← Engineering workspace

Cooling capacity, condenser heat rejection, compressor power and COP

Calculation methods • required data • formulas • worked examples

“CPO” has been interpreted as COP: coefficient of performance.

This is an engineering reference, not a new automatic sizing calculator. Use the linked workspace tools for state properties and design calculations. CoolProp property calculator and manual · Refrigeration cycle · Psychrometric studio · Compressor selection

1. Choose the method that matches your data

Available informationBest useMain limitation
Refrigerant mass flow and inlet/outlet enthalpiesEvaporator and condenser thermal duties; compression energy transferred to refrigerant.Properties alone do not give capacity without mass flow.
Water/glycol flow and entering/leaving temperaturesMeasured evaporator or condenser duty.Use actual fluid concentration, density and specific heat.
Airflow, dry bulb and humidity at both coil facesTotal air-side cooling; sensible dry-condenser heat rejection.Airflow and humidity uncertainty can dominate.
Measured electrical kW plus cooling dutyActual electrical COP and approximate heat rejection.Match measurement time and equipment boundary.
OEM capacity/power mapsCompressor and condenser selection.Match refrigerant, speed, evaporating/condensing conditions, SH/SC and rating basis.
Loads, UA, or expected COP onlyPreliminary design and sensitivity estimates.Not proof of equipment performance.

Use kW for a rate and kWh for energy. Define whether cooling is gross coil duty or net useful cooling, and whether power includes only the compressor or all fans, pumps, controls and other equipment.

2. Refrigerant energy-balance method

Simple steady single-stage cycle: 1 evaporator outlet/compressor inlet; 2 compressor outlet/condenser inlet; 3 condenser outlet/valve inlet; 4 valve outlet/evaporator inlet. This shared numbering assumes negligible connecting-line heat exchange and pressure losses. Otherwise use each component’s actual inlet/outlet states.

Cooling duty: Qe = ṁr(h1 − h4)
Condenser duty: Qc = ṁr(h2 − h3)
Refrigerant compression energy rate: Wr = ṁr(h2 − h1)
Adiabatic throttling valve: h4 ≈ h3
Simple-cycle balance: Qc = Qe + Wr

ṁr is refrigerant mass flow in kg/s; h is specific enthalpy in kJ/kg; results are kW. In imperial units, lbm/h × Btu/lbm gives Btu/h. Neglect kinetic/potential energy changes and storage. Use consistent composition and enthalpy reference. Suction-line heat gain outside the cooled space is not useful cooling.

For an electrical boundary, Qc ≈ Qe + Wcomp,electric only if essentially all that electrical input reaches the condenser through the refrigerant. More generally, for a defined steady boundary: Qc = Qe + Win + Qother,in − Qother,out. Account for compressor-shell/oil-cooler heat loss, line heat exchange and external motor/drive losses. Do not automatically add condenser-fan watts to refrigerant-side condenser duty. Room heat rejection and condenser refrigerant duty are different boundaries. [1]

Mass-flow alternatives: ṁr = Qe/(h1−h4), when Qe is independently known; or a preliminary displacement model ṁr ≈ ηv V̇swept/vsuction. ηv is volumetric efficiency and v is suction specific volume. Displacement alone is not capacity; efficiency varies with pressure ratio, speed and compressor type.

3. Measure heat on the water or air side

Water / glycol

Q = ṁ cp ΔT = ρ V̇ cp ΔT
Evaporator: ΔT = Tin − Tout
Condenser: ΔT = Tout − Tin

ρ in kg/m³, V̇ in m³/s, cp in kJ/(kg·K), ΔT in K gives kW. If flow is m³/h, divide by 3600. With significantly varying properties use ṁ(hout−hin) for heating and the reverse for cooling. For water near common HVAC conditions: Q(kW) ≈ 4.18 × flow(L/s) × ΔT(K). Imperial water approximation: Q(Btu/h) ≈ 500 × US gpm × ΔT(°F); 500 ≈ 60 × 8.33 × 1. It is not a glycol or universal-temperature constant. Locate sensors to exclude unwanted pipe and pump heat. [4]

Air-side total and sensible cooling

Open the complete air-side calculation guide: enthalpy, humidity ratio, RH, SHR, BF, ADP, condensate and worked example →

Total cooling, common approximation: Qtotal ≈ ṁdry-air(hair,in − hair,out)
Sensible cooling: Qsensible ≈ ṁair cp,air(Tin − Tout)
Latent cooling estimate: Qlatent ≈ Qtotal − Qsensible

Air enthalpy is per kg of dry air; use matching dry-air mass flow (V̇/v, with specific volume per kg dry air). Measure humidity as well as temperature for wet-coil total cooling. For a precise wet-coil balance subtract condensate enthalpy flow: Qcoil = ṁda(hin−hout) − ṁcond hliquid,cond, with consistent references and condensate conditions. Standard-air shortcuts: total Btu/h ≈ 4.5 × cfm × Δh(Btu/lb dry air); sensible Btu/h ≈ 1.08 × cfm × ΔT°F. Density, altitude and moisture affect these factors. [3]

For a dry air-cooled condenser use air mass flow × cp × (Tout−Tin), allowing for fan heat between measurement planes. Evaporative condensers/cooling towers require a moist-air and water mass/energy balance; dry-bulb temperature rise alone is insufficient.

4. Different ways to obtain compressor power

MethodFormula / inputsInterpretation
Power analyzerRead real electrical kW at the compressor supply.Preferred field measurement. Use appropriate instrumentation for variable-speed drives and distorted waveforms.
Voltage/current/power factorSingle phase: WkW = Vrms Irms PF /1000.
Balanced three phase: WkW = √3 Vline-line Iline PF /1000.
Use true PF and operating current. For unbalanced supplies measure total real power across phases. RLA/LRA and nominal voltage do not determine running kW; volts × amps alone is apparent power. [5]
Energy meter over an intervalWaverage(kW) = ΔE(kWh)/Δt(h).Includes cycling/standby during that interval; not necessarily on-cycle power.
Refrigerant enthalpy riseWr = ṁr(h2−h1).Equals shaft input only for the assumed adiabatic compressor boundary. With outward compressor heat loss, Wshaft = Wr + Qloss. Do not call Wr measured electrical power.
Isentropic estimateh2s = h(P2,s1); ηis = (h2s−h1)/(h2−h1); h2 = h1+(h2s−h1)/ηis.For a stated adiabatic model. ηis must be validated; calculate Wr from the result. If separate efficiencies are applicable, Welectric ≈ Wshaft/(ηmotor ηdrive). Avoid double-counting losses already included in an OEM efficiency. [2]
Known COP or EERW(kW) = Qe(kW)/COP.
W(W) = Qe(Btu/h)/EER.
Same conditions and boundary required. An assumed COP is an estimate, not an independent check.
OEM performance dataRead capacity and power at the actual rating conditions; use approved interpolation/software.Do not extrapolate beyond the compressor envelope or substitute motor horsepower for refrigeration capacity.

5. COP, EER and heat-rejection ratio

Cooling COP = Quseful,cooling / Winput
Heating COP = Quseful,heating / Winput
Compressor-only COP = Qe / Wcomp,electric
System COP = Qnet / (Wcomp + Wfans + Wpumps + Wcontrols + other included power)
EER [Btu/(W·h)] = 3.412142 × COP
kW/ton = 3.516853 / COP
HRR = Qc/Qe ≈ 1 + 1/COP

Use the same power units in a COP ratio. The HRR expression requires the simple Qc=Qe+W balance and its matching COP; it is not valid with an unrelated seasonal or system COP. COPheating = COPcooling + 1 only for the same loss-free balance and denominator. EER conversion applies to the same operating point and boundary; SEER/IPLV are not interchangeable with instantaneous COP. [6]

For a changing load, interval COP = total useful cooling energy / total input energy over the same interval. Do not take an unweighted average of instantaneous COPs. Carnot cooling limit: COPCarnot = Tcold/(Thot−Tcold), with absolute kelvin reservoir temperatures. It is an ideal upper bound, not an equipment prediction. [2]

6. Design-stage methods

Load calculation: Qrequired = transmission + infiltration/ventilation + product pull-down + people + lights + internal fans + other coincident loads. Transmission commonly uses UAΔT; product sensible cooling uses m cp ΔT/t; freezing also needs latent energy mL/t. Daily energy divided by available runtime gives required average on-cycle capacity, but separately check peak duty, defrost recovery and margins. Required load is not measured compressor output.

Heat-exchanger conductance: Q = U A F ΔTlm, where ΔTlm = (ΔT1−ΔT2)/ln(ΔT1/ΔT2). U is overall heat-transfer coefficient, A its matching area, F a configuration correction, and ΔT1/ΔT2 the positive terminal driving differences. If they are equal, ΔTlm equals that common difference. With U in W/(m²·K), the result is W. U must reflect flow, fouling and geometry. Segment desuperheating, condensation and subcooling, especially for blends; one assumed condensing temperature can misrepresent the exchanger. [7]

Manufacturer maps: a common ten-coefficient form is Y=C1+C2Te+C3Tc+C4Te²+C5TeTc+C6Tc²+C7Te³+C8TcTe²+C9TeTc²+C10Tc³. Use it only when the OEM explicitly specifies this ordering, units, coefficient set and rating basis; Y may be capacity, power or mass flow, each with different coefficients. No coefficients or universal correction factors are supplied here.

Heat-rejection allowance: use actual rated Qe and compressor power rather than a universal percentage. For example, COP=2 implies HRR≈1.50; COP=4 implies HRR≈1.25 under the simplified balance. Low-temperature applications can have substantially different ratios.

7. Worked examples and verification

Illustrative cycle (not a refrigerant rating): ṁr=0.05 kg/s, h1=400, h2=450 and h3=h4=250 kJ/kg. Qe=7.5 kW; Wr=2.5 kW; Qc=10 kW; cycle COP=3; HRR=1.3333. These assumed enthalpies demonstrate arithmetic, not a physical model for a named fluid.

Water-side check: 0.5 kg/s water × 4.18 kJ/(kg·K) × 5 K = 10.45 kW. Wet-coil approximation: 0.8 kg dry air/s × 15 kJ/kg dry air = 12 kW before condensate and boundary corrections.

Electrical check: 230 V × 10 A × PF 0.85 /1000 = 1.955 kW (single phase). System versus compressor COP: 10 kW useful cooling with 3 kW compressor and 0.5 kW auxiliaries gives compressor-only ratio 3.333 and system COP 2.857, provided the cooling numerator has the stated consistent boundary.

Useful uncertainty check: for independent small measurement errors, uQ/Q ≈ √[(uṁ/ṁ)²+(ucp/cp)²+(uΔT/ΔT)²]. A small measured ΔT amplifies temperature-sensor uncertainty. Compare independently measured heat rejection against the balance; a zero residual calculated from the same inputs is not independent validation.

Record refrigerant/composition, operating point, sensor locations, gross/net convention, power boundary, time interval, flow basis, instruments, uncertainty, OEM model/version and all assumptions. Superheat/subcooling and pressure/temperature states alone cannot establish mass flow or kW capacity.

Units and symbols

Q = heat-transfer rate; W = power rate (not the unit watt when used as a variable); ṁ = mass flow; h = specific enthalpy; s = specific entropy; cp = specific heat; ρ = density; V̇ = volume flow; η = efficiency; PF = true power factor; UA = thermal conductance; L = latent heat; t = time; Δ = difference. 1 kW = 3412.142 Btu/h; 1 refrigeration ton = 12,000 Btu/h = 3.516853 kW. 1 mechanical hp ≈ 0.7457 kW shaft power; electrical input also depends on losses. Temperature differences: 1 K = 1°C = 1.8°F difference.

8. Research sources

Reviewed October 17, 2026. Formulas are an engineering synthesis; sources support the named methods, not certification of a particular installation. Manufacturer rating conditions and current applicable test standards govern comparisons.

  1. Copeland — Fundamentals of Refrigeration: condenser heat rejection and compressor input. Search-index text reviewed; PDF retrieval was unavailable during this review.
  2. CoolProp — Simple compression cycle: cycle COP, isentropic efficiency input and Carnot bounds.
  3. Trane TRACE 3D Plus — Load design calculations: air enthalpy and airflow method.
  4. Delta Cooling Towers — Selecting, sizing and design: water-flow heat balance and the 500 factor.
  5. Fluke — Low power factor: real versus apparent power. Fluke — Power harmonics: waveform and instrumentation considerations.
  6. U.S. DOE FEMP — Heating and cooling efficiency definitions: EER and COP.
  7. NPTEL — Heat exchanger temperature difference: LMTD and phase-change considerations.