By Rick Andrew

Reverse osmosis (RO) systems frequently incorporate multiple water treatment technologies. Obviously, they all utilize reverse osmosis; additionally, many include pre-filters and post-filters (filters inline prior to the RO membrane and filters inline downstream of the membrane, respectively). These filters serve distinct purposes. The pre-filters are designed to reduce particulate matter that could potentially foul the membrane. They may also be designed to reduce disinfectants such as chlorine, because these agents may have adverse impacts on the membrane’s durability.

Post-filters may be designed to enhance the taste of the treated water and they may have other capabilities as well. NSF/ANSI 58 recognizes that post-filters may have contaminant reduction capabilities and provides protocols for verifying this performance for VOC (volitile organic compound) reduction. This column will review these protocols and discuss the rationale behind them.

Why not have claims on pre-filters?
Standard 58 contains requirements for VOC reduction claims on post-filters, but not on pre-filters. From time to time, I am asked, “Why can’t a man-ufacturer make claims on the pre-filter of an RO system?”

The answer relates to RO system efficiency. A significant amount of influent water is used to flush contaminants away from the membrane in a crossflow path, ending up as reject water. In fact, a majority of the influent water may never end up as product water. For this reason, it is difficult to correlate product water volumes to volumes treated by a pre-filter. Thus, it would be difficult for consumers to track water usage to know when contaminant reduction capacity of a pre-filter had been exhausted.

Post-filters, on the other hand, treat water after it passes through the membrane. There is a one-to-one ratio of product water to water treated by the post-filter. The result is that VOC reduction claims are allowed on post-filters per Standard 58, but not on pre-filters.

Two different categories of post-filters
There are two distinct positions for post-filter location within an RO system. The post-filter may be located downstream of the membrane but upstream of the storage tank, or it may be located downstream of the storage tank (see Figure 1).

These two different post-filter positions result in two very different usage patterns. The post-filter that is upstream of the storage tank will typically see long periods of flow with very low flow rates. This is because flow will be initiated through the membrane only when the automatic shut-off valve allows flow to refill the tank. Flow rates through membranes are typically on the order of about 0.02 gallons per minute (gpm). So, at this flow rate, it will take an extended period of time with very low flow rates through the filter—often an hour or more—to refill the tank.

The post-filter that is downstream of the storage tank, on the other hand, will see relatively short usage periods with much higher flow rates. This is because flow is initiated when the faucet is opened to draw a glass of water. The pressure in the storage tank provides higher flow rates for a post-filter in this location, typically on the order of about 0.5 gpm to 0.75 gpm. So, the filling of the glass will only take a few seconds.

Different usage patterns leads to different testing protocols
Standard 58 requires that post-filters located upstream of the storage tank be tested for VOC reduction with constant flow at the RO system daily production rate for 16 hours per day. The end point of the test is either 120 percent of the filter’s rated capacity if the system has a performance indication device for the filter, or 200 percent of the filter’s rated capacity in all other circumstances.

In practice, this can be a very long test. At 0.02 gpm, only 1.2 gallons per hour will flow through this filter. Given the 16-hour per day testing period, this amounts to 19.2 gallons per day. For a post-filter with a rated capacity of 500 gallons, even with a PID (performance indication device) it will take 32 days to reach the end of the test. Without a PID, it will take 53 days to complete the test.

Because of the very low flow rates and potentially long test periods that result from testing at an RO system’s daily production rate, the Standard allows for an accelerated test. This option allows the test to be run at any flow rate greater than the RO system’s permeate flow rate with the faucet open. This is a conservative option, because increasing the flow rate through the filter leads to reduced contact time of the water with the filter media, thereby potentially making it a more difficult test for the filter.

Considering the example above, if the flow rate were increased to 0.10 gpm instead of 0.02 gpm, the test would proceed at six gallons per hour or 96 gallons per day. For a post-filter with a rated capacity of 500 gallons and a PID, it will take seven days to complete the test. This is much faster and therefore less labor- and resource-intensive than a 32-day test, making it also much less expensive. The increased flow rate makes it more conservative, too.

Post-filters located downstream of the storage tank, with their relatively higher flow rates and intermittent usage when filling glasses of water as described above, see operating conditions much like typical point of use (POU) filters as covered by NSF/ANSI 53. For this reason, Standard 58 requires that post-filters located downstream of the storage tank are to be tested per Standard 53 for VOC reduction claims. This test involves operating the filter with an on/off cycle, at a flow rate consistent with the highest flow rate delivered by the RO system. For RO systems that use an air charged bladder storage tank, with the air charge set at a given pressure, the highest flow rate delivered by the system will be that achieved with a full storage tank.

Removing the filters for testing
Note that all of the post-filter testing scenarios above refer only to the filter itself. This is because the filters are removed from the RO system for testing. By doing so, any effects on the test from pre-filters or from the RO membrane can be eliminated so only the post-filter is evaluated. Also, the flow rates and on/off periods can be properly controlled by isolating the post-filter.

Filters, filters, everywhere!
The NSF/ANSI DWTU Standards address testing of filters that may be located in a variety of different installations. Some we are very familiar with—POE filters, faucet-mounted filters, countertop filters, under sink filters and gravity feed pitcher filters, to name a few. There are some newer types of filters with unique installation locations that we are becoming more familiar with—refrigerator filters are a prime example. Each of these types of filters has some special considerations for testing, specified in Standards 42 and 53, because of the differences in their usage patterns and operation as dictated by their installation locations.

In this column, we have seen that Standard 58 contains requirements for testing a category of filters that is sometimes overlooked for contaminant reduction claims—RO system post-filters. And like the other types of filters described above, there are unique aspects of the usage pattern for these filters, because of their installation location, that lead to unique testing protocols.

About the author
Rick Andrew has been with NSF International for six years, working with certification of residential drinking water products. He has been the technical manager of the Drinking Water Treatment Units Program for three years. His previous experience was in the area of analytical and environmental chemistry consulting. Andrew has a bachelor’s degree in chemistry and an MBA from the University of Michigan. He can be reached at 1-800-NSF-MARK or email:


Comments are closed.