Saudi Arabia tries to build local desalination water stations to supply water to remote areas. Due to the low cost and energy requirements of reverse osmosis (RO) desalination technology, it has been used to supply fresh water to Arar City in the northeast of Saudi Arabia. In this paper, it is proposed to provide an average of 1000 cubic meters of water per day by using autonomous hybrid renewable energy system (RES). This proposed system contains wind turbines (WTs), photovoltaic (PV), battery, and it is designed to feed the RO system with the energy adequate to produce the required amount of fresh water for the minimum cost and minimum loss of supply probability. The proposed system was designed to generate 2440 kW power to produce this amount of water. Matching study between the site and the best WT among 10 market-available WTs is introduced. Three optimization strategies were used and compared for the design of the proposed system to ensure that no premature convergence can occur. These strategies consisted of two well-known techniques, particle swarm optimization and bat algorithm (BA), and a relatively new technique: social mimic optimization. The simulation results obtained from the proposed system showed the superiority of using a RES for feeding a RO desalination power plant in Arar City, and they also showed that the BA is the fastest and most accurate optimization technique to perform this design problem compared with the other two optimization techniques. This detailed analysis shows that the cost of production of fresh water is $0.745/m3.