THE WHOLE PLANT IMPERATIVE • PART 9
Every deferral in this series was defensible on its own. That is precisely how the number got this large.
Nobody in this series made a bad decision. That is the part worth sitting with.
The condenser cleaning slipped because the unit was running and the megawatts weren’t visible. The catalyst trend review slipped because nothing was out of limits yet. The coating survey slipped because the outage was already oversubscribed with work that was scoped and funded. Each of those was a reasonable call by a competent person under a real constraint. Every one of them would survive a review in isolation.
Which is the actual problem. Maintenance deferral is evaluated one line at a time, and it is never wrong one line at a time. It is only wrong in aggregate — and almost nobody assembles the aggregate.
The number nobody totals
Put this series end to end and the arithmetic changes character. A single 300 MW unit carries roughly $270,000 to $324,000 of lost summer generation from cooling degradation, against an annual program of $80,000 to $150,000. Reactive catalyst management runs 20 to 30 percent above lifecycle cost, with a compliance exposure attached that has no clean dollar figure. Structural corrosion escalates 20 to 50 times on the same defect, decided entirely by when you get to it. Inlet degradation bills through the turbine, which means it has been showing up in the OEM line item for years without ever being named.
None of those sit in the same budget. Cooling is operations. Catalyst is environmental. Coatings is capital. The inlet is turbine maintenance. The plant has no single line where whole-plant deferral is visible, so no one is ever looking at a number that would change a decision — they are looking at four defensible ones.
This is why preventive programs return three to eight times their cost and still lose funding arguments. The return is real. The problem is that it is distributed across cost centers, while the savings from deferral land in one place, this quarter, with somebody’s name on it.
Deferral wins because it is measured. The cost of deferral wins nothing because nobody measures it.
What to bring to the table instead
The case that works is not a longer maintenance list. It is a reframing of what the plant actually is.
Start with the asset life question, because it has quietly changed. Units that were expected to retire are now expected to run for decades — capacity obligations, load growth, and the shortage of dispatchable generation have pushed the horizon out well past what most maintenance budgets were built around. A plant running to 2065 does not make the coating decision once. It makes it forty more times, and at 20 to 50 times escalation, the compounding is not a rounding error.
Then change what gets measured. Availability is a lagging indicator and a poor one — the entire cooling argument is a system that costs you millions while availability reads perfect. Heat rate trend, corrected output against ambient, catalyst conversion efficiency, CEMS data availability, and documented coating condition are leading indicators, and they are the numbers that let somebody act before the expensive version arrives.
And bring the deferral cost as a curve rather than a line. Finance understands escalation. A spreadsheet showing the same defect at $12,000 today, $60,000 in three years, and $350,000 at failure is not a maintenance request. It is a capital argument, and it gets a different hearing.
The condition that makes it possible
There is a practical obstacle worth naming. The whole-plant view is hard to assemble when every system is inspected by a different specialist on a different schedule, each reporting into a different budget, none of them producing a document that talks to the others.
Fragmented vendor coverage is not just a procurement inefficiency. It is the reason the aggregate number doesn’t exist. Nobody is positioned to see the plant whole, so nobody produces the one document that would win the argument.
That is the case. Not that the plant needs more maintenance — that it needs the total to be visible to the people who decide.
The final piece in this series is about what to do with that on Monday.
For more information on how Groome can help make a difference, visit our Gas Turbine Plant Performance Solutions page.
Read Part 10 → WHERE TO START
Start from the beginning → YOUR TURBINE IS FINE. THE ASSETS AROUND IT MAY NOT BE.
