UFH - short cycling

[DoubleSix]

Hi All,

Wonder if you can help with an issue I'm having.

I have been monitoring the behaviour of our UFH and it doesn't seem right. It's JG system only a year old.

I keep it at a consistent 21c through day and eve, dropping back to 19c overnight. Issue is the behaviour of the thermostat, control unit and boiler.

It's goes like this:

- The thermostat reads 20.4c and calls for heat, control unit light goes on...
- delay (1-2mins)
- Pump kicks into life and boiler fires
- Flow to manifold shows hot water on stream

Great!

then...

- Thermostat clicks off within a matter of mins, return still stone cold and the water hasn't even made it's way round the system. And the cycle repeats

At no point is the hot water getting further than a few metres into each UFH loop as far as I can tell. The returns on the manifold remain cold and all that is achieved is some gas burnt.

This behaviour tends to exhibit when the room is close to, but not at, the desired temperature. The only way I can get a consistent pull from the boiler is to jack the temperatures up on the thermostats to a crazy level (but even then the thermostats seem to enjoy clicking off even when some way off target temp).

It's all incredibly frustrating and makes little sense to me. I have used this as an excuse to buy a new gadget (thermal camera) and I can see clearly that the closest loop gets a little heat the further loops very little. I have balanced the system but the issue seems to be the ridiculous behaviour of the thermostats constantly calling then shutting off just as the boiler is taking an interest in doing its job... on, off, on, off....

[Trustmeimadoctor]

Don't use room stats

[hast2]

Mate had a new house built and had problems with his heating - It was caused by the room thermostats being linked to the wrong control valves on the manifold. ie. bedroom room thermostat controlling the living room UFH loop etc.

Figured out which one went where by turning off all the room thermostats and then cranking one up to 30C and seeing which room floor got hot.

Or Is it linked to to the Hot Water cylinder? We have ground Source and the diverter valve for the hot water has a habit of weeping past when the UFH is running and the 25-30C water for the floor going through the Hot Water cylinder coil and cooling the tank, causing it to demand hot water more than normal. On hot days when the system isn't calling for heat for the UFH heating it still kicks in and out every 30mins to top up the heat in the hot water. It's happened a few time now

[hast2]

FYI we have ground source and the room stats are all set at 30C and the house temp controlled by weather compensation and adjusted to maintain about 20C.

[b14]

Is the boiler wired for opentherm, or does it have weather comp? Seems weird that the room stats wouldn't work on a bit more of a binary basis - it sounds like one possible reason could be that the room stats are only part of the puzzle in terms of when the boiler fires. If the room stats were the sole control then they should work on a much more "room at temp, turn off" way. Or it's wrongly wired

Edited by b14 on Friday 18th October 10:56

[Trustmeimadoctor]

If using weather comp you don't need room stats really as when well set up the boiler knows how much it needs to provide and when

Edited by Trustmeimadoctor on Friday 18th October 11:13

[b14]

Also what do you have the mixer temp on the manifold set to? If it is too low then all the hot water from the boiler will just get diverted back to boiler causing it to click off. Perhaps a long shot though.

[andy43]

b14 said:

Also what do you have the mixer temp on the manifold set to? If it is too low then all the hot water from the boiler will just get diverted back to boiler causing it to click off. Perhaps a long shot though.

Could be that and also check if there’s a sensor anywhere that’s designed to protect the slab from over temps that is seeing the temp rise slightly then cutting everything instantly - could be on the flow pipe somewhere. Opening the wiring centre might also find something. I had a similar problem and tweaked the sensor cutoff temp up a bit but mine was in a separate white box.
There could also be dip switches or adjustments in the wiring centre that are set wrong - I think a rad needs a different set up compared to ufh.

ETA looking at your pic you’ve two white cables on the bottom left of the wiring centre. Could be to/fro room stats or one could possibly be to an over temp sensor perhaps?

Edited by andy43 on Friday 18th October 11:45

[OutInTheShed]

DoubleSix said:

Hi All,

Wonder if you can help with an issue I'm having.

I have been monitoring the behaviour of our UFH and it doesn't seem right. It's JG system only a year old.

I keep it at a consistent 21c through day and eve, dropping back to 19c overnight. Issue is the behaviour of the thermostat, control unit and boiler.

It's goes like this:

- The thermostat reads 20.4c and calls for heat, control unit light goes on...
- delay (1-2mins)
- Pump kicks into life and boiler fires
- Flow to manifold shows hot water on stream

Great!

then...

- Thermostat clicks off within a matter of mins, return still stone cold and the water hasn't even made it's way round the system. And the cycle repeats

At no point is the hot water getting further than a few metres into each UFH loop as far as I can tell. The returns on the manifold remain cold and all that is achieved is some gas burnt.

This behaviour tends to exhibit when the room is close to, but not at, the desired temperature. The only way I can get a consistent pull from the boiler is to jack the temperatures up on the thermostats to a crazy level (but even then the thermostats seem to enjoy clicking off even when some way off target temp).

It's all incredibly frustrating and makes little sense to me. I have used this as an excuse to buy a new gadget (thermal camera) and I can see clearly that the closest loop gets a little heat the further loops very little. I have balanced the system but the issue seems to be the ridiculous behaviour of the thermostats constantly calling then shutting off just as the boiler is taking an interest in doing its job... on, off, on, off....

....]

You seem to be saying that the act of firing up the boiler for a short while trips the thermstat into switching off without heating the slab?

That suggests heat is leaking from the pipework to the room thermostat?

Or does the boiler switch off because the flow temperature has got too high?
That can be addressed by increasing flow rate (turn up the pump or open up the valves), setting the target flow temp higher, or maybe some advanced stuff with boiler modulation, hysteresis limits

You may have 'load compensation' in action, because the room temp is very close to the set temp, the controls call for a very low flow temp?
Adjusting some settings there may help.

Do you have headers/mixers which set the flow temp to the slab lower than the flow temp from the boiler?

[toyoda_man]

I had the same problem, with the boiler forever cycling when UFH was being used independently of the CH.

Turned OFF the TPI function which solved it:

[DoubleSix]

toyoda_man said:

I had the same problem, with the boiler forever cycling when UFH was being used independently of the CH.

Turned OFF the TPI function which solved it:

We may have a winner!!!

Am just running tests but immediate the behaviour seems normalised.

What a ridiculous setting to have 'on' by default. Will confirm shortly, and BIG thanks! This has been driving me absolutely mad.

[Trustmeimadoctor]

well not really tpi should bet better than just on and off but see how you get on

[tux850]

DoubleSix said:

What a ridiculous setting to have 'on' by default.

On the contrary; it is an excellent control method (second only to variable flow temperature) for underfloor heating given the high thermal inertia of such a heat emitter and the inherent hysteresis of an on/off thermostat. Without TPI enabled you will likely end up with under- and overshoot of target room temperature which is not only wasteful but can significantly affect perceived comfort.

[DoubleSix]

That may well be.

But on the basis that my boiler has been firing on and off constantly for the last two weeks with zero effect on the room temp I'll need a little persuasion to understand how turning this off hasn't massively improved the performance of my system.

If you digested my original post you'll appreciate that my boiler has been hard at work heating nothing, whilst now I immediately have returns at a perfect 10c lower than flow...

I'm sure TPI has a place but something suggests not here. Is it because my boiler is now getting on 12 years old perhaps?

[b14]

DoubleSix said:

That may well be.

But on the basis that my boiler has been firing on and off constantly for the last two weeks with zero effect on the room temp I'll need a little persuasion to understand how turning this off hasn't massively improved the performance of my system.

If you digested my original post you'll appreciate that my boiler has been hard at work heating nothing, whilst now I immediately have returns at a perfect 10c lower than flow...

I'm sure TPI has a place but something suggests not here. Is it because my boiler is now getting on 12 years old perhaps?

Some thermostats will, alongside having TPI, also ask you to specify the heating they are controlling - our Salus thermostat in this rented house we are in does TPI really well but it has the option to tell it that we're using rads rather than UFH. It feels like yours might be trying to do TPI on rads which react much faster than UFH of course.

[OutInTheShed]

tux850 said:

On the contrary; it is an excellent control method (second only to variable flow temperature) for underfloor heating given the high thermal inertia of such a heat emitter and the inherent hysteresis of an on/off thermostat. Without TPI enabled you will likely end up with under- and overshoot of target room temperature which is not only wasteful but can significantly affect perceived comfort.

These things are only any good if the parameters are close to being correct.

We have load compensation on ours, but the only way we've managed to get it working well is to set the programmer's target temp to about 4 degrees higher than we want, then actually control the temperature with TRVs.
Previous house had weather compensation which the boiler maker's tech guru for the whole of the UK couldn't get to work acceptably.

I've done a bit of work with industrial processes, that's usually easier than central heating because you get to put the sensors where they are needed and you can get the software written properly.
Everybody is trying to be very clever, looking for another % of efficiency for the badge, to the detriment of actually doing what the customer wants.

[tux850]

DoubleSix said:

But on the basis that my boiler has been firing on and off constantly for the last two weeks with zero effect on the room temp I'll need a little persuasion to understand how turning this off hasn't massively improved the performance of my system.

The 'zero effect on room temp' is actually a symptom of success! The whole point of the thermostat is to maintain the target temperature you've set.

DoubleSix said:

If you digested my original post you'll appreciate that my boiler has been hard at work heating nothing, whilst now I immediately have returns at a perfect 10c lower than flow...

Note however that the temperature drop is just one aspect of a system design target, and one that the thermostat isn't all that concerned about - its job is to maintain a target temperature and, ideally, to do so as efficiently and effectively as it is able to do so. Inevitably there can be some conflict and so ultimately it is down to you to decide where your preferences lie.

DoubleSix said:

I'm sure TPI has a place but something suggests not here. Is it because my boiler is now getting on 12 years old perhaps?

Certainly with an old(er) boiler, or more to the point a boiler that can only accept a binary heat/no-heat demand input, you are limited in how good a role it can play in any closed loop control system. It is akin to controlling the altitude of a hot air balloon with only the ability to turn the burners on or off and no option for anywhere in between. Pulsing the burners before you start to fall, whilst learning through experience not to overdo it to avoid rising up, will result in maintaining the steadiest level. It's not perfect (an infinitely variable burner control would be far better) but what else can you do with only an on/off control?

See how you get on with TPI turned off - you might find things are perfectly fine. I wasn't trying to convince you to turn it back on, just trying to explain why it was on in the first place!

Edited by tux850 on Friday 18th October 17:35

[DoubleSix]

“The 'zero effect on room temp' is actually a symptom of success! The whole point of the thermostat is to maintain the target temperature you've set.”

That’s not lost on me. But the room, due to modern insulation has been a consistent 19c. I turned the UFH off for a couple of days… no effect. Turned it back on… still 19c. The only change was some mildly warm flow pipes and a hell of a lot of boiler activity. Changed this TPI setting and we’re immediately at target temp 21c and the boiler has clicked off.

I’m not dismissing your knowledge here, of course a realtime analogue system modulating control would be wonderful thing.

Grateful for all input.

Edited by DoubleSix on Friday 18th October 17:46

Edited by DoubleSix on Friday 18th October 17:53

[DoubleSix]

Just to round this thread out, i have been monitoring behaviour since changing the TPI setting and all seems well.

Rather than endlessly cycling the boiling from dawn ill dusk, the boiler now fires for a full hour or so in the morning, the slab warms and the 150mm insulation keeps the room warm for the rest of the day. It literally doesn’t fire again for the UFH.

Admittedly, we’ve not seen proper low temps at all down here in the SW yet, but i am far happier with this consistent and predictable behaviour. It has to be better for the boiler and, crucially, the room and floor are actually warm.

I’ll monitor this as the temps dive over winter, but for now I’m very happy with change.

[b14]

Great result. Will be interested to see how it fares as things cool down more too.