Every plant has the same argument eventually. Header pressure drops during the soot blowing sequence, the operators notice, somebody writes a work order, and the answer that comes back is that the plant needs another compressor.
Sometimes that is true. More often the plant already has enough compressor and has been running it against a demand curve that sits flat most of the day and then jumps for eight minutes. Buying horsepower to cover eight minutes an hour is an expensive way to solve what is usually a storage problem.
This piece is about the service air side specifically. For the whole picture, including where instrument air fits and why it gets a tighter spec, start with our guide to compressed air for power plants.
What Makes Service Air Different
Instrument air on a running unit is boring in the best way. Actuators hold position, positioners bleed a little, and the flow barely moves all shift. That is why it is straightforward to size and straightforward to keep clean.
Service air is the opposite. On the service side you are feeding soot blowers, ash handling and pneumatic conveying, baghouse and precipitator pulse cleaning, breaker operation, hose stations on the turbine deck, and the maintenance shop. Almost all of it is cyclic, and the cycles are large.
Two things follow from that. First, the average flow number on your meter is close to useless for sizing. Second, the quality spec is genuinely lower than instrument air, and treating the whole plant to instrument quality so that soot blower air is clean is a permanent waste. A refrigerated dryer and ordinary particulate filtration is enough for most service loads.
The Loads, and How They Behave
| Load | Demand shape | What it does to the header |
|---|---|---|
| Soot blowing sequence | Large, sustained for minutes, repeating on a schedule | The biggest single event on most station air systems |
| Ash conveying cycle | Large, cyclic, often overlapping with other loads | Long enough that storage alone will not always cover it |
| Baghouse or precipitator pulse | Very short, very sharp, repeating constantly | Classic storage problem, almost never a capacity problem |
| Breaker and damper operation | Brief, occasional | Minor, but happens at inconvenient moments |
| Shop tools and hose stations | Intermittent, unpredictable | Small in total, annoying because it is unscheduled |
The distinction that matters is duration. A baghouse pulse is milliseconds, so a receiver near it absorbs the whole thing and the compressor never knows it happened. A soot blowing sequence runs for minutes, which is long enough that storage buys you the ride-in but the compressors still have to carry most of it. Ash conveying sits in between and depends entirely on your cycle.
Size storage for the short sharp events. Size compressor capacity for the long ones. Getting those backwards is how plants end up with both too much horsepower and not enough pressure.
Sizing the Ride-Through
The work is mostly measurement, and it is worth doing properly because the alternative is a six-figure guess.
- Log flow and pressure through a full cycle. Not an average over a week. You want the trace through one complete soot blowing sequence and one complete ash conveying cycle, at the resolution to see the ramp. A flow meter with logging is the right tool.
- Find the real minimum pressure. Measure at the most pressure-sensitive device on the service header while it is working, not at the compressor room gauge. That number sets your floor.
- Work out the deficit. Peak demand during the event, minus what the compressors can deliver, times the duration. That is the volume of air storage has to cover, and it is the number that tells you whether this is a storage fix or a capacity fix.
- Convert to receiver volume. The usable air in a receiver is set by the pressure band you can work across, not by the tank size alone. If the compressors hold 120 psi and the header needs 90, you have a 30 psi band to spend. If you are running the tank at 100 and need 90, you have 10, and the same tank gives you a third as much.
- Put the storage near the load. A receiver in the compressor room does not help a soot blower four hundred feet away if the header cannot deliver the flow. A dry receiver local to the event is worth several times the same volume sitting at the source.
That fourth point is where a pressure flow controller earns its place on a station system. Holding the header steady at a lower setpoint while the receivers run higher is what turns tank volume into usable air, and it is usually cheaper than the compressor somebody wanted to buy. Browse air receiver tanks by capacity.
Redundancy Cuts Differently Here
On instrument air, redundancy is about not losing the unit. On service air, it is about not losing the cleaning medium.
If soot blowing air goes away, you do not trip immediately, but you lose your ability to keep the boiler surfaces clean, and that shows up as falling efficiency and eventually as a forced derate. That is why plants running compressed air soot blowing carry redundant compressors, and it is a real cost of the method rather than an optional extra.
Practical arrangement on the service side: enough installed capacity that you can lose one machine and still run the sequence, a tie to the instrument air system with a check valve so service air can back up instruments in an emergency but never the other way, and local storage that covers the gap while a standby machine starts and loads.
Where We Fit
Straight answer, because station scale varies enormously.
Most of the service air side is squarely in our lane. Rotary screw packages, refrigerated and desiccant dryers, filtration, receivers, condensate handling, piping and the parts to keep it running. That covers the great majority of balance of plant work at gas turbine peakers, biomass plants, hydro stations and the maintenance side of any station.
What we do not supply is the very large, very high pressure end of dedicated soot blowing air. Utility scale soot blowing has historically been served by centrifugal machines in the 300 to 500 psig range at flows in the thousands of scfm, and that is purpose-built equipment from a different class of supplier. If your soot blowing air is a dedicated high pressure centrifugal train, we are not the source for the machine itself. We can still supply the treatment, storage, condensate management and consumables around it, and the plant and service air that runs everything else.
Same answer as before on safety-related nuclear procurement. That carries 10 CFR 50 Appendix B and NQA-1 pedigree requirements and a commercial grade dedication process we are not set up for. Non-safety balance of plant air at the same station is a normal commercial purchase.
Things That Bite Plants on the Service Side
Treating service air to instrument quality. Drying the soot blower leg to minus 40F because the instrument header needs it is a permanent energy cost for no benefit. Split the systems and treat each to its own spec.
Storage in the wrong place. Volume at the compressor does not solve a pressure problem at the far end of a long header. Measure where the sag actually is.
Ash in the compressor intake. On coal and biomass units the intake is often in genuinely bad air. A loaded inlet filter raises discharge temperature and pushes oil carryover up, which quietly moves you out of spec. Duct the intake somewhere clean and change elements on differential rather than on the calendar.
Condensate nobody tracks. A station produces a lot of it, and at a plant with an NPDES permit an oily discharge is a compliance issue rather than housekeeping. Run it through an oil water separator and keep records.
Assuming the sequence has not changed. Soot blowing schedules get adjusted, ash cycles get retuned, and the air system was sized against the old numbers. If the complaint is new, check whether the demand is new.
Frequently Asked Questions
How much storage do I need for a soot blowing sequence?
Work from the measured deficit rather than a rule of thumb. Log flow and pressure through one complete sequence, subtract what the compressors deliver from what the sequence draws, multiply by duration, and that volume is what storage has to cover. Then remember that usable air depends on the pressure band you can work across, so a wide differential between receiver pressure and header setpoint makes the same tank go much further.
Is it cheaper to add storage or add a compressor?
For short sharp events like baghouse pulse cleaning, storage wins easily and it is not close. For long events like a full soot blowing sequence or a sustained ash conveying cycle, storage buys you the ramp but the compressors still have to carry the load, so you may genuinely need capacity. The measurement tells you which one you are looking at.
Can service air and instrument air share compressors?
They can share a supply, but they should not share a treatment spec. The usual arrangement is common or partly common compression with separate treatment trains, so instrument air gets desiccant drying and fine filtration while service air runs on a refrigerated dryer. Put a check valve on any tie so service air can back up the instrument header in an emergency and never the reverse.
Does ash handling air need to be dry?
Drier than people assume. Moisture in a pneumatic conveying line mixed with fly ash makes a paste, and that is how conveying lines plug. You do not need instrument-grade dew point, but a properly sized refrigerated dryer and a working condensate drain on that leg prevent a very unpleasant class of problem.
Why does header pressure sag even though the compressors are not fully loaded?
That pattern points at distribution rather than supply. Either the header cannot pass the flow to where the event is, or the storage is all sitting in the compressor room, or a filter on that leg has not been changed and is eating the differential. Measure pressure at the load during the event and compare it to the compressor room gauge; the gap tells you where to look.
The Short Version
Log a full cycle before you buy anything. Put storage near the events rather than at the source, widen the pressure band so that storage is actually usable, and treat service air to service air quality instead of instrument quality. Most of the time the complaint about soot blowing and ash handling air is a distribution and storage problem wearing a capacity problem costume.
