Study of the Structural Design Parameters of an Emergency Rescue Station between Continuous Tunnel Portals in a Single-Tube Railway Tunnel Group

Abstract

In China, emergency rescue stations between continuous tunnel portals have been newly designed and applied in extra-long single-tube mountain railway tunnel groups exceeding 20 km in length. Lower pass rates at emergency exits and slower evacuation speeds on evacuation routes are the major contributors to fire-related fatalities. Therefore, it is very important to use reasonable structural design parameters to ensure the effective evacuation of escapees. This study conducts on-site testing of a railway tunnel evacuation model for 457 people to determine the evacuation speeds. Based on coefficient adjustments, it provides the evacuation speeds of different groups of people in a smoke-filled emergency rescue station. Additionally, a series of simulation evacuation models for 1,500 people in emergency rescue stations between continuous tunnel portals were developed using the buildingEXODUS software. These models simulate the evacuation time under different structural parameters and aim to determine the optimal design parameters by analyzing the relationship between evacuation time and structural factors. The study also investigates the impact of structural design parameters on both evacuation time and evacuation density. The results indicate that, when the width and number of evacuation routes are not optimal, evacuation speed significantly decreases, and more people become stranded in the evacuation paths, which increases the overall risk. This paper recommends that emergency rescue stations between continuous tunnel portals should have at least four evacuation routes, each with a minimum width of 3 m. Additionally, the platform height should be 0.3 m, and the platform width should be 2.5 m.

Keywords

single-tube railway tunnel group emergency rescue station structural parameters evacuation fire safety

China is a mountainous country, with mountainous regions accounting for two-thirds of its total land area. With the rapid development of railway infrastructure, many mountain railway tunnel groups (a group of tunnels with distances between adjacent tunnel portals shorter than the length of a passenger train) have been constructed and rapidly developed in recent years, especially the extra-long mountain railway tunnel groups, each with a length of no less than 20 km ( 1 – 3 ). Currently, for disaster prevention and rescue, the Code for Design on Rescue Engineering for Disaster Prevention and Evacuation of Railway Tunnel TB-10020-2017 stipulates that, for extra-long railway tunnels, there must be at least one emergency rescue station, and the distance between the rescue station and the tunnel portal or adjacent emergency rescue stations should not exceed 20 km ( 4 ). The railway tunnel group can be considered as a facility for disaster prevention and rescue. When constructing an emergency rescue station in an extra-long railway tunnel group, if the terrain and spacing conditions are suitable, the location of the rescue station is preferably chosen at the portal of two adjacent tunnels because of the open-air section. This open-air section not only provides more natural light, enhancing the brightness of the evacuation environment, but also facilitates smoke extraction in case of fire. The emergency rescue station between continuous tunnel portals consists of an open-air tunnel section and two tunnel portal sections on either side of the open-air tunnel section, as shown as a new structure in Figure 1.

Some extra-long mountain railway tunnel groups were constructed as single-tube double-track tunnels rather than double-tube single-track tunnels because of considerations of terrain and geological conditions, engineering construction costs, and other factors ( 5 – 7 ). For example, the Greater Qinling Tunnel Group has a total length of 44,223 m on the Xi’an-Chengdu Railway ( 8 ).

There are three approaches for setting up evacuation routes for the emergency rescue station between the continuous tunnel portals in a single-tube railway tunnel group ( 9 – 14 ). These are shown in Figure 2. The first approach is to set up a pilot tunnel parallel to the main tunnel, with both tunnels connected by transverse alleyways. The second approach involves setting up auxiliary alleyways at the tunnel portal section that connect to the outside, such as shorter inclined wells or horizontal tunnels. The third approach is to establish evacuation trails at the open-air tunnel section. An example of this approach is the Huangtuwan Bridge Emergency Rescue Station on the Dayaoshan Tunnel Group of the Wuhan-Guangzhou Railway, which features two auxiliary alleyways at the tunnel portal section and two evacuation trails at the open-air tunnel section ( 15 ).

The safe evacuation standard for emergency rescue stations is that the required safety egress time (RSET) must be less than 6 min, that is, from the start of the evacuation to the end when all people are evacuated to a safe area ( 16 , 17 ). In a railway tunnel fire accident, major causes of fire deaths include a lower pass rate because of bottlenecks at emergency exits, a lower evacuation speed, and even stomping accidents resulting from crowding on the evacuation route ( 18 ). Therefore, it is crucial to use reasonable structural design parameters, such as a sufficient number of exits and an adequate evacuation route width, to ensure the effective evacuation of escapees.

Recent studies have focused on the structural design parameters of emergency rescue stations in double-tube railway tunnels. These studies have presented and summarized factors such as transverse alleyway spacing and width, as well as platform width and height ( 19 – 30 ). Wang et al. conducted a numerical simulation study on the optimal parameter settings of emergency rescue station evacuation facilities during fire accidents in long railway tunnels ( 31 ). Xu et al. used the orthogonal experimental method to investigate the optimal parameter design of evacuation facilities in emergency rescue stations within tunnels with steep slopes ( 32 ). Li et al. proposed a new type of emergency rescue station structure and, based on factors such as the train fire location, tunnel longitudinal slope, and longitudinal ventilation wind speed, determined the parameters for the evacuation facilities in the new station ( 33 ). Cao et al. studied the impact of the number of cross passages in emergency rescue stations in high-altitude, long tunnels under different fire scenarios on the evacuation process and determined the reasonable number of cross passages in the rescue stations ( 34 ). Li et al. investigated the influence of evacuation facility design parameters in long railway emergency rescue stations on evacuation time and evacuation paths ( 35 ). However, less research has been conducted on emergency rescue stations between continuous tunnel portals, with some recommended values for structural design parameters based on engineering experience, but these studies lack the corresponding theoretical basis ( 4 , 8 , 36 ). At present, there is no systematic calculation of the structural parameters of emergency rescue stations between continuous tunnel portals based on emergency evacuation security. The research results of this study will provide theoretical guidance for the construction of similar projects.

This study aims to fill that gap by conducting an on-site evacuation test involving 457 participants and simulating evacuation scenarios with 1,500 individuals using buildingEXODUS software. The primary goal is to explore the relationship between structural parameters (e.g., number and width of evacuation routes, platform dimensions) and evacuation efficiency, and to propose optimal design recommendations to guide practical engineering.

The remainder of the paper is structured as follows. The next section introduces the structural components and evacuation route options for emergency rescue stations, followed by a section presenting the field evacuation test and the derivation of evacuation speeds. The section after that describes the modeling approach using simulation tools. Then there is a section discussing the simulation results under different parameter conditions. The penultimate section compares and analyzes the influence of various parameters. The final section summarizes the conclusions and outlines future research directions.

Structural Parameters Composition and Evacuation Path

Generally, RSET primarily refers to the movement time during evacuation scenarios, as all passengers wait until the train arrives at an emergency rescue station and then simultaneously begin evacuation as soon as the train door opens. When the number of passengers is determined, the movement time depends on the number of available exits and the width of the evacuation paths in the emergency rescue station. The main evacuation path consists of a train door, a platform, and an evacuation route. When the fire-affected train stops at the emergency rescue station between continuous tunnel portals, the fire-affected carriage is parked in the open-air section, and the passengers disembark the train to the platform from the carriage door close to the platform. Notably, the passengers disembarking on the other side need to cross the track to the opposite platform and evacuate to the evacuation route in the opposite direction of the fire source, as shown in Figure 3. Thus, the key parameters for the emergency rescue station between continuous tunnel portals mainly include the number and width of evacuation routes, as well as the length, height, and width of the platform.

Determination of Evacuation Speed in Emergency Rescue Station

Field Simulation

We conducted an evacuation model test in a specially designed railway tunnel model with a train at Southwest Jiaotong University in China ( 36 ). This tunnel model, which is 80 m long and 6 m wide, contains three train carriages, an emergency exit, and a platform (80 m long and 2 m wide). In this study, college students aged between 19 and 23 years old were selected as experiment participants, with a male-to-female ratio of 6:5. The test process is mainly recorded by multi-point and multi-directional camera, and then data processing and analysis are carried out according to the video. The evacuation test monitored the evacuation process of 457 people from the fire-affected carriage to the emergency exit in the tunnel. The test recorded the movement rate of evacuees between adjacent train carriages, the passage rate of the carriage door during disembarkation, and RSET; these were tested in the emergency station model. The picture of the tunnel model test scene is shown in Figure 4.

The on-site model tests mainly focused on three aspects:

(1) The evacuation situation of passengers after getting off the train, under different evacuation passage widths in the tunnel. The evacuation path diagram is shown in Figure 5.

(2) Figure 6 presents the evacuation scenarios under different numbers of train door openings, including four cases: single-side single door (Figure 6(a)), double-side four doors (Figure 6(b)), double-side two doors (Figure 6(c)), and single-side two doors (Figure 6(d)).

(3) The evacuation situation where passengers do not get off the train, but evacuate from the fire-affected carriage to the adjacent carriage. The evacuation path diagram is shown in Figure 7.

Numerical Simulation (buildingEXODUS Evacuation Model)

In this paper, the buildingEXODUS evacuation model was used to evaluate the evacuation time. To date, this model has been applied to various types of infrastructure (e.g., supermarkets, hospitals, high-rise buildings, schools, train stations, airports, theatres, tunnels) for simulating people movement and behavior under various evacuation conditions ( 37 ). buildingEXODUS software considers the interaction between an individual and the environment or other individuals following a list of rules, and predicts numerous discrete individual (virtual people) movements through 3D space. The geometry of the space under consideration is mapped by a 2D grid that contains nodes that can be connected with eight other neighboring nodes through arcs ( 38 , 39 ). This grid size is generally chosen to represent the area occupied by a single person and can be set to different sizes in different locations. The connections between the grid and exits, including internal exits and external exits, are provided by arcs, the flow rate through exits is a function of its width, and the congestion time at exits can be determined by Equation 1. buildingEXODUS can qualitatively predict the effectiveness of different emergency exit layouts because of its embedded exit selection algorithms ( 40 ).

t_{c} = N / (f \cdot w)

(1)

where

t _c = the congestion time at the exits (s),

f = the flow rate through the exits (per/m·s),

w = the width of the exits, and

N = the number of individuals congested near the exit (per).

Each individual moving under consideration is assigned certain basic attributes, namely, the individual’s height, weight, age, and sex. Additionally, other essential attributes include the free walking speed (which can automatically calculate the up and down stairs speed), jumping speed, and crawling speed through setting percentages; movement flexibility (e.g., leg disability); behavioral characteristics (e.g., across obstacles); movement ability; and the patience required to make queuing choices in the event of an evacuation. Such attributes can be assigned both deterministically and by distribution laws.

Determination Process of Evacuation Speed in Emergency Rescue Station

The evacuation speed of normal young males and females is determined by the mutual authentication of experimental and simulation results. Considering the difference in personnel composition and environment between the field tests environment and the actual tunnel emergency rescue station, several coefficients—age, gender combination, fire smoke visibility, toxic gas concentration, and road roughness—are introduced to obtain more reliable evacuation speed of personnel in the emergency rescue station. The determination process of walking speed with smoke in emergency rescue station is shown in Figure 8.

Comparison of Experimental and Numerical Simulation Results

Comparing the experimental results with the numerical simulation results, the comparison of the number of passengers evacuated at different times during the whole process of evacuation in emergency rescue stations when the fire-affected train stops immediately is shown in Figure 9a; the comparison of the number of passengers having left the fire-affected train at different times when the fire-affected train stops immediately is shown in Figure 9b; and the comparison of the number of passengers evacuated at different times between multiple carriages when the fire-affected train continues running is shown in Figure 9c.

According to the Permanent International Association of Road Congresses (PIARC), the recommended passenger’s evacuation velocity is in a range of 0.5 m/s to 1.5 m/s ( 41 , 42 ). The scenarios of the above model test were simulated, and the sensitivity of the buildingEXODUS numerical parameters was analyzed. With the given information, different velocity settings for young adult males and females have been tried. Based on the trial simulations, 1.44 m/s and 1.2 m/s, respectively, matched the experimental results best.

Correction of Walking Speed during Emergency Evacuation

The main influencing factors on personnel evacuation speed include the composition of the affected people (gender and age), the visibility of fire smoke, concentration of toxic gases, and pavement smoothness.

Effect of Fire Smoke Visibility

The effect of fire smoke on the walking speed is expressed as the extinction coefficient K ( 43 – 45 ). When K > 0.2 m⁻¹, the walking speed can be reduced because of both smoke concentration and irritation. The walking speed can be calculated with the following Equation 2 ( 46 ):

v_{i}^{0} (K) = max {v_{i, min}^{0}, \frac{v_{i}^{0}}{α} (α + β K)}

(2)

where

v_i ⁰(K) = walking speed of a person under the influence of smoke (m/s),

v_i ⁰ = normal walking speed (m/s),

K = extinction coefficient (m⁻¹),

α = empirical coefficient (value 0.706 ms⁻¹),

β = empirical coefficient (value −0.057 m² s⁻¹), and

v ⁰ _i,min = minimum walking speed (value is 0.1v_i⁰).

Influence of Space Toxic Gas Concentration

As the fire products and their components are very different in different scenarios, but CO is the first cause of death, it is simplified to consider only the effect of CO volume fraction on the human body. Taking the simplified CO concentration ξ as 0.08, the influence coefficient of space toxic gas concentration on personnel evacuation is shown in Equation 3 ( 47 , 48 ):

f 1 (ξ) = {\begin{matrix} 1, ξ < 0.02 \\ - \frac{1}{3} ξ^{2} - 0.36 ξ + 1.01, 0.02 \leq ξ < 1.28 \\ 0, ξ \geq 1.28 \end{matrix}

(3)

where

f ₁(ξ) = the coefficient of influence of air toxic gas concentration, and

ξ = the simplified CO concentration (%).

Influence of Pavement Smoothness

People will inevitably walk on uneven terrain or rails during the evacuation because of the constrained evacuation area in the railway tunnel and the condensed escape route. As a result, their speed will be slower than on the flat ground. The Code for Fire Protection Design of Building (GB50016-2014) states that the rate of personnel evacuation on uneven terrain or steps is around 86% higher than the rate of evacuation on the ground ( 49 ). Therefore, the reduction factor of pavement smoothness f₂ is 0.86.

Based on gender and age, the visibility of fire smoke, concentration of toxic gases, and pavement smoothness, the evacuation speed of a fire tunnel was determined, as shown in Table 1.

Table 1.

Evacuation Speed Value of the Different Personnel Types

Personnel type	Adult males	Adult females	Children	Older
Normal walking speed (m/s)	1.44	1.2	0.96	0.86
Walking speed with smoke in emergency rescue station (m/s)	1.18	0.98	0.78	0.7

Establishment of the Evacuation Model

Fixed Parameters

In this paper, a series of simulation evacuation models for the evacuation of an emergency rescue station between continuous tunnel portals were established with buildingEXODUS software to simulate the evacuation time required under different structural parameters (including the height and width of the platform and the number and width of evacuation routes). The tunnel models used the sectional dimension of a single-tube double-track railway tunnel with a design speed of 200 km/h, and the train travel area width is 8 m, as shown in Figure 3b. The length of the platform consisting of a complete open-air tunnel section and one-sided tunnel portal section was equal to the length of the operated train (450 m here), as shown in Figure 3a.

A type 25 fast passenger train with a length of 25 m, width of 3 m, aisle width of 0.5 m, and exterior door width of 0.8 m was selected as the modeling object. The train models consisting of 17 carriages carrying 1,500 people (i.e., 1,480 passengers and 20 staff); among them, adult males accounted for 45%, adult females accounted for 38%, children accounted for 10%, and elderly people accounted for 7% of the total ( 48 ). The passenger distribution is shown in Table 1 and Figure 10 ( 49 ). The evacuation of a fire tunnel in emergency rescue station is shown in Table 2 ( 40 , 49 ).

Table 2.

Number of Fully Loaded Passengers in Different Types of Carriage.

Carriage type	Hard seat carriage	Hard sleeper carriage	Soft sleeper carriage	Dining carriage	Locomotive/loading cargo carriage
Number of fully loaded passengers	128	66	36	90	10
Number of carriages	8	5	1	1	2

Scenario

The model selected a relatively unfavorable parking mode in which the end carriage on fire parked in the open-air tunnel section of the emergency rescue station between continuous tunnel portals and the other carriages parked in the one-sided tunnel portal section. In this parking mode, all the people must evacuate in the opposite direction of the fire-affected carriage, and only evacuation routes on one side of the emergency rescue station between continuous tunnel portals were available for personnel.

The order in which the structural parameters were determined is as follows: the height of the platform, the number of evacuation routes, the width of evacuation routes, and the width of the platform. The height of the platform is the vertical distance from the track surface to the platform plane, and 11 heights, of 0 m, 0.1 m, 0.2 m, 0.3 m, 0.4 m, 0.5 m, 0.6 m, 0.7 m, 0.8 m, 0.9 m, and 1.0 m, were calculated to obtain the speed rate at which people disembark the train to determine the optimal height of the platform. Similarly, the number of evacuation route calculation scenarios was two, four, and six on one side, and the widths of the evacuation route calculation scenarios were 2 m, 3 m, 4 m, 5 m, and 6 m, and the widths of the platform calculation scenarios were 2 m, 2.5 m, 3 m, 3.5 m, and 4 m. A total of 44 models were established. The schematic diagram of structural parameters and the setting of test conditions are shown in Figure 11 and Table 3, respectively.

Table 3.

The Setting of Test Conditions

No.	Scenario	Conditions
A1 to A11	Platform height (m)	0	0.1	0.2	0.3	0.4	0.5	0.6	0.7	0.8	0.9	1
B1 to B3	Number of one-sided evacuation routes	2	4	6
C1 to C15	Number of evacuation routes	2					4					6
C1 to C15	Width of evacuation routes (m)	2	3	4	5	6	2	3	4	5	6	2	3	4	5	6
D1 to D15	Number of evacuation routes	2					4					6
D1 to D15	Width of platform (m)	2	2.5	3	3.5	4	2	2.5	3	3.5	4	2	2.5	3	3.5	4

Calculation Results and Analysis

Height of the Platform

There are three stairs under each exterior door of the carriage, with a total height of approximately 0.5 m, and the vertical distance from the track surface to the carriage floor is 1.0 m. According to the platform height settings, the vertical distance, down which passengers disembarking to the platform from the carriage door (close to the platform) need to jump, is the difference between the vertical distance from the track surface to the step surface closest to the platform and the platform height. Additionally, the vertical distance, down which passengers disembarking from the opposite carriage door need to jump, is the vertical distance from the track surface to the bottom step surface; its value is 0.5 m. Here, the time data of people disembarking in the most unfavorable fire situation was extracted, and the fire situation is that there are two doors on one side of the fire-affected carriage (hard seat carriage, 128 passengers) for people to disembark because of the fire source occurring at the door on the other side of the fire-affected carriage, as shown in Figure 12.

In the most unfavorable fire situation, the evacuation time and speed of passengers disembarking at different platform heights are shown in Figure 13. Because the height that passengers need to jump off in different carriage doors are different, the number of passengers passing through each carriage door is different, as shown in Figure 14. Here, the carriage door next to the platform is Door 1, and the opposite door is Door 2.

Figures 13 and 14 show that the time required for passengers to disembark increases and the disembarking speed decreases as the platform height decreases. The time required for 128 passengers to disembark is 281 s, and the speed of disembarking is only 0.46 persons per second (per/s) when the platform height is 0 m. When the platform height rises to 1.0 m, the disembark time is reduced to 114 s, and the disembark speed is 1.12 per/s. When the platform height is increased to 0.3 m, the speed of passengers disembarking is 0.9 per/s, the number ratio of people passing through Door 1 and Door 2 is approximately 0.7:0.3, and the speed is significantly improved. When the platform height continues to increase, the disembark speed reduction is not obvious. Therefore, by considering the construction cost, it is recommended that the platform height of the emergency rescue station be 0.3 m.

Number of Evacuation Routes

This paper calculated the number of evacuation routes using three assumption prerequisites:

(1) The structure of the emergency rescue station between continuous tunnel portals is basically symmetrical, and the number of evacuation routes in the two sections of the tunnel is the same.

(2) The most unfavorable parking mode: a fire broke out at the end carriage of the train, and the fire-affected carriage stopped at the open-air tunnel section. At this time, there was only one end of the evacuation routes in the emergency rescue station between continuous tunnel portals for people to escape.

(3) The widths of the evacuation routes and the platform were wide enough (8 m selected here) to not hinder the evacuation.

Based on the above calculation conditions, a half-sided model of the emergency rescue station between continuous tunnel portals was established, and the number of evacuation routes were two, four, and six, which were four, eight, and 12, respectively, in the entire emergency rescue station. The calculation model is shown in Figure 15.

RSET under the above three calculation scenarios was estimated and compared with the standard of 6 min to determine whether the number of evacuation routes meets the safety requirements of the evacuation. The results are shown in Table 4.

Table 4.

Calculation Results of Evacuation Time with Different Numbers of Evacuation Routes

Number of one-sided evacuation routes	RSET (s)	Was evacuation time <6 min?
2	354	Yes
4	248	Yes
6	233	Yes

Note: RSET = required safety egress time.

Table 4 shows that RSET decreased as the number of evacuation routes increased. When two evacuation routes were set on one side of the emergency rescue station, RSET was 354 s, which is less than 6 min, and meets the safety requirements of the evacuation. Therefore, the number of evacuation routes on the side of the emergency rescue station between continuous tunnel portals should not be less than two; that is, at least four evacuation routes should be set up. If the evacuation passages can only be set on one side because of terrain limitations, the number on that side should be appropriately increased. If it is possible to set evacuation passages outside the tunnel, the number of passages on each side can be reduced accordingly.

Width of Evacuation Routes

The calculation model of the evacuation route width was still established under the most unfavorable parking mode mentioned above. Assume that the width of the platform is wide enough to not hinder the evacuation. Currently, the regulations stipulate that the evacuation time within stations or emergency rescue stations should be less than 6 min ( 17 ). Therefore, the optimal width of the evacuation routes was determined according to an RSET less than 6 min, and the gathering time was very short or 0 s at the exit under different evacuation route numbers, which were two, four, and six on one side of the emergency rescue station between continuous tunnel portals. The “gathering time” refers to the duration of high-density crowding at the exit during evacuation. When people move toward the exit, the flow rate decreases because the narrowing of the exit passage causes a bottleneck. This results in crowding at the exit. Tregenza stated that, when the average density reaches 3 people/m², congestion and stagnation will occur ( 50 ). When the density reaches 5 people/m², forward movement will stop. Therefore, gathering time can be defined as the duration during which the density exceeds 3 people/m². The calculated widths of the evacuation routes were 2 m, 3 m, 4 m, 5 m, and 6 m. The calculation results of RSET and gathering time at the exit under different evacuation route widths and numbers are shown in Figure 16.

Figure 16 shows that RSET and gathering time decreased as the evacuation route width increased under the same number of evacuation routes. When the number of evacuation routes was two and the width of the evacuation routes was 2 m, RSET was 389 s, which does not meet evacuation time requirements, and the gathering time was 204 s, which is too long to be prone to trampling in this scenario. When the number of evacuation routes was two and the width of the evacuation routes was 3 m, RSET was 354 s, which meets the evacuation time requirements, and the gathering time was 28 s, which is very short in this scenario. Therefore, the optimal evacuation route width was 3 m when the number of evacuation routes was two. When the number of evacuation routes was four and the width of the evacuation routes was 2 m, RSET was 254 s, which meets the evacuation time requirements, but the gathering time was 108 s, which is too long to be prone to trampling in this scenario. Furthermore, when the width of the evacuation routes was 3 m, the gathering time was 28 s, which is very short. When the evacuation route width continues to increase, RSET and gathering time reduction are not obvious or even unchanged. Therefore, considering the construction cost, the optimal evacuation route width was 3 m when the number of evacuation routes was four. Similarly, the optimal evacuation route width was 3 m when the number of evacuation routes was six. The width of the evacuation routes is recommended to be 3 m.

Width of the Platform

The calculation model of the platform width was still established under the most unfavorable parking mode mentioned above. The optimal platform width was determined according to RSET less than 6 min, and the gathering time was very short or 0 s at the exit under a 3 m wide evacuation route and different evacuation route numbers, which were two, four, and six on one side of the emergency rescue station between continuous tunnel portals. The calculated widths of the platform were 2 m, 2.5 m, 3 m, 3.5 m, and 4 m. The calculation results of RSET and gathering time at the exit under different platform widths and evacuation route numbers are shown in Figure 17.

Figure 17 shows that RSET and gathering time decreased as the platform width increased under the same number of evacuation routes. The evacuation route width of all calculation scenarios was 3 m. When the number of evacuation routes was two and the platform width was 2 m, RSET was 368 s, which does not meet evacuation time requirements, and the gathering time was 87 s, which is too long to be prone to trampling in this scenario. When the number of evacuation routes was two and the width of the platform was 2.5 m, RSET was 354 s, which meets the evacuation time requirements, and the gathering time was 30 s, which is very short in this scenario. When the evacuation route width continues to increase, RSET and gathering time reduction are not obvious or even unchanged. Therefore, considering the construction cost, the optimal platform width was 2.5 m when the number of evacuation routes was two. Similarly, the optimal platform width was 2.5 m when the number of evacuation routes was four and six. The width of the platform is recommended to be 2.5 m.

Parameter Comparison

The recommended values from this study are compared with the current regulations and typical tunnel disaster prevention and rescue facility design parameters, as shown in Table 5.

Table 5.

Parameter Comparison

Project	Platform height	Number of evacuation passages	Width of evacuation passages	Platform width
Recommended value (this study)	0.3 m	8 (set on both sides)	≥3 m	2.5 m
Current regulations	≥0.3 m	na	4.5 m	2.3 m
Xinmin-Jinpingyan Tunnel rescue station (Chenglan Line)	0.3 m	6 (set on one side only)	4.5 m	2.3 m
Huangtuwang rescue station (Wuguang Line)	0.1 m	5 (3 set outside the tunnel)	3.5 m (4 passages),2.5 m (1 passage)	2.26 m
Zhi Fang rescue station (Xicheng Line)	0.3 m	6 (2 set outside the tunnel)	4.5 m (4 passages),1.5 m (2 passages)	2.3 m

Note: na = not applicable.

As shown in the table, the platform height and width of the emergency rescue stations at the tunnel entrances of single-bore tunnel clusters are generally consistent with both the recommendations in this study and the current regulations. Specifically, the platform height is 0.3 m, and the platform width is at least 2.3 m. At the Huangtuwang rescue station on the Wuguang Line, because of terrain slope constraints, the platform height is set at only 0.1 m, so evacuation ladders or platforms were provided on the trains accordingly.

The current regulations do not specify or recommend the number of evacuation passages for tunnel entrance emergency rescue stations. Based on calculations, this study suggests that evacuation passages should be symmetrically arranged inside both tunnel entrances, with four passages on each side, making a total of eight passages. This configuration can fully meet the evacuation safety needs in the event of a fire at any part of the train. Currently, the number of evacuation passages at completed tunnel entrance emergency rescue stations is typically five or six. This is mainly because some passages are set outside the tunnel, which provides better evacuation conditions and efficiency than inside passages, so the number has been reduced in practice.

The current regulations recommend using the tunnel interior rescue station parameters for the width of evacuation passages, which is 4.5 m. However, this study finds that, with eight evacuation passages, a width of at least 3 m is sufficient. If the number of evacuation passages is reduced, the width should be increased accordingly to ensure evacuation efficiency. This conclusion has also been verified in the parameters of typical tunnels that have been built.

In conclusion, the recommended design parameters for disaster prevention and evacuation facilities at the tunnel entrance emergency rescue stations of single-bore tunnel clusters proposed in this study are scientifically sound and applicable. These suggestions can serve as a reference for similar projects in the future.

Conclusions

This study focuses on emergency rescue stations in continuous tunnel groups and conducts on-site testing of a railway tunnel evacuation model for 457 people to determine the evacuation speeds. Based on coefficient adjustments, it provides the evacuation speeds of different groups of people in a smoke-filled emergency rescue station. Additionally, a series of simulation evacuation models for an emergency rescue station between continuous tunnel portals was established to simulate the evacuation time required under different structural parameters. The goal was to determine the optimal structural design parameters based on the relationship between evacuation time and these parameters.

The following conclusions can be drawn from this study:

(1) When the width and number of evacuation routes are not optimal, there is a significant reduction in evacuation speed, and more people become stranded in the evacuation paths, which increases the risk.

(2) As the platform height increases, the time required for passengers to disembark decreases, and the speed of disembarking improves. When the platform height is increased to 0.3 m, the speed of the passengers disembarking is significantly improved, and when the platform height continues to increase, the disembarking speed reduction was not obvious. Therefore, by considering the construction cost, it is recommended that the platform height of the emergency rescue station be 0.3 m.

(3) As the number and width of evacuation routes and the platform width increase, RSET and the gathering time at the exit decrease. To ensure that RSET meets the evacuation time requirements and the gathering time is short, at least four evacuation routes should be set in emergency rescue stations between continuous tunnel portals, and each evacuation route should have a width of at least 3 m. The platform width is recommended to be 2.5 m.

As the length of railway tunnels continues to increase, ultra-long tunnels exceeding 20 km are emerging. Emergency rescue stations, as a new type of structure, are being used for personnel evacuation in case of fires and other emergencies in these long tunnels. There is a need to conduct further evacuation behavior experiments involving people of different ages and genders, and to strengthen research on the structural design parameters of emergency rescue stations. This will help improve railway tunnel construction standards and promote the safe operation of railway tunnels.

Footnotes

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: Qi Li, Xue Wang; data collection: Qi Li, Xue Wang; analysis and interpretation of results: Qi Li, Xue Wang, Chang Yang; draft manuscript preparation: Qi Li, Xue Wang, Xiongzhi Cao, Min Li, Peng Lei, Xueyi Xie, Jiale Cheng, Chang Yang. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is funded by the National Natural Science Foundation of China (Grant No. 51908387) and the Sichuan Agricultural University Innovation and Entrepreneurship Training Program Project (Grant No. 23023019).

ORCID iDs

Xue Wang

Xiongzhi Cao

Jiale Cheng

References

Cábová

Apeltauer

Okřinová

Wald

F. A. E.

Application of Fire and Evacuation Models in Evaluation of Fire Safety in Railway Tunnels. IOP Conference Series Materials Science of Engineering, Vol. 236, No. 1, 2017, p. 012080.

Hassannayebi

Memarpour

Mardani

Shakibayifar

Bakhshayeshi

Espahbod

A Hybrid Simulation Model of Passenger Emergency Evacuation Under Disruption Scenarios: A Case Study of a Large Transfer Railway Station. Simulation, Vol. 14, No. 3, 2020, pp. 204–228.

Yonghua

Jun

Gang

, et al. Safe Velocity of On-Fire Train Running in the Tunnel. Tunnelling and Underground Space Technol Incorporating Trenchless Technology Research, Vol. 60, 2016, pp. 210–223.

Code for Design on Rescue Engineering for Disaster Prevention and Evacuation of Railway Tunnel, TB10020-2017. China Railway Publishing House, 2017.

Zhu

Single Double-track Tunnel Tube or Twin Single-track Tunnel Tubes: Case Study on Super-long Yuntunbao Tunnel on Chengdu-Lanzhou Railway. Tunnel Construction, Vol. 34, No. 7, 2014, pp. 661–667.

Arturo

Orlando

Adriana

Daniel

An Experimental Data-Set on Merging Flows in Rail Tunnel Evacuation. Tunnelling Underground Space Technol, Vol. 70, 2017, pp. 155–165.

Liu

Research on the Mode of Single-track Tunnel or Double-track Tunnel of Super-long Railway Tunnel. Journal of Railway Engineering Society, Vol. 35, No. 9, 2018, pp. 49–53+66.

Shi

Harzard Prevention and Rescue Options for Mountainous Passenger-dedicated Railways. Railway Standard Design, No. 8, 2011, pp. 75–79+84.

Kobes

Helsloot

Vries

B. D.

, et al. Building Safety and Human Behaviour in Fire: A Literature Review. Fire Safety Journal, Vol. 45, No. 1, 2010, pp. 1–11.

10.

Xiang

G.-R.

Research on Allocation Method of Emergency Station on High-Speed Railway Bridge in Tunnel Group. China Safety Science Journal, Vol. 24, No. 6, 2014, pp. 99–104.

11.

Sun

Research on the Design of Disaster Prevention and Relief in Large Tunnel of Taiyuan-Shijiazhuang Passenger Dedicated Line. Journal of Railway Engineering Society, Vol. 26, No. 10, 2009, pp. 79–83.

12.

Guo

Yuan

, et al. Numerical Simulation of Smoke Stratification in Tunnel Fires Under Longitudinal Velocities. Underground Space, Vol. 6, No. 2, 2020, pp. 163–172. https://doi.org/10.1016/j.undsp.2019.11.001

13.

Kallianiotis

Papakonstantinou

Tolias

I. C.

, et al. Evaluation of Fire Smoke Control in Underground Space. Underground Space, Vol. 7, No. 3, 2022, pp. 295–310. https://doi.org/10.1016/j.undsp.2021.07.010

14.

Wang

, et al. Discussion on Rescue Facilities Planning in Long Railway Tunnels and Tunnel Groups. Chinese Journal of Underground Space and Engineering, Vol. 13, No. 2, 2017, pp. 567–572.

15.

Code for Design of Metro, GB50157-2013. China Construction Industry Press, 2013.

16.

Code for Design on Rescue Engineering for Disaster Prevention and Evacuation of Railway Tunnel, TB 10020-2017. China Railway Publishing House, 2017.

17.

PIARC Committee on Road Tunnels. Fire and Smoke Control in Road Tunnels. PIARC Committee on Road Tunnels, Paris, 2005.

18.

Carlsona

E.-S.

Kumm

Zakirov

, et al. Evacuation Tests With Elevated Platforms in Railway Tunnels. Fire Safety Journal, Vol. 108, No. 102840, 2019, pp. 1–10.

19.

Fridolf

Nilsson

Frantzich

Fire Evacuation in Underground Transportation Systems: A Review of Accidents and Empirical Research. Fire Technology, Vol 49, No. 2, 2013, pp. 451–475.

20.

Capote

J. A.

Alvea

Abreu

, et al. Analysis of Evacuation Procedures in High Speed Trains Fires. Fire Safety Journal, Vol. 49, No. 49, 2012, pp. 35–46.

21.

Yan

Wang

, et al. Determination of Cross Passage Spacing in Double-Track Tunnel for High Speed Railway under Random Parking Mode. China Railway Science, Vol. 38, No. 6, 2017, pp. 83–88.

22.

Zhang

Tan

Jin

Research on the Technology of Disaster Prevention and Rescue in High-altitude Super-long Railway Tunnel. KSCE Journal of Civil Engineering, Vol. 19, No. 3, 2015, pp. 756–764.

23.

Cao

Yang

Guo

Study on the Designated Disaster Prevention and Rescue in High-altitude Extra-long Railway Tunnel. Railway Standard Design, Vol. 58, No. 10, 2014, pp. 83–88.

24.

Han

Yang

Xia

, et al. Suggestions on Emergency Rescue’s Supplemental Equipment in Railway Tunnel Emergency Rescue Station. Earth and Environmental Science, Vol. 218, No. 012120, 2019, pp. 1–7.

25.

Chen

Research on Disaster Prevention and Rescue Techniques for the Extra-Long Guanjiao Tunnel. Modern Tunnelling Technology, Vol. 52, No. 6, 2015, pp. 14–19.

26.

Yan

Study on Design of Disaster Prevention and Rescue Evacuation System for Dayaoshan Tunnel Group of Wuhan-Guangzhou Passenger Dedicated Line. Journal of Railway Engineering Society, Vol. 8, 2011, pp. 7–12.

27.

Bosco

Lovreglio

Frassoldati

Derudi

Borghetti

Queue Formation and Evacuation Modelling in Road Tunnels During Fires. Chemical Engineering Transactions, Vol. 67, 2018, pp. 805–810.

28.

Di Graziano

Marchetta

Grande

Fiore

Application of a Decision Support Tool for the Risk Management of a Metro System. International Journal of Rail Transportation, Vol. 10, No. 3, 2022, pp. 352–374.

29.

Chen

Wang

Q.-A.

Liu

Research on Intelligent Monitoring of Fire Safety and Fire Rescue Plan for Tunnel Operation Under Quasi-Unattended Background. Buildings, Vol. 13, No. 8, 2023, p. 2110.

30.

Wang

Structural System Reliability Analysis Based on Improved Explicit Connectivity BNs. KSCE Journal of Civil Engineering, Vol. 22, No. 3, 2018, pp. 916–927.

31.

Wang

Study of Design Parameters of Evacuation Facility of Emergency Rescue Stations in Long Railway Tunnel. Tunnel Construction, Vol. 37, No. 3, 2017, pp. 285–290.

32.

Zhu

, et al. Study on Design Parameters and Reliability of Emergency Rescue Station of Long and Large Rack Rail Tunnel with Large Longitudinal Slope. Modern Tunnelling Technology, Vol. 58, 2021, pp. 479–485.

33.

Wang

Guo

, et al. Discussion on Structural Type Emergency Rescue Station Without Evacuation Condition Outside Tunnel. China Safety Science Journal, Vol. 31, No. 9, 2021, p. 184190.

34.

Cao

Xiong

, et al. Simulation Analysis of Safe Evacuation in a Railway Tunnel Emergency Rescue Station. IOP Conference Series: Earth and Environmental Science, Vol. 510, No. 5, 2020, p. 052020.

35.

Liao

Min

, et al. Research on the Characteristics and Regularity of Evacuation in Railway Tunnel Emergency Rescue Station at High Altitude. E3S Web of Conferences, Vol. 233, 2021, p. 03069.

36.

Beard

Carvel

The Handbook of Tunnel Fire Safety. London: Thomas Telford; 2005.

37.

Pelechano

Malkawi

Evacuation Simulation Models: Challenges in Modelling High Rise Building Evacuation with Cellular Automata Approaches. Automation in Construction, Vol. 17, 2008, pp. 377–385.

38.

Galea

E. R.

Perez Galparsoro

J. M.

Pearce

A Brief Description of the EXODUS Evacuation Model, Proceed. 18th International Conference on Fire Safety, Vol. 18, 1993, pp. 149–162.

39.

Ronchi

Testing the Predictive Capabilities of Evacuation Models for Tunnel Fire Safety Analysis. Safety Science, Vol. 59, 2013, pp. 141–153.

40.

Wang

Experimental Study on Evacuation under Fire Mode in Long and Large Railway Tunnel. China Railway Science, Vol. 36, No. 6, 2015, pp. 78–84.

41.

Chen

Xin

. Numerical Analysis of CO Diffusion Law in Tunnel Under Fire Conditions. Highway Transportation Science and Technology, Vol. 12, No. 12, 2016, pp. 107–110.

42.

Wang

Guo

Discussion on CO Concentration Control Standard of Fire Flue Gas in High-Altitude Railway Tunnel. Modern Tunnel Technology, Vol. 59, No. 3, 2022, pp. 40–45.

43.

Frantzich

Nilsson

Utrymninggenomtätrök: beteendeochförflyttning. Evacuation in Dense Smoke: Behaviour and Movement, Department of Fire Safety Engineering and Systems Safety, Lund University Report 3126, 75, 2003.

44.

Fridolf

Andree

Nilsson

Frantzich

The Impact of Smoke on Walking Speed. Fire Mater, Vol. 38, 2013, pp. 744–759. https://doi.org/10.1002/fam.2217

45.

Zhang

Xiang

Zhang

Tang

Experimental Study on Evacuation Behavior and Traffic Capacity of Highway Tunnel Crossing. Journal of Southwest Jiaotong University, Vol. 51, No. 4, 2016, pp. 615–620.

46.

Code of design on building fire protection and prevention: GB 50016,2014.

47.

Research on Reliability of Personnel Safe Evacuation Under High-speed Railway Super-long Tunnel in Fire. Doctoral Thesis. Southwest Jiaotong University, 2018.

48.

Wang

Structure Parameters of Portal Section of Shaft Type Emergency Exit in Railway Tunnel. China Railway Science, Vol. 36, No. 5, 2015, pp. 36–42.

49.

Frantzich

Nilsson

Utrymning Genom TätRök: Beteende Och Förflyttning. Department of Fire Safety Engineering, 2003, pp. 75–76.

50.

Tregenza

P. R.

The Design of Interior Circulation. Van Nostrand Reinhold Company, New York, NY, 1976.