I decided to measure the temperature under my solid uninsulated floor

So far at 250mm down it is 18.75 C and at 450 it is constant 18.5

I am doing a run at 750mm deep over the next 48 hours

With this very slow temperature gradient the rate of heat flow must be low I think and therefore the heat loss much smaller than generally accepted.

its all about your perimeter over area calculation

very little loss over a warehouse, say, but huge loss from the slab of a garden privvy.

what is the footprint of the building you are measuring?

20 C floor area is 80 m^2 and my probe is almost in the middle.

Jacky -- you give me a year and I ll do that.

Posted By: tony
With this very slow temperature gradient the rate of heat flow must be low I think and therefore the heat loss much smaller than generally accepted.

Other way isn't it? A low gradient implies good conductivity. A perfect conductor has zero thermal gradient.

Posted By: CWatters
Other way isn't it? A low gradient implies good conductivity. A perfect conductor has zero thermal gradient.

It would do, but for a given rate of heat flow. Tony is right because the heat flow is determined by the temperature gradient, not the other way about. ...plus of course the conductivity of the material under his floor, and the floor area. The gradient, as the good Doctor says, is likely much higher near the outside walls.

Tony,
18 and 18.5 degC seem surprisingly high at those depths. Isn't average soil temp in the UK around 10 - 12 DegC?
That means you're heating a lot of subsoil. I know you're not an advocate of excessive insulation beneath slabs but here you're surely proving the case for it?
The 750mm depth result will be interesting.
Coincidentally 750mm is the recommended depth for burying water pipes to avoid ground frost damage

What's the temperature 750mm down in the garden? For that matter what's the outside air temperature?

It may be looking dated but I'd have a look at "The thermal performance of ground floor slabs - a full scale in-situ experiment" by Thomas and Rees 1998 (from the Geo-environmental Research Centre, Cardiff) in Building and Environment. Though the technology may have moved on a little I imagine that this will give you many of the pointers that you need.

Mark

Posted By: Saintyou're heating a lot of subsoil. I know you're not an advocate of excessive insulation beneath slabs but here you're surely proving the case for it?

tony's aim is to heat the subsoil, until it's 'filled up' - isn't that right tony? The point is, do these measurements support the hope that it's not dissipating continuously but, having spent a few years 'filling up' is now losing little?

Tony -
To determine the heat flow you'll be needing some idea of the thermal conductivity of the ground that you are measuring the temp difference across.The table of properties below may help. Note the big differences in conductivity for wetter ground - so heat loss near outer walls may be high not just because of the nearness of the outside temp, but also due to the greater conduction if, as I suppose is likely, the ground is damper here than at the centre. Interesting project - all power to your elbow.

Soil Properties (from GS2000) (hope the tabulation survives)


Material.....water......Conductivity....Diffusivity....... Vol Sp Heat
................ %............W/moC ..........mm/s......... Ws/m3.oC x106

Sand dry...... 2..............0.60...........0.42................1.4
.....damp.....7..............1.50...........0.72 ................2.1
...saturated..12.............2.20...........0.72.................3.1
Sandy soil.....2..............1.30...........0.48.................2.7
.................5..............2.50...........0.84.................2.9
.................8..............2.8............1.10.................2.5
Silty loam.....2..............0.60...........0.38.................1.6
................10.............1.30...........0.56..................2.3
................20.............2.3............0.68..................3.4
Silty clay/clay 2.............0.40...........0.33..................1.2
.................10............0.80...........0.48..................1.7
.................20 ............1.10...........0.57..................1.9
Basalt -lwr quartile 1.20 0.60 2.0
- mean 1.60 0.71 2.3
Granite lwr quartile 2.80 0.81 3.5
- mean 3.40 1.28 2.6
Limestone - lwr 2.20 0.98 3.5
- mean 3.40 1.28 2.6
Shale – lwr qtl 1.60 0.49 3.3
- mean 2.40 0.74 3.2

Water.......100..............0.60...........0.14................4.2

Posted By: fostertom

Posted By: Saintyou're heating a lot of subsoil. I know you're not an advocate of excessive insulation beneath slabs but here you're surely proving the case for it?

tony's aim is to heat the subsoil, until it's 'filled up' - isn't that right tony? The point is, do these measurements support the hope that it's not dissipating continuously but, having spent a few years 'filling up' is now losing little?

Humm. I've not heard of a way to change the thermal conductivity of soil by heating it. Not unless you plan to boil off any water it might contain :-)

Posted By: mike7This method is for a wall isn't it? The boundary layer/contact resistance if that's the right term might be quite different for a floor, I'd guess

It is for a wall but there's no reason it shouldn't work for a floor. The "outside" temperature is simply the deep ground temperature (i.e. equal to the annual average air temp). It should be a good first order approximation of the effective resistance of the ground. We used the IR thermometer trick to build the models for hot2000 of my old house - it had been about -12C outside for a couple of days so we felt it would be reasonably accurate.

Paul in Montreal.

Posted By: Paul in Montreal

Posted By: mike7This method is for a wall isn't it? The boundary layer/contact resistance if that's the right term might be quite different for a floor, I'd guess

It is for a wall but there's no reason it shouldn't work for a floor. The "outside" temperature is simply the deep ground temperature (i.e. equal to the annual average air temp). It should be a good first order approximation of the effective resistance of the ground. We used the IR thermometer trick to build the models for hot2000 of my old house - it had been about -12C outside for a couple of days so we felt it would be reasonably accurate.

Paul in Montreal.

I'm thinking that a static layer of air cooler than the general 'ambient' would likely be present over a floor, thicker than the boundary layer against a wall - which would be limited by convection currents. I recon this would skew the result and give a worse picture than actual.

Posted By: Paul in MontrealThe "outside" temperature is simply the deep ground temperature (i.e. equal to the annual average air temp)

But that's modified after pouring heat into the ground for a while. the temp gradient between inside and that primeval deep ground temp gets flatter in gradient as it gets longer in distance, so heat flow diminishes year on year, never quite reaching steady state. It's quite wrong to see the ground as maintaining steady "equal to the annual average air temp" whatever you throw at it, quite wrong to see the ground as a perpetual source of heat loss. tony's right:

Posted By: tonyonce charged up ( = steady state ) then the ammount of heat flow to keep it charged = heat loss = smaller than all calcs give.

except as I say it never quite reaches steady state. All the above ignoring things like groundwater flow.

Posted By: mike7I'm thinking that a static layer of air cooler than the general 'ambient' would likely be present over a floor, thicker than the boundary layer against a wall - which would be limited by convection currents. I recon this would skew the result and give a worse picture than actual.

Maybe - I guess Tony should measure the air temperature just above the floor - also depends how the room is heated as to how much stratification there is.

Posted By: fostertomIt's quite wrong to see the ground as maintaining steady "equal to the annual average air temp" whatever you throw at it, quite wrong to see the ground as a perpetual source of heat loss. tony's right:

I don't disagree with you - but it's probably a close enough approximation for now to calculate the heat flow into the floor. After all, he is measuring a temperature gradient. It might be interesting to measure the temperature of the ground at similar depths that aren't under the floor slab. You're right about the ground not being a perpetual source of heat loss (or gain for that matter) - that's why it's at the annual average air temperature for those places it's not covered by buildings etc. - that's the equilibrium temperature it has come to over a long period of time. I think what could be interesting here is to determine how much umbrella insulation is required to effectively isolate the ground under a house and make it into a usable store (without requiring insulation under it). A lot depends on the water table and transport through the area under the house. I believe this is one of the reasons the Drakes Landing Solar Community is successful - the ground is dry and so the area losses around the edge of the store are small compared to the volume (plus the top is insulated).

Posted By: Mike GeorgePaul, Do you know if is possible to isolate heat loss data in Hot 3000? Say by comparing the same models only one with and one without inulation?

Absolutely. There's a mode where you can calculate the efficacy of an "upgrade" (e.g. basement insulation) in a model. Actually, I've only done this in hot2000, not hot3000 but I have no reason to believe it's any different. Of course, it really depends on how realistic the modelling is. At least test houses have been built to try and validate some of the modelling performed. I'm not sure what was done in the area of slab insulation versus umbrella insulation versus no insulation at all in real test buildings - but there are models for all these cases. From experiments I've performed, underslab insulation doesn't make as much difference as umbrella insulation.

Paul in Montreal.