Does anyone know in plain language the theoretical basis of the P/A floor reistance method in EN 13370? Can anyone slip me a copy?

What does it assume about the isotherm-shape of temp distribution at depth below GL, with seasonal variation, in the subsoil - both under surrounding varying-temp surface of surrounding open ground, and under the constant-temp floor surface of heated building?

There are 2 components to heat loss through floors - the perimeter heat loss and the straight-down heat loss. P/A ratio seems a good basis for calculating the perimeter heat from around the edge of the slab, going down, curving outward and back up to the cold surrounding ground surface. But seems unhelpful in calculating the straight-down heat loss, which should depend straightforwardly on A alone.

The perimeter heat loss is proportional to delta-t - the temp difference between inside and out.
But what is the temp difference to apply to the straight-down heat loss? What is the temp, and at what depth in the subsoil, that heat is being lost straight-down to?

No expert but not very much I almost certain

have a look at this

https://www.youtube.com/watch?v=Jxd6eCmbuW8

not certain he is an expert either but I am sure has more knowledge than me .

The ground under a house will warm up to the average temperature of the house, this can take several years and will follow a seasonal temperature sine wave offset by some weeks or months depending on the insulation levels of the house.

A point under the middle of the slab will loose very little heat as the escape path is the longest, points on the perimeter will loose the most, similar to walls but likely more as any insulation tends to stop too high up.

The p/a is a rough and ready method of determining how little insulation to put under a floor, not for us that. We used to put more under the edges and led under the middle, not so easy as it sounds but more effective use of materials. For very large slabs (commercial) no insulation is needed under the middle the ground has sufficient thermal resistance and ir warms up too.

I use the PA method a fair amount when calculating floor U-Values but I must confess that I use a bit of software to do it and so don't fully understand the maths behind it. I have a copy of that standard you mention so I can check and see if it actually gives a formula.

I also have an excel spreadsheet that I got on the AECB therm course many years ago which allows you to calc the PSI value of a solid ground floor to wall junction based on the PA calc.

I know that definitely makes all sorts of assumptions that may or may not accurately reflect the real building - those may be same as the official U-Value software I use for floors. If I get a spare minute this week I will have a look into it.

Here's an old NRCan paper on basement heatloss with real-world measurements and calculation: http://web.mit.edu/parmstr/Public/NRCan/nrcc23378.pdf

Here's a useful list of tools for building simulation too: https://www.canadianarchitect.com/asf/enclosure_design_tools/digital_tools/digital_tools.htm

NRCan used to provide "Basecalc" as a tool for basement heatloss modelling - it's not available separately anymore but is built into hot2000

Paul in Montreal.

For any who've not heard the obvious caveat before: water. Specifically, the highest water table for the year.

Thanks Timber

Posted By: TimberI have a copy of that standard you mention so I can check and see if it actually gives a formula.

The end-result formula itself is readily available - U = 0.05 + 1.65P/A - 0.6(P/A)^2

Posted By: TimberI also have an excel spreadsheet that I got on the AECB therm course many years ago which allows you to calc the PSI value of a solid ground floor to wall junction based on the PA calc.

I know that definitely makes all sorts of assumptions that may or may not accurately reflect the real building - those may be same as the official U-Value software I use for floors.

That'll be the one(s) in http://www.peterwarm.co.uk/?dl_id=2 ? I agree with yr 2nd para comment.

Thanks Paul, esp for http://web.mit.edu/parmstr/Public/NRCan/nrcc23378.pdf .

I guess the basement model (fig 1) is valid to use for slab-on-grade (N Am parlance) by setting certain terms to zero or as appropriate?

It depends on
'lower thermal boundary at a constant temperature equal to the mean ground temperature'
and
'assumes that there is sufficient groundwater flow to maintain a constant temperature at some depth below the basement floor'.
What do we think of the latter - a bit odd?

Posted By: Ed DaviesFor any who've not heard the obvious caveat before: water. Specifically, the highest water table for the year.

I've been thinking about that, and wonder whether it's an exagerated threat.
http://www.homeintheearth.com/tech_notes/basics-of-earthsheltering/earth/soil-properties/soil-temperature-experiment
shows negligible difference in ground temps at depth, whether surface-sheltered from rain or not. As he says:
'Note: this result is significant because it contradicts the theory found in many earth sheltered umbrella books!'

Not about ground water, I know, but ... Actually it's not just vertical rise and fall of water table that could carry stored heat away, but also horizontal flow of underground water. How to guess what's happening with water in the subsoil beneath your site? Maybe it's not so critical.

First, precipitration - rain and snow-melt percolating down through open ground surrounding the building, which is important because surrounding surface and subsoil temps are what drives the major edge-loss from a slab, esp a high-P/A domestic-size slab as opposed to a low-P/A warehouse-size slab.

First, snow is an excellent insulator, till it melts.
Next, what is the mean temp of rainwater, over the heating season? Above 0C, by definition, and I suspect considerably higher.
Then, what is the rainwater's thermal capacity, totalled over the heating season? How many inches of water? A fraction of the vertical height of the subsoil block that we're interested in. Even allowing that water has 5(?)x the volumetric specific heat of subsoil, by how much will percolating rainfall actually cool the subsoil?

Once water is deep in the subsoil, the stuff of underground horizontal flow and water table vertical movement, why should its temp be any different from the soil it's in? The deeper its source and the longer in the ground, the warmer it should be - it's horribly hot down a mine!

Well, getting the U-value right becomes critically important - but doesn't change anything in principle?

Posted By: ringiRemember you are also trying to stop condensation on the floor in summer.

Say more?

Here's a fantastic resource
https://www.pik-potsdam.de/services/climate-weather-potsdam/climate-diagrams/ground-temperature
Anyone know of similar closer to home? I must go see what ST's favourite Plymouth data has to offer.

At Potsdam, the 12m deep temp seems not all that stable - in fact the 4m deep one seems most constant.

Posted By: fostertomThe end-result formula itself is readily available - U = 0.05 + 1.65P/A - 0.6(P/A)^2

Shome mishtake shurley? The units don't work, and where is the ground conductivity?

Posted By: fostertomI've been thinking about that, and wonder whether it's an exagerated threat.
http://www.homeintheearth.com/tech_notes/basics-of-earthsheltering/earth/soil-properties/soil-temperature-experiment

Interesting link, thanks. Rest of the site looks worth a read, too - bookmarked.

But that's still with the water table deep down and just precipitation soaking through the ground. What about the cases where the water table is close to, at or even above ground levels?

Actually it's not just vertical rise and fall of water table that could carry stored heat away, but also horizontal flow of underground water.

Absolutely - that's mostly what I'd worry about.

How to guess what's happening with water in the subsoil beneath your site?

Well, if it pisses out of the ground on the downhill side of your house and runs across the road that's a bit of a hint.

OK, northern part of Scotland tends to be a bit of an extreme case but, for example, lots of Norfolk and Lincolnshire would be similar with more-or-less continuous water a metre or so down with a steady horizontal flow towards the nearest drainage ditch.

Sure, all of that Ed. But what about

Posted By: fostertomwhat is the rainwater's thermal capacity, totalled over the heating season? How many inches of water? A fraction of the vertical height of the subsoil block that we're interested in. Even allowing that water has 5(?)x the volumetric specific heat of subsoil, by how much will percolating rainfall actually cool the subsoil?

If rainwater's not that cold, and doesdn't cool the ground that much, then what makes it supposedly colder than the ground it's in, once it's joined the water table and/or is flowing horizontally underground?

Posted By: fostertomOnce water is deep in the subsoil, the stuff of underground horizontal flow and water table vertical movement, why should its temp be any different from the soil it's in? The deeper its source and the longer in the ground, the warmer it should be - it's horribly hot down a mine!

Posted By: mike7a conductivity of 1W/mK

Not so far off correct.

Posted By: fostertomOnce water is deep in the subsoil, the stuff of underground horizontal flow and water table vertical movement, why should its temp be any different from the soil it's in? The deeper its source and the longer in the ground, the warmer it should be - it's horribly hot down a mine!

The point is that it's the same temperature as the soil that's not underneath your building, so if it's only 1 m under your building then all you've got between the bottom of your floor and 4°C is a metre of damp earth in winter. You don't get any heating of the earth under your building; it's a thermal short-circuit.