For you my friends I have moved a temperature probe to outside 1m down, 1m away from house initial result 12.7 C.

Perhaps we shouldn't get too hung up on this. Most of the heat loss is sideways around the perimeter. I guess there is a case for lining the sides of foundation trenches with insulation. Interestingly we have that although it was done to prevent heave on clay soil.

CWatters, you've switched - originally

Posted By: CWattersI've not heard of a way to change the thermal conductivity of soil by heating it

now

Posted By: CWattersYou can "fill" something with heat but that doesn't generally change it's thermal resistance

Conductivity is different from resistance. Without changing the conductivity we can certainly change the effective resistance between inside and the primeval "equal to the annual average air temp" subsoil - by pushing further away the boundary where the 'primeval' subsoil meets our expanding bulb of warmed subsoil. That's what 'filling up' means - an expanding bulb of heat spreading from the injection point out through the subsoil, hottest at the injection point, tapering to "equal to the annual average air temp" at the advancing boundary. The temp gradient gets extended over an increasing distance or thickness of subsoil of unchanged conductivity. You can look at it either that an ever-increasing thickness of subsoil is being brought in as an insulant, or that the temp gradient is flattening. Either way, heat loss declines steadily quickly in the first, slower and slower in succeeding years, asymptotically never quite reaching a new steady-state.

Going back to what Mike7 posted, and hoping I have my units right...
Silty loam, 10% water has a conductivity of 1.3 W/mC
So if the stable layer is normally 1m down the U value is 1.3 W/m^2C
To improve that to an equivalent of say 0.2 you would have to push the bubble down to 1.3/0.2 = 6.5 meters?

<blockquote><cite>Posted By: tony</cite>For you my friends I have moved a temperature probe to outside 1m down, 1m away from house initial result 12.7 C.</blockquote>

That compares quite well with my guesstimate sketch which gives 12.2 or thereabouts. Got any figures for the floor nearer the outside wall? Is the soil noticably damper 1m down outside than in?

<blockquote><cite>Posted By: CWatters</cite>Going back to what Mike7 posted, and hoping I have my units right...
Silty loam, 10% water has a conductivity of 1.3 W/mC
So if the stable layer is normally 1m down the U value is 1.3 W/m^2C</blockquote>

That's right, but it is a U value for a bit of floor which is only exposed to a deltaT of 1.2C from Tony's figures. To get the equivalent Uvalue for exposure to the full 10C deltaT, multiply 1.3 by 1.2/10 to get 0.156 W/m2C...... or 0.072 W/m2C using Tony's estimate of conductivity. We're getting warmer.

Tony,

can you give us a few more details of the construction of your floor slab and associated perimeter insulation. How are you digging a hole in the middle of your floor to get the temperature probe in? What sort of probe are you using and how are you doing the probing? Just interested!

Paul in Montreal.

A very useful test case then. What's the length x width ratio that makes up your 80m2, tony? Has it been steadily maintained at 20oC av for years, so well equilibriated? (that's a gd word!)

<blockquote><cite>Posted By: fostertom</cite>Sort of ... and mike7, we've done some basic calcs and it seems that heat loss (after near-equilibrium is reached, after a year or three) drops rapidly as the notional 'bubble' gets bigger, but after a point the the further reduction in loss gets negligible, with increasing 'bubble'. In particular, a 6m cube (single house size) 'bubble' suffers equilibrium loss about twice that of a near-infinite size 'bubble'. That, I think, we can all live with, and proceed to discuss single-house size schemes, knowing that even expanding the 'bubble' to multi-house estate scale doesn't hugely improve matters. Anyone comment on that?</blockquote>

How do your calcs compare with the figures/formulae djh and I came up with? Any chance of a look? Our original conclusion was that there would be a real annual maintenance charge of heat required to maintain the bubble ( or thermal bund wall as I've come to think of it), and that it could be a significant proportion of the heat stored. That leads to the question: "What proportion of the heat delivered to the store needs to be recoverable for the store to be considered viable?"

I'm anxious only to point out that the single house notion is somewhere on a dangerously steep slope between viable and not, so it is vital to know just where "to avoid disappointment". The recent 'Massive Thermal Store' topic by Jon touched on this, and there it seemed to me that you only needed to double the linear scale for the numbers to look quite promising.

I don't want to hijack Tony's interesting thread here - should we continue on the original ....John Holt (shouldn't that be Hait?) thread?

Edit: PS What's a P/A approximation?

Tony, any idea how far down your water table is? Or, for that matter, how far down it'll likely to be at the end of January if we have a wet winter? It seems to me that the moment the "bubble" touches ground with any water flowing through it the losses will shoot up enormously.

Tony could you get some longer drill bits or an extension ? just wondering what the temp is internally near outside wall @ 2meters depth, what an excellent thread.
tom

after all the help I've gained form this forum, I'll gladly make you an extension for an SDS ?
tom
P.S. whisper your postal address, you'll have it next week !

Fostertom, while Tony is otherwise engaged toasting sandwiches:-

<blockquote><cite>Posted By: mike7</cite><blockquote><cite>Posted By: fostertom</cite>Sort of ... and mike7, we've done some basic calcs and it seems that heat loss (after near-equilibrium is reached, after a year or three) drops rapidly as the notional 'bubble' gets bigger, but after a point the the further reduction in loss gets negligible, with increasing 'bubble'.</blockquote>

How do your calcs compare with the figures/formulae djh and I came up with? Any chance of a look? </blockquote>

Specifically, are your calculations for three dimensions?

Also, djh's/my results seem bomb-proof to me - bring on your bomb!

...Aaaand another thing: You say above 'the further reduction in loss gets negligible' - you mean the absolute level of loss is approaching a constant non-zero value? If so, I agree, and that's the figure we're after.

At 0.5m 18.5

3.5m

Which got overlaid