Abstract

This report documents the activities performed and the results obtained for the EPA arsenic removal technology demonstration project at the Town of Taos in New Mexico. The main objective of the project was to evaluate the effectiveness of Severn Trent Services’ (STS) SORB 33™ adsorptive media in removing arsenic to meet the maximum contaminant level (MCL) of 10 g/L. Additionally, this project evaluated 1) the reliability of the treatment system for use at small water facilities, 2) the required system operation and maintenance (O&M) and operator skill levels, and 3) the capital and O&M cost of the technology. The project also characterizes water in the distribution system and residuals generated by the treatment process. The types of data collected include system operation, water quality, process residuals, and capital and O&M cost.

The STS system consisted of a carbon dioxide (CO2) pH control system and three 63-in-diameter, 86-in-tall fiberglass reinforced plastic (FRP) vessels in parallel configuration, each designed for approximately 60 ft3 of SORB 33™ pelletized media. The media is an iron-based adsorptive media developed by Bayer AG and packaged under the name SORB 33™ by STS. The system was designed for a flowrate of 450 gal/min (gpm) (or 150 gpm to each vessel), corresponding to an empty bed contact time (EBCT) of about 3.0 min and a hydraulic loading rate of 6.9 gpm/ft2. The actual amount of media loaded based on freeboard measurements was 216 ft3 (or 72 ft3/vessel), thus resulting in a longer EBCT of 3.2 min even at a higher flowrate of 503 gpm.

Upon approval of engineering plans by the New Mexico Environment Department/Drinking Water Bureau (NMED/DWB) and completion of a pipeline project by the Town of Taos, the APU-450 treatment system began operating on February 14, 2006. From February 14, 2006 through October 23, 2007, the treatment system operated for only 215 days, with an average daily operating time of only 3.9 hr. Frequent and prolonged system downtime occurred during the performance evaluation study, caused primarily by non-system-related issues, such as power outages and facility pipeline leakage. Because the treatment system and booster pump station were not integrated with the Town’s system control and data acquisition (SCADA) system, the operator had to manually operate the system. The labor intensive nature of system operations also contributed to the fewer and shorter daily runs. The system treated approximately 22,977,000 gal of water, or 14,192 bed volumes (BV), which was only 11% of the vendor-estimated working capacity for the SORB 33™ media. Because the system was far from reaching the treatment target of 10 μg/L after about 20 months of operation, a decision was made to discontinue the performance evaluation study.

Source water supplied by Well 8 had an average total arsenic concentration of 16.9 g/L with soluble As(V) as the predominating species at 16.8 μg/L (on average). pH values of source water were high, ranging from 9.5 to 9.8 and averaging 9.6. After some troubleshooting, the pH control system effectively reduced pH values of the water entering the treatment system to 7.3 to 7.4, close to the target value of 7.2. The automatic pH control system used a JUMO pH/Proportional Integral Derivative (PID) to regulate CO2 gas flow with signals received from an inline pH probe. CO2 gas was injected to a side stream of water with a microporous membrane module housed in a sanitary cross.

During the performance evaluation study, total arsenic was reduced to an average of less than 1 μg/L, except for the one spike at 7.4, 7.2, and 8.8 μg/L at the TA, TB, and TC sampling locations, respectively. The exact cause of the spike was unknown. Little or no iron or manganese was measured in raw water and system effluent.

Comparison of the distribution system sampling results before and after the system startup showed no significant differences in concentrations of arsenic and other analytes. This was because the treated Well 8 water contributed only 10% of the capacity in a 1,000,000-gal water tower, from which water was distributed either directly to the distribution sytem or indirectly through a 500,000-gal storage tower.

Backwash wastewater was sampled three times during the performance evaluation study. pH values ranged from 7.4 to 8.1 and averaged 7.7, somewhat higher than that of the treated water used for backwash. The water used for backwash was withdrawn from a 50,000-gal holding tank. Some CO2 degassing likely took place during storage and transit, thereby elevating the pH values. As expected, total suspended solids (TSS) values were low, ranging from 16 to 82 mg/L and averaging 37 mg/L. Concentrations of total arsenic, iron, and manganese ranged from 1.1 to 11.8 g/L, from 0.14 to 8.9 mg/L, and from 0.7 to 64.0 g/L, respectively, with the majority of iron and manganese existing in the particulate form. The unexpectedly high iron concentrations in the backwash wastewater might have been media fines produced during the backwashing process.

The capital investment for the system was $296,644 consisting of $202,685 for equipment, $32,750 for site engineering, and $61,209 for installation, shakedown, and startup. Using the system’s rated capacity of 450 gpm (or 648,000 gal/day [gpd]), the capital cost was $659/gpm (or $0.46/gpd) of design capacity. This calculation does not include the cost of the building to house the treatment system.

The O&M included only incremental costs associated with media replacement and disposal, CO2 supply, electricity, and labor. Although not replaced, the media changeout cost was estimated to be $41,749 for all three adsorption vessels, which would represent the majority of the O&M cost. CO2 cost was $0.29/1,000 gal of water treated, most of the CO2 cost was for the lease of four 380-lb dewars and two 50-lb back-up cylinders.

Contact

Thomas J. Sorg

You will need Adobe Reader to view some of the files on this page.
See EPA's PDF page to learn more.