Multi-objective optimization of hybrid nuclear power plant coupled with multiple effect distillation using gravitational search algorithm based on artificial neural network

Abstract

In this study, multi-objective optimization has been performed for finding the best net electric efficiency, gain output ratio, and fresh water cost in a nuclear co-generation plant. Bushehr nuclear power plant (BNPP) with 1000 MWe nominal power with a MED (multi-effect distillation) desalination system is selected. The BNPP is simulated with Thermoflow software and energy analysis of flow cycle has been carried out. The MED plant with a capacity of 25,000 cubic meters per day is coupled. The desalt water cost was evaluated by DEEP software and the gravitational search algorithm (GSA) has been performed to optimize the co-generation plant. The performance of the gravitational search algorithm has been proven by the Griewank test function compared to the harmony search and particle swarm algorithms. An artificial neural network with 10 thermodynamic inputs related to the power plant, intermediate loop, and desalination system has been used to predict fitness function for the optimization process. The network is created with Levenberg-Marquardt back-propagation algorithm with a 0.27% probable error. Energy analysis showed that the power plant's efficiency before coupling the desalination system is 33.93%. The multi-objective optimization results demonstrated that 0.1123 is the best fitness function found by the algorithm, according to which the minimum cost of desalted water is 1.07 $/m3, gain output ratio is 13, and net electric efficiency is 33.3%.