Safety

Experiments on flame propagation regimes in a turbulent hydrogen jet with velocity and hydrogen concentration gradients have been performed at the FZK hydrogen test site HYKA. Horizontal stationary hydrogen jets released at normal and cryogenic temperatures of 290K and 80 and 35K with different nozzle diameters and mass flow rates in the range from 0.3 to 6.5 g/s have been investigated. Sampling probe method and laser PIV technique have been used to evaluate distribution of hydrogen concentration and flow velocity along and across the jet axis. High-speed photography (1000 fps) combined with a Background Oriented Schlieren (BOS) system was used for the visual observation of the turbulent flame propagation. In order to investigate different flame propagation regimes the ignition position was changed along the jet axis. It was found that the maximum flame velocity and pressure loads can only occur if the hydrogen concentration at the ignition point exceeds 11% of hydrogen in air. In this case the flame propagates in both directions up- and downstream the jet flow whereas in the opposite case the flame propagates only downstream. Such a behavior is consistent with previous experiments according to that the flame is able to accelerate effectively only if the expansion rate σ of the H2-air mixture is higher than a critical value σ* = 3.75 (like for the 11% hydrogen-air mixture). The measured data allow conservative estimates of the safety distance and risk assessment for realistic hydrogen leaks.

Experience makes a superior teacher. Sharing the details surrounding safety events is one of the best ways to help prevent their recurrence elsewhere. This approach requires an open non-punitive environment to achieve broad benefits. The Hydrogen Incident Reporting Tool (www.h2incidents.org) is intended to facilitate the sharing of lessons learned and other relevant information gained from actual experiences using and working with hydrogen and hydrogen systems. Its intended audience includes those involved in virtually any aspect of hydrogen technology systems and use with an emphasis towards energy and transportation applications. The database contains records of safety events both publicly available and/or voluntarily submitted. Typical records contain a general description of the occurrence contributing factors equipment involved and some detailing of consequences and changes that have been subsequently implemented to prevent recurrence of similar events in the future. The voluntary and confidential nature and other characteristics surrounding the database mean that any analysis of apparent trends in its contents cannot be considered statistically valid for a universal population. A large portion of reported incidents have occurred in a laboratory setting due to the typical background of the reporting projects for example. Yet some interesting trends are becoming apparent even at this early stage of the database’s existence and general lessons can already be taken away from these experiences. This paper discusses the database and a few trends that have already become apparent for the reported incidents. Anticipated future uses of this information are also described. This paper is intended to encourage wider participation and usage of the incidents reporting database and to promote the safety benefits offered by its contents.

Computational Fluid Dynamics (CFD) tools have been increasingly employed for carrying out quantitative risk assessment (QRA) calculations in the process industry. However these tools must be validated against representative experimental data in order to have a real predictive capability. As any typical accident scenario is quite complex it is important that the CFD tool is able to predict combined release and ignition scenarios reasonably well. However this kind of validation is not performed frequently primarily due to absence of good quality data. For that reason the recent experiments performed by FZK under the HySafe internal project InsHyde (http://www.hysafe.org) are important. These involved vertically upwards hydrogen releases with different release rates and velocities impinging on a plate in two different geometrical configurations. The dispersed cloud was subsequently ignited and pressures recorded. These experiments are important not only for corroborating the underlying physics of any large-scale safety study but also for validating the important assumptions used in QRA. Blind CFD simulations of the release and ignition scenarios were carried out prior to the experiments to predict the results (and possibly assist in planning) of the experiments. The simulated dispersion results are found to correlate reasonably well with experimental data in terms of the gas concentrations. The overpressures subsequent to ignition obtained in the blind predictions could not be compared directly with the experiments as the ignition points were somewhat different but the pressure levels were found to be similar. Simulations carried out after the experiments with the same ignition position as those in the experiments compared reasonably well with the measurements in terms of the pressure level. This agreement points to the ability of the CFD tool FLACS to model such complex scenarios well. Nevertheless the experimental set-up can be considered to be small-scale and less severe than many accidents and real-life situations. Future large-scale data of this type will be valuable to confirm ability to predict large-scale accident scenarios.

An accidental hydrogen release in equipment enclosures may result in the presence of a detonable mixture in a confined environment. Numerical simulation is potentially a useful tool for damage assessment in these situations. To assess the value of CFD techniques numerical simulation of detonation was performed for two realistic scenarios. The first scenario starts with a pipe failure in an electrolyzer resulting in a leak of 42 g of hydrogen. The second scenario deals with a failure in a reformer where 84 g of hydrogen is released. In both cases dispersion patterns were first obtained from separate numerical simulation and were then used as initial condition in a detonation simulation based upon the reactive Euler's equations. Energy was artificially added in a narrow region to simulate detonative ignition. In the electrolyzer ignition was assumed to occur 500 ms after beginning of the release. Results show a detonation failing on the top and bottom side but propagating left and right before eventually failing also. Average impulse was 500 Ns/m². For the reformer three cases were simulated with ignition 1.0 1.4 and 2.0 seconds after the beginning of the release. In two cases the detonation wave failed everywhere except in the direction of the release in which it continued propagating until reaching the side wall. In the third the detonation failed everywhere at first but later a deflagration to detonation transition occurred resulting in a strong wave that propagated rapidly toward the side wall. In all three cases the consequences are more serious than in the electrolyzer.

In the development of hydrogen technology special attention is paid to the technical problems of hydrogen storage. One possible way is cryogenic storage in liquid form. Generally cryo-technical machines need components with interacting surfaces in relative motion such as bearings seals or valves which are subjected to extreme conditions. Materials of such systems have to be resistant to friction-caused mechanical deformation at the surface low temperatures and hydrogen environment. Since materials failure can cause uncontrolled escape of hydrogen new material requirements are involved for these tribo-systems in particular regarding operability and reliability. In the past few years several projects dealing with the influence of hydrogen on the tribological properties of friction couples were conducted at the Federal Institute for Materials Research and Testing (BAM) Berlin. This paper reports some  investigations carried out with polymer composites. Friction and wear were measured for continuous sliding and analyses of the worn surfaces were performed after the experiments. Tests were performed at room temperature in hydrogen as well as in liquid hydrogen.

In this paper the detonation propensity of different compositions of mixtures of hydrogen propane and methane with air has been evaluated over a wide range of compositions. We supplement the conventional calculations of the induction delay with calculations of the characteristic acceleration parameter recently suggested by Radulescu Sharpeand Bradley(RSB) to characterize the instability of detonations. While it is well established that the ignition delay provides a good measure for detonability the RSB acceleration or its non-dimensionalform provides a further discriminant between mixtures with similar ignition delays. The present assessment of detonability reveals that while a stoichiometric mixture of hydrogen-air has an ignition delay one and two orders of magnitude shorter than respectively propane and methane hydrogen also has a parameter smaller by respectively one and two orders of magnitude. Its smaller propensity for instability is reflected by an RSB acceleration parameter similar to the two hydrocarbons. The predictions however indicate that lean hydrogen mixtures are likely to be much more unstable than stoichiometric ones. The relation between the parameter and potential to amplify an unstable transverse wave structure has been further determined through numerical simulation of decaying reactive Taylor-Sedov blast waves. Using a simplified two-step model calibrated for these fuels we show that methane mixtures develop cellular structures more readily than propane and hydrogen when observed on similar induction time scales. Future work should be devoted towards a quantitative inclusion of the RSB parameter in assessing the detonability of a given mixture.

Stationary pressure vessels for the storage of large volumes of gaseous hydrogen at high pressure (>70 MPa) are typically manufactured from Cr-Mo steels. These steels display hydrogen-enhanced fatigue crack growth but pressure vessels can be manufactured using defect-tolerant design methodologies. However storage volumes are limited by the wall thickness that can be reliably manufactured for quench and tempered Cr-Mo steels typically not more than 25-35 mm. High-hardenability steels can be manufactured with thicker walls which enables larger diameter pressure vessels and larger storage volumes. The goal of this study is to assess the fracture and fatigue response of high hardenability Ni-Cr-Mo pressure vessel steels for use in high-pressure hydrogen service at pressure in excess of 1000 bar. Standardized fatigue crack growth tests were performed in gaseous hydrogen at frequency of 1Hz and for R-ratios in the range of 0.1 to 0.7. Elastic-plastic fracture toughness measurements were also performed. The measured fatigue and fracture behavior is placed into the context of previous studies on fatigue and fracture of Cr-Mo steels for gaseous hydrogen.

Explosion venting is a frequently used measure to mitigate the consequence of gas deflagrations in closed environments. Despite the effort to predict the vent area needed to achieved the protection through engineering formulas and CFD tools work has still to be done to reliably predict the outcome of a vented gas explosion. Most of available data derived from experimental campaigns performed in the past involved homogeneous conditions while especially in the case of a very buoyant gas such as hydrogen the most probable scenario that can follow and unintended release in a closed environment foresee the ignition of a stratified inhomogeneous mixture. University of Pisa performed experimental tests in a 25 m3 facility in homogeneous and inhomogeneous conditions. The present paper is aimed to share the results of hydrogen dispersion and deflagration tests and discuss the comparison of maximum peak overpressure generated in the two scenarios. Description of the experimental set-up includes all the details deemed necessary to reproduce the phenomenon with a CFD tool.

Explosion venting is a prevention/mitigation solution widely used in the process industry to protect indoor equipment or buildings from excessive internal pressure caused by an accidental explosion. Vented explosions are widely investigated in the literature for various geometries hydrogen/air concentrations ignition positions initial turbulence etc. In real situations the vents are normally covered by a vent panel. In the case of an indoor leakage the hydrogen/air cloud will be stratified rather than homogeneous. Nowadays there is a lack in understanding about the vented explosion of stratified clouds and about the influence of vent cover inertia on the internal overpressure. This paper aims at shedding light on these aspects by means of experimental investigation of vented hydrogen/air deflagration using an experimental facility of 1m3 and via numerical simulations using the computational fluid dynamics (CFD) code FLACS

The thermal hazards from ignited under-expanded cryogenic releases are not yet fully understood and reliable predictive tools are missing. This study aims at validation of a CFD model to simulate flame length and radiative heat flux for cryogenic hydrogen jet fires. The simulation results are compared against the experimental data by Sandia National Laboratories on cryogenic hydrogen fires from storage with pressure up to 5 bar abs and temperature in the range 48–82 K. The release source is modelled using the Ulster's notional nozzle theory. The problem is considered as steady-state. Three turbulence models were applied and their performance was compared. The realizable k-ε model showed the best agreement with experimental flame length and radiative heat flux. Therefore it has been employed in the CFD model along with Eddy Dissipation Concept for combustion and Discrete Ordinates (DO) model for radiation. A parametric study has been conducted to assess the effect of selected numerical and physical parameters on the simulations capability to reproduce experimental data. DO model discretisation is shown to strongly affect simulations indicating 10 × 10 as minimum number of angular divisions to provide a convergence. The simulations have shown sensitivity to experimental parameters such as humidity and exhaust system volumetric flow rate highlighting the importance of accurate and extended publication of experimental data to conduct precise numerical studies. The simulations correctly reproduced the radiative heat flux from cryogenic hydrogen jet fire at different locations.