NSF/ANSI 58 and Reverse Osmosis Systems
NSF/ANSI 58 is the certification standard for reverse osmosis drinking water treatment systems. It addresses both health-effects-based contaminant reduction (the performance component) and structural/materials requirements (the safety component). RO systems certified under NSF/ANSI 58 have demonstrated reduction of a specified list of contaminants under defined test conditions — and RO is often promoted as a comprehensive PFAS removal solution by manufacturers and public health communicators alike.
The question is whether NSF/ANSI 58 PFAS certification, like NSF/ANSI 53, is built around PFOA and PFOS performance testing while leaving the broader compound spectrum uncharacterized. The answer is yes — with some nuances that reflect both the relative strength of RO membrane rejection for certain compound classes and the limitations of the specific compounds included in the test protocol.
RO Membrane Rejection Mechanics and Chain-Length Sensitivity
RO membranes remove solutes through a combination of size exclusion (the membrane’s semi-permeable barrier is physically smaller than most PFAS molecules) and Donnan exclusion (the negatively charged membrane surface repels anionic PFAS). For long-chain ionic PFAS, both mechanisms operate effectively: PFOA and PFOS, as large, highly charged anions, are rejected at >90–98% efficiency under standard operating conditions. This performance is genuine and represents a real advantage of RO over adsorption-based technologies for long-chain PFAS.
For short-chain PFAS, RO performance is better than GAC but still not comprehensive. Published studies report RO rejection of PFBS ranging from 75–92% and PFHxA rejection from 80–93%, compared to >97% for PFOA and PFOS. The declining rejection at shorter chain lengths reflects the reduced size exclusion effect (smaller molecules pass through membrane pores more readily) and reduced Donnan exclusion (less hydrophobic charge distribution). For TFA, as previously discussed, RO rejection drops to 50–80% depending on membrane type.
NSF/ANSI 58 PFAS Test Protocol: What Is and Is Not Included
NSF/ANSI 58 PFAS reduction testing, like NSF/ANSI 53, requires demonstration of PFOA and PFOS reduction. The standard does not require testing against PFBS, PFHxA, GenX, or other short-chain and replacement PFAS. A product certified under NSF/ANSI 58 for PFAS reduction has demonstrated adequate removal of PFOA and PFOS; its performance against other PFAS compounds is neither required, tested, nor reported under the certification.
This is a significant gap for consumers and institutions relying on RO systems as a comprehensive PFAS treatment solution. While RO’s inherent rejection mechanism is somewhat more chain-length-agnostic than GAC adsorption, ‘somewhat better than GAC for short-chain PFAS’ is not the same as ‘comprehensively effective across the PFAS spectrum.’ An 80% rejection of PFBS still allows 20% of the PFBS present to pass through — and at 10 ppt influent concentration, that is 2 ppt of PFBS in the treated water, approaching the EPA hazard index threshold.
Brine Management and the Concentration Factor
RO systems address one contamination problem while creating another: the reject stream (brine) concentrates all rejected compounds, including PFAS, by factors that typically range from 3× to 10× depending on system recovery rate. For PFOA and PFOS, a system achieving 97% rejection and operating at 25% recovery produces a brine stream with a PFAS concentration approximately four times higher than the feed water.
Brine disposal is a PFAS treatment design parameter that receives less attention than rejection efficiency but has significant environmental implications. Point-of-use RO systems typically discharge brine to the drain; this does not destroy the PFAS but transfers it to the sewer system and ultimately to wastewater treatment plants, which are generally ineffective at destroying or removing PFAS. Industrial RO systems require active brine management strategies — evaporation, further concentration, thermal destruction, or secure landfill — that add cost and complexity to the overall PFAS treatment system. These downstream implications are absent from the NSF/ANSI 58 certification framework.