Two numbers, and one of them is usually wrong
Specific power is the input kilowatts divided by the delivered cfm, times a hundred. At the defaults, 22 kW delivering 100 cfm is 22 kW per 100 cfm. That figure is what lets two machines be compared honestly, and it is what a compressed air survey is really measuring.
The trap is the flow number. Displacement is what the pump sweeps; delivery is what comes out of the discharge after slip, heating and the aftercooler have had their share, and the two are not close. A specific power worked out from displacement flatters the machine badly. If you do not have a trustworthy delivered figure, a pump-up test against a known receiver volume gives you one: fill from one pressure to another with everything closed, and the delivered flow is the vessel volume times the pressure rise divided by atmospheric, divided by the minutes it took.
What a hundred cubic feet costs
At 22 kW and 100 cfm, a hundred cubic feet takes one minute of running and 0.3667 kWh. At 14 cents that is 5.13 cents while the machine is loaded. Simple, and slightly dishonest, because the machine spends part of its life running and making nothing.
Across a year at 4,000 running hours and 65 percent load, the defaults give 57,200 kWh loaded and 9,800 kWh unloaded, 67,000 kWh and 9,380 dollars in total, against 15.6 million cubic feet of air. That is 6.01 cents per hundred cubic feet all in — a 17.1 percent surcharge for the hours the motor spun and delivered nothing. That is the number to use for everything else, because it is the one the bill agrees with.
The blow gun arithmetic
A tool drawing 15 cfm is using 900 cubic feet an hour, which at 6.01 cents a hundred is 54 cents an hour, or nine tenths of a cent a minute. Nobody chases a cent. Five hundred hours of trigger time a year is 271 dollars, which is a different conversation, and a blow gun wedged open on a bench overnight is 6.50 dollars while nobody is in the building.
The reason to price air per minute is that it makes those two facts sit next to each other. The per-minute figure explains why nobody notices; the per-year figure explains why it matters.
Load share is the cheapest thing to measure
Most machines with any control sophistication have both a run hour meter and a load hour meter, and the ratio between them is the load percentage on this form. It takes thirty seconds to read and it moves the all-in price more than anything else on the page. A machine at 90 percent load carries almost no idling surcharge. A machine at 30 percent is spending more than half of its electricity making nothing, and that is the signature of a compressor sized for a shop that no longer exists, or of a system with no storage to speak of.
The page will not tell you which. Both look identical on a bill, and telling them apart means watching the pressure trace over a shift rather than reading a meter.
What is deliberately not in here
No capital, no maintenance, no oil or separator elements, no filter media, no dryer electricity, no floor space. Surveys that quote a fully loaded cost of compressed air are including those and are answering a different question — whether to have the system at all. This one answers the question that comes up weekly: what does the air itself cost to make, so that a leak, an idle tool or a badly chosen pneumatic actuator can be priced against the alternative in an afternoon.
Questions people ask
How much does compressed air cost per cubic foot?
It depends entirely on your machine and your electricity price, which is why this asks for both. At 22 kW delivering 100 cfm and 14 cents a kWh, a hundred cubic feet costs 5.13 cents while the machine is loaded and 6.01 cents once the idling hours are spread across it. Halve the specific power or double the electricity rate and the answer moves proportionally. A figure quoted without the specific power and the price behind it is not usable.
What is specific power and why does it need a pressure attached?
It is the motor input kilowatts divided by the delivered cfm, times a hundred, so it reads as kW per 100 cfm. It needs a pressure because the same machine asked for a higher discharge pressure delivers less air for the same power. Comparing a specific power measured at 100 psi against one measured at 125 psi is comparing nothing. Take both figures at the pressure you actually run.
Is nameplate horsepower the same as input power?
No. Nameplate horsepower is a shaft rating. Multiplying it by 0.746 gives shaft kilowatts, and the electrical input is above that by whatever the motor efficiency and power factor happen to be, and the machine may not be working at its nameplate point anyway. A clamp meter with a power factor reading at the panel settles it in five minutes, and the difference between the assumed number and the measured one is routinely ten percent or more.
Why does unloaded running cost so much?
Because a load and unload machine keeps the motor turning and often keeps the airend turning against no load, which on the defaults here is 7 kW producing nothing. At 65 percent load across 4,000 hours that is 9,800 kWh and 1,372 dollars a year of electricity that buys no air. It shows up on the page as a 17 percent surcharge on the price of every cubic foot. More storage, a smaller machine or a variable speed drive all attack it differently, and which of those applies is a question about your demand profile rather than about this arithmetic.
Should I include the demand charge?
Include it if the compressor is running during the window your utility uses to set demand, which for a shop on one day shift it almost certainly is. It is entered separately because it behaves differently: it is charged on peak kilowatts rather than on kilowatt hours, so it does not care how much air you made, only that the motor was pulling when the meter was watching. If the machine is off during the peak, set the share to zero and it drops out.