Design-Variable Optimization of Hydropower Tunnels and Surge Tanks Using a Genetic Algorithm

Abstract

Optimal design of tunnels and surge tanks is a major concern in the implementation and operation of large hydropower plants, particularly those in which tunnels are long and water hammer is likely. In this paper, flow simulation and an optimization process have been linked in an iterative manner to develop a general model for the optimal design of hydropower tunnels and surge tanks for the emergency condition of operation (maximum surge pressure). The governing equations for the analysis of transient flow in conduits are solved using a finite difference scheme. Furthermore, a genetic algorithm (GA) optimization technique is utilized to select the optimal diameter for the headrace tunnel, penstocks, and surge tanks. The benefit-cost ratio is considered as the objective function in the optimization analysis. The proposed method was used to design diameters for Marun hydropower tunnels (an in-operation large hydropower dam located in southwest Iran). The results were compared with results from a nonlinear optimization technique. They revealed considerable savings in construction cost, attributable to the application of GA. The calculated pressures across the tunnels with and without surge tank agree with valve closure theory and previous studies. The model is able to optimize diameters for a branching system of tunnels and penstocks with surge tanks.