Thermal Bridge Problem – Internal 9" Brick Wall joining External 9" Brick Wall

[Anonymous]

Can anyone say whether I'm doing the right thing?

I'm converting an agricultural building which has 9″ solid brick walls. I'm building a new dense concrete block/earth brick internal wall to create a 250mm cavity which will be fully filled with mineral wool cavity batts. As some existing internal structural walls are joined to the external walls I'm planning on continuing with the cavity wall an extra 600mm along the internal wall to help reduce heat loss through this connection.

Is this acceptable? I know it's not ideal to have the internal walls directly connected to the outside walls but I have no choice as they are structural. Before anyone suggests it, insulating the outside of the building is not an option.

[Mark Siddall]

Hi Paul,
I'm sorry to say that whilst this solution will certainly improve the surface temperature of the wall (from 13C to ~18C at -5C ext), and thus help to prevent condensation and mould growth at sub zero temps, it will not address really thermal bridging in a meaningful way compared to the U-value as you still have masonry penetrating the thermal insulation it's just further from the external wall. Whilst you may reduce the thermal bridge by 50%

Assuming a base U-value of ~0.17 W/m2K

* No flanking insulation psi-value ~ +0.145 W/mK (locally this is the equivalent to a 185% degredation of U-value)
* 600mm flanking insulation psi-value ~ + 0.054 W/mK (locally this is the equivalent to a 132% degredation of U-value)
* 1000mm flanking insulation psi-value ~0.045 W/mK (locally this is the equivalent to a 126% degredation of U-value)

Mark

[Nick Grant]

Nice Therm work Mark but I'm not sure how usefull % degredation of U value is as a concept as U values are area based and psi are linear. Paul needs to work out the heat loss through the walls and add the length x psi heat loss.

This is easy in PHPP but can do long hand for SAP I assume.

Nick

[Mark Siddall]

Nick,
The %ages are just for some loosely informative appreciation of the degredation. The %age degredation concept would be best applied as as Y-value across whole building evelope and this would be affected by the details selected.

Mark

[Nick Grant]

Agree useful if applied accross whole building.

[Anonymous]

Is there a way of removing this thermal bridge? The building (single storey) is a single room deep and the internal wall helps to hold the front and rear walls.

[Nick Grant]

Paul

First work out what the heat loss is.

Do you have timber going through the internal wall to outside wall? Would be more of a concern for me because of damp/rot issue. Wall is just an energy loss and condensation risk.

Nick

[Anonymous]

Heat loss calculations is something I haven't ventured into so far. I'm working on the principle of insulating as best as I can. Because it is a conversion of an agriculural building the building inspector said he was happy to have a SAP calculation once the building was complete due to the likelihood of materials/design changing as the build progressed.

I've designed the new internal walls so they finish at sole plate level. The cavity will be closed at this level (suggestions as to the best way to close it would be appreciated). Sub-rafters will be used below the purlins to meet the top of the new internal wall. Once boarded over the roof void (rafters + depth of purlins + sub-rafters) will be filled with cellulose fibre. I'm using the cellulose because of its hygroscopic property so that it will protect (hopefully) the timbers. A vapour barrier will be used on the warm side of the roof and plastered into the walls to create the airtight barrier.

The only timber that I can see that could be a potential problem are the trusses (when I say 'only' I mean that in the context that there should not be any other timber penetrating through the internal/external wall rather than the trusses are unimportant!). The trusses sit 150mm lower than the sole plate so would be bridging the cavity wall. Any suggestions as to what I could do about ensuring damp/rot is not an issue?

[David Olivier]

Sounds to me as if the new structure could be designed to hold up the roof. The 225 mm brick skin would revert to a (thick) ornamental outer leaf and the bricks of the internal wall would be recycled into new uses. This also frees up internal floorspace.

Does that work at all?

Unless you have a continuous air barrier, though, the extra air leakage usually dwarfs the thermal bridge losses. From the description of the trusses inside the purlins, it appears you don't yet have the prospect of a sealed air barrier, since to seal a membrane around structural members which have to pass through it is virtually impossible (ref.: Taby Houses, Sweden, 1977). This should be given attention even before the cross wall.

David.

[Anonymous]

Not sure at the moment about making the new internal walls structural to hold roof. I'll think about it to see what the implications would be.

Has anyone got a solution to the trusses penetrating through the air barrier? Someone must have come up against this problem before.

[Mark Siddall]

Paul,
I think that David may be being a little harsh (if the Taby houses are those discussed in Carlsson.) Though the details around such penetration were found to pose some *very* major challenges an airtightness of between 0.7 to 1.4 ahc/hr at 50pa was achieved at the these houses.

Here they used over lapping 0.2mm air barrier memranes with taped joints. The membrane was then taped to the structure (in the case of refurb, rather than using tapes, it may be better to use a flexible sealant and a timber batten which would be used to create a continuous pressure plate and ensure that the tape does not delaminate from the old timber). To allow access and site checks around the penetrations a 200mm wide zone was left clear of the barrier until the pressure tests had been completed. The gap was then closed with boarding.

Hope this helps.

Mark

[Anonymous]

Thanks Mark. That's what I was anticipating having to do. I shall be doing the air barrier membrane myself to ensure it's as airtight as possible.

David, I've thought about making the new internal walls structural and decided it's going to get too complicated. This is because the new internal wall is not continuous around the entire building as some small extensions are going to be built which will mean some of the existing external walls will become internal. I assume you would cut the ends off the trusses if the new internal wall became structurally supporting so that the timbers didn't span the cavity. Looking at where I'd have to cut the ends would seriously weaken the trusses.

[Tom Foster]

Has anyone got a solution to the trusses penetrating through the air barrier?

Is the problem that the top chord of the trusses interrupts the air barrier? The purlins don't do the same because the air barrier runs below them? Presumably the purlins bear on top of the trusses? So the question is how to get the air barrier continuous over the top chord of the trusses? That being not impossible except where the purlins bear on the trusses?

Lay a broad strip of polythene lengthwise over the top of the truss top chord, flaps hanging well down both its side faces. At perimeter of ceiling panel/areas, turn the ceiling air barrier also down to form a flap against the truss side faces. Mechanically clamp the flaps together by screwing a pressure batten to the truss sides. Job done, except the purlins get in the way of the polythene strip. So when installing the strip, laboriously cut the nails holding the purlins to the top of the truss top boom. Jack up the purlin a bit, slip the polythene in, re-nail the purlins through the polythene, with some mastic in the sandwich to seal the nail holes if not a tight clamping joint.

If the purlins are instead tenoned into the sides of the truss, I dunno! Drylining-type light metal angles can be tinsnipped and well screwed to form effective continuous clamping around difficult shapes – as long as you can get at it all round.

Don't even think of resorting to sticky tape of any kind – continuous mechanical clamping only, with mastic between if any doubt about clamping surfaces being tight-fitting. In due course easy/versatile/foolproof products will emerge to make this stuff easy, as happened with through-ventilation, after several years of struggling with battens and flymesh.

[Author]

Posts