Risk and Safety
in
Engineering
Prof. Dr. Michael Havbro Faber
Swiss Federal Institute of Technology
gy
ETH Zurich, Switzerland
Swiss Federal Institute of Technology
Contents of Today's Lecture
• Th organisation
The i i off the
h lecture
l – practical
i l stuff
ff
• Why risk and safety in engineering?
• Decision problems in engineering
• Examples
• The lecture program
Swiss Federal Institute of Technology
Organisation of the Lecture
• Course webpage address is: http://www.ibk.ethz.ch/fa/education/ws_safety/index
• Available on course webpage:
Lecture notes for the entire course ((non-printable
p version))
Exercises
Exercise Solutions
Print edition of lecture notes for the entire course is available for a cost of CHF 45.
PowerPoint presentations for each lecture will be uploaded on the webpage the
latest one day before the respective lecture.
• Support will be available – you are welcome to contact Prof.
Prof Michael Faber in room
HIL E 23.3 or contact Harikrishna Narasimhan in room HIL E 13.1 or by email
(
[email protected] )
Swiss Federal Institute of Technology
Examination
• Th exam for
The f the
th course is
i an orall examination
i ti
- the emphasis of the exam is to ensure that a fundamental understanding of
the area of risk and safet
safety in engineering is acquired
acq ired b
by the st
students
dents
• The confirmation (or “testat”) for admission to the examination is fulfilled by
attending the lectures of the course.
• Doctoral students need to take the oral examination in order to g
get credit points
p
for this course.
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• What do engineers do ?
- Plan, design, build, maintain and decommission
Infrastructure
Roads, water supply systems, tunnels, sewage systems,
waste deposits, power supply systems, channels
Structures
houses, hospitals, schools, industry buildings, dams, power
plants,, wind turbines,, offshore platforms
p p
- Safeguard
people
environment SUSTAINABLE DEVELOPMENT !
assets
from natural and man made hazards
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• What do engineers do?
Is what we are doing of any relevance for society?
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• E
Examples
l off what
h t we help
h l to
t develop
d l
Golden Gate Bridge - USA Øresund bridge - Denmark
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• E
Examples
l off what
h t we help
h l to
t develop
d l
Big Dig Boston/USA
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• E
Examples
l off what
h t we help
h l to
t develop
d l
Hoover Dam - USA
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• E
Examples
l off what
h t we help
h l to
t develop
d l
Hong Kong Island - China
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t natural
t l hazards
h d
Tornados and strong winds
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t natural
t l hazards
h d
Earthquakes
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t degradation
d d ti
Corrosion Fatigue
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t accidents
id t
Fires Explosions
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t lack
l k off knowledge
k l d
Over load Design error
Swiss Federal Institute of Technology
Engineering Decision Making for Society
• H l i tto control
Helping t l risks
i k due
d to
t malevolence
l l
Bomb explosions Airplane impacts
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• What are engineers working with ?
Real problems – the real world - nature
Concrete
Gravity Ice Dynamics
Waves
Snow New materials
Soil Temperature
Water Rocks
Waste
Air Chemicals
Wind Electricity
Steel
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• How do engineers work with the real world ?
Mathematics Physics
Models/hypothesis
We model the real world to the “best” of our knowledge
g
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• How do engineers use knowledge ?
In a p
perfectly
y known world
Models
Decisions
Costs/Benefits
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• How do engineers establish knowledge ?
Data Experience
Real world
Models
Accuracy/Uncertainty
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• How do engineers use knowledge ?
Data Experience
Models
Models are not precise
Uncertainty Data are not sufficient
WHY ? Natural variabilityy
Experience is subjective
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
• How do engineers make decisions
Data Experience
Models
Decisions
C t /B
Costs/Benefits
fit
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
All activities
ti iti are associated
i t d with
ith uncertainties
t i ti
Activities could be:
- Transport
- Work
- Sport
S t
but also
- Production of energy
- Exploitation of resources
- Construction and operation of production
and infrastructure p
projects
j
- Research and development
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Every day
E d we must make
k d
decisions
i i in
i regard
d to activities
i i i associated
i d
with uncertainties
Car driving
Cooking Smoking
Mountain
climbing
Crossing the
Surfing
street
Stock
trading
Every one of these activities is associated with uncertainties
p
We all have an opinion regarding
g g the associated risks
We have gut feelings !
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Disasters and accidents have always occurred
Some examples
Tacoma Narrows
Narrows, Washington,
Washington 1940
Fort Mayer, Virginia, 1908
Open questions
- did we realise the risks ?
- are the consequences acceptable ?
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Disasters and accidents have always occurred
Some examples
Kobe, 1995
Open questions
- did we realise the risks ?
- are the consequences acceptable ?
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Disasters and accidents have always occurred
Some examples
Minneapolis, Minnesota, 2007
Open questions
- did we realise the risks ? N
New Y
York,
k 2001
- are the consequences acceptable ?
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Disasters and accidents have always occurred
Some examples
Open questions
- did we realise the risks ?
Hurricane Katrina, New Orleans, 2005
- are the consequences acceptable ?
Swiss Federal Institute of Technology
Why Risk and Safety in Engineering?
Risk
Ri k assessment, t within
ithi the
th framework
f k off decision
d i i analysis,
l i provides
id a
basis for rational decision making subject to uncertain and / or
incomplete information
Thereby we can take into account, in a consistent manner, the
prevailing uncertainties and quantify their effect on risks
Thus we may find answers to the following questions
- How large is the risk associated with a given activity ?
- How may we reduce and / or mitigate risks ?
- How much does it cost to reduce and / or mitigate risks ?
- What risks can we accept – what can we afford ?
Swiss Federal Institute of Technology
Definition of Risk
Risk is a characteristic of an activity relating to all possible events nE
which may follow as a result of the activity
The risk contribution REi from the event Ei is defined through the product
between
the event probability PEi
and
the consequences of the event CEi
The risk associated with a given activity RA may then be written as
nE nE
R A = ∑ REi = ∑ PEi ⋅ C Ei
i =1 i =1
Swiss Federal Institute of Technology
Decision Problems in Engineering
Uncertainties
U t i ti mustt be b considered
id d in
i the
th decision
d i i making
ki throughout
th h t all
ll
phases of the life of an engineering facility
Planning g and Investigations
g and
feasibility study tests
Uncertainties Design
T ffi volume
Traffic l Idea &
Concept
Loads • Safety
Safety of
ofpersonnel
personnel
Manufacturing
Resistances • Safety
Safety
yy of
ofenvironment
environment
(material, soil,..)
Degradation processes • Economic feasibility
Execution
Service life
Manufacturing costs
Decommissioning Operation &
Execution costs maintenance
Decommissioning
costs
Swiss Federal Institute of Technology
Example – Decommissioning of the Frigg Field
• Th Frigg
The F i Field
Fi ld – built
b il 1972-1978
1972 1978
- TCP2
- TP1
- CDP1
According to international
conventions the structures
must be decommissioned
Each structure :
Weight
W i ht : 250000 t
Costs : 200 – 600 million CHF
• N
None off th
the platforms
l tf were designed
d i d for
f decommissioning
d i i i !
Swiss Federal Institute of Technology
Decision Problems in Engineering
• S
Structural
l Design
D i
Exceptional structures are often associated with structures of
“Extreme Dimensions”
G t Belt
Great B lt Bridge
B id under
d C Construction
t ti C
Concept
t drawing
d i off th
the T
Trollll platform
l tf
Swiss Federal Institute of Technology
Decision Problems in Engineering
• S
Structural
l Design
D i
or associated with structures fulfilling
“New and Innovative Purposes”
Concept drawing of Illustrations of the ARIANE 5 rocket
Floating Production, Storage and Offloading unit
Swiss Federal Institute of Technology
Decision Problems in Engineering
Before During After
Optimal allocation of available Damage reduction/Control Rehabilitation of infrastructure
resources for risk reduction functionality
Emergency help and rescue
- strengthening Condition assessment and
- rebuilding After quake hazards updating of reliability and risks
in regard to possible earth- Optimal allocation of resources
quakes for rebuilding and strengthening
Swiss Federal Institute of Technology
Decision Problems in Engineering
• I
Inspection
i and
d Maintenance
M i Planning
Pl i
Due to
- operational loading
- environmental exposure
structures will always to some
degree be exposed to degradation
processes such as
- fatigue
- corrosion
- scour
- wear
Swiss Federal Institute of Technology
Decision Problems in Engineering
• I
Inspection
i and
d Maintenance
M i Planning
Pl i
For industrial facilities inspection
and maintenance is also an
important issue with regard to:
reduction of production downtime
safety of workers
safeguarding the environment
Swiss Federal Institute of Technology
Decision Problems in Engineering
• I
Inspection
ti and
d Maintenance
M i t Planning
Pl i
For industrial facilities inspection
and maintenance is also an
important issue with regard to:
reduction of production downtime
safety of workers
safeguarding the environment
Swiss Federal Institute of Technology
Decision Problems in Engineering
Swiss Federal Institute of Technology
Decision Problems in Engineering
Swiss Federal Institute of Technology
Decision Problems in Engineering
New emerging challenges
have necessitated
exceptional requirements
with regard to construction
managementt and d safety
f t
Swiss Federal Institute of Technology
The Risk Based Decision Process
Swiss Federal Institute of Technology
Define Context
The Risk Based Decision Process and Criteria
Define System
Risk
Ri k assessmentt supports
t d
decision
i i
making subject to uncertainties
Identify Hazard
Scenarios
- what might go wrong
- how can it happen
-how to control it
The theoretical basis for risk based
decision making is the theory of
Analysis of Analysis of
decision analysis Consequences Probability
The main task is to optimally manage
risks in terms of life safety, economic Identify Risk Analyse
Scenarios Sensitivities
losses as well as potential damages to
the environment
Assess Risks
Monitor and Risk
Review Treatment
Swiss Federal Institute of Technology
The Risk Based Decision Process
The risk assessment can be categorized
according to the degree of detail
Level 1 : Categorization of risk
Analysis of the probabilities of occurence
of critical events assessments is usefull!
Levell 2 :
L
Analysis of the probabilities of occurence
Documents to what detail
of critical events and the corresponding the risk assessment was
consequences
performed!
Level 3 :
As for level 2, but with consideration of
human errors as well as potential loss of
lives – if relevant
Swiss Federal Institute of Technology
Life Quality
• D
Demographical
hi l indicators
i di t
- Gross domestic product (GDP) per capita
35000
30000
pita (USD PPP))
25000
20000
15000
GDP/cap
10000
5000
0
1950 1960 1970 1980 1990 2000 2010
Year
Swiss Federal Institute of Technology
Life Quality
• D
Demographical
hi l indicators
i di t
- Life expectancy at birth
90
80
Expected life at birth
h
70
60
50
40
30
20
10
0
1880 1900 1920 1940 1960 1980 2000
Year
Swiss Federal Institute of Technology
Life Quality
• Lif quality
Life lit can be
b assessed
d through
th h the
th
“Human Development Index” (UNO) or
“Life Quality
Q alit Index”
Inde ” (JCSS)
as a function of GDP,
GDP life expectancy,
expectancy free time,…
time
GDP and life expectancy are important components
10-20% of the GDP is reinvested into life saving activities
Maintenance of infrastructure costs around 10-15
10 15 per cent of the available GDP
Cost efficiency is of tremendous importance!
Swiss Federal Institute of Technology
Individual Risks
Based on statistical information the contribution to life risks of different
activites may be assessed
A ti it /
Activity/course Number
N b off ffatalities
t liti
per hour per 108 persons
Mountaineering (international) 2700
Aviation transport (international) 120
Deep sea travling 59
Auto traffic 56
Mining (coal) 21
Construction work 7.7
Manufacturing/production 2.0
Accidents at home 2.1
Accidents at home (healthy persons) 0.7
Fires at home 0.1
Structural failures 0.002
Swiss Federal Institute of Technology
Individual Risks Causes of death probability/year probability/lifetime
Transport Accidents 1.66E-04 1.28E-02
- Pedestrian 2.13E-05 1.64E-03
• Accidents
A id t accountt only
l for
f - Pedal
P d l cyclist
li t
- Motorcycle rider
2 78E 06
2.78E-06
1.07E-05
2 14E 04
2.14E-04
8.24E-04
4% of all deaths - Car occupant 5.24E-05 4.05E-03
- Occupant of heavy transport vehicle 1.31E-06 1.01E-04
- Bus occupant 1.30E-07 1.00E-05
Illness such
s ch as heart fail
failure,
re - Animal rider or occupant of animal-drawn
animal drawn vehicle
- Occupant of railway train or railway vehicle
4 07E 07
4.07E-07
9.12E-08
3 14E 05
3.14E-05
7.04E-06
cancer and strokes - Air and space transport accidents 3.22E-06 2.49E-04
Non-transport Accidents 1.90E-04 1.47E-02
account for 58% - Falls 5.27E-05 4.07E-03
- Struck by or against another person 1 58E-07
1.58E-07 1 22E-05
1.22E-05
- Accidental drowning and submersion 1.15E-05 8.88E-04
- Exposure to electric current, radiation,
temperature, and pressure 1.51E-06 1.17E-04
- Exposure to smoke, fire and flames 1.16E-05 8.96E-04
- Uncontrolled fire in building or structure 9.38E-06
9.38E 06 7.24E-04
7.24E 04
- Contact with venomous animals and plants 2.14E-07 1.65E-05
- Earthquake and other earth movements 9.82E-08 7.58E-06
- Storm 1.89E-07 1.46E-05
- Flood 1.23E-07 9.48E-06
- Lightning
g g 1.54E-07 1.19E-05
- Alcohol 1.06E-06 8.20E-05
- Narcotics and hallucinogens 2.28E-05 1.76E-03
Intentional self-harm 1.07E-04 8.26E-03
Assault 7.12E-05 5.49E-03
Legal intervention 1.39E-06 1.07E-04
Operations of war 5.96E-08 4.60E-06
Complications of medical and surgical care 1.06E-05 8.18E-04
Swiss Federal Institute of Technology
Individual Risks
• O
Occupational
ti l risks
i k
% of Fatalities per
Occupation
p sector
employees 100 000 employed
100,000
Private industry 90 4.2
- Agriculture, forestry and fishing 14 22.7
- Mining 2 23.5
- oil and gas exploitation 1 23 1
23.1
- Construction 20 12.2
- Manufacturing 10 3.1
- Transportation and public utilities 16 11.3
- Wholesale trade 4 40
4.0
- Retail trade 9 2.1
- Finance, insurance, and real estate 2 1.0
- Services 12 1.7
Government 10 27
2.7
- Federal (including resident armed forces) 2 3.0
Total 100 4.0
Swiss Federal Institute of Technology
Individual Risks
• N t
Natural
l disasters
di t
1 Insured
Victims 2,3 Event Country
losses
300 000 – 1 4 . 1 1 . 1 9 70 S torm and flood catas trophe Banglades h
250 000 – 2 8 . 0 7 . 1 9 76 Earthquake in T angs han (8 . 2 R ichter s cale) C hina
138 000 3 2 9 . 0 4 . 1 9 91 T ropical cyclone Gorky Banglades h
60 000 – 3 1 . 0 5 . 1 9 70 Earthquake (7 . 7 R ichter s cale) Peru
50 000 156 2 1 . 0 6 . 1 9 90 Earthquake
q in Gilan Iran
25 000 – 0 7 . 1 2 . 1 9 88 Earthquake in A rmenia A rmenia, ex-US S R
25 000 – 1 6 . 0 9 . 1 9 78 Earthquake in T abas Iran
23 000 – 1 3 . 1 1 . 1 9 85 V olcanic eruption on Nevado del R uiz C olombia
22 000 233 0 4 . 0 2 . 1 9 76 Earthquake (7 . 4 R ichter s cale) Guatemala
19 118 1063 1 7 . 0 8 . 1 9 99 Earthquake in Izmit T urkey
15 000 100 2 6 . 0 1 . 2 0 01 Earthquake (moment magnitude 7 . 7 ) in Gujarat India, Pakis tan
15 000 106 2 9 . 1 0 . 1 9 99 C yclone 0 5 B devas tates Oris sa s tate India, Banglades h
15 000 – 0 1 . 0 9 . 1 9 78 Flooding follow ing mons oon rains in northern parts India
15 000 530 1 9 . 0 9 . 1 9 85 Earthquake (8 . 1 R ichter s cale) M exico
15 000 – 1 1 . 0 8 . 1 9 79 Dyke burs t in M orvi India
10 800 – 3 1 . 1 0 . 1 9 71 Flooding in Bay of Bengal and Oris sa s tate India
10 000 234 1 5 . 1 2 . 1 9 99 Flooding, muds lides , lands lides V enezuela, C olombia
Swiss Federal Institute of Technology
Individual Risks
• M l
Malevolence
l - terrorism
t i
Victims1 Insured Date Event Country
losses2
T error attack
tt k agains
i t W T C , Pentagon
P t andd
at leas t 3 0 0 0 19 000 1 1 . 0 9 . 2 0 01 US A
other buildings
Bombing of US Marine barracks
300 2 3 . 1 0 . 1 9 83 Lebanon
and French paratrooper bas e in Beirut
300 6 1 2 . 0 3 . 1 9 93 S eries of 1 3 bomb attacks in M umbai India
PanA m Boeing 7 4 7 cras hes at Lockerbie
270 138 2 1 . 1 2 . 1 9 88 UK
due to bomb explos ion
T w o s imultaneous bomb attacks on US
253 0 7 . 0 8 . 1 9 98 K enya
embas sy complex in Nairobi
Bomb attack on government building in
166 145 1 9 . 0 4 . 1 9 95 US A
Oklahoma C ity
Hijacked Ethiopian A irlines Boeing 7 6 7-2 6 0 Indian Ocean
127 45 2 3 . 1 1 . 1 9 96
ditched at s ea
118 1 3 . 0 9 . 1 9 99 Bomb explos ion des troys apartment R us sia
block in Mos cow
100 0 4 . 0 6 . 1 9 91 A rs on in arms w arehous e in A ddis A baba Ethiopia
100 6 3 1 . 0 1 . 1 9 99 Bomb attack on C eylinco Hous e in C olombo S ri Lanka
1Dead or missing 2 Excluding liability losses; in USD m, at 2001 price level
Swiss Federal Institute of Technology
Sources of Risks in Engineering
Any activity carries a risk potential
It is important
p that this potential
p is
fully understood
Only when the risk potential is
fully understood can rational
decisions be identified and
implemented
Swiss Federal Institute of Technology
Sources of Risks in Engineering
Case where
C h th risk
the i k potential
t ti l
was not fully appreciated
The Tjörn bridge in Sweden
S eden
Just after construction
Swiss Federal Institute of Technology
Sources of Risks in Engineering
Case where
C h th risk
the i k potential
t ti l
was not fully appreciated
The Tjörn bridge in Sweden
S eden
A few weeks later
Swiss Federal Institute of Technology
Sources of Risks in Engineering
F il
Failures in
i structural
t t l engineering
i i
3 1 2 2 2 2
100
10
29 21
Retrofit / Disposal
80 39 40
62 Usage
60 Execution
88
40 69 77
58 58
20 37
0
high Industrial Traffic Dams Rest All
Type of Construction rising buildings structures cases
building
Number of samples 362 152 75 51 52 692
Swiss Federal Institute of Technology
Sources of Risks in Engineering
F il
Failures in
i structural
t t l engineering
i i
100 5 4 8
9 11 Rest
80 11 6 20
4 Interrim structures
12 3 6
60 14 Technical
19 2
40 Site and installation
72
44 50 Fitting
20
Structure
0
Structural Cost of Damage of
damage damages persons
692 692 60
Swiss Federal Institute of Technology
Sources of Risks in Engineering
F il
Failures in
i structural
t t l engineering
i i
Primary causes of structural failure
Poor construction procedures 54.3
Inadequate connection elements 47
I d
Inadequate
t load
l d behavior
b h i 42.2
Unclear contract information 23.5
Contravention of instructions 21.8
Unforeseeable events 7.1
Errors in design calculations 2.5
Reliance on construction accuracy 1.8
Complexity of project system 1.2
0 10 20 30 40 50 60
Frequency (%)
Swiss Federal Institute of Technology
Aim of the lecture
• G
General
l introduction
i t d ti to
t risk
i k and
d safety
f t in
i engineering
i i
• Basics and principles of risk based decision analysis
• Theory and technical aspects of risk assessment
• Methods of reliability analysis and introduction to the JCSS probabilistic model code
• Introduction to time variant and systems reliability analysis and assessment of structural
robustness
b t
• Illustration of computer tools for risk and reliability anaylsis
• Applications for design, assessment and maintenance planning
• Understanding of engineering optimization and risk acceptance
Swiss Federal Institute of Technology
Organisation of the Lecture
DATE SUBJECT DATE SUBJECT
16.09.09 Introduction and overview 04.11.09 The JCSS probabilistic model code
- Life quality, risks and decision making - Probabilistic modeling of resistances
- Hazards and causes of failures - Probabilistic modeling
- of loads
23.09.09 Review of p probability y theoryy and statistics 11 11 09
11.11.09 Systems reliability analysis and robustness
- Uncertainties in engineering modeling - Series and parallel system analysis
- Basic data analysis - Structural systems analysis
- Random variables and processes - Robustness assessment of structures
- Engineering model building
18.11.09 Time variant reliability analysis
30.09.09 Engineering decision analysis
- Prior decision analysis - The
Th Poisson
P i andd the
th Normal
N l processes
- Posterior decision analysis - The first excursion problem and mean out crossing rates
- Pre-posterior decision analysis - Approximations in time variant reliability analysis
07.10.09 Risk assessment in civil engineering - Treatment of non ergodic variables and random sequences
- Procedures of risk assessment 25.11.09 Bayesian Probabilistic Nets (BPN) in risk assessment
- Scenario identification and analysis - Basic theory of BPN
- System representation - Application of BPN in risk assessment
- Probabilities and consequences - Application of BPN in large scale risk management
- Multi scale indicator based risk assessment 02.12.09 Reliability based structural design and assessment
14.10.09 Classical reliability analysis - Safety formats of design codes
- Failure
il rate data
d - Calibration of design codes
- The reliability function - Reliability updating
- Updating of failure rates based on data - Assessment of existing structures (and SIA 269)
- Failure rate functions and the bath-tub curve 09.12.09 Risk based inspection and maintenance planning
21.10.09 Structural reliability analysis - The basic problem
- First
Fi t andd Second
S d Order
Od R Reliability
li bilit Methods
M th d - Modeling
d li off degradation
d d i processes
- Monte Carlo and importance sampling - Inspection quality and the PoD concept
28.10.09 Software tools for reliability analysis - Generic approaches to inspection planning
- Presentation of the COMREL program 16.12.09 Optimal decision making and risk acceptance criteria
- Exercises in the PC-lab - Optimality in engineering decision making
- The
Th ALARP principle
i i l for
f acceptability
t bilit
- The Life Quality Index and acceptable life safety
- Societal life saving costs and willingness to pay
Swiss Federal Institute of Technology
