A comparative evaluation of Gravitational Search Algorithm (GSA) against Artificial Bee Colony (ABC) for thermodynamic performance a geothermal power plant

Abstract

Optimizing a complex system/problem under real working conditions with optimization methods means ensuring that they operate more efficiently, economical, and eco-friendly. For this purpose, in order to maximize the exergy efficiency of a thermodynamic model of a real operated geothermal power plant (GPP), two optimization methods, namely Gravitational Search Algorithm (GSA) and Artificial Bee Colony (ABC), have been comparatively evaluated in this study. The selected thermodynamic model is a problem that is highly complex, non-linear and unsolvable through mathematical methods. In order to solve the problem, 17 optimization parameters have been selected on the model. In addition, the selected parameters have been divided into 11 groups according to the system equipment specifications to reduce time loss. The results of the study reported that GSA and ABC maximized the exergy efficiency of the real system from 14.52% to 26.31% and 23.92% respectively. The effects of the optimized parameters on the model are observed, and it has been verified by GPP operators, engineers and researchers that no contrariety to logic and engineering discipline existed. Hence, the results of GSA method for the engineering problem addressed in this study are better than those of ABC method and they responded in a much shorter time. The most effective group in both methods is the G3 group related to the turbines. Besides, the most effective optimization parameters on the system performance are the pressure differences in evaporators and mass flow of the geothermal fluid. (C) 2018 Elsevier Ltd. All rights reserved.