Posted By: goodevansI'm going to stick my neck out here...
A house with 70% humidity at 20 deg C inside when it is 0 deg C outside has ventilation problems - 12g/m3 inside, 4.8g/m3 outside - A continuous ventilation system in all wet rooms would be able to keep the average down to a difference of 2g/m3 (a MVHR does just as well but without the heat loss) - Even trickle vents do a lot better than this scenario (but I accept that occupants can and will do strange things to houses).

A vapour barrier will help the calcs because of the insanely high Sd values on vapour barriers - only valid if there is absolutely no holes - but as most houses have a blower test above zero m3/m2Env at 50Pa you can assume the calcs using those Sd values would not be valid.

Even taking into account the massive internal humidity in the scenario above - if you slice the inner layer of eps again you will find the dew point inside the eps not on the inside surface. In reality there will be so little moisture (because of the high Sd value of the EPS) that there will be no issue - and it gets even more complicated because as the vapour condenses out heat is released into the eps slowing down the condensing action. And once the vapour has turned to liquid it will move through the eps in a different way due to capillary action.

But I could be wrong - is there a rash of EWI problems out there with interstitial condensation that can not be attributed to poor ventilation or from rain ingress reservoirs. (If so I'd like to see some eps sectioned and see the a dew point half way through i.e. one side dry the other side saturated).

Most CRA's I've seen for externally insulated walls show the Dew Point in about 10mm from the external surface (as tony said). I've done loads of External Insulation so I've pushed the probes of the humidity monitor I use for checking the moisture content of wood in through the plaster, I've never found moisture there. This shows that we should worry about airtightness but this water vapour diffusion issue is a bit of a red herring. It receives way too much air play and there has never been a proven case where WVD caused structural decay in a building!

Posted By: WillInAberdeennot a problem if the outside of the wall is more vapour open than the inside, as moisture will permeate out faster than it can build up

Isn't that an old theory, popularised in the now-abandoned 3:1 (or 5:1) rule of thumb (which reading between the lines caused catastrophically accumulating interstitial condensation, law suit and ruin of an up-and-coming eco-Architect pracice in the groundbreaking near-PH primary school at Dartington Devon, now demolished and under conventional rebuild)?

When condensation occurs at the dewpoint within the sandwich, how can that 'permeate outward' when all laYERS OUTBOARD OF THAT ARE EVEN MORE SUSCEPTIBLE TO CONDENSATION, REGARDLESS OF (oops sorry) their relative vapour permeability?

That gradient notion is a complete red herring AFAICS, as I discovered when testing the 3:1 theory through many case-variants in WUFI.