Three-dimensional Multi-Particle Electrochemical Model of LiFePO4 Cells based on a Resistor Network Methodology. (1st February 2016)

• Record Type: Journal Article
• Authors: Mastali, Mehrdad; Samadani, Ehsan; Farhad, Siamak; Fraser, Roydon; Fowler, Michael
• Is Part Of: Electrochimica acta. Volume 190(2016)
• Page Start: 574
• Page End: 587
• Publication Date: 2016-02-01
• Subjects: Lithium-ion battery -- LiFePO4 -- Simplified one-dimensional battery model -- Three-dimensional battery model; Electrochemistry -- Periodicals; Electrochemistry, Industrial -- Periodicals; 541.37
• DOI: 10.1016/j.electacta.2015.12.122
• Languages: English
• ISSNs: 0013-4686

Abstract

Highlights: Two simplified one-dimensional models to predict the performance of lithium-ion batteries with LiFePO4 cathode (applicable to any composite cathode material) are presented. A new three-dimensional method for simulating full-sized lithium-ion batteries is developed and compared with the common approach used in literature. ABSTRACT: An efficient fundamental approach for three-dimensional modeling of LiFePO4 (LFP) battery pouch cells is presented in this paper. First, the standard Newman pseudo two-dimensional (P2D) model is compared with two simplified approaches developed by the authors: a simplified electrochemical multi-particle (SEMP) model, and a homogenous pseudo two-dimensional (HP2D) model. It is shown that the SEMP and HP2D models can predict the operating voltage of LFP half-cells with less than 2.5 % and 1.5 % maximum error, respectively, when compared to the P2D model. It is also shown that the simulation time of these two simplified models are one order of magnitude less than the P2D model, hence, they are then used for three-dimensional modeling of the LFP half-cell. Multiple one-dimensional SEMP models are combined, as a first approach, to form a three-dimensional battery model. It is explained that although this method is adequate for predicting the electrochemical current generation distribution, it may introduce errors in heat generation calculations since it does not consider the electrolyte concentration and potential gradient parallel to the current collectors. Therefore, using the HP2D model, an optimized method is proposed that combines the speed and simplicity of the first approach with three-dimensional simulation of the electrolyte. It is shown that this method is able to predict the mentioned gradients that contribute to the battery heat generation.