A few questions about vapour permeability etc

[Crofter]

Early days yet but I'm starting to turn my attention to future projects (yes, I know).

I'm sketching out some ideas for building something small and cost-effective. This would be for construction under the 'portable building'; rule and thus building regs exempt. Critical factor is cost effectiveness. Application is for glamping accommodation, studio, garden room, etc.

I'm playing around with performance and cost of different materials at the moment, and I'm a big fan of corrugated steel as a cladding material (cheap, easy, fast). I also really like the idea of doing away with the whole VCL/void/battens/PB thing and simply using a ply or OSB interior sheathing layer, sealed with appropriate paint for vapour resistance, and then with all wiring etc surface mounted.

Now with no VCL I think there is a risk of vapour in the structure, even if I do my best to seal the interior. So I'm looking at the vapour permeability. The frame would be most likely 45x95 stick build, fully filled with glasswool. I think I then need some sort of exterior board to hold the wool in, prior to the breather, battens, and steel cladding going on.

I know that I could just use OSB on the outside, but I'm very curious about the woodfibre boards that @ProDave used. A quick glance online shows that the thinner versions of these aren't silly prices. I don't need them for racking strength, as that's what the interior osb layer will do. Question is where is it best to spend the money: beef up the frame to 145mm and just use osb outside, vs keep the 95mm frame and use say 20mm woodfibre boards as EWI. Or, should I use a different material as EWI, e.g. PIR, PUR- my gut feeling is that this isn't vapour permeable enough.

 

Another option would be to use Kingspan's insulated roof sheets as my exterior cladding, but how do you manage vapour permeability with this system?

[A_L]

Corrugated steel O.K., use approx. 50mm battens attached to vertical studs to provide rain-screen type construction with generous gaps top and bottom to provide solar assisted ventilation! The battens would also secure a vapour open and preferably airtight breather membrane which should hold the underlying semi rigid insulation batts in place without further sheathing.

Try to avoid OSB sheathing on the outside,  as some are quite vapour resistant,  this is the riskiest option with regard to interstitial condensation. Cement bonded particle board, MgO board, have low vapour resistances, panelvent dhf if you can get it and even plywood are other possibilities with lower vapour resistances than OSB.

I would only use woodfibre if you think reducing the thermal bridging of the studs is important. Do not use PIR/PUR on the cold side of the insulation as you are right, they have significant vapour resistances

The Kingspan roof sheets in principle have an insulation of sufficient vapour resistance to prevent sufficient water vapour getting to  the cold side, i.e. under the vapour impermeable metal sheet, that condensation does not occur. This not something I am entirely happy about.

[ProDave]

This is certainly doable.

 

Many many older timber framed houses around here have no vapour layer and they don't seem to rot away. The normal was a 4" frame with just plasteboard on the inside. If you were really lucky they sometimes used foil backed plasterboard, but often only in the bathrooms.

 

The designers of my house, the local portable eco building company do not (or to be more accurate did not a few years ago) use a vapour membrane on their own builds. All they did was tape the joints of the OSB inner skin and seal any penetrations.  So what you are talking about seems very possible.  Indeed the vapour analysis they did on my house for the building warrant showed my house the same, with just taped OSB on the inside as the racking layer and air tightness / vapour layer. It was partly the builder who built and erected the frame, and partly this forum that made me decide a separate air tightness membrane on the inside was a good idea.

 

I have 2 layers of 9mm OSB with staggered joints as my racking later. But if you want the OSB to be the finished layer then one sheet of 18mm would probably be better.  I did wire a house where the customer wanted plan OSB as the finish and for that the builders used 18mm OSB sheets, and routed grooves into the edges and joined the joints to adjacent boards with biscuit joiners to ensure a tight straight joint (not sure what happened to the idea of leaving a 3mm expansion gap!!!)

 

For the outside, the wood fibre board is available in anything from 20mm thick upwards. For some strange reason the thinner sheets come as 2400 by 600 sheets somewhat larger than the thicker sheets.  Be aware the thin sheets are not very strong.  I used some on the back wall of the garage where a slight miscalculation meant if we had used 100mm it would have intruded into the reveal of a window on an adjacent wall. Because mine was being rendered, and only supported on studs at 600 centres, I was advised to put an OSB layer on first to give it a bit more strength. But as you are not rendering then it should be okay, just be careful not to knock a hole in it before the cladding goes on.

 

Lastly I can't finish without recommending Knauf Earthwool Frametherm 35 for the frame insulation. It's made for the job. It's somewhat stiffer than most mineral wool types of insulation, so you push it into the frame between the studs, and it just stays there without falling our or slipping down.  I had a test piece inserted between the rafters in my roof/ It was there unsupported for 6 months, it did not move or fall out in that time.  It's also a let less nasty to handle than some other types of mineral wool.  The cheapest place I found to buy it was SIG in Inverness

[MikeSharp01]

56 minutes ago, A_L said:

Do not use PIR/PUR on the cold side of the insulation as you are right, they have significant vapour resistances

Now you have me worried as two of the things yout depreciate are in our design for the house. The wall build up, inside - out, is as follows.

 

Plaster board 12.5mm 

Service void 50mm

Air tight VCL

Frametherm 38 300mm (between I joists)

External OSB sheathing 15 mm

PU insulation 40mm

Outer Tyvek membrane

Counter batten's and batten's 50mm

Marley Thrutone tiles.

 

I think that means I should be concerned about the OSB and the external insulation or does the internal air tight barrier sort this?

[Jeremy Harris]

There is a general rule of thumb that pretty much removes the risk of interstitial condensation, and that is to make sure that there is a gradient in the vapour permeability of the build up of the walls and roof (or ceiling if it's a ventilated cold roof) such that the least vapour permeable layer is on the inside and the most vapour permeable layer is on the outside.

 

Roofing sheets with pre-applied insulation are really for two main applications; well ventilated buildings, often those that are large and where the volume to surface area ratio is high (like warehouses, big DIY stores, etc) or very well sealed structures, where there is no possibility of water vapour getting into the wall/roof/ceiling from either side, or at the edges.

 

Trying to use these materials for a small dwelling isn't a good idea, for all the reasons given.

 

With regard to OSB and vapour permeability, @A_L makes a good point.  The water resistant forms of OSB tend to be less vapour permeable than the non-water resistant types.  OSB2 is fine, as it is fairy vapour permeable in thinner thicknesses (as used on external anti-racking skins), OSB3 a bit less so, as it is a bit less vapour permeable, and some specialist forms of OSB, such as ProPassiv OSB are almost vapour impermeable, and intended for use as a VCL inner skin.

[Crofter]

Thanks all, some good information in there.

So I can discount the Kingspan roof panels- thought as much. Shame because in theory it would do about four layers work in one go.

It may well be an idea to use nothing at all between the frame and the breather. I haven't used the rigid batts myself but if they are stiff enough not to need to be boarded into the frame on both sides, then that would certainly be useful.

I'll have to have a think about how poor I want to allow the thermal performance to get. This application isn't really for permanently inhabited buildings, and low build cost is vital. I totted up my initial buildup idea (18mm osb, 95mm glasswool filled frame, 20mm woodfibre board) and got U=0.3 (not accounting for bridging) which isn't all that shabby really, for what is essentially a shed. If I lose the woodfibre layer it jumps to 0.37 and the bridging gets worse. Up the frame to 145mm and it drops again to 0.25. I really must run these numbers through a calculator which accounts for bridging.

 

Re the inner skin, I'm sure I read somewhere about OSB that was available in a sanded/smooth finish. Can't find it again now. I think it would have to be something like this, or indeed ply, to give an acceptable finish for painting. Rather than using biscuit joints I would probably just use cover strips to hide the nail heads and expansion gap.

 

Scratching my head somewhat for the roof, and specifically how to provide a small slope on what is likely to be a 16x16ft building. For simplicity I think a monopitch works best as it can be done in single lengths of steel sheeting; a shallow duopitch would risk having water blown up under the ridge. The sheet manufacturers don't actually give a minimum pitch for a roof, but I wonder if something like 5deg would be sensible to ensure runoff.

A green roof option would be great, obviously at higher cost. And I'm also intrigued by the EPDM roof that @Tennentslager used on his hut.

[ProDave]

195 frame filled with frametherm, 100mm wood fibre and render outside, OSB, service void plasterboard inside.  U value 0.14 allowing for bridging.

[Tennentslager]

MmEPDM is easy peasy and as it's all one piece you can be pretty confident of staying dry!

I believe it's the liner for most green roofs but that's a whole other proposition with much beefier supporting structures.

Don't be shy with playing around with some design thoughts re the roof...could be a big statement for a small building.

If it's a plain mono roof why not have the run corner to opposite corner?

I had fun with my shed roof and think it works.

I saw your post when I was up at the hut so took some pics of the roof.

This one shows the EPDM lapped up to complete the gable join. Just stuck on to fresh OSB with contact adhesive.

Another junction here using a jointing strip. Really easy, if you've repaired a cycle puncture before, EPDM is a skoosh.

Finally, look up the school at Sagrada Familia to see the fantastic offset wave roof by Guidi...would work well on a little building?

[Crofter]

The sketch that I'm working to is very utilitarian at the moment- minimum cost design. A 16x16ft box, 8ft high with a 4" fall across the roof. Corrugated steel cladding on all surfaces.

It's a fairly brutal look, but stick a grass roof on there and you might get away with it.

Alternatively setting the front wall back a foot or two could help, but it complicates things quite a bit and you lose floor area.

[Tennentslager]

4" fall isn't much...2400 one side and 2100 the other would be better.

12" fall in old money ??

[Stones]

@Crofter Have you sounded out planning regarding your proposals or are you intending to rely on the provisions of the 1960 caravan act / 1961 order relating to crofting land?

[Tennentslager]

10 minutes ago, Stones said:

@Crofter Have you sounded out planning regarding your proposals or are you intending to rely on the provisions of the 1960 caravan act / 1961 order relating to crofting land?

Oh what's that?

[Crofter]

22 minutes ago, Stones said:

@Crofter Have you sounded out planning regarding your proposals or are you intending to rely on the provisions of the 1960 caravan act / 1961 order relating to crofting land?

 

I don't have a specific site in mind yet, it's just a costing exercise at this stage. But you've piqued my interest now...

[Stones]

@Tennentslager Briefly, Part 5 of Schedule 1of the General Permitted Development (Scotland) Order 1992, details how land can be used as a caravan site. It refers you to Schedule 1 of the 1960 Caravan Act which lists the permitted uses, one of which is the use of holdings of 5 acres or more. That particular use can be amended by an order of the secretary of state, and was in 1961 (the full title of which escapes me at present) which reduced the acreage to two acres and increased the period of time caravans could be on the land, provided that land forms part of a registered croft.

[Ferdinand]

21 hours ago, Crofter said:

Another option would be to use Kingspan's insulated roof sheets as my exterior cladding, but how do you manage vapour permeability with this system?

 

Just a note: if that is the corrugated sheet with attached insulation, then perhaps take care to make sure you can work with it first.

 

I had a family member use that to reroof a rear bathroom extension on a terrace and they had difficulty with things such as cutting it and fixing through with long screws etc. It may not have been the Kingspan version, and this was back in the early noughties.

 

One problem was finding a way to cut corrugated metal cleanly with something that could cut a thickness of several inches of insulation at the same time. Another was disposal of offcuts.

 

I am sure it should be fine with the right tools, but they found it awkward to handle and relatively expensive. 

 

Ferdinand

 

Edited September 18, 2017 by Ferdinand

[Crofter]

@Ferdinand I've discounted the Kingspan panels already as I can't really see a way of using them on the outside of a small timber frame domestic building without it creating condensation problems. The other issues you mention are also valid- I think there's maybe a reason why we see these things used to build warehouses rather than houses.

[ProDave]

I believe I heard that a croft can have 3 "caravans" on it's land. this was told to me by the owner of a croft camping site who would take any number of tents, but only ever 3 caravans at a time.

[SteamyTea]

Why not DIY SIPs.

Basically just a couple of sheets of OSB, a sheet of Celeotex or similar, some posts in the right place and a pot of glue.

Knock them up in a garage, tow to site, nail together and then clad in tin.

[Tennentslager]

13 hours ago, Crofter said:

The sketch that I'm working to is very utilitarian at the moment- minimum cost design. A 16x16ft box, 8ft high with a 4" fall across the roof. Corrugated steel cladding on all surfaces.

It's a fairly brutal look, but stick a grass roof on there and you might get away with it.

Alternatively setting the front wall back a foot or two could help, but it complicates things quite a bit and you lose floor area.

I'd like to do one like this but don't think planning would approve ?

Bottom two are the same, top on a variation on the same theme.

Would be fairly easy to construct and would work with 9mm OSB covered in EPDM...* Whistles ruefully while daydreaming*

???

[A_L]

On ‎17‎/‎09‎/‎2017 at 11:35, MikeSharp01 said:

Now you have me worried as two of the things yout depreciate are in our design for the house. The wall build up, inside - out, is as follows.

 

Plaster board 12.5mm 

Service void 50mm

Air tight VCL

Frametherm 38 300mm (between I joists)

External OSB sheathing 15 mm

PU insulation 40mm

Outer Tyvek membrane

Counter batten's and batten's 50mm

Marley Thrutone tiles.

 

I think that means I should be concerned about the OSB and the external insulation or does the internal air tight barrier sort this?

 

Hello @MikeSharp01, sorry for the slow reply, went out after replying to Crofter, came back late and was too tired to assemble this reply, unfortunately things are now going to get a little complex and perhaps a little indecisive.
First here is the BBA cert for Norbord OSB3 which may be what you have http://www.bbacerts.co.uk/CertificateFiles/38/3857PS1i1.pdf
Here is a useful document that will on page 4 enable you to convert the various units that are used in vapour resistance/condensation analysis  http://builddesk.co.uk/wp-content/uploads/2013/01/vapourResistances.pdf 
I have approximated your build-up as follows

Note the high vapour resistance off the polythene VCL
Based on this build-up here is a condensation risk analysis

 

The model, which operates to BS5250 1989 and BS6229 and suggests that there will be no interstitial condensation under the given conditions. If the vapour resistance of the VCL layer was reduced to about 40Mn.s/g it suggests condensation may start to occur at the OSB layer.
Here is a publically available tool to model condensation risk, which works to BS5250 2002 http://www.builddesk.co.uk/software/builddesk-u/condensation-risk/ a 30 day free trial version is available. Just make sure the external/internal temps and humidity you use are realistic
In summary
1) Your build-up as described appears to have no problem with interstitial condensation.
2) If the VCL was degraded sufficiently a risk could arise, it does not 'fail-safe'
3) If your build-up is exactly as you describe it seems to me that VCL is vulnerable.

 

[Jeremy Harris]

Just to add that there is a concern I have with the BS5250 and BS6229 model.  I know that this steady/slowly varying state 2D model is widely accepted, and there isn't really anything better generally available, but it doesn't model dynamic changes.  In the main, this doesn't matter, but when I looked at modelling the vapour transmission rates and rate of change of temperature of details like a SIPs wall to floor junction, it was clear that there was a condensation risk, even with a very good internal VCL.  The risk came from external vapour movement during periods when external humidity and temperature were changing quite quickly, especially the local conditions close to the structure.  A typical example would be a cold, still, night, with moisture condensing on the ground and external surfaces of the walls, followed by a still, sunny, morning that warmed up the ground and walls and created a zone of high humidity close to the structure.

 

What can then happen is that, unless there is an adequate layer of insulation before the first component of the structure with limited vapour permeability, say the OSB skin and timber sole plate, the vapour concentration and temperature of the area around the sole plate can be such as to cause the water vapour to condense.  If the sole plate region stays cool, as a consequence of heat conduction through the foundation (and this is a particular SIPs issue, as both skins need structural support) then there is a chance that there will not be enough sensible heat to cause the water to change phase and move back out as the external humidity drops.

 

The challenge is to come up with a good dynamic model to allow this to be checked as quickly and easily as the existing, reasonably well proven models.  I've not yet seen one, and it would need to be a full 3D model that allowed the actual house structure to be fully modelled.  Not an easy thing to do, but perhaps something that needs to be looked at if using relatively new methods of construction where issues like this could arise.

Edited September 18, 2017 by JSHarris

[Stones]

10 hours ago, ProDave said:

I believe I heard that a croft can have 3 "caravans" on it's land. this was told to me by the owner of a croft camping site who would take any number of tents, but only ever 3 caravans at a time.

 

That's the 1961 order I referred to in my earlier post.

 

I'll be posting in more detail soon.

[SteamyTea]

2 hours ago, JSHarris said:

The challenge is to come up with a good dynamic model to allow this to be checked as quickly and easily as the existing, reasonably well proven models

I would think that a temporal statistical model would do that.

Just run the above model though the different temperature and humidity conditions for an area, then weight those results to local weather data.

[Jeremy Harris]

11 minutes ago, SteamyTea said:

I would think that a temporal statistical model would do that.

Just run the above model though the different temperature and humidity conditions for an area, then weight those results to local weather data.

 

 

Thinking about it a bit more, I think you may well be right, and, as we're only really interested in the worst case (as this is just a risk-reduction model), we could make do with 2D, and some data derived from some sample measurements of a few standard details.  What would be useful would be to have a better idea between the weather and the temperature and humidity at various locations, like outside the base of a wall, inside the lower part of the any external cavity and at the sole plate to foundation junction.

[SteamyTea]

That is really down to just collecting data.

But I think that simply extending the extremes of a model would be good enough.