Safety studies on high-pressure hydrogen vehicle refuelling stations: Releases into a simulated high-pressure dispensing area

If the general public is to use hydrogen as a vehicle fuel, customers must be able to handle hydrogen with the same degree of confidence, and with comparable risk, as conventional liquid and gaseous fuels. The hazards associated with jet releases from leaks in a vehicle-refuelling environment must be considered if hydrogen is stored and used as a high-pressure gas since a jet release in a confined or congested area can create an explosion hazard. As there was insufficient knowledge of the explosion hazards, a study was initiated to gain a better understanding of the potential explosion hazard consequences associated with high-pressure leaks from hydrogen vehicle refuelling systems. This paper describes the experiments with a dummy vehicle and dispenser units to represent refuelling station congestion. Experiments with ignition of premixed 5.4 m × 6.0 m × 2.5 m hydrogen–air clouds and hydrogen jet releases up to 40 MPa (400 bar) pressure are described. The results are discussed in terms of the conditions leading to the greatest overpressures and overall conclusions are made from these.     

The effect of vent burst pressure on a vented hydrogen–air deflagration in a 1 m3 vessel

A series of experiments are conducted to investigate the effect of vent burst pressure on stoichiometric hydrogen–air premixed flame propagation and pressure history in a 1 m³ rectangular vessel in this paper. Pressure buildup and flame evolution are recorded using piezoelectric pressure transducers and a high-speed camera, respectively. The results show typical pressure peaks of three different mechanisms for all vent burst pressures in the experiments. The first pressure peak, generated by the rupture of the vent cover, increases with the vent failure pressure, with the subsequent outflow inertia of combustion products giving rise to a negative pressure. The second pressure peak results from the constant bulk motion of the flame bubble (the Helmholtz oscillation), and the third is produced by the interaction between the combustion waves and the acoustic waves. The time interval between the first pressure peak and the second pressure transient remained nearly constant. The Helmholtz oscillation always appears as the vent ruptures and its magnitude increases with the vent burst pressure. Furthermore, the lower the vent failure pressure, the longer the Helmholtz oscillation is sustained. The peak of the acoustically enhanced pressure always occurs within several milliseconds of the flame front touching the vessel. From a theoretical perspective, Rasbash's equation models the relationship between the maximum reduced explosion overpressure and the vent burst pressure precisely. Also, it is observed that the maximum lengths of the external flames were found to be nearly identical in all tests, but the average propagation rate of the flame front increases with the vent burst pressure. It is interesting that a phenomenon of intense oscillation of internal flame bubble was observed with the increase of vent burst pressure.     

Effects of nitrogen addition on vented hydrogen–air deflagration in a vertical duct

In this paper, experiments were performed to investigate the coupling effects of venting and nitrogen addition ratio (χ) on flame behavior and pressure evolution during hydrogen–air deflagration within and outside a 1-m-high vertical duct with a vent on its top. Experimental results reveal that χ has significant effects on the pressure–time histories in the duct. Helmholtz oscillations of the internal overpressure were observed in all tests, and acoustic type oscillations appears in the tests only for χ = 25% and 30%. For a certain χ, the maximum overpressure (Pmax) increased with the distance to the vent, i.e., the highest overall explosion overpressure was attained near the duct bottom; however, the difference in Pmax between various measuring points decreases with an increase in χ. In all tests, a pressure peak in the duct was observed shortly after external explosion. The maximum internal and external overpressure decreased as χ was increased.     

Consequences of catastrophic releases of ignited and unignited hydrogen jet releases

The possibility of using a risk based approach for the safe installation and siting of stationary fuel cell systems depends upon the availability of normative data and guidance on potential hazards, and the probabilities of their occurrence. Such guidance data is readily available for most common hydrocarbon fuels. For hydrogen, however, data is still required on the hazards associated with different release scenarios. This data can then be related to the probability of different types of scenarios, from historical fault data, to allow safety distances to be defined and controlled using different techniques. Some data on releases has started to appear but this data generally relates to hydrogen vehicle refuelling systems that are designed for larger throughput, higher pressures, and the general use of larger pipe diameters than are likely to be used for small fuel cell systems.The aim of this paper is to report on work that is providing data for informing safety distances for high-pressure components/fuel cell systems and associated fuel storage. Using high-pressure release scenarios, the extent of the clouds, jets and, following ignition, fires and explosions were investigated.     

Duct-vented hydrogen–air deflagrations: The effect of duct length and hydrogen concentration

Experiments on duct-vented explosions of hydrogen–air mixtures in a 12.3 l cylindrical vessel were conducted, and the effects of duct length and hydrogen concentration on the maximum overpressure and flame behavior within and outside the vented enclosure were investigated. The results show that the maximum overpressure in the vessel first increased and then was maintained nearly unchanged with the length of a relief duct increasing to 2 m. For a given duct length, the maximum overpressure first increased and then decreased when hydrogen concentration increased from 20% to 55%. The burn-up in the duct caused the gas mixtures to move in reverse from the duct to vessel, which consequently decreased the venting efficiency. A pressure wave caused by burn-up in the duct was observed, which resulted in a pressure peak in the external pressure–time histories after it traveled outside the duct. The maximum external overpressure first increased and then decreased with an increase in duct length. For a given duct length, the maximum external overpressure increased with an increase in hydrogen concentration.     

Numerical assessment of hydrogen explosion consequences in a mine tunnel

The aim of the work is a numerical estimation of the conditional probability of damage to the mine personnel during an accidental explosion of a hydrogen-air mixture. The methodology for determining the parameters of the gas-dynamic process of the explosion of a hydrogen-air cloud in an open and closed space, taking into account chemical interaction and space clutter, is presented. A computational method based on a probit analysis for determining the damage probability fields of a person exposed to the explosion shock wave has been developed. To automate the computational process, the tabular dependence “probit-function-damage probability” is replaced by a piecewise cubic spline. Numerical studies of the influence of the drift working space clutter by an electric locomotive on the distribution of the overpressure of the gaseous medium and the conditional probability of the eardrums rupture and lethal damage to personnel in the accidental zone of the coal mine have been carried out. It was obtained that the closed nature of the working space and its blockage significantly changes the shape and size of the danger zone and requires consideration by an expert at the stage of deciding on the safety level at the mine. The scientific novelty of the method proposed in the work is in taking into account in the mathematical model of the movement of a multi-component chemically reacting gas mixture the effect of compressibility of flow, complex terrain (space clutter with equipment), three-dimensional nature of the gas-air mixture dispersion process. The model allows obtaining the space-time distributions of the shock-impulse load of the blast wave that is necessary for determining the non-stationary three-dimensional fields of the conditional probability of damage to the staff on the basis of probit analysis. The developed computational method allows analyzing and forecasting in time and space the conditional probability of damage of varying degrees of severity of personnel who are exposed to an explosive shock wave as an indicator of the safety level of a coal mine.     

Experimental investigation on the dynamic responses of vented hydrogen explosion in a 40-foot container

To investigate the structural dynamics of a container subjected to a vented hydrogen explosion, 48 field tests were conducted in a 40-foot container with roof vents and an end vent. The effects of the hydrogen concentration, ignition position, and obstacles on the evolution of the dynamic responses were investigated. Three stages were generally observed for displacements: (1) At the stage of the vent rupture, the displacement could be approximated as a quasi-static response, and there was a linear relationship between the peaks of positive overpressure and displacement. (2) Structural deformation appeared as reciprocating vibration at the stage of Helmholtz oscillation. (3) The structure exhibited relatively weak irregular fluctuation when high-frequency acoustic oscillation occurred. Two types of the structural acceleration with low and high amplitudes resulting from Helmholtz oscillation and acoustic oscillation, respectively, were clearly observed. For the end-vented explosion, multiple peaks were observed for the displacement at the quasi-static stage due to the rupture, discharge, and external explosion. Moreover, the displacement was sensitive to hydrogen concentration, whereas the number of obstacles and the ignition position had significant influences on the peak acceleration for roof venting. This work conducted the fundamental explanation for the evolution law of structural responses induced by vented hydrogen explosions from the perspective of structural dynamics and enriched the experimental accumulation in a large-scale container with congestion in this field.     

A simple and effective approach for evaluating unconfined hydrogen/air cloud explosions

The topic of hydrogen safety assessment has been focused by many researchers. The overpressure evaluation of vapor cloud explosion (VCE), is an important issue for both designing and evaluating on chemical plants, as well as buildings. Unknown flame radius history limits the original acoustic approximation model's application. The objective of this work is to develop an achievable model for hydrogen/air deflagration assessment in engineering applications, and the model should have high computational efficiency. A tentative scheme that starts from flame/piston speed history solving was adopted, and the flame/piston radius and acceleration history will be obtained subsequently. Thus, the overpressure history for far field could be gotten based on the acoustic approximation model. A simplified scheme was employed for the region inside the flame cloud. The model proposed in this paper could be solved in several seconds, because there are no differential equations but only algebraic equations. The model was verified by hydrogen/air deflagration tests from small scale to large scale. Compared with the experimental data, the model appeared well agreements in the medium and large scale cases. In the small scale cases, the model obtained acceptable solutions.     

Effects of sample orientation on nonpiloted ignition of thin poly(methyl methacrylate) sheets by a laser: 2. Experimental results

The effect of the sample orientation angle on frontside (irradiated surface) ignition and subsequent backside (nonirradiated surface) flame appearance over thin poly(methyl methacrylate) (PMMA) sheets having thicknesses of 0.2 and 0.5 mm has been experimentally investigated, using a CO₂ laser as an external radiant source in quiescent normal gravity. The sample orientation angle was varied from θ=−90° (ceiling configuration) to +90° (floor configuration) at intervals of 15° under three different laser powers of 16.0, 17.3, and 26.1 W. The shortest frontside ignition delay time was observed for the ceiling configuration (θ=−90°) and frontside ignition delay time significantly varied with increase in sample orientation angle at a laser power of 16.0 W. As the laser power was increased, frontside ignition was observed at all angles and its delay time became less dependent on the sample orientation angle. The appearance of a backside flame was achieved after the formation of an open hole (due to local consumption of the sample) by two different processes: the onset of laser induced ignition over the backside sample (backside ignition) and a flame traveling from the frontside through an open hole to the backside (backside flame). The former process was observed for a limited number of cases only around the vertical configurations (−30°?θ?30°). The delay time for the appearance of backside flame tended to be longer for sample surfaces facing downward (θ°<0) than for the sample surface facing upward (θ?0°) regardless of the laser power. When the duration of laser irradiation was shortened from 10 to 4 s, as soon as the laser was shut off, the flame on the frontside immediately shrank, moved close to the sample surface, and then traveled rapidly to the backside. Therefore, the delay time of backside flame appearance (about 6 s) became longer with longer duration of laser irradiation after the onset of a frontside flame. The size of the hole (about 4 mm diameter) was large enough for the flame to travel through it, even after 4 s of laser irradiation to sample. These results indicate that the size of the hole appears to be not a critical parameter for the appearance of the backside flame.     

Investigation of small-scale unintended releases of hydrogen: Buoyancy effects

Measurements were performed in small-scale hydrogen leaks to characterize the dimensional properties and flow characteristics of the resulting ignitable hydrogen cloud. The data are intended to provide a technological basis for determining hazardous length scales associated with the formation of ignitable mixtures due to unintended releases. In contrast to a previous study where momentum-dominated releases were considered, the present study focuses on smaller-scale releases at lower flow rates where buoyancy becomes important. A turbulent jet flow is selected as representative of releases in which the leak geometry is circular. Laser-based Rayleigh scattering is used to characterize the hydrogen concentration field downstream of the leak. Particle Image Velocimetry (PIV) is also used to characterize the flow velocity. Time-averaged mean and fluctuating hydrogen concentration statistics are presented and compared with results in momentum-dominated flows to elucidate the effects of buoyancy on the H₂ dispersion process. Over the range of Froude numbers investigated (Fr = 268, 152 and 99), increasing effects of buoyancy are seen as the Fr is reduced and at downstream locations where the influence of buoyancy increases relative to jet momentum. The primary effect of buoyancy is to increase the centerline decay rate of the time-averaged H₂ mass fraction relative to momentum-dominated flows. Acceleration due to buoyancy also results in a slower decay of the time-averaged axial velocity component along the centerline. Radial profiles of the time-averaged H₂ mass fraction also collapse onto the same curves as results in momentum-dominated flows when plotted against the same similarity/scaling variables. While buoyancy is found to have a negligible effect on centerline velocity fluctuations, the maximum H₂ mass fraction fluctuation intensity increases by 70 percent in the buoyant regime and the peak value shifts from the mixing region to the jet centerline. The database presented should provide a good test for the validation of CFD models being developed to predict unintended hydrogen releases under conditions where buoyancy is important.     

Experimental study on the explosion behaviors of premixed syngas-air mixtures in ducts

Explosion characteristics of premixed syngas-air mixtures at room temperature and atmospheric pressure were experimentally reported when the explosion flame propagates in ducts with various heights (H) and lengths (L). The discussion was based on flame morphology and pressure dynamics. The ratio of L/H and the ratio of H₂/CO had a significant effect on the explosion flame behaviors as the explosion occurred in ducts. The structure of the explosion flame changes more drastically, as both the L/H ratio is large. The ratio of L/H affected the flame tip dynamics after the flame reached the duct wall, and the time of flame reaching the duct walls is divinable. For a given duct height, the shorter the duct length is, the faster flame propagates, and the maximum flame tip speed was higher as the duct length was small. For a given duct length, flame tip dynamics showed a nearly same development tendency, but the shorter the duct height, the faster the flame propagated. The venting pressure affected the overpressure dynamics, and the venting pressure increased with the increase of the L/H ratio and the H₂/CO. For a given duct height, the overpressure reached the maximum value almost at the same time, and the longer duct length resulted in the greater maximum overpressure. Finally, for a given duct length, the higher duct height caused the higher maximum overpressure.     

Experimental and numerical investigation of hydrogen gas auto-ignition

This paper describes hydrogen self-ignition as a result of the formation of a shock wave in front of a high-pressure hydrogen gas propagating in the tube and in the semi-confined space, for which the numerical and experimental investigation was done. An increase in the temperature behind the shock wave leads to the ignition on the contact surface of the mixture of combustible gas with air. The required condition of combustible self-ignition is to maintain the high temperature in the mixture for a time long enough for inflammation to take place. Experimental technique was based on a high-pressure chamber inflating with hydrogen, burst disk failure and pressurized hydrogen discharge into tube of round or rectangular cross section filled with air. Two numerical models involving the gas-dynamic transport of a viscous gas, the detailed kinetics of hydrogen oxidation, turbulence model, and heat exchange were used for calculations of the hydrogen self-ignition both in semi-confined space and a tube.     

Thermal and chemical effects of hydrogen addition on catalytic micro-combustion of methane–air

In this paper hydrogen assisted catalytic combustion of methane on rhodium is numerically modeled in steady condition. The aim of the work goes to better understand how the addition of hydrogen affects the combustion of methane–air. For this purpose, a micro flatbed channel is investigated by a three-dimensional simulation including an elementary-step surface reaction mechanism. It is clearly shown through a numerical study that appropriate hydrogen addition increases the conversion of methane and expands the lower limit of burnable equivalence ratio. In addition, the main effect of hydrogen is thermal when the mass fraction of hydrogen addition is less than 0.67%, while not only thermal effect but also chemical effect appears when the mass fraction is more than 0.67%. The sharp decreases of hydrogen fraction appear twice till hydrogen fraction increases from 0.67%. In addition, the first abrupt decline increases Rh(s) coverage to create favorable conditions for adsorption and oxidation of methane and it can suddenly reduce the ignition temperature 15 K and advance ignition distance 3%. Thanks to the second sharp decline, the adsorption–desorption equilibrium of oxygen slowly shifts towards desorption with increasing temperature to increase Rh(s) coverage.     

Combustion characteristics for a hydrogen-fuelled low-pressure-ratio split-cycle engine

Investigation of hydrogen-air combustion characteristics for low-pressure-ratio split-cycle engines is performed for the first time. Ignition delay and combustion duration data are presented for low initial pressures below 5 bars and temperatures from ambient level till 550 °C. A simplified constant-volume combustion chamber experiment with glow-plug ignition and timed intake and exhaust valves is constructed in order to simulate the real operation. Ignition delay is found to be inversely proportional to the temperature, and unexpectedly directly proportional with pressure. Equivalence ratio has a weak influence on the ignition delay. Combustion duration exhibits the same behaviour as the ignition delay and represents an average 40% portion of the total combustion delay. Ignition delay results are bounded to the selected glow plug surface temperature range and complements with data in literature. Based on the results, the engine rotational speed would be limited to a maximum of 4500 RPM for an intake duration of 30° at 550 °C and 3.5 bars pre-combustion conditions.     

Experimental study on spark ignition of non-premixed hydrogen jets

The leaks of pressurized hydrogen can be ignited if an ignition source is within a certain distance from the source of the leaks, and jet fires or explosions may take place. In this paper, a high speed camera was used to investigate the ignition kernel development, ignition probability and flame propagation along the axis of hydrogen jets, which leaked from a 3-mm-internal-diameter nozzle and were ignited by an electric spark. Experimental results indicate that for successful ignition events, the ignition delay time increases with an increase of the distance between the nozzle and the electrode. Ignitable zone of the hydrogen jets is underestimated if using the predicted hydrogen concentration along the jets centerline. The average rate of downstream flame decreases but that of the upstream flame increases with the electrode going far from the nozzle.