Development of Control Methodologies for Energy Storage Systems in Electricity Distribution Networks

Abstract

Prompted by technical issues that have arisen due to the widespread deployment of distributed intermittent renewable generators, rapidly rising peak demand and reductions in battery price, the use of Battery-Based Energy Storage Systems (BESS) in power networks is on the rise. While BESS has the potential to deliver technical benefits, the best possible sizing, location and usage govern the financial viability. Prevailing models of determining BESS size, location and charging patterns have treated them as independent problems and lack explicit dependence on factors such as load growth rate, PV penetration level, network size and structure. Furthermore, the existing literature does not provide any guidelines on possible interaction between network operators and retailers to employ batteries for a distribution network operator’s benefit.To bridge the existing research gaps, a generic approach to select appropriate sizing and siting of BESS for supporting both distribution utilities and customers is developed in this thesis. A method is established to model network upgrade deferral as a function of load growth rate, renewable generation penetration and peak shave fraction. This model is then used for the formulation of an optimisation problem which benefits from multi-period power flow analysis to co-optimise the size, location and dispatch scheduling of BESS for a pre-specified number of units to be deployed in a given distribution network. The proposed approach is implemented on a segment of the Australian medium voltage distribution network under multiple practical and potential future scenarios. Moreover, the developed methodology is utilised to obtain appropriate sizing and siting while controlling a BESS in the rest of the thesis.BESS are widely considered as a potential solution to counter the voltage regulation challenges arising due to solar Photovoltaic (PV) generation in low voltage distribution networks. Although the present approaches of using BESS are promising, the resulting voltage regulation performance and the prolongation of the lifetime of usually costly BESS units are heavily reliant on the underlying control algorithms. The existing BESS control approaches require frequent micro-cycles for voltage regulation and hence, put additional stress on BESS cycle-life. Furthermore, the vast majority of voltage regulation and storage management techniques in the literature lack the essential steps of experimental validation of their proposed approaches under realistic conditions.In this thesis, a new control method is proposed and practically validated to ensure smooth BESS operation amenable for prolonged BESS life without compromising the voltage regulation performance. The approach is based on the finite short-term forecast of PV generation to obtain forecast voltage trajectories. The forecast PV generation in conjunction with calculated feasible BESS charge-discharge trajectories is utilised to regulate voltage response over a finite time horizon that substantially reduces the charge-discharge cycling of BESS.As the uptake of BESS for photovoltaic applications rises, their aggregated use for network voltage regulation is considered as an impending option. Therefore, a generic approach to coordinate distributed BESS of customers with a view to system voltage regulation through the help of a demand response aggregator is developed. The proposed control algorithms are practically validated by using a Hardware-in-the-Loop (HIL) setup comprising of a Real Time Digital Simulator (RTDS) and a dSPACE controller board.