Air tightness should be very good so long as it is joined to the next materials windows, roof etc

Insulation, the thicker the better, aim at a U value of 0.1

Do you mean rammed earth, or cob (different things)?

It's definitely not cob - looks like raw as-dug earth, complete with stones, image below. The timber seems to be the fixing for a door frame (out of shot). If anything was added for stability I guess it would have been lime, but probably nothing. I'll check out the Grand Designs programme in case it provides any ideas, if it's online.

I did some quick research last night and found that:

- the walls are traditionally built off a stone dpc/footing to stop damage from rising damp; I can't see that, but it may well be there.

- as suspected, external insulation is bad idea, as it leads to a build up of moisture in the wall, leading to collapse.

- externally, only lime renders, direct to the wall are advised - or leaving the wall on it's natural state.

- seems that mineral fibre internal insulation is not advised, as it can become saturated from moisture from the wall, in addition to the risk of moisture condensing in it if the house vapour barrier isn't fully effective. Seen a suggestion that hemp + lime may be suitable, as hemp has similar hygrothermal properties to the wall, but haven't had chance to look into this further, yet.

- DPCs + concrete floor slabs are bad, as the moisture that builds up beneath leads to an increase of moisture at the edges where it's likely to migrate into the walls. (Though maybe that could - perhaps - be avoided by ensuring that the top of the slab is below the top of the stone DPC, if it exists). Though I've found a suggestion to use a mix of lime mixed with expanded clay beads, laid over a few hundred mm of expanded glass beads for insulation, and a suggestion that a thick slab (150 to 200mm) + teracotta tile finish is compatible with UFCH.

Clearly a lot more work to do on this, including on how to get it though building control or even find thermal data, so further ideas are still appreciated.

I don't see why

Posted By: Mike1external insulation is bad idea, as it leads to a build up of moisture in the wall

Assuming the EWI is breatheable and so is the interior face, so the cob can re-dry both inward and outward, why should protecting the cob from weather, and from interstitial insulation in its outer layers, lead to moisture buildup? Quite the reverse.

Posted By: Mike1mineral fibre internal insulation is not advised, as it can become saturated from moisture from the wall, in addition to the risk of moisture condensing in it if the house vapour barrier isn't fully effective

Internal insulation makes the whole of the cob, from inside to out, colder than it would have been, while subject to vapour load generated from use as a house, hence 'un-natural' interstitial condensation within its coldness. Then attempt at an internal vapour barrier (whether effective or not) prevents internal insulation from re-drying inward.

If insulation has to be internal, make it relatively weak (say 3" max) so the cob stays a bit warm. And on no account seal it against inward re-drying with an internal vapour barrier .

Ah - so do you think your walls have been 'blocklaid' i.e. blocks cast before placing/mortaring? That could still count as cob - is a standard method of cob repair (using concrete blocks to repair cob is disastrous).

Posted By: fostertomAssuming the EWI is breatheable and so is the interior face, so the cob can re-dry both inward and outward, why should protecting the cob from weather, and from interstitial insulation in its outer layers, lead to moisture buildup? Quite the reverse.

I'll need too look back at my notes to see their reasoning, but a major concern seemed to be ensuring the evaporation of moisture rising from the footings. And apparently 20% moisture content can be enough to cause collapse, so want to get it right (or look for another project).

Internal insulation makes the whole of the cob, from inside to out, colder than it would have been, while subject to vapour load generated from use as a house, hence 'un-natural' interstitial condensation within its coldness. Then attempt at an internal vapour barrier (whether effective or not) prevents internal insulation from re-drying inward.

Agree with the theory.

If insulation has to be internal, make it relatively weak (say 3" max) so the cob stays a bit warm. And on no account seal it against inward re-drying with an internal vapour barrier .

Or would thinner insulation just place the dew point a bit deeper in the wall (which must be about 600mm thick)? Is it better to keep the dew point well within the insulation? Or due to the materials, does moisture migtation make the dew point of little practical help anyway? Agree with avoiding a vapour barrier though, and needs.to be a breathable paint too.

Posted By: djhI think if there are blocks of earth then it's most likely cob as that's a traditional method as Tom says. Rammed earth is just loose earth laid in layers inside shuttering and 'rammed', sometimes with the addition of cement or lime depending on what the local soil is like. Straw in cob adds tensile strength, much like hair or other fibres in lime. Cow dung works in a similar way, with the fibres and proteins in it adding tensile strength and the dung also 'glues' the material together.

Not blocks in this case, and no sign of straw or other inclusions that I'd expect to see with cob.

I agree with Tom as well about insulation. Where did you get the information about external insulation being bad news?

As above, will need to look back at my notes - but it's possible that was not taking account of using natural fibre (eg hemp) external insulation, or problems in using such fibers externally (not sure if they may cause adhesion problems for external lime render, for example, unless the render is on wood or metal laths, though under timber cladding sounds better - on the basis of no research).

BTW, living in a cob farmhouse has convinced me that cob has effectively insulative properties that go beyond Uvalue.

Cob samples, including all voids etc can be tested, or site measured, as historic buildings organisations have done for many kinds of traditional walls, coming up with much better Uvalues than post-war default assumptions.

But I'm suggesting better effective performance than even that. Some complex of moisture transactions and diurnal heat flows. Not to mention the all-pervading hygroscopic buffering, which makes shower room extract almost unnecessary. That must have something to do with it.

An old ambiguity that used to plague the " outboard vapour resist*** must be one third (some said one fifth) of inboard resist*** " rule of thumb - now thankfully discredited and abandoned. Was it meant to be resistance, or resistivity? - no-one could ever say, because no-one knew who had invented the rule.

That first item of French advice to me smells of just such falacy and misunderstood rule-of-thumb. Vapour transport is not analogous to electical resistance etc. For my money, it should refer to resistivity i.e. thickness doesn't come into it. It's the vapour gradient that shouldn't steepen outward.

Or, why shouldn't it? The above one-third/one-fifth rule is ineffective, no substitute for WUFI modelling, or at least WUFI-based experience. Materials don't need to be super-permeable - just a quantum-leap more permeable than a vapour barrier.

Anyway, why should a cob wall be sopping wet? It should have good 'hat and boots' i.e, well sheltered from rain on top and built on a 500mm high rubble stone base wall, which acts as capillary break against any poss of rising damp. If french pise is customarily built straight off the ground, then they have a serious problem anyway.

Posted By: djhIn what way isn't it?

For a start, 'thickness' rarely comes into everyday electricity - an electrical resistance is an encapsulated object, thickness irrelevant, resisting electrons (not to mention impedance and capacitance), whereas in vapour transport thickness is all-important and involves moving massive objects around (if you can call a vapour cloud massive). Also electrons don't suddenly condense into a puddle and drop out of apparent existence. What's the electrical analogy of relative humidity, pressure/density, capillarity?

Posted By: fostertomThe above one-third/one-fifth rule is ineffective, no substitute for WUFI modelling, or at least WUFI-based experience. Materials don't need to be super-permeable - just a quantum-leap more permeable than a vapour barrier.

The same paper I referred to mentions that their research has found that models such as WUFI don't accurately model the performance of rammed earth walls, for example underestimating temperature rise (by 2°C) and time lag (by 4 hours) in a 30cm thick wall. One of the researchers has published a more accurate mathematical model in a conference paper at https://www.researchgate.net/publication/264405140_Etude_de_la_pertinence_des_hypotheses_dans_la_modelisation_hygrothermique_du_pise - though too complex for me to start getting into that. BTW, there's a photo of a house that collapsed from water infiltration on page 11 of some associated slides (at http://ibpsa.fr/index.php?option=com_jdownloads&Itemid=53&view=finish&cid=290&catid=63 ).

Anyway, why should a cob wall be sopping wet? It should have good 'hat and boots' i.e, well sheltered from rain on top and built on a 500mm high rubble stone base wall, which acts as capillary break against any poss of rising damp. If french pise is customarily built straight off the ground, then they have a serious problem anyway.

It seems they are built off a stone base in France too, but with lime mortar that's not going to be 100% water tight - more a throttle than a break - and as djh mentions, the exposure of the wall would of course be a factor too.

Posted By: fostertomAnyway, why should a cob wall be sopping wet?

Looks like another reason is flooding; just come across this photo and report (in French) of flooding of the Saône in 1856 which destroyed dozens of rammed-earth homes (while leaving those built from stone standing) and killed 18: https://www.leprogres.fr/lyon/2016/12/11/le-1er-juin-1856-lyon-victime-du-rhone-en-crue-offre-un-paysage-de-desolation

As you were - a thought. Of course unfired clay is not as other materials - highly hygroscopic i.e stores liquid water not in capillaries between particles, but in the particles themselves, causing them to swell.

So tho vapour behaviour should be standard, liquid behaviour, whether soaked up or by condensation, may be significantly different in absorbtion and/or re-drying, in quality and/or quantity.

Doubt those French recommendations would therefore be right, tho.