Heat pumps · Chillers · Sludge drying systems
Knowledge

How to Read a Heat Pump COP (and Check the Number Yourself)

The rating conditions behind the number

No COP figure means anything without its test point. Two standards dominate, and they are not identical.

Parameter EN 14511 (typical hot water rating) GB/T 21362 (China, hot water units)
Air entering dry bulb 7 °C 20 °C (nominal) or 7 °C (low temperature)
Air entering wet bulb 6 °C —
Water inlet 40 °C (heating) or 30 °C (cooling) 15 °C or 20 °C
Water outlet 45 °C 55 °C
Water flow Nominal, within ±5% Nominal, within ±5%
Steady-state period 30 min minimum 30 min minimum

The consequence is easy to see. A unit rated COP 4.2 at 20 °C air and 15→55 °C water will not produce 4.2 when the air is 2 °C and the water is 40→60 °C. The reduced evaporating temperature and increased condensing temperature together can take 30%–40% off the figure.

How COP moves with the two temperatures

COP is governed by the temperature lift — condensing temperature minus evaporating temperature — not by either temperature alone. A rough Carnot-based estimate for a real machine is:

COP ≈ 0.45 × T_evap (K) ÷ (T_cond (K) − T_evap (K))

where temperatures are absolute. For an evaporating temperature of 0 °C (273 K) and condensing temperature of 65 °C (338 K), the ideal COP is 4.2, and with a second-law efficiency of 0.45 the real figure is about 1.9 — which matches what a heat pump does at −5 °C ambient producing 60 °C water.

Ambient air Water out Typical COP Comment
20 °C 55 °C 3.8–4.3 Nominal rating point
7 °C 55 °C 3.1–3.5 Common winter average
2 °C 55 °C 2.7–3.1 Defrost band, frequent cycles
−5 °C 55 °C 2.3–2.7 Design condition in much of northern China
−15 °C 55 °C 1.8–2.2 Cold-climate models
20 °C 65 °C 2.9–3.3 High-temperature delivery costs COP
20 °C 45 °C 4.4–4.9 Underfloor heating, low delivery

The last two rows are the reason the same unit can be quoted as COP 4.6 or 3.1 depending on the application. A heat pump is not efficient or inefficient in the abstract; it is efficient at a low lift and poor at a high one.

Correcting a published COP to your site

Three corrections turn a datasheet figure into something usable for an energy estimate:

  1. Water temperature. Every 5 K increase in delivery temperature costs roughly 8%–12% of COP. Going from 55 °C to 60 °C is not free, and it is often done to satisfy a legionella regime that could be met by a weekly pasteurisation instead.
  2. Ambient. Use the bin distribution for the site, not the design day. A plant that runs 3,000 hours a year spends most of them between 5 °C and 20 °C, where COP is 3.2–3.8, not at the design point of −5 °C.
  3. Defrost and part load. Deduct 5%–8% for defrost in a temperate winter, and check the part-load performance separately — a unit that is 30% oversized will spend its life at 25% load, where fixed-speed machines lose 10%–15% of COP.

A worked example makes the difference concrete. A 40 kW unit rated COP 3.9 at 20 °C serves a load that needs 55 °C water all year in a city with a 2,800-hour heating season:

Basis COP Annual electricity for 120,000 kWh heat
Datasheet figure 3.90 30,800 kWh
Ambient-corrected bin average 3.30 36,400 kWh
Less 6% defrost allowance 3.10 38,700 kWh

The gap between the first and last row is 26%. An energy-saving estimate built on the datasheet figure will overstate the saving by a quarter, which is exactly the sort of claim that ends up in a dispute two years after commissioning.

Verifying COP on site

A site test does not need a laboratory, but it does need steady conditions and four measurements:

Measurement Instrument Accuracy needed Where
Water flow Ultrasonic or clamp-on meter ±3% Main return
Water inlet temperature Calibrated probe ±0.2 K Before unit
Water outlet temperature Calibrated probe ±0.2 K After unit
Electrical input Power analyzer ±1% Unit supply, all phases

Then COP = 4.187 × flow (m³/h) × ΔT (K) ÷ 3,600 ÷ kW_electrical. For 5.2 m³/h and a 7.8 K rise, heat is 4.187 × 5.2 × 7.8 ÷ 3.6 = 47.2 kW; at 12.1 kW input the measured COP is 3.90.

Four rules make the test meaningful:

  • Steady state first. Ten minutes at constant flow and stable temperatures; a unit still recovering from defrost will read 20% low.
  • Match the rating point or state the deviation. A test at 20 °C ambient and 55 °C delivery is comparable with the datasheet; a test at 4 °C ambient is not, and the result should be reported with the conditions.
  • Account for the pump. If the pump is inside the unit boundary, include its power; if outside, state that the figure is unit COP.
  • Log the defrost state. Note whether the unit defrosted during the test window. A 15-minute test that included a defrost cycle is not a valid steady-state measurement.

Where COP stops being the right metric

COP describes one point. For anything that runs across a season, the part-load metric is the one that predicts the bill — integrated part-load value for chillers, seasonal performance factor or annual efficiency for heat pumps. A unit with a slightly lower nominal COP but a flat part-load curve can use less electricity over a year than a higher-COP machine that falls away at 40% load.

That comparison is set out separately in our guide to COP versus IPLV, which explains why the two numbers can rank the same two machines in opposite orders.

Longkang Environmental Equipment Co., Ltd. publishes capacity and COP at 20 °C, 7 °C and the design ambient for each model, with the water temperatures and flow stated, so the figures can be corrected and later verified rather than taken on trust. Phone: +86 400-878-0280 | +86 130-0200-1280

Questions about COP

Is a higher COP always better?

At the same rating conditions and the same delivery temperature, yes. Across different conditions, no — a COP quoted at 20 °C/45 °C is not comparable with one at 7 °C/60 °C, and the second machine may be doing harder work.

Why is my measured COP lower than the catalog?

Usually because the site conditions differ: colder air, hotter water, lower flow, or a unit that defrosted during the test. Check the water flow first — 20% low flow reads as 20% low capacity and roughly 15% low COP.

Does COP include the water pump?

Usually not; the pump is outside the unit boundary. A circulation pump drawing 1.5 kW on a 30 kW unit adds 5% to the input and drops system COP by about 0.2. Ask for both figures.

What COP should I expect in winter?

At −5 °C ambient and 55 °C delivery, 2.3–2.7 for a standard unit and 1.8–2.2 at −15 °C. If a supplier quotes COP 4 in January at −10 °C, ask for the test conditions in writing.

Is COP the same as energy efficiency ratio?

No. EER is used for cooling and is the ratio of cooling delivered to electricity input. For a reversible unit the two describe opposite modes; the coefficient of performance is the general term.

Can I improve the COP of an existing unit?

Yes, within limits: lower the delivery temperature if the application allows, clean the coil, restore the water flow, insulate the pipes and tank, and stop the unit short-cycling. Each of those is worth 3%–10%, and together they often recover more than a decade of refrigerant technology improvements.

Related reading

Frequently asked questions