Operation of Medium-size Reverse Osmosis Plants with Optimal Energy Consumption

This paper deals with the optimal operational strategy of a reverse osmosis (RO) plant for remote sites. The electricity supply to these plants comes usually from renewable energies (wind and solar), when they are not temporarily available, they are complemented by a diesel generator and batteries. The water demand of small settlements in arid regions suffers strong variations along a day. As the higher demand of water usually occurs when solar energy is more available, the operational expenses can be reduced by considering the RO plant as an active load. A good control strategy, will implement a variable operation point, taking into account the predictions of water demand, the expected variation of the available energy sources and the scheduling of cleaning operations in the RO plant, in order to optimize the energy use. In this paper a hybrid predictive control is proposed to implement this task. Simulations of a specific plant show that an adequate operation reduces the diesel energy consumption, while satisfying the variable water demand.