Drinking water filters in pitchers and refrigerators reduce levels of perfluorinated compounds — PFAS — less effectively than reverse osmosis and two-stage filtration systems, according to a new study conducted by Duke University and NC State University scientists.
Teams lead by professors Heather Stapleton and Detlef Knappe compared the levels of contaminants in raw water and filtered water from 76 point-of-use filters, such as pitchers and in-fridge models. The scientists also tested water from another 13 under-sink reverse osmosis or whole-house systems. The homes were located in central and southeastern North Carolina.
The reverse osmosis and dual-stage systems were “very effective,” Stapleton told reporters today, “and removed at least 90%” of PFAS the scientists tested for. The pitcher, countertop, faucet-mounted and refrigerator filters, which use granulated activated carbon to remove the contaminants, decreased levels by only 50%.
“It’s unclear why,” Stapleton said. The filtration systems are proprietary, and the companies don’t have to release that internal information.
“The whole-house systems were also widely variable,” Stapleton said, “and in some cases actually increased PFAS levels in the water.”
The PFAS-removal efficiency of whole-house systems using activated carbon filters varied widely. In four of the six systems tested, PFSA and PFCA levels actually increased after filtration. Because the systems remove disinfectants used in city water treatment, they can also leave home pipes susceptible to bacterial growth.
“The under-sink reverse osmosis filter is the most efficient system for removing both the PFAS contaminants prevalent in central North Carolina, and the PFEAs, including GenX, found in Wilmington,” said Knappe of NC State. “Unfortunately, they also cost much more than other point-of-use filters. This raises concerns about environmental justice, since PFAS pollution affects more households that struggle financially than those that do not struggle.”
A 10-cup Brita pitcher with two filters runs about $30; replacement filters cost roughly $6 each. But the price of under-sink reverse osmosis systems ranges from $300 to $500, plus installation. Replacement filters run from $80 to $100.
In general, the frequency of filter replacements did not affect the removal or reduction levels. However, Stapleton said that at some homes, the filters did reach “the saturation point.” And in some of those cases, the filtered water contained higher levels of contaminants than the raw water. That could have occurred because the filter membrane essentially sloughed off the excess compounds into the water. It’s important for residents to replace the filters regularly.
“Home filters are really only a stopgap,” said Knappe, whose lab teamed with Stapleton’s to conduct the study. “The real goal should be control of PFAS contaminants at their source.”
The type of PFAS in the water also influenced the filters’ effectiveness. “Long-chain” compounds such as PFOA and PFOS were more thoroughly removed — 70% to 80% — than “short-chain,” which had reduction rates of 40%.
After 3M and DuPont phased out their manufacture of long-chain compounds, like C8 (so- called because they contain eight carbon atoms), the companies replaced them with short-chain compounds. Those contain four to six carbon atoms. GenX is a short-chain compound.