Four scales for one piece of equipment
An air conditioner in the United States can be quoted four ways, and buyers routinely compare numbers from different scales as though they were the same thing.
EER is efficiency at a single operating point: BTU of cooling per watt-hour of electricity, measured at 95 °F outdoors with 80 °F dry-bulb, 67 °F wet-bulb air entering the coil. It answers "how much power does it draw when it is working hard?"
SEER is a weighted average across a standard cooling season, including part-load operation and cycling losses. It answers "how much electricity does it use over a summer?" Because most equipment spends most of its life below full load, and because part-load operation is usually more efficient, SEER is always the larger of the two for the same machine.
SEER2 and EER2 are the same two quantities measured under the Department of Energy's Appendix M1 test procedure, which took effect on 1 January 2023. M1 raises the external static pressure the equipment must work against, so it reflects a realistic duct system rather than a laboratory-perfect one. The same machine scores lower on the new scale — which is why a "14.3 SEER2" unit today is roughly the machine that used to be labelled 15 SEER.
COP and kW/ton are the same information again in engineering units. COP is dimensionless — watts of cooling per watt of input — and is what you will see outside the United States. kW/ton is the reciprocal expressed per ton of cooling and is how commercial chillers are quoted; the chiller kW per ton calculator works in those units throughout.
How each conversion works, and which ones are exact
Not all four conversions have the same standing, and it matters which you rely on.
COP and kW/ton from EER are exact. They are pure unit conversions. One watt is 3.412142 BTU/h, so COP = EER ÷ 3.412142. One ton is 12,000 BTU/h, so kW/ton = 12 ÷ EER. No approximation is involved and no test data is needed.
SEER2 from SEER is an approximation. AHRI publishes conversion factors for translating legacy ratings onto the new scale: multiply SEER by 0.95, EER by 0.95, and HSPF by 0.85 for split systems. These exist so that an installed base of equipment can be discussed on one scale. They are not test results, and the true M1 figure for a given model comes only from running the M1 test.
EER from SEER is an empirical correlation. This is the weakest link, and it has to be, because SEER and EER are separate physical measurements. Two units with identical SEER can have genuinely different EER depending on how their capacity modulation is arranged. The relationship used here — EER = −0.02 × SEER² + 1.12 × SEER — is a curve fitted to residential equipment data and used in building energy modelling. It reproduces the right behaviour: EER rises with SEER but ever more slowly, because a growing share of a high SEER rating comes from part-load performance the EER test never sees.
That curve has a peak. Its maximum is at SEER 28, where EER reaches 15.68, and beyond that the fit turns over. Practically this means the reverse conversion has no solution for an EER above about 15.7: real equipment can be rated there, but you cannot infer its SEER from the curve, and this calculator says so rather than inventing a number.
Worked example: comparing a SEER2 quote against a SEER quote
One contractor quotes a 3-ton system at 15.2 SEER2. Another quotes a 3-ton system at 16 SEER. Which is more efficient?
- Put the SEER2 quote on the old scale. 15.2 ÷ 0.95 = 16.0 SEER. The two quotes are the same efficiency.
- Find the full-load EER. −0.02 × 16² + 1.12 × 16 = −5.12 + 17.92 = 12.80.
- EER on the current scale. 12.80 × 0.95 = 12.16 EER2.
- Coefficient of performance. 12.80 ÷ 3.412142 = 3.751 — the system moves 3.75 units of heat per unit of electricity.
- Kilowatts per ton. 12 ÷ 12.80 = 0.9375 kW/ton.
- Power draw at 3 tons. 3 × 0.9375 = 2.813 kW, or 36,000 ÷ (12.80 × 1,000), which is the same arithmetic.
The comparison is now honest, and it exposes that the two quotes differ on something other than efficiency. Work the running cost from the SEER figure with the air conditioner running cost calculator and you get 2,304 kWh for a 960-hour season on either machine.
Which number to use for which decision
Use SEER or SEER2 for annual energy and utility bills. It is the only one of the four that accounts for the fact that equipment cycles and modulates. It is also the number rebate programmes and building codes are written against.
Use EER or EER2 for peak demand and electrical sizing. Peak power draw determines breaker and conductor sizing, generator capacity, and any demand charge on a commercial tariff. It also determines how the equipment behaves on the hottest afternoon of the year, which is precisely when SEER's part-load credit does not apply. In hot-dry climates the DOE sets separate minimum EER2 requirements for this reason — a machine can post a good seasonal number and still be a poor performer at design conditions.
Use COP when comparing across technologies. Heat pumps, chillers, absorption equipment and heat recovery are all quoted in COP in the engineering literature, and it is the only dimensionless figure of the four. A COP of 3.75 means the same thing in any unit system.
Use kW/ton for commercial and chilled-water work. Central plant efficiency is universally quoted this way, and lower is better — which trips up anyone moving from SEER, where higher is better. 0.9375 kW/ton and EER 12.8 are the same statement.
One caution about the correlation. The EER shown from a SEER input is a modelling estimate. If the AHRI certificate for your model lists an EER2, that figure is a measurement and should be used in its place — particularly for anything involving electrical sizing, where an estimate that runs 10% optimistic is a real problem.
Conversion table across the common rating range
| SEER | SEER2 | EER | EER2 | kW/ton | COP |
|---|---|---|---|---|---|
| 13 | 12.35 | 11.18 | 10.62 | 1.0733 | 3.277 |
| 14 | 13.30 | 11.76 | 11.17 | 1.0204 | 3.447 |
| 15 | 14.25 | 12.30 | 11.69 | 0.9756 | 3.605 |
| 16 | 15.20 | 12.80 | 12.16 | 0.9375 | 3.751 |
| 17 | 16.15 | 13.26 | 12.60 | 0.9050 | 3.886 |
| 18 | 17.10 | 13.68 | 13.00 | 0.8772 | 4.009 |
| 20 | 19.00 | 14.40 | 13.68 | 0.8333 | 4.220 |
| 22 | 20.90 | 14.96 | 14.21 | 0.8021 | 4.384 |
| 24 | 22.80 | 15.36 | 14.59 | 0.7813 | 4.502 |
Read the flattening in the EER column: SEER rises by 11 points across this table while EER rises by 4.2. That gap is the part-load credit, and it is the whole reason the two numbers cannot be treated as one.
Where these conversions go wrong
- Comparing a SEER2 number against a SEER number. The most common error since 2023, and it makes newer, better equipment look worse. Convert first.
- Treating the SEER-to-EER correlation as a definition. It is a fit to a population of equipment. Two units at the same SEER can differ by a point of EER depending on how they modulate.
- Using the 0.95 factor on a heating rating. HSPF converts at approximately 0.85, not 0.95, because the M1 changes bite harder in heating. Applying the cooling factor overstates HSPF2 by about 12%.
- Sizing a breaker from a converted EER. Electrical sizing comes from the minimum circuit ampacity on the equipment nameplate, not from any efficiency rating. Efficiency figures are averages of a tested condition, not worst-case draws.
- Forgetting that lower is better in kW/ton. Moving between SEER and kW/ton inverts the direction of “good”. 0.78 kW/ton is excellent; 0.78 SEER would be nonsense.
- Assuming rated efficiency is delivered efficiency. All four ratings assume correct airflow, correct charge and duct losses within the tested allowance. A poorly installed 20 SEER system can use more electricity than a well installed 14 SEER one.
Key terms
- SEER / SEER2
- Seasonal energy efficiency ratio — total BTU of cooling over a standard season divided by total watt-hours of electricity. SEER2 is the same quantity under the 2023 Appendix M1 test procedure.
- EER / EER2
- Energy efficiency ratio — BTU per watt-hour at a single full-load rating point, 95 °F outdoors with 80/67 °F entering air. EER2 is the M1 version.
- COP
- Coefficient of performance — heat moved divided by work in, both in the same units. COP = EER ÷ 3.412142, so a COP of 1.0 is EER 3.412.
- kW/ton
- Kilowatts of electrical input per ton of refrigeration, the standard efficiency unit for chillers and central plant. Lower is better, and kW/ton = 12 ÷ EER.
- Appendix M1
- The DOE test procedure effective 1 January 2023 that defines SEER2, EER2 and HSPF2. Its main change is a higher external static pressure representing real duct systems.
What the ratings cannot tell you
Efficiency ratings describe a machine in a laboratory. Three things that matter as much are outside them.
Sizing. An oversized system cycles more, dehumidifies less, and delivers worse comfort at higher cost regardless of its rating. Size from a load calculation. A correctly sized 15 SEER system beats an oversized 20 SEER one on both bill and comfort.
Installation. The rating assumes design airflow, correct refrigerant charge and a duct system no worse than the test bench. Field studies of installed residential systems consistently find airflow and charge outside manufacturer tolerance on a large fraction of them, and both cost efficiency directly. Checking the temperature split and the refrigerant charge after installation protects the rating you paid for.
Climate. SEER is weighted for a standard season. If you live where cooling runs 2,000 hours a year, seasonal efficiency dominates and a high SEER pays back. If cooling runs 300 hours but every one of them is at design condition, full-load EER matters more than the season average, which is exactly why the DOE writes separate EER2 minimums for the hot-dry region.
Once you have the ratings on a common scale, the useful next step is money rather than ratios: the running cost calculator turns a SEER into a seasonal bill, and the SEER upgrade savings calculator compares two ratings against the price difference between them.