(539d) System Level Component Models for Electrochemical Power Sources in Hybrid Environments | AIChE

(539d) System Level Component Models for Electrochemical Power Sources in Hybrid Environments


Diwakar, V. - Presenter, Tennessee Technological University
Boovaragavan, V. - Presenter, Tennessee Technological University

A combination of power sources is usually referred as a hybrid power system. Hybrids usually consist of combination of one or more of electrochemical power sources (fuel cell, battery, and super-capacitor) with an electrical motor (and/or other electrical components). Analysis and simulation of hybrid power system require simultaneous series/parallel simulations of a fuel cell, secondary (rechargeable) battery, super capacitor, a DC motor and many other components in reasonable time at low cost of computation and the operating conditions can be charge/discharge at a known load or varying load or a combination of different conditions.

For the purpose of modeling and analyzing electrochemical hybrids, a full-blown PDE model is not ideal because of time constraint and the other components (DC motor, capacitor) are usually described by a dynamic equation in time. A system level dynamic model is a differential equation (in time) model. Material and energy balances of batteries and fuel cells should be ideally written by ordinary differential equations in time. However, spatial variations inside the electrochemical power sources cannot be neglected and if neglected, the models fail at various operating conditions. Since spatial variations cannot be neglected for fuel cells and batteries, there are no existing system level models for batteries and fuel cells that predict the electrochemical behavior of power sources with reasonable accuracy in real time without having to use expensive processors.

This paper addresses the development and implementation of efficient approximate dynamic models for electrochemical power sources for hybrid environment. These models will be simple enough to be used in hybrid environment and at the same time robust enough to predict the electrochemical behavior affected by spatial variations. Preliminary results of the efficient schemes will be presented.