An optimal sizing framework for autonomous photovoltaic/hydrokinetic/hydrogen energy system considering cost, reliability and forced outage rate using horse herd optimization

The components outage of an energy system weakens its operation probability, which can affect the sizing of that system. An optimal sizing framework is presented for an autonomous hybrid photovoltaic/hydrokinetic/fuel cell (PV/HKT/FC) system with hydrogen storage to supply an annual load demand with forced outage rate (FOR) of the clean production resources based on real environmental information such as irradiance, temperature, and water flow. The sizing problem is implemented with the objective of cost of energy (COE) minimization and also satisfying probability of load supply (PLS) as a reliability constraint. The FOR effect of the photovoltaic and hydrokinetic resources is evaluated on the hybrid system sizing, energy cost, reliability, and also storage contribution of the system. Meta-heuristic horse herd optimization (HHO) algorithm with perfect capability on exploration and exploitation phases is used to solve the sizing problem. The results proved that the PV/HKT/FC configuration is the optimal option to supply the demand of an autonomous residential complex with the minimum COE and maximum PLS compared with the other system configurations. The results demonstrated the overlap of hydrogen storage with clean production resources to achieve an economic-reliable power generation system. The findings indicated that the COE is increased and the PLS is decreased due to the FOR increasing because of reducing the generation resources operational probability. The results demonstrated that the hydrogen storage level is increased with FOR increasing to maintain the system reliability level. Also, the sizing results indicated that the FOR of the hydrokinetic is more effective than the photovoltaic resources in increasing the system cost and undermining the load reliability. In sizing of the hybrid PV/HKT/FC system, the COE is obtained 1.57 $/kWh without considering the FOR and is achieved 1.66 and 1.63 $/kWh considering the FOR (8%) for the hydrokinetic and photovoltaic resources, respectively. Moreover, the results cleared that the HHO is superior in comparison with particle swarm optimization (PSO), genetic algorithm (GA), and grey wolf optimizer (GWO) in the PV/HKT/FC system sizing with the lowest COE and higher reliability