A utility evaluating a treatment technology often relies on a demonstration. A pilot skid runs at the plant for a season, the vendor presents a summary of removal rates or fouling behavior, and the results go into a specification. The demonstration is useful evidence. It is not, by itself, a performance commitment, and the gap between those two things is where many procurement disputes begin.
The question is worth raising now because of a leadership change at a nonprofit innovation body. Water Finance and Management reported that Melissa Meeker, chief executive officer of The Water Tower, "recently announced her planned departure," and that the Buford, Georgia, water innovation center will begin a search for new leadership. The report does not describe the center's programs in detail, so this piece makes no claim about any specific pilot or result. The general point holds regardless: when an organization that produces or hosts demonstration evidence changes leadership, buyers should confirm who stands behind the data they intend to use.
What a demonstration can and cannot establish
A pilot records what happened under one set of conditions. Those conditions usually include one influent range, one operating crew, one membrane or media lot, one chemical dose window, and one sampling schedule. The result is a bounded observation. The specification writer needs to know which of those bounds carry over to the full-scale design and which do not. A removal figure reported at a moderate turbidity range says little about performance during a storm event, and a fouling trend observed over a few weeks may not show the seasonal pattern the full plant will see.
The practical test is to rewrite each claimed result as a condition statement. For example, instead of accepting "the unit met the target," the file should read "the unit met the target at the stated influent range, at the stated flow, for the stated period, with the stated analytical method." Any result that cannot be written that way is a claim, not a finding.
Warranty language is a separate record
Performance guarantees usually name a measured parameter, a sampling location, a sampling method, a time window, and a remedy. Pilot summaries rarely match those elements exactly. The reported value may be an average where the guarantee requires a maximum, or the sample may have been taken at a point that differs from the regulatory compliance point.
Ask whether the guaranteed value traces to the demonstration data, whether the guarantee depends on influent conditions that the utility can actually control, and who carries responsibility if the vendor's measurement partner or program sponsor changes. None of these questions assumes bad faith. They are the ordinary terms that determine whether a guarantee can be enforced when the operating record is disputed.
Check the chain of custody for the data
When a demonstration is run or reported by a third party, such as a nonprofit hub, a university group, or a consortium, the utility should know who collected the samples, who analyzed them, who holds the raw files, and who will answer questions years later. Raw laboratory reports, calibration records, and chain-of-custody forms should be requested at the time of the contract, not after a dispute begins.
A change in an organization's leadership does not, by itself, invalidate past data. It can, however, change who is available to explain it, who has authority to release the raw files, and whether the organization's archive is maintained on the same schedule as its public reports. Those are administrative facts, and they are easy to overlook until a regulator or a board asks for the underlying numbers.
Write the contract so the evidence survives staff turnover
Three provisions help. First, the contract should attach the demonstration report and the raw data as exhibits, so the guaranteed performance is tied to a fixed record. Second, it should name the measurement method and the lab or instrument standard that governs compliance testing. Third, it should specify a notice and access clause that requires the vendor, and any third-party partner, to preserve and release the underlying data for a stated period after the warranty ends.
Utility staff turnover matters here as well. The person who negotiated the pilot may not be the person who operates the plant when the guarantee is tested. A short internal memo that records the assumptions behind the specification, the conditions under which the demonstration ran, and the open questions at signing will protect the record far better than a general reference to the vendor's report.
Keep the demonstration in its place
A pilot is a reasonable basis for choosing a technology, but it is not a substitute for a warranty, a verified sampling plan, or a clear chain of custody. The Water Tower news is a reminder that the organizations producing evidence can change. The utility's record should be built so that the evidence, the guarantee, and the explanation of both remain intact when the people who first described them have moved on.