부산대학교 도서관

This paper presents the sizing of run-of-river small hydro-PV hybrid power system using the Non-dominated Sorting Genetic Algorithm (NSGA-II). The two objective functions are the total generated energy and the energy production cost of hybrid system. The total energy has been maximized whereas the energy production cost of hybrid system has been minimized. The nominal turbine flow rate (Q Tr ), the number of hydropower units (n hyd ), and the number of PV modules (n PV ) are considered as decision variables in this problem. The Yeripao site in Benin has been considered as case study. The optimal solutions converge to Pareto front which represent the best trade-offs between total generated energy and energy production cost. The results have shown that energy production cost increases with the total generated energy. Thus, minimizing the energy production cost is contradictory with maximizing the total generated energy. Moreover, the sensitivities analysis of Q Tr and n hyd on the total generated energy and on energy production cost have been conducted in this study. It is relevant to note that the optimal solutions are grouped into four categories according to n hyd = 1, 2, 3 or 4. For each category, the total generated energies, energy production costs and cost per kWh increase with the Q Tr . For n hyd = 1 the lowest values of energy production costs, total generated energies and costs per kWh have been recorded. Moreover, n hyd = 4 match with the solutions that present the highest total generated energies and costs. The lowest overall cost per kWh is € 0.363/kWh. The conducted study can be applied to other sites by using their hydro and solar resources characteristics.