Safety

Hydrogen can be used as a buffer for storing intermittent electricity produced by solar plants and/or wind farms. The MYRTE project in Corsica France aims to operate and test a large scale hydrogen facility for regulating the electricity produced by a 560 kWp photovoltaic plant.

Due to the large quantity of hydrogen and oxygen produced and stored (respectively 333 kg and 2654 kg) this installation faces safety issues and safety regulations constraints that can lead to extra costs. These extra costs may concern detectors monitoring barrier equipments that have to be taken into account for evaluating the system‘s total cost.

Relying on the MYRTE example that is an R&D platform the present work consists in listing the whole environmental and safety regulations to be applied in France on both Hydrogen and Oxygen production and storage. A methodology has been developed [1] [2] for evaluating safety extra costs. This methodology takes into account various hydrogen storage technologies (gaseous and solid state) and is applicable to other ways of storage (batteries etc.) to compare them. Results of this work based on a forecast of the operating platform over 20 years can be used to extrapolate and/or optimize future safety costs of next large scale hydrogen systems for further PV or wind energy storage applications.

The scenario of ignition of fuels by the passage of shock waves is relevant from the perspective of safety primarily because shock ignition potentially plays an important role in deflagration to detonation transition. Even in one dimension simulation of ignition between a contact surface or a flame and a shock moving into combustible mixture is difficult because of the singular nature of the initial conditions. Indeed initially as the shock starts moving away from the contact surface the region filled with shocked reactive mixture does not exist. In the current work the formulation is transformed using time and length over time as the independent variables. This transformation yields a finite domain from t = 0. In this paper the complete spatial and temporal ignition evolution of hydrogen combustible mixtures of different concentrations is studied numerically. Integration of the governing equations is performed using an Essentially Non-Oscillatory (ENO) algorithm in space and Runge-Kutta in time while the chemistry is modeled by a three-step chain-branching mechanism which appropriately mimics hydrogen combustion.

Developers interested in high pressure storage of hydrogen for mobile use increasingly rely on composite cylinders for onboard storage or transport of dangerous goods. Thus composite materials and systems deserve special consideration. History gives interesting background information important to the understanding of the current situation as to regulations codes and standards.<br/>Based on this review origins of different regulations for the storage of hydrogen as dangerous good and as propellant for vehicles will be examined. Both categories started out using steel and sometimes aluminium as cylinder material. With composite materials becoming more common a new problem emerged: vital input for regulations on composite pressure systems was initially derived from decades of experience with steel cylinders. As a result both regulatory fields suffer somewhat from this common basis. Only recent developments regarding requirements for composite cylinders have begun to go more and more separate ways. Thus these differences lead to some shortcomings in regulation with respect to composite storage systems.<br/>In principle in spite of separate development these deficits are in both applications very much the same: there are uncertainties in the prediction of safe service life in retesting procedures of composite cylinders and in their intervals. Hence different aspects of uncertainties and relevant approaches to solutions will be explained.

Safety distances for hydrogen plumes are currently derived using models developed for hydrocarbon releases. It is well known that hydrogen behaves in a significantly different manner to that of hydrocarbons when released to atmosphere. There are two main aspects involved with the development of safety distances for credible hydrogen releases; the intensity of the thermal radiation from such a plume should it be ignited and the distance downwind from the release point to the point where a flammable mixture with air no longer exists. A number of distinct areas of venting behaviour were investigated; Thermal radiation from ignited plumes from vertical open ended vent pipes Far field radiation measurements for direct comparison with models Thermal radiation from ignited plumes from vertical vent pipes terminating in a T-piece Thermal radiation measurements from ignited hydrogen with a 45 vent termination Hydrogen concentration measurements with a T-piece.

Type 1 steel pressure vessels are commonly used for the transport of pressurized gases including gaseous hydrogen. In the majority of cases these cylinders experience relatively few pressure cycles over their lifetime perhaps in the hundreds. In emerging markets such as hydrogen-powered industrial trucks hydrogen fuel systems are expected to experience thousands of cycles over just a few year period. This study investigates the fatigue life of Type 1 steel pressure vessels by subjecting full- scale vessels to pressure cycles with gaseous hydrogen between nominal pressure of 3.5 and 43.8 MPa. In addition engineered defects were machined on the inside of several pressure vessels for comparison to fatigue crack growth measurements on materials sectioned from these pressure vessels. As-manufactured pressure vessels have sustained >35000 cycles with failure while vessels with machined defects leaked before bursting after 8000 to 15000 pressure cycles. The measured number of cycles to failure in these pressure vessels is two to three times greater than predicted using conservative methods based on fatigue crack growth rates measured in gaseous hydrogen.

There are new safety challenges with an increased use of hydrogen e.g. that people may not see dangerous jet flames in case of an incident. Compared to conventional fuels hydrogen has very different characteristics and physical properties and is stored at very high pressure or at very low temperatures. Thus the nature of hazard scenarios will be very different. Consequence modelling of ventilation releases explosions and fires can be used to predict and thus understand hazards. In order to describe the detailed development of a hazard scenario and evaluate ways of mitigation 3D Computational Fluid Dynamics (CFD) models will be required. Even with accurate modelling the communication of risk can be challenging. For this visualization in virtual reality (VR) may be of good help in which the CFD model predictions are presented in a realistic 3D environment with the possibility to include sounds like noise from a high pressure release explosion or fire. In cooperation with Statoil Christian Michelsen Research (CMR) and GexCon have developed the VRSafety application. VRSafety can visualize simulation results from FLACS (and another CFD-tool KFX) in an immersive VR-lab or on a PC. VRSafety can further be used to interactively control and start new CFD-simulations during the sessions. The combination of accurate CFD-modelling visualization and interactive use through VRSafety represents a powerful toolbox for safety training and risk communication to first-responders employees media and other stakeholders. It can also be used for lessons learned sessions studying incidents and accidents and to demonstrate what went wrong and how mitigation could have prevented accidents from happening. This paper will describe possibilities with VRSafety and give examples of use.

Two-dimensional numerical simulations of deflagration-to-detonation transition (DDT) in hydrogen–air mixtures are presented and compared with experiments. The investigated geometry was a 3 m long square channel. One end was closed and had a single obstacle placed 1 m from the end and the other end was open to the atmosphere. The mixture was ignited at the closed end. Experiments and simulations showed that DDT occurred within 1 m behind the obstacle. The onset of detonation followed a series of local explosions occurring far behind the leading edge of the flame in a layer of unburned reactants between the flame and the walls. A local explosion was also seen in the experiments and the pressure records indicated that there may have been more. Furthermore local explosions were observed in the experiments and simulations which did not detonate. The explosions should have sufficient strength and should explode in a layer of sufficient height to result in a detonation. The numerical resolution was 0.5 mm per square cell and further details of the combustion model used are provided in the paper.

This study is driven by the need to understand requirements to safe blow-down of hydrogen onboard storage tanks through a pressure relief device (PRD) inside a garage-like enclosure with low natural ventilation. Current composite tanks for high pressure hydrogen storage have been shown to rupture in 3.5–6.5 min in fire conditions. As a result a large PRD venting area is currently used to release hydrogen from the tank before its catastrophic failure. However even if unignited the release of hydrogen from such PRDs has been shown in our previous studies to result in unacceptable overpressures within the garage capable of causing major damage and possible collapse of the structure. Thus to prevent collapse of the garage in the case of a malfunction of the PRD and an unignited hydrogen release there is a clear need to increase blow-down time by reducing PRD venting area. Calculations of PRD diameter to safely blow-down storage tanks with inventories of 1 5 and 13 kg hydrogen are considered here for a range of garage volumes and natural ventilation expressed in air changes per hour (ACH). The phenomenological model is used to examine the pressure dynamics within a garage with low natural ventilation down to the known minimum of 0.03 ACH. Thus with moderate hydrogen flow rate from the PRD and small vents providing ventilation of the enclosure there will be only outflow from the garage without any air intake from outside. The PRD diameter which ensures that the pressure in the garage does not exceed a value of 20 kPa (accepted in this study as a safe overpressure for civil structures) was calculated for varying garage volumes and natural ventilation (ACH). The results are presented in the form of simple to use engineering nomograms. The conclusion is drawn that PRDs currently available for hydrogen-powered vehicles should be redesigned along with either a change of requirements for the fire resistance rating or innovative design of the onboard storage system as hydrogen-powered vehicles are intended for garage parking. Further research is needed to develop safety strategies and engineering solutions to tackle the problem of fire resistance of onboard storage tanks and requirements to PRD performance. Regulation codes and standards in the field should address this issue.

A two-dimensional (2-D) simulation of spontaneous ignition of high-pressure hydrogen in a length of duct is conducted in order to explore its underlying ignition mechanisms. The present study adopts a 2-D rectangular duct (i.e. not axisymmetric geometry) and focuses on the effects of initial diaphragm shape on the spontaneous ignitions. The Navier-Stokes equations with a detailed chemical kinetics mechanism are solved in a manner of direct numerical simulation. The detailed mechanisms of spontaneous ignition are discussed for each initial diaphragm shape. For a straight diaphragm shape it is found that the ignition occurs only near the wall due to the adiabatic wall condition while the three ignition events: ignitions due to leading shock wave reflection at the wall hydrogen penetration into shock-heated air near the wall and deep penetration of hydrogen into shock-heated air behind the leading shock wave are identified for a largely deformed diaphragm shape.

Hydrogen is seen as an ideal energy carrier for stationary and mobile applications. However the use of high energy density materials such as hydrides comes with the drawback of risks associated to their high reactivity towards air and water exposure. We have developed novel strategies to mitigate these risks. These strategies were evaluated using standard UN tests and isothermal calorimetric measurements. Cycling experiments were conducted to assess the impact of the mitigants on the modified materials derived from 8LiH•3Mg(NH2)2 system. In some cases our results show an improvement in kinetics when compared to the unmodified material.  Effective mitigants were also discovered for aluminium hydride (alane) and lithium borohydride completely inhibiting ignition.

The  storage  of  hydrogen  poses  inherent  weight  volume  and  safety  obstacles.  An  innovative technology which allows for the storage of hydrogen in thin sealed glass capillaries ensures the safe infusion storage and controlled release of hydrogen gas under pressures up to 100 MPa. Glass is a non-flammable material which also guarantees high burst pressures. The pressure resistance of single and multiple capillaries has been determined for different glass materials. Borosilicate capillaries have been proven to have the highest pressure resistance and have therefore been selected for further series of  advanced testing.  The innovative  storage  system  is  finally  composed of  a  variable number  of modules. As such in the case of the release of hydrogen this modular arrangement allows potential hazards to be reduced to a minimum. Further advantage of a modular system is the arrangement of single modules in every shape and volume dependent on the final application. Therefore the typical locations of storage systems e.g. the rear of cars can be modified or shifted to places of higher safety and not directly involved in crashes. The various methods of refilling and releasing capillaries with compressed hydrogen the increase of burst pressures through pre-treatment as well as the theoretical analysis and  experimental results of  the resistance  of glass  capillaries will further  be discussed  in detail.

Vehicular use of hydrogen is the first attempt to apply hydrogen energy in consumers’ environment in large scale and has raised safety concerns in both public authorities and private bodies such as fire services and insurance companies. This paper analyzes typical accident progressions of hydrogen fuel cell vehicles in a road collision accident. Major hydrogen consequences including impinging jet fires and catastrophic tank ruptures are evaluated separately in terms of accident duration and hazard distances. Results show that in a 70 MPa fuel cell car accident the hazards associated with hydrogen releases would normally last for no more than 1.5 min due to the empty of the tank. For the safety of general public a perimeter of 100 m is suggested in the accident scene if no hissing sound is heard. However the perimeter can be reduced to 10 m once the hissing sound of hydrogen release is heard. Furthermore risks of fatalities injuries and damages are all quantified in financial terms to assess the impacts of the accident. Results show that costs of fatalities and injuries contribute most to the overall financial loss indicating that the insurance premium of fatalities and injuries should be set higher than that of property loss.

In this work we evaluate the consequences of the combustion of a stoichiometric mixture of hydrogen-air on a mechanical device which can be considered as a long tube. In order to choose the most dangerous combustion regime for the mechanical device we devote a particular attention to the investigation of the 1D deflagration-to-detonation transition. Then once established the most dangerous combustion regime we compute the reacting  flow and the stress and strain in the mechanical device. Analyses are performed using both semi-analytical solutions and Europlexus a computer program for the simulation of  fluid-structure systems under transient dynamic loading.

In a near future with an increasing use of hydrogen as an energy vector gaseous hydrogen transport as well as high capacity storage may imply the use of high strength steel pipelines for economical reasons. However such materials are well known to be sensitive to hydrogen embrittlement (HE). For safety reasons it is thus necessary to improve and clarify the means of quantifying embrittlement. The present paper exposes the changes in mechanical properties of a grade API X80 steel through numerous mechanical tests i.e. tensile tests disk pressure test fracture toughness and fatigue crack growth measurements WOL tests performed either in neutral atmosphere or in high-pressure of hydrogen gas. The observed results are then discussed in front of safety considerations for the redaction of standards for the qualification of materials dedicating to hydrogen transport.

A reflected shock wave in two-dimensional shock tube is studied numerically using Navier-Stokes equations with the detailed oxy-hydrogen reaction mechanism. The results show detailed process of mild ignition. The interaction between the reflected shock wave and the boundary layer yielded behind the incident shock wave produces clockwise and counter-clockwise vortices. These vortices generate compression waves. The future study related wall conditions (adiabatic or isothermal) will be shown at the conference site.

Hydrogen gas explosions in homogeneous reactive mixtures have been widely studied both experimentally and numerically. However in practice combustible mixtures are usually inhomogeneous and subject to both vertical and horizontal concentration gradients. There is still very limited understanding of the hydrogen explosion characteristics in such situations. The present numerical investigation aims to study the effect of mixture concentration gradient on the process of Deflagration to Detonation Transition and the effect of different hydrogen concentration gradient in the obstructed channel of hydrogen/air mixtures. An obstructed channel with 30% blockage ratio (BR=30) and three different average hydrogen concentrations of 20 % 30% and 35% have been considered using a specially developed density-based solver within the OpenFOAM toolbox. A high-resolution grid was built with the using adaptive mesh refinement technique providing 10 grid points in half reaction length. The numerical results are in reasonably good agreement with the experimental  observations [1]. These studies show that the concentration gradient has a considerable effect on the accelerated flame tip speed and the location of transition to detonation in the obstructed channel. In all the three cases the first localised explosion occurred near the bottom wall where the shock and flame interacted and the mixture was most lean; and the second localised explosion occurred at the top wall due to the reflection of shock and flame front and later develops to form the leading detonation wave. The increase in the fuel concentration was found to increase the flame acceleration (FA) and having a faster transition to detonation.

A large number of natural gas vehicles (NGV) with composite cylinders run in the world. In order to store hydrogen using the composite cylinder has also reached commercialization for the hydrogen fuel cell vehicle (FCV) which is been developing on ECO Energy. Under these increasing circumstances the most important issue is that makes sure of safety of the hydrogen composite cylinder. In case of the composite cylinder a standards to verify the safety of cylinders obey several country's standards. For NGV ISO 11439 has adopted as international standards but for FCV it has been still developing and there is only ISO/TS 15869 as international technical standards. In contents of international standards the fire test is the weakest part. The fire test is that the pressure relief valves (PRD) normally operate or not in order to prevent cylinders bursting when a vehicle is covered by fire. However with present standards there is no method to check the problem from vehicles in local flame. This study includes fire test results that have been performed to establish the fire-test standards.

This paper shows the numerical investigation on the self-ignition behavior of high pressure hydrogen released from the tube. The present study aims to clarify the effect of parameters on the behavior and duration of self-ignition outside the tube using two-dimensional axisymmetric numerical simulation with detailed chemistry. The parameters in this study are release pressure tube diameter and tube length. The strength of the spherical shock wave to keep chemical reaction and expansion are important factors for self ignited hydrogen jet to be sustained outside the tube. The trend of strength of spherical shock wave is enhanced by higher release pressure and larger tube diameter. The chemical reaction weakens due to expansion and the degree of expansion becomes larger as the spherical shock wave propagates. The characteristic time for the chemical reaction becomes shorter in higher release pressure larger tube diameter and longer tube diameter cases from the induction time under constant volume assumption. The self ignited hydrogen jet released from the tube is sustained up to the distance where the characteristic time for chemical reaction is shorter than the characteristic time for the flow to expand and higher release pressure larger tube diameter and longer tube length expand the distance where the tip flame can propagate downstream. For the seed flame which is the key for jet fire the larger amount of the ignited volume when the shock wave reaches the tube exit contributes to the formation and stability of the seed flame. The amount of the ignited volume tends to be larger in the longer tube length higher release pressure and larger tube diameter cases.

Hydrogen behavior at elevated pressures and temperatures was intensively studied by numerous investigators. Nevertheless there is a lack of experimental data on hydrogen ignition and combustion at reduced sub-atmospheric pressures. Such conditions are related to the facilities operating under vacuum or sub-atmospheric conditions for instance like ITER vacuum vessel. Main goal of current work was an experimental evaluation of such fundamental properties of hydrogen–air mixtures as flammability limits and laminar flame speed at sub-atmospheric pressures. A spherical explosion chamber with a volume of 8.2 dm3 was used in the experiments. A pressure method and high-speed camera combined with schlieren system for flame visualization were used in this work. Upper and lower flammability limits and laminar flame velocity have been experimentally evaluated in the range of 4–80% hydrogen in air at initial pressures 25–1000 mbar. An extraction of basic flame properties as Markstein length overall reaction order and activation energy was done from experimental data on laminar burning velocity.

We report the development and testing of a novel hydrogen sensor that shows a very peculiar response to hydrogen exposure due to its micro-structured palladium surface. The fabrication of the wrinkled Pd surface is obtained using an innovative fast and cheap technique based on the deposition of a thin Pd film on to a thermo-retractable polystyrene sheet that shrinks to 40% of its original size when heated. The buckling of the Pd surface induced by shrinking of the substrate produces nano and micro-wrinkles on the sensor surface. The micro-structured sensor surface is very stable even after repeated hydrogen sorption/desorption cycles. The hydrogen sensing mechanism is based on the transitory absorption of hydrogen atoms into the Pd layer leading to the reversible change of its electrical resistance. Interestingly depending on hydrogen concentration the proposed sensor shows the concurrent effect of both the usually described behaviors of increase or decrease of resistance related to different phenomena occurring upon hydrogen exposure and formation of palladium hydride. The study reports and discusses evidences for an activation threshold of hydrogen concentration in air switching the behavior of sensor performances from e.g. poor negative to large positive sensitivity and from slow to fast detection.

Although many studies have looked at safety issues relating to hydrogen fuelling stations few studies have analyzed the security risks such as deliberate attack of the station by threats such as terrorists and disgruntled employees. The purpose of this study is to analyze security risks for a hydrogen fuelling station with an on-site production of hydrogen from methylcyclohexane. We qualitatively conducted a security risk analysis using American Petroleum Institute Standard 780 as a reference for the analysis. The analysis identified 93 scenarios including pool fires. We quantitatively simulated a pool fire scenario unique to the station to analyze attack consequences. Based on the analysis and the simulation we recommend countermeasures to prevent and mitigate deliberate attacks.

Experimental and numerical investigation of hydrogen-air and hydrogen-oxygen detonation parameters was performed. A new detonation model was introduced and validated against the experimental data. Experimental set-up consisted of 9 m long tube with 0.17 m in diameter where pressure was measured with piezoelectric transducers located along the channel. Numerical simulations were performed within OpenFoam code based on progress variable equation where the detonative source term accounts for autoignition effects. Autoignition delay times were computed at a simulation run-time with the use of a multivariate regression model where independent variables were: pressure temperature and fuel concentration. The dependent variable was the autoignition delay time. Range of the analyzed gaseous mixture composition varied between 20% and 50% of hydrogen-air and 50%–66% of hydrogen in oxygen. Simulations were performed using LES one-equation eddy viscosity turbulence model in 2D and 3D. Calculations were validated against experimental data.

This paper provides an introduction to the powered industrial truck application of fuel cell power systems the safety similarities with the automotive application and safety lessons learned. Fuel Cell niche markets have proven their value to many early adopters. How has the automotive market provided a springboard for these niche applications? How are niche markets revealing gaps in current safety approaches? What is different about the powered industrial truck application and what new codes and standards are needed to accommodate those differences?

Liquid hydrogen (LH2) storage is viewed as a viable approach to assure sufficient hydrogen capacity at commercial fuelling stations. Presently LH2 is produced at remote facilities and then transported to the end-use site by road vehicles (i.e. LH2 tanker trucks). Venting of hydrogen to depressurize the transport storage tank is a routine part of the LH2 delivery and site transfer process. The behaviour of cold hydrogen plumes has not been well characterized because of the sparsity of empirical field data which can lead to overly conservative safety requirements. Committee members of the National Fire Protection Association (NFPA) Standard 2 [1] formed the Hydrogen Storage Safety Task Group which consists of hydrogen producers safety experts and computational fluid dynamics modellers has identified the lack of understanding of hydrogen dispersion during LH2 venting of storage vessels as a critical gap for establishing safety distances at LH2 facilities especially commercial hydrogen fuelling stations. To address this need the National Renewable Energy Laboratory Sensor Laboratory in collaboration with the NFPA Hydrogen Storage Task Group developed a prototype Cold Hydrogen Plume Analyzer to empirically characterize the hydrogen plume formed during LH2 storage tank venting. The prototype analyzer was field deployed during an actual LH2 venting process. Critical findings included

• Hydrogen above the lower flammable limit (LFL) was detected as much as 2 m lower than the release point which is not predicted by existing models.
• Personal monitors detected hydrogen at ground level although at levels below the LFL.
• A small but inconsistent correlation was found between oxygen depletion and the hydrogen concentration.
• A negligible to non-existent correlation was found between in-situ temperature measurements and the hydrogen concentration.

The prototype analyzer is being upgraded for enhanced metrological capabilities including improved real-time spatial and temporal profiling of hydrogen plumes and tracking of prevailing weather conditions. Additional deployments are planned to monitor plume behaviour under different wind humidity and temperature conditions. The data will be shared with the Hydrogen Storage Task Group and ultimately will be used support theoretical models and code requirements prescribed in NFPA 2.

The formation of hydrogen jets from pressurized sources and its ignition when hitting hot devices has been studied by many projects. The transient jets evolve with high turbulence depending on the configuration of the nozzle and especially the pressure in the hydrogen reservoir. In addition the length of the jets and the flames generated by ignition at a hot surface varies. Parameters to be varied were initial pressure of the source (2.5 10 20 and 40 MPa) distance between the nozzle and the hot surface (3 5 and 7 m) and temperature of the hot surface (between 400 and 1000 K). The interaction of the hydrogen jets is visualized by high-speed cinematography techniques which allow analysing the jet characteristics. By combination of various methods of image processing the visibility of the phenomena on the videos taken at 15 000 fps was improved. In addition high-speed NIR spectroscopy was used to obtain temperature profiles of the expanding deflagrations. The jets ignite already above 450 K for conditions mainly from the tubular source at 40 MPa. In addition the propagation of the flame front depends on all three varied parameters: temperature of the hot surface pressure in the reservoir and distance between nozzle and hot surface. In most cases also upstream propagation occurs. A high turbulence seems to lead to the strong deflagrations. At high temperatures of the ignition sources the interaction leads to fast deflagration and speeds up- and downstream of the jet. The deflagration velocity is close to velocity of sound and emission of pressure waves occurs.