of the flights,
the first and last bag delivery times were 19 minutes and 36 minutes
respectively, which far surpass the figures of 25 and 43 minutes for Kai
Tak. In the week commencing 31 August 1998, only 296 bags out of a
total of 228,000 departure and transfer bags processed missed their flight.
269
As at today, the baggage handling process can certainly be said to have
attained the world-class standard.


12.54 Chapter 8 sets out briefly the problems which the
Commissioners regard as moderate. This section of Chapter 12 deals
with each of these problems in detail, outlining their causes and remedial
measures. The responsibility for each of these moderate problems is
reviewed in Chapter 15.

[26] Delay in Flight Arrival and Departure

12.55 There were significant delays of incoming and outgoing
flights during the first week of operation of the new airport. The
following table sets out statistics of the delay during the first week of
AOD.

Incoming flights

AOD 7
July
8
July
9
July
10
July
11
July
12
July
Number of flights

213 227 220 240 220 230 235
Early arrival and on time

51%* 32% 34% 46% 47% 38% 50%
Delay within15 minutes

7% 20% 21% 23% 27% 26% 28%
Delay within 30 minutes

23% 34% 35% 36% 41% 44% 37%

36% 48% 53% 47% 49% 53% 45%
Delay more than 60 minutes

13% 20% 13% 7% 4% 9% 5%
Average Delay for Incoming
Flights (Hour)
0.4 hr 0.8 hr 0.6 hr 0.6 hr 0.4 hr 0.6 hr 0.4 hr

270
* Holding time of aircraft on the taxiway is not included

Outgoing flights

AOD 7
July
8
July
9
July
10
July
11
July
12
July
Number of flights

207 227 220 240 220 230 235

0% 7% 6% 15% 16% 22% 13%

3% 15% 25% 36% 41% 55% 47%

13% 38% 66% 75% 77% 86% 81%
Delay more than 60 minutes

87% 62% 34% 25% 23% 14% 19%
Average Delay for all
outgoing Flights (Hour)
2.63
hrs
1.7 hrs 0.9 hr 0.7 hr 0.8 hr 0.6 hr 0.7 hr

12.56 It is clear from the above tables that the greatest element of
flight delay in the week from AOD was in respect of departing flights,
with delays for more than 60 minutes ranging from 87% to 14%. For
incoming flights, delays for more than 60 minutes ranged from 20% to
4%. Since 13 J uly 1998, the average delays for both incoming and
outgoing flight were comparable to those for Kai Tak in J uly 1997. By
end of the first month of operation, the average delays for incoming and
outgoing flights at the new airport were less than the statistics on the
same subject for Kai Tak.

12.57 On AOD, delays became more serious after around 11 am
when traffic was very busy. W23 Mr Alan LAM Tai Chi, General
Manager (Airfield Operations) of AA said the morning on AOD between
7 am and 9 am was not particularly busy as there were only 11 arrival
flights. Traffic began to build up around 11 am and over lunch time.

271
12.58 The statistics for arriving flights on AOD show that on
average aircraft were achieving chocks-on within 24 minutes. The
chocks-on time will not equate with the time at which passengers were
able to disembark. The combined effect of the various causes for the
flight delays described below was that it took much longer than the usual
turnaround time for an aircraft arriving at and departing from the new
airport. This longer turnaround time comprised delays in disembarking
passengers and unloading baggage and cargo from arriving aircraft and
delays in embarking passengers and loading baggage and cargo onto
departing aircraft.

12.59 Delays were also caused to some of the passengers by
reason of the additional time taken to travel to and from aircraft parked
at remote stands at the new airport. Under current operations, an
average of 80% of arrival flights are handled at frontal stands leaving
about 20% of flights being serviced at remote stands. On AOD,
approximately 50 aircraft were serviced at remote bays representing 30%
to 40% of the flights.

12.60 The main cause of the problem was the inefficiency of FIDS
resulting in the lack of accurate flight information. The witnesses from
all three RHOs said that FIDS was the major problem which affected
their ability to render timely and efficient service to their customer
airlines. Other causes contributing to the flight delay as well as the
delay in the disembarkation of passengers included the baggage handling
chaos, the Access Control System (ACS) and Public Address System
(PA) malfunctioning, confusion over parking of planes, malfunctioning
of airbridges, late arrival of tarmac buses and communication problems
experienced by RHOs and the other operators at the new airport. The
latter is caused by the problems with TMR and mobile phones due to
overloading and poor quality of transmission. All the above problems
are discussed separately in the rest of this chapter and in Chapters 9 and
15.

12.61 Another contributory factor to the flight delays on AOD was
the disruption on the cargo apron. For instance, export cargoes were
delayed in being prepared for collection by the RHOs for loading onto
some aircraft and in some cases were not loaded onto the aircraft at all.
272
Furthermore, import cargo were not being processed quickly enough so
that the normal procedure for cargo handover was not able to be adhered
to. These had, to a certain extent, caused delays to incoming and
outgoing flights. The cargo handling chaos is described in more detail
in Chapter 11.


12.62 ACS is one of the most crucial systems for airport security
and its status could affect the aerodrome licence. ACS is a
computerised system that performs three functions, namely, (1)
production of permits to authorised holders to unlocked doors; (2)
verification of permits to identify personnel that are entitled to enter
certain restricted areas; and (3) monitoring movement of personnel
through ACS doors and when doors are opened and closed. ACS doors
will only open on the swiping of a valid swipe card with appropriate
access rights. Sometimes, inputting of a personal identification number
(PIN)is required. The ACS will then check the swipe card against
information that had previously been input and stored in the system.
There are some 5,000 doors in the new airport, of which ACS controls
about 1,505 doors. There are two types of permits or badges at the new
airport. Permanent permits are for employees in the new airport.
They can be used to activate the various doors in certain areas of the new
airport depending on the areas that the respective holders are permitted
to enter. Temporary permits are for visitors, contractors and other
persons on a non-permanent basis. They are merely encoded with the
information specifying the areas that the holders are allowed to enter.

12.63 The ACS stores all card holder information in a computer
centrally. The information is also downloaded to various local
Distributed Access Controllers ("DACs"). Each DAC controls around
12 doors. These DACs operate to prevent a person from gaining access
to an area which he is not permitted to enter.

12.64 The contractor for ACS was Guardforce Limited
("Guardforce). Its scope of work included the design, supply, testing
273
and commissioning of system software, the network sub-system and
head end equipment, the supply and commissioning of the permit
production equipment installed in the Maintenance Headquarters and the
installation works on site, and testing and commissioning of door control
and interface units. Controlled Electronic Management Systems
Limited ("CEM") was the nominated subcontractor of Guardforce for
ACS. The doors, electro magnetic locking and detection devices were
provided by another contractor of AA, the British-Chinese-J apanese
J oint Venture ("BCJ "). The receipt of permit applications, coordination
of application for security vetting purposes, data entry into the ACS
database, applicants photo taking, permit production and issuance were
carried out by AVSECO.

12.65 Due to the delay in the progress of work which will be
described below, ACS had not been completed on AOD although AA
claimed that it was operational. For instance, there had been substantial
slippage to the site acceptance test (SAT) which was supposed to be
carried out in around December 1997. The SAT was only about 60%
complete as at 30 November 1998 and was expected to be finished in
December 1998. Since AOD, various problems relating to ACS had
been reported. There were problems in securing the timely production
of security permits. ACS doors including airbridge doors were not
working resulting in the deactivation of all the airbridge doors for
departing flights from 7 J uly 1998 until 19 J uly 1998 and security guards
were posted to maintain security. There were also allegations of
security risk by reason of the inoperative or incomplete state of ACS.

12.66 The allegations of security risk are dealt with under
paragraph [28] Airside Security Risks below. Another matter about
security risk was mentioned by Guardforce. This was in relation to the
alleged lack of intrusion detection on the North Shore Airfield of the new
airport, which is to the north of the second runway site. According to
Guardforce, the installation work that it had completed for the North
Shore Airfield, which included ducting and foundation works, had been
destroyed by another contractor working in the area. Guardforce
suggested therefore that there was a problem with the intrusion detection
274
on the North Shore Airfield, in that AA would not know if there were
people trespassing on that side of the new airport. AA denied this
allegation and stressed that there was no security risk. According to
AA, the work being undertaken by Guardforce related to a construction
site outside the present operational boundary of the new airport. AA
also claimed that there is a separate fence between the construction phase
of the works and the operational part of the new airport and the fence
was patrolled by AVSECO security staff under the enhanced security
arrangements in place since 13 J une 1998 and with knowledge and
approval of Civil Aviation Department (CAD).

12.67 Despite the allegation, there was no reported incident of
trespassing on the North Shore Airfield. In view of the remedial action
taken by AA and the evidence before the Commission, it appears to the
Commissioners that there is no problem with security on that part of the
new airport.

(a) Delay in permit production

12.68 There were some problems with producing security permits
in a timely manner. This problem was raised in relation to other issues
such as the no tap water in toilet rooms problem under item [15] in
Chapter 9. Lack of security permits would disrupt staff and workers in
carrying out their work. For instance, BCJ complained that no swipe
cards had been issued to it to allow independent access around PTB,
which inhibited their work.

12.69 AA attributed some delay in permit production to the
applicants. According to a survey of AA's business partners, it was
anticipated that about 2,500 temporary permits and 24,000 permanent
permits would be needed (26,500 permits in total). By the end of J une
1998, some 14,000 and 25,000 applications for temporary permits and
permanent permits respectively had been received (totalling 39,000
permits). The applications were also late. AA introduced a day pass
system to relieve the pressure on production of permits. This however
275
created further work for the Permit Office as new application for day
pass had to be made every day.

(b) Problems with the airbridge doors

12.70 On AOD, 11 out of 38 airbridge doors were not working.
AAs evidence was that on the night before AOD, it was discovered that
a number of doors would not open with the swipe card. The
malfunctioned doors were disabled, and where a quick repair was not
possible, guards were deployed to guard these areas. According to W44
Mr Chern Heed, Airport Management Director of AA, there were at least
two incidents on AOD where passengers were unable to exit the
airbridges into PTB. One of the incidents occurred when about 200
arriving passengers from flight CX 722 were trapped in the airbridge at
stand W46 and were unable to gain access into PTB because of the
failure of the ACS door. To gain access, it appeared that these
passengers had the airbridge door released by breaking the emergency
break glass.

12.71 AA decided to disconnect all airbridge doors for departing
flights from ACS on 7 J uly 1998. ACS was disconnected during the
period 7 J uly 1998 to 19 J uly 1998. Security guards were then posted.

12.72 There were five other incidents after AOD where passengers
were unable to exit the airbridges into PTB. There was also an incident
involving the transit passengers of China Airlines Limited (CAL)
getting to the Departures Hall (through an ACS door which was
deactivated at the time) without security check. As a result, the China
Airlines aircraft had to be recalled after take off for security check.
This incident is discussed under item [28] Airside Security Risks below.
AA alleged that two out of the five incidents after AOD were apparently
due to airline staff not arriving at the airbridge in time to open the
airbridge door with a swipe card. Guardforce on the other hand denied
knowledge of any of the incidents of trapped passengers. It argued that
had they been related to ACS, AA should have informed Guardforce
about them. The five incidents are set out as follows:
276

(1) 14:45, Saturday 11 July 1998 CX501 at Boarding Gate No.27. The
time of rectification was not recorded. Passengers were released
subsequently through Departures Level doors.

(2) 09:08, Saturday 19 July 1998 CX460 at Boarding Gate No.3.
Ground staff arrived within eight to ten minutes and opened the
doors at Level 5 immediately. AA claimed that this incident was
due to the late arrival of airline ground staff and not the failure of
ACS.

(3) 20:29, Monday 20 July 1998 CX507 at Boarding Gate No.3.
When the passengers disembarked, the airbridge door was locked
and no airline ground staff were present. AA, Guardforce and
AVSECO staff however were present, testing the swipe card reader.
AVSECO staff therefore activated the door and allowed passengers
to leave the airbridge. AA alleged that this incident was not
caused by any failure of ACS.

(4) 05:42, Wednesday 22 July 1998 CX829 at Boarding Gate No.2.
The AVSECO guards could not open the arrival door with the
swipe card when the passengers arrived. At approximately 05:45,
one of the passengers broke the emergency break glass and
released the door. The door was then temporarily removed by
maintenance staff at 09:45 before being repaired.

(5) 12:00, Thursday 23 July 1998 CX710 at Boarding Gate No.4.
Passengers were unable to exit through the airbridge on their
arrival. The guard on the airbridge immediately opened the
arrival door with a test card after obtaining instructions from the
AVSECO manager.

277
12.73 AA claimed that the above incidents were all of a short
duration with minimal delay and inconvenience to the passengers.

12.74 New Airport Projects Co-ordination Office (NAPCO)
claimed that the failure of ACS precluded access of departure gates on
AOD, necessitating numerous gate changes by AA. AA disagreed and
alleged that wherever there were problems with airbridge doors, staff
would have been sent to open them. W44 Heed also maintained that no
gate change was necessary because of the ACS failure.

(c) ACS doors and other system problems

12.75 Whilst problems with the airbridge doors affected
passengers, other ACS door problems would have an impact on airline
staff and other people working at the airport. The fault report for ACS
showed that there were 440 reported faults with the system between
AOD and 31 August 1998. This figure presumably included both faults
with the airbridge doors and other ACS doors. The AA's Help Desk
summary also provided some figures of reported incidents, which
showed that there were 178 incidents in four weeks from the airport
opening (44 in the first week of AOD, 29 in the second week, 48 in the
third week and 57 in the fourth week). W44 Heed thought that these
incidents were likely to be related to tenants or staff working at the new
airport rather than to passengers. W37 Mr Dominic Alexander Chartres
Purvis, Manager (Customer Services) of Cathay Pacific pointed out that
access was a problem for their staff in PTB in the first month after
opening due to failure of swipe cards.

12.76 According to AA, the major outstanding problems with ACS
as at AOD consisted of the locking up of workstation for permit
production, the DACs sometimes did not receive all records when card
data was loaded into them in bulk, server concentrator failure and
possible communications problems with airfield DACs.

(d) Causes of the ACS problems
278

12.77 AA attributed the delay to the completion of the ACS before
AOD to Guardforce and CEM. According to AA, factory acceptance
tests (FATs), which were supposed to be carried out between J une and
September 1996, did not take place until J uly 1997. In October 1997,
W43 Mr Douglas Edwin Oakervee, Project Director of AA and Mr
Alastair Blois-Brooke, Senior Construction Engineer of AA and a
Guardforce's representative visited CEM in Northern Ireland in order to
expedite CEM's work. There were delays in the delivery of the
software by CEM which arrived on 18 December 1997 instead of early
December 1997. Model tests for the system, which were supposed to
be completed by September 1996, were only commenced in February
1998 and had not been completed as at 30 November 1998 when W47
Mr Graham Morton, Project General Manager of Guardforce gave
evidence. Guardforce argued that the delay to the model tests was
caused by the unavailability of the General Building Management
System and the Building Systems Integration (BSI) System by AA.
SATs were supposed to be done in about December 1997 to J anuary
1998. The progress of SATs continued to slip from March to J une 1998.
Although some software for ACS was delivered in December 1997 and
on 9 May 1998, the final release of the software was not delivered to site
until 2 J uly 1998. W47 Morton confirmed that as at AOD, the
commissioning of ACS, SATs and reliability tests (including the testing
on the head end computers) had not been completed prior to AOD.
SATs on ACS were re-started after AOD in late August or early
September 1998. W47 Morton told the Commission that as at 30
November 1998, the SAT was about 60% complete. According to AA,
SAT was anticipated to be finished in December 1998. Confidence
trials will be carried out between J anuary and March 1999.

12.78 It was obvious that ACS was not ready as at AOD, at least in
the sense that the system as a whole was not tested, although W25 TSUI
King Cheong, Project Manager Electrical & Mechanical Works of AA
testified that the system might have been functioning. W44 Heed also
acknowledged that ACS was handed over to Airport Management
Division (AMD) before it had been accepted by the Project Division of
AA (PD).
279

12.79 AA alleged that the major outstanding problems on AOD
were primarily caused by physical damage to the ACS doors, software
problems with ACS and generally by a lack of resources by Guardforce
or CEM to complete the works on time.

12.80 Guardforce accepted that there were problems with the ACS
software and systems. There was queuing problem with the head end
computers. Where there was too great a backlog of data, the system
could not handle the backlog. There was also a stability problem with
the head ends. There was another problem with the server
concentrators, which took the DACs offline for one or two minutes or an
hour.

12.81 From AAs evidence, 95% of the ACS doors had been
installed by mid-J une, with locks powered and tested. Over the
following three weeks there was a marked increase in the number of
people using PTB. Security arrangements were being tightened up and
an increasing number of doors were locked and connected to ACS. A
number of doors were subsequently forced open for shortcuts, causing
difficulties in testing. According to Guardforce, there were around 900
doors in respect of which ACS equipment including card readers and
break glass units had been damaged by third parties. AA and BCJ had
taken various steps to prevent the vandalism including the employment
of additional security staff to patrol PTB, and the issuing of warning
letters to apprehend the culprits. According to W47 Morton, tenants of
PTB were made aware of the heavy penalties imposed on those who
broke the rules, including the prohibition against using unauthorised
doors for access. AA also implemented the Interim Security Measures
limiting access to and egress from PTB to specific control points with
security guards on duty. W44 Oakervee also said that with 1,505 doors
that they tried to secure and with about 7,500 to 8,000 workers at the
new airport at the time, it was virtually impossible to catch the
wrongdoers despite measures being taken.

280
12.82 According to W47 Morton, since J une 1998, there were
about 8,000 to 12,000 alarms a day. This caused problems in the
communication between the DACs and the head end computers. It
created difficulties in getting the field end computer stable enough to
allow correct testing on the head end. Software problems were
therefore not discovered.

(e) Other remedial measures

12.83 Some alarms were set off due to operational errors. It was
said that operators selected "staff" rather than "passenger" mode so that
the door alarm sounded when the door was held open for too long. To
prevent this, AA gave a series of briefing to airline staff from 17 J uly
1998 to 20 J uly 1998.

12.84 All airbridge doors were tested on a daily basis by AVSECO
as from 26 J uly 1998. A week's confidence trial was carried out during
the first week of August 1998. The system was activated section by
section from 21 J uly 1998 to 27 J uly 1998. No fixed guards are now
assigned to particular airbridges, although various such area are still
patrolled by guards. The physical works on the doors, about 150 in
total, were awaiting repairs as at 27 November 1998.

12.85 Turning to the outstanding problems on AOD mentioned by
AA, Guardforce had successfully downloaded more than 35,000 data of
permit holders to every DAC by 15 J uly 1998. Nevertheless, the
problem was not resolved until the end of September 1998. The rest of
the other outstanding problems were rectified by September or October
1998. The head end computers became stable around mid-September
or the beginning of October 1998. There were some problems with
them as at 30 November 1998, but most software issues had been
resolved. The Tuxedo 6.4 version was loaded on 31 October 1998 to
resolve the queuing problem with the head end computers.

281
12.86 On the issue of permits, AVSECO had made arrangement to
run the Permit Office for 24 hours. Additional printers were installed to
speed up permit production. One-day escorted permits were introduced
by AVSECO to relieve the backlog of permit application. AVSECO
claimed that it was able to process the one-day permit and the three-day
temporary permit within 15 minutes on the day of applications. On
AOD, the Permit Office had issued 1,053 escorted permits to enable
contractors to carry out urgent repair works within the restricted area.

[28] Airside Security Risks

12.87 Airside security is of utmost importance in the overall
context of airport security. Failure to ensure airside security would
jeopardise the safety of passengers and aircraft. This explains why the
Commissioners classify the following four incidents posing airside
security risks as moderate.

(a) Delayed entry of police motorcycles into restricted area

12.88 On 10 J uly 1998, a minor traffic accident inside PTB
Baggage Hall resulted in two workers sustaining slight injuries. Two
ambulance service vehicles were allowed immediate entry to the
Enhanced Security Restricted Area (ESRA) to attend to the injured.
However there was a delay in allowing traffic police on motorcycles
entry into the ESRA. Normally, permits are required for entry into the
airside restricted area. However, section 22 of the Aviation Security
Regulations provides for exemption from these requirements where
disciplined and emergency service vehicles and personnel are responding
to an emergency. AVSECO indicated that it has established procedures
to deal with disciplined or emergency service personnel and vehicles
responding to an emergency. In the event that siren and flashing lights
of such vehicles are activated, the vehicles would be allowed immediate
entry into ESRA. The two ambulance service vehicles which were
given immediate access to ESRA had activated blue lights and sirens
while the police motorcycles had not.

12.89 AVSECO stated that since this incident, and also from the
experience gained since AOD, procedures for dealing with disciplined or
282
emergency service personnel and vehicles responding to an emergency
were fine tuned in conjunction with the Police and other emergency
services operating in the airport. The revised procedures are said to
have worked well so far.

(b) Transit passengers allowed to enter Departures Hall and board
flight without security check

12.90 Upon arrival of CAL flight CI 651 on 25 J uly 1998 at
boarding gate 23, the ground staff of CAL took some 90 transit
passengers direct from the aircraft to the Departures Hall on Level 6 of
PTB without proceeding through Level 5 for security screening at the
designated Transfer Points. The transit passengers boarded the flight
and the aircraft took off but was subsequently recalled by CAL. All
these 90 transit passengers were re-screened before departing Hong
Kong.

12.91 At the material time of the incident, ACS at boarding gate
23 did not function. Had ACS at the gate been operative, there would
have been an effective barrier which prohibited the entry from the
airbridge to Level 6. A security guard of AVSECO was stationed there
to stop arriving passengers from going to the Departures Hall but did not
do so.

12.92 Upon being notified by AOCC, the AVSECO Duty Manager
responded to the scene, but after the CAL staff and the passengers had
left. The AVSECO Duty Manager then contacted the CAL Duty
Manager and asked him to undertake security screening for the
passengers or else CAL had to accept the responsibility for the flight to
proceed with the unscreened transit passengers on board.

12.93 When the CAL Duty Manager decided to conduct screening
on the passengers, the flight had already departed. The AOCC
therefore requested the Air Traffic Control (ATC) to call back the
aircraft. This was not acceded to as such a request should come from
the airline concerned except in an emergency in which case it would
come from designated police officers. CAL then contacted the aircraft
via their company frequency. The pilot advised ATC of the decision to
283
return to Hong Kong International Airport.

12.94 Subsequent to the incident, CAD conducted an investigation
and found that CAL had breached airport security procedure. CAL
apologised for the breach, and undertook to take steps to ensure that
there was no repeat of the incident. AA sent a circular to all airport
organisations reminding them of the need to adhere strictly to airport
security procedures.

12.95 Separately, CAD has written to AA to offer a number of
suggestions to improve security arrangements to prevent recurrence of
the incident. Some of these suggestions have already been
implemented, for example, tensa barriers have been set up in airbridges
to demarcate more clearly the Arrivals and Departures channel within the
airbridges. Other suggestions include putting up clear directional signs
inside airbridges to direct passengers to appropriate ramps and warning
signs prohibiting transit/transfer passengers from going up to Level 6
ramps. Passenger handling is carried out at separate transfer desks
located within the body of PTB.

(c) Unauthorised access to Airport Restricted Area (ARA)

12.96 The Commissioner of Police established that between 6 J uly
and 17 October 1998, a total of 55 cases of breach of ARA were reported
to the Police. A large number of cases involved failing to possess a
permit, failing to carry a permit which has been issued, and using a
permit which belonged to another person. In the early days of
operation of the airport, many people appeared to be confused about the
permit requirements, and the conditions applying to the use of the permit.
Some appeared to have tried to circumvent the regulations by using
permits belonging to others with the intention of carrying out duties at
the airport. By far, the majority of those intercepted were persons
working, either permanently or temporarily, at the new airport. The
greatest number of incidents occurred within the first month after AOD,
but there was a marked decline in the number of incidents over
subsequent months.

12.97 In his witness statement, Mr Sidney FC CHAU, General
284
Manager of AVSECO, explained to the Commission that the majority of
the unauthorised entries were technical in nature devoid of any criminal
intent. These unauthorised access cases were attributed to one or more
of the following:-

(i) permit holders not being familiar with the new airport
environment and physical setting of the different operational
zones of the ARA at the early stages of airport operation.
The size of the new airport, which is seven times that of Kai
Tak, is also relevant;

(ii) sponsoring organisations not giving adequate instructions to
their staff on the conditions of issue of the permit;

(iii) AA not providing sufficient signage during the initial stage
of operation of the airport; and

(iv) less than effective control over unauthorised entry caused by
operational problems arising from deployment of security
guards whilst ACS was under test.

12.98 There was a marked decline in the number of incidents of
unauthorised entry over subsequent months as a result of :-

(i) permit holders getting more familiar with the geographical
layout and security arrangements at the new airport, and the
conditions of issue attached to the permits;

(ii) more signage and warning notices being provided by AA;

(iii) round-the-clock guarding and patrol services being
introduced and maintained by AVSECO to complement
ACS; and

(iv) measures to strengthen ACS.

(d) A KLM flight took off with baggage of two passengers who were
not on board
285

12.99 On 8 J uly 1998, a KLM flight KL888 departed for
Amsterdam with the checked baggage of two passengers on board but
without those passengers. Boarding of passengers commenced 35
minutes behind the flights estimated time of departure. Boarding gate
readers (BGRs) were used to scan through boarding passes (BPs).
It was discovered during the process that BGR was not working properly.
KLM staff hence switched to manual collection and checking of BP
stubs to verify the number of passengers boarded. The result indicated
that 10 passengers were missing which appeared unlikely to the boarding
staff as the flight was already 50 minutes behind schedule. The cabin
crew then conducted a passenger head count. It was reported that the
figure given by the cabin crew was equivalent to the final number of
passengers checked in (ie, 218) after two round of counts. At this time
the flight was nearly an hour behind schedule. On the understanding
that the head count matched with the number of passengers checked in, ,
and that the missing BP stubs could have been due to failure to remove
the stubs from the BP during the rush, the boarding staff formed the
opinion that all passengers were on board. Since there was no report of
any missing passenger, no request was made to the baggage handling
staff for the removal of baggage.

12.100 Two passengers showed up at the boarding gate when all the
doors had been closed and the aircraft was about to take off. It was not
until then that the boarding staff realised that the head count was
incorrect. The two passengers were arranged to depart via another
airline.

12.101 Subsequent to the incident, KLM has taken measures to
avoid a repeat. CAD has issued a letter to airlines to remind them of
the need to ensure compliance with the requirements of the Hong Kong
Aviation Security Programme (HKASP).

[29] Congestion of Vehicular Traffic and Passenger Traffic

12.102 On AOD, there were traffic congestion, congestion at lifts
from Level 3 (ground level) to the PTB and contra-flow movement
among passengers on the down ramp from Arrivals on Level 5 to Level
286
3.

12.103 Members of the public who wish to travel to the new airport
can take the following buses: Airbuses (A buses) which are aimed at
taking air passengers to and from PTB; external buses (E buses) which
are catered for carrying staff and workers to and from the main
employment centres on Chek Lap Kok (CLK) Island; shuttle buses (S
buses) which take passengers from the Tung Chung MTR station to PTB;
there are also overnight routes providing services to the passenger during
late evenings and early hours of the day. A buses stop at the departure
kerb at Level 8 where travellers can walk down to the Departures Hall on
Level 7. E buses and S buses stop at Cheong Tat Road on Level 3
(ground level) outside PTB.

12.104 During the first week of AOD, more than 60,000 curiosity
sightseers per day visited the new airport, many of them taking the E
buses and S buses. Traffic congestion occurred at the section of
Cheong Tat Road near PTB where passengers alighted and got on board
of the E and S buses. The large number of visitors increased the
frequency of buses travelling to and from the airport. It also took
longer for buses to drop off passengers. The situation was aggravated
by the suspension of one of the two bus stops (ie, 15a) at Cheong Tat
Road and the non-completion of pavement work. Vehicles needed to
queue up to pull in or out and this further slowed down the traffic flow at
Cheong Tat Road.

12.105 Passengers who were described as joy-riders did not find
much joy after alighting from the buses. From Cheong Tat Road,
passengers would reach Level 3 (the ground level) of PTB. Passengers
who wish to get into PTB can make use of the six passenger lifts and
escalators in carpark 2 and Level 3. However, none of these facilities
had been put into service on AOD. The only way passengers could get
to PTB was via the two staff lifts, the down ramp leading to the Arrivals
Hall on Level 5 and the two emergency staircases. While people
packed into the staff lifts, this led to lift congestion. Measures were
taken to divert the passengers without luggage and sightseers to use the
down ramp. As the down ramp was originally designed for the arriving
passengers leaving PTB, this led to the contra-flow movement among
287
passengers. Departing passengers who managed to go to Level 5 either
through the down ramp or emergency staircases were confused as to how
to get to Departures Hall from the Arrivals Hall inside PTB due to
insufficient signage because the signs were not designed for this purpose.

12.106 A meeting to remedy the problems was called by the
Transport Department on 8 J uly 1998 and attended by Citybus Limited,
Long Win Company Limited, New Lantau Bus Company (1973) Limited
and AA. The following measures were implemented from 11 J uly 98:
Reduced the number of buses via Cheong Tat Road by re-routing the
outbound routes to go via the Ground Transportation Centre (GTC)
bus terminus instead and by diverting some inbound routes to go via
departure kerb on Level 8. This ensured that there was no more than a
manageable flow of bus services into Level 3. Further, the S buses
and E buses were segregated to observe bus stops 15a and 16a
respectively. These measures were effective in resolving the traffic
congestion on Cheong Tat Road.

12.107 The bus layby at the section of Cheong Tat Road between
carparks 1 and 2 were extended and the outstanding pavement work was
completed. Bus Stop 15a was reinstated on 27 J uly 1998. Availability
of lifts from Level 3 had increased since 12 J uly 1998 until early August
1998.

12.108 AA had deployed additional staff for traffic and crowd
control purposes since AOD. Temporary signs and barriers were
installed to direct arriving passengers.

12.109 With these remedial measures and the number of curiosity
visitors steadily decreasing, the traffic and passenger congestion problem
were resolved.

[30] Insufficient Air-conditioning in PTB

12.110 The Commissioners will deal with the issue in respect of (i)
PTB, and (ii) the tenant areas. The air conditioning system in the PTB
mainly consists of the following:

288
(1) seawater supply for cooling provided by a pump system ;
(2) chilled water supply (cooled by seawater) provided
throughout PTB; and
(3) air-conditioning plant (cooled by chilled water) which
provides cool air to the public areas.

12.111 The chilled water supply is provided by six chillers, which
rely on seawater supplied from six seawater pumps. Only five chillers
had been commissioned as at AOD. The chillers will shutdown or
trip if there is insufficient seawater flow. Generally, three to four
chillers are enough to supply air conditioning to PTB at the design
temperature of 24 degrees Celsius. The Commissioners note that the
setting at 24 degrees Celsius might be a bit high for the general puller in
Hong Kong, especially during summer time.

12.112 The insufficiency of air-conditioning was due to the shutting
down of the chillers as described below. However, some of the
complaints by the public might have been due to the rather high setting
of the design temperature.

12.113 Youngs Engineering Company Limited (Youngs) is the
contractor for seawater pumps whereas AEH J oint Venture (AEH) is
the contractor for the chillers and the air conditioning plant.

12.114 Since AOD, AA has reported 12 occasions of chillers
shutdown, causing disruption to the supply of air-conditioning to the
PTB. Each of the incidents is described below.

12.115 (1) 6 July 1998. On 6 J uly 1998, one to three chillers shut
down during various periods for approximately four hours causing the
temperature in PTB to rise by about 2 to 3 degrees Celsius. Three
chillers (Nos. 3, 4 and 5) were operating, supported by two seawater
pumps (Nos. 4 and 6) running at high speed. The following events
occurred and is depicted in a diagram prepared by W54 Professor Xiren
CAO, one of the Commissions experts at Appendix VIII to this report.

Time Event

289
10:15 am Chiller No. 5 tripped due to a faulty low pressure
switch protection device installed by Carrier. As
a result, the control system, which Youngs
controlled, switched from two high speed seawater
pumps to two low speed seawater pumps (Nos. 1
and 2) because of the reduction in demand of the
seawater.

10:24 am Pump No. 2 tripped due to a flow switch fault,
leaving only one low speed pump running (Pump
No. 1). As a result, Chiller No. 4 tripped due to
insufficient seawater flow. This left only Chiller
No. 3 running, supported only by Pump No. 1.

11:01 am After unsuccessful attempts to contact the pump
house operated by Youngs through mobile phone
or land line telephone, AEH tried to restart Chiller
No. 4.

Upon AEHs attempt to restart Chiller No. 4, the
control system attempted but failed to restart Pump
No. 2, which had a flow switch fault. Essentially,
the bypass valves had opened causing reduced
seawater flow to Chiller No. 3, which had to be
shut down manually.

11:15 am Youngs reset the time delay for Pump No. 2

11:24 am AEH restarted Chiller No. 4 again. The control
system restarted Pump No. 2, and it tripped again
after the preset time delay, indicating a fault in the
flow switch.

Youngs therefore set Pump No. 2 into manual
mode. The control system started another low
speed pump (No. 3) successfully. At this point,
the system without Pump No. 2 ran in auto mode.

290
1:10 pm Chiller No. 3 was started

1:47 pm Chiller No. 4 was started

2:36 pm Chiller No. 4 having been stablised, Chiller No. 1
was started making a total of three chillers which
were sufficient to support the air-conditioning
supply to PTB.

12.116 From the evidence provided to the Commission and
according to W54 Cao, the technical causes of the air conditioning
problem on 6 J uly 1998 were as follows:

(1) A low refrigerant pressure switch fault caused Chiller No. 5
to trip at 10:15 am.

(2) A flow switch fault caused Pump No. 2 to trip. The flow
switch on Pump No. 2 was removed and checked for
possible debris that might have caused the flow switch to
stick.

(3) Inadequate communication between the bypass controller
and pump house control systems which caused the bypass
valves working undesirably, leading to reduced seawater
flow to Chiller No. 3.

(4) A problem in the logic in the control system which tried to
restart a faulty pump. The system might have preferred to
start Pump No. 2 as it had the lowest run time. Changes
have since been made by Youngs to the logic that compares
image run times.

(5) The delay in the start-up of the chillers was caused by
communication difficulties between the chiller plant and the
seawater pump house. AEH had attempted but failed to
reach the pump house control room by mobile phone and by
land telephone line. The latter method failed due to fire
291
alarm tests being conducted on AOD.

12.117 (2) 10 July 1998. On 10 J uly 1998, three chillers were
operating, supported by three seawater pumps running at high speed.
At approximately 9:15 am, one seawater pump tripped causing one
chiller to shutdown due to insufficient seawater flow. The tripping of
the seawater pump was caused by human error and the chiller was
restarted within 30 minutes.

12.118 (3) 12 July 1998. On 12 J uly 1998, four chillers were
operating with four seawater pumps running at high speed. At
approximately 10 am, due to a sudden energisation of a main chilled
water branch, the pressure of the chilled water system was reduced,
causing two chillers to shutdown. Air-conditioning service was
affected due to the shutting down of two of the four chillers. All four
chillers were back in operation within approximately two and a half
hours of the incident. After the incident, AEH was asked to ensure that
all energisation of chilled water pipe work was to be carried out by
opening valves slowly to minimise system pressure fluctuations.

12.119 (4) 13 July 1998. On 13 J uly 1998, four chillers were
operating and four seawater pumps were running at high speed. At
approximately 00:35am, all chillers and secondary chilled water pumps
were shutdown due to voltage fluctuations affecting electrical supply to
the chiller controls. These voltage fluctuations were allegedly caused
by lightning strike. Youngs alleged that although the lightning strike
took place at 10:30 pm on 12 J uly 1998, its engineers did not receive a
call until 4:30 am on 13 J uly 1998 about the interruption.

12.120 AA alleged that although the chillers had tripped and the
demand for seawater ceased, the seawater pumps continued to operate.
This was due to a software programme error with the control logic of the
seawater pumps within the seawater pump house. Youngs rectified the
programming error by approximately 5:30 am on the same day.
Seawater supply from the seawater pump house was not restored until
approximately 6 am and chiller re-starting commenced at around 6:30
am. Three chillers were back in operation by approximately 7:30 am
and the fourth by 9:15 am.
292

12.121 To avoid future fluctuations or losses in the power supply,
uninterruptable power supply (UPS) units had been installed
between 28 September 1998 and 27 October 1998 to the chiller control
panels and the panel serving the seawater controls in the chiller plant
rooms.

12.122 (5) 28 August 1998. At approximately 3:30 pm on 28
August 1998, all chillers tripped due to lightning strike affecting power
supply to the seawater pump house. The pumps were subsequently
re-started and the first chiller resumed operation within 45 minutes.
The remaining chillers resumed within two and a half hours.

12.123 (6) 29 August 1998. At approximately 1:40 pm
(approximately 12:20 pm according to AEH) on 29 August 1998, all
chillers tripped due to loss of seawater supply. Youngs alleged that
there was a loss in all external power supply to the pump house. AA
alleged that the electrical protection setting to the banscreen motors in
the seawater pump house had been incorrectly set and that Youngs
immediately altered the settings to rectify the problem. The first chiller
resumed within one hour 20 minutes and the remaining within three
hours 20 minutes.

12.124 (7) 30 August 1998. At approximately 10:30 pm on 30
August 1998, all chillers tripped due to lightning strike affecting power
supply to the seawater pump house. The pumps were re-started and the
first chiller resumed operation within 45 minutes. The remaining
chillers resumed within two hours 45 minutes. AEH alleged that the
last chiller did not start until approximately 8:45 am on 31 August
1998 ie, a disruption of approximately 10 hours. This was the third
time air-conditioning supply was affected by lightning strike. After the
UPS units had been installed between 28 September 1998 and 27
October 1998 to regulate power distortions from a number of possible
causes, including lightning strikes, there have been no further reported
incidents of interruption in air-conditioning supply due to lightning
strike.

12.125 (8) 8 September 1998. At approximately 2:29 pm on 8
293
September 1998, all chillers tripped (AEH alleged that only two chillers
tripped) due to a power failure caused by the tripping of circuit breakers
on Youngs switchboard. The power was immediately restored by
Youngs and all chillers resumed within one hour.

12.126 (9) 14 September 1998. At approximately 7:00 pm on 14
September 1998, all chillers tripped due to human error whilst the
contractor for the Mechanical Building Management System carried out
testing of that system. All chillers were resumed within four hours.
Signs have been posted on panels to warn staff not to turn off power
supply.

12.127 (10) 12 October 1998. At approximately 3:25 pm on 12
October 1998, air handling units and all chillers (only three chillers
according to AEH) tripped due to a disturbance by the power system of
China Light & Power Company Limited (CLP). CLP alleged that the
incident was caused by third party damage to their underground cable
which is a frequent cause of disruption to utility networks. All chillers
resumed in over two hours (AEH alleged that there was an interruption
of over four and a half hours).

12.128 (11) 22 October 1998. At approximately 1:10 am on 22
October 1998, there was a planned shutdown of the chillers in order to
test an interface with the seawater pump house. Control circuit
modifications were made to Youngs High Voltage motor control centre
(MCC) serving the PTB pumps in order to provide greater flexibility
in the control sequencing of pumps. All chillers resumed within one
hour (AEH alleged that there was an interruption of over two hours).

12.129 (12) 28 November 1998. At approximately 11:30 am on 28
November 1998, all chillers tripped due to a loss in seawater supply.
Youngs alleged that this was in turn caused by an unauthorised isolation
of power supply to the high voltage battery charger and associated UPS.
AA alleged that the UPS unit was incorrectly set to bypass mode which
prevented power backup. All chillers resumed within one hour 10
minutes (over two hours according to AEH). The UPS unit has been set
to standby mode to provide power backup in the event of re-occurrence.

294
12.130 The air conditioning system of the tenant areas functions
similarly to the one in PTB as described in para 12.110 with the
exception that each tenant supplied and fitted its own air conditioning
plant which was then connected to AAs chilled water system. The
tenants air conditioning plant provides cool air to its own individual
areas.

12.131 Before connection to AAs chilled water supply, the tenant
was required to have, amongst other things, completed the tenancy pipe
work installation and submitted to AA a request for chilled water
energistion. Upon receipt of the request, AEH was to energise the
chilled water within three days.

12.132 There were delays experienced in the energisation of
tenants chilled water supply, causing insufficient air conditioning being
supplied to the tenanted areas of the PTB. The delays were mainly
caused by the large quantity of late requests from tenants for connection
to AAs chilled water supply which were in turn caused by the failure to
complete or commission the tenants installation of its air conditioning
system by the tenants contractors. As a result, AEH was faced with a
huge volume of requests in the period immediately before AOD. Other
contributing factors included crossed pipe work in the North and South
Concourse, difficulties of gaining access outside tenant working hours
and restricted access to PTB necessitated by the opening of the airport.

12.133 With increased working hours and labour from both AA and
AEH staff, all requests for chilled water by tenants were processed by 13
J uly 1998.

[31] PA Malfunctioning

12.134 Public announcements at the new airport can be made
centrally or at local gates. When the calls are made centrally, the
person who makes the announcement can select the message to be
broadcast to all or selected areas. Requests for boarding, calls for
passengers and the like are announced locally at the gates. Gate
changes are announced centrally.

295
12.135 The public address system operating within PTB is the
General Coverage Public Address (GCPA) which will be referred to as
the Central PA. Apart from the Central PA, there are consoles just off
the loading bridges at each of the gates mainly controlled by airline
operators and AA staff. They are collectively referred to here as the
Local PA. Hepburn Systems Limited (Hepburn) is the main
contractor for the PA while SigNET (AC) Limited (SigNET) is its
sub-contractor.

12.136 Throughout PTB, there are 22 communication rooms
which service particular geographic areas of the building. The link
between AOCC (from where announcements are made) and the
communication room is via the BSI system and the Voice Routing
System (VRS). As the BSI and VRS systems were not available on
AOD and for some time thereafter, announcements were made through
the use of the manual all zone (MAZ) system.

12.137 The MAZ system operates through a notebook computer in
AOCC. Through the use of this system, the operator can identify the
geographical area of the airport in which an announcement is to be made.
The MAZ system is connected to one of the communication rooms and
information travels from one communication room to another.

12.138 On AOD, the Central PA was down twice for a total of an
hour as alleged by Hepburn, or 46 minutes as alleged by AA. On 7 J uly
1998, the Central PA was down for six times totaling three hours and 37
minutes. One particular downtime lasted two hours and five minutes.
On 8 J uly 1998, five occurrences of downtime were experienced, totaling
two hours and 46 minutes. The Central PA was also reported to be
unserviceable for a few minutes up to one hour five minutes on 10, 14
and 19 J uly and 16 August 1998.

12.139 There were many instances of failure of the Local PAs.
AA and Hepburn had different records of the number of problems.
According to AA, 26 gate consoles experienced problems in the first
week of AOD, 21 reports of PA problem in the second week, 25 in the
third week and 122 in the fourth week. These were significant numbers
of failure considering that there were only about 50 consoles. The new
296
airport still experienced two system failures a day for the Local PA as at
late November 1998. Initially, Hepburn asserted that only about four
consoles were not operational on AOD. Hepburn later agreed that a
total of 12 consoles were not working. According to Cathay Pacific,
announcements for three of its flights on 6 J uly 1998 and one of its
flights on 7 J uly 1998 were not made.

12.140 Problems affecting the PA were ongoing and variable in
nature, including both hardware and software problems. Hardware
problems include

(a) incomplete installation,
(b) human-induced damage to membranes to the consoles and
gooseneck microphones, and
(c) defective consoles due to failure of electrical components.

Software problem comprise

(d) intelligibility,
(e) zoning and priority problems,
(f) slow response time,
(g) overriding,
(h) system instability, and
(i) locking and latching to downright console outages.

(a) Incomplete installation of equipment

12.141 There were four areas in which items of equipment had not
been installed prior to AOD. These are (a) speakers at lift lobbies in the
central concourse on Level 6; (b) speakers at the east hall corridor; (c)
two of the baggage reclaim microphone consoles; and (d) some
equipment in areas within the GTC. Hepburn alleged that AA should
be responsible for the incomplete installation. Hepburn alleged that
items (a) and (b) were not installed because the false ceilings had not
been installed. Item (c) had not been installed because the necessary
increased conduit was not available as at AOD and it was not physically
possible to install the equipment without the conduits. As for item (d),
access to the GTC was not provided until the end of May 1998, some 18
297
months later than the original date, so Hepburn did not have sufficient
time for its installation. Nevertheless, this was not a widespread
problem and it was specific to certain areas.

12.142 The ambient noise-operated amplifier facility (which sets
the volume of PA announcement in accordance with the levels of
ambient noise) was not installed prior to AOD although AA alleged that
this did not prevent the operation of PA.

(b) Human-induced damage

12.143 Gooseneck microphones on a number of the consoles were
damaged as a result of the microphones being bent or knocked prior to
AOD, sometimes short circuiting the electronics below. Membranes
covering the consoles on many of the consoles had been scored,
apparently by coming into contact with pen or other sharp instruments.
However the damage alone did not render the consoles inoperational.
The precise causes of these damages are yet to be found. To rectify the
damage to the membranes to the consoles and gooseneck microphones,
AA and Hepburn went round all the consoles, double checked them and
assisted airline on the actual usage of the system. Such measures,
according to Hepburn, were successful.

(c) Failure of electrical components

12.144 Some of the failure of the consoles were caused by the
failure of particular electronic components within the consoles.
Hepburn alleged that the failure of particular electronic components
within the consoles was usual after they were subject to full use. These
consoles could not immediately be repaired as Hepburn had used up its
stock of the particular spare part controllers required for these consoles.

12.145 According to Hepburn, the majority of the hardware
problems were rectified within a week although it took slightly longer to
replace membranes on the consoles. Repair and replacement of the
defective hardware was largely completed by the end of September 1998.

298
(d) Intelligibility

12.146 On some occasions, users would notice that the
announcement from the local PA was unclear or no announcement could
be made by it. AA claimed that it was not aware of any report of lack
of intelligibility on or after AOD although some reports were received
during airport trials. AA however admitted that some echoing and
volume problems were experienced on AOD in some parts of PTB.

12.147 Final adjustment to feedback and volume could not be made
until the Rapid Assessment of Speech Transmission Index (RASTI)
testing was completed. The purpose of RASTI test was to assess the
rapid speech intelligibility index of the actual sound of PA. The RASTI
testing was dependent on the final configuration of PTB, including the
complete installation of all acoustic related materials, the final fit out of
tenant areas and shops, and any other material that might affect the
acoustic performance of PTB. Prior to AOD, AA agreed with the
contractor that such testing could only realistically be carried out after
AOD. SATs, including the RASTI testing, were completed at the end of
October 1998. Despite outstanding tests to be done, AA maintained
that PA was functional on AOD.

(e) Zoning and priority problems

12.148 NAPCO pointed out that the PA design is such that when an
announcement is made to a selected zone(s), the immediately adjacent
zones are inhibited from making announcements. This will prevent
conflict and lack of intelligibility due to overlapping announcements.
NAPCO alleged that this was not set up or functioning properly as at
AOD thereby causing zoning problems.

12.149 NAPCO also alleged that there were priority problems
within the structure of PA. There is a priority level for different inputs
into the PA. For instance a fire or bomb alert would be rated with a
high priority and instantly take control of all selected zones. Airlines
complained that announcements of a low priority from AOCC might
block out more important messages which airlines and AA staff tried to
announce at the gates.
299

12.150 The priority problem was overcome by resetting the system
to allow everyone to have equal priority.

(f) Slow response time

12.151 For the airline staff to operate the consoles it is necessary
for the staff member to input a 4-digit PIN. When the correct PIN is
entered, a light on the console will indicate that it is operational. On
AOD, Hepburn admitted that there was a relatively slow response time
for this logging on process for some gates, including four gates identified
on 11 J uly 1998. This created an impression on the airline staff that the
console was not working. The slow response time in the logging on
process was rectified by Hepburn in early September 1998.

(g) Overriding

12.152 On 13 J uly 1998, 29 local consoles were not serviceable.
Of these 29, 20 gates were not operational because the MAZ system
overrode the loading gate console. Hepburn corrected the problem
within 30 minutes and claimed that it related to an isolated incident
occurring on that date.

(h) Instability of the Central PA

12.153 On AOD, the MAZ notebook was not operational on two
occasions in the morning and in the afternoon for about 29 and 17
minutes respectively. Subsequently, the same problem occurred on
three or four further occasions after AOD rendering the PA inoperational.
According to Hepburn, the power supply contractor, AEH, was notified
of the problem. MAZ console outage was rectified within 15-30
minutes by resetting the MAZ notebook.

12.154 MAZ notebook outage, as alleged by Hepburn, was caused
by the noisy earth lines between the communication rooms and were
interfering with the data communication between the control room and
the MAZ notebook. W47 Morton of Hepburn agreed, however, that he
had no documentary evidence to support the claim of electrical
300
interference and that the possibility of a noisy earth link was more like a
deduction than an actual finding. Nevertheless, to remove the
possibility of the noisy earth link, technical staff of Hepburn added
isolators to PA at either end of the relevant cable to remove the earth link
and allow the system to float independent of the earth.

12.155 AA denied that the instability of the MAZ console was due
to a noisy earth line. AA claimed in its submission to the
Commission that the cause of the instability in the MAZ console was
unknown to it, and the problem disappeared about two weeks after AOD.

(i) Locking up or latching problem

12.156 Incidents of lock out problem caused by fire evacuation
warning announcement were reported by AA on 10, 24 and 25 J uly 1998.
When the fire alarm was triggered, it could not be turned off unless the
system was manually reset after locking up, thereby affecting other
announcements in the area concerned.

12.157 No actual fire report was recorded. The fire alarms were
probably triggered by people breaking glasses in order to go through the
ACS doors. The problem with the fire announcement was resolved by
a software patch delivered by Hepburn in early September 1998.

12.158 Apart from the hardware and software problems described
above, some problems with the PA were caused by human errors. On
one occasion, AAs engineer forgot to replug the MAZ console back into
its socket after testing, resulting in an inoperative console. Hepburn
admitted that on another occasion, it was Hepburns engineer who did
the same thing. According to Cathay Pacific, some announcements
requested were not made. That might have been because of system
failures, or it might, as pointed out by Cathay Pacific, have been because
either the phone lines were busy or AA staff simply did not make the
announcements.

12.159 The problems with PA are classified as moderate by the
Commissioners because of its terminal-wide use in the new airport and
that a great number of passengers in PTB would have been affected.
301
They also featured significantly in the inquiry because PA was touted as
one of the back-ups and workarounds for the dissemination of flight
information in the absence of an effective FIDS.

12.160 Although the local gate PA had not been a reliable system
since AOD, the overall impact of the deficiency of the Local PA should,
as submitted by AA, be seen in the light of the following matters agreed
by W51 Mr J ason G YUEN, expert for the Commission during his
cross-examination: (1) in Hong Kong it is not common airline practice to
board by rows, so the local consoles would not generally have been used
for that activity on and immediately after AOD; (2) getting passengers to
the gate is not a question of local PAs at all, but is a function of FIDS; (3)
the local gate problem in isolation is not a significant problem. On the
other hand, W51Yuen stated that the Local PA is essential in that it is
used to announce boarding readiness, flight delays, and to page
passengers. He also maintained that boarding instructions at gate
rooms are basic requirements for airport operations. From an airport
and airlines operational point of view, any problem with such a system
should not be allowed to continue for over a month.

12.161 Further, if the Local PA fails, the message can be broadcast
through the Central PA. The Central PA is only comparatively more
stable than the Local PA. If the Central PA fails, there is very little the
airport staff can do other than putting up whiteboards. If the message is
urgent, people have to be sent to spread the message. The
Commissioners note that the last contingency might create a strain on the
airport resources and it would affect the image of the new airport as a
world-class airport.

12.162 W44 Heed of AA disagreed that there would have been utter
chaos for the passengers during the periods of PA outage on AOD,
despite their occurrence at peak periods, because whether there would be
chaos depended on the number of changes required to be announced
through PA at that time.

12.163 According to AA, no passengers missed their flights as a
result of the problems with local announcements, as gate changes were
announced centrally from AOCC. No evidence has been received by
302
the Commission as to whether any passengers missed their flights as a
result of gate changes not being announced during the downtime of the
Central PA.

12.164 Progress on PA fixes was monitored initially on a
twice-daily basis and later on a daily basis in Task Force Meetings. AA
held daily meetings with the contractor to track progress. A meeting
was held 10 days after AOD to develop a programme for completion of
outstanding work. Hepburn provided staff 24 hours per day during the
first week of operation. Since AOD, Hepburn concentrated on
resolving system integration and reliability including software problems,
upgrading the Local PA at gate airline counters, level adjustments,
zoning issues and hardware problems. It claimed that the majority of
the problems were rectified within a week, although it took slightly
longer for the replacement of the membranes on the consoles.

12.165 AA stated that SAT for PA commenced on 4 May 1998 and
the agreement to defer the RASTI testing until after AOD would not
have interfered with the other tests of PA. Hepburn admitted that they
had problem with a Hong Kong sub-contractor, Univision Engineering
Limited, and Hepburn had to change the sub-contractor. This affected
the development of an interface software to the BSI, resulting in a delay
in FAT. FAT was not completed until around the end of J une 1998.
SATs, including RASTI, eventually continued on 1 September 1998 and
were completed at the end of October 1998. Those were apart from the
testing for the maintenance reporting terminal which Hepburn expected
to complete by the end of November. Hepburn also expected that
confidence trials for the Central and Local PAs would be completed by
about March 1999.

12.166 AA maintained that the MAZ console was not stable, until
about two weeks after AOD. Problems with PA still exist although the
system is more stable with fewer faults and failures.

[32] Insufficient Staff Canteens

12.167 Human resources play an important part, alongside with the
various systems, in the operation of an airport. To enable efficient
303
passenger services and air cargo operations, it is necessary to provide a
reasonably comfortable working environment for staff working in the
new airport including adequate and convenient catering facilities for staff.
This would be particularly relevant for the new airport at CLK given that
it is built on an island. The Commissioners have heard criticisms that
staff canteens in PTB were substantially under-provided during the initial
period of airport opening. The alleged problem of lack of sufficient
staff canteens in the new airport has inevitably caused inconvenience to
the staff working there.

12.168 J ATS made specific allegations about the lack of catering
facilities for staff and, on some occasions, staff had to wait for more than
40 minutes for a table and food. During the first two weeks, there was
only one staff canteen in operation in PTB with a seating capacity of 250.
Although there was an alternative for the staff to use the commercial
catering facilities in PTB, these facilities were often very crowded, and
were quite far away from their place of work, apart from being
expensive.

12.169 AA explained that there are at present four staff canteens in
PTB, two on the landside and two on the airside. Altogether, they can
accommodate a total of 954 people at any one time. A breakdown of
these facilities showing their seating capacity and dates of
commencement of business is as follows:

Staff Canteen

Location
Seating
Capacity
Date of
Opening
Sky Bird 1 Baggage Hall, Level 2,
Airside

250 14/6/98
Sky Bird 2 West Hall, Level 5, Airside

422 14/7/98
Sky Bird 3 Level 6, Landside

122 29/7/98
Sky Bird 5 Level 6, Landside 160 15/10/98
Total : 954

304
12.170 According to the Strategic Planning Department of AA, the
number of staff working at CLK is about 44,629 on any one day.
However, this is a figure showing all the staff working on the entire CLK
island. The estimate given by the Department in May 1996 suggests
that on any given day, there would be only about 14,600 people actually
working in PTB in 1998. As at the end of March 1998, there were
26,788 people working at Kai Tak on any given day. There was only one
staff canteen at the passenger terminal building at Kai Tak, capable of
catering for a total of around 560 people. Based on these figures, it
may appear that staff catering facilities at CLK compares favourably to
that at Kai Tak. Moreover, in addition to the four staff canteens, there
are other commercial catering outlets (both restaurants or kiosk style
outlets) within PTB providing special discounts for airport staff. All of
the landside commercial restaurants were contractually required to
provide a staff meal programme to airport staff as a supplement to the
staff canteen facilities. The total seating capacity of these restaurants is
about 2,800. There are also a number of additional staff canteens
provided by AA elsewhere in CLK and its various franchisees such as
Hong Kong Air Cargo Terminals Limited (HACTL), HAS, Hong Kong
Aircraft Engineering Company Limited (HAECO) for their own staff.
J udging from the above, there does not seem to be an obvious case for
insufficient staff canteens.

12.171 The Commissioners, however, note that there are two more
dimensions to the issue. First, it is not appropriate to compare the
situation at CLK with that in Kai Tak because there were a lot of catering
outlets in Kowloon City where the Kai Tak airport was situated. These
facilities were easily accessible to airport staff working at Kai Tak. The
new airport at CLK does not enjoy this advantage. During his oral
evidence before the Commission, W44 Heed agreed to this point.
Secondly, it is obvious from the table above that not all of the four staff
canteens were available to provide service on AOD. From AOD right
up to 13 J uly 1998, there was only one staff canteen in operation and two
others were not opened until later that month. The remaining canteen
only came into operation on 15 October 1998. The situation on AOD
and the few days immediately after was aggravated by the presence of an
overwhelming number of local visitors who came to the new airport for
sightseeing. In her witness statement, Ms Eva TSANG Wai Yi,
305
Manager Retail Planning & Development of AA, quoted about 60,000
casual visitors on a single day in the week following AOD. This was in
addition to the number of airport users. Their presence put pressure on
the commercial catering outlets in PTB and substantially restricted the
use of these facilities by airport staff. In fact, there were only nine
commercial catering outlets offering food to airport staff at discounts on
AOD although more outlets joined the discount programme at a later
stage.

12.172 PD of AA was responsible for making available the
mechanical and electrical services which were required for the operation
of canteens and the availability of these services would dictate the size,
capacity and location of the facilities. Both AAs AMD and
Commercial Division were also involved in the planning of staff catering
requirements. The requirements had also been discussed with the
Airline Operators Committee. In the course of evidence, W43
Oakervee revealed that the original conceptual design was to build a
main staff canteen within the maintenance building alongside PTB and
that would be the biggest one for staff. However, for various reasons,
including cost and profitability, the proposal was eventually rejected.
There appears to be a lack of consensus between PD and AMD as to
what were the reasonable planning requirements for staff catering
facilities.

12.173 In the light of the evidence before them, the Commissioners
are of the view that the problem of insufficient staff canteens existed on
AOD and some time thereafter. In fact, W3 Dr Henry Duane Townsend,
Chief Executive Officer of AA did not deny in his oral evidence that staff
canteen was under-provided. Although the situation has apparently
improved, the problem has not been completely resolved. When
W44 Heed gave evidence before the Commission on 27 November 1998,
he testified that there were still complaints of insufficient catering for
staff but confirmed that a small committee had been set up to look into
the subject with a view to improving the overall situation, especially in
the ramp handling areas.

[33] Radio Frequency Interference (RFI) on Air Traffic Control
Frequency
306

12.174 According to CAD, they have been receiving reports from
airline pilots regarding RFI on air-ground Very High Frequency (VHF)
radio communication channels used by air traffic control since late 1994.

12.175 To address the problem, CAD used spare frequencies to
replace the affected ones for communication in the event that there was
interference with the normal frequency channels. Since 1996, six
additional frequencies were used by air traffic control as extra backup to
further safeguard flight safety.

12.176 RFI signals were being monitored and measured by the
Office of Telecommunications Authority (OFTA) from hill tops in
Hong Kong and from aircraft operating in the vicinity of the Pearl River
Delta area as well as along the coast of Guangdong on a weekly basis.
Results of RFI monitoring were forwarded to the Mainland authorities
monthly to help trace the sources and eradicate the problem.
Investigation by OFTA indicated that the sources of RFI were in the form
of spurious or intermodulation signals originated from some unknown
paging stations along the coastal areas in the Guangdong Province. The
issue has been raised by the Government with relevant Mainland
authorities since December 1994.

12.177 The Mainland authorities have adopted a range of measures
to tackle the problem including dismantling radio transmitters on top of
hills, and closing down offending paging stations. Tighter control
measures on paging stations such as limiting their transmission power
and requiring them to install filters and isolators have also been
introduced in some cities.

12.178 Since May 1998, a Technical Working Group was
established with technical experts from Hong Kong and the Mainland
authorities to step up co-operation in addressing the RFI issue. In
addition, a Task Force has also been formed between operational
personnel of Hong Kong and the Mainland authorities for quick
exchange of RFI information, if necessary. CAD assured the
Commission that with the spare frequencies available, air traffic
operations at the New Airport had not been affected.
307

[34] Aircraft Parking Aid (APA) Malfunctioning: a Cathay Pacific
Aircraft was Damaged when Hitting a Passenger J etway during
Parking on 15 J uly 1998

12.179 At Kai Tak, aircraft parking was done with the assistance of
air marshall. InCLK, APA was introduced. APA is a laser scanning
device which directs the pilot to park the aircraft through a real time
display unit. The APA display unit will give directions to the pilot as to
positioning, steering instructions and when to stop. If the APA
malfunctions for some reason, there is a fail-safe mechanism advising
the pilot to stop. The APA display unit can be installed upon the faade
of the terminal building or on a gantry. In the new airport, there are 28
Building Mounted APAs, nine Gantry APAs installed within the terminal
area and 31 Gantry APAs installed at the remote stands. Safegate
International AB (Safegate) was the contractor for the design and
maintenance of the APA system. APA is operated by a qualified
marshall of AA who will be available to provide manual marshalling if
for any reason the APA cannot be used.

12.180 On AOD, three of the APAs were not functioning.
Safegate agreed that there were occasions prior to 15 J uly 1998 when the
Gantry APAs were unable to give the necessary directions to the pilot.
This, according to Safegate, was due to the height of the Gantry which
affected the laser scanning angle, the stop position and the type of
aircraft in question. Safegate alleged that if there was an error of APA,
the fail safe mechanism would advise of an error. The air marshall
would then give manual directions to the pilot to park the aircraft.

12.181 In reply to Safegates allegation that APA was affected by
the non-standard height of the gantry, AA argued that Safegate was
aware of the dimensions of the gantries since early 1997 and no issue
had been raised about this until after AOD when the problem appeared.

12.182 AA alleged that there was a problem with the detection
software as APA could not display the correct type of aircraft. AA also
said that there was a problem with APAs self calibration function which
failed to detect a sensor problem in the relevant APA. Safegate denied
308
both allegations. Safegate said its records revealed that the system
always correctly displayed the aircraft type specified by the operator.
Also, failure to detect a sensor problem was caused by human error of a
staff number who inadvertently left the auto-calibration function disabled
and this had nothing to do with the system.

12.183 All APAs were suspended from use from 15 J uly 1998.
AA said it was due to an incident on 15 J uly 1998 where a Cathay
Pacific aircraft was damaged during parking, allegedly related to the
malfunctioning of a Gantry APA at a frontal stand. This incident will be
described below. Safegate disagreed that APAs were not functioning
properly. Nevertheless, enhancement was made to the software of
Gantry APAs to increase the effective viewing angle of the laser.

12.184 Five out of 28 Building Mounted APAs had experienced
non-operational incidents after AOD. Safegate suggested that there
were two causes for this: (1) the installation of sponge washable air
filters in the display units which curtailed airflow within the display units;
(2) the unstable voltage experienced at the new airport which caused
thermo fuses and resistors to trip. This was denied by AA who alleged
that Safegate had failed to design the system with suitable cooling and in
compliance with the agreed voltage variation capability.

12.185 To rectify the problem, Safegate removed the washable
sponge filters from the display units and optimised the size of the thermo
fuses and resistors to accommodate the voltage situation in the new
airport.

12.186 Since 12 September 1998, all APAs at frontal stands have
been put back to service. As at 17 September 1998, all Gantry APAs
were successfully tested and were waiting for AAs decision to use them.
Aircraft parking was not seriously affected during the suspension of APA
as parking was directed by air marshalls.

12.187 The accident occurred at 7:41 am on 15 J uly 1998 involving
a Cathay Pacific B-747 aircraft (CX260). The Cathay Pacific aircraft
overshot the stop bar by about six metres during parking. As a result,
its engine came into contact with the passenger jetway and damage was
309
caused to the engine cowling and the lower cover of the fixed ground
power housing on the airbridge. Fortunately, no one was injured.

12.188 The parking was directed by an APA with an air marshall
being present to provide safety monitoring. Apparently, the APA was
not working properly in that the laser sensor was unable to track the type
of aircraft under parking. The inoperative laser sensor would have been
identified by the system but for the inadvertent disabling of the
auto-calibration function by a Safegate staff. When the air marshall
realised that the APA was not working properly, he tried to signal the
pilot to stop. The air marshall could have pressed the emergency stop
button on the control panel of the APA to effect the display of Stop
message on the display unit to direct the pilot to stop. However, the
control panel was outside the reach of the air marshall who would have
to take more than 10 seconds to reach the panel. The air marshall
therefore gave manual emergency stop signals to the pilot. The pilot
apparently misunderstood the signal of the air marshall and the floating
arrows on the APA display unit as a direction to move forward. When
the pilot realised the emergency stop signal and stopped the aircraft, it
had overshot by about six metres and had hit the passenger jetway.

12.189 After the incident, all APAs were suspended from operation
for 16 days as described under paragraph 12.183 above. Safegate also
carried out certain remedial actions: (1) the laser unit concerned was
replaced and tested; (2) all the stands were checked to verify that the
auto-calibration test was properly enabled; and (3) documentation
showing individual stands correct configuration data had been printed
and signed off by Safegate and countersigned by AA. AA was
considering the reinstallation of the control panel of the APA so that
emergency stop could be activated by the air marshall in case of
problem.

12.190 This was a single incident and no further accident was
reported involving the parking of aircraft.

[35] An Arriving Passenger Suffering from Heart Attack not being Sent to Hospital
Expeditiously on 11 August 1998

310
12.191 There were complaints that on 11 August 1998, an arriving
passenger with heart attack on board China Southern Airlines Limited
flight CZ3077 from Hainan to Hong Kong was not sent to hospital
expeditiously. It was alleged that it took about 20 minutes for an
ambulance to reach the patient on board.

12.192 According to the facts gathered by the Commissioners from
the parties involved in the incident, at about 10:56 am on 11 August 1998,
the Fire Services Communication Centre (FSCC) of Fire Services
Department (FSD) received through 999 an emergency call made by
a person using mobile phone on board flight CZ3077 that a passenger
was suffering from heart disease. At that time, it was understood that
the relevant aircraft was already on the apron. The subsequent events
are set out in the following chronology :-

10:57 Ambulance from CLK Fire Station was despatched.

10:59 Airport Main Fire Station Rescue Control (AMFSRC)
was informed to contact ACC and AOCC for arrangement
of escort.

11:00 ACC was informed to provide escort.

11:01 Ambulance arrived at apron gate.

11:06 ACC escort arrived.

11:09 Ambulance arrived at the aircraft and reached the patient on
board.

12.193 According to the above chronology, it took thirteen minutes
for the ambulance to reach the patient, not the alleged twenty minutes.
To improve response time to emergency incidents like this, AA and FSD
are already arranging a direct line between FSCC and ACC to be
installed so that in future, requests for ACC escort vehicle do not have to
go through AMFSRC. AA is also exploring with FSD possibilities of
AA providing training for the Landside Fire Station ambulance crew on
driving in PTB and Cargo Apron areas and qualifying them for driving at
311
airside so as to eliminate delay in waiting for escort at the apron gate.

[36] Fire Engines Driving on the Tarmac Crossed the Path of an
Arriving Aircraft on 25 August 1998

12.194 On 25 August 1998, four fire engines drove across the
runway to attend to an incident of a J AL aircraft without obtaining
permission from the ATC, forcing a Cathay Pacific flight to abort
take-off and a China Eastern Airlines flight to delay landing.

12.195 According to the report of FSD on the incident, ATC
informed the Airport Fire Contingent (AFC) to respond to an incident
whereby a small hatch door of a J AL DC-10 aircraft on a taxiway was in
an open position. Four fire engines were dispatched. The Rescue
Leader of the four fire engines radioed ATC Tower wishing to obtain
clearance to cross the runway. Before he could obtain the necessary
clearance, the driver of the first fire engine speedily drove across the
runway without confirming permission from ATC nor the Rescue Leader.
Upon seeing the first engine crossing the runway at high speed, the
Rescue Leader considered that instructing it to return would only
lengthen the time of the fire appliance staying on the runway which
would further obstruct runway operation. Seeing that the aircraft at the
threshold of the runway was stationary, he quickly followed with the
remaining three appliances and dashed across the runway.

12.196 According to the report on the incident of CAD, ATC saw
the fire engines crossing the runway without permission while the
Rescue Leader reported on radio that they were responding to the request
for inspection of the J AL aircraft. At that time, Cathay Pacific Airbus
A340 aircraft cleared by ATC for take-off had just started its take-off roll
and an incoming China Eastern Airlines Airbus A320 aircraft was
approximately five kilometres from the airport. Instruction was
immediately given by ATC to A340 to abort take-off. A340 stopped at
about 200 metres from its take-off commencement position and by that
time, the fire engines had already crossed the runway at a location some
1,400 metres further down the runway. There was no risk of collision.
A340 was then instructed by ATC to vacate the runway. The Air
Movement Controller judged that the runway would not be available in
312
time for the approaching A320 to land and instructed it to discontinue its
approach. No danger to safety was involved.

12.197 It was later confirmed that the reported unclosed hatch door
of the J AL aircraft was in fact a valve in a proper position.

[37] A HAS Tractor Crashed into a Light Goods Vehicle, Injuring Five
Persons on 6 September 1998

12.198 On 6 September 1998, a tractor of HAS crashed into a light
goods vehicle (a control van), injuring five persons. The driver of the
tractor towing two empty containers and an empty dolly was driving in
the restricted area of the airport from the south towards north. As he
was driving between two lines of containers, his view was partially
blocked on the left while he was going out of the area and he was not
aware of the arrival of the light goods vehicle. The tractor collided with
the control van passing horizontally in front. As a result of the collision,
five persons on the control van sustained injuries. All but two were
immediately discharged after medical treatment and none was
hospitalised.

[38] Tyre Burst of United Arab Emirates Cargo Flight EK9881 and
Runway Closures on 12 October 1998

12.199 On 12 October 1998, United Arab Emirates Airline flight
EK9881, a cargo B747-200 aircraft leased from Atlas Air, Inc, sustained
tyre burst on departure for Dubai, leaving behind tyre debris on the
runway. Tyre fragments covered an extensive area of the runway. The
runway was closed for 40 minutes for removal of the tyre debris.
About one and a half hours after take-off, the aircraft returned to Hong
Kong on a slight hydraulic problem, damaging runway lights on landing.
The runway was then closed twice, 39 minutes and 20 minutes
respectively for inspection of the runway conditions and emergency
repairs to the lights. Further repairs to the lights were made overnight.

12.200 The incidents, which necessitated the closure of the runway
three times in a day, had an impact on the operation of the airport. To
maintain the integrity of the runway for the safe operation of the airport,
313
runway closures were necessary to remove the tyre debris and effect
repairs to the runway lights. During the runway closures, four aircraft
were diverted to alternative airports, 42 arriving flights were delayed
between 15 and 69 minutes and 88 departing flights were delayed
between 15 and 75 minutes.

[39] Power Outage of ST 1 due to the Collapse of Ceiling Suspended
Bus-bars on 15 October 1998

12.201 This item is dealt with in para 11.15 of Chapter 11.

314
314
CHAPTER 13

RESPONSIBILITY -- FIDS

Section 1 : History of Development, Installation, Testing and
Commissioning of FIDS Delays and Problems

Section 2 : What was Wrong with FIDS?

Section 3 : Repairs after AOD

Section 4 : Causes and Responsibility

Section 1 : History of Development, Installation, Testing &
Commissioning of FIDS Delays and Problems

13.1 Flight Information Display System (FIDS) is a highly
sophisticated state of the art system serving a number of important
functions essential for airport operation. A diagrammatic illustration of
FIDS and other airport systems connected to it (excluding certain systems
like the Time of Day Clock which are not relevant for present purposes)
is at Appendix XIV. It can be seen from this illustration that FIDS
interfaces with a number of other systems, which are developed by a
number of contractors or suppliers.

13.2 There are the following contractors, subcontractors and
suppliers:

(a) For FIDS, the main contractor of Airport Authority (AA),
under contract C381, is G.E.C. (Hong Kong) Ltd. (GEC),
and under it there are two main sub-contractors: (i) Electronic
Data Systems Limited (EDS) who is responsible for the
system hardware except the liquid crystal display (LCD)
315
boards and the development of the software, and (ii) EEV
Limited (EEV) who is responsible for supplying the LCD
boards.

(b) Under EDS, The Preston Group Pty Ltd (Preston) is the
subcontractor for the provision of the Terminal Management
System (TMS), an integral component of FIDS. The
monitors were supplied by EDS sub-subcontractor
FIMI-Philips S.r.l. (FIMI).

(c) The main contractor for Baggage Handling System (BHS)
is Swire Engineering Services Ltd (SESL) under contract
C360.

(d) For the Airport Operational Database (AODB), Hughes
Asia Pacific (Hong Kong) Limited is the main contractor
under contract C399, and Ferranti Air Systems Limited the
sub-contractor who designed and developed the software.

(e) The Common User Terminal Equipment (CUTE) network,
used by the airlines, is provided by the Societe Internationale
de Telecommunications Aeronautiques (SITA).

13.3 Another diagram in colour, Appendix XV, shows the four
functions of FIDS, namely, generating in red, processing yellow,
transmitting blue and presenting purple. The main components of FIDS
include the following:

(1) FIDS Workstations - there are 18 PC workstations located at
the Apron Control Centre (ACC), Airport Operations
Control Centre (AOCC) and Baggage Control Room
(BCR) where the Airport Management Division (AMD)
and BHS operators are able to monitor or enter data into
FIDS and TMS through a number of pre-designed
screens/windows or Man Machine Interface (MMI). The
airlines are also able to access FIDS via the CUTE network
workstations provided by SITA (288 check-in desks, 54
316
transfer desks and 86 gate desks), which run a FIDS
software provided by EDS.

(2) FIDS Host Server - the FIDS host comprises an integrated
software performing (A) stand, gate and desks allocation and
(B) flight information display, with a common Oracle
database for both parts of the software.

(A) TMS - this component is a resource allocation system
and performs the following main functions:

(i) Stand allocation - TMS automatically optimises
the allocation of parking stands for aircraft (both
frontal and remote) at the new airport based on
the flight schedules for each day and in
accordance with certain preset rules.
Optimisation means that the allocation of
parking stands for aircraft would be made in a
most efficient and desirable manner, taking into
account their arrival and departure times. This
function is performed using the FIDS
workstations in the ACC. The stand allocation
produced by the TMS optimisation process has
to be confirmed by the ACC operators, who
could also manually override the allocation
produced by the optimiser.

(ii) Gate and Desk Allocation - these functions
optimise the allocation of gates (mainly apron
passenger vehicle (APV) gates) and desks
(both check-in and transfer desks) in a similar
way to the stand allocation function. This part
of the system is operated and accessed through
the FIDS workstations in the AOCC.

(iii) Data Input - since TMS and the FIDS display
software share the same database, stand, gate
317
and desk allocations produced by TMS do not
have to be manually input into the system before
other components of the system or other
connected systems could access the information.
However, TMS is capable of being used for data
input independently of the optimiser function.
Indeed, if the stand allocation optimiser function
is not used for any reason, the only way that the
stand allocations (produced by Stand Allocation
System (SAS) or manually by ACC staff)
could be properly displayed by FIDS is to have
such data manually input through TMS.

(B) Flight Information Display - this part of FIDS is
represented by the small box marked FIDS within
the big yellow box on Appendix XV. As stated
above, the display software shares the same database
with TMS. Thus the stand, gate and desk allocations,
once confirmed in TMS, will be automatically
available to the display software and the various
systems connected with FIDS, like CUTE, AODB and
BHS. FIDS also receives relevant flight or flight
related information from the AODB, CUTE and BHS
and sends data to the display servers for further
transmission to the display devices.

(3) Display Servers - there are altogether 57 display servers
installed in various Communications Rooms throughout the
Passenger Terminal Building (PTB) which drive various
groups of display monitors and LCDs throughout PTB.
These display servers receive flight data from the FIDS host
and determine which data are to be displayed on the display
devices.

(4) Display devices - under the contract there should be some
146 LCDs and 2,057 monitors throughout PTB displaying
relevant flight information at different locations in PTB.
318
According to GEC, on airport opening day (AOD) 142
LCDs and 1,952 monitors were available, of which 137 and
1,913 respectively were working throughout the day.

13.4 The interaction between the various airport systems depicted
in Appendix X can be briefly summarised, as follows:

(a) AODB is the main repository or post office of operation
data, mostly flight-related. AODB receives data from
various sources and distributes them unmodified to the
various connected systems. Flight data from FIDS and
Civil Aviation Department (CAD), for example, are
connected to and made available to each other through the
AODB linkage. One of the most important data supplied
by CAD to TMS is the estimated time of arrival (ETA),
which is available when the aircraft is within the range of the
Radar Tracker, ie, 45 minutes before landing.

(b) The interface between FIDS and BHS enables the exchange
of information relating to baggage handling, like stand
allocations for inbound and outbound flights and lateral and
baggage reclaim allocations for departing and arrival bags.

(c) Another important system connected to the AODB is the
Flight Data Display System (FDDS), which is designed
and built by Hong Kong Telecom CSL Limited (HKT).
End-users of the FDDS (including Flight Display Data Feed
Services (FDDFS)) service receive flight information or
data they require via the FDDS/FDDFS servers, which in
turn receive the information from AODB. It is noteworthy
that amongst the receivers of flight information via FDDS
are the three ramp handling operators (RHOs) and both
cargo terminal operators (CTOs).

13.5 The contract C381 dated 16 J une 1995 made between AA
and GEC as the main contractor was for GEC to provide FIDS, including
software and hardware. From the very beginning, the FIDS software
319
was plagued with problems. Translation of AAs user requirements as
set out in the particular technical specification (PTS) into functional
requirements in the system segment specification (SSS), according to
which the system would be designed and developed, took some 12
months longer than expected. This was due to the admitted ambiguity
of the PTS, and the many subsequent changes to the SSS requested by
AA. The changes were in turn the result of the relatively late
involvement of the present AMD, the user of the system, in the
specification phase. In March 1997, the development of FIDS was
severely set back when AA and EDS agreed that the level of complexity
of AAs user requirements would not be met by EDS either modifying one
of their existing software or by buying a ready product off the shelf. As
a result, EDS had to start from scratch to develop an entirely new
software for AA. This caused a delay of 14 months and additional costs.
An agreement was subsequently reached between AA and GEC to settle
the delay up to 10 December 1997 and for additional costs, at $89.7
million to be payable by AA to GEC. This late development of FIDS
resulted in the time that was originally planned for installation, testing
and commissioning of FIDS, for integration of FIDS with other systems
and for training of operators of the systems to be hard compressed.

13.6 In September 1997, there were discussions amongst AA,
GEC and EDS as to how to recover from the further slippage that was
experienced in the development of the FIDS software. It was then
decided that the system should be divided into various builds, with each
build designed for a particular functionality, namely, build 1.0 for control
and monitoring of external interfaces, build 1.1 for integration with
AODB, build 1.2 to provide the core host server data processing
functionality for integration with AODB display panel, build 1.3 for the
production and maintenance of screen formats to display on the FIDS
display devices, build 1.4 to provide interface with CUTE, build 1.5 to
provide FIDS with seasonal schedule and TMS facilities, and build 2.0 to
incorporate all previous builds, including the MMI functionality. The
development of the FIDS software in builds enabled AA to monitor EDS
progress. EDS delivered each build as soon as it was completed to the
Interface House, a facility of AA used for the testing purposes, so that
each portion of FIDS, providing a certain functionality, would be tested
320
and be available for use for hands-on training of airport operators as soon
as possible. Each build could have been developed to work as a
standalone system, ie, to work independently without being linked up or
integrated with any of the other builds. With standalone builds, AA
would have had a fallback on these unintegrated functionalities for airport
operations, should the integrated FIDS run into problems.

13.7 Unfortunately, the development of the integrated FIDS
proceeded without consideration of a fallback, and when AA and EDS
stopped to consider the re-development of the standalone builds, it was
already too late for meeting the April 1998 AOD. Build 2.0
incorporated and integrated all previous builds, and was delivered to PTB
in December 1997. However, there were numerous problems identified,
termed problem reports (PRs), during the various tests from thence up
to AOD. The final build of FIDS, with a great number of PRs critical to
airport opening fixed, was delivered around 23 J une 1998, and a new
release of TMS having a memory leakage problem corrected was made at
the end of J une 1998.

13.8 Earlier in December 1997, at the time when airport opening
was expected to be in April 1998, there had been a further agreement
between AA and GEC that the factory acceptance test (FAT) of the
FIDS software was to be combined with the site acceptance test (SAT).
FAT was meant to test the software at EDS premises in Hook, England,
so that any PRs revealed could be fixed before the software was to be
delivered to site, ie, the new airport. The software delivered to and
installed at site would then be subject to SAT when there would be fewer
problems, as those identified at FAT would have already been fixed.
SAT was meant to unearth problems encountered when the software was
set up at site, and these PRs would then be resolved. According to the
contract between AA and GEC, FAT was to take place in EDS premises
in Hook, and that would take place before the software was delivered to
the new airport. The combination of FAT and SAT was aimed at
catching up some time that had been lost previously, so that problems
with the software would be identified at one and the same time during the
combined tests and such problems resolved at the new airport collectively.
This agreement was made with objection from EDS which maintained the
321
view that a combined FAT and SAT would add risk to the completion of a
problem-free product as certain software errors were best identified and
rectified in the factory environment.

13.9 At the first airport trial on 18 J anuary 1998, FIDS as
installed at PTB crashed. It was discovered that FIDS and CUTE used
by airlines were not compatible. In the Airport Development Steering
Committee (ADSCOM) meeting on 14 February 1998, W43 Mr
Douglas Edwin Oakervee, the Project Director of AA, explained that the
problem at the time was caused by the fact that SITA, the contractor
supplying CUTE for use by the airline operators, had loaded the software
for CUTE incorrectly. Subsequently, SITA flew in an expert from New
York to resolve the problem. In the ADSCOM meeting on 22 May 1998,
AA reported that the FIDS/CUTE interface had been completed and that
SITA had provided the certification to FIDS in the form of an e-mail
advice on 6 May 1998. However, even in the final airport trial on 14
J une 1998, there were reports about the slow response experienced by
Cathay Pacific Airways Limited (Cathay Pacific) and Hong Kong
Dragon Airlines Limited (Dragon Air) in accessing their Departure
Control System (DCS) through CUTE, which according to Cathay
Pacific was caused by the failure of the communication links between
PTB and SITAs overseas sites. After some system improvements by
SITA, the airlines organised in conjunction with AA a large-scale dry run
of CUTE on 30 J une 1998. Thirty-two or 35 airlines participated in the
exercise and their operators logged in at 80 check-in counters
simultaneously. These operators operated the FIDS/CUTE interface and
DCS. AA operators also tested the standby FIDS, a contingency more
particularly referred to below in case of failure of the main FIDS,
according to the procedures prescribed by AA. No problem was noticed
nor afterwards reported during this dry run.

13.10 Apart from the interface problem between FIDS and CUTE,
FIDS was not running smoothly. It was continuously plagued with
problems and Government kept on reminding AA of the necessity to have
a stand-alone system to work as a fallback in the event of a FIDS
breakdown. At the AA Board meeting on 26 February 1998, the AA
Board finally directed the AA management to procure a separate standby
322
system (standby FIDS) that would function as a fallback if the main or
permanent FIDS could not be delivered on time or failed. At the AA
Board Meeting on 23 March 1998, it was also decided that a standby SAS
should be commissioned as a fallback should TMS fail. SAS is a
stand-alone system and does not interface with FIDS.

13.11 There were a number of tests of FIDS, although they were
carried out later than the contractual timetable or planned dates. Out of
the initial contractual term for commissioning the whole system in three
years, 14 months had been lost in its development by early 1997. The
remaining time for the completion of the work was tight, involving a risk
which was appreciated by EDS. The altered timetable agreed between
the parties for the delivery of the builds and testing and commissioning
was aggressive, as described by the project engineer of EDS, W21 Mr
Michael Todd Korkowski. He told the Commission that there was a
delay in the delivery of each build, which affected the testing programme.
The integrated FIDS, that is FIDS and its interfaces with AODB, BHS,
FDDS, Aeronautical Information Database, Mass Transit Railway
Corporation and Scheduling Committee Computer, was run continuously
for two weeks leading up to AOD, to test the processing of flight data.
The test revealed the need for Central Processing Unit (CPU) upgrade
and additional memory in the system . However, W21 Korkowski was
not sure whether those other systems were in full running mode, eg,
whether BHS was actually moving bags or whether all the check-in desks
were running on the FIDS/CUTE interface. A fully simulated live load
operation was never performed on FIDS and its interfaces and the closest
thing was when FIDS was used during the fifth airport trial that took
place on 14 J une 1998.

13.12 W21 Korkowski told the Commission that the variations
requested by AA and the five airport trials from J anuary to 14 J une 1998
were outside the contractual terms. The variations required testing and
the airport trials needed a lot of preparation by EDS personnel. These
additional requirements diverted EDS concentration on resolving PRs.
He produced a list to show that altogether over 12,000 man-hours had
been taken off their key resources that could not be replaced, and a month
of delay was the consequence.
323

13.13 FIDS is composed of hardware and software. For the
purpose of the inquiry, two hardware components were of significance,
namely, the LCD boards supplied by EEV, and the monitors supplied by
FIMI. The monitors and LCD boards display flight-related information
such as flight times, check-in desks, boarding gate numbers and reclaim
belt assignment to passengers. Most of the flight information displayed
on the monitors and LCD boards would be disseminated through 57
display servers which were fed with the data by two host servers.

13.14 Mr Raymond HO Wai Fu, Chief Assistant Secretary for
Works (Information Technology) of Government, stated that he noted
from the FIDS daily report prepared by AA on the reliability tests carried
out in J une 1998 that there were two incidents of major failures of display
servers, although the accumulated availability of the host servers was
satisfactory during the initial period between 14 J une and 18 J une 1998.
The first one occurred on 18 J une 1998 involving seven servers and the
other one occurred on 19 J une 1998 involving 16 servers. At the time of
the report, AA recognised that the major failure of display servers was a
serious concern and that GEC and EDS and their subcontractors were
trying to resolve the problem as soon as possible. On 22 J une 1998, Mr
Raymond Ho noticed from the daily FIDS reports that there were some
problems on the display servers which failed to update some of the
display devices. Consequently there were inconsistent information
between display monitors.

13.15 The 1,952 monitors and 142 LCD boards were themselves
not free from trouble. According to the contractors, many monitors
were delivered to PTB in 1996 and kept there before the air-conditioning
system was operating. Subject to heat and moisture, many connectors
were oxidised and the voltage level at the end of the cable reduced.
About 20 monitors were damaged by water. Many of the other monitors
were subject to problems such as software errors, operator errors, cable
length, data cable integrity and even lack of power supply. The length
of the cable between the display server and the monitor or LCD board
was not to exceed 90 metres or else the display performance would be
affected. At the current design, 30% of the display devices had cables
324
over 100 metres and 5% over 150 metres. W3 Dr Henry Duane
Townsend said he also knew of this cable length requirement, but by
installing more display servers at various locations, the cable length
would be shortened. On AOD, according to W22 Mr Edward George
Hobhouse, the project manager of GEC at PTB, about 93.8% of LCD
boards and 93% of monitors were working, that is, 32 monitors and 2
LCD boards malfunctioned on AOD. About 120 monitors were
replaced in the three weeks after AOD. However, W22 Hobhouse said
that that was within the expected failure rate of monitors which was about
200 a year, ie, 10% out of the roughly 2,000 monitors installed.

13.16 According to the Daily Report dated 22/6/98, FIDS
resilience tests would be carried out on 25 J une 1998 and stress and
loading tests on 26 J une 1998. These tests would serve to verify if the
system could perform under load as designed. However, the Daily
Report dated 28/6/98 stated that the resilience tests were rescheduled to
30 J une 1998 and the stress and loading tests deferred until after AOD.
The Daily Report dated 2/7/98 recorded that formal resilience tests would
be carried out after AOD. Mr Raymond Ho stated that he had
reservations about the lack of these tests, but considered that it was a
matter for AA, its contractors and consultants to decide the testing
requirements for the commissioning of the systems. He was not advised
by AA of the risks of operating a FIDS which had not been through a
stress test.

13.17 While maintaining that FIDS was operational and functional
on AOD, W21 Korkowski, the project engineer of EDS stationed at PTB,
frankly admitted that no stress test that should have been undertaken
before AOD was performed. Although there was no specific
requirement for a stress test on FIDS in the contract between AA and
GEC, a stress test was industry practice for an important system on which
operations depend, and would have taken EDS three to five days in
preparation and one or two days in carrying out. W21 Korkowski
agreed that with hindsight he should have advised AA or CSE
International Limited (CSE), AAs consultant on systems, of the risks
of not performing a stress test before AOD, which he failed to do. He
also said that AA and CSE should have known the significance of a stress
325
test. As EDS had at the period starting from 9 J une 1998 at least 38 PRs
to rectify, there was no time to carry out the stress test. That reflected
the compressed timetable for testing since the loss of 14 months at the
early stage of the development of the software and continued slippage
thereafter. Mr David J ohn Thompson, a senior coordinator of New
Airport Projects Co-ordination Office (NAPCO) for special systems
from International Bechtel Company Ltd. stated in his witness statement
that he was aware that the FIDS stress test had been deferred at the very
last minute. He believed that there was insufficient time to carry it out.
FIDS exhibited problems related to operational stability and given the
time constraints leading to AOD, he believed that resolving stability
problems and ensuring Day One functionality was of far higher concern
to AA and EDS than stress testing, which in any event would normally
only be performed on a stable system. Mr Thompson claimed that in the
case of FIDS, a stress test would have most probably indicated that the
system was slow in responding to operator input, which in itself would
not necessarily have been catastrophic for Day One operations had there
not been other simultaneous problems. According to W55 Dr Ulrich
Kipper and W56 Professor Vincent Yun SHEN, experts appointed by the
Commission, a stress test would probably have revealed slow response to
operator input and certain deadlock problems in the database. These
two problems hampered ACC and AOCC operations on AOD, especially
when operators were trying to catch up with time lost in the early
morning due to difficulties performing flight swapping. These problems
were in fact known to AA and EDS before AOD, but their reappearance
in a stress test might have highlighted their significance on a fully loaded
system, and thus the need to resolve them before AOD. Bearing in mind
that any test is only as good as the measures taken to resolve the PRs, and
considering the limited time EDS and AA had to solve PRs before AOD,
it is uncertain if a stress test performed on the problem-ridden system and
in compressed conditions would have saved the day. Yet, the
Commissioners note that it was industry practice to carry out the test
before operations. This highlights the serious risks that AA faced with
their operation systems in the build-up to AOD and the dire need for a
global contingency plan.

13.18 W22 Hobhouse placed importance on a confidence trial
326
involving operations staff working on the system before AOD, which
would have given them a measure of confidence in operating FIDS. He
told the Commission that a stress test would take place prior to a
confidence trial, and such a trial should have been carried out before
AOD, but it was not done. He described FIDS as workable on AOD,
with TMS almost there, running as a planning and allocation tool on Day
Three, and that by Day Six operation was as good as at Kai Tak. On the
other hand, he did not deny that there were a lot of problems on AOD and
that FIDS was not operating efficiently then.

13.19 In his witness statement, W28 Mr Anders YUEN Hon Sing,
the Assistant Airfield Duty Manager of AA stationed in ACC, wrote that
at the multi-system reliability tests from 8 to 23 J une 1998, a number of
problems with TMS/FIDS were encountered, namely,

(a) extreme slow functioning of the system;

(b) use of the optimisation command would often cause the
Gantt chart to freeze or shut down;

(c) incorrect display of flight data (TMS would split the display
of a linked arrival and departure, thus showing double the
number of actual flights and sometimes a number of flights
would be missing from the chart altogether);

(d) the inability to print a full Gantt chart (the screen could not
display a full Gantt chart on one view); and

(e) the what if scenario planning function was not available.

13.20 W24 Ms Rita LEE Fung King, FIDS Project Manager of
AAs Information Technology (IT) Department told the Commission
that there were also concerns at the instability of the TMS server. When
giving evidence before the Commission, W28 Yuen, W23 Mr Alan LAM
Tai Chi, AAs General Manager (Airfield Operations) and W24 Lee
clarified that the above problems were to a certain extent fixed before
AOD. However, slow response time improved only slightly and
327
suspicions by AA that there were problems with the TMS server persisted
even on AOD. The what if scenario, which would only be used once
in a month for planning, was not critical for AOD and was to be available
only in October 1998.

13.21 W28 Yuen relayed these problems to W23 Lam on a number
of occasions. W23 Lam took the decision in consultation with W44 Mr
Chern Heed and AAs Mr Thomas Lam to use SAS and not TMS to
produce the preliminary stand allocation schedule. As the SAS and
TMS systems had different software configurations, they produced
different stand allocations, even though they were programmed with the
same stand allocation rules. This decision means that SAS would be
used on AOD for optimisation of the allocation of parking stands for
aircraft, and because SAS did not interface with FIDS, the stands so
allocated by SAS would have to be manually entered into TMS through
the FIDS MMI for dissemination and display on FIDS.

Section 2 : What was Wrong with FIDS?

13.22 From the evidence of the observers of the FIDS displays and
operators of FIDS, as detailed in Section 2 of Chapter 10, the following
visible problems are identified:

(a) The flight information displayed on the FIDS monitors and
LCD boards was incorrect, inconsistent and incomplete, and
the monitors and LCD boards were sometimes blank or
blacked out;

(b) ACC operators were not able to perform the flight swapping
function initially, and had to learn the correct method from
W24 Lee and Preston later on;

(c) Response time was slow with TMS and FIDS MMIs at ACC,
AOCC and BCR, resulting in very slow update of
information into FIDS;

328
(d) The Gantt chart on TMS shut down intermittently; and

(e) Green bars appeared on the Gantt chart on TMS indicating
that ETA was earlier than scheduled time of arrival (STA)
by more than 15 minutes.

13.23 GEC, EDS and Preston, the contractor and subcontractors in
the supply of FIDS including TMS and the monitors and LCD boards
should be the people who had intimate knowledge of the underlying
causes for the problems as observed. According to them, the causes
were as follows:

(a) The monitors and LCD boards displayed obsolete data
because the display servers sometimes locked up, preventing
receipt of updated data. An operator-controlled switchover
would cause the monitor to blank out for about five minutes
while the substitute server was receiving a download of data.
A LCD board would be blanked out completely when the
connection to the display server was lost. There were also
hardware problems that affected the availability of monitors
and LCD boards, such as lack of power, incorrect cable
lengths and networking problems.

(b) They did not know that AA did not use TMS as a stand
allocation planning tool with its optimisation function.
Using TMS to input stand allocations produced by SAS, as
AA staff did on AOD, was not a designed use. Had they
known, they would have raised concerns or advised AA as to
how to update TMS properly by manual input. Using SAS
for optimisation and manually entering the allocation
obtained from SAS into FIDS increased the workload of the
AA operators.

(c) There were two FIDS MMIs in ACC. If the Gantt chart on
one of them shut down, the other could be used to create
another Gantt chart, with little impact on operations.

329
(d) Green bars were a system design in TMS to show that ETA
was earlier than STA by more than 15 minutes. The green
bars appeared in accordance with the design and they did not
indicate a software problem.

13.24 The contractors and subcontractors also said that the
operators of the AA, airlines and BCR were not familiar or experienced
with using TMS and FIDS MMIs and therefore caused problems. AA
did not ask EDS and Preston personnel for assistance immediately upon
AAs staff experiencing difficulty operating TMS. AA did not confirm
the stand for turnaround flights until ETA was received, which was in
some cases about 15 minutes before the landing of the arriving flight,
which resulted in RHOs not knowing where to find the aircraft and in
gates not being able to be assigned as CUTE would not allow the airline
to log on until the stand was allocated. All these had nothing to do with
the integrity of FIDS.

13.25 Except for the cabling problem, W21 Korkowski admitted
that EDS was responsible for the quality and commissioning of the
monitors that were supplied by its sub-subcontractor FIMI. GEC, on the
other hand, was responsible for the quality and commissioning of the
LCD boards that were supplied by its subcontractor EEV. While only
10% of the monitors and LCD boards did not function, for whatever
reasons, on AOD, the impact would not have been serious had FIDS
worked properly and smoothly. This is because each malfunctioning
display device affected only a small group of end-users, who could have
been directed to the nearest working display.

13.26 W23 Lam told the Commission that a decision was made in
J une 1998 that on AOD, TMS would not be used for stand allocation
planning, which would instead be performed on standby SAS developed
by City University (City U). Both SAS and TMS had the function of
optimisation for such planning. While SAS had only the function of
planning stand allocation and was a standalone system unintegrated with
any of the other systems used in the new airport, TMS also had the
functions of allocating gates and check-in and transfer desks, and it was
part of FIDS and therefore integrated with many other systems. This
330
difference meant that stand allocations produced by TMS could be
disseminated through FIDS to users without extra effort, while SAS
allocations had to be manually entered into TMS via FIDS MMI before
the information could be published on FIDS display devices. The
decision to use SAS as the primary stand allocation tool was made mainly
because the performance of TMS during various tests up to the middle of
J une 1998 was unreliable for the reasons cited in paragraphs 13.19 to
13.21. The staff at ACC were all drawn from Kai Tak and they had the
experience of operating Kai Taks stand allocation system that was
similar to SAS. SAS was able to print out the Gantt chart prepared by it
whereas TMS was unable to do so when the decision was taken.
Although the performance of TMS showed improvement when reaching
AOD, the decision was not altered.

13.27 W23 Lam told the Commission that on 5 J uly 1998 at about
4:30 pm, the scheduled flight information was loaded into both SAS and
TMS. From both systems, a Gantt chart was created on the workstations,
and stand allocation planning using the optimisation function was
employed in each, but for different purposes. SAS was used as the
primary tool for stand allocation, and the allocations made by it would be
adopted for operations. On the other hand, the optimisation function of
TMS was used purely for stands to be allocated on its own Gantt chart so
that any allocations that were not identical to those made by SAS would
be altered accordingly from time to time when required by manual input.
The necessity of using TMS in this way was to enable FIDS to be updated
with allocations from SAS and for dissemination of the updated
information to the other systems interfaced with FIDS, with which SAS
had no link.

13.28 W23 Lam further said that as aircraft from Kai Tak arrived at
the new airport as from 9:15 pm on 5 J uly 1998, stands were allocated to
these flights on TMS in accordance with the schedule prepared on SAS.
Updates were input into SAS and similarly into TMS. For all these ferry
flights, ACC staff entered chocks-on time and the registration number in
the FIDS MM1. However, in doing so, W28 Yuen invoked a prompt
linking flights by registration number. This would inhibit flight
swapping by manual linking procedures later on. At about 1 am on
331
AOD, when ACC received flight movement sheets from Cathay Pacific
requiring flight swaps, ACC operators were unable to execute the flight
swapping command because they were unfamiliar with the progression of
the levels of flight linking and did not know how to use the method of
swapping by aircraft registration numbers. Difficulty was also
experienced with attempts to carry out flight swapping on SAS to the
extent that at about 2:30 am, SAS crashed. Thereupon, City U was
contacted for help. With both systems hung, a manual system of stand
allocation was set up by ACC staff. The manual system consists of a
whiteboard set up with a printed copy of the SAS Gantt chart with all the
stands available and with flight numbers and aircraft registration numbers
(or tail numbers) on stickers for adhering to the allocated or assigned
stands. The stands allocated manually would have to be manually
entered into TMS/FIDS through the FIDS MMIs, so as to disseminate the
information through FIDS to other systems of the new airport. When a
City U representative came to ACC with a workaround for SAS at about
8:30 am, ACC staff had gone quite far with using the manual allocation.
However, the problems encountered with TMS remained, and the
operators at ACC had difficulty in inputting the stands allocated,
especially for swapped flights.

13.29 W29 Mr CHAN Kin Sing, a colleague of W28 Yuen on duty
at ACC on AOD, told the Commission that he called the IT Department
of AA when problems with flight swapping were first encountered at
around 2 am. A female person who answered his call told him that she
would go to check the server, but nothing was heard from her afterwards.
W28 Yuen told the Commission that he telephoned W24 Lee at about 3
am and asked for help. After this call, he did not hear from her and was
not able to contact her. On the other hand, W24 Lee testified that the
first time she heard from ACC was at 6 am, before which she was going
round PTB to inspect FIDS displays and the operation of the CUTE
workstations by airline staff at the check-in counters. She immediately
called W21 Korkowski who directed her to call W35 Gordon J ames
Cumming, also of EDS.

13.30 W24 Lee arrived at ACC at about 6:30 am on AOD. She
helped the operators with flight swapping and with confirming stand
332
allocations entered manually into TMS. W23 Lam, who gave evidence
with W24 Lee in a group, was also in ACC from the small hours of the
morning and witnessed what happened at ACC and W24 Lees arrival.
W24 Lee was conversant with IT matters and the operation of FIDS and
TMS. Her ability in IT matters and operation of FIDS was not doubted
by W21 Korkowski.

13.31 According to both W23 Lam and W24 Lee, the ACC staff
had a lot of problems in inputting data into TMS. The updating of flight
schedules could not catch up with the real time situation. The input of
stand confirmation, chocks-on and chocks-off times, and aircraft tail
numbers could not be made in time, such that sometimes the aircraft were
on the ground but the information had not been successfully entered into
TMS. This was mainly caused by the slow response of TMS.
Sometimes, the Gantt chart of TMS disappeared for between 5 minutes to
one and a half hours. The operators had difficulty in confirming a
particular stand, and experienced problems entering the aircraft
registration number to break manual links.

13.32 On the corrupted flight data from CAD which was used by
ACC before 8:30 am, W24 Lee said that while the ETA being earlier than
STA by more than 15 minutes created green bars on Gantt chart, there
were only very few such occurrences and they did not have a major
impact on operations because they did not prevent her from performing
flight swapping or confirmation.

13.33 On the other hand, W34 Mr Peter Lindsay Derrick said that
when he arrived at ACC at 12:30 pm on AOD, he noticed that the Gantt
chart on the TMS workstation showed Gantt boxes overlapping one
another and in conflict. The data showed that the ETAs of the
associated flights were significantly different from the scheduled times.
It was then agreed between him and the AA staff there that the flights
with ETAs should have the ETAs removed, leaving TMS with the original
scheduled times for the affected flights. After this was done, the Gantt
boxes returned to their planned times and the conflict conditions
disappeared. From W34 Derricks evidence, it is very clear that by 2 pm,
the conflict had all been cleared. He said that while the invalid
333
estimated times were in the database, TMS functionality was severely
impacted.

13.34 W34 Derrick attributed the problems to causes either
external or internal to TMS. The external causes were lack of
experience of operators in using TMS Gantt chart, resulting in incorrect
usage of TMS or receipt of incorrect input by TMS. His descriptions are
as follows:

(a) Some manual links reverted to rotation number links. This
was the result of the operators erroneously pressing the
Return key when entering aircraft registrations for departure
flights via the FIDS MMI. This had only a minimal effect
on apron control operations.

(b) CAD sent invalid estimated time information. The CAD
interface was shut down and invalid estimated times were
removed from TMS. TMS functionality was severely
impacted by the invalid estimated times in the database.

13.35 The internal causes described by W34 Derrick were as
follows:

(a) Constraint error when Confirming some stands with TMS.
When the operator, as advised, repeated the Confirm action,
the operation would complete successfully. This had
minimal effect on apron control operations.

(b) Gantt chart shutdown. This happened intermittently
throughout the afternoon. The cause was later found out to
be operators repeatedly pressing the Apply or OK button on
the allocation dialog when the system did not respond as
quickly as hoped. Although this was annoying for the
operators, it had no effect on apron control operations.

(c) TMS did not behave as expected when flights were linked
using the flight swap function, in two situations. Some
334
manual links did not unlink when the aircraft registrations
were input. However, to fix this, the operator could simply
invoke the Unlink functionality on the button bar or
allocation dialog. Another situation was that re-linking
confirmed allocations did not always reconfirm the new
re-linked allocation. This was simply fixed by asking the
operator to reconfirm the allocations after re-linking.

13.36 W34 Derrick described the internal causes as minor internal
issues, having only a negligible effect on apron control operations, and
that they did not stop the operators from entering the planned stand
allocation information into TMS. However, he also noted that some
frustration was expressed as to the performance of TMS during the
afternoon as some of the transactions were taking longer than hoped
given the pressure the operators were under. The ACC staff also pointed
out to him that during the course of the day, the flight swap function
(Link) as delivered was not as efficient as was required for day to day
operations of the new airport. He said that this particular function was
discussed in detail with the operators long before AOD and the
functionality was available to the operators when TMS was installed on
the test systems at Interface House in late 1997. AA had ample
opportunity to review the functionality before AOD and determine its
suitability for airport operations, but no change requests were made to
Preston.

13.37 W34 Derrick also said that the decision to use SAS, instead
of TMS, to plan stand allocation was not within the knowledge of Preston
until 5 J uly 1998, and intimated that this contributed to the problems
encountered at ACC on that day. W21 Korkowski and W34 Derrick
both also stressed that the use of TMS for inputting stand allocation
planned by another system, namely SAS, but not invoking the
optimisation function in TMS for stand planning was not a designed use,
again intimating that such use would or did cause the problems.
However, both W21 Korkowski and W34 Derrick did not deny that stand
allocations produced by optimisation on TMS, could be altered by
manual input and that manual input for alteration or to override the
automatic allocation by optimisation was a proper function of TMS.
335
Moreover, all these persons responsible for FIDS agreed that there was a
software bug with the Oracle database, which was part of FIDS that
affected the performance of TMS. This was sometimes described as the
shared pool memory problem, which was only fixed, either temporarily or
permanently, days after AOD.

13.38 W27 Ms Yvonne MA Yee Fong, an IT Project Manager of
Information Resource Management in AAs IT Department, gave
evidence with W26 Mrs Vivian CHEUNG Kar Fay, the Terminal Systems
Manager of AMD as a group before the Commission. She explained
that the main cause for the slowness of FIDS was the deadlock problem.
A deadlock occurs in a database with multi-user access to shared
resources, like the database shared by FIDS and TMS, to which different
FIDS and TMS applications need access. The database locks a record
that an application needs to access, to prevent another application from
using the same record before the first has completed access. If two
applications, A1 and A2 need to access the same two records R1 and R2
in different sequence to complete their respective transactions, a deadlock
results when A1 is accessing R1 and A2 is accessing R2. Both
applications cannot proceed since the database has not unlocked R1 and
R2 from the first part of their transactions. W27 Ma said that each time
there was a deadlock, the operators had to wait for certain processes to
finish before they could proceed to another, and that took up a lot of CPU
resources.

13.39 The reason for the crashing of SAS, according to W28
Yuens understanding, was that an attempt to carry out flight linking
resulted in illogical times of the arrival and departure of corresponding
flights, ie, the departure time of an aircraft was earlier than the arrival
time of the same aircraft. Usually there were no problems with flight
swaps. However, when for example there were three pairs of arrival and
departure flights on the same day using the same aircraft, the aircraft
would arrive three times and leave three times. One input of a swap
would affect three departure flights, and the second departure flight might
appear to leave before the third arrival time of the same aircraft. The
illogicality was thus transient and would be corrected when the swaps
relating all three pairs of flights were completed. However, SAS was
336
not able to deal with this transient illogicality. This problem was never
encountered with the stand allocation system used at the Kai Tak airport,
but that system was different from SAS. SAS was therefore required to
be altered after the problems encountered on AOD.

13.40 Dr CHUN Hon Wai of City U agreed that the transient
illogicality was not acceptable to SAS. That was because SAS was
designed to prevent operational errors from happening. He stated that
the design was an improvement on the stand allocation system, also
designed by City U, that was used in Kai Tak, so that no potential errors
in SAS could cause other airport systems, which he expected would be
linked with SAS, to crash. As it stood on AOD, SAS did not prevent
flight swapping altogether, but the user had to correct the departure time
before the swap.

Section 3 : Repairs after AOD

13.41 In order to understand the causes for the problems
encountered with FIDS on AOD, the repairs done or solutions employed
are most relevant.

13.42 Within three weeks after AOD, 120 monitors were replaced.

13.43 After AOD, City U altered SAS. The key feature was to
disable the error checking to permit any type of illogical data to be input.

13.44 The problems with FIDS did not end on AOD. A number
of witnesses gave evidence to the Commission that the system response
remained slow for the first several days of operation. Serious problems
also surfaced on Day Five when FIDS experienced a significant amount
of locking and very high CPU utilisation. Major system changes,
Changes 109 and 118, were effected that night to solve the WDUM
problems and TMS locking, after which system performance improved
significantly. Steps were also taken to increase the memory of the FIDS
workstations in ACC, AOCC and BCR over the first few days of airport
opening. W26 Cheung and W28 Yuen testified that FIDS performed
337
efficiently and stably by about a week after AOD. The evidence of
Cathay Pacific is that by around Day Four to Day Five, the information
was largely but not always accurate, and it did not regain full confidence
in the system until the week after AOD. There is similar evidence from
RHOs. The details of the remedial measures taken for FIDS can be
found in Section 3 of Chapter 10.

Section 4 : Causes and Responsibility

13.45 Having carefully considered all the evidence and
submissions of counsel for the Commission and for the parties, the
Commissioners come to the following findings and conclusions on the
causes for the deficiencies of FIDS on AOD and the days thereafter as
well as the responsibility for these causes. The major contributing
causes to the deficiencies of FIDS were as follows:

(a) Compression of software development time.

(b) Insufficient software testing and rectification of software
problems before AOD.

(c) Insufficient training and practice of operators on software
functionalities.

(d) Lack of or late confirmation of stands.

(e) Lack of communication and coordination (i) within AA, (ii)
between AA and other parties, and (iii) between GEC, EDS
and Preston.

13.46 All the above causes are inter-related and cumulatively led to
the problems witnessed on AOD and the first few days of airport opening.
It is important to remember that each of the causes must be viewed in the
context of all the other causes rather than in isolation. There were other
minor contributing factors to the problems on AOD, such as monitors and
LCD boards malfunctioning, SAS hanging and FDDS not fully
338
performing.

(a) Compression of software development time

13.47 The compression of time is probably the single most
significant underlying cause of the problems encountered by FIDS on
AOD. Both W55 Kipper and W56 Shen, experts appointed by the
Commission, extensively considered the importance of the time factor.
The Commissioners agree with their opinion and highlight the following
salient points in their expert opinions and the evidence:

(a) The software development programme for FIDS was
probably an ambitious programme from inception, bearing in
mind the high degree of integration and sophistication in the
system design and the time available in which to develop it.
The contract period was about 30 months, from J une 1995 to
December 1997 if confidence trials are included, or about
two years from J une 1995 to J une 1997 if only the main
software development part is counted. Either way, it was a
very tight programme.

(b) The substantial time spent on agreeing on the SSS was, to
say the least, unfortunate. As a result of the delay, software
development had to start effectively from scratch in about
November 1997, some 17 months after the contract had
commenced. Despite the attempts to save time, including
the agreement to break up the software into separate builds
and combining the FAT with SAT, events proved that the lost
time was never recovered. As W56 Shen incisively pointed
out, history has proven that the time taken for the
development of the software and for the AOD version of
the software to work stably turned out to be very close to the
original plan, which had been so drastically shortened.

(c) In view of the tightness of the programme, AA and the
contractors should have been all the more vigilant to ensure
that no significant slippage would be allowed. However,
339
the evidence clearly shows that from J une 1997 onwards the
programme had been slipping. The revised substantial
completion date, ie Key Day 4, under the settlement
agreement of December 1997 was 6 March 1998 but was
again revised to AOD. However, according to W25 Mr
TSUI King Cheong of AA, that milestone had not even been
achieved by end of October 1998 when he gave evidence.

(d) With a tight schedule, AA ought to have realised that any
delay to the programme would be likely to cause
compression of the tail-end activities, ie, testing, training
and practice. However, the evidence shows that despite
protests and early warnings from AMD to Project Division
(PD) by way of W44 Heeds memo to W43 Oakervee
dated 4 April 1997, the development programme continued
to slip throughout the latter part of 1997 and 1998. W44
Heed had agreed to the progress programme suggested by
W32 Mr J han Schmitz of NAPCO in September 1997 that
there should be a minimum clear period of six weeks after
the system had been fully integrated and tested so as to
enable training to be undertaken. Although W44 Heed did
not agree that without that period of time for training and
familiarisation for the operators of the airport, problems on
AOD would be inevitable or that standard of service would
be lowered, he accepted that it would be prudent to have that
period of time. However, he told the Commission that there
was new development to FIDS and TMS functionalities and
workarounds in the last few days before AOD and TMS
became a usable system only in the last three or four days
before AOD. It is therefore clear that due to the
compression of time, standards were compromised.

13.48 AA suggested that the original progress programme for FIDS
agreed with GEC provided for confidence trials at the end of 1997, giving
a float of four months before the then targetted AOD of April 1998.
This suggestion is quite irrelevant because of the slippages experienced.
The system ought to have been fully integrated, tested and stable for a
340
suitable period of time before AOD so as to enable the operators to have
proper training and familiarisation with the functionalities, which was no
doubt one of the major reasons why the programme was originally
scheduled to complete by end of 1997. In a memo from W44 Heed to
W25 Tsui dated 4 April 1997, W44 Heed strongly protested against the
reduction of trial duration from three months to two months. However,
as W43 Oakervee put it, the three months clear trial period was
evaporating all the way through 1997.

13.49 In September 1997, NAPCO and AMD were both of the
view that a minimum clear period of six weeks after the system had
been fully integrated and tested was required to enable training to be
undertaken. W44 Heeds evidence is that this was already reduced from
the three months he had planned for in April 1997. There is no
conceivable reason to believe that the clear period should not equally
apply at any other time but September 1997.

13.50 Moreover, the first airport trial was planned to take place in
J anuary 1998, when the FIDS was used but effectively failed to work.
Nor did FIDS fare satisfactorily in the second airport trial on 15 February
1998. Thus the original objective of having the system development
substantially completed by end of December 1997 was not achieved, and
whatever float that was planned to have had gone under the drain.

13.51 For the compressed time for the development of the FIDS
software, which badly affected the efficient operation of FIDS on AOD,
AA, GEC and EDS should be responsible. The crucial slippage was the
loss of 14 months after the commencement of the contract C381. In the
Commissioners opinion, this was caused by the users requirements of
the operators of the new airport, ie, AMD, being ascertained too late to be
reflected in the original PTS used for contract tendering and in the
admitted ambiguity of the PTS. From the settlement agreement
resulting in $89.7 million payable to GEC, it appears to the
Commissioners that GEC and EDS were not responsible for this crucial
slippage. AA must therefore be solely responsible for it. However, for
the slippages from the end of 1997 to AOD, the evidence indicates that
both AA and EDS are responsible, although the evidence is not sufficient
341
for the Commission to make any proper apportionment of such
responsibility. This will remain a matter of contractual dispute between
AA and GEC, and probably between GEC and EDS, for there is no
priority of contract between AA and EDS.

(b) Insufficient software testing and rectification of software problems
before AOD

13.52 The slow system response was a major contributing factor to
the problems on AOD and the few days thereafter. Although EDS and
Preston hotly disputed the extent and inception of the slowness, all who
had given evidence before the Commission agreed that FIDS did respond
slowly to the commands of the operators. Mr C K Chan said that TMS
responded slowly when he was attempting flight swapping at about 2 am
on AOD, with a response time between 5 to 10 minutes. W24 Lee said
that when she started to do flight swaps at about 7 am that day, it took
from 20 seconds to 10-15 minutes to carry out a swap. According to her,
the response after the reboot at about 11 am was a bit faster, but not
significantly. She told the Commission that while the response time
varied, the vast majority of inputs, 80-85%, were affected by slow
response. There was consistently slow response to the entry of aircraft
registration numbers on AOD.

13.53 On the other hand, W35 Cumming of EDS said that he was
at PTB throughout the night between 5 and 6 J uly 1998, and he first
noticed slow response at about 6 am. W34 Derrick of Preston claimed
that while he was in ACC from 12:30 pm to 6:30 pm on AOD, the
response time was typically 3-4 seconds, up to 30-45 seconds in the
extreme, and he did not witness any command or operation taking 5, 10
or 15 minutes to execute.

13.54 While it may not be necessary for the Commissioners to
definitely accept one of these conflicting versions on the slowness of the
system on AOD, they consider that the evidence of the AA operators
should be preferred. The response time was not measured by anyone on
AOD with a stop watch, and the evidence from the opposing story-tellers
must have been based on estimates derived from memory of their feeling
342
on that eventful day. However, the following pieces of evidence, which
the Commissioners accept, lend strong support to their finding that FIDS
response time was too slow for normal or reasonable operation on AOD:

(a) EDS accepted in evidence that the system performance was
so slow before the reboot sometime after 10 am that it was
practically unusable.

(b) Other users of FIDS apart from the ACC operators reported
slow response. For instance, at around 8 am and 11 am, it
took 20-25 minutes for the FIDS workstation at AOCC to
allocate a baggage reclaim belt. This is consistent with the
statement of Mr Guy Gerard Summergood of EDS who
reported response time of up to 12 minutes from the FIDS
workstation in the BCR, and SESL representations that
AOCC took over the reclaim assignment from BCR because
it was taking too long for the BCR workstation to execute the
allocations. Cathay Pacific witnesses gave evidence that
the FIDS application on their CUTE workstations was
generally very slow throughout AOD and there was logging
on problems from around 5 am when Cathay Pacific opened
their check-in desks.

(c) Mr Rupert J ohn Edward Wainwright of EDS stated that CPU
usage started to increase from about 6 am on AOD although
the system was functioning. That would be about the time
when users terminal-wide began to use the system
intensively under live conditions. He also said that at about
8 am, he began to receive reports from AOCC that FIDS
MMI users had problems associated with the Oracle shared
memory allocation being too small. The problem slowly
increased to peak at around 10 am. Due to the problem and
the difficulty in getting the number of users on the system
reduced, he recommended shutting down the system to
change the shared memory allocation. However, though the
system was shut down, the change was not made due to a
Unix operating system parameter restriction. After the
343
reboot, Mr Wainwright said that the system performance
improved at first but began to slow down again as more users
logged back on again. This is consistent with the evidence
of W24 Lee and W26 Cheung who both said that the
response time improved a little after the reboot but not
significantly.

(d) Mr Stefan Paul Bennett of EDS said that while he was at
PTB at about 9 am on AOD, there were already shared
memory allocation problems. He also confirmed that the
CPU usage for AOD was and remained close to 100% and
the system performance on Day Two was similar as in the
afternoon on AOD. Both W55 Kipper and W56 Shen
advised that such high usage would cause performance
problems.

13.55 The PTS specifies that operators shall, without exception,
get an initial response to their inputs within 0.5 seconds and the final
response for 90% of updates shall be received within 2 seconds. This
shows beyond doubt that, even if W34 Derrick was entirely right with his
estimation of the extent of the slow response, the FIDS performance on
AOD was very far off the mark and could not reasonably be said to be
acceptable to AA.

13.56 The Commission accepts W55 Kippers explanations of the
technical causes contributing to the slow response in his expert report.
Briefly, the following were the major problems which plagued the FIDS
system performance on AOD and the few days thereafter:

(a) WDUM - The WDUM process is a core application
background process which defines which flight information
has to be sent to the FIDS MMI in the workstations to update
the displays. This process was identified as having the
most serious effect on system performance. There were
two problems with the WDUM:

(i) Excessive CPU utilisation. Mr Wainwright of EDS
344
stated that the CPU utilisation on AOD and Day Two
was on or near 100%. The system change to solve
the problem, Change 109, was carried out in the early
morning of 11 J uly 1998. However, performance
issues related to WDUM were identified more than a
month before AOD but remained open.

(ii) Deadlock problems. The evidence is that the WDUM
process was in deadlock with other user applications in
the ACC, AOCC and BCR workstations. Deadlocks
caused serious performance problems since they
prevented updates and blocked the operators operation
on FIDS workstations. Deadlock problems were
identified but had not been resolved before AOD.

Two workarounds were implemented to overcome the
WDUM problem: the MMI refresh rate was reduced from 6
seconds to 45 seconds in the early hours of Day Three, and
transactions were split to reduce the occurrence of deadlock
situations.

(b) Shared pool memory in the Oracle database - This error has
been described in detail in the witness statements of Mr
Rupert Wainwright (EDS), Ms Susan Wong (AA) and Mr Ian
Cheng (Oracle) and summarised by W55 Kipper in his report.
The shared pool memory parameters were changed in the
early morning of Day Two to enable the system to run faster.

13.57 AA obtained the services of Oracle Systems Hong Kong Ltd
(Oracle Systems) as consultants on Oracle matters, which related to not
only the database of AODB and other systems (for which GEC was not
responsible), but also the FIDS database. However, Oracle Systems was
only contracted to start work in late J une 1998. Oracle Systems made a
list of recommendations on 3 J uly 1998, but the recommendations came
too late to save AOD and even if carried out in time by AOD, would not
have cleared the performance problems with FIDS. Mr Wainwright of
EDS said:
345

Had the outstanding Oracle recommendations been
implemented between 3 J uly and 12 J uly, I do not consider
that performance or stability during this period would have
noticeably improved. Primarily this is because none of the
outstanding recommendations related to the ORA-04031 SGA
issue or deadlocks, locking, WDUM or alerts. Given that
these were the issues that affected performance and stability of
the system during AOD, I consider that none of the outstanding
recommendations from the 3 J uly 1998 list would have made a
significant impact on performance or stability.

13.58 Mr Wainwrights views succinctly bear out the causes that
affected performance and stability of FIDS on AOD.

13.59 Oracle Systems personnel were able to help identify the
causes on and after AOD, but of course, had AA obtained their services
earlier, the problems with the Oracle database of FIDS could have been
identified and rectified well before AOD.

13.60 Similarly, had there been sufficient testing of the FIDS
software prior to AOD, both the nature and the extent of the above
problems would probably have been identified and addressed.

13.61 The root of the problem probably goes back to the tight
timing of the programme. The settlement agreement reached between
AA and GEC in December 1997, whereby in order to catch up some of
the lost time it was agreed that the FAT and SAT would be combined and
to be carried out on site, was a dangerous move. In a fax dated 4
December 1997 from EDS to GEC, EDS stated that this posed a major
threat to the confidence in the final system. EDS was apparently not
consulted before the conclusion of the settlement agreement.

13.62 No doubt because of the time pressure and the consequent
inability to test the software properly before delivery to site, Build 2.0A
which was delivered to Hong Kong on 4 December 1997 was not as good
as expected. FIDS did not work during the first airport trial on 14
346
J anuary 1998. Nor did it perform satisfactorily during the second trial
on 15 February 1998 or, indeed, even at the third trial on 20 March 1998.
The result of the combined FAT and SAT from 23 February to 19 March
1998 was not as good as it might have been hoped. Two further rounds
of Re-SAT had to be carried out: the first Re-SAT from 1 to 7 April 1998
and second Re-SAT from 18 to 29 May 1998. Even then the tests were
not entirely satisfactory.

13.63 In the end, time ran out and the software had not even
reached a sufficiently stable state for proper stress and loading tests
before AOD.

13.64 There is overwhelming evidence that had there been proper
stress and load tests before AOD, of the actual load on AOD or 120% as
opposed to the design load of 200% of the expected load in year 2010, the
performance problems on AOD would probably have been identified.

13.65 AA argued that the problems with WDUM had already been
identified since early J une 1998, and a stress test would do no better.
However, the evidence shows that WDUM problems arose merely on a
few occasions out of numerous PRs identified in early J une 1998. W56
Shen said that if the WDUM problems happened only once or twice,
people had more important things than them to deal with. On the other
hand, a stress test might have detected the deadlock problems very often,
which would immediately raise the priority of attending to these problems.
Agreeing with W56 Shen, W55 Kipper further pointed out that a stress
test, even of a 100% of the actual load on AOD, would not only identify
the WDUM problems, but would have helped identify the reasons for the
problems, aiding in their rectification before AOD.

13.66 W45 Mr Kironmoy Chatterjee, the Head of AAs IT
Department, also accepted that a stress test would have thrown up the
problems haunting FIDS on AOD, in particular the parameters,
configuration and WDUM problems. Other AA witnesses such as W24
Lee and W26 Cheung said that a stress test would bring sufficient
confidence to them in FIDS and should have been carried out before
AOD. Nor did the specialist contractor and subcontractor gainsay the
347
importance of a stress test. W21 Korkowski of EDS told the
Commission that slow performance of FIDS would have been revealed if
the system was put under stress condition. He emphasised that the only
failsafe way of testing a system such as FIDS was to have the whole
system tested under stress conditions. His colleague W35 Cumming
agreed that stress and load test might have brought out the WDUM
problem. W22 Hobhouse of GEC said that even a stress test for one day
or half a day would have been better than none at all.

13.67 Much has been said about using the airport trials as tests
for the FIDS. No doubt, the trials provided some useful input as to the
performance and deficiencies of the system. However, the usefulness of
the trials as tests was limited in the light of the following
circumstances:

(a) There was not even one occasion when the entire integrated
system was tested as a whole. For example, TMS was not
used for optimisation even at the fifth trial on 14 J une 1998.

(b) Not all the daily routines had been tested in the trials. In
particular, flight swapping was never really tested in live or
even semi-live conditions in the trials.

(c) Even at the fifth trial, live data was not used. Instead, 114
flights from the AOD schedule were compressed into a
3-hour period for the trial. Without live data, the system
could not be tested under real live situations which, after all,
was the aim of the trials.

(d) Even after five trials, W26 Cheung, the Manager of Terminal
Systems, was not very confident of the FIDS prior to AOD.

13.68 It was because of the lack of time that the stress test was
deferred till after AOD. The reason was that the time available before
AOD was thought to be better used for resolving the stability problems
and ensuring Day One functionality that were of higher concern to AA
than the stress test. Assuming there had been sufficient time and proper
348
steps taken to rectify problems identified during the tests, the problems
that surfaced on AOD could have been minimised. This again
underlines how crucial the slippages in the development of FIDS were to
the problems on AOD. The responsibility for this has been discussed in
paragraph 13.51 above.

13.69 GEC and EDS as the contractor and subcontractor in the
provision of FIDS must be responsible for the inefficiency of the system
provided for use on AOD. The Oracle database was part and parcel of
FIDS and should also be the responsibility of GEC and EDS. Between
the two, EDS as the developer and supplier of FIDS, should take the
major part of the blame. As far as Preston (supplier of TMS) is
concerned, it is difficult on the evidence to decide whether TMS operated
unsatisfactorily or the inefficiency of TMS was caused by slow system
response and other problems that affected FIDS as an integral whole.
Only a little remedial action was taken on TMS after AOD, as opposed to
FIDS as a whole. Therefore, it appears to the Commission to be more
probable that TMS would not have had too many problems on AOD if not
for the problems of FIDS.

(c) Insufficient training and practice of operators on software
functionalities

13.70 Added to the insufficient testing was insufficient training or
practice of the software functionalities by the operators. The events on
AOD showed beyond peradventure that the training given to the operators
was not sufficiently thorough or adequate. However, this has to be
distinguished from the allegations of EDS and Preston that the problems
encountered on AOD lay more with the operators than the system
response. The allegations made by these subcontractors that the
problems were external to FIDS are summarised as follows:

(a) the use of TMS by ACC operators was not a designed
function, causing problems;

(b) the ACC operators and those of members of the airport
community were inexperienced and unfamiliar with how to
349
operate FIDS;

(c) the invalid ETAs sent by CAD to ACC severely impacted
TMS functionality; and

(d) the FDDS and AODB interface caused problems to
customers of HKT who subscribed for the FDDS service,
which problems had nothing to do with FIDS.

13.71 The Commissioners are of the view that the allegation that
the use of TMS by AA was not a designed function is but an excuse.
The evidence of W23 Lam and W28 Yuen is, and the Commissioners
accept, that the optimisation function of TMS was in fact used in the
afternoon of 5 J uly 1998, albeit not for the purpose of stand planning but
for the purpose of preparation for the TMS allocations to be made
consistent with the SAS allocations, followed by confirmation of the
allocations on TMS to be populated to FIDS displays. Input thereafter
into TMS was difficult and slow, and the stand information disseminated
by TMS/FIDS to the display devices became corrupted, incomplete or
inaccurate. The Commissioners are satisfied that the fault must be
attributed to TMS or FIDS and nothing else. First, the TMS
optimisation was used. Secondly, W21 Korkowski admitted in
testimony that input of flight information to alter the results of
optimisation on the TMS Gantt chart was within the normal operation
design of TMS, and was not an abuse. Since data could not be input or
confirmed or could only be done with delay, TMS or FIDS must be faulty.
The Commissioners accept, however, that the use of SAS in conjunction
with TMS instead of TMS did mean that the energy and time ACC
operators were further drained on AOD by the necessary input into TMS,
when they were already too busily engaged with the many problems
experienced.

13.72 The three representatives of GEC, EDS and Preston who
gave evidence before the Commission, namely W22 Hobhouse, W21
Korkowski and W34 Derrick, all agreed that FIDS including TMS was
experiencing slow response on AOD, although they denied that the slow
response was caused by the bug in the Oracle database. Some work to
350
improve the configuration of the FIDS database was performed in the
small hours of 7 J uly 1998, and some other work to enhance the memory
of FIDS was done on 11 J uly 1998. Anyhow, the slow response of FIDS
on AOD occurred, albeit intermittently, about 80% of the time on AOD
according to W24 Lee. The Gantt chart on TMS also shut down
intermittently through the day. The slow response and the shutdown not
only made the operations at ACC difficult to cope with the large number
of arriving flights that had to be dealt with on AOD, but also caused
doubts to the operators as to the effectiveness and functionality of TMS.
It was just natural, so said W24 Lee, that the operators repeated clicking
on a function when there was no response, invoked other functions or
restarted the computer, thus causing more problems. This should not, in
the opinion of the Commissioners, be treated as faults committed by the
operators. All the operators in ACC had worked in Kai Tak and those
who were supervising them were very experienced. They had also
undergone training on FIDS, TMS and SAS. W28 Yuen, for instance,
who supervised the operators in ACC, demonstrated his knowledge and
proficiency in operating TMS by explaining to the Commission how to
effect flight swaps on a TMS Gantt chart most succinctly and clearly.
Even W35 Cumming and W34 Derrick accepted that W28 Yuen was
proficient with the operation of FIDS. The Commissioners are satisfied
that the operators who worked under W28 Yuen and W29 Chan at ACC
were not as inefficient or unfamiliar with how to operate TMS as alleged
by the FIDS contractor, subcontractor and sub-subcontractor. If they
were in doubt, they had W28 Yuen to advise and help them. However,
even W28 Yuen experienced difficulty with flight swapping. When Mr
C K Chan told him that he could not perform flight swapping with TMS
at around 2 am, W28 Yuen tried to help. From the evidence it appears
that there were two things that W28 Yuen and his colleagues at ACC did
not know:

(a) When an operator entered the chocks-on time and
registration number of an aircraft on FIDS MM1, the system
would prompt the user on whether it should link the
registration number of the arriving flight to a corresponding
departing aircraft with the same registration number. W28
Yuen testified that the operators usually clicked yes simply
351
to avoid having to manually enter the registration number
later on, but were unaware of the causal effect of this, as
explained in (b).

(b) There were three methods of linking, which were progressed
such that an operator could not return to a method lower in
the progression once a higher method had been invoked.
The method of linking that operators were familiar with,
manual linking, was the second level of linking while linking
by registration number was the third level. This meant that
after W28 Yuen had invoked linking by registration number
with the prompt in (a), this inhibited flight swapping by
manual linking that he was going to do later on.
Furthermore, he did not know that the solution was to swap
the flights by aircraft registration number procedures.

13.73 The cause of this absence of knowledge was that the ACC
operators including W28 Yuen and W29 Chan were not trained how to
deal with these problems. However, the Commissioners do not accept
that this absence of knowledge due to lack of training is evidence of
inexperience. This lack of training was the result of the time planned for
training being hard compressed by the delays in the development and
testing and commissioning of FIDS, and the inadequate communication
and coordination within AA and between AA and others. Although the
prompt was not even known to W34 Derrick until AOD, the registration
linking was described in a document entitled Description of Flight
Linking Functionality (TMS Stand) Initial Draft dated 23/4/98, which
specifically mentioned that it would not be possible to manually link two
flights where one or both have progressed to aircraft registration link type.
W24 Lee confirmed in testimony that she was shown the document in
April 1998 but she did not show it to ACC operators and there was no
training given to them on matters contained in it.

13.74 The difficulty with flight swapping affected the operation of
TMS/FIDS in ACC on AOD significantly. W28 Yuen said that there
was no point for him to confirm stands for aircraft until he finished flight
swapping, in order to avoid wasting time and effort to confirm stands for
352
flights which would be affected by the swap. On top of the difficulty
with flight swapping, there was slow response of FIDS. While the stand
allocation was done manually by the use of placing paper stickers on a
whiteboard, the allocated stands would have to be entered into TMS
through FIDS MMIs, and that input was seriously affected by the slow
response. A large backlog of work began to accumulate when W24 Lee
arrived at ACC at about 6:30 am, and she helped with the flight swapping
and the confirmation of stand allocations into TMS. The slow response
was so bad that the backlog could not be cleared until the night between
Day Two and Day Three. This was despite the availability of assistance
from W24 Lee since 6:30 am and from W34 Derrick since 12:30 pm on
AOD. No one who gave evidence ever suggested these two persons did
not know how to do flight swaps or input stands into the FIDS MMI, but
the backlog was not cleared until more than 36 hours later. This
indicates that the slow response time was very serious.

13.75 The Commissioners are not satisfied that the problems with
FIDS on AOD was caused by the inexperience of the AA operators using
TMS and the FIDS MMIs, as opposed to their lack of training. Nor are
the Commissioners satisfied that the BCR operators employed by SESL
were inexperienced. Although Mr Summergood stated that he felt that
SESL operators in BCR were not familiar with the operation procedure,
that could not be the cause of problems inside that room on AOD.
Despite Mr Summergoods presence and assistance in BCR as early as
8:15 am, the delay experienced by arriving passengers with retrieving
their baggage did not quite improve the rest of the day. This leads one
to consider why even with the assistance of Mr Summergood, little
improvement was achieved. W26 Cheung told the Commissioners she
attended BCR at about 8 am in response to a complaint of slow response
of the system in the BCR. As a result, AOCC took over the operation
from BCR to assign reclaim belts at about 10 am. Even after FIDS was
rebooted at 10:45 am, the response time did not improve significantly as
Mr Summergood alleged. In fact as Mr Wainwright clarified in his
statement, no change was effected to the system at 10:40 am. He
aborted the proposed change at 10:45 am. The Commissioners have no
hesitation in preferring the evidence of W26 Cheung and Mr Wainwright
to Mr Summergoods.
353

13.76 W21 Korkowski also alluded to airline operators not being
familiar with logging out from FIDS through the CUTE workstations at
the check-in counters. He even went to the extent as to say that even if
everything in FIDS was working, the users would not have been able to
operate the system. The Commissioners find this hard to believe,
especially in view of the fact that at least some operators, of AA, SESL
and the airlines, must have participated in one or more of the airport trials.
Even if some of the operators using FIDS on AOD were new and had not
gone through the training or the airport trials, their colleagues who had
better knowledge and experience would have been able to help. The
experts from EDS and Preston all assisted in the working of FIDS on
AOD, but that did not help the operation in any significant degree.
There is not sufficient evidence to show that users unfamiliarity or
inexperience contributed towards the problems of FIDS on AOD to a real
extent.

13.77 W21 Korkowski admitted in testimony that one aspect of the
problems was that when the stand of a flight was confirmed on TMS, that
information would not be passed through to the display devices or would
be corrupted or incorrect when disseminated through FIDS to other parts
of the airport. There was a problem with data generation within FIDS
and that was made worse by the slow response of the system. The
response time was required by contract to be half a second, and obviously
that was not available on AOD. W24 Lee and W27 Ma, both IT Project
Managers, described the response time as taking from 5 to 20 minutes.
This was totally unacceptable, when the new airport was operating in full
swing as from a couple of hours before noon. The slowness simply
disabled AA staff from providing prompt dissemination of essential flight
information necessary for the users.

13.78 The details of CAD radar track provision of ETAs to ACC
can be found in paragraphs 10.18 to 10.20 of Chapter 10. The ETAs
from the radar tracker had not been screened before being automatically
fed into TMS. W34 Derrick complained that this had a severe impact on
the function of TMS. His evidence was that when he arrived at ACC at
12:30 pm on AOD he saw a sea of green on the Gantt chart on the
354
monitor. This appears to be consistent with W28 Yuens evidence that
about one half of the Gantt boxes were green. The green bars showed
the user that the times entered on the Gantt chart were invalid, as the ETA
for a flight was more than 15 minutes earlier than the STA. According
to W24 Lee, there were only several flights on the morning of AOD that
were affected by these invalid ETAs. Anyhow, the link between CAD
and ACC was turned off at 8:30 am, and there could not have been any
more invalid ETAs fed into TMS after that time. W34 Derrick was able
to clear the green phenomenon at about 2 pm. Whatever effect the
invalid ETAs had on TMS would have been cleared off by then. Yet
FIDS continued to suffer from slow response time through the rest of the
day. The green bars that occurred and remained before 2 pm cannot
detract from the fact that FIDS was not performing normally or
efficiently.

13.79 There was an allegation made by counsel for EDS in the
cross-examination of W27 Ma that the FDDS and AODB interface caused
problems to the customers of HKT who subscribed for the FDDS service.
Her response was that AA compared the information from AODB to the
FDDS internal table within AODB and there were no discrepancies found.
AA also looked at the system log, and there was no error message found
either. She was then questioned if the AODB database (as distinguished
from the database of FIDS) operated slowly on AOD, and she said that
there was no such sign. She admitted that she saw FDDS displays
frozen in the afternoon of AOD, but after reboot, the display and the
information were restored accurately.

13.80 In its response to the Commission dated 8 August 1998,
HKT accepted that there were some problems with its servers, resulting in
incomplete or incorrect information being displayed on the FDDS
monitors installed at its customers premises, and the customers needed to
restart their terminals to get refreshed information. That in the opinion
of the Commissioners does not, however, detract from the fact that even
if FDDS was operating without fault, the information which it derived
from FIDS through AODB was not reliable. If accurate and complete
flight information had been available from FIDS, the deficiency of FDDS
would only have caused some inconvenience to the customers and not the
355
extent of problems on AOD. The inconvenience could be translated into
the necessity of sending a person to obtain the required flight information
from the FIDS display devices, which would not drain too much time or
resources of the customers. Therefore the deficiency of FDDS cannot
reasonably be considered a major problem.

13.81 Returning to the training of the operators, the evidence is
that the operators were trained on versions of the software which were
under revision and that not all the functionalities were available during
training. W42 Mr NG Ki Sing, AAs General Manager of Terminal
Operations, told the Commission the effect of the encroachment on
training by the lateness of completion of the works. He said that
because the systems were not finished when training had to begin,
training had to be done on older versions or standalone versions. There
were two effects: first, the training was less effective, because the
operators might face a situation where what they learned today might not
be what they would have to do on AOD; secondly the operators might
need retraining on another version, resulting in duplication of time and
effort. The following are obvious examples:

(a) The aircraft registration table within FIDS was not available
until about 10 days before AOD.

(b) The operators were apparently not aware that they could
have populated the table by themselves via the FIDS MMI.

(c) The operators did not know how to properly respond to the
prompt that popped up in the FIDS MMI when registration
of an arrival aircraft was entered, and even W34 Derrick did
not know about this prompt before AOD.

(d) The operators did not know the implications of the
progression of methods of flight linking.

13.82 FIDS with TMS was clearly not error free, and even on the
first few days of operations W34 Derrick was in the ACC to devise
workarounds when problems arose. This shows that TMS was far from
356
workable by the operators on AOD or even the first few days. Even
Preston in its written response considers that testing and familiarity
compression meant that AOD was arguably 2-3 months too early for
systems ... to be completed.

13.83 The Commissioners find that that lack of sufficient training
of the operators was also a major contributing factor to the chaos on AOD.
The situation improved on Day Two and the days thereafter as the
operators no doubt took advantage of the baptism of fire and were more
familiar with system functionalities and workarounds.

13.84 The inadequacy of training of the operators cannot be
blamed on the operators, but rather like most of the problems experienced
on AOD was caused by the lack of time which was consequent upon the
slippages in the development of FIDS. AA must be primarily
responsible for the insufficient training provided to the operators. The
inability to make available for training the versions of functionalities to
be used on AOD may be the responsibility of AA vis-a-vis GEC, EDS
and Preston, but again, the Commissioners will not attempt to apportion
such responsibility.

13.85 Before leaving this topic, it is also relevant to mention the
problem with the crash of SAS which, according to the contractor City U,
was caused by the input of illogical data such as the departure time being
earlier than arrival time. The subject was not covered in the test
scenarios. Since the Kai Tak system, also developed by City U, was
able to handle such illogical data, there would be little reason for the ACC
operators to suspect that SAS could not accept them. It would not be
difficult to see why the operators were at a loss as to what the problem
was when SAS froze and then shut down. This was apparently caused
by the lack of coordination or understanding between AA and City U.
As to how the responsibility for this should be shared between AA and
City U, there is insufficient evidence to enable the Commission to decide.

(d) Lack of or late confirmation of stands

13.86 The Commissioners entertain no doubt that from the point of
357
view of operation, one of the major problems on AOD was the lack of or
delay in confirmation of stand allocations. The non-confirmation of the
stand allocation meant that the users and operators in the airport would
not get the information from the FIDS or FDDS displays. The early
availability of accurate stand allocation is extremely important to airport
operations. Two causes for the non-confirmation or late confirmation of
stands can be identified from the evidence: first, the difficulties
experienced by ACC operators with TMS, and second, the practice from
Kai Tak of not confirming stands until quite close to touching down or
even afterwards.

13.87 It is reasonably clear from the evidence that the problem was
linked to the difficulties experienced in ACC with TMS. To begin with,
before 2 am on AOD, the stands for the ferry flights, the first arrival flight
and the first departure flight on AOD had been confirmed. After Mr C
K Chan had performed the pairing of the flight movements received from
Cathay Pacific and Dragon Air, problems were experienced in the
swapping of some of the Cathay Pacific and Dragon Air ferry flights. It
would appear that from that point onwards, TMS could not be operated in
any significant way by the operators until W24 Lee arrived at the ACC
and began performing flight swapping.

13.88 W24 Lee gave evidence that about the time when she began
to carry out the flight swapping, at around 7:30 am, W26 Cheung from
the AOCC called her and asked for stand numbers. It was then when
she started to confirm stands on the TMS Gantt charts. W24 Lee
testified that she would allocate a stand on the TMS Gantt chart according
to manual allocation by her colleagues but would confirm only when
somebody called her that it was urgent, and she could barely handle it.
Even then, not all confirmations successfully passed through although
fortunately she knew a workaround to solve this particular problem, ie, by
un-confirming and re-confirming again until the confirmation went
through.

13.89 W35 Cumming, on the other hand, remembered that there
was no stand confirmation until about 9 to 10 am, just before the reboot.
W28 Yuen also recalled that confirmations were made at around the same
358
time, but he was not working at the workstations.

13.90 W28 Yuen also gave evidence that even after the reboot, the
system would take minutes to confirm a stand. With planes arriving at
about every 2 minutes, there was no way that the system could catch up at
that sort of confirmation rate.

13.91 W28 Yuen also told the Commission that the ACCs practice
on AOD was only to confirm stands after ETA had been received from
ATC. W23 Lam confirmed that the practice adopted for ACC operation
was only to confirm stands when ETA had been received or the ETA was
considered to be accurate and reliable. Thus there was a delay in the
confirmation of stands under the practice, exacerbated by the late receipt
of ETA. This practice was altered on Day Two to confirming stands an
hour before touch down, and the interval lengthened subsequently to two
hours before touch down. Anyhow, on AOD, with the radar tracker data
through the CAD link the ETAs were supposed to be available about 45
minutes before aircraft landing, but after the CAD link was switched off
at around 8:30 am due to the problems with corrupted ETA data, the ACC
had to rely on telephone calls from ATC to advise them of ETAs, and
from about 1 pm onwards, by fax. Such information only came from
ATC about 5-20 minutes before touch down. The ETA would then be
passed onto the AOCC by ACC staff who would input the information
into the system through the FIDS MMI.

13.92 In the meantime, in the ACC, many calls were received from
RHOs, airlines, ATC and AOCC, etc, and the ACC was very busy with
lots of phones ringing.

13.93 W28 Yuen further said that by about noon on AOD the apron
was full and incoming planes had to queue along the taxiway, waiting to
be directed to the first available stand, wherever it might be. W23
Lams evidence is somewhat different: he said that the apron was full by
around 1 pm and there were about 35 stand changes on AOD compared to
the normal 10-20 but he agreed that the problem with aircraft queuing for
stands on AOD was serious.

359
13.94 In any event, the situation on the apron must have been
chaotic to say the least. In those circumstances, it is not difficult to
imagine the tremendous impact and difficulties caused to RHOs and
airlines.

13.95 From the evidence, it appears to the Commission that the
major cause for the late or absence of confirmation of stands was the
unsatisfactory response time of FIDS with TMS, and the practice of late
confirmation of stands should also share part of the blame. However, it
is difficult to judge how effectively the input of confirmed stands into the
FIDS MMI would be or how effectively such confirmed stands would
have been disseminated, bearing in mind the slow response time and the
other deficiencies of FIDS on AOD. The alteration of the practice after
AOD, however, shows that the practice was not a good one, considering
the importance of sufficient advance knowledge of stands to operators of
the airport community such as RHOs in their serving of aircraft,
passengers, baggage and cargo.

13.96 It is not necessary to repeat where the responsibility should
lie regarding the inefficiency of FIDS and TMS, which is covered by
matters previously discussed in this chapter. AA, and in particular, W23
Lam, who decided to adopt the practice must be responsible for any
problem caused by the delayed confirmation of stands as a result of the
practice.

(e) Lack of communication and coordination

13.97 On the evidence, there were crucial communication and
coordination problems within and amongst the parties which to different
extents affected the situation on AOD.

(i) Within AA

13.98 There is considerable evidence of lack of communication
and coordination within the AA in the course of the software
development.

360
13.99 First, as far back as April 1997, AMD had already raised
their concerns and objections to the delay in the software programme and
the compression of the training and trial time. AMD's concerns were
not fully taken into account or properly addressed by PD. W42 Ng also
told the Commission that AMD was told, rather than consulted, as to the
appropriateness of the revised programme. Normally, there were
discussions between the two Divisions, but in most cases, the project
manager would have the final say as to how the project should be
programmed and proceed. When AMD put forward its comments and
concerns, some were addressed, but others were not. He gave the
examples of the confidence trials and the list of 38 PRs that W26 Cheung
produced as AMDs main concerns about FIDS functionalities, which
were not resolved before AOD.

13.100 The general lack of coordination between PD and AMD on many
matters was also well documented, both in the ADSCOM documents and
in the Booz-Allen Hamilton report, and this aspect is reviewed in Chapter
17.

13.101 Secondly, within the AA, different people apparently had very
different ideas of what was supposed to be done on important matters.
A telling example is that W26 Cheung thought that (i) stress tests were
supposed to be tests of the year 2010 loading of the system and therefore
could be postponed; and (ii) confidence trials were supposed to be carried
out after AOD. She was, of course, wrong on both counts. Yet, she
was the Terminal Systems Manager, and W44 Heed claimed that he relied
on her (and W42 Ng, who had no IT background), for deciding whether
the FIDS was to be used for AOD operation. Another important aspect
of the lack of communication was that W44 Heed did not know that PD
and IT Department had agreed with EDS to postpone the stress and load
tests because of insufficient time.

13.102 Thirdly, IT Department, principally W24 Lee and W27 Ma
realised that it would be important to have EDS support on AOD, the AA
management failed to ensure that proper support would be available to
those who needed it most or promptly.

361
13.103 It is clear from the evidence that those who needed support most
were the operators in the ACC, not so much because they were not
familiar with the systems, but because they had not been trained on using
TMS as an input tool and the system, at least according to W23 Lam,
W44 Heed and Mr Thomas LAM, Airport Systems and User Support
Manager of AA, was not sufficiently stable for use on AOD. Despite
this, W34 Derrick was only asked to make himself available from about 6
am on AOD.

13.104 To make matters worse, W34 Derrick could not get the necessary
permits to get to the ACC where he was most wanted. W27 Ma claimed
that she had liaised with the permits office and was assured that the
permit would be ready. In the end, W34 Derrick was not able to get to
the ACC until about 12:30 pm. Based on the evidence, if W34 Derrick
had been in ACC before 6 am, the situation on AOD would probably have
been considerably better.

13.105 On the same subject of providing assistance, a worse aspect was
that there was a most damnable mis-coordination between personnel of
AMD and IT Department. IT was supposed to assist AMD in the latters
operations. It is understandable that ACC operators did not know that
EDS personnel were on stand-by, probably because they had less personal
contact with the staff of the contractors. W28 Yuens evidence is that
when he had problems he would contact W24 Lee, but apart from the
only occasion at about 3 am when he was able to get her over the phone,
he lost contact with her. W24 Lee, however, told the Commission that
between 2 am and 6 am, she was checking the displays and CUTE
workstations in various parts of PTB where she might not have been able
to receive calls. This, if the Commissioners may say so, creates grave
doubts as to whether in fact IT was assisting AMD or if there was any
coordination at all between the two departments. W44 Heed said that
the ACC operators should have called the maintenance help desk which
was manned 24 hours, but when W23 Lam reported to him at about 5 am
and told him that he (Lam) could not get hold of W24 Lee, it did not
occur to W44 Heed to tell W23 Lam to call the maintenance desk either.

13.106 For such an important matter as the use of TMS to populate the
362
stand allocations downstream, the management of the AA ought to have
ensured that suitable backup support was available at all times. This AA
singularly failed to do.

13.107 Fourthly, although AA did seek advice from Oracle Systems on
database issues, such advice was only sought in late J une 1998. The
WDUM and shared pool problems would probably have been identified if
Oracle Systems had been consulted earlier. The Commissioners agree
with the comments of W56 Shen that It was a clear case of too little, too
late.

13.108 Fifthly, W26 Cheung who made a contingency plan for passing
information to passengers in the event of the failure of FIDS admitted that
there was no similar plan made in consultation with RHOs. She also
told the Commission that apart from her contingency plan for passengers,
there had not been a formal risk assessment on the possible failure of
FIDS. All this reflects a communication breakdown between AMD and
RHOs, the operators whose services and cooperation were required to
save the airport from chaos should FIDS fail. Apparently, according to
the contingency plan designed by W26 Cheung, whiteboards were
deployed as early as 7 am on AOD at the Departures and Arrivals areas to
provide passengers with flight, gate and reclaim belt information.
However, whiteboards were established at the Airport Emergency Centre
(AEC) for RHOs only at about 7 pm and not earlier, although the lack
of correct flight-related information had manifested at various quarters of
the new airport very early on AOD and AAs management had a meeting
to discuss the lack of flight information at around 10 am. Whiteboards
as a contingency had been included in materials supplied to the airport
service providers such as airlines, baggage handlers and RHOs, and they
were used, albeit in a limited way, at some of the airport trials.
Unfortunately, there was no detailed planning or procedure when
whiteboards would be employed, as manifested in their late use at AEC
on AOD. It is obvious that there was insufficient coordination between
AA and other members of the airport service community on contingency
operational procedures in the event of FIDS failure.

13.109 For the lack of communication and coordination within AA itself,
363
it goes without saying that the responsibility lies with the AA
management. The person who should take the most blame for the lack
of coordination and cooperation between AMD and PD is W3 Townsend.
On the other hand, the relevant divisions or departments within AA who
failed to communicate and coordinate adequately should be responsible
for such failure.

(ii) between AA and other parties

13.110 There are a number of areas in which the Commission finds the
communication and coordination between AA and other parties wanting.

13.111 First, the use of the automatic data feed from CAD without
screening. Although there was an agreement or understanding between
AA and CAD that the flight data from the CAD radar tracker should only
be used after authorisation or screening, AA took upon itself to use such
data on AOD without any authorisation. All would have been well if no
problem had occurred on AOD, or the data were accurate and complete.
Anyhow, there was no reason why AA should not have informed CAD
that AA was going to use the data without authorisation. Problems arose
when some of the ETAs from CAD through the radar tracker were
incorrect, creating green bars on the TMS Gantt chart which confounded
ACC operators working on the TMS Gantt chart. For this matter, the
Commission considers that AA should be responsible.

13.112 Second, the use of SAS rather than TMS as the primary
allocation tool. W21 Korkowskis evidence is that EDS was advised by
AA of the possibility that TMS might not be used as the stand allocation
tool. However, Preston (the subcontractor for TMS) was not informed
of the decision made by W23 Lam and W44 Heed of AMD and Mr
Thomas Lam of IT in J une 1998. Although W34 Derrick of Preston was
prepared to limit his criticism of the decision not to use TMS to a minor
one, the decision obviously would create more work for the ACC
operators. At the very least, EDS and Preston should have been
consulted on the risks involved. It is also pertinent to note W24 Lees
evidence that Mr Thomas Lam had asked her to inquire with EDS as to
whether it was possible not to use TMS at all, and the response of EDS
364
was no. AA should be responsible for the failure to consult.

13.113 Third, there was a significant failure on the part of GEC to notify
AA of EDS views on the revised programme in December 1997. The
evidence before the Commission is that EDS was not advised of the
precise terms of the programme and in fact objected to the representation
by GEC to AA that it would take about 10 days to revert from the
integrated mode to the development of the standalone builds. However,
W25 Tsui was only advised of this much later and W43 Oakervees
understanding was that the integrated build could be unraveled within the
space of two to three weeks, providing a relatively ready and sufficient
contingency. It came as a very rude shock to him in February 1998
that the point of no return had passed without his being really conscious
of it. It appears that the representation from GEC was a material
misrepresentation, which affected the AAs judgment on whether to
proceed with the integration mode. For this, GEC should be
responsible.

13.114 Fourth, there was a lack of communication or understanding
between AA and City U that contributed towards the crashing of SAS, as
discussed in 13.85. There is, however, insufficient evidence before the
Commission for it to decide how the responsibility should be shared
between the two.

(iii) between GEC, EDS and Preston

13.115 There was also a lack of communication and coordination
amongst GEC, EDS and Preston.

13.116 Between GEC and EDS: The most glaring communication or
coordination deficiency between GEC and EDS must be the failure of
GEC to seek EDS views on the revised programme in the settlement
agreement of December 1997, in particular as to whether the system
could revert to the standalone builds within a short time. The
discrepancy between GECs representation to AA and EDSs view
contributed to the AA managements mistaken belief that AA could
always have individual stand-alone builds as a fall-back if and when
365
there was any problem in the development and later operation of the
integrated build. For this, as said above, GEC should be responsible.

13.117 Between EDS and Preston: There was a prompt in the FIDS
MMI on which operators were not trained. When this prompt was
answered by the operator with yes, or as a default, the linking by
aircraft registration number would apply between the arriving flight and
the departure flight using the same aircraft. However, this would inhibit
flight swapping by manual linking procedures later on. Even W34
Derrick of Preston, who provided TMS, did not know about this prompt.
It seems obvious to the Commissioners that if a functionality or feature in
the FIDS MMI was going to affect operation of TMS, as between EDS
and Preston, they would and should have consulted with each other to
train operators on the implications and the correct method of usage before
the product was put into the hands of the user. The lack of coordination
here contributed to the problems faced by the ACC operators in the early
hours of AOD. There is, however, insufficient evidence for the
Commissioners to reach a conclusion as to the apportionment of blame
between EDS and Preston.

13.118 What remain to be mentioned in this chapter are the FIDS
display monitors. As 120 monitors were replaced within three weeks
after AOD, it is apparent that a majority of the monitors were defective.
The cable length and connection problems, which caused 10% or less of
the malfunctioning of monitors and LCD boards on AOD are minor
causes. FIMI was EDS sub-subcontractor providing the monitors, EDS
was responsible for their supply and commissioning at PTB, and GEC
was the main contractor. As far as AA is concerned, GEC must be
responsible for the defective monitors in its position as the main
contractor in providing the software and hardware of FIDS, and EDS
should be responsible towards GEC. The Commissioners do not feel
confident enough on the evidence before them to decide whether FIMI or
EDS should be responsible for the malfunctioning monitors. AA should
be responsible for the cable problems that resulted in malfunction or
inoperation of the monitors and LCD boards, but that only contributed in
a minor way to the chaotic situation created by the problems with the
FIDS software.
366

13.119 As far as the public is concerned, AA should be responsible for
failing to ensure that the FIDS, software and hardware included, would
operate smoothly and efficiently on AOD. Due to the problems with
FIDS on AOD, the users of the new airport, the passengers, airlines, ramp
service providers and baggage handlers had difficulties obtaining reliable
flight related information essential for their operations. This affected
them gravely. The Commissioners find that the inefficiency of FIDS
and TMS was the main reason for the start of the chaos in the new airport
on AOD. For this, AA must be primarily responsible. Its responsibility
was to ensure that FIDS, critical to the operation of the new airport,
would work as efficiently and smoothly as it (AA) had assured
Government. Looking from another angle, AA failed to have sufficient
regard to the efficient movement of passengers who were affected by the
deficient performance of FIDS, as it is required to do by the Airport
Authority Ordinance. GEC is also responsible for not providing an
efficient and smooth FIDS, while EDS, the subcontractor of FIDS, is also
responsible to the extent that it supplied FIDS. The Commissioners are
not able to decide if TMS would have worked efficiently if had not been
affected by the problems facing the whole FIDS, or to what extent
Preston (the supplier of TMS) should be responsible. AA should also be
responsible for not following the agreement that had been reached with
CAD in feeding the ETAs into AODB only after authorisation. The
corrupted ETAs from CAD caused green bars to affect almost one half of
the boxes in the TMS Gantt chart that made input into Gantt chart
difficult. Although the link between AODB and the radar tracker was
disconnected at 8:30 am, the incorrect and not fully reliable information,
as CAD had warned, caused interruptions to the TMS operations up till
about 2 pm on AOD. However, these interruptions were relatively
moderate and would not have resulted in all the problems encountered
with TMS, especially after the ETAs were removed from TMS before 2
pm.

13.120 Despite the insufficient preparation and knowledge that industry
practice tests had not been carried out, AA did not make any appropriate
assessment of risks or have sufficient contingency planning. SAS was
itself a contingency measure, which was planned to be used in the case of
367
failure of TMS. There was also a standby FIDS which would be put
into operation if FIDS failed. There was the whiteboard on which
necessary flight information would be displayed for various users in case
FIDS failed. Whiteboards were to be set up as a contingency measure at
Departures Airside Level 6 at the start of the Central Concourse with
flight information on departures written on them. Baggage reclaim
information was also to be on display on whiteboards set up at the
Baggage Reclaim Hall. However, there was little or insufficient
planning as to when these various contingency measures were to be put in
place in case the primary systems failed. For instance, W24 Lee, W21
Korkowski, W34 Derrick and W35 Cumming were all on standby since
the early hours of AOD. Problems with TMS and SAS started to surface
at about 2 am, but W24 Lee only attended ACC to assist at about 6:30 am,
and the assistance from W34 Derrick and W35 Cumming was not sought
by W24 Lee until shortly before that time. Had there been a proper
assessment of the risk involved, these people who were most familiar
with FIDS should have been asked to standby close to if not inside ACC.
W34 Derrick could not even access ACC as soon as he made himself
available because there were problems getting a permit for him to access
ACC on the airside which took hours to resolve. These deficiencies in
planning and risk assessment as well as the instances of failure or lacking
in communication and coordination must be the responsibility of AA and
it alone.

CHAPTER 14

RESPONSIBILITY CARGO HANDLING

Section 1 : The Development of the Cargo Terminal Operators at
the New Airport

Section 2 : Causes for the Problems on AOD - AAT

Section 3 : Cause for the Problems on AOD HACTL
(a) The Alleged Causes
(b) The Opinions of HACTLs Experts
(c) Dust
(d) The Main Causes


Section 5 : HACTLs Best Endeavours Basis

Section 6 : HACTLs Attitude in the Inquiry

Section 1 : The Development of the Cargo Terminal Operators at the
New Airport

14.1 Different from the situation in Kai Tak where there was only
one cargo terminal operator (CTO) which was Hong Kong Air Cargo
Terminals Limited (HACTL), there were two CTOs in the new airport,
with the additional Asia Airfreights Terminal Ltd. (AAT). The cargo
handling capacity of AAT and HACTL was eventually to be respectively
420,000 tonnes and 2,600,000 tonnes per annum, giving the new airport a
full capacity of about 3 million tonnes a year. However, under AATs
franchise agreement dated 12 J anuary 1996, AAT was to be ready for
handling 1,100 tonnes per day in March 1998, whereas under HACTLs

368
franchise agreement dated 18 August 1995, HACTL would have to
provide 5,000 tonnes per day, equivalent to about 75% of its throughput
capacity on 18 August 1998. HACTL would therefore have 80% of the
market share of cargo handling, while AAT the remaining 20%. The
actual market share of AAT was even less as AAT was planning to operate
at approximately 65% of its total capacity (ie. 280,000 tonnes per annum)
and had planned its manpower accordingly. AAT serves about 12
airlines and HACTL serves about 50 in handling cargo.

14.2 AAT is a joint venture company which was incorporated for
the purpose of tendering for a franchise to operate an airfreight terminal
at the new airport. Its largest shareholders are Singapore Airport
Terminal Services Private Ltd and Changi International Airport Services
Pte Ltd, and they have between them more than 50 years of experience in
the air cargo handling industry. Works commenced in April 1996 to
build the terminal on a site of about 40,000 square metres, consisting of
the terminal building itself and a service road, parking lots for a minimum
of 83 vehicles and 54 trucking docks. The terminal building occupies
the best part of the land. The Material Handling System (MHS) is the
key component of the terminal, costing $190 million to design and build.
MHS is fully automated and provides storage and retrieval functions for
bulk and pre-packed cargo. Within MHS there is a 12-level automated
storage and retrieval system with 1,320 storage positions and a 7-level
pallet container handling system with 734 storage positions. Other
equipment or facilities include equipment transfer vehicles and unit load
device (ULD) equipment, storage racks for small shipments and heavy
shipments, tractors and forklifts. There are also 34 workstations for the
build-up and breakdown of cargo. Computers are also used extensively
for both operations and in the processing of documentation. Occupation
permit (OP) for the terminal was issued on 9 J une 1998, later than the
contractual date for readiness to handle 1,100 tonnes of cargo per day.
However, there were no significant adverse consequences save for some
disruption to AATs on-site training schedules.

14.3 HACTL was incorporated in 1971 with J ardine Pacific Ltd
and Swire Pacific Ltd owning respectively 25% and 20% of its shares.
HACTL had been operating as the sole CTO in Kai Tak since 1976.
There were two terminals in Kai Tak. Terminal 1, which was originally

369
designed to handle up to 350,000 tonnes of cargo per annum, was
eventually able to handle 750,000 tonnes with an extension in 1984 and
other renovation work in 1998. Terminal 2 was commissioned in 1991,
providing an additional throughput capacity of 750,000 tonnes every year.
The total capacity was therefore 1,500,000 tonnes per annum, although
HACTL was able to handle 1.7 million tonnes in 1997. Community
System for Air Cargo (COSAC), HACTLs mainframe-based
proprietary system, was first introduced in 1976 and it had eversince
served the needs of HACTL, its customers and the community very
successfully. COSAC was improved in four phases, with the first three
implemented at Kai Tak between 1994 and 1998. The final phase
included applications that were only applicable to SuperTerminal 1
(ST1). W7 Mr Anthony Crowley Charter, the managing director of
HACTL, stated that No other air cargo handler has such a
comprehensive and established system or equivalent expertise in systems
development. HACTLs service standards at Kai Tak were maintained
at an extremely high level with a mishandling rate of merely 1 in 21,000,
achieving that through automation from the previous mishandling rate of
1 in 7,000. This record can be better appreciated when compared with a
mishandling rate at the point of origin of incoming consignments of 1 in
22. The average dwell times also compared extremely favourably with
other air cargo terminals in the world: about 19 hours for exports and 27
hours for imports, and resulted in HACTL achieving an enviable
reputation for efficiency and reliability.

14.4 ST1 is designed to handle 2.4 million tonnes of cargo every
year. It is a 6-storey building, 290 metres long and 200 metres wide,
providing total floor space of 274,000 square metres. The sheer size of
the building requires automation to organise it to cope with the scale of
the operation. Cargo Handling System (CHS) that has been installed
inside ST1 consists of five levels, which are briefly COSAC 2, Resources
Management System (RMS), Logistic Control System (LCS),
Programmable Logic Controller (PLC) and mechatronics. The details
can be found in Chapter 11. The mechatronics comprise Container
Storage System (CSS) and Box Storage System (BSS). Two
identical parts of CSS are situated at the west and east sides of ST1, and
each of the two CSS has six stacker cranes operating within a single aisle,
providing full redundancy capability in the unlikely event of stacker crane

370
failure. The stacker cranes operate the full 36-metre height of the
building, providing fast and simple container storage and distribution.
The two CSS provide more than 3,500 storage positions for containers.
BSS is located in the centre of ST1 and divided into North BSS and South
BSS. Individual consignments are held prior to build-up in export
containers, or after breakdown from import containers. Almost 10,000
storage boxes are provided in the twin BSS rising the full 36-metre height
of the building. Each BSS has six aisles, with two stacker cranes
serving each aisle allowing for full redundancy capability in the event of
stacker crane failure. A total of 24 stacker cranes operated in BSS.
CSS and BSS had altogether about 15,000 sensors and reflectors built in
for the purpose of processing cargo and safety. The design of CSS and
BSS is modular in nature, which means that each portion of the systems is
able to work independently so that failure of one portion would have little
effect on the capability of the others. The throughput capacity of the
main ST1 building is to be 2,400,000 tonnes of cargo per annum.
Adjacent to the main building is the Express Centre, which is dedicated to
the special handling needs of integrated carriers and express and courier
operators, and is designed to handle up to 200,000 tonnes of express
cargo a year.

14.5 As with most construction contracts, the development of the
buildings of both CTOs was delayed. While the delay regarding AATs
terminal had much less impact, the building construction slippages
regarding ST1 were very substantial. An agreement to accelerate the
works was entered into between HACTL and its main building contractor,
Gammon Paul Y J oint Venture (GPY) in April 1998. However, this
was not able to catch up with all the delays already suffered. The
installation, testing and commissioning of CHS inside ST1 were
consequentially delayed, and there was much less time for HACTL to
train its staff and get them familiar with the new environment in ST1 and
in the operation of CHS. AAT was able to obtain an OP for its terminal
on 9 J une 1998 while HACTL had a temporary occupation permit
(TOP) issued for ST1 on 3 J uly 1998, just in time for airport opening
day (AOD) on 6 J uly 1998.

371
Section 2 : Causes for the Problems on AOD - AAT

14.6 The details of the problems witnessed at AAT and HACTL
are set out in Chapter 11. This section will list the alleged causes for the
problems and analyse the evidence to find out what the true causes were.

14.7 W4 Mr SEE Seng Wan, the Chief Executive Officer of AAT
alleged that the problems encountered by AAT on AOD were mainly
caused by the following:

(a) delay in cargo delivery;

(b) lack of co-ordination between AAT and ramp handling
operators (RHOs) and amongst RHOs;

(c) problems in relation to the receipt of flight information
through Flight Information Display System (FIDS) or
Flight Data Display System (FDDS);

(d) effect of shutdown of ST1; and

(e) adequacy or otherwise of dollies.

14.8 W4 See stated that AAT opened on 6 J uly 1998 with no
major technical difficulties with its MHS and computer systems.
Although there were some glitches with MHS and the computer systems,
they were relatively minor and amounted to little more than teething
problems that AAT expected to experience on opening a new facility.
The most significant problem was that on AOD and a few days thereafter,
an enormous backlog of cargo was built up which led to a heavily
congested working environment, both within the terminal and on the
ramp interfacing with it. The backlog seriously hampered the
processing of daily inbound cargo leading to a snowballing of
unprocessed cargo. It was not until arrangements had been made with
Airport Freight Forwarding Centre (AFFC) for transferring the backlog
to AFFC for processing from 18 J uly 1998 that the congestion subsided.

14.9 The problems facing AAT were not serious as compared with

372
those experienced by HACTL. Most of the causes alleged by AAT are
similar to some of those raised by HACTL and will be dealt with in the
latter part of this chapter. It suffices to state that the Commission
considers that the main causes of AATs difficulties on AOD were the
following:

(a) inadequate co-ordination between AAT and RHOs in the
hand-over of cargo from RHOs to AAT; and

(b) AAT staff were not too familiar with handling live loads of
cargo in the new environment and using new equipment
because they did not have adequate training.

14.10 Although AAT alleged that it was in possession of only a few
hand-over forms signed and presented by RHOs on AOD and a few days
after, during cross-examination of W4 See, RHOs produced over 10
duplicate hand-over forms in their possession, and W4 See was unable to
deny that those were signed for receipt by AAT staff. In the
Commissions view, the few hand-over forms kept by AAT does not
demonstrate that AAT staff only received a few pieces of cargo from
RHOs as AAT would try to portray, but rather that AAT staff did not even
keep copies of the 10 odd hand-over forms whose duplicates were kept by
RHOs. Coupled with this evidence, there was an admission from AAT
that the interface between AAT and the ramp was filled with a large
backlog of cargo, and that it needed till 13 August 1998 to clear
completely with the use of AFFC. There must have been something
wrong with the ability of AAT to handle the cargo on the ramp, or else the
backlog would not have taken so long to disappear.

14.11 There is also evidence that training of AAT staff was started
only immediately after OP for AATs building was obtained on 9 J une
1998, which was less than a month before AOD. The combined effect
of these two main causes, in the opinion of the Commissioners, resulted
in a huge backlog of cargo building up at the interface between the ramp
and AATs terminal. For its staffs inadequate training and unfamiliarity
with the environment and equipment, AAT must be responsible. In
respect of the inefficient co-ordination between AAT and RHOs, the
Commissioners consider it more probable that AAT staff were too busy

373
with their work inside the AAT building, and that they were not readily
available at the interface area to receive cargo from RHOs. RHOs were
driving tractors pulling dollies of cargo behind them, and in the
hurry-scurry of AOD, if AAT staff were not at the interface area ready to
receive cargo straight away, as the Commissioners infer should have been
the case, they would simply leave the cargo there. AAT is therefore also
found to be responsible on this score, although RHOs should share a
small part of the responsibility.

Section 3 : Cause for the Problems on AOD HACTL

(a) The Alleged Causes

14.12 The finding of causes of the breakdown of ST1 is much
more complicated. HACTL mentioned software and hardware problems
of its CHS immediately after AOD when it announced embargoes to
restrict imports and exports save for perishables and urgent items.
However, from the start of the inquiry, HACTL alleged a number of
causes which were mostly unrelated to software and hardware problems
with its computer systems. It eventually called two experts to attribute
the breakdown mainly to causes external to CHS. On the other hand, Dr
Ulrich Kipper, an expert for the Commission, also dealt with the causes in
his report in great detail which were, more often than not, different from
the findings of the HACTLs experts. The Commission will examine the
majority of these alleged causes to reach its conclusions.

14.13 On AOD, HACTLs ST1 failed. There were enormous
quantities of cargo scattered at the northern part of ST1 and the
surrounding areas, and on the next day, HACTL made a public
announcement that there was to be an embargo of all inbound cargo. It
is therefore undeniable and not denied by HACTL that ST1 was shut
down for a considerable period after AOD. The event was variously
described as a paralysis, collapse, breakdown or crash of CHS at ST1.

14.14 HACTL stated that ST1 was ready for operation on 3 J uly
1998 when the TOP for the building was issued. As from that day,
outbound cargo were accepted at ST1 for processing before they would
be exported on flights using the new airport. By AOD about 2,000

374
containers had also arrived at ST1 from Kai Tak. These containers had
been moved from Kai Tak to ST1 mainly in the night of 5 and 6 J uly
1998. Some of them were with outbound cargo but most of them were
empty. Outbound and inbound cargo were handled on AOD, but
eventually an embargo was announced by HACTL at 3 pm on 7 J uly
1998. HACTL alleged that there were many factors contributing to the
collapse of ST1, and the major factors can be summarised as follows:

(a) delays in the completion of construction works resulting in
delays in the installation of its machinery, systems and other
facilities;

(b) the delays in (a) also resulted in delay in the testing and
commissioning of HACTLs machinery and systems and in
the training of its staff for operating the machinery and
systems;

(c) the late completion of the construction and related works
(such as fit-out and decoration works) caused contamination
of the environment in ST1 beyond the level expected,
especially in relation to the air-borne dust, that affected the
operation of the machinery;

(d) the circumstances created by the fact that unlike in the Kai
Tak airport where there was only one CTO, i.e., HACTL and
one RHO, i.e., Hong Kong Air Terminal Services Ltd.
(HATS), there were two CTOs, being HACTL and AAT,
and three RHOs, namely, J ardine Air Terminal Services Ltd.
(J ATS), Hong Kong Airport Services Ltd. (HAS) and
Ogden Aviation (Hong Kong) Limited (Ogden), in the new
airport, thus requiring more co-ordination and understanding
between the CTOs and RHOs which was not readily
forthcoming;

(e) RHOs were not too familiar with the geography of the new
airport, the facilities available, and the work required of them,
and they did not follow the procedures, in particular the
hand-over procedures, agreed to be used at the opening of the

375
new airport, thus causing problems;

(f) dollies at the new airport for carrying cargo between ST1 and
aircraft were insufficient; and

(g) FDDS or Flight Display Data Feed Services (FDDFS) was
not providing flight-related information to ST1 as expected or
at all, causing trouble or inconvenience in the operation of
cargo handling.

14.15 It can be noted that all the alleged causes hardly rest on any
fault on the part of HACTL, and if they are established, blame should be
attached to other parties. HACTL, however, did mention
computer-related causes for the breakdown, e.g., CSS-LCS problems and
data mismatch, etc. However, the causes summarised in the preceding
paragraph would have the effect of downplaying the causes internal to
ST1.

14.16 HACTL retained two experts, namely, W52 Mr Max William
Nimmo and W53 Mr J erome J oseph J r. Day. These two experts jointly
produced a report dated 14 November 1998 and also gave evidence
before the Commission. They opined that there was nothing wrong with
CHS and that the throughput capacity of the system was available and
capable to deal with the cargo load on AOD. There was little evidence
in support of this opinion except their stating that they had examined the
system and questioned personnel of HACTL who provided answers
consistent with their findings. They attributed the breakdown of CHS to
various causes which can be divided into external and internal ones,
although they were sometimes intertwined. The internal causes were
those happening within ST1 and related to the operation of HACTL
within ST1, not being caused by any factor outside ST1, whereas the
external causes were those not generating from within ST1 that caused or
substantially contributed to the breakdown. The causes are summarised
below, and whether each is external or internal is included at the end of
the item in square brackets:

(a) The absence of information feed from FDDFS. Two
alternative ways were employed by HACTL in an attempt to

376
overcome this difficulty: (i) obtaining information from
FDDS terminals and (ii) sending staff to the Passenger
Terminal Building (PTB) to read the FIDS display devices
and relate the information through mobile radios to HACTL.
Both of these methods were useless as the displayed
information proved to be either lacking or incorrect.
[external cause]

(b) As RMS was not receiving accurate information from FDDS
through COSAC or through other means, RMS was unable
to provide useful information to LCS, and therefore it was
disconnected. [external cause]

(c) Confusion on the part of airlines about how they were
supposed to use the customs clearance system of the
Customs & Excise Department (C&ED) resulted in delays
in customs clearance notification being received from the
C&ED for automatic input into COSAC via HACTLs Air
Cargo Clearance System (ACCS). [external cause]

(d) Cargo pre-manifests from airlines and shippers arrived late
causing time pressure on the operational staff. [external
cause]

(e) On AOD, faults occurred in the handling of inventory
records at LCS level, and operators were unable to
sufficiently keep COSACs inventory records manually
synchronised with LCS records or the actual situation in
CSS, resulting in serious inventory mismatches that reduced
the overall integrity of the inventory records to an
unacceptable level, and consequently slowed the operation
of CHS as a whole. [internal cause]

(f) The inventory adulteration was caused by the operators
switching from the automatic mode to manual mode of
operation of CSS, and they were not familiar with operating
in manual mode. The reasons for such a switch were that (i)
the equipment was defective; (ii) there were problems

377
arising from live load operations, such as mis-shaped ULDs
running over to the side of a roller bed, causing ULDs to jam
or causing limit switches to stop transfers, or roller beds
mis-aligned causing ULDs to run off to one side, or pieces of
polyethylene wrapping drooping down from ULDs and
blocking light curtain sensors; and for any of such
occurrences, the operator would put the equipment off-line
to manually re-position the ULDs; and (iii) faulty or dirty
sensors producing incorrect interruptions. [internal cause]

(g) Operators wrongly perceived that CHS was running slowly.
This was caused by the operators not knowing clearly how
LCS-CSS operated in automatic mode. When an order to
move cargo was keyed into a workstation, the required cargo
movement would not commence until LCS could schedule a
complete end-to-end cargo movement. In order for the
movement to start, there had to be no equipment on the
movement route off-line or unavailable for LCS to schedule.
This means that movement would seldom commence
immediately, whilst waiting for a complete transfer route to
become available and to be scheduled. However, once the
movement began, it would proceed very quickly end-to-end.
This routing and reachability check function of LCS
would cause the system to appear to be slow to the operator
who would then switch to manual mode. [internal cause]

(h) The congestion at the ramp interface with ST1 impeded the
identification of ULDs. The extreme pressure to release
cargo dollies adversely affected the manual data entry
processes of opening ULD initialisation records in COSAC,
in preparation for loading the ULDs into CSS. This data
entry process was necessary as it enabled the association of
ULD identity information ultimately to consignment
information previously received from the airline shortly after
the aircraft took off from its foreign departure point. It was
also necessary as it would create a computer record of the
ULD location in CSS or the terminal. As RHOs did not
follow the agreed hand-over procedures, partly caused by the

378
absence of FDDFS and FDDS, it was a very slow and labour
intensive process for HACTLs staff to identify ULDs with
the flight on which they arrived. The amount of cargo
mis-handling on AOD simply overwhelmed the capacity of
HACTLs ramp supervisors to keep up with the
ever-deteriorating situation. They could not even get RHOs
to park dolly trains in a fashion that did not block the
removal of other dolly trains that were ready to be moved to
CSS airside. Due to the chaos on the ramp, it was also not
possible for HACTLs ramp supervisors to reach identified
perishable cargo and move it into ST1 for delivery to waiting
trucks. [external cause]

(i) Mistakes such as the following, namely, (i) passenger
baggage being delivered to HACTL; (ii) airmail being
delivered to HACTL; (iii) cargo for ATT being delivered to
HACTL and vice versa; (iv) cargo being delivered to
HACTL on passenger baggage carts which could not be used
for the transfer of ULDs to HACTLs CSS conveyor
machinery; and (v) incomplete cargo loads being delivered
to HACTL. [external cause]

(j) Errors and omissions in the inventory caused by operators
input necessitated by the manual mode of operation
interfered or stopped the operation of LCS in carrying out
cargo movement orders. This resulted in more manual
mode operations, giving rise to a vicious circle. There was
a dire necessity for the inventory to be corrected, and a
decision was made to make a manual inventory check in the
early hours of 7 J uly 1998. However, during the process of
the inventory check, inventory records were inadvertently
deleted due to a human error. As the reason for the deletion
was not known at the time, HACTLs management and
Control Systems Development Group (CDG) suspected
that there might be some fundamental defect in the computer
software, and their confidence in the computer systems was
shaken. It was utterly unpredictable how long it would take
to locate and solve the problem, and a move back to Kai Tak

379
offered the only realistic hope of recovering the situation.
[internal cause]

14.17 To support its case that HACTL should not be criticised for
the breakdown of CHS on AOD, HACTL made the following points in
their closing submissions to the Commission:

(a) HACTL committed additional funds in the Supplemental
Agreement with GPY to accelerate the building works so as
to complete the necessary works ahead of the contractual
completion date of 18 August 1998.

(b) The best endeavours basis had implicit risk involved.

(c) Both Government and AA simply relied on the oral and
written assurances given by W7 Charter in May and
mid-J une 1998 that HACTL would be ready, although
knowing full well that HACTL was facing immense pressure
to complete ST1 by AOD and that serious delays with the
construction work had occurred, which presented a risk of
HACTL being not able to complete ST1 on AOD due to the
overlapping of construction works, testing and
commissioning of HACTLs CHS and training programmes
for HACTLs staff.

(d) The delays in the construction works at ST1 had a significant
impact on HACTLs state of readiness throughout the
construction phase of ST1. In addition, the overlapping of
these construction works with the installation of the
mechatronic cargo handling machinery also disrupted the
commissioning and testing of the entire CHS and also
interfered with HACTLs training programmes.

(e) Disruptions to installation, testing and commissioning of
CHS and to training of HACTLs staff were caused by:

(i) Delay in construction works by the GPY, the main
contractor in the construction of ST1;

380

(ii) Failure of YDS Engineering Ltd. (YDS) to complete
its first fix works relating to ducting and electrical
trunking;

(iii) Delay in providing permanent power supply by GPY;

(iv) Instability of temporary power supply for
commissioning CHS on the west side of ST1;

(v) Five occasions of power interruptions, and three of
these due to leaking water and the other two due to
poor workmanship by YDS (loose or faulty
connections);

(vi) Floor slaps constructed by GPY not up to
specifications so that the floors in various areas were
not even or level: some floors had to be made even,
and sometimes Murata Machinery (HK) Ltd.
(Murata), the contractor for BSS, had to and did
raise the level of the footing of BSS;

(vii) Grinding works required to rectify the uneven floor
creating a lot of dust because GPY did not use wet
grinding method;

(viii) The constant failure of GPY to keep water out of ST1
as it should, adversely affecting the equipment and the
installation of CHS, and resulting in continuous
complaints lodged by Mannesmann Dematic AG
Systeme (Demag), the contractor for CSS, and
Murata;

(ix) The overlap in construction and installation works
which was a result of Government deciding 6 J uly
1998 to be AOD and GPYs delays in the construction
works. Installation of CHS equipment had to take
place before the building was sealed. As

381
construction works were carried on in other parts of
the new airport, the equipment was subject to severe
contamination problems;

(x) Leaking from the roof of ST1;

(xi) The wet weather in the few months before AOD
which damaged the equipment;

(xii) GPYs suspended ceiling contractor, Companion Ltd,
delaying its suspended ceiling works, which seriously
delayed the completion of the fire services
installations for statutory inspections;

(xiii) GPY leaving piles of uncontained sand which spread
throughout the floors and onto and into the equipment;
and

(xiv) Logistical and practical problems as a result of the
large number of workers of various disciplines in ST1.

(g) Flooding by a fire services drencher on 2 J uly 1998 putting
an elevating transfer vehicle and the inner scissor lifts in the
Express Centre out of service on AOD, causing a loss of
some cargo handling facilities to ST1.

(f) Normal day-to-day operations. Today, CHS at ST1
experienced about 200 equipment interruptions a day, whilst
at the same time, performing 60,000 movements, dealing
with 4,000 tonnes of cargo. This is roughly comparable to
the situation of the CSS at the Kai Tak Airport. The
equipment interruptions on AOD were therefore nothing
other than normal and do not indicate that CHS was not
operational.

(h) HACTL was unable to have AOD deferred as it believed that
it was irreversible.

382
(i) There should have been a soft opening for air cargo
operations, but that idea had been rejected by AA and CAD.

14.18 W55 Dr Ulrich Kipper, one of the experts for the
Commission, classified the causes in a very systematic way. He first
assigned the problems identified in the inquiry related to CHS to various
problem areas. He then categorised the problems as initial and
consequential problems, consequential problems being the effect of initial
problems, generally following a sequence of events whereby it can be
recognised that the effect of a particular problem was the cause of another
problem. He also explained the different meaning between a
snowballing and spiralling effects. The phenomenon that one initial
problem is causing multiple consequential problems is described as a
snowballing effect. A spiralling effect is resulted when the effect of a
particular consequential problem is linked with a previous problem,
forming a problem chain (feedback loop). Additionally, the resulting
effect is increasing with each cycle. According to W55 Kipper, due to
the complexity of airport processes, error propagation can be
characterised as a combination of spiralling and snowballing effects. In
order to ensure a smooth airport operation it is most important to keep the
initial problems under control. Once a combination of spiralling and
snowballing effects is established, it is an extremely difficult and
long-drawn-out process to return to normal operation.

14.19 W55 Kipper identified the following problem areas (Px):

(a) P1 : ST1 cargo operations,
(b) P2 : ST1 building and environment,
(c) P3 : CHS software (levels 2 to 5), including commissioning
and testing,
(d) P4 : CHS machinery (levels 1 and 2), including
commissioning and testing,
(e) P5 : training of HACTLs staff, and
(f) P6 : risk assessment, contingency and system fallback
capabilities.

14.20 W55 Kippers categorisation of the identified problems is as
follows:

383

Caused
By
(I)
Initial/
(C)
Conseq.

Identified Problem

Px
- I1 CSS live ULD irregularities

P1
- I2 BSS live load irregularities

P1
- I3 Excessive flight delays

P1
- I4 New C&ED system and procedures (for cargo
handling)

P1
- I5 Dirt on CSS mechatronics

P2
- I6 Dirt on BSS mechatronics

P2
- I7 Lack of marked interface area on the ramp on
the northern part of ST1 for RHOs hand-over
of cargo

P2
- I8 Hostile (ST1) building environment

P2
- I9 CSS-LCS software bug (errors)

P3
- I10 BSS-LCS software bug (errors)

P3
- I11 Electromechanical teething problems

P4
- I12
New and unfamiliar operating environment
(for HACTL and RHO staff)

P5
- I13 Insufficient training of HACTL's operational
and maintenance staff

P5
I14 Interchange Server (IS) as stand-by system P6

384
Caused
By
(I)
Initial/
(C)
Conseq.

Identified Problem

Px
for LCS was not operational on AOD

I1, I9, I10,
I11, I12,
I13 C4, C9
->

C1 Manual operation (of CHS) P1
I3, I7, C4-> C2 Ramp congestion

P1
C4 -> C3 Dolly shortage

P1
I1, I2, I4,
I5, I6, I8,
I9, I10,
I11,I12,
I13 ,C1,C2,
C10, C11->

C4 (Cargo handling) process slowdown P1
I3, I7, I12,
C1, C2 ->
C5 Sloppy ULD hand-over (RHOs HACTL)

P1
C1 -> C6 (Negative effect on BSS through) CSS
(operational) problems

P1
I12,
I13 ,C1 ->

C7 Human fatigue (of HACTL staff) P1
C4 -> C8 Inadequate mobile communication support

P2
C1 -> C9 Inventory mismatch (in CSS/BSS)

P3
C1 -> C10 CSS-LCS insufficient operator feedback

P3
C1 -> C11 BSS-LCS: insufficient operator feedback P3

385
Caused
By
(I)
Initial/
(C)
Conseq.

Identified Problem

Px

(b) The Opinions of HACTLs Experts

14.21 If the Commissioners understood the HACTL experts
correctly, their opinions were that as CHS was operating without any
problem for the last three days prior to AOD, there was no reason why it
could not cope with the live load operation on AOD. W53 Day also
made calculations based on the work processed at ST1 24 hours from 7
am on 5 J uly 1998 to 7 am on 6 J uly 1998 to conclude that the throughput
capacity of CHS was sufficient for the purposes of AOD. W52 Nimmo
and W53 Day maintained that there was nothing fundamentally wrong
with CHS, and what caused difficulty on AOD was that the operators
wrongly perceived that LCS-CSS was operating slowly. This perception
resulted in the inefficient operation of CHS as LCS-CSS was increasingly
operated in manual mode instead of on-line automatic mode. The fact
was, according to these two experts, LCS-CSS was not responding slowly
but only appeared to the untrained eye to be so. The design of the
operation of LCS-CSS in on-line mode was that an order for movement
of cargo would only start when the whole route or path of transfer of the
cargo was clear. The operators, however, did not understand but as they
were subject to heavy pressure of work on AOD they perceived that the
operation of LCS-CSS was slow. There were alternative routes and if
the quickest one was conceived by LCS as blocked, it would try another
one and so on. It was not until all possible paths were conceived to be
unavailable that LCS would stop carrying out the movement order.
According to the two experts, the designed time for starting movement
would be about a few seconds or a few minutes, but sometimes because
all the transfer paths were blocked, the function would never be carried
out. They opined that if 10% of CSS was used in manual mode, there
was a 50/50 chance of LCS starting a transfer order, but if 30% of CSS
was in manual mode, LCS would only have 10% chance of starting a
transfer order. They said that on AOD nearly 30% of CSS was operated

386
in manual mode. The operation in manual mode caused difficulties in
that (i) such operation mode would not be able to provide the throughput
capacity that was required on AOD as the automatic mode would; and (ii)
it created data errors affecting the inventory kept by COSAC on which
LCS operation would need to depend. The data error was either caused
by the data entered being incorrect or by the operators forgetting to input
the data after executing an order in manual mode. The data concerned
included the position of each ULD. When the manual mode of
operation became widespread and when the inventory became more
corrupted, a physical inventory check became necessary. That took
place in the early hours of 7 J uly 1998. However, shortly after the
conclusion of the inventory check, the whole inventory was inadvertently
deleted. That sealed the fate of the ST1 breakdown, for it was gravely
suspected that there were some unknown causes for the loss of the
inventory, involving serious software problems, and thus CHS could not
safely continue to operate. The two experts concluded that the
breakdown was mainly caused by two factors external to HACTLs
systems and equipment, namely, (i) the ramp confusion and the
unfamiliarity or non-compliance of the procedures by RHOs; and (ii) the
lack of flight information from FDDS, a service provided by Hong Kong
Telecom CSL Limited (HKT) to HACTL as a subscriber.

14.22 There are two intrinsic flaws in the expert opinions of W52
Nimmo and W53 Day. First, they addressed the issues which were not
within their professed expertise: for instance, while they said that there
was no problem with CHS either in the mechatronics level or with the
computer systems, they concluded that the breakdown of CHS was
caused mainly by the absence of flight information from FDDS and the
chaos at the ramp. However, they did not profess any expertise in
FDDS in relation to cargo handling and ramp operations save that both of
them said that they had experience and expertise in information systems
and management matters. Secondly, both witnesses relied heavily on
facts that they were told by HACTL staff but those facts were not
supplied to the Commission through documentation or while the HACTL
personnel were giving evidence. For instance, it was alleged by the
experts that HACTL received only 15 Import Hand-over Forms from
RHOs on AOD, based on which they opined that HACTLs staff had
difficulties in matching the data relating to cargo imported on AOD that

387
were not covered by those forms with the data in COSAC. HACTLs
staff were required to match the consignment details contained in the
forms with the data recorded in COSAC so as to handle the cargo for the
consignees. The two experts assumed that since HACTL staff told them
that they only received 15 such forms, RHOs were at fault in failing to
hand-over the other forms covering the rest of the cargo imported for
HACTLs handling on AOD. They did not know, until pointed out, that
when the RHO representatives gave evidence before the Commission,
they testified that their personnel could not find HACTLs staff on the
ramp for the hand-over and that this evidence was not challenged. The
two expert witnesses also relied heavily on what they had been told by
HACTL staff to base their opinion that LCS for both CSS and BSS was
functioning. Moreover, they used information provided by HACTL staff
to assist them to draw the conclusion that LCS did not experience slow
response but only that HACTL operators perceived it to be functioning
slowly. It is unfortunate that the two experts relied on information
provided by HACTL staff that had not been tested before the Commission.
Their independence as perceived and the correctness of their opinions are
thus marred.

14.23 Insofar as the two expert witnesses did not profess expertise
in cargo handling by the use of FDDS or ramp operation, there is little
doubt that their opinion on such matters should be disregarded, as their
only legitimate purpose was to advise HACTL on the fields or areas of
their expertise and to assist the Commission with their opinions on the
same fields and areas. They are not factual witnesses who can offer any
evidence on facts in which the Commission was inquiring. The factual
inquiry and the determination on what facts are reliable and acceptable
are squarely within the purview of the Commission, and no one could
legitimately or justifiably usurp the Commissions function in this inquiry.
The two witnesses investigation into the facts with HACTL staff is futile
and must be declared to be so, save where such facts are identical to those
contained in the evidence received by the Commission. The only safe
approach for the two experts is that they should base their opinions and
conclusions on the evidence already presented to the Commission. They
can rely on parts of the evidence, stating that the evidence is subject to
dispute if such being the case as apparent from the testimonies or
documents given to the Commission. Since their approach was wrong,

388
it is risky to rely on anything they expressed as their opinions unless it is
clearly proved that their views are supported by the evidence presented to
the Commission, and that such views are within their fields of expertise.

14.24 The two experts started on the premise that as far as they
could see, there was nothing wrong with CHS, and the required cargo
handling capacity on AOD was within its throughput capabilities. They
based this conclusion on their alleged examination of the system as well
as from information provided by HACTLs staff they interviewed.
However, they did not even set out any detail as to how they had
examined CHS. From thence forward, they started to assign blame for
the breakdown of ST1 to external factors, which they concluded to be the
lack of complete and correct flight information from FDDS or FDDFS
and the ramp chaos experienced outside ST1 on AOD. They reached
their conclusions on the basis of the alleged facts supplied by HACTL
staff they interviewed without paying sufficient regard to the evidence
already before the Commission. Despite the risk involved in relying on
their opinions on these so-called external causes because such opinions
might have been based on matters not properly before the Commission
and because they are not experts on FIDS, FDDS or ramp operations,
their views of attributing the causes for the breakdown of ST1 to these
alleged external factors should still be examined carefully against
reasonableness and the facts as found by the Commission.

14.25 From the evidence of HACTL staff who had given evidence
before the Commission and the chronology of events on AOD at ST1
prepared by HACTL, it is clear that the problems experienced with CSS
took place very early in the morning. For instance, at 2 am one of the
three operational stacker cranes stopped functioning [item AODH 18 of
the chronology in Chapter 11]; at 2:20 am about 30 CSS orders were
found queuing for being processed due to a LCS error and manual mode
operation started [AODH 20,21]; at 4 am build-up staff needed to search
for the cargo in loaded stacker boxes at 3/F and 4/F [AODH 23]; at 6 am
a lot of units were still waiting at workstations for automatic transfer
vehicle (ATV) pick-up [AODH 24]; at 7:40 am the backlog of cargo
had increased [AODH 25]; and at 9 am most of the stacker cranes were
being operated in manual mode resulting in further inaccuracies of the
inventory [AODH 27]. Due to the slow response of CSS, a large

389
backlog of ULDs had been built up before 7 am.

14.26 The supplemental statement dated 12 December 1998 of Mr
Peter LUI Shui Hing, the General Manager Planning of HACTL, shows
that at various times on AOD, the following ULDs were received and
handled by CHS:

Time on
AOD
Number of
Import ULDs
received from
airlines with
exact flight
identification
(ID) (arrayed
by ATA)
Number of
Export ULDs
accepted onto
aircraft with
exact flight ID
(arrayed by
ATD)
Number of
ULDs
check-in:
Import with
and without
flight ID*
Number of
ULDs
check-out:
Export with
and without
flight ID,
335 with
flight ID
0000
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
0
0
0
0
0
0
45
67
59
79
57
39
48
40
0
0
0
0
0
0
0
3
18
4
32
11
7
0
7
154
123
2
171 720
132
127
4
90
42
28
63
55
101
0
0
1
0
1
1
7
15
5
20
6
34
27
14

390
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
Unknown
33
25
26
19
27
26
93
37
65
1
45
4
27
2
1
17
5
21
3
18
9
24
98
19
80
13
25
27
49
84
42
16
61
21
4
20
12
21
12
23
23
18
Total 831 206 1,552 346
* This includes some prepacked export ULDs checked in over the
landside.

14.27 It is to be noted that the figure of 1,552 ULDs was inclusive
of the empty ULDs that had been relocated from Kai Tak to ST1. From
midnight up to 7 am on AOD, when there was little received by way of
import, 720 ULDs had been checked in. Out of these 720 ULDs, 9 were
with exact flight ID whereas the remaining 711 ULDs were with dummy
flight ID, ie, without actual flight ID. It is therefore clear that, even
accepting these figures from Mr Luis statement without question, from
midnight to 7 am on AOD, a small amount of cargo imported had been
received by HACTL. If CHS found any difficulties in checking in
ULDs, that had nothing to do with the imports, and therefore had nothing
to do with the non-compliance with hand-over procedures of which was
alleged by HACTL and its experts to be one of the two major factors
causing the problems in ST1. These figures indeed contradict the
allegation.

391
14.28 It is therefore pellucid that the so-called ramp chaos that
could only have commenced since the arrival of import cargo could not
have caused the slow response of CSS resulting in its operation in manual
mode. The amount of import cargo (ie, 45 ULDs in total) arriving
before 7 am AOD was negligible. Prior to 7 am that day, the work done
by CSS and BSS could only have been in relation to preparing cargo for
export, and that CSS had already started to go slow without being
bothered by any imported goods. The conclusion that the
Commissioners draw from these pieces of evidence is that the ramp
confusion and chaos on AOD was caused by, and not causing, the
breakdown of CHS. The two HACTL experts views are not in
accordance with the evidence found by the Commission and are illogical.
After some cross-examination, even W53 Day agreed that if the ramp
chaos occurred only after 7 am, it was a consequence of and not a cause
for the breakdown of ST1.

14.29 The second main cause that the two experts opined for the
ST1 breakdown was the lack of correct and complete flight information
from FDDS. Their view is based on the fact that LCS relied on the
flight information so as to prepare the operators for exports and imports.
LCS was used to control the operation of the mechatronics of CSS and
BSS. The flight information was necessary for RMS to make the
necessary planning. The information would convey an order to LCS
which would in turn operate CSS and BSS accordingly. The most
important flight information for export was the estimated time of
departure (ETD). Apart from that, the scheduled time of departure
(STD) was always available to HACTL as the STD was from the
seasonal schedules from airlines and their Societe Internationale de
Telecommunications Aeronautiques (SITA) system with which COSAC
was connected. The STD would not be altered unless there was a 10- to
15-minute difference between it and the ETD. When HACTL had the
STD, unless the ETD differed by more than 10 to 15 minutes, the
preparation for export would not reasonably be affected in any way. The
same applies to the estimated time of arrival (ETA) and the scheduled
time of arrival (STA). Similarly, when HACTL had the STA, it would
not reasonably have been affected by the absence of ETA. As early as
May 1998, AA informed HACTL that the ETA and ETD in FDDS were to
be received from Airport Operations Control Centre (AOCC) and the

392
airlines. If HACTL was truly in need of such flight information from
FDDS on AOD, it could and should have contacted AOCC or the relevant
airlines for the information. W53 Day did not desist from maintaining
that the lack of flight information from FDDS was one of the two main
causes for the breakdown of ST1. He said that HACTL would need to
know the ETD about five hours before the actual time of departure
(ATD) for export. As for import, HACTL would need to know ETA.
However, FDDS supplied neither ETD nor ETA on AOD. Nevertheless,
he agreed that there would only be a difference between five hours (time
ahead of the ETD when planning for export should normally start) and
three hours (when the ETD could be obtained from the FIDS display
service for the public), and therefore there was not much difference in
time between obtaining the ETA from FDDS or from the monitors
displaying ETA through FIDS. Anyhow, RMS, the system that would
make the planning, was disconnected from LCS at about 6 pm on AOD,
whence the planning function of RMS was no longer used. Regarding
imports, if RHOs delivered the cargo to HACTL at the ramp, then the
cargo would be there for HACTL to handle, rendering the fact of HACTL
having no ETA quite irrelevant on AOD.

14.30 The alleged confusion over the C&ED customs clearance
system at the most could only have caused delay in customs clearance.
It had nothing to do with the operation of CHS, especially when RMS
was disconnected.

14.31 HACTL and its experts alleged that the late delivery of
pre-manifests by airlines also caused problems. The force of this
allegation is much diluted by another part of Mr Luis supplemental
statement. Mr Lui stated that there were altogether 54 late pre-manifests
records on AOD, but those were out of a total of 198 flights, meaning that
slightly over 27% of the pre-manifests were late. The impact could not
have been that substantial as to become anything near to a major cause of
ST1s breakdown. There is evidence that flight departures on AOD
were delayed. The inconvenience caused to HACTL by the late delivery
of pre-manifests would have been alleviated if the departure time of the
flight on which the cargo was put was not delayed. Anyhow, these
delays would not have caused any slow response to CHS, which was the
main cause for HACTLs operational staff to turn into the manual mode.

393
It was the switch to manual mode that eventually led to the breakdown of
CHS.

14.32 The other alleged external causes mentioned in paragraph
14.16(i) above were, in the opinion of the Commission, minor matters,
even if true. They collectively would not have caused difficulties in
operating CHS had it been running normally.

14.33 Although the two HACTL experts stated that there was
nothing wrong with LCS-CSS and LCS-BSS, yet in cross-examination,
W52 Nimmo and W53 Day accepted that there were a number of
mechanical problems with CHS. Plenty of examples of mechanical
problems can be found in the evidence, such as imperfect calibration of
the Geotronic system, insufficient torque provided to the invertor drive on
each transfer vehicle, mis-alignment of metal wheels of the right-angle
decks, incorrect spacing between the metal wheels at the edges of the
conveyor decks and the right-angle decks, and excessive rotation of the
metal wheels of the turntable transfer vehicles, etc. Further, they also
admitted that there were some problems with LCS-BSS, but they
disregarded and did not investigate them because they thought that the
most important thing was CSS. CSS had to operate first before the
service of BSS would be required. CSS was to store containers or send
containers for export or delivery to consignees whereas BSS was for
making up cargo into ULDs (containers) in preparation for export, and for
breaking up import cargo from ULDs for delivery or storage before
delivery. Eventually, they agreed that if BSS could not operate for
longer than 24 hours, CHS could not operate.

14.34 It is surprising that the operators were not apprised of the
way of operation of LCS-CSS and thought that it was running slowly.
W53 Day told the Commission that as far as his investigation with
HACTL went, he found that the operators were not sufficiently warned or
trained about the way LCS-CSS operated. The two experts also agreed
that HACTL staff were not familiar with operation in manual mode,
therefore causing a lot of inaccuracies in the inventory. These
inaccuracies culminated and finally an inventory overhaul was required.
An inventory loss following from it eventually led to the breakdown.

394
(c) Dust

14.35 Dust had been maintained by HACTL as one of the causes
for the breakdown of CHS on AOD. It was first presented as a major
problem on 15 J uly 1998 when the top management of HACTL had a
meeting with Government officials led by the Chief Secretary. It was
stressed as a major problem by almost all personnel of HACTL who gave
evidence before the Commission, save perhaps W7 Charter. A lot of
time had been spent at the hearing for this alleged major problem which,
when HACTLs two experts were cross-examined, had eventually been
conceded to be a manageable problem. HACTL had never withdrawn
its allegation that one of the contributing factors for the crash of ST1 on
AOD was dust and contamination of CHS. Dust together with the
presence of water at ST1 allegedly blocked and seriously affected the
15,000 highly sensitive sensors and reflectors installed for the operation
of the mechatronics of CHS. Both CSS and BSS consist of
mechatronics. The mechatronics are the mechanical, electrical and
electronic equipment that handle cargo. There are conveyor belts and
ATVs which move and transfer cargo. The stacker cranes pick up cargo
from the conveyor belts and ATVs, putting it into and retrieving it from
the storage compartment. W2 Mr K K YEUNG, the Deputy Managing
Director of HACTL, emphasised that the mechatronics, being the lowest
arm of the 5-level CHS, was the most important element in the handling
of cargo. Without them, the whole CHS could not work, while they
could work alone even if the higher levels of CHS all failed.

14.36 W11 Mr LEUNG Shi Min gave evidence from Day 17 to
Day 19 of the hearing, namely from 7 October 1998 to 9 October 1998,
together with W10 Mr HO Yiu Wing. W11 Leung was the Maintenance
Manager of CHS of HACTL and admittedly had the responsibility of
arranging and supervising the cleaning of CHS, in particular, the sensors
and reflectors. He stated in his witness statement which he confirmed
on oath that since late April 1998 he had arranged for a team of engineers
seconded by the Engineering Department of HACTL to check CHS
equipment regularly and to clean the sensors and reflectors thoroughly.
As from 18 J une 1998, there were up to about 15 engineers from that
Department undertaking the cleaning operations for the sensors and
reflectors whereas various other parts of the premises were cleaned by a

395
large number of ordinary cleaners.

14.37 W11 Leungs evidence laid great emphasis on the
seriousness of dust and contamination. The Commissioners have come
to the view that he is wholly unreliable in that regard. The
Commissioners observed his demeanour closely during his evidence and
he was always most hesitant and not straight forward when being asked
questions about dust.

14.38 On Day 17 of the hearing, W11 Leung testified that at about
noontime on 6 J uly 1998, ie, AOD, four engineers told him that the
mechatronics of CHS that stopped operation could be restored by wiping
the sensors and reflectors, and that experience tallied with his own when
he helped to rectify problems that morning. He immediately told W20
Mr Tony KWAN To Wah, the General Manager of the Engineering
Department of HACTL, about the dust problem, and W20 Kwan asked
him to look further into it.

14.39 On Day 18 of the hearing, without being questioned, W11
Leung volunteered that he had made a mistake regarding what he had told
the inquiry the day before, in that he did not on AOD tell W20 Kwan
about the dust problem, but only about the interruptions to the operation
of the mechatronics of CHS, without mentioning dust as the cause. He
merely suspected that the cause was dust although he had been told by
four of his engineers that the problem was with dust.

14.40 On Day 17 of the hearing, W11 Leung stated that dust,
according to his reckoning, caused about 30% of the problems
experienced at ST1 on AOD. On Day 18 of the hearing, he produced
some tables setting out figures that purported to show that CHS
equipment was affected by dust, and he estimated it to be responsible for
30% of the problems on AOD. In fact, the estimate was only his
guessing without any contemporaneous document in support.
According to the proper reading of those tables, about 50% of the
problems encountered at ST1 on 6 and 7 J uly 1998 were caused by dust.
Yet he maintained that he only suspected the problems was caused by
dust, and did not tell anyone more senior than him in the HACTL
hierarchy on 6 or 7 J uly 1998 about his alleged suspicion.

396

14.41 W11 Leungs difficulties in explaining as to whom he had
informed the alleged cause of dust and if so when, should be viewed in
the light of the evidence that was recorded in contemporaneous
documents which are outlined as follows:

(a) By a press release of 7 J uly 1998, W7 Charter, the managing
director of HACTL, when announcing a 24-hour embargo on,
inter alia, imports on all passenger flights other than
perishables, told the media that HACTL had encountered
computer system difficulties and that it had to buy time to
rectify these system problems;

(b) On 8 J uly 1998, in another press release, W7 Charter extended
the embargo for 48 hours so as to allow HACTLs engineers
and contractors adequate time to rectify current hardware and
software problems with BSS;

(c) In a press release of 9 J uly 1998, W7 Charter announced a
moratorium till 18 J uly 1998 on all cargo on all aircraft, save
urgent items, and mentioned that the moratorium would
assist the company in rectifying software and mechanical
problems;

(d) On 10

J uly 1998, W7 Charter held a press conference at the
Conrad Hotel, in which he again mentioned that the
moratorium till 18 J uly 1998 would enable HACTL to
address and deal with software and minor electrical and
mechanical equipment problems; and

(e) In the faults summaries compiled by the Engineering
Department covering 5 to 7 J uly 1998, there was little
mention of dust.

14.42 W10 Ho, who gave evidence together with W11 Leung,
faced similar difficulties as W11 Leung. On 21 September, W10 Ho
made a witness statement to the Commission stating that On AOD, the
engineers reported to management that the majority of the equipment

397
faults were caused by dust and other debris on the surfaces of the sensory
equipment. When W10 Ho gave evidence on Day 18, 8 October 1998,
he amended that statement to read Immediately after AOD Despite
the amendment, it does not alter the fact that dust was a problem.
However W7 Charter in his evidence said that top management of
HACTL was only aware of dust being the cause of CHS breakdown
between 10 to 15 J uly 1998. The Commissioners do not accept the
evidence of W11 Leung and W10 Ho that on AOD or immediately
thereafter the engineers reported to them problems with dust as the cause
of the breakdown of CHS. There was no public statement by W7
Charter or anyone of HACTL about dust or contamination being the
cause of the problems at ST1 until 15 J uly 1998 when HACTLs top
management met with the Chief Secretary and other Government officials.
The stress in all the press releases thitherto was on software and system
problems and nothing on dust.

14.43 The trouble created by dust and contamination was
appreciated by W2 Yeung, as early as 21 April 1998, as evidenced by the
minutes of a meeting held on that date. W2 Yeung directed W20 Kwan
to deploy people to deal with dust and contamination. W11 Leung was
the person who was entrusted by W20 Kwan with the task. W11 Leung
told the Commission that he had arranged many cleaners to clean various
parts of ST1 and deployed 15 engineers to deal with the cleaning of the
mechatronics equipment, especially the sensors and reflectors.
Realising that dust caused problems to the operation of the equipment on
AOD, W11 Leung sought the assistance of the engineers on day shift, in
addition to his 15 engineers, to deal with dust. If he had mentioned the
dust problem to W20 Kwan at about noon on AOD, it would have been
extremely unlikely that W20 Kwan would not have reported it to W7
Charter or W2 Yeung, and it would have been impossible for W7 Charter
to have failed to mention openly to the media in the period between 7 and
10 J uly 1998 that dust was the main culprit for the ST1 crash.

14.44 The various versions of the evidence of different witnesses
from HACTL simply do not tally. The Commissioners do not believe
that dust did cause the amount of problems facing ST1 on AOD as
alleged. Dust might have caused some problems, but those could have
been rectified easily by engineers wiping the sensors or reflectors.

398
During cross-examination, W7 Charter stated that the problem of dust
was overplayed. The statement would be puzzling if in fact dust was
responsible for 30% or 50% of the problems encountered by ST1 on
AOD.

14.45 Further, if dust did cause any problem, the Engineering
Department or W11 Leung that allowed the problem to persist till AOD
would be guilty of failure of duty on their part. The excuse that the
severity of the dust problem could not have been foreseen prior to AOD,
put forward by both W2 Yeung and W11 Leung, is unreasonable and not
accepted by the Commission. W9 Mr Gernot Werner, the Senior Project
Manager of Demag, the supplier of CSS, described that dust was always
present in ST1. He said that dust was inside ST1 during the months
when testing and commissioning of CSS were carried out, as it was
present on AOD and for a long period thereafter. If in fact dust was the
culprit on AOD and it was noticed by at least W11 Leung and four
engineers, it would be inconceivable why the faults summaries on AOD
and 7 J uly 1998 hardly identified it. The faults summaries reported
problems by CHS and were prepared by the Engineering Department a
week or two afterwards. On AOD W11 Leung and no less than 15
engineers were assigned the task to deal with dust, and he should have
known as early as noon on AOD that, at least according to the report from
four of his engineers if not through his own experience, that dust did
cause at least part of the problems on AOD. If the crash of ST1 was
caused by software or hardware deficiency, the responsibility would rest
squarely with HACTL, but if the cause was dust as alleged, the fault
should, as must be realised by HACTL, lie on somebody or somewhere
else. While it is appreciated that there must be a lot of confusion to the
extent of a general panic on AOD at ST1, it would be most unlikely that
the culprit of dust, if at all it was a fact, could have escaped the attention
of all the persons in HACTLs middle and top management.

14.46 According to W9 Werner, he met with W20 Kwan of
HACTL on 8 J uly 1998, when he was told that the problems encountered
were data mismatch and file corruption, and that they were mainly high
level software related. W9 Werner carried on to say that on 13 J uly
1998, he had a meeting with HACTLs management when it was
mentioned to him that there were software related problems within

399
HACTLs computer systems and network problems and the problems
were under current investigation. On 14 J uly 1998, W9 Werner put
forward a proposal to HACTL to develop an off-line mode based on the
operations of PLC and the mechatronics so as to enable CSS to operate
only levels 1 and 2 of CHS, cutting the link of PLC to LCS that would
have enabled CSS to be operated in an automatic or on-line mode. His
proposal was accepted and as a result, W9 Werner and his colleagues
made necessary modifications to some 100 PLCs, and assisted in the
training of HACTLs staff on how to use the off-line mode. The off-line
mode was ready and started operation on 18 J uly 1998. During these
meetings that W9 Werner had with HACTL, HACTL never mentioned
any problem with dust. When W9 Werner proposed to modify PLC in
order to enable operation of CSS on off-line mode, no one in HACTL
suggested that the proposal would not work since whatever modification
was done to CSS, its operation would similarly be hampered by dust. In
fact this off-line mode worked and experienced little problem with dust.

14.47 The Commission accepts W9 Werners evidence cited above,
not only because it was not challenged by HACTLs counsel in
cross-examination that the off-line mode of CSS operation was in fact
effected on 18 J uly 1998, but also that if dust was the source of all evils,
W9 Werners proposal should have been rejected by HACTL as futile, or
at least he would have been warned that whatever Demag did, that would
still be subject to the colossal problem of dust. It might be argued that
W9 Werner testified from the motive of saving Demag and tried to shift
the responsibility for the paralysis of ST1 to HACTL, by telling the
Commission that HACTL had problems with its own software or network.
However, even if W9 Werner were to tell the Commission that dust was
mentioned by HACTL to him as the culprit, it would not have adversely
affected Demags interest either: Demag could never be accused to be the
creator of dust. Even though HACTL had cleaned the whole of CHS, in
particular all the sensors and reflectors between 8 and 13 J uly 1998 after
having removed all the cargo from both BSS and CSS, there would have
been no conceivable reason for it to have accepted W9 Werners proposal
to cut the link between PLC and LCS had there been no inherent problem
with LCS or any part of the computer system on higher levels of CHS.

14.48 On Days 21 and 22 of the hearing, ie, 13 October 1998 and

400
15 October 1998, W16 Mr Hiroshi NAKAMURA, W17 Mr Tomonobu
SAEKI and W18 Mr Shin YAMASHITA gave evidence in group. They
were respectively the Project Manager, Project Engineer and Testing and
Commissioning Manager of Murata, the supplier and installer of BSS, the
other main mechatronics component of CHS apart from CSS. These
three witnesses were at ST1 during the installation, testing,
commissioning and maintenance of BSS. They were all working at ST1
in J uly and August 1998. They told the Commission that there were
three problems that affected the operation of BSS on AOD, namely, dust,
mechatronics and LCS. They also explained that the slow response of
BSS was one and the same problem as that caused by LCS. On AOD,
they personally had knowledge of three to four occasions when
interruption to BSS was caused by dust. W16 Nakamura described by
means of a pie chart his impression of the proportion of each of the three
problems that contributed to the troubled working of BSS on AOD. The
pie chart showed that dust represented about 2%, mechatronics about 2%
and LCS covered the rest of the pie. W17 Saeki said that his impression
was the same as W16 Nakamuras, while W18 Yamashita drew another
pie chart showing that dust occupied about 6%, mechatronics about 5%,
with the large remaining portion attributed to LCS. W16 Nakamura told
the Commission and the two other witnesses agreed that the dust situation
was worse in the period between 1 and 5 J uly 1998 than that between 6
and 9 J uly 1998, while it was less serious in the period from 10 J uly to 3
August 1998.

14.49 W18 Yamashita testified that HACTL gave three instructions
dated respectively 16, 18 and 21 J uly 1998 to Murata for modification
works to be done by Murata to the interface between LCS and PLC
regarding BSS. Such works were carried out accordingly and followed
by site tests and operation user training, resulting in the full operation of
BSS since 13 August 1998. Without these works, which were
performed at the request of and had to be paid by HACTL, BSS could not
be operated smoothly. Few of these works were required by reason of
dust.

14.50 There were delays in the construction of ST1 and the
adjoining Express Centre, with consequent delays in the installation of
various facilities and fit-out works. The installation, testing and

401
commissioning of the systems, including the 5-level CHS, and the
training of personnel in the operation of CHS were likewise delayed.
The TOP for ST1 was only granted on 3 J uly 1998 and the OP for the
Express Centre was issued slightly earlier on 27 J une 1998. There was
little doubt that the construction and touch-up works did result in a lot of
dust pervading the air on AOD and thereafter. However, all those
concerned with CHS, namely, HACTL, Demag and Murata, knew full
well at the latest from late April 1998 that dust affected the testing and
commissioning of CHS and, unless controlled, would continue to affect
the CHS functions when it was put into actual operation. Each of them
endeavoured to ensure that the environment would be sufficiently clean
for CHS to work. While the sensors and reflectors installed by Demag
were designed with protection from dust invasion, those provided by
Murata that had been badly affected by dust were duly replaced during
the stage of testing and commissioning. If the sensors and reflectors
were not replaced by Murata, the mechatronics of both CSS and BSS
would not have been treated by all concerned as fully prepared, perhaps
subject to some fine-tuning, for operation on AOD. The Commissioners
accept that there were a few interruptions to CSS and BSS caused by dust
on AOD, but these were insignificant, because the joint efforts of
HACTLs engineers under the direction of W11 Leung and the personnel
of both Demag and Murata had worked to reduce this environmental
impact to a negligible level. Even HACTL accepted that dust did not
cause any problem on AOD or thereafter in the cargo handling by the
Express Centre that was situated next to ST1, although the explanation
proffered was that there were few sensors and reflectors in the cargo
handling equipment of the Express Centre. The Commissioners come to
the view that dust at ST1 on AOD was but a minor problem on any
reasonable reckoning, and was overplayed by all the witnesses from
HACTL, as its Managing Director W7 Charter himself conceded.

(d) The Main Causes

14.51 As said before, nothing in the major causes alleged by
HACTL and its experts seems to attach blame to HACTL. There was a
notable silence on the operation of ST1s CHS save regarding the
functioning of CSS and BSS. The only admissions made by HACTL
regarding the failure of CHS to handle cargo efficiently on AOD were

402
that there was a slow or perceived slow response of the system and there
were some minor problems with LCS.

14.52 LCS is level 3 of CHS which gives orders to PLC for
operating and controlling the mechatronics. LCS applies to both CSS
and BSS. HACTLs witnesses who gave statements to the Commission
and testified at the hearing all stated that there were many contributing
factors to the paralysis of ST1 on AOD. Although all the alleged causes
set out in various paragraphs above were described as problems and
contributing factors, none was freely admitted by HACTL to be the main
cause or the major problem. It can be noted that the responsibility for
all of the alleged causes was attributed to rest with other parties, and, if
correct, HACTL will seem to be free from blame.

14.53 When the group of four witnesses, W12 Mr J ohnnie WONG
Tai Wah, W13 Peter PANG Tai Hing, W14 Ms Violet CHAN Man Har
and W15 Mr Daniel LAM Yuen Hi, all from HACTL, were
cross-examined by Mr Benjamin YU, Counsel for the Commission, W15
Lam said that he was responsible for doing all the on-site integration
testing for CSS and BSS. He was to report the test results to his superior
officers W13 Pang and W12 Wong. As he did not find anything in the
testing that caused him any concern regarding the operation of CSS or
BSS, he so reported to his two superiors.

14.54 W15 Lam testified that he had tested the throughput of CSS
but did not carry out any testing on the throughput of BSS. The testing
of the throughput would be for seeing how many units of cargo could be
handled by CSS and BSS in an hour. The result of his throughput test
with CSS was that CSS could handle 30 containers each zone in an hour.
However, he did not carry out any throughput test with BSS, as that
would have exhausted his manpower. He told the Commission that he
would have to use over 700 boxes or bins to test a zone of BSS, which
according to the user specifications would be required to move that
number of boxes or bins within an hour. He merely relied on the test
that had been conducted by Murata with the result that a zone of BSS was
able to move 720 boxes or bins in an hour. He explained that as
HACTLs Terminal 1 at the Kai Tak Airport was using a similar BSS, it
would be superfluous and unnecessary to test its throughput. Yet

403
Muratas tests were relating to the working of BSS on levels 1 and 2,
namely, the mechatronics and PLC of the system, whereas W15 Lam
should have tested BSS as integrated with the higher levels, namely, LCS,
the RMS and COSAC 2. HACTLs case is that COSAC 2 was an
enhancement of COSAC 1 used in Kai Tak, the RMS was a new
development for ST1 and LCS was a new version of that used in Kai Tak.
As all the softwares had been used in Kai Tak, integrated tests conducted
by HACTL were not necessary. However, Muratas tests did not include
the three higher levels of CHS, and without any test on BSS as integrated
with these higher levels, W15 Lam could not have any sound basis to be
sure that BSS as so integrated would be able to perform the throughput as
expected when it was put into full operation on AOD and thereafter.
Moreover, the results of the on-site integration tests were reported by
W15 Lam to his two superiors, W12 Wong and W13 Pang, but neither of
them made any checking, purely relying on the reported results from W15
Lam. While W15 Lam thought it was superfluous for him to conduct
any throughput integrated test for BSS, his superiors did not notice it. In
actual fact, no witness from HACTL has told the Commission that any
integrated test on the throughput of BSS was ever done before AOD.
The Commissioners find that this was one of the major reasons why on
AOD, when there were many cargo for ST1 to handle, BSS experienced a
slow response.

14.55 In fact, one of the causes alleged by HACTL and accepted
by the Commission for the paralysis of ST1 on AOD and the few days
that following was the loss of inventory in CSS. Apart from some
stacker crane stoppages, CSS also experienced slow response, and so
much so that the staff manning the system switched to manual mode, or
operation of the system without relying on LCS. The normal procedure
for operation in manual mode would require the staff to enter into LCS
the particulars relating to the item of cargo manually dealt with, so that
the base inventory would be updated. If data were not entered, LCS
would lose track of the item and its whereabouts. If CSS operated in
automatic mode, it would need to rely on the inventory being complete
and correct. For instance, if an item had by manual mode been put into
a storage compartment and LCS was not informed about it, LCS might in
automatic mode send a cargo to that compartment for storage, which
could not be done. When the operators who had used the manual mode

404
to handle cargo either made an error in the required data they entered into
LCS or forgot to enter the data into LCS, the base inventory would be
adulterated with mistakes, resulting in a return to the operation of CSS in
automatic mode impossible.

14.56 The significance of having a throughput test for BSS was
overlooked by all those responsible for ensuring that the system was
properly and sufficiently tested before AOD. This tallies with what was
stated in the press release made by HACTL on 8 J uly 1998, when the
embargo mainly on import cargo was announced to be extended further,
as follows:

Since our announcement yesterday of temporary measures to
relieve SuperTerminal 1 from the pressures it was under, we have
now had time to more closely analyse problems

allowing our engineers and contractors adequate time to
rectify current hardware and software problems with our Box
Cargo Storage Systems.

14.57 The Commissioners therefore conclude that one of the major
causes for the breakdown of ST1 was that CHS, especially integration of
BSS with the higher levels, was not sufficiently tested before AOD,
which was a result of the compression of the time required for testing and
commissioning of such sophisticated and complex CHS. This is also the
view of W55 Kipper and W56 Professor Vincent Yun SHEN, the experts
appointed by the Commission.

14.58 Another main cause was identified by W52 Nimmo and W53
Day, HACTLs experts. They stated that the operators working on the
floor of ST1 were not well trained or familiar with operating CSS or BSS
in manual mode, which was supposed only to be used temporarily. That
was the reason why there were so much operators errors in data entry
into CHS that corrupted the inventory database. This is further borne
out by the fact that Demag had to assist in the training of HACTLs
operation and maintenance staff on how to use the off-line mode after
AOD. The experts theory of perceived slowness compelled them also
to accept that the operators were not well trained or familiar with the

405
working of LCS-CSS and LCS-BSS, because if the theory was correct,
the fact that LCS would not normally commence the process of a cargo
movement order until the entire route was clear must have been
unbeknown to the operators. Although the Commissioners do not
accept the theory, the admission of insufficient training and unfamiliarity
with the equipment and machinery may have certain truth in it. The
delay of the construction works must have similarly resulted in lesser and
untimely training of the operation staff as it had compressed the time
required for the proper testing and commissioning of the systems.

14.59 HACTL generally maintained its reticence about what had
been done with CHS by way of rectification after its breakdown on AOD.
This has caused great difficulty to the Commission in identifying what
was precisely wrong with CHS. The evidence of W9 Werner, the Senior
Project Manager of Demag which supplied CSS to ST1, is that his
proposal made on 14 J uly 1998 to develop an off-line mode based on the
operations of PLC and the mechatronics so as to enable CSS to operate
on levels 1 and 2 only was accepted by HACTL. The proposal had the
effect of cutting the link between PLC and LCS that would have enabled
CSS to operate in on-line automatic mode. As a result, Demag made
modifications to some 100 PLCs, and assisted in the training of HACTLs
staff on how to use the off-line mode, which was started on 18 J uly 1998.
W18 Yamashita of Murata, the supplier of BSS, also testified that
HACTL gave instructions to Murata from 16 to 21 J uly 1998 for works to
be done by Murata to the interface between LCS and PLC of BSS,
resulting in the full operation of BSS as from 13 August 1998. It can be
reasonably inferred that there was something wrong with the interface.
The Commissioners find more probable than not that one of the main
causes for ST1s paralysis was that there was something wrong either
with the software of LCS or with the interfaces between LCS and CSS
and between LCS and BSS.

14.60 During the course of the evidence of W7 Charter, he hinted
that HACTL had been operating under pressure to make ST1 ready for
handling cargo on AOD, which was decided by Government without
consulting it and despite the contractual completion date of 18 August
1998. This could be viewed in two stages: before AOD was decided and
thereafter. Airport Authority (AA) and all contractors employed by

406
AA had been operating under the belief that the target date for opening
the new airport for operation was April 1998. Due to the delay in the
construction works in putting up ST1, AA eventually in December 1997
came to the view that the new airport was ready to open in the last week
of April instead of on 1 April 1998. This conclusion can be found in a
letter dated 10 December 1997 from W50 WONG Po Yan, the Chairman
of the AA Board, to the Chief Secretary. However, HACTL was not
informed of this conclusion. Airport Development Steering Committee
(ADSCOM) on the other hand took into consideration mainly the
delays in the ST1 construction works and in the provision of FIDS and
Mass Transit Railway Corporation (MTRC)s insistence that the Airport
Railway (AR), which was later known as the Airport Express, would
only be ready on 21 J une 1998 to reach a decision that AOD should be in
J uly 1998. When this decision was announced on 13 J anuary 1998,
HACTL was, according to W7 Charter and W2 Yeung, relieved and
happy because HACTL would have three further months, from 1 April
1998, the original target date to 6 J uly 1998, the announced AOD, to
make itself ready. The pressure under which HACTL was operating
before 13 J anuary 1998 must have been relieved.

14.61 After the announcement of AOD, HACTL should not have
been subject to any pressure to make ST1 ready for operation on AOD.
This is obvious from the fact that HACTL volunteered that it would be
able to process a throughput of 75% on AOD in place of its previously
promised throughput of 50% by April 1998. W7 Charter also said in
evidence that HACTL did not know that if it was not ready on 6 J uly
1998, AOD could be deferred; this might be treated as a hint that HACTL
was again operating under pressure to be ready by AOD and would not
make any suggestion for the date being deferred. However, this feeling
of pressure was inconsistent with the assurances given by HACTL
continuously right up to the beginning of J uly 1998 that ST1 would be
ready for operation on AOD. There was correspondence whereby
HACTL was urging the Fire Services Department and the Buildings
Department to grant fire safety certificates and occupation permit.
Obviously HACTL was eager to obtain these permits, but this eagerness
could hardly be properly translated into HACTL operating under undue
pressure.

407
14.62 HACTL also mentioned in evidence that it thought that the 6
J uly 1998 was not moveable. It also raised the idea of a soft opening,
meaning to use Kai Tak and the new airport at the same time even after
AOD. The idea of a soft opening was suggested by HACTL at the stage
even before the franchise agreement was reached. By a letter dated 16
August 1995, two days before the agreement was initialed, the then
Financial Secretary wrote to the Chairman of HACTL,

You raised the issue of HACTL operating with partial capacity at
Chek Lap Kok (CLK) and a portion of its operations at Kai Tak.
The Authority is firmly committed to the new airport opening in
April 1998. Although HACTL will only be contractually bound
to a 36 months programme, we expect HACTL to improve on it in
practice with a view to achieving a facility capable of handling a
minimum of 1.2 million tonnes per annum on airport opening.

I can also confirm that in the event that temporary trucking
operations (for also using Kai Tak after AOD) were required
(and we hope that this would never be required), the cost of such
operations would be taken into account in any future scheme of
control arrangements.

14.63 As it eventually transpired, the new airport opening in April
1998 to which the Authority is firmly committed was altered to 6 J uly
1998. It would be unreasonable for HACTL to hold the belief that AOD
could never be deferred. HACTLs position is that it was reasonable for
it to be confident in the ability of CHS to successfully handle the cargo
presented to it on AOD. Had HACTL been less confident with the
readiness of ST1 on AOD and requested either for its postponement or a
soft opening, and put hard facts in support before Government, it would
have been unrealistic of ADSCOM not to accede to one of these
alternatives. Indeed, after the notice of 25 March 1998 to quit Kai Tak
by 5 J uly 1998 had been served on HACTL, HACTL was advised by a
Lands Department letter dated 5 J une 1998 to write if it had any
difficulties in vacating those premises on 5 J uly 1998. However,
HACTL did not take up the matter further. When ST1 was paralysed,
Government and AA never failed to help HACTL in its arrangements to
handle cargo together in ST1 and Terminal 2. The Commissioners are

408
of the view that it was not so much HACTLs belief that AOD could not
be deferred or that soft opening was absolutely unavailable that was the
root of the problem. Rather, it was HACTLs over confidence with its
brainchild, ie the computer systems of CHS and with its ability to have
ST1 ready by AOD that resulted in the chaos in ST1. Everyone was
doubtlessly working under a certain amount of pressure to pull all
available resources together in order to minimise slippages and to ensure
that the new airport would be fully operational on AOD. Government
was trying to impress upon all concerned to work towards a common
target and to maintain the necessary momentum, but it would be against
logic and reason to imagine that Government would continue to insist if it
had been shown that the goal was impossible.


14.64 The responsibility for the problems of AAT should
mainly lie with AAT, although RHOs should also be responsible in a
minor way. AAT must be responsible for its staff who were not too
familiar with the new environment and the working of the new system
installed in the terminal. AAT should be responsible for not giving them
sufficient training and providing them with on-site familiarisation. On
the other hand, while AAT must be responsible for the deficient
co-ordination in the hand-over of cargo on the ramp, RHOs should also
share a small portion of responsibility.

14.65 As to the causes for the paralysis of ST1 on AOD, the
responsible parties that can be identified by the Commissioners as to who
should be responsible are set out in the following paragraphs.

14.66 There are two identified parties who could be responsible for
the delay in the completion of the construction works resulting in delay in
the installation of other facilities and ST1s machinery and systems, ie,
HACTL or GPY. Murata and Demag are not responsible because
HACTL did not maintain any allegation against either Murata, the
supplier of BSS, or Demag, the supplier of CSS, despite the fact that the
installation, testing and commissioning of those machines had been late.
There is also little in the evidence that Murata and Demag should be

409
responsible for those items of delay. While HACTL alleged that GPY
had caused the delays, GPYs case was that the delays were consequent
upon HACTL continuously giving instructions for additional and extra
works. It is impossible, in the short time available to the inquiry, to find
out whether it was one or the other or both who should be responsible for
the delays. In the circumstances, the Commissioners find it suffices to
conclude that there were delays in the construction works, and HACTL
should have known the problems that might and did arise from such
delays, and should not have given the assurances to AA and Government
that ST1 would be ready on AOD. In its written submissions to the
Commission, HACTL argued that it fully appreciated that the building
delays would interrupt the test plan and that not every component of CHS
could be fully tested prior to AOD. Hence, the focus of the integrated
operations testing was upon the components of CHS that would be
essential for successful operation on AOD and upon the throughput
demonstrated by those essential components during testing. Relying on
the throughput achieved by CHS during integrated operations testing,
HACTL maintained that it was reasonable for it to believe that ST1
would be able to successfully process the anticipated throughput on AOD,
on the basis that other essential airport facilities and services would also
be operational on AOD. HACTL further stated that 75% of ST1 was
operational on AOD. The anticipated throughput on AOD, the spare
capacity, system redundancy and modular design of CHS gave HACTL
confidence in its readiness for AOD. While the Commissioners accept
that HACTLs assurances were not lightly given and must have been
based on its top managements honest assessment of the effect, that the
delays in construction had and could have had on the readiness of ST1 in
providing 75% of its capacity throughput, the assessment was incorrect.
It can be said to be an error of judgment, which was mainly based on its
unfailing performance in Kai Tak for over the past decade and consequent
upon it being over confident with the software programmes that it had
developed for operation of cargo handling in ST1 and its having
under-estimated the harmful effect of the delay on the testing of the
programmes and the mechatronics in a fully integrated manner.

14.67 The Commissioners reach the same conclusion regarding the
disruptions to the testing and commissioning of the machinery and
HACTLs own systems. HACTLs allegations in support of the

410
disruptions to the preparation of CHS included mainly problems caused
by the delays in the construction of ST1, in the availability of power
supply to CHS, and in ST1 not having been made watertight. Again, the
Commissioners are not able to decide who should shoulder the blame for
these problems, whether HACTL itself, or GPY and/or other contractors
or subcontractors. HACTL being confronted by all these problems and
difficulties, the fact remains, as the Commissioners have found, that there
was an error of judgment on its part. Despite the shortage of time,
HACTL erroneously believed that all the machinery and systems would
have been sufficiently tested and would have faced little problem when
they were employed to work together in actual operation on AOD.
HACTL was too confident that the tests done by its staff during the
compressed period available before AOD and the experience of its staff in
operating similar systems would enable CHS to cope with a live load
operation, while overlooking that the software programmes developed for
ST1 contained enhancements of or alterations to those that had been used
in Kai Tak and therefore the programmes needed time to be fully and
unmistakably integrated and tested before they could handle live load
operations efficiently and effectively. For this, none other than HACTL
itself should be responsible.

14.68 The Commissioners think that the contamination of the
environment caused by the late completion of the construction and related
works (such as fit-out and decoration works) was expected, and engineers
had been deployed to clean and keep sufficiently clean the sensors and
reflectors of BSS and CSS over a month before AOD. As no problem
was expected from the contamination, there was no mention of it in the
public announcements made by HACTL on 7, 8, 9 and 10 J uly 1998
immediately after the crash of ST1. There should be a small extent of
interruptions caused by the contamination, but the extent of it was
overplayed. The cause of the contamination of the environment was
closely linked with the delays in construction. Similarly, the
Commissioners do not think they can reach a conclusion other than to say
that either HACTL or GPY or both should be responsible for the
contamination.

14.69 The Commissioners acknowledge that there were two CTOs,
being HACTL and AAT, and three RHOs, namely, J ATS, HAS and Ogden

411
in the new airport, and therefore the situation at the new airport was
different from that at Kai Tak. In the Commissions view, these
circumstances did not create much difficulty in co-ordination and
understanding between the CTOs and RHOs. These different
circumstances were known at the planning stage and frequent meetings
had been held among all those involved to design procedures for the
smooth running of the cargo handling services in the new airport. There
was no lacking in the spirit of cooperation amongst all parties in spite of
their being competitors. Instead of a machine having only two
components, the machine is run by several components instead. One
may well expect a short period of time for the slightly larger number of
components to settle down to provide a smooth running. The
Commissioners think it is an exaggeration to say that the slightly larger
number of partners became a contributing factor to the failure of ST1 on
AOD.

14.70 It is to be expected that RHOs were not too familiar with the
geography of the new airport, the facilities available, and the work
required of them, but there is little evidence to show that they did not
follow the procedures agreed amongst themselves and the CTOs to be
used at the opening of the new airport, so much so as to cause ST1 to shut
down. During cross-examination of the CTO and RHO witnesses, it
was alleged time and again that the hand-over procedures were not
followed by RHOs, but no particulars were available as to the manner in
which RHOs were not compliant with the procedures. In the opinion of
the Commissioners, the insurmountable problems were those encountered
by CHS itself, and they spilled over to affect the operations of RHOs.
Admittedly, RHOs were very much distracted by the difficulties they
were facing with their handling of baggage and serving aircraft and
passengers, mainly caused by the malfunctions of FIDS and the delay in
aircraft arrival and departure. However, but for the failure of CHS, the
Commissioners do not think that the problems facing RHOs would have
any noticeable effect on the smooth running of ST1 on AOD. The
Commissioners do not think that RHOs should fairly be held responsible
on this score except for a very small part of the blame for the inadequate
co-ordination between them and CTOs.

14.71 The allegation that dollies at the new airport for carrying

412
cargo between ST1 and aircraft were insufficient is rejected. Had CHS
worked as well as expected by HACTL, there would be little difficulty
with the number of dollies. The slow response of the mechatronics of
CHS and the change to manual mode operation reduced the speed in the
processing of cargo at ST1, resulting in the dollies being detained for
much longer than the agreed turnaround time of half an hour. Thus, the
1,030 dollies that were available in the new airport, as compared with the
about 530 available in Kai Tak, were found to be insufficient for
HACTLs purposes. The dollies being found to be insufficient was an
effect rather than a cause of the failure of ST1.

14.72 HACTL alleged that FDDS or FDDFS was not providing
flight-related information to ST1 as expected or at all, causing trouble or
inconvenience in the operation of cargo handling. The Commissioners
find that this must have to a certain extent adversely affected the cargo
handling by HACTL. However, the impact should, in normal
circumstances, be reflected by HACTL having to deploy several members
of its staff to obtain the necessary flight information from customers,
airlines, the AOCC or others, instead of causing the paralysis of ST1.
The Commissioners consider this as a contributing factor towards the
trouble encountered by ST1 on AOD, but no further. For this failure, the
main culprit must be AA, who failed to provide the necessary flight
information through Airport Operational Database (AODB) from which
the FDDS and FDDFS drew the information. Apart from AA, there may
be other parties responsible, which will be dealt with in the chapter
devoted specifically to FIDS.

14.73 Having considered all the evidence, the Commissioners
make the finding that the main cause of the failure of ST1 was that there
were probably faults in the interface between LCS and BSS and between
LCS and CSS. Those faults manifested on AOD in slow response of
BSS and CSS. The operators of the mechatronics of both BSS and CSS
resorted to manual mode operation, in order to improve on the speed.
As the operators were not used to or well trained in using the manual
mode, or using it on a large scale, they either forgot to input the necessary
data into LCS or input incorrect data into LCS, adulterating the inventory
record kept by LCS. The inventory was eventually corrupted to such an
extent that there must be a manual inventory check to purge it. The

413
personnel doing the manual inventory check unwittingly and
inadvertently conjured up a software programme that had the effect of
deleting the whole inventory. Although the old inventory could
subsequently be found as stored in the computer system, HACTL was not
sure of the reason for the deletion of the inventory, thus severely sapping
HACTLs confidence in the integrity of its computer software, requiring
some investigation to be made. For investigating the true cause for the
slow response, HACTL needed to clear BSS and CSS of all cargo already
loaded into these two systems. The backlog of cargo lying around ST1
was very large, and this backlog must be moved somewhere else before
BSS and CSS cargo could find some place to be put. A decision was
therefore made to transport all the cargo, both the backlog and those to be
removed from BSS and CSS, to Kai Tak so that a thorough investigation
could be undertaken. Repairs or improvements were then done to LCS,
BSS and CSS, especially to ensure that there was no problem regarding
their use in a fully integrated manner. Embargoes were therefore
announced one after another, and the major operations for processing
cargo were brought back to Kai Tak in the meantime. For all these,
HACTL is solely responsible.

14.74 HACTL contended that the LCS-CSS and LCS-BSS
software was of sound design and performed satisfactorily on AOD. No
significant changes have been made to the software since AOD. The
inventory mismatches and the accidental deletion of the container
inventory occurred through human error and they were not software
problems. Although the LCS-CSS and LCS-BSS experienced expected
problems on AOD, those problems did not in themselves lead to a
breakdown of CHS on AOD. HACTL explained that LCS-CSS
throughput testing and integrated operations testing of CHS were
successfully conducted in the period from J anuary to J une 1998. It was
not possible for HACTL to carry out effective simulated live load testing
on CHS (including CSS) prior to opening of the new airport. Such
testing could only be achieved by a soft opening of the cargo handling
operation at ST1. Manual mode testing of CSS was carried out during
the hand-over of the cargo handling machinery from equipment suppliers
to HACTL. HACTLs supervisory operational staff were trained and
familiar with operations in manual mode. In view of the fact that ST1
was not designed for terminal-wide manual operations and that the extent

414
of terminal-wide manual operations on AOD was unforeseeable, training
of lower grade operational staff in full manual operations was not viewed
as essential for successful operation on AOD and consequently not
conducted. The unexpected widespread manual operations of CHS in a
new working environment inevitably led to container inventory mismatch
and a slowdown of the overall performance of CHS.

14.75 The Commissioners are not persuaded by HACTLs
contentions and explanations. It may be useful to look at the causes in a
chronological order. Prior to 13 J anuary 1998, when AOD was
announced, HACTL gave the assurance to AA that it would be ready with
50% throughput capacity in April 1998. After the announcement of
AOD on 13 J anuary 1998, HACTL was relieved that it had three more
months to get ready, and instead of reaching 50% capacity in April and
75% in J une 1998, it started to give assurances to AA and Government
that ST1 would be ready on AOD with 75% throughput. The only
concern that HACTL had was with the completion of the construction
works, and when the TOP for ST1 was issued on 3 J uly 1998, HACTL
was honestly sure that its assurances would be fulfilled. HACTL was
confident with the operational efficiency and effectiveness of its CHS
because the testing and commissioning of BSS and CSS were expected to
have been completed prior to AOD. HACTL did not anticipate that any
major problem would arise when CHS, with BSS and CSS integrated
with the software programmes, started to operate on AOD or a few days
before. The confidence was induced by the good and almost unfailing
record of the software programmes that HACTL installed for ST1.
These software programmes had been used in HACTLs establishment at
Kai Tak and had been tested quite substantially off-site before they were
introduced at ST1. However, HACTL under-estimated the significance
of having the software tested thoroughly when integrated with BSS and
CSS as the software was not the original version as that used in Kai Tak
but had been enhanced for adoption at ST1.

14.76 In its submissions, HACTL denied that there was
under-estimation and alleged that the level of throughput achieved during
the integrated operations testing showed that the higher level computer
systems had been successfully integrated with CSS and BSS by AOD.
The Commissioners consider the submissions unacceptable and against

415
the facts found. HACTL also failed to realise the seriousness of the
delays of the construction works that had substantially reduced the testing
times. The confidence and under-estimation was manifested in HACTL
not having any viable contingency plans for the failure of CHS. The
main contingency plan, as described by W7 Charter and W2 Yeung, was
merely that the 75% throughput capacity was an over-provision for the
amount of cargo that ST1 was expected to handle on AOD which was
assessed to be about 50% of HACTLs throughput capacity. By reason
of this over-provision of capacity and the modular design of CHS,
HACTL was confident that there would not be difficulties in handling the
expected throughput on AOD. However, what seems not to have been
considered is that the modular design was only available to save the day
if there was nothing wrong with LCS, which operated CSS in the same
automatic mode as it operated BSS. If there was problem with the LCS,
as the Commissioners find probably to have been the case on AOD and
weeks thereafter, the automatic mode would all be lost, affecting ST1
terminal-wide. The modular design can only be relied on if part of the
mechatronics (ie, CSS and BSS), as opposed to LCS, fail as they would
only affect ST1 regionally. Nor does the evidence show that HACTL
had made any risk assessment of CHS failing. In their submissions,
HACTL elaborated that it was not feasible (either practically or
commercially) for it to operate Kai Tak and ST1 simultaneously with a
trucking arrangement on AOD. Its contingency plan with regard to the
unavailability of flight information at ST1 proved unworkable on AOD
due to the complete lack of flight information even in PTB. The
Commissioners do not find HACTLs explanation useful in understanding
what their contingency plans, if any, were in case of CHS failure. Even
if such plans did exist, they certainly did not help in ensuring that ST1
functioned smoothly on AOD and the following weeks.

14.77 On AOD, about 2,000 containers had been transferred from
Kai Tak to ST1. In addition, cargo arriving from inbound flights started
to accumulate. HACTLs operation staff began to notice slow response
with both CSS and BSS. That was mainly caused by LCS not operating
PLC and the mechatronics smoothly. W7 Charter admitted on oath that
the fault level of CHS was one of HACTLs primary problems, together
with the difficulties it was having with LCS and that those were the main
reasons for the breakdown on AOD. The absence of flight information

416
from FDDS was a contributing factor but that would only drain HACTLs
workforce by several members of its staff requiring them to get the
necessary flight information. There would be little impact if CHS was
operating normally. Dust was another contributing factor, but both the
failure of FDDS and dust are viewed by the Commissioners as minor as
compared with the main causes for the breakdown. The circumstances
that there were three RHOs and two CTOs and the unfamiliarity of RHOs
with cargo handling work also would not have caused any noticeable
problem had LCS worked properly and smoothly. The slow response of
CHS led HACTLs operators to switch into manual mode, instead of the
pre-set automatic mode. Although this helped cargo processing, it was
still much slower than the automated process. As a result, the
procedures of hand-over of cargo that had been agreed between HACTL
and RHOs could hardly be followed, and inbound cargo were left by
RHOs on dollies outside the airside at the northern part of ST1. Dollies
were detained for much longer than the agreed turn-around time of 30
minutes, and as a result, there was a shortage and RHOs placed the goods
from the dollies onto the ground in order to retrieve the dollies for other
inbound cargo. These matters were not causes for the breakdown of
ST1, but rather consequences.

14.78 After HACTLs operators switched into manual mode in
operating CSS and BSS in many areas of the mechatronics, human errors
in not updating LCS or updating it incorrectly caused the inventory to be
adulterated, so much so eventually that there had to be a physical check
of the inventory. During the course of the physical check, a utility
programme was inadvertently switched on which erased the inventory.
This gave rise to grave concern to HACTL as it had to find out the reason
before there was any meaningful rebuilding of the inventory. At the
same time, investigation had to be made as to why LCS was not operating
as smoothly as expected. All these problems resulted in the embargo
announced in the days following AOD, so that the cargo at ST1 could be
cleared from CHS and moved to Kai Tak for processing. During the
period of the embargo, the cargo were removed out of CSS and BSS, the
equipment was cleaned, and CSS and BSS contractors were instructed to
cut the link between LCS on the one hand and PLC and mechatronics on
the other. Thereafter, CSS and BSS could be operated smoothly in an
off-line or manual mode. In the meantime, HACTL was debugging or

417
enhancing LCS and the software of the higher levels of CHS, leading to
recovery.

14.79 The cause for the deletion of the inventory was found as
early as 8 J uly 1998, but HACTL announced a 9-day moratorium on 9
J uly 1998. Had there been nothing wrong with the computer systems,
HACTL would not have imposed the lengthy moratorium even after the
reason for the deletion of inventory had already been known.

14.80 It is therefore clear, and the Commissioners find on the
balance of probabilities that the following parties are responsible for the
breakdown of ST1 on AOD and in the period of about a month thereafter:

(a) HACTL is responsible for giving the assurances to AA and
Government that ST1 would be ready to provide 75% of its
throughput capacity on AOD;

(b) Either HACTL or GPY or both are responsible for the delay
in the construction works at ST1;

(c) Either HACTL or GPY or both are responsible for the delay
caused to the installation of the machinery and systems at
ST1 and in the testing and commissions of such machinery
and systems;

(d) HACTL knew of the delays in (b) and (c) above, and is
responsible for under-estimating their effects on the
readiness of ST1 to operate efficiently on AOD;

(e) Contamination of the environment on AOD was very minor,
and would pose little difficulty to HACTL in the operation
of its CHS;

(f) Contamination of the environment, anyhow, was known to
HACTL as early as late April 1998, and HACTL is
responsible for not sufficiently clearing the environment for
the proper and efficient operation of CHS;

418
(g) The circumstances of there being three RHOs and two CTOs
were known to HACTL long before AOD, and RHOs
involvement with cargo handling could hardly be described
as an appreciable cause for the breakdown of HACTL;

(h) The ramp chaos and alleged insufficiency of dollies were
consequences of the slow response of CHS in processing
cargo and not the causes of the slow response;

(i) The failure of FDDS or FDDFS (for which AA and others
are responsible) also would not have been a serious threat to
the efficient operation of CHS, as HACTL could have used a
few employees to obtain the necessary flight information;

(j) The late delivery of pre-manifests by airlines and the new
C&ED customs clearance procedures would cause some
inconvenience to HACTL but did not contribute to the
breakdown of ST1; and

(k) The main causes for the breakdown of ST1 were (i) the
faults with CHS which resulted in the inefficiency of LCS in
controlling and operating PLC and the mechatronics, (ii) the
insufficient testing of CHS in fully integrated mode, and (iii)
the insufficient training and unfamiliarity of HACTLs
operation staff with operating CSS and BSS in manual mode;
and for all these HACTL is solely responsible.

Section 5 : HACTLs Best Endeavours Basis

14.81 It has always been HACTLs emphasis that it was not under
any contractual obligation to anyone to provide on AOD a cargo handling
throughput of 75% of the full capacity of ST1 or at all. The 75%
capacity means 5,000 tonnes of cargo a day or about 1,800,000 tonnes a
year. The franchise agreement made between HACTL and AA clearly
and indisputably stipulates that HACTL shall achieve 75% capacity by 18
August 1998, and not on any earlier date. HACTL only promised to use
its best endeavours to be ready with 75% capacity on 6 J uly 1998, the

419
AOD, and this basis was not contractual or obligatory and stemmed
merely from goodwill or a gentlemens understanding or agreement.

14.82 A best endeavours basis is obviously different and has to
be distinguished from a contractual basis. A contractual basis imposes
or results in an obligation the non-fulfillment of which will attract
contractual liability attached. On the other hand, a best endeavours or
best efforts basis involves no contractual obligation and therefore no
contractual liability attached. Contractual liability apart, a promise to
exercise best endeavours requires examination of two elements: first,
whether the promising party has in fact used its best endeavours to
perform the promised task and secondly, whether the promised task has
been satisfactorily performed. If the promising party has used its best
efforts, but the promised task is not performed satisfactorily, no blame
can be attached to it since it has already done what it has promised. If,
however, the promising party has not used its best endeavours as
promised and the task is not performed, then the promise is not kept,
regardless whether any contractual liability arises. In this case, it is not
that HACTL did not use its best endeavours as promised, because the
Commission feels and finds that HACTL did use its best efforts in the
circumstances. The fact that it committed and expended additional
funds in the Supplemental Agreement with GPY to accelerate the
building works so as to complete the works ahead of the contractual date
of 18 August 1998 bears fine witness. What is crucial is that HACTL
represented to AA and Government that ST1 would be ready to produce
75% of its throughput capacity on AOD. This was a representation that
was relied upon by the representees. It might have been a representation
of a future event, but regardless, it was a representation based on an
estimate of the status current at the time when the representation was
made. It was a representation of an estimate that turned out to be wrong,
and a wrong representation that was relied on for its honesty and accuracy
by the representees. If the representation was that ST1 would not be
ready, there could be no contractual obligation or liability, and the
representor could not be blamed either, but the representation of readiness
makes HACTL blameworthy, and for that HACTL must be responsible.
A good illustration of the situation is readily available from what had
happened with MTRC. MTRC was contractually obliged to complete
AR by 21 J une 1998. Despite the fact that Government expected MTRC

420
to be able to gain time during the course of its construction and
installation works, so that the new airport which was expected to be ready
in April 1988 could open for operation in April with the substantial
transportation support provided by AR, MTRC maintained that AR would
not be ready by April 1998. At the request of ADSCOM, MTRC made a
presentation about its progress to ADSCOM in October 1997, and on that
occasion maintained that AR would not be ready until the contractual
completion date. There was no promise of using best endeavours,
because best endeavours would not enable MTRC to abridge the time for
completion of its works. Nor was that any representation that AR would
be ready earlier than the contractual completion date, and no one was
misled. MTRC could not be blamed, nor is there any evidence that
ADSCOM or Government or AA ever at any time blamed it for not being
able to complete AR ahead of time.

14.83 Apparently, HACTL was too confident of its ability and
capacity in the development and commissioning of its CHS, so confident
that even the enormous delays in ST1s construction works did not cause
it to engage in any risk assessment seriously, nor to cause it to realise the
risks of non-readiness sufficiently enough to suggest a deferment of AOD
or to insist on a soft opening. Its continual success and reputation of
efficiency and capacity in the cargo handling field for over a decade
doubtless contributed towards its over-confidence and complacency.
That success and reputation had also lulled AA, New Airport Projects
Co-ordination Office and ADSCOM into placing too much reliance on
HACTLs assurance, to the extent of accepting its words without
engaging in any meaningful and professional monitoring of its systems
development and commissioning.

14.84 After reading through the statements of witnesses and
hearing all oral testimony, the Commissioners have come to the view that
although HACTL was not contractually bound to be ready with a 75%
throughput on AOD, its assurances given to AA and Government that it
would be so ready had given rise to a sense of security to AA and
Government that ST1 would be ready to provide the necessary cargo
handling facility reasonably assessed to be required of the new airport on
AOD. Taking into account HACTLs unfailing success at Kai Tak, there
was nothing that could induce AA or Government to doubt that HACTL

421
would not be as good as its words. The only worry that AA and
Government had about ST1s readiness was that there had been slippages
in ST1 obtaining the OP. When the TOP for ST1 was issued on 3 J uly
1998, there was no longer anything that diluted that sense of security.
HACTLs continual assurances made AA and Government confident that
the decision to open the new airport should not be altered, as far as cargo
handling was concerned.

14.85 When ST1 crashed on AOD, the Commissioners accept that
HACTL did not fail in its contractual obligation. However, the
expectations of AA and Government induced by the assurances were
proven ill conceived and incorrect. Had HACTL maintained its
contractual position that it would only be 75% ready on 18 August 1998
and not earlier, Government would never have made the decision to open
the new airport for operation on 6 J uly 1998 in the first place, and
HACTL could not in all fairness be blamed for not being helpful. The
confidence of HACTL in its newly developed COSAC 2 and computer
software programmes for CHS was too strong, and worse still the
confidence was manifested in the assurances. The confidence was based
on the fact that COSAC 1, from which COSAC 2 was developed, and
most parts of the software programmes that were to be introduced in ST1
had worked in Terminals 1 and 2 in Kai Tak for a long time, without
realising that the small amount of enhancements or alterations made to
these existing software programmes would result in the systems being
less reliable unless and until sufficient tests had been performed and
sufficient time had been used for adapting them to the real live operations
required of CHS once it started to work on AOD.

14.86 The Commissioners conclude that HACTL is responsible for
giving the false sense of security to AA and Government that it was ready
to operate on AOD. It would not be fair for HACTL to cling to the
contractual terms to say that it is not responsible for not being ready on
AOD. Even though this responsibility arose out of goodwill and a mere
gentlemens agreement without any contractual liability, the
Commissioners think that leading AA and in particular Government to
reach the decision on AOD and not to alter that decision is culpable, and
HACTL must fairly be held responsible for that area of decision-making
process and thereafter for either failing to render ST1 ready to deal with

422
the expected tonnage of cargo on AOD as it had promised and over a
month thereafter or failing to strive for a deferment of AOD or to seek a
soft opening timeously.

Section 6 : HACTLs Attitude in the Inquiry

14.87 Over 10 solid days were spent in the hearing of the
Commission for seeking facts and reasons relating to the question of dust,
which had been raised as a major problem by HACTL for ST1s
breakdown on AOD. A lot of effort was used by both counsel for
HACTL and counsel for the Commission as well as the Commissioners in
dealing with dust and its related problems. Had dust been raised as a
minor factor contributing to the breakdown, much less effort and time
would have been spent. If the Commission was empowered by the
Commission of Inquiry Ordinance to award any costs against a party, it
would not have hesitated to make an appropriate order relating to the time
and costs wasted for dust.

14.88 One interesting thing that has come to the notice of the
Commission is that it was only after 15 J uly 1998 that HACTL started to
raise the question of environmental contamination as one of the major
causes for the breakdown of CHS. Before that date, HACTL was very
frank in its press releases and open statements that they imposed the
moratoria in order to deal with software and hardware problems.
When W50 Wong and W49 LO Chung Hing, the Vice-Chairman of the
AA Board, gave evidence, they told the Commission that they paid a visit
to HACTL on 14 J uly 1998 before they attended an AA Board meeting
that afternoon. The visit was not prearranged, but was intended to show
AAs sincerity and readiness to help HACTLs situation. When they met
W7 Charter and W2 Yeung, there were also four to five other persons
there, and W49 Lo knew that one of them was a lawyer. W50 Wong and
W49 Lo were told that HACTLs operation was adversely affected by
things such as the confusion at the ramp and the insufficiency of dollies.
They felt that blame was put on AA and they were not too welcome, so
they left very quickly.

14.89 It appears to the Commissioners that HACTL was apologetic

423
for what had happened on AOD and the moratoria that it imposed on its
customers and it was frank to let the public know what was wrong with
ST1 in the first few days after AOD. After seeing lawyers, HACTLs
top management were obviously advised of their possible legal liability,
and thenceforth, their attitude changed, obviously for fear that legal
liability might attach. HACTL had maintained the same attitude during
the inquiry. Although W7 Charter was honest to admit that the dust
problem was overplayed, other officers of HACTL continued to stress the
major effects of dust, and even attempted to suggest strange interpretation
of W7 Charters admission. W2 Yeung was one of the protagonists of
the dust theory and he also maintained throughout that there was nothing
wrong with the computer systems of CHS. This is in a way
understandable, for the main constituents of the software systems of
COSAC were his brainchild. The evidence of all the HACTL officers
and the two experts was extremely protective of this prodigy of W2
Yeung and HACTL. The result was that the Commissioners were
presented with evidence and arguments that provided them with no
obvious answer as to what the causes were but only what the causes were
not. It is unfortunate that public funds and time had to be wasted for this
uncandid attitude of HACTL.

424
CHAPTER 15

RESPONSIBILITY THE OTHER MAJOR PROBLEM
AND MODERATE PROBLEMS

Section 1 : Major Problem : Baggage Handling



Section 1 : Major Problem : Baggage Handling

15.1 In this Chapter, the Commission recapitulates the various
problems contributing to the baggage handling chaos that are discussed in
Chapter 12 and deals with the responsibility for each of them in turn.

(a) Accumulation of problem bags

[BHS 1] Cathay Pacific Airways Limited (Cathay Pacific) and
Securair Limited (Securair) staff fed about 220 bags from
Kai Tak with no baggage labels into the conveyor system at the
new airport. [See paras 12.19-12.20]

15.2 In [BHS 1], Securair was engaged by all airlines (including
Cathay Pacific) to transport their interline baggage from Kai Tak to the
new airport. The bags were put onto the conveyor belt by Securair and
Cathay Pacific staff. Cathay Pacific claimed that its staff were merely
rendering voluntary assistance to Securair who had been instructed to
send the bags down to the Baggage Hall, either via the out-of-gauge
(OOG) lifts or through the conveyors after fallback tags were put on.
Securair alleged that it was only engaged to deliver the bags to the
check-in counter at Check-in Area B. From there, it handed the bags
425
over to Cathay Pacific baggage staff who was responsible for sending
them to the Baggage Hall.

15.3 Without cross-examination of the relevant witnesses, it is not
possible for the Commissioners to determine between Cathay Pacific and
Securair, who should be held responsible for the injection of the interline
bags into the system. However, there is no dispute that it was the staff
of both Securair and Cathay Pacific who sent the bags to the Baggage
Hall without using the OOG lift or the special fallback tags. Both of
these companies should therefore be responsible, though the Commission
is not able to make a proper apportionment of the blame.

[BHS 2] Airlines checked in bags with incorrect labels or invalid or no
Baggage Source Messages (BSMs). [See
paras12.21-12.23]

15.4 In [BHS 2], from the evidence submitted to the Commission,
there is no identification of the offending airlines except in the case of
some 600 bags from J apan Airlines Company Limited (J AL) and seven
transfer bags from Thai Airways International Public Company Limited
(Thai Airways). In the result, the Commissioners are unable to find
which of the airlines using the new airport should be responsible except
those two.

[BHS 3] Airlines checked in about 2,000 bags with invalid flight
numbers. [See para 12.24]

15.5 In [BHS 3], the airlines did not inform Airport Authority
(AA) or Swire Engineering Services Ltd (SESL) of the extra flight
numbers which did not appear from the flight schedule and of the
requirement for a separate flight lateral for the onward destination and
should therefore be primarily responsible. Canadian Airlines
International Limited (Canadian Airlines) and Virgin Atlantic Airways
Limited (Virgin) (together with Ansett had admitted responsibility for
the respective incidents referred to in paragraph 12.24 of Chapter 12.
Apart from that, the evidence received by the Commission does not
enable it to identify the particular offending airlines who should be
responsible for the other problem bags.
426

[BHS 4] Aviation Security Company Limited (AVSECO) staff
rejected a large number of bags at Level 2 security screening,
putting pressure on Level 3 screening, lengthening baggage
handling time and causing more problem bags. [See para
12.25]

15.6 [BHS 4] is obviously a matter of familiarity and caution.
Despite their training, AVSECO staff were working in a new environment
with a new system. Obviously they took longer to examine the bags
shown on the security screen at Level 2, and erred on the side of caution.
The Commissioners do not consider that should be blameworthy. After
all, it would be better for them to take slightly longer to pass a baggage,
which was apparently the case, than to take chances that might create a
security risk. No one should be blamed for this. The problem would
not have mattered too much or at all but for the fact that there were other
problems that compounded its effect.

[BHS 5] Ramp handling operators (RHOs) delivered transfer bags
from inbound flights into Baggage Handling System (BHS)
after connecting flight laterals had been closed. [See para
12.26]

15.7 W30 Mr Ben Reijers, Senior Design Engineer of AA, saw
[BHS 5] as a moderate to major problem. This was obviously the fault
of RHOs concerned. However, there is ample evidence to show that the
late delivery of transfer bags to BHS was mainly caused by the delayed
delivery of baggage to the Baggage Hall, due to the various difficulties
faced by RHOs as a result of the deficiency of the Flight Information
Display System (FIDS). The problem was exacerbated by frequent
stoppages of the system including intermittent stoppages of three out of
four laterals. The Commissioners think that the problem with the
offending RHOs was caused by too many things they needed to do at the
time and they were not too familiar with the geography or the new system.
There is, however, no sufficient evidence to identify which of the three
RHOs should be responsible.

[BHS 6] RHOs did not clear bags from departure laterals in time,
427
resulting in full lateral alarms, which caused subsequent bags
to go to the problem bag area. [See para 12.27]

15.8 In respect of [BHS 6], again the evidence does not indicate
which RHOs contributed to this problem and the Commissioners are
unable to decide which of the three RHOs should be responsible.

[BHS 7] One of RHOs, Ogden Aviation (Hong Kong) Limited
(Ogden), put about 230 arrival bags from a KLM flight No.
887 onto transfer laterals. [See para 12.28]

15.9 [BHS 7] was an isolated incident caused by human error for
which Ogden had accepted responsibility.

(b) System stoppages

[BHS 8] Bags that could not be safely conveyed were not put in tubs
and OOG bags were fed into the conveyor system instead of
being sent down to the Baggage Hall via the OOG lift. [See
paras 12.31-12.32]

15.10 Whilst the Commissioners have little doubt that [BHS 8] did
happen, the evidence does not show precisely which airlines should be
held responsible.

[BHS 9] Too many erroneous emergency stops led to numerous
disruption and system downtime. [See paras 12.33-12.34]

15.11 From the evidence received, the Commissioners are unable
to determine who had pressed the emergency buttons, and whether
accidentally or deliberately. The emergency buttons could also have
been pressed to ensure safety of the staff working around the area. The
person or persons pressing the buttons, albeit causing system stoppage,
should in such circumstances not be held responsible. There is evidence
that the protruding design of the emergency button accords with
international safety regulations. In order to avoid accidental activation,
SESL has subsequent to airport opening day (AOD) introduced of a
protective glass box to enclose the button. The Commissioners are of
428
the view that the design is proper in that the button should be easily
accessible to the operators of the laterals and conveyor belts to facilitate
activation at once in case of danger. Such a design that incidentally
increases the chances of accidental activation by persons not too familiar
with the area should not properly be considered as a blemish on the
design. There is no evidence as to the identity of the persons who
activated the buttons to cause the stoppages on AOD and the Commission
is not able to find out more about the actual circumstances surrounding
the stoppages.

[BHS 10] Communication difficulties between operators in the Baggage
Hall due to Trunk Mobile Radio (TMR) overload and
unavailability of other means of communication resulted in
longer time for the system to be reset each time it was stopped.
[See para 12.35]

15.12 When an emergency stop is activated, only a certain part of
BHS will stop and the rest of the system continues to operate. On AOD,
when a stop occurred, BHS operators would have to find out why, and in
most cases they would visit the place where the stop took place or where
the emergency button had been activated. When they found out that it
was an accidental activation or the problem that required the application
of the emergency stop had been cleared, they would need to notify the
Baggage Control Room (BCR) to restart the system in the relevant area.
On AOD, however, it was not always easy to communicate through TMR
which, because of the extreme high demand by various users, was
overloaded. The resulting delay in contacting BCR was translated into
delay in restarting the affected part of BHS. If another stop occurred
nearby, the staff who had seen the first stoppage being cleared could not
be contacted by BCR to inspect the second area, and he needed to return
to BCR to take the order. While it is not able to find out the
responsibility for the stoppage, as discussed in the immediately preceding
paragraph, the TMR problem is dealt with in Chapters 9 and 16 of this
report.

(c) Delays and confusion in handling arrival baggage

[BHS 11] RHOs had no reliable flight information from FIDS and had
429
communication difficulties due to the overloading of TMR and
mobile phones and unavailability of other fixed lines of
communication. [See paras 12.37-12.38]

15.13 Flight information is crucial for the operation of RHOs in
their handling of baggage. The most important information is the
estimated time of arrival (ETA) of the aircraft that they are serving as
well as the parking stand. With these two pieces of information, they
are able to plan the deployment of their baggage tractors to the allocated
stand to await the arrival of the flight to unload the baggage and transfer
it to the Baggage Hall. Similarly, the estimated time of departure
(ETD) and parking stand are important for RHOs to send baggage to
the flight. Due to the unreliability of FIDS on AOD, TMR was used by
RHO personnel to pass these kinds of flight information. When TMR
was overloaded, some RHOs resorted to their own mobile phones, only to
find that they were also overloaded. There were hardly any telephones
installed in the Baggage Hall close to the laterals and not all the
telephones for the RHO offices were completely installed or functional.
It is not difficult to imagine that delays in baggage handling resulted.
The problems regarding TMR, mobile phones, conventional phones as
well as FIDS are dealt with in Chapters 9, 10, 13 and 16 of this report.

[BHS 12] RHOs did not use both feedlines of the reclaim carousels.
[See para 12.39]

15.14 According to Hong Kong Airport Services Ltd (HAS), the
rear feedline would only be used in exceptional circumstances with
particularly heavy baggage demands. Whilst Ogden and J ardine Air
Terminals Services Ltd (J ATS) also confirmed that they were aware of
the additional feedline, it is not clear from the evidence why it was not
used. J ATS in its submission quoted the evidence of W6 Mr Samuel
KWOK King Man, Business Support Manager of HAS, who testified that
only one lateral was working. However, W6 Kwok was merely
referring to the transfer laterals which did not seem to relate to the
conveyors for the reclaim carousels at which passengers were to reclaim
their luggage. HAS claimed that the failure to use the rear feedline
would not have slowed down the baggage handling process. Although
the use of both feedlines might have expedited the despatch of baggage to
430
passengers at the reclaim belts, the Commissioners consider that the time
that could have been saved would be slight. This problem would have
been negligible but for the other problems surfacing on AOD. This
appears to be a familiarisation problem with RHOs, to which the
Commissioners will return later.

[BHS 13] RHOs did not know the assigned lateral for arrival bags. [See
paras 12.40-12.42]

15.15 Without a usable pre-allocation template in [BHS 13], RHOs
needed to find out the allocated lateral by other means. The liquid
crystal display (LCD) board situated at the entrance to the Level 2
Baggage Hall that should indicate the match of flights to laterals was not
working, and RHOs arriving at the Baggage Hall with baggage from
aircraft could not know at a glance the allocated lateral. There were
inadequate back-up measures to address the lack of accurate information
for RHOs at the Baggage Hall. AA admitted that no whiteboard was
placed at the entrance to the Baggage Hall to direct RHOs to the correct
arrival carousel conveyors or laterals until Day Two. According to
discussions before AOD between AA and SESL on baggage handling
procedures in the event of FIDS failure, SESL was to arrange for fallback
signage at the Baggage Hall. Given the knowledge that FIDS might not
be available in the Baggage Hall on AOD, AA and SESL should have
ensured there were sufficient whiteboards to give RHOs the necessary
information.

15.16 While AA and SESL must be responsible for not having
whiteboards or fallback signage made available at the Baggage Hall to
direct RHOs to the proper areas and laterals, SESL might not reasonably
expect a FIDS failure. It is difficult to apportion the blame on the
evidence.

15.17 Counsel for the Commission submitted that SESL should be
responsible for change in the allocation since it had not followed the
templates it had given to RHOs for the pre-assigned lateral allocation.
While this argument is attractive, the Commissioners do not feel that the
evidence is sufficient to hold that it was unreasonable or improper for
SESL to disregard the template in the hope of facilitating better use of the
431
laterals, in accordance with the actual flight times. Anyhow, SESL
quickly returned to use the information on the template. The crux of the
problem was the deficient operation of FIDS, which was the cause for the
LCD board to go blank and for the difficulties experienced by the SESL
operators in BCR.

[BHS 14] RHOs abandoned unit load devices (ULDs) around arrival
baggage feedlines, causing congestion and confusion in the
Baggage Hall. [See para 12.43]

15.18 Ogden thought that this matter was more the effect of
problem baggage accumulation rather than the cause, for which no RHO
was responsible. HAS attributed this problem to AAs failure to provide
RHOs with a baggage staging area for temporary storage of problem bags,
despite many requests by RHOs before AOD. HAS alleged that on
AOD, the large number of problem bags were sorted and loaded in ULDs.
Because there was no baggage staging area, the only place available for
the temporary storage of ULDs was the space around the arrival baggage
feedlines. The Commissioners consider the arguments of both Ogden
and HAS sound. This matter was an effect rather than a cause. Insofar
as there were too many problem bags, which was the situation on AOD,
they would cause congestion, regardless of where they would be put,
unless there was a large staging area to accommodate them. However,
they were put in the ULDs which were placed around the feedlines,
which obviously caused inconvenience and even inaccessibility of the
feedlines. As there is no evidence that any other place in the Baggage
Hall was available to accommodate these ULDs, the Commissioners feel
that it might be unfair to criticise any one of RHOs. This is rather a
matter caused by the insufficient contingency planning of AA.

[BHS 15] FIDS workstation in BCR performed slowly and hung
frequently. [See para 12.44]

15.19 [BHS 15] relates to the inefficiency of FIDS which has been
dealt with in Chapter 13.

[BHS 16] There was no reliable flight information displayed on the LCD
in the Baggage Reclaim Hall (BRH). [See para 12.45]
432

15.20 When W21 Mr Michael Todd Korkowski, Site Project
Manager of Electronic Data Systems Limited (EDS) gave evidence, he
said that information from FIDS to the baggage laterals was incomplete
because of problems with the BHS/FIDS interface. This resulted in
arriving passengers not having clear, correct and timely information for
reclaiming their baggage. EDS alleged that SESL staff in BCR did not
always correctly progress the baggage status, so that information about
baggage reclaim was cleared off the display LCD boards too early, or was
delayed in reaching the LCD boards. This was denied by SESL who
attributed the problem to slow and unstable performance of FIDS.
There is ample evidence that FIDS did suffer from slow system response
as early as 6 am on AOD, and the operation on the FIDS Man Machine
Interface (MMI) in BCR was so slow (about 20 minutes for a function
to be carried through) that eventually, according to W26 Mrs Vivian
CHEUNG Kar Fay, Terminal Systems Manager of AA, the function was
taken away from BCR to be performed by staff in Airport Operations
Control Centre (AOCC). In the circumstances, the Commissioners
prefer the evidence of SESL to that of W21 Korkowski. The matter,
concerning FIDS, is also discussed in Chapters 10 and 13.

(d) Stretching of RHOs resources

[BHS 17] On AOD, RHOs had inadequate manpower deployed at the
problem baggage area to remove the large number of bags
going there. [See paras 12.46-12.48]

15.21 W30 Reijers considered that [BHS 17], namely the
insufficient resources of RHOs, was a major problem as it was clear that
there were not enough people to remove the problem bags from the
problem bag area.

15.22 HAS handled about 52% of total air traffic movements. It
had 240 staff on roster in the baggage team, the majority of whom had
been working for Hong Kong Air Terminal Services Ltd (HATS) (the
sole RHO in Kai Tak) or Cathay Pacific at Kai Tak before joining HAS
and should thus be experienced in airport operations. HAS considered
the number to be sufficient as HATS at Kai Tak had 300 staff to handle
433
100% of all the traffic movements there, taking into account the larger
airport at Chek Lap Kok (CLK) and possible disruptions at a new
airport. J ATS claimed that with the exception of the Managing Director
and one other manager, the whole of the operational arm of the company
is ex-Kai Tak. Ogden operated 15 passenger and six freighter flights per
day or approximately 9% of the total ramp handling market at the new
airport. On AOD, Ogden handled 13 departing and 13 arriving flights
out of a total of about 250 flights, or 5% of the flights. Of 230
employees, 58 were previously hired at Kai Tak and 12 by airlines at Kai
Tak or other airports. Employees without extensive experience were
trained, albeit under considerable access constraints to the Baggage Hall.

15.23 Having considered the figures in the preceding paragraph,
the Commissioners find it improbable that there would have been a
manpower shortage with RHOs, had the problem bags not been of the
unexpectedly large number on AOD. The drain on their manpower was
caused by the inefficient operation of FIDS and the other many problems
that occurred in a vicious cycle and a downward spiral on AOD. The
Commissioners do not feel that RHOs should be criticised for not
removing the problem bags in time, which came in at the rate of 15 per
minute instead of one per minute under normal circumstances. The
large number of the problem bags, in the Commissions view, could not
reasonably have been foreseen by RHOs.

[BHS 18] The Remote or Hot Transfer System, although available, was
not used to handle transfer baggage with the result that all
transfer baggage was handled only by the Central Transfer
System in the Baggage Hall, which slowed down operations.
[See para 12.48]

15.24 W30 Reijers thought that the effect of not using the remote
transfer system in [BHS 18] was minor. The Commissioners agree.
But for the other problems that surfaced on AOD, this matter would have
been negligible and might very well not have been noticed.

(e) Inexperience or unfamiliarity of airline, RHO and SESL staff

[BHS 19] Inexperience or unfamiliarity of airline, RHO and SESL staff.
434
[See para 12.50]

15.25 W30 Reijers considered the inexperience of airline, RHO
and SESL staff to be a moderate problem.

15.26 RHOs experience has been described under [BHS 17] above.
HAS and Ogden initially claimed that AA was primarily to blame for the
inexperience of their staff with operating the new system and in a new
environment. They said AA denied them adequate access to the apron
and baggage basement, ie, the Level 2 Baggage Hall, and did not provide
sufficient opportunities for them to train their operators. J ATS
maintained that their staff were well-trained and were absolutely ready
for performance on AOD. W8 Mr MacKenzie Grant, Managing
Director of Ogden, testified to claim that given the limitations of time and
the state of construction that the new airport was in, Ogden staff had been
adequately trained. Bus service and tractor markings on the apron were
in place towards the end of May 1998, giving some opportunity to train
before the final airport trial on 14 J une 1998. At the trials of 2 May and
14 J une 1998, aircraft serviced by HAS were flown into the new airport
and HAS staff simulated dummy baggage runs from aircraft to Baggage
Hall with tractors. There had been sufficient liaison meetings between
AA, RHOs and baggage handling working groups.

15.27 Counsel for the Commission submitted that had AA
organised more or more realistic trials, the staff of the airlines, RHOs and
SESL would have been better prepared for AOD. As W6 Kwok of HAS
pointed out, there were only several hundred pieces of baggage between
all three RHOs used for the trials. The relatively few bags would not
have caused a system jam. They would not pile up at the end of the
conveyor, nor activate the stop buttons. The trials did not sufficiently
simulate a real life situation where the conveyors were full of bags and
people were turning round, pulling bags off, and so on.

15.28 Viewing the evidence as a whole, the Commissioners find it
improbable that the staff of RHOs were inexperienced, because many, if
not most, of them had done the same kind of job in Kai Tak. The same
finding should also apply to the staff of SESL. SESL was the builder
and developer of BHS and probably their staff deployed to operate in
435
BCR as well as in the Baggage Hall would have been well trained for the
job and would have ample opportunities to get hands-on training with the
system during and after the system testing stage. On the other hand,
there is evidence to show that not too many or readily available
opportunities were given by AA to RHO staff to get familiar with BHS
and the environment and geography of the Baggage Hall. There must
also be a certain amount of truth in the evidence that airline staff (and of
course, Securair staff) were not too familiar with handling OOG and
baggage that needed to be put in tubs, because BHS and the working
environment was new to them. The lack of complete or accurate flight
information over the FIDS monitors and LCD boards must also have
caused them great trouble in having to answer numerous enquiries from
passengers, inconvenience and anxiety. The Commissioners come to the
view that it was more because of unfamiliarity rather than the lack of
experience or training on the part of the airline, RHO and SESL staff that
caused the problems with baggage handling on AOD, though undoubtedly
more hands-on training would have resulted in more familiarity in
operations. There were a number of incidents of human errors, such as
items [BHS 1], [BHS 5], [BHS 6], [BHS 7], [BHS 8] and [BHS 9], which
indicate the unfamiliarity with BHS and the environment. However, it
was the concatenation of such problems, which was not and could not
reasonably have been foreseen, that caused the chaos and great
inconvenience to the passengers on AOD and the few days thereafter.

15.29 The Commissioners consider that AA should bear some
responsibility for the lack of familiarity of the airline, RHO and SESL
staff with baggage handling procedures, with BHS and the working
environment. There were slippages of the construction works in respect
of the Passenger Terminal Building (PTB), and the Baggage Hall was
not made available for the familiarisation process of RHOs on sufficient
occasions. The Commissioners also opine that the unfamiliarity might
not have been so serious had AA planned and worked out with RHOs the
required resources for coping with baggage handling in case of FIDS
failure. This is a matter of lack of sufficient coordination, for which AA
should be responsible.

15.30 According to W30 Reijers, [BHS 1] to [BHS 4] combined to
create a major problem. The evidence shows that [BHS 1] involved 220
436
bags, [BHS 2] involved 600 bags, [BHS 3] involved 2,000 bags and
[BHS 4] involved 6,705 bags going for level 3 screening. Counsel for
the Commission submitted that [BHS 1] to [BHS 4], [BHS 5] to [BHS 7]
and [BHS 18] together created a major problem. The Commissioners
feel that it would not be productive to attribute to any particular item or
series of items as the major cause for the baggage chaos on AOD. Each
of the problems compounded its effect with that created by the others. It
is difficult to estimate the proportion of blame in respect of each. For
example, the Commissioners may be correct to find that 99 emergency
stoppages of BHS on AOD [BHS 9] as a major cause, but that finding
cannot fairly or properly be made without any evidence as to how long
and how serious the disruptions were to the operation of BHS. Each of
the items contributed to the problem which was serious and widespread,
but each of them in itself would not have caused a major problem and
could have been handled satisfactorily by the system and operators.

15.31 It is important not to lose sight of the fact that the
inefficiency of FIDS on AOD drained heavily on the resources of RHOs
in obtaining the necessary stand and time information, resulting in delay
in their baggage handling activities. FIDS is connected with the items
[BHS 5], [BHS 10], [BHS 11], [BHS 13], [BHS 15] and [BHS 16]. The
lack of necessary flight information also affected RHOs who had to
service passengers, such as HAS having to provide mobile steps and
tarmac buses to disembark and transport passengers at remote stands,
while all RHOs were affected in their deployment of operators for
operating airbridges for planes parked at frontal stands. The full apron
situation from midday to 5 pm and from 8 to 11 pm, requiring planes to
wait for parking stands, also affected RHOs services. Had the FIDS
problems not occurred on AOD, RHOs would have had more staff
available to assist in sorting out problem bags, and might have alleviated
or even eliminated the baggage chaos.

15.32 In the examination of the baggage handling problems,
references are made to airlines and RHOs. For the avoidance of doubt,
the Commission would point out that save where expressly indicated, no
attempt was made to pinpoint any particular airline or RHO who was
involved in a particular problem or who should be responsible for it.
The reason is that whilst the evidence shows that airlines or RHOs were
437
involved, the Commissioners do not have sufficient evidence to single out
any particular party. The Commissioners do not consider that such
approach would be unfair to any one of the airlines or RHOs. It is
appreciated that they are business competitors in their respective fields,
and any finding against anyone might damage it in its competitiveness.
Nonetheless, insofar as airlines and RHOs are found in this inquiry as a
class to be involved, the Commissioners are not persuaded that any
particular one of them will be prejudiced or should feel unfairly treated.
After all, the task of the Commission is to find out the truth as borne out
by the evidence, and the publics interest in getting to know such truth
should override any claim of prejudice suffered by a particular group or
class in the community.


15.33 The following provides easy reference and the responsibility
for each of the moderate problems will be dealt with in turn:

[26] Delay in flight arrival and departure: paras 8.27 and
12.55-12.61;

[27] Malfunctioning of the Access Control System (ACS):
paras 8.28 and 12.6212.86;

[28] Airside security risks: paras 8.29 and 12.8712.101;

[29] Congestion of vehicular traffic and passenger traffic:
paras 8.30 and 12.10212.109;

[30] Insufficient air-conditioning in PTB: paras 8.31 and
12.11012.133;

[31] Public Address System (PA) Malfunctioning: paras
8.32 and 12.13412.166;

[32] Insufficient staff canteens: paras 8.33 and 12.16712.173;

438
[33] Radio frequency interference (RFI) on air traffic
control frequency: paras 8.34 and 12.174-12.178;

[34] Aircraft Parking Aid (APA) malfunctioning: a Cathay
Pacific aircraft was damaged when hitting a passenger
jetway during parking on 15 J uly 1998: paras 8.35 and
12.179-12.190;

[35] An arriving passenger suffering from heart attack not
being sent to hospital expeditiously on 11 August 1998:
paras 8.35 and 12.191-12.193;

arriving aircraft on 25 August 1998: paras 8.35 and
12.194-12.197;

[37] A HAS tractor crashed into a light goods vehicle, injuring
five persons on 6 September 1998: paras 8.35 and 12.198;

and runway closures on 12 October 1998: paras 8.35 and
12.199-12.200; and

[39] Power outage of SuperTerminal 1 (ST1) due to the
collapse of ceiling suspended bus-bars on 15 October 1998:
paras 8.35 and 11.15.


[26] Delay in Flight Arrival and Departure [see also paras 8.27 and
12.5512.61]

15.34 Flight delay was a consequential problem caused by a
number of factors, such as the deficient FIDS, the baggage handling
chaos, the ACS and PA malfunctioning, confusion over parking of planes,
malfunctioning of airbridges, late arrival of tarmac buses, communication
problems experienced by RHOs and the other operators at the new airport
439
and the cargo handling chaos on AOD. The parties responsible for these
matters are described separately in the rest of this chapter and other parts
of the report.

[27] Malfunctioning of ACS [see also paras 8.28 and 12.6212.86]

15.35 Please refer to the paragraphs under item [27]
Malfunctioning of ACS in Chapter 12 for the nature and causes of the
problems. A lot of conflicting allegations and issues were raised by AA
and the contractor under contract C396 for ACS, Guardforce Limited
(Guardforce). Some of the allegations are set out below prior to the
Commission dealing with the question of responsibility.

(a) Delay in permit production

15.36 The Commissioners noted that in the original response from
Guardforce dated 14 September 1998, it said it would not describe the
delay to the production of permits arising out of late changes as serious.
W47 Mr Graham Morton, Project General Manager of Guardforce, stood
by this statement and testified that at all times they were able to provide
passes which allowed the system and the airport to continue operation.

15.37 Guardforce also claimed that although there had been
downtime of the system at the Permit Production Office for about 31
hours, out of a total operating time of 1,440 hours, it considered this not
to be a serious problem. It was alleged to be minimal in comparison to
the time that ACS had been operational. The system downtime was
caused by the instructions to include Chinese text on the permanent
permits (sometimes called badges) and teething problems with the ACS
software. In order to include Chinese text, the system had to change to
Windows NT 4.0 which had a known software bug that occasionally
prevented the users from being able to log onto the system temporarily.

15.38 Guardforce received AAs instructions to change the design
for the permanent staff security permits to include Chinese characters. It
then instructed Controlled Electronic Management Systems Limited
440
(CEM) to develop the necessary software. There is a discrepancy as
to the time when such instructions were issued. Guardforces reply to
the Commission stated that the instructions were not given until February
1998. At the inquiry hearing, W47 Morton however accepted that the
instructions were received in November 1997. On the other hand, CEM
claimed that it did not receive a confirmation of the order until the end of
April 1998. Guardforce claimed that the absence of Chinese text
software should not have prevented the issue of permits, as AA could still
issue effective permits without any Chinese text.

15.39 AA alleged that the CLK computer system for permit
production was not operational until 7 August 1998. Prior to August,
Guardforce used a standalone system situated at Kai Tak as an interim
measure. This Kai Tak system and its printing equipment broke down
frequently in May 1998 and the Permit Office staff had to perform tasks
by hand to complete permits which caused delay in the permit production
process. A second Permit Office was opened in J une at CLK, but the
server for the Kai Tak system could not cope with the two locations and
further breakdowns occurred. The numerous breakdowns in the
computer system in Kai Tak and at the new airport were evidenced by the
schedule of breakdown exhibited to the witness statement of Mr J oseph
WONG, Deputy General Manager of AVSECO, and a letter of complaint
from AVSECO to Guardforce about the downtime on 20 August 1998.

15.40 W 47 Morton disagreed that the CLK permanent permit
system was in use only on 7 August 1998. This appears to be
inconsistent with his evidence that the first permits with Chinese text
were produced on 8 August 1998. He agreed that the system was slow
on occasions due to a software bug. This was due to the use of the
commercial off-the-shelf product required for the Chinese software (ie,
Window NT 4.0) which had a known software bug. Also on two
occasions, power failures or power changeovers caused the downtimes.

15.41 According to CEM, there was bound to be downtime to run
the test. The system had not been fully tested before it started for permit
441
production. Also, the amount of alarms coming through the system
caused considerable downtime, which was beyond the control of CEM.

15.42 AVSECO accepted that there were some delays in the
average processing time for permanent and temporary permits. The
delay as alleged by AVSECO was caused by last minute rush of
unexpected large number of applications for permits by business partners
and the frequent breakdown of ACS. AA also attributed some delay in
permit production to the late application for permits by the applicants.

15.43 There were also problems with a lack of ink and paper for
the permanent permits between middle J une and the end of August 1998,
which put a strain on the temporary permits production as more had to be
issued. These materials could only be obtained through Guardforce
from an overseas supplier. AA's version was that an order was placed
for the paper in early J une. On 9 J une 1998, W47 Morton reported that
the shipment had gone astray and he placed a new order, delivery to be on
11 J une 1998. No delivery came. W47 Morton agreed when giving
evidence that one shipment of paper ordered in early J une went astray,
but claimed that AA did not put in a fresh order until about 22 J une 1998,
though he was not too certain about this date.

15.44 The Commissioners find that the development and
installation of ACS had been plagued by delays and various problems,
which contributed to the delay in the production of permits. While some
of the problems will be dealt with below in more detail, the
Commissioners findings regarding the delay in permit production are as
follows:

(a) AA issued the instructions for Chinese text in November
1997 and not February 1998 as at one time alleged by
Guardforce or April 1998 as alleged by CEM. CEM might
have received instructions for Chinese text in April 1998, but
that should be the responsibility of Guardforce and not AA.
CEM had warned Guardforce that these instructions might not
442
be completed by AOD, but Guardforce apparently did not
pass on the warning to AA. Again, Guardforce was at fault.
While AA should have imposed the requirement of Chinese
text in the contract or have issued the instructions for the
Chinese text earlier than November 1997, it would be unfair
that all the blame should be attached to AA. The
Commissioners feel that Guardforce should be mainly
responsible.

(b) For the breakdown of the printing equipment in Kai Tak,
Guardforce being the contractor to provide the software and
hardware of ACS must be responsible. Similarly it must be
responsible for the breakdowns caused by the failure of the
server at Kai Tak. Guardforce should also be responsible for
the lack of ink and paper.

(c) Guardforce should not be responsible for the two
occasions of downtime which were caused by power failures
or power changeovers. There is insufficient evidence for the
Commission to reach a finding if AA should be responsible
for these downtimes.

(d) The large number of last minute rush applications for
permits by business partners of AA cannot reasonably be the
responsibility of Guardforce and must be the responsibility of
those business partners, and possibly AA. AA should have
planned to avoid such late applications, and should not allow
them to disrupt the normal permit issuing process. However,
there is no sufficient evidence before the Commission for it to
make a finding that AA failed to make such a plan or that the
plan was not followed through by AA. So there should not
fairly be a finding that AA should be responsible.

(e) The questions about ACS not having been fully tested
and the amount of alarms causing disruption will be dealt with
below.

(b) ACS doors and other problems
443

15.45 On the progress of testing, AA alleged that site acceptance
test (SAT) was not completed prior to AOD because the installation had
not been completed. There were also problems with damaged doors by
contractors workers such that the system was not sufficiently stable for
testing as well as problems with the servers when large number of permits
were downloaded onto the system. W47 Morton agreed that SAT had to
be stopped so that a software development fix to the queuing problem
(Tuxedo version 6.4) with the head end computers could be loaded on the
system. It is clear from the evidence that ACS was not complete as at
AOD. The deactivation of all airbridge doors after AOD showed that
ACS was not able to perform its verification of permit function and the
monitoring function.

15.46 Guardforce alleged that model tests were delayed because
AA had failed to make available the General Building Management
System (GBMS) and Building Systems Integration (BSI).
Guardforce also alleged that the slippage of the programme was primarily
due to the late issuance of various instructions by AA, damage to
Guardforce's installed works by third parties and late completion of work
by other contractors. On the delay in the production of permits,
Guardforce put forward AA's late issuance of instructions outside its
scope of contract and the system downtime as the contributing reasons.
These instructions are as follows:

(a) In October 1997, Guardforce received AA's instructions to
design and install a temporary system at Kai Tak for the
production of security permits. CEM was instructed to
design the necessary software and systems.

(b) In November 1997, AA gave instructions to include
Chinese text in the permits. This question has been dealt
with under delay in permit production above and will not be
repeated here.

444
(c) On 2 J une 1998, Guardforce received AA's third instructions
to provide software and support to transfer data from the
temporary permit system at Kai Tak to ACS at the new airport.
Guardforce immediately instructed CEM to develop the
software.

(d) In J une 1998, AA issued further instructions to Guardforce to
increase the size of the permit system by providing five
additional computer terminals complete with installation,
software configuration and with additional printers.

15.47 AA alleged that the temporary system was required because
the permanent system, which Guardforce was to set up, was unavailable
at the time. Guardforce disagreed and alleged that at no time were they
asked to have the main system up and running before AOD. The
Commissioners consider that if Guardforce had felt that these instructions
were outside contract C396, they could have either refused to accept the
instructions or have warned AA of the risk of disruption. However,
Guardforce failed to do either. The instructions were issued as early as
October 1997, some eight months before AOD. If Guardforce accepted
the instructions, which it did, it must provide additional resources to
complete the work without allowing it to cause difficulty or disruption to
the C396 works. The Commissioners therefore find that blame should
not be attached to AA in this regard.

15.48 W47 Morton agreed that the need to transfer data from Kai
Tak to CLK was foreseen. The Kai Tak system was a temporary system,
and so there was a need to modify the software. The instructions
included taking data, which had already been transferred, to the new
system and putting it into the revised software which provided the
Chinese text. Under cross-examination, W47 Morton conceded that it
was the instructions for the inclusion of Chinese text that increased
substantially Guardforces work. The Chinese text problem has been
dealt with earlier. The Commissioners find that it is Guardforce who
should be responsible for the disruptions, if any, caused by the transfer of
445
data from Kai Tak to CLK that it had foreseen.

15.49 W47 Morton also accepted that the instructions in J une 1998
for five further computer terminals took 10 to 12 days to configure. He
agreed that the order for additional printers were in April, not J une 1998,
and that it did not cause complications. In the premises, this matter
should not reasonably be considered as a factor contributing to the ACS
problems. Moreover, the added computer terminals would presumably
have helped quicker production of permits and should not have been
treated as a problem, in particular, if Guardforce had sufficient resources
to comply with the instructions.

15.50 Another issue was raised that doors and related equipment
were damaged. AA and Guardforce alleged that physical damage to
doors and wrongful activation of break glass release buttons was a main
contributory factor to the delays in completion of ACS contract. The act
triggered alarms in the system which hampered Guardforce's ability to
test and stabilise the system. Thousands of emergency break glass had
to be replaced, sometimes with strong plastic to deter further breakage.
The Commissioners accept these pieces of evidence. These matters
certainly caused disruption to the installation and testing of ACS, but
there is no evidence as to who was the culprit of the vandalism.
Guardforce should not be responsible. The responsibility for the
damaged doors and related equipment should clearly be assigned to those
people who committed such irresponsible acts of vandalism. Those
people cannot be identified as AA was unable to catch any.
On the other hand, it may be said that AA did not provide a secure and
safe place for Guardforce to carry out its works and have its works
preserved when completed. However, there is evidence that AA did use
a lot of efforts to prevent vandalism, for instance:

(a) The problem of vandalism was recognised by AA as early as
November 1997 and steps were taken in conjunction with
British-Chinese-J apanese J oint Venture (BCJ ) to guard
against these acts by, inter alia, various steps listed in BCJ s
letter to the Commission dated 3 December 1998, including
the procurement through BCJ of a total of 230 security staff to
446
patrol PTB.

(b) AA gave instructions to AVSECO to attempt to apprehend
the culprits causing the damage by written instructions dated
respectively 20 May, 29 May, 5 J une, 12 J une and 22 J une
1998. Tenants of PTB were aware that heavy penalties
might be imposed on those who broke the rules, including the
prohibition against using unauthorised doors for access.

(c) AA issued instructions to contractors on 1 April 1998
advising the implementation of the Interim Security Measures,
which clearly stated that access to and egress from PTB were
limited to specific control points with security guards on duty.

(d) AA also instructed Guardforce to provide guards for the
communications rooms for the period from J uly 1997 through
to J une 1998.

15.51 Notwithstanding these steps, it was virtually impossible to
catch the offenders. The Commissioners feel that it may be
unreasonable to find AA responsible for not having taken sufficient steps
to prevent vandalism.

15.52 There are other problems that caused disruption to the works
on ACS:

(a) W47 Morton agreed that in some cases, Guardforce had
incorrectly installed its apparatus although he maintained that
this would not have affected the operation of ACS.
Guardforce should be responsible for these errors.

(b) W47 Morton alleged that half of the problems with the
447
airbridges were to do with the door holders and door closing
magnets, which were not within the scope of work of
Guardforce. BCJ was contracted to provide for the door
holders. On the other hand, BCJ had attributed the problems
with the door holders to changes in the AA's design intent.

(c) On the issue of late completion of works by other contractors,
Guardforce alleged that they had to wait for repairs by third
parties to door lockings, hinges and other mechanical items.
Nevertheless, W47 Morton later accepted that the delays in
such work was not as serious as previously suggested.

(d) Some alarms were set off due to operational errors. It was
said that operators selected staff rather than passenger
mode such that the door alarm sounded when the door was
held open for too long.

15.53 Other than physical damage to ACS doors, AA attributed the
causes of the major outstanding problems on AOD to software problems
with ACS and generally to the lack of resources on the part of Guardforce
or CEM to complete the works on time. Guardforce accepted that there
were various software problems, such as the queuing problem with the
head end system where the system could not handle the backlog of data.
There was a server concentrator problem which took the Distributed
Access Controllers (DACs) offline for one or two minutes or an hour.
There was also a stability problem with the head end systems although
this together with the queuing problem and the server concentrator
problem did not seriously affect the overall satisfactory operation of ACS.
W47 Morton admitted that Guardforce was responsible for the software
problem although he alleged that it was minor in nature and that large
amount of damage to their installed work, the late instructions from AA
and the late completion of other work had contributed to the difficulties
of Guardforce in finalising the software and in completing the testing.
Guardforce also admitted that it was responsible for the queuing problem
with the head end computers which was subsequently resolved by loading
a software fix (Tuxedo version 6.4) onto the system. For these software
problems, the responsibility is squarely on Guardforce.
448

15.54 W47 Morton also accepted that there were difficulties in
downloading the data to the DACs. Problems arose on AOD when the
system had to deal with some 40,000 permits being downloaded in a
single tranche. Guardforce experienced difficulties in getting the
network up and running and the data downloaded to the DACs. W47
Morton said the problem was not so much that the data had to be
downloaded in one tranche, but that the data was not consistent across all
the 200 DACs at the airport. Some DACs might have between 2 to100
cards missing. He said he would have advised AA to download it in
smaller tranches if there had been time. It was AA's choice to download
in one single tranche. He agreed however that there was less chance of
losing data if the downloading was in one tranche.

15.55 Guardforce was able to successfully download data of more
than 35,000 permit holders to every DAC only on 15 J uly 1998 and the
downloading problem was not resolved until the end of September 1998.
Guardforce alleged that the problem with data downloading was caused
by the numerous alarms of 8,000 to 12,000 a day, which took priority
over downloading information.

15.56 Subject to the observations in paragraph 15.59 below, the
Commissioners find that Guardforce should be responsible for the
downloading problems.

15.57 On the lack of resources, W47 Morton accepted that one of
the problems Guardforce faced was a lack of resources on the part of
CEM and accepted the complaints as to CEM's lack of staff were
legitimate. AA alleged that Guardforce was under a contractual
obligation to ensure that there were sufficient resources to complete the
contract before AOD. CEM denied the allegation and attributed the
problem to the late instructions from Guardforce. In the opinion of the
Commissioners, Guardforce should not shirk responsibility for the
inadequate resources that it had in performing contract C396. The lack
of resources had been raised by AA with Guardforce:

449
(a) At a meeting on 15 May 1998, between AAs staff W43 Mr
Douglas Edwin Oakervee, Project Director, W25 Mr TSUI
King Cheong, Project Manager Electrical & Mechanical
Works and others and Mr Ted Devereux (the Chief Executive
Officer (CEO) of Guardforce) and W47 Morton of
Guardforce, AA raised concern as to the lack of progress of
C396.

(b) AAs complaints as to CEMs lack of resources were
reflected in letters from AA to Guardforce, one of 6 May 1998
and another one of 10 November 1998.

15.58 Even as late as 30 November 1998, only 60% of SAT had
been completed. The Commissioners find that Guardforce and CEM
probably had resources problems, and they should be responsible therefor
respectively.

15.59 Although the major portion of the responsibility must be
Guardforces, the Commissioners have the following observations:

(a) AAs instructions would not have caused serious delays in the
C396 contract works. However, they must have caused
some hindrance to Guardforces work.

(b) The late application for permits by the business partners of
AA did cause added difficulty to ACS.

(c) Guardforce was hampered by the delay in completing and
repairing the mechanical parts of the doors of ACS.
According to BCJ , the problems with the defective door
holders were attributable to the design changes of AA. The
fact that Guardforce was awarded extensions of time would
indicate that it was affected by such delay.

450
(d) Guardforce was disrupted in its work by the damage to its
equipment caused by other contractors in PTB. Perhaps, had
the system not been loaded with so many alarms, Guardforce
would have been able to detect software problems before
AOD. For those alarms that were set off due to operational
errors, it is not clear from the evidence whether these were
caused by lack of training of the staff by AA or the airlines or
whether they were caused by the operators own faults.

(e) Some delay was also caused by AA, which did not provide in
time GBMS and BSI for the purposes of the model tests of
ACS, for which AA should be responsible.

(f) Guardforce should not be responsible for the disruptions and
delays due to the unfinished state of the construction works
and the damage to the works. AA, as the overall coordinator
of the works, should bear some responsibility for the delay in
the construction, which meant that Guardforce could not carry
out its work on a system where fitting out had finished and
vandalism was not so rampant.

(g) AA should have recognised that there would be problems with
opening doors on AOD and should have assigned staff to be
ready with keys and other means of opening locked doors.
This would have avoided the incidents of passengers being
trapped, although the incidents were more an inconvenience
than a security risk.

15.60 In the course of his evidence, W47 Morton mentioned that
the works of Guardforce under C396 for the North Shore Airfield Works
were damaged by another contractor working in the same area, and stated
that if you have people trespassing on that side of the airport, you would
not have known. This raised concern instantaneously because security
of the new airport is such an important issue. However, based on the
451
following evidence, the Commissioners do not accept that there is a
security risk as adumbrated by W47 Morton:

(a) The works being undertaken by Guardforce relate to a
construction site outside the present operational boundary of
the airport;

(b) The area where the works are being carried out is the north of
the second runway site on CLK island;

(c) Behind the line of the works and to the south of the works is a
separate fence between the construction phase of the works
and the operational part of the airport; and

(d) That fence is patrolled by AVSECO security staff under the
enhanced security arrangements in place since 13 J une 1998
and with the knowledge and approval of the Civil Aviation
Department (CAD).

[28] Airside Security Risks [see also paras 8.29 and 12.87-12.101]

(a) Delayed entry of police motorcycles into restricted area

15.61 AVSECO maintained that this incident was an isolated one
with no security risk involved. According to them, their procedures in
dealing with emergency service vehicles responding to an emergency are
clear. Those with their siren or flashing lights turned on to indicate the
urgent nature of their duties would be given immediate access. In this
incident, misunderstanding might have arisen as the siren and flashing
lights of the police motorcycles were not on. The Commissioners feel
that in ensuring prompt and effective response to an emergency, there
should be no room for misunderstanding among the parties involved of
what the correct procedure is. In this particular incident, either there
was ambiguity in AVSECOs procedures, or there had been a failure of
communication between AA and the Police. The Commissioners are
glad to learn that the relevant procedures have been fine tuned after the
incident, and that the revised procedures have worked well.
452

(b) Transit passengers allowed to enter Departures Hall and board
flight without security check

15.62 Both the Hong Kong Aviation Security Programme
(HKASP) and the Hong Kong International AirportAirport Security
Programme (HKIAASP) clearly require airline operators to ensure
security screening of their transit passengers. In this incident, China
Airlines Ltd. (CAL) clearly breached the security procedure which
requires transit passengers to be security screened before proceeding to
departure. CAL admitted this breach and apologised for it.

15.63 The Commissioners opine that CAL should also be faulted
for its failure to stop the flight in time to carry out remedial security
screening, resulting in the recall of the aircraft after it took off. CAL
argued that this was due to late instructions from AVSECO. They said
AVSECO did not give CAL a clear decision on what remedial measure
should be undertaken.

15.64 At first, the Duty Manager of CAL was only requested to
submit a written report on the incident. It was only at a later stage that
AVSECO was unequivocal about the need to recall all the transit
passengers for security screening. AVSECO however maintained that
their staff had made prompt and concerted effort to rectify the situation
by requesting CAL ground staff to recall passengers for security
screening, but the latter adopted an uncooperative attitude. Despite
obtaining verbal undertaking eventually from CAL that the aircraft would
be held up pending the off-loading and screening of the passengers, the
AVSECO Duty Security Manager noticed that the aircraft was being
pushed back. He therefore immediately requested the Air Traffic
Control Centre (ATCC) through AOCC to recall the aircraft. On
balance, the Commissioners prefer AVSECOs evidence in this respect
over CALs.

15.65 CAL further argued that they should not be the only party to
be blamed since at the time of the incident :-

(a) ACS at boarding gate 23 did not function. Had ACS been
453
operative, there would have been an effective barrier to
prohibit entry to Level 6; and

(b) the AVSECO guard stationed at the airbridge did not stop the
transit passengers from proceeding to Level 6.

On (a), Guardforce did not accept that they were responsible for the
incident because they were responsible for installing the ACS, but not for
the day to day operation of the system. On (b), AVSECO explained that
due to difficulties with ACS, a security guard was positioned at the door
(which was not locked) inside the airbridge at gate 23 connecting Arrivals
Level 5 with Departures Level 6 to prevent unauthorised access from the
Arrivals to the Departures levels. A tensa barrier was also placed across
the airbridge passage connecting Levels 5 and 6. When the AVSECO
guard saw CAL ground staff leading the transit passengers towards the
door to Level 6, he directed them to proceed to Level 5. CAL staff
however ignored him and began to dismantle the tensa barrier. The
supervisor of the guard on duty went to assist but could not stop the flow
either. According to AVSECO, the guards quite rightly refrained from
the use of force, which would be undesirable and also would have had a
potentially disastrous effect in the confined space of the airbridge. AA
concluded that the guards had done all they could to stop the passengers
from proceeding to Level 6.

15.66 The Commissioners views on Guardforces role and
responsibility in the malfunctioning ACS are set out under item [27]
above. On the performance of the guards in question, the
Commissioners accept the argument by AVSECO that the tasks of the
guards are mainly to prevent breach of security by unauthorised persons
not familiar with the HKASP and HKIAASP requirements, rather than
by properly authorised airline staff who decided to pay little regard to the
security requirements. The Commissioners also acknowledge that the
guards were outnumbered by the transit passengers. Despite these
points, the Commissioners are disappointed with the fact that the two
guards failed to intervene effectively to stop the CAL staff and transit
passengers. This is after all a situation posing serious security risk to the
airport. The guards should have adopted a more robust approach by, for
example, asking for immediate help from more guards and/or the top
454
management to intervene. This incident highlights the importance of
ACS to the security of the airport. Had the ACS door not malfunctioned
and been locked, the incident, with the resultant security risk, would
probably not have occurred.

(c) Unauthorised access to Airport Restricted Area (ARA)

15.67 The Commissioners agree generally to the analysis of
Mr Sidney CHAU, General Manager of AVSECO about the causes of the
55 cases of unauthorised entry, and are glad that the number of cases has
dropped to an insignificant level three months after AOD. Nevertheless,
the Commissioners believe it is necessary to hold AVSECO responsible
for the failure to prevent the 55 cases of unauthorised entry into ARA in
the first place, and AA for not putting up sufficient signage to indicate
boundaries of ARA. Some holders of permits of ARA are also
responsible for inappropriate use of such permits, resulting in
unauthorised entry.

(d) A KLM flight took off with baggage of two passengers who were
not on board

17.68 The incident was investigated thoroughly by CAD.
According to the investigation report, the boarding process of KLM
involved the comparison of information from the boarding gate reader
(BGR) with that in the Departure Control System (DCS) from the
check-in counter to check the number of passengers boarded and identify
any missing passenger. There was no linkage between the software of
the BGR system and that of DCS used by KLM which meant that the
boarding process was not fully automatic. The agent obtained the
number of passengers checked in at the time from DCS and input it to
BGR to set the control limit. This figure needs to be updated until the
check-in counter is closed. Upon boarding, the boarding pass (BP)
will be screened through BGR and the number of passengers boarded will
be compared with the control limit to determine whether all passengers
have boarded. The BP number of any missing passenger will be shown
in the BGR system. The agent will key in this number to DCS to trace
the passengers name for paging. In this particular case, the agent was
unable to update the control limit. As a result, the boarding process was
455
disturbed. It could not be established whether it was a human error or
the malfunctioning of the BGR system at the time.

15.69 KLM was found by CAD to be in breach of the requirements
of the HKASP for airlines to ensure that where a passenger has checked
in baggage for a flight and does not board the aircraft, his baggage is
removed from the aircraft before its departure. However, there is no
material security implication in this case because :

(a) all checked baggage of the flight was x-ray screened before
being loaded on the aircraft; and

(b) the two passengers showed up at the boarding gate when the
flight was about to take off. They had no intention of not
boarding the aircraft.

15.70 The requirement in the HKASP for airlines to remove the
baggage of a passenger who does not board the aircraft is an additional
safeguard for passengers safety. All passenger baggage is security
screened to comply with the international standard.

15.71 The Commissioners concur fully with the result of CADs
investigation. They are satisfied that this is an isolated case of failure to
comply with the HKASP requirement on passenger and baggage
reconciliation caused by human error, for which KLM should be
responsible.

[29] Congestion of Vehicular Traffic and Passenger Traffic [see also
paras 8.30 and 12.102-12.109]

15.72 The Transport Department is responsible for approval of
design, and monitoring of the operations of the transport facilities.
Citybus Limited and Long Win Co Ltd, the franchised bus companies, as
well as AA claimed that the huge and overwhelming number of sightseers
was not foreseeable. AA also alleged that it could not control the
number of buses or the number of visitors coming to the airport.

15.73 The relocation of the new airport was much publicised
456
before AOD and the public was eager to see the new airport.
Accordingly, it must be foreseeable that curiosity visitors would visit
the airport on AOD and the days thereafter. It appears that the parties
concerned, in particular, the Transport Department, did not make
sufficient planning relating to the traffic on Cheong Tat Road. From the
effectiveness of the remedial measures taken from 11 J uly 1998, such as
reducing the number of buses going via Cheong Tat Road, and
segregating shuttle buses and external buses to use different bus stops,
such measures could have been put in place before AOD, had there been
better traffic planning by the Transport Department.

15.74 Those people who were not travelling passengers, were
expected to take the E and S buses which stopped at Cheong Tat Road.
If the passenger lifts at the nearby carparks and the escalators were put
into service on AOD and if temporary signs were put in place to avoid
confusion of passengers getting into PTB via the down ramp, the
crowding problem could have been alleviated. In this respect, AA

[30] Insufficient Air-conditioning in PTB [see also paras 8.31 and
12.110-12.133]

15.75 The Commissioners will deal with the issues in respect of the
PTB area first before those relating to the tenant areas. There are
altogether 12 incidents in which air-conditioning in PTB was affected.
Each of the incidents will be dealt with separately.

(a) In PTB

15.76 (1) 6 July 1998 Carrier Hong Kong Limited (Carrier)
admitted that they were responsible for the low refrigerant pressure
switch fault, which caused chiller No.5 to trip. One of the
Commissions experts, W54 Professor Xiren CAO however stated in his
report that this type of problem might be considered as normal.
Furthermore, Carrier alleged that this should not have caused the problem
on AOD as it was a self-contained situation which did not affect the
remainder of the system. It alleged that the real problem was the
shutdown and inability to restart the chillers due to the loss or reduction
457
in sea water flow. Youngs Engineering Company Limited (Youngs)
admitted that they were responsible for the flow switch fault, which
caused pump No.2 to trip. W54 Cao also stated that that this type of
problem might be considered as normal. On the issue of communication
between the chiller plant and the pump house control systems, AEH J oint
Venture (AEH) stated that this is carried out via the bypass controller,
over which AEH has control of its functioning. AEH also stated that a
pump did start, a signal was given and increased seawater flow was
observed for a short time but shortly after the pump tripped due to a
faulty flow switch in the seawater pump house. The seawater pump
house control system is the responsibility of Youngs. Youngs denied
this allegation and stated its system did not give a signal that any
additional pumping had started up for chiller No.4 and that AEHs bypass
valve was supposed to have a logic that prevented it from modulating
until such a signal was given by Youngs control system. Without
examining the system in detail, the Commissioners are not in a position to
come to any view on which party should be responsible.

15.77 Youngs confirmed that the logic (for which it is responsible)
did try to start a faulty pump (Pump No.2) which seems to suggest that
there was problem with the control logic. The inefficient oral
communication between the pump house control room (for which
Youngs is responsible ) and the chiller rooms (for which AEH is
responsible) was caused by the poor reception of the mobile phone.
According to Youngs, contact through land telephone line was hampered
by the fact that the telephone could not be heard due to a fire alarm test
being conducted at the time.

15.78 Interface testing between the pumps and chillers were
conducted between 12 and 30 J une 1997, well before AOD. However,
the testing and commissioning had revealed the inability of the seawater
system to control and balance the seawater flow provided by the pumps to
match the needs of the chillers. In September 1997, a remedy was
devised but the order for the necessary equipment was not made until
December 1997. Delivery and installation took five months mostly due
to the time it took to acquire the valve actuators. Testing of the system
did not take place until 12 J une 1998. The timing was dictated by AA as
the testing would potentially involve a complete shutdown of the chillers.
458
Further improvement was still required and AA asked Mott MacDonald
Limited to carry out a further review which was not completed until
September 1998. The late timing of the testing meant that Youngs was
not provided sufficient opportunities to test the logic of the control
system nor did it get to test scenarios of tripping during the J une tests.
AA alleged that AEH had failed to submit complete documentation to
complete the commissioning stage and had failed to carry out sufficient
testing of the chillers. AEH denied this and submitted that the chillers
were operating 24 hours a day, 7 days a week between September 1997
and 5 J uly 1998 and had performed reliably. On the evidence, and
without cross-examining the relevant witnesses on the allegations, the
Commission simply cannot decide.

15.79 Nonetheless, in the Commissioners opinion, it appears that
the problem on 6 J uly 1998 was one of interfacing between Youngs
seawater pumps and AEHs chillers. AA should bear the responsibility
for failing to coordinate and organise sufficient interface testing between
the systems of Youngs and AEH.

15.80 (2) 10 July 1998 Youngs admitted that the tripping of the
seawater pump was due to the error of one of its pump operators. It is
therefore clear that Youngs must be responsible.

15.81 (3) 12 July 1998 AEH admitted that it was responsible for
the sudden energisation of a main chilled water branch, which could have
been avoided if it opened the valves slowly.

15.82 (4) 13 July 1998 Whilst lightning strike had caused the
chillers to trip, Youngs admitted that there was a small error in the
control logic, which was due to a missed line from the software
programme. This accordingly should be Youngs responsibility.

15.83 (5) 28 August 1998 This incident was caused by lightning
strike affecting power supply to the chillers. AA alleged that since the
incident in (4) above, to avoid fluctuation or loss in power supply,
uninterrupted power supply units (UPS units) had been installed
between 28 September 1998 and 27 October 1998 to the chiller control
panels and the panel serving the seawater controls in the chiller plant.
459
However, the instructions were issued by AA on 17 J uly 1998, some six
weeks before the incident on 28 August 1998. Had AA issued
instructions for the installation of UPS units or had them completed
earlier, or had other precautionary measures taken much earlier, this
incident and the other lightning incidents referred to in items (4) above
and (7) below might have been avoided. The Commissioners find that
AA should bear some responsibility for the late instructions.

15.84 (6) 29 August 1998 Although Youngs alleged that the loss
of power was not within its control, it should be responsible for ensuring
that the electrical protection setting was set correctly.

15.85 (7) 30 August 1998 This incident could have been avoided
had AA organised for UPS units to be installed or taken other
precautionary measures much earlier. The comments under item (5)
apply here.

15.86 (8) 8 September 1998 Youngs admitted that it was
responsible to the extent that the system was vulnerable due to critical
control circuits not being on a dedicated supply.

15.87 (9) 14 September 1998 This was a single incident of
human error to which no responsibility should be assigned except to the
person who committed the error. If that person was an employee of the
contractor for the Mechanical Building Management System, then that
contractor should, in the opinion of the Commissioners, be responsible.

15.88 (10) 12 October 1998 The damage to China Light & Power
Company Limiteds (CLPs) underground cable was caused by a third
party contractor which has not been identified.

15.89 (11) 22 October 1998 As this was a planned shutdown, no
one should be responsible.

15.90 (12) 28 November 1998 There is inconclusive evidence as
to who should bear responsibility for this incident. It may be that AEH,
as the contractor who is responsible for the installation of the UPS units,
should have correctly set the UPS unit in the appropriate mode of
460
operation. Youngs alleged that the unauthorised isolation of power
supply was due to the failure of Airport Management Division (AMD)
to provide more stringent access controls to the seawater pump house.
AA contended that it had not decided to implement more stringent
methods of access control as the cause of the incident was not known
and that if the current monitoring exercise identified a cause, appropriate
action would be taken.

(b) In tenant areas

15.91 Much of the responsibility for this problem should lie with
tenants themselves. Late applications for the connection to AAs chilled
water supply and non-compliance with the procedure for this connection
by tenants led to a large volume of requests for connection in the few
days immediately prior to AOD. As a consequence AEH was unable to
respond within the usual time frame. The Commissioners do not think
such tenants should or should be in a position to complain about the late
supply of air conditioning to their premises. The Commissioners are not
able to find sufficient evidence to [hold AA responsible] for not having
had a closer coordination with the tenants or a more effective
management over the tenants in this respect.

[31] PA Malfunctioning [see also paras 8.32 and 12.134-12.166]

15.92 As described in Chapter12, there were hardware and
software problems with PA. The responsibility for the software
problems will be dealt with first. For the Central PA, the problems with
the stability of the manual all zone (MAZ) notebook and the locking up
and latching were the most serious. Most software problems were
caused by the required tests not being performed by AOD. For example,
the intelligibility problem could have been eradicated had there been a
Rapid Assessment of Speech Transmission Index (RASTI) test for
assessing the rapid speech intelligibility index of the actual sound of PA.
But this test could only be usefully and meaningfully done after PTB was
completed with acoustic related materials, and such materials had not
been installed prior to AOD. The result was that AA and Hepburn
Systems Limited (Hepburn) agreed that RASTI tests should be deferred
after AOD. This should not be considered as a fault of AA or Hepburn.
461
PTB was not completed in such a way as to usefully have the tests, and
that was caused by the various slippages of the construction works.
There is no evidence to show that the slippages were the fault of AA as
opposed to the various contractors who were responsible for various
pieces of works to build PTB. For this reason, the Commissioners are
not prepared to hold either AA or Hepburn responsible for the problem of
intelligibility of PA.

15.93 The zoning and priority problems only required minor and
quick adjustments to be made, and they alone can be considered to be
minor and teething problems. Again these problems could have been
uncovered and remedied had there been more tests and trials with airlines
that were going to use the PA. Based on the evidence that Hepburn
delayed in its work on PA, Hepburn should be primarily responsible.

15.94 The slow response time of the consoles that inconvenience
the users, being a software problem, must be the responsibility of
Hepburn.

15.95 Similarly, the overriding problem, which is a one-off
incident, was a matter of software for which Hepburn should be
responsible.

15.96 There were different allegations by Hepburn and AA over
the MAZ console outages on AOD which prevented the Central PA from
functioning. MAZ notebook outage, as alleged by Hepburn, originated
from the earthing problems of the communications rooms. The earth
lines between the rooms were electrically noisy and were interfering
with the data communication between the control room and the MAZ
notebook. W47 Morton of Hepburn agreed, however, that he had no
documentary evidence to support the claim of electrical interference and
that the possibility of a noisy earth link was more like a deduction than an
actual finding.

15.97 Hepburns allegation was denied by AA. AEH, the power
supply contractor, stated that it had no record and no recollection of AEH
being asked after AOD to investigate specific problem with power
supplies to the PA equipment and that no rectification work had ever been
462
done.

15.98 AA claimed that the MAZ outage was mainly a software
problem. Inadequate testing, insufficient resource and fire alarm
latching problem were the causes.

15.99 The locking up or latching problem was apparently not
related to the fire alarm system itself. The Fire Services Department
(FSD) used the Audio and Visual Advisory System (AVAS) which
provided audio/visual indication of safe direction of egress from the area
affected by fire. AVAS has an interface with PA. FSD had conducted
extensive inspections on the AVAS since October 1997. A final
inspection was conducted on 26 J une 1998 and the result was found
satisfactory.

15.100 AA alleged that the locking problem was not just a one-off
manifestation of the problems with PA but rather the cause of all the
failures of the Central PA. Hepburn denied that. W47 Morton said the
locking problem did not appear until late J uly and in any event did not
stop PA from operating. According to Hepburn, the system was putting
out about 270 calls per day from AOD. W47 Morton conceded,
however, that the locking problem was caused by a software problem to
which Hepburn was responsible. In the cross-examination of W47
Morton by AA, it would appear from an internal e-mail of AA dated 28
J une 1998 from Mr Peter W H WONG, Senior Design Engineer, to Mr
Alastair Blois-Brooke, Senior Construction Manager, and copied to W25
Tsui of AA that Hepburn and SigNET (AC) Limited (SigNET) were
aware of the locking problem prior to AOD.

15.101 Despite the conflicting allegations, the Commissioners are of the
view that the system instability problems and those relating to locking
and latching causing console outages, being software problems as
accepted by W47 Morton, should be the responsibility of Hepburn.
These problems on the evidence were not probably caused by the noisy
earth lines.

15.102 The hardware problems tended to be more localised. As far
as the physical damage to the membranes and gooseneck microphones are
463
concerned, Hepburn alleged that AA should ensure that PTB was a secure
area in the period leading up to AOD. AA however maintained that the
activities described could only been the result of random and gratuitous
vandalism by unknown persons, and submitted that the responsibility for
such damage must rest primarily with the persons who inflicted it on the
local consoles. The Commissioners do not fully agree. While the
damage might have been caused by vandalism, and the vandals must be
responsible in such a case, operators could have carelessly inflicted the
damage, such as using a ball-pen to poke at the membrane. The
Commissioners think that AA should send an advice and warning to all
users and possible users on how to use the consoles properly, so that
unknowing damage of this kind can be prevented.

15.103 Hepburn also alleged that AA should be responsible for the
incomplete installation of some PA equipment. AA admitted that the
ambient noise-operated amplifier facility was not installed prior to AOD
although it claimed that the absence of this facility did not prevent the
operation of PA. This goes back to the lack of time and PTB not being
absolutely ready for the installation of PA equipment, referred to in
paragraph 12.141 of Chapter 12 above.

15.104 AA alleged that Hepburn failed to ensure the completion of
the testing and commissioning of PA before AOD. Hepburn argued that
over a number of months before AOD, it had been agreed between
Hepburn and AA that SAT would be carried out after AOD (but no time
or date was fixed). In addition, the conditions for SAT to take place did
not exist prior to AOD and BSI and the Voice Routing System (VRS)
were only available at the end of October 1998. Hepburn alleged that
SAT could not be carried out without BSI and VRS being available.
However during cross-examination, W47 Morton qualified the above
statement by saying that, while RASTIs were agreed to be postponed
until after AOD, those would not constitute the majority of SAT testing.
W47 Morton admitted that they had problems with a Hong Kong
sub-contractor, Univision Engineering Limited, and Hepburn had to
change the sub-contractor and it affected the development of an interface
software to BSI. AA alleged that this problem resulted in a delay of
factory acceptance test (FAT) which was only completed at the end of
J une 1998. Indeed, SATs, including the RASTI testing, were only
464
completed at the end of October 1998.

15.105 Complaints were made against Hepburn and SigNET for
lack of resources and shortage of specialist engineers thereby creating
delay and resulting in insufficient testing. As reported in AA Board
Paper 194/98 dated 14 J uly 1998, the confidence level of AAs
Management in Hepburn was low and consideration was given to appoint
another contractor to install an alternative Local PA. On the other hand,
AA was accused of frequent change of instructions and poor
coordination.

15.106 Hepburn denied AAs allegation that it was under-resourced
but W47 Morton of Hepburn later admitted that on two occasions,
software specialists were not available even upon the request of AA.

15.107 SigNET, the subcontractor, denied that resource was ever
withheld from the project or that maximum resource was not applied at
all times in accordance with the needs of the project. SigNET, however,
admitted that it is a small company with finite limits to some areas of
expertise. SigNET entered the project at a relatively late stage and it
had to contend with frustration due to late responses from Hepburn or AA,
frequent changes of priorities and instructions, and insufficient support
from Hepburn. Due to the specialist knowledge and experience required,
it was impossible for SigNET to expand quickly to meet the requirements
with all these additional constraints.

15.108 From the evidence, the Commissioners find that while a
small part of the delay in the commissioning and testing of PA could have
been caused by the late readiness of PTB, the major delay was caused by
Hepburn in failing to keep the contractual deadlines. There might have
been frequent changes of instructions given by AA to Hepburn, but that
should normally have been covered by extensions of time granted.
Hepburn was awarded an extension of time, but even with the extension,
the revised completion date for the system was 15 April 1998. The
delay can be easily appreciated when SATs were only completed in late
October 1998, long after AOD. The main reason seems to be the
inadequate resources that Hepburn and its subcontractor SigNET had
assigned to the contract. For the delays and inadequate resources,
465
Hepburn and SigNET should be responsible.

15.109 There is no evidence that the problems with PA caused
passengers to miss their flights. However, the general picture is that the
PA problems added to the confusion on AOD, contributing to the
impression that the new airport was not ready for opening.

15.110 CSE International Ltd (CSE), AAs consultants on systems,
had repeatedly flagged up problems in the development of the PA. For
instance, in its software evaluation report of 26/3/98 it had pointed out
that the delivery of the remaining SigNET software and formal onsite
testing continued to slip. AA knew PA would not be completed or
completely tested before AOD, and W44 Mr Chern Heed, Airport
Management Director of AA, admitted in evidence that prior to AOD, he
knew that PA had not gone through SATs and that there were problems
with the Local PA.

15.111 PA always featured with the AA management as a
contingency measure in the case of FIDS failure. It would be used as a
backup for the dissemination of information in such a case. However,
W44 Heed admitted that he had not considered the problems of
disseminating gate change information in the case of a FIDS failure and a
full apron, not thinking that this situation would arise. He agreed that if
PA failed at that time there would be problems. He had not planned for
the possibility that PA (for making central announcements) might not
work at the same time that FIDS did not work. This failure of overall
contingency planning is dealt with in Chapter 17.

[32] Insufficient Staff Canteens [see also paras 8.33 and 12.167-12.173]

15.112 The problem of insufficient staff catering facilities does not
impact on the operations of the new airport directly. However, since a
large number of people working at the new airport are affected, the
Commissioners classify it as a moderate problem.

15.113 The Commissioners consider that AA should be criticised for
its poor planning of staff catering facilities. There does not seem to
have been a scientific and realistic assessment of the requirements taking
466
into account the number of staff working there. In his expert report,
W51 Mr J ason G YUEN, expert for the Commission, opines that the
planning ratio of 19 to 1, assuming 15,000 people (14,600 actual) as
against 800 (since mid-October 1998, 954) seats in staff canteens,
appeared to be very low. Though some members of the people working
in the airport might bring their own food or might go out to Tung Chung
for their meal, still the ratio of persons to a seat being 15 to 1 (14,600
persons to 954 seats) appears high. One would have to remember that if
staff do not bring their own food, a trip from the new airport to nearby
Tung Chung and back on bus will take sometime, causing some
inconvenience to the staff.

15.114 While the Commissioners have not found relevant evidence
to determine if the planning ratio is reasonable or not, the fact that more
catering capacity was once planned by AA at some stage but was
somehow dropped due to unknown reasons goes some way to reinforce
the suspicion that the existing provisions might not be adequate.
Even if the total capacity of the existing four staff canteens is sufficient to
cater for the actual daily needs, AA should be responsible for not having
been able to ensure that all the four planned canteens could open for
business on AOD. This is particularly so since the large number of
sightseers on AOD is certainly something which should have reasonably
been envisaged. When all kinds of problems surfaced on AOD, the
inadequacy of staff canteens would certainly have caused difficulty to the
staff of the airport community who were already exasperated with the
chaotic conditions.

[33] Radio Frequency Interference on Air Traffic Control Frequency
[see also paras 8.34 and 12.174-12.178]

15.115 On the basis of the information provided to the Commission,
the Commissioners find that both the Hong Kong and Mainland
authorities attach great importance to flight safety and strenuous efforts
are being made with a view to eliminating radio frequency interference
completely. No finding is therefore called for in respect of this problem.

[34] APA Malfunctioning: a Cathay Pacific Aircraft was Damaged
when Hitting a Passenger J etway during Parking on 15 J uly 1998
467
[see also paras 8.35 and 12.179-12.190]

15.116 Generally, on the causes of the malfunctioning of APA, there
were crossed allegations between AA and Safegate International AB
(Safegate). J udging from the evidence, and in particular having
regard to the fact that Safegate had to take some remedial measures
towards APAs in the new airport after AOD, the Commissioners come to
the view that Safegate should be responsible for these general causes.

15.117 In respect of the accident on 15 J uly 1998, AA and Cathay
Pacific both alleged that it was caused by the malfunctioning of the
particular APA. This was denied by Safegate who said that the APA
system was at that time still being tested. Safegate, however, accepted
that the auto-calibration system of the particular APA had been
inadvertently disabled by its staff. As a result, the problem with the
inoperative laser sensor was not identified in the auto-calibration process.
This is the root cause of the accident. First, the laser sensor was not
working. Secondly, the auto-calibration process could have revealed the
sensor fault and warned the air pilot, but this checking process had been
disabled by a Safegate staff during testing who apparently forgot to
re-activate it after testing. Safegate should be responsible.

15.118 The air marshall was too far away from the operator panel to
switch on a signal on the APA to show the pilot to stop. He was in a
dilemma, because it would take him some time to reach the operator
panel, when the aircraft was still moving forward. He chose to make a
hand signal to the pilot to halt.

15.119 On the location of the operator panel, Safegate said it had
previously advised AA to install the panel within reach of the air marshall.
AA denied this but agreed that the control panel should be repositioned.
It appears from this allegation that it was contractually a matter for
Safegate to advise. However, there is insufficient evidence before the
Commission for it to reach a conclusion whether Safegate had the duty
and if so, whether it had in fact advised AA, or whether AA ignored the
advice, or whether AA should have made a correct decision even without
any advice from Safegate. On this matter, the Commissioners can make
no decision. The happy news is that AA agrees that the control panel
468
should be repositioned, obviously to enable the air marshall to make use
of it even while he is standing on the ramp manually directing aircraft
parking.

15.120 It was alleged by AA that the pilot was also suspected of not
being familiar with the docking routine of how APA operates. He
allegedly misinterpreted the floating arrow signals of the APA and the
emergency stop marshalling signal as move ahead signal, such that
he continued to taxi forward until overshooting the stop bar by six metres.
Similar allegation was made by Safegate against the pilot.

15.121 Cathay Pacific responded by alleging that the air marshall
and the pilot were both victims of a degraded, deceptive and poorly
configured guidance system. It further alleged that the stop signal might
not have been properly given by the air marshall as it was a rather frantic
attempt by him to intervene once he realised the APA was not working.
The pilots evidence was that the air marshall arrived from a position
below the right hand side of the aircraft very shortly before the aircraft
stopped, which was estimated to be 1.5 metres prior to coming to stop.
AA had a different version of this. It alleged that the air marshall gave
the emergency stop signal when the aircraft was about 12 metres away
from final stop position.

15.122 Cathay Pacific admitted that the pilot had no experience with
parking at a frontal stand at the new airport as in this case. With the
difference in perspective between a remote stand, being very close to the
bay number board, and a frontal stand where the terminal is further away,
the pilots perspective would be different.

15.123 Neither the pilot nor the air marshall have been summoned to
give evidence before the Commission, because of time constraint.
When a situation described in the allegations occurs, without the
assistance of oral testimony, it is impossible to judge which version, if at
all, is true. It was alleged either the air marshall was giving incorrect
signals or the pilot did not have a full understanding of such hand signals.
The Commissioners do not believe either allegation being probable.
Both of them are experienced in their own field and had gone through
rigorous training with constant refresher courses. Apart from this
469
observation, the Commissioners do not feel they can make any other
finding on the conflicting allegations as to how the accident actually
occurred.

[35] An Arriving Passenger Suffering from Heart Attack not being Sent
to Hospital Expeditiously on 11 August 1998 [see also paras 8.35
and 12.191-12.193]

15.124 The Commissioners note that the cabin crew did not notify
the Apron Control Centre (ACC) or AOCC about the sick passenger on
board before landing. Hence, no arrangement had been made to put an
ambulance on standby on arrival of the aircraft. China Southern Airlines
confirmed that ambulance service was called after the aircraft had landed.
The thirteen minutes spent waiting for the ambulance to arrive could have
proved critical to patients requiring immediate emergency treatment. It
was fortunate that in the case in question, this did not result in a major
incident. Nevertheless, the Commissioners have to hold the China
Southern Airlines responsible for failing to notify the airport about the
sick passenger before landing. AA informed the Commission that after
this incident, airlines had been reminded that if a passenger was taken ill
on an inbound flight, the flight crew should notify the airport before
landing so that an ambulance can be standing-by on arrival of the aircraft
at its parking stand.

15.125 Thirteen minutes is not an unreasonable response time in this
particular incident. Indeed, both the ambulance and the ACC escort
arrived at their destinations within their normal response time. However,
the Commissioners found that if there had been better coordination and
communication between Fire Services Communication Centre (FSCC)
and ACC, the response time in emergencies like this can be further cut
down. In the chronology set out in paragraph 12.192 of Chapter 12, the
ambulance had to wait at the apron gate for five minutes for the ACC
escort vehicle to arrive. This was partly because:-

(a) FSCC contacted ACC indirectly through Airport Main Fire
Station Rescue Control (AMFSRC) for an escort. Had
FSCC contacted ACC directly, the response time could have
been reduced by at least one minute; and
470

(b) ACC was not contacted for an escort immediately upon the
CLK Fire Station being alerted to the dispatch of an
ambulance.

15.126 On (a), it has already been mentioned in paragraph 12.193 of
Chapter 12 that arrangement is being made for a direct line between
FSCC and ACC to be installed. On (b), FSD explained that this
procedure was laid down because, according to FSCC standard
operational procedure, a single and straightforward ambulance case
should be handled by one FSCC console operator who would carry
through all despatch and information dissemination action. This
arrangement was considered the most effective due to the large number of
emergency ambulance calls received by FSCC each day. It would help
minimise possible omissions. While the Commissioners appreciate the
reason for the arrangement, they suggest that FSCC should contact ACC
for an escort immediately upon the Chek Lap Kok Fire Station being
alerted to dispatch an ambulance. This may involve some changes to
the existing mode of operation but should help achieve a better
coordination between AA and FSD and even better response time in an
emergency.

Arriving Aircraft on 25 August 1998 [see also paras 8.35 and
12.194-12.197]

15.127 The procedure for vehicles entering the runway is clear and
unmistakable. All relevant communication equipment was functioning
properly and was not a contributing factor to the incident. It was the
Rescue Leader and the driver of the first fire engine who were responsible
for the failure to obtain clearance from ATC before crossing the runway.
They had been admonished and disciplined by FSD subsequent to the
incident. The staff of the rescue appliances involved in the incident
have all been warned and reminded that the airside safety driving
regulations should be strictly adhered to at all times. FSD also reminded
its personnel of the proper procedures for appliances to seek permission
from ATC before entering the runway. The Commissioners opine that
FSD had taken appropriate follow-up action on the incident.
471

Persons on 6 September 1998 [see also paras 8.35 and 12.198]

15.128 The Commissioners consider this incident of a moderate
nature not only because five persons were injured but also because it was
a traffic accident inside the restricted area of the new airport. It is
necessary to maintain the new airport as a safe place, and traffic accidents
within the restricted area may give rise to an impression to the public that
the airport itself is not running safely and smoothly. The incident report
of HAS found that the driver of the tractor had not followed the proper
driving procedures in stopping his tractor to ensure road clearance in
front when he was driving between two lines of containers. He was
regarded as having failed to pay due care and attention. As a result of
Police investigation into the incident, the driver was prosecuted for
careless driving. The Commissioners are satisfied that HAS and the
Police have investigated into the incidents thoroughly and have no further
comment on the incident.

Runway Closures on 12 October 1998 [see also paras 8.35 and
12.199-12.200]

15.129 The Commissioners treat this incident as a moderate one
because it led to closure of the runway on three occasions and affected a
large number of flights. At the time of the incident, the relevant
freighter aircraft was operating under a wet lease agreement between the
United Arab Emirates and the Atlas Air, Inc (Atlas Air) and was fully
controlled by the Atlas Air crew. Accordingly, Atlas Air has to be held
responsible for the incident.

[39] Power Outage of ST1 due to the Collapse of Ceiling Suspended
Bus-bars on 15 October 1998 [see also paras 8.35 and 11.15]

15.130 This matter has been dealt with in paragraph 11.15 of
472
Chapter 11.

473
CHAPTER 16

RESPONSIBILITY
TEETHING AND MINOR PROBLEMS

Section 1 : Teething and Minor Problems


Section 1 : Teething and Minor Problems
16.1 In this chapter, the Commission recapitulates what the
teething and minor problems are and reviews who should be responsible
for them. Details of all teething and minor problems can be found in
Chapters 8 and 9. The following provides easy reference and the
responsibility for each of the problems will be dealt with in turn:
[1] Mobile phone service not satisfactory: paras 8.9 and
9.2-9.8;
[2] Trunk Mobile Radio (TMR) service not satisfactory:
paras 8.9 and 9.9-9.16;
[3] Public telephones not working: paras 8.9 and 9.17-9.22;
[4] Escalators breaking down repeatedly: paras 8.10 and
9.23-9.30;
[5] Insufficient or ineffective signage: paras 8.11 and
9.31-9.35;
[6] Slippery and reflective floor: paras 8.12 and 9.36-9.39;
[7] Problems with cleanliness and refuse collection: paras
8.13 and 9.40-9.49;
[8] Automated People Mover (APM) stoppages: paras 8.14
and 9.50-9.59;
[9] Airport Express (AE) ticketing machine
malfunctioning: paras 8.15 and 9.62-9.64;
473
[10] AE delays: paras 8.15, 9.60-9.61 and 9.65-9.68;
[11] Late arrival of tarmac buses: paras 8.16 and 9.69-9.74;
[12] Aircraft parking confusion: paras 8.17 and 9.75-9.78;
[13] Insufficient ramp handling services: paras 8.18 and 9.79;
[14] Airbridges malfunctioning: paras 8.19 and 9.80-9.84;
[15] No tap water in toilet rooms and tenant areas: paras 8.20
and 9.85-9.95;
[16] No flushing water in toilets: paras 8.20 and 9.85-9.95;
[17] Urinal flushing problems: paras 8.20 and 9.96-9.110;
[18] Toilets too small: paras 8.21 and 9.111-9.122;
[19] Insufficient water, electricity and staff at restaurants:
paras 8.22 and 9.123-9.136;
[20] Rats found in the new airport: paras 8.23 and
9.137-9.139;
1998: para 9.140;
[22] Traffic accident on 28 August 1998 involving a fire
engine, resulting in five firemen being injured: para 9.141;
[23] A maintenance worker of Hong Kong Aircraft
Engineering Company Limited (HAECO) slipped on the
stairs inside the cabin of a Cathay Pacific aircraft on 3
September 1998: para 9.142;
flights: para 9.143; and
on 1 October 1998: paras 9.144.

16.2 In respect of some of the problems set out above, it is not
possible to make a finding on responsibility. This is where the evidence
received by the Commission does not enable the Commissioners to come
to a conclusive view on the question of responsibility and because the
Commissioners decided from the commencement of the inquiry that
valuable time should be better used for investigating other problems
474
whose impact on airport opening day (AOD) was more serious. In
such cases, the Commission will set out the allegations of the parties
involved, and highlight the key allegations. In other cases, the problem
itself may simply be an accident or fact of life for which no one should be
held responsible. Nevertheless, whatever the determination on the
issues is, if a determination can be reached, the Airport Authority (AA)
must be primarily responsible for the inconvenience and inefficiency
suffered by airport users in the Passenger Terminal Building (PTB) on
AOD and a few days thereafter, because after all, AA has under the
Airport Authority Ordinance its functions to discharge in operating the
new airport and operate it in an efficient manner.
[1] Mobile Phone Service Not Satisfactory [see also paras 8.9 and
9.2-9.8]

16.3 The evidence shows that SmarTone Mobile Communications
Limited (SmarTone), Hutchison Telecommunications (Hong Kong)
Limited (Hutchison) and Hong Kong Telecom CSL Limited (HKT)
as the mobile phone network operators did not provide an efficient or
adequate mobile phone network for their users at the new airport on AOD.
However, that is entirely a matter between these three operators and their
own customers, with which the Commission should not be concerned.
In their written submissions, all three operators relied on the unforeseen
breakdown of Flight Information Display System (FIDS) to refute their
responsibility for inadequacy of network capacity. There is of course
certain truth in the argument that the problem has been caused mainly by
the deficiency of FIDS, the TMR systems being overloaded and the
inoperation of about two-thirds of conventional public telephones planned
for PTB on AOD. These and the other events resulted in a sudden
upsurge in the public demand for mobile phone services. Although it
might have been foreseen that there would be many sightseers visiting
PTB on AOD, the Commissioners agree, after viewing the evidence, that
the concatenation of the many problems facing AOD that might further
increase the demand for mobile phone service was not properly
foreseeable by the mobile phones operators. The Commissioners feel
that it would be unfair, in the circumstances, to hold any of the three
operators responsible for underestimating the required capacity of their
networks. The fact that SmarTone added channels to the Common
475
Antenna System (CAS) on 7 J uly 1998 may mean there was
insufficient capacity in the CAS as a whole. This evidence is, however,
insufficient to enable the Commission to come to the conclusion that the
antenna system capacity is inadequate for shared use by the mobile phone
network operators. SmarTone has initiated coordination with all the
mobile phone operators for expansion of their equipment capacity in
order to supplement existing networks.
16.4 AA should have given advance warning to the mobile phone
operators of the possibility of heavy demand on the use of mobile phones
in the event of FIDS failure. Such advance warning could have enabled
the operators to take this into account in designing their respective
network capacity and, could therefore have avoided the problem. In any
case, the problem of inefficient mobile phone service was short-lived and
vanished from the second day of airport operation. In terms of its own
impact on airport operation, it should only be a minor problem.
[2] TMR Service Not Satisfactory [see also paras 8.9 and 9.9-9.16]
16.5 TMR is an important component of the communications
network of the new airport and has an impact on the overall efficiency of
airport operation. As with the mobile phone network, the TMR network
was overloaded because of the huge demand for TMR service, which
arose from the lack of flight-related information from FIDS. The
Commissioners are of the opinion that the inadequate capacity of the
TMR due to the unexpectedly high volume of usage was not reasonably
foreseeable by TMR operators. It might be argued that the problem of
TMR overloading was not a general phenomenon affecting all operators
because AA did not seem to have a problem with its TMR on AOD.
Apart from that, however, there is no other evidence sufficient for the
Commissioners to alter their aforesaid opinion. On the evidence,
therefore, both Hutchison and China Motion United Telecom Limited
(CMT) should not be held responsible for the insufficient network
capacity of their TMR services on AOD.
16.6 There is evidence to suggest that the Hutchison network
suffered from weak signals. Due to the delay in the completion of the
antenna farm by AA, Hutchison located its main base station outside the
476
airport perimeter in Tung Chung which resulted in weaker signal for its
TMR users. This was further affected by the inoperation of the common
antenna system for Hutchison and CMT TMR due to the inoperation of
the link between common antenna system and the Tung Chung base
station. Hutchison attributed the problem to its contractor in supplying
an inappropriate connector. The Commissioners believe that the
problem of weak signal could have been anticipated by Hutchison and
adequate counter measures should have been provided to overcome the
problem prior to AOD. Moreover, Hutchison should be primarily
responsible for the failure to put in place a link between common antenna
system and the Tung Chung base station for operation on AOD.
16.7 On the part of AA, the Commissioners consider that, as the
operator of the airport, AA is responsible for the delay in completing the
antenna farm for use by Hutchison and CMT. According to Mr Edmund
SIN Wai Man, Director of Engineering of Hutchison, AA wrote to
Hutchison as early as March 1997 with the idea of having an antenna
farm but then advised it in February 1998 that the facility would not be
ready for use until late October or even December 1998. As the
provision of the antenna farm would have facilitated operational
efficiency of the TMR system, AA should have taken steps to ensure that
the facility was available prior to AOD. In its written submission, AA
argued that there was delay because no interested parties came forward
immediately when it sought business plans in late 1997. It was not until
February 1998 that HKT agreed to construct and operate the antenna farm
and an agreement to that effect was signed in August 1998. AA also
defended its rejection of Hutchisons request to install antennae on the
Cathay Pacific Catering Services building for outdoor TMR coverage on
the basis that the rejection was consistent with AAs policy to have all
antennae located in the antenna farms. AA had not received any further
similar request from Hutchison until some time in August 1998.
However, the Commissioners do not consider that these facts should in
any way reduce AAs responsibility to ensure the timely completion of the
antenna farms for use on AOD.
16.8 AA should have forewarned airport operators and the two
TMR providers of the possible heavy demand on the use of TMR in the
event of FIDS failure. As part of the contingency or workaround
477
measures for FIDS, AA should have advised the TMR operators to take
this into account in designing the TMR capacity of their respective
systems.
[3] Public Telephones Not Working [see also paras 8.9 and 9.17-9.22]
16.9 Both AA and International Computers Limited (ICL)
accepted that there was delay in the completion of the cabling and
jumpering work. This hampered the progress of the work of New World
Telephone Limited (NWT), including the testing of cabling circuits and
the payphone network. This resulted in about two thirds of the planned
number of public telephones not being made available on AOD. As to
the other operational problems with those phones that were working, such
as phones not accepting coins, NWT did not deny responsibility.
Nevertheless, the effect of the problem was minimal in view of the small
number of phones involved.
16.10 Close to AOD, there was a saturation of the originally
designed backbone cabling system which became insufficient to cope
with users demand. This resulted in the late issue by AA in May 1998
of variation and further instructions for additional cables in PTB and for
tenant jumpering work. ICL alleged that the late instructions effectively
doubled its work and stretched its resources. This also did not take into
account the lead time needed to procure materials. AA accepted that
some of the instructions were made in the last two months before AOD
but contended that this was due to the late request by tenants for
additional cables during that period. AA also alleged that delay was
caused by the late order of materials by ICL.
16.11 ICL alleged that despite being requested to do so since
November 1997, AA did not issue tenant jumpering requirements until
April 1998. When it did issue these requests, they were not in the
agreed form and were often made directly to staff on site. Information
was therefore often duplicated or altered, causing delay to completion.
16.12 It appears to the Commissioners that the substantial increase
in the volume of tenant jumpering work immediately before AOD was
foreseeable. Back in November 1997, it was agreed between AA and
478
ICL that pre-emptive jumpering should be carried out before the end of
J anuary 1998 to reduce the expected volume of work. Once completed,
the tenants request for jumpering work could be done simply by
allocating the circuits. AA alleged that ICL did not complete the work
until May 1998. This adversely affected ICLs resources and timetable.
AA also complained that there was a backlog of jumpering records to be
updated by ICL and that this affected the progress of the work as AA had
no way of knowing exactly what had been completed on site.
16.13 AA had accepted responsibility for the problems relating to
cabling and jumpering vis--vis NWT. W25 Mr TSUI King Cheong,
Project Manager E&M Works of AA, agreed that AA was responsible for
the failure to provide cable connections to NWT. However, it is not
clear to what extent (if any) should ICL be responsible for the cabling
problem, by reason of the allegations made against it by AA. Without
hearing all the witnesses from AA and ICL on this issue, it would not be
possible for the Commissioners to attribute responsibility between AA
and ICL. Their conflicting allegations were various and would require
the detailed examination of witnesses, which the Commission could ill
afford to do in the limited time available. They are matters of a
contractual dispute which should only be properly resolved through
arbitration or litigation.
16.14 AA also alleged that the late submission by tenants of their
cabling requirements contributed to the problem. The Commissioners
are of the view that the tenants should not be responsible for problems
despite their late submissions. The reason is that the tenants could not
reasonably have imagined that their late submissions would result in a
large number of public phones not being connected. Their late
submissions might reasonably result in their requirements not being
satisfied in time for AOD. On the other hand, AA should be responsible
for its failure to coordinate and oversee the cabling work and to ensure
that prompt remedial action was taken, eg, promptly instructing a
different or another contractor to help when the delay and the effect of
which was foreseen. It was not until late May to J une 1998 that AA
instructed HKT and two other contractors to assist in the cabling and
jumpering works. Alternatively, AA should have given priority to
having all the public telephones installed first before catering for the
479
tenants requirements, as the tenants could not reasonably blame AA for
the delay since they were late in submitting their requirements in the first
place.
[4] Escalators Breaking Down Repeatedly [see also paras 8.10 and
9.23-9.30]
16.15 Constructions Industrielles De La Mediterranee SA
(CNIM) is responsible for the first year maintenance of the installation
of escalators which work was overseen by AAs maintenance team, and
AA is responsible for the operation of the escalators. CNIM claimed
that the problem and the huge number of visitors could not be foreseen
before AOD; hence prevention was not possible.
16.16 Although the actual live load requirements of the escalators
could not have been precisely foreseen, the Commissioners feel that the
sensitivity level of the protective device of the escalators could have been
set properly had sufficient tests been carried out before AOD. For this,
Airport Management Division (AMD) of AA and CNIM should both be
responsible for not having sufficient tests or for failing to take
precautionary measures. On the other hand, the users pressing
emergency stops unnecessarily and foreign objects jamming the steps are
facts of life, and are unlikely to have been avoided. For that, no one
16.17 Had the systems for automatic control and monitoring of
smooth maintenance services that would enable staff of AAs Engineering
& Maintenance Department and Airport Operations Control Centre
(AOCC) staff to respond quickly to any breakdown of the escalators
been completed before AOD, the disruptions caused by the breakdowns
would have been reduced. However, these monitoring systems were
considered to be non-AOR critical, and were not completed before AOD
apparently because of a shortage of time. The unavailability of these
systems was patched up by AA sending duty staff of terminal operation
carrying keys to restart escalators after visual inspection to clear any
evident jams.
16.18 A more lenient view is that this is a teething problem, which
480
was cured very easily and quickly. After AOD, the safety devices were
promptly adjusted to the appropriate levels so as to match the actual
working conditions and passenger load. This measure had proved
successful. AA stated that on 8 J uly 1998 the escalators were working
more smoothly and that a special team was on stand-by to deal with
problems. After the adjustments made in the days following the AOD,
stoppage rate has dropped to normal level of around 0.2 per escalator
each month, which is considered to be a normal rate of operation.
[5] Insufficient or Ineffective Signage [see also paras 8.11 and
9.31-9.35]
16.19 The Commissioners are of the view that members of the
public visiting such a large building as PTB for the first time will
necessarily go through a period of familiarisation with the new
environment. The complaints about inadequate signage do not seem to
accord with the facts. In the light of evidence, conflict of allegations
between the Board of Airline Representatives and AA cannot be resolved.
In any case, airport operational readiness does not reasonably include
signs for airline offices. After considering the evidence very carefully,
the Commissioners cannot make a finding that the problem of inadequate
signage existed. Even if it did, it was but a teething problem which was
quickly remedied by the additional signs installed in J uly and August
1998. The Commissioners also accept the expert opinion of W51 Mr
J ason G YUEN that it is quite common among major airports to have
signage additions, revisions and refinement after the terminal has been
put to actual use . It is also difficult to find fault with the design
philosophy that is based on the logical flow of passengers within various
parts of PTB. Further installation of signs will generate too many signs
and they can be confusing and aesthetically offensive. Moreover, any
perceived problem regarding signage in PTB was mitigated by the fact
that extra AA staff wearing yellow sashes were on duty on AOD, trying to
guide passengers around the landside. Accordingly, the Commissioners
do not believe any organisation or person should be held responsible for
the signage issue.
[6] Slippery and Reflective Floor [see also paras 8.12 and 9:36-9.39]
481
16.20 In the light of evidence, the problem of slippery and
reflective floors is nothing but minor in nature having regard to its impact
on the operations of the new airport. There were only five reported
incidents of people slipping on the granite floors from AOD right up to
the end of August 1998 and the figure is considered insignificant when
compared with approximately six million users of PTB during that period.
In two of the reported cases, granite flooring might not be the real culprit
as water on the floor may be the major factor for its slipperiness.
16.21 The British-Chinese-J apanese J oint Venture (BCJ ) as the
contractor for the PTB superstructure and Grant Ameristone Limited as
the nominated sub-contractor responsible for the laying of the granite
flooring do not have any role to play in the problem. They were not
involved in the selection of the types of granite and the specification of
the surface finish for the PTB floors. All the flooring works were
completed to the architectural specifications laid down in the relevant
contracts. In fact, choice of granite as well as workmanship for the
laying of the PTB flooring are not crucial factors but, rather,
post-installation testing and the subsequent remedy to problems identified
may be the key to the alleged problems. Having reviewed the
background leading to the issue, the Commissioners are of the opinion
that AA has failed to take prompt and speedy remedial action to address
the problem prior to AOD. The problem of slippery black granite floors
at PTB was brought to the attention of AA as early as September 1997.
Despite the fact that complaints were received about the slipperiness and
reflectiveness of the granite floors following the airport trial in J anuary
1998, there did not appear to have been any noticeable progress thereafter
in seeking a solution to the problems. Possible courses of remedial
actions were still being considered in May and J une 1998 and at one stage
AA even contemplated the possibility of bringing in an outside contractor
to fix the problem. Although an effective remedy was eventually
identified and necessary works were carried out through the nights shortly
before AOD, the whole process was not completed until some time after
airport opening. Although the problem itself is minor, the
Commissioners do not accept that it is part of the usual teething
difficulties since it is something that was identified and anticipated at an
early stage and could have been eradicated before AOD. In this regard,
AA and, in particular, its Project Division which was responsible for
482
overseeing the design, installation and completion of the PTB flooring
should be responsible for their lack of inefficiency in taking remedial
actions. Should AA have tackled the matter more promptly, the problem
of slippery and reflective granite floors would not have been allowed to
develop into an issue in the opening of the new airport.
16.22 In his visits to the new airport, the Chairman of the
Commission noticed that the black granite flooring was quite reflective
and might cause embarrassment to persons wearing skirts. This matter
had not been substantially raised in the inquiry, and the Commissioners
would only request AA to employ measures to deal with it.
[7] Problems with Cleanliness and Refuse Collection [see also paras
8.13 and 9.40-9.49]
16.23 The problem of rubbish build-up was a short-lived one and
had been quickly remedied after AOD. Although the number of airport
users affected might be substantial, the problem is only minor in terms of
the degree of seriousness. Having reviewed the evidence, the
Commissioners consider that the principal cause of the problem lies with
the delay of PTB tenants fitting out works and the unscrupulous way of
disposing of fitting out refuse, for which both the PTB tenants and the
contractors employed by them for the works are undoubtedly culpable.
Should the tenants have completed their fitting out works earlier and
should they have been more self-disciplined in removing their refuse, the
problem could have been far less acute than it had been on airport
opening. On the other hand, AA as the management authority should
also be responsible for its failure to ensure timely completion of
fitting-out works by tenants and proper removal of the debris by them and
their contractors.
16.24 Regarding the refuse collection system at PTB, the
Commissioners have reasons to believe that there are inherent
deficiencies in the design. As pointed out by Pearl Delta WMI Limited
(Pearl), the refuse rooms were not adequate to handle the volume of
refuse in some areas. This point was refuted by the Mott Consortium
(Mott) which stressed that the design was consistent with the
appropriate standards. While the truth may lie somewhere in between,
483
the Commissioners consider that more weight should be attached to the
evidence of Pearl which is a reflection of the actual experience of the
ultimate user of the waste collection system. Also, the refuse chutes
between Level 3 and Level 5 are not continuous and this necessitates
manual transportation of refuse along the walkway on Level 4.
Obviously, this additional trip has an effect on the overall operational
efficiency of the system. The absence of a refuse room for restaurant
operators located at the area on top of the chutes seems to be more a
design problem than anything else. In this regard, blame should be laid
on both AA and Mott as parties responsible for the design. On the
matter of late issue of permits and passes, there is insufficient evidence to
enable the Commissioners to attribute responsibility.
16.25 From the sequence of events unveiled, it is quite clear that
both AA and its various contractors uttered their best endeavours to
overcome the problem before and after AOD. At the request of AA,
BCJ employed 300 additional workers on top of its usual force of 200 for
the three days prior to AOD and they worked round the clock in an
attempt to clear up the debris accumulated. It is also evident that both
Los Airport Cleaning Services Limited and Reliance Airport Cleaning
Services Limited hired a large number of extra staff for the weeks
surrounding AOD to clear away the refuse from areas like the apron and
the shopping mall. Unfortunately, because of the enormous amount of
waste, from the tenants as well as from the large number of sightseers,
these contractors were simply unable to cope with the work within a short
time. As testified by W42 Mr NG Ki Sing during his oral testimony, the
rate of waste build-up was so fast that it was actually more than anyone
could handle. The Commissioners are convinced that both AA and its
contractors did make great efforts to clear away the refuse although the
fact remains that the problem of rubbish build-up persisted for a couple of
days after AOD. In this regard, the Commissioners are not prepared to
put any blame on the contractors which took part in the removal of waste.
Despite the fact that there were things which were beyond the control of
AA, the Commissioners, consider that AA, being the party responsible for
the overall management of PTB, is responsible for its failure to ensure
that adequate cleaning service was provided for the premises when the
airport opened on 6 J uly 1998.
484
[8] Automated People Mover (APM) Stoppage [see also paras 8.14
and 9.50-9.59]
16.26 In its submission, Mitsubishi Heavy Industries, Ltd. (MHI)
pointed out that it achieved service availability of 98% in the Confidence
Trial, demonstrating completion of works in accordance with the
particular technical specifications (PTS) of Contract 350. The APM
system operated effectively in all incidents to ensure the safety of
passengers. Train stoppages triggered by matters already discussed
conformed with the relevant design specifications of the system. Given
the nature of APM as a mass transportation system, safe operation should
be accorded first priority. MHI suggested that the alleged problems
were in fact not problems but part of the safe and reliable operation of the
APM system.
16.27 The Commissioners note that all of the problems identified
with the APM system resulted in only slight disruption of train service
and, consequentially, some degree of passenger inconvenience. In terms
of the magnitude of their impact on the overall operation of the new
airport as well as passenger safety, these problems are considered to be
minor in nature. As far as the problem of occasional train stoppages is
concerned, the Commissioners are of the view that both the train
passengers and MHI should be held responsible. Passengers should be
blamed for their improper behaviour in forcing themselves through
closing vehicle doors or attempting to pry open doors. It is possible that
most of the problems could have been avoided if individual components
of the system such as sensitivity of the door control unit had been tested
and modified to suit actual operational needs prior to AOD. From the
evidence made available by MHI, the Commissioners have no reason to
dispute that the APM system had been properly tested since the
contractual target of 98% of service availability was achieved during the
Confidence Trial prior to AOD. However, it is evident from the
remedial actions taken so far that the causes of door-related failures and
train overshooting and undershooting were technical or mechanical in
nature and could be rectified by modifications to the door mechanism and
other replacement and adjustment work. Although such problems can
well be regarded as inevitable start-up difficulties which will disappear
after fine tuning, the Commissioners hold the view that if more thorough
485
testing and proper modification works had been done before
commissioning of the system, the frequency of their occurrences could
have been minimised.
16.28 The trapping of passengers on 20 J uly 1998 was fortunately
a minor and isolated incident. The actual causes of the incident are
considered to be more related in nature to behaviour than system.
Responsibility for the incident should primarily lie with the improper
behaviour of the five persons involved, if they failed to take proper heed
of the train announcement and boarded the train at the West Hall
departures platform against the announced advice. However, the
Commissioners cannot conclude that the announcements against boarding
were in fact made or audible to those persons. What is quite sure is that
the persons ignored the advice of the APM operator to wait for the
assistance of the APM maintenance staff. Instead, they resorted to the
emergency door release valve to open the vehicle door and this eventually
led to the deactivation of the train. They are, therefore, partly
responsible for causing their incarceration. On the other hand, the
Commissioners find that, prior to the incident, AA failed on its part to
provide a sufficient number of station attendants at the West Hall
departures platform to ensure that trains arriving there were properly
cleared and no one would attempt to get on board. Although AA
claimed that it had taken reasonable measures to the same effect, the
Commissioners do not find the argument convincing because, apart from
the repeated announcements to advise train passengers to alight at the
West Hall departures platform and the advice of the APM operator sent
through the intercom, no effective means of preventive control can be
found in the evidence put before the Commission. As a matter of fact,
had AA stationed sufficient attendants at the APM platforms after the
system had been put in use, most of the train stoppage problems relating
to door failures could have been avoided. As APM is a driver-less
system, the need to station a sufficient number of attendants at the
platforms for the purpose of keeping order is something that can and
should be envisaged. As such, AA should be criticised for failing its
duty to ascertain correctly the actual operational needs in terms of the
keeping of the order of passengers while boarding or alighting from trains
and to put in place sufficient attendants to attend to train problems at the
platforms. J udging from the chronology of events, MHI as the operation
486
contractor has adequately fulfilled its duty and has taken speedy remedial
actions to restore the train service. In this regard, no blame is attached
to MHI. The Commissioners also consider that the lack of effective
communication means between AOCC and the APM maintenance staff
while attending to emergencies remains an area which both MHI as the
APM maintenance contractor and AA as the management should jointly
look into in order to further enhance the speed of response in future
incidents. The rescue action in the incident on 20 J uly 1998 could have
been much quicker if the APM maintenance personnel had efficient use
of the TMR provided by AA.
[9] Airport Express (AE) Ticketing Machine Malfunctioning [see
also paras 8.15 and 9.62-9.64] and
[10] Airport Express Delays [see also paras 8.15, 9.60-9.61 and
9.65-9.68]

16.29 In light of the evidence received, the Commissioners are
satisfied that there were coin handling problems with the AE ticketing
machines during the initial period of operation. Mass Transit Railway
Corporation (MTRC) as the operator of the service should be held
accountable for the failure to provide problem free machines. It is quite
evident that the software problem associated with the machines could
have been detected earlier if, during the development and testing of the
software, the range of parameters of coins was set at a level close, if not
identical, to that in actual operation. Early detection of the problem
would have allowed the contractor to look for effective ways to remove it
well before AOD. MTRC has apparently failed to ensure that the
ultimate product is capable of meeting fully the needs of users when in
actual operation. Nonetheless, the problem is only a minor one having
regard to its negligible impact on the operation of the new airport.
MTRC is responsible for the train disruptions which the Commissioners
accept were startup problems. Such problems have not recurred since
early August 1998.
[11] Late Arrival of Tarmac Buses [see also paras 8.16 and 9.69-9.74]
16.30 A number of the problems were caused by the inefficiency of
FIDS which resulted in lack of accurate flight information, serious flight
487
delays, difficulties in coordination between boarding gate assignment and
the location of aircraft. There were problems with TMR and airbridges
which compounded to affect the tarmac bus service. The
responsibilities of the relevant parties for the problems relating to FIDS,
TMR and airbridges are set out in Chapter 13 and items [2] and [14] in
this chapter respectively.
16.31 For the insufficient number of security cards issued by AA,
this is part of the problems generally encountered by AA with the delay in
the issue of security cards to airport staff which is described in item [27]
on Access Control System in Chapter 16.
16.32 Under the relevant franchise agreement, Hong Kong Airport
Services Ltd. (HAS) had to provide 22 new buses, 19 for operation and
three in reserve. On AAs allegation that HAS had provided insufficient
bus drivers and passenger buses necessary for AOD and subsequent days,
HAS stated that on AOD, 38 drivers and 19 buses were made available,
more than the planned 23 drivers and 12 buses previously calculated by
HAS based on the aircraft parking stand allocation plan provided by AA
on 18 J une 1998 in which AA had anticipated 16 flights scheduled to
arrive at remote stands which needed to be serviced.
16.33 HAS had apparently admitted to being 30% understaffed on
AOD in a memorandum prepared for the Bussing Review meeting
scheduled for 16 J uly 1998. HAS disagreed with AAs interpretation of
the alleged admission which, as HAS contended, merely compared the
current required staffing levels with the original planned staffing levels
for HAS bussing service. HAS maintained that its resources were in
excess of its calculated needs and that the problem was mainly caused by
the lack of accurate flight information.
16.34 AA further contended that HAS had failed to make any
allowance for off-schedule flight operations.
16.35 According to the evidence presented by AA, there was only
one bus available in reserve, instead of three as agreed between AA and
HAS. HAS reply was that as it had 19 buses running on AOD, it meant
seven more buses available than the calculated figure of 12 based on the
488
aircraft parking stand allocation plan provided by AA. While it is
unnecessary to decide whether it was proper for HAS to rely on AAs plan
in estimating the necessary driver and bus force for operation on AOD,
and it is not within the Commissions terms of reference to determine
contractual liability, it is clear that had two more buses been made
available as reserve on AOD and Day Two, the added 10% of buses
would have helped to alleviate the situation. This alleviation was not
attained because HAS only provided 20 buses on those days. The
responsibility for this situation must be attributed to HAS.
16.36 The main cause for the inefficient and late tarmac bus
service was the deficiency of FIDS resulting in the lack of accurate and
prompt flight information that should have been available to HAS. The
problems with the airbridges and the TMR overloading compounded the
difficulties. HAS manpower was strained as they were required to
locate exactly where the aircraft were. The inability of using TMR to
relate messages meant that buses were required to return to PTB to
receive instructions and information of the location of aircraft. Delays
resulted in serving the aircraft in a timely way and extra trips were
necessary for the buses. The Commissioners are of the view that AA
should be responsible for all these factors which caused the late arrival of
tarmac buses.
[12] Aircraft Parking Confusion [see also paras 8.16 and 9.75-9.78]
16.37 Aircraft parking confusion is a problem consequential upon
the problems relating to FIDS and ACCs practice of confirming stand
allocation. The responsibilities of the relevant parties for these initiating
problems are described in Chapter 13.
[13] Insufficient Ramp Handling Services [see also paras 8.18 and 9.79]
[14] Airbridges Malfunctioning [see also paras 8.19 and 9.80-9.84]
16.38 The delay in providing mobile steps for passengers to
disembark from aircraft parked at remote stands was similar to that in the
provision of tarmac bus service discussed in item [11] above. However,
all three ramp handling operators, instead of HAS alone, were involved in
489
serving the passengers of the airlines they contracted respectively to serve.
The causative problems involved, such as the deficiency of FIDS, were
also the same as those relating to tarmac buses, and the conclusions of the
Commission on responsibility in paragraph 16.36 above are adopted here.
16.39 The Commissioners find that PT. Bukaka Teknik
Utama-RAMP J oint Venture is responsible for the programming error that
caused the auto-leveller alarms. Only B737 mockups were used in
operation tests since there was no requirement by AA to have tests
conducted with B747 mockups. The extent of testing was of course a
matter for AA and Bukaka Ramp to decide. The Commissioners are of
the view that the problems might well have been identified if more varied
or extensive testing or trials had been conducted. For this, AA and
Bukaka Ramp or one of them should be responsible.
16.40 The Commissioners do not believe that the problems
experienced with the airbridges were caused by operator error. First the
auto-leveller system contained a software error resulting in alarms.
Second, all operators had been certified by AA before they were allowed
to operate the airbridges single-handedly. Third, airbridge operation
was not a difficult process. While operator error may have triggered
airbridge faults on occasions, it would not have resulted in recurrent
problems. The Commissioners therefore do not think that the operators
should be blamed.
[15] No Tap Water in Toilet Rooms and Tenant Areas [see also paras
8.20 and 9.85-9.95]
[16] No Flushing Water in Toilets [see also paras 8.20 and 9.85-9.95]
(a) Problems relating to Tank Rooms 3 and 8
16.41 The AEH J oint Venture (AEH) and Rotary (International)
Limited (Rotary) were responsible for the installation of the valves that
failed before AOD. As a result, Tank Rooms 3 and 8 had to be operated
manually on AOD. Whatever the reason for the malfunctioning of the
valves, the interruption in the supply of potable water was caused by the
inability of Rotarys staff in obtaining security permits to gain access to
490
the tank rooms in order to maintain their operation.
16.42 According to Rotary, due to new security arrangements, their
staff were denied access to Tank Rooms 3 and 8, which are in a restricted
area. W25 Tsui of AA said that there was a change in access control
procedures a few days before AOD as well as on AOD. He also said
that there were some problems experienced by some contractors in
obtaining permits in time before AOD although he did not mention the
names of those contractors. However, AA denied that there was any
unexpected change in security arrangements. AA alleged that AEH
should have been aware of the need for permits to allow access to the
tank rooms on AOD and that they should have ensured Rotary, their
subcontractors, were likewise aware of these requirements. AA
maintained that there were delays on the part of business partners and
contractors in seeking proper permits into restricted areas.
16.43 Mr Shafqat Tariq, Project Manager - Hydraulics of Rotary
said in his witness statement that at around 10 am on AOD, he called Mr
Wallace TANG (electrical superintendent of AMD, AA) and explained to
him the problem with filling the water tanks and asked him to put AA
staff in the tank rooms to man them. Mr Tang however had a different
story. Whilst he remembered the conversation with Mr Tariq, he did not
remember Mr Tariq asking him to arrange for operation of the tank rooms.
He recalled Mr Tariq telling him that there had been a problem with
filling the tanks, which Mr Tariq had rectified by making manual
adjustments to the valves. Since that problem had been fixed, Mr Tang
thought there was no need to take any remedial action.
16.44 As the Commission has had no opportunity to hear the oral
testimony of these two gentlemen and other relevant witnesses on these
issues, it is not possible for the Commissioners to come to a proper
conclusion as to who or which of the versions should be believed.
Whilst there is contradictory evidence regarding the difficulty of Rotarys
staff in gaining access to the tank rooms and whether Rotary had asked
AMD to operate the tank rooms on AOD, it appears to the
Commissioners that there was lack of coordination between AA, AEH
and Rotary to ensure that Rotary would be allowed access to the tank
rooms on AOD, for which AMD of AA should reasonably be responsible
491
as manager of the new airport.
(b) Problems relating to Tank Room 2
16.45 Nishimatsu Construction Company Limited (Nishimatsu),
the contractor providing drainage service, acknowledged that the flooding
was caused by blockage in the length of the pipework for which they
were responsible. It was found on 18 J uly 1998 that a section of the
pipe was broken. Nishimatsu acknowledged that it was responsible for
the clearance of the blockage and the repair of the damaged pipe.
However, Nishimatsu maintained that it was only notified of the flooding
problem on 5 J uly and therefore it was not able to rectify the problem
until 13 August 1998. AA admitted that they had only instructed
Nishimatsu to investigate the matter after the flooding in Tank Room 2 on
7 J uly.
16.46 Flooding did not occur for the first time after AOD: there
had been flooding of tank rooms since late May 1998. During the
flooding on 30 May, temporary pumps were installed to pump dry the
tank room. AA admitted that they foresaw the flooding and had taken
preventive measures by instructing BCJ , the main contractor responsible
for construction of pipes underneath the tank rooms, to remove any
blockages in the pipe to Tank Room 2. BCJ carried out these
instructions. Nishimatsu was also asked to employ a high-pressure
water jet company to clean the pipe work in this area. AA, however, did
not ask Nishimatsu to deal with these early flooding problems as it
thought that it was the responsibility of BCJ who was primarily
responsible for the pipe work directly underneath the tank room. In
relation to the flooding which occurred on 7 J uly, AA alleged that BCJ ,
having discovered that the blockage was not in their part of the pipework,
should have known that it must have been in that part for which
Nishimatsu was responsible. Accordingly BCJ should have notified
Nishimatsu. However, BCJ alleged that it did not have responsibility
nor authority to give notice to AAs other contractors with respect to
defects in others works.
16.47 Despite the steps taken by AA, further flooding continued
to occur on 23 J une, 29 J une and 5 J uly. Notwithstanding the few
492
reported incidents on flooding and before the cause of the flooding had
been identified, AA did not see fit to instruct Rotary to install a pump to
prevent further flooding or take other preventative measures.
16.48 Having considered the evidence, the Commissioners come to
the conclusion that the flooding in Tank Room 2 was foreseeable and
preventative measures, for example, the installation of sump pumps to
control the flooding, should have been taken by AA and AEH prior to
AOD. Indeed, AA admitted that no preventive measure was put in place
before AOD. The Commissioners are also of the view that there was
also a lack of coordination amongst AA, BCJ , Nishimatsu, AEH and
Rotary, for which AA as manager of the new airport should be primarily
responsible.
[17] Urinal Flushing Problems [see also paras 8.20 and 9.96-9.110]
(a) Flow of flushing water
16.49 One of the experts appointed by the Commission, W54
Professor Xiren CAO, is of the opinion that the design of the flushing
system was not appropriate, given the poor condition of the seawater.
On the evidence, the Commissioners find that AEH should be responsible
for the following matters:
(a) failing to provide a satisfactory flushing system for the
urinals, and in particular, for using inappropriate flushing
valves which were subsequently replaced at its own
expense; and
(b) failing to install weirs to stop sand and dirt getting into the
water pipes which would have been a remedial measure for
alleviating the accumulation of sediment in the valves.
16.50 AEHs failure to complete the testing and commissioning of
the hydraulic system could well be another cause of the problem, but
there is insufficient evidence before the Commission to make a proper &
fair finding on this issue.
493
16.51 The Commissioners accept W54 Caos expert view that there
was some design problem, for which AA should be responsible. AA
should also bear some responsibility in not taking prompt and sufficient
remedial actions to prevent or alleviate the flushing problem since the
problem had been identified in early 1998.
16.52 On the alleged two outstanding problems, namely, corrosion
of the solenoid valves and the pressure setting, AA alleged that AEH had
failed to resolve these problems and AA was carrying out its own
remedial work on the corroded valves. There is conflicting evidence as
to whether there remains a problem with the setting of the pressure of the
valves and the Commissioners are not able to come to any view on this
issue.
(b) Problem with sensors
16.53 There is conflicting evidence as to who is responsible for the
correct setting of the sensors. AA alleged that AEH had not taken into
account the fact that the sensor would be installed on a part of the wall
which was about 6 inches further back than the lower portion of the same
wall on which the urinals were mounted, making the gap between the
user and the sensor greater than 18 to 24 inches as preset by the
manufacturer. However, Rotary, AEHs subcontractor, stated that the
sensor was not preset and that it could be adjusted to the optimum range
between 18 and 36 inches within 20 minutes of power being switched on
the system. The Commissioners tend to prefer the evidence provided by
Rotary as it would be difficult to understand why the sensors could have
been designed with the rigid operational range as alleged by AA.
However, as there is conflicting evidence as to which party is responsible
for the correct setting of the sensors, no firm conclusion can be reached.
W51 Yuen, another expert appointed by the Commission, opined that
public misuse is a normal occurrence in a busy airport. The
Commissioners agree, and given that there were so many people visiting
the airport in the early days of airport opening, the Commissioners are of
the view that the blame for the damage to sensors should not be attached
to AA, AEH, Rotary or Los.
(c) Blockages of urinals
494
16.54 Blockages were mainly caused by public misuse and the
problem was exacerbated by the huge number of visitors to the airport
when it first opened. Los, the cleaning contractor, was responsible for
clearing rubbish in the urinals and therefore preventing blockages in
urinals. As described below, this task was made more difficult by other
problems such as the disruption to flushing and potable water to toilets.
(d) Cleanliness of toilets
16.55 Los was responsible for keeping the toilets clean. Whilst
there might have been staff training and supervision issues, the cleaning
task was hampered by other problems such as the flushing problems, lack
of potable and flushing water and blockage of urinals. The problem was
exacerbated by the sheer number of curiosity visitors and stranded
passengers during the first few days of operation of the new airport. On
the whole, the Commissioners are of the view that the toilets were not
sufficiently clean due to the shortage of manpower. This could have
been caused by the requirements of labour imposed by AA being too low,
or Los failure to deploy sufficient people to perform the task. However,
on the evidence received, the Commission is not able to decide.
16.56 The shortage of manpower was, according to the
Commissioners view, partly caused by the question of permits for
restricted areas of PTB. In reply to Los allegation that it had problems
in obtaining permits, Aviation Security Company Limited (AVSECO)
said that Los had been issued with 309 permits out of 660 applications
from Los. The remaining permits were not issued because, according
to AVSECO, the staff of Los had failed to turn up for photo-taking or for
collection of permits. On this matter, again the Commission considers
the evidence received not sufficient for it to make any proper or fair
finding. This matter, however, relates to coordination and operation
between contractors working within PTB, and AA as their employer and
manager of the new airport should primarily be responsible.
[18] Toilets Too Small [see also paras 8.21 and 9.111-9.122]
16.57 The Commissioners accept that a larger number of smaller
toilet blocks rather than fewer but larger blocks is a proper and reasonable
495
design concept for the new airport because of the sheer size of PTB.
Having regard to the fact that the actual toilet provisions have already
exceeded the prescribed standards laid down in the British Airport
Authority guidelines, the view of counsel for the Commission was that
complaints about the lack of toilets was not fully borne out by the
evidence. The Commissioners accept this view. The scenario at the
Kai Tak airport, which had very big toilet blocks, may have had a
psychological impact on airport users, creating and reinforcing their
impression that toilets in the new airport might be too few and too small.
16.58 Concerning the issue of trolley accessibility to toilets, the
Commissioners accept the views of Mr Barry Ball, Senior Architect
Interiors of AA, that the decision of not allowing trolleys into toilet
blocks is a correct one. The reason for allowing trolleys in would be
primarily to allow passengers to keep sight of their baggage on trolleys
while using the toilet facilities. However, since it was never a realistic
option to allow baggage trolleys inside the cubicles, trolleys would still
have to be left unattended for certain periods. The Commissioners also
accept the opinion of W51 Yuen in his supplemental expert report of 1
December 1998 that AAs policy of not allowing baggage trolleys into
toilet rooms is common amongst many airports.
16.59 W51 Yuen, however, pointed out that the passageway to the
toilet room might be too narrow and did not allow two people carrying
hand baggage to pass each other easily. He went on to add that toilet
rooms in many airports had two passageways, one for entrance and the
other for exit so as to avoid the problem. While accepting that toilet
provisions in the new airport have been designed in accordance with
recognised industry standards and are in line with those in most hotels
and public buildings, the Commissioners are inclined to conclude that
toilets and their passageways could perhaps have been widened slightly
for the convenience of airport users. Even W3 Dr Henry Duane
Townsend agreed during his testimony that Mott could have adopted
more generous standards in planning toilet provisions. As a matter of
principle, consideration of commercial rental revenue should never take
priority over public convenience in the design of PTB. There is,
however, no hard evidence before the Commission to suggest that AA has
inappropriately trimmed down toilet facilities in the new airport in order
496
to maximise the commercial rental space in PTB. Nevertheless, it
remains a fact that public expectations have not been fully met in this
respect and more generous allowances for space in toilets could have
been provided.
[19] Insufficient Water, Electricity and Staff at Restaurants [see also
paras 8.22 and 9.123-9.136]
(a) Water and electricity supply
16.60 The Commissioners are of the view that both AA and the
relevant tenants probably contributed to the water and electricity
problems. The tenants and AA had a part to play in the electricity
problems related to the system upgrade of the electricity system.
However, due to limited time available to this Inquiry and without
investigating further into this matter, it is not possible for the Commission
to apportion responsibility.
16.61 The length of the electricity outage on 7 J uly 1998 should
properly be attributable to the AA or AVSECO because AAs maintenance
personnel and contractors staff were denied access by a security guard to
effect remedial work. As the outage was caused by improper loading
settings of the installation of the tenant concerned, the tenant is probably
responsible for causing the problem although it is not possible for the
Commissioners to assign responsibility.
16.62 It is not clear who is responsible for the electricity outage on
17 J uly 98. Although AA suspected that the negligence of a contractor
of Cathay Pacific might have caused the outage, there is no substantial
evidence before the Commission. According to Cathay Pacifics
contractor, they had no record of the alleged incident.
16.63 In relation to the disruption to the water supply on AOD and
the few days thereafter, please refer to the discussions under items [15]
and [16] above.
497
(b) Staffing problems
16.64 AVSECO alleged that the delays in the processing of permits
were for reasons out of their control, including the massive last-minute
rush for permit and the frustrating regular breakdown of ACS and the
permit computer system. It also stated that it had taken a number of
contingency measures, including the flexibility given to and prompt
processing of escorted and three-day temporary permits and additional
temporary staff deployed by the Permit Office to ensure 24-hour and
seven-day a week service to the permit applicants. Without hearing oral
testimony, it is impossible for the Commissioners to reach a fair
conclusion on these issues.
16.65 The sheer number of sightseers also exacerbated the problem,
especially when the staff were not familiar with the new environment.
The tenants should also be responsible for ensuring that a reasonable
level of service was provided to the public and that the staffing should be
sufficient and well trained.
[20] Rats Found in the New Airport [see also paras 8.23 and
9.137-9.139]
16.66 Reports in newspapers and on television on the problem of
rat infestation were probably exaggerated through the media process.
The Commissioners are satisfied that it is but a minor problem and is
under control. AA has implemented a range of measures to contain the
problem. Although it is not certain whether rats would be eradicated in
the new airport, provided AA can keep up with its rodent control
programmes, the situation will no doubt continue to be under control.
[21] Emergency Services Failing to Attend to a Worker nearly Falling
into a Manhole while Working in PTB on 12 August 1998 [see also
para 9.140]
16.67 After reviewing the evidence on the case, the Commissioners
consider that upon the first call for help, an ambulance as well as a fire
engine with trap rescue equipment should have been despatched. The
Commissioners find that either the caller who made the first call to
498
request for ambulance service was not accurate in providing necessary
information or the receiver of the call had not asked the appropriate
questions, resulting in only an ambulance being sent. Had it been made
known at the time of the first call that special service operation crew was
required to save the injured worker, the actual rescue would not have
been delayed by 21 minutes. On the basis of the information available,
the Commissioners are unable to ascertain who is responsible for this
delay. This is however only a minor incident.
Resulting in Five Firemen being Injured [see also para 9.141]
16.68 The Police investigated into the accident immediately upon
its occurrence but did not find sufficient evidence for further action to be
taken. Later, the Traffic Accident Inquiry Board of Fire Services
Department also made investigation within the Department and found
that the accident could be attributed to the drivers misjudgement on the
prevailing traffic situation, road configuration and the weather condition.
The driver was suspended from driving duties until he successfully
passed a driving re-examination. He was also held responsible for
paying the repair cost of the damaged vehicle. Since the accident has
been thoroughly investigated by both FSD and the Police, the
Commissioners do not have anything further to add save to agree with
FSDs findings.
Cabin of a Cathay Pacific Aircraft on 3 September 1998 [see also
para 9.142]
16.69 This is an accident and no one should be held responsible for
it.
in Lifts and on the APM as well as Delaying Two Flights [see also
para 9.143]
16.70 The Commissioners regard this as a minor incident. Since
investigation into the incident is ongoing and not all materials are
499
available, the Commissioners cannot take any view on the question of
responsibility.
October 1998 [see also paras 9.144]
16.71 According to AA, missed approaches are not infrequent
occurrences in an airport. Based on the material supplied by AA, it
appears to the Commission that the incident was handled safely,
efficiently and in accordance with laid down procedures. No
responsibility should be attached to anyone.

500
CHAPTER 17

RESPONSIBILITY OF THE AIRPORT AUTHORITY

Section 1 : AAs Obligations under the Airport Authority
Ordinance

Section 2 : Co-ordination and Communication

Section 3 : Overview of What Went Wrong

Section 4 : Misstatements and Responsibility for Them
(a) FIDS
(b) ACS

(a) W3 Townsend
(b) W48 Lam
(c) W43 Oakervee
(d) W44 Heed
(e) W45 Chatterjee
(f) The AA Board

Section 1 : AAs Obligations under the Airport Authority Ordinance

17.1 According to the Airport Authority Ordinance, Airport
Authority (AA) has the following functions, duties and objectives:

(a) to provide, operate, develop and maintain the new airport
[the preamble to the Ordinance] and to provide such facilities,
amenities or services as are, in its opinion, requisite or
expedient [s 5(1)(b)];

501
(b) to maintain Hong Kongs status as a centre of
international and regional aviation [s 5(1)(a)];

(c) to conduct its business according to prudent commercial
principles [s 6(1)]; and

(d) in conducting its business and performing its functions, it
shall have regard to safety, security, economy and operational
efficiency and the safe and efficient movement of aircraft, air
passengers and air cargo [s 6(2)].

17.2 For the purpose of the inquiry, one should note that AAs
business includes the operation of the new airport and that in conducting
such operation it shall have regard to the safe and efficient movement of
air passengers and air cargo. That being part of AAs statutory duties
and functions is therefore indisputable. The problems encountered on
airport opening day (AOD) reveal that AA did not have sufficient
regard in these respects when opening the airport for operation on 6 J uly
1998. While little blame should be attached to AA for the teething
problems which are inevitably facts of life, AA must be responsible for
the other and major problems that created the chaos on AOD because no
or insufficiently efficient movement of air passengers and air cargo was
provided. AA employed contractors in the discharge of its duties, and
when the work performed by the contractors or their subcontractors did
not come up to standard giving rise to a problem, the contractors and
subcontractors, if they can be ascertained, must be primarily responsible.
The issues involving the relationship and responsibility between AA on
the one part and its contractors and business partners on the other are
reviewed in other chapters. It is also necessary to examine matters
relating to the internal organisation, working and action of AA and its key
officers that caused or contributed in causing the problems, and this
chapter deals with the problems identified through the evidence that relate
to co-ordination and communication within AA and between AA and
Government.

502
Section 2 : Co-ordination and Communication

17.3 Issues of co-ordination and communication were identified
in the documentary evidence. The deficiency of co-ordination in AA
was noted by the New Airport Projects Co-ordination Office (NAPCO)
in the Airport Development Steering Committee (ADSCOM) Paper
34/97 dated 19/9/97 prepared by it. A number of passages dealt with the
organisation and co-ordination aspects of AA. NAPCOs comments on
the organisation structure of AA and questions of co-ordination and
communication were as follows:

The matrix organisational split of AOR (airport operational
readiness) responsibilities between the various AA Divisions is
not functioning efficiently, and information and decision-making
bottlenecks exist. We also find that programme and other
essential information is not fully shared between the AA and
Business Partners. We recommend herein that a single-point
responsible high-powered executive manager be vested with
requisite and clear authority to direct the AOR process and dictate
action inclusive of all participants AA (including all Divisions),
Business Partners and Government. We further recommend that
a full open book approach to co-ordination and information
sharing be implemented immediately inclusive of all participants.

17.4 NAPCO also found that co-ordination within the AA itself,
particularly amongst the following divisions, the Airport Management
Division (AMD), the Project Division (PD) and the Commercial
Division, as well as co-ordination and cooperation between the AA, its
business partners, Government and all others required intensified
attention and immediate improvement. The co-ordination and
cooperation between AMD and PD was particularly important from about
the last quarter of 1997, as the new airport was transitioning from the
construction stage to the operation stage, from the care and responsibility
of PD to those of AMD. While PD was concentrating and prioritising
on the works side, AMD was eventually to be handed over the works and
systems developed under the auspices of PD, and AMD had to use the
services and facilities so provided to develop and implement
familiarisation and training programmes and trials, to review

503
AOR-related issues, and to operate and run the new airport.

17.5 The point was stressed in the 170
th
ADSCOM meeting on 20
September 1997, when the then Director of NAPCO, W48 Mr Billy LAM
Chung Lun, was recorded as saying: AMD should be in the driving seat
of the airport project at this point in time, but because of the personalities
involved, it was being pushed round parameters set by PD and had yet to
gear itself up. The Deputy Director of NAPCO advised that W3 Dr
Henry Duane Townsend should, but did not, quickly and firmly resolve
this problem.

17.6 At the 171
st
ADSCOM meeting on 13 October 1997, W36
Mrs Anson CHAN, the Chief Secretary for Administration and the
Chairman of ADSCOM (the Chief Secretary), enquired with AA
officers present about the relationship between PD and AMD. The
Divisional Manager (Planning & Scheduling) of AA responded that their
relationship was getting better every day. W3 Townsend added that
AMD had more people working on site then. The Chief Secretary stated
that she wanted the two divisions to work in concert towards the target
and requested W3 Townsend to keep a close watch on the situation and to
sort out any difficulties between them promptly. The Director of
NAPCO, on the other hand, reported that the relationship between PD
and AMD had much improved.

17.7 The notes of the ADSCOM special meeting on 7 November
1997 also recorded Director of Civil Aviation (DCA) as saying that he
had no faith in the top management of AA. He said that the project was
driven by the Project Director, W43 Mr Douglas Edwin Oakervee who
always tried to bulldoze his way through. W3 Townsend was not in
control and the organisation was not functioning as it should. DCA gave
an example of the problems on the software side. Build 1.5 failed to
arrive on 3 November as scheduled, and the AA management could not
give him a date on which it would arrive. DCA was worried about
systems integration within Flight Information Display System (FIDS)
and about its integration with other airport systems. There had to be
definitive contingency measures in case of failure, but so far AA had only
developed crude contingency plans. In his view, for the airport to
operate, the Airport Operational Database (AODB) system had also to

504
be interfaced with the Aeronautical Information Database. The latest
schedule of 15 February 1998 was too close to the opening date for
comfort. There would not be enough time to put in contingency
measures and for sufficient re-training. DCA was also unsure of the
software implications of the standalone systems, and he failed to obtain
any reassurances from W3 Townsend. At the same meeting, the
Director of NAPCO revealed that the joint study carried out by NAPCO
and AA on AAs airport operational readiness (AOR) programme had in
effect forced AMD and PD to start talking to each other, which was
something they should have done months ago.

17.8 There was difficulty in Government obtaining information
from AA. NAPCO recorded the lack of cooperation from AA in its
Weekly Site Report of 7/3/98. NAPCOs attempts to find out what was
going on regarding systems integration during the period were
continually thwarted because AA staff were warned not to say anything.
No wonder that NAPCO started to distrust AA. NAPCO reported that
AA would claim that, all the scheduled tests were completed, however,
the reality was that the system could not yet display flight information at
a number of locations.

17.9 In its Weekly Situation Report of 1/5/98, NAPCO reported
that it had still not received the AAs quantification of additional
requirements for the contingency plan in case of FIDS failure, as
promised.
.
17.10 Another week passed by, NAPCO again reported that AA
claimed to have corrected many of the FIDS critical software issues and
resolved the Societe Internationale de Telecommunications Aeronautiques
(SITA) (Common User Terminal Equipment (CUTE)/FIDS interface
problems with implementation at site continuing). However, a number
of software issues, which AMD stated as critical, were still outstanding
and this raised concerns as to AAs ability to establish Day One operating
scenario. AA was developing the contingency FIDS with GEC (Hong
Kong) Ltd (GEC) and Hong Kong Telecom CSL Limited (HKT) but
the time available for development was short. Work to interface FIDS
with other systems such as AODB, Baggage Handling System (BHS)
etc continued and updates to AODB software was due in mid May.

505
NAPCO had been chasing AA but had still not received its quantification
of additional data transfer requirements [Weekly Situation Report,
8/5/98].

17.11 In the ADSCOM Chairmans brief prepared by NAPCO for
the 183
rd
meeting of ADSCOM on 22 May 1998 and in the minutes of
that meeting, NAPCO pointed out that by opening, the airport systems
would largely operate in standalone mode. It was clear from the AA
report that lots of integration were still underway and programmed for
completion by the end of May 1998. ADSCOM had been assured that
systems existed for manual data transfer. However, as most systems had
to be operated on a standalone basis, more staff, procedures, etc, had to
be organised. The quantification of what this involved in terms of
equipment, staff, changed procedures, training, etc which NAPCO had
been after for months had yet to be forthcoming from AA. In the
Summary of Critical CLK Issues, dated 19/6/98, NAPCO again stated
that the demonstration of the viability of workarounds, schedule and
procedures of installing enhancements, system status etc were all
expected in a detailed report which was still not yet received. NAPCO
had yet to receive from AA the quantification of additional data transfer
requirements under the contingency scenario.

17.12 As late as May 1998, the co-ordination between AMD and
PD still caused concern. In his Weekly Site Report for the week ending
23/5/98, NAPCOs Mr David Thompson, Senior Coordinator for Special
Systems, reported that in order to accommodate the new back up system
to FIDS, AMD needed to have some more workstations, without which
there would be problems for system development and training functions.
In answer to a NAPCO question concerning the reason why five
additional workstations had not simply been purchased, it appeared that
PD was not willing to spend money and AMD did not have access to
funds.

17.13 There was also a co-ordination problem regarding the testing
of Government entrusted works. In a memo dated 28/5/98 from W33
Mr KWOK Ka Keung, Director of NAPCO, it was noted that the
continuing delays in testing and commissioning of Civil Aviation
Department (CAD) systems were the result of ongoing AA installation,

506
testing and commissioning problems with the primary AA systems.
Thus, until the primary AA systems were fully functional and operational,
CAD systems which were dependent upon the AA master system could
not be adequately tested or commissioned.

17.14 Eventually, however, the co-ordination between NAPCO and
AA improved. W31 Mr J ames WONG Hung Kin, Project Manager in
NAPCO, testified on this matter before the Commission. The very
detailed internal project reports prepared by AA were originally only
supposed to be available to the AA Board members. That practice was
changed in mid-1996. After that, AA was much more open to
Government and shared with NAPCO its internal reports. From those
working level reports, NAPCO staff on the site knew a lot more about the
true picture in addition to having day-to-day contact with AAs working
level staff. The relationship gradually improved a lot, particularly
towards the end of the project. In the half year before AOD, AA was
quite open towards NAPCO by allowing NAPCO staff to take part in the
site acceptance tests (SATs) and to visit Interface House which was
previously quite closed to outsiders, including NAPCO. W31 Wong
said that towards the end of the project, NAPCO generally had quite a
good feel about the progress of a wide spectrum of the AA works, and
focussed their attention on FIDS by reason of its apparent difficulties and
also to interfaces with Government entrusted works, because Government
departments had to have available a lot of facilities at the new airport.

17.15 W43 Oakervee (Director of PD), W44 Mr Chern Heed
(Director of AMD), W45 Mr Kironmoy Chatterjee (Head of Information
Technology (IT)) and W46 Mrs Elizabeth Margaret Bosher (Director of
Planning and Co-ordination), all of AA, gave evidence before the
Commission as a group. In the course of their evidence, they all denied
that there was insufficient communication or co-ordination amongst
themselves or their Divisions and Departments. All along, the
organisation structure of AA was such that W43 Oakervee and W44 Heed,
the Directors of the two most important Divisions, PD and AMD,
reported to W3 Townsend, the Chief Executive Officer (CEO), who
would co-ordinate and decide on the important issues, especially those
regarding the transition from projects to airport operation. As pointed
out in the ADSCOM meeting on 7 November 1997, W43 Oakervee

507
bulldozed his way through, W3 Townsend was not in control, and the
organisation was not functioning in the way it should. As a result, AMD
was afforded lower priority to PD although AMD was going to operate
the airport eventually and the works and systems would have to be
suitable and fit for operation according to the operators requirements.
NAPCO advised in ADSCOM Paper 34/97 dated 19/9/97 that a
single-point responsible high-powered executive manager be vested with
requisite and clear authority to direct the AOR process and dictate action
inclusive of all participants AA (including all Divisions), Business
Partners and Government.

17.16 Several issues were involved, namely, the organisational
structure of AA, the co-ordination between PD and AMD, the lesser
importance placed on operational requirements than on the works
programme, and personalities of the senior management. These issues
were quite intertwined. W44 Heed admitted that it was partly true that
he allowed himself to be pushed around and W3 Townsend was not
backing him up. When called back to give evidence again, W3
Townsend explained the preponderance he placed on the works carried
out by PD by saying that PD was the major part of the organisation up to
AOD representing about three-fourths of the total organisation whereas
AMD did not really start to grow and expand until the latter part of 1997.
When W48 Lam was appointed as the Deputy CEO of AA in J anuary
1998 to be in charge of the AOR programme, W44 Heed reported to him
while W43 Oakervee continued to report to W3 Townsend. W48 Lam
told the Commission that when he became the Deputy CEO, W3
Townsend intimated that W43 Oakervee would continue to report to him
(W3 Townsend). While admitting that W43 Oakervee continued to
report to him instead of to W48 Lam after the latters appointment, W3
Townsend pointed out that W48 Lam reported to him and when W48 Lam
needed help or direction, W3 Townsend was there helping W48 Lam.
He further added that that W43 Oakervee worked closely with the AOR
programme by having Mr Alistair Ian Thompson, W43 Oakervees
number two man attending the AOR meetings chaired by W48 Lam. By
all these, W3 Townsend implied that there was no lack of co-ordination.

17.17 W48 Lam used to be the Director of NAPCO, occupying that
position from 22 March 1993 until 5 J anuary 1998 when he was seconded

508
to AA as Deputy CEO. He was one of the contributors who commented
on AAs organisational structure and the personalities of its top
management in the NAPCO papers and ADSCOM meetings between
September and November 1997. At a site inspection on 12 March 1998,
W48 Lam received a serious injury to his leg and as a result he was
hospitalised for a week and had his leg in plaster cast for six weeks
thereafter. He was given sick leave which lasted till 15 J une 1998.
However, he resumed his duties at AA before the end of his sick leave but
needed to go for physiotherapy almost daily, and sometimes he had to
excuse himself from meetings. His duties in AA were to be in charge of
two main matters, the AOR programme and the planning and preparation
for relocation from Kai Tak to Chek Lap Kok (CLK). He spent about
half of his working time in AA for the relocation programme. After
W48 Lam joined AA as its Deputy CEO, he established weekly meetings
on AOR issues (AOR Meetings) so as to pull the staff from various
Divisions together in order to improve co-ordination amongst them and
ensure focus on critical AOR issues. Apart from the AOR Meetings,
W48 Lam also gave examples of how he helped co-ordination. On
several occasions, W48 Lam asked W44 Heed and his AMD staff to put
their operational requirements and outstanding problems in writing to
W43 Oakervee and PD senior staff, so that the situations could be
remedied. However, apparently operational management demands were
still deferred to project requirements. W44 Heed told the Commission
that he knew that when the systems were handed over to his AMD, they
had not been fully tested and commissioned. He knew that for whatever
reason projects were delayed he had to take the consequence. He felt
that he had no alternative and had a lot of frustration. W3 Townsend
mentioned that the construction activities were moving very fast and that
testing and commissioning required very strong control, which according
to him, was certainly provided by W43 Oakervee. He and W43
Oakervee got along very well together in terms of reporting relationships.
W43 Oakervees comment on the organisation structure of AA was that
AA adopted a matrix management structure which was fine for
communication expected between one division and another, but he
preferred a hierarchical management. It is difficult to understand why
W43 Oakervee favoured hierarchical management, which supposedly
means that the higher rung on the hierarchy should control and give
orders to those on the lower rungs. That would have been accomplished

509
if W3 Townsend had been in control, and if the Commissioners
understand W43 Oakervee correctly, he must have meant that one needs a
masterful personality to be the CEO. That would be another way of
agreeing with NAPCOs advice that W3 Townsend was not in control.
When questioned about the implication of Oakervees evidence of
hierarchical management that a masterful person should be at the helm,
W3 Townsend agreed that it might have been the way as W43 Oakervee
envisaged, but added that the AA management relied very heavily on
delegation whereby instructions given were followed, people were held
accountable for their activities, and support was given to them.

17.18 J udging from the evidence of W3 Townsend, W43 Oakervee
and W44 Heed, and having carefully observed the demeanours of these
three witnesses, it appears to the Commissioners that the difficulties
encountered by W44 Heed and his AMD were not a consequence of the
organisational structure. It rather boils down to a matter of personalities
and the interaction of personalities amongst the top echelon of the AA
management. While W3 Townsend was correct in saying that he had
W43 Oakervee to exercise strong control over the construction works,
their progress and the testing and commissioning of the systems, the
assertive and imposing character of W43 Oakervee greatly influenced W3
Townsend, relatively milder in personality, in placing the too much
significance and priority on PD and giving less support to AMD in its
planning and preparation for the operation and management of the airport
in the making than AMD rightfully deserved. W44 Heed, a soft-spoken
and less resolute personality, took whatever was on offer, well knowing
that he would be facing great difficulties when operating the new airport
after the systems were handed over to AMD from PD with the degree of
testing and commissioning leaving much to be desired.

17.19 When the question was put whether he considered that there
was a leadership problem with W3 Townsend, W48 Lam said that it was
not so much a question of leadership but rather a question of personality
as well as a question of emphasis. W48 Lam believed that W3
Townsends emphasis with his engineering background was on the
engineering side, on the works side, and to complete the project on time
and within budget, and airport operation not being his forte or specialty,
W3 Townsend might have tended to overlook that aspect. W48 Lam

510
agreed that there was an element of interaction of personalities at the top
management of AA. W48 Lams opinions agree generally with the
Commissioners views.

17.20 In fact, as early as 29 May 1997, the AA Board approved the
establishment of a working group to review AAs organisation and
management structure post airport opening. The working group was
tasked to choose a suitable consultant to carry out the proposed study.
The working group was headed by W49 Mr LO Chung Hing (the
Vice-Chairman of the AA Board) and had W3 Townsend, Director of
NAPCO and three other members of the Board as members. The
working group selected Booz-Allen & Hamilton as the consultants to
undertake the study. The consultants made a report dated 20 October
1997 and their recommendations were eventually approved by the Board.
When W49 Lo and W50 Mr WONG Po Yan (Chairman of the AA Board)
gave evidence together, they were asked about the contents of the report,
which revealed deficiencies in the leadership and teamwork of the senior
management and competence of some senior managers. While W49 Lo
said he did not know who the senior managers referred to in the report
were, W50 Wong told the Commission that he realised that the AA
management had the problems as identified. Both witnesses said that
apart from introducing measures as recommended in the report, to
strengthen leadership and improve co-ordination, such as asking
Government to second the Director of NAPCO, W48 Lam, to be the
Deputy CEO, they could not possibly afford to change any member of the
top management. At that stage, which was barely about six months
before the Boards target date of April 1998 for airport opening, making a
change of the senior management would be too risky and the personnel
problem could not be resolved in the midst of the transition from building
stage to the operation stage. The Commission accepts this as a
reasonable explanation and does not attach any blame to the Board.

17.21 The evidence given by the senior AA management also
identifies a problem regarding the proper allocation of resources. W43
Oakervee stated that although his PD was able to get funds available for
the works, it was more difficult for AMD to get resources at the early
stage. W44 Heed told the Commission that AMD was a very thin
organisation. His explanation was that AMD was going to be an

511
ongoing organisation whereas PD was there for the project, and so there
was greater scrutiny on him to keep his staffing levels and costs as low as
possible. A lot of what was expected of him was not achievable in a
sense because there was not enough management staff to go around to do
all the things required. An example of the consequence given by him
was that AMD was not able to effectively oversee the preparations that
ramp handling operators (RHOs) were making for AOD. However,
W50 Wong and W49 Lo told the Commission that there had not been any
indication to the Board about the resources problem experienced by
AMD.

17.22 While the senior officers of AA denied that there was any
co-ordination and communication problem within AA, specific instances
of lack of co-ordination can be identified from the testimony of W44
Heed. W44 Heed did not know that the 98.7% reliability of FIDS
reported to him at the meeting of AMD general managers on 19 J une
1998 only related to the availability of the host servers. He was never
told before AOD that the figure did not refer to the reliability of the
system. His understanding was that the system was 98.7% reliable, save
perhaps the display devices and the Terminal Management System
(TMS) which was still experiencing problems. Nor was he informed
that PD and GEC had agreed to defer the stress and load test of FIDS till
after AOD. He was not advised about the risk of not having the stress
and load test conducted before the system was put into use. Had he
known these two matters, he would have inquired further with his staff
and he would have realised that the risks involved in using FIDS on AOD
were increased. He would have paid more attention to the reliability of
the standby FIDS and the contingency plans in the case of a FIDS failure.
In the Commissioners view, there was a lack of communication and
co-ordination between W44 Heed and his colleagues in his own Division
and PD.

17.23 As AMD Director, W44 Heed only assessed the reliability of
each of the systems to be used for AOD separately, and he never assessed
the risks involved in a global manner in case of more than one of the
systems failing. Had there been a global assessment, he would have
realised that the contingency plan in case of a FIDS failure would require
reliance on Public Address System (PA), Trunk Mobile Radio (TMR)

512
System and most probably other communication systems such as the
telephone and facsimile. Whiteboards were set up shortly after 7 am at
the Departures Hall and Baggage Reclaim Hall for the passengers, but the
monitors and liquid crystal display (LCD) boards that were working
provided incorrect or outdated information. The incorrect flight
information shown on the monitors and LCD boards remained a source of
confusion to passengers: either they relied on the incorrect information or
they did not know that they should look at the whiteboards for accurate
information, unless they fortuitously asked airport attendants they
happened to see. PA would be required for notifying passengers of
flight information, especially gate changes. TMR, conventional and
mobile phones would be required by people like RHOs for operational
communications. In case of a FIDS failure, flight information could
only be obtained by operators of the airport community telephoning
Apron Control Centre (ACC) or Airport Operations Control Centre
(AOCC) to seek it. If they could not get through, they could attend
Airport Emergency Centre (AEC) to obtain the flight information
shown on a whiteboard, an arrangement not pre-planned before AOD but
only established as late as 7 pm on AOD. According to W5 Mr Allan
KWONG Kwok Hung, the Operations Manager of J ardine Air Terminal
Services Ltd. (J ATS) (one of the three RHOs), he attended an AA
meeting at AOCC at 4 pm on AOD when it was decided the first time that
whiteboards were to be put up at AEC. When the person obtained the
information at AEC, he would have to use a means of communication,
either telephone, TMR, mobile phone or fax, to relay that information to
his own company. The demand for use of such communications systems
would therefore be heavy, and arrangements should have been made
before AOD to ensure that such systems were effective and efficient
without overloading. W44 Heed, as Director of AMD, obviously failed
to have an overall risk assessment, especially in view of the history of
unreliability of FIDS, and did not work out a sufficiently careful
contingency plan with members of the airport community who were
required to be fully prepared in case of a FIDS failure. For this, W44
Heed must be responsible.

17.24 W48 Lam was also asked if he had made any overall risk
assessment as the person in charge of the AOR programme in AA. He
told the Commission that he had discussed the requirement of an overall

513
risk planning with W44 Heed. W44 Heed told W48 Lam that he had
contingency plans for each of the crucial AOR issues, and assured W48
Lam that in his expert opinion that was adequate. As W44 Heed was in
overall charge of operations, W48 Lam accepted his assurance.
Moreover, as there were a lot of things to be done at the time, W48 Lam
did not see any justification for setting aside resources to conduct an
overall risk assessment. The Commissioners feel that in the
circumstances it would not be fair to hold W48 Lam responsible.

17.25 The fact that W44 Heed did not know that the 98.7% only
related to the availability of the host servers and not the whole FIDS, and
that the stress and load test had been deferred till after AOD involving
consequential risk demonstrates a fault in communication and
co-ordination within AA, contributing towards the insufficiency of
planning and risk assessment in case of a FIDS failure. The two matters,
if W44 Heed had known, might have served as a good reminder to him as
to the importance of ensuring the availability of sufficient communication
channels for the use of the airport community. The responsibility for the
fault should properly be attributable to staff of AMD as a whole. That
said, as the Director of AMD in charge of the operation of the new airport,
W44 Heed himself must be responsible for his insufficient planning and
risk assessment in case of a FIDS failure, in particular in failing to assess
and analyse the reliability of the systems to be used at the Passenger
Terminal Building (PTB) on AOD as a global whole, which he
admitted in evidence.

17.26 Another blatant lack of co-ordination was in the
arrangements for the expert personnel of the Electronic Data Systems
Limited (EDS) (the subcontractor for FIDS) and the Preston Group Pty
Ltd (Preston) (the subcontractor providing TMS in FIDS) to be present
at the new airport on AOD to assist in case of trouble being encountered
by the AA operators of FIDS. Both W44 Heed and W45 Chatterjee
realised that because of the late commissioning of FIDS, the on-the-job
training and familiarisation for AA staff was compressed. W45
Chatterjee recommended that the contractors be asked to station
dedicated resources to support AMD and IT Department whenever
problems arose and to assist in the settling down of daily work. As a
result, EDS staff were on standby on AOD at additional cost. However,

514
there was no arrangement for the experts on FIDS and TMS to be
stationed at the crucial ACC where it turned out that operators did
experience difficulties in performing flight swapping with TMS and with
input into FIDS. Either W44 Heed, as Director of AMD responsible for
running the new airport, or W45 Chatterjee who was Head of IT, or both
of them, should have made satisfactory arrangements, but neither did so.
Indeed, what happened was that W35 Mr Gordon J ames Cumming was
staying all night at the EDS office close to AOCC, and W34 Mr Peter
Lindsay Derrick joined him at 6:30 am, but W34 Derrick was only able to
go to ACC late in the morning. On the other hand, when the operators at
ACC were experiencing difficulties, W28 Mr Anders YUEN Hon Sing
contacted W24 Ms Rita LEE Fung King at about 3 am but lost contact
with her until sometime around 6 am. When W34 Derrick was to attend
ACC, he was not able to do so because he did not have the required
access permit which was eventually made available to allow his
attendance at ACC at about 12:30 pm. Neither W44 Heed nor W45
Chatterjee could proffer any reasonable explanation for this failure of
co-ordination, for which both of them must be responsible.

Section 3 : Overview of What Went Wrong

17.27 After the discussions in Section 2 on the organisational
structure of AA and some specific instances of lack of co-ordination and
communication within AA, it may be beneficial to sum up the points that
the Commissioners see as being the causes internal to AA that were
responsible for the chaos on AOD and the days after. A number of these
causes were suggested by counsel for the Commission and have been
adopted by the Commissioners after careful consideration.

17.28 There was a major problem with personalities in the top
management of AA. The characters of W3 Townsend, the CEO, W43
Oakervee, a top engineer and Director of PD, and W44 Heed, an
established airport manager with over 30 years of experience and Director
of AMD, and their interaction have been dealt with in Section 2. These
three persons played the vital role of getting the new airport ready for
operation, from scratch to the construction stage, and from the completion
of the works and systems to the operational sphere. Due to the

515
preponderance given to the works side, insufficient heed was paid to the
requirements of the operators who were eventually to run the new airport
on AOD using the systems and facilities provided, still not fully tested.
The leadership and co-ordination problem were unfortunately only
exposed as late as the end of October 1997 when the Booz-Allen &
Hamilton report drew the attention of the members of the AA Board to
the problem. The report was commissioned because the Board decided
to have a review of the organisation and management structure for the
post-opening stage, not for AOR. By then, even if the Board knew
which part of the senior management should be replaced, it is not possible
to judge that they should have made the replacement and run the
substantial risk of disrupting the senior management, and adversely
affecting the transition from the works phase to the operational stage.

17.29 The delay in the construction phase had hard compressed the
time that was necessary for training the operators and allowing them to be
familiar with the facilities and systems that they were going to operate.
The delay in the commissioning of the systems, in particular, created
deficiency in the training and familiarisation, for the operators were from
time to time trained with systems that were subject to change and
improvement, and had to be trained again after the change or
improvement had been implemented. This is also linked to the fact that
many of the operators were still required to work in Kai Tak, while being
given time off for training on the different operational equipment in CLK.

17.30 The involvement of AMD and IT Department in system
development should have started much earlier. AMDs requirements
were not given high priority until sometime in 1997 whereas IT
Department, which used to be part of the Commercial Division, only
became involved from late 1997 when the new airport was due to open
for operation in April 1998. Had IT Department and AMD joined in the
planning and design of the systems to be used for the new airport much
earlier, correct or clearer functional requirements according to the users
needs would have been incorporated into the functional design
specifications in the systems contracts or fed into the systems at an early
stage. That would have resulted in less changes being required to be
made to the systems such as FIDS, saving the development of the systems
from delays for variations and providing more time for operators training

516
and familiarisation.

17.31 There was no planning to ensure a smooth transition from
the construction stage to the operational phase of the new airport, and no
experts or consultants had been engaged for that purpose. Such experts
might have helped in identifying the issues that needed to be resolved and
measures that needed to be implemented for a smooth transition. W44
Heed pointed out that there was no such expert, but AA had obtained
some experience from a few of its staff who had been involved in
opening other airports and sent staff to other countries to witness
preparation for such opening. However, during the crucial stage, which
should be around a year before the scheduled opening of the new airport,
no expert help was engaged to concentrate on a comprehensive
examination of the necessary measures to effect a smooth transition.
The Commissioners accept that there may not be a single person who is
qualified for such a job. A firm of consultants could have been tried or a
group of experts from different sources could have been formed for the
task. W43 Oakervee in evidence pointed to a firm that he was aware of,
and when the Commission was looking for experts to assist the inquiry, it
had quite a number of names supplied to it. W51 Mr J ason G YUEN
had experience in reviewing the transition from construction to operation
of airport facilities, and gave names of a few firms of consultants who
were known to him to be doing this kind of work, although he did not
vouch for their competence.

17.32 When the works and systems projects were completed late,
and when the testing and commissioning of the systems encountered
problems, there should have been an overall risk assessment. Although
there were various contingency plans made, they were directed at
addressing the failure of each individual system. There was insufficient
examination of the negative aspects of the interaction of the failures or ill
performances of more than one system. For instance, there was a
contingency plan for standby FIDS to be invoked in the event of the
failure of the main FIDS. However, when the main FIDS was not
operating smoothly and speedily and standby FIDS was not resorted to,
there were no adequate contingency measures to ensure that the RHOs
would be provided promptly with necessary flight information. TMR,
mobile phones and telephones were mostly working over capacity on

517
AOD, resulting in an inefficient passing of vital flight information.
There was also insufficient co-ordination by AA with the RHOs and
possibly airlines as to how to react when FIDS was not performing as it
should, causing confusion and delay in the provision of services to
aircraft and passengers.

17.33 Despite the lack of effective co-ordination amongst PD,
AMD and IT Department, all those who were working towards preparing
the new airport for operation on a fixed target date worked overtime and
were fully stretched. This was mainly caused by delays in the
completion of the works and systems severely compressing time left to
the target date. This had at least two consequences: the yearn of those
involved for achievement overbore their sense of risk and forced upon
them an over-optimism, and they were left with little time to spare to step
aside to look at the negative side or risks involved. The involvement of
consultants on the transition from works to operation might have
identified issues that could have been resolved and measures that could
have been implemented before AOD to avoid the sort of situation that
was the chaos on AOD. If they had advised that a smooth transition
would need more time, that would have instilled a required sense of risk
and insecurity into AA which would doubtless be compelled to seek a
deferment of AOD. The over-confidence was not only limited to what
AA could itself accomplish, but it also applied to AAs monitoring of
Hong Kong Air Cargo Terminals Limited (HACTL)s readiness. AA
took the assurance from HACTL that the latter would be ready, without
having any expert opinion on the correctness of the assurance based on an
examination of HACTLs cargo handling systems. AA was mainly
relying on HACTLs world reputation as one of the most efficient cargo
operators, gained from HACTLs long operation as such at Kai Tak and
the fact that HACTL had its own reputation and business interest to look
after for making SuperTerminal 1 (ST1) ready. Had consultants been
engaged to monitor HACTLs testing and commissioning of its systems,
this would not only have assisted AA in assuring itself that HACTL was
ready, but would certainly have helped HACTL to re-examine its own
assurance more carefully.

17.34 W49 Lo and W50 Wong were asked if AA had anyone
possessing the necessary expertise to monitor HACTLs systems, and

518
both of them did not believe so. However, they told the Commission
that the question was never raised in the Board and the AA management
had never indicated that there were or should be doubts regarding the
implementation of the HACTL systems. The Board was relying on
HACTLs assurance and its position as the experienced and reputed
expert in the field, and so did everyone else including the senior
management of AA, NAPCO and ADSCOM. The Commissioners feel
that the failure in engaging a consultant to monitor HACTLs systems was
primarily the responsibility of the AA management, for they should have
to satisfy themselves that HACTL was as good as its words. The AA
management should have raised the point with the AA Board, and the
failure to do so resulted in the non-specialist Board not considering the
possibility. The responsibility of NAPCO and ADSCOM in this respect
is dealt with in Section 4 of Chapter 5.

Section 4 : Misstatements and Responsibility for Them

17.35 There were two misstatements identified in the course of the
inquiry, the first was that the reliability of FIDS as a whole was 98.7%
available and the other was that Access Control System (ACS) had
been tested successfully. Although these misstatements were not direct
causes for the chaos encountered on AOD, the Commissioners think that
they had a significant bearing on the perceptions of top management of
AA as discussed in this chapter. The misstatements might also have
created a false sense of security in ADSCOM.

(a) FIDS

17.36 The representation made by AA to ADSCOM in ADSCOM
Paper 34/98 dated 23/6/98, prepared by AA for the ADSCOM meeting on
24 J une 1998 on the reliability of FIDS raised concerns on ADSCOM
having been misled. The following passage is taken from the Paper:

Reliability tests on the present version of FIDS (Version 2.01C)
commenced on 14 J une and were completed on 20 J une using live
data from Kai Tak through the AODB. The reliability of the
system as a whole has been 98.7% available; the reasons for

519
unavailability of some monitors and LCD boards at the 24 J une
trial have been identified and the problems are being rectified.

17.37 Both W45 Chatterjee and W43 Oakervee admitted that the
passage conveyed false ideas, while W44 Heed did not have sufficient
technical know-how as to comment. The truth of the matter is as
follows:

(a) The 98.7% was of the availability of the host servers, a
hardware and not a software, and not any other part of FIDS,
let alone FIDS as a whole;

(b) There is a slight difference between availability which
means the time when a system is operational as opposed to
downtime, and reliability which relates to the soundness and
consistency of the system; and

(c) The mention of the unavailability of some monitors and
LCD boards implied that the only problem with FIDS
preventing the achievement of 100% reliability was the
monitors and LCD boards, and this implication was false.

17.38 These false ideas misled ADSCOM, for its members all
understood that the 98.7% referred to the reliability of FIDS as a whole
system. However, because of the prior knowledge of ADSCOM
members on the continual unreliability or instability of FIDS during the
various tests up to that date, they placed greater reliance on the standby
FIDS that had been reported to have been successfully tested on 30 J une
1998 in case of a failure of the main FIDS. The false ideas therefore had
not, in the Commissioners opinion, caused too much mischief. Yet this
is something that should never have happened, as the Paper must have
gone through the heads of the Divisions of AA before it was sent to
ADSCOM, and the untrue statements should never have been allowed to
slip through.

17.39 The genesis of the misstatement was revealed in the oral
testimonies of W43 Oakervee, W44 Heed, W45 Chatterjee and W46
Bosher of AA, who together gave evidence before the Commission as a

520
group, and by a letter from AA dated 18 November 1998 in reply to the
Commissions queries regarding the matter. The letter enclosed
documents in support of AAs answers, but the documents were mainly
regarding the history of the AA Board Paper 183/98 dated 23/6/98 (the
same date as ADSCOM Paper 34/98) which contained a similar but not
identical statement, as follows:

Version 2.01C was loaded on 4 J une as the operational system
and can be used at airport opening. Progressive updates to the
base FIDS system that eliminate problem reports will be installed
in accordance with AMDs requirements. The reliability testing
was completed on 20 J une with 98.7% reliability.

17.40 As from 14 J une 1998, there had been reliability tests of the
host servers of FIDS, and the results showed that these servers were
98.7% available during the whole course of the tests. In a draft Board
Paper prepared by or under the auspices of W45 Chatterjee on 18/6/98, it
was correctly stated: Reliability and resilience testing has been
conducted. Stability of the system has improved to over 95%
availability. AA Board Paper 179/98 dated 23/6/98 for discussion at
the Board meeting on 25 J une 1998 and prepared by PD contained a more
particularised statement: The seven day system reliability test started at
9 am on Sunday 14 J une. After 4 days the commutative (sic, cumulative)
availability of the host servers and display servers was in excess of 98%.
While W45 Chatterjee admitted that there was a slight mistake in that
only the availability of the host servers and not the display servers was
tested, the passage was substantially in accordance with the facts and
correct.

17.41 The first draft of AA Board Paper 183/98 was produced by
Ms Pratima Patel of AMD on the basis of information provided by
relevant departments and the draft was submitted to W46 Bosher. From
the documentary evidence, it is clear that a draft of the relevant paragraph
was based on the said 18/6/98 draft prepared by W45 Chatterjee which
reads: Good progress has been made. Progressive updates to the base
FIDS system that clear operational deficiencies have been loaded.
Reliability and resilience testing has been conducted. Stability of the
system has improved to over 95% availability. Ms Patel circulated a

521
draft paper incorporating W45 Chatterjees draft passage to W46 Bosher
and W44 Heed and PD for comments between 19/6/98 and 20/6/98.
W45 Chatterjee also made further comments on 19/6/98 and 20/6/98, but
the relevant passage remained unaltered. On 20/6/98, Mr Nigel Milligan,
on behalf of W43 Oakervee suggested certain changes to the passage,
stating as follows:

Version 2.01C was loaded on 4 J une. This version can be used
at airport opening. Between now and 6 J uly progressive versions
that eliminate operational deficiencies will be installed in
accordance with AMDs requirements. The reliability tests were
re-started on 14 J une and will run to 20 J une.
(the penultimate version)

17.42 The penultimate version was also a correct statement, and
there was no mention of the percentage. There was another suggestion
of amendment of the draft from Mr Milligan on 22/6/98, but the
penultimate version was untouched. The draft substantially
incorporating the penultimate version was sent to W44 Heed, W46
Bosher and W3 Townsend for review and comments by Ms Patel on
23/6/98. From the drafts returned by W44 Heed and W46 Bosher, as
produced by AA, although some amendments had been made to the drafts,
the penultimate version again remained quite intact. From the
documentary evidence, which were contemporaneous, W48 Lam seemed
to have not been sent the drafts. The last person to whom the draft was
sent was W3 Townsend. While W46 Bosher could not find the draft
containing W3 Townsends comments, she said in response to AAs
counsel in re-examination:

I did make inquiries, Mr Ribeiro, but I have to say perhaps the
fault is mine here, it was not my practice to keep drafts. What
usually happened is that after Dr Townsend had looked at the
paper and made any changes that he wanted to, he would simply
walk into the next door office where I was and hand me the paper
with his amendments on it.

17.43 W46 Bosher maintained on oath that the inclusion of the
relevant passage was made at the very last stage of the drafting when the

522
Paper was cleared by herself, W44 Heed and W3 Townsend, and did not
appear to have come either from IT or PD. The relevant passage in the
finalised version of the AA Board Paper 183/98 is set out under paragraph
17.39 above. The penultimate version was altered in substance in two
respects: (a) operational deficiencies became problem reports; and (b)
The reliability tests were re-started on 14 J une and will run to 20 J une
was revised to become The reliability testing was completed on 20 J une
with 98.7% reliability.

17.44 The approval of ADSCOM Papers was generally
co-ordinated by W46 Bosher. On some occasions, W46 Bosher
produced a first draft of parts of the paper based on her understanding of
the position and circulated it to persons with relevant knowledge of the
particular subject for their comments, while on other occasions, relevant
individuals would produce the first draft of various paragraphs
comprising the paper and submit them to W46 Bosher for editing and
compilation. W46 Boshers practice was to circulate drafts to W43
Oakervee, W44 Heed, W45 Chatterjee and W48 Lam for their agreement
before the paper was sent to W3 Townsend for final approval. The
paper was then submitted to the Secretary for ADSCOM.

17.45 W46 Bosher disclosed that there was a draft of the
ADSCOM Paper 34/98, where the crucial sentence appeared in this
manner, The reliability of the (software ?) system as a whole has been
98.7% available; the reasons for unavailability of some monitors and
LCD boards at the 14 J une trial have been identified and the problems are
being rectified. It was she who put the question in parenthesis in the
draft. Having gone through W43 Oakervee, W44 Heed, W45 Chatterjee
and W48 Lam, and after their comments were collected, this draft was
sent to W3 Townsend before it was finalised in the form as quoted in
paragraph 17.36 above. It will be noted that the (software ?) which
was raised by W46 Bosher was omitted in the finalised form. None of
the four witnesses, W43 Oakervee, W44 Heed, W45 Chatterjee and W46
Bosher could remember what exactly happened to the draft and why the
offending statement appeared in the manner as finalised or who was the
person who was responsible for the finalised version.

17.46 W48 Lam was not able to identify the author of either

523
ADSCOM Paper 34/98 dated 23/6/98 or Board Paper 183/98 dated the
same date. He denied himself being the author. His involvement in the
day-to-day activities regarding the systems was limited, partly because he
had no technical knowledge and partly because his sick leave only ended
on 15 J une 1998. As far as he was concerned, he did not know the exact
meaning of the offending passage Reliability of the system as a whole
has been 98.7% available, nor did he know the difference between
reliability and availability. He said that as a layman, to him the two
words were perhaps just the same thing, but reliability meant more
certainty. He merely relied on the reports about the progress of the
systems from officers of AA and the attendees at the AOR Meetings and
the System Hand-Over Meetings which were mainly chaired by him.
Nothing in the records of these meetings could indicate that the offending
passage in the ADSCOM Paper or the Board Paper was wrong. On the
contrary, in the notes of the 22/6/98 AOR Meeting, it was stated that As
FIDS is presently over 90% reliable it will be used at airport opening.
W48 Lam thought that percentage related to FIDS as a whole, and he
thought it meant FIDS was over 90% okay. W48 Lam remembered that
at the time of the AOR Meeting, as the latest progress of FIDS was 90%
okay, it was considered that FIDS would be used on AOD instead of the
standby FIDS. With that knowledge in mind, W48 Lam would not have
been able to discern that the ADSCOM Paper or Board Paper presented
any incorrect information.

17.47 It appears from the notes of the 22/6/98 AOR Meeting that
the term 90% reliability started to creep in, and it found its way into the
finalised version of the AA Board Paper 183/98 but with a higher
percentage. When W3 Townsend was recalled to be given an
opportunity to deal with this matter, he accepted that it was possible that
he was the author of the finalised version of the Board Paper. He further
said that almost always those (the drafts) would be passed forward to me
before they would be sent to ADSCOM or the Board. This tallies with
the evidence of W46 Bosher and W48 Lam. From the evidence on the
genesis of the Board Paper, it seems clear to the Commissioners that the
finalised version of that Paper was made by W3 Townsend, because all
the draft versions that had gone through W43 Oakervee, W44 Heed, W45
Chatterjee and W46 Bosher did not use the term of 98.7% reliability.
W48 Lam appeared from the evidence not to have been sent the draft at

524
the last stage, because Ms Patel addressed the penultimate draft to only
W46 Bosher, W44 Heed and W3 Townsend. The penultimate version in
fact did not even mention any percentage, and apparently, the finalised
version was incorporating the results of the tests up to 20 J une 1998 on
the host servers being of 98.7% availability. The finalised version of
98.7% reliability without the qualification that it related to host servers
(and the mistakenly included display servers) was misleading to the AA
Board, and apparently also found its way into the ADSCOM Paper 34/98,
although there is little evidence as to how.

17.48 There was a distinct suggestion by W3 Townsend that W45
Chatterjee, being Head of IT, would have been involved in drafting and
reviewing ADSCOM Paper 34/98. W45 Chatterjee told us in evidence
that he did not remember whether he drafted the offending paragraph in
the ADSCOM Paper. However, in the drafts for Board Paper 183/98
that W45 Chatterjee admitted to have been originally prepared by him,
the following was repeated, Reliability and resilience testing has been
conducted. Stability of the system has improved to over 95%
availability. He never used 98.7% and he maintained the word
stability. The Commissioners do not believe that W45 Chatterjee was
the author of the offending passage in either the ADSCOM Paper or the
Board Paper.

17.49 From all the evidence, the Commissioners are satisfied on a
balance of probabilities that the finalised versions of the AA Board Paper
183/98 and the ADSCOM Paper 34/98 were made by W3 Townsend, and
he must be personally responsible for putting on paper the misstatement
to the AA Board and ADSCOM.

17.50 W50 Wong and W49 Lo gave evidence before the
Commission on their understanding of the relevant paragraph in the AA
Board Paper 183/98. Both of them understood the sentence to mean that
the whole FIDS was reliable to the extent of 98.7%. They had been
advised by the senior management of AA that 95% reliability would be
acceptable for operation. Of course, Board Paper 179/98, which was
available to the AA Board at the same time as Board Paper 183/98,
mentioned that tests had been conducted and after four days, the
commutative (sic, cumulative) availability of the host servers and display

525
servers was in excess of 98%. W3 Townsend pointed out that if the
Board members had read and asked questions, they would have been
aware of the reference to host servers and display servers in the context of
availability in excess of 98%. W50 Wong was not at the meeting on 25
J une 1998, which was chaired by W49 Lo. W49 Lo told the
Commission that he treated the two Board Papers to be talking about
different things. He understood that the host servers and display servers
were hardware of FIDS, but he felt comforted by the 98.7% reliability
referred to in Board Paper 183/98 because he had been told that 95%
reliability would be acceptable for operating the new airport.

17.51 There were two ADSCOM meetings following the service of
the ADSCOM Paper 34/98 before AOD, on 24 J une and 4 J uly 1998.
W45 Chatterjee clarified that the ADSCOM meeting on 24 J une 1998
took the form of a walk around as the ADSCOM members were visiting
PTB to observe the building and facilities on the ground, and there was
no specific discussion about the offending paragraph. Indeed, apart
from some tables and lists prepared for ADSCOM members, no notes or
minutes as to what was said and by whom were kept for that meeting.
Both W45 Chatterjee and W46 Bosher who attended the ADSCOM
meeting on 4 J uly 1998 told the Commission they did not remember the
figure of 98.7% being mentioned at the meeting. The evidence tallies
with the memory of W36 the Chief Secretary. On the other hand, W3
Townsend said when he first gave evidence on Day 10 that after the Paper
was provided to ADSCOM, the matter was mentioned subsequently with
ADSCOM and he did not think ADSCOM was misled. This he
subsequently withdrew when asked on Day 48 at his recall to the witness
box. Anyhow, the notes of the ADSCOM meeting on 4 J uly 1998
recorded no reference to this figure. Rather, the following was the only
record on FIDS:

5. On FIDS, HIT/AA (ie, W45 Chatterjee) reported that the
permanent FIDS continued to be stable. FIDS had been running
continuously since 22 J une and with Airport Management
Divisions permission, there had been controlled bring-downs to
update the software. HIT/AA said that there would be
workarounds when a function of the system went down, and the
workarounds had been tested and found to work well.

526
6. The switch over from permanent FIDS to the standby had
also been tested on Thursday last with the assistance of 35 airlines.
Within 30 minutes, most displays were switched on. HIT/AA
confirmed that that was acceptable from the operational point of
view. During the switch over, the information displayed on the
LCD boards and monitors would become out-dated. To remedy
the situation, the Public Address system could be used to
disseminate up-to-date information.
7. HIT/AA explained that during the workarounds, the system
would be in the permanent FIDS environment. AA would try to
re-boot the system. Meanwhile, white boards and extra hands
would be available to help with directing the passengers in the
problem area. Such happenings were not uncommon in an
operating airport.

17.52 The statements recorded as made by W45 Chatterjee
concentrated more on the reliability of the standby FIDS and the
workarounds, and there was no mention of the 98.7% figure. This
reasonably contemporaneous record of the meeting is consistent with
what W36 the Chief Secretary told the Commission from her memory
and corresponds with her evidence that she did not place too much
reliance on the figures, but focussed more on the success of the testing of
the standby FIDS that was reported to have taken place on 30 J une 1998.
Coupled with the evidence of both W45 Chatterjee and W46 Bosher that
they did not remember the figure being mentioned at the 4 J uly meeting,
the Commissioners are satisfied that the evidence of W36 the Chief
Secretary, W45 Chatterjee and W46 Bosher is to be preferred to that of
W3 Townsend who, when cross-examined on Day 10 with what was said
by way of explanation of the figure, could not point out anything specific.
At his returning to be questioned again on Day 48, W3 Townsend
withdrew his previous evidence that the matter was subsequently
discussed. The Commission finds that the misstatement that FIDS was
98.7% reliable as a whole was made to ADSCOM and no clarification of
its true meaning was ever proffered to ADSCOM members. W3
Townsend was probably the author. He was present at the ADSCOM
meeting on 4 J uly 1998 and knew that this statement was contained in
AAs ADSCOM Paper 34/98 and that it was either misleading or untrue,
but he did not disabuse ADSCOM members of the false meaning. W45

527
Chatterjee who was also at the meeting should have appreciated the
falsehood in the offending passage. He would certainly have read the
paper before attending the important meeting and been aware of the vast
discrepancy between the wording and meaning of the offending passage
and those contained in his own drafts for the current Board Paper. After
all, he had the specific responsibility to monitor the progress of FIDS.
Yet he did not point out the mistakes to the meeting either.

17.53 The Commissioners have considered hard as to whether
there was intent to mislead on the part of W3 Townsend. He testified
that it was more a matter of editing the paper than any deliberate attempt
to present any confusing information. The Commissioners come to the
conclusion that the evidence is not weighty enough for an inference to be
drawn that there was clearly an intent on W3 Townsends part to mislead
ADSCOM. They are of the view that at least both W3 Townsend and
W45 Chatterjee were grossly negligent in allowing the misstatement to
remain unexplained at the ADSCOM meeting on 4 J uly 1998. From the
evidence of W36 the Chief Secretary, it appeared that she placed more
reliance on the availability of the successfully tested standby FIDS, and it
appears that even if she and other members of ADSCOM had been told
the true meaning of the 98.7% figure, little difference would result in the
deliberation by her and her colleagues on the readiness of the new airport
to open on AOD. Nonetheless, had they known the true meaning of the
figure, ADSCOM members might have considered to impress upon AA
that the standby FIDS should be used for AOD instead of the main FIDS,
in view of the relative reliability of the former system. That is, however,
a purely hypothetical matter.

(b) ACS

17.54 ACS is the acronym for the Access Control System. The
184
th
ADSCOM meeting on 6 J une 1998 was chaired by the Financial
Secretary and attended by W3 Townsend, W44 Heed and W45 Chatterjee,
amongst others of the senior management of AA. The notes recorded
that after W45 Chatterjee reported to the meeting that tests of some of the
systems had been successfully concluded,

CEO/AA (ie, W3 Townsend) added that the four key safety and

528
security systems access control, fire alarm, closed circuit
television and public address system had also been successfully
tested. They were at the moment busily engaged in issuing
access cards.

17.55 In fact, in the PDs Construction Monthly Report for May
1998, but with information up to 8 or 9 J une 1998, it was reported on
ACS that

The installation process has been improved with more doors
installed. A programme with forecast door security energisation
dates from the PTB is now being reviewed. There is still
however a serious concern at the lack of engineering resources to
commission to core system and to resolve engineering problems
which neither Guardforce nor their subcontractor seem willing to
address.

17.56 W43 Oakervee explained to the Commission that although
Guardforce Limited was the ACS contractor, the software was provided
by the subcontractor, Controlled Electronic Management System Limited
in Belfast whom AA had difficulty in getting to come out to Hong Kong
to address the engineering problems. W43 Oakervee said that ACS was
a serious concern because it was such an important subject that Mr Heed,
Mr Siegel and I and Billy Lam all had our attention and minds focused
completely on it. Dr Townsend would have known also. it was a
key aspect that it hinged about the issue with the aerodrome
licence. By the time of the ADSCOM meeting on 6 J une 1998, the
problems had not yet been fixed, and indeed, up to the day when the four
senior officers of AA gave evidence together before the Commission,
ACS problems had not yet been fully rectified. In view of the clear
concern expressed in the said Construction Monthly Report, W43
Oakervee agreed that the statement of W3 Townsend to ADSCOM at the
meeting of 6 J une 1998 was incorrect. W44 Heed also testified that the
falsity of W3 Townsends statement at the meeting was also obvious to
him. W44 Heed said in evidence that ACS related to security within the
airport for which his AMD had overall responsibility and that ACS was
not in a position to be used yet and he knew that there were a lot of
problems with the system. W45 Chatterjee was at the meeting but did

529
not utter anything openly or do anything in any inconspicuous way to
correct W3 Townsend there and then, nor afterwards intimated to him that
he had made a mistake. The untruth was allowed to be unclarified
before all the attendees of the meeting. This is inexcusable. W45
Chatterjee explained that he was not specifically responsible for ACS as
opposed to FIDS, and he had not followed ACS to that level of detail.
At the said ADSCOM meeting, he had just finished explaining about
FIDS when W3 Townsend remarked about the successful tests of ACS
and he therefore did not make a mental connection with ACS.

17.57 W48 Lam also attended the ADSCOM meeting held on 6
J une 1998. However, his knowledge on ACS at the time was quite
limited. The following is a summary of the information relevant to ACS
he received at the System Hand-Over Meetings and the AOR Meetings at
the end of May and early J une 1998, as evident from the notes of those
meetings:

(a) At the AOR Meeting on 25/5/98, it was reported that fire
detection system tests were ongoing, that ACS was 95%
ready and still on target for hand-over at the end of the
month, and that Aviation Security Company Limited was
working alongside at workstations on training. It was also
recorded that W48 Lam said that it was absolutely essential
that ACS be ready for testing of the permit system by 1 J une
for the Enhanced Security Restricted Area to start function
by the 14 J une airport trial.

(b) At the System Hand-Over Meeting on 28/5/98, a
representative of PD confirmed that major software
problems encountered with ACS had been overcome and the
SAT would be carried out the week following to be
completed by 5 J une.

(c) At the AOR Meeting on 1/6/98, it was reported that takeover
of ACS was scheduled for 1 J une.

(d) At the System Hand-Over Meeting of 5/6/98, a representative
of PD confirmed that the SAT for ACS could be completed by

530
8 am on 8/6/98, and was confident that majority of the system
functionality would be ready for operation.

It can be seen that with the above information in mind, it did not appear
to W48 Lam that the information on ACS given by W3 Townsend at the
ADSCOM meeting of 6 J une 1998 was misleading.

17.58 During cross-examination, W44 Heed never denied that he
was at the ADSCOM meeting of 6 J une 1998. He admitted that he knew
that what W3 Townsend said about ACS was not true, but he did not
correct him. His explanation was that he did not think that it was his
place to speak up on that occasion, that it would border on
insubordination, that there were others at the meeting who knew the true
situation, and that his role at the ADSCOM meeting was to support W3
Townsend if he was asked any questions by ADSCOM members. W44
Heed said that he merely let the matter pass, not having a private word
with W3 Townsend, nor did he think it necessary to do so for W3
Townsend should have known the situation. The following exchange
between counsel for the Commission and W44 Heed is important and
indicative of the witnesss attitude:

Q: So it does not matter that the Financial Secretary was misled
because other people did not bother to correct Dr Townsend;
is that right?
HEED: Yes, that is right, yes.

17.59 W3 Townsend was recalled by the Commission to be given
an opportunity to answer the allegation of falsehood against him. He
agreed that from the vantage of hindsight I probably should have
elaborated more on that particular point and been more specific, and I
feel that perhaps further detail may have been appropriate. His
explanation was that time was valuable in those days, and generally we
would go through the programmes at the ADSCOM meetings in a timely
way, so it lent itself towards trying to summarise various points that had
been recorded in the papers. He denied having an intent to give false
information.

17.60 As with the misrepresentation of the 98.7% reliability of

531
FIDS as a whole, the Commissioners have not been able to find
sufficiently weighty evidence to sustain a finding of wilful intent on W3
Townsends part to mislead ADSCOM about the progress of ACS.
Nonetheless, the Commissioners find that W3 Townsend must be the
main culprit in making the misstatement to ADSCOM. As far as W45
Chatterjee is concerned, after evaluating his testimony, the Commission
feels that it might be unfair to him to hold that he should have disabused
ADSCOM of the misleading statement. Quite unlike his position
vis--vis FIDS, he was not specifically responsible for the development
of ACS and he was concentrating on explaining to the meeting about
other systems, immediately after which, W3 Townsend made the untrue
remark about ACS. There may be some truth in W45 Chatterjee saying
that he was not mentally alert about ACS in the circumstances.

17.61 The admitted involvement of W44 Heed in not providing
any clarification of W3 Townsends misleading statement at the
ADSCOM meeting on 6 J une 1998 took a strange turn when W48 Lam
gave evidence. W48 Lam said that it seemed to him that W44 Heed was
not there at the meeting. As a result, the notes of the ADSCOM meeting
were carefully checked and it was discovered that while W45 Chatterjee
was present when W3 Townsend made the statement, W44 Heed joined at
a later juncture upon W45 Chatterjee leaving the meeting. Had this
been pointed out to W44 Heed when he was giving evidence, he would
not have been subjected to the cross-examination, which revealed his
attitude towards the matter. As he was not at the meeting when the
statement was uttered by W3 Townsend, W44 Heed cannot be responsible
for not making any attempt to point out the incorrectness or mistake of
the statement to W3 Townsend or more importantly to ADSCOM. What
is damnable is that he saw himself at such a meeting merely to support
the CEO, W3 Townsend and to respond to questions put, but not
bothering if ADSCOM was misled. In the opinion of the
Commissioners, he was unbecoming of his position as a member of the
senior management of a large organisation such as AA. W44 Heed told
us that W3 Townsend must know that the statement was incorrect, and it
was therefore not for him (W44 Heed) to point out the mistake. From
the answers given by W44 Heed, there were perhaps several explanations
for his taking such a stance. He might be (a) too submissive to his boss;
(b) too loyal to him; or (c) too embarrassed to point out his mistake.

532
Whatever the reason, his attitude is reproachable, because he would allow
his boss to state an untrue fact to ADSCOM in the course of its being
apprised of AOR critical issues, and did not mention to W3 Townsend
that he had made a mistake at or after such an important meeting. Even
W3 Townsend did not seem to approve, as he had this to say: I am also, I
might add, rather surprised that Mr Heed did not speak up at the
meeting, because it was a special review meeting, people were normally
encouraged to express different opinions. Both W49 Lo and W50
Wong said that W44 Heeds attitude was inappropriate in the
circumstances. Albeit W3 Townsend was the main culprit in the act,
W44 Heeds own evidence exposed an attitude unbefitting of his senior
position, and not worthy of the trust that ADSCOM must have placed in
him such that it invited him to attend its meetings from time to time.
His attitude makes it doubtful that he should be entrusted with the
important task of being in charge of the management and operation of the
new airport. One may argue, however, that being a very experienced
airport manager, he must be up to the job. That may very well be the
case, but the doubt as to his appropriateness relates to his integrity and
attitude towards his responsibility which were only revealed because he
did not deny that he was present at the ADSCOM meeting on 6 J une 1998
at the crucial moment. At the very least, his attitude makes it doubtful
whether he could properly handle matters in a crisis or delicate situation.

17.62 The Commissioners have also considered the two misleading
statements from a broader perspective. In evidence W43 Oakervee,
W44 Heed and W45 Chatterjee all viewed that 6 J uly 1998 was a target
date that all had to work towards, although that date was not irreversible
in case any major item of AOR could not cope. Everybody was working
extremely hard with a view to making that date successful, and focussing
on all the AOR critical issues. None ever thought of having the date
deferred. All were thinking of fulfilling AOR on AOD and bearing the
burden or pressure while aiming at making the opening a success. They
admitted that because of all these factors in the prevailing circumstances,
as W43 Oakervee described their being in the box, they might have
been over optimistic as to what they had achieved, rather than critically
examining the risks involved in the things that they had not completed.
They were too involved with the goals in their own sphere, driving
ahead with each of their own domains (W43 Oakervees words) and

533
because of the workload did not have the opportunity to step outside
their bounds (W45 Chatterjees evidence). They never took a step aside
to look at the situation as a critical outsider. They subconsciously
viewed the facts known to them on the bright side, to bolster up their
confidence and belief that AOD was manageable, and this confidence had
a vital influence on their presentation of the facts and their views to
ADSCOM. They did not feel they were painting too rosy a picture for
ADSCOM, for they were lulled by their own self-induced confidence and
sense of achievement rather than dishonestly misleading ADSCOM.
The Commissioners feel that there is a certain ring of truth in this, and
agree that W3 Townsend should be responsible for his truly held false
confidence and belief in allowing ADSCOM to be misled about the
98.7% and about the status of ACS, rather than for wilfully misleading
ADSCOM. The others like W44 Heed and W45 Chatterjee were also
too immersed in the drive to accomplish and too imbued with the sense of
optimism somewhat forced upon them by the circumstances to pay
sufficient heed to reality. They were thus susceptible to adopting and
acquiescing in views that were swayed from the facts.

17.63 The misrepresentations to ADSCOM and to the AA Board
made by W3 Townsend as the most senior person in the AA management
are most deplorable. It is clear that he put the misrepresentation of
98.7% reliability in the AA Board Paper, and more probable than not that
he put the misstatement of 98.7% as being the reliability of FIDS as a
whole in the ADSCOM Paper that AA presented to ADSCOM.
Moreover, his making the untrue statement about ACS face to face with
ADSCOM members has been proved beyond reasonable doubt. As the
CEO of AA, who had the duty to plan, develop and operate the new
airport, and therefore to make it ready for operation on AOD, he betrayed
the trust that was reposed in him by both the AA Board and ADSCOM,
especially regarding AOR critical issues. While it is unfortunate that
none of his subordinates were able to correct his mistakes, it remains that
he must personally be held responsible for what he has done. Indeed,
viewing it from another angle, he had created the circumstances which
dragged W44 Heed and W45 Chatterjee, his unwitting and unwise
subordinates, into the blame and disrepute that he should otherwise
properly face all by himself.

534


17.64 Whilst Sections 2, 3 and 4 deal with an overview of the
responsibilities of various persons under their respective headings, this
section specifically deals with the responsibility of the top AA
management and the AA Board.

17.65 The Commission finds that the AA management failed to
maintain a right balance between PD and AMD in two ways. First,
AMDs participation in project and systems development was not
provided for in an early stage. W43 Oakervee told the Commission that
PD did not have a client, in the sense that PD did not work towards the
requirement or satisfaction of anyone. W44 Heed as Director of AMD
lamented that that client should have been there in 1992 not 1994,
emphasising the lateness of AMDs involvement in the project
development stage. W48 Lam, the Deputy CEO who joined AA in
J anuary 1998, also agreed. He said:

we should have somebody from the users point of view, from
the operational point of view, who could feed back the users
requirements to the Project staff from the very start that means
there is no consistent stream of feedback from the users angle,
and that partly means that some of the users requirements or
operational requirements were not fed into the system or were not
given loud voices at the very beginning.

17.66 AMD was the ultimate operator of the new airport, and it is
obvious that the operators input ought to have been sought from the
beginning, but it was not done. This resulted in AMDs requirements
not being fully taken into account during systems development. A
glaring example is that the particular technical specifications (PTS)
used for the tendering of the contract for FIDS were not prepared by
AMD or in full consultation with AMD. As a result, the PTS based on
which GEC and its subcontractor EDS took up the contract did not
represent fully or sufficiently AMDs needs, and the FIDS software
programme had to be written from scratch. This caused a delay in the
development of FIDS of about 14 months and payment of $89.7 million

535
to the contractor for the aborted and variation works.

17.67 Secondly, the personalities of the persons occupying key
posts caused problems. This has been discussed in detail in paragraphs
17.18 and 17.28 above.

17.68 For the purpose of the inquiry, the acts and omissions and
therefore the responsibilities of the following persons in the top AA
management have been examined in detail, namely, W3 Townsend (the
CEO), W48 Lam (the Deputy CEO), W43 Oakervee (Director of PD),
W44 Heed (Director of AMD) and W45 Chatterjee (Head of IT).

(a) W3 Townsend

17.69 Under the AA, the CEO was responsible for the general
management and administration of AAs affairs: the Airport Authority
Ordinance section 15(1)(b)(i). Prior to AOD, that duty must include
both construction and operational readiness of the airport, as W50 Wong
and W49 Lo stated in evidence. One would therefore reasonably expect
that the CEO of AA, the person in charge of the construction and
operational readiness of the airport, would be sufficiently experienced in
airport construction or at least airport management to enable him to
discharge that duty properly.

17.70 However, as the Commission has heard from W3 Townsend,
prior to his appointment as the CEO (then with the Provisional Airport
Authority) he had no experience with airport management or construction.
There is therefore some truth in W48 Lams assessment that W3
Townsends engineering background and lack of experience in airport
management would be a reason why operations were overlooked.

17.71 From the totality of the evidence presented to the
Commission, it is more probable than not that W3 Townsend was not up
to the task entrusted to him. The relevant comments by various persons
as recorded in documents are set out hereunder for ease of reference:

(a) At the ADSCOM meeting on 7/11/97, it was recorded:
DCA had no faith in the top management of AA. The

536
project was driven by PD/AA who always tried to bulldoze
his way through. CEO/AA was not in control and the
organisation was not functioning as it should.

(b) In the Chairmans Brief for the 7/11/97 ADSCOM meeting,
W48 Lam (then Director of NAPCO) reported that: On the
other hand, Chern Heed still allows himself to be pushed
around and CEO/AA is not backing him up. There is talk
to get Clinton Leeks to take over the training and trial
programme and Howard Eng to underpin Chern Heed; this
has yet to be confirmed.

(c) At the same ADSCOM meeting, the following discussion
was minuted: On systems, there was no one within AA
who was experienced in this field. On operations, despite
some experienced airport management staff, especially
those from Kai Tak, there was no one within AA who had
experienced the transition of an airport from the
construction to the operational phase. NAPCO had some
(International Bechtel Company Ltd.) staff with such
experience and he was thinking of seconding them to AA,
to help take things forward. DCA pointed out that there
was staff in AA who had worked in Kai Tak. The
unfortunate thing was that these staff had no clout to ensure
that things that should be done were in place.

17.72 In ADSCOM Paper 34/97 of 19/9/97, NAPCO made the
following recommendations to ADSCOM:

We find that co-ordination within the AA itself, particularly
between AMD and the Project and Commercial Divisions, as
well as co-ordination and cooperation between the AA,
Business Partners, Government and all others requires
intensified attention and immediate improvement. The
matrix organisational split of AOR responsibilities between
the various AA Divisions is not functioning efficiently, and
information and decision-making bottlenecks exist. We also
find that programme and other essential information is (sic) not

537
fully shared between the AA and Business Partners. We
recommend herein that a single-point responsible high-powered
executive manager be vested with requisite and clear authority
to direct the AOR process and dictate action inclusive of all
participants - AA (including all Divisions), Business Partners
and Government. We further recommend that a full open
book approach to co-ordination and information sharing be
implemented immediately inclusive of all participants.

17.73 The Commission accepts the submission of counsel for the
Commission that the very fact that NAPCO had to make such
recommendations showed that:

(a) W3 Townsend was not in control of and not able to
co-ordinate the various divisions;

(b) The AA management, under W3 Townsend, was not able to
co-ordinate and cooperate with business partners and
Government;

(c) W3 Townsends management style did not work; and

(d) Although he was the CEO, W3 Townsend was not able to
play the role of the single-point responsible high-powered
executive to direct the AOR process efficiently.

17.74 The evidence therefore points quite clearly that W3
Townsend was not in control of the management, resulting in lack of
co-ordination between the PD and AMD. He did not give sufficient
priority and adequate support to operational requirements of AMD,
especially since the end of 1997 when more preponderance should have
been accorded to AMD in the transitioning of the project stage to the
operation sphere. He did not assign sufficient resources to AMD at an
early stage, and failed to give sufficient support to W44 Heed, who was
frustratingly left with a FIDS that was not fully ready and with
compressed time for training and familiarisation for his staff. He did not
engage an expert to monitor HACTLs systems (see paragraph 17.34).
All these ultimately resulted in the deficiencies in the operational

538
readiness of the airport. In addition to his general inability to maintain
proper control over the management, as the CEO, W3 Townsend must
bear overall responsibility for the failings of the senior management. In
particular, he must be responsible for failing to have any or any proper
global assessment of AOR (including ensuring sufficient contingency
measures had been put in place) or the risks involved in opening the
airport on AOD with incomplete critical systems. He is further
responsible for the misstatements he made to the AA Board and
ADSCOM referred to under Section 4 above.

(b) W48 Lam

17.75 W48 Lam has been found by the Commission not to be
responsible for the problems witnessed on AOD, or for the lack of
communication and co-ordination or for the misrepresentations. The
details of the review of his involvement in various issues can be found in
paragraphs 17.24, 17.42, 17.46, 17.47 and 17.57 above. In fact, he was
disabled by an accident on 12 March 1998 and did not resume his duties
as the Deputy CEO until 15 J une 1998 (see paragraph 17.17). He was in
charge of the mammoth relocation exercise, the major phase of which
took place in the night between 5 and 6 J uly 1998, and the Commission
has not received any complaint on this score.

(c) W43 Oakervee

17.76 W43 Oakervee showed himself to be a straightforward
witness. His strong and aggressive character is borne out not only in the
documentary evidence but also when he was giving evidence before the
Commission. Though unfortunately these attributes of W43 Oakervee
operated unfavourably towards the composition of the senior AA
management and the interaction amongst the personalities occupying the
AA top posts, the Commissioners have the impression that they worked
very well for him in the position of the PD Director. However, there
were various slippages of the construction and systems programmes,
which even W43 Oakervee was not able to eliminate.

17.77 In a letter dated 8 J anuary 1999 to the Commission, W44
Heed responded to various allegations against him. At paragraph 5 of

539
the letter, he stated as follows:

the responsibility to provide the airport facilities and
systems and have them ready for AOD resides with the
Authoritys Project Division. None of the facilities and
systems on AOD were accepted from the contractor. Although
official hand-over for use by AMD for some systems, i.e. lifts,
escalators, etc. had taken place, others were being used by
AMD because they were required. However, the
responsibility for these systems, i.e. ACS, PA, chillers, etc.
rested with Project Division and their contractors.

17.78 Reading this quote from W44 Heeds letter and with W43
Oakervees express acceptance of responsibility in his testimony, it is
clear to the Commission that W43 Oakervee and PD which he heads must
be primarily responsible for the slippages and the unreadiness of facilities
and systems. Nonetheless, the Commissioners consider that slippages in
construction programmes are almost unavoidable, and in view of the fact
that there is no evidence that W43 Oakervee failed in his duties as
Director of PD, the Commissioners do not think that too much blame
should be attached to him.

(d) W44 Heed

17.79 The AMD, being responsible for the management and
operation of the airport, is primarily responsible for the problems and
shortcomings witnessed on AOD. W44 Heed, as the Director of AMD,
must take the major share of the blame, despite his pleas in the
above-mentioned letter dated 8 J anuary 1999. The unreadiness of the
facilities and systems will be discussed later. W44 Heeds personality
was too weak as compared with W43 Oakervees and he did not have the
support of W3 Townsend. His inadequacies and weakness contributed
to the problems encountered on AOD.

17.80 First, as the director of AMD, W44 Heed ought to have stood
firm vis--vis PD, in particular his counterpart W43 Oakervee, to ensure
that AMD would have sufficient time to get properly prepared for AOD.
When he was cross-examined by counsel for the Commission, he

540
conceded that it was partly true that he had allowed himself to be pushed
around and that the CEO was not backing him up. His weakness in
itself would not have been too much of a problem and might even have
helped his relationship with his subordinates, but was problematic when
interacted with W43 Oakervees strong character. This was apparent
from the following exchange in W43 Oakervees cross-examination on
the hand-over of the FIDS, ACS and PA from PD to AMD:

COUNSEL Q: But you heard Mr Heed say that he had no
alternative. Do you blame him for taking over the systems?
Do you blame Mr Heed for taking over the systems when
they were not really fully tested and commissioned?
OAKERVEE: I do not blame anybody. It was entirely Mr
Heeds choice as to whether he did it or not.
Q Would you say that he should speak up and say: I am not
going to accept them?
A I cannot speak for Mr Heed.
DR CHENG: If you were in his position, would you have done
that?
A Bearing in mind that I have no knowledge of running an
airport, I may have been a bit more aggressive, yes.

17.81 W48 Lam also told the Commission of occasions when W44
Heed was too shy to make his points across to W43 Oakervee, and W48
Lam had to call up W43 Oakervee to intervene and ask W44 Heed to put
down his views in writing.

17.82 Secondly, W44 Heed failed in his duty to ensure that he was
kept properly informed of the progress of the FIDS development so as to
enable him, as head of the AMD, to make an informed assessment as to
the readiness of the FIDS for AOD. In this connection, he has failed in
at least three major respects:

(a) He failed to ensure that AMD would be consulted on major
decisions and stages in the programme which might have
repercussions on operational readiness of the airport. For
example, he let the point of no return (having stand-alone
builds or one integrated build of FIDS) pass by without even

541
knowing it. W42 Mr NG Ki Sing, the General Manager
(Terminal Operations), said that AMD did not find out until
February 1998 that the point of no return had already past.

(b) He failed to ensure that he had a proper and accurate
understanding of the statistics about the reliability or
availability of FIDS given to him towards the end of J une
1998 when a decision had to be taken as to whether to use
FIDS on AOD. He thought that figures of over 90% and
98.7% reliability reported by IT referred to the reliability of
the whole system, whereas in fact the figures only referred to
the up-time of the host servers and did not include the
software. As a result of his ignorance, he made an erroneous
report to W48 Lam on the reliability of FIDS by stating in his
memo dated 19/6/98 that At yesterdays systems meeting a
review of the reliability of the FIDS was discussed.
Although the FIDS reliability tests indicated 97-98%
reliability, the TMS stand allocation module is not to the
standard for operational use. His decision to use FIDS on
AOD was therefore made on a wrong basis.

(c) He did not even know that a decision had been made
sometime between 19 and 22 J une 1998 to defer the stress test
for FIDS. Needless to say, he was not advised of the risks
involved in going ahead with the main FIDS without having a
proper stress test.

17.83 The result of the above failures on W44 Heeds part
translated into the absence of any proper assessment of the risks involved
in using FIDS on AOD. As the head of AMD, he cannot possibly escape
responsibility.

17.84 Thirdly, as an experienced airport management professional,
he ought to have ensured that an appropriate overall risks assessment was
carried out during J une at the latest so as to assess the risks involved in
proceeding with the opening as scheduled and the sufficiency of
contingency measures. In his letter of 8 J anuary 1999, he prayed in aid
the fact that the facilities and systems that had been handed over to AMD

542
for operation on AOD had not yet been accepted by AA from the
contractors, to show that these matters were the responsibility of PD and
not AMD. While the Commissioners accept this submission, the
unprepared states of the facilities and systems highlight the importance of
having a very careful overall risk assessment and global contingency plan.
W44 Heed admitted that no such assessment had been undertaken, as it
ought to have been, according to W48 Lam. W44 Heed also admitted
that there was no global contingency plan. Had such exercises been
carried out, the state of unreadiness of FIDS, the insufficiency of
contingency measures and lack of co-ordination with other operators like
RHOs might well have been revealed and remedied or at least reduced
before AOD. As a result, when FIDS failed on AOD, vital lines of
communication were either not available or overloaded, the airlines and
RHOs found themselves completely lost without vital flight information,
and chaos ensued.

17.85 In his letter of 8 J anuary 1999, W44 Heed laid great
emphasis on the insufficiency of resources available to AMD. He said:

I point this out to put in perspective the amount of resources
that were available to accomplish the workload as undertaken
and the limitations to take on much more.

He also revealed that when four senior experienced general managers
were taken away from AMD in 1997, which was disruptive to the staff
and impeded AMDs ability to maintain the momentum on the many
initiatives under underway, his expressed concerns led to the assignments
of Mr Howard ENG as Deputy Director and Mr K W TONG as General
Manager, Engineering and Maintenance. If there was truly a resources
problem that caused AMD to be unable to carry out the tasks entrusted to
it effectively, concern should have been raised by W44 Heed with W3
Townsend or with the AA Board for provision of adequate resources, and
if they were not forthcoming, then it would be for W44 Heed to warn
them of the high risks in attempting to operate the new airport on AOD or
even suggest a postponement of AOD. To support this warning, it
would behove W44 Heed to have an overall risk assessment, or to make a
global contingency plan in case the warning was not heeded. W44 Heed
admitted that he had done neither. There is little evidence to show that

543
these two exercises would have demanded too many resources.
Moreover, W50 Wong and W49 Lo told the Commission that they were
never apprised of lack of resources for AMD to carry out its functions.
It is now too late after the events on AOD for W44 Heed to harp on
inadequate resources which should have been boldly mentioned by him to
the CEO or the Board at the right time. The absence of documentary
evidence on urgent and serious requirement of resources for AMD also
goes to show W44 Heeds weak character referred to in paragraph 17.18
above.

17.86 In the end, the Commission finds that W44 Heeds weakness
and deficiencies deprived Hong Kong of the chance of a smoother and
more efficient airport on AOD. Additionally, his failures to discharge
his duties materially contributed to the mayhem and confusion witnessed
on AOD.

17.87 W44 Heeds integrity is doubted in his attitude towards
ADSCOM. This matter is covered in paragraph 17.61 above.

(e) W45 Chatterjee

17.88 IT, headed by W45 Chatterjee, served a supportive role to
PD and AMD. IT was actively involved in the FIDS programme as
from about December 1996 when PD required support on testing and
commissioning of the systems contracts. A task force was set up around
20 December 1996 to support PD in the testing and commissioning of the
systems, including FIDS. This task force reported to W44 Heed and Mr
Raymond LAI (Director, Financial and Commercial). Its role was to
act as AMDs expert technical representatives working with the PD to
ensure that the technical operational aspects of the infrastructure systems
were fully tested.

17.89 The Commission finds that W45 Chatterjee, as Head of IT,
failed in his duties in two respects: (1) not properly assessing the risks
involved in deferring the stress test for the FIDS; and (2) not properly
advising the AMD of the risks involved in not undergoing such test
before AOD.

544
17.90 The lack of proper testing of the FIDS in detail can be found
in Chapter 13. The evidence of W21 Mr Michael Todd Korkowski,
W35 Cumming and Mr Rupert J ohn Edward Wainwright of EDS, W22
Mr Edward George Hobhouse of GEC, W34 Derrick of Preston, W55 Dr
Ulrich Kipper and W56 Professor Vincent Yun SHEN (the Commissions
IT experts) all points to the importance of testing in software
development and commissioning, and emphasis is laid on a stress test
being able to reveal problems. The evidence is also clear that the stress
test for FIDS was deferred because of the lack of time. There were 38
problem reports (PRs) identified in early J une 1998, and the witnesses
from the parties including those from NAPCO seemed all to agree that
the time remaining up to AOD should better be used to rectify the PRs
and that FIDS was not in a stable enough state to be subject to a stress
test.

17.91 It is necessary to decide first whether W45 Chatterjee ought
reasonably to have appreciated the risks involved. From the evidence, it
appears that the risks were clear and significant, yet they were not
recognised by W45 Chatterjee fully or at all. The reasons in support are
as follows:

(a) It is industry practice to carry out stress testing for an
important system like FIDS. The deferment of the stress
test was therefore a major deviation from that practice.
The evidence from various witnesses who were IT
professionals and experts, including W45 Chatterjee
himself, is that a stress tests would probably have revealed
the performance problems of FIDS and the extent of such
problems.

(b) The major reason given by witnesses on the postponement
of the stress test was that FIDS was not stable enough to
undergo a stress test. Hence, the danger of using the
system for Day One operation must have been evident to
someone with W45 Chatterjees IT background. Yet
apparently this risk did not receive the attention it deserved.

(c) Given the less than smooth software development and

545
testing of the FIDS, W45 Chatterjee should have been on
the alert to ensure that the system would be up to scratch
for live use on AOD. In particular, as there were many
open PRs up to AOD with some 38 major outstanding PRs
having been identified in J une 1998, W45 Chatterjee must
have been aware that for AOD, the system was at best
functional on workarounds if at all. Moreover, W45
Chatterjee should have known that not having conducted a
proper stress test due to the lack of time, problems which
would otherwise have been revealed might crop up only
during live operation on AOD. Bearing these two factors
in mind, it is difficult to see how W45 Chatterjee could
have failed to appreciate the very serious risks involved in
going ahead with FIDS on AOD.

(d) If and insofar as W45 Chatterjee did not have the necessary
expertise to provide a proper assessment of the risks
involved, outside expertise, for example CSE International
Ltd (CSE), the systems consultants retained by the AA
management, should have been sought.

17.92 As to the second question, namely, whether W45 Chatterjee
should have advised AMD of the risks involved in not undergoing the
stress test before AOD, the answer must be clear. AMD being the
operator of FIDS on AOD needed a full picture of all relevant factors so
as to be able to come to an informed decision on whether to use FIDS on
AOD, and to plan for the necessary contingency measures in the event of
serious problems impeding operation.

17.93 W45 Chatterjee confirmed in testimony that he did not
advise W44 Heed or even W48 Lam of the difference between a formal
stress test and the fifth airport trial as a test of the loading on AOD, nor
the decision to defer the stress test. Thus neither W44 Heed, as director
of AMD, nor W48 Lam, the Deputy CEO and chairman of AOR Meetings,
was advised of the risks involved.

17.94 No doubt AMD shared in the responsibility in that it should
also have taken steps to find out for itself the relevant information, but

546
W45 Chatterjee as Head of IT and being in charge of the task force to
report to AMD on testing must take the major portion of the blame for not
advising AMD properly. This is another demonstration of lack of
co-ordination within the AA management.

17.95 Counsel for the Commission also submitted that W45
Chatterjee made two misrepresentations, namely,

(a) The misrepresentation contained in ADSCOM paper 34/98
where it was reported that The reliability of the system as
a whole has been 98.7% available

(b) That he reported at the ADSCOM meeting of 4/7/98 that
the standby FIDS had been satisfactorily tested with 35
airlines on 30 J une 1998.

17.96 The first misrepresentation has been dealt with in great detail
under Section 4 of this chapter, and the Commissioners find that W45
Chatterjee was grossly negligent in allowing it to remain unexplained at
the ADSCOM meeting on 4 J uly 1998 (see paragraph 17.53).

17.97 On the other hand, there is no sufficient clear evidence for
the Commissioners to find that the report made by W45 Chatterjee about
the successful or satisfactory test of the standby FIDS on 30 J une 1998
contained untruth when what he told ADSCOM is carefully analysed.
Counsels allegation is mainly based on airlines responses to the
Commission that they did not know standby FIDS was tested at the trial
on 30 J une 1998 in which they took part. The Commissioners consider
that such evidence does not necessarily falsify W45 Chatterjees
statement to ADSCOM. First, the airlines and their handling agents did
take part in a trial held on 30 J une 1998. Secondly, towards the end of
J une, there was an e-mail sent by AA to the airlines or their handling
agents in which it was mentioned that the FALLBACK FIDS would be
used at the trial. Although the airlines and their handling agents might
not have known that the fallback FIDS, which was another term that
could reasonably be used interchangeably with standby FIDS, would be
tested at the trial, they should reasonably appreciate that that standby
system was to be used. Thirdly, the evidence from the AA staff and

547
NAPCO personnel all pointed to the fact that the standby FIDS was in
fact tested at the trial. The participation of the airlines and their
handling agents at the trial which, albeit unbeknown to them, was also for
carrying out a test of the standby FIDS, should not reasonably render
W45 Chatterjees report to ADSCOM of a successful test having been
conducted on 30 J une 1998 as false.

17.98 Admittedly, the report might have given a sense of security
to members of ADSCOM who were relying heavily on the standby FIDS
in the light of the history of instability of the main FIDS. However,
insofar as there is insufficient evidence to qualify W45 Chatterjees report
as a misrepresentation, it would not be fair to condemn him on this score.

17.99 Rather, the lack of full knowledge on the part of the airlines
and their handling agents that standby FIDS would be tested, as opposed
to merely used, at the 30 J une trial, indicates that AMD did not plan the
test well and failed to co-ordinate adequately with the participants. The
test required the participation of the airlines, but AMD failed to let them
know that there was such a test. There was no meeting with the airlines
or any details of the test and its procedure in writing for the airlines
beforehand. This is another illustration of a failing in the AA
management.

(f) The AA Board

17.100 The AA Board has overall responsbility for the problems on
AOD because the duty for developing and operating the new airport,
being part and parcel of the functions of AA and within the care and
management of the affairs of AA, is placed squarely on it by section 4 of
the Airport Authority Ordinance, which provides:

Subject to the provisions of this Ordinance, the affairs of the
Authority shall be under the care and management of a board
whose functions shall comprise such care and management.

17.101 The responsibility to discharge the functions of developing
and operating the new airport remains with the AA Board, although it is
allowed by sections 9 and 15 of the Ordinance to delegate its functions to

548
a CEO and management.

17.102 The Commissioners do not accept counsels submission that
the AA Board should be responsible for W3 Townsends acts and
omissions or the acts and omissions of the AA management, nor that the
Board should be professionally qualified.

17.103 W3 Townsend must be responsible for his own acts and
omissions. He was appointed before the Airport Authority Ordinance
came into force on 1 December 1995 when the present AA Board was
constituted. The AA Board might or might not have realised W3
Townsends deficiencies after the publication of the Booz-Allen &
Hamilton report at the end of October 1997. Even if the Board felt that
W3 Townsend might not be fully up to his job, it would be too risky to
have him replaced at the time, bearing in mind that the new airport was to
open within a matter of six months, the target then being April 1998.
Replacing W3 Townsend required looking for a replacement and settling
him in within a very short time without causing further disruption to
management. Even if a replacement was immediately available, it is
difficult to judge that the disruption to management at that juncture would
not carry enormous risks for the preparation for opening the airport in a
few months time. All these matters are highly speculative, and the
Commissioners cannot come to any reasonable conclusion that replacing
W3 Townsend was a clear alternative open to the Board.

17.104 It is not disputed that the professional aspect of the work of
AA could have been entrusted to a professionally qualified AA Board.
The Board concerned consists of official members who are mainly
high-ranking administrative officers of Government. The non-official
members are mainly community leaders. There are a few professionals
such as the Secretary for Works, but their professional fields did not cover
IT. If full reliance were to be placed on the Board, it would need to be
filled with

(a) professionals and experts in both construction and systems
when the new airport was built; and

(b) professionals in business management and airport

549
management for the operation of the new airport after it
was built.

17.105 However, neither of the aforesaid is usual, and they do not
normally occur in boards of directors of public companies where the
interests of shareholders including members of the public are at stake.
The rationale behind is apparent. A board such as the AA Board and
boards of public companies are more normally constituted of people of
prominence, respectability or capability than necessarily with
professional or expert knowledge. Members of such boards are usually
expected to decide on policies and provide overall steers, insight and
perhaps foresight to the management instead of delving into operational
details. If a board does not consist of such professionals and experts as
are required for the job to be undertaken, as in the present case, it could
and should have retained consultants to advise it on the project, especially
its progress. In the case of the AA Board, however, that would involve
expending public funds. As the AA management had already retained
systems consultants such as CSE, any employment by the Board of
another firm of consultants to oversee the progress of systems
development would be duplicating efforts and resources, and would be
questionable deployment of public funds. It is in these circumstances
and from this perspective that the Commission agrees with W51 Yuens
view that it would be better for the AA Board to have consultants to
advise it instead of the AA management employing experts to advise the
management but not the Board. The benefit would be at least three-fold.
On the one hand, the Board would have expert assistance in
understanding the works on construction and systems and their progress,
so that it would be able to discharge its functions imposed on it by the
Ordinance, and satisfy itself as to the quality and progress of the works.
In case anything untoward was reported by the consultants, the Board
would be able to warn the management of it and instruct the management
to take remedial or improvement measures. At the same time, the Board
would have independent advice from the external consultants, instead of
having no choice but to take the reports from the management on trust.
The consultants reports would be a source of information and advice
additional to that provided by the management, operating as a check and
balance that would be required for such an enormous project as the
development of the new airport. As the consultants advice would be

550
made available to the Board and passed onto the management by the
Board, the management could do away with retaining its own consultants
on the same subject. Hence, there would not be duplication of efforts
and expenses.

17.106 Apart from the eventual responsibility to bear in making the
new airport ready on AOD, as it had assured ADSCOM, the AA Board
should also be responsible for not having appointed outside consultants to
advise itself, instead of allowing the AA management to have such
consultants. However, this view may be derived from the wisdom of
hindsight, which might not have been clear to the AA Board at the
material time. The AA management consisted of various kinds of
professionals, and it had the assistance from outside experts on systems
development and progress. The AA Board might not have felt the
benefit of having external consultants to advise it on the same subjects,
which benefit would be more readily appreciated with hindsight. The
Commissioners therefore consider that this failure of the AA Board
should not be over-stated.

551

552
CHAPTER 18

CONCLUSIONS

Section 1 : The Decision to Open the Airport

Section 2 : Extent of Readiness and the Problems

Section 3 : Causes of the Problems and Responsibility


Section 5 : Responsibility of AA

Section 6 : The Present Situation

Section 7 : Could the Chaos and Confusion have been Avoided?

Section 8 : Lessons Learned

Section 1 : The Decision to Open the Airport

18.1 The decision to open the airport on 6 J uly 1998 was taken by
the Airport Development Steering Committee (ADSCOM) in J anuary
1998. The target date for the opening of the new airport was originally
scheduled for April 1998. It was always understood that as a target date,
it would require confirmation by a formal announcement nearer the time
by Government in conjunction with the Airport Authority (AA), in the
light of the overall airport readiness achieved and the prospect of the
Airport Railway (AR), later known as Airport Express (AE), being
ready ahead of time. AR had a completion date of 21 J une 1998 but
there was expectation that progress on AR could probably be accelerated
to support airport opening in April 1998.

553

18.2 In AAs franchises with its franchisees, AA was obliged to
give a three-month advance notice to the franchisees of the date of
opening of the new airport. ADSCOM was mindful of the importance
to fix an airport opening date well in advance so that the public as well as
all concerned parties would know this date for their own purposes and
planning. It was therefore necessary for Government to take a decision
on a firm airport opening date at least three months ahead of April 1998.

18.3 AAs optimism that the airport would be ready for opening in
April 1998 was not shared by ADSCOM for mainly two reasons: (a) in
respect of the Passenger Terminal Building (PTB), the works
programmes had slipped and the plan for systems training was tight; and
(b) in respect of the Hong Kong Air Cargo Terminals Ltd (HACTL),
there were delays in the construction works of SuperTerminal 1 (ST1).
In October 1997, Mass Transit Railway Corporation (MTRC) made a
detailed presentation to ADSCOM explaining why it was not able to
advance the completion of AR from J une to April 1998. Given the
doubts about the adequacy of the transport arrangements pending
completion of AR and the state of readiness of airport systems and
HACTL, ADSCOM decided in early J anuary 1998 that airport opening
should be deferred, with the aim of producing on airport opening day
(AOD) a world class airport supported by efficient transport facilities.
1 J uly 1998 was decided to be the date for the airport ceremonial opening
to allow more time for AR to get ready and for public relations reasons
(since it coincides with the first anniversary of the Hong Kong Special
Administrative Region). In addition, ADSCOM accepted Monday, 6
J uly 1998 to be the date for the operational opening of the airport because
a few days would be needed between the airport ceremonial and
operational opening for the critical phase of the airport relocation exercise.
Opening the new airport on a Monday would offer the advantage of the
night move taking place when road traffic was light and when a big
spectator turnout would be unlikely. Air traffic was also lighter on a
Monday.

18.4 The Chief Secretary then explained to the Chief Executive in
Council why ADSCOM had decided to defer the opening date to J uly
1998. He endorsed the decision and agreed that the Executive Council

554
(ExCo) should be informed of ADSCOMs recommendation. The 1
J uly date was eventually altered to 2 J uly 1998 for the ceremonial
opening, obviously in order to prevent a clash of the airport opening
ceremony with the activities anticipated for commemoration of the first
anniversary of Hong Kongs reunification with the Mainland, while the
day for the operational opening was unaltered to be on 6 J uly 1998.

18.5 The Commissioners find that ADSCOM acted cautiously and
wisely in deciding 6 J uly 1998 as the operational opening day for the new
airport. ADSCOM had considered all relevant matters very carefully
and diligently after being provided with the necessary information by its
executive arm, the New Airport Projects Co-ordination Office
(NAPCO), and AA. NAPCO was acting as an overall monitor over
AOR. AA furnished reports and information on the progress of the
development of the new airport to NAPCO. NAPCO critically
examined such material and advised ADSCOM accordingly.

18.6 Having examined all the evidence very carefully, the
Commissioners find it clear that the Chief Executive in Council was not
involved in any way in the decision making of the opening of the airport,
although he approved that decision. The decision was taken by
ADSCOM which was then reported to him by the Chief Secretary and
also reported to the ExCo at its meeting on 13 J anuary 1998.

18.7 In deciding on the operational readiness of the new airport,
the policy consistently adhered to by ADSCOM was to have the new
airport operating safely, securely, efficiently and smoothly. The
Commissioners find that ADSCOM did not make any mistake in deciding
that 6 J uly 1998 should be the date for the operational opening of the new
airport. Indeed, ADSCOM members had exercised great care and
diligence in reaching that decision. The main reason for ADSCOMs
selecting J uly 1998 was to await the completion of AR, and that was
despite AAs insistence that all critical AOR items would be ready by late
April 1998. The added time of over two months between April and J uly
would moreover provide a comfortable float to PTB and HACTL projects.
The Commissioners conclude that it was a proper and wise decision.
There was no evidence whatsoever to suggest that the decision to open
the airport in J uly 1998 was a result of any political consideration or

555
ulterior motive. During the period between J anuary 1998 after the
decision was made up till AOD, ADSCOM exerted no less effort and care
regarding the progress of AOR issues. Numerous reports were required
to be supplied by AA, numerous reports were made by NAPCO and many
meetings were held by ADSCOM when very often AAs top management
was invited to attend to explain various matters. The continuous
assurance given by AA and HACTL that PTB and ST1 respectively
would be ready had lulled ADSCOM members into a false sense of
confidence and security, resulting in their not revisiting the opening date.
Indeed, once decided, AOD should not be changed lightly, for it was a
decision creating the certainty on which many people relied.
Nonetheless, if sufficiently weighty material was proffered, the
Commission has no doubt that ADSCOM would certainly consider
whether a deferment was necessary. No one ever suggested a deferment
or put situations before ADSCOM that would, at the time, justify a revisit
of the decision. All concerned were taken by surprise by the chaotic
situations that occurred on AOD. The Commissioners hence feel that it
would be unreasonable to hold ADSCOM or any of its members
responsible for not appreciating the risks of keeping AOD in the then
prevailing circumstances.

Section 2 : Extent of Readiness and the Problems

18.8 When AOD was considered and eventually decided, there
had been delays in the construction works and systems works relating to
PTB. The construction works on HACTLs ST1 also suffered slippages.
The additional time between the original target opening in April 1998 and
6 J uly 1998 was a cushion to ensure both PTB and ST1 would be ready.
When eventually occupation permit was issued for PTB on 29 J une 1998,
temporary occupation permit (TOP) was obtained by ST1 on 3 J uly
1998, and aerodrome licence was issued for the new airport on 1 J uly
1998, everything seemed to be ready that would provide Hong Kong with
a safe, secure, efficient and smooth airport. No evidence has been
received by the Commission that raises concern about the safety and
security of the new airport on AOD.

18.9 The problems occurred on AOD related to efficiency. Two

556
of the critical AOR issues on which everybody concerned focussed were
the readiness of the Flight Information Display System (FIDS) and ST1.
FIDS was considered to be critical for the operation of the new airport,
for the flight-related information to be provided by it was essential for
airport operations. ST1s readiness was important because HACTL
would be required to handle about 80% of all of Hong Kongs air cargo at
the new airport. The deficiency of FIDS and the paralysis of ST1 were
the major problems encountered on AOD that rendered movements of air
passengers and air cargo inefficient.

18.10 There were also many other problems on AOD, mostly
consequential upon the deficiency of FIDS which failed to provide
prompt and correct flight-related information to various operators of the
new airport. There were the baggage handling problems, where
handling operators (RHOs) were unable to have stand allocation
information and flight times readily available through FIDS. These
baggage problems delayed baggage reclaim by passengers and flight
departures. Other problems that resulted from the deficiency of FIDS
included late arrival of tarmac buses, insufficient ramp handling services,
aircraft parking confusion, delay in flight arrival and departure, etc,
which were most noticeable by the users of the airport.

18.11 Problems that were not caused by the deficiency of FIDS
also occurred. They either arose individually, or they were a
consequence of other problems. The Commission has classified all the
problems that occurred since AOD into three categories: teething or
minor, moderate and major. The categorisation was made in accordance
with the opinion of experts appointed by the Commission and the
Commissioners own views as to the seriousness or otherwise of the
nature of each problem. They are set out below:

Teething or Minor Problems:


557
[8] Automated People Mover (APM) stoppages
[9] Airport Express (AE) ticketing machine malfunctioning
[10] AE delays
1998
[22] Traffic accident on 28 August 1998 involving a fire engine,
resulting in five firemen being injured
[23] A maintenance worker of Hong Kong Aircraft Engineering
Company Limited (HAECO) slipped on the stairs inside the
cabin of a Cathay Pacific Airways Limited (Cathay Pacific)
aircraft on 3 September 1998
flights
on 1 October 1998

Moderate Problems:

[27] Malfunctioning of the Access Control System (ACS)
[31] Public Address System (PA) malfunctioning

558
[33] Radio frequency interference (RFI) on air traffic control
frequency
[34] Aircraft Parking Aid (APA) malfunctioning: a Cathay
Pacific aircraft was damaged when hitting a passenger jetway
during parking on 15 J uly 1998
[35] An arriving passenger suffering from heart attack not being
sent to hospital expeditiously on 11 August 1998
[37] A Hong Kong Airport Services Ltd. (HAS) tractor
crashed into a light goods vehicle, injuring five persons on 6
September 1998

Major Problems:

[40] FIDS malfunctioning
[41] Cargo Handling System (CHS) malfunctioning
[42] Baggage handling chaos

18.12 About 30 of the 42 listed problems occurred on AOD.
Although the three major problems caused the greatest adverse effect on
the operating of the new airport on AOD and for a period thereafter, all
the other 27 problems occurred on AOD. Relating to airport operational
efficiency, each of most of these 27 problems would not have raised
concern or even been noticeable by itself . It was the concatenation of
all these problems that created the chaos on AOD. In anyones standard,
the new airport was not ready to open on AOD.

Section 3 : Causes of the Problems and Responsibility

18.13 The Commission is tasked to find out the causes for the
problems with the new airport since AOD and where the responsibility

559
lies. In such an attempt, the Commission wrote numerous inquiry letters
to persons or parties who might be concerned with each of the problems,
and held an inquiry to hear preliminary matters, evidence and
submissions by parties for 61 days. Within the time allowed by its terms
of reference, the Commission is only able to find out the causes of and
responsibility for most of the problems, but not all. A summary of the
causes and responsibilities are set out in the following paragraphs.

Teething or Minor Problems

[1] Mobile Phone Service Not Satisfactory

[2] TMR Service Not Satisfactory

[3] Public Telephones Not Working

18.14 On AOD, only about one third of the public telephones
planned for airport opening were operational at PTB and the other two
systems of communication experienced different degrees of capacity
overloading problems. TMR users also encountered reception and
coverage problems at various places in the new airport. Both the
malfunctioning of FIDS and the presence of a large number of curiosity
visitors and stranded passengers during the first few days after airport
opening are identified as factors contributing to unforeseen huge demand
on the use of the three communication systems around that time.

18.15 The Commissioners consider that, as the concatenation of
the many problems occurring on AOD that might increase the demand for
mobile phone service was not properly foreseeable, it would not be fair to
hold any of the operators responsible for the capacity overloading
problems. On the other hand, AA should, however, be responsible for
not giving advance warning to the operators of the possibility of heavy
demand on the use of mobile phones in the event of FIDS failure.

18.16 As regards the use of TMR, the Commissioners hold the
view that Hutchison Telecommunications (Hong Kong) Ltd (Hutchison)
being one of the TMR operators should have foreseen the problem of
weak signal transmitted from its base station located at Tung Chung and,

560
therefore, put in place adequate counter measures to overcome the
problem prior to AOD. Moreover, Hutchison should be held
accountable for its failure to provide an operational link between its base
station and the TMR Distributed Antenna Network for PTB. As for AA,
the Commissioners find that it is the sole party responsible for the delay
in completing the outdoor antenna farm for use by TMR operators. The
delay inevitably had an impact on the operational efficiency of the TMR
system for use on AOD. Also, since the use of TMR was part of the
contingency or workaround measures for FIDS, AA should have
forewarned the TMR operators as well as other airport operators of the
possible heavy demand for service in the event of FIDS failure.

18.17 As to the subject of public telephones, both AA and
International Computers Limited (ICL) accepted that there was delay in
the completion of the cabling and jumpering work which resulted in more
than 60% of the planned telephones being not ready for service on AOD.
However, without hearing all the witnesses from AA and ICL on the
issue, it would not be possible for the Commissioners to attribute
responsibility between AA and ICL. In any case, AA should be held
responsible for failing its duty in coordinating and overseeing the cabling
work and ensuring that prompt remedial action was taken when the delay
and the effect of which was reasonably foreseen. As a matter of fact,
AA did accept the responsibility for the cabling and jumpering problems
encountered by New World Telephone Limited (NWT) as the
contractor for the supply and installation of the public telephones at PTB.
For those telephones that worked on AOD, there were other operational
problems, such as coin acceptance difficulties and, for these problems,
NWT did not deny responsibility. The Commissioners note that, by mid
J uly 1998, almost all public telephones were in operation.

18.18 In any event, the problems that plagued the three
communication systems were short-lived and were rectified very quickly
after AOD.

[4] Escalators Breaking Down Repeatedly

18.19 On AOD, in respect of the 59 escalators in operation, there
were 20 incidents of stoppage on that day and 19 such incidents on the

561
following day. While there are several contributing factors, the
stoppages were mainly caused by the protective devices of escalators
being set at too sensitive level, with the result that even slightly heavier
loads would trigger a stop. This problem was remedied promptly after
AOD by adjusting the safety devices to the appropriate levels so as to
match the actual working conditions and passenger load. As a result of
the adjustment, the stoppage rate dropped to around 0.2 per escalator each
month which is considered to be normal. There were also incidents in
which stoppages were caused by foreign objects jamming the steps or
people pushing the emergency stop button for unknown reasons. These
are, however, normal occurrences at airports or in public buildings.

18.20 Constructions Industrielles De La Mediterranee SA
(CNIM) is responsible for the first year maintenance of escalators and
AA is responsible for their operation. Although the actual live load
requirements of the escalators could not have been precisely foreseen, the
Commissioners consider that, had sufficient tests been carried out before
AOD, the sensitivity level of the protective devices could have been set
properly. For this, both AA and CNIM should be responsible. The
Commissioners further note that the unavailability of the automatic
control and monitoring systems for maintenance services prevented staff
of AA from responding quickly to breakdown of escalators. These
systems were considered to be non-AOR critical and were not completed
before AOD, apparently because of a lack of time. In any case, this
seems to be a teething problem which was cured very easily and quickly
after AOD. Operation of escalators in fact stabilised since the first week
of airport opening.

[5] Insufficient or Ineffective Signage

18.21 Insufficient and ineffective signage has been cited as one of
the problems plaguing the new airport during its initial period of
operation. AA acknowledged that among more than 1,500 directional
signs within PTB, a sign with single arrow within the Meeters and
Greeters Hall pointed in the wrong direction and that misdirected sign
was corrected in one day. Also, on AOD, an unanticipated number of
passengers and visitors used the external buses instead of Airbuses and
crowded at Cheong Tat Road which led them to Level 3 (ground level).

562
As a result, passengers starting at Level 3 without luggage and visitors
were diverted to the Departures Hall through the Arrivals Hall. This
caused some confusion under the one-way flow signage system as these
passengers presumably saw signs intended for arriving passengers rather
than for departing passengers. Having reviewed the evidence, the
Commissioners agree that members of the public visiting PTB for the
first time will necessarily go through a period of familiarisation with the
new environment. Complaints about inadequate signage do not seem to
be borne out by factual evidence. In the light of evidence, conflict of
allegations between the Board of Airline Representatives and AA cannot
be resolved. In any case, airport operation readiness does not
reasonably include signs for airline offices. Even if there was a problem,
it was but a teething problem that was quickly remedied by means of the
additional signs installed in J uly and August 1998. The Commissioners
accept the expert advice of W51 Mr J ason G YUEN, an expert appointed
by the Commission, that signage additions, revisions and refinement is
quite common among major airports after the terminal has been put to
actual use. In this regard, the Commissioners do not intend to attach any
blame to any party.

[6] Slippery and Reflective Floor

18.22 There were criticisms about the black granite floors which
were allegedly both slippery and very reflective causing potential
embarrassment to female airport users wearing skirts. According to
evidence, a total of five incidents of people slipping on floors in the
public areas of PTB were recorded between AOD and 31 August 1998.
However, it is noted that none of the incidents occurred on the black
granite floors and, also, wet floor was a contributing cause in two of the
incidents.

18.23 The problem of slippery and reflective floors came up during
the first airport trial held on 18 J anuary 1998. As a result of feedback
from the trial participants, AA carried out remedial actions to raise the
slip resistance of the polished surfaces by means of honing. The task
however proved to be extremely time consuming. After research, AA
decided to carry out non-slip surface treatment to all black granite
surfaces but the whole operation could not be completed before AOD.

563
All treated floor surfaces meet the standard of the American Society of
Testing and Materials for use by disabled persons.

18.24 Having reviewed the history leading to the problem, the
Commissioners do not consider that it is part of the usual teething
difficulties since it is something that was identified and anticipated at an
early stage. AA should be blamed for its failure to take prompt and
speedy remedial action to eradicate the problem prior to AOD. The
overall problem is only very minor in nature as the rate of incidents of
people slipping on floors does not appear to be out of the ordinary in view
of the approximately six million people using PTB during the same
period. However, should AA have tackled the matter more promptly, the
problem would not have been allowed to develop into an issue on airport
opening.

[7] Problems with Cleanliness and Refuse Collection

18.25 There were problems of cleanliness and refuse build-up in
some parts of PTB immediately before AOD and shortly thereafter.
While there were various causes for the problem, it was evident that both
AA and its various cleaning contractors tried their best endeavors to
overcome the problem. Unfortunately, because of the enormous amount
of waste, from both PTB tenants and the large number of sightseers and
stranded air passengers, AA and their cleaning contractors were simply
unable to cope with the required work within a short time.

18.26 From the evidence available, the following factors are
possible causes contributing to the problem:

(a) PTB tenants were late in taking up their premises and, hence,
completing their fitting-out works. As a result, their
relocation exercises started later than anticipated and this
eventually led to large volumes of construction refuse to be
removed within a short time. What made the situation even
worse was that some tenants failed to comply with the
proper disposal procedures and dumped their rubbish in the
surrounding premises.

564
(b) There were a number of design or equipment related
deficiencies impacting on the operational efficiency of the
disposal system. For instance, the design of the refuse
room is inadequate to cope with the demand in some areas.
Also, refuse chutes between Level 5 and Level 3 are not
continuous, thus requiring waste to be pushed along a
walkway on Level 4. Another problem is that some refuse
rooms and refuse compactor stations were not ready for use
on AOD.

(c) Much longer time was needed by Aviation Security
Company Limited (AVSECO) to issue security permits for
both workers and vehicles of cleaning contractors and this
severely affected the latters ability to deploy adequate
resources to work within the restricted areas.

(d) The coordination between AA and its cleaning contractors
was insufficient resulting in the failure to provide adequate
cleaning service. In one incident, one of the contractors,
Los Airport Cleaning Services Limited (Los), failed to
undertake an order instructed by AA due to difficulties in
communication. There was also an allegation from AA
claiming that the contractors stuck rigidly to their respective
boundaries of work.

(e) The presence of a large number of curiosity visitors and
stranded passengers at the new airport shortly after AOD
undoubtedly aggravated the problem of rubbish build-up.

18.27 The problem lasted only a few days after AOD. By 10 J uly
1998, all rubbish was substantially cleared and there is now sufficient
manpower inside the restricted areas to carry out cleaning services. On
the issue of responsibility, the Commissioners find reasons to believe that
both the PTB tenants and AA as the management authority should be held
responsible for not ensuring timely completion of fitting-out works and
proper removal of the fit-out debris. As to the design deficiency, AA
together with The Mott Consortium (Mott) as the design contractor
may both be accountable for the resulting difficulties. The Commission

565
does not have sufficient evidence to apportion responsibility in relation to
the late availability of security permits for the cleaning workers and
vehicles. While the Commissioners are not prepared to lay any blame
on the cleaning contractors in view of their efforts made in tackling the
problem, AA should be criticised for the overall failure in the provision of
adequate cleaning service.

[8] APM Stoppages

18.28 There were stoppage problems with the operation of APM
during its initial period of operation. While most of the incidents were
related to door-related problems, passengers were trapped inside an APM
train and unable to leave for about 50 minutes in the incident that
occurred on 20 J uly 1998.

18.29 The investigation conducted by Mitsubishi Heavy Industries,
Ltd. (MHI), the operation and maintenance contractor for APM,
revealed that most stoppages were caused by vehicle door failure,
platform door failure or train overshooting. There were incidents in
which passengers forced a door open and this effectively disrupted the
closing movement of train doors, causing the train in question to stop.
The friction of door equipment with surrounding mechanical parts and the
failure of local door control circuit were also identified as the causes
leading to train stoppages. A number of remedial measures had been
taken to tackle the problems. Measures will continue to be undertaken
to maximise the vehicle stopping accuracy. Following the series of
incidents, AA also took steps to provide station attendants at each of the
four APM platforms to assist in passenger control.

18.30 In the incident on 20 J uly 1998, one passenger accompanied
by four airline staff members were trapped inside a train and were unable
to leave until 50 minutes later. Before the APM maintenance staff
arrived to restore the train, the group of passengers attempted to pry open
the door by turning the emergency door release valve and eventually got
onto the emergency walkway. For safety reasons, the APM operator
immediately shut down the traction power in the tunnel and the five
persons were eventually escorted to the West Hall departures station.

566
18.31 The above problems identified with the APM system
resulted in only slight disruption of train service and, consequently, some
degree of passenger inconvenience. For most of the occasional stoppage
incidents and, in particular, the trapping incident on 20 J uly 1998,
passengers should primarily be blamed for their improper behavior in
forcing themselves through closing doors or attempting to pry open doors.
While the door-related failures may be regarded as part of the start-up
difficulties that will disappear after fine tuning, the Commissioners hold
the view that, should more thorough and proper modification works have
been done by the contractor prior to commissioning of the system, the
frequency of occurrences could have been minimised. Also, the
Commissioners consider that AA should be responsible for its failure to
ascertain correctly the actual operational needs and to put in place from
AOD sufficient attendants to attend to train problems and for keeping of
the order of passengers at the platforms. Furthermore, the
Commissioners are concerned with the apparent lack of an effective
communication means between Airport Operations Control Centre
(AOCC) and the APM maintenance staff while attending to
emergencies. As revealed by the incident on 20 J uly 1998, the
maintenance staff did not have access to the use of the TMR system of
AA. It was possible that the rescue action in the incident could have
been much quicker if the maintenance personnel were provided with
radios for communication with AOCC.

[9] Airport Express (AE) Ticketing Machine Malfunctioning

[10] AE Delays

18.32 When AE went into operation on AOD, the coin
management system on all ticketing machines was not in service due to
some software problems and, as a result, they would accept notes only.
The problem had in fact been identified in loading tests carried out prior
to AOD and, to cope with actual needs upon commissioning of AE,
MTRC put in place a series of counter-measures to facilitate passengers.
The problem did not last long and, by 14 J uly 1998, the software problem
was completely solved and all ticketing machines have been working
properly since 24 J uly 1998.

567
18.33 As to the issue of disruption of service, MTRC decided prior
to AOD that AE should open for passenger operations on 6 J uly 1998 at a
service interval less than the design capacity for full operation and with
the journey time longer than the scheduled time of 23 minutes. This was
because of the highly complex nature of integration of the many systems
involved and the need to regulate both the Tung Chung Line service and
the AE service that operated on the same pair of tracks for the most part
of the length of the railway. As a result, the AE service would be run at
12-minute intervals when it went into operation. According to records
available, there were minor train service disruptions on 9, 11, 14, 23 and
27 J uly 1998 and in some incidents passengers were transferred from one
train to another, the most serious incident was that on 23 J uly 1998.

18.34 The Commissioners accept that the problem relating to
ticketing machines is only minor, particularly in the light of the effective
counter-measures put in place by MTRC from AOD. Nonetheless, the
fact remains that there were coin handling problems with the machines
and, for this, MTRC is responsible for its failure to ensure that problem
free machines were available for use on commissioning of AE. As
regards service disruptions, MTRC is also responsible although it is
accepted that these were start-up problems and have not recurred since
the end of J uly 1998. Since October 1998, AE has operated at the
original performance specification of 8-minute service intervals with a
journey time of 23 minutes.

[11] Late Arrival of Tarmac Buses

18.35 HAS is the sole franchisee for the provision of airside bus
service, commonly known as tarmac buses, for the transportation of
passengers and airside staff between PTB and remote stands. On AOD
and the following day, there was significant delay in the disembarkation
of arriving passengers, both at the frontal stands of PTB and at remote
stands. In some incidents, the delay lasted up to two hours.

18.36 It has been revealed that while the delay at the frontal stands
of PTB docking bays were primarily caused by problems pertaining to
airbridges, the delay in the disembarkation of arriving passengers at the
remote stands was due to a combination of factors. In essence, the

568
breakdown of FIDS and the overloading problems of TMR and mobile
phone networks were all contributing causes impacting on the efficiency
of the operation of tarmac buses. Also, there was a greater utilisation of
remote stands for parking of aircraft due to serious flight delays and this
put more pressure on the demand for tarmac buses. Furthermore, flight
delays and a full apron on occasions created difficulties in coordination
between boarding gate assignment and the location of aircraft. This in
turn resulted in increased travelling time due to the longer distance
between PTB and some remote stands. Another relevant contributing
factor is that, due to the insufficient number of security cards made
available by AA, arriving passengers and airline staff could not gain
admittance to PTB on some occasions resulting in busdrivers having to
act as doormen. Since AOD, a number of remedial measures had been
taken to improve the operational efficiency of the tarmac bus service and,
by 13 August 1998, bussing operation was able to achieve service
standards in over 90% of the assignments.

18.37 In the light of evidence, the Commissioners accept that the
main cause for the inefficient and late tarmac bus service was the
deficiency of FIDS resulting in the lack of accurate and prompt flight
information to HAS. The problems with airbridges, TMR and mobile
phone systems compounded the difficulties and, as a result, HAS
manpower was strained for locating arriving aircraft. AA as the terminal
management must be responsible for all these factors in causation. On
the other hand, the Commissioners find that there was only one bus
available in reserve, instead of three as agreed between AA and HAS.
Without going into the contractual liability between the two parties
concerning planning of resources, it is clear that had two more buses been
made available as reserve on the first two days, the added 10% of
resources would have helped alleviate the situation. In this regard, the
responsibility must be attributed to HAS.

[12] Aircraft Parking Confusion

18.38 On 6 and 7 J uly 1998, aircraft stand allocation had to be
performed by staff of the Apron Control Centre (ACC) manually due to
the problems with SAS and TMS. Problems of FIDS and TMS around
that time also hampered the ability of ACC to perform timely allocation

569
of parking locations for departing and arriving flights. Furthermore,
extended stay of aircraft due to flight delays eroded parking capacity and
made the allocation task more difficult. The problem was compounded
by other airport problems such as the malfunctioning of some airbridges,
failure of some ACS doors, communication difficulties encountered by
operational staff, insufficient towing tractors due to the amount of aircraft
repositioning required, non-familiarity of push-back procedures by some
tractor drivers, and the unfamiliarity of pilots with the apron, taxiways
and remote stands.

18.39 The Commissioners find that aircraft parking confusion is
basically a consequential problem resulting directly from problems
relating to FIDS and the operation of ACC. With FIDS, together with
stand allocation, now back in proper operation, and as a result of the
improvement measures to passenger, baggage and ramp handling services,
significant improvements have been achieved in terms of aircraft parking.

[13] Insufficient Ramp Handling Services

18.40 The cause of the delay in providing mobile steps for
passengers at the remote stands was similar to that in the provision of
tarmac bus service described in item [11] above, although all the three
RHOs, instead of HAS alone, were involved in serving passengers.
With the exception of too few tarmac buses, which relates solely to the
issue of late arrival of tarmac buses, the conclusions of the Commission
as to the causative problems including the deficiency of FIDS, and
responsibility are identical as those relating to tarmac buses. Following
improvements in the performance of FIDS and TMR and in the operation
of airbridges, ramp passenger services have greatly improved. As to the
servicing of passengers disembarking through airbridges, the relevant
problems and causes are summarised under item [14] below.

[14] Airbridges Malfunctioning

18.41 On AOD, four of 74 airbridges were out of service for one to
two and a half hours. From AOD to Day Five, there were 19, 30, 30, 30
and 34 faults calls respectively and, up to the end of J uly, there were in
total 576 fault calls. Many of the faults related to auto-leveller failure

570
alarms and there were also problems in the extension and retraction of the
airbridges to and from the aircraft. To deal with the problems, two
airbridge teams were formed on Day Three by AA and PT. Bukaka
Teknik Utama-RAMP J oint Venture (Bukaka Ramp) to restore service
promptly and, usually, service was restored in no more than five minutes.
The unusually high number of auto-leveller failure alarms was caused by
a programming error in the software for controlling airbridges. The
error was identified on 11 J uly 1998 and solved on the following day.
Refresher training was also provided to RHO staff and, since then, the
problem with airbridges has not recurred.

18.42 The Commissioners find that Bukaka Ramp as the
installation contractor should be responsible for the programming error
that caused the auto-leveller alarms. Also, had there been more varied
or extensive testing or trials of the equipment, the problem might have
been detected and rectified prior to AOD. For this, Bukaka Ramp and
AA or one of them should be responsible. On the other hand, the
Commissioners do not accept that the problems experienced with the
airbridges could be attributed to operators errors since all the operators
had been certified by AA before they were allowed to operate the
equipment single-handedly. In this regard, no blame should be attached
to the operators.

[15] No Tap Water in Toilet Rooms and Tenant Areas

[16] No Flushing Water in Toilets

18.43 The AEH J oint Venture (AEH) is the contractor employed
by AA in respect of the installation of the systems which provide flushing
and potable water to toilets in the public areas and valved connections to
the boundary of the tenant areas in PTB. The supply, installation,
testing and commissioning of the related electrical and hydraulic works
were carried out by its subcontractor, Rotary (International) Limited
(Rotary).

18.44 On AOD and the few days thereafter, there were problems
with the supply of flushing and tap water in certain areas of PTB. The
primary causes of the problem were basically related to difficulties with

571
the functioning of Tank Rooms 2, 3 and 8. In the morning of 7 J uly
1998, flooding occurred in Tank Room 2 and caused the control panel
that operated the pumps to be switched off. This resulted in the
suspension of water supply from the tank room. Staff of both AEH and
Rotary attended to the problem and deployed temporary pumps to pump
dry the area. Water pumping from Tank Room 2 was resumed at about
7:45 am on 8 J uly 1998. The flooding was later found to have been
caused by blockage in the pipe work for which Nishimatsu Construction
Co., Ltd. (Nishimatsu) was responsible. On 18 J uly 1998, a section of
the pipe was found to have been broken. Remedial work was
subsequently carried out and the pipe was reinstated on 15 August 1998.

18.45 As to the Tank Rooms 3 and 8, it was known immediately
prior to AOD that the valves which regulated water flow into the water
tanks were not functioning properly. To ensure an adequate level of
water, they had to be manually operated by Rotary on a 24-hour basis.
However, staff of Rotary were denied access to the tank rooms on AOD.
As a result, no one was there to operate the tank rooms and the water in
the tanks ran dry. Water supply was not restored until the morning of 7
J uly 1998 when Rotarys staff were allowed access to the tank rooms.
The tank rooms were under manual operation as late as mid-September
and there has not been any further interruption of water supply.

18.46 The Commissioners note from evidence that there was
contradictory evidence regarding the causes of difficulty of Rotarys staff
in gaining access to the tank rooms. Irrespective of who should be
responsible for the causes, it appears that there was a lack of coordination
between AA, AEH and Rotary to ensure that Rotary would be allowed
access on AOD. For this, Airport Management Division (AMD) of
AA should be responsible as the manager for the new airport. As to the
flooding incident, there was evidence to show that the problem of
flooding was foreseeable since there had been flooding of tank rooms
since late May 1998. AA did admit that they foresaw the flooding
problem and had instructed the British-Chinese-J apanese J oint Venture
(BCJ ), being the main contractor responsible for construction of pipes
underneath the tank rooms, to remove any blockages in the pipe to Tank
Room 2. AA, however, did not ask Nishimatsu to deal with the problem
of flooding and, despite the steps taken, further flooding continued to

572
occur in the later part of J une and on 5 J uly 1998. Even so, AA did not
arrange with Rotary to install a pump to prevent further flooding or take
other preventive measures before AOD. In this regard, the
Commissioners conclude that there was a lack of coordination amongst
AA, BCJ , AEH, Nishimatsu and Rotary, for which AA as manager of the
new airport should be primarily responsible.

[17] Urinal Flushing Problems

18.47 From the evidence available, there are four problems
identified with the urinals in the new airport, namely (a) difficulties in
controlling the flushing water flow; (b) maladjustment of the infrared
sensors which activate the flushing valves; (c) blockage of urinals caused
by rubbish; and (d) cleanliness of toilets. The conclusions of the
Commission in relation to the causes of and responsibility for the
problems are summarised below:

(a) The desired flow rate of flushing water through the flushing
valves should be sufficiently high to self clean the valve of
seawater sediment whilst at the same time not causing
splashing. However, the poor quality of seawater and a low
flow rate caused the build-up of sediment in the flushing
valves of urinals. This problem was identified in early
1998 but it was only until mid-J uly 1998 when AA
eventually accepted the recommendation of the
subcontractor, Rotary, to install hoods and an amended
piston within the valves. The Commissioners consider that
AEH being the contractor should be responsible for failing
to provide a satisfactory flushing system for the urinals and
to install weirs to stop sand and dirt from getting into the
water pipes. AA should bear some responsibility in not
taking prompt and sufficient remedial actions to prevent or
alleviate the flushing problem. The Commissioners also
accept W54 Professor Xiren CAOs expert view that there
were some design problems, for which AA may be held
responsible.

(b) Not all the infrared sensors had been correctly set to detect a

573
person standing at normal usage distance from urinals.
Also, some users mistakenly pressed the sensor cover plates,
believing this to be a flushing button and this either damaged
the sensors or affected their setting. To avoid the
misconception, a label reading Do Not Push was affixed to
each sensor cover plate. Replacement of the damaged
sensors was effected by the end of August 1998 and they
were fitted with more substantial fixtures to prevent
interference and damage. There is conflicting evidence as
to which party is responsible for the correct setting of the
sensors, and no firm conclusion can be reached. As to the
damage to sensors, the Commissioners note the view of W51
Yuen that public misuse is a normal occurrence in a busy
airport, and consider that, in view of the large number of
visitors in the early days of airport opening, no one should
be blamed for the problem.

(c) Blockages in drains were caused by users disposing of
rubbish into urinals and the problem was exacerbated by the
huge number of visitors present at the airport. It was
alleged that the plastic waste strainers in urinals were not
fixed and this allowed rubbish to get into the system thereby
causing blockages. The Commissioners note in this context
that regular attendance by cleaners was required to clear
rubbish in the urinals and therefore prevent blockages in
urinals. Los as the cleaning contractor is responsible for
keeping the toilets clean.

(d) While a number of matters such as staff training and
supervision issues, the flushing problems, lack of both
flushing and potable water and urinal blockages were cited
as contributing causes, the problem of cleanliness might well
be attributed to the sheer number of curiosity visitors and
stranded passengers during those days. In any event, the
Commissioners are of the view that the toilets were not
sufficiently clean simply because of the shortage of
manpower but are unable to decide on the responsibility in
the light of the evidence received . The shortage of

574
manpower might have been caused by the difficulties
encountered by Los in obtaining permits from AVSECO for
its staff to enter restricted areas but the evidence received
does not allow the Commissioners to reach a fair conclusion
as to the responsibility. However, this matter relates
principally to coordination and operation amongst
contractors working within PTB and AA as their employer
and manager of the new airport should primarily be
responsible.

18.48 All the necessary rectification work by contractors were
completed by mid-October 1998 and, since then, substantial improvement
has been achieved since then.

[18] Toilets Too Small

18.49 There were criticisms about the size of the toilets in PTB
which was allegedly too small causing inconvenience to airport users.
In particular, air passengers could not get their baggage trolleys into
toilets. The Commissioners find that the design of toilets at the new
airport was based on the planning guidelines of the British Airport
Authority (BAA) and that the actual provisions exceeded the BAA
requirements for some facilities. Given the sheer size of PTB, AA
adopted an approach that was different to that in Kai Tak to provide
strategically a large number of smaller toilets, so as to enable passengers
to locate them easily. As to the accessibility of trolleys, it was a
deliberate decision of AA not to allow trolleys into toilets having regard
to the travelling habits of passengers. Even so, suitable circulation
space around the hand basins and urinal stalls was available to
accommodate a trolley if it was brought into the toilet.

18.50 It is also noted that AA has taken on board the comments
received from the various airport trials and made efforts to modify and
improve the design of toilets. The efforts resulted in provision of
additional lighting, installation of hand dryers and widening of the dry
shelves. In particular, the height of the cubicle doors which originally
stretched from floor to ceiling was reduced, so as to alleviate the
claustrophobic feeling on the part of users. New larger toilets were also

575
constructed in the meeters and greeters area for the convenience of airport
users.

18.51 The Commissioners accept the rationale behind the design of
toilets for the new airport. The Commissioners also accept the view of
W51 Yuen that AAs policy of not allowing baggage trolleys into toilet
rooms is common amongst many airports. However, it is the view of the
Commissioners that toilets and their passageways could perhaps be
widened slightly to convenience airport users. While there has been a
suggestion that the consideration of commercial rental revenue might
have affected the provisions for toilets, the Commissioners do not find
any hard evidence to conclude that AA has inappropriately trimmed down
the facilities in order to maximise the commercial rental space in PTB.
Although it is true that the existing provisions meet the BAA standards, it
remains a fact that public expectations have not been fully met in terms of
the size and more generous allowances for space in toilets could have
been provided.

[19] Insufficient Water, Electricity and Staff at Restaurants

18.52 There were complaints about the service of restaurants at the
new airport. This is attributed partly to problems with water and
electricity supply to restaurants in the first few days after AOD and partly
to some staffing difficulties experienced by some operators.

18.53 According to AA, the problems relating to inadequate
utilities were caused by tenants who took possession of their premises at
the last possible moments and, hence, were late in their submission of
applications for connection of water and power supply. Also, some
tenants failed to carry out their work according to the required standards,
causing delay in water supply. There were a number of occasions of
power outage which, as alleged by AA, were mainly caused by faults in
the electrical installation put up by tenants. In an incident on 7 J uly
1998, the power failed for 2 hours and 40 minutes. The outage was
caused by improper loading setting in the installation of a tenant but the
maintenance staff of both AA and the contractor was refused access by
AVSECO to the switch room to carry out remedial work. In a separate
incident on 17 J uly 1998, the outage lasted for about four hours and AA

576
suspected the cause to be related to a contractor staff of Cathay Pacific
working on the CX lounge who left a fire hose reel running resulting in a
short circuit across the terminal. AA also revealed that the actual
demand for electricity from tenants was out of its expectation and, as a
result, the overall power system had to be upgraded. On the other hand,
tenants complained that they had difficulties in getting security permits
promptly to enable their contractors to carry out work in restricted areas.

18.54 For restaurants in the restricted area, there were problems
with their staff not receiving security passes by AOD and these prevented
them from attending to duties. On the landside, the large number of
visitors, in excess of 60,000 per day during the first week of airport
opening, taxed the facilities beyond expectations. As a result,
restaurants experienced problems of long queues, lack of food variety and
inability to operate long service hours. To address the issue, AA
reminded all catering licensees to comply with the service standards in
the licence agreements and made improvement to the permit issuing
process by introducing a new type of temporary permit from mid-J uly
1998.

18.55 The problems with restaurants was generally short-lived.
After the first week following AOD, the problems have substantially
subsided as the number of sightseers gradually decreases. The
Commissioners consider that both AA and the relevant tenants had a part
to play in the problems of utilities supply. Both parties have contributed
to the problem relating to the upgrade of the power system but, without
the benefit of time to investigate further into the matter, the
Commissioners are unable to apportion responsibility in this respect. As
to the power outage incident on 7 J uly 1998, the tenant concerned is
probably responsible for causing the problem although it is not possible
for the Commissioners to assign responsibility. Also, AA or AVSECO
should be held responsible for the delay in effecting the remedial work of
this electricity outage. For the incident on 17 J uly 1998, there is
however no substantial evidence before the Commission to pinpoint
clearly the culprit despite the allegation of AA. As to staffing, while the
tenants are responsible for ensuring that their staff were sufficient and
well trained so as to provide a reasonable level of service, the
Commissioners are unable to determine who should be responsible for the

577
late issuance of access permits to restaurant staff working on the airside
because of the massive late-minute rush for permit and the regular
breakdown of ACS and the permit system.

[20] Rats Found in the New Airport

18.56 It was reported in the media towards the end of August 1998
that thousands of rats were pestering the new airport. Allegedly, parts of
PTB and the aircraft maintenance facilities were affected.

18.57 The Commissioners find from evidence that AA arranged for
the employment of a full time professional pest control contractor to
provide pest control service for the common areas of PTB and the ground
transportation system in as early as October 1997. As a result, an
intensive 120-day rodent eradication programme was implemented with
effect from 1 May 1998. In addition, an in-house pest control team was
employed to carry out rodent control work in various areas of the new
airport and the work area covered the airfields, aprons and small airport
ancillary buildings when the airport went into operation in J uly 1998.
As to the airport tenants, they are required under their tenancy agreements
to implement their own pest control programmes. Periodic
environmental audits are also being performed in tenant areas to ensure
the adequacy of these programmes.

18.58 In the light of the evidence, the Commissioners are satisfied
that it is but a minor problem. Although it is not certain as to whether
rats could be completely eradicated, the situation appears to be under
control and will continue to be under control provided that AA keeps up
its efforts in this regard.

[21] Emergency Services Failing to Attend to a Worker Nearly Falling
into a Manhole while Working in PTB on 12 August 1998

18.59 On 12 August 1998, a worker nearly fell into a manhole in a
cable tunnel L3 near Gate 61 in PTB and sustained minor injuries as a
result. In the incident, it took 17 minutes for ambulance service to reach
the scene and locate the injured. It was discovered after the arrival of
the ambulance that special service operation crew was required to save

578
the injured worker. Therefore, another call had to be made to the Fire
Services Communication Centre (FSCC) through AOCC and that call
was only made 21 minutes after the first report.

18.60 The Commissioners are of the view that, on the first call, an
ambulance as well as a fire engine with trap rescue equipment should
have been despatched in view of the nature of the request for assistance.
There was apparently a misunderstanding between the caller and the
receiver of the call. However, the Commissioners are unable to
ascertain who should be responsible for the delay of rescue on the basis
of information available. This is, however, only a minor incident.

Resulting in Five Firemen Being Injured

18.61 On 28 August 1998, a Fire Service Vehicle lost control and
hit the kerb embankment whilst travelling along the slip road of the
Airport Road towards Tung Chung. Upon impact, the vehicle ran down
a slope and five Fire Services Department (FSD) personnel were
injured in the accident.

18.62 While the investigation of the Police did not reveal sufficient
evidence for further action to be taken, FSD concluded from its
investigation that the accident would be attributed to the drivers
misjudgement and FSD suspended the driver from driving duties. The
driver was also held responsible for paying for the repair cost of the
damaged vehicle. The Commissioners note that there has been thorough
investigation by both the Police and FSD over this incident and agree
with the findings of FSD.

Cabin of a Cathay Pacific Aircraft on 3 September 1998

18.63 A maintenance worker of HAECO fell from a flight of
staircase inside the cabin of a Cathay Pacific aircraft while at work on 3
September 1998. The worker sustained minor injuries in the incident.
The Commissioners consider that it is only an isolated accident for which
no one should be held responsible.

579

in Lifts and on the APM as well as Delaying Two Flights

18.64 From the press reports, the Commissioners note that, on 8
September 1998, passengers and airport staff were trapped in lifts and
APM trains for several minutes as a result of power failure. Two flights
were also delayed in the incident. Investigation into the incident was
made by Rotary but no conclusive evidence as to the cause has been
made available so far. Investigation is still ongoing. In any event, this
is only a minor incident.

October 1998

18.65 On 1 October 1998, a China Eastern Airlines flight MU503
was instructed to carry out missed approach when a Cathay Pacific
Airbus was unable to vacate the runway in time for the landing of MU503.
According to AA, missed approach procedures are safe and standard
manoeuvres published in the Aeronautical Information Publication for
pilots and, also, missed approaches are not infrequent occurrences in an
airport. In the light of evidence, the Commissioners agree that the
incident was handled safely, efficiently and in accordance with laid down
procedures. No responsibility can be attached to anyone.

Moderate Problems

[26] Delay in Flight Arrival and Departure

18.66 There were significant delays of incoming and outgoing
flights during the first week of operation of the new airport. On AOD,
incoming flights and outgoing flights experienced an average delay of 24
minutes and 2.63 hours respectively. The delays became more serious
after around 11 am when traffic was very busy. These delays were not
however problems in themselves but, rather they were the results and
consequences of other airport problems such as the inefficiency of FIDS,
difficulties in baggage handling, airbridge malfunctioning, confusion over
parking of planes, late arrival of tarmac buses, problems encountered by

580
RHOs and other operators in the use of TMR and mobile phones, and
malfunctioning of PA and ACS. Another contributory factor on AOD
was cargo handling chaos which caused delays in the processing of cargo.
The combined effect of all these factors was that it took much longer than
the usual turnaround time for an aircraft arriving at and departing from
the new airport.

18.67 In the light of the evidence, the Commissioners accept that
flight delay is only a consequential problem resulting from a combination
of other problems. The cause of and the responsibility for them can be
found under the respective items.


18.68 ACS is a computerised system that performs 3 functions,
namely, production of permits, verification of permits and monitoring
movement of personnel through ACS doors.

18.69 Guardforce Limited (Guardforce) was the contractor for
ACS. Controlled Electronic Management Systems Limited (CEM)
was the nominated subcontractor of Guardforce, mainly to provide
software works for ACS. The British-Chinese-J apanese J oint Venture
(BJ C) was another contractor of AA to provide doors, electromagnetic
locking and detection devices. The processing of permit applications
were carried out by AVSECO.

18.70 ACS had not been completed on AOD, although AA claimed
that it was operational. There had been significant slippage for site
acceptance test (SAT) which was supposed to be carried out in
December 1997. As at 30 November 1998, SAT was only about 60%
complete.

18.71 Since AOD, there were various reported problems with ACS.
There were problems with the timely production of security permits.
The lack of security permits affected staff and workers in carrying out
their work. Some of the ACS doors including airbridge doors were not
working. On AOD, 11 out of 38 airbridge doors were not working.
There were reported incidents both on and shortly after AOD of

581
passengers being trapped in the airbridges because of the malfunctioning
of the ACS doors. This resulted in the deactivation of all the airbridge
doors for departing flights from 7 J uly to 19 J uly 1998 and the security
guards being posted to maintain security. The malfunctioning of ACS
doors also had an impact on airline staff and other people working at the
new airport.

18.72 Delay in permit production. The Commissioners find that
the development and installation of ACS had been plagued by delays and
various problems, which contributed to the delay in permit production.
In particular, the Commissioners have come to the following conclusions:

(a) Guardforce should mainly be responsible for the delay
caused by the instructions to include Chinese text in the
permanent permits, although AA should have imposed the
requirement of Chinese text in the contract or issued the
instructions earlier.

(b) Guardforce should be responsible for the breakdown of
the printing equipment, breakdown of the permit system
caused by the failure of the server at Kai Tak and the lack of
ink and paper for permit production.

(c) Guardforce should not be responsible for the two
occasions of downtime in the Permit Production Office
caused by power failures or power changeovers. There is
insufficient evidence for the Commission to reach a finding
whether AA should be responsible for these downtimes.

(d) The business partners of AA, and possibly AA, should be
responsible for the large number of last minute rush
applications for permits. There is no sufficient evidence to
conclude that AA failed to make planning to avoid late
applications or that such plan was not followed through by
AA. Accordingly, AA should not be responsible.

18.73 On ACS doors and other problems relating to the disruption
of the works under the ACS contract C396, a great number of allegations

582
were raised by the parties concerned. After considering the evidence
very carefully, the Commissioners make the following findings:

(a) On AAs instructions to set up a temporary system at Kai
Tak for permit production, if Guardforce had felt that the
instructions were outside its original scope of work, it could
have either refused to accept the instructions or have warned
AA of the risk of disruption. Guardforce failed to do either.
If Guardforce accepted the instructions as it did, it must
provide additional resources to complete the work without
allowing it to cause disruption to the contract works.

(b) On AAs instructions to transfer the data from Kai Tak to
Chek Lap Kok (CLK), Guardforce should be responsible
for the disruption caused by these instructions, since the need
to transfer data was already foreseen by Guardforce.

(c) The alleged late instructions for five further computer
terminals should not reasonably be considered as a factor
contributing to the ACS problems. The added computer
terminals would presumably have helped quicker production
of permits and should not have been treated as a problem, in
particular, if Guardforce had sufficient resources to comply
with the instructions.

(d) The Commissioners accept that physical damage to doors
and wrongful activation of alarms caused disruption to the
installation and testing of ACS. Those people who
committed such irresponsible acts of vandalism should be
responsible. There is evidence that AA did make a lot of
efforts to prevent vandalism. It may be unreasonable
therefore to find AA responsible.

(e) Guardforce should be responsible for its incorrectly
installed apparatus.

(f) Guardforce should be responsible for the software
problems of ACS.

583

(g) Guardforce should be responsible for the downloading
problems.

(h) Guardforce and CEM probably had resources problems,
and they should be responsible for having inadequate
resources in performing contract C396.

(i) AAs allegedly late instructions would not have caused
serious delays in the C396 contract works. Nevertheless,
they must have caused some hindrance to Guardforces work.

(j) The late application for permits by the business partners
of AA did cause added difficulty to ACS.

(k) Guardforce was hampered by the delay in completing and
repairing the mechanical parts of the doors of ACS.
According to BCJ , the problems with the defective door
holders were attributable to the design changes of AA. The
fact that Guardforce was awarded extensions of time would
indicate that it was affected by such delay.

(l) Guardforce was disrupted in its work by the damage to its
equipment caused by other contractors in PTB. It seems that
had the system not been loaded with so many alarms,
Guardforce would have been able to detect software problems
before AOD. For those alarms that were set off due to
operational errors, it is not clear from the evidence whether
these were caused by lack of training of the staff by AA or the
airlines or whether they were caused by the operators own
faults.

(m) Some delay was also caused by AA, which did not
provide in time the General Building Management System
and Building Systems Integration package for the purposes of
the model tests of ACS, for which AA should be responsible.

(n) AA, as the overall coordinator of the works, should bear

584
some responsibility for the delay in the construction, which
meant that Guardforce could not carry out its work on a
system where fitting-out had finished and vandalism was not
so rampant.

(o) AA should have recognised that there would be problems
with opening doors on AOD. AA should have assigned staff
to be ready with keys and other means of opening locked
doors. This would have avoided the incidents of passengers
being trapped, although the incidents were more an
inconvenience than a security risk.

[28] Airside Security Risks

18.74 Airside security is of utmost importance in the overall
context of airport security. The Commissioners however have evidence
to find that there were airside security risks at the new airport as reflected
in the following four incidents, although the first one in fact did not
present any such risks.

(a) Delayed entry of police motorcycles into restricted area.
On 10 J uly 1998, a minor traffic accident inside Baggage
Hall resulted in two workers sustaining slight injuries.
While two ambulance service vehicles were allowed
immediate entry to the Enhanced Security Restricted Area to
attend to the injured, the traffic police on motorcycles
experienced a delay because their siren and flashing lights
had not been switched on. According to section 22 of the
Aviation Security Regulations, the requirement for permits is
exempted where disciplined and emergency service vehicles
respond to an emergency. The established procedures of
AVSECO provide that these vehicles would be allowed
immediate entry if their siren and flashing lights are
activated. The Commissioners consider that, in ensuring
prompt and effective response to an emergency, there should
be no room for misunderstanding among the parties involved
of the correct procedures. As highlighted by this particular
incident, either there was ambiguity in the AVSECO

585
procedures, or there had been a failure of communication
between AA and the Police. The procedures have been
fine-tuned after the incident and the revised procedures have
worked well.

(b) Transit passengers allowed to enter Departures Hall and
board a flight without security check. This incident
happened on 25 J uly 1998 when staff of China Airlines
Limited (CAL) took approximately 90 transit passengers
from aircraft to the Departures Hall directly, without going
through the required security screening. These transit
passengers boarded the aircraft which took off but was
subsequently recalled by CAL for re-screening of the
passengers. At the material time of the incident, the ACS
door at the relevant boarding gate, which would have been
an effective barrier prohibiting access from the airbridge to
departures Level 6, did not function. Also, the AVSECO
guard stationed at the airbridge failed to stop the transit
passengers from proceeding to Level 6. Although, upon
notification, the AVSECO Duty Manager requested the CAL
Duty Manager to undertake security screening for the
passengers, the flight had already departed by the time the
CAL Duty Manager decided to do so. Both the Hong Kong
Aviation Security Programme (HKASP) and the Hong
Kong International Airport-Airport Security Programme
clearly require airline operators to ensure security screening
of their transit passengers. In the incident, CAL breached
the relevant requirement and this was admitted by CAL.
Furthermore, CAL should be blamed for its failure to stop
the flight in time for security screening, necessitating recall
of the aircraft after it had taken off. The responsibility for
the malfunctioning of the ACS door in relation to the
incident is set out under item [27]. As to the performance
of the guards stationed at the airbridge, the Commissioners
accept that they were outnumbered by the transit passengers
but, nonetheless, they failed to intervene effectively to stop
the group. The guards could have adopted a more robust
approach to intervene. All in all, had the ACS door not

586
malfunctioned, the incident with the resultant security risk
would probably not have occurred. Following investigation
into the incident, Civil Aviation Department (CAD)
offered a number of suggestions to improve security
arrangements to prevent recurrence. Some of these
suggestions have already been implemented and they include
putting up appropriate signs, setting up tensa barriers in
airbridges to demarcate more clearly the arrival and
departure channel within airbridges and location of separate
transfer desks within the body of PTB.

(c) Unauthorised access to Airport Restricted Area ("ARA").
According to the records of the Police, there were a total of
55 reported cases of breach of ARA between AOD and 17
October 1998. In some of these incidents, the offenders
failed to bring with them their own permits or used
colleagues permits for convenience. According to the
evidence received from AVSECO, the majority of
unauthorised entries were technical in nature devoid of any
criminal intent. The causes were attributed to permit
holders not being familiar with the new environment at the
new airport, inadequate instructions being given to them,
insufficient signage during the initial stage of operation and
the less than effective control over unauthorised entry whilst
ACS was under test. Following implementation of some
improvement measures to the signage and the overall
security system, there was a marked decline in the number of
such incidents in the subsequent months. While the
Commissioners agree that the majority of the incidents were
technical in nature, they are of the view that AVSECO
should nonetheless be held responsible for its failure to
prevent the 55 cases of unauthorised access into ARA in the
first place. Some ARA permit holders were responsible for
the inappropriate use of their permits resulting in
unauthorised entry. Also, AA should be blamed for not
putting up sufficient signage to indicate boundaries of the
area.

587
(d) A KLM flight took off with baggage of two passengers who
were not on board. On 8 J uly 1998, a KLM flight departed
with the checked baggage of two passengers on board but
without the passengers. The incident arose from some
functional difficulties with the boarding gate readers
(BGRs) which were used to scan boarding passes (BPs).
This necessitated manual collection and checking of the BPs
stubs so as to verify the number of passengers on board.
Upon verification, 10 passengers were found missing. The
cabin crew then conducted a passenger head count but,
unfortunately, the head count was incorrect leading to
boarding staff having the impression that all passengers were
on board. It was not until the two missing passengers
turned up at the boarding gate that the staff realised the
mistake in the head count. But the aircraft was about to
take off at that time. The two passengers were
subsequently arranged to depart via another airline. The
investigation of CAD failed to establish whether it was
human error or the malfunctioning of the BGR system at the
time that had caused the incident. However, KLM was
found to be in breach of the requirements in the HKASP to
remove the baggage of a passenger who does not board the
aircraft. This requirement is an additional safeguard for
passengers safety since all passenger baggage is security
screened to comply with international standards. The
Commissioners fully concur with the investigation results
and are satisfied that the incident is an isolated case of
failure to comply with the HKSAP procedures for passenger
and baggage reconciliation. This was caused by a human
error in the head count, for which KLM should be
responsible.

[29] Congestion of Vehicular Traffic and Passenger Traffic

18.75 On AOD, there were problems with traffic congestion,
congestion at lifts from Level 3 (ground level) to PTB and contra-flow
movement among passengers on the down ramp from Level 3 to Arrivals
Hall on Level 5. During the first week of AOD, more than 60,000

588
curiosity sightseers per day visited the new airport and many of them took
the external buses and shuttle buses which stopped at Cheong Tat Road
on Level 3 outside PTB. Traffic congestion occurred at the section of
the road where passengers alighted and got on board of the buses. The
situation was aggravated by the suspension of one of the two bus stops
there and the non-completion of pavement work. The problem greatly
inconvenienced users of the airport for a short period since AOD.

18.76 Passengers who wish to go into PTB after alighting at
Cheong Tat Road can make use of the six passenger lifts and escalators in
carpark 2 and Level 3. However, none of these facilities were put into
service on AOD. As a result, the passengers had to make use of the two
staff lifts, the down ramp leading to the Arrivals Hall and the two
emergency staircases. This resulted in lift congestion and contra-flow
movement among passengers along the down ramp.

18.77 To remedy the situation, the Transport Department
introduced measures to re-route and divert some of the routes. AA also
deployed additional staff for traffic and crowd control and installed
temporary signs and barriers to direct arriving passengers. Availability
of lifts from Level 3 had also increased since 12 J uly 1998. With these
measures, the problems of congestion were resolved, particularly when
the number of visitors gradually subsided after AOD. The
Commissioners note the view of W51 Yuen that extraordinary increase in
traffic on opening of major airport facilities is a common occurrence due
to drivers circulating the roadways to find their destinations. However,
on evidence, the Commissioners consider that the Transport Department,
as the party responsible for approving the design and monitoring of the
operation of transport facilities, should take more precautionary steps in
traffic planning for the opening of the new airport. For such a
significant event, the large number of curiosity visitors could be
foreseeable. As for AA, the Commissioners find that it should be
responsible for the insufficient signage and inadequate lift service on
AOD, which eventually led to undesirable congestion of people.

[30] Insufficient Air-conditioning in PTB

18.78 The air-conditioning system in PTB mainly consists of the

589
following:

(a) the pump system for supplying seawater for cooling purpose;

(b) chillers for the supply of chilled water throughout PTB; and

(c) the air-conditioning plant for the supply of cool air to the
public areas.

18.79 The Youngs Engineering Company Limited (Youngs) is
the contractor for seawater pumps whereas AEH is the contractor for the
chillers and the air-conditioning plant.

18.80 There were a number of incidents of air-conditioning failure
inside PTB with different length of duration and varied causes. The
insufficiency of air-conditioning was mainly due to frequent tripping of
the chillers in PTB and a number of these incidents were caused by power
failure, human error or technical faults. The systems design setting of
temperature at 24C instead of a more acceptable 22C may also be a
contributory factor accounting for the public perception of inadequate
air-conditioning during the summer months in which the new airport
opened for operation. In the tenant areas, there were delays experienced
in the energisation of tenants chilled water supply, causing insufficient
air-conditioning supply to these areas. The delays were mainly caused
by the large quantity of late requests from tenants for connection to
chilled water supply and the failure on the part of tenants to complete or
commission their air-conditioning installations. With increased working
hours and labour from AA and AEH, all tenant requests for chilled water
supply were processed by 13 J uly 1998.

18.81 Evidence received has revealed altogether 12 reported
incidents of chillers shutdown causing disruption to the supply of
air-conditioning to PTB. These incidents are briefly described below:

(1) 6 July 1998. On 6 J uly 1998, one of the three chillers
shut down during various periods for approximately five hours
causing the temperature in PTB to rise by about two to three
degrees Celsius. There are cross allegations from AA, AEH

590
and Youngs as to the technical causes for the event and,
without examining the system in detail, the Commissioners are
not in a position to come to any conclusive view on the issue of
responsibility. Youngs, however, did admit that there was a
faulty flow switch in the seawater pump house for which it was
responsible. In the Commissioners opinion, it appears that
there was also a problem of interfacing between the systems in
pump house control room (for which Youngs is responsible)
and the chiller room (for which AEH is responsible) and, in the
that respect, AA should bear the responsibility for failing to
coordinate and organise the necessary interface testing between
the systems of the two contractors concerned.

(2) 10 July 1998. On 10 J uly 1998, one seawater pump
tripped causing one of the three chillers running to shutdown
due to insufficient seawater flow. The incident was caused by
human error and, for this, Youngs is responsible.

(3) 12 July 1998. Arising from a sudden energisation of a
main chilled water branch, the pressure of the chilled water
system dropped causing two of the four operating chillers to
shutdown on 12 J uly 1998. AEH was responsible for the
occurrence of sudden energisation, which could have been
avoided if the valves were opened slowly to minimise the
system pressure fluctuations.

(4) 13 July 1998. On 13 J uly 1998, all the four chillers
running and the secondary chilled water pumps were shutdown
due to power voltage fluctuations that, allegedly, had been
caused by lightning strike. To avoid future fluctuations or
loss in power supply, uninterrupted power supply (UPS)
units were installed to the chiller control panels and the panel
serving the seawater controls in the chiller plant rooms
between 28 September and 27 October 1998. In the event, it
also came to light that although the chillers had tripped and the
demand for seawater had ceased, the seawater pumps
continued to operate. Youngs admitted that this was
attributed to a small error in the control logic of the seawater

591
pumps within the pump house. The error was rectified on the
same day.

(5) 28 August 1998. All chillers tripped on 28 August 1998
due to lightning strike affecting power supply to the pump
house. This incident was similar to the one at item (4) above.
The instructions for installation of the UPS units were issued
by AA on 17 J uly 1998. Had AA issued the instructions
earlier or had the UPS units been installed earlier, or had other
precautionary measures been taken much earlier, this incident
and the other two incidents described in items (4) above and (7)
below might have been avoided. In this context, the
Commissioners opine that AA should bear some responsibility
for the late instructions.

(6) 29 August 1998. On 29 August 1998, all chillers tripped
because of loss of seawater supply, resulting from power
failure. Also, as alleged by AA, the electrical protection
setting to the banscreen motors in the seawater pump house
had been incorrectly set and Youngs immediately altered the
setting to rectify the problem. Although Youngs alleged that
the loss of power was not within its control, it should be
responsible for the incorrect electrical protection setting.

(7) 30 August 1998. Similar to items (4) and (5) above, all
chillers tripped on 30 August 1998 due to lightning strike
which affected the power supply. AA should be responsible
for the late instructions for installation of UPS units.

(8) 8 September 1998. On 8 September 1998, all chillers
(two chillers only, as alleged by AEH) tripped due to a power
failure caused by tripping of circuit breakers on Youngs
switchboard. Youngs admitted its responsibility to the extent
that the system was vulnerable due to critical control circuits
not being on a dedicated supply.

(9) 14 September 1998. On 14 September 1998, all chillers
tripped due to human error whilst the contractor for the

592
Mechanical Building Management System carried out testing
of that system. In the opinion of the Commissioners, no
responsibility should be assigned except to the person who
committed the error.

(10) 12 October 1998. All chillers (three chillers only, as
alleged by AEH) and the air handling units tripped as a result
of a disturbance by the power system of China Light & Power
Company Limited (CLP). CLP alleged that the incident
was caused by third party damage to their underground cable.
However, that responsible third party could not be identified.

(11) 22 October 1998. The incident on 22 October 1998 was
a planned shutdown to enable testing on an interface with the
seawater pump house to be carried out and no party should be
blamed for the incident.

(12) 28 November 1998. On 28 November 1998, all chillers
tripped due to a loss of seawater supply. The loss of water
supply was caused by an unauthorised isolation of power
supply to the high voltage battery charger and the associated
UPS unit. Also, the UPS unit had been incorrectly set to
bypass mode which prevented power backup in the incident.
There is inconclusive evidence as to which party should bear
the responsibility for the power interruption.

18.82 Although the Commissioners do not have the benefit of time
to examine all the issues involved in greater detail, it is, however, evident
that AA, AEH and Youngs should bear certain degrees of responsibility
for the failure of the air-conditioning system in some of the incidents.
As to the problem in tenant areas, the Commissioners consider that much
of the blame should lie with the tenants themselves because of their late
applications for connection as well as non-compliance with the
connection procedure.

[31] PA Malfunctioning

18.83 Public announcements at the new airport are made either

593
through Central PA or Local PA. In the case of the former,
announcements are made centrally from AOCC which may broadcast to
all or selected areas. Local PA comprises of consoles near the boarding
gates controlled by airline operators and AA staff.

18.84 AOCC is linked to the communications rooms throughout
PTB via the Building System Integration (BSI) package and the Voice
Routing System (VRS). As both BSI and VRS were not available on
AOD and sometime thereafter, announcements were made through the
manual all zone (MAZ) system, which operates through a notebook
computer in AOCC and which is also connected to one of the
communications rooms.

18.85 Hepburn Systems Limited (Hepburn) is the main
contractor for PA and SigNET (AC) Limited (SigNET) its
sub-contractor.

18.86 Between AOD and 16 August 1998, Central PA was down on
several occasions. Most notably, on 7 J uly 1998 Central PA was down
six times, totaling over three hours with one downtime lasting over two
hours. Local PA also experienced a significant number of problems.
In the first four weeks since AOD, AA recorded 194 problems out of
about 50 consoles. While some problems were caused by human errors,
others were caused by hardware and software problems.

18.87 Hardware problems included the late and incomplete
installation of equipment, including speakers, consoles and the ambient
noise-operated amplifier facility at various locations throughout PTB and
the Ground Transportation Centre. This was mainly due to the lack of
time and PTB not being absolutely ready. Other hardware problems
included human induced damage to membranes covering the consoles
and gooseneck microphones and defective consoles due to the failure of
electrical components.

18.88 There were some problems with the intelligibility of
announcements on some occasions such as unclear or no announcements
made, and problems relating to echoing and volume. Adjustments to
feedback and volume could not be made until the Rapid Assessment of

594
Speech Transmission Index (RASTI) testing was completed which in
turn could not be carried out until acoustic related materials were
installed in PTB.

18.89 There were other software problems experienced. For
instance the function which prevented overlapping announcements was
not set up (zoning problem), low priority announcements blocked out
more important messages (priority problem), the slow response time in
the logging on process for some gates, MAZ system overriding the
loading gate console causing the gates to be inoperable, instability of
Central PA resulting in MAZ notebook outages, and incidents of lock
out problems caused by fire evacuation warning announcements.

18.90 Most software problems were caused by the required tests
not being performed by AOD, largely due to inadequate resources from
Hepburn and SigNET. Hepburn admitted that they had a problem with a
subcontractor, Univision Engineering Limited, which affected the
development of an interface software to the BSI. This resulted in a
delay of factory acceptance test which was not completed until the end of
J une 1998. There was also significant delay in performing SATs, which
only began in May 1998 and completed in October 1998, long after AOD.
AA and Hepburn agreed to defer RASTI testing after AOD.

18.91 A remedial programme was developed 10 days after AOD
for the completion of outstanding work. Hepburn provided staff 24
hours a day during the first week of operation. Since AOD, Hepburn
concentrated on resolving system integration and reliability including
software problems, upgrading Local PA, level adjustments, zoning issues
and hardware problems.

18.92 PA was one of the major back-ups and workarounds for the
dissemination of flight information in the absence of an effective FIDS.
AA is responsible for the lack of effective contingency planning. AA
knew PA would not be completed or completely tested before AOD and it
also knew that PA had not gone through SATs and that there were
problems with the Local PA. Further it did not plan for the possibility
that both PA and FIDS may not work at the same time. Fortunately
there was no evidence that any passengers missed their flight as a result

595
of gate changes. Hepburn claimed that PA was putting out about 270
calls per day from AOD despite all its problems.

18.93 Whilst the incomplete installation of some PA equipment
was partly due to the late readiness of PTB, the Commissioners find that
the major delay was caused by Hepburns failure to meet its deadlines.
Although there may have been frequent changes of instructions from AA,
there were extensions of time granted to Hepburn. The primary reason
for the delay was the inadequate resources that Hepburn and SigNET had
assigned to the contract.

18.94 In relation to the physical damage made to the membranes
and gooseneck microphones, it is the Commissioners view that whilst the
damage might have been caused by vandalism, the damage could have
been caused by careless use by operators. All console users should be
advised of the proper use of the consoles so as to prevent unknowing
damage.

18.95 As contractor of PA, Hepburn is responsible for the defective
consoles due to failure of electrical components.

18.96 In relation to the intelligibility problem, the Commissioners
find that neither AA nor Hepburn should be responsible. Without
having the acoustic related materials ready in PTB, the necessary tests
would not have been useful or meaningful.

18.97 As Hepburn had delayed in its work on PA, the
Commissioners are of the view that they should primarily be responsible
for the zoning and priority problems. Had there been more tests and
trials, these problems could have been uncovered and remedied.

18.98 The slow response time of the consoles and the overriding
problem are software problems for which Hepburn should be responsible.

18.99 The issues of MAZ outages and locking problems, the
Commissioners find that these were software problems, as accepted by
W47 Mr Graham Morton, and accordingly Hepburn should be
responsible.

596

18.100 Problems with PA still exist although the system is more
stable with fewer faults and failures. The slow response time of PA in
the logging on process was rectified early in September 1998. Repair
and replacement of defective hardware was largely completed by the end
of September 1998. Problems with the fire announcement were fixed
around 15 October 1998.

18.101 SATs, including RASTI, were completed at the end of
October 1998 while testings for the maintenance reporting terminal were
scheduled to be completed by the end of November 1998. Hepburn
expect confidence trials for Central and Local PAs to be completed by
about March 1999.

[32] Insufficient Staff Canteens

18.102 There were complaints from airport staff about the lack of
sufficient staff canteens. There was a specific allegation that staff had to
wait for more than 40 minutes for a table and food on some occasions.
The new airport has a working population of about 45,000, with about
14,600 people working daily in PTB. A total of four staff canteens were
planned to cater for the need of staff and they altogether provide 954
seats. The Commissioners find from the evidence that, right from AOD
up to 13 J uly 1998, there was only one staff canteen in operation. Two
others were not opened until later that month and the last one came into
service only on 15 October 1998. Therefore, the full planned canteen
capacity was not available during the initial period of airport opening.
Furthermore, owing to the large number of visitors around that time, the
alternative to use the commercial catering facilities at PTB which were
crowded with visitors did not help the situation.

18.103 Upon review of the evidence, the Commissioners note that
the original concept was to build a main staff canteen within the
maintenance building alongside PTB but, for various reasons including
cost and profitability, the proposal did not materialise. Also, there did
not seem to have been a scientific and realistic assessment of the catering
requirements for staff working at the new airport. In this regard, the
Commissioners consider that AA should be responsible for its poor

597
planning of staff catering facilities. The planning ratio of 15 to 1,
assuming 14,600 people as against 954 seats in staff canteens, appears to
be on the low side. Furthermore, AA should be blamed for not ensuring
that all the four planned canteens could open for service right from AOD.

[33] RFI on Air Traffic Control Frequency

18.104 CAD has been receiving reports from airline pilots regarding
RFI on air-ground Very High Frequency radio communication channels
used by air traffic control since late 1994. To address the problem, CAD
used spare frequencies to replace the affected ones for communication in
the event of interference and had brought in six additional frequencies as
extra backup for air traffic control since 1996 to safeguard flight safety.

18.105 Investigation by the Office of Telecommunications Authority
showed that the sources of RFI were in the form of spurious or
intermodulation signals originated from some unknown paging stations
along the coastal areas in the Guangdong Province. The Mainland
authorities have adopted a range of measures to tackle the problem
including dismantling radio transmitters on top of hills, and closing down
offending paging stations. Some cities have also introduced tighter
control measures on paging stations such as limiting their transmission
power and requiring them to install filters and isolators. Since May
1998, a Technical Working Group was established with technical experts
from Hong Kong and the Mainland authorities to step up cooperation in
addressing the RFI issue. A Task Force has also been formed between
operational personnel of Hong Kong and the Mainland authorities for
quick exchange of RFI information, if necessary.

18.106 The Commissioners are satisfied that both the Hong Kong
and Mainland authorities attach great importance to flight safety and
strenuous efforts are being made to eliminate RFI completely. No
finding is therefore called for in respect of this problem.

[34] APA Malfunctioning: a Cathay Pacific Aircraft was Damaged when
Hitting a Passenger J etway during Parking on 15 J uly 1998

598
18.107 APA is a laser scanning device that directs the pilot to park
the aircraft through a real time display unit. On AOD, three out of the
68 APAs (comprising 28 Building Mounted APAs, 40 Gantry APAs) at
the new airport were not functioning. In an incident on 15 J uly 1998, a
Cathay Pacific aircraft was damaged during parking, allegedly as a result
of the malfunctioning of a Gantry APA at a frontal stand. As from that
day, all APAs were suspended from use. Prior to that, there were
occasions on which the Gantry APAs were unable to give necessary
directions to the pilot. According to Safegate International AB
(Safegate), the contractor for the design and maintenance of the APA
system, problems with the Gantry APAs was due to the height of the
Gantry which affected the laser scanning angle, the stop position and the
aircraft type in question. Another contributing factor was that Safegate
staff had inadvertently disabled the auto-calibration function of the
system, which could have detected a sensor problem. To address the
problem, enhancement was made to the software of Gantry APAs to
increase the effective viewing angle of the laser. As for the Building
Mounted APAs, there were incidents of non-operational problems. The
problems were caused by: (a) curtailed airflow within the display units as
a result of installation of sponge washable air filters; and (b) the unstable
voltage experienced at the new airport. To rectify the problem, Safegate
removed the sponge filters and optimised the size of the thermo fuses and
resistors to accommodate the voltage situation. Since 12 September
1998, all APAs at frontal stands have been put back to service and as at
17 September 1998, all Gantry APAs were successfully tested before
being put to use.

18.108 In the incident on 15 J uly 1998, the APA was apparently not
working properly and the air marshall had to give hand signal the pilot to
stop. Unfortunately, the pilot apparently misunderstood the signal of the
marshall as a direction to move forward. When the pilot realised the
emergency stop signal and stopped the aircraft, it had overshot by about
six metres and had hit the passenger jetway. The marshall could have
pressed the emergency stop button on the control panel of the APA to
effect the display of Stop message on the display unit so as to direct the
pilot to stop. However, the control panel was outside the reach of the
marshall at that time and the marshall had therefore to resort to manual
signalling.

599
18.109 There were cross allegations between AA and Safegate as to
the general causes for the malfunctioning of APAs. However, judging
from the evidence and having regard to the fact that Safegate had to
undertake some remedial measures after AOD, the Commissioners come
to the view that Safegate should be responsible. As for the incident on
15 J uly 1998, Safegate should be held responsible for the malfunctioning
of the laser sensor and, also, the inadvertent act of its staff to disable the
auto-calibration function of the system. The Commissioners find
comfort in that AA has agreed to reposition the control panel, obviously
to enable air marshall to reach it whilst working on the ramp.

[35] An Arriving Passenger Suffering from Heart Attack not being sent
to Hospital Expeditiously on 11 August 1998

18.110 There was a complaint that on 11 August 1998, an arriving
passenger with heart attack on board China Southern Airlines flight
CZ3077 from Hainan to Hong Kong was not sent to hospital
expeditiously. The evidence shows that it took thirteen minutes for the
ambulance to reach the patient after FSCC received an emergency call
through 999 at 10:56 am. In between, five minutes have been spent
by the ambulance in waiting at the apron gate for the ACC escort vehicle
to arrive.

18.111 The Commissioners note that when FSCC received the
emergency call, the relevant aircraft was already on the apron.
Apparently, the cabin crew did not notify the ACC or AOCC about the
sick passenger on board before landing. Hence, no arrangement had
been made to put an ambulance on standby on arrival of aircraft.
Though the delay did not result in a major incident, the Commissioners
have to hold the China Southern Airlines responsible for failing to notify
the airport about the sick passenger before landing. After the incident,
AA has reminded airlines that the flight crew should notify the airport
before landing if a passenger was taken ill on board.

18.112 Although both the ambulance and the ACC escort arrived at
their destinations within their normal response time, the Commissioners
find that better coordination and communication between FSCC and ACC
could have helped to cut down the response time in this incident. From

600
the evidence, the five-minute waiting time for the ambulance at the apron
gate was to some extent caused by the need for FSCC to contact ACC
indirectly through the Airport Main Fire Station Rescue Control
(AMFSRC) for an escort. In the light of the incident, AA and FSD
are arranging a direct line to be installed between FSCC and ACC so that,
in future, requests for ACC escort vehicle do not have to go through
AMFSRC.

Arriving Aircraft on 25 August 1998

18.113 On 25 August 1998, four fire engines drove across the
runway to attend to an incident of a J apan Airlines Company Limited
(J AL) aircraft without obtaining permission from the Air Traffic
Control (ATC), forcing a Cathay Pacific flight to abort take-off and a
China Eastern Airlines flight to delay landing.

18.114 According to the report of FSD on the incident, the four fire
engines were despatched to respond to an incident of a J AL aircraft. The
Rescue Leader of the four engines radioed the ATC tower for clearance to
cross the runway. Before he could obtain the necessary clearance, the
driver of the first fire engine speedily drove across the runway without
confirming permission from ATC nor the Rescue Leader. As the first
engine was crossing the runway at high speed, the Rescue Leader
considered that instructing it to return would only lengthen the time of the
fire appliance staying on the runway, further obstructing runway
operation. Seeing that the aircraft at the threshold of the runway was
stationary, he quickly followed with the remaining three appliances and
dashed across the runway.

18.115 According to CADs report, the Rescue Leader of the fire
engines only reported on radio that they were responding to the request
for inspection of the J AL aircraft without asking for specific permission
to cross the runway. When ATC saw the fire engines crossing the
runway, it immediately instructed a Cathay Pacific Airbus A340 aircraft
which had just been cleared for take-off to abort take-off . An incoming
China Eastern Airlines Airbus A320 was also instructed to discontinue its
approach. No danger to safety was involved.

601

18.116 The procedures for vehicles entering the runway is clear and
unmistakable. All relevant communication equipment was functioning
properly in the incident. The Rescue Leader and the driver of the first
fire engine were responsible for the failure to obtain clearance from ATC
before crossing the runway. They had been disciplined by FSD
subsequent to the event. FSD also reminded its personnel of the proper
procedures for appliances to seek permission from ATC before entering
the runway. The Commissioners opine that FSD has taken appropriate
follow-up action on the incident.

Persons on 6 September 1998

18.117 On 6 September 1998, a tractor of HAS crashed into a light
goods vehicle (a control van), injuring five persons. The driver of the
tractor towing two empty containers and an empty dolly was driving in
the restricted area of the airport. As he was driving between two lines of
containers, his view was partially blocked on the left while he was going
out of the area and he was not aware of the arrival of the control van.
The tractor collided with the control van passing horizontally in front.
As a result of the collision, five persons on the control van sustained
injuries. All but two were immediately discharged after medical
treatment and none was hospitalised.

18.118 The Commissioners consider this incident of a moderate
nature not only because five persons were injured but also because it was
a traffic accident occurring inside the restricted area of the new airport.
It is necessary to maintain the new airport as a safe place, and the incident
may give rise to an impression to the public that the airport itself is not
running safely and smoothly. The incident report of HAS found that the
driver of the tractor had not followed the proper driving procedures in
stopping his tractor to ensure road clearance in front when he was driving
between two lines of containers. As a result of Police investigation,
prosecution was made against the driver for careless driving. The
Commissioners are satisfied that HAS and the Police have investigated
into the incident thoroughly.

602
Runway Closures on 12 October 1998

18.119 On 12 October 1998, United Arab Emirates flight EK9881, a
cargo B747-200 aircraft leased from Atlas Air, Inc. (Atlas Air),
sustained tyre burst on departure for Dubai, leaving behind tyre debris on
the runway. Tyre fragments covered an extensive area of the runway.
The runway was closed for 40 minutes for removal of the tyre debris.
About one and a half hours after take-off, the aircraft returned to Hong
Kong because of a slight hydraulic problem, damaging runway lights on
landing. The runway was closed twice, 39 minutes and 20 minutes
respectively, for inspection of the runway conditions and emergency
repairs to the lights. Further repairs to the lights were made overnight.

18.120 The incident, which necessitated the closure of the runway
three times a day, had an impact on the operation of the airport. During
the runway closures, four aircraft were diverted to alternative airports, 42
arriving flights were delayed between 15 and 69 minutes and 88
departing flights were delayed between 15 and 75 minutes. It is on this
basis that the Commissioners treat this incident as a moderate one. At
the time of the incident, the relevant freighter aircraft was operating under
a lease agreement between the Emirates Airline and the Atlas Air and was
fully controlled by the Atlas Air crew. Accordingly, Atlas Air has to be
held responsible for the incident.

[39] Power Outage of ST1 due to the Collapse of Ceiling Suspended
Bus-bars on 15 October 1998

18.121 On 15 October 1998, there was a structural failure of a
section of ST1s power distribution system linking ST1 to a local
substation causing disruption to ST1s operation. A large section of the
ceiling suspended bus-bars and cables used in the distribution of
commercial power to certain parts of the building collapsed around 6 am
on 15 October 1998, cutting power to mainly the eastern half of the
terminal building. This resulted in a number of airline offices losing
power, limiting their ability to communicate with their counterparts who
were required to be informed of cargo movement, both in and out of ST1.
The power failure affected ST1s operational efficiency and slowed down

603
the processing time for some types of cargo. The Express Centre and
the Perishable Cargo Handling Centre were operating normally.
Temporary power was restored to the airline offices by early evening of
the same day and, for operating computers, telephones and facsimile
machines, some 12 hours after the failure. Temporary measures were
also employed to restore power to other affected areas. Permanent
power for air-conditioning and full lighting in the offices was restored on
20 October 1998 and all other affected areas of ST1 were connected with
permanent power on 22 October 1998.

18.122 There is insufficient evidence to determine who should be
responsible for this incident, whether it was HACTL or its contractor who
installed the ceiling bus-bars, or the power company.

Major Problems:

[40] FIDS Malfunctioning

18.123 Flight-related information is the driving force of operations
and of movement of passengers in all airports, particularly for any highly
sophisticated and busy airport like the one in CLK. After touching
down at the new airport, the arriving aircraft is directed to either a frontal
stand bordering PTB or a remote stand for parking and disembarkation of
passengers. The stand is allocated by the ACC based on flight schedules
and estimated time of arrival (ETA) obtained from the ATC tower
operated by CAD. With knowledge of the allocated stand, the RHO
serving the arriving aircraft can then proceed to the stand to disembark
passengers and unload baggage to be deposited on the reclaim lateral
assigned to the flight. At the stand, the plane can be serviced and
prepared for its departing flight at the scheduled time of departure.
Departing passengers check in with their baggage at either the check-in
desks on Level 7 of PTB, or at the In-Town Check-in facility at two of the
major AR stations. Checked-in passengers are notified of boarding gate
numbers, boarding and departure times. Their bags are injected into the
Baggage Handling System (BHS) which are automatically sorted to
departure laterals allocated according to flight schedules and aircraft
stands. With knowledge of aircraft stands, the relevant RHOs collect the
bags from the departure laterals and load them onto the aircraft.

604

18.124 All these steps are connected in the cyclical process of
arriving and departing flights. For each step, relevant information
should be available to the relevant operators in advance to give them time
for preparation and to maintain the timeliness of the entire process.
Delay in one step inevitably sets back the next. The information
required for operations can be provided by computerised means or by
more manual means of allocating aircraft stands on paper and relaying
flight-related information by conventional means of communication.

18.125 FIDS at the new airport is a highly computerised means of
generating, processing, disseminating and displaying flight-related
information. FIDS was designed with a high level of integration
between the various systems that supply and use flight information. The
ultimate user and owner of FIDS, as of most systems at the new airport, is
AAs AMD. G.E.C. (Hong Kong) Ltd. (GEC) was the main contractor
for the delivery of software and hardware. The FIDS software was
developed by GECs subcontractor, Electronic Data Systems Limited
(EDS).

18.126 In the eyes of the public, FIDS crashed or broke down
on AOD and had problems for about a week or so thereafter.
Specifically, meeters and greeters, and arriving and departing passengers
saw blank liquid crystal display (LCD) boards and FIDS monitors, or
incomplete, inaccurate or outdated information displayed. 137 out of
142 available LCD boards and 1913 out of 1952 available monitors
worked without interruption on AOD. While there were some hardware
problems and display server problems that affected the availability of
devices and the update of information displayed, the lack of reliable flight
information was mainly caused by problems with the FIDS software, as
discussed in this report.

18.127 For flight-related information to be displayed, operators and
systems must first generate it. Operations on the apron are at the heart
of airport operations and stand allocations generated by ACC on the TMS
for arriving (and therefore departing) aircraft are vital flight information.
TMS is a resource allocation software sharing a database with FIDS and
interfaced with FIDS and FIDS Man Machine Interface (MMI). TMS

605
was supplied by The Preston Group Pty Ltd (Preston), EDS
sub-contractor. Difficulties allocating stands for arriving aircraft in
ACC in the small hours on AOD resulted in delays in aircraft landing and
parking. This triggered a series of delays, which threw apron operations
and eventually the whole airport into a vicious cycle of delays. A
consequence of delayed allocation of stands that was visible by noon on
AOD was a congested apron, which became full from about noon to 5 pm
and again from 8 pm to 11 pm. Aircraft had to queue up for the next
available stand. RHOs work plans were thrown to the winds by the
changes in aircraft arrival times and stands, and they had to adjust
manpower and vehicles to meet the real situation. FIDS was unable to
provide RHOs with real-time information on arrival times and the
location of the aircraft, causing delay to the provision of service to the
arriving aircraft. At an emergency meeting at 4 pm on AOD, AA, and
RHOs agreed to set up a whiteboard at the Airport Emergency Centre
(AEC) displaying up-to-date flight-related information. This put
strain on RHOs resources who had to send its staff to AEC to obtain the
vital information to support its operation. Everyone tried to convey
information through Trunk Mobile Radio (TMR) or mobile phones.
However, they experienced difficulty in getting a channel or line due to
the unusually high demand on the TMR and mobile phones network.

18.128 The peculiar events in the ACC on AOD triggered the
situation of delay upon delay on the apron. Difficulties in allocating
stands started at around 1 and 2 am on AOD, when ACC operators were
unable to execute the flight swapping function on TMS. The
Commissioners have identified the main problem in this respect as the
lack of training on the part of ACC operators on the proper flight
swapping functionalities of TMS and lack of assistance present in the
ACC at that crucial moment, resulting from poor coordination within AA
and between GEC, EDS and Preston.

18.129 At about 10 pm the night before AOD, operators unfamiliar
with a system prompt for linking arriving and departing flights by
registration number mistakenly executed the link for ferry flights arriving
from Kai Tak. W34 Mr Peter Lindsay Derrick of Preston gave evidence
that he too was not familiar with this prompt. When flight movement
sheets were received at about 1 am from Cathay Pacific, ACC operators

606
started to carry out the required flight swapping. ACC operators were
not trained on all methods of flight linking and their progression, and
encountered difficulties executing the command on TMS and later Stand
Allocation System (SAS). SAS hung up and could not be used.
Assistance was late in coming, with W24 Ms Rita LEE Fung King of the
Information Technology (IT) Department of AA only arriving at ACC
about 6:30 am because she was uncontactable since 3 am, W34 Derrick
arrived at ACC at 12:30 pm, due to a six-hour delay for AA to obtain an
access permit for him and to escort him to ACC. W34 Derrick was able
to clear most flight swapping problems.

18.130 Meanwhile, ACC had resorted to manual allocation of stands
on paper, followed by entry and confirmation of the allocations into TMS
through FIDS MMI, which would enable FIDS to disseminate and
display the information on FIDS display devices. The main problem
W24 Lee and operators encountered in flight swapping and inputting and
confirming stands on TMS was the slow system response. There were
other problems that annoyed operators but did not hamper their
operations significantly, such as the intermittent shutdowns of the TMS
Gantt chart, and green bars on the TMS Gantt chart indicating invalid
ETAs that were earlier than STAs by more than 15 minutes, caused by the
incorrect ETAs obtained from the CAD radar tracker processor.

18.131 While trying to resolve flight swapping difficulties and
confirming stand allocations, W24 Lee would only confirm a stand when
this was requested urgently by Airport Operations Control Centre
(AOCC) or an airline on the phone. There was also a practice
inherited from Kai Tak of confirming a stand allocation only upon receipt
of ETA, which when combined with late receipt of ETA, meant late
display of allocations to RHOs and delayed allocation of gate and
boarding gate desks for airlines. These delays had obvious knock-on
effects for RHOs and airline operations resulting in the chaos on AOD.

18.132 Slow system response was the most serious problem that
plagued FIDS generally. There were conflicting versions from AA and
Preston on the response times, ranging from three seconds to 15 minutes.
W24 Lee said that there was slow response of workstations in the ACC
about 80% of the time on AOD. W26 Mrs Vivian CHEUNG Kar Fay

607
reported that from 8 am to 11 am on AOD, it took 20 to 25 minutes for
the FIDS workstation at AOCC to allocate a baggage reclaim belt.
System slowness persisted after AOD. Serious problems surfaced on
Day Five when the FIDS database experienced frequent locking and very
high CPU utilisation. Major system changes were effected that night to
solve the WDUM problem and locking in TMS. System performance
improved significantly after that. Measures were also taken to increase
memory in FIDS workstations in ACC, AOCC and Baggage Control
Room (BCR) in the first few days of AOD.

18.133 SAS hung up or crashed at about 2:30 am on AOD when
operators tried to do flight swapping because a swap to one pair of
departure and arrival flights caused transient illogicality to other pairs of
flights of the same aircraft, in that departure time was earlier than arrival
time. City University (City U) designed the system not to accept
illogical states so as to prevent operational error. City U explained that
the system could have been used for flight swapping if the operator had
adjusted the departure time thus removing the illogical state, before
carrying out the swap. City U carried out modifications to SAS after
AOD to permit input of illogical data.

18.134 To sum up, the specific causes for the deficiency of FIDS
were as follows:

(a) Software problems that were manifested in slow system
response time.

(b) Lack of comprehensive training on the part of ACC
operators on TMS flight linking functionality.

(c) Assistance was not readily available to the ACC operators
when they started to experience difficulties with flight
swapping.

(d) Practice of ACC operators to confirm stands only after ETA
was received which in some cases was 15 minutes before
landing of aircraft.

608
(e) Green bars on the TMS Gantt chart caused by ETAs from
CAD being earlier than STAs by more than 15 minutes,
though this was not a major problem.

Broad Causes and Parties Responsible

18.135 In Chapter 13, the Commissioners have analysed the
evidence presented to it on factors that contributed to the malfunction of
FIDS into five broad areas: compression of software development time,
insufficient software testing and rectification of software problems before
AOD, insufficient training or practice of operators on software
functionalities, lack of or late confirmation of stands and poor
communication and coordination (within AA, between AA and its
contractors and between contractors and subcontractors). Each cause
should be considered in combination with the other four broad causes, as
well as contributing factors and specific causes discussed in the events
that took place on AOD.

(a) Compression of software development time

18.136 The development of the FIDS software got off to a
dangerous start when discussions on detailed functional requirements, as
set out in the system segment specification (SSS) were prolonged for
an extraordinary 14 months. The Commissioners find that this loss was
the most crucial slippage and was caused by the late involvement of
AMD in the negotiations and also to the ambiguity of the particular
technical specifications (PTS) drafted by AAs consultants in the early
days, with which the SSS had to comply. AAs Project Committee
approved a payment of HK$89.7 million to GEC for delays and variations
up to 10 December 1997. It appears from the justification for this
payment that AA, and not GEC or EDS, must be responsible for this
crucial set back.

18.137 The outcome of clarification of AMDs requirements was an
agreement by AA and EDS that EDS would develop the software from
scratch some 17 months after the contract commenced, as software
already developed for other airports could not be modified to the extent
required by the SSS. The option to buy a ready made product developed

609
by other companies was briefly considered and dismissed. The new
timetable for delivery, testing and commissioning of builds of software
was aggressive and no further slippage could be afforded.

18.138 Delays continued in the delivery of builds and various tests
revealed many problem reports (PRs) that had to be resolved for FIDS
to work efficiently. In J une 1998, AA was aware of at least 38 major
operational PRs that would affect the efficiency of FIDS. As far as
possible they cited workarounds for each PR, but some PRs involved
system bugs that AA and EDS had little time to cure. W44 Mr Chern
Heed told the Commission that there were new developments to FIDS
and TMS functionality and workarounds right up to a few days before
AOD and that TMS was usable only three or four days before AOD.

18.139 From the evidence, it appears to the Commissioners that on
AOD, the new airport was dependent on a critical system that was at best
workable on workarounds. At least, this put operators under additional
work pressure, and it certainly created risks that the system might suffer
significant failures. The Commissioners are of the opinion that
compression of software development time was the most fundamental
and significant cause for the problems encountered with FIDS, as it
forced testing, problem resolution and training for operators to be
severely compromised. FIDS should have been completely integrated,
tested and stable for a suitable period of time before AOD to enable
training and familiarisation for operators. The responsibility for
slippage in the development of FIDS from the end of 1997 to AOD lies
with both AA and EDS, though it remains a contractual matter between
AA and GEC and between GEC and EDS.

(b) Insufficient testing and rectification of software errors before AOD

18.140 The Commissioners find that FIDS responded slowly,
against PTS requirements of initial response to operator input of 0.5
second and final response to 90% of updates of 2 seconds. On the
evidence which was corroborated by expert opinion, slow response was
caused by problems with the WDUM process and a small pool memory in
the Oracle database. The two problems with WDUM, excessive CPU
utilisation and deadlocks had been identified before AOD but obviously

610
not resolved.

18.141 The Commissioners note that AA should have engaged
assistance from Oracle consultants on resolving problems with the Oracle
database much earlier than they did on 3 J uly 1998. These consultants
were able to identify and resolve the problems after AOD, and no doubt
could have done so before AOD.

18.142 The Commissioners also note that sufficient testing and
rectification of software errors revealed in testing was indispensable for
an efficient FIDS on AOD. From the evidence, it appears that testing
was compromised and often sacrificed at the altar of meeting AOD, with
the consequence that problems that could well have been avoided with
proper testing and rectification of errors, were encountered on AOD.
The parties were aware of the kinds of tests that should have been carried
out before the system could properly be said to be ready, but their
judgement and willingness to forego these tests in the face of an
imminent AOD were affected by the compression of time available to
them.

18.143 One of the most important tests to prepare FIDS for live
operations was the stress and load test contemplated in late J une 1998.
However, this test was deferred to after AOD, because the time remaining
to AOD would better be used to rectify the problems already identified
and FIDS was not stable enough to receive such a test. Witnesses from
AA, EDS and GEC and the Commissions experts all agreed that a stress
test would have thrown up problems with FIDS relevant to AOD. AA,
GEC and EDS are responsible for this decision.

18.144 Bearing in mind that any test is only as good as the measures
taken to resolve the PRs revealed by it, and considering the limited time
EDS and AA had to solve PRs before AOD, it is uncertain if a stress test
would have saved the day. Yet, the Commissioners note that it was
industry practice to carry out the test before operations. This highlights
the serious risks that AA faced with their operation systems in the
build-up to AOD and the dire need for a global contingency plan.

18.146. The Commissioners are of the opinion that GEC and EDS as

611
contractor and sub-contractor for the supply of FIDS are both responsible
for the problems encountered with FIDS, including problems with the
Oracle database, with EDS being mostly to blame. As between EDS
and Preston, the Commissioners find it difficult to decide on
responsibility. The evidence is not clear whether the software problems
of FIDS was attributable to those relating to the FIDS software developed
by EDS or the TMS software developed and supplied by Preston.

(c ) Insufficient training and practice of operators on software
functionality

18.147 The Commissioners find that the problems encountered in
the ACC with flight swapping and entering data and confirmations into
TMS were problems with TMS/FIDS and were not the fault of ACC
operators. The Commissioners are not satisfied on the evidence that any
of the problems in the ACC, BCR or at Common User Terminal
Equipment (CUTE) workstations were caused by the inexperience or
error of operators. The ACC operators were experienced, but
unfortunately they were not trained as to the implications of the
progression of the methods of flight linking and were not familiar with
the prompt linking flights by registration numbers.

18.148 The Commissioners find that operators were trained on old
versions of software that were subsequently revised or had functionality
added to them. This rendered training less effective, and resulted in
some duplication of time and effort in training. The Commissioners find
that the inadequate training was a major contributing factor to problems
on AOD. Operators, through no fault of theirs, needed to be trained to
know the functionalities and workarounds. The Commissioners find
that the inadequate training was caused by the compression of time
caused by continued slippage in the development of FIDS. AA must be
primarily responsible for the resultant inadequate training, while some of
the responsibility may be apportioned to GEC, EDS and Preston for not
providing all functionalities in training.

18.149 Finally, the Commissioners find that there was a lack of
coordination or understanding between AA and City U that SAS was

612
programmed not to accept illogical departure and arrival data, causing
further difficulties to ACC operators. There is, however, insufficient
evidence for the Commission to decide the apportionment of
responsibility between AA and City U on this issue.

(d) Lack of or late confirmation of stands

18.150 It is clear to the Commission that the failure to promptly
confirm stand allocations resulted in delayed information to users and
operators, slowing down their operations. ACC operators could not
confirm allocations promptly because they were hampered by difficulties
with TMS and because of the practice of not confirming allocations until
ETA was received.

18.151 W24 Lee gave evidence that she was so occupied with
resolving problems with flight swapping since she arrived at ACC at 6:30
am on AOD that she would confirm allocations made manually by the
other operators only when she received an urgent request. Even then,
not all confirmations successfully passed through the first time, requiring
her to unconfirm and reconfirm. This was not quick enough for the
situation on the apron, which was quickly filling up. The same findings
on the inefficiencies of FIDS and TMS apply here.

18.152 Delays in confirmation were also the result of the practice of
confirming an allocation only after ETA was received, compounded by
late receipt of ETA. AA and in particular W23 Mr Alan LAM Tai Chi
should be responsible for adopting this practice, which was changed after
AOD.

(e) Lack of communication and coordination

Within AA

18.153 The Commissioners find that AMDs concerns about the
delay in the software programme and compression of training and testing
were not fully taken into account and not properly addressed by Project
Division (PD), a reflection of the general lack of coordination between
PD and AMD as documented in ADSCOM documents and the report of

613
Booz-Allen & Hamilton.

18.154 It appears that W44 Heed was not informed of the decision
by PD and IT department to postpone the stress and load test.

18.155 Poor coordination within AA was also the cause of the
unavailability of expert and other IT support for operators in the ACC in
the early morning of AOD. This was especially condemnable given the
prior knowledge of all concerned that TMS was not stable and that a
method of operations using SAS that doubled the effort required would
be adopted. The delay in issuing W34 Derrick with a permit to access
ACC was inexcusable and regrettable.

18.156 Better coordination and planning might have procured the
assistance of Oracle Systems Hong Kong Ltd consultants early enough to
save the day.

18.157 Poor coordination was also evident in the contingency plan
consisting of whiteboards, which was not clear to operators as to when
these should be employed. This meant that when FIDS could not be
relied on for information, the operators of the airport community had no
clear knowledge as to when and where the whiteboards would be put up,
and what they were required to do.

18.158 The Commissioners find that W3 Dr Henry Duane
Townsend, the Chief Executive Officer (CEO) of AA, and the rest of
AA management and relevant departments and divisions are to be
blamed.

Between AA and other parties

18.159 The Commissioners find that AA is to be blamed for not
informing CAD that it would use ETA from the radar tracker without
prior authorisation or screening, and thus for the problems caused by the
invalid ETA, causing green bars on the TMS Gantt chart.

18.160 The Commissioners also find that AA must take the blame
for not consulting EDS and Preston before AOD on the merits of using

614
TMS only to input stands allocated, and not to use TMS as an
optimisation stand allocation tool.

18.161 The lack of communication between AA and City U also
contributed to the development of SAS not to accept illogical states,
resulting in the system hanging up in the morning of AOD, though the
Commission do not find sufficient evidence to apportion responsibility
between the two.

Between GEC, EDS and Preston

18.162 GEC is responsible for not communicating with EDS and
thus misrepresenting to AA that it would take only a short time to revert
to development of standalone builds.

18.163 The Commissioners find that both EDS and Preston are
responsible for not ensuring that ACC operators were aware of the
implications and the correct method of usage of the prompt linking flights
by registration numbers, resulting in problems for ACC operators in the
early morning of AOD, which triggered a series of delays on the apron
and in the airport in general. On the evidence available, however, the
Commissioners are not able to apportion blame between them.

Other matters

18.164 The Commissioners find that GEC as main contractor must
be responsible to AA for the defective monitors and LCD boards, while
AA is responsible for cable problems that caused display devices to
malfunction.

18.165 To the public, AA is responsible for failing to ensure that
FIDS worked for smooth and efficient airport operations on AOD and the
week after. As a result, the efficient movement of passengers was not
achieved and airlines and service providers were seriously affected in
their operations.

18.166 120 of the 2,057 monitors required under the contract were
replaced in the three weeks after AOD. According to W22 Mr Edward

615
George Hobhouse, the expected failure rate of monitors was 10% a year.

18.167 W26 Cheung and W28 Mr Anders YUEN Hon Sing testified
that FIDS performed efficiently and stably from about a week after AOD.
Representatives of Cathay Pacific gave evidence that information from
FIDS was largely accurate by Day Four to Five and that they regained
confidence in FIDS a week after AOD. RHOs gave similar evidence.
There was apparently nothing fundamentally wrong with FIDS, and it has
worked efficiently and smoothly since late September 1998.

[41] CHS Malfunctioning

18.168 Cargo operation is one of the critical elements in the
operation of the new airport and the efficient and speedy handling of air
cargo has played a vital role in maintaining the vibrant economic growth
of Hong Kong. Section 6(2) of the Airport Authority Ordinance
expressly provides that AA shall have regard to the safe and efficient
movement of air cargo. HACTL, the only cargo terminal operator
(CTO) at Kai Tak, had established itself through the last two decades as
the largest and one of the most efficient cargo handling operators
(CHOs) in the world. The US$1 billion ST1 at the new airport is one
of the worlds most sophisticated cargo terminals. In CLK, cargo
operation was shared between HACTL and Asia Airfreight Terminal
Company Limited (AAT), who were assessed to cater for about 80%
and 20% of the expected cargo capacity of the new airport respectively.
Due to their importance, the readiness of the two cargo handling facilities
had always been considered by AA and Government as a critical AOR
issue.

18.169 In the HACTL franchise, HACTL agreed to provide 75%
throughput capacity out of the full capacity of 2,400,000 tonnes (not
including the annual capacity of 200,000 tonnes of the Express Centre) by
18 August 1998. This throughput capacity was raised by HACTL from
50% for April 1998 to 75% on AOD. Although the contractual date was
over a month after AOD, HACTL had promised to use its best endeavours
to get ST1 and the CHS ready for operation on AOD to provide a certain
percentage of its yearly throughput capacity.

616
18.170 While AA and Government were always concerned about the
constant slippage in the construction programme of ST1, there was no
suspicion that CHS had any problems. In fact, HACTL never reported
any problem with CHS. Despite the delay, HACTL obtained a TOP for
ST1 on 3 J uly 1998.

18.171 On AOD, cargo operation at the new airport can be described
as nothing but chaotic. Ramp space at the northern boundary of both
ST1 and AATs building was full of cargo. For AAT, whose system is
much less sophisticated than that in ST1, the difficulties experienced and
the impact thus caused were relatively small and manageable than
HACTLs. That is not to say that there was nothing wrong with AATs
operation. On AOD and a few days thereafter, an enormous backlog of
cargo was built up which led to a heavily congested working environment,
both within the AATs terminal and on the ramp interfacing with it. The
backlog seriously hampered the processing of daily inbound cargo
leading to a snowballing of unprocessed cargo. One of the main causes
of AATs difficulties on AOD was the inadequate coordination between
AAT and RHOs in the hand-over of cargo from RHOs to AAT.
Furthermore, AAT staff was not too familiar with handling live loads of
cargo in the new environment and using new equipment because they did
not have adequate training. For this AAT should be responsible,
although on the inadequate coordination, RHOs should share a small part
of the responsibility.

18.172 A detailed description of ST1s CHS is set out in Chapters 11
and 14. Suffice to say that it consisted of 5 levels, namely, Level 5
Community System for Air Cargo (COSAC 2), Level 4 the Resources
Management System, Level 3 the Logistic Control System (LCS),
Level 2 the Programmable Logic Controller (PLC) and Level 1 the
mechatronics of CHS which perform the work of cargo handling. The
main components on Level 1 are the Container Storage System (CSS)
and the Box Storage System (BSS). Both CSS and BSS have
stacker-cranes which pick up cargo from automatic transfer vehicles
(ATVs) or conveyor belts and lift it to the assigned compartment for
storage and retrieve it from the compartment whenever needed. The
equipment are mechanical, electrical and electronic and are called
mechatronics. Although LCS is a single computer system, it is linked

617
separately to CSS and BSS, giving orders through PLC for the two
systems to perform work independently or collectively. CSS was built
by Mannesmann Dematic AG Systeme (Demag) on the east and west
sides of ST1, and on AOD, the whole of the west side, namely W1, W2
and W3 were to be used together with a part of the east side, E1. BSS,
on the other hand, was built by Murata Machinery (HK) Ltd (Murata)
and is divided into north and south of ST1. The design of CSS and BSS
is modular in nature, which means that each portion of the systems is able
to work independently so that failure of one portion would have little
effect on the capability of the others.

18.173 A chronology of the problems with ST1 and CHS is set out
in paragraph 11.10 of Chapter 11. In summary, HACTL experienced
difficulty with the operation of ST1 in the early morning on AOD.
Operation of CSS was turned into manual mode. The manual mode of
operation created inventory inaccuracies for the upper levels of CHS due
to operators keying in inaccurate information of the location of the unit
load devices (ULDs), or their delay or omission in inputting the data.
There was a backlog of unprocessed cargo as a result of the slow
operation of the system in manual mode. The manual mode also
adulterated the inventory of and record of locations of ULDs,
necessitating a physical check of the inventory. In the small hours of
Day Two, when HACTL conducted an inventory check, the inventory
records were inadvertently deleted. This gave rise to serious suspicion
that there was something wrong with the systems. At about 3 pm on
Day Two, HACTL announced a 24-hour embargo on export bulk cargo
and import cargo on passenger flights except urgent items. On Day
Three, HACTL imposed a 48-hour embargo except urgent items. On
Day Four, a 9-day moratorium on all cargo on all aircraft (except inbound
and outbound urgent items) was announced. It was also announced that
cargo at ST1 would be moved to Terminal 2 for storage and distribution.
On 16 J uly, HACTL announced a four-phase recovery programme for air
cargo services using both ST1 and Terminal 2 in Kai Tak. By 24 August
1998, HACTL was handling all cargo at ST1, some 8 days ahead of the
recovery programme. Details of the recovery programme are
summarised in paragraph 11.13 of Chapter 11.

18.174 Prior to 15 J uly 1998, HACTL had announced having

618
computer system difficulties or software problems in its press releases
and public statements. However, since 15 J uly 1998, HACTL has
changed its emphasis to electrical and mechanical faults caused by the
environment and that the computer software problem was remarkably
downplayed.

18.175 HACTL attributed the causes of the problems to a number of
factors, such as the delay in the completion of construction works, dust
contamination and shortage of dollies. Few of the alleged causes could
be attributed to HACTLs own fault. In support of its case, HACTL
appointed two experts, namely W52 Mr Max William Nimmo and W53
Mr J erome J oseph J r. Day who produced a report in this inquiry. They
were of the opinion that there was nothing wrong with CHS and that the
throughput capacity of the system was available and capable to deal with
the cargo load on AOD. In short, they were of the view that as CHS was
operating without any problem on the last three days prior to AOD and
there was sufficient throughput capacity based on the figures obtained for
this 3-day period, there was no reason why it could not cope with the live
load operation on AOD. HACTLs experts opined that the increased
operation of CHS in manual mode as opposed to automatic mode was due
to the operators wrong perception that LCS-CSS was operating slowly
when in fact it was not. Heavy pressure faced by the operators on AOD
contributed to their perceived slowness of the system. Various external
and internal causes of the problems were identified by them but they
came to the view that the breakdown of ST1 was mainly caused by two
factors, namely, (i) the ramp confusion and the unfamiliarity or
non-compliance of the procedures by RHOs; and (ii) the lack of flight
information from the Flight Data Display System (FDDS) or the Flight
Display Data Feed Services (FDDFS).

18.176 The Commissioners consider the opinion of HACTLs
experts to be flawed. The alleged two main factors were in respect of
fields not within their professed expertise. Further, rather than basing
their opinions on the evidence already presented to the Commission, W52
Nimmo and W53 Day based their report on facts that they were told by
HACTL staff but those facts were not supplied to the Commission and
had not been tested before the Commission. It is therefore risky to rely
on anything they expressed as their opinions unless it is clearly proved

619
that their views are supported by the evidence presented to the
Commission, and that such views are within their fields of expertise.

18.177 Despite the apparent deficiency in their opinion, the
Commissioners had, as a matter of caution, examined W52 Nimmo and
W53 Days views on the factors that were said to have caused the
breakdown of ST1 against reasonableness and the facts as found by
Commission. Contrary to the view of HACTLs experts, the evidence
shows that the ramp confusion and chaos on AOD were caused by, and
not causing, the breakdown of CHS. As to the second main cause
identified by the two experts, namely, the lack of correct and complete
flight information from FDDS or FDDFS, this simply could not stand in
the light of the evidence before the Commission. Another reason that
was put forward was the late delivery of pre-manifests by airlines. The
Commissioners take the view that even if this did occur, the impact could
not have been that substantial as to become a major cause of the
breakdown. It certainly would not have caused the slow response to
CHS, which was the main cause for the HACTLs staff to go for manual
operation which eventually led to ST1s breakdown.

18.178 Dust had been maintained by HACTL as one of the culprits
for the breakdown of CHS on AOD. It was first presented as a major
problem on 15 J uly when HACTL began to downplay the computer
software problem. It was stressed as a major problem by almost all
employees of HACTL who testified before the Commission, except the
Managing Director, W7 Mr Anthony Crowley Charter. HACTL alleged
that dust together with the presence of water at ST1 blocked and seriously
affected the 15,000 highly sensitive sensors and reflectors installed for
the operation of the mechatronics of CHS on Level 1, which was the most
important element in the handling of cargo. The witnesses from
HACTL, notably W11 Mr LEUNG Shi Min and W10 Mr HO Yiu Wing,
who were working in ST1 on AOD told the Commission that a substantial
cause of the problems with CHS on AOD was dust. However, from all
the public statements made by HACTL up to 15 J uly 1998, that alleged
substantial cause was not mentioned. The evidence of W9 Mr Gernot
Werner of Demag and W16 Mr Hiroshi NAKAMURA, W17 Mr
Tomonobu SAEKI and W18 Mr Shin YAMASHITA of Murata did not
support the dust theory either.

620

18.179 On the contrary, their evidence suggested that the problem
lay with the software of the LCS. Upon HACTLs instructions, on about
18 J uly 1998, Demag cut the link between PLC and LCS that enabled
CSS to be operated on an off-line mode. Murata, on the other hand, had
received three instructions from HACTL on 16, 18 and 21 J uly 1998 to
modify the interface between LCS and PLC regarding BSS resulting in
BSS going back to full operation since 13 August 1998.

18.180 When W2 Mr YEUNG Kwok Keung, the Deputy Managing
Director of HACTL, gave evidence before the inquiry, he admitted that
the trouble created by dust and contamination were appreciated by him as
early as 21 April 1998 and that teams of engineers were deployed to deal
with the cleaning of the mechatronics of CHS. If in fact dust was the
culprit on AOD and was noticed by W11 Leung and four engineers, it
would be inconceivable why the faults summaries regarding problems
experienced by CHS on AOD and 7 J uly that were prepared by the
Engineering Department a week or two afterwards hardly identified it.
The Commissioners do not believe that dust did cause the amount of
problems facing ST1 on AOD, which according to W11 Leungs evidence
might be as much as 30% or 50% of the problems encountered by ST1 on
AOD. Even W7 Charter stated that the dust problem was overplayed.
The Commissioners also rejected the allegations of W2 Yeung and W11
Leung that the severity of the dust problem could not have been foreseen
prior to AOD.

18.181 The Commissioners regret to say that much time and costs
had been spent in this inquiry on dust, which, towards the end of
hearing HACTLs evidence, HACTL conceded to be not a major factor.
When its two experts gave evidence, they said that this problem was
manageable.

18.182 Various factors had been put forward by HACTL or its
experts as being contributed to the chaos. These factors have been dealt
with in Chapter 14 in this report. It is interesting to note that few of
these problems could be said to be the responsibility of HACTL. When
W12 Mr J ohnnie WONG Tai Wah, W13 Mr Peter PANG Tai Hing, W14
Ms Violet CHAN Man Har and W15 Mr Daniel LAM Yuen Hi, all from

621
HACTL gave evidence in this inquiry, W15 Lam, Operations Computer
Project Manager, said that he had tested the throughput of CSS but not
the throughput of BSS. The test would be used for assessing how many
units of cargo could be handled by CSS and BSS in one hour. He
merely relied on the test that had been conducted by Murata
notwithstanding that the latter was not an integrated test involving the
higher levels.

18.183 Having considered the evidence, the Commissioners arrive
at the conclusion that one of the major causes for the breakdown of ST1
was that CHS, especially BSS integrated with the higher levels, was not
sufficiently tested before AOD due to the compression of the time
required for testing and commissioning of such a sophisticated and
complex cargo handling system. This was one of the major reasons why
on AOD, BSS experienced a slow response in dealing with many cargo in
live load operation. In a press release made by HACTL on 8 J uly 1998
announcing an extension of the 24-hour embargo for another 48 hours, it
was stated that:

Since our announcement yesterday of temporary measures to
relieve SuperTerminal 1 from the pressures it was under, we have
now had time to more closely analyse problems

allowing our engineers and contractors adequate time to
rectify current hardware and software problems with our Box
Cargo Storage Systems.

18.184 Looking at the evidence, such as the work performed by
Demag in de-linking PLC from LCS regarding CSS to enable CSS to
operate in an off-line mode as well as Muratas work to the interface
between LCS and PLC of BSS, the Commissioners find more probable
than not that one of the main causes for ST1s paralysis was that there
was something wrong either with the software of LCS or with the
interfaces between LCS and CSS and between LCS and BSS.

18.185 Another main cause for the ST1s breakdown was the
insufficient training and unfamiliarity of HACTLs operation staff with
CHS, particularly the operators of CSS and BSS. This is mainly caused

622
by the delays of the construction works. HACTLs experts, W52
Nimmo and W53 Day, effectively agreed to this. In support of their
theory that there was perceived slowness of the system by the operators,
the two experts stated that the operators working on the floor of ST1 were
not well trained or familiar with operating CSS or BSS. Otherwise, the
operators would have known that LCS would only commence a process
of cargo movement until the entire route was clear and therefore would
not have perceived the system to be slow. W52 Nimmo and W53 Day
also attributed the high level of operators error in data entry of inventory
record when the CHS was in manual mode to the lack of training and
unfamiliarity of HACTLs operators.

18.186 W7 Charter, in his evidence, hinted that HACTL had been
operating under pressure to make ST1 ready for handling cargo on AOD,
which was decided without consulting it and despite the contractual
completion date of 18 August 1998. HACTL also alleged that the 6 J uly
date was cast in stone and was not moveable, even if HACTL were not
ready. It also intimated that a soft opening was requested by it but
rejected. The evidence however pointed to the contrary. After the
announcement of AOD in J anuary 1998, HACTL volunteered the
information that it would be able to process a throughput of 75% on AOD
instead of 50% as it had previously promised to achieve by April 1998.
The feeling of pressure was also inconsistent with the assurances given
by HACTL continuously right up to the beginning of J uly 1998 that ST1
would be ready for operation on AOD. The Commissioners are of the
view that the root of the problem was not so much HACTLs belief that
AOD could not be deferred or soft opening was absolutely unavailable.
Rather, it was HACTLs over confidence with its brainchild, ie, the
computer systems of CHS and with its ability to have ST1 ready by AOD
that resulted in the chaos at ST1.

18.187 For the delay in the completion of the construction works,
which resulted in delay in the installation of the CHS, the responsibility
should lie between HACTL and Gammon Paul Y J oint Venture (GPY).
Blames were put by one on the other for the delay. However, the time
available in this inquiry does not permit the Commission to investigate
into such complex construction disputes between HACTL and GPY.

623
18.188 Despite the change in HACTLs emphasis on the
contamination problem, the Commissioners consider that it must have
caused a small extent of interruption to CHS. Similar to problem of the
delay in the construction and related works, which is closely related to the
problem caused by contamination, the Commissioners are unable to reach
a conclusion save as to say that either HACTL or GPY or both should be
responsible for the contamination.

18.189 The Commissioners do not accept that the increase in the
numbers of CTOs and RHOs at the new airport contributed to the
paralysis of ST1 on AOD. On the confusion with the procedures in the
handing over of cargo between HACTL and RHOs on AOD, the
Commissioners opine that it was the problem of CHS itself that spilled
over to affect the operations of RHOs. For this, RHOs should not be
held responsible. There was a shortage of dollies for delivery of cargo
on AOD. However, the Commissioners consider that this was an effect
rather than a cause. The slow response of CHS led HACTLs operators
to switch onto the manual mode which slowed down the whole process.
As a result, the hand-over procedures for cargo can hardly be followed.
It resulted in dollies being detained for much longer than the agreed
turn-around time of 30 minutes. HACTL should be responsible for the
shortage of dollies.

18.190 The Commissioners find that FDDS or Flight Display Data
Feed Services (FDDFS) not providing flight-related information to ST1
as expected or at all did, to a minor extent, cause trouble or
inconvenience in the operation of cargo handling. For this, AA should
be mainly responsible for failing to provide the necessary flight-related
information through Airport Operational Database (AODB) from which
FDDS and FDDFS drew the information.

18.191 HACTL should also be responsible for the delay in the
testing and commissioning of the machinery and HACTLs own systems
and in the training of its staff for operating the machinery and systems.
There was an error of judgment on HACTLs part that despite the
shortage of time, all the machinery and systems would have been
sufficiently tested and would face little problem when they were
employed to work together in the actual operation on AOD.

624

18.192 HACTLs confidence in its computer system as its brainchild
and the under-estimation of the significance of having the software tested
thoroughly when integrated with BSS and CSS was manifested in
HACTL not having any contingency plans for the failure of CHS. The
Commissioners do not accept that HACTL had made any risk assessment
or any real and workable contingency plan.

18.193 Having considered the evidence in this inquiry, and having
examined the witnesses testimonies, it is clear to the Commissioners as
to the major causes of the chaos on Day One. On AOD, about 2,000
containers had been transferred from Kai Tak to ST1. In addition, cargo
arriving from inbound flights started to accumulate. HACTLs operation
staff began to notice slow response with both CSS and BSS due mainly to
LCS not operating PLC and the mechatronics smoothly. The slow
response of CHS led HACTLs operators to switch into the manual mode,
instead of the pre-set automatic mode. Although this helped processing
of the cargo, it was still much slower than the automatic process. As a
result, the procedures of hand-over of cargo that had been agreed between
HACTL and the RHOs could hardly be followed, and inbound cargo was
left by the RHOs on dollies outside the airside at the northern part of ST1.
The dollies were detained for much longer than the agreed turn-around
time of 30 minutes, and as a result, there was a shortage and the RHOs
placed the goods on the dollies onto the ground in order to retrieve the
dollies for other inbound cargo. Unprocessed cargo started to build up
outside ST1 during the course of AOD and the northern part of ST1 and
the surrounding area was congested with an enormous number of cargo.
The backlog of cargo, in turn, seriously hampered the processing of cargo.
Insufficient training of the operators, particularly in the manual mode,
resulted in human errors in not updating LCS or updating it incorrectly.
This act or omission caused the inventory to be adulterated, so much so
eventually that there had to be a physical check of the inventory. During
the course of the physical check in the small hours of AOD, a utility
programme was inadvertently switched on which erased the inventory.
This gave rise to grave concern to HACTL as it had to find out the reason
before there was any meaningful rebuilding of the inventory. At the
same time, investigation had to be made as to why LCS was not operating
as smoothly as expected. All these resulted in the 24-hour embargo

625
announced on 7 J uly, which was effectively extended for 48 hours on 8
J uly and 9 days on 9 J uly, so that the cargo at ST1 could be cleared from
CHS and moved to Kai Tak for processing. During the period of the
embargo, the cargo was removed out of CSS and BSS, the equipment was
cleaned, Demag was instructed to cut the link between LCS on the one
hand and PLC and mechatronics of CSS on the other. Murata was
instructed to do some improvement work on the LCS-BSS interface.
Thereafter, CSS and BSS could be operated smoothly in an off-line or
manual mode. In the meantime, HACTL was debugging or enhancing
LCS and the software of the higher levels of CHS. HACTL announced
a four-phase recovery programme on 16 J uly 1998, during which HACTL
took the opportunity to test and commission its computer system. Cargo
was processed at both Kai Tak and the new airport during recovery
leading to the resumption of the full cargo handling process at ST1 on 24
August 1998.

18.194 Under the HACTL franchise, HACTL is not under obligation
to provide any particular capacity by AOD for the contract only obliged
HACTL to provide by 18 August 1998 a cargo handling throughput of
75% of the full ST1 capacity, ie, 5,000 tonnes of cargo a day. Whether it
is a matter of goodwill or a gentlemens agreement, HACTL promised to
use its best endeavour to be ready by 75% capacity on AOD. The
Commissioners have no trouble finding that HACTL did use its best
efforts in the circumstances. However, the question was HACTL
represented to AA and Government that ST1 would be ready to produce
75% of its throughput capacity on AOD. This representation was relied
on by AA and Government whom might have been induced in doing so
by the success and reputation of HACTL as the top CHO in the world.
The events on AOD and the days after have proved the representation to
be ill conceived and incorrect. Had HACTL maintained its contractual
position that it would only be 75% ready on 18 August 1998 and not
earlier, Government would never have made the decision to open the
airport on 6 J uly 1998. A good example is the case of AR where MTRC
represented to ADSCOM that AR would not be ready before the
contractual date in J une 1998. Government did not insist on AR
opening before J une and the date for opening the new airport was
postponed from April to 6 J uly 1998 in J anuary 1998.

626
18.195 The Commissioners conclude that HACTL is responsible for
giving the false sense of security to AA and Government that it was ready
to operate on AOD. It would not be fair for HACTL to cling to the
contractual terms to say that it is not responsible for not being ready on
AOD. Even though this responsibility arose out of goodwill and a mere
gentlemens agreement without any contractual liability, the
Commissioners think that leading AA and in particular Government to
reach the decision on AOD and not to alter that decision is culpable, and
HACTL must fairly be held responsible for that area of the process of the
decision-making and thereafter for either failing to render ST1 ready to
deal with the expected tonnage of cargo on AOD as it had promised or
failing to strive for a deferment of AOD or to seek a soft opening
timeously.

18.196 To summarise, the Commissioners find, on the balance of
probabilities, that the following parties are responsible for the breakdown
of ST1 on AOD and in the period of about a month thereafter:

(a) HACTL is responsible for giving the assurances to AA and
Government that ST1 would be ready to provide 75% of its
throughput capacity on AOD.

(b) The main causes for the breakdown of ST1 were (i) the
faults with CHS which resulted in the inefficiency of LCS in
controlling and operating PLC and the mechatronics, (ii) the
insufficient testing of CHS in fully integrated mode, and (iii)
the insufficient training and unfamiliarity of HACTLs
operation staff with operating CSS and BSS in manual mode;
and for all these HACTL is solely responsible.

(c) Either HACTL or GPY or both are responsible for the delay
in the construction works at ST1.

(d) Either HACTL or GPY or both are responsible for the delay
caused to the installation of the machinery and systems at
ST1 and in the testing and commissioning of such machinery
and systems.

627
(e) HACTL knew of the delays in (c) and (d) above, and is
responsible for under-estimating their effects on the
readiness of ST1 to operate efficiently on AOD.

(f) Contamination of the environment on AOD was very minor,
and would have posed little difficulty to HACTL in the
operation of its CHS.

(g) Contamination of the environment, anyhow, was known to
HACTL as early as late April 1998, and HACTL is
responsible for not sufficiently clearing the environment for
the proper and efficient operation of CHS.

(h) The circumstances of there being three RHOs and two CTOs
were known to HACTL long before AOD, and the RHOs
involvement with cargo handling could hardly be described
as an appreciable cause for the breakdown of HACTL.

(i) The ramp chaos and alleged insufficiency of dollies were
consequences of the slow response of CHS in processing
cargo and not the causes of the slow response.

(j) The failure of FDDS or FDDFS (for which AA and others
are responsible) also would not have been a serious threat to
the efficient operation of CHS, as HACTL could have used a
few employees to obtain the necessary flight information.

(k) The late delivery of pre-manifests by airlines and the new
Customs and Excise Department customs clearance
procedures would cause some inconvenience to HACTL but
did not contribute to the breakdown of ST1.

18.197 One other matter should be mentioned. Over 10 solid days
had been spent in the hearing of the Commission for seeking facts and
reasons relating to the question of dust, which had been raised as a major
problem by HACTL for ST1s breakdown on AOD. Had dust been
raised as a minor factor contributing to the breakdown, much less effort
and time would have been spent.

628

18.198 To deal with the backlog of cargo that congested its terminal
and the area outside the interface with the ramp, AAT made arrangements
with the nearby Airport Freight Forwarding Centre (AFFC) to use the
latter for breakdown, storage and collection of the backlog cargo, which
was cleared by 13 August 1998. Since the use of AFFC, the severe
congestion at AATs terminal started to abate.

18.199 HACTLs four-phase recovery programme was completed
on 24 August 1998, and ST1 has appeared to have operated normally.

[42] Baggage Handling Chaos

18.200 The BHS is an important system at the new airport. It
affects flight departures and the time in which arriving passengers can
collect their baggage. As the baggage handling chaos on AOD and the
few days afterwards show, problems with BHS can have a huge
ramification that can affect the efficient operation of the new airport.
The chaos had a direct and significant impact on passengers, arriving or
departing, causing delays and inconvenience to them.

18.201 There was a serious problem in the handling of baggage on
AOD. Some 6,000 to 10,000 of 20,000 departure and transfer bags
processed on AOD missed their flights. Some departure bags were
loaded onto flights late, adding to delays in flights departing. Departure
baggage handling started getting unmanageable by about 9 am on AOD.

18.202 On the first week of AOD, arrival passengers experienced
significant delays in reclaiming their baggage. From Days Three to
Seven, arrival passengers had to wait for an average of 1 hour 41 minutes
to collect their bags. There was also some confusion as to where bags
were to be picked up. Passengers were inconvenienced and the
standards previously achieved at Kai Tak were not met at the new airport
until about the second week. The effect of the baggage handling
problem was compounded by the other problems happening on that day,
in particular, the FIDS problem. Flights were delayed, and there was
confusion over stand and gate allocation and parking of planes. There
were also problems in the allocation of reclaim carousels at the Baggage

629
Reclaim Hall (BRH) and in the display of carousel numbers.

18.203 It is clear that the problems were caused by a number of
separate and discrete matters, including human error. Some problems
were the effect of other problems encountered in airport operations, eg,
with FIDS and TMR. Each problem had a significant impact if not by
itself, certainly when combined with the other problems encountered. In
Chapters 12 and 15, the Commission has identified 19 factors leading to
the chaos under [BHS 1] to [BHS 19]. Not one single factor, by itself,
can be said to have caused the chaos. However, it is clear that baggage
operation was seriously hampered by the large number of problem bags
on Day One (about 30% of all the bags processed). There were also
some 500 system stoppages on AOD, one even lasted for a few hours.
Airline staff had to transfer bags from one conveyor belt to another.
Stoppages in turn led to the accumulation of more late and problem
baggage. As the problem bags became unmanageable, BHS started to
die back up to the infeed points. Accordingly, system stoppages and
problem baggage caused a vicious cycle which eventually led to extreme
delays in baggage handling. Due to the problem of FIDS, the resources
of the three RHOs were fully stretched which affected their ability to deal
with all the problems that arose on AOD. There was a lack of
familiarity of the staff of airlines, RHOs and Swire Engineering Services
Limited (SESL) with baggage handling procedures, with BHS and the
working environment. Because of the sheer volume of problem bags
and difficulties faced by RHOs, problem bags were not sorted and dealt
with in time. Many of the departure and transfer bags missed their
flights.

18.204 It is important not to lose sight of the fact that the
inefficiency of FIDS on AOD drained heavily on the resources of RHOs
in obtaining the necessary stand and time information, resulting in delay
in their baggage handling activities. Had the necessary flight
information been available, RHOs resources could have planned and
allocated their resources more efficiently which might have alleviated or
even eliminated the baggage chaos.

630
18.205 The 19 problems that have been identified by the
Commission in Chapters 12 and 15 and the parties responsible for them
are set out below.

18.206 [BHS 1] Cathay Pacific Airways Limited (Cathay Pacific)
and Securair Limited (Securair) staff fed about 220 bags from Kai Tak
with no baggage labels into the conveyor system at the new airport.
These bags were among some 420 interline bags that were brought to the
new airport on AOD. These bags should be brought down to the
Baggage Hall either by the out-of-gauge lift or by using the fallback
tags. As these bags were put into BHS without any baggage labels,
BHS identified them as problem bags and rightly diverted them to the
problem bag area. Whilst it was clear that these bags were injected into
the system by Cathay Pacific and Securair staff, the Commission is not
able to apportion the blame without the benefit of cross-examining the
relevant witnesses.

18.207 [BHS 2] Airlines checked in bags with incorrect labels or
invalid or no Baggage Source Messages (BSMs). BHS sorts
departure and transfer baggage automatically to the correct laterals
through the reading of the 10-digit bar-coded licence plate number on the
baggage label printed by airlines and by looking up the corresponding
BSM in the BHS Sort Allocation Computer (SAC). On AOD, some
departure and a large percentage of transfer bags bore labels with bar
codes that were not recognisable by BHS, or were given BSMs of an
incorrect format. J AL and Thai Airways International Public Company
Limited (Thai Airways) have respectively accepted that they were
responsible for some 600 and seven of these bags. In another case, the
wrong prefix (J L instead of EG) was programmed for recognition by
SAC in BHS for the bags of J apan Asia Airways. BHS was therefore
unable to recognise EGs bags which were sent to the problem bag areas.
This problem was rectified within a few days after AOD. Other than
J AL and Thai Airways, the Commissioners are unable to find out which
were the offending airlines.

18.208 [BHS 3] Airlines checked in about 2,000 bags with invalid
flight numbers. Some airlines entered flight numbers for baggage labels
and BSMs that were different from those listed in the flight schedule, and

631
were thus not recognisable by BHS. These bags therefore became
problem bags. Other than Canadian Airlines International Limited and
Virgin Atlantic Airways Limited who admitted responsibility for a very
small portion of these bags, there is insufficient evidence for the
Commissioners to identify other offending airlines.

18.209 [BHS 4] AVSECO staff rejected a large number of bags at
Level 2 security screening, putting pressure on Level 3 screening,
lengthening baggage handling time and causing more problem bags.
The Commissioners find that this is a matter of familiarity and caution for
AVSECO staff in security screening of bags. No one should be blamed
as a more cautious approach should be preferred over any lesser standard
which might create security risk.
18.210 [BHS 5] RHOs delivered transfer bags from inbound flights
into BHS after connecting flight laterals had been closed. Whilst this
was clearly the fault of RHOs concerned, the Commissioners find that
this was due to the various difficulties faced by RHOs as a result of the
lack of accurate flight information caused by FIDS failure. It is,
however, not possible, from the evidence before the Commission, to
identify the offending RHOs.

18.211 [BHS 6] RHOs did not clear bags from departure laterals in
time, resulting in full lateral alarms, which caused subsequent bags to go
to the problem bag area. This created about 600 problem bags for which
RHOs should be responsible. Again, the Commissioners do not have
sufficient evidence to hold any particular RHO responsible for this.

18.212 [BHS 7] One of RHOs, Ogden, put about 230 arrival bags
from a KLM flight No.887 onto transfer laterals. Some delay and
inconvenience were caused to the arriving passengers on that flight
although most of them received their bags on the same day. This was an
isolated incident caused by human error for which Ogden had accepted
responsibility.

18.213 [BHS 8] Bags that could not be safely conveyed were not
put in tubs and OOG bags were fed into the conveyor system instead of
being sent down to the Baggage Hall via the OOG lift. Unconveyable
bags such as soft bags that will roll along the conveyors and rucksacks

632
with straps must be put in tubs before being injected into the system.
Otherwise, the conveyors might be jammed and thus causing system
stoppage. There were about 200 to 250 bag jams on AOD under this
category which contributed to the baggage handling chaos. Whilst the
Commissioners have little doubt that this did happen, the evidence does
not show precisely which airlines should be held responsible.

18.214 [BHS 9] Too many erroneous emergency stops led to
numerous disruption and system downtime. The emergency stop
buttons were pressed some 99 times on AOD. The buttons might have
been pressed deliberately, as when bags had to be manually removed
from the system by SESL, RHOs or AVSECO. In other cases, it could
have been activated for safety reason. The Commissioners find that the
person or persons pressing the buttons, albeit causing system stoppage,
should in the latter case, not be held responsible. The Commissioners
are also of the view that the protruding design of the emergency button is
not flawed as it should be easily accessible to the operators to facilitate
activation immediately in case of danger. There is no evidence before
the Commission of the identity of the persons who pressed the emergency
buttons.

18.215 [BHS 10] Communication difficulties between operators in
the Baggage Hall due to TMR overload and unavailability of other means
of communication resulted in longer times for the system to be reset each
time it was stopped. When part of the system stopped, Baggage
Handling Operators (BHOs) would be sent to the scene to investigate
the cause and to rectify the problem. BHOs would then have to notify
the BCR to restart that part of the system. However, due to the
overloading of TMR, the operators were not able to communicate with
BCR effectively. The communication difficulties exacerbated the
problems caused by system stoppages because operators had difficulties
communicating with each other and resets of the system which could
have taken one to two minutes had taken 10 minutes instead.
Responsibility for system stoppages has been dealt with above and the
TMR problem is covered in under item [2] TMR in Chapter 16 in this
report.

633
18.216 [BHS 11] RHOs had no reliable flight information from
FIDS and had communication difficulties due to the overloading of TMR
and mobile phones and unavailability of other fixed lines of
communication. There were delays in collecting bags from aircraft and
transferring them to the Baggage Hall by RHOs due to the snowball
effect of delays on the apron caused by a number of factors. For
example, stand allocation by ACC was delayed. The apron was full on
AOD from about midday to 5 pm and from 8 pm to 1 pm and arriving
aircraft had to queue up at the taxiway for the next available stand to be
allocated. There was a lack of accurate flight information being
disseminated to RHOs, which increased ground time for handling arriving
passengers and baggage. Flight information was not displayed via
FDDS. Coupled with the TMR and mobile phones problem, RHOs
operations were seriously hampered because they had difficulties
knowing the time and at which stand the aircraft would park. These
problems are dealt with in other parts of this chapter.

18.217 [BHS 12] RHOs did not use both feedlines of the reclaim
carousels. An allegation was made against RHOs that they did not
maximise the use of the feedlines of carousels as each arrival carousel
could be fed by two conveyors. This increased despatch times and thus
slowed down the baggage handling process. The Commissioners
consider that the time which might have been saved in using both
feedlines would be slight. This problem would have been negligible but
for the other problems surfacing on AOD.

18.218 [BHS 13] RHOs did not know the assigned lateral for arrival
bags. The usual practice is that reclaim laterals are assigned by SESL
according to a pre-arranged allocation, which is distributed to RHOs and
BHOs on a template the preceding night. However, on AOD, SESL
reallocated laterals on a real time basis in order to optimise their use.
The new lateral allocations were displayed for passengers in BRH.
Unfortunately, RHOs did not receive the information as the FIDS LCD
boards in the Baggage Hall were not working and there were inadequate
back-up measures to relay the information to RHOs. While AA and
SESL must be responsible for not having whiteboards or fallback signage
made available at the Baggage Hall to direct RHOs to the proper areas
and laterals, SESL might not reasonably expect a FIDS failure. It is

634
therefore difficult to apportion the blame on the evidence. The
Commissioners do not feel that the evidence is sufficient to hold that it
was unreasonable or improper for SESL to disregard the template in the
hope of facilitating better use of the laterals, in accordance with the actual
flight times. The problem was apparently resolved when SESL was told
by AA to revert to the original fixed schedule and stop real time
reallocation at about 8 am on AOD. The crux of the problem was the
deficient operation of FIDS.

18.219 [BHS 14] RHOs abandoned ULDs around arrival baggage
feedlines, causing congestion and confusion in the Baggage Hall. The
reason given by RHOs was that the problem bags were loaded in ULDs
and since there was no baggage staging area, the only place available for
the temporary storage of ULDs was the space around the arrival baggage
feedlines. There is no evidence that any other place in the Baggage Hall
was available to accommodate these ULDs. The Commissioners
therefore accept that this was the effect whereas the insufficient
contingency planning of AA was the cause.

18.220 [BHS 15] FIDS workstation in BCR performed slowly and
hung frequently. On AOD, the FIDS workstation in BCR was recorded
to have hung up at 10 am and frequently at other times. Evidence was
given before the Commission that at times, it took 20 to 25 minutes to
make one reclaim belt allocation. This resulted in either no or delayed
displays of reclaim belts to RHOs and to passengers. At about 10 am
the performance of the FIDS workstation in BCR was so slow that
AA/EDS decided to reconfigure the parameters and the reclaim belt
allocation was taken over by AOCC which had more workstations to
switch around. This problem is dealt with in Chapter 13.

18.221 [BHS 16] There was no reliable flight information displayed
on the LCD boards in BRH. This caused problem to passengers who did
not know where to pick up their bags. The Commissioners accept the
evidence of SESL which attributed the problem to slow and unstable
performance of FIDS. The matter, concerning FIDS, is also discussed in
Chapters 10 and 13. To fill in or supplement missing information, AA
put whiteboards with necessary information written on them at BRH on
Level 5 early in the morning on AOD.

635

18.222 [BHS 17] Stretching of RHOs resources. As described
above, this was caused by the lack of essential information and other
means of effective communications as well as the extraordinary number
of the problem bags. For instance, runners had to go between AOCCs
whiteboard and staff on passenger and cargo ramps to pass on
information that should have been available from FDDS. RHOs had
difficulties knowing where to send staff to pick up or to load baggage.
The build up of problem bags meant that RHOs manpower was severely
stretched with manually sorting these bags. Additional manpower was
deployed by RHOs to cope with the situation on AOD and by about Day
Three, the situation had improved significantly and baggage operation
began to normalise. Having considered the evidence, the
Commissioners find it improbable that there would have been a
manpower shortage with RHOs, had the problem bags not been of the
unexpectedly large number on AOD. The drain on their manpower was
caused by the inefficient operation of FIDS and the other many problems
that occurred in a vicious cycle and a downward spiral on AOD.
Accordingly, RHOs should not be criticised for not removing the problem
bags in time.

18.223 [BHS 18] The Remote or Hot Transfer System, although
available, was not used to handle transfer baggage with the result that all
transfer baggage was handled only by the Central Transfer System in the
Baggage Hall, which slowed down operations. The Commissioners take
the view that the effect of this was minor and this might not have been
noticed but for the other problems.

18.224 [BHS 19] Inexperience or unfamiliarity of airline, RHO and
SESL staff. Many of the actions of airline, RHO and SESL staff
demonstrated their inexperience or unfamiliarity with their operation at
the new airport. For instance, the airlines incorrect method of
introducing unconveyable bags into the system. However, viewing the
evidence as a whole, the Commissioners come to the view that it was
more because of unfamiliarity rather than the lack of experience or
training on the part of the airline, RHO and SESL staff that caused the
problems with baggage handling on AOD, though undoubtedly more
hands-on training would have resulted in more familiarity in operations.

636
Furthermore, the unfamiliarity might not have been so serious had AA
planned and worked out with RHOs the required resources for coping
with baggage handling in case of FIDS failure. AA should therefore be
responsible for lack of sufficient coordination.

18.225 During the days after AOD, there was improvement to the
performance of FIDS, and the direct and consequential problems it
created gradually subsided. On Day Two, the number of bags left over
was 6,000 out of a total of 24,000 bags processed. This was reduced to
2,000 (out of 26,000 bags), 1,400 (out of 27,000 bags) and 220 (out of
27,000 bags) on Day Three, Day Four and Day Five respectively. RHOs
were able to return to normal operation by about Day Three to Day Four.
RHOS, passenger handling entities and airlines had worked with AA to
put more logic into the assignment of gates to minimise the amount of
travelling time around ramps. Further, as staff and operators became
more experienced and familiar with the system and operation, baggage
handling at the new airport improved significantly.

18.226 AAs statistics showed that by Week 2 of AOD, the average
figures for first and last bag delivery times were similar to figures for Kai
Tak, and were improving. The latest statistics published by AA show
that during the period from 1 December 1998 to 3 J anuary 1999, for 90%
of the flights, the first and last bag delivery times were 19 minutes and 36
minutes respectively, which far surpass the figures of 25 and 43 minutes
for Kai Tak. In the week commencing 31 August 1998, only 296 bags
out of a total of 228,000 departure and transfer bags processed missed
their flight. As at today, the baggage handling process can certainly be
said to have attained the world-class standard.


18.227 There is no evidence received by the Commissioners to
justify a finding that there was any lack of coordination or
communication between ADSCOM and NAPCO in relation to
ADSCOMs decision to open the airport on 6 J uly 1998 or in NAPCOs
overall monitoring of AOR issues.

637
18.228 W36 Mrs Anson CHAN, the Chief Secretary for
Administration and the Chairman of ADSCOM (the Chief Secretary)
said in evidence that NAPCOs monitoring role was that of a critical
observer which the Commissioners accept. The role of the critical
observer is to critically examine and evaluate the progress of various
AOR critical issues through AAs reports, as well as through observation
by NAPCOs own professional staff. The Commissioners opine that
NAPCO failed in two aspects in the performance of its function. First, it
should have inquired with AA whether it had the necessary expertise in
monitoring HACTLs progress relating to the installation, testing and
commissioning of ST1s 5-level CHS equipment and systems, but it did
not do so. Secondly, it should have checked whether AA had plans and
contingency measures and should have had an overall assessment
whether such plans and measures were adequate in view of the then
prevailing circumstances. As a corollary, NAPCO should also examine
if AA had an overall risk assessment.

18.229 The evidence shows that ADSCOM had the duty of an
overall monitor and it had delegated the duty of the overall monitor of the
progress of AOR to its executive arm, NAPCO, and directed it to
discharge the duty. The public looks upon ADSCOM, as opposed to
NAPCO, to discharge the duty as the overall monitor. On this premises,
ADSCOM is ultimately responsible for that duty not having been
satisfactorily discharged by NAPCO.

18.230 There was difficulty in Government obtaining information
from AA which showed a lack of cooperation. From mid-1996 onward,
AA became more open to Government. It shared its internal reports
with NAPCO and allowed NAPCO to take part in system tests.
Towards AOD, coordination and cooperation between AA and NAPCO
improved significantly that NAPCO was no longer complaining.

18.231 AAs business includes the operation of the new airport. In
conducting such operation, it shall have regard to the safe and efficient
movement of air passengers and air cargo. The problems encountered
on AOD revealed that AA did not have sufficient regard in these respects
when opening the airport for operation on 6 J uly 1998. AA should
therefore be responsible.

638

18.232 Coordination and cooperation between AMD and PD was
particularly important from about the last quarter of 1997 since the new
airport was in a transition from the construction stage to the operation
stage. The coordination between AMD and PD continued to cause
concern up to mid-1998. The coordination problem is caused by several
factors, namely:

(a) Less importance being placed on operational requirements as
compared with the works programme.

(b) PD was the major part of the AA organisation up to AOD
representing about three-fourths of the total organisation
whereas AMD did not really start to grow and expand until
the later part of 1997.

(c) W3 Townsends engineering background leads him to place
more emphases on the works side and to complete the project
on time and within budget. Airport management was not
W3 Townsends specialty, and he might have tended to
overlook this aspect.

(d) W43 Mr Douglas Edwin Oakervee, an assertive and imposing
character, greatly influenced W3 Townsend, relatively milder
in personality, in placing too much significance and priority
on PD and giving less support to AMD. W44 Heed, a less
resolute personality, took whatever was on offer, well
knowing that he would be facing great difficulties when
operating the new airport after the systems were handed over
to AMD from PD with the degree of testing and
commissioning leaving much to be desired.

18.233 A consultant report dated October 1997, commissioned for
the purpose of advice on management structure post-AOD, revealed
deficiencies in the leadership and teamwork of the senior management
and incompetence of some senior managers. It is unfortunate that such
important deficiencies were exposed at such a late stage. At that time,
barely about six months before the AA Boards target date of April 1998

639
for airport opening, it would be too risky to introduce a change of the
senior management. The Commission accepts this as a reasonable
explanation and does not attach any blame to the AA Board. The Board
however introduced measures recommended in the report to strengthen
leadership and improve coordination, such as asking Government to
second the then Director of NAPCO, W48 Mr Billy LAM Chung Lun, to
be the Deputy CEO.

18.234 Notable examples of lack of coordination among Divisions
within AA include:

(a) W44 Heed did not know that the 98.7% availability of FIDS
reported to him at the meeting of AMD general managers in
J une 1998 only related to availability of host servers, and not
the whole FIDS.

(b) W44 Heed was not informed that the stress and load test of FIDS
was agreed between PD and GEC to be deferred after AOD,
nor was he advised about the risks of not having the stress and
load test conducted before the system was put into use.

(c) Neither W44 Heed, Director of AMD, nor W45 Mr Kironmoy
Chatterjee, Head of IT, made satisfactory arrangements for
experts of EDS and Preston to be stationed on AOD in the
crucial ACC where it turned out that operators did experience
difficulties in performing flight swapping with TMS and with
input into FIDS.

18.235 The major causes of the problems within AA can be
summarised as follows:

(a) Problems with personalities of the top management of AA.

(b) Late involvement of AMD and IT Department in the system
development. AMDs requirements were not given high
priority until sometime in 1997 whereas IT Department,
which used to be part of the Commercial Division, only
became involved from late 1996.

640

(c) There was no planning to ensure a smooth transition from the
construction stage to the operational phase and no experts or
consultants had been engaged for that purpose. Such experts
might have helped in identifying the issues that needed to be
resolved and measures that needed to be implemented for a
smooth transition.

(d) Insufficient examination of the negative aspects of the interaction
of the failures of more than a single system, resulting in a lack
of overall risk assessment. This lack of overall risk
assessment is especially unfortunate in view of the history of
unreliability of FIDS.

(e) AAs failure to engage a consultant to monitor HACTLs systems.
Had such consultant been appointed to monitor HACTLs
testing and commissioning of its systems, it would not only
have assisted AA in ensuring itself that HACTL was ready,
but would certainly have helped HACTL to re-examine its
assurances of readiness on AOD more carefully.

Section 5 : Responsibility of AA

18.236 The Commission finds that the AA management failed to
maintain a right balance between PD and AMD in two ways. First,
AMDs participation in project and systems development was not
provided for in an early stage. Secondly, the personalities of the persons
occupying key posts caused problems. This was discussed under section
4 above.

18.237 For the purpose of the inquiry, the acts and omissions and
therefore the responsibilities of the following persons in the top AA
management have been examined in detail:

(a) W3 Townsend

18.238 From the totality of the evidence presented to the Commission, it

641
is clear that W3 Townsend was not in control of the management,
resulting in lack of coordination between the PD and AMD. He did not
give sufficient priority and adequate support to operational requirements
of AMD, especially since the end of 1997 when more preponderance
should have been accorded to AMD in the transition from the project
stage to operation sphere. He did not assign sufficient resources to
AMD at an early stage, and failed to give sufficient support to W44 Heed.
He did not engage an expert to monitor HACTLs system. All these
resulted in the deficiencies in the operational readiness of the airport.
He must be responsible for failing to have any or any proper global
assessment of AOR. He is further responsible for the misstatements he
made to the AA Board and ADSCOM referred to below.

(b) W48 Lam

18.239 W48 Lam has been found by the Commission not to be
responsible for the problems witnessed on AOD, or for the lack of
communication and coordination or for the misstatements.

(c) W43 Oakervee

18.240 There is no evidence that W43 Oakervee has failed in his duties
as Director of PD, although he should be primarily responsible for the
slippages in respect of the construction and systems works vis--vis AMD
which caused the time necessary for training and familiarisation of AMD
operators on the systems to have been compressed.

(d) W44 Heed

18.241 W44 Heed, as the Director of AMD, must take the major share of
blame of the problems and shortcomings witnessed on AOD. First, his
personality was too weak. He ought to have stood firm vis--vis PD, in
particular his counterpart W43 Oakervee, to ensure that AMD would have
sufficient time to be properly prepared for AOD. Secondly, he failed in
his duty to ensure that he was kept properly informed of the progress of
the FIDS development so as to enable him, as head of AMD, to make an
informed assessment as to the readiness of FIDS for AOD. Thirdly, he
failed to ensure that an appropriate overall risk assessment was carried

642
out to assess the risks involved in proceeding with the opening and the
sufficiency of the contingency measures that were in place. He admitted
that there was no global contingency plan. W44 Heeds weakness and
deficiencies deprived Hong Kong of the chance of a smoother and more
efficient airport on AOD.

18.242 He was not involved with either of the misstatements made by
W3 Townsend referred to below. However, in his response to questions
asked about the misstatement on ACS, his attitude was exposed. The
matter is dealt with below.

(e) W45 Chatterjee

18.243 The Commission finds that W45 Chatterjee, as Head of IT, had
failed in his duties in two respects. First, he did not assess properly the
risks involved in deferring the stress test for FIDS. Secondly, he did not
advise AMD properly of the risks involved in not undergoing such test
before AOD. He was also grossly negligent in allowing the
misstatement contained in the ADSCOM Paper about the reliability of
FIDS unexplained at the ADSCOM meeting when the Paper was
discussed. The matter is dealt with below.

(f) AA Board

18.244 The AA Board is ultimately responsible for the problems which
occurred on AOD because the duty for developing and operating the new
airport is placed on it by Section 4 of the Airport Authority Ordinance
which provides:

Subject to the provisions of this Ordinance, the affairs of the
Authority shall be under the care and management of a board
whose functions shall comprise such care and management.

The responsibility to discharge the functions of developing and operating
the new airport remains with the AA Board, although it is allowed by
Sections 9 and 15 of the Ordinance to delegate its functions to a CEO and
management.

643
18.245 The Commissioners do not accept the arguments that the AA
Board should be responsible for W3 Townsends acts and omissions or
the acts and omissions of the AA management, nor that the Board should
be professionally qualified. However, the Board may be criticised for
not having appointed outside consultants to advise itself on the progress
of important projects such as FIDS, instead of allowing AA management
to have such consultants. However, this view may be derived from the
wisdom of hindsight, which might have not been clear to the AA Board at
the material time. This failure of the AA Board should not therefore be
overstated.

Misstatements and Responsibility for Them

18.246 Two misstatements were identified during the inquiry. One
was the reliability of FIDS as a whole was 98.7% available and the other
was that ACS had been tested successfully. Although these
misstatements are not related to any direct cause for the chaos on AOD,
they had significant bearing on the top management of AA. They might
also have created a false sense of security in ADSCOM.

18.247 On FIDS, the representation made by AA to ADSCOM in
ADSCOM Paper 34/98 dated 23/6/1998 for the ADSCOM meeting on 24
J une 1998 has the following passage-

Reliability tests on the present version of FIDS (Version 2.01C)
commenced on 14 J une and were completed on 20 J une using live
data from Kai Tak through the AODB. The reliability of the
system as a whole has been 98.7% available; the reasons for
unavailability of some monitors and LCD boards at the 24 J une
trial have been identified and the problems are being rectified.

18.248 A similar, but not identical statement, was found in an AA
Board paper 183/98 dated also 23/6/98.

18.249 Both W45 Chatterjee and W43 Oakervee admitted that the
passage conveyed false ideas, while W44 Heed did not have sufficient
technical know-how as to comment. The truth of the matter is as
follows:

644

(a) The 98.7% was the availability of the host servers, a hardware
and not a software, and not any other part of FIDS, let alone
FIDS as a whole;

(b) There is a slight difference between availability which means
the time when a system is operational as opposed to downtime,
and reliability which relates to the soundness and consistency
of the system; and

(c) The mentioning of the unavailability of some monitors and
LCD boards implied that the only problem with FIDS causing
the achievement of reliability of 98.7%, as opposed to 100%,
was the monitors and LCD boards, and this implication was
false.

18.250 These false ideas did mislead ADSCOM, for its members all
understood that the 98.7% referred to the reliability of FIDS as a whole
system. However, because of the prior knowledge of ADSCOM
members about the continual unreliability or instability of FIDS during
the various tests up to that date, they placed greater reliance on the
standby FIDS that had been reported to have been successfully tested on
30 J une 1998 in case of a failure of the main FIDS. The false ideas
therefore had not, in the Commissioners opinion, caused too much
mischief.

18.251 After examining all the evidence, the Commissioners are
satisfied on a balance of probabilities that the finalised versions of the
ADSCOM paper and the AA Board paper were made by W3 Townsend,
and he must be personally responsible for uttering the misstatement to the
AA Board and ADSCOM, although the evidence is not weighty enough
for an inference to be drawn that there was clearly an intent on W3
Townsends part to mislead ADSCOM.

18.252 W45 Chatterjee is also found by the Commission to have been
grossly negligent in not pointing out the misstatement to ADSCOM or
disabuse ADSCOM members when he attended two ADSCOM meetings
subsequent to the provision of the paper to ADSCOM.

645

18.253 At the ADSCOM meeting on 6 J une 1998, W3 Townsend,
W44 Heed and W45 Chatterjee attended. The notes recorded that

CEO/AA (ie, W3 Townsend) added that the four key safety and
security systems access control, fire alarm, closed circuit
television and PA had also been successfully tested. They were
at the moment busily engaged in issuing access cards.

18.254 The fact, however, is that at the time of the ADSCOM
meeting in question, problems regarding ACS had not yet been fixed and
indeed, up to the day when the four senior officers of AA gave evidence
together before the Commission, ACS problem had not yet been fully
rectified. The statement of W3 Townsend to the ADSCOM meeting on
6 J une quoted above is obviously incorrect. W3 Townsend denied
having an intent to give false information. W44 Heed said he merely let
the matter pass, not having a private word with W3 Townsend, nor did he
think it necessary to do so for W3 Townsend should have known the
situation. He admitted that it did not matter if members of ADSCOM
was misled.

18.255 The Commissioners have not been able to find sufficiently
weighty evidence to sustain a finding of wilful intent on W3 Townsends
part to mislead ADSCOM about the progress of ACS. Nevertheless, W3
Townsend must be the main culprit in making the misstatement to
ADSCOM.

18.256 As far as W44 Heed is concerned, it transpired that he had
not actually joined the meeting at the juncture when the misleading
statement was uttered. However, he revealed his attitude on the matter
merely being in support of the CEO, W3 Townsend and as a respondent
to questions when put, but would not bother if ADSCOM was misled.
The Commissioners find this attitude reproachable, especially in view of
the trust reposed in him by ADSCOM by inviting him to attend its
meetings. The attitude makes it doubtful whether he could properly
handle matters in a crisis or delicate situation.
Section 6 : The Present Situation

646
18.257 The new airport has experienced and undergone a host of
operational and management problems during its initial period of
operation. However, looking at the evidence received by the
Commission and excluding problems not yet surfaced (which are outside
the Commissions work), it can be said that the new airport has
completely come out of the pit of problems to attain the standard of a
world-class airport.

18.258 There have been remarkable improvements to the operation
of the new airport. This is illustrated in the following comparison of
some of the operational statistics between Kai Tak and the new airport, as
provided by AA.

Kai Tak New Airport

Period 26 October
1997
to
27 March
1998
6 J uly 1998
to
31 J uly 1998
1 December
1998
to
3 J anuary
1999
Average delay for incoming
flights
(excluding early & on-time
flights)

30 minutes

30 minutes

24 minutes
Average delay for all outgoing
flights
24 minutes 30 minutes 18 minutes
Time of arrival of first bag in
baggage hall after aircraft
landing
- Average
- time for 90% of flights first
bag to arrive

25 minutes
(service
pledge)

25 minutes
40 minutes

13 minutes
19 minutes
Time of arrival of last bag in
baggage hall after aircraft
landing
- average
- time for 90% of flights last
bag to arrive

43 minutes
(service
pledge)

39 minutes
60 minutes

25 minutes
36 minutes

647

The figures represent a very high standard of service which compare
favourably with those for the Kai Tak airport. A brief summary of some
airport operational statistics can be found at Appendix XVI to this report.

18.259 A more efficient operation of FIDS has enhanced the operation
of other facilities at the airport, such as the operation of RHOs. RHOs
have achieved 100% on-time docking of airbridges for arriving flights,
allowing door opening within 2 minutes. Frontal stands which connect
to PTB via airbridges are now being assigned to about 85% of passenger
flights. Stands are allocated 2 hours before scheduled arrival times so
that airlines, RHOs and other operators have sufficient time to plan their
operation and to allocate their resources. The 68 aircraft parking aids on
the apron have been comprehensively improved and other than a single
incident on 15 J uly 1998 involving the parking of a Cathay Pacific
aircraft that is dealt with above, 25,000 aircraft have conducted parking
safely using the automated system of APAs.

18.260 Cargo operation has normalised. A working group has been
established to resolve operational and communication issues. An
interface agreement has been reached by the cargo terminal operators and
RHOs on their respective roles and responsibilities, freighter handling
procedures, the exchange and utilisation of equipment, the establishment
of a cargo interface area and the procedure in using it and the
establishment of times for cargo movements to and from aircraft and the
cargo complex.

18.261 Since AOD, AA has improved the signage in PTB by adding 300
directional signs and 2,000 information signs. Incidents where people
had to climb up the steep staircase due to escalators breakdown have been
minimised. Improvement work has been carried out on the bus and taxi
stations for the convenience of the public. Lighting, ventilation and
cleanliness of toilets have also been improved. On airside, attendants
have been put in place on the platforms for the APM to prevent people
from boarding train at the terminal station where everyone should alight
and to minimise human intervention in trying to pry open train door while
it is closing. Security appears to be satisfactory and there was no recent
report of major security breach. The overall security at the new airport

648
has in fact been regarded as excellent by the Federal Aviation
Administration, the United States governments aviation safety regulatory
body.

18.262 The Commissioners are encouraged to hear that the new airport
has recently won the annual Critics Choice award of an American
premier travel industry publication which commended the new airports
fast baggage delivery times, the high speed rail link and the attractive
PTB. The new airport has also been selected as one of the top 10
Construction Achievements of the 20
th
Century by the major
CONEXPO-CON/AGG Exposition to be held in March 1999 in the
United States.

Section 7 : Could the Chaos and Confusion have been Avoided?

18.263 The most likely question to be asked after the opening of the
airport and the conclusion of the inquiry seems to be: Could the chaos
and confusion on AOD have been avoided? There are two most
apparent alternative approaches to provide the answer, namely, (a) could
anything have been done to prevent the chaos by way of better planning
and working before AOD? or (b) should AOD be deferred and if so, for
how long the deferment should be?

18.264 W51 Yuen, a consultant of the San Francisco International
Airport, said that the risks of not having an efficient and smooth opening
would only become more apparent when AOD was closely approaching.
The Commissioners agree to this view in that even though a number of
testing of the systems required for AOD could have exposed problems in
operating the systems from time to time, ways and means would be
considered and implemented to resolve the problems whenever they
occurred rather than treating them as endangering AOD until at a late
stage. When the problems persisted, AMD should have a
comprehensive risk assessment, and implement measures that were
considered necessary in the event of the failure of all the systems in PTB.
A comprehensive risk assessment would have identified that in the case
of failure of FIDS, the various means of communication for the
dissemination of flight information would need to be ensured or their

649
capacities increased. Contingency plans consequent upon such an
assessment would have been developed by AA in conjunction with all
other necessary airport operators, like the airlines, RHOs, BHO, line
providers and CTOs. If all these had been done, then the chaos in PTB
could have been alleviated if not eliminated altogether. Moreover,
regarding ST1, if there had been effective monitoring by AA of the
readiness of HACTLs CHS, HACTL might have been warned against its
over-confidence. There might have been more testing of its systems and
their operation in an integrated manner. If the airport trials had been
prepared in such a way as to be much closer to a live situation of
operation and participated by HACTL, it would have also helped expose
problems in CHS. There are, however, grave doubts whether there
would have been sufficient time to do all these things when the risks of
not having a smooth and efficient operation became apparent.

18.265 All involved, those in AA, HACTL and other airport operators
were working extremely hard to achieve the target, and there is no doubt
about it throughout the evidence, oral and documentary, and at least they
had been putting in everything they could since the end of 1997, if not
earlier. It is because of this drive and spirit to focus on attaining the goal
rather than taking a negative attitude of the task being unachievable that
pushed everyone involved to exert himself or herself up to the limit.
And it is this kind of drive and spirit that kept everybody from translating
difficulties experienced in the early stages after the announcement of
AOD into discouragement or warning that AOD could not be met.
When time was getting closer, when problems with regard to a number of
systems in PTB, notably FIDS, ACS, PA and telephones persisted, it
would realistically be the first time that they should consider whether the
risks justified a reconsideration of AOD. That would be too late for all
the required risk assessment to be made or contingency measures to be
planned and fully coordinated. That would only leave those involved
with a Hobsons choice: to defer AOD.

18.266 While a postponement of AOD would prevent the chaos and
confusion, it must be understood that it would not have helped if AOD,
when it was announced in J anuary 1998, was not 6 J uly 1998 but
sometime later. The reason is that the risks affecting a smooth AOD
would only have surfaced close to AOD. Had a later AOD been

650
announced right from J anuary 1998, the added time would not have
exposed the risks at an early stage. W51 Yuen suggested that serious
consideration of a deferment should have happened about two weeks
before AOD. The Commissioners accept this view because it was at this
juncture that the risks could properly and reasonably have been realised.
W51 Yuen said that in hindsight, seeing that most of the problems were
resolved satisfactorily within about two weeks after AOD, but taking
account that if there was a deferment of AOD there would be a loss of
momentum, the safe guide would be to triplicate the time needed for
resolving the main difficulties, which would mean a postponement of six
weeks. However, this estimate does not take into consideration
HACTLs problems and that there could not be any actual live operation
barring a real opening due to the fact that no simulated trial could create
situations identical or as useful as live operation. W55 Dr Ulrich Kipper
and W56 Professor Vincent Yun SHEN opined that two to three more
months would be required to make FIDS run efficiently and four to six
months would be needed to make ST1s operation smooth, taking into
account the inevitable loss of momentum when a deferment was
announced. The Commissions think that the experts estimates were too
conservative, and did not take sufficient account of the hefty financial
implications and the effect of a further loss of momentum that a long
postponement would produce. With the full benefit of hindsight and
having examined all the evidence, the Commissioners feel that if a
deferment were sought and considered about a fortnight before 6 J uly
1998, airport operation commencement should be deferred for about two
months. The Commissioners recognise that momentum would certainly
be impacted by a deferment, and HACTLs confidence of the readiness of
its CHS would still be there. However, if AA and HACTL were
impressed with the importance of making everything ready by the
deferred date, the loss of momentum could well be reduced. Further,
HACTL would then have its contractual deadline of 18 August 1998 to
keep, while AA would know that its previous promised readiness target of
April 1998 had been allowed to slip further. The added time would
certainly be used by HACTL to have further testing done with CHS,
allowing its staff to be better trained and getting more familiar with how
to operate CHS, and having a better contingency planning with the
implementation of contingency measures. On PTB side, the added time
would be employed for a more widespread and intensive trial, say about

651
six weeks before the deferred AOD, leaving sufficient time to eradicate
the problems exposed by the trial. With more time, FIDS could have
gone through a stress test and the means of communication could have
been better prepared.

Section 8 : Lessons Learned

18.267 Although the Commissioners are not tasked to make
recommendations as to how to address the problems encountered on AOD
or oblivious of the expectation that no further airport will be built in
Hong Kong in decades, they think that something would be amiss if they
did not state their views as to what lessons they have learned from what
has been revealed through the heavy work that they and all persons
involved in the inquiry have gone through.

18.268 From the top, there is ADSCOM with NAPCO as its
executive arm. Government was rightly concerned with the
development and opening of the new airport, not only as the major
shareholder of the statutory corporation of AA from a financial
perspective, but also for the public good to ensure that public funds were
spent in a worthwhile manner and to maintain Hong Kong as an
international and regional hub of civil aviation as well as Hong Kongs
reputation of high efficiency. ADSCOM took upon itself the task of
deciding AOD, for all the above good reasons and because of its
involvement in the other nine huge infrastructure projects of Airport Core
Programme (ACP). It was therefore proper for ADSCOM to have an
overview of AAs progress and performance. NAPCO was tasked with
coordinating all the 10 ACP projects, and also monitoring the progress
AAs work relating to AOR. This monitoring role is nebulous because,
at times as W36 the Chief Secretary pointed out, NAPCO was a critical
observer, but when problems were noticed with FIDS NAPCO adopted a
more proactive attitude in getting more information than a critical
observer would. This was perfectly fine for all concerned save that it
would unwittingly lay a trap for AA whose Chairman and Vice-Chairman,
ie, W50 Mr WONG Po Yan and W49 Mr LO Chung Hing, thought, albeit
perhaps unjustifiably, that AA could rely on NAPCOs monitoring. This
had unintentionally given AA a sense of security which should have been

652
avoided, either by reminding AA of its statutory functions and obligations,
or by telling it in no uncertain terms that NAPCO was purely working for
ADSCOM. The involvement of an organisation like NAPCO could also
have conjured up a false idea in the public that Government was to ensure
that the work for which AA was solely responsible would be satisfactorily
performed. In other words, getting more involved than its position
required in a project which is the sole responsibility of a statutory
corporation might give rise to a misunderstanding that the success or
otherwise of the project is a Government responsibility.

18.269 While intervening more than NAPCO should in its overall
monitoring of FIDS, its role regarding HACTLs systems is viewed by
way of comparison. NAPCO is criticised for failing to inquire if AA had
the required expertise in monitoring HACTLs systems. As a critical
observer, NAPCO had, according to the Commissioners opinion, failed
to satisfy itself that AA had such necessary expertise. Its reliance on
HACTLs good reputation and past record is not a reasonable excuse and
its assumption that AA had the expertise was not proper, for the
assumption could have been clarified with simply a question or a letter.
If NAPCO, and in particular AA, were correct in relying on HACTLs
high repute in the cargo handling field, there would have been no
necessity whatever to monitor HACTLs progress, which was part and
parcel of AAs statutory duties to have regard to provide efficient cargo
movement at the new airport.

18.270 Within AA itself, the main lessons that have been learned are
three-fold. First, whatever the organisational structure of a company, the
most important aspect is the fitness of the personality and character of the
persons occupying key posts, which must be viewed not only whether the
persons fit the posts alone, but the interaction of the personalities of those
occupying such posts should be considered carefully. W3 Townsend
might have been fine if he had been given the position of Director of PD.
This does not mean that he would have done a better job than W43
Oakervee in that post, but he would not be required to strike a proper
balance between the requirements of progress or lack of it of the works
and those of the operation side. On the other hand, W44 Heed as
Director of AMD was too weak and irresolute a character to work
alongside W43 Oakervee in getting what was required for operating the

653
new airport. His position differs from that of a director of airport
management in another airport because the situation of CLK in transition
from a construction phase to an operational sphere seldom happens
elsewhere. There seems to be nothing wrong with W43 Oakervee
discharging his functions as Director of PD, but while his personality
ensured that the requirements of his Division received top priority, it
overbore on W44 Heed who strove with whatever was offered but did not
seek, let alone find, substantially what he wanted from either W43
Oakervee or W3 Townsend.

18.271 Secondly, for a large project or in a large organisation, the
eventual user should be given an early, if not a first, opportunity to work
with the provider of the services. Had AMD and IT Department been
involved in the planning stage of the projects, and the development of the
systems in particular, there would certainly have been less changes to the
systems because of the late notice of the operational requirements.

18.272 Thirdly, there should always be a global and comprehensive
risk assessment, especially when various risky factors occurred
incessantly during the development process. Most members of the
senior management of AA knew that there was a risk in FIDS failing.
What they had done was to have a contingency plan for that scenario.
The contingency plan was merely to cover the situation when FIDS failed
to display and distribute information to airport users, but not when such
required information was inaccurate or incomplete. The scenario
envisaged was not bad enough, or at least not as bad as that experienced
on AOD. The only substantial contingency was to commission a
standby FIDS which was however tested very late in the day on 30 J une
1998. There was simply no assessment of how to react to a case if both
of the systems could not function on AOD. There was no overall
planning for the effective and efficient dissemination of the flight
information needed by so many airport operators, and there was no
concrete agreement as to what each involved party should do in the case
of failure of both FIDS and standby FIDS. The availability and
capacities of the means of communication were never considered in this
light. HACTL fell into the same error and only relied on the modular
nature of its systems and equipment as a full fallback position, which
would only have been available if parts of the mechatronics, as opposed

654
to software of its computer system, failed.

18.273 Connected with the lack of global risk assessment and
preparation for the worst is the over-confidence of the key players,
namely, AA and HACTL. The personnel of both organisations were
working extremely hard towards achieving the goal on the target date.
AAs top management were concentrating on what they could achieve
within time and budget but unable or unprepared for sparing a little time
to step outside their bounds to look at what they could not possibly
achieve as a critical outsider. On the other hand, HACTL was focusing
on developing its own intellectual protg and was imbued with its
hitherto success and reputation in running the cargo handling service in
Kai Tak. HACTLs senior and junior management found it hard to
believe that when every physical aspect of the works required for AOD
had been completed there could be something wrong with the integrated
use of CHS. Both of these groups of people in AA and HACTL had
tried so hard and been so immersed in their work that they had failed to
provide for the worst scenario. The over-confidence that had resulted in
AA not seeking any deferment of AOD had similarly caused HACTL to
reiterate the assurance of its readiness instead of even considering at a
late stage to retract it or asking for a soft opening by retaining resort to
Kai Tak, which it eventually did but only after AOD.

18.274 Delay with a deadline is always risky. The benefit of a
deadline is that it will bring pressure to bear on people involved to use
their best efforts and keep up the momentum. However, the pressure
might cause the people to suffer a breakdown, or worse still might lull
them into a false sense of confidence or even achievement, when they tell
themselves that they have already done their best and everything is fine or
everything else is a matter for luck or Providence. The accomplishment
of the task might be at risk, for those who imbued themselves with the
false sense of confidence and achievement would not be able to tell the
faults in their own work. To prevent this from happening, it is necessary
for those who are required to accomplish by a deadline to have a
conscientious risk assessment of the situation and make comprehensive
contingency plans to cater for various eventualities when delay is
experienced.

655
18.275 On a more positive note, it is heart-warming that most
people involved in the chaos and confusion on AOD drew themselves
together in a most cooperative and congenial manner to help solve
problems. The RHOs are an example, because they pulled their forces
together in dealing with the problems with baggage handling and clearing
the huge backlog of problem bags by employing the utmost of their
resources. The witnesses told the Commission that many of them and
others were working very hard, each to resolve the problems that he or
she could help to resolve. Most if not all of them stayed overnight on
AOD and some even several nights after AOD to contribute their share in
the joint efforts. No one was thinking as to who should be responsible
or who should properly be doing what. This applies to the many
members of the airport community as well as civil servants. The
Commissioners are deeply touched by the attitude and spirit that surfaced
at the time of adversity, which they fervently hope will be infectious and
available to help maintaining Hong Kong as a successful and happy
community!

Appendix I
I
Page 1
PARTIES IN THE INQUIRY

Public Hearing of Evidence from 7 September 1998 up to 31 December 1998

Parties
Airport Authority

Asia Airfreight Terminal Company Limited

AEH J oint Venture
AEH
Airlines Operators Committee

British-Chinese-J apanese J oint Venture
British-Chinese-J apanese
656
657
Board of Airline Representatives

Cathay Pacific Airways Limited

Electronic Data Systems Limited
Appendix I
I
Page 2
Gammon-Paul Y. J oint Venture

G.E.C. (Hong Kong) Ltd.

Hong Kong Air Cargo Terminals Limited

Hong Kong Airport Services Limited

Hong Kong Dragon Airlines Limited

658
Hong Kong Special Administrative Region
Government

[The major government departments and
bodies concerned with the new airport are:

Airport Development Steering Committee

New Airport Projects Co-ordination Office

Economic Services Bureau

Civil Aviation Department

Airport Consultative Committee
Appendix I
I
Page 3
]
Hong Kong Telecom CSL Limited

Hutchison Telecommunications (Hong Kong)
Limited

659
660
J ardine Air Terminal Services Limited

Mass Transit Railway Corporation

Ogden Aviation (Hong Kong) Limited
()
Ove Arup & Partners Hong Kong Ltd

Swire Engineering Services Ltd

LEGAL REPRESENTATIVES OF PARTIES IN THE INQUIRY

Public Hearing of Evidence from 7 September 1998 up to 31 December 1998

659

Legal Representatives
Appendix II
II
Page 1
660
The Commission

Mr Benjamin YU SC, Mr J AT Sew Tong
and Ms Yvonne CHENG
(instructed by Messrs Baker and
McKenzie)

()
Parties

Airport Authority

Mr Robert Ribeiro SC, Mr J oseph FOK
and Mr Paul SHIEH
(instructed by Messrs Allen & Overy)

()
Asia Airfreight
Terminal Company
Limited

Mr Robert Whitehead
(instructed by Messrs Simmons &
Simmons)

()
Appendix II
II
Page 2
AEH J oint Venture
AEH
Messrs Slaughter & May

Airlines Operators
Committee

-
British-Chinese-J apan
ese J oint Venture
British-Chinese-J apan
ese
Mr Louis K Y CHAN
(instructed by Messrs Masons Solicitors)

()
Board of Airline
Representatives

-
661
Cathay Pacific
Airways Limited

Mr Adrian Huggins SC and
Mr Anselmo Reyes
(instructed by Messrs J ohnson Stokes &
Master)
Adrian Huggins

()
Electronic Data
Systems Limited

Mr Simon Westbrook

(
Appendix II
II
Page 3
Gammon-Paul Y. J oint
Venture

Mr Denis K L CHANG SC
and Mr J ason POW

()
GEC (Hong Kong)
Limited

-
662
663
Hong Kong Air Cargo
Terminals Limited

Mr J ohn Griffiths SC, Ms Teresa CHENG
and Mr Pat Lun CHAN
(instructed by Messrs Deacons Graham &
J ames)

Pat Lun CHAN
()
Hong Kong Airport
Services Limited

Mr Geoffrey MA SC and
Ms Lisa K Y WONG
(instructed by Messrs Wilkinson & Grist)

(
Hong Kong Dragon
Airlines Limited

Mr Adrian Huggins SC
and Mr Anselmo Reyes
(instructed by Messrs J ohnson Stokes &
Master)
Adrian Huggins

(
Appendix II
II
Page 4
Hong Kong Special
Administrative Region
Government

[The major
government
departments and
bodies concerned with
the new airport are:
Airport Development
Steering Committee
664
New Airport Projects
Coordination Office
Economic Services
Bureau
Civil Aviation
Department
Airport Consultative
Committee

]
Mr Ronny TONG SC,
Mr Ambrose HO and
Mr Eugene FUNG
(instructed by Department of J ustice)

()
Hong Kong Telecom
CSL Limited

Mr Nigel Kat and Ms J ulia LAU
(instructed by Messrs Clyde & Co)

(
665
Hutchison
Telecommunications
(Hong Kong) Limited

Mr Michael Bunting
(instructed by Messrs Denton Hall)
Michael Bunting
Appendix II
II
Page 5
()
J ardine Air Terminal
Services Limited

Mr Clive Grossman SC
(instructed by Messrs Mallesons Stephen
J aques Solicitors)

(
Mass Transit Railway
Corporation

Messrs Deacons Graham & J ames

666
Ogden Aviation (Hong
Kong) Limited
(
)
Messrs J ohnson Stokes & Master

Ove Arup & Partners
Hong Kong Ltd

Messrs Simmons & Simmons

Swire Engineering
Services Ltd

Mr Robert Whitehead
(instructed by Messrs Simmons &
Simmons)

(

LIST OF EXPERTS IN THE INQUIRY

Experts Commissioned
Appendix III
III
Page 1
664

665
The
Commission

Professor Xiren CAO
Professor Cao has a doctorate in Applied
Mathematics, majoring in control and optimization,
from the Harvard University. He is currently the
Professor of the Department of Electrical and
Electronic Engineering of the Hong Kong
University of Science and Technology. As an
expert in mechatronics, Professor Cao has
extensive industrial experience with major
information technology and manufacturing
corporations in the US and in the Mainland.
(engaged by the Commission)

()

666

Appendix III
III
Page 2
667
()

Dr Ulrich Kipper
Dr Kipper has a doctorate in Physics. He
received his tertiary education from the J ohann
Wolfgang University, Frankfurt. As an expert
of information technology and
telecommunications in airport operation, Dr
Kipper is currently the Senior Project Manager
with the Frankfurt Airport. He was also closely
involved with the planning and design of the
information technology and telecommunications
system of the new Athens International Airport.
He has extensive experience in air traffic control
system as well as a wide range of operational
matters in airport management.

Ulrich Kipper
Kipper J ohann Wolfgang

668

Professor Vincent Yun SHEN
Professor Shen received his M.A. and Ph. D.
degrees in Electrical Engineering from Princeton
University in 1967 and 1969 respectively. He
taught at the Computer Sciences Department of
Purdue University from 1969 to 1985. He also
held visiting positions at Tsing Hua University
(Taiwan) and IBM Corp. (California) during that
period. Professor Shen joined the
Micro-electronics and Computer Technology Corp.
in 1985 to work on problems related to large-scale
software systems development. He later directed
the companys Software Technology Program.

Professor Shen joined the Hong Kong University
of Science and Technology in 1990 as Founding
Head of the Computer Science Department. He
served as Associate Vice President for Academic
Affairs at the university from 1996 to 1997 before
returning to teaching and research at the
department in 1997.

19671969
19691985
()IBM (
)1985
(Microelectronics and Computer
Technology Corp.)
Appendix III
III
Page 3

Mr J ason G YUEN
Mr Yuen received his Bachelor degree in
Architecture from the University of California,
Berkeley. He has served as an airport planning
and construction consultant for thirty years. He
has extensive experience in providing technical
and management related advice on airport
management and design. The airport projects
he has worked on spanned from those in North
America to Asia. In the last six years, Mr Yuen
was heavily involved in the San Francisco
Airport, USA where he chaired boards and
committees ranging from airport construction
programme to computerised airport systems.

30

()
Appendix III
III
Page 4
669
Appendix III
III
Page 5
Parties

670

671
Hong Kong Air
Cargo Terminals
Limited (HACTL)

Mr Max William Nimmo
Mr Nimmo received his Bachelor degree in
Electrical Engineering from the Auckland
University, New Zealand in 1969. As an
experienced manager of technology based
companies in industrial automation,
communication and computer markets, Mr
Nimmo has managed engineering development,
engineering production and sales departments.
His experience covers engineering design,
project management, sales management,
marketing, manufacturing logistics, quality
assurance and financial management. Since
April 1998, Mr Nimmo has been contracted as a
Senior Technical Consultant and Project Manager
for The Coca Cola Amatil Embedded Software
group.
(engaged by HACTL)

Max William Nimmo
Nimmo 1969
19984Nimmo
The Coca Cola Amatil
Embedded Software Group

672
Day 1959 (Holy
Cross College) ()
1962
(Wharton School, University of Pennsylvania)
19721983
1983
1997
3

Mr J erome J oseph J r. Day
Mr Day received his Bachelor degree in Physics
from the Holy Cross College in 1959 and his
MBA degree from the Wharton School,
University of Pennsylvania in 1962. He taught
in the MBA Programmes at The Chinese
University of Hong Kong from 1972 to 1983 and
joined the Hong Kong Baptist University to
establish a Computing Studies teaching
programme in 1983. Mr Day undertook various
jobs at the Hong Kong Baptist University and
has retired from the university since December
1997. Mr Day also served as Chairman of the
Hong Kong Management Associations Hong
Kong Telecommunications Users Group for three
annual terms in the late 1980s.
(engaged by HACTL)

J erome J oseph J r. Day
Appendix III
III
Page 6
Appendix IV
IV
Page 1

LIST OF WITNESSES IN THE INQUIRY

Hearing Day

Date

Witness

Organisation & Position

Day 1

Mon

07/09/98 W1 Mr Richard Alan Siegel

CAD, Director of Civl Aviation
,

Day 2

Tue

08/09/98 W1

Mr Richard Alan Siegel

CAD, Director of Civil Aviation
,

W2 Mr YEUNG Kwok Keung

HACTL, Deputy Managing
Director
,

Day 3

Wed

09/09/98 W2 Mr YEUNG Kwok Keung

Director
,

Day 4

Thu


Director
,

Day 5

Mon


Director
,

Day 6

Tue


Director
,

Day 7

Thu

17/09/98 W3 Dr Henry Duane
Townsend

AA, Chief Executive Officer
,
Day 8

Fri

Townsend

,

670
671

Hearing Day

Date

Witness


Appendix IV
IV
Page 2
Day 9

Mon

Townsend

,
Day 10

Tue

Townsend

,
Day 11

Thu

Townsend

,
W4 Mr SEE Seng Wan

AAT, Chief Executive Officer
,

W5 Mr Allan KWONG Kwok
Hung

J ATS, Assistant General
Manager - Operations
,
-

Day 12

Fri

25/09/98 W5 Mr Allan KWONG Kwok
Hung

J ATS, Assistant General
Manager - Operations
,
-

W6 Mr Samuel KWOK King
Man

HAS, Business Support
Manager

,

Day 13

Mon

28/09/98 W6 Mr Samuel KWOK King
Man

HAS, Business Support
Manager

,

W7 Mr Anthony Crowley
Charter

HACTL, Managing Director
,

Day 14

Tue

29/09/98 W7 Mr Anthony Crowley
Charter

,

672

Hearing Day

Date

Witness


Appendix IV
IV
Page 3
W8 Mr Mackenzie Grant

Ogden, Managing Director
()
,

Day 15

Wed

Charter

,

Day 16

Mon

Charter

,

W9 Mr Gernot Werner Demag, Senior Project
Manager Controls
(),
-

Day 17

Wed

07/10/98 W10 Mr HO Yiu Wing (with)
()
HACTL, Project Manager
Controls
,

W11 Mr LEUNG Shi Min

HACTL, Maintenance Manager
Cargo Handling System
,
-

Day 18

Thu

()
Controls
,


,
-

Day 19

Fri

()
Controls
,

673

Hearing Day

Date

Witness


Appendix IV
IV
Page 4

,
-

W12

Mr J ohnnie WONG Tai
Wah (with)
()

HACTL, General Manager
Operations
,

W13 Mr Peter PANG Tai Hing
(with)
()
HACTL, Manager - Projects
and Administration Operations
,

W14 Ms Violet CHAN Man
Har (with)
(
HACTL, ST1 Systems
Manager
,

W15 Mr Daniel LAM Yuen Hi

HACTL, Operations Computer
Project Manager
,

Day 20

Mon

12/10/98 W12 Mr J ohnnie WONG Tai
Wah (with)
()
Operations
,

(with)
()
,

Har (with)
()
HACTL, ST1 Systems
Manager
,


Project Manager
,

674

Hearing Day

Date

Witness


Appendix IV
IV
Page 5
Day 21

Tue

13/10/98 W16 Mr Hiroshi NAKAMURA
(with)
()

Murata, Project Manager
(),

W17 Mr Tomonobu SAEKI
(with)
()

Murata, Project Engineer (Site
Construction Manager)
(),
()

W18 Mr Shin YAMASHITA

Murata, Testing and
Commissioning Manager
(),

Day 22

Thu

15/10/98 W16 Mr Hiroshi NAKAMURA
(with)
()

Murata, Project Manager
(),

W17 Mr Tomonobu SAEKI
(with)
()

Murata, Project Engineer (Site
Construction Manager)
(),
()

W18 Mr Shin YAMASHITA

Murata, Testing and
Commissioning Manager
(),

W12 Mr J ohnnie WONG Tai
Wah (with)
()
Operations
,

(with)
()
,

Har (with)
()
HACTL, ST1 Systems
Manager
,

675

Hearing Day

Date

Witness


Appendix IV
IV
Page 6

Project Manager
,

Day 23

Fri

16/10/98 W12 Mr J ohnnie WONG Tai
Wah (with)
()
Operations
,

(with)
()
,

Har (with)
()
HACTL, ST1 Systems
Manager
,


Project Manager
,

W19 Mr TSUI Shek Chiu

HACTL, Shift Manager
,

W20 Mr Tony KWAN To Wah

HACTL, General Manager -
Enginneering
,

Day 24

Mon

19/10/98 W20 Mr Tony KWAN To Wah

HACTL, General Manager -
Enginneering
,

W21 Mr Michael Todd
Korkowski
EDS, On Site Project Manager
of FIDS
EDS, -

676

Hearing Day

Date

Witness


Appendix IV
IV
Page 7
Day 25

Tue

20/10/98 W21 Mr Michael Todd
Korkowski
of FIDS
EDS, -

Day 26

Thu

Korkowski
of FIDS
EDS, -

Day 27

Fri

Korkowski
of FIDS
EDS, -

W22 Mr Edward George
Hobhouse

GEC, Project Director
,

W23 Mr Alan LAM Tai Chi
(with)
()
AA, General Manager
(Airfield Operations)
,
()

W24 Ms Rita LEE Fung King

AA, IT Project Manager
, -

Day 28

Mon

26/10/98 W23 Mr Alan LAM Tai Chi
(with)
()
AA, General Manager
,
()


, -

Day 29

Tue

27/10/98 W23 Mr Alan LAM Tai Chi
(with)
()
AA, General Manager
,
()

677

Hearing Day

Date

Witness


Appendix IV
IV
Page 8

, -

W25 Mr TSUI King Cheong

AA, Project Manager
Electrical & Mechanical Works
, -

Day 30

Thu

29/10/98 W25 Mr TSUI King Cheong

AA, Project Manager
, -

Day 31

Fri

30/10/98 W25 Mr TSUI King Cheong

AA, Project Manager
, -

W26 Mrs Vivian CHEUNG
Kar Fay (with)
()
AA, Terminal Systems Manager
,
-

W27 Ms Yvonne MA Yee Fong

AA, Project Manager
Information Resource
Management
,
-

Day 32

Mon

02/11/98 W28 Mr Anders YUEN Hon
Sing (with)
()
AA, Assistant Airfield Duty
Manager
,

W29 Mr CHAN Kin Sing

AA, Assistant Airfield Duty
Manager
,

W26 Mrs Vivian CHEUNG
Kar Fay (with)
()
,
-

678

Hearing Day

Date

Witness


Appendix IV
IV
Page 9

AA, Project Manager
Management
,
-

Day 33

Tue

03/11/98

W26

Mrs Vivian CHEUNG
Kar Fay (with)
()

,
-


AA, Project Manager
Management
,
-

Day 34

Thu

05/11/98 W26 Mrs Vivian CHEUNG
Kar Fay (with)
()
,
-


AA, Project Manager
Management
,
-

W30 Mr Ben Reijers AA, Senior Design Engineer
,

Day 35

Fri

06/11/98 W30 Mr Ben Reijers AA, Senior Design Engineer
,

W31 Mr J ames WONG Hung
Kin (with)
()
NAPCO, Project Manager
,

W32 Mr J han Schmitz (with)
()
NAPCO, Deputy Consultant
Project Manager
,

W33 Mr KWOK Ka Keung NAPCO, Director, NAPCO
,

679

Hearing Day

Date

Witness


Appendix IV
IV
Page 10
Day 36

Mon

09/11/98 W31 Mr J ames WONG Hung
Kin (with)
()
,

()
Project Manager
,

W33 Mr KWOK Ka Keung

NAPCO, Director, NAPCO
,

Day 37

Tue

Kin (with)
()
,

()
Project Manager
,


,

Day 38

Thu

Kin (with)
()
,

()
Project Manager
,


,

Day 39

Fri

Kin (with)
,

()
680

Hearing Day

Date

Witness


Appendix IV
IV
Page 11
()
Project Manager
,


,

W34 Mr Peter Lindsay Derrick
(with)
Preston, Project Manager
Preston,

W35 Mr Gordon J ames
Cumming
EDS, Sub-contract Manager
EDS,

Day 40

Mon

16/11/98 W36 Mrs CHAN Fang Anson

Government, Chief Secretary
for Administration
,

Day 41

Tue

17/11/98 W34 Mr Peter Lindsay Derrick
(with)
Preston, Project Manager
Preston,

W35 Mr Gordon J ames
Cumming
EDS, Sub-contract Manager
EDS,

W37 Mr Dominic Alexander
Chartres Purvis (with)
()
Cathay Pacific, Manager
Customer Services
,

W38 Mr Victor WONG Chu
King (with)
()
Cathay Pacific, Systems
Manager Airport
,

W39 Mr Albert LO Sze Wai
(with)
()
Cathay Pacific, Manager Cargo
Services
,

W40 Mr Peter LEE (with)
()
Cathay Pacific, Manager
Business Improvement
,
681

Hearing Day

Date

Witness


W41 Ms Vanessa LI Chui Fung

Cathay Pacific, Chek Lap Kok
Development Co-ordinator
,

Day 42

Thu

19/11/98 W34 Mr Peter Lindsay Derrick Preston, Project Manager
Preston,

W42 Mr NG Ki Sing

AA, General Manager
Terminal Operations
,
-

Day 43

Fri

20/11/98 W42 Mr NG Ki Sing

AA, General Manager
Terminal Operations
,
-

W43 Mr Douglas Edwin
Oakervee (with)
()

AA, Project Director
,
W44 Mr Chern Heed (with)
()
AA, Airport Management
Director
,

W45 Mr Kironmoy Chatterjee
(with)
()
AA, Head of Information
Technology
,

W46 Mrs Elizabeth Margaret
Bosher

AA, Planning and
Co-ordination Director
,

Day 44

Mon

23/11/98 W43 Mr Douglas Edwin
Oakervee (with)
()

,
()
Director
,

Appendix IV
IV
Page 12
682

Hearing Day

Date

Witness


Appendix IV
IV
Page 13

(with)
()

Technology
,
Bosher

AA, Planning and
,

Day 45

Tue

Oakervee (with)
()

,
()
Director
,

(with)
()
Technology
,

Bosher

AA, Planning and
,

Day 46

Thu

Oakervee (with)
()

,
()
Director
,

(with)
()
Technology
,

Bosher

AA, Planning and
,

Day 47

Fri

Oakervee (with)
()
,

683

Hearing Day

Date

Witness


()
Director
,

(with)
()
Technology
,

Bosher

AA, Planning and
,

Day 48

Mon

Townsend

,
W47 Mr Graham Morton

Guardforce, Project General
Manager
,

Day 49

Tue

1/12/98 W48 Mr Billy LAM Chung
Lun

AA, Deputy Chief Executive
Officer
,

W47 Mr Graham Morton

Guardforce, Project General
Manager
,

Day 50

Thu

3/12/98 W49 Mr LO Chung Hing
(with)
()

AA, Board - Vice Chairman
,

W50 Mr WONG Po Yan

AA, Board - Chairman
,

W51 Mr J ason G YUEN

Expert for the Commission

Day 51

Fri

4/12/98 W51 Mr J ason G YUEN


Appendix IV
IV
Page 14
684

Hearing Day

Date

Witness


Appendix IV
IV
Page 15

Day 52

Mon

7/12/98 W52 Mr Max William Nimmo
(with)

Expert for HACTL

W53 Mr J erome J oseph J r. Day

Expert for HACTL

Day 53

Tue

8/12/98 W52 Mr Max William Nimmo
(with)

Expert for HACTL

W53 Mr J erome J oseph J r. Day

Expert for HACTL

Day 54

Wed

9/12/98 W54 Professor Xiren CAO


W55 Dr Ulrich Kipper (with)


W56 Professor Vincent Yun
SHEN


Day 55

Thu

10/12/98 W55 Dr Ulrich Kipper (with)


SHEN


Day 56

Fri

11/12/98 W55 Dr Ulrich Kipper (with)


SHEN

V
(1)

(1998 6 )
MEMBERS OF THE BOARD OF THE AIRPORT AUTHORITY
(AS AT JUNE 1998)

Mr WONG Po Yan

Chairman

Mr LO Chung Hing

Vice Chairman

Dr Henry Townsend, the Chief Executive Officer

Mr Stephen IP, the Secretary for Economic Services

Miss Denise YUE, the Secretary for the Treasury

Mr KWONG Hon Sang, the Secretary for Works

Mr KWOK Ka Keung, the Director, NAPCO

Mr Richard Siegel, the Director of Civil Aviation

556

Mr J oseph YAM, the Chief Executive, HK Monetary Authority

Mr HO Sai Chu

Mr Anthony LEUNG

V
(2)

Mr Vincent LO

Miss Maria TAM

Dr Peter WONG

Dr Philip WONG

557
CHANNEL OF DOCUMENTARY COMMUNICATION TO ADSCOM

1. Draft ACP monthly
progress reports
2. Monthly progress
reports
3. AAs monthly
construction
reports
4. HACTLs ST1
monthly progress
reports
ADSCOM
Papers and
other
documents
prepared by
AA
Biweekly
Chek Lap
Kok issues
reports
Weekly site reports
from senior engineers
to Chief Coordinator
Situation
Report
on AOR
at CLK
1. Weekly situation
reports
2. ACP monthly
progress reports
3. ADSCOM Papers
and other
documents
prepared by
NAPCO
ADSCOM
NAPCO
Staff on site
Works Bureau
NAPCO
AA
Bechtel
(professional staff
seconded to NAPCO)
6
8
7
A
p
p
e
n
d
i
x

V
I

ORGANIZATION CHART OF THE AIRPORT AUTHORITY (AS AT 31 JULY 1998)

BOARD Chairman
WONG Po Yan

Head of Project
Monitoring
KWEI See Kan

Chief Internal Auditor

William YEUNG

Deputy
Chief Executive Officer
Billy LAM

Legal Director
J osiah KWOK

Planning &
Co-ordination
Director
Elizabeth Bosher

Airport
Management
Director
Chern Heed

Finance &
Commercial
Director
Raymond LAI

Project Director
Douglas Oakervee

6
8
8
Head of
Information
Technology
Kiron Chatterjee

Corporate
Development
Director
Clinton Leeks

Human Resources
Director
Sophia KAO

General Manager -
Chairmans Office
Peter TAM

Chief Executive Officer
Henry Townsend

A
p
p
e
n
d
i
x

V
I
I

V
I
I

Appendix VIII
DIAGRAMMATIC PRESENTATION
OF AIR-CONDITIONING PROBLEMS ON AOD

6 July 1998 Chillers Pumps

low pressure
switch fault

10:15 am
demand :
reduced

flow switch fault

10:24 am insufficient
seawater

11:01 am Restart Chiller No. 4

demand increased

flow switch fault

bypass valve
reduced seawater flow

11:24 am Restart Chiller No. 4

demand increased

2 3 4 5 6 1
1 2 3 4 5
1 2 3 4 5
6
1 3 4 5
4 5
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5
1 2 3
4 5
1 2 3 4 5
5 4 3 2 1
1 2 4 3 5 6
2
3
2 1
5
set Pump No. 2 to manual mode
L L
H H L L L
6
5
4 3 2 1

Legends

L =low speed pumps H =high speed pumps
The yellow boxes indicate that the machine was running.
DIAGRAM SHOWING THE INTER-LINK BETWEEN FIDS AND OTHER SYSTEMS
Aiport

Flight and Airport
Information System
Building
Management
System
Gate
Allocation
System
Baggage
Handling
System
Airport
Invoicing
System
Airspace
Information
System
Airlines &
Handling
Agents
Ramp
Management
System
SITA
Message
Server
Cargo
Information
System
Flight
Information
Display
System
Departure
Control
System
Environmental
Monitoring
System
ATC
SITA
Network
Airport
Operational
Database
7
0
0
A
p
p
e
n
d
i
x

X
I
V

DIAGRAM SHOWING USE OF FIDS

Daily
flight
schedule
Allocation
of
baggage
laterals
Stand
allocation
1 Confirmation
of stands and
frontal gates
2 Input chocks
on/off and
registration
mark
1 Update Flight
Progression
Status
2 Input of ETA,
ATA, ETD,
ATD
3 Confirmation
of check-in
desk allocation
4 Confirmation
of APV Gate
1 Allocate
reclaim belts
2 Progress
baggage status
(unallocated,,
done)
Within radar coverage
(45 mins)
ATA/ATD
ETA
Stand
allocation
Progress flight status
- open/close the flight at
check-in desk
- gate open/boarding/final
call/closed at gate
- desk open/closed at transfer
desk
Seasonal
Schedule
STA/ETA/ATA
STD/ETD/ATD
Origin/Destination
Stand/Gate No.
Aircraft type/
Registration No.
Baggage spur
allocation
Flight information
dissemination via 57 display servers
Flight information
STA/ETA/ATA
STD/ETD/ATD
Origin /Destination
Gate/Exit gate no.
Baggage reclaim belt no.
Check-in desk no.
Check-in area/aisle
Flight status
(Flight plan)
ATA/ATD
ETA
AODB
FDDS
Hong Kong Telecom
Airlines
RHOs
A
p
p
e
n
d
i
x

X
F FI ID DS S
TMS
Check-in Desk Transfer Desk
/ Gate Desk
CUTE/AIRLINES
WORKSTATIONS
Operator
Workstations
AOCC
PTB Displays
(Monitors/LCD)
Passengers
CAD/AODB
Gateway
ATC
Radar
Tracker
Flight information
Operator
Workstations
BHS
Operator
Workstations
ACC
SAC/BHS
AIDB
CTOs
Cathay Pacific
Scheduling Committee
Computer (SCC)
MTRC
Airport Express
6
9
1
FLIGHT INFORMATION EXPERIENCES OF VARIOUS FDDS USERS AND MTRC

Cathay
Pacific
Catering
Services
China
Aircraft
Services Ltd.
J apan Airlines
Company
Limited
Kwoon Chung
Motors
Kowloon Hotel LSG Lufthansa
Sky Chefs
Regent Hotel Swiss Air
Transport Co.
Ltd.
Mass Transit
Railway
Corporation
Purpose of
use of FDDS
To enable
servicing of
aircraft
To enable
servicing of
aircraft
To enable
servicing of
aircraft
To plan bus
times for travel
agency service
For guests and
own operations
To enable
servicing of
aircraft
For guests For own flights
and connecting
flights, parking
position,
baggage
delivery
NB: not a
customer of
FDDS;
information
straight from
AODB.

Information for
inflight check in
(HK, Kowloon),
passengers (Tsing
Yi), meeters (all),
seat back display
Nature of
problems
aside from
inaccurate or
incomplete
information

Disconnection
without notice
No testing data
available pre-
AOD
Satisfactory
service
Shrinkage of
screen
6
9
2

A
p
p
e
n
d
i
x

X
I

P
a
g
e

1

Sky Chefs
Transport Co.
Ltd
Mass Transit
Railway
Corporation
Cathay
Pacific
Catering
Services
China
Aircraft
Services Ltd.
J apan Airlines
Company
Limited
Kwoon Chung
Motors
Types of
information
subscribed to
(those in bold
italics are
those
incompletely
received on
airport
opening)
STA
ETA
ATA
STD
ETD
ATD
Flight
number
Flight status
Destination
Bay
Gate
Aircraft type
Handling
agent
STA
ETA
ATA
STD
ETD
ATD
Flight
Aircraft type
Origin
Destination
Bay
Flight status

STA
ETA
ATA
STD
ETD
ATD
Flight
number
Origin
Destination
Bay
Gate
Baggage
reclaim
Check in aisle

Generally not
updated; cannot
recall exactly
which types of
data not
completely
received
STA
ETA
ATA
STD
ETD
ATD
Flight
number
Origin
Destination
Gate
Check in
aisle
Hall
Remarks

Generally not
updated;
cannot recall
exactly which
types of data
not completely
received

STA
ETA
ATA
STD
ETD
ATD
Flight
number
Origin
Destination
Gate
Check in
aisle
Remarks

Generally not
updated; cannot
recall exactly
which types of
data not
completely
received
STA
ETA
ATA
STD
ETD
ATD
Flight number
Flight status
Destination
Bay
Gate
Aircraft type
Handling agent

STA
ETA
ATA
STD
ETD
STD
Status
ETA
ETD
Bay
Connecting
flights bay
Baggage
reclaim

Screen blank
on Day 1.
Otherwise,
wrong
information
STA
ETA (if long delay)
STD
ETD (if long delay)
Flight number
Origin
Destination
Check in desk
status
Time now
Remarks

[No ATA or ATD]
6
9
3

A
p
p
e
n
d
i
x

X
I

P
a
g
e

2

Cathay
Pacific
Catering
Services
China
Aircraft
Services Ltd.
J apan Airlines
Company
Limited
Kwoon Chung
Motors
Sky Chefs
Transport Co.
Ltd.
Mass Transit
Railway
Corporation
Dates of
incomplete
receipt
Weeks 1 & 2 Days 1 & 2 -
nothing at all
for Day 1 and
Day 2 am
First few days Weeks 1, 2, 3 Week 1 Week 1:
incorrect
information
Week 2:
information
after plane
arrived
3 months First few days Day 1: no
information until
1000
19.9.98: outdated
information from
1430 to 1730
28.9.98: outdated
information from
1719 to 1915
Dates of
resumption of
receipt
Week 4 Day 2 pm.
Accuracy
improved
from about
Day 10
Unsure; about
Week 2
around Day 8 Week 2 or
Week 3
29.9.98

Week 2
How
information
received in
meantime
Phone/fax
from airlines;
physical
check
Phone/fax
from airlines;
physical
check
ATC, AOCC,
AA FIDS
control room in
PTB

Fax from HKT Calling airlines Fax from HKT
four times a
day between
Day 1 and Day
10
Calling airlines Engineers
monitored
ATC frequency
then called
office

Whether HKT
advised
rebooting
Yes No reboot
needed
Yes System always
required
rebooting [not
specified
whether HKT
advised]
Yes No No Yes
Whether
rebooting
helped
No NA Sometimes No NA NA Did not try Yes but not
always
6
9
4

A
p
p
e
n
d
i
x

X
I

P
a
g
e

3

Cathay
Pacific
Catering
Services
China
Aircraft
Services Ltd.
J apan Airlines
Company
Limited
Kwoon Chung
Motors
Sky Chefs
Transport Co.
Ltd.
Mass Transit
Railway
Corporation
HKT_s
explanation of
problems
None HKT
responsible
only for
transfer of
data, not
responsible
for source of
data
Did not ask Kwoon Chung
only have a
dial up line,
which is
unstable.
Normal and
essential to
reboot. Unless
get leased line,
stability not
guaranteed
FDDS itself
received
outdated data
from CLK
database

Screen
shirnkage:
Kowloon Hotel
chose wrong
URL (web site
address) and
obtained wrong
display format
None None Flight
information
from AA
database

No No No No No No Whether
problems
foreseen
Noticed shortage
of phone/data
lines on 5.7.98
and therefore
thought there
might be
problem

A
p
p
e
n
d
i
x

X
I

P
a
g
e

4
6
9
5

DIAGRAM SHOWING CHS OF HACTL WITH ITS FIVE LEVELS

Goods Vehicles COSAC Engineering
Operations Planning Rostering
Machinery/Field Devices
Machinery Control
Logistics Control
Personnel Level 5

Level 4
Level 3
Level 2
A
p
p
e
n
d
i
x

X
I
I

Level 1

Appendix XIII
(Page 1)

PICTURES SHOWING THE PROBLEM BAG AREA
IN THE BAGGAGE HALL

Appendix XIII
(Page 2)

Appendix XIII
(Page 3)

DIAGRAM SHOWING THE INTER-LINK BETWEEN FIDS AND OTHER SYSTEMS
Aiport

Flight and Airport
Information System
Building
Management
System
Gate
Allocation
System
Baggage
Handling
System
Airport
Invoicing
System
Airspace
Information
System
Airlines &
Handling
Agents
Ramp
Management
System
SITA
Message
Server
Cargo
Information
System
Flight
Information
Display
System
Departure
Control
System
Environmental
Monitoring
System
ATC
SITA
Network
Airport
Operational
Database
7
0
0
A
p
p
e
n
d
i
x

X
I
V

Appendix XV

DIAGRAM SHOWING THE ALLOCATION
OF FIDS RELATED COMPONENTS
OF AIRPORT OPERATIONS INTO PROBLEM AREAS

MMI
Database
TMS FIDS
ACC
ACC
AOCC
AOCC
CUTE
CUTE
SAS
SAS
Display-
Server
Display-
Server
Display-
Server
Display-
Server
Monitors and LCD Boards
AODB
AODB
FDDS-
Server
FDDS-
Server
FDDS Users
CAD
Gateway
CAD
Gateway
AIDB
AIDB
SSC
SSC
Radar
Tracker
Radar
Tracker
P4
P3
P2
P1
FIDS Clients
FIDS-
Host
BHS
BHS

Legends:
P1 Data Generation
P2 Data Processing
P3 Data Transmission
P4 Data Display

BRIEF SUMMARY OF SOME AIRPORT OPERATIONAL STATISTICS
Kai Tak New Airport at CLK
Day 1
Mon
Day 2
Tue
Day 3
Wed
Day 4
Thu
Day 10
Wed
28 December 98
to 3 J anuary 99
1 Week in J uly 6 J uly 1998 7 J uly 1998 8 J uly 1998 9 J uly 1998 15 J uly 1998 (average)
Number of Flights 1997
- Incoming 213 227 220 240 225 227
- Outgoing 207 227 220 240 225 227
Incoming Flights - Actual Time Vs Scheduled Time

- Early Arrival & On Time 51%
*
32% 34% 46% 48% 46%
- Delay
(1) Within 15 Minutes 7% 20% 21% 23% 27% 29%
(2) Within 30 Minutes 23% 34% 35% 36% 41% 41%
(3) Within 60 Minutes 36% 48% 53% 47% 49% 48%
(4) More Than 60 Minutes 13% 20% 13% 7% 3% 6%
Average Delay for Incoming Flights (Hours)
[Excluding Early & On-Time Flights]
0.4 hr 0.4 hr 0.8 hr 0.6 hr 0.6 hr 0.4 hr 0.5 hr
Outgoing Flights Actual Time Vs Scheduled Time

- Delay Within 15 Minutes 0% 7% 6% 15% 47% 78%
- Delay More Than 60 Minutes 87% 62% 34% 25% 6% 5%
Average Delay for All Outgoing Flights (Hours) 0.5 hr 2.6 hr 1.7 hr 0.9 hr 0.7 hr 0.4 hr 0.3 hr
Total Passengers (In +Out) 86,000 84,000 91,000 84,000 86,000
Number of Departure Bags Left at the End of the Day
(i.e. bags that missed their flight on that day)
5,000 6,000 2,000 1,400 108 -
Time of First Bag Arrival in Baggage Hall After Aircraft Landing (Random Samples)

- Earliest 20 mins 17 mins 8 mins 4 mins
- Average 50 mins 47 mins 24 mins 12 mins
- Latest 1 hr 20 mins 2 hr 4 mins 55 mins 22 mins
Time of Last Bag Arrival in Baggage Hall After Aircraft Landing (Random Samples)

- Earliest 40 mins 41 mins 10 mins 10 mins
- Average 1 hr 7 mins 1 hr 16 mins 33 mins 25 mins
- Latest 1 hr 30 mins 2 hr 33 mins 1 hr 25 mins 42 mins
Total Number of Departure Bags Processed (Originating and Transfer)
[Excluding Arrival Bags]
20,000 24,000 26,000 27,000 32,000 -
Airport Cargo Throughput (HACTL, AAT, Express Cargo)

2,699 tonnes

3,579 tonnes
A
p
p
e
n
d
i
x

X
V
I

*Holding time of aircraft on the taxiway is not included.

New Airport Report

Uploaded by

Copyright:

Available Formats

New Airport Report

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

New Airport Report

Uploaded by

Copyright:

Available Formats

Report of the Commission of Inquiry on

the New Airport

Chief Internal Auditor

You might also like