Pressure effects of an ignited release from onboard storage in a garage with a single vent

https://doi.org/10.1016/j.ijhydene.2018.07.130Get rights and content

Highlights

  • Overpressure due to ignited hydrogen release in a garage examined.

  • Pressure due to ignited and unignited release in enclosure compared.

  • Pressure from ignited release two orders of magnitude greater than unignited.

  • Phenomenon should be considered by regulators and engineers.

Abstract

A numerical study has been performed comparing the hazards, in particular overpressures, arising from the sustained unignited and ignited release from an onboard hydrogen storage tank at 700 bar through a 3.34 mm diameter orifice, representing a thermally activated pressure relief device (TPRD) in a small garage with a single vent equivalent in area to small window. It has been demonstrated how the overpressure predicted in the case of an unignited release using both CFD and an analytical model is in the region of 0.55 kPa and thus unlikely to cause structural damage. However, the overpressure predicted for the ignited release is two orders of magnitude greater, reaching over 55 kPA in less than 1 s and thus potentially causing destruction of the structure.

It has been shown that whilst the overpressures resulting from the unignited release are unlikely to cause harm, the garage is engulfed by a flammable atmosphere in less than 1 s and the oxygen is depleted to levels dangerous to people within this time. In the case of the ignited release, whilst the resultant overpressures are the primary safety concern, it has been shown how the thermal effects resulting from the release extend almost 9 m from the jet in 1.5 s.

Introduction

The commercial use of fuel cell and hydrogen (FCH) technologies is becoming more widespread and they will soon form an essential part of our built environment. There is clear need with emerging technologies that a safety level is maintained which is at least the same as those in existing fossil fuel applications. The number of hydrogen-powered vehicles in use worldwide is growing, and commercialisation is fast approaching a reality, leading to a growth in the necessary indoor use of FCH technologies e.g. material handling, forklifts etc. or parking of these vehicles i.e. in a garage or car park. By understanding the hazards arising due to placement of hydrogen-fuelled vehicles in confined environments, steps can be taken towards reduction of associated hazards and risks by inherently safer design. Indeed, safe indoor use has been the focus of recent investigations, in particular within the HyIndoor Project [2], [3].

In the majority of passenger cars hydrogen is commonly stored as a compressed gas in tanks. Typical storage pressures for vehicle tanks are in the region of 350 bar–700 bar. Onboard hydrogen storage tanks are required by regulation to be equipped with pressure relief devices (PRDs) [4]. These are fitted to the fuel tank and function by releasing the fluid in the event of an abnormally high temperature, e.g. in conditions of fire. Current PRDs provide rapid release of the hydrogen, thus minimising the possibility of catastrophic failure of the tank during exposure to fire. Existing TPRDs intend to vent the hydrogen before this catastrophic rupture occurs preventing disastrous explosions. High mass flow rates from TPRDs are potentially acceptable outdoors, where the buoyancy of hydrogen is an advantage in aiding dispersion below the lower flammability limit. However, from a safety perspective a number of hazards arise following a high mass flow rate release, characteristic for current TPRDs, in a confined space containing a vent. Previous work on this topic by the authors has focused on the overpressure development within an enclosure due to an unignited release. Preliminary numerical and analytical modelling work on this topic by the authors focused on a hypothetical scenario, with a constant mass flow rate release [5] and the phenomenon of a rapid rise in pressure following the unignited release of hydrogen through a “typical” TPRD (diameter 5.08 mm) in an enclosure with a small vent was discovered and explained. It was demonstrated, how for a constant release of 0.39 kg/s of hydrogen into a 30.4 m3 garage with a single vent the size of one brick the overpressure within the enclosure resulting from the injection of hydrogen reaches a level of 10–20 kPa, capable of destroying the garage, within only 2 s. The high volumetric flow rate of hydrogen results in these significant overpressures even without combustion. For the chosen scenario, if the enclosure does not rupture first, the pressure within the garage, reaches a maximum level in excess of 50 kPa for 350 bar storage and 100 kPa for 700 bar. This maximum pressure then drops off and tends towards a steady state value, an order of magnitude lower, and equal to that predicted by the simple steady state estimations. For the very specific case examined it was clear that unacceptable levels of overpressure above 10–20 kPa are reached within a short timeframe of 1–2 s and subsequent work considered “safe” vent sizes for the “pressure peaking” occurring for the same unignited release rate and enclosure size [6]. The latter work focusing on more “realistic” scenarios, utilising a blow-down model developed at Ulster University to account for mass flow rate decay [7] and suggested “safe” TPRD diameters for enclosures of different volumes with different natural ventilation levels. This previous work, specifically the geometry and scenario considered “safe” for unignited releases forms the basis of the work presented here. The pressure peaking phenomenon for unignited jets, and the predicted overpressures has been validated against laboratory scale experiments [8]. It should be emphasised that the pressure peaking phenomenon is not evident with other, heavier, fuels such as propane. Whilst a small pressure rise is evident for e.g. methane for a comparable leak with minimal ventilation, the peak is almost two orders of magnitude lower than that resulting from the hydrogen leak and unlikely to cause structural damage [5].

To date the work and recommendations concerning pressure peaking have focused solely on unignited releases. In the case of TPRD activation, the most likely cause is fire. Hence, when considering a high mass flow release from a TPRD in an enclosure, the scenario of an ignited release cannot be ignored. There has been limited published work on hydrogen fires in enclosures, and none on numerical simulation of ignited pressure peaking phenomenon, however, it should be noted that there have been many studies on free jet fires as discussed by Molkov and Saffers [9]. Recent work at Ulster has focused on analytical modelling of the problem [8]. In the case of TPRD release, scenarios with an initially high mass flow rate are most representative. However, the most recent experiments on hydrogen fires in enclosures have focused on laboratory scale releases [10] and are being used by the authors for validation of the CFD approach to enclosure fire modelling, with an emphasis on overpressure prediction. This approach has then been applied to a hypothetical scenario where pressure peaking is expected to occur. Whilst recommendations have been presented for TPRD diameters which will prevent dangerous over pressures in the case of an unignited release [6], it is highly likely that the pressure resulting from these “safe” diameters will be significantly higher in the event of an ignited release. Hence, this work is driven by the need to understand the hazards resulting from the rapid ignited release of hydrogen from onboard storage tanks through a TPRD inside a garage-like enclosure with low natural ventilation i.e. the consequences of a jet fire which has been immediately ignited, delayed ignition is not considered here. The resultant overpressure is of particular interest. The pressure peaking phenomenon for an ignited release had not been studied numerically and compared with that for an equivalent unignited release and it should be emphasised that this work has relevance beyond a TPRD scenario and may occur anywhere a momentum driven release occurs in an enclosure with minimum ventilation.

Section snippets

Problem description

Both an unignited and ignited hydrogen release, from typical onboard hydrogen storage tanks at 700 bar through a TPRD, are considered in a garage type enclosure with a single vent. The hypothetical events involve conservative “quasi-steady” constant mass flow rate releases of 0.2993 kg/s through a TPRD diameter of 3.34 mm, pressure drop in the storage tank was not considered. The TPRD opening time was taken as 0.01 s. This diameter is lower and hence potentially “safer” than typical 5.08 mm

Methodology

CFD was used for both the unignited and ignited cases and the phenomenological model for pressure peaking was used in the unignited case.

Results

The CFD results for both the ignited and unignited release are discussed in the following section, and the pressure dynamics of the unignited case is compared with the analytical prediction.

Conclusions

A hypothetical scenario has been considered where hazards have been investigated, in particular overpressures which arise from the sustained unignited and ignited release from an onboard hydrogen storage tank at 700 bar through a 3.34 mm diameter orifice, representing the TPRD in a small garage with a single vent equivalent in area to small window. The work has focused on the overpressures arising purely as a result of the ignited and unignited release and the primary fire which may have

References (17)

There are more references available in the full text version of this article.

Cited by (11)

  • Numerical study of hydrogen leakage, diffusion, and combustion in an outdoor parking space under different parking configurations

    2023, Renewable and Sustainable Energy Reviews
    Citation Excerpt :

    In previous studies, the effects of release pressure and orifice size on hydrogen distribution in semi-closed spaces have been considered for risk mitigation, which could be useful in assessing safety issues in mining-like tunnels [18]. Preliminary numerical and theoretical models have been focused on the pressure-peaking phenomenon caused by momentum-dominated hydrogen release in enclosures with minimal ventilation [19,20]. However, the pressure-peaking phenomenon is not caused by hydrogen release from typical TPRD orifice diameters in parking spaces, despite minimal ventilation requirements [21].

  • Thermal and fire characteristics of hydrogen jet flames in the tunnel at longitudinal ventilation strategies

    2021, Fuel
    Citation Excerpt :

    ANSYS FLUENT is a commercial CFD code and provides comprehensive modeling capabilities for a wide range of incompressible and compressible, laminar and turbulent fluid flow problems [42,43]. It has been applied in fire scenarios such as garages, ordinary rooms, tunnels, and so on [44–46]. Among them, researchers have revealed the accuracy and usefulness of FLUENT modelling in fire scenarios through comparing full-scale experimental data with simulation results [47,48].

  • Dispersion of hydrogen release in a naturally ventilated covered car park

    2020, International Journal of Hydrogen Energy
    Citation Excerpt :

    It was demonstrated how a 0.5 mm diameter TPRD resulted in a considerably smaller flammable cloud than that produced by either a 2 mm or 3.34 mm diameter TPRD. Concerns have been previously highlighted for the use of “typical” TPRD diameters in enclosures with relation to the pressure peaking phenomena [10]. This work indicates that safety concerns also exist for “typical” diameters in enclosure scenarios where the ventilation is sufficient to ensure that pressure peaking is not of concern.

View all citing articles on Scopus
View full text