Publications

In this work real scale experiments involving hydrogen dispersion inside a road tunnel have been modelled using the Computational Fluid Dynamics (CFD) methodology. The aim is to assess the performance of the ADREA-HF CFD tool against full-scale tunnel dispersion data resulting from high-pressure hydrogen leakage through Thermal Pressure Relief Device (TPRD) of a vehicle. The assessment was performed with the help of experiments conducted by the French Alternative Energies and Atomic Energy Commission (CEA) in a real inclined tunnel in France. In the experiments helium as hydrogen surrogate has been released from 200 bar storage pressure. Several tests were carried out examining different TPRD sizes and release directions (upwards and downwards). For the CFD evaluation two tests were considered: one with downwards and one with upwards release both through a TPRD with a diameter of 2 mm. The comparison between the CFD results and the experiments shows the good predictive capabilities of the ADREA-HF code that can be used as a safety tool in hydrogen dispersion studies. The comparison reveals some of the strengths and weaknesses of both the CFD and the experiments. It is made clear that CFD can contribute to the design of the experiments and to the interpretation of the experimental results.

The development of hydrogen production technologies and new uses represents an opportunity to accelerate the ecological transition and create a new industrial sector. However the risks associated with the use of hydrogen must be considered. Mitigation of a hydrogen explosion in an enclosure is partly based on prevention strategies such as detection and ventilation and protection strategies such as explosion venting. Even if applications involving hydrogen probably are most interesting for vented explosions in weak structures the extreme reactivity of hydrogen-air mixtures often excludes the use of regular venting devices such as in highly constrained urban environments. Thus having alternative mitigation solutions can make the effects of the explosion acceptable by reducing the flame speed and the overpressure loading or suppressing the secondary explosion. The objective of this paper is to present experimental studies of the mitigation of hydrogen-air deflagration in a 4 m3 vented enclosure by injection of inhibiting powder (NaHCO₃). After describing the experimental set-up the main experimental results are presented for several trial configurations showing the influence of inhibiting powder in the flammable cloud on flame propagation. An interpretation of the mitigating effect of inhibiting powder on the explosion effects is proposed based on the work of Proust et al.

High-accuracy gas dispersion models are necessary for predicting hydrogen movement and for reducing the damage caused by hydrogen release accidents in chemical processes. In urban areas where obstacles are large and abundant computational fluid dynamics (CFD) would be the best choice for simulating and analyzing scenarios of the accidental release of hydrogen. However owing to the large computation time required for CFD simulation it is inappropriate in emergencies and real-time alarm systems. In this study a non-linear surrogate model based on deep learning is proposed. Deep convolutional layer data-driven autoencoder and batch normalized deep neural network is used to analyze the effects of wind speed wind direction and release degree on hydrogen concentration in real-time. The typical parameters of hydrogen diffusion accidents at hydrogen refuelling stations were acquired by CFD numerical simulation approach and a database of hydrogen diffusion accident parameters is established. By establishing an appropriate neural network structure and associated activation function a deep learning framework is created and then a deep learning model is constructed. The accuracy and timeliness of the model are assessed by comparing the results of the CFD simulation with those of the deep learning model. To develop a dynamic reconfiguration prediction model for the hydrogen refuelling station diffusion scenario the algorithm is continuously enhanced and the model is improved. After training is finished the model's prediction time is measured in seconds which is 105 times quicker than field CFD simulations. The deep learning model of hydrogen release in hydrogen refuelling stations is established to realize timely and accurate prediction and simulation of accident consequences and provide decision-making suggestions for emergency rescue and personnel evacuation which is of great significance for the protection of human life health and property safety.

In confined spaces hydrogen released with low momentum tends to accumulate in a layer below the ceiling; the concentration in this layer rises and can rapidly enter the flammability range. In this context ventilation is a key safety equipment to prevent the formation of such flammable volumes. To ensure its well-sizing to each specific industrial context it is necessary to dispose of reliable engineering models. Currently the existing engineering models dealing with the buoyancy-driven H2 dispersion in a ventilated enclosure mainly focus on the natural-ventilation phenomenon. However forced ventilation is in some situations more adapted to the industrial context as the wind direction and intensity remains constant and under control. Therefore two existing wind-assisted ventilation models elaborated by Hunt and Linden [1] and Lowesmith et al. [2] were tested on forced ventilation applications. The main assumption consists in assuming a blowing ventilation system rather than a suction system as the composition and velocity of the entering air are known. The fresh air enters the down opening and airhydrogen mixture escapes through the upper one. The adapted models are then validated with experimental data releasing helium rather than hydrogen. Experiments are conducted on a 1-m3 ventilated box controlling the release and ventilation rates. The agreement between both analytical and experimental results is discussed from the different comparisons performed.

In this paper we study the dispersion ignition and flame characteristics of blended jets of hydrogen and methane (as a proxy for natural gas) at near-atmospheric pressure for a fixed volumetric flow rate which mimics the scenario of a small-scale unintended leak. A reduction in flame height is observed with increasing hydrogen concentration. A laser is tightly focused to generate a spark with sufficient energy to ignite the fuel. The light-up boundary defined as the delineating location at which a spark ignites into a jet flame or extinguishes is determined as a contour. The light-up boundary increases in both width and length as the hydrogen content increases up to 75% hydrogen at which point the axial ignition boundary decreases slightly for pure hydrogen relative to 75% hydrogen. Ignition probability a key parameter regarding safety is computed at various axial locations and is also shown to be higher near the nozzle as well as non-zero at further downstream locations as the hydrogen content in the blend increases. Planar laser Raman scattering is used in separate experiments to determine the concentration of both fuel species. Mean fuel concentrations well below the lower flammability limit are both within the light-up boundary and have non-zero ignition probabilities.

Hydrogen jet fire occurs with high probability when hydrogen leaks from high-pressure equipment. The hydrogen jet fire is characterized by its high velocity and energy. Computational Fluid Dynamics (CFD) numerical analysis is a prominent way to predict the potential hazards associated with hydrogen jet fire. Validation of the CFD model is essential to ensure and quantify the accuracy of numerical results. This study focuses on the validation of the hydrogen jet fire model using Fire Dynamic Simulation (FDS). Hydrogen release is modeled using high-speed Lagrangian particles released from a virtual nozzle thus avoiding the modeling of the actual nozzle. The mesh size sensitivity analysis of the model is carried out in a container-size domain with 0.04m – 0.08m resolution of the jet. The model is validated by comparing gas temperatures and heat fluxes with test data. The promising results demonstrated that the model could predict the hazardous influence of the jet fire.

Quantitative Risk Assessment (QRA) is a risk-informed approach that considers past performances and the likelihood of events and distinguishes must-haves from nice-to-haves. Following the approach applied for the HyRAM code developed by the Sandia National Laboratories a QRA toolkit for hydrogen systems was developed using MATLAB by Canadian Nuclear Laboratories (CNL). Based on user inputs for system components and their operating parameters the toolkit calculates the consequence of a hydrogen leak from the system. The fatality likelihood can be estimated from the severity of a person’s exposure to radiant heat flux (from a jet fire) and overpressure (from an explosion). This paper presents a verification and validation exercise by comparing the CNL model predictions with the HyRAM code and available experimental data including a QRA case study for a locomotive. The analysis produces risk contours recommending personnel (employees/public) numbers time spent and safe separation distances near the incident (during maintenance or an accident). The case study demonstrated the importance of hydrogen leak sensors’ reliability for leak detection and isolation. The QRA toolkit calculates a more practical value of the safe separation distance for hydrogen installations and provides evidence to support communication with authorities and other stakeholders for decision-making.

The large deployment of hydrogen technologies for new applications such as heat power mobility and other emerging industrial utilizations is essential to meet targets for CO2 reduction. This will lead to an increase in the number of hydrogen installations nearby local populations that will handle hydrogen technologies. Local regulations differ and provide different safety and/or separation distances in different geographies. The purpose of this work is to give an insight on different methodologies and recommendations developed for hydrogen (mainly) risk management and consequences assessment of accidental scenarios. The first objective is to review available methodologies and to identify the divergent points on the methodology. For this purpose a survey has been launched to obtain the needed inputs from the subtask participants. The current work presents the outcomes of this survey highlighting the gaps and suggesting the prioritization of the actions to take to bridge these gaps.

This paper presents a comprehensive review of existing explosion mitigation techniques for tunnels and evaluates their applicability in scenarios of hydrogen tank rupture in a fire. The study provides an overview of the current state of the art in tunnel explosion mitigation and discusses the challenges associated with hydrogen explosions in the context of fire incidents. The review shows that there are several approaches available to decrease the effects of explosions including wrapping the tunnel with a flexible and compressible barrier and introducing energy-absorbing flexible honeycomb elements. However these methods are limited to the mitigation of the action and do not consider either the mitigation of the structural response or the effects on the occupants. The study highlights how the structural response is affected by the duration of the action and the natural period of the structural elements and how an accurate design of the element stiffness can be used in order to mitigate the structural vulnerability to the explosion. The review also presents various passive and active mitigation techniques aimed at mitigating the explosion effects on the occupants. Such techniques include tunnel branching ventilation openings evacuation lanes right-angled bends drop-down perforated plates or high-performance fibre-reinforced cementitious composite (HPFRCC) panels for blast shielding. While some of these techniques can be introduced during the tunnel's construction phase others require changes to the already working tunnels. To simulate the effect of blast wave propagation and evaluate the effectiveness of these mitigation techniques a CFD-FEM study is proposed for future analysis. The study also highlights the importance of considering these mitigation techniques to ensure the safety of the public and first responders. Finally the study identifies the need for more research to understand blast wave mitigation by existing structural elements in the application for potential accidents associated with hydrogen tank rupture in a tunnel.

To achieve the net zero carbon emissions goals by 2050 the transition to cleaner forms and carriers of energy should be accelerated without though jeopardizing the public safety. Although hydrogen has been deemed to play significant role in the energy transition for years now there are still concerns for its risks that hamper its widespread implementation in several applications like for instance automobile applications. Hydrogen-powered vehicles raise concerns about their safety especially inside confined spaces like tunnels and thus research on that topic has been intensified during the last years. In this context experiments have been conducted by UK HSE within the EU-funded project HyTunnel-CS to examine hydrogen dispersion and deflagration inside a scaled tunnel resulting from fuel cell car bus and train release.<br/>In this work that was also carried out within the HyTunnel-CS we present the Computational Fluid Dynamics (CFD) simulations of the HSE unignited experiments. Blowdown tests related to high-pressure hydrogen releases through Thermal Pressure Relief Device (TPRD) installed in car and in train were modeled using the ADREA-HF code. The scope of these simulations was two-fold: a) contribute to the design of the experiments (e.g. indicate sensor positioning ignition point etc.) and the interpretation of hydrogen behavior and b) validate the CFD code. For the former pre-test simulations preceded the experiments to provide design recommendations. When the experiments were conducted the measurements were used for the code validation. Overall the CFD results are in satisfactory agreement with the experiments. Finally simulations with different ventilation rates and with model vehicles inside the tunnel were conducted to examine their effect on mixture dispersion and tunnel safety.