ADDIS ABABA INSTITUTE OF TECHNOLOGY

SCHOOL OF GRADUATE STUDIES

POSTGRADUATE PROGRAM IN RAILWAY ENGINEERING
ADDIS ABABA UNIVERSITY

MASTER'S DEGREE THESIS

ON

TRAIN OPERATION SYSTEM DEVELOPMENT FOR ADDIS

ABABA LIGHT RAIL TRANSIT

By
Befikadu Diro

Advisor
Dr.Birhanu Beshah

Submitted to Addis Ababa University, Addis Ababa Institute of

Technology, School of Mechanical and Industrial Engineering, in partial
fulfillment for the award of the Degree of Master of Science
In Mechanical Engineering (Railway Engineering Stream).

September, 2014
 ADDIS ABABA INSTITUTE OF TECHNOLOGY
SCHOOL OF GRADUATE STUDIES
POSTGRADUATE PROGRAM IN INDUSTRIAL ENGINEERING
ADDIS ABABA UNIVERSITY

TRAIN OPERATION SYSTEM DEVELOPMENT FOR

ADDIS ABABA LIGHT RAIL TRANSIT

By: Befikadu Diro

Approved by Board of Examiners

_____________________ __________ _________

Chairman of Department Signature Date

Graduate Committee (DGC)

Dr.Birhanu Beshah_ __________ _________

Advisor Signature Date

________________ ________ _________

Internal examiner Signature Date

________________ ________ _________

External examiner Signature Date

ii
DECLARATION

I hereby declare that the work which is being presented in this thesis entitled
“TRAIN OPERATION SYSTEM DEVELOPMENT FOR ADDIS ABABA
LIGHT RAIL TRANSIT” is original work of my own, has not been presented for
a degree of any other university and all the resource of materials used for this
thesis have been duly acknowledged.

_______________________ _____________

Befikadu Diro Date

This is to certify that the above declaration made by the candidate is correct to the
best of my knowledge.

____________________ ________________

Dr.Birhanu Beshah Date

(Advisor)

iii
Acknowledgements

First and for most, I would like to give my magnificence and respect to the Almighty GOD for
his giving me strength, patience, invaluable cares and support throughout the course of my life
and completion of this work, and frontward too.

Next, I would like to express deepest gratitude to my advisor Dr. Birhanu Beshah for his constant
motivation, guidance, support, supervision and constructive suggestion throughout this thesis
work. I want also to thank Dr-Ir. Eshetie Berhan Director of Research, Technology Transfer and
Industry Linkage for his valuable advice on simulation model development for the train schedule
analysis. On the other hand my glad goes to my parents and friends for supporting me useful idea
and suggesting priceless comments to keep the research on going.

I gratefully acknowledge the staff of Ethiopia Railway Corporation and Addis Ababa LRT
project office for providing me the necessary data and information for the input of this thesis
work.

iv
Abstract

This research paper mainly deals with the train operation system development for railway
transport of Addis Ababa LRT. The prime justification for the introduction of a Light Rail
System in Addis Ababa is the provision of major improvements to public transport in the city,
particularly with reference to greatly increased carrying capacity, reduced pollution and
improved the service of operation. To achieve this, system analysis approach is presented for the
investigation of train operations in railway stations and routes based on demand, network,
timetable (train schedule), and rolling stock and crew analysis data. Network and train schedule
analysis are the main focus of this paper which are solved using a simulation model programmer
language Arena5. The simulation model is formulated in the way that it represents the
appropriate actual working system of the Addis Ababa LRT. For the constructed simulation
model, the “Program” is well made and run. The simulation model combines the two routes
together and the number of waiting time result shows the stations operations are busy from
Lideta to Meskel Square including the route between both stations where the two routes separate
to east and south lines. Therefore, these route and stations are the bottle neck of the Addis Ababa
LRT corridors. Testing the real operation by simulation, Delay minimization and resource
utilization, waiting time for each station and the route bottle neck identification are the major
output of the research. This research is done depending on the actual data for the Addis Ababa
light rail transit and its application is more for the future operation coming soon. Finally, the
operation data are verified and the simulation model output is valid.

Keywords: Train schedule, operations, station, route, speed, dwell time

v
Table of Contents
Acknowledgements ................................................................................................................................ iv
Abstract................................................................................................................................................... v
Tables .................................................................................................................................................... ix
Figures .................................................................................................................................................... x
Notation ................................................................................................................................................. xi
CHAPTER ONE ......................................................................................................................................... 1
Introduction............................................................................................................................................. 1
1.1. Concepts of Train and Its Operations ........................................................................................ 1
1.2 Statement of the Problem ......................................................................................................... 2
1.3 Objective ................................................................................................................................. 3
1.4 Scope of the study .................................................................................................................... 4
1.5 Significance of the Study.......................................................................................................... 4
1.6 Structure of the Thesis .............................................................................................................. 5
CHAPTER TWO .................................................................................................................................... 6
Literature review ..................................................................................................................................... 6
2.1 Railway stations ........................................................................................................................... 8
2.2 Operational Areas of railway station ............................................................................................ 9
2.2.1 Station platform.................................................................................................................. 10
2.2.2 Depots ................................................................................................................................ 11
2.2.3 Track Switch, Turnout and Crossing ................................................................................... 11
2.2.4 Passenger Information ........................................................................................................ 12
2.3 Public Rail Transport Planning and operations ........................................................................... 12
2.3.1 Line planning Problem ................................................................................................... 13
2.3.2 Train Timetabling Problem ............................................................................................. 14
2.4 Single track and Double track operations.................................................................................... 15
2.5 Arena Overview ......................................................................................................................... 17
2.5.1 The Arena window ............................................................................................................. 17
2.5.2 Arena Modules ................................................................................................................... 19
2.5.3 Using Arena ....................................................................................................................... 19
2.5.4 Arena Hierarchy and Analyzers .......................................................................................... 20

vi
2.6 Railway Safety System .............................................................................................................. 21
2.6.1 Railway Accident ........................................................................................................... 21
2.7 Summary of the literature review ............................................................................................... 25
CHAPTER THREE ............................................................................................................................... 27
RESEARCH METHODOLOGY ........................................................................................................... 27
3.1 Research Process ....................................................................................................................... 27
3.2 Research Approaches ................................................................................................................. 28
3.3 Data Collection Approaches ....................................................................................................... 29
3.4 Software used ............................................................................................................................ 29
3.5 Data Analysis Method ................................................................................................................ 30
CHAPTER FOUR ................................................................................................................................. 32
Analysis and Discussion ........................................................................................................................ 32
4.1 Addis Ababa East-West and North-South Corridor ................................................................. 32
4.2 Passenger Transport Operations Analysis ............................................................................... 35
4.2.1 Demand Analysis ........................................................................................................... 36
4.2.2 Rolling Stock Analysis ................................................................................................... 36
4.2.3 Crews Analysis............................................................................................................... 37
4.2.4 Network analysis ............................................................................................................ 38
4.2.5 Train Schedule Analysis ................................................................................................. 39
4.3 Making Simulation model ...................................................................................................... 40
4.3.1 Simulation Model Results and Discussions for North to South and West to East line....... 47
4.3.2 Simulation Model Results and Discussions for South to North and East to West line....... 51
4.3.3 Verification ........................................................................................................................ 55
4.3.4 Validation of the model ...................................................................................................... 55
CHAPTER FIVE ................................................................................................................................... 56
Conclusions and Recommendation ........................................................................................................ 56
5.1 Conclusions ................................................................................................................................... 56
5.2 Recommendation ........................................................................................................................... 57
5.3 Future work ................................................................................................................................... 57
5.4 Contribution of the Thesis .............................................................................................................. 58
References............................................................................................................................................. 59
Appendix .............................................................................................................................................. 59

vii
Appendix A. North-South and West-East Line Simulation Results ..................................................... 59
Appendix B. South-North and East-West Line Simulation Results ..................................................... 66

viii
Tables

Table 1, Hourly Distribution of the Demand Based on the LRT Corridor Data .......................... 36
Table 2, Technical data for Addis Ababa LRT........................................................................... 41
Table 3, Distance between stations for N-S line ......................................................................... 42
Table 4, Distance between stations for E-W line........................................................................ 43
Table 5, Passenger flow of Addis Ababa Light Rail N-S in morning peak hours. ....................... 44
Table 6, Passenger flow of Addis Ababa Light Rail E-W in morning peak hours. ...................... 45
Table 7, some simulation results for the N-S and W-E direction ................................................ 49
Table 8, some simulation results for the reverse direction .......................................................... 53
Table 9, Result summary of waiting time. ................................................................................. 59
Table 10, Result summary of total waiting time. ........................................................................ 60
Table 11, Result summary of number of train waiting. .............................................................. 61
Table 12, Result summary of resource utilization. ..................................................................... 62
Table 13, Result summary of time per Entity. ............................................................................ 63
Table 14, Result summary of train in and out. ........................................................................... 64
Table 15, Result summary of Queue waiting time. .................................................................... 65
Table16, Result summary of waiting time. ................................................................................ 66
Table 17, Result summary of total waiting time. ........................................................................ 67
Table 18, Result summary of number of train waiting. .............................................................. 68
Table 19, Result summary of resource utilization. ..................................................................... 69
Table 20, Result summary of time per Entity. ............................................................................ 70
Table 21, Result summary of number of train in and out............................................................ 71
Table 22, Result summary of Queue waiting time. .................................................................... 72

ix
Figures

Figure 1, Operational areas of railway station ............................................................................ 10

Figure 2, the Arena window ...................................................................................................... 18

Figure 3, Ethiopian new rail network plans................................................................................ 23

Figure 4, Addis Ababa LRT ...................................................................................................... 24

Figure 5, Research process frame works .................................................................................... 28

Figure 6, Railway operation factors ........................................................................................... 30

Figure 7, Overall Schematic Diagram of Addis Ababa LRT ...................................................... 34

Figure 8, Route corridors conceptual model for Addis Ababa LRT ............................................ 39

Figure 9, Simulation model over view ....................................................................................... 40

Figure 10, Passenger distribution ............................................................................................... 46

Figure 11, Simulation Model from Piazza to Kalit and Torhyloch to Ayat ................................. 48

Figure 13, Simulation Model from Kalit to Piazza and Ayat to Torhyloch ................................. 52

x
Notation

ERC - Ethiopian Rail Way Corporation

ILPTP – Integrated Line Planning and Timetabling Problem

LRT - Light Rail Transit

LRV - Light Rail vehicles

Mc - Motor car

MIP - mixed integer programming

MoTaC – Ministry of Transport and Communications

O-D – Origin Destination matrices

PESP – Periodic Event Scheduling Problem

PIS - Passenger Information System

PID - Passenger Information Display

SAMFOST – Simplified Analytical Model for Single tracks

SIMAN – Simulation Model Analysis

TNV – Trein -NummerVolgsystemen ( The Dutch train detection data tools)

Tp - Trailer without driver’s cab and with pantograph

TVEM – Timetable Variant Evaluation Model

xi
CHAPTER ONE

Introduction

Ethiopia has launched the construction of a 5,000 km railway network which aims to link the
capital, Addis Ababa, to various regions of the country which is part of the country’s five-year
transformation plan (Corporation, 2011). The new railroad network is planned to have at least 8
main routes that extends to all compass points. The Addis Ababa LRT is also the part of the plan
which is now under construction. Therefore studying about and preparing for the operation is
indispensable to make the city transportation system modern and efficient.

1.1. Concepts of Train and Its Operations

A train is defined as one or more railway vehicles capable of being moved. It may consist of a
locomotive (sometimes more than one) to provide power with various unpowered vehicles
attached to it. It may also consist of a multiple unit, i.e. several vehicles formed into a fixed
formation or set, which carry their own power and do not require a locomotive. A train may be
only a locomotive running light (deadheading) to a point elsewhere on the railway. It can be
classified as passenger carrying, freight or, rarely nowadays, mixed. The Addis Ababa train is
low-floor articulated 6-axle modern trams, consisting of three modules, bi-directional driving
vehicles. Two tramcars shall be able to operate with double heading. The Addis Ababa train
formation is Mc+Tp+Mc, where Mc module is motor car with driver’s cab and Tp module is
trailer without driver’s cab and with pantograph (Limited C. R., Technical Specification of
Vehicles, 2013)

The system development of train operation analysis approach is presented for investigation of
train operations in railway stations based on infrastructure (network) analysis, schedule
(timetable) analysis, and rolling stock analysis and crew analysis data. The estimated blocking

1
times, buffer times and track occupancies are compared with real operations data recorded
automatically by track circuits. Statistical analysis of train operations clearly reveals that the
trains operate at lower than design speed. Furthermore, the dwell times at platform tracks are
systematically extended due to hinder by other trains and behavior of railway personnel (Hansen,
2005)

The scheduled headway between arrival and departure of some pairs of trains at critical route
nodes running times at junctions are generally exceeded and often leads to route conflicts. The
quality of timetable design and train operations throughout the route would be improved
significantly if the feasibility of the scheduled arrival and departure times at major transfer
stations was proven by a detailed estimation of the blocking and buffer times based on observed
running times and delays during operations. The buffer time at junctions and level crossings
should reflect the distribution of real train speeds and blocking times. Therefore, it is important
to study and research about train operation system development for the new coming of Addis
Ababa light rail transit.

1.2 Statement of the Problem

Even though there is no actual train operation currently in Ethiopia, it is important doing on the
railway operation system development for the railway personnel and society before the real
operation started. In train operation root conflict, train delay and human behavior are the most
common problems in railway industry. Therefore the problem areas that were studied in this
paper are: train delay, unimpressive utilization of resources, train traffic control, Crew
unpunctuality and testing train operation by simulating for Addis Ababa LRT. Therefore, the
purpose of this study is to identify train operation factors, minimize train delay, identify the route
bottle neck and give effective transport service and to answer the following basic research
questions in the case Addis Ababa LRT.
 What are the factors that cause train schedule delay?
 How to use the schedule or time table effectively?
 How train operation punctuality increase?
 Can the Addis Ababa LRT operate with the design operation data (simulation checkup)

2
  The need of understanding modeling and simulation from concept to expected result
 Well communication with simulation package-Arena Software

1.3 Objective

General objectives

The general objective of this research is to analysis and develops the train operation systems for
Addis Ababa Light Rail Transit.

Specific objectives

The specific objectives of this study are:

1. To model and simulate the two Addis Ababa LRT corridors (Piazza to Kaliti and
Torhyloch to Ayat) using the programmer language software Arena5.

2. To identify the factors that cause train delay throughout the Addis Ababa LRT operation
systems.

3. Entail the appropriate train traffic control system and man power for the system
operations, and to raise awareness regarding operation safety issue.

3
1.4 Scope of the study

The Addis Ababa LRT covers the East-West and North-South transport corridors. The initial
phase of the project is 34km length and expected to extend up to 74km in the long term plan. The
East-West corridor passes through the city center from the eastern peripheries of the city to the
west. The initial phase of this route is 17.1km length and expected to extend up to 37.7km in the
long term plan. The North-South corridor is a heavily-used route; it is utilized to access Africa’s
largest open market at Merkato. The initial phase of this route is 16.97km length and expected to
extend up to 36km in the long term plan. The study covers all systems of the east-west line from
Ayat to Torhyloch and the north-south line from Piazza to Kaliti. Therefore the paper of this
study focuses on the initial routes of the Addis Ababa LRT.

1.5 Significance of the Study

Railway will produce economically feasible construction, environmental friendly, safe,

comfortable and competent mode of transportation system in the world. The objective of good
train operation is to use the route, the rolling stock and crews in the most effective way since
train is an expensive piece of things and the infrastructure of a railway is its most expensive one.
The operation system development in this work is analyzed by using the Arena5 simulation
model, station stopping time and route travel time ,maximum and minimum train speed,
identifying the route bottle neck, inter station distance and time between consecutive train
travelling and resource allocation were mostly accessed to make the Addis Ababa LRT train
operation full. Therefore this paper is aimed to show the Addis Ababa LRT operation by
simulating and identifies the factors those can affect the train operation and cause delay. The city
Light Rail Transit design operation data are also checked by simulation check up.

4
1.6 Structure of the Thesis

This research is composed of five chapters. The first chapter is the introduction part that
contains: introduction and justification of the study, the problem statement, research questions,
the objective, scope, the significance of the study, and structure of the research. In the second
chapter, the literature reviews of related train operation, network analysis, timetable scheduling
analysis and the software Arena5 tools are briefly discussed. In the third chapter, the
methodology that is used for doing the research is introduced. In this chapter, the research
process, research approaches and data synthesis and analysis method are discussed. In the fourth
chapter, the operation factors, system development analysis and the results of the study are
briefly presented with its corresponding discussions. Finally in chapter five, the conclusion and
recommendation for train operation system development, future work, and contribution of the
thesis are briefly deliberated.

5
 CHAPTER TWO

Literature review

Railways are complex technical systems. They are often considered to be static, stiff and
inflexible. As long as only constant, non-dynamic parts, such as the infrastructure, are
considered, the system is quite easy to understand. But the reality for train operation system is
different. The variance in different parameters makes the system difficult. Some examples:

 The infrastructure is adjusted and complemented all the time. Most of the adjustments are
minor, but these changes nonetheless imply that important factors for operation vary over
time.
 The timetable creates a well defined structure. However, the capacity is utilized
differently every day since the actual timetable varies from one day to another due to
delays, extra trains and cancelled trains. Therefore the planned timetable is modified from
time to time. The principles for this capacity allocation also change over time.
 The available capacity, which is an important condition for the timetable, varies over
time. Failures, construction works, accidents and delays all make the available capacity
vary over time.
 The railway system is used and operated by humans. Human behavior varies naturally
from one time to another. Train crews, dispatchers and passengers all contribute to this
variance.

All these variances make the railway system complex and interesting to analyze. Modeling the
railway system means modeling all these variables. To find general relations in this noise of
superposed variances is not easy. That means systematic method needs to understand the train
operation system development. Moreover, strong interactions between factors can be expected,
that make the task of analysis even more difficult. (Stockholm, 2010)

6
Train delays are one of the most important performance indicators of railway operations, which
result from many related factors, such as the layout of the infrastructure, the occupancy of the
infrastructure, the variation of running times, arrivals and departures, and the traffic control.
First, the infrastructure must provide enough capacity and smooth routes. The timetable is the
basic plan for train operation. On the one hand, the timetable should ensure the efficient and
balanced use of the infrastructure. On the other hand, it should be flexible enough to cope with
stochastic disturbances during operations. Suitable recovery times in a timetable are very
important to reduce train delays, whereas buffer times can decrease delay propagation. Stochastic
disturbance of operations cannot be eliminated and will always occur, but its statistical
distribution can be grasped by analyzing historic and actual train operation data. To some extent,
this will help to forecast and prevent train delays. Traffic control is the final means to supervise
and improve the train punctuality. The effectiveness of traffic control in case of perturbations,
however, depends on the quality of the actual schedule, the time and precision of conflict
detection and the available means for disposition.

The interdependencies among the related factors are strengthened with the increase of traffic
density and train speed, not only, which requires a further enhancement of the precision of time
calibration a more comprehensive and accurate analysis of train operations. Therefore Railway
operation planning, train operation and the safety issues are the most indispensable one which
have to integrate in the train operation system analysis. Most modern research into railway
operation concerns single track and double-track systems or entire networks of double-track lines
and stations that connect them. The others are focus on timetable scheduling and train detection
data. That means the train operation in railway stations is investigated systematically based on
network, timetable and train detection data. Train detection data and train delays at stations can
be analyzed with a high precision of about a second by means of the recently developed tools
TNV-Prepare and TNV-Filter (Goverde R.M.P, 2001a)Earlier publications (Goverde, 2001b)
deal with the distribution of train delays in some stations of the Dutch Railways.

7
2.1 Railway stations

Railway station is the selected place on a railway line, where trains halt for one or more of the
following purposes: (ARORA, 2008A)

 For exchange of passengers.

 For exchange of goods.
 For control of train movements.
 To enable the trains on a single line track to cross from opposite directions.
 To enable the following express trains to overtake.
 For taking diesel or coal and water for locomotives.
 For detaching engines and running staff.
 For detaching or attaching of compartments and wagons.
 For sorting of bogies to form new trains, housing of locomotive in loco-sheds etc.
 It emergencies in ease of dislocation of track due to rains or accidents etc.

Railway stations and routes play a vital role for passengers, non-travelling users, the staff and the
communities in which they are located. They serve the growing needs of an increasingly mobile
population and are used by a wide range of users. With many journeys made by rail each day,
Network Rail recognizes the importance of its customers and the need to put passengers first. To
ensure that the expectations of its customers are met in full, the Guide places passengers at the
forefront of the design process, acknowledging that all elements of station design should be
considered to help meet their needs. In order to understand fully the perceptions of these users it
is important to engage with them from the outset of the design process to ensure that their
requirements are understood and integrated. Stations are primarily a means of access to the
national rail network, in turn acting as a gateway to or from other destinations. They have
become increasingly important as destinations in their own right as places to shop, work, or
simply to meet family and friends. Most stations also function as interchange hubs, providing
connections between different transport modes. (Goggin, 2011)

8
2.2 Operational Areas of railway station

There are four main functional areas typically housed in most stations as a general. They are;
core, transition, peripheral, and administrative areas. The main users of these four functional
areas are generally passengers, staff, guest and visitors.

 Core areas: focus on processing passengers. It includes ticketing, information, baggage

handling, reclaiming, and waiting.
 Transit areas: connect transit facilities in the core areas to the transportation modes. It
includes restrooms, telephones, and commercial spaces.
 Peripheral areas: support circulation outside the main buildings. They often include
platforms, tracks, and vehicle service spaces.
 Administrative area: control both traffic and station management. These areas can be
isolated from other facilities or inserted among them.

These operational areas of railway station are mostly used in interregional big stations where it is
also impartial applicable for light rails transits.

9
Figure 1, Operational areas of railway station

2.2.1 Station platform

Train stations comprise of the building(s), site access, parking, tracks, platforms, and all
appurtenances necessary to conduct transportation. Station platform is a raised structure or area
within a station providing access to or from a train. Station platforms shall be located on straight
track unless the particular geographical characteristics of the site and the characteristics of the
railway at the proposed location of the platform do not provide a reasonable opportunity for

10
achieving. The Addis Ababa LRT platforms (stations) will be located at-grade, below ground
and on structures. The East-West route has 22 platforms, among which 5 are elevated platforms,
1 underground platform and 16 ground platforms. On the other hand the North-South route has
22 platforms, among which 9 are elevated platforms （5 common platforms at the common line,
2 underground platforms and 11 ground platforms. (Limited C. R., Preliminary Design for
Ethiopia Addis Ababa LRT phase I, 2009)

2.2.2 Depots

The Depots are located near the extremities of the East-West line, and at the southern end of the
North-South Line of Addis Ababa LRT. The Depots will provide facilities for the servicing,
repair, maintenance and stabling for the light rail vehicles (LRVs). The main workshop will be
equipped with specialized machinery to maintain vehicle components such as wheels, bogies,
brakes and electronic controls. The Depot areas should allow additional space for an increase in
the stabling sidings as the number of LRVs increases. (Limited C. R., Technical Specification of
Vehicles, 2013)

2.2.3 Track Switch, Turnout and Crossing

Switch: It is a pair of moveable rails with their fastenings and operating rods, providing a
connection over which to divert the movement of rolling stock and other on-track equipment.
Switches may be operated by power operated switch and lock movements; electrically locked
hand-operated machines; or hand-operated trail able switch stands, depending on the location
and purpose of the switch.

Turnout: It is a particular grouping of two tracks joined together with a frog and switch so
arranged to allow for the transfer of rolling stock and on-track equipment to cross from
one track to another. For a turnout on main track, its rail type shall be identical with main

11
track. (Gebeyehu, 2012) This switch and turnout are located at the start points (Piazza,
Torhyloch, Ayat and Kaliti) and Lideta and Meskel Square where the two line merges and
separate for the common route use.

2.2.4 Passenger Information

Information systems on stations are variously referred to as a Passenger Information System

(sometimes referred to as PIS) or Passenger Information Display (PID). Passenger information
systems are essential for any railway. One of the most common complaints by passengers on
railways is the lack of up to date and accurate information. There are two types of information
constant and instant. Constant information can be described as that which describes the services
and fares available and which changes only a few times a year or less. For instant systems, it can
be assumed that passengers require knowing: (ARORA, 2008A)

 The time now

 The destination and expected time of arrival of the next train
 The stations served by this train
 Where the train will stop - for variable length trains
 Other destinations served from this station and from which platform

2.3 Public Rail Transport Planning and operations

Public rail transport planning and operations are a highly complex task according to Thomas
Lindner, [2000]. Too many objects interact with each other to be manageable simultaneously.
Various sub problems of different nature like network design, scheduling or routing occur, and
the solutions of most of those sub problems depend on the solutions of the other sub problems.
Analytical models get more and more important in supporting managerial decision-making. The
fundamental mathematical model he used here is the Periodic Event Scheduling Problem
(PESP).

12
Thomas also divides the operation planning in to several steps due to the tremendous size of the
public rail traffic system. In a first step, the passenger demand has to be analyzed. As a result, the
amount of travelers wishing to go from certain origins to certain destinations is known. This is
important, to determined lines or routes. The second step is train schedule planning. All arrival
and departure times of the lines are fixed, now engines and coaches have to be assembled to
trains, which are assigned to lines. The third step is the crew management, which means the
distribution of personnel in order to guarantee that each train is equipped with the necessary
staff.

2.3.1 Line planning Problem

The line planning problem is defined as follows: for a given railway infrastructure (given as a set
of interconnected blocks and stations), traveling times on each block and passenger demand
determine the set of lines and frequencies so as to satisfy some operational constraints, such as
demand satisfaction, and optimize some operational goal, such as costs and service levels.

(Bussieck. M.R, 1996) Present a MIP model for choosing a set of lines and their frequencies so
as to maximize the number of direct travelers. The authors assume that passengers between any
two stations use a shortest path with respect to travel time or travel distance. The lines are chosen
so that on each segment a predetermined frequency of trains is ensured. The model does not take
into account the operational costs of the lines. These costs are somewhat bounded by the
frequency constraints, namely, the total number of trains that can be run is bounded. The direct-
travelers variables represent the number of direct travelers between each pair of stations on each
line, this leads to a huge model which is unsolvable for real-world instances. The model size is
reduced significantly by changing these variables to represent only the number of direct travelers
between each pair of stations, without taking into account what line is used (assuming that the
distribution of travelers to lines is carried out later). The reduced model is solvable in a
reasonable time after introducing some cutting planes. The solution of the reduced model is a
lower bound on the optimal solution of the original problem. In addition, a feasible solution for
the original problem can be derived from this optimal solution.

13
 2.3.2 Train Timetabling Problem

The train timetabling problem is defined as follows: for a given set of lines, frequencies and
possibly departure times of each train, determine the arrival and departure time at each block and
station such that safety constraints are satisfied and possibly other goals are achieved. Train
timetabling is divided in to two. The first one is cyclic timetabling which is mainly used in
passengers' trains scheduling where passengers' demands can be estimated far in advance,
journey times vary between a few minutes to a few hours and the frequencies of lines are
relatively high. In a cyclic timetable, each trip is operated repeatedly once or a few times during
each cycle. From the passengers' point of view, such timetables are more convenient because all
they need to remember are the times in a cycle when trains arrive at their station. The second one
is non-cyclic timetabling which is mainly used in freight trains scheduling where the demand for
trains is irregular and may be unknown in advance and the travel time of each train may vary
from a few hours to a few days.

(W.Ukovich, 1989) Propose a general mathematical model for scheduling activities of periodic
type. The Periodic Event Scheduling Problem (PESP) is to set the times (in a period) of a given
set of periodic events so as to satisfy a set of periodic constraints. Each periodic constraint
represents the relationship between an ordered pair of periodic events. The events are scheduled
for one period (cycle time) so that later this period can be repeated over the planning horizon. In
addition to the general model, the authors present more specific cases on which the model can be
applied such as: periodic job-shop, transportation scheduling and traffic light scheduling. The
authors conclude with a suggestion to adapt the general model to specific problems in order to
improve computational efficiency.

14
 2.4 Single track and Double track operations

Infrastructure configuration, timetable design and delays play important roles in the
competitiveness of railway transport operation. This is especially true on single-track lines where
the run times and other timetable related parameters are severely restricted by crossings (train
meetings). The crossings also make the lines’ operation more sensitive to disturbances. Double-
tracks with mixed traffic also show these features.

(E.R, 1974) Presents a simple model that can be used to describe the scheduled delay as a
function of the traffic intensity. The main timetable assumption, that makes the study general, is
that trains’ departure times are independent random variables that are uniformly distributed over
the defined time period. Given this randomized timetable, the time costs (delays) for crossings
and over takings are calculated. One conflict at a time is identified and resolved and so the trains
are treated pair-wise. (A.J, 1987) Presents a method to find the best locations of crossing stations
for a given timetable and line alignment. The method focuses on frequent small delays that can
be managed by longer crossing stations (i.e. partial double-tracks) and time supplements. Longer
delays are handled by secondary crossing stations. The study concludes that single-tracks work
quite well, as long as infrastructure and timetable are coherent and delays limited. (Higgins,
1995) Address the overall timetable reliability in terms of schedule adherence. They present a
model that can be used to prioritize investments designed to improve timetable reliability. Using
their model it is possible to compare effects of track, station and rolling stock projects.

Double-track lines operated with mixed traffic show properties similar to those of single-tracks.
In this case overtaking implies scheduled delays as well as risk of delay propagation. (Wendler
E., 2008) Provides a general introduction to capacity analysis based on queuing theory. He
considers both scheduled delays and knock-on delays and emphasizes the usefulness of queuing
theory for long and medium term studies, where the requested train paths are not known in detail.
(R.J, 2001) Provide a model to forecast knock-on delays due to differences in speed of trains on
double-track line sections. These speed differences may be caused by different train services, or
by primary delays. Train arrivals are modeled as a stochastic process, requiring train frequencies

15
of each service only. Running time distributions for each train service is then obtained solving a
system of linear differential equations. The model seems to be useful for analysis of isolated
track sections between overtaking stations and the approach is well suited for long term planning
where no detailed timetable data is available. All together, these contributions help to enhance
future railway operation and thus increase the competitiveness of the railway even more.

Stockholm, 2010 shows the advantages of simple models for analysis of railway operation. He
presents two tools for infrastructure and timetable planning. He shows how the infrastructure can
be analyzed through fictive line designs, how the timetable can be treated as a variable and how
delays can be used as performance measures by using a simple analytical model SAMFOST and
a combinatorial, mathematical model TVEM. A simplified analytical model SAMFOST
(Simplified Analytical Model for Single-tracks), has been developed to calculate the crossing
time as a function of infrastructure configuration, vehicle properties, timetable and delays for
two crossing trains. TVEM, Timetable Variant Evaluation Model, is a generic model that
systematically generates and evaluates timetable variants on double-track lines. This method is
especially useful for mixed traffic operation where the impact of the timetable is considerable.

(Kaspi, 2010) presented the train planning problem by integrating the line planning problem and
timetabling using ILPTP (integrated line planning and timetabling problem) and O-D (origin–
destination matrices). He divided the operation planning problems into several sub-problems like
Line Planning, Timetabling, Plat forming, Rolling Stock Circulation and Crew Planning. In his
study, he focuses on quality of service aspects derived by decisions made during the line
planning and timetabling phases. The quality of service is measured in terms of total time spent
by the passengers in the railway system, including waiting time at the stations of origin,
connections and travel time. He formulate an integrated Line planning and Train Timetabling
model, devise methods to encode a feasible solution and to quickly evaluate it. He then applies
the Cross Entropy meta-heuristic in order to solve the problem by taking the Israeli railway
system as a test bed.

16
(Hansen, 2005) Presented the investigation of train operations in railway stations based on
network, timetable and train detection data. The tools TNV-Prepare and TNV-Filter were applied
for the analysis of train detection data at the stations. The estimated blocking times, buffer times
and track occupancies are compared with real operations data recorded automatically by track
circuits according to their research. Their paper presents the objectives and approach applied for
a detailed analysis of the variations of train operations at railway stations. It describes the way of
estimating the scheduled and real use of track capacity by means of blocking time graphs in
order to determine the critical track sections and the resulting time margins. Then, a detailed
analysis of arrival and departure delays and speed is performed on the basis of automatically
recorded train detection data in order to estimate the corresponding distributions in the case of a
major Dutch station.

2.5 Arena Overview

Arena is a flexible, powerful modeling and simulation software tool that allows the user to
construct a simulation model and run experiments on the model. The software generates several
reports as a result of a simulation run.

2.5.1 The Arena window

After Arena starts, the computer screen shows the Arena window. Figure 2 below shows what
the Arena window typically looks like. There are three main regions that can be identified in the
main Arena window:

17
 Figure 2, the Arena window

 The Project Bar, located on the left side of the Arena window and below the tool bars.
The project bar contains three panels: the Basic Process panel, the Report panel, and the
Navigate panel. Every panel contains several modules that are used in constructing
simulation models.
 The Model window flowchart view, located on the right side of the Arena window and
below the tool bars. This view is actually the workspace for the simulation model. It will
contain all the model graphics: the flowchart, animation, and other drawings of the
model.
 The Model window spreadsheet view, located on the right-hand side and below the
flowchart view. The spreadsheet view shows the model data.

18
 2.5.2 Arena Modules

The construction of simulation models with Arena, involves using modeling shapes, called
modules, from the Basic Process panel. These modules are used as the building blocks in
constructing a simulation model. There are two types of modules on a panel:

 Flowchart modules. The user places flowchart modules in the model window and
connects them to form a flowchart, which describes the logic of model. The most
common flowchart modules are: Create, Process, Decide, Dispose, Batch, Separate,
Assign, and Record.
 Data modules. The user can edit these modules in the spreadsheet interface. These
modules are not placed in the model window. The most common data modules are:
Resource, Queue, Variable, Schedule, and Set.

A model is constructed in the model window flowchart view, which is the workspace of the
model. From the Project bar, the user drags the flowchart modules needed into the model
flowchart and connects the modules together. To edit a flowchart module, the user either double-
clicks on the module shape and fills in the form, or edits its data in the spreadsheet. The
definition of a data module is carried out by clicking on the shape of then module in the Project
bar to activate its spreadsheet. Then the user can edit the data of the module. Flowchart modules
exist both in the model workspace and as a row in the spreadsheet. Data modules exist only in
spreadsheet view. Data modules can be added or deleted via the spreadsheet, while flowchart
modules can only be added or deleted by placing the object or removing the object from the
model workspace.

2.5.3 Using Arena

To build a simulation model and to carry out simulation runs with Arena, a user performs the
following steps:

19
1 Construction of a basic model. Arena provides the model window flowchart view, which is a
flowchart-style environment for building a model. The user selects and drags the flowchart
module shapes into the model window and connects them to define process flow of the
model.
2 Adding data to the model parameters. The user adds actual data (e.g., processing times,
resource demands, others) to the model. This is done by double-clicking on module icons and
adding data.
3 Performing a simulation run of the model. The user runs the simulation and examines the
results.
4 Analysis of the simulation results provided by the automatic reports of Arena. The user can
expand the statistics.
5 Modifying and enhancing the model according to the user needs.

2.5.4 Arena Hierarchy and Analyzers

In Arena modules are defined utilizing other modules. The base modules in Arena hierarchy
represent SIMAN language. A specific set of modules forms a panel. A number of panels are a
template (e.g. SIMAN, Arena, Basic process, etc.)

Arena Input Analyzer

The Input Analyzer is a standard component of the Arena environment. It is used to determine
the quality of fit of probability distribution functions to input data. It generates sets of random
data. The data files processed by the Input Analyzer typically represent the time intervals
associated with a random process

Arena Process Analyzer

The Process Analyzer assists in the evaluation of alternatives presented by the execution of
different simulation model situations. This is useful to simulation model developers, as well as
decision makers. The Process Analyzer is focused at post-model comparison of models. The role

20
of the Process Analyzer then is to allow for comparison of the outputs from validated models
based on different model inputs.

Interpreting Results/Reports in Arena

There are ten reports provided by Arena. They are: category overview, category by replication,
activity areas, entities, processes, queues, resources, transfers, user specified and frequencies. In
general, Arena is a very versatile integrated simulation development tool. Constructing a
simulation model involves identifying one or more flow objects known as entities that flow
through the system and then building a flowchart of the model using Arena’s flowchart modules.
The model is then enhanced by editing some of the data modules. Arena provides a simple
method to setup the simulation parameters and the model input parameters. A variety of reports
with the corresponding simulation results is produced. (Garrido, 2009)

2.6 Railway Safety System

The railroad is still undeniably one of the safest modes of transport in levels of safety compared
with the other mode of transport. However, as complex systems, despite the high technical and
organizational effort to prevent accidents, sometimes such accidents happen with catastrophic
consequences. It is recognized that such accidents rarely have simple causes and behind the
obvious or direct causes, such as human errors or indirect concealed causes. The main objective
of the safety work is therefore to avoid these causes, if possible, or to control them. Improving
network and effective operations are the main technological improvements to railroad safety.

2.6.1 Railway Accident

Accident is an occurrence in the course of working of railway which does or may affect the
safety of the railway, its engine, rolling stock, permanent way and works, fixed installations,
passengers or servant or which affect the safety of others or which does or may cause delay to
train or loss to the railway. The types of railway accident that may occur are:

21
  Derailment- may be at plain track, curves or junctions.

 Collision- may be head-on, rear-end, and grade (level) crossing

 Fire and violent eruption (including sabotage or terrorism) and

 Others (falls from train, collision with people on tracks)

The main causes of such accident can be divided into three groups
a. Inadequate technical activities.

b. Ineffective communication and organization (management).

c. Weaknesses in safety culture.

Example of such causes include

 Mechanical problems (signal failures, equipment failures and train control)

 Human factors (train operation)

 Infrastructure defect (track, roadbed and structure)

 Environment or external factors (weather-related, obstruction)

Therefore railway operation safety is the most indispensable one that everybody (the staff,
society, management and technical persons) have to give attention and fulfill their responsible as
much as required. This is important for the railway operation of Ethiopia as it is on construction.

Ethiopia is constructing a new national wide rail networks which contains both single and double
tracks even if most of the lines are single tracks according to the government plan. But the Addis
Ababa LRT is totally double tracks. Now Ethiopia has launched the construction of a 5,000 km
railway network in two phases which aims to link the capital, Addis Ababa, to various regions of
the country and 34km Addis Ababa LRT which is part of the country’s five-year transformation
plan and the newly established Ethiopian Railway Corporation is responsible for the supervision
of the construction. The new railroad network is planned to have at least 8 main routes that
extends to all compass points. The rail line will expected to link not less than 49 urban centers,

22
where railway stations are to be established. The proposed rail line crosses the borders of all
regions, except Gambella. The network connects, among others, the Chartered Cities Addis
Ababa & Dire Dawa, 7 of the 9 State capitals, and towns bordering Sudan, Kenya and Djibouti.
The construction of the railroads required high finance, while creating ‘job opportunity for
several hundred thousands of people. Though the construction of the new railroad system is to be
conducted in two phases, it is basically one program which is pivotal to the renaissance of the
nation. The Addis Ababa LRT also conducted in two phases. The initial stage covers 34km while
the long-term will cover 74km in two main lines namely East-west and North-South.
(Corporation, 2011)

Figure 3, Ethiopian new rail network plans

23
Figure 4, Addis Ababa LRT

24
 2.7 Summary of the literature review

From the comprehensive literature review it becomes clear that there exist only very few number
of publications, which take into consideration train operation system development at railway
stations and routes. Different researches have been done on railway planning, operation and
safety management for different countries in the railway transport history which is the main
important in the railway operation at different times as discussed above. (R.M.P, 2000b)
analyzed train detection data and train delays at stations using the recently developed tools TNV-
Prepare and TNV-Filter. (Lindner, 2000) deals about Public rail transport planning and
operations by dividing in to three steps as discussed above by using fundamental mathematical
model the Periodic Event Scheduling Problem (PESP). (Bussieck. M.R, 1996) present a MIP
model for choosing a set of lines and their frequencies so as to maximize the number of direct
travelers. (Stockholm, 2010) shows the advantages of simple models for analysis of railway
operation using the analytical model SAMFOST and TVEM. (Hansen, 2005) presented the
investigation of train operations in railway stations based on network, timetable and train
detection data using the tools TNV-Prepare and TNV-Filter. (Kaspi, 2010) presented the train
planning problem by integrating the line planning problem and timetabling using ILPTP
(integrated line planning and timetabling problem) and O-D (origin–destination matrices).

These researches were done to solve their particular research problems and recommend the
future work also to fulfill the gaps in the railway operation. None of the above researchers use
the tool Arena5 simulation model in the analysis of train operation. Railway operation includes
network, timetable, rolling stock and man power integration. Different countries have their own
operation planning depending on their transport demand and capacity. This paper deals with the
train operation system development for railway transport of Ethiopia, Addis Ababa LRT. Addis
Ababa is a multi nation with dense population where there is a huge passenger transport needs
and economic development which requires modern transportation system. To do this a system
analysis approach is presented for investigation of train operations in railway stations and routes
based on network, timetable and train allocation data. The network and train schedule analysis
are evaluated using the simulation model software Arena5 for the investigation of Addis Ababa
LRT.

25
The main objective of this research is to analysis and develops the train operation system for
Addis Ababa Light Rail Transit based on infrastructure; timetable and identifying delay factors
that may face the rail transport of Ethiopia. The train operation system analysis for the railway
transport of Ethiopia is important because operational designed schedule failure causes train
delay. Train delays and route conflicts are one of the most important performance indicators of
railway operations, which result from many related factors, such as the layout of the
infrastructure, the occupancy of the infrastructure, the variation of running times, arrivals and
departures, resource limitation and the traffic control which is the research gap of this paper.
Therefore the knowledge of train operation and factors that cause timetable delay is very
important to avoid the schedule conflicts. This can be done by showing how the infrastructure
can be analyzed through fictive line designs, how the train schedule can be treated as a variable
and how delays can be used as performance measures using the Arena5 simulation model. All
together, these contributions help to enhance future railway operation and thus increase the
competitiveness of the railway transport of Ethiopia even more. The paper concludes providing
the most important standard train operation for the Addis Ababa LRT and recommends for the
future works especially when the train actual operation will start.

26
 CHAPTER THREE

RESEARCH METHODOLOGY

Research Methodology is a procedure that describes clear and systematic steps to guide
researcher in the process of completing the research. Its purpose is to achieve the objective of the
study.

3.1 Research Process

This paper is done through the scientific research process and focus on the Addis Ababa Light
Rail Transit (LRT) train operation system development case study. To do this, first different
literatures, reference books, internet web-sites, articles and international journals those focus on
train operations were accessed and the standardized railway operation is understood. Then the
research problem and objective of the research were discussed. On the other hand scope of the
research, significance of the study and the thesis structure were argued. The passenger transport
operation factors are identified and discussed. The route corridors conceptual model is made.
Then the input data were identified and made ready to the software usage. The probably
assumption for the passenger flow of the Addis Ababa routes are fitted using the Poisson
distribution summary from the input analyzer. The simulation software is opened and the input
functions were filled at necessary place for the modules Create, Assign, Station, Process and
Route. The model window flowchart is made and the simulation run setup is done. Then the
simulation model is run and examines the results. The result is well discussed and the simulation
model is verified and validated. Finally the paper work is concluded and the necessary
recommendation and future work is mentioned.

27
 Reviewing the Literatures, reference books, internet web-sites, articles and international
journals and understanding the standardized railway operation.

Identification of research problem areas

Making research design for Addis Ababa LRT and collecting the case study data.

Railway operation analysis

Demand Network Train schedule Rolling Crew

analysis analysis stock analysis
Model development
analysis
Making simulation

Figure 5, Research process frame works

3.2 Research Approaches

The case study of this research is” Addis Ababa Light Rail Transit “and the study mainly
concerned on train operation system development for the two main routes: from piazza to kaliti
and torhyiloch to Ayat through the common route “Lideta to Meskel square”. To make train
operation system development, and to achieve the stated objective, both quantitative and
qualitative methods were used for data collection and data analysis during the researching. The

28
rolling stock analysis and crew analysis are discussed qualitatively while the network analysis
and train schedule analysis are analyzed using Arena5 simulation model software and the report
results are discussed.

3.3 Data Collection Approaches

The data for analysis were taken from the Ministry of Transport and Communications (MoTaC)
and Ethiopian Railways Corporation (ERC) for the case of Addis Ababa Light Rail Transit.
These data are: hourly demand distribution on and off peak-hour, train average speed, train
loading capacity per trip, number of tramcar assigned for the route, route length, operation hour
per day, maximum waiting time, number of stations and platforms, average length between
platforms. These data are the Addis Ababa LRT design feasibility studies and it is the operation
planning data. Thus train operation planning problems i.e. Line Planning, Timetabling, Plat
forming, Rolling Stock Circulation and Crew Planning are the most common important train
operation factors. In this study, we focus on the network and train scheduling phases because it
will be vast if we focus on the all operation factors so this paper is limited to study on the
network analysis and train scheduling for Addis Ababa Light Rail Transit (LRT) in detail using
the Arena simulation model software even if the others are also discussed. These improve the
quality of services since the quality of service is measured in terms of total time spent by the
passengers in the railway system, including waiting time at the stations of origin, connections
and travel time.

3.4 Software used

The software used for the network and train schedule analysis in this paper is Arena5. Arena is
the most effective when analyzing complex, medium to large-scale projects related to supply
chain, manufacturing, processes, logistics, transportation, distribution, warehousing, and service
systems. It is a software package used for graphically describing Simulation Analysis (SIMAN)
models. It is the preeminent solution for better business decisions with simulation. It is an easy-

29
to-use, powerful tool that allows you to create and run experiments on animated models of our
systems. Arena modeling constructs, called modules, are grouped into templates for arrangement
into hierarchical model diagrams. Arena’s modules represent types of data and commands within
the software. These modules effectively represent a vendor-specific simulation language.

3.5 Data Analysis Method

In this phase the demand analysis for on and off-peak hour, network analysis, train scheduling,
rolling stock analysis and crew management analysis were done.

Figure 6, Railway operation factors

For the train scheduling data analysis and the network analysis the conceptual and simulation
model were developed. The simulation model was done by the software package arena5. The
input data were added to the software by function in the Arena modules. The passenger

30
distribution probably assumption is fitted using Poisson distribution summary and assigned in the
function. Then the software program is well done and run after the run setup is fixed. The
software output report is discussed and the input parameters which affect the train operation were
identified. The verification and validation of the process proposed were made. Finally, the
paper is concluded and the appropriate train operation without delay is recommended.

31
 CHAPTER FOUR

Analysis and Discussion

4.1 Addis Ababa East-West and North-South Corridor

The Addis Ababa LRT covers the East-West and North-South transport corridors. The initial
phase of the project is 34km length and expected to extend up to 74km in the long term plan.

The East-West corridor passes through the city center from the eastern peripheries of the city to
the west. In its final phase, the East-West Major residential and real estate development locations
at Ayat, CMC, Gurdshola and the area west of Torhailoch are accessed through this route.
Meskel Square is the major public gathering location in Addis Ababa and Meganagna Square is
an important transit hub and commercial centre. The East-West corridor crosses the ring road
twice at Megenagna and Torhailoch. In its final phase, it is proposed that the LRT route will
connect to the future national railway hubs at Legetafo and Sebeta. The initial phase of this route
is 17.1km length and expected to extend up to 37.7km in the long term plan.

The North-South corridor is a heavily-used route; it is utilized to access Africa’s largest open
market at Merkato. The initial phase of the project starts from Giorgis (Piazza) in the north and
heads a little distance to the west to access Merkato and Autobus Tera (the Intercity Bus
Terminal). Afterwards, the route heads south until it joins the East-West route at Lideta. From
Lideta to Meskel Square both services have a common route until the North-South route
bifurcates to the south heading to the new Gotera interchange, Saris and Kaliti. In its final phase
the route will cross the ring road at Kaliti interchange and extend to a new railway hub at Akaki.
The northern end of the route will be extended from Giorgis to Shiromeda. The initial phase of
this route is 16.97km length and expected to extend up to 36km in the long term plan.

The LRT route will connect eight major transit interfaces to provide an integrated transit
network. With transfers at these locations, passengers will be able to move quickly from
suburban areas to the city centre and vice versa. The major transfer points are:

32
 Megenagna taxi/bus station Autobus Tera (Intercity Bus
Stadium-Lagar taxi/bus Terminal station)
station Lideta station
Mexico Square Meskel Square
Giorgis taxi/bus station Gotera Interchange

In addition to these major connection points, there are other stations where heavy traffic flows
are anticipated such as Torhailoch, Urael, Hayahulet, Ayat, Saris and Kaliti. Moreover, stations
will be located at reasonably tolerable walking distances which will facilitate foot access.

The Addis Ababa LRT stations will be located at-grade, below ground and on structures. Despite
the variety of the locations, standard details and a common theme should be introduced to link all
the station structures to the LRT system. The exciting possibilities for station sites when
developed as commercial enterprises, with resulting financial advantages for the LRT system,
particularly at “Transit Hubs”, is considered during the stations design. The heavily used at-
grade stations and all stations on viaducts will be reached by overhead footbridges. Underground
stations will have subway accesses. The at-grade stations will also have pedestrian crossing
access for use by the handicapped, the infirm and others. All ‘major’ LRT stations will be
provided with either footbridge or subway access. It should be well designed, pleasing to the eye,
comfortable and convenient for the passenger as well as efficient in layout and operation.
Therefore it must be properly managed and maintained and must be operated safely. Eventually
it is expected that all stations will be equipped with dedicated access for safety reasons.

33
Figure 7, Overall Schematic Diagram of Addis Ababa LRT

34
 4.2 Passenger Transport Operations Analysis

Nowadays, public rail transport operations are a highly complex task. Too many objects interact
with each other to be manageable simultaneously. Various sub problems of different nature like
network design, scheduling or routing occur, and the solutions of most of those sub problems
depend on the solutions of the other sub problems. Due to severe competition from other
transportation modes, the rail industry is eager to improve its operational efficiency and
rationalize its operations and decisions. Analytical models get more and more important in
supporting managerial decision-making.

Due to the tremendous size of the public rail traffic system, the Operation planning process is
divided into several Steps. In a first step, the passenger demand has to be analyzed. In order to
establish a customer-oriented transportation service, the passenger demand or traffic volume also
must be given or estimated. As a result, the amount of travelers wishing to go from certain
origins to certain destinations is known. As a subsequent task, lines where trains run are
determined. Also, the frequencies for the lines are determined. Afterwards, in the train schedule
planning step, all arrival and departure times of the lines are fixed. This has to be done subject to
the periodicity of the system. Now the trains and coaches have to be assigned to lines. This is
called planning of rolling stock. A similar task is the crew management, which means the
distribution of personnel in order to guarantee that each train is equipped with the necessary
staff.

A problem of the decomposition is that the optimal solution for one part serves as fixed input for
the subsequent problem. One cannot expect an overall optimal solution in the end. It is even
possible that at some point, former decisions have to be changed, and a part or the complete
process has to be repeated. Nevertheless, this art graphic provides the inter dependency of traffic
operation planning problem into manageable tasks.

35
 4.2.1 Demand Analysis

The major part of the proposed LRT alignment route is located along existing streets which are
currently serviced by buses, and taxis. According to Ministry of Transport and Communications
(MoTaC), and Ethiopian Railways Corporation (ERC) (2009) the recorded average demand
hourly distribution of Addis Ababa LRT in the corridor is shown in the following table.

From(hour) To(hour) Approximate percent of

Trips in the corridor
0:00 4:00 0:00%
4:00 6:00 2:00%
6:00 9:00 30:00%
9:00 12:00 10:00%
12:00 13:30 10:00%
13:30 15:30 10:00%
15:30 21:00 32:00%
21:00 22:00 4:00%
22:00 24:00 2:00%
Total per 24 hours 100:00%

Table 1, Hourly Distribution of the Demand Based on the LRT Corridor Data

4.2.2 Rolling Stock Analysis

The trips established by the train schedule must be performed by some vehicles (motor unit,
coaches) and crews (like engine drivers, conductors etc.). Operation analysis methods for vehicle
scheduling in public transportation are very important. Since the dispatch of rolling stock and
personnel has the main influence on the overall transport service costs, optimization methods are
essential at this step, and other parts of the operation analysis may have to be revised. In the
rolling stock circulation problem, one has to determine an appropriate allocation of rolling stock

36
units to the trips to be operated. Relevant objectives to be pursued are service to the passengers,
efficiency, and robustness. Service to the passengers’ means that on each trip the provided
capacity should be sufficient to transport the expected numbers of passengers according to given
norms.

The Addis Ababa LRT will have 41 tram car units and out of these 21 is assigned for East-West
corridor while the left 20 are assigned for North-South corridor. Since each corridor has two start
and end points, the tram car units will be assigned equally at the end four stations. They will
operate at an average speed of 30 to 40 km/hr with the average dwelling time of 30 seconds at
each station. The maximum speed of the tramcar is 80km/hr and one unit of the tramcar is
arranged to be operated with the capacity of 286 persons in one tramcar (with 64 seats and 6
standing persons per m2). This operation will be continuing for 16 hours a day. The vehicles
shall be able to operate under the ground, on the ground or on elevated lines. It is ensured that
the vehicles be able to safely operate at lower speeds with Grade 8 wind (20.7m/s) when
operating on the ground and elevated lines and that the empty vehicles be able to safely park on
the lines with Grade 9 wind (24.4m/s).The vehicles shall be maintained and checked inside the
on-ground depot and parked outside the depot.

4.2.3 Crews Analysis

A set of tasks, which can either be passenger train movements, empty train movements, or
shunting activities, must be assigned to train drivers such that each task is covered and each train
driver has a feasible duty. In this case, a duty is a sequence of tasks after each other that can be
carried out by a single employee on a single day. A duty is feasible if a large set of constraints
are fulfilled, e.g. the length of the duty does not exceed the maximum spread time, and there is a
meal break in a duty with a certain minimum length, and so on.

Crew management not only consists of dispatching train crews, but also local staff like cleaning
staff or ticket office staff. Often, there are complex constraint systems for such duties due to
union contracts for break regulations or working times. The management of the railroad

37
operation should ensure that the railroad is operated smoothly and safely. Regular rules and
training sessions should be provided to ensure that the qualification of operating crews is
maintained and to expand the operating personnel pool. Operational management should
consistently allow the operation of trains and associated activities to proceed safely, smoothly,
and efficiently by ensuring that the equipment and tools needed are serviceable and available,
that there are sufficient numbers of qualified operating personnel, and that all personnel involved
are properly coordinated. There should seek uniformity across all train operations, so that
training is simplified, operating personnel are entirely interchangeable within various tasks, and
misunderstandings about procedure are minimized. Railways must ensure that any type of
resource remains in a consistent cycle and balance of locomotives and crews in a corridor have to
check up.

4.2.4 Network analysis

The Addis Ababa LRT is divided in to two main route category namely as East-West corridor
and North-South corridor. Each corridor has 22 stations with minimum inter stations distance of
525m and 1210m maximum distance. The average interval distance is 815m according to the
ERC design data shows. Line is given by a route and a corresponding frequency. The route is
given by a path in the railroad track network. The frequency determines how often this line is
served in accordance to the schedule period. This is what means by Line planning mean to select
lines from a set of feasible lines subject to certain constraints and pursuing certain objectives.
Some possible constraints are that there must be enough lines (or trains respectively) to carry all
passengers, the capacity of tracks must not be exceeded or that the required trains must be
available. Common objectives are minimization of costs or maximizing the number of travelers
with a direct connection. The periodicity of the schedule has to be kept in mind when designing
line plans.

The line plan serves as direct input for the train scheduling problem, where arrival and departure
times for the lines have to be found. Furthermore, the line plan determines which travelers have

38
to change a train during their trip and thus need acceptable connection times. The line or route
conceptual model for the Addis Ababa LRT is as shown below.

Figure 8, Route corridors conceptual model for Addis Ababa LRT

4.2.5 Train Schedule Analysis

The train schedule is the backbone of public rail transport planning and operation, i.e. the
generation of train schedules is the core subject for train operation. A train schedule consists of
the arrival and departure times of the lines at certain points of the network. Depending on the
required resolution, these points are stations (low resolution) or even switches and important
signal points (high resolution). In general, schedules for public transport are periodical, i.e. the
schedule is repeated after a basic time period or, for short period. The schedule fixes arrival and

39
departure times for lines and thus train delay time has to be more or less minimized or avoided
for all trains of the line. An individual train corresponds to a trip of the line. The following table
shows the Addis Ababa LRT routes with its specific justifications.

4.3 Making Simulation model

Modeling methodologies include network representations, process concept, and the entity-
attribute-set approach. The model representation is formulated in the way that it represents the
appropriate actual working system of the Addis Ababa LRT. The model is encoded by using a
simulation programming language using Arena5. Arena icons are connected in a flowchart to
represent entity flow. The conceptual model is translated into modeling software. For the
constructed simulation model, the “Program” is well made and run.

Figure 9, Simulation model over view

40
Technical Data input for Addis Ababa LRT corridors.

The following table shows the input technical data from which the simulation function input are
calculated. The minimum and maximum operation speed is given as 20 and 70 km/hr, and the
distance between each station is also given therefore the route time is calculated and added to the
software as a uniform function.

S/N specification value

1 Test speed 80 km/hr
2 Maximum operation speed 70 km/hr
3 Minimum operation speed 20 km/hr
4 Average dwell time 30 second
5 Train travel interval 6 minute
6 Rated Passengers （Persons/tramcar） 286
7 Train allocation 41
8 Number of stations / platforms 39
9 Number of common stations / platforms 5
10 Operation hours per day 16 hours

Table 2, Technical data for Addis Ababa LRT

S/N station Value in meter

1 NS1- NS2 653
2 NS2- NS3 741
3 NS3- NS4 945
4 NS4- NS5 609
5 NS5- NS6 612
6 NS6- NS7 861
7 NS7- NS8 744

41
8 NS8- NS9 725
9 NS9- NS10 733
10 NS10- NS11 691
11 NS11- NS12 569
12 NS12- NS13 435
13 NS13- NS14 908
14 NS14- NS15 481
15 NS15- NS16 610
16 NS16- NS17 555
17 NS17- NS18 1971
18 NS18- NS19 908
19 NS19- NS20 481
20 NS20- NS21 540
21 NS21- NS22 840
22 NS22- Exit 223

Table 3, Distance between stations for N-S line

S/N station Value in meter

1 EW1- EW2 390
2 EW 2- EW3 776
3 EW 3- EW4 725
4 EW4- EW5 733
5 EW5- EW6 691
6 EW6- EW7 569
7 EW7- EW8 435
8 EW8- EW9 650
9 EW9- EW10 583
10 EW10- EW11 675
11 EW11- EW12 952

42
12 EW12- EW13 770
13 EW13- EW14 746
14 EW14- EW 15 802
15 EW15- EW16 804
16 EW16- EW17 1083
17 EW17- EW18 970
18 EW18- EW19 724
19 EW19- EW20 860
20 EW20- EW21 863
21 EW21- EW22 1092
22 EW22- Exit 182

Table 4, Distance between stations for E-W line

The number of passenger at a time is calculated by dividing the station passenger distribution at
peak hour for thirty because peak hour is three hours duration at morning (6:00-9:00 am) and the
train goes ten times an hour. This is shown in the table below.

S/N station Unit: person

No. of passengers No. of No. of passengers No. of
getting on trains at passenger on getting off trains passenger off
peak hours at a time at a time
1 NS1 0 0 1675 55.83
2 NS2 10 0.33 1485 49.5
3 NS3 43 1.43 1386 46.2
4 NS4 121 4 1345 44.83
5 NS5 120 4 735 24.5
6 NS6 213 7.1 708 23.6
7 NS7 325 10.833 706 23.5

43
8 NS8 866 28.87 574 19.13
9 NS9 965 32.17 788 26.27
10 NS10 994 33.13 1038 34.6
11 NS11 1174 39.13 1223 40.77
12 NS12 1073 35.77 1083 36.1
13 NS13 654 21.8 730 24.33
14 NS14 646 21.53 622 20.73
15 NS15 682 22.73 427 14.23
16 NS16 917 30.27 372 12.4
17 NS17 1152 38.4 273 9.1
18 NS18 1475 49.17 196 6.53
19 NS19 1411 47 81 2.7
20 NS20 918 30.6 18 0.6
21 NS21 977 32.57 4 0.133
22 NS22 1480 49.33 0 0

Table 5, Passenger flow of Addis Ababa Light Rail N-S in morning peak hours.

S/N station Unit: person

No. of passengers No. of No. of No. of
getting on trains passenger passengers passenger off
at peak hours on at a time getting off trains at a time
1 EW1 0 0 2284 76.13
2 EW2 16 0.53 1742 58.07
3 EW3 38 1.27 1122 37.4
4 EW4 76 2.53 1072 35.73
5 EW5 146 4.87 963 32.1
6 EW6 376 12.53 1105 36.83
7 EW7 580 19.33 904 30.13

44
8 EW8 675 22.5 680 22.67
9 EW9 716 23.87 547 18.23
10 EW10 958 31.93 564 18.8
11 EW11 845 28.17 679 22.63
12 EW12 1513 50.43 1617 53.9
13 EW13 1409 46.97 1909 63.63
14 EW14 756 25.2 612 20.4
15 EW15 380 12.67 149 4.97
16 EW16 435 14.5 117 3.9
17 EW17 686 22.87 142 4.73
18 EW18 936 31.2 115 3.83
19 EW19 1193 39.77 68 2.27
20 EW20 1210 40.33 23 0.77
21 EW21 1511 50.37 6 0.2
22 EW22 1635 54.5 0 0

Table 6, Passenger flow of Addis Ababa Light Rail E-W in morning peak hours.

(Source: Preliminary Design for Ethiopia Addis Ababa E-W & S-N（Phase I）Light Rail
Transit Project)

45
The probably assumption for the passenger flow of the Addis Ababa LRT is fitted using the
Poisson distribution summary as shown in figure below by analyzing the morning peak hour
passenger flow as the input analyzer data file and assigned in the simulation model.

Figure 10, Passenger distribution

46
 4.3.1 Simulation Model Results and Discussions for North to
South and West to East line

In this section the results of the simulation run for the queuing analysis, utilization of resources,
bottleneck analysis, delays, and maximum number of trains in service and throughout analysis at
each station and route would be discussed below.

 For this route the starting point is Piazza and Torhyloch station and the train operation
also starts here for the North-South and West-East corridor.
 The two routes meet each other at Lideta station and use common line until they
separate at Meskel Square station to Kaliti and Ayat stations.
 The service priority uses “First come first served” queuing principle at all stations.
 The Holding principle is applied on the routes before Lideta station by scanning for
condition of route busy.
 The trains assigned for their route decided to move on their proper direction at Meskel
Square.
 The trains are assigned as the resource for the simulation analysis.
 Each platform stopping time and the route travel time is considered during the
simulation.
 The number of Replications of the Run set up is 41 and its Replication length is 200
minutes while the run base time unit is hour.
 The animation speed factor (time units per frame) is 48.6 and the animation frame
delay is 0.01seconds.
 The guided transporter matrix size is 30.

47
Figure 11, Simulation Model from Piazza to Kalit and Torhyloch to Ayat

48
Table 7, some simulation results for the N-S and W-E direction

49
Discussion of the North-South and West-East Simulation results

From the simulation model the stopping time (dwell time) is taken by the triangular function
making thirty second the average stopping time. The route speed is changed into the time using
the distance between stations for minimum (20km/hr) and maximum (70km/hr) operation speed
and feed to the software as uniform distribution and the software simulated. Simulation is a
technique for imitating the behavior of a situation or system using an analogous model, situation,
or apparatus, to gain information more conveniently or to train personnel. It is the process of
numerically evaluating a system model and estimating variables of interest. It is also the make
fun of the operation of a real-world process or system over time. That means the above
simulation executes the Addis Ababa light Rail Transit model from Piazza to Kaliti and from
Torhyloch to Ayat as the real operation. This is used to understand the pre-operation design
testing.

After all the simulation run completed, the reports performance measurements listed in
Appendix-A shows the valuable results for the train operation performance indicators which is
shown in the tables form and briefly discussed below. This performance report is used to define
the logical and physical components of a system. Standard features in simulation model include
the description of resource utilization; queues detail summary, process logic, and system data.
From the result table above, the category by replication shows the time per entity for the
minimum, average and maximum waiting time per entity, total time per entity, number of train
waiting at each station. The process detail summary shows the value added time and the waiting
time at each station, the number of train in and out at each station for the simulation model run
set up. On the other hand the queue detail summary reports the waiting time for each station
operation queue. This simulation model combines the two routes together and the number of
waiting time result shows the stations operations are busy from Lideta to Meskel Square
including the route between both stations where the two routes separate to east and south lines.
Therefore, these route and stations are the bottle neck of the Addis Ababa LRT corridors.

50
 4.3.2 Simulation Model Results and Discussions for South to
North and East to West line

In this section the results of the simulation run for the queuing analysis, utilization of resources,
bottleneck analysis, delays, and maximum number of trains in service and throughout analysis at
each station and route would be discussed below.

 For this route the starting point is Kaliti and Ayat stations and the train operation also
starts here for the South-North and East-West corridor.
 The two routes meet each other at Meskel Square station and uses common line until
they separate at Lideta station to Piazza and Torhyloch stations.
 The service priority uses “First come first served” queuing principle at all stations.
 The Holding principle is applied on the routes before Meskel Square station by
scanning for condition of route busy.
 The trains assigned for their route decided to move on their proper direction at Lideta
station.
 The trains are assigned as the resource for the simulation analysis.
 Each platform stopping time and the route travel time is considered during the
simulation.
 The number of Replications of the Run set up is 41 and its Replication length is 200
minutes while the run base time unit is hour.
 The animation speed factor (time units per frame) is 48.6 and the animation frame
delay is 0.01seconds.
 The guided transporter matrix size is 300.

51
Figure 12, Simulation Model from Kalit to Piazza and Ayat to Torhyloch

52
Table 8, some simulation results for the reverse direction

53
Discussion of the South-North and East-West Simulation results

From the simulation model the stopping time (dwell time) is taken by the triangular function
making thirty second the average stopping time. The route speed is changed into the time using
the distance between stations for minimum (20km/hr) and maximum (70km/hr) operation speed
and feed to the software as uniform distribution and the software simulated. Simulation is a
technique for imitating the behavior of a situation or system using an analogous model, situation,
or apparatus, to gain information more conveniently or to train personnel. It is the process of
numerically evaluating a system model and estimating variables of interest. It is also the make
fun of the operation of a real-world process or system over time. That means the above
simulation executes the Addis Ababa light Rail Transit model from Kaliti to Piazza and from
Ayat to Torhyloch as the real operation. This is used to understand the pre-operation design
testing.

After all the simulation run completed, the reports performance measurements in Appendix-B
shows the valuable results for the train operation performance indicators which is shown in the
tables form and briefly discussed here. This performance report is used to define the logical and
physical components of a system. Standard features in simulation model include the description
of resource utilization; queues detail summary, process logic, and system data. From the result
table above, the category by replication shows the time per entity for the minimum, average and
maximum waiting time per entity, total time per entity and number of train waiting at each
station. The process detail summary shows the value added time and the waiting time at each
station, and the number of train in and out at each station for the simulation model run set up. On
the other hand the queue detail summary reports the waiting time for each station operation
queue. This simulation model combines the two routes together and the number of waiting time
result above shows the stations operations are busy from Meskel Square to Lideta including the
route between both stations where the two routes separate to west and north lines. Therefore,
these route and stations are the bottle neck of the Addis Ababa LRT corridors.

54
 4.3.3 Verification
 It is well seen in debugging the program
 Since the program runs almost as it is intended in the simulation model ,it is
Verified
 The result mach with the real system of the station condition
 It runs and verified for the check of validity

4.3.4 Validation of the model

 Since the result shows an acceptable level of confidence in the performance at each
station, the processing proposed model is valid.
 The proposed model is indeed an accurate representation of the real system.
 The other important element is that the model is well credible and the results bear us
decision-making process.
 Since the model and simulation out-come is validated, a model can be used to investigate
a wide variety of “what if” questions about the real train operation.

55
 CHAPTER FIVE

Conclusions and Recommendation

5.1 Conclusions

The railway operation problem consists of several interrelated sub-problems. These are problems
related to demand, infrastructure, time table schedule, rolling stock assign and skilled man
power. This paper dealt with all of them but two of the main sub-problems are more discussed
and software analyzed; the network problem and the train scheduling problem. Each of these
problems has received great attention in the literature. Since both problems are hard to solve,
they are typically solved using the simulation model software. Modeling methodologies include
network representations, process concept, and the entity-attribute-set approach. The model is
encoded by using a simulation programming language using Arena5. Arena icons are connected
in a flowchart to represent entity flow. The conceptual model is translated into modeling
software. The model representation is formulated in the way that it represents the appropriate
actual working system of the Addis Ababa LRT. From the simulation model the stopping time
(dwell time) is taken by the triangular function making thirty second the average stopping time.
The route speed is changed into the time using the distance between stations for minimum
(20km/hr) and maximum (70km/hr) operation speed and feed to the software as uniform
distribution. For the constructed simulation model, the “Program” is well made and run. The
operation design data are verified and the Addis Ababa Light Rail Transit operation test is
checked up for both corridors and the common routes by the simulation software. After all
simulation, the reports performance measurements are shown as results for the train operation
performance indicators in the table form and briefly discussed. The simulation model combines
the two routes together and the number of waiting time, train stations in and out etc results in the
table shows the stations operations are busy from Lideta to Meskel Square including the route
between both stations. Therefore, these route and stations are the bottle neck of the Addis Ababa
LRT corridors where route conflict is very high if the train scheduled is missed. Finally the
operation data are verified and the simulation model output is valid.

56
 5.2 Recommendation

In the global and local transport modes competition, to own the potential pull of more market
share and business profit margins from the transport service within the due date, the Ethiopian
Railway Corporation should have to prepare himself for the skilled manpower and improve the
operation of Addis Ababa LRT service for the future from time to time. Train operations and
traffic control have to keep in the modern way by the introduction of a comprehensive
management information system concerning the distribution and reasons of primary and
consecutive train operation problems. Such a data base could be incorporated into an advanced
decision support tool for dispatchers and provide realistic models of the propagation of actual
trains operation in stations and routes. Through properly train operation schedule and control, it
is possible to avoid delay or route conflict. The data collected for this research are preliminary
design data since there is no train operation yet. But if the data was on operational stage, a more
realistic result could be obtained. Therefore the Ethiopian Railway Corporation is recommended
to perform schedule analysis during operation stage. Finally the ERC is recommended to use the
Arena tool to control the train operation.

5.3 Future work

A system analysis approach is presented for investigation of train operations in railway routes
and stations based on network, timetable and train detection data. The analysis of train operations
clearly reveals that the trains operate at lower than design speed. Furthermore, the dwell times at
platform tracks are systematically extended due to hinder by train’s operation factors. This paper
is done for the Addis Ababa train operation analysis system development depending on the
preliminary design data. Therefore it is possible to analysis the train operation depending on the
actual data recorded after the real operation will start and can do further research on train
scheduling problem for the Addis Ababa Light Rail Transit if any.

57
 5.4 Contribution of the Thesis

This paper is focused on the train operation system development for the Addis Ababa Light Rail
Transit. To do this the researcher referred so many research papers, internet web sites,
international journals and article which are related to train operation and the like. This is used to
understand how to coordinate and write scientific research on this work and contributes good
experience. Meanwhile the thesis enables to understand the major factors that influence the train
operation, how delay is the train operation performance indicator and how it can be minimized.
In other case the paper also academically contributes to study and apply the Simulation Model
(SIMAN) software programmer language Arena5 for the Addis Ababa LRT as it is actual
operating using the design input data. This simulation is well done and run to test the city train
operation, identify the bottleneck of the route and provides necessary train information like
number of train transfer, waiting time, resource utilization, number in and number out at each
station etc for the number of replication animation setup. Finally the paper will very useful for
the Ethiopian Railway Corporation as it focus on train operation because after infrastructure
completion the knowledge of operation and maintenance is the very important one to continuous
the successful transport service. Therefore this paper contribute to the body of knowledge on the
Addis Ababa Light Rail Transit operation process and verify the designed operation data by
checking with the Arena5 simulation software.

58
References

A.J, P. E. (1987). Design of a single track rail line for high speed trains.

ARORA, S. S. (2008A). Text book of railway engineering.

Bussieck. M.R, P. K. (1996). Optimal Lines for Railway Systems.

C.F., B. (2005). Practical Railway Engineering. Singapore: Imperical College.

Corman, F. (2010). Real-time Railway Traffic Managment.

Corporation, E. R. (2011). The National Railway Network of Ethiopia. Addis Ababa.

E.R, P. (1974). Over the road transit time for a single track railway.

G.Kroon, D. H. (2005). Operations Research in Passenger Railway Tranisportation .

Garrido, J. (2009). Object oriented simulation.

Gebeyehu, Y. (2012). Standardization of guidelines for railway track infrastructure subsystem

for railway system of Ethiopia. Addis Ababa.

Goggin, M. (2011). Guide to station planning and design.

Goverde R.M.P, H. G. (2001a). Statistical Analysis of train traffic.

Goverde, R. H. (2001b). Delay Distributions in railway stations.

Hansen, L. N. (2005). System Analysis of train Operations and Track Occupancy at railway
stations.

Higgins, A. K. (1995). Modelling delay risks associated with train schedules. .

I.A., G. R. (2000a). Analysis of Dutch Railway Operations Based on Train Detection Data.

Kaspi, M. (2010). Service oriented train timetabling.

59
L., B. (2005). Railway Safety Management Systems.

Limited, C. R. (2009). Addis Ababa LRT Project East-West Line Project Study. Addis Ababa.

Limited, C. R. (2009). Addis Ababa LRT Project North-South Line Project Stud. Addis Ababa.

Limited, C. R. (2009). Preliminary Design for Ethiopia Addis Ababa LRT phase I. Addis Ababa.

Limited, C. R. (2013). Technical Specification of Vehicles. Addis Ababa.

Lindner, T. (2000). Train schedule optimization in public rail transport.

Mar'oti, D. H. (2007). Railway Time Tabling From Operation Research Perspective.

P.C, M. (2009). Railroad customer safety handbook.

R.J, H. T. (2001). Running times on railway sections with heterogeneous train traffic.

R.M.P, G. (2000b). Delay Estimation and Filtering of Train Detection Data.

Stockholm. (2010). Railway operation analysis.

Transport, E. R. (2009). Bankable study for Addis Ababa LRT. Addis Ababa.

W.Ukovich, S. P. (1989). A mathimatical model for periodic event scheduling problems.

Wendler E., H. I. (2008). Railway timetable and traffic.

Yuan J., G. R. (2002). Propagation of train delays in stations.

60
Appendix
Appendix A. North-South and West-East Line Simulation Results

Table 9, Result summary of waiting time.

59
Table 10, Result summary of total waiting time.

60
Table 11, Result summary of number of train waiting.

61
Table 12, Result summary of resource utilization.

62
Table 13, Result summary of time per Entity.

63
Table 14, Result summary of train in and out.

64
Table 15, Result summary of Queue waiting time.

65
Appendix B. South-North and East-West Line Simulation Results

Table16, Result summary of waiting time.

66
Table 17, Result summary of total waiting time.

67
Table 18, Result summary of number of train waiting.

68
Table 19, Result summary of resource utilization.

69
Table 20, Result summary of time per Entity.

70
Table 21, Result summary of number of train in and out.

71
Table 22, Result summary of Queue waiting time.

72