(748d) Reducing the Cost of Inland Desalination: Pushing the Conventional Design Limits in EDR and EDR / RO / NF Hybridization Efforts

The shortage of drinking water is an emerging crisis in many parts of the world. After surface water resources, groundwater supplies are a very important source of drinking water in many parts of the world as well as in the Southwest region of the United States.  Reverse Osmosis (RO), Nanofiltration (NF), and Electrodialysis Reversal (EDR) are the dominant technologies for inland desalination of brackish groundwater, and each technology has its own unique characteristics, strengths, and weaknesses.  Electrodialysis Reversal is usually capable of achieving higher recovery under fouling (COD) and scaling (hardness, CaSO4) conditions, and is resistant to silica scaling and chlorine residual.  In EDR the salt removal level is flexible and depends on design and operating conditions.  In RO and NF, CapEx is lower and does not strongly depend on feed chemistry and/or operating conditions.  Energy consumption is relatively insensitive to feed water chemistry; very high salt removal is common in RO systems while NF strongly rejects divalent species. Hybrid RO/EDR or NF/EDR systems potentially provide benefits beyond what the individual technologies can ensure.  The most important benefit would be higher combined recovery rate which contributes to lower concentrate disposal cost.  Lower energy cost per unit of desalinated water is also another benefit. 

Extensive studies were conducted using full sized pilot-scale EDR equipment by Institute for Energy and the environment (IEE), New Mexico State University (NMSU).  The EDR pilot was designed and built by GE Water & Process Technologies (WPT), where the electrodes, spacers, and membranes were sized identical to commercial systems with capacity of 10 gpm, while the number of cell pairs was typically 40 per stage, compared to 600 at most commercial sites.  The experiments operated with single/double electrical/hydraulic stage(s) using CR67 cation-exchange membranes, AR204 / AR908 anion-exchange membranes, and MkIV-2 spacers with an effective membrane surface area of 3,000 cm².  A 2-stage RO/NF pilot system was also designed and built by GEWPT with capacity of 18 gpm using 4” elements.  Different types of RO and NF membranes were characterized.  The experiments were performed at the Brackish Groundwater National Desalination Research Facility (BGNDRF) in Alamogordo, New Mexico, using the various sources of brackish water with conductivity of 1700-6600 µS/cm available on site.  Various hybridization scenarios were evaluated using EDR and RO simulation tools, and the most promising ones were demonstrated using the pilot system.  In addition, the effects of operating EDR above its conventional design limits with respect to salt removal per stage and allowed supersaturation levels were explored.  Water chemistry, temperature, recovery, velocity, volts/amps, pressure, saturation level of different salts, and recovery rates were monitored as experimental variables.  Several water samples were collected for complete water analysis.

Data analysis demonstrated that hybrid systems in certain scenarios can reduce the cost of treated water by reaching higher recovery rates.  In the case of inland brackish water desalination, higher combined recovery rates result in minimizing the concentrate disposal issue.  Operation above the conventional design limits of EDR was also demonstrated on a practical level, contributing to both lower capital cost and higher recovery.