By Ron Marshall
George ran the compressed air system at his plant for eleven years. If you would have asked him last spring how it was doing, heβd have shrugged and said, βFine.β Two screw compressors, sharing the load, humming along the way they always had. Nobody calls a meeting about compressed air until something breaks, and nothing was breaking. Fine was good enough.
Then the audit team showed up with their data loggers.
They clamped current transducers on both machines, pressure transducers on the header, installed a flow meter and let the systems talk for a week. George figured heβd get a clean bill of health. Instead, he got a number that made him ask the auditor to repeat it: 45 kW per 100 cfm. He didnβt need an explanation of why that was bad. He just needed to watch the auditorβs eyebrows go up when he said it out loud.
A well-controlled system should land somewhere around 18 to 22 kW/100 cfm (it can even be lower). Georgeβs system was running close to double that. Two lubricated screws, both set in modulation, both throttling their inlet valves to chase demand. Hereβs the part nobody tells you about modulation: it throttles flow, not power. Choke a compressor down to half its output and it might still be eating 70-80% of full-load energy. Thatβs like driving with one foot on the gas and the other on the brake, then wondering why the fuel bill looks high.
And he had two of them doing it at once! Both partway throttled, both burning more than they delivered, neither one willing to actually unload and shut off, even at 2 a.m. when the plant barely needed any air.
George started running numbers. First option: load/unload control. Nothing fancy. When the compressor isn’t needed, it unloads, idles and then shuts off instead of strangling itself at the inlet. Pair that with enough storage to slow down the compressor cycles, and his modeled specific power dropped into the high 20s. Better. Still not where he wanted to be.
The number that really got his attention came from a different combination: one fixed-speed compressor base-loaded on load/unload, and a slightly larger VSD machine trimming on top of it. A VSD doesnβt throttle. It just slows down to match demand, the way cruise control holds a steady speed on the highway instead of stop-and-go through downtown traffic. Run the two together, and the blended specific power for the whole system came in under 20. Same plant. Same air demand. Half the energy per cfm.
George ran the payback math three times, mostly because he didnβt believe it the first two. The numbers held up, and thatβs before factoring in the maintenance savings on new compressors that arenβt fighting their own throttle valves around the clock, and the utility incentives. What stuck with George wasnβt βmodulation is the enemy.β It was that heβd run a system for over a decade assuming it was fine, simply because nothing had ever told him otherwise.
If you donβt know your specific power number, it is worth finding out before anything else. The data doesnβt care what you assumed.
Want to learn how to measure specific power and spot control problems like Georgeβs before they cost you a decade of wasted kilowatts? Thatβs exactly what we dig into in Compressed Air Challengeβs Fundamentals and Advanced Compressed Air Systems training.
