Quick background for those who have not seen my introductory thread in the newbie section:

We are building a 1.5 storey home with 6.5m ridge height and 38 degree roof pitch. We have tried to keep the roof construction as slim as possible to minimise the impact on headroom under the sloping ceiling. For Building Control the architect had designed the roof to use steel ridge beams and 150mm deep rafters cut on site. If I picked him up correctly, this is the construction most likely to be queried structurally and if an alternative such as trusses are later specified, the supplier will supply calculations for submission to BC.

The original warm roof plan had a predicted U value of 0.12 (SAP report calculations suggest an average of 0.14) and hopefully minimal thermal bridging. From the top down it was made up as follows:

roof slates, battens over counter battens, breathable membrane, 100mm PIR insulation, 150mm rafter full filled with open cell spray foam, plasterboard. Ignoring the slates, battens and plasterboard as I imagine they will remain constant in any application, the intermediate depth is 250mm.

We were planning on using "Foam-Lok", the SAP assessor entered "Thermloc" which may or may not account for the discrepancy in U values.

Any potential wiring was to be fed through conduit before the foam is installed and provision for a vapour barrier below the rafter may be needed for Building Control. If I am correct, the vapour barrier would need to be kept off the plasterboard by a batten adding at least a further 25mm to the depth so 275mm+ in total.

The structural engineer's report showed 10 steel beams were needed to support the roof, some with rather ornate looking endplates and connections. This sounded costly (plus the labour intensive nature of a cut roof to begin with) and as I'm not keen on having steel up there in the first place have been looking into the alternatives.

A quote in from a roof truss manufacturer has specified a 222mm (I will call this 225 for easy maths!) deep rafter instead of the 150mm we had planned for with the cut roof, which increases the intermediate depth to 350mm. After phoning them to discuss, they say this depth is needed for the roof to be stable structurally. To avoid interstitial condensation I don't think we could fill the whole rafter with foam even if the budget allowed so there is no real benefit in performance.

Other options I am looking into currently include:

302mm I beam rafters with glulam ridge beams, full filled with cellulose, grey bead or fibreglass and sheeted with OSB or membrane as appropriate. Not a lot of depth saved by the time you factor in possibly 2 sheets of OSB, but at least the entire 302mm is insulation so U value should be better. Apart from looking a lot more expensive so far (structural timber alone is £3k more than trusses), it is also likely to be more labour intensive than the spray foam.

The truss manufacturer can add a 75mm deep timber on top of the trusses for about £1k, effectively making it 300mm deep which again could be full filled, most likely with fibreglass. I imagine there could be more of a thermal bridging effect though as solid timber would effectively pass through the entire depth.

A third option could be to use the current truss design, but this time add a 75mm counter batten below on site (sandwiching the vapour control layer between if this is the right place for it) and raise the wallplate to compensate. Filling the truss with 125 then 100mm fibreglass and putting 75mm fibreglass between the battens. This in my eyes would reduce thermal bridging to the points where the battens intersect the rafters and the battens could hold up the first 2 layers of insulation while any cables/conduit are installed. If the 75mm could house any services, the full 300mm would be insulated and with methods/materials I could install myself to save money.

If there are any obvious flaws in the aforementioned please point them out or if anyone can put together a different solution I am open to hearing your suggestions:)

Have you looked at SIP panels or the like for the roof construction.

I'm using this firm and having a bungalow with a vaulted roof with no purlins.

http://www.flighthomes.co.uk/hybrid-sips

Posted By: slidersx200
The structural engineer's report showed 10 steel beams were needed to support the roof, some with rather ornate looking endplates and connections. This sounded costly (plus the labour intensive nature of a cut roof to begin with) and as I'm not keen on having steel up there in the first place have been looking into the alternatives.

Send a copy of your plans to U-Roof in Yorkshire - they should be able to do away with the costly on-site labour cut roof and spin out the whole roof structure in light galv steel panels and joists. The rafters will help your height limitations as they will only be 100mm deep. Half what your proposed ones are. Warm roof with OSB deck as Tom suggests. Hopefully they will be able to engineer out some of the more expensive steels the others have specced and come in with something that is a lot more efficient.

There may be an economic truss based solution that the truss makers have not so far considered. Rather than using chunky 225mm rafters in the truss they could make a "parallel chord scissor truss". Here the "rafter" timber is two lighter timbers strengthened by webs, and only horizontal is at collar tie height. Stuff (full fill) the depth, and between the webs and bracing, with glassfibre wool batts.

Not sure this makes the roof as "slim as possible" as asked for in the OP, but there are no taping issues for air tightness. This works well structurally (depending on span of course), but is a non-standard shape and the truss manufacturers don't have a button for it, so you may have to encourage them to model such a design.

Posted By: slidersx200The only difficulty I could foresee with making the air tightness layer at the horizontal ceiling with any method is that we will have ducting etc. for MHRV in the void above so access hatches are presumably necessary which may be difficult to seal.

MVHR being fitted as I type. Ceiling level air tightness, both in the sloped ceiling attic room and in the other rooms, being provided by polythene VCL above the plasterboard, and no service cavity. Lack of service cavity generally is causing me some concerns now I can see how much wiring will penetrate the VCL and need to be sealed, but the MVHR vents come with nice grommets and plates that will make a seal easy.

No hatches planned for the small attic room flat ceiling, not enough space up there in the void to move around in. Using long MVHR pipe runs with only join being at the vent and tested before the ceiling and VCL are in place. Much like the plastered walls, the intention is not to go in there once fitted.

Posted By: slidersx200You may not have been suggesting it, but if the insulation level for the void was at the flat ceiling could there also be issues with condensation occurring in ducting or unwanted cooling to the air inside?

Not sure what you are saying here. The MVHR pipes will be insulated, and surrounded in insulation (stuffing the void with fibreglass), so don't see any condensation or heat loss issues.

If there is a flaw in any of this someone let me know because we are about to do it!

Posted By: slidersx200The only difficulty I could foresee with making the air tightness layer at the horizontal ceiling with any method is that we will have ducting etc. for MHRV in the void above so access hatches are presumably necessary which may be difficult to seal.

Do MVHR ducts need to go in this space? Better to keep them inside the air barrier. Could they run in the floor void below &/or run vertically in internal walls & terminate in "wall throw" directional air valves which use the 'coanda effect' ?

David

Posted By: davidfreeboroughThanks for sharing the picture, looks great. How deep are the parallel chord rafter sections? What are the span & roof covering?

I have to admit I am rather proud of that truss design, it has worked out a treat. Far better than the "super chord" truss solution that some manufacturers offered. Parallel chord sections are 400mm (122 outer, 97mm inner timbers), could have been shallower but I wanted the space for insulation, 37.5° pitch, span 5m between wall plates, stressed for 650kg dead load and zone 4 wind. We have used natural slate as covering so it is well under loaded.

Its only impractical if you use a trussed rafter form of construction & want to fix linings directly to the trusses. If you have a ridge beam or purlins then the air barrier membrane can follow the line of the underside of the rafters. Any ducts, pipes or cables could run in a batten space &/or false ceiling hung from ridge beam/purlins, but most would be coming from below anyway & could be just as easily run in the eaves &/or floor space.

David

Posted By: davidfreeboroughIf you have a ridge beam or purlins then the air barrier membrane can follow the line of the underside of the rafters.

Yes, can do that too with a parallel cord scissor truss.

Any ducts, pipes or cables could run in a batten space &/or false ceiling..

I think that battened service cavities, false ceilings etc. (big enough to take 3" pipes) are key to a do it all inside the air barrier design.

DJH I hope that your plan works out, very interesting to see your strategy. For me having a bungalow rules out 1). We are doing 2), lots of standard lamps. Use of corners and ducting 3) has limitations given house shape, and IMHO visible boxing means that the pipes/cables are only half concealed. Shape of my house means 4) not viable. I like the timber box idea of 5), but also some tapes are made to plaster in just as well (I hope).

The ultimate I guess would be an inner service cavity shell within the air tightness barrier, within the insulated wrapper. Then again how do you get to the air tightness layer if there is a leak?

Sorry slidersx200 if we have highjacked your thread, but hope it is of interest.

That sounds workable. Its not ideal having the vapour check part way through the build-up, but with three quarters of the insulation outside the vapour check you shouldn't get condensation problems.

There are a number of ways of supporting a false ceiling when using a ridge beam & I beam rafters. The most straightforward is to fix a ceiling perimeter batten through the air barrier to the rafters & then fix ceiling joists to this. If the span is too wide for a 100x47mm ceiling joist then support it at the centre with a strap to the ridge beam.

If you use two purlins instead of a ridge beam then these can be placed at false ceiling height & the ceiling joists hung directly off them.

Alternatively there are lots of low cost metal framing systems available.

David

I can really recommend the method of using 18mm T+G OSB/3 (PU wood glue in joints) as the airtight layer - very quick/easy and all the services goes inside it. I got a load of cheap PIR (SIPs panel factory waste), so I used that, but would consider using blown-in warmcel on the inner portion (PIR or EPS outside) if I were doing it again (without the cheap PIR).