Understanding Lithium-Sulphur (Li-S) cell performances under different configurations and temperature environments is crucial for the design of batteries that are compatible with space applications. To do so, scientists and engineers rely on computer models that describe and predict a battery cell’s behaviour. Currently computational models of Li-S cells exist, that characterize the complete cell behaviour on a physical level. These however have two drawbacks: Firstly, they require great computational effort and secondly, they rely on the use of chemical and physical parameters that can be difficult to obtain.
For engineering applications, simple computational models are essential as they provide quick, practical solutions with less computational power and utilize easily obtainable measurements. Although these reduced models give less insight into the chemical, physical and electrochemical processes of Li-S, they are more believable and provide other more application relevant properties such as voltage, state of charge (SoC) and measures of battery health. For the development of an improved Li-S battery in the context of ECLIPSE, two computational Li-S models have been developed by Imperial College London.
The first, an equivalent-circuit network (ECN) model, which was developed in collaboration with Cranfield University, captures battery resistance at various currents and temperatures. ECN models are simplified models that reproduce the transient behaviour of a battery with a circuit of electrical components. Using pulse discharge experiments at different current rates and temperatures for a Li-S cell with known electrical and thermal properties, the model can capture battery voltage with high accuracy, although unable to account for the SoC drift.
To help model the SoC drift, a zero-dimensional (0D) model was developed to compliment the ECN model. The 0D model predicts many of the behavioural features observed in an Li-S cell during charge and discharge by tracking the amount of different chemical species in the battery, thereby providing an accurate computation of the battery’s SoC.
Coupling both models into a single Li-S model should offer further insights into the operation of Li-S batteries and parametrization and validation of the new model should eventually provide the basis for defining and optimizing system-level control laws for space battery management systems and space power subsystems.
Please see the following research papers for more information: