Abstract. In this work we present and discuss a modelling framework for the basic discharge process which occurs in simple electrochemical battery cells. The main purpose is to provide a setting for analyzing delivered capacity, battery life expectancy and other measures of performance. This includes a number of deterministic and stochastic variations of kinetic battery models. The primary tool is a novel phase plane analysis of the balance of nominal and theoretical capacity. In particular, we study spatial versions of such models which lead to a linear diffusion equation with Robin type boundary conditions under scaling. Explicit solutions are obtained by considering reflected Brownian motion.

In this paper we review several approaches to mathematical modeling of simple battery cells and develop these ideas further with emphasis on charge recovery and the response behavior of batteries to given external load. We focus on models which use few parameters and basic battery data, rather than detailed reaction and material characteristics of a specific battery cell chemistry, starting with the coupled ODE linear dynamics of the kinetic battery model. We show that a related system of PDE with Robin type boundary conditions arises in the limiting regime of a spatial kinetic battery model, and provide a new probabilistic representation of the solution in terms of Brownian motion with drift reflected at the boundaries on both sides of a finite interval. To compare linear and nonlinear dynamics in kinetic and stochastic battery models we study Markov chains with states representing available and remaining capacities of the battery. A natural scaling limit leads to a class of nonlinear ODE, which can be solved explicitly and compared with the capacities obtained for the linear models. To indicate the potential use of the modeling we discuss briefly comparison of discharge profiles and effects on battery performance.

No abstract