Designing a simple UFH system

[Nickfromwales]

1 hour ago, SimonD said:

I know this is a debate between you and @Nickfromwales 

Debate is good. Fill reading this can then make more of an informed decision based on the balance of the boatload of information exchanged here (BH).

 

Plus I’m quite stubborn, so good to reflect on my own ideas when challenged, more so when it’s backed by evidence (someone’s done it, moved in, and is living with it for eg) which is gold dust. 

[JohnMo]

1 minute ago, Nickfromwales said:

Not really? 
 

Radiators have thermostats in each room / space, in the form of thermostatic radiator valves, meaning each space can be controlled individually as per the temp set on ‘the stat’, so that’s actually converse to your favoured methodology? 

Ask @JamesPa and others how the thermostatic valves he has in his retrofit system. Does he have issues running weather compensation - no. Think he has them in bedrooms only so he can run those rooms cooler than MCS design standards. So pretty much my favoured methodology.

[SteamyTea]

If you had a local solar gain prediction device, how easy would it be to integrate into the WC curve control in most controllers?

[Nickfromwales]

1 minute ago, JohnMo said:

Ask @JamesPa and others how the thermostatic valves he has in his retrofit system. Does he have issues running weather compensation - no. Think he has them in bedrooms only so he can run those rooms cooler than MCS design standards. So pretty much my favoured methodology.

So he needs / has individual room control, which requires no human input other than telling the stat to be no hotter or cooler than the set point.

 

Without them the rooms (bedrooms in your stated case) would heat up or become too cool, influenced / affected by external or internal factors such as sunshine.

 

I think in open spaces and thoroughfares it’s relatively easy to control, but in rooms that have doors which are routinely kept closed, I’m really not sold on any notion of a) doing away with FF heating, or b) having absolutely zero control over the temp of the space; other than attempting to manage temp via flow rate, which most wouldn’t be able to do, or be even slightly enamoured with the idea of doing. 

[JohnMo]

7 minutes ago, Nickfromwales said:

So he needs / has individual room control, which requires no human input other than telling the stat to be no hotter or cooler than the set point

No stats anywhere other than, I believe bedrooms, but not even convinced they are in active control.

 

9 minutes ago, Nickfromwales said:

other than attempting to manage temp via flow rate, which most wouldn’t be able to do

My influence is flow temp via settings WC curve only. ASHP algorithm does everything else. Don't even have the ASHP controller in house; as that's built in to ASHP. The heat pump will modulate flow rate as it sees fit. Even have the UFH loop flow meters fully open to assist the heat pump looking after itself. Just designed the loops and rooms to match target temp at the same dT for every loop.

 

Gas bills initially drove the changes, way higher than expected. Final leap to no thermostat was the electricity meter going smart with whole house battery and being able to run whole house on cheap rate electric 24/7/365, this drove me away from batch charging floor via a single thermostat, to pure weather compensation for heating and effectively a fixed flow temp for cooling.

[JohnMo]

23 minutes ago, SteamyTea said:

If you had a local solar gain prediction device, how easy would it be to integrate into the WC curve control in most controllers?

May need it for radiators, but you just don't need to bother with low temperature UFH.

 

Pretty easy really, solcast integrates into home assistant etc so you can get hourly forecasts. In shoulder seasons I actually ramp up day time heating to charge floor, on excess solar. Get good day time cop instead of night time cold temperatures. Heating just reverts back to WC mode when sun no longer generating enough. But floor generally full of energy to heat house as temperature drop. Completely the opposite of what we are being informed to do.

[SimonD]

2 hours ago, JohnMo said:

Been there done that with UFH biggest disaster of all the options I tried and tested (Atag boiler). Floor timescale for change is just too long, great for radiators, utter rubbish with UFH (unless very screed and high flow temps used).

 

The main reason I chose UFH, as opposed to radiators, where high flow temp means they do not self regulate. As this topic is about simple UFH, not sure radiators are relevant or room compensation.

 

If your room is 23 (solar gains etc) and floor temp is 22, there is zero output to the room. A thermostat does nothing except close of flow routes available to the heat pump, generally not good. The floor stays charged up ready to release heat as room temp drops. 

 

There are obviously two views, but if I knew in 2020, what I know now, I wouldn't have a cupboard full of needless tatt, that I have removed from the heating system. 

 

Subject is simple UFH.

Simple is no mixers, no more pumps than absolutely required. Open loop, no thermostats. Run a simple weather compensation curve. Have switch that allows cooling to be active if needed. ASHP will run only as required, set the logic so the circulation pumps are off during compressor off periods to get a great CoP. Plus a suitable system water capacity, to ensure decent run times. So about 30L to 40L of system water always engaged per min output kW of the heat source.

 

The heat pump will modulate it's self, as gets closer to target flow temp, so you don't need outside influence to do that.

 

If you don't do UFH, maybe you should explore further. Set it running it on weather compensation it looks after it's self come rain or shine. You don't need external control.

 

Lets back up a bit - yes, the OP is about designing a simple UFH, which you can do with both gas and heat pumps and you can make them both efficient. Radiators are relevant because they need to be there in the design library because UFH may not always be the best option for the given situation, whether you like that or not. The reality is that every house is different, it has its own climate context and it has its own users, all of which play into the design of the heating system.

 

I designed my first system just a little before you, which was for my own house. I designed it with UFH in mind and then changed to radiators for several reasons. Mine is currently on gas and will be a heatpump in the next month or so. I also chose not to go down the route of pure weather compensation, but have actually implemented 2 main zones with room influence using Opentherm. My system has actually been extremely efficient and has this year, for example, used about 5000kWh less than estimated at design and that's even with a thermal store on PDHW sitting at 75C at the top for dhw. It's still a simple system and needs some changes, but this is quite good.

 

Without blowing my own trumpet, I found I had a knack for heating systems and now run my own business doing them. And one area I've made quite a lot of money from is fixing UFH systems to run efficiently, especially with gas boilers  - so I have a little knowledge about how they work and don't work.

 

One of the things you don't seem to be aware of is that room compensation in controls can apply just as much to UFH as it can to radiator systems. Many heatpump manufacturers provide built in room compensation within their controls, including Grant, Vaillant, Stiebel Eltron, to just name a few. This is because it works well in unison with weather compensation which can and does have some limitations. The room compensation provides, as I said, additional real-time data for the heatpump to modulate output.

 

This isn't to say that your solution isn't valid - I've already acknowledge I understand what you've done and why. It's just that this isn't about being a one trick pony, but acknowledging that there are other solutions out there. Your experience with your gas system is unfortunate and a reality for many, but that doesn't meant to say that you can't design and implement a gas system with ufh that functions well and is efficient and can be simple. Same goes for heatpumps, of course!

 

Even in your case, there is a limit to how much self-regulation there is in the system and how flexible it is because as you say, the response times of UFH particularly when sitting in a big slab, are pretty atronomical.

 

[JohnMo]

17 minutes ago, SimonD said:

Your experience with your gas system is unfortunate and a reality for many, but that doesn't meant to say that you can't design and implement a gas system with ufh that functions well and is efficient and can be simple.

My final version of gas powered UFH, wasn't pure weather compensation as boiler just wouldn't modulate down far enough at low temperature and cycled too much to be efficient. So ran weather compensation at a fixed flow temp and a setback. Used a simple 0.1 hysterisis thermostat in the hall and let the system bounce of that a flow temp of 36, heating demand set back off and no heating demand setback on. This gave the boiler two running temps, 36 or 26, so in setback it never fired up. The 36 temperature allowed the boiler to run with no cycling. It started and stopped based on the thermostat. So a -9 day it ran for about 12 hrs, a 7 degree day a few hours. Efficiency was great. Generally running once per day.

[mattgibbs]

So my setup will be ASHP and UFH downstairs. In my situation i dont see how having zones and thermostats would help. If there were some solar gain then this would have a quick effect on heating up a room. If there were a thermostat in that room which then stopped heating that zone what would that actually do? the slab would presumably remain a similar tempreture for hours and have no short term effect on the room tempreture. 

 

So long as the system is balanced and rooms are heated evenly, is any more control other than weather compensation in the heat pump really needed?

[SimonD]

16 hours ago, mattgibbs said:

So long as the system is balanced and rooms are heated evenly, is any more control other than weather compensation in the heat pump really needed?

 

Yes, if there are significant heating inputs that are not measured by weather compensation alone, and/or if the occupant want some control in rooms such as bedroom. In our house, for example, we can have the triple glazing on the north face still covered in ice while the upstairs has already been heated to 23C by the sun. In this instance, even if the MW-AT delta T is such that the radiators then don't theoretically add heat into that space, it seems a bit silly not to modulate the heatpump using room temp feedback as the home heat load has reduced. In ours this is easily by 50%. 

 

The idea around room influence is that an internal thermostat provides input to the heatpump to modulate. This can also be achieved using TRVs on radiators without modulating the heatpump, or with electronic mixers. Most manufacturers will provide 1 or 2 of these room stats but companies like Stiebel Eltron can build a system with up to 6 and it also provides the ability to weight system control input between weather compensation and room. 

 

Nobody here is saying that a heatpump installed with open loop and weather compensation isn't the simplest solution. It's just that get a heating system to work for its occupants may NOT be that simple, especially in highly insulated homes. Room influence does not make the system more complex. It can infact make it simpler for both the occupant and the installer.

 

Room influence does not necessarily shut down flow and therefore reduce system open volume, except for at the extremes. 

 

Zoning, using relay type controls is a different matter as this does reduced system open volume and therefore the system becomes more complex.

 

For me the issue here is the understanding of controls, what they, do and how they work, and then when to implement them - but that's a headache for the whole industry right now which is still learning. But most good decent installers will go for open loop with some room influence baked into the system in current designs.

Edited yesterday at 11:19 by SimonD

[Nickfromwales]

51 minutes ago, SimonD said:

 

Yes, if there are significant heating inputs that are not measured by weather compensation alone, and/or if the occupant want some control in rooms such as bedroom. In our house, for example, we can have the triple glazing on the north face still covered in ice while the upstairs has already been heated to 23C by the sun. In this instance, even if the MW-AT delta T is such that the radiators then don't theoretically add heat into that space, it seems a bit silly not to modulate the heatpump using room temp feedback as the home heat load has reduced. In ours this is easily by 50%. 

 

The idea around room influence is that an internal thermostat provides input to the heatpump to modulate. This can also be achieved using TRVs on radiators without modulating the heatpump, or with electronic mixers. Most manufacturers will provide 1 or 2 of these room stats but companies like Stiebel Eltron can build a system with up to 6 and it also provides the ability to weight system control input between weather compensation and room. 

 

Nobody here is saying that a heatpump installed with open loop and weather compensation isn't the simplest solution. It's just that get a heating system to work for its occupants may NOT be that simple, especially in highly insulated homes. Room influence does not make the system more complex. It can infact make it simpler for both the occupant and the installer.

 

Room influence does not necessarily shut down flow and therefore reduce system open volume, except for at the extremes. 

 

Zoning, using relay type controls is a different matter as this does reduced system open volume and therefore the system becomes more complex.

 

For me the issue here is the understanding of controls, what they, do and how they work, and then when to implement them - but that's a headache for the whole industry right now which is still learning. But most good decent installers will go for open loop with some room influence baked into the system in current designs.

B I N G O. 👌

[JohnMo]

Really the simplest and most pragmatic is - fully open system with room by room monitoring via a good UFH controller. Then if you have issues it relatively easy to plug in an actuator or two. It also assists with system balance, setting up WC and general history of what happens on different days.  If you get one with an outside sensor it will also record outside temp, various room temps, so you can overlay the information.

 

Only thing with Stiebel Eltron, you need a second mortgage to buy one, they demand a buffer, so you end up with at least one, possibly multiple additional pumps. All affect real CoP achieved and most are powered from a different power supply, so don't get added to the real CoP advised.

 

2 hours ago, SimonD said:

it seems a bit silly not to modulate the heatpump using room temp feedback

But when you delete buffers, the heat pump sees this reduced heat load, without any outside assistance, by a changes in return temp, so it modulates naturally to maintain its defined target dT.

 

Even flowing a fixed temp in cooling, the heat pump senses when it needs to do more work or less, purely from return temp, so output and cycle times change to reflect that. Nothing magic is needed, in fact adding stuff kills this self modulation. Cooler rainy day today heat pump has only fired up once ran for 32 mins, then sat idle except for DHW. No input from anything external.

 

2 hours ago, SimonD said:

TRVs on radiators

Radiators? UFH is the discussion and TRV don't work on UFH the time to heat or cool is just way to long

 

19 hours ago, mattgibbs said:

remain a similar tempreture for hours and have no short term effect on the room tempreture. 

Exactly

[marshian]

On 17/08/2025 at 12:33, Nickfromwales said:

So he needs / has individual room control, which requires no human input other than telling the stat to be no hotter or cooler than the set point.

 

Without them the rooms (bedrooms in your stated case) would heat up or become too cool, influenced / affected by external or internal factors such as sunshine.

 

I think in open spaces and thoroughfares it’s relatively easy to control, but in rooms that have doors which are routinely kept closed, I’m really not sold on any notion of a) doing away with FF heating, or b) having absolutely zero control over the temp of the space; other than attempting to manage temp via flow rate, which most wouldn’t be able to do, or be even slightly enamoured with the idea of doing. 

 

I'm doing whole house (inc bedrooms) with radiators without TRV intervention - Yes I have TRV's right now but They are set to a stupid upper limit so actually have no functionality - I could just as easily replace them with decorators caps - Weather compensation combined with individual radiator flow rate control works perfectly

[SimonD]

1 hour ago, JohnMo said:

fully open system with room by room monitoring via a good UFH controller.

 

Can you explain what you mean by this?

[SimonD]

4 hours ago, JohnMo said:

by a changes in return temp, so it modulates naturally to maintain its defined target dT.

 

Just to add - yes, but if the weather compensation is calling for heat and the heatpump sees an increase in return temperature, it will increase the flow temperature to maintain target Dt, which is one reason why a buffer, for example is less efficient when there's too much distortion. Just seeing an increase in return temperature is not enough - the heat pump needs additional feedback. This is another reason why room influence is useful in circumstances where there is additional heat input into rooms.

 

Hence your description of the heat pump modulating down when it sees a higher return temperature, is not necessarily, nor universally, correct. It depends on controls.

[JohnMo]

1 hour ago, SimonD said:

but if the weather compensation is calling for heat and the heatpump sees an increase in return temperature,

Certainly not what I have experienced. Isn't WC always a calling for heat or cooling? Isn't that whole point its always ticking away adding/removing heat as required.

 

In heating if target flow temp is 26 and dT reduces as a result of higher return temp, the flow temp will increase but only if heat pump cannot modulate down to to manage target flow temp, if it cannot modulate further downwards flow temp does increase, but is limited by compressor shut off hysteresis, so may increase up to to about 0.5 Deg or less and then the compressor cycles off.

 

The only way the ASHP doesn't modulate to min output and the then shutoff, is if you running an algorithm that drives the heat pump count deg minutes, which allows a fair degree of temp overshoot. Running basic WC doesn't do that, and can rely on return temp feedback without additional controls

 

2 hours ago, SimonD said:

the heat pump needs additional feedback

Sorry it doesn't. Two different days this week, doing cooling but the same rules apply as for heating.

First shot a couple of days ago, pretty sunny all day - big spike is DHW heating - does 27kWh of cooling

 

Second today, very little sun does 9kWh of cooling

There is no control other than what the heat pump senses, nothing in the house except UFH pipes.

 

A week or so ago, plenty sun and heat, does 44kWh of cooling

 

 

 

[SimonD]

13 hours ago, JohnMo said:

Certainly not what I have experienced. Isn't WC always a calling for heat or cooling? Isn't that whole point its always ticking away adding/removing heat as required.

 

In heating if target flow temp is 26 and dT reduces as a result of higher return temp, the flow temp will increase but only if heat pump cannot modulate down to to manage target flow temp, if it cannot modulate further downwards flow temp does increase, but is limited by compressor shut off hysteresis, so may increase up to to about 0.5 Deg or less and then the compressor cycles off.

 

The only way the ASHP doesn't modulate to min output and the then shutoff, is if you running an algorithm that drives the heat pump count deg minutes, which allows a fair degree of temp overshoot. Running basic WC doesn't do that, and can rely on return temp feedback without additional controls

 

Sorry it doesn't. Two different days this week, doing cooling but the same rules apply as for heating.

First shot a couple of days ago, pretty sunny all day - big spike is DHW heating - does 27kWh of cooling

 

Second today, very little sun does 9kWh of cooling

There is no control other than what the heat pump senses, nothing in the house except UFH pipes.

 

A week or so ago, plenty sun and heat, does 44kWh of cooling

 

 

 

 

There just seems to be some bizarre obfuscating circle going on here John,

 

1. You are running a very low temperature UFH system that also utilises a specifically designed slab arrangement.

2. You have spent countless hours/days/months/years tinkering with your system to get it working right for you.

3. You are using 1 anecdotal example, which is your own, to base a general conclusion that is being suggested as suitable for all and everyone.

4. You have some very particular views on indoor temperatures such as your view on bedroom temperatures for example.

 

Your system, because it utilises the self-balancing effect, actually manages to self-regulate independently of the controls in your system - that is what these systems are known for. But, it requires a significant amount of pre-planning and integrated design, and if being applied in a retrofit situation, a hugely costly upgrade process. Your graphs therefore don't reflect the nature of the heapump and it's controls but this thermodynamic behaviour which is independent of it, but nevertheless impacts the behaviour of the system as a whole. This is a great solution if you have the low heat loads required and all areas of the house have similar heat demand (e.g. there's no area where there is a significantly higher demand), as a basis to allow it to work.

 

These are highly significant caveats to what you are proposing and I believe people reading posts on this forum need to be aware of this and that there are so many other factors involved, such as floor coverings etc.

 

I'm not saying anywhere that your approach is wrong, I'm suggesting that other approaches may be more suitable for other people considering designing their own system, and whether you accept this or not, is neither here nor there as it's something that good designers and installers of systems are utilising more and more from their vast experience of putting these system in.

[JohnMo]

30 minutes ago, SimonD said:

and if being applied in a retrofit situation, a hugely costly upgrade process

But this conversation is about a new build, not a retrofit. All the preplanning I did was design the UFH with loopcad to get a close balance of temperature out the box. Nothing magic there. Followed the rules installed thermostats, tried room compensation. Had a mixer and pump on UFH manifold. The more I stripped away, the lower my gas bills became and the more stable the house temperature became. Doing it again, I would leave all the gunk removed, in the shop and the money in the bank. KISS.

 

32 minutes ago, SimonD said:

This is a great solution if you have the low heat loads required and all areas of the house have similar heat demand

So like the OP of this thread.

 

33 minutes ago, SimonD said:

there's no area where there is a significantly higher demand

That is definitely not true, our lounge is one wall all glass 6m tall. It's heat demand trumps the rest of the house.

 

Not the first to take the approach of a fully open system and leaving the controller to manage what going on. Isn't that a heat geek mantra. It's been done for years, I chose my Atag boiler because it was capable of it, as running pretty similar to a heat pump with similar dT etc.

39 minutes ago, SimonD said:

obfuscating

Certainly not trying make something less clear and harder to understand, especially intentionally. You just don't like my view point, which fine. As title of the thread says simple UFH. 

1. Step one design loops to match room heat loss, OP is on that.

Etc.. all said above. No point repeating it.

 

 

[SimonD]

46 minutes ago, JohnMo said:

That is definitely not true

 

Research suggests differently and I'm more likely to trust a balance of trusted research.

 

47 minutes ago, JohnMo said:

Not the first to take the approach of a fully open system and leaving the controller to manage what going on.

 

And you're not doing that John, so you shouldn't pretend that you are.

 

48 minutes ago, JohnMo said:

Isn't that a heat geek mantra.

 

Not quite, it's more subtle than that. The mantra is to use open-loop managed with pure weather compensation on retrofits and less well insulated homes (including some newbuilds) but in highly insulated and airtight homes use weather compensation together with room influence. The balance between weather compensation and room influence is dependant of specific design elements and context of the house.

 

51 minutes ago, JohnMo said:

You just don't like my view point,

 

Come one John, I've already said that open-loop is the preference many times. What I've been saying, yet again, yawn, is that the specific circumstances of the building must be considered and that building in room influence is preferrable when dealing with well built, highly insulated homes, with good air tightness, especially if they experience large inputs of energy outside the heating system. This isn't complicated either.

 

Occupants might also like more control in their bedrooms....

 

55 minutes ago, JohnMo said:

Certainly not trying make something less clear and harder to understand, especially intentionally.

 

 Err,  so below you're suggesting what? Not room influence by any chance?

 

22 hours ago, JohnMo said:

Really the simplest and most pragmatic is - fully open system with room by room monitoring via a good UFH controller. Then if you have issues it relatively easy to plug in an actuator or two.

 

 

[mattgibbs]

On 19/08/2025 at 12:08, SimonD said:

Room influence does not necessarily shut down flow and therefore reduce system open volume, except for at the extremes. 

 

Zoning, using relay type controls is a different matter as this does reduced system open volume and therefore the system becomes more complex.

@SimonD what would be an example of room influence controls and how does it differ from zoning? Will the MVHR not just negate any differences in room temperature anyway.

 

For context I'm purely looking at downstairs UFH at low temps, air tight house with mostly large open plan areas.

 

Bedroom heating, im hoping will come from heat rising from below combined with the MVHR distributing air evenly and being topped up from electric radiators in the adjoining en-suite's

 

 

[JohnMo]

38 minutes ago, SimonD said:

Research suggests differently and I'm more likely to trust a balance of trusted research.

Isn't that what I inferred, our lounge has heat loss and solar gain way out of proportion to the rest of the house. Ours house doesn't have a very even heat loss or heat gain. Most is nearly south facing with plenty of glass which is mostly in the lounge. Back of house gets best part of zero sun.

 

As the OP pointed out, once the room is getting solar gain, switching the heating off make no difference, you just cause other issues later and just end up chasing getting the floor again, and while it does, room temperature drops or the heat pump has to run harder. Leave warm water flowing through the floor, as the solar gain diminishes floor just picks up heating again.

 

38 minutes ago, SimonD said:

but in highly insulated and airtight homes use weather compensation together with room influence

UFH and room influence, maybe not, true for radiators, but that isn't the discussion here. A decent new build will be flowing sub 32 in any outside temperature, (depending on screed depth) floor surface temps a couple of degrees above room temperature at design condition. Floor itself does the room influence work it doesn't need additional tools to do it.

 

2 hours ago, SimonD said:

You have spent countless hours/days/months/years tinkering

That is very true, but also spend countless hours on here, to assist others from making the same mistakes I did. But the summary of the tinkering is

Remove unnecessary items from the system.

Ensure I have a suitable water volume to give good run times, so minimum 30L per min kW output of heat source.

Train my brain, that simple is actually best, in most circumstances 

 

38 minutes ago, SimonD said:

,  so below you're suggesting what? Not room influence by any chance?

Start with the simple system, no bells or whistles, a simple system that can monitor room temperature to aid commissioning. Only if needed start to add complication, not the converse as I did.