Compressor performance acceptance worksheet

What does the worksheet calculate?

The worksheet compares what was promised with what was measured, on the same basis. It takes the contracted or specified duty, the test readings and any permitted corrections, then produces four results: delivered flow against required flow, specific power against the declared figure, pressure behaviour against the control band, and a final disposition of accept, accept with conditions, hold or reject.

The calculator panel on this page handles the arithmetic. The explanation below sets out what to enter, how the method works and how to read the answer, because most acceptance disputes start with mismatched definitions rather than arithmetic errors.

• Flow deviation: measured FAD minus required FAD, as a value and as a percentage of the requirement.
• Specific power: package input power divided by FAD, in kW per m³/min.
• Pressure check: the minimum pressure seen at the agreed point while the required flow is drawn.
• Evidence completeness: whether every mandatory reading, document and condition is recorded.

Define the requirement before anyone tests

Acceptance is only as good as the requirement it tests against. Write the requirement in measurable terms and agree it with the buyer, the supplier and whoever will witness the test. Changing the comparison after the readings are known turns acceptance into negotiation.

Keep one record per machine and per configuration. A replacement package, a new machine and a machine returning from overhaul are tested against different expectations: the first two against a declared performance, the third against its own baseline or the restored-condition criteria set before the work started.

How the method works

For packaged rotary and reciprocating machines, flow and power are normally measured and corrected by the method of ISO 1217, using the annex that matches the machine type. The standard defines FAD at the compressor inlet conditions and explains how readings taken at a different speed or inlet condition may be corrected. On site, full ISO 1217 conditions are rarely achievable, so the worksheet records exactly which method was used and whether the result is a contractual test or an indicative check.

1. Identify the machine: make, model, serial number, airend serial, motor or engine rating and software version.
2. Record the configuration: dryer, filters, receiver, cooling arrangement and any accessories inside the test boundary.
3. Stabilise the machine at the agreed pressure until discharge temperature, oil temperature and pressure are steady. A normal oil-injected screw usually runs with a discharge temperature of about 75–95 °C.
4. Measure inlet temperature, inlet pressure (barometric) and relative humidity.
5. Measure flow with a calibrated method suitable for the size: nozzle, orifice, or a calibrated flow meter with adequate straight pipe.
6. Measure package input power with a power analyser on all three phases, not by clamp current alone.
7. Repeat readings over an agreed period and record the spread.
8. Apply only the corrections allowed by the agreed method, and show each one.
9. Compare with the requirement and apply the decision rule agreed in advance.

Core formulas

The calculations are simple; the discipline lies in keeping units and reference conditions consistent.

• Flow deviation (%) = (FAD measured − FAD required) ÷ FAD required × 100.
• Specific power (kW per m³/min) = package input power (kW) ÷ FAD (m³/min).
• Electrical input (three-phase) = √3 × line voltage × line current × power factor ÷ 1,000, for a balanced supply; use a power analyser where the supply is unbalanced or a variable-speed drive (VSD) is fitted.
• Unit conversion: 1 m³/min ≈ 35.31 cfm; 1 bar = 14.5 psi.
• Pressure sensitivity: as a rule of thumb, each extra bar of discharge pressure costs about 7% more power, so a test at a higher pressure than the duty point will understate efficiency at the duty point and vice versa.

Never correct a measured value towards the declared value using a factor the agreed method does not include. If inlet conditions fall outside the range the method allows, the result is indicative only and the worksheet should say so.

Worked example

A plant buys an electric oil-injected screw compressor for a duty of 20.0 m³/min (about 706 cfm) at 7.0 bar(g) (about 102 psig), with a declared package specific power of 6.4 kW per m³/min at that pressure. The figures here are illustrative only.

Whether −2.5% flow is acceptable depends entirely on the tolerance agreed beforehand. If the contract references the tolerance bands of the test standard and the result sits inside them, the numerical check passes. If the contract demands the full 20.0 m³/min with no tolerance, it fails. The worksheet does not decide which rule applies; it records the rule and applies it consistently.

Then check the rest of the evidence. If the dryer outlet dew point, the pressure band or a safety trip was not demonstrated, the disposition is hold, even though the flow figure looks satisfactory.

How to interpret the result

The accepted readings become the baseline. Later, when a service engineer finds the machine drawing more power for the same air, the comparison is against this record, not against a brochure.

Limits of a site acceptance test

• Site tests rarely reach laboratory conditions. High Indian summer ambient (often 40–45 °C in Telangana and Andhra Pradesh) affects inlet density, cooling and therefore results; record it and use only the corrections the method allows.
• Flow meters need the straight pipe lengths and installation the maker specifies. A meter fitted close to an elbow or valve can be badly wrong.
• Clamp-meter current readings do not give true power, especially with VSDs or poor power factor.
• Measurement uncertainty must be stated. A 2% deviation measured with instruments uncertain to 3% does not prove a shortfall.
• Air-quality acceptance needs its own sampling method; flow and power tests do not prove ISO 8573-1 class.
• Diesel portable machines need fuel and engine checks as well as air delivery, and should be tested with the hose and tool arrangement the job will use where that matters.

Acceptance after overhaul or airend rebuild

An overhauled machine is not judged against a new-machine catalogue unless that was the agreed target. Before the work, record the pre-overhaul FAD, specific power, discharge temperature and vibration where possible. After the work, test at the same pressure and similar conditions and compare like with like. A rebuild that restores flow but leaves discharge temperature high or vibration rising still needs investigation before release.

For rental machines, the same logic applies in a lighter form: confirm the delivered pressure and flow suit the tools on site, check that safety shutdowns such as high discharge temperature work, and record running hours at hand-over so the service schedule can be tracked.

What an acceptance test proves, and what it does not

A performance acceptance test answers one narrow question: does this compressor, as installed, deliver the agreed free air delivery (FAD) at the agreed pressure for no more than the agreed power, within the agreed tolerance? It is evidence for releasing payment, closing a warranty claim or signing off an overhaul. It is not a life test, a reliability test or an air-quality test, and a pass today says nothing about how the machine will perform after a summer of dusty intake air. Keep the scope of the record matched to what was measured, and record separately anything that was only observed, such as noise, vibration or leaks.

Two tests are often confused. A factory test is done on the maker’s test bed under controlled conditions and instruments, normally to ISO 1217. A site test is done in the plant with the instruments that can be fitted there. A site test can confirm a factory figure within its own uncertainty, but it rarely achieves the uncertainty of a test bed, which is why the method and its uncertainty must be agreed before the test.

Measuring FAD on site: three methods compared

Pump-up worked example (illustrative). A compressor rated 10.0 m³/min fills a 2.0 m³ receiver and 0.3 m³ of isolated pipe, total V = 2.3 m³, from 5.0 to 7.0 bar(g). Site barometric pressure is 0.955 bar(a). The free air pumped in is V × Δp ÷ p(atm) = 2.3 × 2.0 ÷ 0.955 = 4.82 m³. At 10.0 m³/min this takes 4.82 ÷ 10.0 × 60 = 28.9 seconds. A one-second timing error is already 3.5% of the result, a 10% error in the estimated pipe volume is 1.3% of the flow (0.3 of 2.3 m³), and the air in the receiver warms during filling, which makes the compressor look better than it is unless the receiver temperature is measured and corrected for. This is why pump-up is a screening method for large compressors, not an acceptance method.

Correcting measured flow and power to rated conditions

Site conditions never match the rated conditions exactly. ISO 1217 sets out the permitted deviations and the corrections for packaged, electrically driven displacement compressors in Annex C (fixed speed) and Annex E (variable speed). The correction steps below are the ones most often needed on site; apply only those your agreed method allows, and keep the test pressure as close as possible to the rated pressure so that no pressure correction is needed.

Worked example (illustrative). A fixed-speed oil-injected screw compressor is rated 10.0 m³/min FAD at 7.0 bar(g) (7.0 × 14.5038 = 101.5 psig), with a declared package specific power of 6.30 kW per m³/min at a rated motor speed of 2,970 rpm. The test is run at 7.0 bar(g).

1. Pressure. FAD is stated at inlet pressure, so the normal flow is expanded to 0.955 bar(a): factor 1.01325 ÷ 0.955 = 1.0610.
2. Temperature. From 0 °C to 38 °C inlet: factor (38 + 273.15) ÷ 273.15 = 1.1391.
3. Humidity. FAD includes the water vapour drawn in at the inlet, but the meter sees air after the water has been condensed out. Saturation pressure of water at 38 °C is about 0.0663 bar, so the vapour pressure at 60% RH is 0.6 × 0.0663 = 0.0398 bar. Factor 0.955 ÷ (0.955 − 0.0398) = 1.0435.
4. FAD at test conditions. 7.72 × 1.0610 × 1.1391 × 1.0435 = 9.736 m³/min.
5. Speed. For a displacement compressor, flow is close to proportional to speed: 9.736 × 2,970 ÷ 2,955 = 9.785 m³/min. Power at constant pressure is also close to proportional to speed: 63.0 × 2,970 ÷ 2,955 = 63.32 kW.
6. Results. Flow deviation (9.785 − 10.0) ÷ 10.0 = −2.15%. Specific power 63.32 ÷ 9.785 = 6.471 kW per m³/min, which is (6.471 − 6.30) ÷ 6.30 = +2.7% against the declared value. In cfm, 9.785 × 35.315 = 345.6 cfm against a rated 353.2 cfm.

Leaving out the humidity step alone would have understated the FAD by about 4%, enough to turn a pass into a dispute. In monsoon months the vapour pressure is higher still, so always record inlet humidity.

Tolerance bands and the pass or fail decision

The tolerance table in Annex C of ISO 1217 is the reference most contracts use for fixed-speed packaged compressors. The bands below are the values commonly applied from it, with the flow bands converted from the litres-per-second limits in the standard (8.3, 25 and 250 L/s); the tolerances tighten as machine size increases. Confirm the edition and the bands quoted in your contract.

The example above falls in the 1.5 to 15 m³/min band: −2.15% flow is inside ±5% and +2.7% specific power is inside ±6%, so it passes on both counts. Variable-speed machines are judged at several load points, using the tolerances set out in Annex E.

The harder cases are results near a limit. Use this order:

1. Check the evidence is complete: stable conditions, all readings logged, instruments in calibration, and the agreed method followed. If not, the result is a hold, whatever the numbers say.
2. Agree before the test whether the tolerance is taken to cover measurement uncertainty or whether a guard band applies. If a site method is weaker than the standard’s method, a guard band is fair to both sides.
3. If the result is clearly inside the tolerance after allowing for uncertainty, accept.
4. If the result is outside the tolerance but the uncertainty range reaches back inside it, the test cannot decide: treat it as a hold and repeat with a better method, such as a calibrated in-line meter or a nozzle, rather than arguing over the first figure.
5. If the result is outside the tolerance even after allowing for uncertainty, reject, record the deviation and agree the remedy in writing.

In the example, a flow meter uncertainty of ±3% of reading puts the true flow between 9.785 × 0.97 = 9.491 and 9.785 × 1.03 = 10.079 m³/min, that is between about −5.1% and +0.8% of the rating. The lower end lies just beyond the −5% limit, so if the contract requires a guard band, a retest with a better meter would be the right next step before final acceptance.

A test record template

A result that cannot be traced to its readings will not survive a warranty dispute. Record at least the following for every acceptance test.

How Reddy Compressors helps

Reddy Compressors has worked with industry from its office at Tarnaka, Secunderabad, since 2002. As an independent multi-brand specialist (not an authorised dealer of any brand), we commission and check machines of ELGi, Atlas Copco, Ingersoll Rand, KAESER, Chicago Pneumatic and other makes, whether we supplied them or not. We have tie-ups with all major and reputed brands.

• Pre-test planning: help to write a measurable requirement, choose the measurement points and agree the decision rule.
• Site checks: pressure, temperature, power and flow measurement using suitable instruments, with readings and conditions recorded.
• Post-overhaul and post-repair verification against the agreed baseline.
• Commissioning of new and used/refurbished machines we sell, and of rental machines from our own fleet (50–2,000 cfm, 2–30 bar, diesel and electric).
• Service teams of over 20 service engineers and technicians covering the whole of Telangana and Andhra Pradesh, with equipment and spares supplied across India.

We describe what was measured and how; where a contractual test to a standard needs an independent witness or laboratory, we say so.

What to send us

• Make, model, serial number and a nameplate photo of the compressor (and dryer, if in scope).
• The contracted or specified duty: flow, pressure, power and any air-quality class.
• The supplier’s data sheet and any test certificate supplied with the machine.
• Site layout or photos showing where pressure and flow can be measured.
• Whether the test is for new supply, replacement, overhaul or a dispute.

Frequently asked questions

How do I check if my new compressor is giving the rated cfm?

Measure free air delivery at the rated pressure with a calibrated flow method, correct it only as the agreed standard allows, and compare with the data-sheet FAD at that same pressure. Nominal or displacement cfm is not the same figure. Record inlet temperature, barometric pressure and the pressure measurement point, because a reading taken at a different pressure or after the dryer changes the answer.

What is a good specific power for a screw compressor?

There is no single good figure because specific power depends on pressure, size, cooling and control type. Compare the measured package value, in kW per m³/min at the duty pressure, with the maker’s declared value for the same model and pressure. Remember that roughly 7% more power is needed for each extra bar, so compare at equal pressure only.

Which standard is used for compressor performance testing?

ISO 1217 is the widely used standard for measuring the flow and power of displacement compressors, with annexes for different machine types and for simplified acceptance tests. Contracts may also reference a manufacturer’s declared-data scheme. The important point is to name the method and tolerance in the purchase document, so the site test is judged against an agreed rule.

Can a compressor fail acceptance even if the flow is correct?

Yes. Acceptance usually covers more than flow. A machine can deliver the right air yet fail on power input, pressure control, dew point after the dryer, noise, a safety trip that does not operate, or missing documents. If required evidence is incomplete, the correct result is hold until it is provided, not acceptance based on one favourable number.

Is a clamp meter good enough to measure compressor power?

Not for acceptance. A clamp meter reads current, and power also depends on voltage, power factor and phase balance. On machines with variable-speed drives the waveform makes simple calculations unreliable. Use a three-phase power analyser at the package supply terminals, record the readings over a stable period and note the instrument’s calibration status on the worksheet.

How should an overhauled compressor be accepted?

Agree the target before the work starts: restore to the machine’s own earlier baseline, or to a stated flow and power. After the overhaul, test at the same pressure and similar conditions as the earlier readings, check discharge temperature and vibration, and record the results as the new baseline for future maintenance and comparison.

What is the difference between a factory test and a site acceptance test?

A factory test is run on the maker’s test bed, normally to ISO 1217, with fixed instruments, controlled pipe runs and accurate flow measurement such as calibrated nozzles. A site acceptance test is run after installation with the instruments that can be fitted in the plant, usually a flow meter in the discharge line and a power analyser. The site test shows how the machine performs as installed, including the effect of local inlet conditions, but its uncertainty is normally larger. Agree the method, the corrections and how uncertainty will be treated before either test starts.

What is the tolerance on compressor FAD under ISO 1217?

For fixed-speed packaged compressors, the tolerances commonly applied from the table in Annex C of ISO 1217 depend on size: ±7% on flow for machines below 0.5 m³/min, ±6% from 0.5 to 1.5 m³/min, ±5% from 1.5 to 15 m³/min and ±4% above 15 m³/min. Specific energy tolerances are one percentage point wider in each band. Variable-speed machines are judged at several load points using Annex E. Always confirm which edition and which bands your contract or data sheet refers to.

Why does a compressor deliver less air in summer?

FAD is stated as volume at inlet conditions, so a well-maintained compressor delivers roughly the same volume of inlet air all year. What falls in hot weather is the mass of air in that volume: at 45 °C there is about 8% less air by mass in each cubic metre than at 20 °C, and humid air carries water vapour that is later condensed out. The plant receives less usable dry air even though the compressor meets its rating. High inlet temperature can also cause overheating trips. Correct test results to inlet conditions before concluding that the machine is short.

Related pages

• Compressor commissioning and final acceptance: the 40-checkpoint method
• FAD and reference condition conversion tool
• Specific power comparison calculator
• Measurement uncertainty worksheet
• Motor input and electrical measurement worksheet
• Compressed-air commissioning, start-up and acceptance
• Airend post-rebuild validation and acceptance