Assessment of Energy Throughput for Improved Availability and Reduced Degradation of Battery Energy Storage Systems for Frequency Support Applications in Great Britain

UoM administered thesis: Phd

Abstract

Investing in low or no-carbon means of energy production can dramatically reduce the greenhouse gas emissions in the energy sectors. Despite their significant benefits, renewables challenge the stability of the power networks due to their intermittent nature. To ensure the security of supply advanced power balancing techniques, such as energy storage, are essential. Battery energy storage systems (BESS) are widely used for grid frequency support. They can assist the grid in balancing any fluctuations and thereby ensure a reliable and secure supply. This project will investigate a commercial nickel manganese cobalt (NMC) chemistry, Siemens SieStorage BESS, from the perspective of a BESS operator, at a system level and develop a state of charge (SoC) control technique to use the system more efficiently in terms of availability while providing Dynamic Moderation (DM) frequency response services. This thesis presents the results of a performance test schedule to evaluate the round-trip efficiency and power losses across the full operating range of the SieStorage system. The schedule evaluates over 237 operating conditions; ten different real power set-points, and 24 SoC ranges across the 5% to 90% SoC operating window. A round-trip efficiency and instantaneous power loss map is then produced. Historical frequency data for the UK grid have been analysed using the dynamic firm frequency response (FFR) and DM frequency response profiles. An algorithm to calculate the output power, energy, SoC, and predict the availability of a storage system for the above frequency response profiles is developed. The instantaneous power losses of the system are included in the algorithm in a form of a lookup table. The algorithm is then validated against the SieStorage while providing DM service on the most energy demanding day. The algorithm accurately estimates the energy usage of the SieStorage with an error of 2.94%. A novel control algorithm is then proposed, where a small percentage of the system†s power rating is dedicated to manage its SoC level. The algorithm offers a 16.67% reduction of the peak SoC managing power and 92.5% slower ramp rate when compared to the dead-band SoC control, and 98.5% slower when compared to the maximum ramp rate of 5% of the contracted power for the Dynamic Containment service. The proposed control achieves a reduction of 76.44% in the time that the system was unavailable to respond when compared to the dead-band SoC control, while avoiding high SoC regimes where the battery degradation is more severe.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award date1 Aug 2022