Rick Andrew is director of Global Business Development for Water Systems at NSF International. Andrew can be reached at [email protected] or 734.913.5757.
Mar 02, 2006

No Small Task

Testing water treatment products for bacteria removal

It sounds easy, doesn’t it? It seems like a testing laboratory could just challenge a water treatment product with some bacteria-contaminated water and make sure the product kills or removes the bacteria by testing the treated water. It’s consideration of the details that reveals the complexity of trying to develop test protocols to support manufacturers’ claims of bacteria removal.

The issues raised by these details begin to crop up when one asks questions such as:

  • What bacteria should be used for testing?
  • What about viruses and cysts?
  • When should samples be collected?
  • What should be the chemistry of the test water?
  • When should the test be considered completed?
  • What are the criteria for passing the test?

This article will examine these questions and provide some thoughts, information and rationales to explain why certain choices are made. Although currently there are no national standards for verification of bacteria removal claims except for UV (NSF/ANSI 55) and distillation (NSF/ANSI 62) systems, there are efforts underway to develop them for other types of systems. In the meantime, manufacturers of other technologies who want to establish support for their bacteria claims are compelled to use certain existing protocols (perhaps modifying them to be consistent with current thinking). A similar thought process will be outlined in this article.

UV and Distillation

The protocols for testing low-pressure mercury UV systems and distillation systems are well established and have been in place for years. Low-pressure mercury UV has been studied under enough different conditions with different organisms that its effectiveness can be based on the UV dose. The UV dose is a complex function of the intensity of the UV source and the contact time between the UV source and the water. It is well known and accepted that a dosage of 16 mJ/cm2 is sufficient to inactivate nuisance bacteria, and a dosage of 40 mJ/cm2 will inactivate or kill pathogenic bacteria, Cryptosporidium, Giardia and pathogenic viruses (except adenovirus).

UV dosage is commonly measured with Bacillus subtilis or MS-2 coliphage, each of which has very linear responses to UV dosage in terms of log reduction over the range of interest. NSF/ANSI 55 uses these two organisms for testing of UV systems.

Distillation as a purification technology has also been studied very well. These studies have led to the use of B. subtilis as an appropriate indicator organism. A distillation system capable of killing B. subtilis as it operates with sampling throughout production of a batch, or with repeated samples collected from automatic systems, will also be capable of rendering water safe from other bacteria and viruses.

Bacteriostatic Claims

There is an optional test method and claim under NSF/ANSI Standard 42 for bacteriostatic effects. The claim and test method establish that a product does not contribute to bacteria loading in the water being treated. In other words, the water leaving the product is no higher in bacteria counts than the water entering the product. This method does not establish that the product kills or removes bacteria.

The Terrible Trio

Bacteria, viruses and cysts represent the three types of microbiological contaminants that may be found in water. Bacteria and viruses tend to go together in that the environmental conditions and sources are common to both. Protozoan cysts may or may not be associated with bacteria and viruses; however, they all tend to be associated with livestock waste.

Cysts are unique because they are resistant to chlorine. Bacteria and viruses are easily controlled with chlorine, but the protective shells of dormant-stage cysts allow them to cruise right through chlorinated water distribution systems, only to spring into their active state once they have been consumed.

Following this occurrence pattern, the NSF Task Group on Microbiological Standards has proposed a requirement that bacteria reduction claims are accompanied by tested claims of virus reduction and cyst reduction. Claims of cyst reduction only, independent of virus and bacteria reduction, are currently offered under NSF/ANSI 53 and 58, based on the fact that cysts may be present in public water supplies that are safe from contamination by bacteria and viruses.

Test Conditions and Protocols

Test conditions, such as inlet pressure, flow rate, sampling points, water chemistry, use conditions and test end points, can dramatically affect bacteria reduction test results. There is also the question of what bacteria to test. Certainly it is not practical to challenge a water treatment product with all types of potentially occurring waterborne pathogenic bacteria. The approach is to find “surrogate” bacteria that can serve as an indicator to how other bacteria would fare when subjected to the product treatment.

There are several criteria to consider when selecting surrogate bacteria (see Table 1). Once surrogate organisms are selected, the parameters related to the test protocol must be considered (see Table 2). A robust test protocol creates challenging conditions for the product within the reasonable limits of what might be encountered during the use of the product. If the presence of organic matter negatively affects the product’s performance, it would not be appropriate to conduct a test without the presence of organic matter. This analysis must be conducted for each parameter of the protocol as it is developed, which is no small task.

With the protocol fully fleshed out, the remaining question is, “How are the results evaluated?”

How Much is Enough?

Everyone would agree that 99% of anything is significant. Likewise, 99% reduction of bacteria may seem to be quite effective. After all, a certain well-known soap boasts of being 99.44% pure, which sounds pretty good to me. But in the case of bacteria reduction, an even higher standard of performance is required. In some cases, just a few pathogenic bacteria can be enough to cause infection. Because of their potentially very low infectious dose, the U.S. Environmental Protection Agency established in its 1987 “Guide Standard and Protocol for Testing Microbiological Water Purifiers” that a 6 log (99.9999%) reduction of bacteria was required to demonstrate product performance.

This level of performance is still considered appropriate today. Various states, as well as the NSF Task Group on Microbiological Standards, consider 6 log reduction of bacteria a must for products making such claims.

The Guide Standard requires a 4 log (99.99%) reduction of viruses and a 3 log (99.9%) reduction of cysts. The standard is consistent with current thinking with respect to viruses, but 3.3 log (99.95%) reduction of cysts is now considered necessary in standards 53 and 58.

A Complicated Topic

I commend you for sticking with this complicated article all the way to the conclusion! Developing protocols and testing claims of water purification are significant endeavors reserved for individuals with great interest and stamina. And to be honest, this article only scratches the surface of the true level of effort required to evaluate water treatment products for purification capability.

For any of you who have had a mild degree of interest in the topic, I hope this article was enough of a primer to provide you a broad basis of understanding. In addition, it may have also shown you enough of the complexity of the issue to point you toward a simpler endeavor, such as designing microchips, getting into brain surgery or specializing in rocket science.

About the author

Rick Andrew has been with NSF International for seven years, working with certification of residential drinking water products. He has been the technical manager of the Drinking Water Treatment Units Program for four years. Andrew is a member of the Editorial Advisory board of Water Quality Products. He can be reached at 800.NSF.MARK, or by e-mail at [email protected].