An improved single-particle model with electrolyte dynamics for high current applications of lithium-ion cells. (1st September 2021)

• Record Type: Journal Article
• Authors: Mehta, Rohit; Gupta, Amit
• Is Part Of: Electrochimica acta. Volume 389(2021)
• Publication Date: 2021-09-01
• Subjects: Lithium-ion cells -- Single-particle model -- Electrochemical model -- Electrolyte transport; Electrochemistry -- Periodicals; Electrochemistry, Industrial -- Periodicals; 541.37
• DOI: 10.1016/j.electacta.2021.138623
• Languages: English
• ISSNs: 0013-4686

Abstract

Highlights: Development of an improved single-particle model for high C-rate applications. Effect of non-linear spatial distribution of electrolyte potential, overpotential and open-circuit potential incorporated, leading to significantly improved prediction of the cell voltage. Ease of implementation for battery management systems as a linear system of differential-algebraic equations in state-variables need to be solved. Criteria for series-truncation of a semi-analytical model for electrode particle diffusion proposed. Model validated against pseudo-2D and improvement to previous reduced-order models demonstrated. Graphical abstract: Abstract: In this work, an isothermal, reduced-order polynomial-based model is proposed for lithium-ion cells. The correction incorporated into the single-particle model considers the effect of electrolyte dynamics leading to improved model predictions at high current applications without adding any significant computational complexity. The model includes the spatial distribution of the electrode and electrolyte potentials and the electrolyte lithium concentration on the resulting cell voltage. A novel approach to account for the spatially varying overpotential and open-circuit potential is proposed. A linear system of differential-algebraic equations is obtained in the state variables and the cell voltage is computed using a non-linear expression in these states. The resulting linear state-space system makes the model easily implementable for state-estimation algorithms in battery management system applications. Additionally, a semi-analytical model is incorporated for the diffusion of lithium in electrodes and a criterion for truncating the infinite-series solution is proposed for achieving a balance between accuracy and complexity for cells subjected to fluctuating currents. The proposed model results in a decrease in error for cell voltage prediction by a factor of five when compared with existing enhanced single-particle models.