Design optimisation and diagnostics for lithium-ion battery fast charging
File(s)
Author(s)
Campbell, Ian D.
Type
Thesis
Abstract
This thesis begins with a literature review of lithium-ion battery behaviour. Emphasis is placed on thermodynamics, kinetics and the reversible lithium plating reaction. Advancements made to the battery simulation software LIONSIMBA, resulting in the creation of v2.0, are presented.
With LIONSIMBA as one component, a framework is developed and demonstrated for the optimal balancing of cell energy and power density in electric vehicles. Vehicle simulations are performed for acceleration and fast charging. The framework thus spans from the vehicle level to that of a single electrochemical cell layer. The framework employs layer reconfiguration to produce energy-density-maximised cell designs in substantially shorter time frames and at lower cost than is possible by empirical design. Moreover, the designs are inherently protected against lithium plating and the inadequacy of cell design solely for discharge is highlighted. The framework’s ability to support common battery pack module design is demonstrated, facilitating faster, cheaper electric vehicle development. Solid-phase lithium diffusivity is shown to be a limiting property not only for fast charging, but ultimately for vehicle range.
An in situ lithium plating detection and quantification technique is augmented. The technique is a hybrid one relying on capacity fade and discharge voltage plateaus. The application scope of the technique is shown to be wider than previously believed, and its successful use is demonstrated for fast charging at temperatures encountered by electric vehicles. Invasive experimental methods are used to debunk a misconception that an absence of a voltage plateau indicates an absence of plating. Furthermore, the technique’s detection reliability is improved with differential voltage analysis and both its sensitivity and accuracy for quantification increased. The possibility of false negative diagnostic results with even medium rate discharge is shown. Finally, it is concluded that long voltage plateaus are indicative of relatively safe plating.
With LIONSIMBA as one component, a framework is developed and demonstrated for the optimal balancing of cell energy and power density in electric vehicles. Vehicle simulations are performed for acceleration and fast charging. The framework thus spans from the vehicle level to that of a single electrochemical cell layer. The framework employs layer reconfiguration to produce energy-density-maximised cell designs in substantially shorter time frames and at lower cost than is possible by empirical design. Moreover, the designs are inherently protected against lithium plating and the inadequacy of cell design solely for discharge is highlighted. The framework’s ability to support common battery pack module design is demonstrated, facilitating faster, cheaper electric vehicle development. Solid-phase lithium diffusivity is shown to be a limiting property not only for fast charging, but ultimately for vehicle range.
An in situ lithium plating detection and quantification technique is augmented. The technique is a hybrid one relying on capacity fade and discharge voltage plateaus. The application scope of the technique is shown to be wider than previously believed, and its successful use is demonstrated for fast charging at temperatures encountered by electric vehicles. Invasive experimental methods are used to debunk a misconception that an absence of a voltage plateau indicates an absence of plating. Furthermore, the technique’s detection reliability is improved with differential voltage analysis and both its sensitivity and accuracy for quantification increased. The possibility of false negative diagnostic results with even medium rate discharge is shown. Finally, it is concluded that long voltage plateaus are indicative of relatively safe plating.
Version
Open Access
Date Issued
2019-05
Date Awarded
2019-11
Copyright Statement
Creative Commons Attribution-ShareAlike 4.0 International License
Advisor
Offer, Gregory
Marinescu, Monica
Sponsor
Engineering and Physical Sciences Research Council
Jaguar Land Rover
Grant Number
14220106
Publisher Department
Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)