A building products company purchased 90 kW and 45 kW compressors a few years ago to run its plant. The staff had heard that it was not wise to purchase a VSD compressor that was the same size or larger than the base compressors, so the 90 kW unit was selected as the trim VSD unit. The purchasers did not realize by choosing these sizes, they were setting themselves up for an unexpected efficiency problem.
Both compressors were required for daytime full shifts, with flow reaching 900 cfm — but during nights and weekends the flow dropped off to about 400 cfm. The 45-kW unit, producing slightly less than 300 cfm was always fully loaded, and the 90-kW unit would swing between full load at plant peak all the way down to minimum output of 100 cfm at the lowest weekend flow.
A compressed air auditor monitored the system and pointed out a problem with the control of the two compressors. He explained with a large size mismatch, and VSD being much bigger, the smaller base compressor will get locked on, unable to unload, with the result that both compressors were running during light loads, rather than having just one compressor feeding the load.
When a VSD is lightly loaded it runs in a less efficient part of its curve (indicated by the yellow band in the graphic). If the smaller base unit was turned off, the VSD alone could carry the full plant in a more efficient manner (green bar). Operation at this low flow is not desired for VSD compressors, as water tends to form in the lubricant due to the small amount of heat generated during the light compression duty. Oil carryover is also often a problem at low flows.
The auditor pointed out that changing the compressor control band coordination, or setting up an operating schedule, could allow the smaller compressor to unload and turn off during light loads, making for more efficient operation and better air quality.
The auditor also pointed out that the more than 400 cfm weekend load (no production) was excessive. The average flow during all hours was measured at only 470 cfm. If the weekend flow can be assumed to consist of mostly leaks, this would mean that 88% of the air the compressors produce is wasted. Further investigation by the embarrassed maintenance department is ongoing.
It is indeed amazing how many air leaks are ignored. It is usually because they leave no pool of oil or water that the amount of leakage is not understood, at least that is what I see. Perhaps an examination of compressor run power use while the place is shut down would be educational.