A novel diagnostic technique for the quantification of lithium-ion battery degradation modes under in- operando conditions
File(s)
Author(s)
Prosser, Ryan
Type
Thesis
Abstract
Reliable and frequent diagnostics of lithium-ion batteries is critical to extending battery life and by extension reducing the total cost of ownership of electric vehicles. State-of-the-art
quantitative diagnostic techniques almost all require open circuit voltage versus state of charge profiles for them to be used; unfortunately, these measurements are rarely available
in normal device operation. Therefore, this thesis contains the theory and experimental validation of a novel diagnostic technique which is usable in conditions seen under normal device operation. The technique uses cell heat generation rate versus time as its diagnostic signal and can predict electrode capacities and the lithium inventory of the cell. From experimental validation of the diagnostic technique, its predicted outputs are compared to a state-of-the-art open circuit voltage model-based technique. Thorough analysis of the identifiability of the parameters shows that the novel diagnostic technique is capable of robustly predicting the positive electrode capacity to a similar accuracy as the state-of-the-art technique for high-powered and high energy cells if a reduced order diffusion model is included in the technique. Not only has a diagnostic technique capable of quantifying positive electrode capacity been
developed, but a framework for evaluation of any diagnostic technique’s efficacy has also been developed. This work is missing from the diagnostic literature and can hopefully be used by the academic community to help build better diagnostic techniques which will in turn help extend battery lifetime and reduce the total cost of ownership of electric vehicles.
quantitative diagnostic techniques almost all require open circuit voltage versus state of charge profiles for them to be used; unfortunately, these measurements are rarely available
in normal device operation. Therefore, this thesis contains the theory and experimental validation of a novel diagnostic technique which is usable in conditions seen under normal device operation. The technique uses cell heat generation rate versus time as its diagnostic signal and can predict electrode capacities and the lithium inventory of the cell. From experimental validation of the diagnostic technique, its predicted outputs are compared to a state-of-the-art open circuit voltage model-based technique. Thorough analysis of the identifiability of the parameters shows that the novel diagnostic technique is capable of robustly predicting the positive electrode capacity to a similar accuracy as the state-of-the-art technique for high-powered and high energy cells if a reduced order diffusion model is included in the technique. Not only has a diagnostic technique capable of quantifying positive electrode capacity been
developed, but a framework for evaluation of any diagnostic technique’s efficacy has also been developed. This work is missing from the diagnostic literature and can hopefully be used by the academic community to help build better diagnostic techniques which will in turn help extend battery lifetime and reduce the total cost of ownership of electric vehicles.
Version
Open Access
Date Issued
2022-06-29
Date Awarded
2026-03-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Patel, Yatish
Offer, Gregory
Publisher Department
Department of Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)