Safety

Computational Fluid Dynamics (CFD) is applied to investigate the near exit jet behavior of high pressure hydrogen release into quiescent ambient air through different types of orifices. The size and geometry of the release hole can affect the possibility of auto-ignition. Therefore the effect of release geometry on the behavior and development of hydrogen jet issuing from non-axisymmetric (elliptical) and expanding orifices is investigated and compared with their equivalent circular orifices. A three-dimensional in-house code is developed using the MPI library for parallel computing to simulate the flow based on an inviscid approximation. Convection dominates viscous effects in strongly underexpanded supersonic jets in the vicinity of release exit justifying the use of the Euler equations. The transport (advection) equation is applied to calculate the concentration of hydrogen-air mixture. The Abel-Nobel equation of state is used because high pressure hydrogen flow deviates from the ideal gas assumption. This work effort is conducted to fulfill two objectives. First two types of circular and elliptic orifices with the same cross sectional area are simulated and the flow behavior of each case is studied and compared during the initial stage of release. Second the comparative study between expanding circular exit and its fixed counterpart is carried out. This evaluation is conducted for different sizes of nozzle with different aspect ratios.

Catastrophic rupture of onboard hydrogen storage in a fire is a safety concern. Different passive e.g. fireproofing materials the thermally activated pressure relief device (TPRD) and active e.g. initiation of TPRD by fire sensors safety systems are being developed to reduce hazards from and associated risks of high-pressure hydrogen storage tank rupture in a fire. The probability of such low-frequency highconsequences event is a function of fire resistance rating (FRR) i.e. the time before tank without TPRD ruptures in a fire the probability of TPRD failure etc. This safety issue is “confirmed” by observed recently cases of CNG tanks rupture due to blocked or failed to operate TPRD etc. The increase of FRR by any means decreases the probability of tank rupture in a fire particularly because of fire extinction by first responders on arrival at an accident scene.<br/>This study of socio-economic effects of safety applies a quantitative risk assessment (QRA) methodology to an example of hydrogen vehicles with passive tank protection system on roads in London.<br/>The risk is defined here through the cost of human loss per fuel cell hydrogen vehicle (FCHV) fire accident and fatality rate per FCHV per year. The first step in the methodology is the consequence analysis based on validated deterministic engineering tools to estimate the main identified hazards: overpressure in the blast wave at different distances and the thermal hazards from a fireball in the case of catastrophic tank rupture in a fire. The population can be exposed to slight injury serious injury and fatality after an accident. These effects are determined based on criteria by Health and Safety Executive (UK) and a cost metrics is applied to the number of exposed people in these three harm categories to estimate the cost per an accident. The second step in the methodology is either the frequency or the probability analysis. Probabilities of a vehicle fire and failure of the thermally activated pressure relief device are taken from published sources. A vulnerability probit function is employed to calculate the probability of emergency operations’ failure to prevent tank rupture as a function of a storage tank FRR and time of fire brigade arrival. These later results are integrated to estimate the tank rupture frequency and fatality rate. The risk is presented as a function of fire resistance rating.<br/>The QRA methodology allows to calculate the cost of human loss associated with an FCHV fire accident and demonstrates how the increase of FRR of onboard storage as a safety engineering measure would improve socio-economics of FCHV deployment and public acceptance of the technology.

Hydrogen sensors are recognized as an important technology for facilitating the safe implementation of hydrogen as an alternative fuel and there are numerous reports of a sensor alarm successfully preventing a potentially serious event. However gaps in sensor metrological specifications as well as in their performance for some applications exist. The U.S. Department of Energy (DOE) Fuel Cell Technologies Office published a short list of critical gaps in the 2007 and 2012 Multiyear Project Plans; more detailed gap analyses were independently performed by the Joint Research Centre (JRC) and the National Renewable Energy Laboratory (NREL). There have been however some significant advances in sensor technologies since these assessments including the commercial availability of hydrogen sensors with fast response times (t90 < 1 s which had been an elusive DOE target since 2007) improved robustness to chemical poisons improved selectivity and improved lifetime and stability. These improvements however have not been universal and typically pertain to select platforms or models. Moreover as hydrogen markets grow and new applications are being explored more demands will be imposed on sensor performance. The hydrogen sensor laboratories at NREL and the JRC are currently updating the hydrogen safety sensor gap analysis through direct interaction with international stakeholders in the hydrogen community especially end users. NREL and the JRC are currently organizing a series of workshops (in Europe and the United States) with sensor developers end-users and other stakeholders in 2017 to identify technology gaps and to develop a path forward to address them. One workshop was held on May 10 in Brussels Belgium at the Headquarters of the Fuel Cell and Hydrogen Joint Undertaking. A second workshop is planned at NREL in Golden CO USA. This paper reviews improvements in sensor technologies in the past 5 to 10 years identifies gaps in sensor performance and use requirements and identifies potential research strategies to address the gaps. The outcomes of the Hydrogen Sensors Workshops are also summarized.

Deflagration-to-Detonation Transition Ratio (DDTR) is an important parameter in measuring the hazard of hydrogen detonation at given thermodynamic conditions. It’s among the major tasks to evaluate DDTR in the study of hydrogen safety in a nuclear containment. With CFD tools detailed distribution of thermodynamic parameters at each instant can be simulated with considerable reliability. Then DDTR can be estimated using related CFD output. Forstochastic or epistemic reasons uncertainty always exists in input parameters during computations. This lack of accuracy can finally be reflected in the uncertainty of computation results e.g. DDTR in our consideration. The analysis of the influence of the input uncertainty is therefore a key step to understand the model’s response on the output and possibly to improve the accuracy. The increase of computational power makes it possible to perform statistics-based sensitivity and uncertainty (SU) analysis on CFD simulations. This paper aims at presenting some ideas on the procedure in safety analysis on hydrogen in nuclear containment. A hydrogen recombiner case is constructed and simulated with CFD method. DDTR at each instant is computed using a semi-empirical method. RBD-FAST based SU analysis is performed on the result.

Consequence modelling of high potential risks of usage and transportation of cryogenic liquids yet requires substantial improvements. Among the cryogenics liquid hydrogen (LH2) needs especial treatments and a comprehensive understanding of spill and spread of liquid and dispersion of vapor. Even though many of recent works have shed lights on various incidents such as spread dispersion and explosion of the liquid over land less focus was given on spill and spread of LH2 onto water. The growing trend in ship transportation has enhanced risks such as ships’ accidental releases and terrorist attacks which may ultimately lead to the release of the cryogenic liquid onto water. The main goal of the current study is to present a computational fluid dynamic (CFD) approach using OpenFOAM to model release and spread of LH2 over water substrate and discuss previous approaches. It also includes empirical heat transfer equations due to boiling and computation of evaporation rate through an energy balance. The results of the proposed model will be potentially used within another coupled model that predicts gas dispersion]. This work presents a good practice approach to treat pool dynamics and appropriate correlations to identify heat flux from different sources. Furthermore some of the previous numerical approaches to redistribute or in some extend manipulate the LH2 pool dynamic are brought up for discussion and their pros and cons are explained. In the end the proposed model is validated by modelling LH2 spill experiment carried out in 1994 at the Research Centre Juelich in Germany.

In the present study the diffusion process of hydrogen leaking from a FCV (Fuel Cell Vehicle) in an underground parking garage is analyzed by numerical simulations in order to assess the risk of a leakage accident. The temporal and spatial evolution of the hydrogen concentration as well as the flammable region in the parking garage was predicted numerically. The effects of the leakage flow rate and an additional ventilation fan were investigated to evaluate the ventilation performance to relieve the accumulation of the hydrogen gas. The volume of the flammable region shows a non-linear growth in time and rapidly increases eventually. The present numerical analysis can provide a physical insight and quantitative data for safety of various hydrogen applications.

The efficiency of gas sensor application for facilitating the safe use of hydrogen depends to a considerable extent on the response time of the sensor to change in hydrogen concentration. The response and recovery times have been measured for five different hydrogen sensors three commercially available and two promising prototypes which operate at room temperature. Experiments according to ISO 26142 show that most of the sensors surpass much for a concentration change from clean to hydrogen containing air the demands of the standard for the response times t(90) and values of 2 to 16 s were estimated. For an opposite shift to clean air the recovery times t(10) are from 7 to 70 s. Results of transient behaviour can be fitted with an exponential approach. It can be demonstrated that results on transient behaviour depend not only from investigation method and the experimental conditions like gas changing rate and concentration jump as well as from operating parameters of sensors. In comparison to commercial MOS and MIS-FET hydrogen sensors new sensor prototypes operating at room temperature possesses in particular longer recovery times.

This work presents a parametric study on the similitude between hydrogen and helium distribution when released in the air by a source located inside of a naturally ventilated enclosure with two vents. Several configurations were experimentally addressed in order to improve knowledge on dispersion. Parameters were chosen to mimic operating conditions of hydrogen energy systems. Thus the varying parameters of the study were mainly the source diameter the releasing flow rate the volume and the geometry of the enclosure. Two different experimental set-ups were used in order to vary the enclosure's height between 1 and 2 m. Experimental results obtained with helium and hydrogen were compared at equivalent flow rates determined with existing similitude laws. It appears for the plume release case that helium can suitably be used for predicting hydrogen dispersion in these operating designs. On the other hand – when the flow turns into a jet – non negligible differences between hydrogen and helium dispersion appear. In this case helium – used as a direct substitute to hydrogen – will over predict concentrations we would get with hydrogen. Therefore helium concentration read-outs should be converted to obtain correct predictions for hydrogen. However such a converting law is not available yet.

The aim of this study was to validate a model for predicting overpressure arising from accidental hydrogen releases in areas with limited ventilation. Experiments were performed in a large-scale setup that included a steel-reinforced container of volume 14.9 m³ and variable ventilation areas and mass flow rates. The pressure peaking phenomenon characterized as transient overpressure with a characteristic peak in a vented enclosure was observed during all the experiments. The model description presented the relationship between the ventilation area mass flow rate enclosure volume and discharge coefficient. The experimental results were compared with two prediction models representing a perfect mix and the real mix. The perfect mix assumed that all the released hydrogen was well stirred inside the enclosure during the releases. The real mix prediction s used the hydrogen concentration and temperature data measured during experiments. The prediction results with both perfect mix and real mix showed possible hazards during unignited hydrogen releases.

Unintended releases can occur during the production storage transportation and filling of liquid hydrogen which may cause devastating consequences. In the present work liquid hydrogen leak is modeled in ANSYS Fluent with the numerical model validated using the liquid hydrogen spill test data. A three-layer artificial neural network (ANN) model is built in which the wind speed ground temperature leakage time and leakage rate are taken as the inputs the horizontal diffusion distance and vertical diffusion distance of combustible gas as the outputs of the ANN. The representative sample data derived from the detailed calculation results of the numerical model are selected via the orthogonal experiment method to train and verify the back propagation (BP) neural network. Comparing the calculation results of the formula fitting with the sample data the results show that the established ANN model can quickly and accurately predict the horizontal and vertical diffusion distance of flammable vapor cloud relatively. The influences of four parameters on the horizontal hazard distance as well as vertical hazard height are predicted and analyzed in the case of continuous overflow of liquid hydrogen using the ANN model.