Vendor literature for PFAS treatment is full of removal figures, and many of them cannot be compared with one another. A single number may describe a different influent, a different compound list, a different reporting limit, and a different run length than the number beside it. Before any pilot is scored, the buyer needs to know which conditions produced the result. A federal effort to coordinate technology work does not remove that problem by itself, but it gives utilities and dealers a reason to ask for a common basis.
The occasion for this piece is the launch described by Water Finance and Management, which reports that the EPA introduced its first agency-wide PFAS Technology, Evaluation, Coordination, and Harmonization strategy. The outlet's summary says the new effort "will help springboard technology innovation on PFAS," a phrase that describes intent, not performance. The summary does not describe the evaluation protocol, the test matrices, or how results would be published, so none of those points can be assumed here.
Why a Removal Figure Needs Its Conditions
A removal result is a ratio between an influent measurement and an effluent measurement. Each side depends on how the sample was collected, how it was held, and which laboratory method reported it. If the effluent sample came back below the reporting limit, the calculated removal is bounded by that limit, not by a true value of zero. Two vendors who both report strong removal may have run pilots where the effluent floor sat at very different levels, and their headline figures will look alike. The compound list matters as much. A report covering a short target list says less about total PFAS load than one covering a wider set, and a vendor may quote the favorable subset without saying so.
Run length and media age change the picture too. Adsorbent media loses capacity as it loads, so early pilot data can overstate what a unit delivers near the end of its service interval. Breakthrough behavior often differs between short-chain and long-chain compounds, which means an average across a mixed list can hide a weak spot. None of this makes a vendor dishonest. It means the figure is a statement about one configuration, on one water, under one set of dates.
What a Common Basis Has to Fix
A pilot file that can be compared across vendors needs fixed fields. Record the raw water source and any pretreatment, the full analyte list with the method used, the reporting limit for each compound on each side of the unit, the sampling points and holding times, the flow and loading rate, the media lot and its age, and the run length measured in service time rather than calendar time. Report non-detects as non-detects with their limits. Do not average them as zeros or as half the limit unless the protocol says so. Keep the vendor's own claims in a separate column from measured results so the two cannot blur.
The same fields serve a utility's board packet and a household's file. A dealer who sells a point-of-use unit and a vendor who bids on a plant addition answer the same question with different tolerances for error. Installers and service firms, including independent regional dealers such as Jones Air & Water, hold service records that reflect real household use. If those records use the same fields as pilot data, they become comparable evidence rather than a separate set of anecdotes.
What the Federal Strategy Does Not Settle
A federal technology strategy sets a direction for research, evaluation, and coordination. The summary does not indicate that the strategy sets a treatment standard, a compliance schedule, or an approved list of units for any water system. Those obligations still come from the rule and permit record, and a state order or permit still governs what a unit must achieve at a given site. Read the strategy as a source of future protocol, not as present authority. Until a protocol is published and a regulator adopts or references it, a pilot result is evidence only under the terms the pilot used.
Writing the Pilot File
Start the file before the vendor arrives. Name the compounds the utility or dealer needs to see, the reporting limits the laboratory must reach, and the conditions under which a result will count. Ask each vendor to report against that template, including runs that failed or were stopped early. Mark which figures came from the vendor and which came from an independent laboratory. A comparison is only as sound as its weakest field, and the weakest field is usually the one nobody wrote down.