Jump to main content.


  Arsenic Removal from Drinking Water by Adsorptive Media
U.S. EPA Demonstration Project at Rollinsford, NH
Final Performance Evaluation Report
(EPA/600/R-09/017) February 2009

This report documents the activities performed and the results obtained from the arsenic removal treatment technology demonstration project at Rollinsford, New Hampshire. The objectives of the project were to evaluate: 1) the effectiveness of AdEdge Technologies’ AD -33TM media in removing arsenic to meet the new arsenic maximum contaminant level (MCL) of 10 μg/L; 2) the reliability of the treatment system; 3) the required system operation and maintenance (O&M) and operator skills; and 4) the capital and O&M costs of the technology. The project also characterized water in the distribution system and process residuals produced by the treatment system.

The Rollisford, NH demonstration project consisted of two study phases. The source water for both studies consisted of water from two wells having a flow capacity of 95 to 112 gal/min (gpm). Phase 1 of the study utilized an Arsenic Package Unit (APU)-100 system designed for a flowrate of 100 gpm. Because higher flowrates up to 112 gpm were experienced in Phase 1, a 120-gpm APU-RWS system was designed and installed for Phase 2 of the study. Both packages units contained the AdEdge AD-33TM media, which is an iron-based adsorptive media developed by Bayer AG under the brand name of Bayoxide 33.

The Phase 1 APU-100 system consisted of two 36-in-diameter, 72-in-tall pressure vessels in parallel configuration, each initially containing 27 ft3 of AD-33™ media supported by a gravel underbed. Empty bed contact time (EBCT) for the system was approximately 4.0 min per vessel. Hydraulic loading to each vessel based on a design flowrate of 100 gpm was approximately 7 gpm/ft2. The Phase 2 APU-RWS system consisted of two 48-in-diameter, 72-in-tall pressure vessels in parallel configuration, each initially containing 30 ft3 of AD-33™ media also supported by a gravel underbed. EBCT for the APU-RWS system was approximately 3.7 min based on a media volume of 30 ft3 per vessel. Hydraulic loading to each vessel based on a design flowrate of 120 gpm was about 4.8 gpm/ft2.

The APU-100 system included a carbon dioxide (CO2) injection module with manual controls for pH adjustment prior to arsenic adsorption. Contributing, in part, by mechanical problems, the CO2 system failed to consistently adjust pH to the target value of 7.0. Attempts were made to upgrade the manual pH control system for automatic operation to provide for better control for Phase 2; however, the system automation was never completed because the CO2 injection membrane was subject to fouling that could not be resolved. As a result, pH adjustment was not performed during the Phase 2 study.

Two system performance runs were conducted in the Phase 1 APU-100 treatment system. Run 1 operating from February 9, 2004, through October 27, 2004, and Run 2 from November 3, 2004, through January 15, 2005. The replacement system, APU-RWS, was evaluated under Phase 2 from June 13, 2005, through May 8, 2006. During Phase 1, the system was sometimes operated with only one supply well to reduce the flowrate to the system, thereby reducing the inlet pressure and differential pressure (p) in order to extend the time between backwash events. During Phase 2, the system also was operated with one supply well to reduce the flowrate to the system to try to improve arsenic removal performance.

Run 1 of the system treated approximately 11,926,000 gal of water based on totalizer readings from each vessel, operating 11.6 hr/day with an average flowrate of 95 (with both supply wells operating) or 60 gpm (with one supply well operating). Run 2 of the system treated approximately 3,921,000 gal of water operating 10.5 hr/day with an average flowrate of 112 gpm (with both supply wells operating). During Phase 2, the APU-RWS system treated approximately 12,881,000 gal of water, operating 9.7 hr/day with an average flowrate of 97 (with both supply wells operating) or 58 gpm (with one supply well operating). The EBCTs for Run 1 in each vessel ranged from 3.0 to 7.0 min with both wells running and from 4.3 to 9.5 min with only one well running. During Run 2, the EBCTs ranged from 2.5 to 3.9 min. The EBCTs in Phase 2 ranged from 4.0 to 5.6 min with both wells running and from 4.0 to 10.0 min with only one well running.

During Phase 1, higher than normal system p and inlet pressures were experienced. Consequently, the operator conducted frequent backwashes and worked with the vendor to troubleshoot, modify, and replace several system components. The aggressive and frequent backwashing resulted in high media loss – up to 46 to 59% by the end of the study. The system design for Phase 2 successfully addressed the elevated pressure and eliminated the need for frequent backwashes.

Total arsenic concentrations in source water ranged from 28.7 to 52.4 μg/L with As(III) comprising a significant portion of the total soluble arsenic, with concentrations ranging from 7.6 to 28.8 μg/L. After one and one half months of Run 1 operation, the preexisting chlorine injection system was used to prechlorinate the source water and effectively oxidized the As(III) to As(V). The prechlorination step continued throughout the remainder of the study.
Breakthrough of total arsenic at concentrations above the 10 μg/L target MCL was first observed after the APU-100 system had processed between 12,500 and 17,000 bed volumes (BV) of water, representing about 17 to 23% of the estimated working capacity of 74,000 BV. The media performed similarly during the Phase 2 operation of the APU-RWS system. Although the re-design of the system helped alleviate both inlet pressure and Δp problems, it did not improve the media performance in terms of its run length.

Backwash wastewater contained soluble arsenic concentrations ranging from 9.5 to 33.8 μg/L. Soluble iron and soluble manganese concentrations ranged from <25 to 115 and 3.3 to 75.7 μg/L, respectively. As expected, total arsenic, iron, and manganese concentrations were considerably higher than the soluble concentrations, indicating the presence of particulate material in the backwash wastewater. Particulate arsenic might be associated with either iron particles filtered out by the media beds during the service cycle or the media fines. Based on the total suspended solids (TSS) values, approximately 8 lb of suspended solids would be produced in 1,890 gal of backwash wastewater from the vessels for Phase 1 and approximately 25 lb of solids would be produced in 4,200 gal of backwash wastewater for Phase 2.

The spent media passed the Resource Conservation and Recovery Act (RCRA) Toxicity Characteristic Leaching Procedure (TCLP) test for all metals, with only barium showing detectable concentrations ranging from 0.95 and 0.96 mg/L. The average arsenic loading on the spent media based on the inductively coupled plasma-mass spectrometry (ICP-MS) results was 1.88 mg/g or 0.188%. This 1.88 mg/g loading compared well (98%) with the average adsorptive capacity of 1.93 mg/g measured by dividing the area between the influent and effluent breakthrough curves by the amount of dry media in each tank.

Distribution system water samples were collected before and after the installation of the treatment system to determine any impact of arsenic treatment on the lead and copper level and water chemistry in the distribution system. However, because the distribution system in place was a looped system that included water from a third untreated well (General Sullivan Well), the impact of the treated water could not be exactly determined.

The capital investment cost for the re-designed APU-RWS system was $131,692, which included $105,805 for equipment, $4,672 for engineering, and $21,215 for installation. Using the system’s rated capacity of 120 gpm (172,800 gal/day [gpd]), the capital cost was $1,097/gpm ($0.76/gpd). These calculations do not include the cost of a building to house the treatment system. The unit annualized capital cost is $0.20/1,000 gal, assuming the system operated 24 hours a day, 7 days a week, at the system design flowrate of 120 gpm. The system operated only 10 hr/day, producing 21,243,000 gal of water per year with both wells operating. At this reduced usage rate, the unit annualized capital cost increased to $0.59/1,000 gal.

The O&M cost for the APU-RWS system was estimated at $3.59/1,000 gal, which included media replacement and disposal, electricity consumption, and labor. Chlorination was not included in the O&M cost calculation because it was part of the existing treatment system.


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.

Office of Research & Development | National Risk Management Research Laboratory

Local Navigation

Jump to main content.