By Jane Wilson

On Oct. 31, 2001, the U.S. Environmental Protection Agency (USEPA) finalized a new maximum contaminant level (MCL) of 10 parts per billion (ppb) for arsenic in drinking water. This action effectively lowered the previous arsenic MCL of 50 ppb by 80 percent. An implementation date of January 2006 has been established for public water systems to meet the new arsenic regulation. Some 4,000 public water systems, the majority of which serve fewer than 10,000 customers, will need to institute new treatment procedures to meet the new arsenic standard.1

Residential drinking water treatment units (DWTUs) are expected to play a significant role in the implementation of the lowered arsenic MCL. For some water systems, use of residential drinking water treatment units may represent a low-cost treatment option for meeting the more stringent arsenic regulation. Theses units also provide a means for consumers using private water supplies with significant arsenic contamination to reduce the arsenic concentration in their water to below recommended levels. Also, given the lengthy implementation period, residential treatment will provide a means for public water supply consumers to reduce their exposure to this contaminant in advance of the date their water utility is required to comply with the new standard.

Arsenic Task Group work
The NSF Arsenic Task Group has devoted considerable time to development of arsenic reduction protocols for the NSF DWTU Standards. A proposed test protocol for addition to ANSI/NSF 53 Drinking Water Treatment Units-Health Effects is expected to be under ballot by the DWTU Joint Committee by the end of the first quarter of 2002, along with a revision to the current arsenic removal requirements in ANSI/NSF 58 Reverse Osmosis Drinking Water Treatment Systems. These first proposals address pentavalent arsenic reduction — also known as As V, As(+5) or arsenate — for treatment technologies falling under these two standards.

A pentavalent arsenic reduction claim is being established for Standards 53 and 58 since this treatment function will likely address the majority of treatment needs for consumers on public water supplies that use chlorination for residual disinfection. The presence of a detectable free chlorine residual has been demonstrated to effectively convert trivalent arsenic present in the water—also known as As III, As(III), As(+3) or arsenite—to the pentavalent form.2 Many groundwater sources that have undergone arsenic speciation (separate quantitation, or measure, of the trivalent and pentavalent species) have also been demonstrated to predominantly contain pentavalent arsenic.

Proposed Standard 53 requirements
The proposed testing protocol for Standard 53 includes two alternative challenge concentrations, 300 and 50 ppb. These concentrations were determined by task group analysis of U.S. Geological Survey data for water sources used for drinking water containing arsenic. The 300 ppb challenge is equivalent to that used in the current Standard 58 test protocol. For the 50 ppb challenge level, this corresponds to the previous MCL value, so public water systems are already at a maximum of 50 ppb. The challenge water chemistry has been designed to mimic an average groundwater composition in terms of ionic species present. The test will be performed using test waters of pH 6.5 and 8.5, like other metals reduction tests in Standard 53. Other test conditions, such as cycle time and operating pressure, will be the same as already established in this standard.

Because of the complexity of arsenic chemistry, detailed labeling and literature requirements also are being proposed. Basic information about the device’s pentavalent arsenic removal function will be required to appear on product packaging, and a separate section of the Performance Data Sheet will be devoted to detailed information about the arsenic removal function.

Validation study
A four laboratory, round-robin study was administered by the task group to assess the repeatability and reproducibility of the proposed Standard 53 protocol for pentavalent arsenic removal. Each laboratory was provided with a number of identical single-housing, undersink point-of-use devices containing an iron-based activated alumina media. The laboratories tested the units according to the proposed protocol and reported both arsenic removal performance and general water quality characteristics back to the task group for review.

The reported treatment capacities (as defined by arsenic breakthrough exceeding 10 ppb) for the pH 8.5 tests were consistent among the four laboratories. More variability was observed in reported treatment capacities for the pH 6.5 test. The task group attributed this variability in part to variation in the timing of effluent sample collection procedures among laboratories. To minimize this variable, the task group is defining specific requirements regarding the timing of effluent sample collections, particularly after overnight rest periods when some temporary recovery of arsenic treatment capacity may occur.

The reported results for the general water quality parameters indicated the arsenic challenge concentration and pH of the challenge water are stable under conditions of the test. All four laboratories did report at least one general water quality parameter outside of the limits specified in the protocol. Parameters presenting the most difficulty were calcium and phosphate. Since calcium chloride tends to absorb water, the task group has recommended it be dried prior to use in formulating the challenge water, so the appropriate weight is added. Some phosphate analysis methods are subject to interferences from arsenate and fluoride, which are also in the challenge water.3

The Arsenic Task Group concluded the results of the validation study support use of the proposed protocol in evaluating pentavalent arsenic removal performance of treatment technologies under Standard 53. The test protocol and the proposed literature and labeling requirements are undergoing a final review by the task group prior to balloting by the DWTU Joint Committee.

Proposed Standard 58 revisions to claims
Standard 58 is also being revised based on some of the proposed changes to Standard 53. Currently, Standard 58 contains an arsenic reduction claim that’s limited to use on water systems with a detectable free chlorine residual. This claim is being revised to a pentavalent arsenic reduction claim for use on water systems having a detectable free chlorine residual, or that have otherwise been determined to contain only pentavalent arsenic (e.g., through arsenic speciation testing or pre-treatment with another oxidant that is effective in converting trivalent arsenic to pentavalent arsenic2). It’s also proposed that the 50 ppb challenge level be included under Standard 58 in addition to the current 300 ppb test. This will accommodate high flux or other RO systems that may not be able to achieve the 97 percent arsenic reduction required by the current test, but that achieve arsenic reduction to at least 10 ppb at a lower influent challenge level. These units may be suitable for the municipal treatment market where influent arsenic concentrations will be 50 ppb or less. No changes to the other general water quality characteristics of the challenge water are proposed at this time. The same new labeling and literature requirements proposed for Standard 53 will also be instituted for RO systems.

While the Arsenic Task Group work on pentavalent arsenic reduction is drawing to an end, the group remains committed to designing a total arsenic (trivalent and pentavalent) removal claim for drinking water treatment units as well. This additional work will include performance evaluation of oxidative technologies that convert trivalent arsenic to the pentavalent species. Designing a challenge water containing trivalent arsenic is considerably more complicated from a technical standpoint, since oxidation of trivalent arsenic must be controlled to ensure the stability of the challenge concentration throughout the test. The testing of oxidative technologies also requires different challenge water characteristics, such as the addition of iron and manganese, two common ions that may interfere with the oxidative process when present in source waters. The material safety requirements of all the ANSI/NSF DWTU Standards will also be updated to recognize the new arsenic MCL as the appropriate extraction limit.


  1. USEPA, “Arsenic Fact Sheet,” EPA 815-F-01-010, October 2001.
  2. USEPA, “Laboratory Study on the Oxidation of Arsenic III to Arsenic V,” EPA/600/R-01/021, March 2001.
  3. Standard Methods for the Examination of Water and Wastewater, APHA, 20th edition.

About the author
Jane Wilson is the senior project manager for Water and Environmental Standards in NSF International’s Standards Department. She has been with NSF since 1990. Her bachelor’s degree in medical technology and a master’s degree in public health (MPH) are both from the University of Michigan. NSF is located in Ann Arbor, Mich. Wilson can be reached at (800) 673-6275 or email: [email protected].


Comments are closed.