Reliability-oriented Control of Battery and Inverter in Isolated Microgrids

Abstract

Isolated photovoltaics (PV) – diesel generators (DG) microgrids are implemented to provide dependable energy supply in diverse practical applications, including telecom towers, islands, military bases, and remote industrial locations. Such microgrids usually require a battery, to handle PV power intermittency, and an inverter to supply the AC load. The battery and inverter are pivotal in ensuring secure operation of isolated systems. Both components also constitute a significant portion of the system's overall cost. Thus, improving their lifetime and reliability can result in significant cost savings and increased system dependency. Previous studies have limited considerations for the reliability aspects in isolated microgrids. Additionally, no prior research has specifically aimed at enhancing the lifespan of both the battery and power converters simultaneously. In this work, a supervisory adaptive controller is proposed to improve the reliability and extend lifetime of the battery and inverter in an isolated PV – DG – battery AC microgrid.

To develop the envisioned controller, it is needed to realize the stressors that affect the degradation of the components in consideration. These stressors are acknowledged in the literature using experimental life testing, which correlates the stressors to their degrading effect leading to mathematical lifetime models of the components. By deeply analyzing these lifetime models, the information obtained can be used in a higher level of study, the supervisory control level. In this work, it is established that for the inverter, delivering the same amount of energy at lower power, for extended time, results in reduced degradation on its switches. This principle, along with the known information that the battery degradation being dependent on its state of charge (SOC) cycling, are used to design a supervisory adaptive energy management controller (AEMC) for an isolated PV – DG – battery AC microgrid. The controller implements a power sharing approach based on battery state of charge (SOC), DG operational constraints and the mission profile input data of PV generation (temperature and solar irradiance) and electrical load. It works on operating the inverter and DG in droop sharing mode for longer periods, hence minimizing the burden on the inverter as per the established principle. This is done by sharing the load between the inverter and DG with variable adaptive sharing ratios calculated to reduce inverter loading, while also attempting to extract as much PV generation as possible to save DG fuel cost. AEMC also reduces battery stresses by avoiding its overcharging using a SOC threshold that controls battery operation either to charge or actively participate supplying the load. Also, two methods for detecting sudden PV intermittencies are implemented. It is an important function that helps to reduce either PV curtailment or the stress on the inverter switches.

Mission profile-based reliability is adopted for the system to compare the proposed controller to traditional self-consumption controller in two different geographical sites. Both case studies demonstrate considerable improvement in reliability for the battery and the inverter IGBT power modules.

Date of Award	15 Dec 2023
Original language	American English
Supervisor	Khalifa Alhosani (Supervisor)