underfloor heating time constant

[news]

A tenant of my boss's is kicking up about the gas fired underfloor heating, he keeps on fiddling with the program which basically has 6 time settings and I had left it to drop to 15C from 9:00 to 15:00 then rise to 23 till 21:00 and22:00 then drop back to 18 till 5:00 when it rises to 21 C again.

He switches it off in the 9:00-15:00 and 22:00-5:00. This means bungalow is cold when he gets up at 7:30 and overshoots to 26C mid evening.

Never having been involved with underfloor apart from this building what is the likely time lag on a 50mm screed over 100mm celotex in a building drylined with 50mm celotex and 200mm fibreglass in the roof?

AJH

[geoff]

In message , [e-mail] writes

I think you have to measure it - somewhere in the order of an hour, IWHT

[Dave Liquorice]

Best way something that can record the temp every 5 mins then export that data and plot it to see how fast the place can be made to warm up.

I think it will be some what slower than that. It might make 0.5C/hour on the room temp. 50mm of screed has a lot of thermal mass, as indicated by the massive overshoot.

In my view underfloor heating is good to provide the baseline heat required in a longish time scale of many hours. Rather than what many now expect switch on the heating at 15C and be nice and be a cosy 20C in less than an hour. Because underfloor heating is not very good at such "instant heat" you do need something else that will heat the room(s) quickly should there be a need.

[Tim Lamb]

In message , [e-mail] writes

I take it the tenant is paying the gas bill.

My understanding of wet under floor systems is that they are best left running with local roomstats to control the temperature. Obviously this requires a manifolded system with individual control valves.

I would say that the overshoot indicates an over high water temperature.

Could you do some sums based on the building thermal losses and show how little gas is saved by this interference?

[harryagain]

There are long time lags on under floor heating buried in concrete. (hours) It is fairly pointless turning the heating off at all on a daily basis. The best you came expect is a swing of three or four degrees. A lot depends on how thick the concrete is and how much insulation is below it. The only alternative is to use the floor heating as background heat and some other source as an intermittant top up.

[Richard Russell]

Exactly, that's what we have here - 6 independent zones with their own roomstats (and clocks that need changing twice a year!). They are mostly programmed for 17C overnight and 20.5C during the day, but the system is never 'off'.

Richard.

[Bill Taylor]

I installed UFH in a renovated shower room earlier this year and I put

1-wire sensors in various places. The floor dropped from 27C at 21:30 last night to 21C at 06:00 this morning. Yesterday it rose from 19C to 27C in 4.5 hours. The room temperature dropped by about 1C overnight. This is with about 80mm screed over 100mm PU insulation.

Sounds as if you need to fit a dummy controller!

[Steve Firth]

wrote: [snip]

With a similar set up and 37C water supply to the UFH with a Heatmiser manifold and zone control it takes about 90 minutes to get to "erindoors" comfortable with an external temperature below 9C. This can be improved on by programming the zones to warm up bedrooms and bathrooms first then the living areas.

Like you I find it best to keep all of the areas around 16 when not in use. You might consider dropping the night time temperature.

Tell the silly bugger to stop pissing about or to sling his hook.

[Martin Brown]

On 29/10/2012 23:52, Dave Liquorice wrote: > On Mon, 29 Oct 2012 22:50:17 +0000, geoff wrote: >

You probably only need to measure it every half hour to get a decent value for the time constant. It will be sl-o--w. If there are thick carpets on top of the floor then it could take even longer.

If you tell from a cold start to aim for ambient temp +10C then the time to get to +7C will be a rough approximation to the time constant.

My guess would be 2-3 hours of thermal inertia in the floor mass.

Understatement if ever there was one. It is useless at fast heating. You warm up the air inside the room for that.

[The Natural Philosopher]

6-8 hours

Really the only way to run that mass of screed is on 24x7 settings. That way it doesn't overshoot and will actually use less fuel!

[AlanD]

I have wet UFH in my conservatory and it takes 2-3 hours to get up to temperature, longer on COLD days. It's on it's own zone with an intelligent thermostat, which helps however.

[news]

Thanks all of you, Tim I'll check if he has fiddled with the injection temperature.

When I laid the coils in the floor I always intended to fit a fan coil in the kickspace of the kitchen for a rapid warm up. The boss's regular plumber then went his own way.

AJH

[Peter Parry]

Several hours. We have ours aiming for 21deg starting at about 14:00 but going off to 16deg at about 20:00 - that gives plenty of heat in the evening and the temperature starts to fall about 22:30. It aims for 19 deg starting at about 05:00 and getting there about 07:00. At

0800 it drops back to 16 deg. With warm floors the room temperature doesn't need to be high to be comfortable in the morning when people tend to be moving about. The thermostat is a Honeywell one and doesn't switch the system off - just controls it to varying temperatures.

[news]

This too is a Honeywell one which acts similarly.

Boss told the chap not to fiddle with the controls and, thanks to Tim's suggestion, have backed the mixer control from 50(Max) to 40,

Everything seemed hunky dory by 17:00

AJH