(424e) Simulation of the Rapid Charging of a Metal Hydride Hydrogen Storage System

Authors: 
Ebner, A. D., University of South Carolina
Wang, Y., University of South Carolina
Ritter, J. A., University of South Carolina


Recent results from a two-to-the-four full factorial design corresponding to 16 fast charging experiments carried out with the SRNL metal hydride storage system showed that on average it took over 180 min to fully charge the system with just over 3000 liters of H2 at STP. In one of the best cases, this could be achieved in just over 60 min. However, under certain circumstances, over 1000 liters of H2 at STP could be charged to the system in under 1 min. These results suggested that it may indeed be possible to fully charge this system in the very desirable 5 to 10 min time frame. Therefore, the objective of this study is to use a mathematical modeling approach to find the conditions that foster the most rapid charging of this metal hydride hydrogen storage system in an attempt to achieve a fully charged state in under 10 min.

This objective will be achieved by formulating a two-dimensional model of the metal hydride H2 storage system using the COMSOL MultiphysicsTM software package. This model will be calibrated to the fast charging experiments and then used to study the fast charging behavior of this metal hydride H2 storage system over a vast range of conditions. The model will consist of a system of equations that includes mass, energy and momentum balances, as well as an LDF kinetic approach to follow the dynamics of the pressure, temperature, gas velocities and metal hydride loadings within a cross sectional area of the H2 storage bed.

To describe the experimental metal hydride H2 storage system developed by the SRNL, the model will consider four homogeneous but different regions within the bed. One region will contain metal hydride. Another region will not contain metal hydride. The third region will contain a varying number of U-tube heat exchanger tubes that will use water as the coolant. The fourth region will contain the tube for charging and discharging H2 to and from the bed. The bed cross sectional area will also be assumed to remain the same along the bed, that is, in the direction of the heat exchanger tubes and hydrogen discharge tube.

All the mass, energy and momentum balances within each region will be assumed to be limited to the plane containing the bed cross sectional area except for the hydrogen charge tube. The flow within the hydrogen charge tube will be allowed to vary perpendicular to this plane. This flow will also be assumed to vary linearly with distance and to access the hydrogen tube according to well-known choked or subsonic relationships defined for compressible gas.

The fast charging experiments will be used to obtain the effective thermal conductivity of the metal hydride bed and the mass transfer coefficient of H2 into the metal hydride. A linear loading dependence will be assumed for the bed effective thermal conductivity. Sieder-Tate correlations will be used to evaluate the heat transfer coefficients at the heat exchanger walls. This SRNL H2 storage system utilizes the commercially viable Lm1.06Ni4.96Al0.04 metal hydride for which pressure-composition-temperature data were available and already fitted to a dual Langmuir model.

Several parameters will be studied to evaluate the conditions that foster fast charging of this metal hydride bed. These parameters will include the heat exchanger water temperature and flow, the number of heat exchanger tubes, their geometric placement within the bed cross-section, the heat exchanger to bed cross sectional ratio and the bed thermal conductivity. The linear variation of water temperature and flow will also be investigated. The results of this study will be presented and discussed.