In order to implement efficient and cost-effective membrane desalination, one must combat membrane scaling by mineral salts and minimize the volume of concentrate generated. Pilot studies of Colorado River desalting indicate that water recovery (with the use of antiscalants) must be below about 85% to avoid scaling. Field and laboratory studies with of brackish groundwater water in the San Joaquin Valley (>3000 mg/L TDS) suggest that recovery would be limited to about 50-75% to avoid membrane surface scaling. In this range of recovery, concentrate volume is significant and represents a loss of a valuable water resource and major disposal challenge. Clearly, a higher product water recovery is desired to reduce concentrate management issues and make membrane desalination economically feasible for potable water production.

Accelerated chemical precipitative softening of primary RO concentrate, in combination with secondary RO treatment, is a potentially effective method of reducing RO concentrate volume and increasing product water recovery for desalination of brackish surface water and groundwater. Initial studies by the Metropolitan Water District (MWD) of Southern California and researchers at the WaTeR Center have explored various precipitation sequencing strategies that included the addition of NaOH, Ca(OH)₂, Na₂CO₃, as well as CaCO₃ crystal seeds.  Removal of ~80-95% was achieved for divalent cations, such as calcium, barium, and magnesium, from a concentrate stream produced from primary RO desalting of Colorado River (CR) water and San Joaquin Valley brackish groundwater. This removal level was shown to reduce the saturation level (associated with scaling salts) to avoid membrane surface scaling. To date, collaborative studies by WaTeR center researcher and center affiliates have shown that product water recovery of up to 98% can be attained using accelerated precipitation softening (APS) and optmized  secondary RO desalting.

Although seeded crystallization softening of RO concentrate (or as pretreatment for high salinity brackish water feed) has proven feasible, general design guidelines for optimal crystal seed loading rate and size, solution pH, reactor mixing, and residence time are lacking. Efficient clarification of the APS treated stream is imperative prior to secondary (or primary) RO desalting, and enables crystal seed recovery. Therefore, control of particle size (in the APS process) is important to enable efficient filtration. It is important note that precipitation kinetics are lacking for complex multi-ion systems where co-precipitation of mineral salts of strontium and barium as well as silica can take place.  Therefore, in order to adapt APS for large-scale desalting processes, a systematic study has been undertaken by WaTer Center researchers to determine the optimal process operating conditions, configurations, and economics of the APS process for the range of expected brackish water feed compositions.   The potential application of APS for seawater desalination is also under investigation. A design model for APS is currently under development using experimental kinetic data on mineral salt crystallization.  The modeling tool will then serve to optimize process conditions and reactor scale-up for high recovery brackish water and seawater desalination as well as aid in evaluating the overall process economics and integration of APS with other feed pretreatment strategies.

Scanning electron micrograph images (SEC) of calcium carbonate crystal seeds (a) and  (b) precipitate formed upon accelerated precipitation treatment of RO concentrate (85% recovery) produced from primary RO desalting of brackish water. (c) Integrated laboratory RO unit with a temperature controlled crystallizer reactor/feed reservoir, spiral-wound RO module and microfiltration along with a computerized data acquisition system [top right of (c)].

 (Left) Concentration and percent removal of mineral salt scale precursors for AP treatment of primary RO concentrate produced from primary desalination of brackish water. (Right) Calcium ion removal kinetics and pH change during accelerated precipitation with calcium carbonate seeding.