Improving recovery in reverse osmosis desalination of inland brackish groundwaters via electrodialysis

As freshwater resources are limited and stressed, and as the cost of conventional drinking water treatment continues to increase, interest in the development of non-traditional water resources such as desalination and water reuse increases. Reverse osmosis (RO) is the predominant technology employed in inland brackish groundwater desalination in the United States, but the potential for membrane fouling and scaling generally limits the system recovery. The general hypothesis of this research is that electrodialysis (ED) technology can be employed to minimize the volume of concentrate waste from RO treatment of brackish water (BW) and thereby improve the environmental and economic feasibility of inland brackish water desalination. The objective of this research was to investigate the performance sensitivity and limitations of ED for treating BWRO concentrate waste through careful experimental and mathematical analysis of selected electrical, hydraulic, and chemical ED variables.
Experimental evaluation was performed using a laboratory-scale batch-recycle ED system in which the effects of electrical, hydraulic, and chemical variations were observed. The ED stack voltage showed the greatest control over the rate of ionic separation, and the specific energy invested in the separation was approximately proportional to the applied voltage and equivalent concentration separated. An increase in the superficial velocity showed marginal improvements in the rate of separation by decreasing the thickness of the membrane diffusion boundary layers. A small decrease in the nominal recovery was observed because of water transport by osmosis and electroosmosis. Successive concentration of the concentrate by multiple ED stages demonstrated that the recovery of BWRO concentrate could significantly improve the overall recovery of inland BWRO systems. A mathematical model for the steady-state performance of an ED stack was developed to simulate the treatment of BWRO concentrates by accounting for variation of supersaturated multicomponent solution properties. A time-dependent model was developed that incorporated the steady-state ED model to simulate the batch-recycle experimentation. Comparison of the electrical losses revealed that the electrical resistance of the ion exchange membranes becomes more significant with increasing solution salinity. Also, a simple economic model demonstrated that ED could feasibly be employed, especially for zero-liquid discharge.