By Rick Andrew
When testing water filtration products, obviously water must be used. To the uninitiated, it might seem that any tap water should be adequate for testing these products. Experienced water treatment industry professionals, however, recognize that water characteristics play a key role in the performance of filtration products. The NSF International Joint Committee on Drinking Water Treatment Units (DWTUs), responsible for the development of the ANSI/NSF DWTU standards, is comprised of a group of experts from the areas of public health, product users and specifiers, and industry professionals. This group certainly recognizes the importance of water characteristics in specifying the test waters used in these standards.
Tap water for testing
It’s quite convenient to test water filters using a public water supply. This water is readily available and economically priced. To control testing costs and allow for convenient testing, many of the tests in ANSI/NSF Standards 42 and 53 currently use, or use as a starting point, publicly available water supplies.
One of the goals of the DWTU Joint Committee is to provide standards that can be used by qualified testing bodies, whether these bodies are independent laboratories, manufacturers or others. As more qualified testing bodies test to the standards, cases have occurred in which different laboratories obtained different results when testing the same products. Often, these different results were traced back — through the efforts of technical task groups — to differences in the water chemistry of the public water supplies at different test facilities. In these cases, the groups determined what differences in water chemistry were responsible for the different results. This work has resulted in additional test water specifications for the important water characteristics for these tests. These additional specifications mean that the public water supply must sometimes be modified for specific tests. In certain cases, the public water supply must be treated by reverse osmosis (RO) and deionization (DI) and then “built back up” to a specific water chemistry.
Currently, there are 11 different test waters specified in Standard 42, and 10 different test waters specified in Standard 53. Table 1 
describes these different waters.
Test water development
Standard 53 requires that metals reduction performance claims must be tested under two different pH conditions, 6.5 and 8.5. Because metals are more soluble at lower pH, the pH 6.5 challenge water represents primarily soluble metals, whereas the pH 8.5 challenge water may include some colloidal or particulate forms of specific metals.
Several years ago, some laboratories began to have problems when conducting lead reduction testing with the pH 8.5 challenge water. Some filters were clogging part way through the test. As a result, laboratories couldn’t complete the test to the required volume to verify the manufacturer’s claimed capacity for these filters. The problem wasn’t that lead was getting through the filters, but that the filters would clog prematurely.
This issue was brought to the attention of the DWTU Joint Committee. The committee formed a task group to resolve the issue, with the directive that the standard be modified if necessary to help alleviate these clogging problems.
The task group was made up of laboratories and industry experts who knew the clogging was caused by the lead precipitating from solution and forming lead hydroxide solids. Still, not all laboratories suffered the clogging to the same degree. The reason? The public water supplies for different laboratories in different areas had varying water chemistries. The group found that very subtle differences in water chemistry make big differences in the solubility of lead at pH 8.5.
The group came up with an innovative solution to the problem. Rather than zero in on the specific water chemistry parameters that caused the lead solubility problems and specify these parameters in the challenge water requirements, they took a different approach. They developed a specific challenge water from “scratch” using RO/DI water. Because RO/DI water is essentially free of all dissolved species, it’s the same regardless of the local public water supply chemistry. It’s “pure” water in a certain sense. This pure water then has various reagents added to it in specific amounts, order and times to achieve a tightly controlled challenge water that will be the same regardless of where and by whom it’s prepared. The procedure for making this water is so specific that it’s actually called a “recipe.”
It’s more expensive and time consuming to treat water by RO/DI and then add reagents to it to achieve a specific chemistry than it is to use a public water supply for testing. In this case, though, the extra work and expense are necessary to achieve the goal of comparable test results among laboratories.
Other test water specs
Total organic carbon (TOC) has a large influence on the organic contaminant reduction performance of activated carbon filters. Activated carbon filters have a much greater capacity for organic contaminant reduction when there’s very little TOC present than when there’s a high level. The effect of TOC on organic contaminant reduction has been recognized by the DWTU Joint Committee. Currently, the Standard 53 chemical reduction general test water requires that a public water supply with TOC content of at least 1.0 milligram per liter (mg/L) be used for conducting these tests.
Because of this open-ended specification, there have been differences in the organic contaminant test results for specific products tested at different laboratories. This has been especially true when testing for chloroform reduction — the surrogate chemical for the Standard 53 VOC and total trihalomethanes (TTHM) reduction claims. Chloroform is only weakly adsorbed by activated carbon, which is why it’s used as the surrogate compound for these tests. A task group of experts has been formed by the DWTU Joint Committee to investigate this phenomenon and modify Standard 53, if necessary, to foster the goal of having comparable test results achieved at different laboratories. This group has determined that the composition of the TOC in the challenge water — whether it’s naturally occurring TOC or laboratory-added TOC — and the amount have an impact on the performance of activated carbon filters for chloroform reduction. The composition of naturally occurring TOC in public water supplies can vary dramatically based on the public water supply source, water treatment technology and the local climate.
This group is currently working to develop a specific synthetic TOC formulation that will be added to RO/DI water. This will result in chemical-reduction, general test water that will have an identical composition regardless of the local public water supply chemistry. This uniform chemical reduction general test water will improve the comparability of organic contaminant reduction test results generated at different laboratories.
As in the case of lead at pH 8.5, the result will be water that’s more expensive and time consuming to prepare than using the public water supply. Yet, the extra work and expense will be necessary to achieve the goal of comparable test results among laboratories.
Conclusion
It would be simple to just use the local public water supply for conducting testing of drinking water treatment devices. Due to water chemistry variations that can cause differences in results obtained from different laboratories, however, this isn’t always possible. As the DWTU Joint Committee continues its work to improve and fine-tune the standards, development of the various test waters continues. The trend has been to test waters that are “built up” from RO/DI water through a specific recipe to achieve a tightly defined and highly reproducible chemistry appropriate for the test water. This approach may be more time consuming and expensive than testing with tap water, but it aids in the goal of having consistent test results at different laboratories. After all, for a standard to be true, the test results of one qualified testing body on a specific product should agree with the results of another.
About the author
Rick Andrew is the technical manager of the Drinking Water Treatment Units program at NSF International, of Ann Arbor, Mich. He has been with NSF for 3-1/2 years. His previous experience was in the area of analytical and environmental chemistry consulting. Andrew has a bachelor’s degree in chemistry and a master’s degree in business administration from the University of Michigan. He can be reached at email: [email protected].