Every compressed air efficiency conversation eventually lands on a single figure: dollars per CFM per year. It's a useful shorthand: it lets you take a measured leak rate and turn it into a number a finance team can weigh against a repair quote. The trouble is that a generic $/CFM figure, quoted without its assumptions, is close to meaningless. Two sites with identical leaks can have genuinely different costs per CFM, and the gap is usually two or three variables, not one.
Start with the compressor's efficiency, not the leak
The first variable is how much electrical energy your compressor actually spends to produce one CFM of free air delivery at your operating pressure. A commonly used rule of thumb for a rotary screw compressor at 100 PSI (6.9 bar) is somewhere in the range of 4 to 5 CFM per kilowatt of input power. That range exists because real compressors don't sit at one efficiency point: a fixed-speed screw compressor running part-loaded is less efficient per CFM delivered than the same machine running fully loaded, and a variable-speed drive unit can hold efficiency more consistently across a wider load range.
If your plant runs one large compressor near full load most of the time, you're likely closer to the efficient end of that range. If you're running an older fixed-speed machine that spends a lot of time loading and unloading to track a variable demand, your real efficiency is worse than the rule of thumb suggests, which means your actual cost per CFM is higher than a generic calculator will tell you.
Then get the tariff right, not just the usage rate
The second variable, and the one most cost-per-CFM figures get wrong, is the electricity rate itself. Australian commercial and industrial tariffs typically carry two components: a usage charge in cents per kilowatt-hour, and a demand charge based on the site's peak kilowatt draw during a billing period. A quoted usage rate on its own (sometimes in the high teens or low twenties in cents per kWh) understates what a kilowatt-hour actually costs once a fair share of the demand charge is allocated across consumption.
For a mid-size Australian manufacturer, that effective all-in rate commonly lands somewhere between 28 and 36 cents per kWh. Below is what a defensible $/CFM/year figure looks like once you run both variables (compressor efficiency and effective rate) through the calculation, at 5,000 hours a year (a typical two-shift site):
| Compressor efficiency | Effective rate | $/CFM/year |
|---|---|---|
| 4.0 CFM/kW | $0.28/kWh | $350 |
| 4.5 CFM/kW | $0.32/kWh | $356 |
| 5.0 CFM/kW | $0.36/kWh | $360 |
The narrow spread in that table is close to coincidence: the efficiency and rate assumptions happen to roughly offset each other across the range. Push either variable to its real extreme for your site (an inefficient part-loaded compressor on a high demand-charge tariff, or an efficient VSD unit on a favourable contract) and the $/CFM figure moves a lot more than that table suggests. The point isn't the specific numbers: it's that the figure only means something once you've pinned down where your site actually sits on both axes.
A $/CFM figure without its assumptions attached is a guess wearing a decimal point.
Running hours: the variable everyone gets wrong in the same direction
The third variable is how many hours a year the figure should be multiplied by, and it's not the same as how many hours the plant operates. A leak runs for every hour the compressed air system is pressurised, which for most sites is closer to the compressor's run hours than the production schedule. A site that runs one shift on the production line but keeps compressed air live around the clock for climate-sensitive processes, pilot air, or standby equipment is accumulating leak cost on a 8,760-hour basis, not a 2,500-hour one, more than three times the annual cost for the exact same leak.
This is the single easiest place to understate a site's real leak cost, because the instinctive assumption is to use production hours. If your air system doesn't get isolated or shut down outside production, your leak-cost basis should reflect that.
Where a single site-wide figure breaks down
Even once the three variables above are pinned down, a single $/CFM figure applied uniformly across a site can still mislead, because it assumes every CFM lost displaces the same marginal cost. On a site with two compressors (a base-load machine running near full efficiency and a lag machine cutting in only at peaks), a leak large enough to push total demand past the point where the second compressor starts is far more expensive than the same-sized leak on a site with headroom on one machine. The lag compressor typically runs less efficiently at partial load, and its capital and maintenance cost gets spread over fewer hours, which a simple $/CFM/year figure doesn't capture at all.
This matters most on sites sitting close to a capacity threshold. If a plant's lag compressor has started cutting in more often over the past year with no real change in production, that's a sign leak load specifically has been pushing total demand toward that threshold, and the marginal cost of the next CFM lost is higher than the average $/CFM figure suggests, because it's the CFM that tips the system into needing a second machine.
Putting a real figure together
A defensible cost-per-CFM figure for your site needs three inputs pinned to your actual operation, not a generic default: your compressor's realistic efficiency at its typical load point, your effective all-in electricity rate including a fair demand-charge allocation, and the actual hours your air system stays pressurised. Get those three right and a measured leak rate (in CFM, from an actual survey rather than an estimate) turns into a genuinely useful annual dollar figure for deciding what gets fixed first.
Where the $/CFM figure remains an estimate no matter how carefully the assumptions are built is the CFM number itself. A modelled leak size from an orifice chart is directional. A measured leak rate from an acoustic survey, using the sound level and distance to source rather than an assumed hole diameter, is the number that removes the last variable from the equation.
Why the figure is still worth calculating, imprecision and all
None of this means a $/CFM figure isn't worth having: it means it's worth having with its assumptions attached rather than quoted as a bare number. A figure built on your site's actual compressor efficiency, your actual effective rate, and your actual running hours is a legitimate tool for comparing repair priorities against each other, even if the absolute number carries some uncertainty. What it shouldn't be used for is comparing your site against a generic industry average pulled from a chart, because that average was built on assumptions that may have nothing to do with your tariff, your compressor age, or your operating pattern.
The most useful version of this figure is one your own maintenance team can defend line by line: this is our compressor's realistic efficiency, this is our effective rate including demand charges, this is how many hours our air system actually stays pressurised, rather than one borrowed from somewhere else.
Try the assumptions against your own numbers in the ROI calculator, or see how the CFM figure itself gets measured rather than estimated.