ADDIS ABABA UNIVERSITY

ADDIS ABABA INSTITUTE OF TECHNOLOGY

SCHOOL OF CIVIL AND ENVIRONMENTAL ENGINEERING

GEODESY AND GEOMATICS PROGRAM

GIS Based Network analysis for Locating Fire Stations and Emergency
Response: A Case of Addis Ababa City, Ethiopia

By
Temesgen Abuhay Hagos
ID No. GSR/2283/12

Advisor:
Dr. TIBEBU KASSAWMAR

A Thesis Submitted to the School of Graduate Studies of Addis Ababa University in

Partial Fulfillment of the Requirements of the Degree of Master of Science in
Geodesy and Geomatics Program (Specialization in Geomatics)

February, 2022
 Approval sheet

GIS Based Network Analysis for Locating Fire Stations and Emergency
Response: A Case of Addis Ababa City, Ethiopia

By

Temesgen Abuhay Hagos

ID: GSR/2283/12
Email: [email protected]

Approved by the Board of Examiners

Dr. Tibebu Kassawmar -------------------------- --------------------

Advisor Signature Date

Dr .Andenet Ashagrie ------------------------ ------------------

Internal Examiner Signature Date

Dr. K.V. Suryabhagavan ----------------------------- -----------------

External Examiner Signature Date

Dr. Ing. Mebruk Mohammed --------------------------------- -------------------

Chairperson Signature Date
 DECLARATION

I certify that this research work entitled “GIS Based Network Analysis for Locating Fire
Stations and Emergency Response: A Case of Addis Ababa City, Ethiopia” is my own work
and has not been presented for a degree in any other university. Where material has been used
from other sources has been properly acknowledged.

Temesgen Abuhay ------------------------------- ----------------------

-

Name of Candidate Signature Date

As Master research advisor, I hereby certify that I have read and evaluated this MSc thesis
prepared under my guidance.

Dr. Tibebu Kassawmar -------------------------- --------------------

Advisor Signature Date

MSc Thesis Page I

 Acknowledgements

First of all, thanks to Almighty God for His unlimited blessings and for giving me the strength to
complete this study. I would like to express my heartfull gratitude to my advisor, Tibebu
Kassawmar (Ph.D.) for his encouragement, guidance, and valuable comment from draft to final
document:

I would like to acknowledge the Addis Ababa Fire and Emergency Prevention and Rescue
Agency office for supporting me by providing the necessary data and information that helps to
accomplish this study paper. Also, I also wish to thank the help provided by Addis Ababa City
Government Plan and Development Commission. I would also like to pass my gratitude to the
different organizations and the managers, and different levels staff of the organizations that have
been the sources of the data for this study.

Finally, my heartful thanks go to my whole family for their help, motivation, and support
throughout my study work from beginning up to end and people surrounding in this study who
helped me in anyway.

MSc Thesis Page II

 Table of Contents
DECLARATION............................................................................................................................ I
Acknowledgements ...................................................................................................................... II
List of Figures ............................................................................................................................... VI
List of Tables ............................................................................................................................... VII
List of Abbreviations ................................................................................................................. VIII
Abstract ......................................................................................................................................... IX
CHAPTER ONE ........................................................................................................................... 1
INTRODUCTION......................................................................................................................... 1
1.1 Background of the study ....................................................................................................... 1
1.2 Statement of the problem ...................................................................................................... 2
1.3 Objective ............................................................................................................................... 3
1.3.1 General objective ............................................................................................................ 3
1.3.2 Specific objective ........................................................................................................... 3
1.4 Research questions ................................................................................................................ 4
1.5 Scope of the study ................................................................................................................. 4
1.6 Significance of the study ....................................................................................................... 4
1.7 Limitation of the study .......................................................................................................... 5
1.8 Organization of the thesis ...................................................................................................... 5
CHAPTER TWO .......................................................................................................................... 6
REVIEW OF LITERATURE ...................................................................................................... 6
2.1 Definition of Terms and Concepts ........................................................................................ 6
2.1.1 Fire stations..................................................................................................................... 6
2.1.2 Fire incident .................................................................................................................... 6
2.2 Emergency response using GIS............................................................................................. 9
2.3 Fire response time ............................................................................................................... 11
2.4 ArcGIS network analysis ................................................................................................... 13
2.6 Location-Allocation ............................................................................................................ 14
2.6.1 Maximum coverage in ArcGIS..................................................................................... 15
2.7 Multi criteria analysis – analytical hierarchy process ......................................................... 16
2.8 Conceptual framework ........................................................................................................ 17

MSc Thesis Page III

 CHAPTER THREE .................................................................................................................... 18
MATERIALS AND METHODS ............................................................................................... 18
3.1 Description of the Study Area ............................................................................................. 18
3.1.1 Location ........................................................................................................................ 18
3.1.2 Population......................................................................................................................... 19
3.1.3 Climate ............................................................................................................................ 19
3.1.4 Topography ...................................................................................................................... 21
3.1.5 Road Transportation in Addis Ababa ............................................................................... 22
3.2 Data and Data Source .......................................................................................................... 22
3.2.2 Secondary data ............................................................................................................. 25
3.4 Software and instrument used ............................................................................................. 29
3.5 Methodology ....................................................................................................................... 30
3.5.1 Data Collection and preparation ................................................................................... 32
3.5.2 Data processing............................................................................................................. 32
3.5.3 Creation of geodatabase................................................................................................ 32
3.5.4 Building network topology ........................................................................................... 32
3.5.5 Building network dataset .............................................................................................. 32
3.5.6 Performing network analysis ........................................................................................ 34
3.6 Data analysis ....................................................................................................................... 35
3.6.1 Road network accessibility............................................................................................... 35
3.6.2 Network analyst in ArcGIS .......................................................................................... 36
3.6.3 Spatial relationship of fire incidents with population density ...................................... 39
3.6.4 Spatial analyst tools ...................................................................................................... 40
3.6.4.1 Kernel density analysis .............................................................................................. 40
3.6.4.2 Candidate site selection for new fire stations ............................................................... 41
3.6.4.3 Factors used for this study ......................................................................................... 42
3.6.5 GIS Based Multicriteria decision analysis ...................................................................... 46
3.6.7 Weighted linear combination (WLC) in ArcGIS ......................................................... 51
3.6.8 Descriptive analysis ...................................................................................................... 53
3.6.8.1Statistical analysis of fire incidents data..................................................................... 53

MSc Thesis Page IV

 CHAPTER FOUR....................................................................................................................... 54
RESULTS .................................................................................................................................... 54
4.1 Service area coverage ...................................................................................................... 54
4.2 Best route result ............................................................................................................... 56
4.3 Closet facility area ........................................................................................................... 58
4.4 Finding optimum sites based on location allocation ........................................................... 60
4.2.5 After Adding additional fire stations ............................................................................ 64
CHAPTER FIVE ..................................................................................................................... 65
DISCUSSION .......................................................................................................................... 65
5.1 Comparative analysis ....................................................................................................... 65
CHAPTER SIX ........................................................................................................................... 68
CONCLUSION AND RECOMMENDATIONS ...................................................................... 68
6.1 Conclusion........................................................................................................................... 68
6.2 Recommendations .............................................................................................................. 69
6.2.1 Suggestion for further research..................................................................................... 69
References .................................................................................................................................... 70
Appendices ................................................................................................................................... 77
Appendix 1 :- ............................................................................................................................... 77
Appendix 2 ................................................................................................................................... 77

MSc Thesis Page V

 List of Figures

Figure 2.1 Cascading event chart for an emergency situation ...................................................... 12

Figure 2.2 Conceptual frameworks of the study ........................................................................... 17

Figure 3.1 Location map of the study area ................................................................................... 18

Figure 3.2 Population census of Addis Ababa city in year 2020 .................................................. 19
Figure 3.3 Monthly temperature variation .................................................................................... 20
Figure 3.4 Average monthly rainfall in the study area ................................................................. 20
Figure 3.5 Elevation map ............................................................................................................. 21
Figure 3.6 Fire Incidents locations map of the year 2020-2021 ................................................... 23
Figure 3.7 Addis Ababa City Road Network Map ....................................................................... 25
Figure 3.8 Existing fire station location map ............................................................................... 26
Figure 3.9 Land use pattern of Addis Ababa city ........................................................................ 28
Figure 3.10 Flow chart of the methodology.................................................................................. 31
Figure 3.11 Road network dataset of the study area ..................................................................... 33
Figure 3.12 Population density map of sub cities in Addis Ababa ............................................... 39
Figure 3.13 8-Minute Service Coverage Overlapped with Population Density ......................... 40
Figure 3.14 Fire Incident Density Rate Value map ...................................................................... 41
Figure 3.15:Reclassified thematic maps: Distance from existing fire stations ( a ), inaccessible
area ( b ) , Population density (c ) , Proximity of main road( d ) , Land-use map (e) , and
incident density rate (f) ................................................................................................................. 45
Figure 3.16 Candidate sites selection procedure using model builder for new fire stations ........ 51
Figure 3.17 Methodological flow chart of fire stations Site Selection of Addis Ababa city ........ 52
Figure 3.18 Plot Chart Between Sub Cities Versus Number of Fire Incidents........................... 53

Figure 4.1 Service Area Coverage of Existing Fire Stations ........................................................ 54

Figure 4.2 Served population and area covered with their percentage ........................................ 56
Figure 4.3 Best route map from arada sub city. .......................................................................... 57
Figure 4.4 Best route directional windows ................................................................................... 57
Figure 4.5 Closest fire stations within 8 minutes to fire incidents................................................ 58
Figure 4.6 Closest Facilities Analysis Result from Arada Fire Station to FI_272 ....................... 59
Figure 4.7 Closest facilities direction result from Arada Fire Station to FI_272 ........................ 60
Figure 4.8 The selected candidate sites(a) & Proposed sites to locate additional fire stations (b) 61
Figure 4.9 The Ground Truth of Chosen Sites for Fire Stations to The Study Area Taken from
Google Earth. ................................................................................................................................ 62
Figure 4.10 New service area created after adding new fire stations ........................................... 64

MSc Thesis Page VI

 List of Tables

Table 2.1: Fire Incidents report on each sub city in addis ababa 2020-2021.................................. 9
Table 2.2 : Different fire response time standards for different countries .................................... 12

Table 3.1 Functional Road hierarchy Detail ................................................................................ 25

Table 3.2 Existing land use of Addis Ababa city.......................................................................... 27
Table 3.3 Data and data source ..................................................................................................... 29
Table 3. 4 Software and instrument used for the study................................................................ 30
Table 3.5 Important attribute in addis ababa city road network dataset ....................................... 35
Table 3.6 Factors used for candidate site selections for new fire stations ................................... 42
Table 3.7 Scale of relative importance of analytic hierarchy process (ahp) based on saaty. ........ 47
Table 3.8 Preference values of criteria pair wise comparison matrix using AHP ........................ 48
Table 3.9 Normalized pairwise comparison matrix result ............................................................ 48
Table 3.10 The final relative weight of each criterion .................................................................. 49
Table 3.11 Determination of eigenvalue for the comparison matrix ............................................ 50
Table 3.12 Random stability index (Saaty, 2000)........................................................................ 51

Table 4.1 Fire stations coverage within 8 minutes in Addis Ababa city ..................................... 55
Table 4.2 Location allocation analysis using travel time (in minutes) as impedance from 10
optimum fire stations locations. ................................................................................................... 62

MSc Thesis Page VII

 List of Abbreviations

AACFEPRA Addis Ababa City Fire and Emergency Prevention and Rescue Agency

AACRA Addis Ababa City Road Authority

AHP Analytical Hierarchy Process

AU African Union

CMSGHS Central Medical Stores of the Ghana Health Service

ESRI Environmental System Research Institute

FPASA Fire Protection Association of South Africa

GHS Ghana Health Service

GIS Geographic Information System

GPS Global Positioning System

MCDMA Multi-Criteria Decision-Making Analysis

NMA National Metrological Agency

NDRMC National Disaster Risk Management Commission

OSM Open Street Map

WLC Weighted Linear Combination

WHO World Health Organization

NFPA National Fire Protection Association

UNECA United Nations Economic Commission for Africa

UN DESA United Nations Department of Economic and Social

Affairs

UTC Coordinated Universal Time

MSc Thesis Page VIII

 Abstract

Urban fire is one of the most common problems not only for developing countries but also for
developed countries. Particularly in Addis Ababa, Ethiopia. On this year ,2020/21 totally 274
incidents were occurred in the span of two years. Fire stations in Addis Ababa city is one of the
major issues as most of increasing of population and urbanization expansion due to poor and
inefficient access of fire services stations. This study aims GIS based network analysis for
locating fire station and emergency response analysis. It was conducted in Addis Ababa city and
it has a total area of 519.458 sq. km2 and focused on two years fire incident report. The
methodology of the study was integrated with roads network analysis and locating optimal fire
stations for the study area using GIS software. The network analysis process has been applied
and search optimal route between fire incident location and fire station by applying Dijkstra’s
algorithm to the road network of the addis ababa city in connection with this service area was
applied. The served population from existing fire stations is 2,817,315 (66.62%) out of
3,689,001estimated population which is distributed over the area 276.10km2 (53.15 %) of the city
are served out of a total of 519.458 km2. The closest facilities from fire station to the fire incident
eight minutes travel time, a total of 230 route was generated with maximum travel time of 7.99
min and length of 11890.42m as well as using travel time as an impedance factor this study
determines the best route from fire stations to the incident location. And , multi criteria analysis
model has been developed to choose suitable sites for the inaccessible place together with factor
identifications. This study employed six different criteria are considered for selecting candidate
sites . From them the more influenced criteria are inaccessible area, population density and land
use with associated weights of 36% , 24% and 17% respectively and AHP method was used for
selecting 35 candidate sites for the study area. Among of the location allocation problem type ,
maximize coverage has chosen ten new optimum location of fire stations to the study area.
Finally, the present study suggests 10 new fire stations which helps to maximize the coverage of
the city as well as provides services during emergency occurrence. After the suggested 10 new
fire station location together with existing fire stations , the current service coverage the study
area has increased into 87.44%.

Keywords: Urban fire, Network analysis ,GIS , AHP , Dijkstra’s algorithm, and Location
allocation

MSc Thesis Page IX

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER ONE

INTRODUCTION

1.1 Background of the study

Fire is a form of disaster that can occur anytime and anywhere and to some instances without
notice (Edupta, 2011). Fire can be a great threat to human beings, the built-environment and
nature, often resulting in injuries, deaths, economic loss and pollution Yao et al., (2019).
According to the International Association of Fire and Rescue Services Brushlinsky et al.,
(2017) among the reported 31 countries there were on average about 3.5 fires per 1000
inhabitants in 2015, resulting in 18,454 deaths and over 44,000 injuries. Urban fire is one of the
most common problems not only for developing countries but also for developed countries
Nisanci., (2010). Particularly in Addis Ababa, Ethiopia. According to the Addis Ababa Fire and
Emergency Preventation and Rescue Agency (AACFERA) report for the years of 2020 and
2021 a total of two hundred seventy-four (274) firehouse incident has occurred in the span of two
years from this accident so many properties was damaged and there was also loss lives in the
city. The main cause the accident was improper use of power line, poor installation of electric
line , explosion of stoves that work with gas and oils. Firefighting and rescue services or fire
stations are among the most crucial government services for protecting people , property ,and the
environment from fire and other emergencies.

Fire stations are one of the most essential and necessary land uses in urban area that guarantee
the life and safety of urban residents. So, the location and number of the stations is very
significant for the efficient coverage of the area which intended to be protected. Furthermore, the
expanded of urbanization growth has multiplied the increase in population of urban area in the
developing countries which the risk of hazardous fire area its impact of the city became more
obvious Habibi et al., (2008). As a result, emergency response services are critical, especially in
the case of fire disasters, to prevent the country from losing valuable resources Forkuo and
Quaye-ballard., (2013). The success of emergency response depends on procedures, resources,
and Employees. Services are required during the response, preparedness, and rehabilitation times
and must be defined according to the responding agencies and the types of emergencies
experienced. In addition, the right resource control processes, such as the implementation of an
emergency response plan, an incident. As a result of being trained to use the tools and practice

MSc Thesis Page 1

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

the systems in multiple emergency situations, the emergency management staff act as the link
between the systems and resources. In other words, these professional workers are the ones that
are on the site security problems and on the field at hand during crises. firemen and Ambulance
staff are few examples of qualified emergency service personnel Subramaniam and Ali, (2012).
Right now, in addis ababa city ten existing fire stations are found which provide a service for
emergency response occurrence for the entry city. In order to determine the best location to the
study area this study used GIS. Geographic information system (GIS) is one of the best
techniques for determining the best location for a fire station, and it can integrate both spatial and
non-spatial data Şen et al., (2011). There are two spatial data models being used in GIS today:
vector and raster.

The present study was carried out to solve the problem of inequality distribution of fire
stations throughout the city and develop a method of finding optimized routes for emergency
response analysis. For that reason, an integrated method was used such as GIS techniques of
Analytical Hierarchy Process(AHP) to propose additional candidate sites of fire stations to be
added in the study area and network analysis techniques was conducted. Among of location
allocation analysis, Maximize Coverage has been utilized to determine the best optimum location
of fire stations service in the Addis Ababa City.

1.2 Statement of the problem

Fire incidents is a one of the key problems in Addis Ababa city which occurs frequently due to
growth of population in parallel with the presence of dense housing which causes the number
incident tends to increase year by year. On this year, in 2020 and 2021 one hundred twenty
(120) and one hundred fifty-four (154) firehouse accidents occurred respectively.

Most of the time fire accident are caused by human carelessness, unexpected accidents and
natural disasters. Fire is among the most destructive problem in Ethiopia. There is usually an
incidence of fire in Ethiopia which is mainly in January, February and March before the onset of
the rainy season (Yetmgeta, 2013) . The data from Addis ababa city fire and emergency
prevention and rescue agency (AACFEPRA) indicates that among 274 major fire house incidents
occurred in 2020/21, 66 incidents are occurred in the above stated months. The cities need to
enhance and upgrade emergency services and constantly improve on response time since seconds
make differences in terms of saving lives and minimizing damage to property (Baloyi, 2017).

MSc Thesis Page 2

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

But the other is ,increased traffic flow in the city , inconvenience of roads will delay the response
of time to reach the scene of fire this may result in economic failure of the victim.

Due to poor and inefficient access of fire stations service in Addis Ababa city as well as
increasing of population and urbanization expands the fire incidents has increased year to year.
Based on this it was discovered that unequal fire stations facilities distribution has been the
problem in Addis Ababa city.

In the present day, Addis Ababa city has ten existing fire station services from the size of the city
this is not enough to cover the entire city which is actually shown that the facilities are not
equally distributed as per living population. The stipulated target response time of Addis Ababa
city fire and emergency prevention and rescue agency is 8 minutes. However, most parts of
Addis Ababa city fire stations cannot reach to the incident locations at the specified local
standard time response. As the expansion of the city and growing population is increasing, the
need for fire stations also increases. However, the number of fire stations has not increased yet
accordingly. Therefore, to overcome this problem the study develops GIS based network analysis
approach which is used for emergency response analysis. Additionally, Analytical hierarchy
process was used to propose additional fire stations in order to minimize the deficiency of fire
station service to the uncovered area which helps to reduce the huge property damage and loss of
human lives.

1.3 Objective
1.3.1 General objective
The general objective of this study is GIS based network analysis for locating fire stations and
emergency response analysis in Addis Ababa City.

1.3.2 Specific objective

❖ To evaluate the location of service area coverage of existing fire stations based on local
deriving time response.
❖ To find the best route and closest facilities analysis from the nearest of fire station to the
fire incident location.
❖ To choose optimum sites of fire stations to the study area by using location allocation
problem types particularly maximize coverage.

MSc Thesis Page 3

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

1.4 Research questions

In parallel to these objectives, this study raised the following basic research questions.

✓ Which areas are there under the service coverage of existing fire stations based on local
context deriving time response?
✓ What are the methods used for finding the shortest routes and closest facilities from fire
station to the fire accident place?
✓ Which method is used for finding the optimum location of the fire station to the
maximize coverage of the study area?

1.5 Scope of the study

Addis Ababa is one of the biggest cities in Ethiopia and it has a total area of 519.458 sq. km2 .
The scope of the study was GIS based network analysis for locating fire stations and emergency
response analysis for Addis Ababa city. This study mainly performs finding additional candidate
sites and using advanced network analysis extensions tools for emergency response analysis
including service coverage, best route, closest facilities, and location allocation (Maximize
coverage) in Addis Ababa city. In connection with this the study was focused in the year
between 2020 & 2021fire incident reports were used for further analysis.

1.6 Significance of the study

The importance of this study is not limited only for this paper but also the concerned
stakeholders they look at will be beneficial . For example, the result of this study is very much
useful for Addis Ababa city fire and emergency prevention and rescue agency(AACFEPRA) and
it will assist in finding well thought-out locations for constructing new fire stations in the city. In
connection with this the information generated by this study is also believed to provide useful
information for Addis Ababa city government plan and development commission while
designing public facilities and infrastructure in relation to population growth. In addition to this
the methodology and procedures used in the study will be used as a reference to conduct further
research in other places for the other researcher.

MSc Thesis Page 4

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

1.7 Limitation of the study

In the study there was some limitation some of them are listed:

✓ The data obtained from Addis Ababa Fire and Emergency Prevention and Control
Agency was not readily available in spatially compatible format. This made the study
difficult to find and collect coordinate location of fire incident places.
✓ The climate condition during GPS data collection for the fire incident location was rainy
season that made the study not to collect the data as per scheduled time.
✓ There was a shortage of data quality and it was very time consuming to clean the data .

1.8 Organization of the thesis

This thesis consists of five chapter which are arranged as follows: Chapter One : Introduction :
this chapter includes introduction of the research , statement of the problem, research objective,
research questions, significance of the study, limitation of the study, scope of the study and
structure of thesis format. Chapter Two : present literature review of the previous literature
appropriate for this study topic also discuss their common term and related terminology.
Chapter Three: Methodology and Material: elaborate the data methodology and general
implementation to achieve the objective of the study. Chapter Four: Data Analysis , Result and
Discussion : shows that the main data analysis undertaken for the study, detailed discussion and
interpretation of the result. The candidate site selection and advanced network analyst extension
tools are applied in this chapter. The final section is Chapter Five: Conclusion and
Recommendation : provide answers for research questions, generalizing the obtained results and
indicate future work to another researcher. And Reference : a list of sources used in this study
attached at the end of this paper.

MSc Thesis Page 5

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER TWO

REVIEW OF LITERATURE

In this section the study reviews the previous literature related with the present study of GIS
based network analysis for locating fire stations and emergency response analysis are reviewed
and presented under the following sections in addition to this some standard practiced across
world has been reviewed. This chapter attempts to highlight various applications of ArcGIS
integrated with Analytical Hierarchy Process for finding optimum location for fire stations to the
study area. Along with this the reason for doing this study is in the city of Addis Ababa fire
occurs frequently in order to save this event it is intended to establish a new fire station.

2.1 Definition of Terms and Concepts

2.1.1 Fire stations

Fire is among the causes of loss of life and property, therefore, to eliminate possible destructive
outcomes there should be timely interventions. The fire service mission is to protect lives and
property from fire. Therefore, emergency services in cities need to be planned in such a manner
that they cater for all areas within their jurisdiction within the required response time.
Traditionally, the purpose of the fire service departments has been to prevent fires and to respond
to fires that do occur and to put them out. Although most fire departments are the primary
providers of preventative fire protection, all fire departments are organized to react promptly to
crisis situations such as road accidents and flooding Drake et al., (1972)

Worldwide, fire stations play the most important role in case of any type of emergency. the fire
service mission is to protect life, property and natural resources from fire and other emergencies
with the help of their existing vehicles and equipment and manpower skill ESRI, (2007). Fire
station is an around the clock complement of a firefighter together with fire suppression vehicles
such as an engine or a ladder Badri et al., (1998).

2.1.2 Fire incident

Fire incident is unexpected events can have its origin either in nature or in human activities.
Whatever the case, it is a phenomenon that usually involves a process of occurrence, spreading,
human countermeasures and then disappearance Haque, (2001). Fire usually occurs due to

MSc Thesis Page 6

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

improper use of utilities or unaware human behavior. It can spread in any direction if
inflammable materials are present, and it can last anywhere from a few minutes to days or weeks
depending on the fire load and successful countermeasures Maniruzaman and Haque, (2007).

2.1.2.1 Globally

Globally, Fire accidents have become very common in the world . Because of the health and
climate impacts of fires and the fact that fire regimes have been changing, fire prediction has
received increasing attention. Each year, fire causes about 300,000 deaths globally and most of
these occur in the home Zhang et al., (2006). This can be mitigated by taking appropriate
precautions, which are made easier by sophisticated systems and information technologies. Since
GIS can process large amounts of data and respond quickly to spatial queries, it can be used to
analyze data about urban fires Nisanci, (2010).

The possibilities of urban fire are heavily linked to the area’s level of urbanization Wang et al.,
(2011). By 2007, for the first time in history , more than half of the world’s population 3.3 billion
people were lived in cities Khouri et al., (2018), urbanization is a worldwide phenomenon
which is at its peak. Currently, 54 percent of the world’s population lives in cities , with that
number predicted to rise 66 percent by 2050 UN DESA, (2014). The increase of population will
facilitate the fire incident occurrence in the city. According to World Health Organization
(WHO) statistics, fire-induced burns cause over 300,000 deaths per year among over 95 percent
of these deaths occur in low and middle-income countries. Building fires are a major threat to the
safety of building residents and are mostly caused by people‘s activity. Fires, especially in big
residential complexes with many occupants, can result in increased injury or death as a result of
emergency evacuation difficulties Shokouhi et al., (2019).

2.1.2.2 Africa
The Fire Protection Association of South Africa (FPASA) compiles data on fire from brigades
across the country but this is not without its challenges, including incompletely documented or
unreported fires Potts, (2012). Consequently, their data typically underestimates incidence data.
In 2016, FPASA recorded 41,873 fire callouts nationally, which translates to about 115 fires per
day. Between 2003-2016, the number of fires increased between 1.5-1.7% per annum with the
number of fatalities rising between 5-10% per annum.

MSc Thesis Page 7

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Zimbabwe is among the most fire vulnerable countries in Africa south of the Sahara
Chinamatira et al., (2016). Per year more than 1 million hectares of land destroyed by wild land
fires during the fire season span from July to November .There are several causes of wild land
fires. Chinamatira argued that recent increase in fire incidences has attributed to newly resettle
smallholder farmers.

Several key historical documents including some dating back to the independence of Ghana were
lost because of the fire incidence. A fire in early 2015 destroyed the Central Medical Store of
Ghana Health Service (CMSGHS) in Tema , which housing medical supplies for distribution to
medical facilities nationwide. Between January and September 2016 , GHC 86 million worth of
property were lost .Every year Ghana cause huge damage due to fire disasters. Critical sectors of
nations economy and infrastructures are consumed by fire and the cost of the item damaged has
916,355 GHS in Ghana Lost GHC86 Million to Fires between Jan-Sept.2016, (2016). The
country was hit by yet another large fire in late 2016, this time at the Ghana International Trade
Fair Center which resulted in six human deaths Ibrahim, (2016).

2.1.2.3 Ethiopia

In 2010-2012 , considering that there were only 159 fire disaster events were recorded all over
Ethiopia cities during the 2-year period Osuteye et al., (2017) In Ethiopia , Fire accidents occur
frequently NDRMC, (2016), Fire accident, caused by either humans or nature, can pose a hazard
to people, properties, and the environment, possibly resulting in psychological damage, physical
injuries, death and significant economic losses. For example, there were about 9,755 fires during
2000-2016 in Ethiopia, involving 853 deaths and more than 5,421 non-fatal casualties and above
615-million-birr economic costs of property damage caused by about 9,755 fires in 2000-2016.

However, recently in Ethiopia 2020/2021, associated with the expansion of the cities, there has
been an increase in fire accidents , road traffic and construction-related accidents. Among them,
many fire incidents have been registered throughout Ethiopia cities including Addis Ababa,
Gondar, Woliyta Soddo, Bahir Dar, Dire Dawa, and others, resulting in extensive property
damage and loss of human lives.

MSc Thesis Page 8

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

2.1.2.4 Addis Ababa

Addis Ababa city has the largest number of business centers such as Marcato. Moreover, due to
the existence of densely residential and commercial centers as well as the increasing population
from day to day the vulnerability to fire incidents becomes widespread. The study used a two-
year report of fire incident data i.e.,2020/2021 acquired from AACFEPRA which was
categorized based on sub cities occurrences reported with estimated property loss of
1,065,785,980 (ETB) as presented in Table 2.1. Beside this there were also the loss of lives and
injuries

Table 2.1: Fire Incidents report on each sub city in addis ababa 2020-2021

Fire Incidents in 2020/21 Property Damage In Both

Sub City Name
2020 2021 Total Percentage Years
Addis Ketema 16 26 42 15.33 85,348,400
Akaki-Kality 8 9 17 6.69 103,216,900
Arada 13 19 32 11.68 118,431,910
Bole 21 27 48 17.52 121,426,910
Gulele 11 12 23 8.39 113,930,510
Kolfe Keranyo 12 20 32 11.68 125,457,510
Lideta 6 5 11 4.01 108,888,510
Kirkos 6 9 15 5.47 84,432,510
Nifas Silke Lafto 15 14 29 10.58 109,878,910
Yeka 12 13 25 9.12 94,773,910
Total 120 154 274 100 1,065,785,980

( Source : AACFEPRA, Sept 2020- Aug 2021 fire report )

2.2 Emergency response using GIS

In Bhagat and Sharma, (2013) the authors develop a GIS desktop-based emergency response
management system in Delhi, India. In this system, a detailed transportation network and real-
time traffic data were maintained and integrated. GIS capabilities, such as network analysis,
Origin–Destination (OD) cost matrix, proximity analysis, and buffer analysis were also
performed.

MSc Thesis Page 9

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

In Hoyos et al., (2015) , the authors classify Emergency management activities can be classified
into five phases; planning, mitigation, preparedness, response and recovery. Mitigation Phase
associated with analytical modeling and reveals the natural spatial variation in hazard, prone and
ultimately risk.

In (Nguyen, 2006) , the authors states that creating the fire emergency response database,
preparedness in emergency planning is a crucial factor that can impact the lives of people. When
emergency preparedness is properly planned and executed, human lives can be saved and some
of the economic losses in areas affected by fire can be mitigated. However, if not properly
planned or implemented timeously, the consequences can be dire and could cost human lives

In Forkuo and Quaye-ballard, (2013) , the authors introduced a GIS based fire emergency
response service in Kumasi Metropolis, Ghana where the Ghana National Fire Service (GNFS)
can select the best route from its location of any fire incident. The best route was modeled based
on the travel time, travel distance, the slope of the roads and the delays in travel times.

In Bhagat and Sharma, (2013) , the authors based on GIS in Delhi, India to develop a desktop-
based emergency response management system. In this system, a detailed transportation network
and real-time traffic data were maintained and integrated. GIS capabilities, such as network
analysis, Origin–Destination (OD) cost matrix, proximity analysis, and buffer analysis were also
performed.

In Hay et al., (2019) , the authors implemented the GIS-Based Fire Emergency Response System
for Mandalay Road Network using the Dijkstra routing algorithm. The paper emphasizes the
selection of the best route between two locations on the road network data and then finding the
nearest emergency service providers and fire stations to an accident site based on traffic
conditions.

In Sayed et al., (2017) the authors developed an enhanced GIS-based network analysis that was
performed and applied to the Greater Cairo Road network. It focused on finding the best route
between two locations on the road network and identifying the closest healthcare service
providers to an incident location based on the response travel time and used historical traffic data
in the analysis. It shows an improvement in the travel time with 20% to 22%, according to the

MSc Thesis Page 10

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

travel distances. However, the system did not consider the other factors that cause a delay when
traveling through the road network.

2.3 Fire response time

While it is important that fire stations are properly situated to maximize coverage , they must
also be carefully placed to minimize response time fire scene Liu et al., (2006). A fire
department’s response time the length of time it takes a fire truck to arrive at a fire scene starting
from the time the call was made to the fire department Flynn, (2009). Response time is important
for firefighting .Changes in land use and urbanization request continual evaluation of the spatial
spreading of fire stations and fire risk within the framework of costs/benefits. Since these issues
have a spatial element, GIS systems are being utilized for solving and supporting them
MartinLandré, (2008)

NFPA (National Fire Protection Association) 1710: requires countries to respond to fire incidents
within an eight-minute response time (excluding call-processing time intervals) to 90% of
incidents similarly provides essential benchmarks. Total response time, which is the time it takes
from the instant a call is received at the public safety answering point until the first unit reaches
the scene of an emergency occurrences is a baseline for fire department. For all due units and
effective response time should be measured and reported . Total response time is consisting of
the following terms :-

I) Dispatch time: The point at which the emergency alarm is received at the public safety
answering point to the point where sufficient information is known to the dispatcher and
applicable are notified of the emergency.

II) Turnout time: The time that begins when units acknowledge notification of the emergency to
the beginning point of response time. One minute (80 seconds) for turn out time

III) Response time: The time that begins when units are in route to the emergency incident and
ends when units arrive at the scene. The response time should be less than 240 seconds with less
than 80 seconds of the turnout time. According to NFPA 1221 requires that 95 percent of alarms
shall be answered in 15 seconds, 99 percent of alarms shall be answered in 40 seconds, and the
dispatch of the emergency response agency shall be finalized within 60 seconds 95 percent of the
time (ESRI, 2007). In accordance with the Standards for the Construction of Urban Fire Stations

MSc Thesis Page 11

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

(2017) and the development situations of domestic and foreign cities, the standards for the
construction of fire stations can be the site selection of fire stations should follow the principle
that firefighters should be able to reach the margin of their jurisdictions within five minutes after
receiving the set off instructions (the drive time of fire trucks should be within four minutes).
Habibi et al. (2008) suggested a response time of 3 minutes to 5 minutes be applied in a region of
high population density and residential land uses. The use of a 5-minute response time aids in the
fight against fire before it reaches flashover. Flashover occurs when the room temperature rises
between 825° F and 1,000° F and this takes place within 5 to 8 minutes from the initial ignition.
To successfully combat the fire, an effective fire attack must be carried out before 8 minutes
John r. Waters, (1999) . On the other hand, the National Fire Protection Association referred to
its pointing to five minutes limit as the suitable response time goal NFPA, (1710).

Figure 2.1 Cascading event chart for an emergency situation

Source : Oh et al., (2019)

Table 2.2 : Different fire response time standards for different countries

Country Name Fire response times standards

United kingdom 5,8, 10 & 20 minutes (based on risk type)
Ghana 5 minutes
South Africa 8 minutes (commercial& industry)
India 5 minutes
Ethiopia, AA 8 minutes

Source : Table was adjusted by Author using different source (Boakye, 2017; Green et al., 2014; Tali et
al., 2017) to align with the study needs.

MSc Thesis Page 12

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

2.4 ArcGIS network analysis

ArcGIS Network Analysis is an ArcGIS extension that allows performing a spatial analysis of
the network, including routing between nodes, location and allocation of facilities and definition
of service area among others. One of the major advances of the tools is the possibilities of using
the actual condition of the road network such as direction , speed limits and blocks of any kind.
Users can use ArcGIS network analysis to dynamically model realistic network conditions at
different times of the day including turn restriction , speed limits , Hight constraints and traffic
conditions Kumar and Kumar, (2016).

In Sui Min, (2012) , authors implied that network analysis is one of the main functions of GIS,
that plays an important role in vehicle routing, traffic tourism, urban planning, electric power,
communication and closest facility. This network is used, among other things, to build an
immediate, rapid and efficient emergency fire response system for the study area. The
computation of the optimum path between different locations on a network is the most
fundamental and the most crucial problem in network analysis Sui Min, 2012; F. Zhang et al.,
(2009).

In Reveshti et al., (2015), the authors also describe that network analysis program is one of most
useful applications of GIS. It has various applications in different natural and urban fields. One
of the urban applications of this program is finding the shortest and best route for timely
response in emergencies. International standards recommend 5-km radius for fire stations as well
as 3-5 min response time for reacting to fires.

In El & Badri, 2016; Nagib & Ali, 2010; J. Zhang, (2005), the authors have common sense about
ArcGIS network analysis, It uses the standard Dijkstra's algorithm to calculate the least
accumulated cost between the destination node and every other node and details of this algorithm
and its assessment can be found in.

In Tishi, (2015), the authors stated that the service area includes the area in which the firefighting
vehicle is supposed to reach at a place of fire incident within the standard response time within 5
minutes. Designated coverage area of a station should not be larger than the standard service area
of that station.

2.5 The role of GIS for fire stations

MSc Thesis Page 13

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

In Meyer., (2011), the author has described that Geographic Information Systems also plays a
vital role for locating fire stations. In the area of urban cities, GIS can effectively assist in the
design and operation of successful emergency response. Habibi et al., (2008) the authors stated
that GIS is an important and efficient tool that can be used by local administrations to reduce the
impact of natural disasters. Thus, GIS technologies have been used in fire analysis related to the
optimum location of fire stations, the researcher has made spatial analysis of urban fire stations
in Tehran, using an analytical hierarchy process and GIS. The authors stated that, using models
and software in urban planning has become prevalent in response to the complex dimension of
the urban issues and the role of many different indicators in this field.

In Yang et al., (2004) also carried out studies concerning the selection of fire station locations
using GIS. In Jasso et al., (2009), the author has carried out the location information of fires
from 911 emergency calls that could not be identified accurately. By matching linking address
information with coordinate information, GIS aids in determining the location of fires or
accidents in the least amount of time. The literature reveals an increasing use of GIS in the fire
service in the last decade Corcoran et al., (2007).GIS technology has become a powerful tool for
improving all aspects of fire service delivery systems .With all the challenges confronting
emergency crews today, effective response requires good planning, risk management,
comprehensive training, and intelligent deployment through preparation ESRI, (2006).

In Derekenaris et al., (2000), the authors also carried out Geographical Information System (GIS)
which is used for modeling of road networks offering algorithms to analyze and find the shortest
or minimum route through a network. GIS can be used to calculate the distance between sets of
origin and destination, meanwhile location-allocation can be utilized to establish site locations
and assign demand to sites. Street addresses can be converted to map coordinates (address
geocoding). A comprehensive GIS-based fire station location analysis can serve as the
foundation for a fire station master plan. This plan can demonstrate both the efficiencies and
insufficient of present fire stations coverage for a predefined travel time and provide a model for
future fire station coverage using the predefined or other travel time standards ESRI, (2007).

2.6 Location-Allocation

Only six different problem types are available in the ArcGIS location allocation analysis layer to
address a particular type of question which are minimize impedance , maximize coverage ,

MSc Thesis Page 14

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

minimize facilities, maximize attendance , maximize market share and target market share ESRI,
(2010). Location-allocation is a modeling technique , integrated into ArcGIS Network analysis to
solve problems of equating facility supply with demand using sets of criteria’s and constraints
Church,(1910). According to Drezner,(1995) describes that location-allocation model as the
process in which distribution of demands are taken into consideration to select the optimal
location for facilities from a set of available locations. Location-allocation is a public sector
planning tool used to attain the optimum location services and facilities within a given urban
area. including emergency services (police and fire stations or ambulances service), schools,
libraries, depots, warehouses, hospitals ,grocery stores and other public utilities are examples of
such services. In Bolouri et al., (2020) , the authors describe that location–allocation problem
aims to select the appropriate locations from a number of possible supply centers to provide the
appropriate services for customers and decision makers in connection to the customer allocation
(demand centers), based on given criteria for the facilities.

In Habibi and Lotifi, (2008) , the authors developed a model for fire station locations in district
6 of Tehran using the Geographic Information System (GIS). For the optimal location of fire
stations in Kuwait City Algharib, (2018), use the location–allocation analysis in a GIS
environment. Meanwhile, In Bolouri et al., (2018), the author employed both genetic and
simulated annealing algorithms to overcome a variety of location and allocation problems for fire
stations in district 11 of Tehran as multi objective problems; comparison finding for the
algorithms showed the superiority of the genetic algorithm.

2.6.1 Maximum coverage in ArcGIS

The issue of maximum spatial coverage was first introduced by Revelle and Church. It has been
confirmed that this issue, in terms of practical benefit and technical competence, has a significant
contribution to location analysis and modeling. A lot of commercial software programs based on
spatial information systems such as ArcGIS and Trans CAD, have also been synchronized for
general use and the paper gives an overview of the problem of fire station maximum coverage.
Revelle and Hogan, (1988). To implement emergency rescue center location like fire station ,
police station in order to cover maximum demand , maximum coverage type problem is used. If
the objective is to cover 100% demand for emergency support this solution will help to model it
ARIFIN, (2011). Maximum coverage location problem seeks to maximize the total number of

MSc Thesis Page 15

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

covered demands within a maximum service distance or time by locating the candidate facilities
near locations with high population density. Emergency service location models have the
objective to maximize the coverage within a standard response time or distance. One of the most
popular models that was applied to locating emergency service facilities is the maximal covering
location model. This model was often applied to selecting the best location for emergency
service facilities such as fire stations, ambulance and police stations Algharib, (2011)

2.7 Multi criteria analysis – analytical hierarchy process

Analytic hierarchy process is an effective, flexible and simple method to be used in certain cases
on which contradictory criteria make the decision-making process very difficult. This multi –
criteria evaluation method was proposed by Thomas L.saati in 1980 and since then had several
applications in different sciences.

Multi-criteria decision-making analysis (MCDMA) is an effective approach to support decision

making process in which different options are based on a set of explicit objectives measured by
weighted criteria. GIS-based MCDMA is powerful for spatial related decision-making process.
Which has enhanced multicriteria decision making process since 1980s and is a process that
combines geographical data and value judgments Estoque, (2011) Analytic Hierarchy Process is
a common method in the field of fire risk assessment. Its major goal is to solve the problem of
target selection and achieve the objective of optimization. The weight computation is a crucial
stage in determining whether or not the AHP will succeed.

Analytic Hierarchy Process is a method integrated with GIS useful for locating new fire
stations to asses all criteria simultaneously which analyzes complex decision problems under
multiple criteria, and helps the decision makers to set priorities and make the best decision
Abstract, 2014; Chhetri and Kayastha, (2015) and it’s one of the most applied approaches of
Multi Criteria Decision Analysis (MCDA) for spatial analysis decision makers. For site selection
process it starts with the identification of the pertinent site selection factors.

In Chaudhary et al., (2016), the author has proposed a method that combines analytical hierarchy
process and geographic information system to provide a decision maker for suitable fire site
selection, a case study from Kathmandu metropolitan city, Nepal.

MSc Thesis Page 16

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

In summary, GIS based network analyst and AHP are very important techniques when planning
for locating fire stations. This chapter discusses the notion of GIS based network analysis and
emergency response analysis and reviews the international and local context of time response.
From the literature, it can be noted that GIS techniques and procedures integrated with analytical
hierarchy processes have an important role to play a decision-making problem when locating fire
stations. This study seeks to fill a gap within the present literature by analyzing the impact of an
application in cities as well as highlight the usage of more advanced network analysis which
helps for emergency response analysis within GIS.

2.8 Conceptual framework

Emergency Response Analysis & Proposing new fire stations

Network Analysis AHP Analysis

Factor map 1. Inaccessible Area

Addis Ababa Road network
2. Proximity of Main
Road
ArcGIS Network Analysis Reclassified 3. Population Density
4. Land use
5. Distance from
Finding Best Route & Closest Weighted overlay
existing fire stations
Facilities Analysis
6. Incident density rate

Location Allocation Suitable sites

Final optimum
sites

Figure 2.2 Conceptual frameworks of the study

(Source : Prepared by Author ,2021)

MSc Thesis Page 17

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER THREE

MATERIALS AND METHODS

3.1 Description of the Study Area

3.1.1 Location
Addis Ababa is geographically located at the heart of the nation between 90 2' N latitude and
380 45' E longitude with time zone of UTC + 3 (East Africa Time), it occupies a total of 519.458
km2 land area surrounded by mountainous landscape and the average population densities of the
city is 7102.43 people/km2 (Fig 3.1). As the diplomatic Centre of Africa Addis Ababa hosts a
number of international organizations including the headquarters of African Union (AU) and the
United Nations Economic Commission for Africa (UNECA) (Abnet et al., 2017).

Figure 3.1 Location map of the study area

MSc Thesis Page 18

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.1.2 Population

According to the Central Statistical Agency of Ethiopia 2007 population census conducted by
the Ethiopian national statistics authorities, Addis Ababa has a total population of 2,739,551
people, including in the city and in the countryside. 662,728 households were counted in 628,984
housing units in the capital city, resulting in an average of 5.3 people per household. In
connection with this the population of Addis Ababa city is increasing from year to year because
of the migration of people due to economic and social problems, the population greatly increased
in the city. Based on census conducted by (CSA, 2020) the population figure is estimated to be
3,689,001 which comprises 1,743, 000 Male and 1,946,000 Female (Fig 3.2).

577,467
600,000
466,885
500,000 416,128 425,935
Number of Population

400,000 343,807 360,334

284,836 297,949 271,618
300,000 244,042
200,000 T.Popualtion
100,000 Area Km2
0 P.Density

Sub City Name

Figure 3.2 Population census of Addis Ababa city in year 2020

3.1.3 Climate

Addis Ababa has subtropical highland climate with a constant moderate temperature of roughly
21.5oC average high and 14oC average low throughout the year. Climatically, the maximum
temperature recorded in May around 25oC being its hottest months and the lowest average
temperature recorded was around 14°C in Nov & Dec, which is the coldest month (Fig 3.3).

MSc Thesis Page 19

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Avg.Temperature °C Max.Temperature °C Min Temperature in °C

30

25
Temprature (0c)

20

15

10

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Month

Figure 3.3 Monthly temperature variation

(Source: https://www.world-climates.com/city-addis-ababa-ethiopia-africa/)

300

250

200
Rainfall (mm)

150

100

50

0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Month

Figure 3.4 Average monthly rainfall in the study area

Addis Ababa city has a rainfall peak between July and August and minimum rainfall (driest
month) between November and December. There is 5 mm of rainfall in both months. The
greatest amount occurs in August, with an average of 245 mm, out of which close to 83.9% has a
chance of rain (Fig 3.4).

MSc Thesis Page 20

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.1.4 Topography

The average altitude of Addis Ababa city is 2,400 m above sea level, with the highest elevations
at Entoto Hill to the north rising over 3000 m (Fig 3.5). This makes Addis Ababa one of the
high-altitude capital cities of the world and takes the first rank from African cities. The city has
experienced spatial spread mostly towards the southern, eastern, and southwestern parts. The
northern part of the city, which is the foothills of the Entoto Mountains has a higher elevation
whereas the southern part of the city, which is the Akaki Kality area has a lower elevation up to
2052 m. The elevation map of the study area is extracted from earth explorer website Landsat
image with 30m spatial resolution (https://earthexplorer.usgs.gov/).

Figure 3.5 Elevation map

MSc Thesis Page 21

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.1.5 Road Transportation in Addis Ababa

Urban transport plays an important role in transferring people and goods from a given origin to
some other destination within a city or beyond. The construction of the first roads in Addis
Ababa dates back to the foundation of the city in November 1886 by Emperor Menilik II
and Etegie Taitu. The first roads served to link villages within bigger settlements by non-
motorized modes of transport. The construction of modern roads was intensified during the
beginning of the reign of Emperor Haile Selassie.

The construction of roads during that period was carried out by the public work department of
the municipality of Addis Ababa. The road construction had further been strengthened during and after
the brief Italian occupation. The total length of roads in the before 1983 was 1503 km while the
share of asphalted roads was below 20%. Following the establishment of the addis ababa city
road authority (AACRA) in 1998 large scale road construction had commenced. The total length
of roads in the city almost double to 3731 km between 1992 and 2012 and road density reached
12.9%. Today there are different mode of public transportation are there in addis ababa which
provide service for the community including Lionbus/Anbessa
buses,Shegerbuses,minibus,midibuses (Higerbuses),and etc. (Addis Ababa City Planning Project
Office., 2017)

Taking into account Addis Ababa‘s fast growth and to enable the transport sector to play its
required role, the Government has invested a huge resource to construct roads so as to expand
the road network. An effort has been made to improve the transport service provisions.
Preparatory actions required have also been taken to introduce a light rail transit (Transport
Policy of Addis Ababa., 2011).

3.2 Data and Data Source

3.2.1 Primary Data
The primary source of data is direct collection of fire incident spatial location using handheld
GPS with 3m horizontal accuracy level involved in input data for routing analysis. Based on
this the primary datasets that are used to compile this study work mostly: fire incidents locations,
road network. Additionally, oral interview between stakeholders were also done to get enough
information required for the study.

MSc Thesis Page 22

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.2.1.1 Fire incidents locations

This study made the use of primary fire incidents data and its attribute information was acquired
from Addis Ababa Fire and Emergency Prevention and Rescue Agency (AACFEPRA ) report
for 2020/21 stored in Excel Sheet format. This contained sub city name, the special name,
district, accident type, sex, date and time the fire engine started to move and the time the fire
engine arrived at the incident location, and others. The collected field data of fire incident
locations were imported into ArcMap. These are projected into Adindan UTM zone of 37 N
coordinate system and stored in shapefile as shown in figure 3.6. See Appendix 2 for data the
source.

Figure 3.6 Fire Incidents locations map of the year 2020-2021

MSc Thesis Page 23

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.2.1.2 Road network

The road network consists of a system of interconnected paved/unpaved carriageways which are
designed to carry buses, cars and goods vehicles; the road network generally forms the most
basic level of transport infrastructure within urban areas, the function of a road network is to
facilitate movement from one area to another. as such, it has an important role to play in the
urban environment to facilitate mobility. For this study the road networks data was obtained
from OpenStreetMap(OSM) (https://extract.bbbike.org/). After downloading all road segments
were classified by using Field Calculator Parser VB script and coded based on ACRA Standard
and Esri format into five functional class within GIS.

1. Expressway : The city roadways with full or partial control access and grade separation at
intersections that are reserved for motor vehicles. These are often built to provide a direct link
between key traffic generation points in an industrial, commercial, or business districts.

2. Arterial: It is a type of urban road that is commonly encountered within the cities. They are
designed to carry through traffic on continuous route. Arterial roads are divided highways that
have full or partially controlled access. Parking, loading and unloading are carefully regulated.

3. Sub Arterial : the city street which provide a lower level of travel mobility than arterial
roadway. Their spacing may vary from central business districts to sub-urban areas. typically
loading and unloading are carefully regulated.

4. Collector: this are the city roads that are constructed to collect and distribute traffic from and
to local street as well as to provide an access to arterial and sub-arterial streets. These roads can
be found in residential, business and industrial areas.

5.Local Street : this are city roads that offer access to residences, businesses, and other
structures. The traffic that is carried either originates or end on local streets. A local street might
be residential, commercial, or industrial, depending on the importance of the nearby areas.
Pedestrians are free to move on local streets.

The design speed limit used for this study is described in Table 3.1 which helped to calculate the
travel time of each road segment.

MSc Thesis Page 24

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 3.1 Functional Road hierarchy Detail

Road type Speed limit

Expressway 90 km/h
Arterial 80 km/h
Sub Arterial 60 km/h
Collector 50 km/h
Local 50 km/h

Source : (Addis Ababa City Planning Project Office., 2017)

Figure 3.7 Addis Ababa City Road Network Map

3.2.2 Secondary data

The secondary source of data were collected from a various source such as organization record,
institution and individual experts were very important to generate the desired result of this study.

MSc Thesis Page 25

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

The most widely used secondary data for this study work are existing fire station locations, land
use, study area boundary and population data. In addition to this official report, books, journal
articles, and internet sources were also used in order to have fully understanding and solving the
research problem at hand.

3.2.2.1 Existing fire station locations

The exact geographic coordinate of all existing fire stations in decimal degree was taken from
Google Map. On this year, 2020/21 Addis Ababa Fire and Emergency Prevention and Rescue
Agency (AACFEPRA) have ten existing fire stations namely(Fig3.8): Addis Ketema Sub City
Fire Station, Nefas Silk Lafto Sub City Fire Station , Kirkos Sub City Fire Station , Akaki Kaliti
Fire Station, Bethel Fire Station, Tigist Fire Fighting Equipment, Addis Ababa Fire Stations
Headquarters, Gulele Sub city Fire Station(Fire Accident), Addis Ababa Fire and Emergency
Management Authority and Arada Fire Station (https://wwwd.addismap.com/fire-stations).
Appendix 1 for the data the source.

Figure 3.8 Existing fire station location map

MSc Thesis Page 26

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.2.2.2 Elevation
The elevation map of the study area was generated from DEM in the Arc GIS environment
using Spatial Analyst tools. Therefore, based on the map shown, the elevation of the study area
ranges from the highest point of 3049 m to the lowest point of 2052 m above sea level as it is
indicated in the Figure 3.5 (https://earthexplorer.usgs.gov/).

3.2.2.2 Land use

Land use map is one of the major criteria used to select potential sites for fire stations in Addis
Ababa city. The land use map was collected from addis ababa city government plan and
development commission. The area coverage and percentage of each land use type indicated in
the following (Table 3.2)

Table 3.2 Existing land use of Addis Ababa city

No Land use type Area (sq.km) Area (%)

1 Administrative Services 8.35 1.60

2 Commerce 7.60 1.46
3 Environmental protection 158.20 30.41
4 Historical Construction and Site 0.18 0.03
5 Infrastructure Service 8.01 1.54
6 Manufacturing & Storage 30.99 5.96
7 Mixed Residence 221.85 42.64
8 Municipal Service 5.90 1.13
9 Social Service 15.03 2.89
10 Special Project 7.20 1.38
11 Street network 36.37 6.99
12 Transport 11.53 2.22
13 Urban Agriculture 9.09 1.75
Sum 520.29 100

The existing land use map of addis ababa is analyzed based on the data collected on July 2013
E.C (Addis ababa city planning project office, 2017). Among the various uses, thirteen
significant land use types have been mapped in an ArcGIS Figure (3.9) . The land use analysis of

MSc Thesis Page 27

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

the study area helps us to know which areas are more preferable or not for locating fire station
based on various reviews. Based on the above (Table 3.2), the largest part of the study area is
covered by Mixed Residence with the area covered of 221.85 km2 (42.64%) while the smallest
land use of the city is occupied by Historical Construction and Site with the area covered of 0.18
km2 (0.03%).

Figure 3.9 Land use pattern of Addis Ababa city

3.2.2.3 Study area boundary

Previously the city administration of Addis Ababa was divided into ten sub cities. Right now,
the city administration adds one more sub city namely Lemi Kura sub city so that Addis Ababa
has eleven total sub cities. With the help of those in this profession information, the boundary of
this sub city is taken from between Yeka and Bole sub city boundary. However due to unable to
get the boundary of the new sub city this study performs the entire analysis of this paper by using

MSc Thesis Page 28

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

previous boundary of the city (i.e.,10). The boundary of the study area was acquired from
Geospatial Information Institute (GII) in Esri shape file format as presented in the Figure 3.1

Table 3.3 Data and data source

Data Set Data Source Format Data Purpose

Study Area Boundary Geospatial Information Institute .shp For defining the study
(GII) location

Parcel Map of Addis Ababa Addis Ababa City Administration .shp To extract residential
Land house
Management Office
Population data CSA .pdf For giving weight for
maximize coverage
Land Use map of Addis Ababa Addis Ababa city government plan .shp Generate land use
and development commission classification
Road Network of Addis Ababa OSM .shp Input data for network
analysis
Existing Fire Stations https://www.addismap.com/fire- Lat and Lon For defining existing
stations service coverage
Fire Incident information AACFEPRA . xlsx Was used for closest
facilities and spatial
analysis
Demand points Generated from residential .shp Was used one of the
housing block parameters for location
allocation
DEM EarthExplorer (usgs.gov) Grid Generate Elevation
3.4 Software and instrument used

Some of material and software used for data collection ,management, and processing which are
used further data analysis applied in this study are(Table 3.4).

MSc Thesis Page 29

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 3. 4 Software and instrument used for the study

Type Source Purpose

Arc GIS v. 10.4.1 Self Network Analysis, Propose Optimum sites&
others
Google Earth Self To show the ground truth of the selected optimum
sites
IDRIS SELVA 17.0 Self For pair wise comparison
MS-Office Self To prepare the study text and to preform statistical
analysis.
Handheld GPS AAiT Collect the X and Y coordinate of Fire Incidents
Computer 8 GB Ram (Toshiba) Self Processing of raw data
External Hard Disc Self Data backup

3.5 Methodology
The methodology of the study was integrated with roads network analysis and locating optimal
fire stations for the study area using GIS software as shown in Figure 3.10. Six stages of the
process have been applied for network analysis process, beginning with collecting and preparing
the data that was used in the analysis (primary and secondary data), Then processed the data
which were used for further analysis. Following this, the creation of a Geo-database that were
store the prepared data. Then building both the network topology and the network dataset. And
then , the network analysis process has been applied and then it search optimal route between fire
incident location and fire station by applying Dijkstra’s algorithm to the road network of the
addis ababa city in connection with service area analysis of existing fire stations also applied.
Consequently , multi criteria analysis model has been developed to choose suitable sites for the
inaccessible place together with factor identifications. Based on this, the final optimum sites for
new fire stations was chosen by applying location allocation analysis particularly maximize
coverage. This were performed after all candidates site and residential housing polygon are
converted into point feature. 35 Candidate sites as Facilities, and population census 2020 as
Demand Points here population data was assigned as weights were used. Finally, using solver
tools 10 optimum sites was generated. The methodology adopted for this study is shown by a
flow chart in figure 3.10. Following this, the steps of methodology that have been described in
the subsequent sub-sections.

MSc Thesis Page 30

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Data Collection and Preparation

Primary Data
Data Processing
Secondary Data

Creation of Geodatabase

Building Network Topology

Service area analysis

Building Network Data Set

Best Route Analysis

Performing Network Analysis Closest Facilities

Analysis

Suitability Analysis

Location allocation

Input Facilities Input Demand

35 Candidate sites Population Census 2020 as Weight

Load and Display Data

Maximize Coverage

New fire stations

to the study area

Figure 3.10 Flow chart of the methodology

MSc Thesis Page 31

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.5.1 Data Collection and preparation

This phase includes collecting the study area map, preparing the road network data used for
further analysis. First of all , the road network data of addis ababa city was downloaded from
Open Street Map (OSM). The road network data are created with attributes such as functional
class to store the name of each road segment their length, travel time and speed of each road
segment was stored in ArcGIS.

3.5.2 Data processing

After data collection and preparations data processing stage is essential , since it is one the
procedure of transforming raw data into a compatible format that is more suitable for further
analysis and interpretable for user. For this study, the collected primary and secondary data were
stored in Excel format (.xlsx) has been imported into Esri shape file format(.shp).

3.5.3 Creation of geodatabase

The Geo-database is the native data structure used in ArcGIS and is the fundamental data format
used for both editing and management of the data. It provides the fundamental data structure for
managing and editing the data. A Geodatabase can be personal, file, or enterprise. In this
proposed method, a personal Geo-database has been created using ARCGIS. A personal Geo-
database is a database that can store, query, and manage both spatial and non-spatial data. It is
used to handle road network, existing fire stations, and parcel data.

3.5.4 Building network topology

To get good analysis and results, it is necessary to build a topology of the road network to
discover whatever errors in the data. These errors were verified and appropriate correction was
conducted using topology editing tool, and ultimately error-free digital road shapefile was
generated that could be used to create road network dataset . This was performed by applying
some topology rules including : Must Not Be Overlap, Must Not Have Dangle, Must Not
Intersect, and Must Be Single Part and they have also their own error correction method.

3.5.5 Building network dataset

To make a network analysis within ArcGIS, a network dataset is needed to be created . After
correcting the road network errors, it is ready for being used in building the network dataset that
is ready for use in the network analysis. The created geo-database network dataset with attributes

MSc Thesis Page 32

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

such as hierarchy, meters, minutes, one way, road class, travel time, and so on. For this study
Addis Ababa road was used to create a network dataset and it is well suited to model
transportation network. It consists a set of edges that represent the links over which agents will
travel, a set of junctions that connect edges and facilitate navigation from one edge to another, a
set of turns: that represents to store information that can affect movement between two or more
edges. As a result, shows in figure 3.11 the network dataset of the study area with 67152
junctions and 186772 edge for 42998 road segments.

Figure 3.11 Road network dataset of the study area

3.5.5.1 Network attributes

Network attributes are properties of the network elements that control travers ability over the
network. Examples of attributes include the time to travel a given length of the road, which
streets

MSc Thesis Page 33

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

are restricted for which vehicles, the speeds along a given road, and which streets are one-way.
These network attributes have basic properties including name, usage type and use by default .
This study also considers on usage type which specifies how the attributes will be used during
analysis. Here are some of network attribute applied in this study such as Connectivity and
Cost.

o Connectivity
To create network dataset choices that determine which edge and junction elements are created
from the source of the features. Ensuring that edges and junctions are formed correctly is
important for accurate network analysis results. Connectivity in a network dataset is based on
geometric coincidences of line endpoints, line vertices, and points as well as applying
connectivity rules set as properties of the network dataset. In this study Any vertex connectivity
policy method was implemented.

o Cost
Certain attributes are used to measure and model impedance such as travel time .These attributes
are distributed along an edge. During the calculation of a path ( also known as finding the best
route) network analysis usually involves the minimization of cost (also known as impedance).
Finding the fastest route (minimizing travel time) or the shortest routes (minimizing distance )
are two common examples. Travel time and distance are also cost attributes of the network
dataset. The Cost attribute which was presented in this study is based on data attached to the
shapefile consists of travel time and distance by car hence the data in the shapefile was measured
in minutes and meters respectively and One way restriction was ignored.

3.5.6 Performing network analysis

The road network analysis has been implemented using ArcGIS Network Analyst Extension. It is
a powerful extension of ArcGIS that provides network based spatial analysis, including route
analysis, travel directions, closest facility analysis, and service area analysis (Dabhade et al.,
2015) It allows users to dynamically model realistic roads network factors, such as turn
restrictions, speed limits, and traffic conditions at different times of the day. The ArcGIS
Network Analyst Extension uses the standard Dijkstra’s algorithm to calculate the least
accumulated cost between the destination node and every other node in the network. Based on

MSc Thesis Page 34

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

this well-known Dijkstra’s algorithm, there are two types of network analysis: best route analysis
and closet facilities analysis.

3.6 Data analysis

This study did both GIS and descriptive analysis. Meanwhile the following sections discuss and
illustrates that how the objective this study was achieved one by one.

3.6.1 Road network accessibility

This study began with the development of the road network data via model transportation (i.e.,
Network Analyst extension). For creation of new network dataset, addis ababa city
OpenStreetMap road dataset was used as the base. In addition, the attributes consist of time (in
minutes), distance (in meters), and speed (kmph ) set to support the source dataset attribute
fields. This dataset was calculated on the basis of the Addis Ababa city road authority
agency(AACRA) report for vehicles operating at speeds of up to 90 km/h as indicated in table
3.1. Time or distance is commonly used and effectively utilized to carry out accessibility
analysis. In the network dataset three important attribute were required and employed in this
study in order to perform accessibility analysis of the study including distance, travel time
and speed(Table 3.5).

Table 3.5 Important attribute in addis ababa city road network dataset

Attribute Description
Distance Length of the road in meter
Travel Time Time to travel from the beginning to the end of the road
Speed Speed of the road in kilometer per hour

I) Distance calculation: it was one of the fundamental cost attributes used as impedance factors
for road network analysis setting. It measures the distance of each road segment so the study
calculates the distance of all road segment conducted by Calculate Geometry Tool and it’s
measured in meters.

II) Calculation of Travel Time (Minutes)

Travel time is the time from when the first piece of equipment leaves the fire station till the time
when it arrives at the incident location. Travel time is the basic variable taken into account in
calculating the impedance of each road segment. In this study the travel time of all road network

MSc Thesis Page 35

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

segments was calculated by using the formula of : length of all road segments divided by speed
limit that segment and inserting its values depend on the type of roads (Table 3.2 ). It was
calculated by using a Field Calculator Tool and it’s measured in Minutes.

Length
Travel Time =( ) * 0.06 minutes (3.1)
Speed

For example, if one vehicle traveled in the arterial road with designed speed of 80kmph and
length of the segment is 1000m it will be calculated as:

1000
Travel Time = ( ) * 0.06 = 0.75 minute
80

Based on this formula the study received a value that has minutes as a measured unit. Therefore,
knowing the estimated time helps to examine the service coverage of existing fire stations as per
city standard and helps to determine the coverage of the population on the entire services .

III) Speed calculation: Required for calculating emergency travel times was added according to
the type of roads. This attribute was required for accurately designing and building the street
network to compute travel time and it is measured in meters per minutes.

Speed = Distance/ Travel Time (3.2)

3.6.2 Network analyst in ArcGIS

Network analysis is employed in this study for the spatial distribution pattern of the present fire
stations for finding the areas out of the coverage service and optimal site selections. The road
network analysis has been implemented using ArcGIS Network Analyst Extension. This study was
made the following Network Analyst Tools including: service area coverage , Best route , closest
facilities, and location allocations.

3.6.2.1 Service area coverage analysis

Service area is a function of accessibility. It is a region that encloses all accessible streets within
a given time and restriction. Service Area is used to evaluate the service coverage extent of a
facility within specified impedance and it is established by Network Analyst to examine
accessibility by exposing any gap in facility service coverage within a given time. In ArcGIS
Network Analysis, accessibility is measured in terms of travel time, distance, or any other
impedance on the network.

MSc Thesis Page 36

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Drive time polygons are ring-like polygons that display all accessible streets within a particular
travel- time from that point of origin. It can also be described as the defining measure of travel
time in the study area. The new service area network analysis tool can be used to create driving
time-based analysis showing regions that encompasses accessible areas within a specific time
window using the road network.

This study uses Network Analysts Service Area tool and road network emergency data with
predefined impedance cut off time of eight minutes response time. As the analysis sets were put
to obtain time served as the value of impendence, break value placed as eight minutes, path fixed
as away from U-turns, and facility enabled, the existing fire station location was used as a
reference to set a graphic select. For the return of detailed polygons, aggregation attributes and
line generations, impedance (distance and time) , additional settings were placed.

The service area analysis of this study was evaluated by use of the 10-fire station’s location and
road layers. Network analysis was performed to estimate the coverage of service areas by
utilizing the attribute information of each of the road line segments consists of length (meters),
travel time( minutes), and speed (kilometers per hour ) per road segment were calculated. It was
utilized by the five road types and their respective speed limits – local (50km/h), arterial (80
km/h), sub arterial (60 km/h) , collector (50km/h) and expressway (90km/h).

3.6.3.2 Best route analysis

Depending on the impedance this analysis finds the quickest and shortest path of the street so
that if the optimal route is the quickest route and shortest route the impedance factors are time
and distance respectively. As a result , the best route might be described as the route with lowest
impedance or the one with least cost where the impedance is chosen based on the city putted
standard. When choosing the optimum route any cost attribute can be employed as an
impedance. In this study time impedance factor was used for determine best route analysis as
shown in figure 4.3.

3.6.3.3 Closest facilities analysis

The closest facility solver calculates the distance in terms of time and distance of driving
between facilities and incidents or vice versa and determines which facility is closest to
incidents. The closest facility solver displays the best routes between facilities and incidents,

MSc Thesis Page 37

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

records their travel distance, response time and driving directions. This helps in emergency
situations to know the closest facilities that can be reached from the incident location, which in
turns reduces time, effort, resources and saving people life (Gubara et al., 2014). This network
analyst extension tools makes it simple to set the analysis parameters for the closet facilities
analysis such as the impedance factor, the start time, the period to reach the closet facilities, the
number of facilities to find, and the directions of travel (from incident to the facility or from the
facility to the incident). This study used travel time as impedance factor from facility to incident
for data analysis. Meanwhile as the facility (existing fire station) was added and as incident (fire
accidents report) were added as the incident sites and best route was generated within a
predefined response time of 8 minute . Then, by using the network analyst extension solver the
closest fire stations to the location of the fire incident were solved as shown in Figure 4.5.

3.6.2.4 Location allocation analysis

The ArcGIS location – allocation network analysis tool was used to determine the best location
for a new fire station in order to get the most maximum coverage throughout the deficiency
region. This analysis has been applied in this study to suggest new fire stations optimum sites
and allocating the location to it. After the road network was created the location allocation model
was employed. Impedance value (distance or time) and weight attribute, impedance cutoff,
number of facilities (facilities to choose), and location-allocation problem type are among the
parameters contained in the Network Analyst tools. There are seven problem types of location-
allocation to consists of (ESRI, n.d.; Hakimi, 1964) such as : minimize impedance, maximize
coverage , maximized capacitated coverage, minimize facilities, maximize attendance ,
maximize market share and target market share . Among of them the maximize coverage
method was applied in this study to overcome the problem of an 8-minute coverage deficiency in
the study area. (Rahman & Smith, 2000) with the formula.

Maximize {F=∑ 𝑎𝑖 𝑋i} (3.3)

where I is the group of demand points or the time duration or the population as a weight, and xi
is the location-allocation parameter with a value of either 0 or 1. After setting the parameters of
the analysis layer and selecting the “maximize coverage” location-allocation type, the spatial
locations of services (facilities) and the demand points were putted as residential housing
represented by point layers in the Network Analysis window. 35 candidate sites were inserted as

MSc Thesis Page 38

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

facilities and the selection of the optimum sites from candidates was conducted by using
maximize coverage. After that the location allocation analysis was solved to obtain the result
and ten new fire stations were suggested for the study area.

3.6.3 Spatial relationship of fire incidents with population density

The map shown in figure 3.13 illustrate that fire stations coverage and fire incidents occurrence
corresponding with population density in a study area. The population density for each sub city
of the study area was calculated by dividing the population data of each sub city for the year of
2020 and respective area of the sub city as represented in the following equation and graphically
indicated in figure 3.12

PD = P/A (3.4)

where P is Population of sub-city, A is Area of sub-city, PD is Population density.

The population density map layer was processed and had a range of values from 1979 to 39869
representative of the number of people per sub-city area in sq.km

Figure 3.12 Population density map of sub cities in Addis Ababa

MSc Thesis Page 39

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Figure 3.13 8-Minute Service Coverage Overlapped with Population Density

3.6.4 Spatial analyst tools

Spatial analysis is the process of examining the locations, attributes, and relationships of features
in spatial data through different analytical techniques in order to address a question or useful
knowledge. Spatial analysis extracts or creates new information from spatial data.

3.6.4.1 Kernel density analysis

It is one of the spatial analyst tools in GIS environment for making density analysis of features in
a neighborhood around those features. It was used to show the spatial availability and to find the
accident probability zone of fire incidents hot spot information in the study area. To calculate
magnitude per unit area from point features using kernel density to fit a smooth taped surface to
each point. In this study kernel density analysis has been utilized depending on the point where
an incident occurred within two years (2020/2021).This analysis evaluates the number of events

MSc Thesis Page 40

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

across a given surface by calculating the overall quantity of fire incident located within a given
search radius from a target point as presented (Fig3.14) .

Figure 3.14 Fire Incident Density Rate Value map

3.6.4.2 Candidate site selection for new fire stations

Suitability analysis was conducted to select candidate sites for proposing and allocating
additional fire stations. Site suitability is the process of allocating new and candidate sites by
examining the existing structure, pattern and condition of existing sites using a set of specific
criteria. For this commitment, the present study had prepared a suitable model for determining
appropriate location of fire stations by integrating six factors such as : inaccessible area
,population density, land use, proximity of main roads, distance from existing fire stations and
incident density rate of the area. All these maps have been converted into raster so that each
pixel can detect a score. It is important in suitability analysis to set a score for each category at 1-

MSc Thesis Page 41

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

5 or 1–9 point scale as per their suitability. For this study 1-5 scale and the reciprocal is 1,1/2
…,1/5. has been utilized when the row criterion is less important than the column criterion. This
decision approach supported to propose 35 candidate sites for allocating new fire stations
facilities for the study areas. Among of this the most optimum for allocating new fire stations
was determined by maximize coverage.

An intensive literature review was carried out in collaboration with this study to identify the
criteria that have significant influence on the candidate site selection for fire stations, six criteria
has been identified that have a significant role in planning for an optimum fire station within the
scope of this thesis is shown in the table (3.6).

Table 3.6 Factors used for candidate site selections for new fire stations

Factor Literature Source

Population Density (Nyimbili & Erden, 2020)
Land Use (Chaudhary et al., 2016)
Incident density of the area (Chiedu, 2011)
Proximity of Main Road (Habibi et al., 2008)
Distance from Existing Fire Stations (Nyimbili & Erden, 2020)
Inaccessible Area > 8 Minute -

3.6.4.3 Factors used for this study

The mentioned factors applied in this study (Table 3.6) are discussed in the following section in
a concise and precise manner.

I) Distance from existing fire stations

The existing fire stations were determined by Euclidean distance according to the international
standard 5km are the most suitable to propose additional stations. This study applied this
standard to find candidate sites for the study area. Moreover, new fire stations must be far from
the existing fire stations existing , making sure that newly fire stations are situated in the area
where there are no fire stations. Its reclassified map is shown in Figure 3.15 (a)

II) Inaccessible area

MSc Thesis Page 42

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

This is another major important factor in the areas that don't serve with predefined response
time within eight minutes during emergency time. This study employed the service coverage
greater than 8 minutes deriving response minutes determined by road network data which used
during emergency response and travel time used as impedance factor solved by network analysis
extension of service coverage. This analysis result has played a major role in proposing
additional candidate sites for fire stations in the deficiency area. Its reclassified map is shown in
figure 3.15 (b)

III) Population density

Population density is another important factor for locating fire station location. This study was
need to consider the population density of the area of the site to be selected. The densely
populated areas are usually housing on a relatively densely house area and close to each other
which increases the probability of the occurrence of fires and the speed of its spread. The quick
response to these areas and ensure the delivery of fire department services to them has a great
importance in order to avoid unwanted huge damages of firehouse incident. This study
performed point density map using by dividing the total population collected from CSA to
existing residential housing accordingly their sub city. This analysis was applied due to the
inability to find enough information about how many people lived in each census block.
Therefore, the population density has been classified into five class and its reclassified map is
shown in figure 3.15 (c).

IV) Proximity of main road

For the need of quick access by car and emergency service, proximity to main roads is being
considered as one of the most important criteria in fire stations site selection ;however, a new fire
station should be located close to the main road city as far as possible . Close accessibility is
taken as the best preference for fire station selection. it will offer higher accessibility in term of
travel time during emergency occurrence response. The proximity of main road criterion had a
value range from 0 to 300 m indicating the distance from the main road segment layer
determined by Multiple Ring Buffer analysis with incremental distance values in meters of 0−
60 m, 60−120 m, 120−180 m, 180−240 m and 240−300 m (Habibi et al., 2008). Its reclassified
map is shown in figure 3.15 (d).

MSc Thesis Page 43

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

V) Land-use

Land use maps are another important factor for locating fire stations. Based on the land use map
prepared by Addis Ababa city government plan and development commission (Figure 3.9. For
this study four land use classes were considered more suitable to locate fire stations namely
mixed residence, commerce and urban agricultures has been utilized. Its reclassified map is
shown in figure 3.15 (e)

VI) Incident density rate

Incident density rate area is another important factor to locate optimum location for fire stations
hence this factor helps to identify the more frequent occurrence of the incident area in the study
area with the help of spatial analysis tools of kernel density. Therefore, the more incident
density rate location was well suited to propose new fire stations. Its reclassified map is shown
in Figure 3.15 (f). All the reclassified map of this study are shown in the following figures.

MSc Thesis Page 44

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Figure 3.15:Reclassified thematic maps: Distance from existing fire stations ( a ), inaccessible area ( b )
, Population density (c ) , Proximity of main road( d ) , Land-use map (e) , and incident density rate (f)

MSc Thesis Page 45

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.6.5 GIS Based Multicriteria decision analysis

This study adopted Multicriteria decision making analysis that is called the Analytic Hierarchy
Process (AHP), and the ArcGIS Model Builder was utilized to decide the reasonable locations of
fire stations in Addis Ababa City. This analysis helps to find out the candidate sites to the study
area. Analytical Hierarchy Process solution approach is well suited to the decision problem of
candidate site selection. The solution process of the AHP decision model is begin with the
definition of the criteria for site selection which are organized in a hierarchy from a general goal
to multiple attributes in progressive levels. The weights of the site selection criteria are computed
using qualitative factors and analytical pairwise comparisons to determine their relative
importance in respect to the various opinions of the decision-makers. These final criteria weights
are then used in the final analysis results to recommend the optimal sites for constructing fire
station facilities (Yang & Lee, 1997).

Criteria definitions

The first stage of multicriteria decision analysis(MCDA) is concerned with combining

information and data from several criteria for decision-making purposes (Abudeif et al., 2015).
The weight preference was applied by six criteria using AHP method including: Inaccessible
Area (Ia), Population Density(Pd) ,land use(Lu), distance from existing fire stations(Defs) ,
proximity of main road (Pmr) and incident density rate (Idr).

Criteria weighing

The process of assigning weights to the criteria is one of the most sensitive points in this type of
analysis. The analytic hierarchy process (AHP) was used for criteria weights, as this method
proved its high efficiency in solving complex problems(Akinci et al., 2013). It was introduced by
(Saaty, 1977) and is very helpful in decision-making. The AHP language scale has 9
importance, as shown in Table 3.7

MSc Thesis Page 46

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 3.7 Scale of relative importance of analytic hierarchy process (ahp) based on saaty.

Intensity of Importance Definition Explanation

1 Two activity contribute equally to the
Equally Important
objective
3 Experience and judgment slightly
Moderately More Important
favor one activity over another
5 Experience and judgment strongly
Strongly More Important
favor one activity over another
7 An activity is strongly favored and its
Demonstrate More Important
importance demonstrated in practice
9 The evidence favoring one activity
Extremely More Important over another is of the possible order of
affirmation.
1/2, 1/3, 1/4., 1/5, 1/6, The reciprocal of the above-
1/7, 1/8 and 1/9
mentioned importance

The AHP underwent successive stages of generating a pairwise comparison matrix with criteria
values from 1 to 9 as shown in Table 3.7. The diagonal must equal 1(See formula 2) as it
represents the comparison of one criterion against itself, and values above the diagonal are the
inverse of those below it. After the numerical comparison the sum of the columns is obtained and
then divided by the total sum to get the relative weight. To obtain the final relative weight of
each factor this study follows five consecutive stages as discussed below.

First Stage: Assigning criteria preference values

The preferences are determined in the AHP based on pairwise comparisons which are based on
evaluating each element against all other criteria in a specific hierarchical level. The criteria are a
specific character such as a variable or evaluation parameter and the reference comparison point
is an element at the top of the hierarchical chain.

𝑎𝑖𝑗 : where i, j = 1, 2, 3, …n (3.5)

𝑎𝑖𝑗 = 1 for i= j (3.6)

MSc Thesis Page 47

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

𝑎𝑖𝑗 = 1/𝑎𝑖𝑗 for i≠ j (3.7)

The preference matrix is a result of pairwise comparisons of all criteria at a specific hierarchy
level. Formula (3.5) shows a matrix of n × n dimensions where n denotes the number of criteria
being compared. Formula (3.6) represents the principle of preference in which two identical
criteria are not differentiated by preference when compared to each other and All values of the
elements on the diagonal of the matrix are equal to 1 because the difference in preferences is
conveyed by the number 1 (Table 3.8). Formula (3.7) represents the principle of preference value
that has inverse importance.

Table 3.8 Preference values of criteria pair wise comparison matrix using AHP

Criteria Ia Pd Lu Defs Pmr Idr

Ia 1 3 2 4 3 5
Pd 0.33 1 2 3 4 5
Lu 0.5 0.5 1 2 3 5
Defs 0.25 0.33 0.50 1 4 4
Pmr 0.33 0.25 0.33 0.25 1 3
Idr 0.20 0.20 0.20 0.25 0.33 1
Sum 2.61 5.28 6.03 10.5 15.33 23

Second Stage : Computing normalize pairwise comparison matrix

In this stage the normalized pairwise comparison can be calculated as all the elements of the
column are divided by the sum of the column(Table 3.9).

Table 3.9 Normalized pairwise comparison matrix result

Criteria Ia Pd Lu Defs Pmr Idr

Ia 0.383 0.568 0.332 0.381 0.196 0.217
Pd 0.126 0.189 0.332 0.286 0.261 0.217
Lu 0.192 0.095 0.165 0.190 0.196 0.217
Defs 0.096 0.063 0.083 0.095 0.261 0.174
Pmr 0.126 0.047 0.055 0.024 0.064 0.132
Idr 0.077 0.038 0.033 0.024 0.022 0.043
Sum 1 1 1 1 1 1

MSc Thesis Page 48

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Third Stage: The relative weight of preference value percentage for each criteria

The percentage of preference values between two parameters where one is in a column and the
other is in a row can be determined by using the formula(3.8)

𝑎𝑖𝑗
%pv = ∑𝑛 (3.8)
1=𝐼 𝑎𝐼𝑗

where %𝑝𝑣 is the percentage of preference value and 𝑎𝑖𝑗 is the value of preference between

two parameters in column and row respectively and ∑𝑛1=𝐼 𝑎𝐼𝑗 represents the total sum of the
column criteria. The relative weight percentage for all criteria were computed and presented in
Table 3.10. Finally , the relative weight values were determined by using the formula(3.9)

∑𝑛
𝑤𝑖 = 1=𝑐 𝑎𝑖𝑐 * 100 (3.9)
𝑚

𝑛
where 𝑤𝑖 is the value of the relative weight for the row parameter, ∑1=𝑐 𝑎𝑐𝑖 is the sum of

preference values for a row parameter, and m is the total number of criteria.

Table 3.10 The final relative weight of each criterion

Factors Relative Weight ≈ Percentage (%)

Inaccessible area 0.357 36
Population density 0.238 24
Land use 0.172 17
Distance from existing fire stations 0.125 13
Proximity of Main Road 0.069 7
Incident density rate 0.038 4
Total 1.00 100

Fourth Stage: Verification index for calculating consistency index

In this stage, for the determination of consistency ratio the eigenvalue of comparison matrix has
been done first. This was computed by multiplying the preference value of each column (Table
3.8) with respect to their relative weight. After that summing the values in the row then dividing
it with its relative weight which is called weighted sum. Finally, eigenvalue was calculated by
taking the average of the criteria weight as shown in the Table 3.11

MSc Thesis Page 49

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 3.11 Determination of eigenvalue for the comparison matrix

Criteria Weighted sum Relative wight Weighted sum/ Relative wight

Ia 2.312 0.357 6.476
Pd 1.502 0.238 6.310
Lu 1.117 0.172 6.494
Defs 0.807 0.125 6.456
Pmr 0.42 0.069 6.086
Idr 0.292 0.038 7.684
Average 6.584

Therefore 𝜆𝑚𝑎𝑥 = 6.584 and m =6 then Consistency index (CI) is computed using the

formula:

𝜆𝑚𝑎𝑥 −𝑚 6.584 -6
CI = = ≈ 0.116 (3.10)
𝑚−1 6-1

where 𝜆𝑚𝑎𝑥 is the eigenvalue of the comparison matrix and m = the total number of criteria

Five Stage : Calculating the consistency ratio(CR) percentage

In final , the consistency ratio percentage was calculated by using the formula (3.11)

CI 0.116
CR = = (3.11)
RI 1.25

= 0.09 < 0.10 , Pass the consistency test

where CR Consistency ratio determined from consistency index by random value

RI is the random consistency index based on the number of criteria taken from Table 4.1.8

The consistency ratio is 0.09: the result of consistency value for this study is less 0.1 which is
acceptable value. The consistency values must be in a range not exceeding 0.1 (10%), as values
higher than 0.1 mean more conflict in consistency at this time we have to reconstruct the
pairwise matrix as presented in figure 3.17.

MSc Thesis Page 50

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 3.12 Random stability index (Saaty, 2000)

N 1 2 3 4 5 6 7 8 9 10
R 0 0 0.52 0.89 1.11 1.25 1.3 1.4 1.45 1.49

The value of the random consistency index based on the above table is equal to 1.25 because the
number of criteria used is 6. The percentage of the consistency ratio was 0.09*100 = 9%. After
assigning weights to each criterion AHP layer was formed in the ArcGIS and by analyzing the
weighted layer the final site selection layer was confirmed .

3.6.7 Weighted linear combination (WLC) in ArcGIS

Weighted Linear Combination (WLC) is a type of multi-criteria evaluation method in GIS

environment used to evaluate the suitability of fire stations sites. This allows the decision maker
to assign weights according to the relative importance of each suitability map and combines the
reclassified maps to obtain an overall suitability score. WLC implemented with the following
four steps. The first step is determination of data, second step is development of attributes of
criteria and their standardization, the third step is determination of the relative importance
weights, and at final step is overlay analysis/combining of all criteria (Malczewski, 2004).
Finally, to find out the candidate sites for fire stations to this study the study has been conducted
the following procedures (Fig 3.16).

Figure 3.16 Candidate sites selection procedure using model builder for new fire stations

MSc Thesis Page 51

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Objective of the Research

Criteria Preparation Creation of raster feature for each criterion

Assigning preference value of each criteria Reclassification each raster feature

criteria

Calculation of normalized Eigen vector each

criteria
False

If CR< 0.1

True

Preparation of the relative weight map of each

criteria

Overlay the weight maps

New fire stations map

Figure 3.17 Methodological flow chart of fire stations Site Selection of Addis Ababa city

MSc Thesis Page 52

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

3.6.8 Descriptive analysis

Descriptive analysis has been conducted by Microsoft Excel used to make simple statistical
arithmetic calculations as well as for computing Sum , Percentage and Total and then the result
was presented in descriptive tools such as tables and charts.

3.6.8.1Statistical analysis of fire incidents data

According to the AACFEPRA report data shows commercial and residential areas are the
highest common areas in which fire accidents occurred on average for two years. Which was the
highest crowded condition in these areas has contributed more than the other infrastructure.
Because of that exchange of goods could take place in a commercial area and a lot of people are
there around the residential place. From the report indicated that the highest fire incident was
occurred in bole sub city numerically 50 but the lowest is occurred lideta sub city numerically 11
from the total of both years as shown in table 2.1 and graphically represented as shown in Fig
3.18.

Addis Ababa Fire Incidents

Sept 2020 - Aug 2021
Number of Fire Incidents

42 48
32 23 32 29 25
60 17 11 15
40 2020
20
0 2021
Total
Percentage

Sub Cities

Figure 3.18 Plot Chart Between Sub Cities Versus Number of Fire Incidents

( Source : AACFEPRA, Sept 2020- Aug 2021 fire report)

MSc Thesis Page 53

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER FOUR

RESULTS

4.1 Service area coverage

The analysis process was undertaken using a travel time analysis based on the city standard. The
travel time analysis determines the level of accessibility of existing fire stations in Addis Ababa
city. To find the service area of the existing fire stations service area tool of network extension
has been used. Therefore, service coverage area of existing fire stations in Addis Ababa city was
examined using eight minute of travel time as indicated in Figure 4.1.

Figure 4.1 Service Area Coverage of Existing Fire Stations

MSc Thesis Page 54

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Table 4.1 Fire stations coverage within 8 minutes in Addis Ababa city

Coverage within 8 minutes

Name of Fire Stations
Population % Area (km2) %
Addis Ketema Sub City Fire Station 496353 20.20 18.22 6.60
Nefas Silk Lafto Sub City Fire Station 254949 10.37 41.69 15.10
Kirkos Sub City Fire Station 375186 15.27 32.20 11.66
Akaki Kaliti Fire Station 148678 6.05 56.26 20.38
Bethel Fire Station 319773 13.01 36.25 13.13
Tigist Fire Fighting Equipment 104678 4.26 24.08 8.72
Addis Ababa Fire Stations Head Quarters 453853 18.47 28.61 10.36
Gulele Sub city Fire Station(Fire Accident) 1880 0.08 0.87 0.32
Addis Ababa Fire and Emergency Management Authority 228349 9.29 21.87 7.92
Arada Fire Station 73744 3.00 16.04 5.81
Total 2,457,443 100 276.10 100

From the Table 4.1 indicate that service area time response shows that the served population
from existing fire stations is 2,817,315 (66.62%) out of 3,689,001estimated population which is
distributed over the area 276.10km2 which corresponds to 53.15 percent of the city are served out
of a total of 519.458 km2. Along with this the remaining 243.358 km2 which corresponds to
46.85 percent of the city are unserved. Each fire stations service area is a much more
comprehensive way to determine the most vulnerable location during any kind of fire incidents
.The finding shows that the area of existing fire stations coverage to which the firefighter
vehicles responded within eight minutes and the available of roads network during emergency
response time.

Therefore, it shows that existing fire stations do not cover the whole city area so that it is
necessary to make proper plans to protect the un-service area by establishing new fire stations. It
is obvious that the fire fighter vehicle has to travel more than prescribed time to reach fire
incidence in un-served areas as shown in the map which may be disastrous both in terms of life
and property.

MSc Thesis Page 55

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

No. of Served Population

500000
400000
300000
200000
100000
0

Population
Precent
Area
Precent
Name of Fire Stations

Figure 4.2 Served population and area covered with their percentage

4.2 Best route result

The best route analysis result generates the best route between two locations based on travel
time which depends on the traffic conditions with parameter such as travel time and road
direction available on the network. The network analyst extension makes it easy to set the best
route analysis parameters such as the travel time as an impedance factor for emergency response.
Finding the optimal path means that selecting either the shortest path or the path having
minimum travel time Eleiche, (2011). GIS intends to find the optimal route between the given
origin and destination involving shortest distances as well as shortest travel time. After adjusting
the best route analysis settings, the start location is Arada_FS location which is main fire
stations and FI_253 location as the destinations, and then by applying solver tool it generate the
best route between these two locations. The integrating of all the impedances(time and distance),
the best route to reach the incident location is displayed in (Figure 4.3). Its total distance is
8593.8m and the whole time takes is 6 minutes.

MSc Thesis Page 56

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Figure 4.3 Best route map from arada sub city.

Figure 4.4 Best route directional windows

MSc Thesis Page 57

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

4.3 Closet facility area

The closet facilities analysis finds the nearest facilities that can be reached in a specific period
from an incident location based on travel time and traffic condition available. To apply this
analysis, it is required to make a closet facility analysis layer and its analysis properties.
Analyzing closet facilities needs to perform with parameters such as impedance factor, start time,
period to reach closet facilities, number of facilities to find, the direction of travel and
accumulation consists of travel time (Minutes) and Distance (Meters). This layer is useful in
determining the closet facility or facilities to an incident based on a specified network. This
study finds the closet facilities that can reach in a specific period from an incident location based
on 8-minute travel time. Figure 4.5 illustrates the analysis started with existing fire station
location as a facility and ends with all the recorded fire incidents locations as an incident and
then by applying solver tool it generates the shortest path between all locations.

Figure 4.5 Closest fire stations within 8 minutes to fire incidents

MSc Thesis Page 58

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Solving for this analysis the finding within eight-minute closest facilities. It generate a total of
230 routes and maximum travel time of 7.99 min, minimum travel time of 0.16 min and
maximum length of 11890.42 m , and minimum length of 151.68 m.

In similarly way, the following map explain that the closet facility analysis result and directional
window between Arada fire station and FI_272. This was obtained by, firstly all the existing fire
stations are inserted as a facility as illustrated in Fig 4.6 black outlined by a rectangular shape as
well as FI_272 as incident. And then by hitting the solver tools in network analyst. The system
gets Arada fire station as the nearest facilities among of ten fire stations. Then, generate the
optimum route from Arada fire station to FI_272. Its total distance is 3865.6m and the whole
time taken to reach the incident place is 3min in Fig 4.7.

Figure 4.6 Closest Facilities Analysis Result from Arada Fire Station to FI_272

MSc Thesis Page 59

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Figure 4.7 Closest facilities direction result from Arada Fire Station to FI_272

4.4 Finding optimum sites based on location allocation

In this study running location allocation using maximize coverage problem types has been
conducted to identify optimum sites for additional fire stations. 35 potential candidate sites were
identified in the previous sections using AHP method which can offer fire stations service in
poor deficiency area within this region to be covered and will have service coverage for
emergency situation . Since it was chosen by using analytical hierarchy processes and GIS as
stated in previous data analysis sections. Among 35 sites , ten optimum sites have been chosen
by location allocation analysis particularly maximize coverage by giving population as weight.
Then, the final optimum sites are shown in figure 4.2.8(b)

MSc Thesis Page 60

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

a b

Figure 4.8 The selected candidate sites(a) & Proposed sites to locate additional fire stations (b)

As shown in the figure 4.8 (b) above. The following screenshot image was taken from Google
Earth search engine which indicates that what their location looks like the selected optimum site
on the ground.

Site-1 Site-2

MSc Thesis Page 61

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Site- 3 Site-4

Site-5 Site-6

Site-6

Site-5

Site -7 Site-8

Site- 10
Site-9
Figure 4.9 The Ground Truth of Chosen Sites for Fire Stations to The Study Area Taken from Google Earth.
Table 4.2 Location allocation analysis using travel time (in minutes) as impedance from 10 optimum fire
stations locations.

MSc Thesis Page 62

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Name Facilities Demand Count Demand Total travel time Total Weight
Type (Number of residential) Weight (In Minute) Time (In Minute)
(Population)
Site -1 Chosen 40456 160733 6.70 271176.68
Site -2 Chosen 23726 69520 6.61 156764.17
Site -3 Chosen 10538 30876 5.77 60833.55
Site -4 Chosen 70172 203909 5.66 397354.39
Site -5 Chosen 86351 283726 6.01 518916.00
Site -6 Chosen 33075 91728 6.14 203394.88
Site -7 Chosen 44389 121760 5.08 222572.95
Site -8 Chosen 43552 118461 6.25 271999.55
Site -9 Chosen 29815 86151 4.89 145795.35
Site -10 Chosen 5316 13348 6.92 36791.73

The result Table 4.2 shows that the new optimum site for fire station in site-10 had the lowest
allocated demand with 5316 residential polygon and a total of approximately with just 13348
people allocated to it and they will be accessible within 6.92-minute total response time.
However, site-5 had the highest allocated demand with 86351 residential polygon and a total
approximately with 283726 people allocated to it and they will be accessible within 6.01 total
minutes response .

From Table 4.2 the output depicts that in term of total travel time , the least amount of travel
time were obtained in site-9 with will provide fire emergency response time within 4.89 minutes
whereas the maximum amount of travel time were generated in site-10 will also provide fire
emergency response within 6.92 minutes.

MSc Thesis Page 63

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

4.2.5 After Adding additional fire stations

After adding 10 more additional fire stations to the study area. It made the service area of fire
stations coverage can be extended from 53.15% up to 87.44 % within 8- minute impedance
travel time based on the emergency response time of Addis Ababa city. Those chosen sites help
to determine the area of improvement in response time and coverage extent.

Figure 4.10 New service area created after adding new fire stations

MSc Thesis Page 64

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER FIVE

DISCUSSION

5.1 Comparative analysis

In the recent years, many researchers have undertaken similar study on locating fire station in
different city. However, Tali et al. (2017), estimated the fire station needs in Karnataka, India,
based on 5 min of travel distance coverage area. The research used simple GIS network analysis
and population data for assessing optimal sites for new fire stations. Habibi et al., (2008), the
researcher used an integrated AHP and index overlay (IO) logic for locating best location of fire
stations had chosen population density, accessibility, proximity to hazardous point and size of
plots as selection factors for fire station location the zone 6 of Tehran, Iran. Oh et al., (2019) ,the
researcher proposed a nonlinear programming model to determine the locations of fire stations
in the city of Kingsville, Texas. The main objectives of the research were to minimize response
time with optimal fire stations and cover the maximum number of inhabitants, in terms of the
geographical area and population. Bolouri et al., (2018), the researcher proposes multi-objective
location–allocation model for Region 11, Tehran aimed to minimize the distance and time as
well as maximizing the coverage of the city. Chevalier et al., (2012) developed a decision-aid
tool for locating fire station in Belgium. Their decision-making involved occurrence of fire
hazards, high potential risk areas, and fire response capacity.

This study taken into account six criteria have been chosen to find the optimum location of fire
station in Addis Ababa city i.e., population density, Proximity of main roads, Land use ,
Distance from existing fire stations, Distance from existing fire stations, and Incident density rate
based on local standard of 8-minute response travel time along with this the study propose ten
new fire stations using AHP and network analysis was applied . However ,many researchers such
as Wei et al., (2011) had chosen population density, building loss after catching fire and road
distance to nearest fire station as selection criteria for preparation of the layout of fire stations
within Beijing 3rd ring road by combining GIS and AHP together. Erden and Coşkun, (2010)
had chosen population density, proximity to main roads, distance from existing fire stations,
distance from hazardous material facilities, wooden building density, distance from area

MSc Thesis Page 65

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

subjected to earthquake risks as selection factors to ensure optimal site location(s) for fire
stations in Istanbul, Turkey.

Chaudhary et al. (2016) had chosen distance from roads, land-cover, distance from rivers
population density as selection factor has been used to find out high suitable site for fire station
locations. Uddin and Warnitchai, (2020) , the researcher develops a model for locating optimal
sites for new fire stations with three integrated method such as geographical information system,
multi-criteria analysis, and set cover algorithm to solve the location allocation problem of fire
station in complex urban. Idhoko et al. (2015) evaluates the digital road network of Oyo-Town,
Oyo State, Nigeria for management of emergency service as long as perform query to select the
best route in case of emergence occurrence and access the closest facilities. The main objectives
of the research were designing a database, spatial data acquisition, database creation and spatial
analysis for the town. Sarker and Ghosh (2021) determines the potential site of fire stations
based on the criteria such as distances between fire stations, number of industries and the level of
risk in the city coverage area, road accessibility and Arc-GIS network analysis as well as
evaluate the existing fire station in 5-minute response time this helps the researcher to identify
the deficiency coverage in Gazipur industrial area. Nyimbili and Erden (2020) used the
integration of fuzzy analytical hierarchy process and GIS to obtain optimal site for new fire
station in Istanbul region, Turkey and the author had chosen six the criteria to select the best sites
i.e. wooden building density, density of hazardous material, distance from existing fire stations,
distance from the areas subjected to earthquake risk, proximity to main roads , and high
population density using this the study propose 34 new fire station. Anna-hajnalKa, (2018), the
research paper had focused on developing an optimal route which is used during an emergency
situation, the time and distance has been used as an impedance factor , to evaluate between
accident and emergency units. Mezaal et al. (2020), focused on the sample of Puchong fire
station in Malaysia based on the response time Using Network Analysis Extension some of them
utilized are Service Areas and Closest Facility. The service area of the research was evaluated
by 10-minute response time along with 325 fire incident data that were used for closest facility
evaluation by response time of 10-20min , <20min and > 20min. Hay et al., (2019) , the major
objective of the research were finding the best route by considering traffic data for Mandalay,
Ghana, develop a system in case of emergency response to identify the best routes from fire

MSc Thesis Page 66

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

incident to the nearest place such as emergency services, hospitals and police stations and 5-
minute response time were used to find the closest facilities.

Forkuo and Quaye-ballard (2013) present to establish a GIS based fire emergency response
services where Ghana National Fire Service can identify the optimal route from its location to
any fire incident, its the optimal route model were developed and utilized by ArcGIS Network
Analysis based on time of travel, the distance of travel, the slope of the roads and the delays in
travel times and also used a system which comprises geodatabase of fire hydrate and queries to
locate fire incident to the particular service area.

Kc et al. (2018), the study employed the location model i.e. maximum coverage location model
to examine the spatial coverage of current and the future proposed new fire stations aligned
with population growth , estimated by 14-minute response time based on local context standard
in Brisbane, Australia. The researcher had chosen a new fire station location based on the input
data of fire incidents, the road network, and areas of significant anticipated population. Davoodi
(2019) , the researcher develops a hybrid model based on network analysis and set covering
location problem for the site selection of new fire stations in Tehran, Iran. The road network data
of the city had used to determine the service coverage of existing fire stations by 3-minute
standard response time. Even though all the above studies are similar in some points to the
present study. Meanwhile the present study is unique in terms of selecting holistic features as
well as taking various steps to solve the problem of fire stations deficiency in the study area by
combining advanced network analysis and analytical hierarchy processes with ArcGIS.

MSc Thesis Page 67

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

CHAPTER SIX
CONCLUSION AND RECOMMENDATIONS

6.1 Conclusion

In this study, GIS-based network analysis was implemented and applied to the Addis ababa road
network. It focuses on finding the optimal route between two locations on the road network and
finding the nearest fire stations service providers to an incident location based on local context
deriving time response. Furthermore, the proposed method integrates the collected fire incidents
data of 2020/21 to be used in the analysis, which in turn to produces more accurate results that
are suitable for realistic road networks. The Dijkstra algorithm used in this study by using the
functions of ArcGIS software’s Network Analyst extension is best method for network analysis ,
especially in the crowded city such as Addis ababa. In connection with this, from the result in the
best route analysis it can be concluded that it emphasizes on selecting the best route direction
between two given locations on the road network data and their total distance is 8593.8m with
time taken of 6minutes. On the other hand, from result closest facilities it can be concluded that
some of the fire incident locations that occurred in the study are outside of the predefined
response time of 8 minutes which implies that those incidents point in which the fire fighter
vehicle not reaching the incident location within cutoff travel time during emergency time.
Associated with this, the service coverage of the study area was examined using ArcGIS network
analyst tools within in 8-minute response time. From the result it was concluded that 276.10km2
which corresponds 53.15 percent of area is served whereas the remaining 243.358 km2 which
corresponds 46.85 percent of the city are unserved area.

Therefore, in order reduce spatial disparity distribution of fire stations for the city, proposing a
new fire station is extremely important. To solve this problem this study applies multi criteria
decision analysis such as AHP method which identifies 35 candidate sites using six factors such
inaccessible area, land use, population density, distance to existing fire stations, proximity to
main roads , and incident density rate. Among of them the more influenced criteria are
inaccessible area, population density and land use with associated weights of 36%, 24% and 17%
respectively. A number of candidate sites were introduced as the best location for establishment
of a new fire station to the study area. Using location allocation particularly maximize coverage
has chosen 10 new optimum location of fire stations to the study area. This study concluded that

MSc Thesis Page 68

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

a total of twenty fire stations ( i.e., the existing fire stations together with the suggested new fire
stations) are more than enough to cover the entire city within 8-minute.

Finally , the present study suggests 10 new fire stations which helps to maximize the coverage
of the city. After the suggested 10 new optimum fire station together with existing fire stations ,
the current service coverage level of the study area has increased into 87.44%. The ground truth
of those locations was taken from Google Earth search engine.

6.2 Recommendations
Based on the finding of the study. This study has made the following recommendations for the
study area.

• This study highly recommended that collecting and storing the spatial location of fire
incidents at the time of emergency occurrences using handheld GPS is very important for
the further interpretation and for researchers to do their study easily.
• Addis Ababa city fire and emergency prevention and rescue agency should increase as
much as possible the number of fire stations more than the existing and including the
proposed fire stations in this study in order to rescue long term sustainable and viabilities
property .
• This study recommended that the future development of the city and the population
growth has to be considered during the development of new fire stations. Since its
obvious that increasing population will leads fire incident occurrence.
• Addis Ababa city fire and emergency prevention and rescue agency should consider the
development of a GIS database for fire stations to capture all the necessary information
during emergency situations.

6.2.1 Suggestion for further research

For the possible future research. This study suggests the following recommendation for further
study.

✓ The development Web GIS based fire emergency response system database to locate fire
incident dynamically shall be investigated which provide quick and easy access.
✓ Taking into account the fire hydrate location will helps to achieve better results for
proposing new fire stations.

MSc Thesis Page 69

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

References
Abnet, G. B., Dawit, B. E., Imam, M. H., Tsion, L. M., & Yonas, A. S. (2017). City profile ADDIS
ABABA. In Ethiopian Institute of Architecture Building Construction and City Development..
https://doi.org/10.1016/S0264-2751(03)00056-8
Abstract, A. A. (2014). Application of the Analytic Hierarchy Process (AHP) for locating fire stations:
Case Study Maku City. Merit Research Journal of Art, Social Science and Humanities, 2(1), 1–010.
http://www.meritresearchjournals.org/assh/index.htm
Abudeif, A. M., Abdel Moneim, A. A., & Farrag, A. F. (2015). Multicriteria decision analysis based on
analytic hierarchy process in GIS environment for siting nuclear power plant in Egypt. Annals of
Nuclear Energy, 75, 682–692. https://doi.org/10.1016/J.ANUCENE.2014.09.024
Addis Ababa City Planning Project Office. (2017). Addis Ababa City Structure Plan Summary Report.
December, 1–234.
Akinci, H., Özalp, A. Y., & Turgut, B. (2013). Agricultural land use suitability analysis using GIS and
AHP technique. Computers and Electronics in Agriculture, 97, 71–82.
https://doi.org/10.1016/J.COMPAG.2013.07.006
Algharib, S. M. (2011). DISTANCE AND COVERAGE: AN ASSESSMENT OF LOCATION-
ALLOCATION MODELS FOR FIRE STATIONS IN KUWAIT CITY, KUWAIT (Vol. 16, Issue 22).
Kent State University.
Algharib, S. M. (2018). Distance and Coverage: an Assessment of Location-Allocation Models for Fire
Stations in Kuwait City, Kuwait. International Journal of Geological and Environmental
Engineering, 12(8), 1035–1041.
Anna-hajnalKa, K. (2018). Road network analysis using gis techniques in the interest of finding the
optimal routes for emergency situatons. case study: CLUJ-NAPOCA (ROMANIA). 1, 57–70.
ARIFIN, M. S. (2011). Location allocation problem using genetic algorithm and simulated annealing: a
case study based on school IN ENSCHEDE.
Badri, M. A., Mortagy, A. K., & Ali, C. (1998). A multi-objective model for locating fire stations.
Baloyi, E. (2017). Evaluating the location of fire stations using Geographic Information Systems GIS-
based accessibility analysis: a case study in Ekurhuleni Name: (Vol. 0002, Issue August).
JOHANNESBEURG.
Bhagat, A., & Sharma, N. (2013a). GIS-Based Application for Emergency Preparedness and
Management Accelerating Response System through GIS. http://en.wikipedia.org/wiki/Delhi
Bhagat, A., & Sharma, N. (2013b). GIS-Based Application for Emergency Preparedness and
Management Accelerating Response System through GIS Abstract : About the Author : 1–7.
Boakye, K. A. (2017). Boakye, Kwadwo Adu. Emergency Fire Response in Ghana: The Case of Fire
Stations in. university of north Texas.
Bolouri, S., Vafaeinejad, A., Alesheikh, A. A., & Aghamohammadi, H. (2018). The ordered capacitated
multi-objective location-Allocation problem for fire stations using spatial optimization. ISPRS
International Journal of Geo-Information, 7(2). https://doi.org/10.3390/ijgi7020044
Bolouri, S., Vafaeinejad, A., Alesheikh, A., & Aghamohammadi, H. (2020). Minimizing response time to
accidents in big cities: a two ranked level model for allocating fire stations. Arabian Journal of

MSc Thesis Page 70

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Geosciences, 13(16). https://doi.org/10.1007/s12517-020-05728-6

Brushlinsky, N., Ahrens, M., Sokolov, S., & Wagner, P. (2017). World fire statistics. Center of Fire
Statistics22. www.academygps.ru
Chaudhary, P., Chhetri, S. K., Joshi, K. M., Shrestha, B. M., & Kayastha, P. (2016). Application of an
Analytic Hierarchy Process (AHP) in the GIS interface for suitable fire site selection: A case study
from Kathmandu Metropolitan City, Nepal. Socio-Economic Planning Sciences, 53, 60–71.
https://doi.org/10.1016/j.seps.2015.10.001
Chevalier, P., Thomas, I., Geraets, D., Goetghebeur, E., Janssens, O., Peeters, D., & Plastria, F. (2012).
Locating fire stations: An integrated approach for Belgium. Socio-Economic Planning Sciences,
46(2), 173–182. https://doi.org/10.1016/J.SEPS.2012.02.003
Chhetri, S. K., & Kayastha, P. (2015). Manifestation of an analytic hierarchy process (AHP) model on
fire potential zonation mapping in Kathmandu metropolitan City, Nepal. ISPRS International
Journal of Geo-Information, 4(1), 400–417. https://doi.org/10.3390/ijgi4010400
Chiedu, C. (2011). The Use of GIS for Effective and Efficient Emergency Management: A case study of
the Cardiff Fire Services in South Wales. (Issue August).
Chinamatira, L., Mtetwa, S., & Nyamadzawo, G. (2016). Causes of wildland fires, associated socio-
economic impacts and challenges with policing, in Chakari resettlement area, Kadoma, Zimbabwe.
Fire Science Reviews, 5(1), 1–11. https://doi.org/10.1186/s40038-016-0010-5
Church, R. (1910). The Maximal Coverage Location Problem. 101–118.
Corcoran, J., Higgs, G., Brunsdon, C., Ware, A., & Norman, P. (2007). The use of spatial analytical
techniques to explore patterns of fire incidence: A South Wales case study. Computers, Environment
and Urban Systems, 31(6), 623–647. https://doi.org/10.1016/j.compenvurbsys.2007.01.002
CSA. (2020). Central Statistical Agency.
Dabhade, A., Kale, K., & Gedam, Y. (2015). Network Analysis for Finding Shortest Path in Hospital
Information System. Undefined.
Davoodi, M. (2019). A GIS based Fire Station Site Selection using Network Analysis and Set Covering
Location Problem (Case study: Tehran, Iran). December 2018, 433–436.
https://doi.org/10.6084/m9.figshare.10053329.v1
Derekenaris, G., Garofalakis, J., Makris, C., Prentzas, J., Sioutas, S., & Tsakalidis, A. (2000). Information
system for the effective management of ambulances. Proceedings of the IEEE Symposium on
Computer-Based Medical Systems, 269–273. https://doi.org/10.1109/cbms.2000.856910
Drake, A. W., Ralph, L. K., & Philip, M. M. (1972). Analysis of Public Systems. 64–74.
Drezner, Z. (1995). Facility Location: A Survey of Applications and Methods. Springer Verlag.
https://doi.org/10.1007/978-1-4612-5355-6
Edupta, B. . (2011). Fire Prevention as an aspect of environmental Management in Nigeria. Tropical Built
Environment Journal, 1(4), 141–148.
El, S., & Badri, A. (2016). Development of a GIS-Based Monitoring System for Road Network. April
2012.
Eleiche, M. (2011). Network Analysis Methods for Mobile GIS. Ilex.Efe.Hu, University of West Hungary,
2–16. http://ilex.efe.hu/PhD/emk/mohamedanwar/angol.pdf

MSc Thesis Page 71

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Erden, T., & Coşkun, M. Z. (2010). Multi-criteria site selection for fire services: The interaction with
analytic hierarchy process and geographic information systems. Natural Hazards and Earth System
Science, 10(10), 2127–2134. https://doi.org/10.5194/nhess-10-2127-2010
ESRI. (n.d.). Arc GIS, version 10X help. Retrieved October 29, 2021, from
https://desktop.arcgis.com/en/arcmap/latest/extensions/network-analyst/location-allocation.htm
ESRI. (2006). GIS Technology and Applications for the Fire Service. 1–28. www.esri.com
ESRI. (2007). GIS for Fire Station Locations and Response Protocol GIS for Fire Station Locations and
Response Protocol An ESRI White Paper. J, January.
ESRI. (2010). Location -allocation analysis( In ArcGIS Desktop 10 Helped).
Estoque, R. C. (2011). GIS-based Multi-Criteria Decision Analysis (in Natural Resource Management).
https://doi.org/10.4324/9781315235189
Flynn, J. D. (2009). Fire Service Performance Measures. National Fire Protection Association Fire
Analysis and Research Division, November, 1–43. www.nfpa.org
Forkuo, E. K., & Quaye-ballard, J. a. (2013). GIS Based Fire Emergency Response System. International
Journal of Remote Sensing and GIS, 2(1), 32–40.
Ghana lost GHC86 million to fires between Jan-Sept.2016. (2016).
https://www.ghanaweb.com/GhanaHomePage/business/Ghana-lost-GHC86-million-to-fires-
between-Jan-Sept-2016-496679
Green, C., Mans, G., Schmitz, P., Mckelly, D., & Water, M. (2014). Planning for emergency services
using GIS-based geographic accessibility analysis. Town and Regional Planning, 64(0), 53–64.
Gubara, A., Amasha, A., Ahmed, Z., El Ghazali 4 Global, S., & Cairo, G. (2014). DECISION SUPPORT
SYSTEM NETWORK ANALYSIS FOR EMERGENCY APPLICATIONS.
Habibi, K. S.Lotifi, M. . K. (2008). Spatial Analysis of Urban Fire Station Location by Integerating AHP
Model and IO Logic Using GIS (A Case Study of Zone 6 of Tehran ). Journal of Applied Sciences 8
(9), 3302–3315.
Habibi, K., Lotfi, S., & Koohsari, M. J. (2008a). Spatial analysis of urban fire station locations by
integrating AHP model and IO logic using GIS (A case study of zone 6 of Tehran). Journal of
Applied Sciences, 8(19), 3302–3315. https://doi.org/10.3923/JAS.2008.3302.3315
Habibi, K., Lotfi, S., & Koohsari, M. J. (2008b). Spatial analysis of urban fire station locations by
integrating AHP model and IO logic using GIS (A case study of zone 6 of Tehran). Journal of
Applied Sciences, 8(19), 3302–3315. https://doi.org/10.3923/JAS.2008.3302.3315
Hakimi, S. L. (1964). Optimum Locations of Switching Centers and the Absolute Centers and Medians of
a Graph. Operations Research, 12(3), 450–459. https://doi.org/10.1287/OPRE.12.3.450
Haque, Md, Q. F. (2001). Urban fire hazard modeling in Dhaka city using GIS.
Hay, S., Shwe, M., & Oo, H. N. (2019). GIS Based Fire Emergency Response System for Mandalay.
International Journal of Open Information Technologies, 7(12), 39–45.
Hoyos, M. C., Morales, R. S., & Akhavan-Tabatabaei, R. (2015). OR models with stochastic components
in disaster operations management: A literature survey. Computers and Industrial Engineering, 82,
183–197. https://doi.org/10.1016/j.cie.2014.11.025

MSc Thesis Page 72

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Ibrahim, A. (2016). Trade Fair gas explosion: 6 burnt to death, 12 in critical condition -.
https://www.myjoyonline.com/trade-fair-gas-explosion-6-burnt-to-death-12-in-critical-
condition/?param=
Idhoko, K. E., Aguba, W., Emefeke, U., & Nwanguma, C. (2015). Development of a Geographic
Information Systems Road Network Database for Emergency Response ; A Case Study Of Oyo-
Town, Oyo State, Nigeria. International Journal of Engineering Science Invention, 4(12), 34–42.
www.ijesi.org
Jasso, H., Hodgkiss, W., Baru, C., Fountain, T., Reich, D., & Warner, K. (2009). Using 9-1-1 call data
and the space-time permutation scan statistic for emergency event detection. Government
Information Quarterly, 26(2), 265–274. https://doi.org/10.1016/j.giq.2008.12.005
JOHN R. WATERS. (1999). FIRE DEPARTMENT RESPONSE TIMES VS. FLASHOVER - Fire
Engineering. https://www.fireengineering.com/leadership/fire-department-response-times-vs-
flashover/#gref
Kc, K., Corcoran, J., & Chhetri, P. (2018). Spatial optimisation of fi re service coverage : a case study of
Brisbane , Australia. Geographic Research, 270–284. https://doi.org/10.1111/1745-5871.12288
Khouri, R., Doytsher, Y., & Mclaren, R. (2018). Rapid Urbanization and Mega Cities:The Need for
Spatial Information Management Rapid Urbanization and Mega Cities : International Federation of
Surveyors, December, 11–16.
Kumar, P., & Kumar, D. (2016). Network Analysis using GIS Techniques: A Case of Chandigarh City.
International Journal of Science and Research (IJSR), 5(2), 409–411.
https://doi.org/10.21275/v5i2.nov161143
Liu, N., Huang, B., & Chandramouli, M. (2006). Optimal Siting of Fire Stations Using GIS and ANT
Algorithm. Journal of Computing in Civil Engineering, 20(5), 361–369.
https://doi.org/10.1061/(asce)0887-3801(2006)20:5(361)
Malczewski, J. (2004). GIS-Based Land-Use Suitability Analysis A Critical Overview. Progress in
Planning, 62.
https://www.scirp.org/(S(351jmbntvnsjt1aadkposzje))/reference/ReferencesPapers.aspx?ReferenceI
D=2122500
Maniruzaman, K., & Haque, Q. (2007). Fire Hazard in Dhaka City: A Case Study of the Service Area of
Mohammadpur Fire Station. Urbanization in BangladeshPatterns, ….
http://www.bip.org.bd/SharingFiles/journal_book/20130718121437.pdf
MartinLandré. (2008). GIS in Response Time Analysis. https://www.gim-
international.com/content/article/gis-in-response-time-analysis
Meyer, E. (2011). Performing Location Allocation Measures with a GIS for Fire Stations in Toledo, Ohio
(Issue December).
Mezaal, M. R., Shaker, A., & Noori, A. M. (2020). Response Time Analysis of Firefighting Operations
Using ArcGIS Network Analyst in Puchong Fire Station Malaysia. Solid State Technology,
63(January 2021), 18932–18945.
Nagib, G., & Ali, W. G. (2010). Network routing protocol using Genetic Algorithms. International
Journal of Electrical & Computer Sciences IJECS-IJENS, 10(02), 40–44.
NDRMC. (2016). National Disaster Risk Management Commission (NDRMC), Early Warning and
Response Directorate: Early Warning and Response Analysis - April 2016.

MSc Thesis Page 73

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

https://reliefweb.int/report/ethiopia/national-disaster-risk-management-commission-ndrmc-early-
warning-and-response
NFPA. (1710). Standard for the Organization and Deployment of Fire Suppression Operations,
Emergency Medical Operations, and Special Operations to the Public by Career Fire Departments.
https://catalog.nfpa.org/NFPA-1710-Standard-for-the-Organization-and-Deployment-of-Fire-
Suppression-Operations-Emergency-Medical-Operations-and-Special-Operations-to-the-Public-by-
Career-Fire-Departments-P1441.aspx
Nguyen A. (2006). University Campus Emergency Evacuation Plan. CP255: Advance GIS. University of
California, Berkeley. http://www.anhnses.com/uploads/Projects/ UCB_Evac_WP.pdf. (June
%0A10, 2010).
Nisanci, R. (2010). GIS based fire analysis and production of fire-risk maps: The Trabzon experience.
Scientific Research and Essays, 5(9), 970–977. https://doi.org/10.5897/SRE.9000531
Nyimbili, P. H., & Erden, T. (2020). GIS-based fuzzy multi-criteria approach for optimal site selection of
fire stations in Istanbul, Turkey. Socio-Economic Planning Sciences, 71(July 2019), 100860.
https://doi.org/10.1016/j.seps.2020.100860
Oh, J.-Y., Hessami, A., & Yang, H. J. (2019). Minimizing Response Time with Optimal Fire Station
Allocation. Studies in Engineering and Technology, 6(1), 47. https://doi.org/10.11114/set.v6i1.4187
Osuteye, E. ., Johnson, C., & Brown, D. . (2017). The data gap: An analysis of data availability on
disaster losses in sub-Saharan African cities. International Journal of Disaster Risk Reduction,
26(September), 24–33. https://doi.org/10.1016/j.ijdrr.2017.09.026
Potts, D. (2012). Whatever Happend to African’s Rapid Urbanisation ? American Journal of Obstetrics
and Gynecology, 7(5), 1–16. https://doi.org/10.1016/S0002-9378(24)90028-7
Rahman, S. U., & Smith, D. K. (2000). Use of location-allocation models in health service development
planning in developing nations. European Journal of Operational Research, 123(3), 437–452.
https://doi.org/10.1016/S0377-2217(99)00289-1
Revelle, C., & Hogan, K. (1988). A reliability-constrained siting model with local estimates of busy
fractions. Environment & Planning B: Planning & Design, 15(2), 143–152.
https://doi.org/10.1068/b150143
Reveshti, M. A., Ahmadi, T., & Teymouri, R. (2015). 5 Identifying Areas outside the Response Range of
Fire Stations Services Using Network Analysis (Case Study: Fire Stations in Tabriz). 1(2), 2014–
2016.
Saaty, T. L. (1977). A scaling method for priorities in hierarchical structures. Journal of Mathematical
Psychology, 15(3), 234–281. https://doi.org/10.1016/0022-2496(77)90033-5
Saaty, T. L. . (2000). Fundamentals of Decision Making and Priority Theory With the AnalyticHierarchy
Process.
https://books.google.com.et/books?hl=en&lr=&id=wct10TlbbIUC&oi=fnd&pg=PT1&dq=Saaty,+T.
L.+The+Fundamentals+of+Decision+Making+and+Priority+Theory+with+the+Analytic+Hierarchy
+Process,+2nd+ed.%3B+RWS+Publications:+Pittsburgh,+PA,+USA,+2000%3B+Volume+VI,+p.+
478&ots=_D3tWY1GDb&sig=glK7eMC0gZL2czxPBI44Uc2s_is&redir_esc=y#v=onepage&q&f=f
alse
Sarker, M. Ghosh, S. (2021). RESPONSE TIME ANALYSIS OF FIRE SERVICE IN GAZIPUR
INDUSTRIAL AREA : RESPONSE TIME ANALYSIS OF FIRE SERVICE IN GAZIPUR

MSc Thesis Page 74

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

INDUSTRIAL AREA : AN APPLICATION OF GIS. May.

Sayed, A., Romani, F. I., & Hesham, A. H. (2017). GIS-Based Network Analysis for the Roads Network
of the Greater Cairo Area. Proceedings of the International Conference on Applied Research in
Computer Science and Engineering ICAR’17, Lebanon.
Şen, A., Önden, I., Gökgöz, T., & Şen, C. (2011). A GIS approach to fire station location selection.
Gi4DM 2011 - GeoInformation for Disaster Management, January 2015.
https://doi.org/10.13140/2.1.2568.4804
Shokouhi, M., Nasiriani, K., Cheraghi, Z., Ardalan, A., Khankeh, H., & Fallahzadeh, H. (2019).
Preventive measures for fire-related injuries and their risk factors in residential buildings: a
systematic review. Journal of Injury and Violence Research, 11(1), 1–14.
https://doi.org/10.5249/jivr.v11i1.1057
Subramaniam, C., & Ali, H. (2012). Initial emergency response performance of fire fighters in Malaysia.
January. https://doi.org/10.1108/09513551211200294
Sui Min, W. W. (2012). Application of GIS best path algorithm in Harbin Roads. World Rural
Observations 4(1), 86–90.
Tali, J. A., Divya, S., & Nusrath, A. (2017). Location–Allocation Model Applied to Urban Public
Services: Spatial Analysis of Fire Stations in Mysore Urban Area Karnataka, India. International
Journal of Advanced Research and Development, 35(2), 201–207.
https://doi.org/10.22146/IJG.25365
Tishi, T. R. (2015). a Study on Frequency of Fire Incidents and Fire Fighting Capacity in Different Land
Use Categories of Dhaka Metropolitan Area.
TRANSPORT POLICY OF ADDIS ABABA. (2011). Transport Policy of Addis Ababa. August, 1–47.
https://chilot.me/wp-content/uploads/2011/08/transport-policy-of-addis-ababa.pdf
Uddin, M. S., & Warnitchai, P. (2020). Decision support for infrastructure planning: a comprehensive
location–allocation model for fire station in complex urban system. Natural Hazards, 102(3), 1475–
1496. https://doi.org/10.1007/s11069-020-03981-2
UN DESA. (2014). World’s population increasingly urban with more than half living in urban areas.
https://www.un.org/en/development/desa/news/population/world-urbanization-prospects-2014.html
Wang, J. H., Sun, J. H., Lo, S. M., Gao, L. J., & Yuen, R. K. K. (2011). Statistical analysis on the
temporal-spatial characteristics of urban fires under typical urbanization features. Procedia
Engineering, 11, 437–444. https://doi.org/10.1016/j.proeng.2011.04.680
Wei, L., Li, H. L., Liu, Q., Chen, J. Y., & Cui, Y. J. (2011). Study and implementation of fire sites
planning based on GIS and AHP. Procedia Engineering, 11, 486–495.
https://doi.org/10.1016/j.proeng.2011.04.687
Yang, B., Viswanathan, K., Lertworawanich, P., & Kumar, S. (2004). Fire Station Districting Using
Simulation: Case Study in Centre Region, Pennsylvania. Journal of Urban Planning and
Development, 130(3), 117–124. https://doi.org/10.1061/(asce)0733-9488(2004)130:3(117)
Yang, J., & Lee, H. (1997). An AHP decision model for facility location selection. Facilities, 15(9/10),
241–254. https://doi.org/10.1108/02632779710178785
Yao, J., Zhang, X., & Murray, A. T. (2019). Location optimization of urban fire stations: Access and
service coverage. Computers, Environment and Urban Systems, 73(November 2017), 184–190.

MSc Thesis Page 75

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

https://doi.org/10.1016/j.compenvurbsys.2018.10.006
Yetmgeta, T. (2013). GIS AS A TOOL FOR FIRE RISK PREVENTION DATA ANALYZING AND
MAPPING IN THE CASE OF ADDIS ABABA. ADDIS ABABA UNIVERSITY.
Zhang, F., Qiu, A., & Li, Q. (2009). Improve on dijkstra shortest path algorithm for huge data. 313–316.
http://www.isprs.org/proceedings/XXXVIII/7-C4/313_GSEM2009.pdf
Zhang, G., Lee, A. H., Lee, H. C., & Clinton, M. (2006). Fire safety among the elderly in Western
Australia. Fire Safety Journal, 41(1), 57–61. https://doi.org/10.1016/j.firesaf.2005.08.003
Zhang, J. (2005). Optimal Path Queries in Very Large Spatial Databases. Waterloo, Onterio,Canada.

MSc Thesis Page 76

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

Appendices
Appendix 1 :- Existing Fire stations Geographic Coordinate points
S.no Fire Stations Name Easting Northing
1 Addis Ketema Sub City Fire Station 9.033437 38.724256
2 Nefas Silk Lafto Sub City Fire Station 8.948606 38.7649
3 Kirkos Sub City Fire Station 8.986628 38.749243
4 Akaki Kaliti Fire Station 8.871546 38.788999
5 Bethel Fire Station 9.006456 38.681882
6 Tigist Fire Fighting Equipment 8.973223 38.760958
7 Addis Ababa Fire Stations Head Quarters 9.037573 38.75321
8 Gulele Sub city Fire Station(Fire Accident) 9.058938 38.712672
9 Addis Ababa Fire and Emergency Management Authority 9.05759 38.713312
10 Arada Fire Station 9.037687 38.753776

Appendix 2 : 100 Sample Fire Incidents Coordinate points out of 274 in 2020/2021

No. Sub City The Special Place Easting Northing

1 A/ketema Amanuel Church 469706.26 997721.91
2 Bus ordinary 470062.27 998513.24
3 Millennium Health Center 470293.79 998082.71
4 Bus ordinary 470872.63 998809.35
5 Abebe Bakele Stadium 469728.38 999585.27
6 Kolfe Electric Power 469462.68 998432.16
7 Eshet school 469574.12 999400.54
8 South Mercato Hotel 471510.95 998190.38
9 Melata Meda 468522.4 1000276.15
10 Mesalemiya 470397.68 998163.84
11 AK_kality Chefe Condominum 475951.21 982723.8
12 sertie Mariyam No.1 476234.83 983494.7
13 Saris Health Center 473846.79 988838.17
14 Koye Nefas silek 479180.29 983871.79
15 koye feche medhaniyalem 482367.11 993331.93
16 Abo Mazoria Fafa Factory 474704.64 989214.95
17 Cheralia Mosque 984872.34 984872.34
18 Federal Prisons 472517.99 985150.44
19 Gelan Condominium Old 476123.3 988824.37

MSc Thesis Page 77

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

20 Urban development 477086.35 979565.76

21 Arada St. Mary's Church Almi Building 474123.48 998156.9
22 Giorgis Church 472485.84 998999.6
23 Piazza Mall 473103.07 998478.22
24 Kurtu Building 471548.16 999227.93
25 degate wube 471176.83 999000.46
26 4Kilo Examinations Agency 473956.54 998920.06
27 Aware Women's Square 473486.33 999259.67
28 4Kilo Bash wolde 473627.19 998262.59
29 Tourist Hotel 998607.1 998607.55
30 Main Post Office 472855.13 997213.42
31 Bole Bora Children's Entertainment 479611.35 993747.03
32 Gerji Giorgis Church compound 475271.87 994705.09
33 behind Gerge kadisco hospital 479742.57 994311.98
34 No.49 square (medhaniylem shopes) 484213.49 996471.51
35 School of Tommorrow 476083.1 993272.05
36 Summit Condominium Block 2 Second Gate 487360.65 994464.15
37 Figa electric power 480790.91 994610.9
38 Figa Court 482194.47 995889.43
39 Gakrose Guna 480451.28 995914.84
40 Bole ayer abma school 475695.62 993116.16
41 Gulele Shiro Meda Welega sefer 473874.22 1001699.91
42 Retirement Bridge 470556.62 1002220.11
43 Gulele Police Department 471835.11 1000981.82
44 Shegole kidane mihurte 468772.26 1002076.74
45 Shiro Meda Qusquam Mariam Tabot Madari 473106.39 999945.38
46 kechene 8 No. Rotation 472809.36 1001091.96
47 Medre 7 469127.99 1001156.4
48 Menen Police Station 473329.01 1000757.81
49 Mulugeta Park 468322.06 1001347.54
50 Kechene Medhaniylem 472770.06 1000904.38
51 Kirkos Addis Ababa stadium Rotation 473213.78 996492.81
52 legahare Common Corporation 472559.43 996390.33
53 Denble back 474149.98 995173.54
54 Meskele Flower Ezra Hotel 474300.6 993716.17
55 Awash Bank National Theater 472852.5 996572.81
56 Sarbet Oromia compound 471899.24 994820.15
57 behind east powedr fabrica 474434.73 993565.1
58 kirkose legahare exit 472881.118 996111.454
59 nearest to arba miche fishe house 472837.962 994061.738
60 around leba fana school 474849.97 996875.32

MSc Thesis Page 78

 GIS Based Network Analysis for Locating Fire Stations and Emergency Response : A case of Addis Ababa City, Ethiopia

61 Lideta African Union Taboo Monopoly 471434.28 995055.74

62 Lideta Balcha 472282.59 996290.72
63 Police Hospital 472004.96 996290.59
64 TekeleHaymnote Hospital 471981.94 998100.2
65 Lideta Balcha 472149.48 996781.24
66 Police Hospital 472033.98 996566.99
67 around welda berka 467409.55 1000052.39
68 around gola micheal 471971.61 997829.35
69 40 kebelshe office 470535.98 996764.04
70 Gola micheal 472083.9 997946.38
71 N_silke Sosofia Mall Zeryakobe School 472048.48 991905.61
72 Sosofia Mall Zeryakobe School 472074.56 991766.08
73 Mekanisa 470837.52 992866.2
74 Bisrat Gabriel Home Depo 469760.96 993992.25
75 Chefe Mekanisa 470354.42 991842.04
76 Haile Garment 479580.68 991454.79
77 Jemo Michael 469580.68 991454.79
78 Bisrat Gabriel 468807.21 993030.82
79 Adey Ababa Zenbaba Hotel 474195.72 989911
80 Meri Ordinary 472668.43 992585.82
81 Yeka Megenagna Betelehem Building 477327.07 997281.14
82 Kotebe Emmanuel Hospital 483272.27 999569.04
83 Denbroa Hospital 477250.84 996928.76
84 Signal Adwa Bridge 476086.16 997695.05
85 Tiru Eder Temezha Yaze 479715.51 997557.34
86 Embassy of Israel 488838.87 997320.17
87 Were taboot dormitory 485292.76 999222
88 Yeka Michael Bano Sefer 477823.25 997987.18
89 Yeka Adado Woreda 13 483113.39 997574.14
90 Leme Hotel 476247.15 996582.31
91 Kolfe World Bank Square 464567.36 995003.22
92 Bethel Alpha School 466171.63 994941.51
93 Alpha University School 465223.58 994721.97
94 Kirkos Hotel Air Health 466618.09 993100.65
95 Debre Abay Arsema Church 465780.87 995107.01
96 shewa bread 467484.37 995797.61
97 China Bridge 464781.25 997106.75
98 Asco High School Campus 467427.69 1000192.29
99 Rep Medhaniylem 467156.69 996567.07
100 Kedinae Merhurt church 465506.92 994086.99

MSc Thesis Page 79