Spray foam on cold bridge steels

[NCXo82ike]

We have most of the steelwork up in our wraparound extension and loft conversion. 1905 mid-terrace, solid wall construction. We have picture frame steels sucking heat into the ground and steels bearing on to (and/or supporting) cold brick walls. I'm looking to minimise the impact of cold bridging.

 

The wraparound is brick and block with mineral wool cavity, pitched vented roof. The loft conversion is timber, vented cold roof (unfortunately necessary).

 

Priority 1 is to avoid condensation, priory 2 to reduce heat loss. No cold steel near bathrooms, but plenty near the kitchen in the wraparound.

 

The builder suggested PIR or mineral wool around the beams, but I worry if we insulate without a vapour barrier against the steels it will worsen interstitial condensation? I wonder if spray foam is a sensible way to do this, cutting the cold bridge with a continuous vapour barrier I can tape against other membranes. 

 

I've sent the horror stories but get the impression the issue is where critical ventilation e.g. cold roof is compromised by the spray foam.

 

Other relevant measures are:

-internally insulating solid walls (planning on vapour-open with thermactive plus wood fibre).

-windows uprated

-We'll be getting as airtight as reasonably possible

-MVHR system fitted which I will control based on internal humidity

 

Grateful for any thoughts- couldn't see similar in the forum surprisingly

Edited January 22 by NCXo82ike

[JohnMo]

Steel, connected to ground will be a cold surface. Wrap in PIR and tape the joints, this gives vapour barrier. 25mm is normally enough but the more the merrier. Or insulation and wrap and tape a vapour barrier. Condensation will only form from constant contact with air

[NCXo82ike]

27 minutes ago, JohnMo said:

Steel, connected to ground will be a cold surface. Wrap in PIR and tape the joints, this gives vapour barrier. 25mm is normally enough but the more the merrier. Or insulation and wrap and tape a vapour barrier. Condensation will only form from constant contact with air

Thanks John, on that basis my question is whether there's a reason not to use spray foam since there's a lot of complex junctions, joists etc

[JohnMo]

I would do all to can in PIR and where needed apply spray foam to fill gaps.

[Super_Paulie]

I have a goalpost type setup with the post flush to the block. I attached a sheet of aerogel to that and foamed the edges before plasterboarding, attached with foam rather than traditional adhesive. The upright leads to a comically oversized beam that has been left exposed over 8m. 

 

I can tell you, the beam is cool but never cold and it seemingly has very little effect on the rooms temperature which I have at 23° daytime and 21 evening.

 

Aerogel might seem expensive but could be an option?

 

Edited January 22 by Super_Paulie

[Nickfromwales]

@NCXo82ike

 

Have you not thermally decoupled the steel from the concrete?

 

I use different products to achieve this, with Bosig Phonotherm on the most recent clients project.

 

If the steels are already in, then defo look at a spray foam, but box the areas you wish to insulate, using release agent, and fill the voids. Then remove the boxing in to make way for bonding on Marmox boards to then accept plaster skim coat + paint. 

[NCXo82ike]

On 25/01/2026 at 23:01, Nickfromwales said:

@NCXo82ike

 

Have you not thermally decoupled the steel from the concrete?

 

I use different products to achieve this, with Bosig Phonotherm on the most recent clients project.

 

If the steels are already in, then defo look at a spray foam, but box the areas you wish to insulate, using release agent, and fill the voids. Then remove the boxing in to make way for bonding on Marmox boards to then accept plaster skim coat + paint. 

No decoupling, a lesson for next time around. Not something discussed by the architect or engineer, and by the time I understood it we were too far along. Thanks for the recommendation- boxing in and spray foam sounds great.

 

 

[NCXo82ike]

Some joists come off the steels, so these will have cool ends- the spray foam will hopefully stop vapour getting to condensation spots.

 

We have the same situation with old joists in the solid walls. I've seen some recommendations to avoid internal insulation at the joist pockets to keep these warm, and others to insulate and tape around the joists for a continuous VCL.

 

 

[Nickfromwales]

9 minutes ago, NCXo82ike said:

Some joists come off the steels, so these will have cool ends- the spray foam will hopefully stop vapour getting to condensation spots.

 

We have the same situation with old joists in the solid walls. I've seen some recommendations to avoid internal insulation at the joist pockets to keep these warm, and others to insulate and tape around the joists for a continuous VCL.

 

 

For floors at ground floor level, or for first floor joists? 

[NCXo82ike]

14 minutes ago, Nickfromwales said:

For floors at ground floor level, or for first floor joists? 

First floor. We've got a new insulated slab over the whole ground floor

[Nickfromwales]

49 minutes ago, NCXo82ike said:

No decoupling, a lesson for next time around. Not something discussed by the architect or engineer, and by the time I understood it we were too far along. Thanks for the recommendation- boxing in and spray foam sounds great.

 

 

So many architects haven’t got a scooby-doo about these things. 
 

Even the SE’s tame associated, clueless. Constantly sense checking with Obi-Wan back at HQ, and just no confidence in anything he was looking at.

 

Client rang me on speakerphone whilst the chap was on site, so I got to speak to him about the steelwork I spec’d and Obi-Wan had agreed over the phone; we made some slight changes, but beefed up certain elements to (over) compensate, and he was head-melted. 
 

After a lot of gibberish and flannel I just got a bit annoyed and said to go back to the office and just pass the bloody thing ffs. Myself 30+ years on the tools, and the steel fabricator & welder 35+ years, so I just said we’ve done more than a few of these over the years…..!

 

Obi-Wan saw the adaptations, and signed it off without further comment.

 

We put Bosig Phonotherm under the feet of the steels, where they sat on padstones, very nice product to work with and cuts / drills like a very dense particle board. Router / planer zoomed through it where I made profiles and recesses in it for the Norrsken doors and sliders etc.

 

Blasted it with liquid DPC all around before then screwing it all into place and then blasting it with more.

 

Available in sensible sizes too, unlike Compacfoam. 

[Nickfromwales]

4 minutes ago, NCXo82ike said:

First floor. We've got a new insulated slab over the whole ground floor

In that case, up at that height, you can happily consider the steel and wood to be equalised for temp etc, imho, as they’re well with the heated and airtight envelope.

[NCXo82ike]

6 minutes ago, Nickfromwales said:

So many architects haven’t got a scooby-doo about these things. 
 

Even the SE’s tame associated, clueless. Constantly sense checking with Obi-Wan back at HQ, and just no confidence in anything he was looking at.

 

Client rang me on speakerphone whilst the chap was on site, so I got to speak to him about the steelwork I spec’d and Obi-Wan had agreed over the phone; we made some slight changes, but beefed up certain elements to (over) compensate, and he was head-melted. 
 

After a lot of gibberish and flannel I just got a bit annoyed and said to go back to the office and just pass the bloody thing ffs. Myself 30+ years on the tools, and the steel fabricator & welder 35+ years, so I just said we’ve done more than a few of these over the years…..!

 

Obi-Wan saw the adaptations, and signed it off without further comment.

 

We put Bosig Phonotherm under the feet of the steels, where they sat on padstones, very nice product to work with and cuts / drills like a very dense particle board. Router / planer zoomed through it where I made profiles and recesses in it for the Norrsken doors and sliders etc.

 

Blasted it with liquid DPC all around before then screwing it all into place and then blasting it with more.

 

Available in sensible sizes too, unlike Compacfoam. 

It's a shame this level of detailed knowledge isn't common

[NCXo82ike]

5 minutes ago, Nickfromwales said:

In that case, up at that height, you can happily consider the steel and wood to be equalised for temp etc, imho, as they’re well with the heated and airtight envelope.

Thank Nick, does that mean we shouldn't insulate the beams then? They have external solid walls bearing on them so I (as a total amateur) consider them a cold bridge/condensation risk?

 

Perhaps all academic given that we're getting crittall-style doors as a massive thermal bridge 

[Nickfromwales]

Just now, NCXo82ike said:

Thank Nick, does that mean we shouldn't insulate the beams then? They have external solid walls bearing on them so I (as a total amateur) consider them a cold bridge/condensation risk?

 

Perhaps all academic given that we're getting crittall-style doors as a massive thermal bridge 

To be accurately providing information, we’d really need to see detailed section drawings (or a really shit sketch on the back of an envelope) to say for sure.

 

Whats the cavity insulation like?