Download files
Abstract
According to the recent developments, the application of Renewable Energy Sources (RESs) such as wind, solar, biomass and hydro has become popular especially in the distribution systems. Improved reliability, higher flexibility, less cost and lower emission are some of the main advantages of utilising RESs. Renewable and distributed energy sources such as solar photovoltaic, fuel cells, wind turbines, micro turbines and energy storage devices are expected to play an important role in the electricity supply industry and low carbon economy of near future. Nevertheless, the high utilisation of distributed generating resources and high penetration level of Electric Vehicles (EVs) in the new power systems has caused new challenges from both operation and management perspectives. A part of these challenges is addressed in the form of microgrids (MGs). This thesis aims to investigate the optimal operation and management of renewable MGs in the presence of EVs. In this regard, different RESs, EVs and storage devices are modelled and analysed as integrated parts of a MG. Here, the concept of vehicle2grid (V2G) technology is used to change the EVs role from only mobile loads into mobile storages which can benefit the main grid. To make the analysis more practical, the battery degradation costs due to V2G are included as additional term in the objective function. To improve the MG performance through a more effective scheduling of units, optimal reconfiguration strategy is employed to alter the feeders’ paths and thus change the power flow direction in the system. The simulation results show that reconfigurable MGs can support the electric consumers with lower power losses and costs while providing better power dispatch of units. In the case of congestion, optimal reconfiguration can also help to release the feeder capacity by transferring heavy-loads to other feeders with light-load. The uncertainties associated with MG components such as load demand prediction error, wind turbine output power variations, solar photovoltaic output power forecast errors, EVs charging and discharging characteristics, and market price forecast error are considered using a stochastic framework based on Unscented Transform (UT) method. The simulation results show that the proposed stochastic framework can provide more realistic results while showing better scheduling of units in the MG with lower costs. Finally, the proposed optimisation method could overcome the traditional methods such as genetic algorithm and particle swarm optimisation in the literature. The feasibility and satisfying performance of the proposed methods are shown on several tests systems.