Balancing radiators - temperature drops?

Hi all,

My HP has been running for about 18 months and the bathroom is rarely warm enough despite opening the valves fully (it’s a combined towel rail & radiator, our only radiator without a TRV).

I decided that I should try to balance the 23 radiators, which were originally set up with 2x LPG combi boilers on the same heating circuit; these potentially put out so much heat I don’t suppose balancing mattered much!

So I turned all TRVs to fully open and ran the heating. Once the flow temperature had stabilised I went round each radiator and using my trusty laser dot temperature gadget attempted to get temperatures from each end of each radiator. Then I would know which side is flow and return for each radiator and what their respective temperatures were on this occasion (I have done it twice now).

I encountered two problems:

  1. a previous occupant has painted all the pipes with emulsion which seems to mask the temperature so I couldn’t get consistent readings from the pipes so I had to take it off the valve bodies. Some of these are chromed and don’t give nice readings, but I managed to get something seemingly worth having.
  2. having got the readings, typical YouTube advice is to have a temperature drop of 10-12c. This is obviously for boilers however my temperature drops ranged from 0c - 5.4c (with HP Flow 46.3c, outside 2.6c).

I would appreciate any advice on how best to get these readings and what temperature drop I should aim for.


I’m going through the same balancing act with my system, and it’s a major challenge as I don’t have a buffer tank. Therefore I’m fighting issues with flow, delta-T at the rads and system behaviour as TRVs start to act. As well as UFH loops of various sizes all being actuated by different room controllers. I’m fighting getting the right balance, trying to get the ASHP to run well without cycling, and also at it’s most efficient.

We’d need to know a lot more to help you

Do you only have rads ?
Any UFH ?
Whats your flow temp from the heatpump, and what’s the return ?
If there a buffer tank ?
Are your radiators / house ‘zoned’ by thermostats ?

It’s a balance between ensuring the heatpump has enough circulation (a problem when the TRVs start to close) and ensuring you get even heat across the radiators. You should be in a position to monitor at a very least your electricity consumption in realtime from your heatpump, as this will tell if the compressor is cycling (bad) or not.

Balancing ashp systems is almost a black art.

While not strictly about balancing, this is a good read and gives all the things you may need to consider.

AFIAK, getting a delta-T with Rads on a ASHP of around 3-4 degrees is good, especially at lower flow temps (I run mine at 35c - running at 45C you could get a better delta-T at the cost of more energy spent).

These are just my own opinions after months of playing and installing much monitoring (e.g. montoring on all my UFH circuits, monitoring on the heatpump temps, electricity, defrost function, backup heater operation and a host of other things).

All said it’s complicated.

So if you have a delta t of 5.4 that’s what you should be aiming for on all your rads, I’m assuming that it is a larger property to have 23 of them. A couple of questions has the heat pump been sized correctly and have your rads been upsized to match the output. Do you have a secondary pump fitted because if you haven’t that could be one of your problems. Give some details of your system eg size of heatpump, cylinder size, setting, timings anything that helps build a picture

Thanks both.

I have a 16kW Samsung Gen6 heat pump, slightly oversized. 23 radiators, no UFH.
House is 5 bedroom 2 reception, 280 sq m, ~24000kWh heat demand although there’s only 2 of us here now.
The heat pump was fitted around an existing custom Newark cylinder, 250l with 3 coils & 2 immersion heaters. HP is connected to 2 coils & 1 immersion heater.
There’s a 50l Cordivari buffer after the cylinder with 4 connections to flow/return.
There is a pump on the heatpump circuit and a second on the heating after the buffer. It runs at 24.x litres/minute, weekly or if it drops to 24.0 I clean out the strainer & filter.
The heating soon tees off in 2 directions, one doing the original house (15 radiators) and the second doing the extension (8 radiators). The radiators are the original ones, not replaced for the heat pump. The guy who refurbished our house basically put a radiator under every window plus a few more where there weren’t any windows.
We only have 1 thermostat, currently a Tado, in the main reception room. Usually, the Tado heats in relatively short bursts which doesn’t look great but is surprisingly kind of okay.
I have emonCMS monitoring electricity consumption, flow/return of heat pump, cylinder & heating.
We have a system diagram but I think it must be in my wife’s email as I can’t find it.

For the above test, all TRVs were fully open and the heating had been on for over an hour for the flow temperature to stabilise at 46.3c. Flow temp is of course variable from weather compensation. I did another test at another flow temperature and will do another when I can run the heating on free electricity from solar.
Looking at tonight, we have:

  • Heatpump flow 40.1
  • Heatpump return 32.3
  • Heating flow 35.0
  • Heating return 31.1
  • So I think my future checks and any attempt to adjust radiators I need to compare against the heating flow/return values which need to be steady.
    I can aim for the same temperature drop in the radiators as I am monitoring on the heating side of the buffer.

    Thanks both!

    So the obvious thing that sticks out is the fact you are losing 5degrees from your heating buffer tank immediately, so what ever temp your heat pump flow is should be what’s your heating flow is leaving the buffer. If the radiators haven’t been upsized you will struggle getting your rooms up to temperature. So if there is only 2 of you in the property you should be only using about 90ltrs of hot water so heating a 250ltr cylinder is not only costing you money but reducing the efficiency of your heat pump. Try not to use your Emerson heater except for a legionaries cycle (which I doubt you need anyway) but regs are regs. Where are you measuring your flow rate from heat pump side or heating side from buffer. If it is the heating circuit it looks a little low for the size of the system.

    This is great information and it’s great you’ve got that level of monitoring.

    The problem statement is the bathroom doesn’t get warm enough.
    What temperature does that rad/towel rad get to, and how does that compare to the others ?
    Does it get up to temp at roughly the same speed ?
    Basic balancing is to get the rads at the same temp at the same speed, and normally achieved by constraining a little the ones nearest the heat source.

    I’m wondering is this just a cold room with high heat loss and the heat going into the room from the rad isn’t sufficient to compensate.

    Ref the buffer tank comments, I’d suggest you get accurate reading with a multimeter thermocouple probe (hold in direct contact to copper pipe) or a known accurate thermal imaging device. I say this, as my attempt at strapping probes to pipes led to the readings being out by as much as 8 degrees (despite being tightly strapped and under insulation). Unrelated but some good information on that is here.

    1 Like

    Yes totally agree with you, also you will never get the same amount of heat out of a towel rad as a normal rad, so it might be a case of replacing that one with something suitable. There are no myths around heatpumps, it’s just the case of setting them up correctly to be most efficient. Like most things in life we buy have all been tested in perfect conditions and not the real world.

    1 Like

    IME towel rads are never very effective at warming the room and if they do it is always warm at the wrong time. Hot air blower used sparingly is my suggestion. Probably cheaper than running the HP (with associated losses).

    I’ll second this.

    Except in these cases they are often designed and fitted by numpties which magnifies the problems.

    Absolutely right Brian, MCS accreditation isn’t worth the paper it’s written on. There are so many poor designers and installers and see it all the time.

    1 Like

    It has no outside walls but its ceiling is under an inaccessible poorly insulated loft space (there’s some insulation near the hatch so I assume it’s all the way through but is from the noughties and not very thick.
    There’s a Velux window with a thermal blind which has been kept closed for a couple of months. Walls & floor are tiled.

    Temperature’s a bit funny tonight as the Tado is now doing short bursts. The sitting room with 2 outside walls has 3 radiators which have got to 26.6c just now, we’re comfy and the room is 19.9. I can read 28.8c off the bathroom one, it doesn’t feel too bad in there tonight.