A machine that is unavailable for an hour has lost an hour of potential operation. That statement is clear, but it does not tell a wind-business leader everything needed to judge the consequence. The hour occurs at a particular point in a resource pattern, under particular operating and grid circumstances. Two identical totals of unavailable hours may therefore describe materially different service outcomes. An annual availability measure can remain useful while being insufficient for the decision beside it. The difficulty begins when its apparent precision makes the missing context seem less important than the number that is easiest to compare.
The hour has a place in the year
I made that mistake as chief operating officer in 2016. A common maintenance calendar simplified staffing and our commercial account. Safiya asked whether the local wind pattern justified the same window at every site. I accepted the argument that the annual hours measure gave us a sufficiently coherent basis for the offer. At a six-machine customer cluster, planned shutdowns then occupied fourteen days of a strong wind period. The point is not that I lacked access to a more complicated technical report. I had heard a relevant objection and treated it as detail to be handled after the main approval.
The subsequent review estimated 1.8 GWh of foregone generation against an operating counterfactual. That estimate required assumptions about what the machines would have produced had they been available. It was not a second meter reading from a parallel version of the same fortnight. We also accepted ₹1.4 crore in contractual customer credits. Those credits were determined by the commercial arrangements, not by multiplying the energy estimate by an assumed tariff. Keeping the two figures distinct matters. A financial consequence can be real without becoming a measurement of every physical effect the review is trying to understand.
The first temptation after an error like this is to replace the old measure with a more sophisticated one and declare the problem solved. I think that response can repeat the original mistake in a more impressive form. A resource-weighted account has its own assumptions and may be useful for one purpose without settling another. A customer needs an understandable commitment. An operating team needs a workable schedule. Engineers and safety owners need the conditions required for the work. Finance needs a defensible account of consequences. Leadership has to make those accounts meet; it cannot choose one measure and require every participant to express the whole problem through it.
Planning does not remove the weather
There is also a credible argument for standardisation. A company serving many sites cannot reinvent every planning process for every machine. Common methods can make responsibilities clearer, allow scarce specialists to be organised and expose exceptions that would otherwise remain hidden. My error was not wanting a repeatable process. It was allowing the common calendar to become the conclusion before we had examined the exception Safiya raised. A sound standard should tell us when additional site-specific consideration is necessary. It should not make a local difference disappear simply because acknowledging it complicates the service proposal.
We funded ₹1.8 crore of revised scheduling and service coverage. The work did not give us authority to conduct maintenance whenever a revenue forecast preferred it. Safe working conditions, technical requirements and the accountable owners' decisions remained essential. Nor could a revised schedule make wind predictable. The practical change was to examine resource periods, customer commitments and operating constraints together before fixing the calendar. Where those accounts conflicted, the conflict had to reach someone able to make the commercial and organisational choice. It could no longer be left to the site team after a simple promise had already been made.
I would resist describing that change as proof that a specified amount of energy was recovered. We did not run a controlled comparison in which the only difference was the calendar. Weather, equipment condition and the work itself can vary between periods. The reliable account is narrower: the company changed its planning and approval process, paid credits and funded additional coverage. Later performance requires its own evidence. A leader does not strengthen a case by adding an outcome merely because it is the outcome the correction was intended to improve. The intention can be sensible while the eventual effect remains difficult to isolate.
What the board should receive
For a board, the useful question is not whether directors should redesign maintenance schedules. They should not need to become the operating team to understand what the company has promised. They can ask which measure carries the main service account, what consequential differences sit outside it and how management handles those differences before committing. They can also ask whether a favourable operating metric is being used to displace a less favourable customer or energy result. That line of inquiry respects specialist responsibility while testing the completeness of the executive explanation.
The language used after a shortfall is especially revealing. Saying that the annual availability figure was achieved may answer a contractual or operational question. It does not automatically answer whether the timing served the customer's need well. Equally, an unfavourable energy outcome does not establish that every operational choice was wrong. The board needs both the measured result and a bounded explanation of its relationship to management decisions. I want to distinguish what the evidence supports, what depends on a counterfactual and what remains unknown. Those distinctions are not excuses for inaction; they help identify which next action can responsibly follow.
The lesson has changed how I listen to a concise portfolio account. When a colleague raises a site condition that complicates the proposed standard, I ask what consequence the complication represents before asking how quickly it can be removed from the presentation. Sometimes it is genuinely immaterial. Sometimes it is the fact that changes the promise. In 2016 I chose convenience too early, and the customer consequence made that choice visible. A mature operating business needs measures it can use consistently, but it also needs leaders willing to recognise when consistency has become a way of declining to hear the particular problem in front of them.