I have recently spent the best part of a day adjusting the ΔT on the UFH downstairs and radiators upstairs which has improved the compressor run time considerably. I had a couple of short UFH loops which had the flow rate set too high and the ΔT between flow and return was only 2 deg. Reducing the flow increased ΔT of the loop and also slowed down the gradual increase in LWT/EWT at the pump.
What I haven’t been able to resolve is the use of a couple of towel rails. These are plumbed into the radiator circuit. No matter how low the flow in the rail, it acts as a short circuit with identical flow and return temperatures. Any flow which is sufficient to warm the rail results in a more rapid rise of LWT/EWT and subsequent shortening of the compressor on cycle.
There is no direct electrical heating element in the rails but I’m considering fitting some and isolating the lockshield valves. That would also allow their use when space heating is not required.
I was wondering if anyone knew of an alternative to the standard ladder rail that gave any significant ΔT or had any ideas on how they could be made to work with a heat pump.
I did think that plumbing them in parallel with the DHW circuit would probably work but that is not something I could do without a huge amount of work.
Return temperature limiter (like a TRV but senses the return temperature rather than the room temperature - this also works well for small, always on, underfloor zones such as bathrooms if set to 22C ISH (floor temp)
Plumb the towel rail in full flow (feed the underfloor manifold for a few zones - or indeed the bathroom UFH zone - through the towel rail first)
For a small bathroom with wet UFH I would take flow from heat pump through the towel rail, through the floor loop, Pas a return temperature limiter set at 22C ISH, then back to the heat pump.
Ohh I didn’t know these existed.
I also have a couple of towel rails that can heat the small bathrooms each is in but getting any kind of delta T is a struggle.
Not sure how I’d decide what temperature to set one of these to when the flow will be varying with outdoor temperature.
Do you have any more info on the type of towel radiators you have: physical size, nominal output, maybe a photo? Are they the only heat source in the rooms they’re in or are they supplementing other radiators?
I have towel radiators as the only radiators in my installation (the UFH does most of the heating) and they don’t seem to cause any issues - I’m typically getting dT of 5 for the whole system; I’d have to do some experiments to check what dT I get for the radiators by themselves.
My towel radiators happen to be stainless steel, which I understand is a bit better for heat transfer (than chrome plated mild steel) and have quite ‘chunky’ bars hence a higher water volume than some alternative designs (and more volume is A Good Thing).
If you’re open to swapping yours out you might be able to find alternatives that behave more like your other radiators - or re-piping them to be ‘in series’ with other emitters as Marko suggests should definitely work too.
I have two, they are both I think 1.8-1.9m tall. One seems much better than the other as it is much more chunky with square section horizontals the other is a fairly standard with circular horizontals.
The two rooms are next to each other and I was thinking of sorting out the pipe run to the furthest one anyway (goes through a bedroom and overheats it sometimes, also lots of pipe creaking) so I could maybe pipe them up in series then.
With underfloor heating you set them at the floor temperature that you want.
With towel rails set them at approximately the return temperature that you want, seeing on the low side, and adjust them perhaps three times over winter.
The setup works best with a towel rail (at “full” flow temperature with minimal deltaT) feeding into the bathroom underfloor circuit (sitting on a return temperature limiter)
I also have a towel rail with a heat pump in a quite small bathroom which is heated nicely by it. All my radiators are in parallel and fed by a UFH manifold in the basement.
What make and model heat pump do you have? How many cycles per hour do you see at which outdoor temperature? How is internal temperature controlled - purely by weather compensation or an indoor room thermostat?
I don’t think you need to change anything. There really isn’t a good reason to target DT5, as @Zarch has explained in a number of posts on the forum. The heat pump supplies power to the water circuit which heats it up. Your radiators put that heat into the rooms, hence cooling down the water. DT is simply a consequence of flow rate and power delivered into the rooms. If you have a continuously rising flow temperature, this means your emitters cannot fully shed that heat into the rooms. Restricting flow to increase DT will not change this - it’s purely a question of energy balance.
Actually, you will even need a slightly higher flow temperature at DT5 compared to say DT2, because the radiator temperature - which determines its heat output - is approximately equal to the mean of flow and return temperature. At a fixed flow temperature DT5 yields a lower mean temperature than DT2. For example, with 30°C flow temperature and DT2, the radiators are at 29°C, whereas at DT5 they are at 27.5°C. To get the same radiator temperature, you’ll have to raise your flow temperature to 31.5°C. This will hurt COP a bit. Overall, this could perceivably reduce cycling, because your heat pump is now working less efficiently for longer times instead of more efficiently for shorter times.
Just reducing DT by restricting flow would make your house colder, but depending on the thermal inertia this will take a while to manifest. Truly judging whether any of those subtle changes makes a difference can only be done when the system is in its steady state, which can take a few days even to reach. To find two comparable steady state scenarios pre- and post change is another issue. Take it from me who also tinkered massively with flow rates and DT - it’s very easy to fall into the trap of mistaking momentary changes in heat pump behavior for a true change in steady state behavior.
There might be some specifics in how your heat pump’s control algorithm works, hence my initial question on heat pump model. Independent of that, cycling fundamentally indicates that power output from the heat pump into the water exceeds heat loss of the house to it’s surroundings.
Do the math for a low dT on the towel rail and a high tdT on the radiators and the resulting “blended” dT that the heat pump sees.
The risk is that excess flow through the towel rail sees your radiators operating at dT 6 whilst the towel rail is dT 1 and the overall dT is say 2. Your “effective system volume” also gets reduced by the bypassing towel rail.
I expect a high flow rate is needed in the towal rail to prevent a layer of cooler water “sticking” to the metal work with the warmer water going down the center of the wide tubes. This will therefore always act as a bypass unless used as the feed for a manifold.
If the towel rails are connected in 22mm, is the resistance low enough to have all the heating flow going var them?
Personally I would replace them with normal radiators with towal hooks above them.
Does the return temperature limiter give an even consistent flow, or is it letting a large high flow pulse of water out when the return temperature drops, then stopping flow until temperature drops again?
My heat pump is a Riello NHXM008 which is identical to Midea M Thermal R32. (it’s a rebadge). I am running open loop and no buffer. I do have TRVs but all currently set to max.
I’m not going to use WC until I have a good idea of what flow temps to use, and currently I have the LWT set to 30 which works well at outside temps down to 3C. I’m waiting for colder weather to determine how much to increase this.
Before I had any detailed monitoring, I thought the heat pump controller would attempt to maintain the flow temperature at the LWT set point, but this only happens in the initial part of the heating cycle.
At the start of the heating cycle the overall flow rate is the maximum that the heating pipework allows. This is maintained until the flow temperature is 0.5K above the set point AND the return temperature is within 4.5K
Then the pump controller lowers the flow rate to limit the overall pump ΔT to 4.5K.
If flow through the towel rails is allowed, their return temperature has a close to zero ΔT. Likewise, when any other individual radiator or heating circuit has a low ΔT it tends to lower the overall ΔT between LWT/EWT. If the heat pump controller has lowered the overall flow rate to the minimum it will then increase the LWT to maintain the LWT/EWT delta. When this happens, there is an increase in LWT/EWT until the upper LWT hysteresis limit is reached causing the compressor to turn off. The circulation pump continues running until the LWT drops to the lower hysteresis limit. This effect was causing the compressor to turn off after an hour and taking several hours to turn back on. By adjusting flows so each circuit has at least 4.5K ΔT I can now have over 12 hours without the compressor turning off.
A similar effect happens if a lower LWT is used, say 28C, but for the reason you described.
“Your radiators put that heat into the rooms, hence cooling down the water. DT is simply a consequence of flow rate and power delivered into the rooms. If you have a continuously rising flow temperature, this means your emitters cannot fully shed that heat into the rooms.”
I know from using an IR camera that the UFH loops are at best 150mm, so not the most efficient emitters, but I have no intention of changing that so I think I’ll stick with a min LWT of 30 and leave the towel rails off for now.
This data is on a gas boiler that is cycling (heat load of the house is ~40 kWh/day or 1.7 kW with it about 7C outside) but you can see that for the “full flow towel rail feeding 4m2 of underfloor heating” meter (yes, it was an experiment) the flow is stable whilst the pump is running AND that valve isn’t really moving open/closed between boiler cycles.
The visible dip this morning (morning of 10th in data) was sunshine warming the room in the morning and the return temperature limiter responding by nipping down the flow through the underfloor zone.
