D1B.3B Preliminary work plan for explosion testing in gas stations

The starting point in the transition from natural gas to hydrogen is the need for the distribution and use of hydrogen to be at least as safe as natural gas. This document describes a literature review on realistic scenarios for the occurrence of flammable gas mixtures in and around gas stations. After completing the literature review, attention turned to the question of how to gain a better insight into these realistic scenarios, to the probability of the occurrence of flammable gas mixtures (in gas stations) and also to the hazardous situations that might arise as a result. This document is part research report (Section 4, literature review) and part research proposal (Section 5 et seq.). The first three sections are of a general nature and introduce readers to both the literature review and the research proposal.

The object of the process outlined above is to answer the following question: ‘How probable is the occurrence of an explosion in or close to a gas station?’. In other words, the probability of ignition and, as such, the occurrence of a hazardous situation (a flash fire, fire or explosion) too. The ultimate aim is to formulate recommendations for follow-up actions like follow-up research or the amendment of standards.

The objective of the research is to gain an insight into the distribution pattern of natural gas and hydrogen in the event of a leak in a gas station and also into the risk (probability and impact) that arises if an ignition source with sufficient ignition energy is added. This objective will be achieved by answering the following sub-questions:

1. Which scenarios in which a flammable mixture occurs in a gas station are realistic? Section 6.2 describes the relevant parameters.
2. Which impact needs to be taken into consideration in these scenarios? The objective of the research described in the research proposal is to qualitatively describe this impact. The research data obtained must be sufficient to facilitate quantitative analyses (including calculations on issues like pressure waves and heat radiation) in follow-up research, if required.

An inventory has been made of the information available from national and international research on explosions in gas stations for natural gas and hydrogen and from research on ignition sources. It reveals that much research has been done on the explosion behaviour of gas stations and the ignition potential of the equipment commonly used. However, the parameters used in the explosion behaviour studies are different to the parameters applicable in realistic scenarios. The leak sizes in most studies conducted previously were chosen with the object of facilitating an explosive gas-air mixture, in order to study the effects of explosions of gas cabinets. The focus was not on proving that the leak sizes chosen were in fact realistic or on performing tests with realistic leak sizes. Also, much research on ignition sources did not involve the ignition sources likely in the vicinity of a gas station (for example, a hair dryer or a bread toaster). This proposal has been written on the basis of the useful information obtained from these research projects. The results obtained from the literature review are described in Section 3.

The research described in this proposal consists of the following three steps:

1. Identify realistic scenarios in which ignition/explosion situations could arise in or near gas stations/gas cabinets. This will be done by organising an expert meeting in which all of the various relevant parameters are discussed. These parameters will include realistic leak sizes, weather conditions and potential ignition sources.
2. Perform Computational Fluid Dynamics (CFD) calculations. In this step, leaks in gas stations will be modelled on the basis of the scenarios identified in Step 1. Calculations will be performed for each scenario, to establish whether and where an ignitable gas mixture will occur. One advantage of CFD calculations in comparison with practical tests is their ability to dictate specific scenarios, making it possible to manipulate input parameters to compare specific interests (wind speed and wind direction, for example). This facilitates the formulation of a shortlist of worst-case (or most-probable-case) scenarios to be tested. CFD calculations also make it possible to maintain the stability of parameters that are difficult to control during testing (wind speed and wind direction, for example). This improves the certainty of the outcomes of these calculations, particularly the probability of the occurrence of an ignitable mixture. This method also enables researchers to study the build-up of gas concentrations in and around a gas station or gas cabinet.
3. Experiments. A shortlist of CFD calculation outcomes, including the worst-case scenario, will be verified by simulating the conditions modelled in a controlled experiment. An ignition source with sufficient ignition energy will be added to gain an insight into whether or not the mixture will actually ignite in the situations modelled and what the consequences are if it does.