Can we think more about what happens, moisture-wise in the insulation-filled underfloor voids we're discussing in http://www.greenbuildingforum.co.uk/forum114/comments.php?DiscussionID=8571&page=2

To recap, I'm proposing two variant/cases; a newbuild version, and a retrofit one;

The newbuild one (simplified): 35mm heavy skin (tiles on 25 Cement Particle Board) on 3x7 treated joists (on ground beams) surrounded with Leca incl 35mm under the joists so 205 of Leca at the floor centre, increasing a little as the DPM covered subsoil slopes outward for drainage. The CPB is a VCL. Around the perimeter, the EPS EWI is carried down in narrow-bucket trench to 1m below FL and the Leca buts up against the EPS's inner face. So the filled void is isolated by DPM from subsoil below; has VCL above, but around the perimeter is vapour-open to the surrounding subsoil outboard of the floorplan, via the EPS. No through-ventilation of the free-air spaces between the Leca granules.

The retrofit one: Leca is simply pumped into an extg building's underfloor void, in direct contact with subsoil and surrounding foundation walls, and up to u/s of floorboards, surrounding the joists. Before even thinking of doing this, any liquid water penetration wd have to be cured, and any rot dealt with. So the filled void is vapour-open to top, to subsoil, and around the perimeter to the surrounding subsoil outboard of the floorplan, via the surrounding found walls and any EPS EWI (optional) carried down to found level. No through-ventilation of the free-air spaces between the Leca granules.

So in both cases there's exposure of the unventilated void's insulation fill, to potential semi-permanent source of evaporative moisture; perhaps slight in the newbuild case, maybe considerable (incl subsoil surface) in the retrofit case.

Here's the question: what effect does that exposure have on the RH of the free-air spaces between the Leca granules? (unventilated, remember). Does the free-air eventually become fully saturated throughout?

The free-air's upper levels (close to floor deck underside) will be much warmer than its lower levels (and than its perimeter, if there's no EWI downstand), even when steadily mixed by buoyant convection. Does that mean that any saturation (and condensation) will be limited to the colder lower levels (and perimeter), and that RH will decrease upward as it gets warmer?

Or does it mean that the whole will be at high RH, meaning that abs water content of the free-air will be higher where warmer?

Does the level of airtightness of the floor make a difference? I think if it is fairly air tight (Chipboard floor & lino v Floorboards and carpet) then there must be a different risk of condensation (not sure which way though!).

Leca has low capilary action but what happens if dirt is washed into it?

Posted By: CWatterswhat happens if dirt is washed into it?

In the newbuild version, open only to the perimeter, geotextile in the EWI trench. In the retrofit version, where's the dirt coming from? We've already stipulated all liquid water ingress to be fixed first.

I assume the Leca is pretty vapour permeable. If so then the partial pressure of water vapour will equalize throughout the underfloor area given that the vapour flows should be very slow. Therefore the density of water vapour (g/m³) will be about the same throughout as well: a little higher where it's cold and a little lower where it's warm. The relative humidity in the warm bits will consequently be lower.

Really, though, you'd need to put some numbers to the permeability of the Leca, EPS, soil and so on to have much confidence.

Why only 35 mm of Leca under the middle of the floor?

Posted By: Ed DaviesWhy only 35 mm of Leca under the middle of the floor?

It's 170 joist depth + 35 = 205 of Leca, equiv insulation-wise to 60 of EPS.

From http://www.greenbuildingforum.co.uk/forum114/comments.php?DiscussionID=8571&page=1

Posted By: fostertomthe block of subsoil under the floorplan is enclosed in 200 EPS downstanding 1m, and that subsoil block is in thermal contact with the interior, via Leca of varying thickness under the CPB and tiled finish. This is an attempt to make an effectively heavy solar-absorbing floor (because tall wide glass area to SE) in contact with its subsoil (protected from edge-loss) but with a modicum of insulation to ensure that floor surface keeps cosy.

'Cosy' means near enough at internal air temp, thanks to the modicum of insulation under it, rather than pulled down (slightly) to the probable year-round 17C of the top of the subsoil.

Posted By: fostertomThis is an attempt to make an effectively heavy solar-absorbing floor...

Ah, good, that makes sense.

Posted By: fostertom

Posted By: Ed Daviesthe partial pressure of water vapour will equalize throughout the underfloor area ... Therefore the density of water vapour (g/m³) will be about the same throughout as well

Does it mean that?

Yes, I believe it is the partial pressure which is uniform in the absence of anything significantly impeding the movement of molecules. Density and relative humidity can then be derived from the pressure and temperature - as the temperature will not be uniform under this floor the density and RH will not be either. The RH is much more sensitive to temperature than the density so will be a lot more non-uniform.

I haven't done any formal physics since A-level so don't take my word alone but this is pretty basic stuff and I'm moderately confident of it. Still, like I said above, I think you need to run the numbers on the whole setup.

OK I need to run the nos - about what RH the air in a void will reach when exposed to a supply of water by evaporation from (some of) the void's walls.
Is that a fair statement of the question? (please don't say 'dunno, it's your question')
I need help with what the controlling parameters wd be.
Obviously air temp, temp of the water-laden evaporative surface - what else? something about the permeability of the evaporative surface?
Behind it, I'm asking why shouldn't such an evaporative surface make the air in a sealed void completely saturated? Why wd the air's RH be less than 100%? (I'm hoping it wd be less).

As a second step, I want to clarify what the effect is on that moist air, if the air is warmer at top and colder below (due to temp of adjacent surfaces).

Posted By: fostertomBehind it, I'm asking why shouldn't such an evaporative surface make the air in a sealed void completely saturated? Why wd the air's RH be less than 100%

Because the whole environment is trying to get into equilibrium. If you just think about the water vapour, and ignore all the other gases in the atmosphere, it's easy to think of it trying to equalize pressure everywhere ( as Ed says, pressure, not RH). And for long time scales, slow movement, that's a fair estimate. So the pressure under the floor will be roughly the average (and ST can tell us exactly what kind) of the water vapour pressure in the house, in the soil and in the local atmosphere. If the permeability of the route to/from atmosphere is dominant, it will have most influence. So in winter when the pressure is low (but the RH is typically high) the pressure under the floor will be low and if it is warmer than outside then the RH will be low. In summer the outside pressure is much higher though the RH may be lower. But underneath the floor, it may well be cooler than outside, so the RH will be higher and there may be condensation.

There's lots of stuff about condensation in crawl spaces. I'm also pretty sure there are papers about the RH variation through the thickness of permeable insulation but I don't remember any titles at the moment. Maybe for straw bales or wood fibre panels.

Posted By: fostertomED, ST or anybody, seems to me that the PVP of a water vapour sample at fixed volume and fixed water density g/m3 will be higher if warmed and lower if cooled. Correct, or not?

Yes. Water vapour is just a gas like any other and follows (to reasonable approximation - no gas is ideal) the normal gas laws (PV/T is a constant, etc).

If correct, then if the sample is temp-stratified i.e. made warmer at the top and cooler at the bottom, and if it's PVP that will equalise throughout the sample, then water density g/m3 will have to be lower at the top and higher at the bottom.

That's the only way, I think, that the warmer top vapour can equal the PVP of the cooler bottom vapour.

Yes.

If PVP is equalised throughout, by varying density inversely with temp, then I think that means that RH is also equalised throughout.

This doesn't follow. Because the other gases in the air (nitrogen, oxygen, etc) also follow the gas laws the ratio of the densities of the water vapour and dry air at the top and bottom of the volume will also be the same. However, that ratio is not the RH. See ST's link above:

http://hyperphysics.phy-astr.gsu.edu/hbase/kinetic/relhum.html

The “saturated vapour density” (I think that's a confusing term and much prefer the equivalent but less misleading term “equilibrium vapour density” but let's not get sidetracked) is much more strongly dependent on temperature than the simple density of the air at normal sorts of temperatures. For a 10°C increase in temperature the saturated vapour density will about double but the densities of the water vapour and of the dry air will only decrease by about 1/30th (at around 300 K).

This is important, to get a qualitative picture before hoping to 'run the nos'. It makes quite a difference, in principle, to the feasibility. I can't find the answer by googling.

Posted By: Ed DaviesPV/T is a constant

Says it all, Avogardro would be proud. As would Hooke, Boil, Einstein and Robert Brown knew a bit about it too.

As air pressure will vary relatively little, just concentrate on the temperatures for now.

As Ed pointed out earlier, dew can starts at 95%RH, so use that as a figure (your doing a risk analysis).
Treat the Leca as a solid and find out the temperature gradient (from the Conductance). You will need to know the seasonal ground and air temperatures and RH figures.
Not sure about what effect earth will have, though I suspect it will act as a damper that has the ability to soak up a semi infinite amount of moisture.

Hmmm I disagree!

The way PV= znRT works for your sample is: if T increases then V increases (expansion) and P doesn’t have to change. Your sample has a fixed total pressure (=atmospheric), not a fixed volume. Atmospheric air pressure does not change linearly with temperature, as we know from the weather forecast!

Posted By: fostertomED, ST or anybody, seems to me that the PVP of a water vapour sample at fixed volume and fixed water density g/m3 will be higher if warmed and lower if cooled. Correct, or not?

Incorrect in your case. The PVP is (defined as) the total air pressure, roughly 1 bar, multiplied by [the fraction of the air that is water vapour]. This does not change along the temperature gradient, because you said the absolute humidity in g/m3 of water in your sample is fixed.

Aside: to work out this sum, the AH should not be expressed in g/m3, it has to be expressed in mole fraction. (a below-ground construction joke there, incase you missed it). To convert from g/m3 to mole fraction, multiply by 0.0013. (only for water in air).

Lets say you had 10 g/m3 of water vapor. The AH would be 10x0.0013 = 0.013 mole fraction. The Pp of water would be 1bar x 0.013 = 0.013 bar = 13 millibar. Irrespective of temperature.


This story would change if there was also some liquid water present. Then temperature would start to have an effect, because increasing the temperature would encourage the liquid water to evaporate, which would increase the AH hence the Pp. The evaporation would continue until the Pp equaled the Saturated VP of water (which is temperature dependent).

Diffusion in air is driven by vapour pressure, not RH, so the water will diffuse until its PVP and AH is the same everywhere, irrespective of any temperature gradients.

Diffusion through hygroscopic materials like wood is not driven by VP, but by lots of other stuff like capillary tension. I dont know if LECA is hygroscopic or not. Assuming that it is not:

At equilibrium, your floor will have the same water vapour pressure everywhere, so the same absolute humidity everywhere. But it will have a temperature gradient, so the relative humidity will be low in the warm spots, and high in the cold spots. (Like you intended)

If the cold spots get so cold that RH=100%, then liquid water will condense , locally reducing the AH and forcing water vapour to diffuse from the warm spots to restore eqbm, so reducing the AH of the entire floor. (how a dehumidifier works). This drying effect (beneficial) can only work if there is condensation happening somewhere (maybe not beneficial).

If there is inflow of water vapour through porous floors, walls, subsoil then that will increase the AH.

These two effects will compete, and an eqbm AH will be formed somewhere inbetween the two. The exact eqbm point will shift according to daily weather, seasons, etc.


If there is ventilation or air ingress at a different AH than what was already there, then that will also shift the eqbm, might make it so dry that condensation stops. Ventilation could come from the house (warm, high AH) or from the outside air (usually lower AH).

Theoretical physics sometimes doesnt seem to apply in the real world!

Sorry, got a rush of enthusiasm there! Feeling a bit dizzy now.
I know a little physics but not much bout buildings, thats why I'm here...
all good sport, Will

I think I will have that little sleep, come back refreshed. I see your point about vol being elastic in this case, therefore total gm of water vapour present - but please shortcut me - does this make a great difference? Does does it change

Posted By: fostertomif density goes down as temp goes up, then RH will go down considerably - something like a square function?

That sounds great because it means that in my temp-stratified void, colder bottom may have high RH, where it doesn't matter. But whatever the bottom RH is, the warmer top, where the timber is, will have lower RH than that.