GSHP with UFH

[Ed Davies]

1 hour ago, Jeremy Harris said:

of specific heat

 

What would non-specific heat be? Or do you mean sensible heat?

 

“Specific” in contexts like “specific heat capacity” means “per unit”, which can be taken to mean per unit mass unless otherwise, err, specified.

[Jeremy Harris]

34 minutes ago, Ed Davies said:

 

What would non-specific heat be? Or do you mean sensible heat?

 

“Specific” in contexts like “specific heat capacity” means “per unit”, which can be taken to mean per unit mass unless otherwise, err, specified.

 

 

Sorry, typos, should have been sensible heat. . .

 

I'll edit it so it makes sense.

[SteamyTea]

Anyone wish to comment on @Nickfromwales's comment about a thin slabs needed higher temperatures for longer to counteract thermal instability.  Just does not ring true to me.

 

[ProDave]

1 minute ago, SteamyTea said:

Anyone wish to comment on @Nickfromwales's comment about a thin slabs needed higher temperatures for longer to counteract thermal instability.  Just does not ring true to me.

 

Previous house and current house heated by UFH on a timber floor in a 25mm biscuit mix, so I guess that qualifies as a "thin slab"

 

No problems at all. Just heat it with low temperature water and there is no significant overshoot,  the house maintains a steady 20 degrees barely fluctuating by more than +/- half a degree according to the  thermometers in each room.  And I only use basic mechanical room thermostats which imho work perfectly well..

 

Upstsairs temperature varies a bit more as there is no heating there at all so in a cold spell may be 2 degrees lower than downstairs.

 

[Ed Davies]

1 hour ago, SteamyTea said:

Anyone wish to comment on @Nickfromwales's comment about a thin slabs needed higher temperatures for longer to counteract thermal instability.  Just does not ring true to me.

 

Doesn't ring true to me, either. I'd think all floors, other things being equal, would need to take the same amount of heat long term but a low-heat-capacity one would benefit more from having lower-temperature water for more time. If he can present a plausible reason why that might not be so we can at least have a sensible discussion.

 

1 hour ago, ProDave said:

Previous house and current house heated by UFH on a timber floor in a 25mm biscuit mix, so I guess that qualifies as a "thin slab" … And I only use basic mechanical room thermostats which imho work perfectly well..

 

That make sense as far as it goes. It doesn't follow that a basic mechanical room thermostat would work so well with a thick slab, though.

[ProDave]

1 minute ago, Ed Davies said:

 

That make sense as far as it goes. It doesn't follow that a basic mechanical room thermostat would work so well with a thick slab, though.

Why not?

 

The main criticism given of mechanical room stats is hysteresis.  but if properly wired they have a built in "accelerator heater" (just a resistor) that overcomes the hysteresis and makes them pretty accurate and tight.

[Jeremy Harris]

What's the practical difference between a thin floor with UFH and radiators, as far as this goes?

 

Stability is going to be governed by the thermal time constant of the house, I'd have thought.

[Ed Davies]

18 minutes ago, ProDave said:

The main criticism given of mechanical room stats is hysteresis.  but if properly wired they have a built in "accelerator heater" (just a resistor) that overcomes the hysteresis and makes them pretty accurate and tight.

 

Agreed. The mechanical thermostat in my living room holds the temperature there to within about 0.25 °C. That's not the problem.

 

With a thick slab, though, won't there be a possibility that too much heat will be dumped into the slab before the thermostat begins to see a rise in temperature resulting in an overshoot? Whether that's actually a problem in practice is a separate question but not one answered by your quoted experience.

[ProDave]

A heating system is just a form of a "control system", something that was a main part of my employment in previous years.  At it's most basic, any control system that overshoots has too much gain in the system somewhere (most of my control system work was to do with motion control)   That may well be with a heating system that the slab is heated too hot, putting too much heat into the room for instance.

[Jeremy Harris]

1 hour ago, Ed Davies said:

 

With a thick slab, though, won't there be a possibility that too much heat will be dumped into the slab before the thermostat begins to see a rise in temperature resulting in an overshoot? Whether that's actually a problem in practice is a separate question but not one answered by your quoted experience.

 

Yes, our experience is that this can happen. 

 

I originally started off by trying to control the slab temperature, on the basis that if that was held to a target value relative to the outdoor temperature it would effectively self-regulate.  Despite my best endeavours I never managed to get this control system to work OK.  I think I know why, and it was probably related to where I chose to fit the slab temperature sensor.  I embedded a DS18B20 sensor in the slab, under an internal stud wall and pretty much equidistant from UFH pipes running either side.  My thought at the time was that this should make the sensor relatively immune from variations in heat loss from the surface of the slab.  The snag seemed to be that the time lag between heat being applied to the floor and the sensor measuring a temperature rise was just too long.  By the time the sensor started to register an increase, the UFH had already pumped too much heat into the slab, so the temperature would overshoot.  The overshoot could be controlled by reducing the flow temperature, which reduced the amount of sensible heat that went into the floor before the system shut off from the temperature rise.

 

I ended up removing my control system completely and reverting to a wall thermostat.  By using one that has a hysteresis of 0.1°, together with careful control of the flow temperature, I've found that that the overshoot can be kept down to under half a degree, which seems OK.

[joe90]

On a previous thread someone controlled their UFH by mixing the flow temp to 1 or 2 degrees above required room temp. This was self regulating but required a mixer that could deliver a low temp with small hysterisis. My wall thermostat has too big a hysterisis, must get round to changing it.

[Ed Davies]

1 hour ago, ProDave said:

A heating system is just a form of a "control system", something that was a main part of my employment in previous years.  At it's most basic, any control system that overshoots has too much gain in the system somewhere (most of my control system work was to do with motion control)   That may well be with a heating system that the slab is heated too hot, putting too much heat into the room for instance.

 

Talking about thick slabs is a bit tangential (sorry for the distraction) but mildly illuminating. Yes, the real control to make a wall thermostat work is that you don't heat the slab too hot. Fine with ASHP, perhaps, but might well be more awkward with, say, a non-modulating boiler (e.g. an oil boiler) unless you add a buffer tank as well or have a control system somewhat like @TerryE's where you count how many joules you're putting into the slab.

 

Which brings us back to the @SteamyTea vs @Nickfromwales discussion: why might it be better to put higher-temperature water into a thin slab for controllability? I think we need to know more about the control system which @Nickfromwales has in mind if we're going to get anywhere. I'm having a hard time imagining one which would work better with hotter water rather than water just a few degrees over the desired slab temperature.

Edited February 14, 2020 by Ed Davies

[joe90]

As far as I can see, Fir a given thickness of slab, hotter water means quicker response but bigger chance of overshooting, I heat our house at a set temp of 21’ (if the wall stat had a decent hysteresis).

[Dan1983]

Thanks for all of the advice guys.

I think air source heat pump is going to push me over budget as being told it will add approx 10k to the quote.

Really not a fan of combi boilers so may have to look at an unvented system boiler with 200ltr tank