A wind plant's compliance files, audited claim by claim

A utility-scale wind plant, in service for about five years in a market with its own ride-through rules, had 45 compliance claims in its owner's registers, nearly all marked complete. Its folders held more than 400 compliance documents: protection studies, model verification reports, facility ratings, training certificates, procedures and test forms.

It was one of the better-documented plants in the fleet gap analysis of about 70 plants. That is what made it the right plant for a deep dive.

For every standard the plant claims, each requirement needs a dated document that states the required result for this plant, is inside its re-verification cycle, and has no later settings or equipment change on record. A corporate procedure that never names the plant, a template, a draft or an undated page does not count.

We read all of the plant's roughly 430 compliance documents against every requirement of every standard it claimed: 41 due at the time of the review, four not yet due. For each claim we recorded a verdict, the reason, the document, the page and the date, about 700 citations for this one plant, and listed the fix for every gap.

Five claims were proven, about one in eight: a facility rating sheet, a protection coordination study, a relay loadability study, relay setting philosophies and a communications training certificate, each dated and naming the plant.

Five more were sound on their own date but due for re-verification: a corporate cyber security policy on a 15-month review cycle, last reviewed six years earlier; a generator capability verification whose five-year cycle had lapsed two weeks before the review; two model verification reports superseded by a plant change recorded in 2023; and an annual model data submittal dated 2019.

Twenty-eight claims fell short: a maintenance program summary filed without the maintenance records behind it; a template filed as evidence; a corporate event-reporting procedure that never names the plant; a ride-through test report that answers one of the three items the rule asks for; an undated document; and a dozen standards where the folder held related material but no page stated the required result.

Three claims had nothing on file at all.

The plant's deep dive shipped inside the fleet review, two days after approval. Each row tells the plant team exactly what to file and where: the in-cycle version of five studies and submittals, dates and completed records for the 28 that fell short, and a confirm-or-accept decision for the three with nothing on file.

The plant was not badly run. Its studies were real and its consultants were competent. The gaps were in dating, cycles, source records and plant-specific statements, the things an auditor reads for. The same remedies apply across the fleet.