Tightening regulations on concentrate disposal from reverse osmosis (RO) and nanofiltration (NF) plants are pushing the membrane industry to find more ways to improve operating efficiencies. The large volume of concentrate brine generated, typically 15% to 50% of the total, can inhibit future growth of the industry unless new methods are developed to optimize recovery rates and reduce waste.
Ion Exchange & Membranes Unite
In the past, ion exchange and membranes were regarded as competitive technologies rather than providing complementary functions. The new reality is that synergistic benefits of boosting RO recoveries and minimizing waste are indeed possible, as evidenced by the recent introduction of technology referred to as cyclic ion-exchange softening. This green technology uses shallow shell cation resins to soften the feedwater to the RO and then uses the very dilute reject brine generated by the RO to regenerate the resin.
The process can effectively regenerate the resin utilizing brine concentrations as low as 0.5%—20 times lower than the typical 10% concentration used by conventional softeners. No supplemental commercial salt is needed except in cases of extreme variability of feedwater quality. This synergy between ion exchange and membranes opens up new possibilities for reducing membrane treatment costs while minimizing impact on the environment.
Of great interest is the capability of the process to efficiently reduce hardness and barium leakages to sub-ppm and -ppb levels, respectively. Potential membrane scaling from sparingly soluble salts (e.g., barium sulfate, calcium sulfate and calcium carbonate) is better controlled. With improved control over scaling potential, RO recovery rate can be increased, provided no limitations are imposed by other contaminants in the water (e.g., silica, organic matter and colloids). For cases in which the latter contaminants are not limiting or are adequately controlled by other pretreatment methods, it is quite possible to design the RO plant for recovery rates of 90% to 95%.
Because brine concentrations as low as 0.5% (5,000 mg/L) can be used to regenerate the resin in this new process, both brackish and semi-brackish waters can be softened. Such waters will typically have total dissolved solids (TDS) ranging from 500 mg/L (0.05%) and upward, and at 90% recovery the reject brine concentration would be adequate to regenerate the resin effectively.
The new cyclic ion-exchange technology promises to eliminate the handling and feed of hazardous acids, including sulfuric and hydrochloric acid, which typically are used to reduce feedwater pH and control potential calcium carbonate scaling of the RO membranes. A drawback of feeding acid that is not experienced with ion-exchange softening is the formation of carbon dioxide (CO2) gas from neutralization of bicarbonate present in the water by the acid. The resultant CO2 passes through the RO membrane, requiring additional capital and operating cost for a degasser tower to liberate the CO2 gas as well as the feed of caustic soda for post-pH adjustment before the water is distributed.
Anti-scalants, more commonly used than acid, provide good control over a wider variety of potential scaling and fouling compounds (e.g., calcium sulfate, barium sulfate, calcium carbonate, iron and silica). When cyclic ion-exchange softening is combined with anti-scalant dosing, a unique synergy takes place, allowing for higher RO recovery rates than achievable using either technology alone. The rewards are reduced concentrate volume and disposal cost as well as reduced consumption of scarce water supplies.
While the benefits of this green technology may be obvious, it is important to assess whether the technology is competitive with the established alternatives of acid or anti-scalant dosing. From a cost standpoint, cyclic ion-exchange softening eliminates three of the major cost components that have inhibited wider use of conventional ion-exchange softening: commercial salt, the associated labor for storage and handling and the increasing cost and complexity for disposal of spent brine. Water consumed in regeneration is also a fraction of that used for conventional water softeners because no salt dilution water is needed and rinse volumes are much lower due to the lower brine concentration. Costs for pumping of the feedwater and for disposal of wastewater are lower, too, because higher recovery rates are achieved.
For comparing scale control pretreatment options, a brackish water typical of that found in the southwestern U.S. was used in a desktop evaluation to compare three options:
1. Anti-scalant dosing only;
2. Anti-scalant plus acid dosing; and
3. Anti-scalant plus cyclic ion exchange.
A water analysis showed a TDS of 1,254 mg/L, a pH of 7.6 and barium and hardness levels that can significantly limit recovery rates. A permeate flowrate of 100 cu meters/hour was chosen. The comparative operating and amortized capital costs to produce 1 cu meter of permeate, or 264 gal, were then determined.
Software from an independent anti-scalant vendor was used to determine the appropriate anti-scalant dosage. For option No. 1, using only anti-scalant dosing, RO recovery was limited by potential of calcium carbonate scale formation to 84%. For option No. 2, using anti-scalant and acid dosing, recovery was limited to 86% by potential for barium sulfate scale formation. Option No. 3, using cyclic ion-exchange softening plus anti-scalant dosing at a reduced rate, allowed maximum permeate recovery of 95%—limited not by scaling potential but by silica fouling potential. Softening using the new technology predicted reduction of barium and calcium to less than 0.02 mg/L and 2 mg/L, respectively.
In this example, the $46,000 cost of the resin was spread over five years; the capital cost for the ion-exchange vessels was assumed to be $167,000 and amortized over 10 years using straight-line depreciation. The cost of water purchases and cost for disposal of reject water were combined and assumed to be low at 50 cents per cubic meter, or about $2 per 1,000 gal. Acid cost was assumed to be 33 cents per kilogram, or 15 cents per pound, while anti-scalant cost was assumed to be $11 per kilogram, or $5 per pound.
The comparison shows that option No. 3 with cyclic ion-exchange softening plus reduced anti-scalant dosing was at least 40% lower in overall cost per cubic meter of permeate produced. Savings, when producing 100 cu meters/hour of permeate and operating continuously, amounted to $43,000 annually.
Additional savings include the smaller size and lower cost for the feedwater train and the lower pumping cost for the feedwater. Reject water cost was the largest cost component for all three options. The higher the cost of water, the greater the savings realized with cyclic ion exchange. Because the cost of water and the cost of disposal varies by region, the technology will be more relevant to geographies where water supplies are scarce or where disposal costs are high.
While the long-term cost for cyclic ion-exchange softening may be lower than that of competitive alternatives, the extra capital outlay and space requirements needed for implementation may be considered drawbacks, whether for new membrane projects or retrofit of existing plants. But these factors should not be considered in isolation, as water supplies are becoming scarcer and concentrate disposal regulations are increasing in scope and complexity.
Implementation of environmentally friendly technology, such as cyclic ion-exchange softening, should be an ongoing part of the strategy to reduce overall cost for desalinated water while minimizing impact on the environment.