Thermal modelling and control of lithium-ion batteries

Licentiate thesis, 2024

The urgent need to reduce greenhouse gas emissions has thrust electrification to the forefront of sustainable solutions. Electric Vehicles (EVs), powered by lithium-ion batteries (LiBs), offer a promising pathway to reducing the transport sector's carbon footprint, which accounts for one-quarter of global CO2 emissions. However, these LiBs, which suffer from the so-called ''Goldilocks syndrome'', exhibit complex nonlinear behaviour and their functionality is strongly influenced by temperature. This necessitates a sophisticated thermal management system capable of controlling the battery temperature within desired limits, regardless of operating conditions. The challenge lies in balancing the trade-off between minimising thermal gradients within the cell and maintaining a low average temperature rise. Achieving this balance requires an optimal combination of tab (terminal) and surface cooling methods to leverage their unique individual strengths. 

In this thesis, we present a new modelling framework for battery cells of different geometries by integrating Chebyshev spectral-Galerkin method and model component decomposition. Consequently, a library of reduced-order computationally efficient two-dimensional battery thermal models is obtained, characterised by different numbers of states. The proposed models allow for the independent control of tab and surface cooling channels for improved thermal performance optimisation. Evaluations under real-world vehicle driving and cooling scenarios demonstrate that these models accurately predict the battery's spatially resolved temperature distribution with minimal errors. Remarkably, the one-state model proves to be both more accurate and computationally efficient than the widely studied and commercially utilised two-state thermal equivalent circuit (TEC) model. Consequently, the proposed model can readily replace the TEC model in existing battery management system applications for enhanced safety and lifetime management. As the developed models enable targeted cooling control to any side of the cell, they are particularly suitable for battery temperature estimation and control in complex cooling scenarios. Furthermore, using these models, the thesis formalises the optimal integration of tab and surface cooling strategies as an optimal control problem and solves it using the model predictive control (MPC) framework. The evaluation of the MPC scheme demonstrates superior thermal performance compared to conventional side and base battery cooling methods. Ultimately, our proposed model and optimal scheme not only enhance immediate thermal performance but also address long-term concerns regarding battery lifespan, safety, and economic viability, representing a valuable advancement in EV battery thermal management.