Episode 76: Air Compressor Control: Matching Output to Demand

Jason Reed

[matter-of-fact] Welcome to the show. Lisa, I’ve seen plants spend real money chasing the wrong problem, because the compressor looked fine, the pressure gauge looked fine, and meanwhile that machine was burning power and beating itself up just trying to keep up with a demand pattern nobody had actually measured.

Lisa Saunders

[curious] And that phrase right there -- “looked fine” -- that’s the trap, right? Because if the header pressure still lands where people want it, they assume the control strategy must be working. But behind that, you can have short cycling, pressure swings, and a motor starting way too often.

Jason Reed

[reflective] Exactly. Air compressor capacity control is not just on/off logic. It’s output matched to demand. That’s the job. And when that match is bad, the costs show up in ugly places: wasted energy, unstable pressure, overheated motors, moisture from equipment not staying hot long enough, and parts wearing out before they should.

Lisa Saunders

[questioning tone] So when somebody says, “Just tell me the BEST control method,” your answer is... what, “that depends”? Which I know people hate hearing. [laughs]

Jason Reed

[chuckles] They do hate it. But yeah -- it depends on the job. If you’re on a drilling rig, for example, simple and reliable may matter more than squeezing every last kilowatt-hour out of the machine. In a plant with variable demand and high power costs, energy efficiency might be the whole game.

Lisa Saunders

[reflective] That drilling rig example is useful, because it gets at the real tension. Reliability versus efficiency. Not always, but often. People want one answer when the honest answer is more like, “What failure hurts you more -- wasted power or unstable operation?”

Jason Reed

[firm] Right. And I’d add one more thing. Bad controls can hide behind “good enough” pressure for a long time. A compressor can still hit pressure while idling inefficiently, cycling too fast, or running in a control mode that just does not fit the demand profile. So the machine looks obedient... while the electric bill tells the truth.

Lisa Saunders

[energized] Okay, let’s do the practical walk-through. If somebody’s standing in front of a compressor room and trying to make sense of the options, start with the simplest one.

Jason Reed

Start/stop. Dead simple. The pressure switch says go, it starts. Pressure reaches the setpoint, it stops. That works for low duty cycles. Small reciprocating compressors, classic case. But if that unit is starting and stopping constantly, that’s short cycling -- or rapid cycling -- and now you’re heating up the motor, increasing wear, and setting yourself up for moisture problems because the airend never stays hot long enough.

Lisa Saunders

[skeptical] So “simple” can turn into expensive pretty fast. If it’s banging on and off all day, that simplicity is fake.

Jason Reed

That’s exactly it. Then you’ve got load/no load. Three states: loaded, unloaded, off. In the unloaded state, the inlet valve is closed, but on a fixed-speed unit the motor is still spinning at full speed. It’s not making air. It’s just idling.

Lisa Saunders

And the number people need to remember there is 40%. An idling compressor can draw up to 40% of the power it uses at full load... while doing basically nothing useful. That one sticks with me.

Jason Reed

[emphatic] Yep. Up to 40%. That’s why load/no load can be brutal if it’s not applied right. It works best when the machine is near full capacity most of the time -- like a base-load unit in a multi-compressor setup. But without enough flow control and storage, it can become the most inefficient option on the floor.

Lisa Saunders

Okay, now here’s where I think people get whiplash. Because they hear that and go, “Fine, then modulation.” And modulation does sound smarter.

Jason Reed

[matter-of-fact] Smarter in some cases, yes. Modulation changes output by opening and closing the inlet valve, so now you’ve got full load, no load, off, and partial load -- roughly 20% to 99%. The upside is reliability. It protects the compressor from short cycling because the motor keeps running while the inlet valve does the work.

Lisa Saunders

But the partial-load penalty is real. When you partially close that inlet valve, you create vacuum at the inlet, which means more horsepower to compress the air. So this is where I push back on people who act like modulation is automatically outdated. It’s not elegant on energy, but it can be exactly the right answer when you need steady, rugged operation.

Jason Reed

[responds quickly] I’ll push back on your pushback a little. Modulation is reliable, yes. And in the wrong -- or maybe I should say the real-world -- sizing situation, where the compressor isn’t properly sized and storage is thin, it may be absolutely necessary. But if you’ve got the right application for a VFD, I’m taking the VFD every time.

Lisa Saunders

[laughs] There it is. Knew you were going there. All right -- make the VFD case, but make it honest.

Jason Reed

VFDs, or VSDs, adjust motor speed based on system sensors. So instead of wasting energy unloading or throttling, they slow the motor to match demand. In the RIGHT application, they can cut energy use by about 33%. They also hold header pressure very tight -- within 1 PSIG -- and they give you soft starts and stops with very little inrush current. That’s why they make excellent trim compressors.

Lisa Saunders

The 1 PSIG part matters. If you’ve lived with a system bouncing all over the place, “within 1 PSIG” sounds almost... civilized. [chuckles] But you said “right application” in caps, basically. What’s the catch?

Jason Reed

The catch is they’re not magic. Oil-flooded VFDs are acceptable around 21% to 40%, ideal around 41% to 70%, and then good above that. Oil-free units really want to live higher -- acceptable around 41% to 70%, ideal around 71% to 100%. So if somebody expects perfect efficiency down at the basement of turndown, that’s not how it works.

Lisa Saunders

And environment matters too. Dirt and dust can clog cooling passages, high ambient temperature stresses components, wet environments invite corrosion. So this is not “install drive, walk away forever.”

Jason Reed

Correct. Same goes for variable-capacity or variable-displacement airends. Turn valves, spiral valves, poppet valves -- different ways of changing the compression chamber so the machine compresses only the air needed. That avoids a lot of modulation’s inefficiency. At 81% capacity, a variable-capacity compressor might use 83% of full-load power, versus 96% for a modulating unit.

Lisa Saunders

Eighty-three versus 96 -- that’s not rounding error. That’s the kind of number that changes a utility bill. But there’s a maintenance angle too, right?

Jason Reed

Yeah. Those valves can stick. And when that happens, pressure drops, airflow drops, the unit can overheat. Sometimes the tell is amperage. You’ve gotta actually measure what the compressor is pulling, not just guess from how it sounds.

Lisa Saunders

So the practical takeaway so far is kind of unglamorous: there is no perfect mode. There’s only fit. Start/stop for the right low-duty situation. Load/no load when a base-load machine is staying busy. Modulation when reliability and anti-cycling matter more than part-load efficiency. VFD when the demand band and environment make sense. And variable-capacity if you understand the mechanism and maintain it.

Jason Reed

Now here’s the part people skip, and honestly it changes everything: storage. Most of these control methods depend on receiver tanks more than folks want to admit. If you don’t have enough storage, demand hits fast, everybody calls for air at once, and the compressor starts reacting instead of operating.

Lisa Saunders

[reflective] I love this point because it’s boring -- and because it saves people money. The rule of thumb from CAGI is eight to ten gallons of storage for every CFM from the largest compressor. Eight to ten gallons per CFM. That is not a tiny tank tucked in the corner.

Jason Reed

Nope. And there isn’t really storage unless there’s a pressure difference between the tank and the demand side. That’s why pressure-flow control matters. The flow controller sits between supply and demand, and when there’s a fast event, it releases supplemental air from storage to keep production pressure stable.

Lisa Saunders

So let me say it back and you tell me if I’m mangling it. [pauses] The compressor is too slow to react to sudden demand, but the tank is not. The flow controller lets the tank protect the plant while the compressor catches up.

Jason Reed

[approving] That’s it. The flow controller maintains system pressure; the compressor maintains tank pressure. Once you understand that, a lot of bad control decisions start to look obvious.

Lisa Saunders

And it also explains why “we have a VFD, so we don’t need storage” is just... wrong. A drive still can’t teleport air into the header. It reacts, but it does not react instantly.

Jason Reed

Exactly. Now layer in multiple compressors. Base-load, trim, backup. A master controller can network them and hold system pressure in a tight band -- sometimes as low as plus or minus 2 PSIG -- which is great compared with the old cascading setup where units sit at staggered pressure levels and waste energy.

Lisa Saunders

But plus or minus 2 PSIG is not a permission slip to stop thinking. Because if your lead machine fails right when you’re at the bottom of that band, pressure can fall fast while the standby unit is still coming online and running through its startup sequence.

Jason Reed

[serious] That’s the risk. And if critical processes are close to shutdown pressure, you can lose production before anybody in the room even understands what happened. Master controls are powerful. They are not “fix it and forget it.”

Lisa Saunders

Which brings us back to the least flashy truth in the whole conversation: a smart compressed air system is balanced. Control method, storage, actual demand pattern, and the kind of pressure stability the process really needs. Not the kind people assume it needs.

Jason Reed

[calm] Yeah. If I had to leave one thing with people, it’s this: the smartest capacity strategy usually isn’t the fanciest control package. It’s the one that keeps pressure stable, avoids short cycling, and uses storage to do the fast work the compressor physically can’t do.

Lisa Saunders

[warmly] And if you’re choosing between a clever answer and an honest demand profile, pick the demand profile. That’s where the truth is.

Jason Reed

That’s the show.