ICF versus SIP

[RichS]

So, I was decided upon ICF as my shell. I then thought to use SIP panels for the roof.

 

I sent a request to a few SIP suppliers re. the roof and invariably a couple have tried to get the whole job.

 

So what are the feelings / knowledge of ICF v SIP.

 

Having read some old threads on ebuild it seems that SIP have a considerable problem with cold bridging.

 

Thoughts please

[najem-icf]

SIPs are a great product for their thermal efficiency.  Using them on a roof is perfect.  The issue with SIPs usually is the lack of sound attenuation that ICF can give you.  SIPs is a timber product so will react to our forever changing seasons.  As a result the airtightness blower test will always deteriorate as time goes on.

 

ICF is a monolithic build - the concrete is poured in one go per storey (or higher if you are brave enough or have a competent, experienced contractor) so essentially goes off in one go and is, for all intents and purposes, one piece of concrete per storey and usually with reinforcement in there.

 

Recently, PIR/PUR prices have shot through the roof and this has had a direct impact on the price of SIPs.  The Grenfell disaster is also making people take notice of what they are putting in their walls, many are now insulating with EPS instead of PIR/PUR.

 

My views have changed over the years - from being a complete tree hugger, I now firmly believe that concrete is the answer.  We have been building with it for years, the Romans, Greeks, Egyptians all built with the stuff - and its all still there.

 

Hope this helps  - athough with this post being so old I am curious to know what you decided on.

[Jeremy Harris]

SIPs had a few problems, many of which have been gradually addressed by the manufacturers.  Roof panels invariably needed internal joists, which added thermal bridges.  A lot of the readily available panel sizes were aimed to "just" meet building regs (for the mass market developers) so weren't great for those looking for good thermal performance.  The insulation used, like all foam products, has a low decrement delay value, so doesn't help in terms of building a house with a long internal thermal time constant and, probably the biggest issue, is the detailed design of the wall to floor junction, because both skins of a SIP need support and that means careful design to both avoid a nasty thermal bridge and, more importantly, address the potential for interstitial condensation at the sole plate, driven by external changes in humidity and temperature.

 

There was an interesting exchange on this forum's predecessor, Ebuild, where Kingspan became involved, at first saying their standard detail was fine, then accepting that there was a potential sole plate thermal bridging and interstitial condensation issue, and agreeing that adding a peripheral skirt of external insulation could partially mitigate the issue.  I take it that is the old thread on Ebuild you may have read, @RichS

 

One option we explored in some detail, was to use a 140mm SIPs panel for the structure, then add external, relatively high decrement delay wood fibre insulation externally, to both improve the U value to a reasonable level (our target was less than 0.12 W/m².K for the walls, 0.1 W/m².K for the roof and a decrement delay factor of greater than 6 hours).  A passive concrete slab foundation, with a raised EPS upstand to join the external wood fibre layer mitigated both the wall to floor thermal bridge and eliminated the interstitial condensation risk.  Our problem was that we struggled to get the builder who was offering this system to provide a quote - they were hopeless and repeatedly gave quotes for a spec that was different to that we'd agreed in several meetings with them.

Edited September 14, 2017 by JSHarris

[najem-icf]

I had a very detailed discussion with Kingspan guys some time ago at a show.  I put a challenge to them to air test one of their best builds over a period of five years - if there was a degradation in performance between year 1's test and year 5's test, would they give the home owner their money back?  They categorically said absolutely not - it is an organic product and over time it will open up and adhesive sticking plasters will eventually fall off.  They declined, however I would have loved to have that kind of data just for sake of education and improvement of housing stock.  Innovate UK, a government agency, published a report back in February 2016 that stated there was no relationship between design and deployment - ie whatever is specified and expected in a design is seldom what appears in the finished product.

 

A summary of that report is here:

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/497757/Domestic_Building_Performance_Summary_2016.pdf

Note recommendation on page 3 to "Designers/homeowners/policymakers" that states:

Quote

"Remember that airtightness often deteriorates over time – sometimes by more than a third. So the tested airtightness on day 1 is unlikely to endure for the home’s lifetime."

 

I understand what you were trying to achieve with your detail - what did you end up doing given the contractor could not give you a quote for what you wanted?  If you have achieved the uValues stated - you will be running your home on pocket change!!

[jack]

Yes, those values were achieved, but not with SIPs. @JSHarris was one of the first in the UK to use MBC Timberframe, an Irish company, which a number of us have since used to build our frames. They do a cellulose-filled twin-stud on insulated raft construction, which achieves the U-values above while avoiding thermal bridging.

 

There's a lot (and I mean a LOT!) of detail about Jeremy's build in his blog.

[Jeremy Harris]

I must get around to updating that blog soon!  I've been too tied up fitting out the "perfect workshop", something that I'm sure will be impossible to achieve..............

 

What I can say is that airtightness deterioration with a timber lined construction, be it SIPs or some other form of completely insulation filled design, like our MBC twin stud, is not an issue at all.  For example, have a lot of boat building experience with using the PU adhesive that is used to bond SIPs splines to the panels, and it simply doesn't degrade or debond, even under very harsh conditions.  Somewhere I have some test samples of different types of wood bonded with the stuff that have been outside for years, and I'm sure one of them is a bit of OSB3.  Last time I checked none had debonded at all, even when left wet, all that happens is that excess adhesive exposed to sunlight starts to degrade as the UV breaks down the molecular bond in the surface layer.  Even then the bond strength doesn't seem affected.

 

Frankly, airtightness is a red herring in this context, anyway, as what is far more critical is vapour permeability.  The inner surface of any form of well-insulated construction needs to have a low vapour permeability, and there should be a gradation in vapour permeability, such that it is greater on the outer layers of the construction than it is in the inside layer.  OSB is faily vapour permeable, so a SIPs build with OSB panel skins (which is the standard material used) will have a vapour permeable inner skin.  This means there needs to be a vapour control layer internally, to prevent the migration of water vapour from the high concentration inside the warm house to the low concentration outside on a very cold day.  Failure to prevent this vapour flow may lead to interstitial condensation, where the water vapour concentration reaches a temperature point inside the structure where it can condense to liquid water.  Once condensed, it needs a lot of thermal energy to cause the water to phase change back to vapour so that it can migrate out, and in an insulated structure it may well be hard to get enough thermal energy into the structure to allow this to happen.  Essentially this is what caught out a well-know major house builder in the 1970's, when they introduced timber frame construction in England, without fully understanding the interstitial condensation risk.  They ended up with frames that were rotting out at the base within ten years or so, IIRC.

 

So, in essence, it doesn't really matter if the SIPs panels move a bit, as any form of construction will with changes in temperature, as the VCL should have either enough elasticity, or be subject to such small temperature variations, that airtightness will be maintained, anyway.

 

As an aside, a friend built a block and brick house a few years ago, around the time that permeability testing was first introduced.  When I went to look at his build he'd failed two air tests, and they were struggling to find where the leakage was.  It turned out that 90% of it was coming through the concrete block walls - they were surprisingly porous.  The fix was easy, just parge coat the inside of all the external walls.  Once this was done the house passed the air test easily.

Edited September 15, 2017 by JSHarris

[najem-icf]

That blog is awesome! Wish more people did that...

 

Are you doing a regular blower test @JSHarris? I would be curious to see the results - it would be a good advert for MBC if they have solved the issues most stick builds face.

 

[Jeremy Harris]

47 minutes ago, najem-icf said:

That blog is awesome! Wish more people did that...

 

Are you doing a regular blower test @JSHarris? I would be curious to see the results - it would be a good advert for MBC if they have solved the issues most stick builds face.

 

 

I've done one additional test since the original, and the permeability had reduced a by about a third from the original test figure, but that improvement was largely down to me adjusting the French windows and squirting some thick silicone into the centre of the lock barrels, as all three were leaking air through the keyhole like mad on the negative pressure test.  MBC guarantee that every build will meet or exceed 0.6 ACH, and there's a hefty final payment (20% in our case) that isn't paid until the air test result comes in under this figure.  AFAIK, every build passes first time, as they've got an inherently good build system, in terms of airtightness, high decrement delay insulation and absence of thermal bridging.  The combination of their passive slab and modified Larsen truss "twin wall" structure pretty much guarantees no thermal bridging, as long as there are no unusual design features that get in the way of this.

 

The airtightness won't degrade, I'm sure, as 99% of the fabric airtightness comes from the thick layer of highly compressed cellulose insulation, blown in under pressure, which pretty much guarantees the long term airtightness of the frame, as well as having a pretty good decrement delay factor.  The "airtightness" tape used to seal the joins in the internal vapour control skin isn't there to enhance the airtightness, it's just there to ensure that the internal vapour permeability is very much lower than that on the OSB outer skin (the inner skin is a vapour impermeable board).  The only significant cause for airtightness degradation is likely to be from the doors and windows, where some adjustments are bound to be required through life to endure the seals remain tight after a period of use.