Hydrogen-powered vehicles and buses are already on the road, and major car companies have announced the commercialization to begin in 2015. Volume of Type 3 and Type 4 on-board hydrogen storage tanks is ranging from 7 to 170 litres. They have maximum storage pressure either 35 MPa or 70 MPa. Similarly, stand-alone high pressure storage tanks can be found as a part of the emerging hydrogen infrastructure, e.g. at refuelling stations. The volume of a stationary storage tank could be of a few cubic meters up to tens of cubic meters and pressure could be as high as 100 MPa. There is a number of knowledge gaps in hydrogen safety engineering [1]. One of these gaps is how to calculate a deterministic separation distance from either a vehicle or a stationary storage tank when they rupture in a fire. There is widely spread opinion that in the case of high pressure hydrogen tank rupture, hydrogen combustion would not contribute to the blast wave as shock is so strong and propagates so fast that the combustion being slower process does not contribute to the blast wave strength. However, this opinion is not supported by detailed analysis of available data.
The existing methodology of a blast wave decay calculation suggested by Baker et al. [2] was applied in this study to reproduce experimental data by Weyandt obtained in bonfire tests with a stand-alone and an under-vehicle tanks with similar experimental parameters [3]. The methodology does not account for the contribution of combustion to the blast wave. It is thought the reason why the prediction of experimental data [3] by this existing methodology was poor. This is especially valid for the under-vehicle test. It is worth noting that the mechanical energy coefficient 2 was applied for prediction of measured overpressures due to the hemispherical symmetry of the tests (lower value of the coefficient 1.8, which is accounting for ground cratering, gives close result).
A new model to calculate blast wave decay that takes into account the combustion of hydrogen in air is developed and presented. The model is able to reproduce closely published experimental data. Due to highly non-ideal behaviour of hydrogen at pressures up to 100 MPa the Brode’s equation [3] used in the methodology by Baker et al [2] was changed in our model to the one for real gas (the equation for mechanical energy was derived using Abel-Noble equation of state for real gas).
The chemical energy of combustion is introduced into the model dynamically in the assumption that a fraction of hydrogen reacts with air behind the shock instantly to feed the blast. This fraction of chemical energy is added to the mechanical energy of compressed gas to calculate the energy-scaled non-dimensional distance. The values of total mechanical energy in test 1 (stand-alone tank) and test 2 (under-vehicle tank) were 5.23 MJ and 5.95 MJ respectively. The values of the stored total chemical energy were 198 MJ in test 1, and 230.9 MJ in test 2.
The completeness of released hydrogen combustion is accounted for by the empirical coefficient, which value is in the range 0.05-0.15 that is within previously published estimates. Thus, the amount of chemical energy released is comparable to the amount of mechanical energy stored in compressed hydrogen. This empirical coefficient accounts for both the completeness of hydrogen combustion behind the shock (in time comparable with shock propagation time) and the losses on thermal radiation from the fireball. As was expected, the higher value of this coefficient was observed for under-vehicle tank bonfire test. This is partially due to more intensive mixing of hydrogen with air by the vehicle frame, and partially due to slower shock propagation because of significant losses of the mechanical energy spent to damage the vehicle body and move it significantly by 22 m [2]. While the mechanical energy coefficient for the stand-alone tank test was as expected 1.8-2.0, in the case of under-vehicle test the value of the coefficient is less by more than an order of magnitude due to losses in the presence of the vehicle above the tank.
The model was able to reproduce experimental data on the rupture of a stand-alone stationary hydrogen tank and a tank under the vehicle in bonfire tests, in contrast to existing models. The model can be applied for the estimation of overpressure and impulse in a blast wave from typical hydrogen applications using the published information about harm effects from blast waves on humans and civil structures. Thus, it can be used as a tool for hydrogen safety engineering to assess deterministic separation distance from a vehicle in case of catastrophic failure of on-board storage tank or a stand-alone storage tank.
References
- V. Molkov, Fundamentals of Hydrogen Safety Engineering. bookboon.com, free download eBook, Part 1 and Part 2, 2012.
- W. E. Baker, P. A. Cox, P. S. Westine, J. J. Kulesz, and R. A. Strehlow, Explosion hazards and evaluation. Elsevier Scientific Publishing Company, 1983.
- N. Weyandt, Analysis of Induced Catastrophic Failure Of A 5000 psig Type IV Hydrogen Cylinder, Southwest Research Institute report for the Motor Vehicle Fire Research Institute, 01.06939.01.001, 2005.
- H. L. Brode, Blast wave from a spherical charge, Phys. Fluids, Vol. 2, No. 2, 1959.
