ABSTRACT

The two-stage forward osmosis (FO)-reverse osmosis (RO), and reverse osmosis (RO)-brackish water reverse osmosis (BWRO) processes were evaluated for seawater and brackish water desalination, respectively. Three different salts, NaCl, MgCl2, and MgSO4, were used as draw solutions. The performances of the FO, RO, and BWRO regeneration processes were simulated using pre-developed software. The simulation results showed that the water and solute fluxes across the FO membrane decreased as the FO recovery rate increased. For brackish water desalination, the membrane flux decreased as the FO recovery rate increased while the solute flux increased as the FO recovery rate increased. The highest water flux of 20.7 L/m2h was observed with the 1.2 M NaCl draw solution, and the differences between MgCl2 and MgSO4 were insignificant. For seawater desalination, the total power consumption in the RO process (3.66 kWh/m3) was lower than that in the FO-RO process (8.37 kWh/m2). However, as the RO L/m2h recovery rate increased, the difference in the total power consumption between the RO process and the 0.65 M MgCl2 FO-RO process was insignificant. The lowest permeate total dissolved solid (TDS) was achieved with the 0.65 M MgCl2 FO-RO process, which was attributed to the high rejection rate of MgCl2 by the RO regeneration unit. In the case of brackish water desalination, the lowest specific power consumption and permeate TDS was in the case of the conventional BWRO process, followed by the 0.32 M MgSO4, 0.22 M MgCl2, and 0.33 M NaCl FO-BWRO processes. However, the specific power consumption of the RO-BWRO process dropped to less than that in the conventional BWRO process when an NF membrane was used for MgSO4 regeneration and water extraction. The major drawback of the FO-NF process was the high permeate TDS, which as not suitable for drinking purposes, but could be used for other domestic applications.