What a Compressor Is Actually Doing to the Air

Why this matters

A tech who thinks a compressor "makes air" has no way to answer the three questions that come up on every air-room call: why the discharge pipe is too hot to touch, why raising the setpoint costs more than the ten pounds suggests, and why the machine that does the same job in two stages costs more and uses less. All three fall out of one fact. A compressor does work on a gas, and almost all of that work turns into temperature before any of it turns into anything useful.

Get that straight and the rest of the air room stops being a set of memorised rules.

Work on a gas, not air out of nowhere

Air goes into the intake at atmospheric pressure and comes out of the discharge at line pressure. Nothing was added. The same molecules now occupy a smaller volume, and squeezing them into it took work supplied by the motor.

That work has to land somewhere, and the first place it lands is internal energy, which you observe as temperature. The gas leaves the compression element hot, every time, on every machine type. What happens to that heat next is the design difference between families, and where it ends up in the plant is covered in the article on why most of the energy leaves as heat. This article is about the compression itself.

The relationship that governs is the ratio, not the pressure

Compression work and compression heating both scale with the ratio of absolute discharge pressure to absolute intake pressure, not with the gauge difference. Absolute means gauge plus atmospheric, roughly 14.7 psia at sea level; the gauge zero and why it moves with elevation belong to the article on measuring pressure and what the reference is, and this article uses 14.7 psia throughout as a sea-level value.

For an ideal gas compressed with no heat removed during the compression, the temperature ratio follows the pressure ratio raised to the power of (k-1)/k, where k is the ratio of specific heats. For dry air near room temperature k is about 1.4, making the exponent about 0.286. Those conditions matter and they are not decoration: ideal gas, constant specific heats, and crucially no cooling during the compression stroke. A real machine violates the last one, deliberately, and the section after the example says by how much.

Temperatures in this relationship are absolute. Add 460 to Fahrenheit to get Rankine. That offset is a rounding of 459.67, which is close enough everywhere except a radiation calculation.

Worked example: one stage to 125 psig on paper

Intake air at 70 F, sea level. Target 125 psig.

• Intake absolute: 14.7 psia. Intake absolute temperature: 70 + 460 = 530 R.
• Discharge absolute: 125 + 14.7 = 139.7 psia.
• Pressure ratio: 139.7 / 14.7 = 9.50.
• Temperature ratio: 9.50 raised to the 0.286 power = 1.90.
• Discharge absolute temperature: 530 x 1.90 = 1008 R.
• Discharge temperature: 1008 - 460 = 548 F.

That is the paper answer for a single stage taking atmospheric air to 125 psig with no heat removed on the way. It is not a number you will read on a gauge, and it is not meant to be. It is the size of the problem the machine has to solve.

Two things fall out of it immediately. First, discharge temperature is set by the ratio and the intake temperature, so a machine drawing hot air off a ceiling is starting the multiplication from a bigger number: intake at 110 F instead of 70 F gives 570 x 1.90 = 1083 R, which is 623 F, a rise of 75 F in the discharge for a 40 F rise at the intake. Second, nothing about this depends on the machine being a screw or a piston. It is the gas, not the mechanism.

The failure mode of ignoring it. A shop that treats discharge temperature as a machine characteristic rather than a ratio consequence will chase a high-discharge alarm through the cooler, the fan and the thermostat, and never look at the intake the customer moved into a boiler room, or the setpoint somebody raised from 110 to 135 psig to chase a symptom at a tool.

Why the real machine runs far cooler than the paper one

Every practical design removes heat during or immediately after compression, which is why you will read a few hundred degrees at the airend discharge of an oil-injected screw rather than 548 F.

The oil-injected rotary screw injects a large flow of oil directly into the compression chamber. The oil is there to seal the rotor clearances and to lubricate, but its dominant duty is thermal: it absorbs heat as the gas is compressed, so the compression is closer to isothermal than the uncooled case, and the mixture leaves the airend far below the paper figure. That injected oil then has to be separated from the air and cooled, which is why the package has a sump, a separator element and an oil cooler that between them are physically larger than the airend.

An oil-free machine has no such heat sink inside the element. It handles the same physics by staging and intercooling instead, which is not an option but a requirement, and it is why oil-free packages above small sizes are two-stage as a rule.

State the condition with the number: 548 F is the uncooled single-stage result at a 9.50 ratio from 70 F. Any real reading below it is the cooling working. A reading approaching it is a cooling failure, not a normal variation.

What staging buys, under what condition

Split the same 9.50 ratio evenly across two stages and each stage sees the square root, 3.08.

• Temperature ratio per stage: 3.08 raised to the 0.286 power = 1.379.
• First stage out: 530 x 1.379 = 731 R, which is 271 F.
• Intercool that back to 100 F, which is 560 R, before the second stage.
• Second stage out: 560 x 1.379 = 772 R, which is 312 F.

Both stages land under about 320 F against 548 F for one stage. That is the whole reason two-stage machines exist at these pressures, and it is a materials and lubrication argument before it is an energy argument.

The energy argument follows from the same figures. Ideal compression work per stage is proportional to the inlet absolute temperature times (ratio to the 0.286 power, minus one):

• Single stage: 530 x 0.90 = 478 units.
• Two stage: 530 x 0.379 = 201, plus 560 x 0.379 = 212, totalling 413 units.
• 413 / 478 = 0.864, so about 14 percent less ideal work.

The condition on that 14 percent is the intercooling temperature of 100 F. Intercool perfectly back to the 530 R intake temperature and the saving grows to about 16 percent; intercool badly, back to 620 R, and the second stage costs 235 units, the total is 436, and the saving falls to about 9 percent. The intercooler is not an accessory on a two-stage machine. It is where the saving lives, which is why a fouled intercooler shows up as energy before it shows up as an alarm.

None of the three figures above is a machine specification. They are ideal-gas results that ignore mechanical losses, leakage past the rotors or rings, and pressure drop through the intercooler itself, all of which move the real number the wrong way. Take the machine's actual specific power off its performance data sheet, tested to ISO 1217 or published on a CAGI data sheet, and use these numbers to understand which direction a change will move it.

What this decides on a service call

Intake location is a live variable, not a detail. Cooler, cleaner, drier intake air lowers discharge temperature proportionally in absolute terms and reduces the water load the rest of the system has to deal with. An intake ducted from outdoor shade beats one breathing the hot ceiling of the room the compressor is heating.

A pressure increase is a ratio increase. Going from 100 to 120 psig moves the ratio from 7.80 to 9.16, about 17 percent, and both the work and the heating follow it. It is never the cheap fix it sounds like when someone says "just give it ten more pounds."

Discharge temperature is a diagnostic of the cooling, read against the ratio. Before condemning a cooler, take the intake temperature and the actual discharge pressure and work out what the uncooled result would be. The gap between that and what you read is what the cooling is doing.

How to verify you got this right

Take three readings on one machine: intake air temperature at the filter, discharge pressure at the machine, and discharge air temperature at the aftercooler outlet. Compute the ratio, compute the uncooled temperature, and write all four numbers on the ticket. Repeat on the same machine on a hot afternoon. The ratio should be essentially unchanged and both temperatures should have risen by close to the intake rise. If the aftercooler outlet rose much more than the intake did, the cooler is the term that moved, not the compression.

Every reading above is taken on the outside of the equipment with a surface or air probe, on a running machine, and nothing is opened. Discharge piping, the aftercooler and the separator body reach temperatures that burn on contact, so probe them with a non-contact or clamp-on sensor and keep sleeves, leads and hands off bare pipe. Do not throttle, valve or restrict a running compressor's discharge to observe an effect: the relief device is the only thing between you and the stored energy, and 29 CFR 1910.169(b) exists specifically because a valve in the wrong place defeats it. If a reading requires opening the package or any pressurised component, that is a different job: stop the machine, isolate and lock its energy sources under 29 CFR 1910.147, vent the sump and the receiver to zero, and confirm zero on a gauge you trust before a fastener moves.

References

• 29 CFR 1910.147, control of hazardous energy, for isolation and stored-energy release before opening a compressor package
• 29 CFR 1910.169(b), air receivers, installation and equipment requirements including safety devices
• ISO 1217, displacement compressors acceptance tests, and CAGI performance data sheets, which bind through the purchase specification rather than by law
• See related: Measuring Pressure and What the Reference Is; Why Most of the Energy Leaves as Heat; The Compressor Types and What Each One Is Suited To