PLANNING, OPERATION AND CONTROL OF AN ISOLATED MICROGRID WITH RENEWABLE ENERGY AND BATTERY STORAGE

Abstract

This thesis focuses on the optimal microgrid (MG) planning, operation, and voltage control of an isolated microgrid under various system conditions and the available constraints. Microgrids (MGs) are the promising archetype for the penetration of renewable energy sources (RESs) into the system for reducing carbon footprints. While dealing with optimal planning and operation of an isolated MG system, the research work mainly focuses on the system reliability considerations that incorporate into the system as a percentage of total load supply requirements or as a desired loss of load into the MG system. The penetration levels of RESs have also been considered while solving the formulated planning and operation problem in this thesis. The proposed work presents a stochastic cost-based optimization model for the MG planning problem, which is solved using a sequential programming approach to obtain the optimal size of distributed energy resources (DERs). Uncertainties in the renewable output and the load demand data are considered using multiple scenarios with random variations, which are based on beta distribution. The overall objective function is formulated as a cost-based multi-variable constrained nonlinear (MVCNL) programming problem through various costs associated with the MG system. These costs are system planning costs, daily operational costs such as unit commitment costs, switching costs associated with the dispatchable units, the cost associated with the losses in the storage system, etc. The formulated problem is simulated under the MATLAB environment. Quantitative results indicate the impact of the operational paradigm along with the planning of an MG in the form of cost analysis with different reliability conditions. An assessment of results in tabular form is presented, which incorporates the optimal sizing of the MG under the different system reliability conditions and constraints. Some additional cases pertain to the renewable penetration, and the load demand variations have also been included in the study. Further, an optimal day-ahead dispatch problem has been formulated with the help of the earlier developed MVCNL problem through an operation cost-based objective function. The day-ahead renewable and load demand data are predicted using the regression learning approach. Solar and wind energy sources are considered as RESs. Historical data pertaining to solar irradiance, wind speed, and the load demand for a specific geographical location is considered for training purposes. Thereafter random scenarios are added to the obtained i predicted data using beta distribution. The formulated problem is solved for the optimal solutions through the sequential quadratic programming (SQP) approach, and a comparative assessment of results with a hybrid particle swarm optimization (hPSO) approach has also been presented. The optimal dispatch of the DERs connected to the MG is obtained under various reliability conditions and the available system constraints. It is found that there is a trade-off between the results obtained from the sequential programming approach and the hPSO approach with respect to system economic and computational aspects. However, since the operation models incorporate a short time horizon with smaller time intervals, the SQP approach seems to be a more appropriate choice. In the prescribed work, a robust voltage control strategy to control the inverter output voltage has also been proposed for the isolated MG with renewable penetration and a storage system. For this purpose, an adaptive PID controller is designed, and its performance is checked under various system conditions. Four cases have been considered to examine the effectiveness of the proposed controller. These cases are related to variations in MG system load conditions as well as the temporary fault case. The obtained results are compared with the other controllers, viz conventional PID and the model reference adaptive controller (MRAC). It is found the proposed controller performs better to control the voltage as compared to the other two controllers. The results are also verified with a Hardware-in-loop (HIL) setup using a real-time digital simulator (RTDS) with certain test conditions.