Why compressor sizing is a duty-cycle problem, not an addition problem
The single most common way to buy the wrong compressor is to add up the nameplate CFM of every air tool in the building. A shop with six 5 CFM tools does not need 30 CFM, because nobody runs six tools continuously; and a shop with one 25 CFM sander genuinely does need close to 25 CFM whenever that sander is running, even though the shift average is far lower. The two errors point in opposite directions, and only a duty-cycle calculation separates them.
So the sizing question splits in two. Average demand — the sum of each tool's consumption weighted by the fraction of time it flows — determines the compressor's capacity and, more importantly, its energy bill. Peak demand — what happens when several tools open at once — determines whether the header pressure holds, and it is solved with storage rather than with capacity. That is why a properly designed system pairs a modestly sized compressor with a generously sized receiver: the receiver rides through the peaks so the compressor only has to cover the average.
This calculator handles the average side, which is the part that determines what you buy and what you pay to run. Size the storage that covers the peaks with the compressed air receiver tank sizing calculator, which works from the pressure band you are willing to give up and the size and duration of the demand event.
The formula, term by term
Time-averaged demand. For each tool, multiply its rated consumption by its duty cycle and add the results. A 20 SCFM grinder at 50% contributes 10 SCFM; a 5 SCFM continuous purge contributes its full 5. The rated consumption must be the tool's average figure at its rated pressure, which is what manufacturers publish; the instantaneous surge when an impact wrench first spins up is several times higher and is a storage problem, not a capacity problem.
Leak allowance. Leakage is real, continuous demand that no tool list contains. It runs 24 hours a day if the compressor does, which is why it dominates the energy bill in older plants far more than it dominates the capacity requirement. Do not guess it if you can measure it: shut every tool off, let the system come to pressure, and time the load/unload cycles. Leak flow equals compressor capacity multiplied by the fraction of the cycle spent loaded. The compressed air leak cost calculator turns a measured or estimated leak rate into an annual dollar figure.
Growth allowance. Straight headroom for equipment you expect to add. Keep it honest — an oversized compressor spends its life at part load, where load/unload and inlet-modulation controls are markedly less efficient per cubic foot than at full load.
Standard versus actual volume. This is where most spreadsheet sizing goes wrong. SCFM is a mass flow dressed up as a volume: it is the volume that quantity of air would occupy at a fixed reference of 14.696 psia and 68 °F. ACFM is the volume the compressor's inlet actually swallows at the conditions where it sits. At altitude the inlet air is thinner, so the same mass of air occupies more volume, and the machine must be bigger. The correction is the ideal gas law applied twice: multiply by the pressure ratio 14.696/p and by the absolute temperature ratio T/527.67, with temperature in degrees Rankine (°F + 459.67).
Barometric pressure from elevation. The calculator computes site pressure from the US Standard Atmosphere troposphere model, p = 14.696 × (1 − 6.87535×10⁻⁶·h)^5.2559 with h in feet. At sea level that returns 14.696 psia; at 5,280 ft it returns 12.100 psia.
Power. Package specific power — total input kilowatts per 100 cfm delivered at full load — is published on the CAGI data sheet for every compressor sold in North America, and it is the only honest way to compare two machines. Multiply it by the flow in hundreds of cfm. Then apply the discharge pressure factor: the widely used field rule is that each 2 psi of additional discharge pressure changes compressor power by about 1%, so running a header at 120 psig instead of 100 psig costs roughly 10% more energy for the same air.
Worked example: a three-tool fabrication shop
A shop runs a 20 SCFM sander half the time, a 10 SCFM impact wrench a quarter of the time, and a 5 SCFM continuous air-operated pump. Leaks are estimated at 15% of tool demand, and 10% growth headroom is wanted. The shop is at sea level with a 68 °F intake, runs a 100 psig header, and operates 2,000 hours a year at $0.12/kWh. The compressor under consideration has a CAGI specific power of 18 kW per 100 cfm.
- Time-average each tool. 20 × 0.50 = 10.0; 10 × 0.25 = 2.5; 5 × 1.00 = 5.0.
- Tool demand. 10.0 + 2.5 + 5.0 = 17.5 SCFM. Note that the nameplate total is 35 SCFM — exactly twice the real figure.
- Leaks. 17.5 × 0.15 = 2.625 SCFM, so average consumption is 17.5 + 2.625 = 20.125 SCFM. Leaks are 2.625/20.125 = 13.0% of everything the compressor actually makes.
- Growth. 20.125 × 1.10 = 22.14 SCFM of design capacity.
- Site correction. At sea level p = 14.696 psia and the intake is at the 68 °F reference, so both ratios are 1.000 and ACFM = SCFM = 22.14.
- Machine to buy. Rounding up to the next 5 cfm gives a 25 ACFM requirement, which in catalogue terms is a 5 hp class rotary screw or a well-chosen 7.5 hp reciprocating unit.
- Full-load power. (22.14 ÷ 100) × 18 kW × 1.00 = 3.98 kW, or 3.98 ÷ 0.7457 = 5.34 hp of electrical input.
- Annual energy. Energy follows the average consumption of 20.125 SCFM, not the design figure: (20.125 ÷ 100) × 18 × 2,000 = 7,245 kWh, which at $0.12 is $869.40 a year.
Now change one thing: fix the leaks. Dropping the leak allowance from 15% to 5% takes average consumption from 20.125 to 18.375 SCFM, and annual energy from 7,245 kWh to (18.375 ÷ 100) × 18 × 2,000 = 6,615 kWh. That is 630 kWh, or $75.60 a year, from a shop with only three tools — scale it to a plant with a 500 SCFM demand and the same proportional fix is worth over $1,800 a year.
How to read the result
Compare like with like before you buy. The required SCFM figure is a mass flow at a stated reference; the ACFM figure is an inlet volume at your site. Manufacturers quote capacity both ways and do not always say which. If a data sheet says "acfm" it usually means inlet volume measured on the test stand under ISO 1217 conditions, which for a site at altitude is not the same volume your machine will need to swallow. Take the ACFM figure from this calculator to the vendor and ask them to confirm the machine delivers your SCFM at your elevation.
Treat the recommended rating as a floor, not a target. It is the average demand plus your own allowances, rounded up to the next 5 cfm. Peaks are handled by receiver volume, and a common and expensive mistake is to buy capacity to cover a peak that a $2,000 tank would have covered instead.
Read the leak line as a purchasing decision. The calculator reports leakage in SCFM and as a share of what the compressor actually produces. That flow is capacity you are already paying for and getting nothing from, and unlike every other line in the calculation it can be reduced without buying anything.
Treat the energy figure as a lower bound. It assumes input power tracks flow linearly, which is close to true for a well-set-up variable-speed drive machine and optimistic for load/unload or inlet-modulating control. A modulating compressor at half flow can still draw a large majority of its full-load power, so if your machine uses that control scheme, the real bill will be higher than the number here.
Check the pressure you actually need. The discharge pressure factor makes this visible: dropping a header from 110 psig to 100 psig cuts compressor power by about 5% by the same rule that raised it. If one machine forces the whole plant up, a dedicated booster for that machine is usually cheaper than the whole-plant penalty.
Altitude correction: barometric pressure and inlet volume multiplier
| Elevation (ft) | Barometric pressure (psia) | ACFM per SCFM at 68 °F |
|---|---|---|
| 0 | 14.696 | 1.000 |
| 1,000 | 14.173 | 1.037 |
| 2,000 | 13.664 | 1.075 |
| 3,000 | 13.171 | 1.116 |
| 5,000 | 12.228 | 1.202 |
| 7,000 | 11.339 | 1.296 |
| 10,000 | 10.106 | 1.454 |
The multiplier is simply 14.696 ÷ p. A hot intake multiplies it further by T°R ÷ 527.67: a 100 °F day adds (559.67/527.67) − 1 = 6.1% on top of the altitude figure.
Input power and annual energy at 18 kW per 100 cfm, 100 psig
| Average demand (SCFM) | Input power (kW) | Input power (hp) | Annual kWh | Annual cost |
|---|---|---|---|---|
| 10 | 1.80 | 2.41 | 3,600 | $432 |
| 25 | 4.50 | 6.03 | 9,000 | $1,080 |
| 50 | 9.00 | 12.07 | 18,000 | $2,160 |
| 100 | 18.00 | 24.14 | 36,000 | $4,320 |
| 200 | 36.00 | 48.28 | 72,000 | $8,640 |
| 500 | 90.00 | 120.69 | 180,000 | $21,600 |
Horsepower here is electrical input, kW ÷ 0.7457, not motor nameplate shaft horsepower. Scale any row for a different specific power by the ratio of the two figures.
Sizing mistakes that cost real money
- Adding nameplate ratings without duty cycles. In the worked example above that doubles the requirement, from 17.5 to 35 SCFM, and an oversized machine then spends its life at inefficient part load.
- Sizing capacity to cover a short peak. A momentary demand event is a storage problem. Receiver volume is far cheaper per unit of ride-through than compressor capacity.
- Ignoring altitude. A machine specified in SCFM at sea level and installed at 5,000 ft delivers about 1/1.202 = 83% of the mass flow you expected from the same inlet volume.
- Leaving leaks out of the calculation. They are continuous demand, they grow every year, and they are the only term on the list you can reduce for the price of a soapy-water bottle and an afternoon.
- Running the header higher than the plant needs. By the 1%-per-2-psi rule, an unnecessary 20 psi costs about 10% of the compressor's entire energy bill, every hour it runs.
- Comparing motor horsepower instead of specific power. Two 50 hp machines can differ substantially in the cfm they deliver per input kilowatt; the CAGI sheet is what settles it.
- Confusing tool inlet pressure with compressor discharge pressure. The tool needs its rated pressure at the coupling. Piping, filters, dryers and hose all take their share, so the compressor must discharge higher.
Which reference conditions this uses
The SCFM figures on this page are referenced to 14.696 psia and 68 °F (20 °C), dry. That is not universal. ISO 1217, the international standard for displacement-compressor acceptance testing and the basis of most published compressor data, references 1 bar absolute (14.5 psia) and 20 °C; some North American tool literature uses 14.7 psia at 60 °F, and ASME's definition adds 36% relative humidity. The differences are a few per cent, which is small next to a duty-cycle error but large enough to matter when you are comparing two machines within 5% of each other. Check the reference stated on the data sheet before you treat two capacity figures as comparable.
What this calculator deliberately leaves out
Simultaneity across departments. A large plant rarely runs every department at its own peak at the same moment, and a diversity factor below 1.0 is sometimes applied across zones. This calculator does not apply one, because duty cycles already capture time-averaging and applying both is double-counting. If you have measured department-level diversity, apply it to your entered duty cycles rather than adding a second factor.
Dryer and filtration losses. A heatless desiccant dryer diverts a share of its rated flow as purge air, which is genuine continuous demand. Enter it in the "other continuous demand" field — it is not in any tool list.
Piping pressure drop. The calculator asks for the discharge pressure the compressor must produce, and it is your job to have added distribution losses to the tool's requirement. Undersized headers and quick-connect couplings are the usual culprits.
Part-load control behaviour. Modelling load/unload cycling, inlet modulation and variable-speed control properly requires the machine's own performance curve. The linear assumption here is a floor.
Once the machine is chosen, the rest of the plant-room arithmetic follows: the motor full load amps calculator sizes the feeder for the compressor motor, the electrical power calculator converts kW to amps at your service voltage, and the bearing L10 life calculator covers the rotating-equipment maintenance side of the same plant room.
Key terms
- SCFM
- Standard cubic feet per minute — a mass flow expressed as the volume it would occupy at a fixed reference pressure and temperature. Independent of where the compressor sits.
- ACFM
- Actual cubic feet per minute — the volume of air at the conditions where it is measured, normally the compressor inlet. Rises with altitude and with intake temperature for the same mass flow.
- Duty cycle
- The fraction of the working period during which a tool actually flows air. The bridge between a nameplate rating and a real demand.
- Specific power
- Total package input kilowatts per 100 cfm delivered at full load, published on the CAGI data sheet. The efficiency figure that lets two compressors be compared directly.
- Receiver
- The storage tank that supplies short demand peaks from stored volume so the compressor only has to meet average demand.