Accuracy of COP measurements

I am a retired cruise ship chief engineer and at present just an interested observer.

Am I right that in the public dashboard, the COPs are calculated from heat output divided by total electrical energy input?

It’s fairly easy to monitor electrical energy with good accuracy, but how is the heat supplied measured? I assume this is based on water temperature before and after the heat pump, or is it measured before and after the radiators or underfloor coils which be a better measure of actual heat delivered to the house.

Some of the SCOPs seem rather optimistic to me.

1 Like

My understanding is that ideally the measurement is done using a calibrated (to a certain tolerance) heat meter which measures both flow rate and delta T between the flow and return coming from the heatpump.

The dashboard does not indicate what measurement principle is used - in my case I don’t have a heat meter installed and am calculating a value based on Delta T - with two probes which I have calibrated to 0 Deg C using ice in water and 100 DegC using Boiling water - as well as checking their ability to track each other as said boiling water is allowed to naturally cool - I believe they are accurate to within 0.2 DegC. My flow rate is an assumed value which I have arrived at by drawing a known volume of domestic hot water at a known temperature - and measuring how many KW/hr it takes to return the DHW to its starting condition - then calculating back to work out what the flow through the DHW coil must have been. Its not perfect but the only method I have thought of so far.

I am a (nearly retired) process control/automation engineer - so can guess the reason for your question - I suppose the SCOPs have to be taken in the context of a wide range of different installations with varying degrees of calibration of the various sensors involved.

It probably more useful as a way to get a feel for what conditions tend to lead to higher performance.

Thanks, what COP are you achieving in the colder weather? GSHP or ASHP?
Did you mean kw per hour? Or kW.h?
You could fit a hot water flow totaliser that would give a running total of m^3. Or litres and record it hourly.

Hello @kelper

We do need to highlight which monitoring equipment is used for each of those in the list. But I can say with certainty that all three at the top of the list with highest COP’s have class 1 metering on electric and class 2 heat meters for heat output (all sontex heat meters). The electric metering also includes the central heating pump and indoor controller in addition to the outside unit & compressor. Most of the heat pumps in the list have class 1 electric metering (usually an SDM120) and class 2 heat metering (either sontex, kamstrup or sharky class 2 heat meters), apart from a couple measuring heat output with other equipment e.g (an emontx4 and sika VFS flow meter).

1 Like

Here’s the product page for the sontex 440 (this is the model that I have) Sontex Superstatic 440 Im number 3 in the list :slight_smile:

Thanks, I’ll look at that link. Is your installation on the public dashboard? Roughly how accurate do you think your COP is +/- %

Am I right, that a Class II heat meter has a maximum error around 7%?

I wonder if, for those of us who have the data, it would be useful to provide historic gas use? SCoP is all very well, but if we’re trying to understand the relative economics of a HP vs gas boiler, then really what matters is what electricity you consumed to maintain your house at the desired temperature compared to how much gas you used to use (then anyone can estimate the cost benefit from current tariffs). Ideally you’d want an average of several years’ gas use (which I happened to have, but not everyone will). We will then have fiscal-metering data as the absis for the comparison. Obviously there might might be some inaccuracy of you were cooking with gas int he past, but I doubt it’s very significant.


Perhaps a popup when you click one of the lines with more details (remove some of the columns) and a COP confidence score based on where the data is coming from?

This BEIS report on heat meter accuracy is probably as detailed as you can get:

the overall conclusion being … it’s complicated!

“Over the majority of tests and operating conditions of the laboratory test program, the overall MID
MPE for the heat meters was between 4 and 7%” - page 115.

This page from Kamstrup is also useful European Standard EN 1434 ensures repeatability

Yes, it’s this one:

1 Like

Thanks. Yes, I had read that very document earlier today! I’m impressed with these heat meters, not something we use on ships, yet.

1 Like

While not exactly the same, “equivalent gas use” can be calculated from the heat output divided by 90%. If the home is being heated to the same temperature, then this is how much gas would have been consumed, assuming that a gas boiler system is 90% efficient - this has been described as “generous”, as many systems won’t be anywhere near this good, but it’s a reasonable number for comparisons.

I have an experimental dashboard that shows running costs of heat pumps (for current UK prices).

See this post for the analysis of running costs I did a little while back.

1 Like

To be careful with at the same time : on moving out some ordinary tenants and moving in myself, with no change at all to the heating hardware (a gas combi boiler plus a small single room air source heat pump), I immediately cut gas use from “usual” to “hardly any at all” by mainly changing the heating policy to “gas normally OFF” which is often sufficient for me in this well-insulated mid-terraced house. That is, household policy decisions as simple as “put on a jumper, stupid!” can sometimes make huge savings too.

My using the air source heat pump for six hours in the daytime has not replaced 20 to 40 kWh/day of a tenant’s gas use. But it is plenty for me. From advertised CoP and a plug-in energy monitor at the three pin plug, I cannot have gathered more than 10kWh(heat) yesterday, and I used nothing else to heat the house.


I did see your comparison earlier and thought it interesting

I guess my thinking is that the EPC figure is a guide based on many assumptions including occupant behaviour. Whereas if you have your own gas readings and you consider the level of comfort prior to ASHP to be the same as after, then it ought to be a more meaningful comparison.

Ill do it at the end of April when i have a full year of data!


1 Like