Parametric optimization of steam cycle in PWR nuclear power plant using improved genetic-simplex algorithm

Abstract A parametric optimization framework for the steam cycle of a typical pressurized water reactor (PWR) nuclear power plants (NPP) (i.e. Daya Bay nuclear plant) is proposed. The approach determines the optimal operation parameters of the steam cycle to maximize the power output with primary loop remaining unchanged. Thermodynamic model of the steam cycle system has been established to determine the power output under different operating conditions, and the availability and accuracy of the model are investigated. In addition, an improved genetic-simplex algorithm (IGSA) is put forward by using modified search strategies and integrating genetic algorithm (GA) with simplex algorithm (SA). Performance comparison between the improved algorithm and the original ones is performed by solving the optimization test problems. The improved algorithm enables the handling of nonlinear constrained optimization problem because of dramatically enhanced search capability. Based on the model, the parametric optimization for the steam cycle of Daya Bay PWR nuclear plant is carried out using the IGSA, GA and SA, respectively. In the optimized schemes it has been found that the IGSA can yield increase in power output as much as 23.8 MW by an average of 236 iterations over current practice, and the corresponding thermal efficiency increases from 33.87% to 34.69%. In comparison, even more iterations have been executed, the maximum power output increments gained by the GA and SA are 21.4 MW and 18.7 MW, respectively. The results show that the new algorithm IGSA is superior to original ones. It also demonstrates that the proposed approach can effectively increase the efficiency of the PWR nuclear power plant without burning any additional fuel or replacing the devices of the system.