By Rick Andrew

The quality of bottled water is assured through many approaches and regulatory requirements, varying to some degree by geographical jurisdiction. In general, bottled water is treated as a food product, subject to oversight similar to other packaged foods and beverages. Typically, the quality of the water is assured through analysis for trace contaminants that may be present due to contamination of the source water or through impacts related to treatment and processing. Additionally, there is a focus on production facility cleanliness, similar to the approach used in production facilities for other types of packaged food and beverages. Given this background, it might be a little surprising to find that NSF/ANSI 53 Drinking Water Treatment Units – Health Effects addresses bottled water filters.

Why bottled water filters under NSF/ANSI 53?
Requirements for bottled water filters were added to NSF/ANSI 53 in the late 1990s. Basically, manufacturers of these filters were seeking a way to provide buyers with assurance that their filters were of high quality and would perform under conditions typical to a bottled water treatment and packaging facility. The cyst reduction claim was seen to be an appropriate one for proving performance, especially given the high-percent reduction requirements, small physical size and acute health effects associated with it. The test operation under NSF/ANSI 53 for typical POU and POE systems making cyst reduction claims, however, was not consistent with operation in a bottled water facility. So, a new protocol was developed.

Differences between operations
There are two significant differences in operation between typical POU/POE and operation in a bottled water facility:

Reduction in flowrate versus continuous flowrate

  • Typical plumbed-in mechanical filtration POU/POE systems are driven by line pressure. As the filters become loaded and begin to clog, the pressure drop increases and the flowrate decreases.
  • Bottled water plants are typically operated at a constant flowrate, so the filling of the bottles is consistent and the bottling line can be more easily controlled. As pleated filters used in bottled water plants become loaded and begin to clog, adjustments in pressure can be made to keep the water flowing at a constant rate and therefore keep the bottling line moving at a constant speed.

On/off cycling versus flowing continuously

  • Most POU/POE systems are used sporadically, for short periods of time, as end users draw treated water from them. This is accounted for in the standard’s contaminant reduction testing procedures by requiring on/off cycling of flow. This is very important because the sudden initiation of flow at the beginning of the on cycle can strain seals and media within the system.
  • Bottled water plants are operated under continuous flow for long periods of time in order to keep the bottling line running smoothly. On/off cycling would not be appropriate for a testing protocol for this end-use application.
    These differences in operation have been taken into account in the cyst reduction test protocol for bottled water plant filters in NSF/ANSI 53.

Details of the test protocol
Two bottled water filters per the manufacturer’s instructions using the general test water are described in Figure 1. Testing begins using the cyst microsphere challenge water (see Figure 1) at the rated service flow specified by the manufacturer with a dynamic test manifold inlet pressure of up to 620 kPa (90 psig) with continuous flow. The manufacturer’s rated service flow ± 10 percent is maintained throughout the test using a control valve located downstream of the test filters. The cyst microsphere challenge water is introduced until the collection of the start-up sample is completed.

The laboratory then changes the water being introduced to the filters from the cyst microsphere challenge water to the test dust loading water (see Figure 1). This water is introduced to the filters until the pressure drop across the filters is 25 percent of the manufacturer’s maximum recommended pressure drop. At this point, the laboratory changes to the general test water for 10 minutes. It then changes to the cyst microsphere challenge water for 20 minutes. At the end of the 20-minute period, a pressure pulse is administered to the filters by causing a rapid interruption and resumption of flow, typical of a fast-acting valve located downstream of the filters under test. The pressure pulse is included in the test protocol to assure that the filter integrity is sufficient to withstand this kind of pulsation and still provide 99.95 percent removal of cysts. After administering the pressure pulse, challenge water and filtered water samples are collected.

After sampling, the laboratory changes the water being introduced to the filters from the cyst microsphere challenge water to the test dust loading water. The introduction of the test dust loading water is continued until the next sampling point, at which point the procedure is repeated starting at the point of the laboratory changing to the general test water for 10 minutes.

The influent (cyst microsphere challenge water) and effluent (filtered water) samples are collected and measured at the start of the test and at 25, 50, 75, 100 and 150 percent ± 10 percent of the manufacturer’s recommended maximum pressure drop at the rated service flow. Filters that have maintenance procedures (such as re-use, backwashing, cleaning, sterilization, etc.) are tested a second time after the manufacturer’s maintenance procedures are followed. These filters must pass the test both times.

A note on microspheres
In order to make a claim of cyst reduction, bottled water filters must be capable of removing 99.95 percent of three µm polystyrene microspheres at all sample points when evaluated under the protocol described above. The microspheres themselves must have a tightly controlled size specification, with 95 percent of them in the range of 3.00 ± 0.15 μm. Additionally, the spheres must have a low surface charge to assure that the mechanism of removal is purely mechanical filtration, as opposed to adsorption. For this reason, the spheres must have a surface charge content of less than two uEq/g. The microspheres must also contain a fluorescein isothiocyanate (FITC) dye or equivalent, so that they are visible under an epiflourescent microscope. Each sample of challenge water and filtered water must be examined under the microscope so the number of microspheres per liter can be accurately counted. In addition to successfully passing the test for cyst microsphere reduction, bottled water plant filters are evaluated for material safety per the standard’s extraction testing requirements.

Housings for testing
Certification of bottled water filters under NSF/ANSI 53 is limited to filters with -222 or -226 double O-ring seals or a similar redundant sealing mechanism. This type of seal is industry standard and the double O-ring redundant seal helps assure sealing with industry standard housings. Housings for testing purposes must be appropriate to the configuration of the filter (fin end or flat end, etc.) and also sized according to the sizing on the double O-ring seal of the filter being tested (i.e., -222 or -226 size).

The bottom line is that the NSF/ANSI DWTU Standards are tools to help assure quality of products. These criteria and methods for evaluation of bottled water plant filters for cyst reduction performance and safety for contact with drinking water actually help assure quality at two levels in the supply chain. First, it allows manufacturers of these filters to assure the quality of their product and second, it helps bottlers to assure the quality of their product by using filters that conform to these requirements.

About the author
Rick Andrew is NSF’s Director of Global Business Development–Water Systems. Previously, he served as General Manager of NSF’s Drinking Water Treatment Units (POU/POE), ERS (Protocols) and Biosafety Cabinetry Programs. Andrew has a Bachelor’s Degree in chemistry and an MBA from the University of Michigan. He can be reached at (800) NSF-MARK or email: [email protected]


Comments are closed.