Modeling of 2LiBH₄+MgH₂ Hydrogen Storage System Accident Scenarios Using Empirical and Theoretical Thermodynamics

It is important to understand and quantify the potential risk resulting from accidental environmental exposure of condensed phase hydrogen storage materials under differing environmental exposure scenarios. This paper describes a modelling and experimental study with the aim of predicting consequences of the accidental release of 2LiBH₄+MgH₂ from hydrogen storage systems. The methodology and results developed in this work are directly applicable to any solid hydride material and/or accident scenario using appropriate boundary conditions and empirical data.
The ability to predict hydride behaviour for hypothesized accident scenarios facilitates an assessment of the risk associated with the utilization of a particular hydride. To this end, an idealized finite volume model was developed to represent the behaviour of dispersed hydride from a breached system. Semi-empirical thermodynamic calculations and substantiating calorimetric experiments were performed in order to quantify the energy released, energy release rates and to quantify the reaction products resulting from water and air exposure of a lithium borohydride and magnesium hydride combination.
The hydrides, LiBH₄ and MgH₂, were studied individually in the as-received form and in the 2:1 “destabilized” mixture. Liquid water hydrolysis reactions were performed in a Calvet calorimeter equipped with a mixing cell using neutral water. Water vapor and oxygen gas phase reactivity measurements were performed at varying relative humidities and temperatures by modifying the calorimeter and utilizing a gas circulating flow cell apparatus. The results of these calorimetric measurements were compared with standardized United Nations (UN) based test results for air and water reactivity and used to develop quantitative kinetic expressions for hydrolysis and air oxidation in these systems. Thermodynamic parameters obtained from these tests were then inputted into a computational fluid dynamics model to predict both the hydrogen generation rates and concentrations along with localized temperature distributions. The results of these numerical simulations can be used to predict ignition events and the resultant conclusions will be discussed.