Cold weather COP of air source heat pumps is about 3.1

The COP of air source heat pumps during cold weather is still about 3.1 even allowing for defrosting energy? Jez BRE, BrendanUys – Heacol


Dear Clavertons – what is the cold day COP of a heat pump including all defrosting energy?

Dave Andrews


In my experience it is not the defrosting energy that impacts on the efficiency of an air source heat pump. Rather it is the effect of the gradual frost build up on the evaporator which mostly reduces the efficiency. The data from my system (with a design flow temperature of 35°C) gives the typical electric energy for a reverse cycle defrost at between 15 and 20 Wh for my 8kW Nibe. For a typical compressor cycle in typical foggy sub zero UK conditions my heat pump will be running at with ~40 minute long compressor cycles at around 2kW electric input for a heat output that starts at around 7 to 8 kW at the start of the cycle, falling to around 6kW at the end of the cycle. The typical daily COP (SPFH2) for my system on a winters day averaging 0°C outside temperature is 3.3.

There is plenty that manufacturers could do to improve frosting and defrosting behaviour for air source heat pumps. Better frost detection sensors and algorithms for a start, so the system only goes into defrost when it needs to. There is research going on to develop better evaporator fin designs and coatings that reduce frost build up. There is also work to develop air pre-conditioning units that remove the moisture before the air enters the evaporator. Some of these technologies will make their way onto products eventually.

The plot below shows a typical compressor cycle for my heat pump on a cold moist winter’s day in November. The outside temperature over the period is between -1°C and -2°C. The pink is the heat output and the purple/blue the electric input (left hand axis in W). The yellow line is the external temperature (right hand axis). The little blue blip at around 10:15 is the electric used by the reverse cycle defrost. For the period shown in the graph, integrating the data gives the heat output of 4.51 kWh for 1.43 kWh electric input (including the defrost) – giving a COP for the full cycle of 3.15.


Please note that I work part time. My normal work days are Wednesday, Thursday and in the morning on Friday. If you contact me outside of these times, I will not be able to provide a response until I am next in the office.

Dr Jez Wingfield

Senior Technician – Physical Building Performance

UCL Energy Institute

Central House, 14 Upper Woburn Place, London WC1H 0NN

email:  withheld

“In God we trust; all others must bring data” W.E. Deming

3 comments on “Cold weather COP of air source heat pumps is about 3.1

  1. Hi Mark

    I must say, I dissagree with your assunption that an air source has an anual average performance of 3. Non of mine do and I have installed a lot. In my view the only way you can get those levels of performance is eaither very poorly designd equipment or very poor system design. On larger installations we find that the performance levels are in the regeon of SCOP H5 4.5 (this includes all pumping, disrtibutoon and controll costs to the heated area) on both air and groundsource, and better on watersource. It is not difficult to achieve but installation costs are higher and plant costs are generally higher as I get the air source units designed specifically for our temperate climate, therefor ther is very little frosting problems and better performance at lower temperatures. If you use the standard cooling optimised units, you are just asking for trouble. with groundsourse we increase the loop lenght to raise the temperature or collect solar thermal modules to it to improve performance (I have some trials running as er speak). The installation costs are higher but with the current energy prices they pail in to insinicance when looking long term.

    please can you elaborate on your table, I cannot get the maths.

    Thank you
    Kind Regards
    Brendon Uys

  2. Brendon, how does one install heat pumps properly so that they have good COPs round the year and particularly not too low during cold spells?

    Yes, it is possible, and it is simpler than most think. Here is a list of the points:

    Continuous heating.
    Correct emitter seizing (size radiators to 45 Deg C and under floor to 35 Deg C), not difficult with the right calculator.
    High flow rates, at least 2 l/min / kw
    System volume must be at least 15 L / minimum output of the unit. If insufficient, a volumiser must be used.
    Open zone except in the sleeping areas.
    Do not use on-off thermostats, if thermostats are essential, then use modulating units.
    The internal temperature / flow temperature controlled by weather compensation and or load compensation.
    No buffer tanks
    No system separation
    If large heat storage is required, use large volumisers, NOT buffer tanks.
    This will work on all heat pumps, air or ground, and will facilitate the lowest possible running costs for that system.

    Thank you
    Kind Regards
    Brendon Uys

  3. David,

    Understand. There is a drop in efficiency of a heat pump but its not as much as down to Williams sCOP= 1

    I posted ages ago a study that showed exactly what the drop was right down to -15°. It was something like 3 down to 2.2. a 26% reduction for the heat pumps in that study.

    5kW @ 3= 1.7kW electrical input (-4 °C outside)

    5kW @ 2.2 = 2.3kW electrical input (-15 °C outside)

    So on the absolute lowest temp day (doesn’t apply to all regions and in a warming world?) that’s a +35% increase in electrical power required. Current design is based on around 9 days’ worth of the design temp at -4° for London.



    Steve Harper
    Group M & E Manager

    t : 0208 418
    m : 07917 615 I

Comments are closed.