Jump to main content.


  Arsenic Removal from Drinking Water by Adsorptive Media, U.S. EPA Demonstration Project at Rimrock, AZ, Final Performance Evaluation Report (EPA/600/R-08/008) March 2008

This report describes the activities and results of the arsenic removal treatment technology demonstration at the Arizona Water Company (AWC) facility in Rimrock, Arizona. The objectives of the project were to evaluate the:

  • Effectiveness of AdEdge’s Arsenic Package Unit-100 (APU-100) AD-33 adsorptive media system in removing arsenic to meet the maximum contaminant level (MCL) of 10 micrograms per liter (µg/L)
  • Reliability of the treatment system for use at small water facilities
  • Simplicity of the required system operation and maintenance (O&M) and operator skill level
  • Capital and O&M costs of the technology

The project also characterized water in the distribution system and residuals produced by the treatment process. The types of data collected included system operation, water quality, process residuals, and capital and O&M costs.

The APU-100 treatment system consisted of a 25-micrometer bag filter assembly, two 3-foot-by-6-foot composite fiberglass pressure tanks, a backwash wastewater recycling system, associated piping and Fleck controller valves, and an instrument/control panel. Each tank contained 22 cubic feet (ft3) of Bayoxide E33 iron-based adsorptive media developed by Bayer AG and branded AD-33 by AdEdge. Due to the loss of one of AWC’s production wells, the system flow rate was reduced from 90 to 30 gallons per minute (gpm), prompting a change in system configuration from parallel to series (lead/lag). The reconfigured APU-100 system had a design capacity of 45 gpm and began operation on June 24, 2004. Actual flow rates through the system averaged 30 gpm, corresponding to an empty bed contact time of 5.4 minutes per tank and a hydraulic loading rate of 4.2 gallons per minute per square foot (gpm/ft2).

Source water contained 43.8 to 81.4 µg/L of total arsenic, with arsenic (V) as the predominant species. Prechlorination, though not required for oxidation, was performed to provide disinfection throughout the treatment train and the residuals within the distribution system. Concentrations of iron, manganese, silica, orthophosphate, and other ions in source water did not appear to affect arsenic removal. The system operated for 12 or 24 hours per day on a timer with 1 to 2 percent downtime for repairs and media replacement. After treating 52,150 bed volumes or 17,164,000 gallons of water during media run 1A, based on 44 ft3 of media in the lead and lag tanks, the system effluent reached the 10-µg/L arsenic MCL on August 9, 2006. Only the lead tank was rebedded because the media in the lag tank still had about 50 percent of adsorptive capacity remaining. After rebedding, the tank positions were switched with tank B containing partially exhausted media in the lead position and tank A with virgin media in the lag position. Media run 1B thus began on November 27, 2006. Personnel monitored the system until March 28, 2007, to ensure normal operations after the media changeout.

Comparison of the distribution system sampling results from three residences before and after system startup showed a decrease in the average arsenic concentration from 48.8 to 19.3 µg/L. However, samples of the distribution system water exhibited higher arsenic concentrations than those of the treatment system effluent due to blending of the treated water with untreated water from other source wells. Alkalinity, pH, iron, manganese, lead, and copper concentrations did not appear to be affected by the system operation.

Backwashing of the media was initially conducted automatically, but due to several unscheduled backwash events and the need to obtain operational data and backwash wastewater samples, the programming was changed to manual initiation once every 30 days. Backwash frequency was eventually decreased to quarterly due to minimal differential pressure increase across the tanks between backwash events. Backwash was performed using source water for 15 minutes per tank at approximately 47 gpm, or 6.6 gpm/ft2. Backwash wastewater from the lead tank generally contained higher concentrations of all analytes than that from the lag tank, most likely because the lead tank removed most of the particulates from source water. A piping loop, a recycle tank, and a metering pump enabled the system to reclaim nearly 100 percent of the wastewater produced by blending it with intake after prechlorination but before arrival at the adsorption tanks, at a rate of 0.5 gpm.

The capital investment for the system was $88,307, including $63,785 for equipment, $11,372 for site engineering, and $13,150 for installation. Using the system’s rated capacity of 45 gpm (or 64,800 gallons per day [gpd]), the capital cost was $1,962 per gpm (or $1.36 per gpd). The capital cost was converted to a cost of $8,335 per year based on a 7 percent interest rate and a 20-year return period. During the first year, the system produced approximately 8,505,000 gallons of water, so the unit capital cost increased to $0.98 per 1,000 gallons. These costs do not include the cost of the system enclosure and backwash recycling system.

O&M costs, estimated at $0.86 per 1,000 gallons, included costs for media replacement and disposal, electricity, and labor.


Tom Sorg

See Also

Arsenic Research

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.