Analysis of variable reverse osmosis operation powered by solar energy

The use of solar photovoltaic (PV) to power reverse osmosis (RO) plants and produce water will enhance the sustainability of water supplies in several dry remote coastal areas. Varying the operating power level of the RO plant has been proposed in the literature as a solution to accommodate intermittent PV power sources. Such variable operation is intended to match the RO load to the available PV power. Nevertheless, such operation has not been used outside research laboratories and small pilot plants. In this work, we used different case studies to evaluate the benefit of using variable operation and its effects on system design, system operation and levelized cost of water (LCOW). A simulation model for the optimal operation of the system is developed using three-dimensional dynamic programming (DP) to determine the power levels of the battery, diesel generator, and RO plant while optimal sizing of these plants and associated water tanks and PV generators was solved using an ordinal optimization (OO) approach. The use of OO permitted the examination of a large design search space quickly but exhaustively using a simple model. We then ranked the different designs in increasing cost order and assessed a reduced number of these using an accurate model to simulate the system on an hourly basis for all the days of a year. This approach relies on the fundamental tenet of OO: "order is robust to the noise introduced by the simple model". Different power modulation strategies are investigated, and their implications on the hy-draulic operating parameters are presented. In this respect, we investigate the operation of the RO system at varying power levels and different sizes of backup systems (battery and diesel generator). This ability to vary the RO operating level helped in a better matching of the system load to the available, yet variable, PV power, even when the backup and storage systems were at a minimum. Operating an RO plant with PV and backup systems is found to be far more cost effective than operation without backup systems, reducing costs by 37-55% for the case studies considered.