Safety

Recent modeling efforts of non-equilibrium effects in detonations have suggested that hydrogen-based detonations may be affected by vibrational non-equilibrium of the hydrogen and oxygen molecules effects which could explain discrepancies of cell sizes measured experimentally and calculated without relaxation effects. The present study addresses the role of vibrational relaxation in 2H2/O2 detonations by considering two-bath gases argon and helium. These two gases have the same thermodynamic and kinetic effects when relaxation is neglected. However due to the bath gases differences in molecular weight and reduced mass differences which affect the molecular collisions relaxation rates can be changed by approximately 50-70%. Experiments were performed in a narrow channel in mixtures of 2H2/O2/7Ar and 2H2/O2/7He to evaluate the role of the bath gas on detonation cellular structures. The experiments showed differences in velocity deficits and cell sizes for experimental conditions keeping the induction zone length constant in each of the mixtures. These differences were negligible in sensitive mixtures but increased with the increase in velocity deficits while the cell sizes approaching the channel dimensions. Near the limits differences of cell size in two mixtures approached a factor of 2. These differences were however reconciled by accounting for the viscous losses to the tube walls evaluated using a modified version of Mirels' laminar boundary layer theory and generalized Chapman-Jouguet theory for eigenvalue detonations. The experiments suggest that there is an influence of relaxation effects on the cellular structure of detonations which is more sensitive to wall boundary conditions. However the previous works showed that the impact of vibrational non-equilibrium in a mixture of H2/Air is more visible due to the effects of N2 in the air slowest to relax. Previous discrepancies suggested to be indicative of relaxation effects should be reevaluated by the inclusion of wall loss effects.

The development of reliable hydrogen sensors is crucial for the safe use of hydrogen. One of the main concerns of end-users is sensor reliability in the presence of species other than the target gas which can lead to false alarms or undetected harmful situations. In order to assess the selectivity of commercial of the shelf (COTS) hydrogen sensors a number of sensors of different technology types were exposed to various interferent gas species. Cross-sensitivity tests were performed in accordance to the recommendations of ISO 26142:2010 using the hydrogen sensor testing facilities of NREL and JRC-IET. The results and conclusions arising from this study are presented.

The minimum distances between exposures and bulk liquid hydrogen listed in the National Fire Protection Agency’s Hydrogen Technology Code NFPA 2 are based on historical consensus without a documented scientific analysis. This work follows a similar analysis as the scientific justification provided in NFPA 2 for exposure distances from bulk gaseous hydrogen storage systems but for liquid hydrogen. Validated physical models from Sandia’s HyRAM software are used to calculate distances to a flammable concentration for an unignited release the distance to critical heat flux values and the visible flame length for an ignited release and the overpressure that would occur for a delayed ignition of a liquid hydrogen leak. Revised exposure distances for bulk liquid hydrogen systems are calculated. These distances are related to the maximum allowable working pressure of the tank and the line size as compared to the current exposure distances which are based on system volume. For most systems the exposure distances calculated are smaller than the current distances for Group 1 they are similar for Group 2 while they increase for some Group 3 exposures. These distances could enable smaller footprints for infrastructure that includes bulk liquid hydrogen storage tanks especially when using firewalls to mitigate Group 3 hazards and exposure distances. This analysis is being refined as additional information on leak frequencies is incorporated and changes have been proposed to the 2023 edition of NFPA 2.

Throughout the history of the mining petroleum process and nuclear industries continuous efforts have been made to develop and improve measures to prevent and mitigate accidental explosions. Over the coming decades energy systems are expected to undergo a transition towards sustainable use of conventional hydrocarbons and an increasing share of renewable energy sources in the global energy mix. The variable and intermittent supply of energy from solar and wind points to energy systems based on hydrogen or hydrogen-based fuels as the primary energy carriers. However the safety-related properties of hydrogen imply that it is not straightforward to achieve and document the same level of safety for hydrogen systems compared to similar systems based on established fuels such as petrol diesel and natural gas. Compared to the conventional fuels hydrogen-air mixtures have lower ignition energy higher combustion reactivity and a propensity to undergo deflagration-to-detonation-transition (DDT) under certain conditions. To achieve an acceptable level of safety it is essential to develop effective measures for mitigating the consequences of hydrogen explosions in systems with certain degree of congestion and confinement. Extensive research over the last decade have demonstrated that chemical inhibition or partial suppression can be used for mitigating the consequences of vapour cloud explosions (VCEs) in congested process plants. Total and cooperation partners have demonstrated that solid flame inhibitors injected into flammable hydrocarbon-air clouds represent an effective means of mitigating the consequences of VCEs involving hydrocarbons. For hydrogen-air explosions these same chemicals inhibitors have not proved effective. It is however well-known that hydrocarbons can affect the burning velocity of hydrogen-air mixtures greatly. This paper gives an overview over previous work on chemical inhibitors. In addition experiments in a 20-litre vessel have been performed to investigate the effect of combinations of hydrocarbons and alkali salts on hydrogen/air mixtures.

For the successful deployment of the Heavy Duty (HD) hydrogen vehicles an associated infrastructure in particular hydrogen refueling stations (HRS) should be reliable compliant with regulations and optimized to reduce the related costs. FCH JU project PRHYDE aims to develop a sophisticated protocol dedicated to HD applications. The target of the project is to develop protocol and recommendations for an efficient refueling of 350 500 and 700 bar HD tanks of types III and IV. This protocol is based on modeling results as well as experimental data. Different partners of the PRHYDE European project are closely working together on this target. However modeling approaches and corresponding tools must first be compared and validated to ensure the high level of reliability for the modeling results. The current paper presents the benchmark performed in the frame of the project by Air Liquide Engie Wenger Engineering and NREL. The different models used were compared and calibrated to the configurations proposed by the PRHYDE project. In addition several scenarios were investigated to explore different cases with high ambient temperatures.

Three different types of symmetrical tilt grain boundaries Ȉ3 Ȉ11 and Ȉ27 were constructed to study the dislocation behavior under the indentation on bi-crystal nickel. After hydrogen charging the number of hydrogen atoms in the Ȉ3 sample is the smallest and gradually increases in Ȉ11 and Ȉ27 samples. The force-displacement curve of indentation shows that the deformation resistance of the Ȉ3 sample is significantly higher than that of Ȉ11 and Ȉ27 samples. With the presence of grain boundaries the deformation resistance of Ȉ11 and Ȉ27 samples is significantly improved while the deformation resistance of the Ȉ3 VDPSOH is weakened. The indentation depth during the formation of dislocations in single  crystals  is  significantly  greater  than  that  of  bi-crystals.  Grain   boundaries  slow  down  the dislocation  propagation speed.  Compared  with  the  bi-crystals   without  hydrogen  the  presence  of hydrogen reduces the deformation resistance and accelerates the dislocation propagation.

The development of safe dispensing equipment for the fueling of heavy duty (HD) vehicles is critical to the expansion of this newly and quickly expanding market. This paper discusses the development of a HD dispenser and nozzles assembly (nozzle hose breakaway) for these new larger vehicles where flow rates are more than double compared to light duty (LD) vehicles. This equipment must operate at nominal pressures of 700 bar -40o C gas temperature and average flow rate of 5-10 kg/min at a high throughput commercial hydrogen fueling station without leaking hydrogen. The project surveyed HD vehicle manufacturers station developers and component suppliers to determine the basic specifications of the dispensing equipment and nozzle assembly. The team also examined existing codes and standards to determine necessary changes to accommodate HD components. From this information the team developed a set of specifications which will be used to design the dispensing equipment. In order to meet these goals the team performed computational fluid dynamic pressure modelling and temperature analysis in order to determine the necessary parameters to meet existing safety standards modified for HD fueling. The team also considered user operational and maintenance requirements such as freeze lock which has been an issue which prevents the removal of the nozzle from LD vehicles. The team also performed a failure mode and effects analysis (FMEA) to identify the possible failures in the design. The dispenser and nozzle assembly will be tested separately and then installed on an innovative HD fueling station which will use a HD vehicle simulator to test the entire system.

In  a  liquid  hydrogen  storage  tank  hydrogen  vapor  exists  above  the  cryogenic  liquid.  A  common modeling  assumption  of  a  liquid  hydrogen  tank  is  thermodynamic  equilibrium. However  this assumption may not hold in all conditions. A non-equilibrium storage tank with a pressure relief valve and a burst disc in parallel was modeled in this work. The model includes different boiling regimes to handle scenarios with high heat transfer. The model was first validated with a scenario where normal boil-off from an unused tank was compared to experimental data. Then four abnormal tank scenarios were explored: a loss of vacuum in the insulation layer a high ambient temperature (to simulate an engulfing fire) a high ambient temperature with a simultaneous loss of vacuum and high conduction through the insulation layer. The burst disc of the tank opened only in the cases with extreme heat transfer to the tank (i.e. fire with a loss of vacuum and high insulation conductivity) quickly releasing the hydrogen. In the cases with only a loss of vacuum or only external heat from fire the pressure relief valve on the tank managed to moderate the pressure below the burst disc activation pressure. The high insulation conductivity case highlights differences between the equilibrium and non-equilibrium tank models. The mass loss from the tank through the burst disc is slower using a non-equilibrium model because mass transfer from the liquid to gas phase within the tank becomes limiting. The implications of this model and how it can be used to help inform safety codes and standards are discussed.

Building  the  international  hydrogen  supply  chain  requires  the  large-scale  liquefied   hydrogen(LH2) carrier. During shipping LH2 with LH2 Carrier the tank is pressurized by LH2 evaporation due to heat ingress from outside. Before unloading LH2 at the receiving terminal reducing the tank pressure is essential for the safe tank operation. However pressure reduction might cause flashing leading to rapid  vaporization  of liquefied  hydrogen  liquid  leakage.   Moreover  it  was  considered  that  pressure recovery phenomenon which was not preferred in terms of tank pressure management occurred at the beginning  of  pressure  reduction.  Hence  the   purpose  of  our  research  is  to  clarify  the phenomenon inside  the  cargo  tank  during   pressure  reduction.  The  CFD  analysis  of  the  pressure  reduction phenomenon was conducted with the   VOF based in-house CFD code utilizing the C-CUP scheme combined  with  the  hybrid   Level  Set  and  MARS  method.  In  our  previous  research  the  pressure reduction  experiments   with  the  30  m³  LH2  tank  were  simulated  and  the  results  showed  that  the pressure   recovery  was  caused  by  the  boiling  delay  and  the  tank  pressure  followed  the   saturation pressure  after  the  liquid  was  fully  stirred.  In  this  paper  the  results  were re-evaluated  in  terms  of temperature.  While  pressure  reduction  was  dominant  the  temperature  of  vapor-liquid  interface decreased.  Once  the  boiling  bubble  stirred  the  interface  its  temperature  reached  the  saturation temperature  after  pressure  recovery  occurred.  Moreover  it  was  found  that  the  liquid  temperature during  pressure  reduction  could  not  be  measured  because  of  the  boiling  from  the  wall  of  the thermometer. The CFD analysis on pressure reduction of 1250 m³ tank for the LH2 Carrier was also very could occur in the case of the 1250 m³ tank in a certain condition. These results provide new insight into the development of the LH2 carrier.

NREL has been developing compliance verification tools for allowable hydrogen levels prescribed by the Global Technical Regulation Number 13 (GTR-13) for hydrogen fuel cell electric vehicles (FCEVs). As per GTR-13 FCEV exhaust is to remain below 4 vol% H2 over a 3-second moving average and shall not at any time exceed 8 vol% H2 and that this requirement is to be verified with an analyzer that has a response time of less than 300 ms. To be enforceable a means to verify regulatory requirements must exist. In response to this need NREL developed a prototype analyzer that meets the GTR metrological requirements for FCEV exhaust analysis. The analyzer was tested on a commercial fuel cell electric vehicle (FCEV) under simulated driving conditions using a chassis dynamometer at the Emissions Research and Measurement Section of Environment and Climate Change Canada and FCEV exhaust was successfully profiled. Although the prototype FCEV Exhaust Analyzer met the metrological requirements of GTR-13 the stability of the hydrogen sensor was adversely impacted by condensed water in the sample gas. FCEV exhaust is at an elevated temperature and nearly saturated with water vapor. Furthermore condensed water is present in the form of droplets. Condensed water in the sample gas collected from FCEV exhaust can accumulate on the hydrogen sensing element which would not only block access of hydrogen to the sensing element but can also permanently damage the sensor electronics. In the past year the design of the gas sampling system was modified to mitigate against the transport of liquid water to the sensing element. Laboratory testing confirmed the effectiveness of the modified sampling system water removal strategy while maintaining the measurement range and response time required by GTR-13. Testing of the upgraded analyzer design on an FCEV operating on a chassis dynamometer is scheduled for the summer of 2021.