THE WHOLE PLANT IMPERATIVE • PART 6
The only system on the plant whose failure gets blamed on something else — usually on the most expensive asset you own.
Output is off about two percent. Heat rate has crept. Nobody can point to an event, so the water washes get scheduled a little closer together, somebody notes that the machine has some hours on it, and the degradation goes into the file as compressor fouling. Which it is. The question nobody asks is why the compressor keeps fouling.
The inlet sits upstream of everything. Every megawatt the plant produces begins as air pulled through that house, and a large frame machine pulls it by the hundreds of thousands of cubic feet per minute. Whatever is in that air goes into the compressor. Which means the inlet is the one system on the plant whose failure never announces itself under its own name. It shows up as somebody else’s problem — and the somebody else is usually the turbine.
Filtration you own and don’t have
Filters are a solved problem. Plants buy good ones, track differential pressure, and change them on schedule. The failure mode is almost never the media.
It is the path around it. Inlet expansion joints, housing seams, and access doors sit in a structure that heats, cools, and vibrates every day of its life, and the plant now cycles far more than the design basis assumed. Fasteners loosen. Sealing surfaces relax. A joint that has quietly opened is drawing unfiltered air directly into a compressor that has no idea it is being bypassed — and every hour it stays open, you are paying for filtration you own and are not receiving.
What arrives is not dust in the general sense. It is silica and salts that stick to airfoils and cost you compressor efficiency; harder particulate that erodes leading edges; and in coastal or industrial settings, chlorides that corrode. Fouling is recoverable with a wash. Erosion and corrosion are not. They come out of the hot section at overhaul, and by then the invoice is written against the turbine — not against the seam.
The turbine wasn’t degrading. It was being fed.
The used filter is a report nobody reads
Most plants treat a spent filter as waste, which is understandable and expensive. That filter is a physical record of everything the plant ingested since the last change — the only direct evidence available about what the compressor has actually been breathing.
Sent to a lab, it answers questions that instrumentation cannot. What is the particulate, and where did it come from — traffic, agriculture, a neighboring industrial source, salt, or the cooling tower drift returning to the intake? Is the loading even across the bank, or concentrated in a pattern that indicates a bypass path rather than normal service? Is the class of filter installed the right one for the site as it exists now, or the one specified when the site looked different?
The answers change what you buy and where you look. A plant that discovers its loading is chloride-heavy has learned something no differential pressure reading would ever have told it, and it has learned it while the damage is still cheap. Filter forensics is one of the few diagnostics in this series that costs almost nothing and is skipped almost universally.
June 2023, and the intake nobody rechecked
When Canadian wildfire smoke settled over the Northeast and Midwest in June 2023, the fleet kept running. Air quality indices hit numbers most of those states had never recorded. Plants stayed online — the load was there, and the machines were fine.
Inlet filtration across that whole region absorbed a particulate event nothing in its service history had prepared it for, in the space of a few days. The systems that were watching differential pressure saw it. The systems that were on a calendar did not. And the ones that ran filters to their scheduled change date afterward were, in effect, operating on filtration that had already done its life’s work.
The lesson generalizes past smoke. A construction project upwind, a nearby facility’s upset, harvest season, a coastal storm driving salt inland — the intake experiences all of it, and none of it appears on a maintenance schedule. The discipline is simple and almost nobody has it written down: after any regional environmental event, the inlet gets inspected. Not at the next scheduled interval. Then.
Why this one hides so well
Every other system in this series eventually declares itself. The HRSG leaks. The condenser gives back megawatts you can measure. The catalyst produces a number a regulator reads. The inlet does none of that. It has no permit attached to it, no event code, and no failure of its own to report — which is exactly how it gets away with costing so much.
It bills through a different asset. The loss lands as turbine degradation, gets absorbed into the OEM conversation, and confirms the belief this series started by challenging: that the turbine is where the risk lives. Some of it does. Some of it is being handed to the turbine by a joint upstream that costs a fraction of a day’s revenue to reseal.
Next in the series: two systems that fail almost never — and cannot be recovered quickly when they do.
For more information on how Groome can help make a difference, visit our Gas Turbine Plant Performance Solutions page.
Start from the beginning → YOUR TURBINE IS FINE. THE ASSETS AROUND IT MAY NOT BE.
