Optimal design of grid-connected photovoltaic (PV) power plants using optimisation algorithms

Photovoltaic (PV) technology has increasingly become an important source of energy. Considering the recent drop (up to 86%) in PV module prices from 2010 to 2017, many countries have shown investment interest in PV plants to meet their energy demand. In this case, improving the levelised cost of electricity (LCOE) and producing energy for PV power plants is a complicated design tradeoff. This involves several parameters, such as meteorological data variation, nonlinear operation of the PV plant components inverter types, and PV module efficiency. Hence, this thesis presents a detailed methodology for the optimum configuration of large-scale PV power plants ground-mounted connected to the electrical grid. In this proposed procedure, the mathematical modelling includes several important parts of the PV system and many design parameters, including the number of PV modules connected in series and parallel, number of PV module lines in each PV row, PV modules tilt angle, inter-row distance, PV modules orientation, PV module and inverter type, to improve the PV plant performance and to address the accuracy of the results. Besides, the minimum LCOE, maximum annual energy, and minimum cost are the objective functions in this methodology. The design problem is solved using recently proposed metaheuristic optimisation techniques to determine the global optimum. A sensitivity study is performed to investigate the effect of important parameters on PV plant performance. To examine the PV power plant performance, a cost-effective structure is investigated based on comparing the most used PV power plant topologies. Several scenarios are presented to assess the impact of many key factors variation, such as ambient temperature fluctuation, solar irradiance variation, and geographic location on the PV power plant optimum configuration. The proposed PV power plant sizing methodology is implemented in MATLAB software. The results demonstrate that using the proposed optimisation method is effective and suitable to design PV power plants. LCOE using thin-film (CdTe) PV modules has the lowest value compared to crystalline silicon (c-Si) PV modules. Metaheuristic optimisation techniques show effectiveness and efficiency to measure the optimum solution. However, it is found that hybrid optimisation techniques have better performance than single optimisation techniques in determining the optimal design of the PV power plant. In economic terms, LCOE-based design optimisation presents lower installation, maintenance, and operation costs compared to maximum energy and minimum cost objectives. Furthermore, the PV plant using central topology presents the lowest cost of energy. The results also show that the geographic location, a change in meteorological condition levels, and an increase or decrease in the available area require the re-design of the PV plant. A change in inverter size and PV module type has a significant impact on the configuration of the PV plant, leading to an increase in the cost of energy. The predefined objectives and proposed optimisation methods can affect the PV plant design by producing completely different structures.