Abstract

Utilizing existing natural gas pipelines to transport hydrogen-blended natural gas is a primary strategy for achieving cost-effective, long-distance, and large-scale hydrogen transportation. However, blending hydrogen with natural gas alters its physical properties, resulting in changes in leakage and diffusion characteristics and the affected range. To illustrate this, we focus on the Jingxi Third Line natural gas long-distance pipeline and develop a buried hydrogen blended natural gas pipeline model to analyze the concentration distribution of hydrogen-blended natural gas and the temporal variation of gas velocity at the leakage point. We explore the influence of various factors, including pressure, leak orifice size, wind speed, and hydrogen-blending ratio, on the diffusion range of hydrogen-blended natural gas. The research findings demonstrate that in the vicinity of the leakage point, the methane concentration significantly exceeds the upper explosive limit while the hydrogen concentration remains within the explosive limit range. The hazardous range of hydrogen-blended natural gas leakage and diffusion is slightly larger than that of natural gas alone. Furthermore, both the vertical and horizontal hazardous ranges of hydrogen-blended natural gas leakage and diffusion exhibit positive correlations with pressure and leak orifice size. Additionally, as wind speed increases, the maximum impact distance in the vertical direction gradually decreases, while it gradually increases in the horizontal direction.

1 Introduction

The central and western regions of China are rich in renewable energy resources. However, due to limitations in the “West-East Power Transmission” capacity, transmitting all the generated renewable energy to the eastern regions poses a challenge. Renewable energy generation is characterized by intermittency, volatility, unpredictability, and difficulties in regulating peak loads. The issue of curtailment and integration of renewable energy is urgently needed to be addressed [1]. To address these challenges, scholars have proposed utilizing renewable energy for hydrogen production and blending hydrogen into existing natural gas transmission systems. This approach enables low-cost, long-distance, and large-scale transportation of hydrogen, while also mitigating the waste issues associated with renewable energy [2]. In recent years, many countries and regions have actively conducted feasibility studies on the utilization of operational natural gas pipelines and related facilities for transporting hydrogen-blended natural gas [3,4]. These studies involve evaluating the suitability of hydrogen-blended natural gas for residential and commercial users, testing the effects of hydrogen blending ratios on key equipment, pipelines, and terminal devices in the pipeline system [5,6], and conducting safety research on hydrogen-blended natural gas transportation systems [79].

Hydrogen is both colorless and odorless, possessing a lower density in comparison to natural gas. It demonstrates a higher diffusion coefficient and exhibits a wider range of explosion limits. When hydrogen leaks in an environment rich in oxygen under high pressure conditions, its detection becomes challenging, and it becomes prone to self-ignition. In the event of leakage within confined or semiconfined spaces, the accumulation of combustible hydrogen may result in delayed ignition and the formation of hydrogen clouds, thereby giving rise to hydrogen cloud explosions and even detonations, posing a significant threat to pipeline safety [10,11]. Additionally, the introduction of hydrogen into pipelines and storage tanks can lead to hydrogen embrittlement [1214], further accentuating the hazards of combustion and explosions subsequent to gas leakage [15]. In recent years, hydrogen fuel has witnessed widespread adoption in Europe and the U.S. [36,16], prompting scholars to extensively investigate hydrogen and hydrogen-blended natural gas leakage and diffusion through the utilization of numerical simulation methods.

The Environmental Research Laboratory of the National Center for Scientific Research “Demokritos” in Greece [17] employed computational fluid dynamics (CFD) to simulate leaks and diffusion of hydrogen gas. They conducted a comparative analysis with the hydrogen accident that occurred in Stockholm in 1983, demonstrating the suitability of CFD for simulating hydrogen leakage accidents [18,19]. Wilkening and Baraldi [20] utilized CFD to investigate the characteristics of accidental leaks in natural gas and hydrogen gas pipelines. The results indicated that hydrogen, due to its density differences, is more affected by buoyancy compared to methane. Hydrogen exhibited significantly higher leakage rates, and hydrogen clouds were observed to travel greater distances from the ground or buildings. Su et al. [21] used ansysfluent software to simulate the process of leakage and they determined the diffusion status and affected range of natural gas and hydrogen gas following pipeline leakage and obtained the thermal radiation impact distance caused by combustion and explosion accidents diffusion of hydrogen-blended natural gas in a confined space. The findings revealed that, with a constant leakage rate, an increase in the hydrogen blending ratio reduced the time required to reach the lower explosive limit, thereby enhancing the combustibility and explosiveness of the leaked gas. Liu et al. [22] developed a model for simulating leaks and diffusion of high-pressure hydrogen gas and natural gas in pipelines. Through numerical simulations, they identified distinct leakage and diffusion characteristics between hydrogen gas and natural gas. The results indicated that the hazardous cloud formed by high-pressure hydrogen gas exhibited a larger and more concentrated shape, with a faster increase in the maximum diffusion height compared to natural gas. Additionally, the hazardous consequences of hydrogen gas leakage and diffusion near the ground were found to be smaller than those of natural gas. Zhao et al. [23] conducted a simulation analysis of pipeline leakage accidents with varying severity using Det Norske Veritas Process Hazard Analysis Software Tool (dnvphast). They assessed the diffusion status and extent of both natural and hydrogen gas after a pipeline leak and determined the distance of thermal radiation impact resulting from combustion and explosion accidents. Li et al. [24] simulated the leakage and diffusion process of hydrogen-blended natural gas at different proportions in an enclosed container. Their study revealed that, as the hydrogen blending ratio increased, the area and duration of the flammable zone gradually expanded. Furthermore, owing to its lower density, hydrogen gas exhibited more rapid diffusion into the lower section of the enclosed space. Jia et al. [25] conducted a simulation analysis to examine the impact of hydrogen blending ratio, wind speed, leak orifice diameter, and leak orifice direction on the leakage and diffusion of hydrogen-blended natural gas within a valve chamber. The study revealed that hydrogen-blended natural gas diffused in the direction of the leak orifice, following a sequence from the boundaries to the center of the valve chamber. When the hydrogen gas volume fraction remained below 15%, conventional combustible gas detectors were capable of detecting gas leakage before the hydrogen gas concentration reached the explosion limit. The Naturalhy project [26,27] investigated indoor leakage and accumulation of hydrogen-blended natural gas at the Advantica test site of Spadeadam company in Northern England. The study observed that the leakage and accumulation characteristics of hydrogen-blended natural gas were similar to those of natural gas. The concentration and volume of hydrogen-blended natural gas increased as the hydrogen gas content increased, albeit the increase was minimal when the hydrogen gas content was maintained below 50%. The Gas Technology Institute in the U.S. assessed the risks associated with the addition of hydrogen to natural gas pipelines at various levels [28]. The findings indicated that incorporating hydrogen into the natural gas pipeline would increase the risk of leakage, although the increase in leakage risk was relatively modest when the hydrogen blending ratio remained below 20%. Hormaza Mejia et al. [29] investigated the leakage rates of hydrogen and natural gas in low-pressure infrastructure. The study determined that in typical low-pressure gas infrastructure, the leakage rates of hydrogen and natural gas were similar or comparable, and proposed measures for reducing gas leakage. Zhu et al. [30] devised and constructed a large-scale experimental system to simulate the diffusion behavior and concentration distribution of small-hole leaks in high-pressure hydrogen-blended natural gas buried pipelines. This research contributed to the provision of technical support for risk assessment and emergency response in the transportation of hydrogen-blended natural gas through buried pipelines. Drexler et al. conducted cyclic loading tests on three pipeline steels at pressures of 5.5 MPa and 34 MPa. The results indicated that the fatigue crack growth rates of these steels in a hydrogen environment were significantly higher than those in air, and this effect increased with pressure [31]. Amaro et al. established a phenomenological model to describe the fatigue crack growth behavior of different steels under various hydrogen pressures. When comparing their model with experimental data, they found the error to be within 10% [32].

Presently, researchers have utilized numerical simulation methods to compare and investigate the outcomes of hydrogen and natural gas pipeline leakage accidents. However, due to the high leakage rates of hydrogen and natural gas in pipelines, which approach the local speed of sound, and the intricate establishment of soil models, the impact of soil cover above the pipeline on the leakage process is often disregarded during the research, resulting in an overestimation of simulation results. In this study, the CFD method was employed to numerically simulate the process of leakage and diffusion in buried pipelines containing hydrogen-blended natural gas. The analysis involved examining the concentration distribution of hydrogen-blended natural gas across soil and atmosphere, and the temporal variation of gas velocity at the leakage point. The dnvphast software was employed to develop a model for leakage and diffusion in buried pipelines carrying hydrogen-blended natural gas, allowing for a quantitative investigation of the influence of factors such as pressure, leakage orifice size, wind speed, and hydrogen blending ratio on the diffusion range of hydrogen-blended natural gas. The research findings hold significant importance in comprehending the characteristics of leakage and diffusion in hydrogen-blended natural gas pipelines under soil-atmosphere coupling conditions. Moreover, they provide insights for monitoring, early warning systems, and personnel evacuation in the event of leakage incidents.

2 Physical Model and Numerical Methods

In this study, we focuses on the Jingxi Third Line natural gas transmission system to investigate the safety issues associated with transporting hydrogen-blended natural gas using an existing natural gas pipeline, with Fig. 1 illustrating the pipeline's route. The Jingxi Third Line consists of two parts: the main line from Jingbian Compressor Station to Yongle Distribution Station and the Zhidan branch line. The total length of the main line and the branch line is approximately 575.5 km and 50.5 km, respectively. The entire pipeline includes six main stations, including the Jingbian Compressor Station. The pipeline is designed with a pressure of 8.0 MPa and an average burial depth of 1.5 m. The total designed gas transportation capacity is 3300 × 104 m3/d. The hydrogen blending is implemented at the Anbian Compressor Station, with hydrogen blending ratios ranging from 0% to 20%.

Fig. 1
Jingxi Third Line trend map
Fig. 1
Jingxi Third Line trend map
Close modal

2.1 Physical Model.

The Jingxi Third Line natural gas long-distance pipeline has a design pressure of 8 MPa and a pipe diameter of DN900. The operational data of the pipeline are presented in Table 1.

Table 1

Pipeline operation data

Diameter (mm)Pressure (MPa)Temperature (°C)Minimum annual output (104 Nm3/d)Maximum annual output (104 Nm3/d)
DN9004–7.632–553701428
Diameter (mm)Pressure (MPa)Temperature (°C)Minimum annual output (104 Nm3/d)Maximum annual output (104 Nm3/d)
DN9004–7.632–553701428

Taking the Anbian Compressor Station inlet pipeline of the Jingxi Third Line natural gas long-distance pipeline as a case study, a physical model for buried hydrogen-blended natural gas pipelines is established. The diffusion process of hydrogen-blended natural gas in the soil-atmosphere system is analyzed. The simulation conditions are set as follows: soil depth of 1.5 m, dimensions of the atmospheric computational domain: length × height = 200 m × 350 m, pipeline diameter of 900 mm, pipeline length of 200 m, circular leakage orifice located at the 12 O'clock position of the pipe cross section, with a diameter of 50 mm, and a vertical upward leakage direction. The center of the leakage orifice is 100 m away from the origin. The pressure is 4 MPa, and the transported medium is natural gas blended with 10% hydrogen. The physical model for the leakage and dispersion of hydrogen-blended natural gas pipelines under soil-atmosphere coupling is shown in Fig. 2.

Fig. 2
Physical model of leakage and diffusion in buried hydrogen-blended natural gas pipeline
Fig. 2
Physical model of leakage and diffusion in buried hydrogen-blended natural gas pipeline
Close modal

2.2 Mesh Generation.

Given the relatively large scale of our physical model, using a two-dimensional model allows us to reduce computational workload while maintaining accuracy. Relevant research has indicated that terrain fluctuations have almost no impact on diffusion [20]. Diffusion remains consistent across each two-dimensional cross section in space, suggesting that considering gas diffusion only in the horizontal and vertical directions is feasible [33,34].

The partitioning of the computational domain into grids using ICEM CFD 19.2 is a crucial step in conducting numerical simulations in the soil-atmosphere environment. To ensure the accuracy of the simulation results, it is necessary to refine the grids vertically above the leakage port and horizontally near the ground surface. Additionally, we incorporated five boundary layers at the soil-atmosphere coupling interface. As the hydrogen-blended natural gas disperses outward, the flow pattern gradually decelerates. While still meeting engineering standards, the grids in the outer computational domain can be progressively coarsened to enhance the efficiency of the solution process, as depicted in Fig. 3. The average skewness within the computational domain grids is 0.12, the aspect ratio is 0.88, and the orthogonal quality is 0.87. Furthermore, the mesh growth rate in the leak orifice area is set to 1.1.

Fig. 3

2.3 Boundary Conditions

  1. Leakage orifice conditions

    In the hypothetical scenario of a hydrogen-blended natural gas pipeline leak, we assume an adiabatic process. The leakage orifice is subjected to a pressure inlet boundary condition, set to match the operational pressure of the pipeline. The leakage occurs in a vertically upward direction. At the point of leakage, the volume fractions of methane and hydrogen are specified.

  2. Soil domain conditions

    Following the leakage in the pipeline, the hydrogen-blended natural gas first disperses within the adjacent soil. The interface between the soil and the atmosphere is designated as the soil-atmosphere coupling boundary. The soil layer is treated as an isotropic porous medium, filled with air. The local soil type is sandy loam, with an average summer soil temperature of 19.5 °C and an average winter soil temperature of 11.2 °C. The soil density is 2034 kg/m3, thermal conductivity is 1.472 W/(m K), and the soil moisture content is 6%. The corresponding soil resistance coefficients are provided in Table 2.

Table 2

Soil resistance coefficients

Soil typeAverage particle diameter (mm)PorosityViscous drag coefficient (m–2)Inertial drag coefficient (m–2)
Sandy loam0.500.252.16 × 10103.36 × 105
Soil typeAverage particle diameter (mm)PorosityViscous drag coefficient (m–2)Inertial drag coefficient (m–2)
Sandy loam0.500.252.16 × 10103.36 × 105
  1. Atmospheric domain conditions

    In the computational domain representing the atmosphere, a pressure outlet boundary condition is implemented. In the presence of wind, the left side is specified as a velocity inlet, as outlined in Table 3. At the initial time, the volume fractions of methane and hydrogen within the atmospheric domain are initialized to 0, while the volume fraction of air is set to 1. The pressure is fixed at 101.325 kPa, and the temperature is set to 20 °C.

Table 3

Model boundary condition settings

PositionBoundary nameBoundary type
Left side of computational domain (windy)WindVelocity-inlet
Left side of computational domain (no windy)Outlet-leftPressure-out
Right side of computational domain (no windy)Outlet-rightPressure-out
Top of computational domain (no wind)Outlet-upPressure-out
SurfaceGroundWall
Pipeline leakage orificeInletPressure-inlet
Leakage spaceAirspaceFluid
SoilSoilFluid-porous zone
PositionBoundary nameBoundary type
Left side of computational domain (windy)WindVelocity-inlet
Left side of computational domain (no windy)Outlet-leftPressure-out
Right side of computational domain (no windy)Outlet-rightPressure-out
Top of computational domain (no wind)Outlet-upPressure-out
SurfaceGroundWall
Pipeline leakage orificeInletPressure-inlet
Leakage spaceAirspaceFluid
SoilSoilFluid-porous zone

2.4 Mathematical Model and Governing Equation.

The pressure of long-distance hydrogen blended natural gas pipeline is very high, a leak in the pipeline can lead to the rapid diffusion of gas at the leak orifice, forming a jet, so gas pipeline leakage can be regarded as multicomponent gas turbulence. The diffusion process of this gas in both soil and air adheres to the principles of mass conservation, momentum conservation, species transport, and turbulent flow.

Compared to other turbulence models, the standard kε model offers good stability, cost-effectiveness, and computational accuracy. It is particularly suitable for large-scale model computations and can provide accurate predictions of the flow and jets surrounding a plane [17,35,36]. Research has revealed that standard kε model has better accuracy in its ability to calculate the diffusion of hydrogen and methane during the leakage process, and the result is consistent with experimental data [24,37].

Consequently, we have chosen to employ the standard kε model for simulating the leak and diffusion of hydrogen-blended natural gas. The expressions for standard kε model are as follows.

The continuity equation is
ρt+(ρui)xi=0
(1)
The momentum equation is
(ρui)t+(ρuiuj)xj=pxi+τijxj+Fi
(2)
The species transport equation is
(ρcs)t+xi(ρuics)=xi(Dairsρcsxi)
(3)

where ρ is the fluid density; t is the time; ui is the velocity in i direction, i=x,y,z; τij is the shear stress; Fi is the mass force in i direction; p is the pressure; s denotes gas species; cs is the volume concentration of species s; Dairs is the diffusion coefficient of species s in the air. According to the literature [10] Dairs=7.6×105 m2/s and DairCH4=1.6×105 m2/s.

The standard kε turbulent model with wall functions is applied in this study [38], which includes the turbulent kinetic energy equation and the turbulent dissipation rate equation
t(ρk)+xi(ρkui)=xi[(ui+utσk)kxi]+GK+GbρεYM+Sk
(4)
t(ρε)+xi(ρεui)=xj[(ui+μtσε)εxj]+C1εεk(GK+C3εGb)C2ερε2k+Sε
(5)

where Gk is the generation of turbulent kinetic energy due to the mean velocity gradients; Gb is the generation of turbulent kinetic energy due to buoyancy; YM is the contribution of fluctuating dilatation in compressible turbulence to the overall dissipation rate; the constants C1ε=1.44, C2ε=1.92, σk=1.0, and σε=1.3.

2.4.1 Species Transport Model and Material Definition

  1. Species transport model

    The diffusion process of hydrogen-blended natural gas in both soil and air can be regarded as a process of movement. In order to accurately simulate this phenomenon, a species transport model that does not account for chemical reactions is chosen, taking into consideration the impact of buoyancy effects.

  2. Material definition

    In order to streamline the calculation, the analysis focuses solely on methane and hydrogen as the considered components. These materials are defined as a mixture comprising methane, hydrogen, and air. The influence of buoyancy is taken into account for the flow of this mixture within the computational domain. At the initial time, the atmospheric computational domain is initially filled with air.

2.4.2 Convergence Judgment.

The diffusion process of hydrogen-blended natural gas leakage in the pipeline is characterized by a nonsteady-state behavior. To solve the equations, the PISO algorithm is selected. A time-step of 0.1 s is set, with 300 iterations performed for each time-step. Through iterative experimentation, it has been determined that convergence can be achieved within 300 iterations. The convergence criterion is set at an iteration accuracy of 10−3.

When conducting simulations using ansysfluent, there are primarily two approaches to assess convergence:

  1. Difference in mass flow rate between the inlet and outlet is less than 0.2%;

  2. Target values exhibit no further changes.

In this study, the second method is employed to evaluate convergence, with the concentrations of methane and hydrogen serving as the target variables. The default value of 0.001 is commonly used for the convergence precision.

2.4.3 Grid Independence.

The physical model is divided into a structured grid using ICEM CFD 19.2. Four different grid number are employed: 201,458; 425,136; 1,001,586; and 1,984,254 grids. The grid quality for all cases surpasses 0.8. To assess the grid quality, the variation of the diffusion height of hydrogen-blended natural gas leakage over time is examined. The results indicate that, under the same leakage conditions, the gas diffusion range exhibits minimal variation when the grid quantity increases from 200,000 to 2,000,000, as depicted in Fig. 4. Hence, it can be concluded that a grid size of 200,000 is sufficient to fulfill the requirements.

Fig. 4
Grid independence verification
Fig. 4
Grid independence verification
Close modal

3 Leakage and Diffusion Characteristics

3.1 Concentration Distribution.

Following the release of hydrogen-blended natural gas from the pipeline, the gas initially disperses into the soil layer. The diffusion rate is relatively slow due to the resistance of the soil. The concentration distribution of hydrogen-blended natural gas in the soil at different leakage times of 5 s, 8 s, 10 s, and 15 s is illustrated in Figs. 5 and 6.

Fig. 5
Contour plots of methane concentration distribution at different moments
Fig. 5
Contour plots of methane concentration distribution at different moments
Close modal
Fig. 6
Contour plots of hydrogen concentration distribution at different moments
Fig. 6
Contour plots of hydrogen concentration distribution at different moments
Close modal

Based on the analysis of Figs. 5 and 6, it is evident that the concentrations of methane and hydrogen exhibit symmetric distributions within the soil domain above the leakage orifice due to the isotropic nature of the porous soil medium. The pattern of leakage and accumulation of hydrogen-blended natural gas resembles that of natural gas, without any discernible separation of hydrogen from the mixture. During the initial phase of leakage, the diffusion of hydrogen-blended natural gas encounters significant resistance from gravity, soil viscous resistance, inertial resistance, and capillary pressure. This resistance is considerable, leading to substantial losses in kinetic energy and a significant decrease in diffusion velocity as it reaches the upper boundary of the soil. Consequently, the gas does not jet vertically upward but diffuses outward in a “semicircular” pattern. After a certain duration of leakage, the concentration of hydrogen-blended natural gas in the soil reaches the lower explosive limit, and the hazardous range gradually expands. At 10 s, influenced by surface tension, the hydrogen-blended natural gas overflows from the soil layer, rapidly accumulating a substantial amount of combustible gas near the leakage point within a short period.

Beyond 10 s, the hydrogen-blended natural gas permeates through the soil and enters the atmosphere. With reduced air resistance and buoyancy force surpassing gravity, the diffusion velocity increases rapidly. The concentration distribution of hydrogen-blended natural gas in the atmosphere is illustrated in Figs. 7 and 8.

Fig. 7
Distribution nephogram of methane component diffusion concentration
Fig. 7
Distribution nephogram of methane component diffusion concentration
Close modal
Fig. 8
Distribution nephogram of hydrogen component diffusion concentration
Fig. 8
Distribution nephogram of hydrogen component diffusion concentration
Close modal

Within the time range of 10–50 s, a significant pressure gradient exists between the interior and exterior of the pipeline, resulting in a rapid leakage rate of hydrogen-blended natural gas. Influenced by the buoyancy of the surrounding air, the gas ascends swiftly, forming a concentrated cloud in the atmosphere. The concentration distribution of the hydrogen-blended natural gas leakage at 50 s is depicted in Fig. 9, indicating a symmetrical mushroom-shaped pattern of the cloud around the central axis of the leakage orifice. Despite the continuous gas supply from the ground, the leakage rate from the pipeline gradually diminishes. Consequently, there is a noticeable concentration difference between the gas cloud and the surrounding air, leading to the rapid diffusion of the cloud into the air while it ascends. As the diffusion time surpasses 50 s, the internal pipeline pressure decreases, resulting in a slower leakage rate. The cloud in the atmosphere gradually expands to the surrounding area, with only a high concentration of hydrogen-blended natural gas remaining above the leakage orifice.

Fig. 9
Contour plots of diffusion concentration for hydrogen-blended natural gas leakage
Fig. 9
Contour plots of diffusion concentration for hydrogen-blended natural gas leakage
Close modal

Based on the observation from Fig. 9, it is evident that following the leakage of hydrogen-blended natural gas from the pipeline, the concentration of methane reaches the explosive limit at a specific diffusion distance. At the leakage orifice, the hydrogen concentration falls within the range of 4–10%, which is within the explosive limit range. However, the methane concentration exceeds the upper explosive limit of 15%. The explosive limit range of the hydrogen component can extend beyond the explosive limits of the methane component, thereby elevating the risk of explosion at the leakage orifice.

3.2 Temporal Variation of Gas Flow Velocity.

The temporal variation curve of the gas flow velocity at the leakage orifice of the hydrogen-blended natural gas pipeline is presented in Fig. 10. During a leakage event in the hydrogen-blended natural gas pipeline, the gas flow velocity at the leakage orifice initially experiences a rapid increase, reaching a maximum value of 48 m/s. Subsequently, it swiftly decreases to 11.6 m/s, displaying minor fluctuations within a narrow range and gradually declining until the gas leakage within the pipeline ceases.

Fig. 10
Temporal variation of gas flow velocity at leakage orifice: (a) gas flow velocity at leakage orifice in the first 10 s and (b) gas flow velocity at leakage orifice within 400 s
Fig. 10
Temporal variation of gas flow velocity at leakage orifice: (a) gas flow velocity at leakage orifice in the first 10 s and (b) gas flow velocity at leakage orifice within 400 s
Close modal
  • Gas flow velocity at leakage orifice in the first 10 s.

  • Gas flow velocity at leakage orifice within 400 s.

3.3 Temporal Variation of Gas Volume Fraction.

Two monitoring points were established to track the volume fraction of hydrogen-blended natural gas. Monitoring Point 1, located at coordinates (100 m, 1 m), is positioned in the atmospheric region 1 m above the leakage point. The temporal variation of the volume fraction of hydrogen-blended natural gas at Monitoring Point 1 is displayed in Fig. 11. During the initial 10 s, the diffusion of hydrogen-blended natural gas occurs primarily in the soil, resulting in negligible concentrations of methane and hydrogen at Monitoring Point 1. However, after 10 s, the hydrogen-blended natural gas diffuses from the soil into the atmosphere, leading to a rapid rise in the volume fraction of methane and hydrogen at Monitoring Point 1, followed by stabilization. The maximum methane concentration recorded at Monitoring Point 1 reaches 68.7%, while the maximum hydrogen concentration approximates 7%.

Fig. 11
Temporal variation of the volume fraction at Monitoring Point 1
Fig. 11
Temporal variation of the volume fraction at Monitoring Point 1
Close modal

Monitoring Point 2, positioned at coordinates (101 m, 1 m), is located in the atmospheric region 1 m above the leakage point at an inclined angle. The variation in the volume fraction of hydrogen-blended natural gas at Monitoring Point 2 is illustrated in Fig. 12. Distinct peaks in CH4 and H2 volume fractions were observed within the time intervals of 7–13 s, 16–32 s, and 46–69 s. The temporal trends of CH4 and H2 volume fractions were generally consistent, with CH4 reaching a maximum concentration of 8.5% and H2 reaching a maximum concentration of 0.94%.

Fig. 12
Temporal variation of the volume fraction at Monitoring Point 2
Fig. 12
Temporal variation of the volume fraction at Monitoring Point 2
Close modal

Comparing the variation patterns of hydrogen blended natural gas volume fractions at Monitoring Point 1 and Monitoring Point 2, we observed that after 10 s, the values at Monitoring Point 1 tended to stabilize, whereas the values at Monitoring Point 2 exhibited more fluctuations and only gradually stabilized after 80 s. The reason for this difference can be attributed to the following factors: Monitoring Point 1 is located 1 m directly above the leak orifice, and the variations in the volume fractions of hydrogen blended natural gas are directly influenced by the gas flow velocity at the leak orifice. Monitoring Point 2 is situated 1 m to the right of Monitoring Point 1, and the volume fractions of hydrogen blended natural gas at Monitoring Point 2 are also affected by the gas flow velocity at the leak orifice. However, due to the presence of a wind from the left in the atmospheric domain, the values at Monitoring Point 2 are also influenced by the gas concentration at Monitoring Point 1.

4 Leakage and Diffusion Range

Given the substantial computational demands associated with simulating the diffusion of hydrogen-blended natural gas using ansysfluent, it is only practical to monitor the concentration of methane or hydrogen as separate components rather than the explosive limits of the mixture. Furthermore, the time-dependent nature of the leakage process presents challenges in determining the maximum diffusion range of hydrogen-blended natural gas using ansysfluent. To address these limitations, the dnvphast software is utilized to evaluate the extent of hazard resulting from the diffusion of hydrogen-blended natural gas. Influential factors and values of hydrogen blending natural gas diffusion range are provided in Table 4.

Table 4

Influential factors and values of hydrogen blending natural gas diffusion range

Variable namePressure (MPa)Orifice diameter (mm)Wind speed (m/s)Hydrogen blending ratio (%)
Variable value4, 5, 6, 7, 85, 25, 50, 80, 100, 150, 200, 4002, 4, 6, 8, 100, 5, 10, 15, 20
Variable namePressure (MPa)Orifice diameter (mm)Wind speed (m/s)Hydrogen blending ratio (%)
Variable value4, 5, 6, 7, 85, 25, 50, 80, 100, 150, 200, 4002, 4, 6, 8, 100, 5, 10, 15, 20

4.1 Influence of Pipeline Pressure on the Diffusion Range.

By considering a middle orifice leakage with a diameter of 25 mm, the impact of operating pressure on the maximum hazardous range of hydrogen-blended natural gas following a leak is examined, with the lower explosive limit of the gas as the criterion. The study considers a pressure range of 4–8 MPa, hydrogen-blending ratios ranging from 0% to 20%, a temperature of 25 °C, a wind speed of 1.5 m/s, and a burial depth of 1.5 m. The mass flowrate of hydrogen-blended natural gas leakage at different pressures is depicted in Fig. 13. Furthermore, the hazardous range of leakage diffusion is presented in Table 5 and Fig. 14.

Fig. 13
The mass flowrate at different pressures
Fig. 13
The mass flowrate at different pressures
Close modal
Fig. 14
Maximum diffusion distance at different pressure: (a) horizontal direction and (b) vertical direction
Fig. 14
Maximum diffusion distance at different pressure: (a) horizontal direction and (b) vertical direction
Close modal
Table 5

Dangerous diffusion range of hydrogen-blended natural gas under different pressures

Pressure (MPa)Horizontal hazard range (m)Vertical danger range (m)
44.88–5.614.46–15.31
55.58–6.3615.61–16.49
66.29–7.0817.25–17.85
76.98–7.8218.23–19.33
87.71–8.519.45–20.27
Pressure (MPa)Horizontal hazard range (m)Vertical danger range (m)
44.88–5.614.46–15.31
55.58–6.3615.61–16.49
66.29–7.0817.25–17.85
76.98–7.8218.23–19.33
87.71–8.519.45–20.27

Under identical conditions, the density of natural gas decreases when blended with hydrogen. Consequently, when a pipeline experiences a leak, the mass flowrate of natural gas surpasses that of hydrogen-blended natural gas. For leaks of the same aperture size, the mass flowrate shows a positive correlation with pressure. The introduction of hydrogen into natural gas results in a decrease in density. Hence, under equivalent aperture size and pressure, the mass flowrate decreases as the hydrogen-blending ratio increases.

  • Horizontal direction

  • Vertical direction

Based on the findings depicted in Fig. 14, it is evident that the diffusion range of hydrogen-blended natural gas varies under different pressures during a leakage event. Due to the faster diffusion rate of hydrogen in the air compared to methane, the diffusion of hydrogen-blended natural gas accelerates following a leak. Within the range of up to 20% hydrogen blending, the distances at which the concentration of hydrogen-blended natural gas reaches the explosion limit exhibit slight increases in both the vertical and horizontal directions. Specifically, when studying the leakage from a high-pressure natural gas pipeline with 20% hydrogen blending, the vertical hazardous distance expanded by approximately 0.75 m, representing an increase of approximately 10%, while the horizontal hazardous distance increased by approximately 0.85 m, indicating a 5% increase.

4.2 Influence of Leakage Orifice Diameter on the Diffusion Range.

Using the Anbian Compressor Station's Inlet Pipeline as a case study, this research investigates the maximum distances of impact in both the vertical and horizontal directions for the diffusion of hydrogen-blended natural gas under various leakage orifice diameters. The pipeline, with a diameter of DN900, operates at a pressure of 4 MPa and a temperature of 25 °C. The leakage orifice diameters range from 5 to 400 mm, with a wind speed of 1.5 m/s and a burial depth of 1.5 m. The hydrogen-blending ratios range from 0% to 20%. The hazardous diffusion range of hydrogen-blended natural gas is provided in Table 6.

Table 6

The diffusion range of hydrogen-blended natural gas at different leakage orifice diameter

Leakage orifice diameter (mm)Horizontal hazard range (m)Vertical danger range (m)
51.84–2.17.21–7.4
254.88–5.614.46–15.31
5010.79–12.225.44–27.05
8018.8–20.7534.42–38.98
10024.34–26.5344.59–47.31
15039.63–41.7861.28–64.99
20055.85–57.4776.8–81.39
40080.58–85.43225.67–242.66
Leakage orifice diameter (mm)Horizontal hazard range (m)Vertical danger range (m)
51.84–2.17.21–7.4
254.88–5.614.46–15.31
5010.79–12.225.44–27.05
8018.8–20.7534.42–38.98
10024.34–26.5344.59–47.31
15039.63–41.7861.28–64.99
20055.85–57.4776.8–81.39
40080.58–85.43225.67–242.66

The leakage mass flowrate of hydrogen-blended natural gas pipelines, with leakage orifice diameters ranging from 5 to 400 mm, is illustrated in Fig. 15. When natural gas is blended with hydrogen up to 20%, there is a slight decrease in the leakage mass flowrate. This decrease is mainly influenced by the diameter of the leakage orifice and the leakage mode. For leakage orifice diameters smaller than 150 mm, continuous leakage is observed, and the leakage mass flowrate for hydrogen-blended natural gas remains relatively constant. On the other hand, for orifice diameters of 200 mm and 400 mm, instantaneous leakage occurs, and the maximum flowrate values are plotted.

Fig. 15
Leakage mass flowrate at different leakage orifice diameter
Fig. 15
Leakage mass flowrate at different leakage orifice diameter
Close modal

Figure 16 illustrates the maximum horizontal and vertical diffusion distances of hydrogen-blended natural gas leakage for different leakage orifice sizes ranging from 5 mm to 400 mm. The graph shows that the diffusion range of hydrogen-blended natural gas varies depending on the size of the leakage orifice. For smaller orifice sizes, the vertical diffusion range of the hydrogen-blended natural gas pipeline falls between 1.84 m and 2.1 m, while the horizontal diffusion range lies between 7.21 m and 7.4 m. As the leakage orifice size increases, the maximum diffusion distance of the hydrogen-blended natural gas leakage grows exponentially. In the case of pipeline rupture, the vertical diffusion range extends from 80.58 m to 85.43 m, while the horizontal diffusion range ranges from 225.67 m to 242.66 m.

Fig. 16
Maximum diffusion distance at different leakage orifice diameter: (a) horizontal direction and (b) vertical direction
Fig. 16
Maximum diffusion distance at different leakage orifice diameter: (a) horizontal direction and (b) vertical direction
Close modal
  • Horizontal direction

  • Vertical direction

4.3 Influence of Wind Speed on the Diffusion Range.

Examining the scenario of an intermediary orifice leakage (d = 25 mm), this study explores the maximum diffusion range in the vertical and horizontal axes for a hydrogen-blended natural gas under various wind velocities. The parameters taken into consideration are: a pressure of 4 MPa, a temperature of 25 °C, wind velocities ranging from 2 to 10 m/s, a burial depth of 1.5 m, and a hydrogen-blending ratio varying from 0% to 20%.

The mass flowrate of the leaking hydrogen-blended natural gas is depicted in Fig. 17. Following the introduction of hydrogen into natural gas, a marginal reduction in density is observed. The leakage mass flowrate exhibits a direct linear relationship, decreasing progressively with an increasing ratio of hydrogen blend. For instance, a gas blend with 5% hydrogen exhibits a leakage mass flowrate of 51.5 kg/s, whereas a blend with 20% hydrogen presents a leakage mass flowrate of 46 kg/s. Figure 18 provides a visual representation of the diffusion range of the hydrogen-blended natural gas at disparate wind velocities, given a hydrogen-blending ratio of 10%.

Fig. 17
Leakage mass flowrate at different hydrogen-blending ratios
Fig. 17
Leakage mass flowrate at different hydrogen-blending ratios
Close modal
Fig. 18
Diffusion range at different wind speeds (hydrogen-blending ratio 10%)
Fig. 18
Diffusion range at different wind speeds (hydrogen-blending ratio 10%)
Close modal

Conducting the study at a pressure of 4 MPa and with a leakage orifice diameter of 25 mm, the diffusion scope of the hydrogen-blend natural gas escaping from a pipeline was examined at varying wind speeds, spanning from 2 to 10 m/s. The perilous diffusion extent of the leaked hydrogen-blend natural gas is presented in Table 7. Meanwhile, Fig. 19 graphically represents the maximum impact distances in both the horizontal and vertical axes.

Fig. 19
Maximum diffusion distances at different wind speeds: (a) horizontal direction and (b) vertical direction
Fig. 19
Maximum diffusion distances at different wind speeds: (a) horizontal direction and (b) vertical direction
Close modal
Table 7

Dangerous diffusion range of hydrogen-blended natural gas under different wind speeds

Wind speed (m/s)Horizontal hazard range (m)Vertical danger range (m)
26.51–7.4214.9–15.71
47.77–8.8410.44–10.97
68.43–9.548.38–8.77
88.67–9.797.16–7.46
108.81–10.016.31–6.58
Wind speed (m/s)Horizontal hazard range (m)Vertical danger range (m)
26.51–7.4214.9–15.71
47.77–8.8410.44–10.97
68.43–9.548.38–8.77
88.67–9.797.16–7.46
108.81–10.016.31–6.58
  • Horizontal direction

  • Vertical direction

In the vicinity of the leakage orifice, the hydrogen-blended natural gas demonstrates an elevated initial leakage velocity, whereas the impact of wind speed on the direction of the gas jet is comparatively mild. As the height of the jet escalates, the velocity of the gas diminishes due to the effects of gravity and air resistance. Conversely, the lateral wind expedites the downward diffusion of the hydrogen-blended natural gas. As per Fig. 19, at a wind velocity of 2 m/s, the vertical diffusion distance of the gas reaches its peak, while the horizontal diffusion distance remains minimal. With increasing wind speeds, the maximum vertical diffusion distance shrinks, while the maximum horizontal diffusion distance expands. At a wind velocity of 10 m/s, the minimum vertical diffusion distance falls within the range of 6.31–6.58 m, while the maximum horizontal diffusion distance lies between 8.81 and 10.01 m.

5 Conclusion

This study used simulations to examine the diffusion process of hydrogen-blended natural gas in soil and air. It scrutinized the gas's concentration distribution in the soil and air over varying time periods, and the fluctuation of gas velocity at the leakage point. The investigation considered elements such as operating pressure, size of the leakage orifice, and environmental wind speed to evaluate their influence on the dangerous range of hydrogen-blended natural gas diffusion. Consequently, the research yielded the following conclusions:

  1. Upon the occurrence of a leakage in a hydrogen-blended natural gas pipeline, the methane concentration hits the explosive limit at a particular distance. Near the leakage point, the methane concentration significantly exceeds the upper explosive limit of 15%, while the hydrogen concentration stays within the explosive limit range. The hydrogen component compensates in areas where the methane component is below the explosive limit. If a leakage incident transpires during the transportation of hydrogen-blended natural gas via a natural gas pipeline, it heightens the explosion risk at the leakage point.

  2. Given hydrogen's more rapid diffusion rate in air compared to natural gas, the hazardous range of hydrogen-blended natural gas leakage diffusion slightly augments relative to pure natural gas. Increasing the hydrogen blending ratio leads to an expansion of the hazardous areas in both the vertical and horizontal directions.

  3. The vertical and horizontal hazardous ranges of hydrogen-blended natural gas leakage diffusion show a positive correlation with pressure and the size of the leakage orifice. Influenced by crosswinds, the maximum vertical impact distance reduces incrementally as wind speed intensifies, whereas the maximum horizontal impact distance shows a gradual increase.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding Data

  • “Research on the Impact of Hydrogen-blended Natural Gas on Downstream Users and Terminal Facilities” project of China Petroleum Engineering and Construction Corporation (Grant No. XN23D/JS/WX147).

Data Availability Statement

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.

References

1.
Administration, N. E.
,
2020
, “
Transcript of National Energy Administration's Online Press Conference in the First Quarter of 2020
,” National Energy Administration, Beijing, China, accessed Aug. 2023, http://www.nea.gov.cn/2020-03/06/c_138850234.htm
2.
Jin
,
X.
,
Zhuang
,
Y.
, and
Wang
,
H.
,
2019
, “
Feasibility Analysis Research on Abandoning Wind and Solar Energy With Hydrogen Energy Storage Technology
,”
Electrotechnics Electr.
, 39(
4
), pp.
63
68
.
3.
Deymi-Dashtebayaz
,
M.
,
Ebrahimi-Moghadam
,
A.
,
Pishbin
,
S. I.
, and
Pourramezan
,
M.
,
2019
, “
Investigating the Effect of Hydrogen Injection on Natural Gas Thermo-Physical Properties With Various Compositions
,”
Energy
,
167
, pp.
235
245
.10.1016/j.energy.2018.10.186
4.
Yıldırım
,
A.
,
2011
, “
NATURALHY (The Potential of Existing Natural Gas Network for Hydrogen Delivery) Project Objectives and Some Results
,”
Proceedings of the 18th International Petroleum and Natural Gas Congress and Exhibition of Turkey
, Ankara, Turkey, May 11–13, p.
cp-377
.10.3997/2214-4609-PDB.377.155
5.
Anonymous
,
2014
, “
GRHYD Project to Demo Solid-State H2 Storage
,”
Mod. Power Syst.
,
34
(
2
), p.
39
.
6.
Patel
,
S.
,
2020
, “
WindGas Falkenhagen: Pioneering ‘Green’ Gas Production
,”
Power
,
164
(
9
), pp.
30
31
.
7.
Farzadi
,
A.
,
2015
, “
Gas Pipeline Failure Caused by In-Service Welding
,”
ASME J. Pressure Vessel Technol.
,
138
(
1
), p.
011405
.10.1115/1.4031443
8.
Zheng
,
J.
,
Li
,
L.
,
Chen
,
R.
,
Xu
,
P.
, and
Kai
,
F.
,
2008
, “
High Pressure Steel Storage Vessels Used in Hydrogen Refueling Station
,”
ASME J. Pressure Vessel Technol.
,
130
(
1
), p.
014503
.10.1115/1.2826453
9.
Xie
,
P.
,
Wu
,
Y.
,
Li
,
C.
, and
Jia
,
W.
,
2021
, “
Research Progress on Pipeline Transportation Technology of Hydrogen-Mixed Natural Gas
,”
Oil Gas Storage Transp.
,
40
(
4
), pp.
361
370
.
10.
Li
,
J.
,
Su
,
Y.
,
Zhang
,
H.
, and
Yu
,
B.
,
2021
, “
Research Progresses on Pipeline Transportation of Hydrogen-Blended Natural Gas
,”
Nat. Gas Ind.
,
41
(
4
), pp.
137
152
.
11.
Shang
,
J.
,
Lu
,
Y.
,
Zheng
,
J.
, and
Sun
,
C.
,
2021
, “
Research Status-In-Situ and Key Challenges in Pipeline Transportation of Hydrogen-Natural Gas Mixtures
,”
Chem. Ind. Eng. Prog.
,
40
(
10
), pp.
5499
5505
.10.16085/j.issn.1000-6613.2020-2140
12.
Slifka
,
A. J.
,
Drexler
,
E. S.
,
Amaro
,
R. L.
,
Hayden
,
L. E.
,
Stalheim
,
D. G.
,
Lauria
,
D. S.
, and
Hrabe
,
N. W.
,
2017
, “
Fatigue Measurement of Pipeline Steels for the Application of Transporting Gaseous Hydrogen
,”
ASME J. Pressure Vessel Technol.
,
140
(
1
), p.
011407
.10.1115/1.4038594
13.
Yamabe
,
J.
,
Itoga
,
H.
,
Awane
,
T.
,
Matsuo
,
T.
,
Matsunaga
,
H.
, and
Matsuoka
,
S.
,
2015
, “
Pressure Cycle Testing of Cr–Mo Steel Pressure Vessels Subjected to Gaseous Hydrogen
,”
ASME J. Pressure Vessel Technol.
,
138
(
1
), p.
011401
.10.1115/1.4030086
14.
Somerday
,
B. P.
, and
Barney
,
M.
,
2015
, “
Measurement of Fatigue Crack Growth Relationships in Hydrogen Gas for Pressure Swing Adsorber Vessel Steels
,”
ASME J. Pressure Vessel Technol.
,
137
(
2
), p.
021406
.10.1115/1.4028349
15.
Liang
,
J.
,
Li
,
Y.
,
Liu
,
C.
, and
Zhu
,
J.
,
2019
, “
Experimental Study on Leakage Characteristics of Buried Gas Pipelines
,”
CIESC J.
,
70
(
4
), pp.
1635
1643
.10.11949/j.issn.0438-1157.20180944
16.
Alberta
,
E. R.
,
2020
, “
Fort Saskatchewan Hydrogen Blending Project
,” ATCO, Calgary, AB, Canada, accessed Aug. 2023, https://gas.atco.com/en-ca/community/projects/fort-saskatchewan-hydrogen-blending-project.html
17.
Olvera
,
H. A.
, and
Choudhuri
,
A. R.
,
2006
, “
Numerical Simulation of Hydrogen Dispersion in the Vicinity of a Cubical Building in Stable Stratified Atmospheres
,”
Int. J. Hydrogen Energy
,
31
(
15
), pp.
2356
2369
.10.1016/j.ijhydene.2006.02.022
18.
Rigas
,
F.
, and
Sklavounos
,
S.
,
2005
, “
Evaluation of Hazards Associated With Hydrogen Storage Facilities
,”
Int. J. Hydrogen Energy
,
30
(
13–14
), pp.
1501
1510
.10.1016/j.ijhydene.2005.06.004
19.
Rosyid
,
O. A.
,
Jablonski
,
D.
, and
Hauptmanns
,
U.
,
2007
, “
Risk Analysis for the Infrastructure of a Hydrogen Economy
,”
Int. J. Hydrogen Energy
,
32
(
15
), pp.
3194
3200
.10.1016/j.ijhydene.2007.02.012
20.
Wilkening
,
H.
, and
Baraldi
,
D.
,
2007
, “
CFD Modelling of Accidental Hydrogen Release From Pipelines
,”
Int. J. Hydrogen Energy
,
32
(
13
), pp.
2206
2215
.10.1016/j.ijhydene.2007.04.022
21.
Su
,
Y.
,
Li
,
J.
,
Yu
,
B.
, and
Zhao
,
Y.
,
2022
, “
Numerical Investigation on the Leakage and Diffusion Characteristics of Hydrogen-Blended Natural Gas in a Domestic Kitchen
,”
Renewable Energy
,
189
, pp.
899
916
.10.1016/j.renene.2022.03.038
22.
Liu
,
Y.
,
Xu
,
P.
,
Zheng
,
J.
, and
Zhao
,
Y.
,
2008
, “
Numerical Simulation on the Dispersion of Hydrogen and Natural Gas Due to High Pressured Pipeline Leakage
,”
Acta Energ. Sol. Sin.
,
29
(
10
), pp.
1252
1255
.
23.
Zhao
,
B.
,
Zhu
,
M.
,
Peng
,
Y.
, and
Wang
,
M.
,
2017
, “
Simulation of Leakage From Hydrogen and Natural Gas Pipelines Based on PHAST
,”
Petrochem. Ind. Technol.
,
24
(
5
), pp.
48
50
.
24.
Li
,
H.
,
Cao
,
X.
,
Du
,
H.
,
Teng
,
L.
,
Shao
,
Y.
, and
Bian
,
J.
,
2022
, “
Numerical Simulation of Leakage and Diffusion Distribution of Natural Gas and Hydrogen Mixtures in a Closed Container
,”
Int. J. Hydrogen Energy
,
47
(
84
), pp.
35928
35939
.10.1016/j.ijhydene.2022.08.142
25.
Jia
,
W.
,
Wen
,
C.
,
Yang
,
M.
, and
Huang
,
J.
,
2021
, “
Study on Leakage and Diffusion of Hydrogen Mixed Natural Gas in the Valve Chamber
,”
Pet. New Energy
,
33
(
6
), pp.
75
82
.
26.
Lowesmith
,
B. J.
,
Hankinson
,
G.
,
Spataru
,
C.
, and
Stobbart
,
M.
,
2009
, “
Gas Build-Up in a Domestic Property Following Releases of Methane/Hydrogen Mixtures
,”
Int. J. Hydrogen Energy
,
34
(
14
), pp.
5932
5939
.10.1016/j.ijhydene.2009.01.060
27.
Tiekstra
,
G.
, and
Koopman
,
F.
,
2008
, “
The NATURALHY Project: First Step in Assessing the Potential of the Existing Natural Gas Network for Hydrogen Delivery
,”
IGRC
,
Paris, France
.
28.
Melaina
,
M. W.
,
Antonia
,
O.
, and
Penev
,
M.
,
2013
, “
Blending Hydrogen Into Natural Gas Pipeline Networks: A Review of Key Issues
,”
National Renewable Energy Laboratory
, Golden, CO.
29.
Hormaza Mejia
,
A.
,
Brouwer
,
J.
, and
Mac Kinnon
,
M.
,
2020
, “
Hydrogen Leaks at the Same Rate as Natural Gas in Typical Low-Pressure Gas Infrastructure
,”
Int. J. Hydrogen Energy
,
45
(
15
), pp.
8810
8826
.10.1016/j.ijhydene.2019.12.159
30.
Zhu
,
J.
,
Pan
,
J.
,
Zhang
,
Y.
,
Li
,
Y.
,
Li
,
H.
,
Feng
,
H.
,
Chen
,
D.
,
Kou
,
Y.
, and
Yang
,
R.
,
2023
, “
Leakage and Diffusion Behavior of a Buried Pipeline of Hydrogen-Blended Natural Gas
,”
Int. J. Hydrogen Energy
,
48
(
30
), pp.
11592
11610
.10.1016/j.ijhydene.2022.10.185
31.
Drexler
,
E. S.
,
Amaro
,
R. L.
,
Slifka
,
A. J.
,
Bradley
,
P. E.
, and
Lauria
,
D. S.
,
2018
, “
Operating Hydrogen Gas Transmission Pipelines at Pressures Above 21 MPa
,”
ASME J. Pressure Vessel Technol.
,
140
(
6
), p.
061702
.10.1115/1.4041689
32.
Amaro
,
R. L.
,
White
,
R. M.
,
Looney
,
C. P.
,
Drexler
,
E. S.
, and
Slifka
,
A. J.
,
2018
, “
Development of a Model for Hydrogen-Assisted Fatigue Crack Growth of Pipeline Steel
,”
ASME J. Pressure Vessel Technol.
,
140
(
2
), p.
021403
.10.1115/1.4038824
33.
Ebrahimi-Moghadam
,
A.
,
Farzaneh-Gord
,
M.
, and
Deymi-Dashtebayaz
,
M.
,
2016
, “
Correlations for Estimating Natural Gas Leakage From Above-Ground and Buried Urban Distribution Pipelines
,”
J. Nat. Gas Sci. Eng.
,
34
, pp.
185
196
.10.1016/j.jngse.2016.06.062
34.
Yu
,
Q.
,
Hou
,
L.
,
Li
,
Y.
,
Chai
,
C.
,
Liu
,
J.
, and
Yang
,
K.
,
2022
, “
Numerical Study on Harmful Boundary of Above-Ground Section Leakage of Natural Gas Pipeline
,”
J. Loss Prev. Process Ind.
,
80
, p.
104901
.10.1016/j.jlp.2022.104901
35.
Ebrahimi-Moghadam
,
A.
,
Farzaneh-Gord
,
M.
,
Arabkoohsar
,
A.
, and
Moghadam
,
A. J.
,
2018
, “
CFD Analysis of Natural Gas Emission From Damaged Pipelines: Correlation Development for Leakage Estimation
,”
J. Cleaner Prod.
,
199
, pp.
257
271
.10.1016/j.jclepro.2018.07.127
36.
Cheng
,
Y.
,
Lien
,
F. S.
,
Yee
,
E.
, and
Sinclair
,
R.
,
2003
, “
A Comparison of Large Eddy Simulations With a Standard k–ε Reynolds-Averaged Navier–Stokes Model for the Prediction of a Fully Developed Turbulent Flow Over a Matrix of Cubes
,”
J. Wind Eng. Ind. Aerodyn.
,
91
(
11
), pp.
1301
1328
.10.1016/j.jweia.2003.08.001
37.
Bagheri
,
M.
, and
Sari
,
A.
,
2022
, “
Study of Natural Gas Emission From a Hole on Underground Pipelines Using Optimal Design-Based CFD Simulations: Developing Comprehensive Soil Classified Leakage Models
,”
J. Nat. Gas Sci. Eng.
,
102
, p.
104583
.10.1016/j.jngse.2022.104583
38.
Choi
,
S.-K.
, and
Kim
,
S.-O.
,
2012
, “
Turbulence Modeling of Natural Convection in Enclosures: A Review
,”
J. Mech. Sci. Technol.
,
26
(
1
), pp.
283
297
.10.1007/s12206-011-1037-0