2002. Dynamical models of lead-acid batteries: implementation issues

Abstract—This paper explains how the lead-acid models de-scribed in a previous paper can be utilized in practice. Two main issues are opened by that paper: 1) The paper does not supply detailed information on how to identify the several parameters of the proposed models, and 2) it defines a whole family of models, but does not discuss which model of the family is adequate for a given purpose. These two issues are tackled in this paper. For the first issue, the more complex one, two options are proposed and discussed: 1) a complete identification procedure involving exten-sive lab tests and 2) a simplified one that combines information from lab tests and data supplied by the manufacturer. In addition, further simplifications applicable in cases of batteries belonging to the same family are presented. Index Terms—Batteries, modeling, parameter estimation. I. PROPOSED MODEL I N THIS PAPER, the lead-acid battery model family de-scribed in detail in [11] is rapidly reported for reader’s convenience. The model can be represented by an equivalent electric net-work reported in Fig. 1, along with the dynamic equations rep-resenting the charge storage process and the electrolyte heating where so-called “extracted charge, ” i.e., the charge that has been actually extracted from the battery starting from a battery completely full (battery full when); and battery thermal capacitance and resistance, re-spectively; heating power generated inside the battery by conversion from electrical or chemical energy. It has to be remembered that the resistances and ca-pacitances shown in Fig. 1 are function of the battery state-of-charge and electrolyte temperature. The equations defining these parameters are reported in the Appendix. II. IDENTIFICATION OF MODEL PARAMETERS BY LAB TESTS In the diagram of Fig. 2 the battery model is globally de-scribed as having the environment temperature and the bat