Skip to main content
PM Assist

Chiller efficiency converter

COP, EER and kW per ton are three ways of saying the same thing, and UK engineers meet all three. Convert between them, or work the efficiency out from a measured duty and absorbed power.

COP

4.5000

Dimensionless

EER

15.355

Btu/h per W

kW/ton

0.7815

Lower is better

Method and sources

Equations used

COP = cooling output (kW) ÷ power input (kW)

Dimensionless.

EER = COP × 3.412 1416

EER is cooling output in Btu/h per watt of input, so the factor is Btu(IT)/h per kW ÷ 1000.

kW/ton = 12 ÷ EER

12,000 Btu/h per ton ÷ 1,000 W per kW. The only one of the three where a lower number is better.

kW/ton = 3.516 852 842 ÷ COP

The same relationship expressed against COP.

Constants assumed

SymbolValueSource
1 ton of refrigeration12,000 Btu(IT)/h = 3,516.852 842 WNIST SP 811 App. B.9 (prints 3.516 853 E+03 W); derived from the exact Btu(IT)
1 Btu(IT)1,055.055 852 62 J (exact)International Table definition; NIST SP 811 App. B.9
1 kW3,412.1416 Btu(IT)/hDerived from the Btu(IT) definition

Sources

  • •NIST Special Publication 811, Appendix B.9 — conversion factors.
  • •The ton of refrigeration and the Btu(IT) are definitions, not measurements, so the conversions above are exact rather than approximate.

What this calculator deliberately does not do

IPLV and NPLV part-load values are not calculated.

Sources give two incompatible forms — an arithmetic weighting (0.01A + 0.42B + 0.45C + 0.12D) and a reciprocal one — which differ by around 6% on a realistic chiller. ANSI/AHRI 550/590 is not openly published and could not be consulted to settle which applies to the current edition, so no part-load figure is offered rather than one that might be quietly wrong.

These conversions are exact arithmetic between defined units. What they cannot tell you is whether two efficiency figures are comparable: chiller efficiency depends on the water and condenser temperatures it was rated at, so always check the rating conditions before comparing machines.

Frequently asked questions

What is the difference between COP, EER and kW/ton?
They are three ways of stating the same thing. COP is cooling output divided by power input, both in kilowatts, so it is dimensionless. EER is cooling output in Btu/h divided by power input in watts, which makes it COP multiplied by 3.4121416. kW/ton is the electrical input in kilowatts for each ton of refrigeration produced — the only one of the three where lower is better. UK datasheets usually quote COP; US ones quote EER or kW/ton, and any chiller selected through a US manufacturer's software will report all three.
How many kW is a ton of refrigeration?
3.516853 kW. A ton of refrigeration is defined as 12,000 Btu(IT) per hour, and the International Table Btu is exactly 1,055.05585262 joules, so the figure is exact rather than measured: 12,000 × 1,055.05585262 ÷ 3,600 = 3,516.852842 W. NIST Special Publication 811 prints it as 3.516853 × 10³ W. A value of 3.5168525 circulates widely and is wrong in the eighth significant figure.
Why does this converter not calculate IPLV?
Because the sources disagree on the formula and the disagreement is material. The arithmetic weighting — 0.01A + 0.42B + 0.45C + 0.12D — appears in a professional-engineer-stamped compliance submission citing ARI 550/590 section 5.4.1, while current commentary gives a reciprocal form, 1 ÷ (0.01/A + 0.42/B + 0.45/C + 0.12/D). Both cannot be right for a COP or EER input; they differ by roughly 6% on a realistic chiller. ANSI/AHRI 550/590 is not published openly, so the current edition could not be checked. A part-load figure that is quietly 6% out is worse than no part-load figure, so this tool does not offer one.
Is a higher COP always better?
At the same operating conditions, yes — but a COP quoted without its conditions is not comparable. Chiller efficiency depends heavily on the chilled water and condenser temperatures it was measured at, so a machine rated at one set of conditions can look far better than a machine rated at another while performing worse on your site. Check what conditions the figure was taken at before comparing two machines, and remember that full-load efficiency says little about how a chiller behaves at the part load it spends most of the year at.