A multi-objective optimization approach for a fault geothermal system based on response surface method

Abstract

A fault geothermal system is in operation with a multi-objective optimization for maximum heat extraction, minimum reservoir impedance and minimum construction cost (these are three objectives in this study). Its heat extraction efficiency heavily depends on the interaction of fluid flows in wells and faults. So far, this multi-objective optimization problem is usually solved through single-objective analysis. This study proposes a multi-objective optimization approach for a fault geothermal system based on the geological conditions in Western Anatolia, Turkey and response surface method. First, a numerical simulation model for heat extraction is established by using high permeability faults as main flow paths. Second, a surrogate model is proposed for the prompt response of the three-objective optimization to largely reduce the huge computation load in direct optimization. Third, a reasonable group of simulation runs are designed with an I-optimal method and the surrogate model is determined through the least square fitting on the simulation run results. Finally, the Pareto front of the three objectives is obtained through a multi-objective evolutionary algorithm and an optimal operation scheme is obtained. These results show that the experiment design and the response surface method can practically replace the complex numerical simulations in a surrogate model, thus largely reducing the computation load in the multi-objective optimization for a fault geothermal system.