If the construction is as shown then the insulation on the solid wall will help as far as it goes. That is you will get benefit up to the end of the insulation after which you have the u value of the solid wall.

I don't think that putting insulation in the corner as shown will have any greater benefit than putting it anywhere else on the solid wall because the cold bridge is the whole of the solid wall and it is no different in the corner.

Just IMO

Sorry... my diagram wasn't entirely clear.

On the other side of the solid wall from the insulated portion is inside space, not the outside. The cold bridge that needs to be dealt with is indicated by the green arrow:

Posted By: gyrogear

Posted By: Peter_in_Hungarybecause the cold bridge is the whole of the solid wall and it is no different in the corner.

Not to contradict PiH, but French eco-renovation recommendation I have seen on MOOC was for 60 cms on the crosswall as a rule of thumb.

The reasoning being that part-insulating the crosswall lengthens the path that heat has to travel to reach cold.

This appears to be confirmed by this other source I have just dug up...

http://www.ajena.org/renovact/media/technique_ponts-thermiques.pdf

At top of page five, it says, "use any technically simple and economically accessible solution: unless treating the entire thermal bridge, at least 60 centimeters of the crosswall should be insulated"

Same reasoning applies beneath a floor slab.

FWIW,

gg

Thanks, that's useful.

I think PiH's answer was based on a misunderstanding of my original diagram, which wasn't quite clear.

+1

Posted By: SigaldryThe other side of that wall past the new wall will be cold (with a risk of surface condensation and mould growth) whether you insulate as shown or not.

This is true, however it doesn't really change the situation compared to how it is now. The room on that side will have to deal with this risk in the same way that it does at the moment - heating, and adequate ventilation relative to the normal level of humidity to be expected in that room.

The new room, that is the one enclosed by the (insulated) extension walls, in this case will be subject to a high level of humidity and therefore I want to be careful about cold spots near that wall junction.

Of course the ideal situation would be to retrofit the whole house with proper insulation, however that is often well outside the scope of extension projects so you have to just concentrate on making the extended part as well insulated as reasonably possible.

Tapering seems to make sense - but in fact effectively shortens the width of the return-wall board by half.

It all depends on forcing a heat-flow path length along the masonry, long enough that the resistance of that path is similar to the much smaller 'length' (i.e. thickness) of the IWI. That's done by the return-wall board almost completely cutting off any more direct shortcut by the heat flow.

Once you've decided what the length of the heat flow path should be, why not enforce it for that length, instead of progressively weakening that enforcement?

It's false thinking, to say that towards the inboard end of the return-wall board, the temp differential between wall and room has become so small it's hardly worth cutting off.

No - the heat-flow path length along the masonry is a straight temp gradient, linear from max to zero. Tapering the return-wall board turns the gradient into a curve which can only be shorter than the straight one.

Maximise the length of the temp gradient either by maintaining the return-wall board's thickness,
or if you want to taper it for cosmetic reasons (valid), by making it extend much further back into the room.

Posted By: fostertomTapering seems to make sense - but in fact effectively shortens the width of the return-wall board by half.

It all depends on forcing a heat-flow path length along the masonry, long enough that the resistance of that path is similar to the much smaller 'length' (i.e. thickness) of the IWI. That's done by the return-wall board almost completely cutting off any more direct shortcut by the heat flow.

Once you've decided what the length of the heat flow path should be, why not enforce it for that length, instead of progressively weakening that enforcement?

It's false thinking, to say that towards the inboard end of the return-wall board, the temp differential between wall and room has become so small it's hardly worth cutting off.

No - the heat-flow path length along the masonry is a straight temp gradient, linear from max to zero. Tapering the return-wall board turns the gradient into a curve which can only be shorter than the straight one.

Maximise the length of the temp gradient either by maintaining the return-wall board's thickness,
or if you want to taper it for cosmetic reasons (valid), by making it extend much further back into the room.

My logic would be (with made-up numbers)

The new insulated wall is 300mm thick and has a U-value of *something*.

Say 900mm of masonry would have the same U-value, then insulating (perfectly) 600mm back along the return-wall means that the temperature of the wall where the insulation stops is going to be similar to the temperature of the inside of the new-build wall.

Where the return-wall insulation meets the newbuild wall, the heat path is through 300mm of masonry then X thickness of return-wall insulation. Say I decide that 150mm thick insulation here gives a heat path through 300mm masonry then 150mm insulation and this achieves something similar to the U-value of the newbuild wall.

Then if I taper the insulation from 150mm at that point, down to 0 at the 600mm along-the-wall point, then doesn't that broadly make sense? At the half-way point, the heat path will be through 600mm of masonry then 75mm of insulation. It seems plausible that this will give an effective U-value something similar-ish to 300mm masonry then 150mm insulation.

You can say why not just run the 150mm insulation the full 600mm length and where it makes no odds then you might as well. But where space is tight, you might not want to lose that 150mm except where you really need to.

If space is tight, you could run 75mm for the whole length instead. That would lose you the same amount of space as the tapered option. But surely comparing those two options, the tapered one is preferable?

WilllnAberdeen - I am interested when you say….'If it's a rubble filled stone wall, outdoors air will percolate through the gaps in the core of the wall, this is necessary to keep it dry'

Are the gaps in the rubble filled stone wall necessary? I presumed that you can full fill a stone wall without detriment to the walls performance ~(maybe better for some things?). Am i wrong?

Posted By: lineweightWhen space is very tight, the extra 100 or so mm that can be the difference between insulated plasterboard, and, say, plaster direct onto the brick (or even the brick left exposed) can be significant.

But surely the space we're talking about here is part of the new extended space? In which case the thing to do is design the extension with enough space for the insulation?

<blockquote><cite>Posted By: WillInAberdeen</cite>
Also sorry if this is a sidetrack of interesting thread, lineweight's bricks are themselves porous so they work in a different but related way. There are previous threads, search for 'rubble'?</blockquote>

Indeed I have come up against similar concerns with brickwork... anyone interested can read this thread:

http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=14973&page=1

Posted By: WillInAberdeenThe existing wall might in any case need enough room to be insulated at floor and ceiling level to control thermal bypass upwards/downwards?

It will, yes. In this case it's a two storey extension which means that this applies at ceiling level in the upper floor and at floor level in the lower floor.