Assessing railway accident risk through
event tree analysis
Cite as: AIP Conference Proceedings 2138, 030023 (2019); https://doi.org/10.1063/1.5121060
Published Online: 21 August 2019
Nur Izyan Mustafa Khalid, Nur Farah Najeeha Najdi, Nurul Faziera Khairul Adlee, Masnita Misiran, and
Hasimah Sapiri
ARTICLES YOU MAY BE INTERESTED IN
Numerical approximation of centred fan field region for the indentation of granular material
by a smooth rigid wedge
AIP Conference Proceedings 2138, 030008 (2019); https://doi.org/10.1063/1.5121045
Mathematical solution for controlled drug release from a swelling hydrogel to the small
intestine
AIP Conference Proceedings 2138, 030037 (2019); https://doi.org/10.1063/1.5121074
Phase diagrams of the Potts model on the Cayley tree with competing binary interactions up
to the third nearest-neighbour generations
AIP Conference Proceedings 2138, 030011 (2019); https://doi.org/10.1063/1.5121048
AIP Conference Proceedings 2138, 030023 (2019); https://doi.org/10.1063/1.5121060 2138, 030023
© 2019 Author(s).
Assessing Railway Accident Risk through Event Tree
Analysis
Nur Izyan Mustafa Khalid1, a), Nur Farah Najeeha Najdi2, b), Nurul Faziera Khairul
Adlee3, c), Masnita Misiran4, d), Hasimah Sapiri5, e)
School of Quantitative Sciences, Universiti Utara Malaysia, 06010 UUM Sintok, Kedah, Malaysia
1,2,3,4,5
a)
Corresponding author: [email protected]
b)
[email protected]
c)
[email protected]
d)
[email protected]
e)
[email protected]
Abstract. Railway safety is an important issue since the safety of passengers, employees, road users and residents in the
area of railway can be affected when railway accidents happen. However in Malaysia, the risk management for railway
safety is still new, thus few studies available in the literature. In this study, measurement models involving descriptive
analysis and event tree analysis for Keretapi Tanah Melayu Berhad (KTMB)’s railway accident risk are developed by
considering the primary and secondary data from KTMB. The finding shows collective risk of 0.2406, in which a strong
0.217 is influenced by human error of not following the predetermined standard operating procedures, and only 0.0232 of
this risk is caused by the systematic failure. The determinant factors of such accidents include carelessness, conductor’s
effect, mechanical failure, high speed, human negligent, and track’s obstruction.
INTRODUCTION
Although rail transport is regarded as one of the safest transport modes compare to other type of transport, there
are still possibilities that accidents can occur. Rail safety management needs to be enhanced further. The risk in
railway sector can be described as occurrences or events that can cause catastrophe to others be it staffs, officers,
passers-by or passengers, and also the stability of transportation itself [1]. Proper railway maintenance and their
safety measures should be put in place [2]. This is to ensure that the train moves safely and economically. [3]
highlighted inverse proportional relationship between the numbers of accidents versus the level of safety. In the
current years, the rate of train derailment is reported to be quite high in the world, where in freight railway transport
industry, derailment is considered as a common accident that frequently happen ([4], [5]). The train speed, track
class, traffic density in certain stations and carriages and also the method of operation in management of the railway
organization are among factors that contribute to the accident risk in railway [6]. In the field of management and
engineering, risk-based safety management is a progressive theory. Risk-based safety management can secure the
safety of railway system efficiently [7].
In Malaysia landscape, railway safety issue has encountered breakneck changes for instance in structural reform
and interoperability circumstances in prevalent railway system [8]. In Malaysia, Keretapi Tanah Melayu Berhad
(KTMB) or Malayan Railways Limited is the largest railway company in Malaysia, with other minor railway
companies include Sabah State Railway and Express Rail Link. Currently, KTMB is targeting its risk-based safety
management to handle increased hazard factors on railway safety [9]. Conferring to the preceding Chief Executive
Officer (CEO) of KTMB on safety on the community’s railways, railway operators and infrastructure managers
The 4th Innovation and Analytics Conference & Exhibition (IACE 2019)
AIP Conf. Proc. 2138, 030023-1–030023-8; https://doi.org/10.1063/1.5121060
Published by AIP Publishing. 978-0-7354-1881-3/$30.00
030023-1
should get government’s authorization for their safety management plan, which is disposed by the risk evaluation
result. Consequently, Malaysian railway and federal research organizations begin to build up the general hazard
study and models of risk assessment for the systems of railway in Malaysia since 2007, though the stage is still in its
fundamental [10].
Bestowing via the accident categorization of national regulation, the accident scenarios in railway systems were
detached into five major scopes consist train concussion, level crossway, derailment of train, railway calamity
accident and also train fire [11]. As the number of passengers are significant, a better risk assessment management
should be done to reduce risk in this mode of transportation. TABLE 1 shows the number of passengers in KTMB’s
annual passenger traffic.
TABLE 1. KTMB’s Traffic Statistics, 2008-2017
Year Passenger
2008 36,557,000
2009 34,683,000
2010 34,995,000
2011 35,510,000
2012 34,847,000
2013 43,942,000
2014 46,957,000
2015 49,690,000
2016 41,469,000
2017 37,274,000
In present-day, the importance of railway safeness has been displayed in diverse analysis. For instance, in
Malaysia, heavy vehicles such as trucks crashing with trains and lead to disastrous destructions to the trains and the
people aboard. Several passengers also experienced minor injuries when ETS train heading towards Ipoh collided
with another train soon after leaving the station in Kuala Kubu Baru [12]. Other than that, one man was killed after
his car collided with a train at the rail crossing in Sabah [13]. Such horrific incidents signal the need of counter
measure of this problem. Among challenges of providing one-for-all solution is because of the current practice in
railway system in Malaysia, in which the safety precautions in railway transports falls upon individual jurisdictions.
Each jurisdiction is in charge of its own initiatives such as in enhancing the level crossing infrastructure and the
development of modelling techniques to pinpoint dangerous locations, and its preventive measures.
In this study, we outline three forms of advantages affiliation to financial and socio-economic perspectives of
assessing risk in railway management. First is to diminishing (if not entirely eradicating) collisions between trains
and any vehicles. Second is to reduce the deaths, injuries, property damage and human suffering correlated with
accidents. Third is that of curtailing the delays to both rail and road traffic as a consequence of dictated speed
stipulations on rail operations and of enormous barrier closure times against traffic. Based on the literature, we
highlight common influencers of risk assessments – collisions, trespasser, electrocuted and vandalism. We use event
tree analysis (ETA) to further investigate the level of collective risk in KTMB. ETA has been used to investigate the
railway system risk assessments worldwide ([14], [15], [16], [17]). ETA can be utilized at backside train collision
accident and models were created to figure the probability of rear end train collisions [18]. Other than that, ETA was
used to point out the risk of train crashes on city rail transit frameworks [19]. We use event tree analysis (ETA) to
make use of its advantage, where this analysis is able to present the events graphically [20]. Furthermore, this
analysis can help in identifying any countermeasures which are not effective [21]. In order to estimate risk of system
failure, comprehensive knowledge of the outcomes of each series of events is not essential [22], though we
acknowledge limitation of this analysis, where it can only analyse one incident (or event) at certain time [23].
METHODOLOGY
This study will make use on the primary and secondary data obtained from the Department of Transportation and
Keretapi Tanah Melayu Berhad (KTMB). The primary data were collected through an interview with a staff from
KTMB Arau, although the scope of secondary data collection covers data across the country. The data includes
030023-2
issues that are frequently occur, the contributing factors, and the impacts of railway incidents. Meanwhile, the
secondary data were collected from KTMB website, as more detailed data is nondisclosed to the public. The data
consisted of type of accidents with their frequencies involving railway transports in Malaysia. Data collected from
KTMB are coded for the purpose of analysis. The measurement model is estimated by using descriptive analysis and
event tree analysis (ETA). The data of the year 2016 and 2017 are analyzed. ETA is used to evaluate the risk on
railway accidents. To generate the models of KTMB railway risk assessment, numerous risk management steps are
examined, for instance the International Standard Organization (ISO) manual and policy of risk management that
will be adapted to this study. It is similar to the normal approach of risk management in Malaysia [24]. The scheme
in Figure 1 displays three aspects consist railway structure, system’s life process and the risk management’s
processes.
FIGURE 1. Risk Management Approach [25]
The risk management began from the structure design phase. They are the basic form for this system [26], and is
adaptable in various industry, such as in technology [27], finance [28], and manufacturing [29]. Accordingly, to the
organization participating in the construction phase, the outcomes of the risk control in the system for design phase
will be disseminated and as a loop, analogous method will be enforced to the subsequent phase. Figure 2 shows the
hazard analysis and procedure of risk assessment, in particular to the railway management.
FIGURE 2. The Procedure of Risk Assessment of Railway [25]
By assembling different accident reports, material and associating with railway safety experts, the hazard
identification of railway accidents will be executed. A hazardous event infers one that can possibly lead specifically
to damage or death. Before hazardous events, the accidents regarding railway appearance situations allude to the
event procedures of mischances. For recurrence assessment on railway accidents, the scenarios give the fundamental
of event tree analysis (ETA) model design.
030023-3
DATA ANALYSIS AND RESULT
Descriptive Analysis
This section describes the secondary data which displays the profiles of accident across the year 2016 and 2017
associating with KTMB railway.
TABLE 2. Statistics of Accident at KTMB Railway
Type of Accident 2016 2017
Collision With animals 79 22
With road vehicles 3 1
Trespasser Injured 4 0
Fatal 24 6
Electrocuted Injured 2 1
Vandalism Stone pelting - incident 23 11
Stone pelting – injured 2 0
Stealing communication cable 33 3
Sabotage / stealing track 2 1
component / rolling stock
component
TABLE 2 shows the accidents that occurred during the year 2016 and 2017 involving KTMB railway. The
accidents are divided into four types – collision, trespasser, electrocuted and also vandalism.
FIGURE 3. Type of Accidents in 2016 FIGURE 4. Type of Accidents in 2017
Figure 3 – 4 showed the decrease in type of accidents that occurred between 2016 and 2017, with collisions (with
road vehicles and animals) consistently the highest in both years, with being electrocuted the least, signaling a
proper electrical management in placed.
Event Tree Analysis (ETA)
We follow [30] and construct ETA that can represent accident scenario in KTMB. The assessment is made by
considering the data in one month obtained from KTMB Arau.
030023-4
TABLE 3. Decision and Probability Table
Decision Frequency Total Possible Probability
Frequency
Workers respond to alarm 128 160 Failure = 0.2
Success = 0.8
Workers follow safety 154 160 Failure = 0.04
procedure Success = 0.96
Automatic Shutdown 8 160 Failure = 0.05
Begins Success = 0.95
FIGURE 5. The Event Tree Analysis for Railway System Breakdown
From the developed accident scenarios using ETA, the collective risk can be calculated. The risk is indicated as
the fatalities and injuries (FI) per year. The risk correlated with a specific hazardous event is computed as
CR = F×C,
where CR is the collective risk, F is frequency and C is consequences. Collective risk is calculated by dividing
the mean number of FI with year. Frequency is calculated by counting the mean frequency where incident occurs
while consequence is calculated by dividing the event by the number of FI. Based on Figure 5, we can see that the
collective risk for the system to fail is 0.2406, with a strong 0.217 (0.017 + 0.2 from Figure 5) is based on human
error (decision not to follow the determined SOP), and only 0.0232 is based on systematic failure. Referring to the
ETA, a conclusion has been made that human’s choices contribute massively towards safety precautions in the
workplace. In terms of safety results, the probability for success does obviously increases in the presence of safety-
minded employees.
Next, we highlighted some determinants that can cause railway accidents.
Determinants and Impacts of Railway Accidents
Railway accidents happen because of derailment, collisions with different trains and vehicles, and the obstruction
of observers or vehicles at train intersections. Table 4 represents some of the determinants involved, in particular in
KTMB Arau.
030023-5
TABLE 4. Table of Causes and Impacts of Railway Accidents in Malaysia
Causes Brief
• Carelessness is assessed in Arau Railway Station, railroad
representatives, Arau station administration, and equipment
manufacturers.
• Include faulty operation of an intersection arm or the failure of a
Carelessness signal light to give sufficient warning.
• On 22nd March 2018, KTMB was fined MYR 60,000 due to the poor
maintenance of the rail and its carriages. It results in the derailment of
the train.
• There are numerous occasions of out-dated railway safety innovation.
• Human error is dependably a potential factor in any accident, in
particular KTMB's maintenance workers or conductor.
Conductor’s Effect
(error, • Poor judgement and impeded responses have contributed such a large
inexperience, or number of avoidable train disasters.
fatigue)
• In KTMB case, no conductor error is presence as they use automated
system. Conductors only need to monitor train movements.
Mechanical Failure • Notwithstanding the KTMB railway’s earnest attempts to keep up the
infrastructure of the tracks, trains, and operating systems, mechanical
failure, however, can result in a train derailment or impact.
• The faster the train is moving, the more awful the consequences are if
it crashes. Crashes are more certain if the train is going at high speeds
and the seriousness of injuries are exponentially expanded.
High Speeds
• In Malaysia, KTMB has implemented the Global Tracking System
(GPS) at 120 locomotives in 2015. The control centre can observe and
control the train’s speed limit.
Human Negligent • In uncommon cases, a walker or vehicle might be on or close to the
(irresponsible or railway tracks at the wrong time, causing the KTMB conductor to
negligent respond all of a sudden or the train to derail.
pedestrian or motor
vehicle operator) • On January 2018, a man was almost dead when his car was crossed
and hovered to the commuter platform as he drove near the railway
track.
• At times, foreign objects, for example, vehicles or equipment, might
be left incidentally in the way of an approaching train. If the KTMB
conductor fails to see them or is going too fast, they will most likely
be unable to abstain from striking them, causing an accident.
Track’s • On 7th November 2018, the KTMB train service was disrupted due to
Obstruction fallen trees in Sungai Buloh, Selangor. The incident caused delayed in
train services for about 90 minutes and no injury or fatality reported.
• Most of the obstructions on the KTMB railway tracks are due to
natural disaster. This is because the tracks are mostly being neatly
controlled.
030023-6
CONCLUSIONS
In this study, we assess accident risk for KTMB railway system by using the event tree analysis. The finding
suggest that the success of the implemented standard of procedures (SOP) when the system showed breakdown is
0.73, given all tiers are being followed diligently. The collective risk posed in this system is 0.2406, where 0.217 is
based on human error, and only 0.0232 is based on systematic failure. In term of accident, zero risk is expected, thus
the need for more efforts to be made. This study also revealed challenges in standardize counter measure to combat
railway accident risks since the railways are under individual jurisdictions, where each stations are responsible each
to their railway tracks, thus dampening more thorough approach of proper risk management. We also highlighted
some determinants for the railway accidents that include carelessness, conductor’s effect, mechanical failure, high
speed, human negligent, and track obstructions. As managing accident risk is focusing on zero accident rate, the
performance of the current system needs to be improved. We recommend a standardization of railway maintenance
and management, and the training to human resources for the needs to follow predetermined SOP, in which a proper
risk management system and counter measure can become more effective.
REFERENCES
1. V. Profillidis, Railway Management and Engineering. Railway, (2016).
2. A. Jabbar, Lateral Resistance of Railway Track. Reliability and Safety in Railway, (2012)
3. P. Stankaitis and A. Iliasov, Theories, Techniques and Tools for Engineering Heterogeneous Railway
Networks. Reliability, Safety, and Security of Railway Systems. Modelling, Analysis, Verification, and
Certification Lecture Notes in Computer Science, (2017).
4. C.P.L. Barkan, C.T. Dick and R.T. Anderson, Transportation Research Record 1825, 64-74 (2013).
5. X. Liu, M. R. Saat and C.P.L. Barkan, Journal of Hazardous Materials 276, 442-451 (2014).
6. X. Liu, Optimal strategies to improve railroad train safety and reduce hazardous materials transportation risk,
(2013).
7. Y. Zhang, J. Guo, N. Xiao and Y. Zhao, Y., Risk-based safety management on railway system. Second
International Conference on Transportation Engineering, (2009).
8. M. Jacyna and J. Szkopiński, WIT Transactions on Ecology and the Environment 193, 399-409 (2015)
9. I. Loh, (2018). KTM probing train collision. <https://www.thestar.com.my/news/nation/2018/10/19/ktm-
probing-train-collision-were-looking-into-technical-issue-and-human-error-says-ceo/>
10. N.W. Chan, Impacts of disasters and disaster risk management in Malaysia: The case of floods. in Resilience
and Recovery in Asian Disasters, Springer, Tokyo, 2015, pp. 239-265.
11. A. Gupta, (2014). The world’s worst train disasters. <https://www.railway-technology.com/features/featurethe-
worlds-deadliest-train-accidents-4150911/>
12. Retrieved November 7, 2018, from The Straits Times: https://www.straitstimes.com/asia/se-asia/several-
passengers-suffer-injuries-after-2-trains-collide-in-malaysia
13. One killed in crash with train. (2018). Retrieved December 2, 2018, from The Star:
https://www.thestar.com.my/news/nation/2018/09/27/one-killed-in-crash-with-train/
14. Z. Peng, Y. Lu, A. Miller, C. Johnson and T. Zhao, Quality and Reliability Engineering International 32(1),
181-194 (2016).
15. B. Leitner, Procedia engineering 187, 150-159 (2017).
16. S. Dindar, S. Kaewunruen, and M. An, Journal of Risk Research 21(8), 974-995 (2018).
17. Q. Xu and K. Xu, Curr. Sci, 12, 2530-2538 (2018).
18. Y.F.J. Li, J. Mi, H. Z. Huang, S. P. Zhu, and N. Xiao, Maintenance and Reliability 15(4), 403-408 (2013)
19. M.Y. Wang, H. Wang, and Z.G. Liu, International Journal of Business and Social Science 5(8), 128-134
(2014).
20. Y. Xiaomin and T. Fulvio, T., Risk Analysis 32(2), 330-344 (2011).
21. F. Refaul, K. Faisal and S. Rehan, Risk Analysis 31(1), 86-107 (2010).
22. M.A. Zentner, Application of Event Tree/Fault Tree Modeling Approach to the Evaluation of Proliferation
Resistance. Volume 14: Safety Engineering, Risk Analysis and Reliability Methods, (2007).
23. M. Rausand, Risk Assessment: Theory, Methods, and Applications. John Wiley & Sons, (2013).
030023-7
24. Rail e-Services, (2014). <http://www.spad.gov.my/transport-operators/rail>
25. A. Berrado, E. El-Koursi, A. Cherkaoui, and M. Khaddour, M. Open Transportation Journal, 19p, (2010).
26. N. Bahr, System Safety Engineering and Risk Assessment: A Practical Approach (2nd ed.). CRC Press, 2014.
27. Y. Salim, T.E. Chang, S. Hamzah, M. Misiran and H. Sapiri, Journal of Advanced Research in Business and
Management Studies 8(1), 66-75 (2017).
28. K.C. Iyer and D. Purkayastha, Journal of Structured Finance 22(4), 17 (2017)
29. S. Albery, D. Borys, and S. Tepe, Safety science 89, 180-189 (2016).
30. B. Ahmed, M. Hainin, and I. Ahmed, Jurnal Teknologi 65(3), (2013).
030023-8
