Global Journal of Science Frontier Research: I

Interdisciplinary
Volume 23 Issue 1 Version 1.0 Year 2023
Type: Double Blind Peer Reviewed International Research Journal
Publisher: Global Journals
Online ISSN: 2249-4626 & Print ISSN: 0975-5896

By Razieh Khaksari & Zambri Harun

University of Malaysia
Abstract- One of the critical transportation infrastructures is road tunnels and fire safety is one of
the important aspects of their operation. Interaction between the fire, tunnel users, traffic, and fire
safety measures influences fires in road tunnels. Therefore, a complex model is required to
analyse the risk and quantify the consequences. In this paper, a novel quantitative risk analysis
model developed for UK road tunnels is presented consisting of a quantitative consequence
analysis model and a quantitative frequency frequency analysis model. The proposed
quantitative consequence analysis model is provided through three sub-models; queue model,
distribution model, and egress model. The frequency analysis is via an event tree that takes into
account the tunnel fire rate in UK road tunnels. After a brief description of this model, the
proposed method is illustrated through a case study of an urban road tunnel. The effect of
different emergency ventilation systems on societal risk and sensitivity of the model to pre-
movement time, accident frequencies involving Heavy Good Vehicles (HGVs), tenability threshold
temperature, and different burning vehicles were studied in this case study.
Keywords: fire safety, quantitative risk analysis, consequence analysis, frequency analysis.
GJSFR-I Classification: LCC: TE228.3

DevelopmentofaNovelQuantitativeRiskAssessmentToolforUKRoadTunnels
Strictly as per the compliance and regulations of:

© 2023. Razieh Khaksari & Zambri Harun. This research/review article is distributed under the terms of the Attribution-Non
Commercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0). You must give appropriate credit to authors and reference this
article if parts of the article are reproduced in any manner. Applicable licensing terms are at https://creative-
commons.org/licenses/by-nc-nd/4.0/.
 Razieh Khaksari α & Zambri Harun σ

Abstract- One of the critical transportation infrastructures is and fire in Yanhou in China (2014; 40 fatalities) have
road tunnels and fire safety is one of the important aspects of raised awareness of the safety issues of the tunnels,
their operation. Interaction between the fire, tunnel users, and that of the road tunnel users. Therefore, risk

2023
traffic, and fire safety measures influences fires in road analysis has become one of the essential tools to
tunnels. Therefore, a complex model is required to analyse the
improve and optimize the safety level of road tunnels.

Year
risk and quantify the consequences. In this paper, a novel
quantitative risk analysis model developed for UK road tunnels There are several methods of risk analysis to determine
is presented consisting of a quantitative consequence analysis risks and evaluate the effectiveness of safety measures.
291
model and a quantitative frequency frequency analysis model. One of these methods is quantitative risk analysis
The proposed quantitative consequence analysis model is (QRA), which has been one of the explicit requirements

Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

provided through three sub-models; queue model, distribution under the European Union (EU) Directive (2004/54/EC)
model, and egress model. The frequency analysis is via an [5]. QRA includes event tree analysis and consequence
event tree that takes into account the tunnel fire rate in UK analysis models. Various QRA models for fire in road
road tunnels. After a brief description of this model, the tunnels have been developed by different countries,
proposed method is illustrated through a case study of an
such as the TuRisMo model of Austria, the TUNPRIM
urban road tunnel. The effect of different emergency ventilation
systems on societal risk and sensitivity of the model to pre- model of the Netherlands, the Italian risk analysis model,
movement time, accident frequencies involving Heavy Good and the OECD/PIARC model [6]. Societal risk
Vehicles (HGVs), tenability threshold temperature, and represented graphically in the form of a frequency/
different burning vehicles were studied in this case study. number of fatalities (FN) curve, is the main output of
Keywords: fire safety, quantitative risk analysis, these models. The calculated societal risk must be
consequence analysis, frequency analysis. evaluated by comparison with the risk acceptance
criteria, i.e. must be less than a specified minimum and
I. Introduction acceptable value (threshold) as presented in the case

T
studies, for instance, Diamantidis, 2005 [7]; Botschek et
he number of road and rail tunnels constructed al., 2007 [8]; Kohl and Zibert, 2010 [9].
worldwide within the last 15 years has increased. In the UK, many of the 48 road tunnels have
This increase is due to the considerable rise in been constructed when transport conditions,
traffic volume. Not only has the number of tunnels knowledge, and technical possibilities were in their
increased, but the type and complexity of fire safety infancy and very different from the present. For instance,
systems used in them have also changed. Tunnels can current traffic volume and its composition are not the
relieve congestion in developed urban areas and reduce same as decades ago. There are still tunnels under
travel times in urban areas. construction and also innovations today that permit the
However, the growth of using tunnels results in use of safety systems and devices that were improbable
an endogenous problem, the severity of accidents that before.
may occur [1]. Although the accident rate in tunnels is Although the UK’s design manual BD78/99,
lower than on open roads as shown in [2], accidents which was published in August 1999 and then
can lead to severe consequences due to tunnel nature. superseded by CD 352 in March 2020, identifies risk
Fire is one of these disastrous consequences, which analysis as a practical method for determining the risk
can cause human losses and structural damage [3]. level of all road tunnels in the United Kingdom, it must
Much research about tunnel fire dynamics and be improved and the lack of a quantitative risk model, a
influencing parameters has been performed within the comprehensive system-based risk assessment tool to
last 20 years. [4] summarized the main research in that cover fires in road tunnels, is felt.
field. Fire in Mont Blanc in France (1999; 39 fatalities), In other industries, the risk assessment system
fire in Frejus in France (2005; 2 fatalities and 21 injuries), or software is widely used to be as a decision-support
tool [10,11]. Therefore, it is important to develop a
Author α: London Bridge Associates Ltd. Southwark Street, London SE1
robust and flexible QRA software tool that is suitable for
1TY, UK. e-mail: [email protected] road tunnels.
Author σ: Faculty of Engineering and Built Environment National As far as research work and further
University of Malaysia, UKM Bangi 43600, Malaysia. development of methods are concerned, efforts should

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

focus on developing a systematic approach to assessment tool to cover fires for road tunnels in the
performing a quantitative risk analysis (QRA) model for UK. Therefore, it was decided to develop the most
the UK road tunnels. Along with both legal regulations appropriate quantitative risk analysis model for safety in
and requirements defined in the UK, the risk analysis UK road tunnels, leading to improved and consistent
model can be an additional tool to estimate the level of decisions on the grounds of safety in the design and
safety that meets the respective requirements. operation of road tunnels, LBA Quantitative risk analysis
Consequently, a model of quantitative risk analysis model (LBAQRA).
concerning safety in road tunnels called LBAQRA has Due to the specific features of each tunnel, risks
been developed to perform a risk analysis as required must be evaluated considering the tunnel geometry and
by UK regulations. its infrastructures, fire safety measures, as well as
The LBAQRA performs a quantitative risk equipment and management procedures. In addition, it
assessment and includes quantitative frequency requires the analysis of many complex factors and
2023

analysis and quantitative consequence analysis. processes related to human behaviour, such as pre-
Year

This paper first presents a brief description of evacuation times, interactions between occupants,
the sections of this model and how it derives the societal interactions between occupants and smoke, etc. These
301 risk. In the second part, an illustrative case study was aforementioned parameters are taken into account in
carried out to study the sensitivity of this model to this model.
various risk indicators, and the risk reduction potential of This model consists of two main sections:
Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

different ventilation strategies. quantitative consequence analysis and quantitative

Risk assessment of fire incidents in road tunnels frequency analysis.
includes complicated scenarios as they constitute
a) Quantitative Consequence Analysis
interactions among the fire, tunnel users, and safety
The number of fatalities is calculated via
measures such as fire detection, tunnel alarm, or
quantitative consequence analysis which comprises
emergency ventilation. Thus, many risk indicators, which
three parts: queue model, distribution model, and
affect risks for tunnel users, describe the scenario. In
egress model.
order to evaluate the LBAQRA and included risk
The queue model calculates the length of the
indicators, Southwick Tunnel with 490 m length is
vehicle queue behind the fire by considering traffic
investigated. The tunnel is an urban tunnel with
density, percentage, and type of vehicles in each lane,
approximately 46900 vehicle/day traffic density in the
and a comparison between tunnel closure time and
Eastbound bore. A total of 94 fire scenarios is included
saturation time. Then the number of vehicles queueing
in this project. Under-study scenarios have different heat
in each lane and consequently the number of exposed
release rates (HRR), fire source locations, and traffic
tunnel users is estimated by queue length and traffic
conditions. The influences of different emergency
density.
ventilation strategies including various activation times,
ventilation velocities and operational strategies, and
detection times on societal risk are studied. In each
scenario, the possibility for safe egress is estimated by
3D fluid dynamics simulation via measuring tenability
thresholds through the evacuation path. The sensitivity
of the model to pre-movement time, accident
frequencies involving HGVs, tenability threshold
temperature, and different burning vehicles were studied
as well.

II. Methodology
The specific consequences of tunnel fires such
as oxygen concentration decrease, and toxic gases'
concentration increase make them one of the
catastrophic hazards in road tunnels. Consequently,
assessment of risk either quantitatively or qualitatively
should be considered an influential tool to reduce
such hazards. Although quantitative risk analysis
models have been developed in countries such as
Austria, Italy, and Singapore to improve the validity of
qualitative risk analysis, there is a lack of a quantitative
risk model, a comprehensive system-based risk

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

2023
Year
Figure 1: The Process Sequence of Calculating the Number of People in the Queue. 311
After calculating the number of potentially continue the evacuation process based on the

Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

exposed people, the distribution model is used to divide considered tenability thresholds. In this section, the
the queue into cells of the same size and then distribute evacuation time is compared with the time when the
tunnel users homogeneously. The total distance of the tenability thresholds exceed their limits, ASET.
evacuation path is calculated by the distribution model Three indicators are measured at humans’
taking into account both the longitudinal and lateral height at the centre of each cell to assess whether
shares in the evacuation route. Exposed tunnel users occupants can travel the evacuation path and reach a
distributed in each cell start the evacuation path from place of safety.
their initial position and they travel cell by cell towards a
The Tenability Criteria used for this Model are:
place of safety i.e., an emergency exit or the tunnel
portal. • Fractional Irritant Concentration (FIC)<1.
It should be mentioned that the structure of the • Fractional Effective Dose (FED) < 1.
queue formation model and the distribution model has • Gas temperature<80 °C.
been inspired by [12]. The total number of casualties is defined as the
The egress model is a four-section timeline sum of fatalities in each cell where the tenability
model. The first stage is detection which depends on thresholds are exceeded (RSET>ASET). The total
the safety equipment in the tunnel (𝑡𝑡𝑑𝑑𝑑𝑑𝑑𝑑 ). The second number of causalities in the whole tunnel is determined
stage is the alarm stage which is the time between by the sum of fatalities in each lane.
detection and the time when the alarm system is
activated (𝑡𝑡𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 ). The third stage is the pre-movement b) Quantitative Frequency Analysis
stage including recognition time, response time, and the The quantitative frequency analysis is an event
time to exit the vehicle (𝑡𝑡𝑝𝑝𝑝𝑝𝑝𝑝 ). The distance users have tree. The first column of the frequency event tree is initial
fire frequency which has been obtained from the
from the accident and the severity of the fire influence
historical statics of fire incidents in England road
the recognition time. The variation of response time is
tunnels. Then this initial fire rate is updated by
opposite to that of the recognition time as those who are
considering the length, traffic volume, and gradient of
actively evacuating influence those still in their car. The
the under-studied tunnel. Then the updated fire rate is
last stage is traveling, which depends on the movement
distributed across event tree branches including “time of
speed and the distance to the emergency exit (𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ).
the incidents”, “traffic condition”, “type of fire incidents”,
In this model, the relationship between walking speed
“vehicle type”, and “fire source location”.
and people density is used to estimate walking speed.
The results of quantitative consequence
The presence of other people can influence walking
analysis and quantitative frequency analysis are
speed during evacuation.
presented in terms of social risk, as an FN curve. Risk
The total time of these stages defines RSET
acceptance is obtained using the ALARP (As Low as
(required safe escape time). According to this
Reasonably Practicable) criterion in the UK.
description, RSET was calculated through Eq.(1) as
below: III. Case Study
𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 = 𝑡𝑡𝑑𝑑𝑑𝑑𝑑𝑑 + 𝑡𝑡𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝑡𝑡𝑝𝑝𝑝𝑝𝑝𝑝 + 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 (1) In order to achieve the purpose of this study
The next step is the verification of users’ egress mentioned in the introduction section, a specific tunnel
process to find out if evacuees can start and then was investigated. This tunnel, Southwick Tunnel, is

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

located on the A27 between the junction of the ventilation system strategy on the FN curve. According
Holmbush interchange (A27), portal A and the to the Regulatory Reform (Fire Safety) Order 2005, all
Hangleton interchange (A293), portal B, UK. tunnel operators in the UK are obliged to take
The structure is a twin-bore curved uni- precautions to ensure that the safety risk in the tunnel is
directional road tunnel. Each bore is approximately 490 as low as reasonably practicable (ALARP). The ALARP
meters long and carries two lanes of traffic. The bores limit, acceptable and unacceptable limits, was adopted
are connected by 3 cross passages, approximately by Moonis et al. (2008) [13]. In this study, three different
100m apart and 100m from either portal.This tunnel has traffic conditions including free flow, congested, and
a positive longitudinal slope of 3% from portal A to B. stoppage were considered. FN curve of scenarios
The tunnel equipped with a longitudinal without ventilation was located at the unacceptable
ventilation system consists of 14 jet fans (7 pairs) region which means it could not meet the safety target
installed at the ceiling of each bore. The system (Figure 2 (a)). The comparison between scenarios
2023

operates automatically to remove pollution from vehicle without ventilation (V= 0m/s) and other ventilation
Year

exhausts and also to help manage smoke in the event of conditions in Table 1 was presented in Figure 2 (b, c,
a fire. Monitoring systems are considered to consist of and d). According to FN curves, when the ventilation
321 the Tunnel Control Centre (TCC) with CCTV video system was activated at 4 min after ignition, there was
surveillance, air quality monitoring, and automatic fire no obvious difference between FN curves of V=0m/s
detection systems. and V=3.2 m/s with 4 min activation time (Figure 2 (b,
Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

This tunnel has an annual average daily traffic of c)). It means as the ventilation system was activated
about 46900 vehicles/day traffic density in the very late, it could not cope with the influences of fire and
Eastbound bore with an average percentage of 77% smoke. On the other hand, when the ventilation system
passenger car, 0.2% bus, 19% truck, and 3.8% HGV. was activated 2 min after ignition (Figure 2 (d)), the FN
Three different traffic conditions, free fluid, curve was lower than in scenarios without ventilation
congested, and stoppage, were considered. 158 MW, which showed the positive effect of the activation time of
47 MW, and 30 MW fire scenarios at three locations the ventilation system. Although it was still in the
including 0.3L, 0.5L, and 0.8L (L is tunnel length) are unacceptable region.
studied. Overall, 94 scenarios were simulated via Fire For scenarios with a 3.5 m/s ventilation velocity,
Dynamic Simulator (FDS) to analyse different jet fans were activated by the detection system when the
emergency ventilation systems for the aforementioned temperature was above 68 ˚C. The activation time was
fire loads. between 2.67 and 3.17 min for HGV fires and between 6
FN curves as the output of quantitative risk and 7.17 for bus fires. As the detection time of these
analysis and sensitivity analysis of this model taking into scenarios depended on the linear heat detection
account pre-movement time, accident frequency system, same as ventilation system activation, and this
involving HGVs, tenability threshold temperature, and time was the first stage of the evacuation timeline,
different types of vehicles were carried out for the people started to evacuate later than scenarios with 1
scenarios understudied. min detection time (Scenario 1-4 – Table 1). Although for
30 MW fire scenarios, the activation time was between 2
IV. Results and 3 min, the positive effect of higher ventilation
a) FN Curve velocity, 3.5 m/s, could make a more tenable situation
Five different ventilation conditions listed in during evacuation and alleviate the negative effect of
Table 1 in the case of 158 MW, 47 MW, and 30 MW fire late activation of jet fans (Figure 2 (e)).
scenarios were studied to investigate the effect of
Table 1: Under Study Ventilation Systems

Ventilation Ventilation activation time after

Scenario No. No. Fans
velocity (m/s) ignition
1 0 _ _
2 3.2 80% 4 min
3 3.2 100% 4 min
4 3.2 80% 2 min
5 3.5 80% Activation by detector at 68˚C

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

2023
a) V=0 b) Scenario 1 and 2

Year
1
33

Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

c) Scenario 1 and 3 d) Scenario 1 and 4

e) Scenario 1 and 5
Figure 2: FN Curve of Different Emergency Ventilation Systems
b) Sensitivity Analysis expected damage value, EDV, which is calculated via
Basic recognition and response times, and time Eq.(2), was calculated for each mode.
taken to leave the vehicle Sensitivity analysis was carried
out for 30 MW, 47 MW, and 158 MW fire scenarios 𝐸𝐸𝐸𝐸𝐸𝐸 = ∑ 𝑁𝑁𝑠𝑠𝑠𝑠 × 𝐹𝐹𝑆𝑆𝑆𝑆 , (2)
where 80% of jet fans were activated 2 min after ignition
Where 𝑁𝑁𝑠𝑠𝑠𝑠 is the number of fatalities for each fire
with a 3.2 m/s velocity. Recognition time, response time,
scenario and 𝐹𝐹𝑆𝑆𝑆𝑆 is the cumulated frequency of each fire
and time to leave the vehicle are determined as pre-
scenario.
movement times of tunnel users ‘evacuation, a time
Table 2 illustrates the variation of EDV for half
before the evacuation process toward the emergency
and double pre-movement parameters in comparison
exits begins. In this model, these pre-movement times
with its value in the standard configuration.
have basic values, 𝑡𝑡𝑟𝑟𝑟𝑟𝑟𝑟 ,𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 ,𝑡𝑡𝑟𝑟𝑟𝑟𝑟𝑟 ,𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 ,and 𝑡𝑡𝑣𝑣𝑣𝑣ℎ (each type
of vehicle). The higher these values, the more likely that
the harmful effects of the fire incident reach the users
before they start the evacuation. Three modes, basic
values of basic pre-movement parameters, half of the
basic pre-movement parameters, and double the basic
pre-movement parameters, were considered to study
the sensitivity of this model to pre-movement times. The
© 2023 Global Journals
 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

Table 2: The Variation of EDV for Half and Double Pre-Movement Parameters in Comparison with its Value in the
Standard Configuration

Half Parameters Reference Parameters Double Parameters

Zone 1, Zone 2, Zone 3 Zone 1, Zone 2, Zone 3 Zone 1, Zone 2, Zone 3

𝑡𝑡𝑟𝑟𝑟𝑟𝑟𝑟 ,𝑖𝑖𝑖𝑖𝑖𝑖 (min ) 1,1.5,2 2,3,4 4,6,8

𝑡𝑡𝑟𝑟𝑟𝑟𝑟𝑟 ,𝑖𝑖𝑖𝑖𝑖𝑖 (min ) 2,1.5,1 4,3,2 8,6,4

4 for PC, 8 for van, 30 for 8 for PC, 16 for van, 60 16 for PC, 32 for van, 120
𝑡𝑡𝑣𝑣𝑣𝑣ℎ (s) bus, and 2 for truck and for bus, and 4 for truck for bus, and 8 for truck
HGV and HGV and HGV
2023

EDV 3.54E-01 5.89E-01 9.04E-01

Year

_
Variation % -40.00% 53.35%
341
By halving these parameters, the EDV without ventilation decreased by a factor of 10. That
leads to lower cumulative frequencies and a lower FN
Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

decreased by 40%. On the other hand, by doubling

these parameters, the EDV increased by 53.3% which curve compared to the initial FN curve of the same
indicated greater pre-movement time and consequently, scenarios (without ventilation) (Figure 3). These results
a higher number of tunnel users implemented were explained when considering that the frequency of
evacuation process later and were in the damage occurrence of each scenario depended directly on the
intolerable condition for too long. Borghetti et al. [12] value of the incidental frequencies.
also carried out the same analysis and showed the
same variation.
c) The Accidents Frequency Involving HGVs
To study the influence of incident frequency
involving HGV, the frequency of HGV fire scenarios

Figure 3: Sensitivity Analysis of the Model to the Influence of Incident Frequency Involving HGV, HGV FN Curves of
Scenarios Without Ventilation
d) Tenability Threshold Temperature temperature, and double temperature. It is reasonable to
The tenability threshold temperature defines to expect that increasing the threshold temperature results
what temperature exposed tunnel users can resist the in a lower EDV or fewer fatalities, and vice versa. Table 3
extended damage of fire and smoke before considered shows the EDV variation in relation to the standard
as being dead. The initial tenability threshold configuration, halving, and double values. Doubling this
temperature in this model is 80˚C, the tolerable temperature made the EDV decrease by about 35%
maximum temperature for 15 min exposure during a while halving it caused the EDV to increase by 17.8%.
tunnel fire according to PIARC (1999). To study the
sensitivity of results to this temperature, three cases
were considered, initial temperature, halved

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

Table 3: The EDV Variation in Relation to Tenability Threshold Temperature of the Standard Configuration, Halving,
and Double Values
Halved paramer (˚C) Reference parameter (˚C) Doubled parameter (˚C)
𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 40 80 160
EDV 5.44E-01 4.62E-01 3.00E-01
Variation % 17.80% _ -35.04%

e) Different Types of Burning Vehicles on the for HGV scenarios was probably due to more effect of
Evacuation Time HGV fire scenarios on pre-movement time. The basic
A comparison between the HGV and bus fire recognition time decreased for more severe accidents

2023
scenario located in the middle of the tunnel length and and consequently, people became aware of the
80% of the tunnel length for three traffic conditions was threatening situation within a shorter time. The

Year
carried out. In this model, the exposed tunnel users are percentage of dead people and ASET are reported in
distributed homogeneously into cells. The time required Table 4 for the various scenarios. Results clearly show
for each cell evacuation via emergency exits was that in the case of a bus fire, all people might safely 1
35
between 192 and 695 s for the HGV fire and between evacuate the tunnel except in one scenario. In the event

Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

590 and 705 s for bus scenarios. Lower evacuation time of HGV fires, most or all people might be at risk.
Table 4: The percentage of Dead People and ASET
Fire Location People at Risk ASET Range
Traffic Ventilation Velocity
(%L) (%) (S)
Bus Free fluid 0.5 0 0% 616-681
Bus Congested 0.5 0 0% 603-698
Bus Stoppage 0.5 0 3.15% 625-705
Bus Free fluid 0.8 0 0% 638-684
Bus Congested 0.8 0 0% 600-690
Bus Stoppage 0.8 0 0% 590-700
HGV Free fluid 0.5 0 100.00% 633-703
HGV Congested 0.5 0 90.91% 192-690
HGV Stoppage 0.5 0 72.00% 599-651
HGV Free fluid 0.8 0 57.00% 623-693
HGV Congested 0.8 0 45.00% 600-698
HGV Stoppage 0.8 0 81.54% 598-692

V. Conclusion frequency analysis and quantitative consequence

Fire is the most important critical event for road analysis. The frequency of each fire scenario has been
tunnel users' safety. Although the UK’s road tunnels determined based on the basic fire rate of UK road
design manual identifies risk analysis as a practical tunnels per 100 million veh-km. This basic fire rate is
method for determining the risk level of all road tunnels updated by taking into account the length, traffic, and
in the United Kingdom, it must be confessed that there gradient of the tunnel under study. Then the effect of
is a need for a comprehensive system-based time of fire incident, traffic condition, accident type,
quantitative risk assessment tool to cover fires in road vehicle type, and the fire source location on the updated
tunnels. fire rate is considered through branches of the
Risk assessment techniques play a very frequency event tree.
valuable part in tunnel safety decision-making. The quantitative consequence model comprises
Consequently, a model of quantitative risk analysis three parts, the queue formation model that estimates
concerning safety in road tunnels called LBAQRA the number of potential tunnel users for each lane of a
has been developed to perform a risk analysis as multi-lane tunnel separately, the distribution model that
required by UK regulations for complex fire scenarios. calculates the evacuation distance in two dimensions,
The proposed model performs a quantitative and the egress model that calculates the required
risk assessment to effectively and efficiently evaluate the egress time.
risks for UK road tunnels and includes quantitative

© 2023 Global Journals

 Design and Implementation of a Quantitative Risk Assessment Model for UK Road Tunnels

This paper first presents a brief description of safety evaluation”, Tunnelling and Underground
the sections of this model and how it derives the societal Space Technology, 31, 33-50, 2012.
risk. In the second part, an illustrative case study was 4. Haddad, R. K., Maluk, C., Reda, E., Harun,
carried out for the Southwick Tunnel to study the Z.,“Critical velocity and backlayering conditions in
robustness of this model, and the risk reduction rail tunnel fires: State-of-the-art review”,Journal of
potential of different ventilation strategies was assessed. Combustion, 2019.
A study of various emergency ventilation strategies 5. European Parliament and the Council of the
shows that when the ventilation system is activated very European Union. Directive 2004/54/EC of the
late, it cannot cope with the influences of fire and European parliament and of the council of 29 April
smoke. For scenarios with a 3.5 m/s ventilation velocity 2004 on minimum safety requirements for tunnels in
when jet fans are activated by the detection system, a the Trans-European Road Network. Official Journal
higher ventilation velocity can make a more tenable of the European Union, 167, 39–91, 2004.
2023

situation during evacuation and alleviates the negative 6. PIARC Technical Committee C3.3 Road Tunnel
Year

effect of late activation of jet fans, although the Operation. Risk Analysis for Road Tunnels. (2008).
activation time is between 2 and 3 min. [Online]. Available: http://publications.piarc.org/
361 ressources/publications_files/4/2234,TM2008R02-
The Sensitivity Analysis Shows that: WEB.pdf
• By halving pre-movement time parameters, the EDV 7. Diamantidis, D., "Risk analysis versus risk
Global Journal of Science Frontier Research ( I ) Volume XXIII Issue I V ersion I

decreases. By doubling these parameters, the EDV acceptability in major European tunnel projects",
increases, which indicated higher pre-movement Proceedings 1st Asia Pacific conference on risk
time and consequently, a higher number of tunnel management and safety, Hong Kong, 2005.
users who implemented the evacuation process 8. Botschek, K., Kohl, B., and Hörhan, R.,"Austrian
later and were in the intolerable environment for too Risk Analysis for Road Tunnels–Development of a
long and became incapacitated. New Method for the Risk Assessment of Road
• When the frequency of HGV fire scenarios Tunnels", International Conference Tunnel Safety
decreases by a factor of 10, that leads to lower Forum for Road and Rail, 2007.
cumulative frequencies and a lower FN curve.This 9. Kohl, B., Zibert, M., “Risk analysis study for
indicates that the frequency of occurrence of each Slovenian motorway tunnels”, Proceedings of
scenario depended directly on the value of the Slovenski Kongres O Cestah in Promeu, Portoroz,
incidental frequencies. Slovenia, 2010.
• An increase in the threshold temperature results in a 10. Hadjimichael, M., “A fuzzy expert system for aviation
lower EDV or fewer fatalities, and vice versa risk assessment”, Expert Systems with Applications,
because the threshold temperature is the main 36, 6512–6519, 2009.
tenability criterion that determines intolerability 11. Roisenberg, M. C., Schoeninger, R., and da Silva,
conditions. Rodrigues,“A hybrid fuzzy probabilistic system for
• Different types of burning vehicles result in different risk analysis in petroleum exploration prospects”,
evacuation times.because the basic recognition Expert Systems with Applications, 36, 6282–6294,
time decreased for more severe accidents such as 2009.
HGV fire and consequently, people became aware 12. Borghetti, F., Cerean, P., Derudi, M. and Frassoldati,
of the threatening situation within a shorter time. A., “Road Tunnels: An Analytical Model for Risk
Analysis”, Springer International Publishing, 2019.
Acknowledgment 13. Moonis, M., Wilday, J., Wardman, M. and Balmforth,
We would like to appreciate for UKM’s grant H., “Assessing the safety of delivery and storage of
GUP-2020-015, and London Bridge Associates Ltd. hydrogen”, Health & Safety Laboratory Report
PS/08/01, 14, 2008.
References Références Referencias
1. Beard, A., and D, Cope., “Assessment of the safety
of tunnels”. Brussels, Belgium: European
Parliament, 2008.
2. Caliendo, C., De Guglielmo, M.L., “Accident rates in
road tunnel and social costs evaluation”, In: SIIV –
5th International Congress – Sustainability of Road
tunnels Infrastructures, 166–177, Rome, 2012.
3. Caliendo, C., Ciabeli, P., Gugliemo, M.L., et al.,
“Numerical simulation of different HGV fire
scenarios in curved bidirectional road tunnels and

© 2023 Global Journals