 |
Treatability Studies for Perchlorate TreatmentLos Alamos, NMSource: Worland, V. Peter, 2001: "Pilot Scale Tests to Remove Perchlorate from the RLWTF Effluent by Ion Exchange", FWO-WFM, January 30, 2001 Project Summary: The following text was excerpted from Worland, V. Peter, 2001: "Pilot Scale Tests to Remove Perchlorate from the RLWTF Effluent by Ion Exchange", FWO-WFM, January 30, 2001: Perchlorate is present in the influent to the Radioactive Liquid Waste Treatment Facility (RLWTF) at Los Alamos National Laboratory (LANL) at concentrations up to several thousand parts per billion (ppb). Perchlorate affects hormone production in the human thyroid and is a suspected, but not proven, carcinogen. The California Department of Health Services has issued a health advisory limit of 18 ppb for perchlorate in drinking water. Pilot tests to remove perchlorate from the RLWTF effluent are being conducted. This report describes the results from ion exchange tests performed to remove perchlorate from the current RLWTF effluent. Testing began on September 6, 2000, and continuted through publication date of January 2001. Perchloric acid is the major source of perchlorate ion in the RLWTF influent. This acid is used in analytical chemistry labs and for metal dissolution. Influent perchlorate concentrations to the RLWTF range from several 100 to several 1,000 ppb. The range of perchlorate concentration in the Raw Weekly Composite (RWC) to the RLWTF was found to range from just over 3 mg/L (3,000 ppb) for the week ending August 20, 2000 to less than 0.1 mg/L (100 ppb) for the week ending November 12, 2000. If the reverse osmosis (RO) treatment unit is not used to produce effluent from the RLWTF, then effluent concentrations of perchlorate are essentially equal to the influent perchlorate concentrations. Even when the RO unit is used to make effluent, the RLWTF effluent normally exceeds 18 ppb health advisory limit for drinking water. The RLWTF effluent is discharged to the environment into Mortandad Canyon via a National Pollutant Discharge Elimination System permitted outfall. Perchlorate concentrations in shallow groundwater monitoring wells (from one to forty feet below the surface) in Mortandad Canyon are found to be in the 80 to 220 ppb range. Perchlorate has been detected 646 feet below the floor of Mortandad Canyon in a perched groundwater zone at a concentration of 12 ppb. The upper surface of the regional groundwater aquifer is at a depth of 900 feet below the floor of Mortandad Canyon. No perchlorate contamination of the drinking water aquifer underlying Mortandad Canyon has been found. However, the potential for groundwater contamination of perchlorate in Mortandad Canyon due to RLWTF effluent does exist. (Perchlorate-contaminated groundwater at LANL is addressed in PERC0052). Based upon suggestions from IX experts from US Filter, a Type I, strong base anion (SBA) exchange resin was selected for removal of perchlorate from the RLWTF effluent. The resin is USF A-464. It is a porous gel anion resin. The polymetric matrix is composed of styrene divinylbenzene activated with quaternary amines for strong base characteristics. Two 0.25 cubic foot resin vessels were obtained of this resin. The resin in one vessel had hydroxide as the exchangeable anion (OH-SBA), the exchangeable anion in the other vessel was chloride (CL-SBA). Each vessel is 15.25 centimeters (cm) in diameter and 38 cm in length. Batch Pilot Tests For the batch pilot tests, flow into each vessel was 2.5 liters/minute (lpm). The hydraulic loading rate was 3.4 gpm/ft2. The actual contact time of the water with the resin was 40 seconds. The RLWTF discharges a 5 lpm stream of effluent to the environment for 50 minutes per day; this stream was split between the two resin vessels, so that each received a 2.5 lpm stream for 50 minutes per day. A bed volume (BV) is the volume of space taken up by both the resin beads and the interstitial space between the resin beads. The empty BV of each vessel is 7.14 liters, thus 17.5 BV of water passed through each vessel per discharge. At the end of each discharge period, 500 ml samples were collected of the plant effluent (resin vessel influent) and the treated effluent frome each vessel. Use of the vessels was initiated on September 6, 2000, and the last samples analyzed were obtained on January 9, 2001. Thus nearly 1,100 BV (7,900 liters) of water have passed through each column in the time period, and 62 discharges were sampled and analyzed. In the batch pilot tests, the concentration of perchlorate in the feed to the resin volumes varied dramatically, from 100 to 1,600 ppb, due to fluctuating perchlorate concentrations in the influent to the RLWTF, and depending upon whether or not the Reverse Osmosis (RO) is being operated after the tubular ultrafilter. Perchlorate was completely removed from 1,069 BV of the RLWTF plant effluent by both the chloride and hydroxide resins, which each contained 1.87 gallons of resin. The 1.87 gallons of resin thus removed perchlorate from 1,999 (1.87 x 1,069) gallons of RLWTF effluent. The perchlorate concentrations in the effluent from the two columns were generally at the detection limit of 4 ppb. BVs to breakthrough of several anions was determined for each resin, and is given below.
Continuous Pilot Tests For the continuous pilot tests, two additional IX columns were obtained to be run in a continuous mode. By operating the new columns in continuous mode, feeding 500 BVs per day, information on breakthrough and exhaustion of the columns could be obtained more quickly than in batch mode. For the continuous mode tests, feed to the columns was from the north or south 20,000 gallon steel effluent tanks. The columns were fed by a peristaltic pump at 2.5 lpm for the CL-SBA resin and at 1.0 lpm for the OH-SBA resin. With this configuration, the CL-SBA resin was fed >4,000 BV in 15 days. Breakthrough of perchlorate (column effluent concentrations >18 ppb) was noted between 1,800 and 2,100 BVs), and complete exhaustion of the column with respect to perchlorate was noted at just over 4,000 BV. The influent perchlorate concentration to the resins had varied during the test. From BVs 400-700 and 1,200-1,500, the perchlorate concentration ranged from 80-120 ppb. During BVs 3,200-3,500, the perchlorate concentrations were 1,300 ppb. During the reamining period of CL-SBA operation, the perchlorate concentrations were in the 500-800 ppb range. Breakthrough observations of major anions revealed chloride C/C0 greater than 1.0 due to the desorption of the chloride anion by sulfates, nitrates, and perchlorates. Fluoride breakthrough occurred almost initially. Breakthrough of nitrate, sulfate and perchlorate was seen at 200, 800, and 2,000 BVs, respectively. At 4,000 BVs, nearly 14,000 mg perchlorate had been loaded onto the 0.25 cu. ft. or resin, or 56 gms perchlorate per cu. ft. resin. For the OH-SBA continuous mode tests, the feed rate was reduced to 1.0 lpm to assess the kinetic rate of flow. Breakthrough of perchlorate occurred after feeding 1,750 BVs of RLWTF effluent. The average perchlorate feed concentration to the OH-SBA resin was greater than the continuous feed to thee Cl-SBA resin. Total exhaustion of the OH-SBA resin occurred at 3,400 BVs. Both breakthrough and exhaustion of not only perchlorate, but also sulfate and nitrate, occurred at lower BVs in the OH-SBA resin than in the Cl-SBA resin; this may be due to the higher levels of perchlorate, sulfate, and other anions in the feed concentration rather than the slower feed rate. For the OH-SBA test, 20,000 mg/0.25 cu. ft. of resin, or 80 gms perchlorate per cu. ft. of resin, was loaded onto the resin. Thus, higher perchlorate concentrations in the feed result in greater loadings of perchlorate onto the resin. In summary, pilot tests indicate that both the CL-SBA and OH-SBA US Filter A-464 resins can remove 100% of the perchlorate ion over a feed volume of 1,700 bed volumes. Conservatively, additional perchlorate removal (50%) continues for another 1,700 bed volumes until the resin is exhausted with respect to perchlorate. Feed solutions to the columns were actual RLWTF effluent, containing 400 to 600 mg/L of other dissolved solids including the anions: sulfate, chloride, nitrate, bicarbonate, and fluoride. Applying conservative design principles for scale up, two columns operated in series could remove 100% of the perchlorate from 5,100 BV of RLWTF feed, using the first column as a "roughing" column and the second column as a "polishing" column, and rotaing the polishing column into the roughing column mode upon breakthrough in the original roughing column. Two IX vessels with a capacity to hold 28 cu. ft. of resin each are available at the RLWTF, providing capability of treating 532,950 gal. of RLWTF effluent. After treating 5 million gallons in a typical year of RLWTF operation, exhausted resin would fill 36 55 gal. Drums, and would likely be a low level waste that would be shipped to an off-site incinerator for thermal processing. Cost analyses were performed for the RLWTF at LANL. Recommendation is made to pilot test two additional resins (Rohm & Haas IMAC HP555 and Sybron Chemical SR-7) that are suspected to have higher selectivity for perchlorate than the US Filter A-464 resin pilot tested at the RLWTF. The Rohm & Haas resin is projected to remove three times the amount of perchlorate that is removed by USF-A-464, but the resin is three times as expensive. However, the three-fold decrease in waste volumen, and less manpower and related expenses to remove the resins and prepare them for offsite shipment could result in a cost savings if the resin performs as expected. Recommendation is made to incinerate rather than to chemically regenerate the perchlorate loaded resins to eliminate handling of a secondary highly concentrated perchlorate stream. Additional Info Source: Worland, V. Peter, 2001: "Pilot Scale Tests to Remove Perchlorate from the RLWTF Effluent by Ion Exchange", FWO-WFM, January 30, 2001 |
|
|||||||||||||||||||
      
|
|||||||||||||||||||||