Techno-economic optimization and modelling of grid-connected photovoltaic and battery energy storage system

• Main Author: Gopinath Subramani
• Format: text::Thesis
• Language: English
• Published: 2023

Currently, commercial and industrial operations face increasingly high costs for electricity billing especially peak or maximum demand charges. These costs affect their business and for some businesses, it can severely affect their ability to operate profitably. Peak or maximum demand (MD) shaving is one of the energy storage applications that has a huge potential in future smart grid. The goal of MD shaving is to reduce the maximum demand during peak hours over an entire billing period. For instance, based on the Malaysian electricity tariff, the maximum demand charges are high compared to net consumption charges on commercial and industrial customers. This indicates the need to reduce the maximum demand, which has caught the attention of many recent studies. In the current energy trends, solar PV and battery storage technologies are widely used by large energy users to reduce energy consumption, decrease peak demand, and improve the reliability of electricity supply. However, higher costs of batteries and incorrect sizing of solar PV and battery systems may lead to higher costs for installation, operation and maintenance which eventually delays the period of return on investment. In this thesis, a Maximum Demand Reduction (MDRed) model is designed and created as an effective algorithm to size the solar PV and battery. The model focuses on maximizing the maximum demand and net consumption savings based on the electricity tariff scheme with an ideal MD limit. The maximum demand limit is the criteria for a battery management system to maintain a specific limit with or without support from the solar PV generation. Maximum demand limit will be the key to determine the capacity and overall integration of solar PV-battery system. In this thesis, the MDRed algorithm will be used for the optimization of the solar PV-BESS. Optimization was performed via MATLAB using particle swarm optimization (PSO) and Genetic Algorithms (GA) techniques. Three (3) types of scenarios are selected to verify the MDRed model and to achieve the optimal techno-economic sizing of the overall system. The three (3) scenarios are solar PV-battery, solar PV only and battery only. Optimization results from GA and PSO on MDRed modeling shows that the solar PV-BESS will be the best option for higher MD reduction. Based on the case study, optimization of solar PV-BESS sizing via MDRed modelling revealed that the energy savings for net consumption, MD up to 30% and 16% with less than 14 years of Return on Investment (ROI) in regard to technical and economic aspects. Besides that, the challenges in MD shaving are mainly the integration of solar PV system and utility grid with battery ON/OFF operation to respond quickly to the MD limit activation command. In this thesis, MATLAB Simulink modeling toolbox is used for simulation of the solar PV-BESS design and verification of the MDRed modeling via variations in MD limit and solar PV generation load. Apart from that, the maximum demand supervisory control algorithm is developed via MATLAB Stateflow® for the battery management system (BMS) to monitor and regulate the load to maintain the net load consumption below MD limit with or without solar PV generation. In summary, the Maximum Demand Reduction (MDRed) model has been developed using GA and PSO optimization tools to attain an idealsizing of the solar PV-BESS, thus reducing the costs of the overall system with maximum energy savings for any commercial and industrial customers. The simulation with the master MD controller shows the effectiveness of the proposed controller on the BMS to cater for MD reductions during solar irradiance and load pattern variations.