Automatic Optimization of Metal Hydride Storage Tanks and Novel Designs

Wednesday, November 10, 2010: 10:40 AM
SnowBird Room (Marriott Downtown)
Stephen L. Garrison1, Mikhail B. Gorbounov2, David A. Tamburello3, Bruce Hardy3, Claudio Corgnale4, Daniel A. Mosher5 and Donald Anton6, (1)Savannah River National Laboratory, Aiken, SC, (2)Thermal Management, United Technologies Research Center, East Hartford, CT, (3)Computational Sciences, Savannah River National Laboratory, Aiken, SC, (4)Hydrogen Technology Center, Savannah River National Laboratory, Aiken, SC, (5)Applied Mechanics, United Technologies Research Center, East Hartford, CT, (6)Energy Security, Savannah River National Laboratory, Aiken, SC

A key technical hurdle to an energy economy focused on clean-burning hydrogen over non-renewable petroleum is on-board storage of hydrogen for automotive vehicular applications. Metal hydrides provide one possible solution to the need for high gravimetric and volumetric capacities. However, metal hydrides release significant amounts of heat during hydrogen adsorption, on the order of 0.1 GJ. The subsequent rise in storage temperature results in decreases in maximum hydrogen storage capacity and loading rate at the short fill times mandated by the Department of Energy (DOE) targets. Integrated tank heat exchangers, e.g., cooling tubes and fins, are needed to maximize heat transfer out of system to maximize storage rate and capacity. Given the high cost of experiments (especially at higher pressures), the non-commodity nature of some metal hydrides, and the possibility of air- or water- reactivity, computational modeling provides a significant contribution to a comprehensive methodology for the design, evaluation, and modification of hydrogen storage systems. While many literature studies focus on maximum capacity assuming an (impossible) infinite fill time, full, detailed modeling of the coupled heat and mass transfer and chemical reaction kinetics is needed. The focus of this presentation will be our work developing detailed numerical models that couple reaction kinetics with heat and mass transfer via the general purpose finite element solver COMSOL Multiphysics® for general metal hydride beds and the automatic optimization of the designs by linking the COMSOL models with routines in MATLAB®. Optimization of novel heat exchanger designs will also be presented.

Extended Abstract: File Not Uploaded