The current approach to achieve compliance with federal and state drinking water regulations for perchlorate is treatment by either ion exchange or biodegradation, with ion exchange favored for drinking water applications. An alternative regenerable ion exchange process that uses weak base anion (WBA) resin was developed and patented by Applied Research Associates, Inc. and the Purolite Company under ESTCP project ER-200312 and demonstrated on a pilot scale for the treatment of groundwater at Redstone Arsenal, Alabama.

The objective of this project was to demonstrate at the pilot scale and full scale (~1,000 gpm) the use of the WBA resin, ion exchange system for treatment of drinking water. The pilot scale system was demonstrated at Fontana, California and was completed in 2008. The goals of the pilot scale demonstration were to achieve complete perchlorate removal, show efficient and complete WBA resin regeneration, demonstrate a “zero-discharge” perchlorate scavenger process, and produce treated water that met all drinking water quality guidelines.

The full scale system was demonstrated at Rialto, California, wellhead #3. The goal of this phase was to demonstrate and validate the technology at a full scale application in order to obtain permitting and certification from the California Department of Public Health as an approved perchlorate treatment technology. After the WBA resin process is permitted in the state of California, this technology can then be deployed at other contaminated sites.

The ion exchange process takes advantage of the pH dependent nature of WBA resins. At low pH, functional groups on these resins are ionized (R-NH₃⁺) and capable of performing anion exchange. However, at high pH, the resin functional groups lose a proton and are uncharged (R-NH₂) (i.e., no longer attracting the counter anion and enabling efficient and complete regeneration). The pH dependent nature of WBA resins enables efficient regeneration, minimizing the amount of regeneration chemicals and resulting in an economical process. The WBA ion exchange process has two primary modes: operation and regeneration. During operation, perchlorate is removed from the contaminated water. Once the resin has reached its exchange capacity for perchlorate treatment and is considered "spent," the resin must be regenerated before it can be returned to the operational mode.

Pilot Scale System

Six test periods were conducted during this demonstration. The minimum treatment rate was 24 bed volumes (BV) per hour or 3 gpm/ft³ (a surface loading rate of 9.7 gpm/ft²). Four test periods were breakthrough tests (1, 2, 5, and 6). During regeneration of the spent column, the lag column remained online and treated water in a single column. The remaining two test periods (3 and 4) were short-cycle tests. In short-cycle tests, columns were regenerated after approximately one week on-line and before breakthrough. These short-cycle tests were conducted to maximize the number of regenerations per column and minimize the duration of the demonstration. The short-cycle tests also were used to evaluate perchlorate removal efficiency at a higher specific flow rate of 4 gpm/ft³ (a surface loading rate of 12.9 gpm/ft²). Regeneration of spent resin and treatment of the spent regenerating solution using the zero-discharge scavenger process were conducted on-site.

The treatment capacity determined from this demonstration was 9,700 BV. The treated water was below the method report limit for perchlorate (<0.10 ppb) using ion chromatography-tandem mass spectrometry. Nitrosamines were analyzed using EPA Method 521. NDMA was 2.6 ppt with a detection limit of 2 ppt. All other nitrosamines analyzed (including NDEA, NDBA, NDPA, NMEA, NMOR, NPIP, and NPYR) were below the detection limit. A “dial in” capability for controlling residual alkalinity of the treated water in the post treatment process was demonstrated by varying the pH and using a combination of air/membrane stripping and calcite contacting. Treated water had a Langelier Saturation Index (LSI) near zero, which indicated that it had neither corrosive nor scaling tendencies. Five resin regenerations were accomplished using 3 bed volumes of regenerant solution, or approximately 0.03% of the treated water. The spent regenerating solution was successfully treated using the zero-discharge scavenger resin approach to remove perchlorate to below method reports limits. 

Full Scale System

During the full scale demonstration, a total of 14,950 BV (39.15MG) of groundwater were treated over four test periods. The perchlorate concentration of all treated water samples was below the detection limit for reporting of 4.0 ppb. During start up, NDEA and NPIP were detected at <5 BV of water treated, but did not appear after this point. All testing was performed at flow rate of 800 gpm (2.29 gpm/ft²), which was the highest possible flow rate due to equipment and pressure limitations. The first and second test periods were designed to be short cycle tests (1,339 BV and 2,261 BV) where the lead vessel was regenerated after only seven days online and well before perchlorate breakthrough. These tests were designed to improve resin performance by executing more regenerations per vessel to condition the virgin resin. The third test period was designed to operate the system to approximately 50% of resin capacity (4081 BV), while the fourth test period was designed to operate the system to perchlorate breakthrough. Test period four treated 7,269 BV, but perchlorate breakthrough was not achieved due to operational delays and budgetary constraints. Based on previous ESTCP field demonstrations and models using site groundwater characteristics, the lead vessel will treat ≥9,000 BV of water before significant perchlorate breakthrough is observed.

Resin was regenerated at the end of the first three test periods. No detectable perchlorate bleed was observed when the regenerated vessel was placed back online as the lag vessel. The spent regenerant volume was limited to 0.07% of the total water treated during testing, which resulted in concentrating the perchlorate to over 35,000 ppb. A strong base anion (SBA) scavenger process effectively lowered perchlorate in the spent regenerant to non-detectable levels (<2.5 ppb).

Implementation of this technology is straightforward. Commercial, large-scale, ion exchange equipment for WBA resin technology is commonplace. However, the application of this technology to other sites will require additional engineering to meet site-specific requirements based on groundwater characteristics and onsite needs and/or restrictions. Bulk chemicals (i.e., acid, caustic, and soda ash) will be required onsite for application. Parameters that directly affect implementation of the WBA ion exchange technology are groundwater alkalinity, perchlorate groundwater concentration, and treated water alkalinity. The amount of acid required to achieve operating pH is directly proportional to feed water alkalinity. Perchlorate concentration directly affects the amount of scavenger resin required, which can also increase cost. The amount of acid used in pre-treatment and the desired alkalinity of the treated water affects soda ash requirements for neutralization, which, in turn, affects neutralization cost. The cost of each of these drivers is affected by fluctuating market prices.

As perchlorate concentration in groundwater increases, regenerable resins offer significant cost savings and longer service life over single-use resins. The regeneration approach using WBA resin is up to 50 times more efficient than regeneration of strong base anion resins, which typically require a large excess of salt brine. In addition, for many applications, the treated spent regenerating solutions generated contain no perchlorate. (Pilot-Scale Completed – 2008; Full-Scale Completed – 2012)