Assessment of genetic algorithm selection, crossover and mutation techniques in power loss optimization for a hydrocarbon facility

In this paper, different selection, crossover including deferential evolution and mutation techniques are considered for optimizing the electrical power loss in real hydrocarbon industrial plant using genetic algorithm (GA). The subject plant electrical system consists of 275 buses, two gas turbine generators, two steam turbine generators, large synchronous motors, and other rotational and static loads. The minimization of power losses objective is used to guide the optimization process. Eight GA selection, crossover and mutation techniques combination cases are simulated for optimizing the system real power loss. The potential of power loss optimization for each case versus the base case will be discussed in the results. The results obtained demonstrate the potential and effectiveness of the proposed techniques combination cases in optimizing the power consumption. Also, in this paper a cost appraisal for the potential daily, monthly and annual cost saving associated with the power loss optimization for each case will be addressed.