THE WHOLE PLANT IMPERATIVE • PART 7
Two systems that almost never fail — and can keep a plant off the grid for a year when they do.
Every maintenance program in the industry is built on frequency. What breaks often gets attention, a spare on the shelf, and a line in the budget. What almost never breaks gets a walk-down and a good record. The logic is sound and the plants that run it are usually well run.
It has one blind spot, and it is the expensive one. Frequency tells you how often you will be interrupted. It tells you nothing about how long the interruption lasts. There are two systems on a combined-cycle plant where those two numbers point in opposite directions — where the odds in any given year are genuinely low, and the recovery, when it comes, is measured in seasons.
The failure that arrives with the fuel
Pipeline gas is not a constant. Composition shifts with the source, the season, and the pressure at the interconnect, and what leaves the supplier as dry gas does not always arrive that way. When heavier hydrocarbons drop out of the stream, the plant’s fuel conditioning equipment is the only thing standing between that liquid and the combustion system.
Liquid carryover into a DLN combustor is the highest-consequence failure mode in the fuel system, and it is fast. Modern lean premix combustors are built to burn a gas with a known heating value inside a narrow window. Introduce a slug of liquid and the local energy release goes far outside anything the design contemplated. The damage is not to a filter or a valve. It is to combustion hardware and the hot section behind it, and it happens in seconds, across the machine, before any operator can respond.
Which is why the equipment that prevents it deserves more attention than its cost suggests. Coalescing filters and separators with a maintained element schedule. Heaters that actually hold superheat margin above the hydrocarbon dew point at winter conditions, not just at design conditions. Drain systems that are proven to work rather than assumed to. And the safety and shutdown functions — double block and bleed integrity, vent path verification, leak detection — that only earn their keep on the day something has already gone wrong.
Fuel conditioning is a small line item protecting the largest single asset on site. That ratio is the whole argument.
The relevant question was never how often it fails. It’s how long you’re down when it does.
Where the market stops being the constraint
The electrical side of the plant is the clearest version of the same problem. A main power transformer or the switchgear that connects it does not fail often. When it does, the plant is not waiting on a repair crew. It is waiting on manufacturing.
Replacement lead times on large power transformers and associated equipment have run 12 to 18 months and longer, driven by global demand for the same units, constrained production capacity, and a queue that does not care about your capacity obligations. There is no expediting a position in that line and no premium that reliably buys one. Every other failure in this series can be solved by spending money faster. This one cannot.
What makes that acceptable is that these assets are among the most diagnosable on the plant. Dissolved gas analysis reads the fault chemistry inside a transformer long before anything external is visible, and the trend matters more than any single sample. Infrared thermography finds developing connection problems while they are still connections rather than failures. Switchgear testing — insulation resistance, contact resistance, protective relay verification, breaker timing — confirms that equipment sitting idle for years will still operate on the one occasion it is asked to.
None of that is exotic and none of it is expensive relative to what it protects. It is simply work with no visible payoff, performed on equipment that appears to be fine, which is the definition of the maintenance that gets deferred first.
A different kind of math
Deferral on most systems is a bet you can lose gracefully. Skip the condenser cleaning and you give back megawatts you can quantify and recover next season. Skip the catalyst trending and you eventually get a letter. Both are bad outcomes with a floor under them.
Fuel gas and electrical have no floor. The probability is genuinely low — low enough that a plant manager who defers this work will most likely be right, and right again the year after. But the consequence isn’t an outage measured in days. It is a plant that cannot produce, cannot meet its capacity commitments, and cannot buy its way back to service on any schedule it controls.
These systems don’t reward attention. They just quietly remove the outcome you cannot survive.
Next in the series: the systems that sit at the bottom of every priority list — and carry the steepest cost curve on the plant.
For more information on how Groome can help make a difference, visit our Gas Turbine Plant Performance Solutions page.
Read Part 8 → COATINGS, INSULATION, AND STRUCTURE
Start from the beginning → YOUR TURBINE IS FINE. THE ASSETS AROUND IT MAY NOT BE
