By Rick Andrew
NSF/ANSI 42 Drinking Water Treatment Units–Aesthetic Effects and NSF/ANSI 53 Drinking Water Treatment Units–Health Effectsare the relevant standards for testing contaminant reduction performance of POU filtration systems. Both of these standards require on/off cycling for this type of testing. This means that the flow of water to the test systems is not continuous, but rather consists of repeated periods of flow and no flow. On/off cycling is representative of actual usage by consumers, because each time they use a POU filtration system, they turn the faucet on to dispense drinking water and then turn it back off. So, from this perspective, it is consistent to require on/ off cycling when testing POU filters for contaminant reduction performance. On/off cycling also results in several different impacts on the performance of POU filters. These impacts are significant, since testing with on/off cycling can often provide very different results from testing with steady flows.
Activated carbon filter systems and chemical reduction
Chemical reduction performance of activated carbon systems is significantly affected by on/off cycling. The kinetics of the contaminant adsorption are such that contaminant reduction performance can be measurably better when there are periods of no flow of water during the off portion of on/off cycling. This difference in performance can be dramatic when testing reduction of certain organic contaminants—typically those that are relatively low molecular weight and relatively water soluble. In fact, this difference can be dramatic enough that when some activated carbon filter systems are tested with a 50/50 on/off cycle, meaning that in a 20-minute cycle there is flow for 10 minutes and no flow for 10 minutes, the result is a failure. When the same system is tested with a 10/90 on/off cycle, meaning that in a 20-minute cycle there is flow for two minutes and no flow for 18 minutes, the result is a pass. The additional time of no flow during the 20-minute 10/90 cycle allows adsorption to continue and contaminant reduction performance to be better for these contaminants. A logical extension of this observation is that if continuous flow were to be used instead of on/off cycling, the contaminant reduction performance of activated carbon systems for these organic contaminants would be poorer than when tested with a 50/50 on/off cycle.
Mechanical filtration systems
Although on/off cycling results in better contaminant reduction performance compared to continuous flow for activated carbon systems treating organic contaminants, the opposite effect is seen when testing mechanical filtration performance of POU filtration systems. In this case, the sudden increase in flowrate at the beginning of the on cycle causes significant stress on both the filtration media and the seals within the POU filtration system. This phenomenon is recognized and addressed in both NSF/ANSI 42 and NSF/ANSI 53. The standards are very specific that the on/off cycling valves must be fast opening, such as air- activated solenoids. Slow moving, ball-type valves do not provide a sufficient hydraulic stress to be a legitimate test. Additionally, the standards require that the tubing of the test apparatus downstream of the POU system must be minimized. The treated water samples must be collected based on calculating the water volume of this downstream tubing, letting it pass through, and immediately sampling, such that the first water filtered at the beginning of the on cycle is collected for analysis. It is this water that is most likely to have contaminants that have passed through either the media itself or the seals during the period of hydraulic stress at the beginning of the on cycle.
Further, mechanical filtration contaminant reduction testing is conducted until the filters have become clogged with particulate matter. The sudden onset of flow when the filter is clogged can cause even more hydraulic stress on the sealing mechanisms than when the filter is not clogged, resulting in the increased potential to have a blowout on the seals even as the filter cake has already established on the media. Given these factors, it is no surprise that on/off cycle testing produces very different results from steady-state, flow-type testing for mechanical filtration performance. Under steady-state flow, the repeated stress on the sealing mechanisms by multiple on cycles and initiation of flow repeatedly is avoided, and the filtration media can build up a significant filter cake, often improving filtration performance as the filter cake builds. Steady-state, flow-type testing is much gentler and does not tend to produce the instances of seal blowouts that on/off cycle testing can produce.
Consumer operation of POU filtration systems involves a large number of usage cycles, typically initiated from a state of no flow by rapidly opening a faucet valve fully. The fact that usage of these systems does not involve steady-state flow, but rather the hydraulic stress of repeated on/off usage cycles, is significant when considering their design and construction. This fact is also significant when considering appropriate methodology for testing these systems for contaminant reduction performance. A summary of the impacts, issues and approaches related to on/off cycling under NSF/ANSI 42 and NSF/ANSI 53 for testing POU filtration systems is included in Figure 1. As is typical of the NSF/ANSI Drinking Water Treatment Unit Standards for POU and POE systems, a first read of the test methodologies may lead to questions regarding why the testing is specified in such a manner. Hopefully, columns such as this one, describing the issues and rationale for on/off cycling, contribute to greater understanding and appreciation for this well-developed and thoughtful rationale.
About the author
Rick Andrew is the General Manager of NSF’s Drinking Water Treatment Units (POU/POE), ERS (Protocols) and Biosafety Cabinetry Programs. He has previously served as the Operations Manager, and prior to that, Technical Manager for the program. 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].