Battery Energy Storage Systems: Ageing-Aware Trading and System Diagnostics

Li-ion battery energy storage systems (BESSs) are a mature technology for grid-scale applications, characterised by scalability, modularity, high energy density, high round-trip efficiency, fast response, and continuously falling costs. BESSs are increasingly leveraged by developers to enable renewable integration, enhance grid stability, and serve as a profitable source of income, given their potential to participate in various electricity markets. Despite the technological readiness level, interest from investors and developers has only emerged in recent years, resulting in limited practical knowledge of the technology that forces BESS developers to rely on technical support from suppliers or consultancies. Consequently, enhancing their understanding of Li-ion BESS behaviour and performance is pivotal to boost investor confidence for future project development.
This industrial Ph.D. thesis aims to bridge the gap between academic research and industrial practice by enhancing the knowledge of BESS developers during both the planning and operational phases of a BESS project. The BESS topic is explored considering the needs of developers and asset owners, thus without the need to delve deeply into the Li-ion chemistry. Building upon the findings of four research articles, this Ph.D. thesis addresses the following two research questions: (i) How can BESS developers consolidate the business case for new projects by accounting for degradation aspects?, and (ii) What are the capabilities and limitations of BESS developers for system diagnostics of grid-connected assets?
The first research topic of the thesis examines the strategic operation of a 300 MW/600 MWh BESS. It begins by investigating the optimal operation of the BESS when co-located with a 1.3 GW wind power plant (WPP). It compares the economic performance of trading the WPP and the BESS individually in the day-ahead (DA) market, or using the BESS to minimise the resulting wind imbalance volumes, subject to the imbalance price. The findings suggest that the profit from trading the two assets independently in the DA market is higher than that collected by using the BESS to minimise imbalance volumes of the WPP. Consequently, by focusing on the stand-alone operation of the BESS, the investigation contributes to science by providing a techno-economic framework to link short-term market profits with long-term BESS degradation. By simulating ageing-aware BESS operations throughout the entire lifetime until its end-of-life (EOL), the analysis identifies the most profitable operational strategy, thereby supporting investment and operational decision-making for new BESS investors. In this regard, even though the variable cost per state-of-health (SOH) fade is the most profitable strategy in the majority of cases, it also depends on the chosen inputs of the analysis. Indeed, results differ when the BESS size, EOL criteria, electricity prices, or ageing model are changed. Further assessments are needed to identify any clear trends and relationships between electricity prices and trading strategies. However, the author suggests employing the variable cost per SOH fade as a standard approach for future business cases, with the option of freely trading the BESS only function of market prices if the predicted electricity prices are high and volatile.
The second research topic of the thesis explores the opportunities for BESS operators regarding system diagnostics. The analysis starts by focusing on the electrical modelling of a grid-connected 300 kW/652 kWh BESS without dismantling, using only the data available from the energy management system (EMS) and the battery management system (BMS), and minimising the cost related to laboratory equipment and qualified personnel. The analysis provides a satisfactory estimate of the dischargeable energy capacity of the BESS, the internal impedance and open-circuit voltage of the installed cells, as well as the efficiency of the power converters. Afterwards, a grid-connected BESS operating in a micro- grid is considered. First, the online SOH estimation based on Coulomb counting and partial discharges is explored, followed by an assessment of the influence of operational temperature on the resulting SOH. Finally, a remaining useful life-time (RUL) analysis is applied to predict when the BESS reaches EOL. The results show that the operational temperature can affect the online SOH estimation based on operational data. Consequently, one possibility for reducing the impact of temperature is to improve state-of-charge (SOC) estimation using an extended Kalman filter. However, this would require support from the BESS supplier. Finally, the RUL prediction enables the assessment of the asset’s lifetime availability, which can be used to predict future maintenance and reassess BESS business cases throughout the asset’s operation.
Overall, this Ph.D. project provides insights and robust tools for BESS developers to enhance trading strategies for improving business cases of future projects, as well as to consolidate system diagnostics understanding, for both modelling and estimating SOH of the asset once grid-connected and in operation.