Posted By: Mike GeorgeI thought the counter battening had some effect on the way in which vapour migrates to the outside air? Also there shouldn't really be any significant volume of water to run off whether membrane pulled tight or not. I'd be interested in reading a BBA which details this if you have a link as like I said it is an assumption [probably wrong] on my part

If the membrane is pulled tight or fully supported then wind driven rain finding it's way through the tiles will not be able to run off under the tile battens. So they need to be lifted up 12mm with counter battens.

You're not wrong, the Tyvek Supro BBA refers to this in Section 13.2 as a a drainage and "vapour dispersal space"
http://construction.tyvek.co.uk/Tyvek_Construction/en_GB/assets/downloads/certificates/roof_bba_tyvek_supro_in_warm_non_ventilated_cold_ventilated_roofs.pdf

David

Thanks David, that makes sense.

Posted By: fostertomDavid, what's the difference between your c) and your d)?
c) and d) [i] seem to support sinnerboy. I'm depressed!

c) is without sarking boards/insulation & d) is with sarking boards/insulation.

d) [i] is what I'm doing & what you would be doing with woodfibre sarking insulation. It requires batten cavity ventilation above or below the draped breather membrane, I would ventilate above the membrane for windtightness.

d) [ii] is what you would be doing with EPS sarking insulation. It does not require batten cavity ventilation.

David

Regarding..

b) forming a ventilated gap between top of insulation and underside of a breathable membrane,

It's not just some BCO that insist on a gap. Some manufacturers of breathable membrane also specify a small unventilated gap. I believe there are two issues..

a) The first has already been mentioned - the need for drape so that rain can run down under tile battens without the need for counter battens.

b) a possible issue with the membrane "wet clinging" to the insulation, forming some kind of chemical reaction or perhaps simply not drying as quickly as it should when wet.

I can't find much info on b) but Celotex say...

"With some breathable membranes, the cold side of the insulation may be placed in direct contact with the membrane (check with membrane supplier prior to specification)."

Knauf say...

"Knauf Breatheline is non-wicking and so can be laid in direct contact with insulation and roof boarding with no risk of moisture penetration."

I'm sure I've seen other references that imply a potential issue. So perhaps best check that the insulation and membrane are compatible before putting them in contact with no gap.

From BS 5250 2002 - As Amended 23 Dec 2005

"Roof coverings and batten spaces
Research has shown that moisture conditions in the space between the outer roof covering and the underlay
are dominated by the external environment, the air openness of the outer covering and the vapour openness
of the underlay. If an HR " ( underlay has a water vapour resistance greater than 0.25 MN·s/g. ) " underlay is used there will be relatively little moisture transfer from within the building to the batten space. In the case of an LR " (underlay has a water vapour resistance of less than or equal to 0.25 MN·s/g.) " underlay more moisture will be transferred: in order to avoid a build up of moisture in the batten space it is good practice to ensure adequate air movement through the void between the outer covering and the underlay. Most traditional unsealed slating and tiling methods are sufficiently air open. "

There you go Tom . Your POV is reinforced here .

" Should it be necessary to determine if the roof covering is sufficiently air open, it should be tested using a
rig as described in Annex L of BS 5534:2003 but with equipment designed to measure low pressure
differences down to 2 Pa. If the airflow at 2 Pa is, or could be reasonably expected to be, greater than 7.8Ar (in m3/h, where Ar is the effective test rig area in m2) the roof covering will allow sufficient air movement in the batten space without additional ventilators.

If the airflow is not, or could not reasonably be expected to be, greater than 7.8Ar (in m3/h), it is necessary to provide either:
a) ventilation openings to the batten space which are equivalent to a continuous slot 25 mm wide in the eaves and 5 mm wide at the ridge and 25 mm counterbattens; or
b) ventilation openings to the roof void or air void below the underlay as specified in 8.4.2.2.3.
In case of uncertainty, the air tightness of the roof covering should be tested, as above, or ventilation
openings to the batten space or loft installed. "

So - if in doubt - test as described above ( as if ! ) or provide vent slots / strips into the batten void .

Posted By: CWatters

b) forming a ventilated gap between top of insulation and underside of a breathable membrane

It's not just some BCO that insist on a gap. Some manufacturers of breathable membrane also specify a small unventilated gap. I believe there are two issues..

a) The first has already been mentioned - the need for drape so that rain can run down under tile battens without the need for counter battens.

This is a good point. However, counter battens achieve this without the risk of unwanted air movement below the sarking membrane, while allowing the insulation to protect the full depth of the rafter. So I would always use counter battens.

Posted By: CWattersb) a possible issue with the membrane "wet clinging" to the insulation, forming some kind of chemical reaction or perhaps simply not drying as quickly as it should when wet.

Most breathable sarking membranes are classed as "non-tenting". So liquid water will not pass through if in contact with another surface & there will be no surface tension type effects. Any breathable membrane which is classed as suitable for use with sarking boards/insulation will be non-tenting. There are some punctured-polyethylene based products which may be vulnerable to this, but I wouldn't use them & you've got to question how good they'd be at protecting the rafters they're in contact with from liquid water.

The best way to keep the rafters dry is to keep them warm (do not insulate below, insulate to full depth & use sarking insulation) & not allow them to come into direct contact with air which may cause condensation. By keeping a non-tenting breathable membrane between the rafters & the ventilated air gap as in d), you prevent liquid water from condensation of the ventilating air coming into contact with the rafters. This has got to be a good thing.

David

Very good, everyone. Getting some clarity here (I think!).

No, I think its its probably bue to 'the 50% rule' which states that 50% of the surface area of timber should be open to evaporation [at least where non-breathable insulations are used] More discussion of this on page 1 here http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=4698&page=1

I don't follow you? In the case of a 50 x 100 rafter. 50mm [say PUR] insulation between would cover 100mm rafter surface area plus 50mm on the inside face. This leaving 150mm open to the air gap. Ie 50%.

I'm not sure but this is where I think the requirement originates though perhaps it has been relaxed somewhat to allow for greater thickness of insulation between wider rafters. I think the the reason for the gap though is a good one where materials such as PUR between are used.

Yes you're right. Just my thought on where this originated.

Very informative thread. Well done all. Moving back to the summary of possible ventialtion requirements in various pitched roof constructions and the principles we just collectively (largely) agreed on, can we consider the case where the construction is a flat roof? How do options a-d change?

The physical principles and surroundings (inside/outside) are just the same. The only change is the angle and the outer covering(s), and the major point that without a significant slope air movement rates may be much reduced (no warm air rising/cool air falling), and arranging ventialtion at both ends is very difficult anyway.

Normally there will not be tiles and thus no battens, but some type of membrane. Either old fashioned waterproof/vapour-proof felt, or more newfangled materials which are all waterproof but may or may not also be vapour-tight. (e.g. cefil (vapour permeable), sarnafil (not sure), polisystem (vapour tight)).

My reading of all the above for this context suggests
a) Don't do it - even worse idea on flat roofs than pitched ones.
b) I'd generally prefer to avoid this construction too but do we reckon a gap is still needed for b[ii]?
c) I don't think this ever occurs as a flat roof design effectively always has sarking boards.
d) This is the interesting case as that's the sensible way to build a modern flat roof. Does anyone think that a ventilation gap is still useful here? I've never seen a design where one is fitted, whether the sarking insulation or meberane is vapour-tight or not. Clearly the gap for drainage is not relevant, but the comments about vapour dispersal still apply, and of course the upper membrane is going to be much more airtight than tiles.

So for a pitched roof a ventilation gap for d[i] was suggested/recommended. For d[ii] it is not needed. For a flat roof we don't have a gap in either case. Is that because without battens there is no need for ventilation to protect the battens themselves? Or perhaps because there is really no such thing as d[i]. Even the 'vapour permeable' flat-roof membranes are in fact not very permeable really, so we are always in case d[ii] in practice?

Thoughts?

d) for a flat roof is in effect a "warm deck" construction. The recommendation here is to keep all of the insulation above the joists & deck. The deck will typically be of non-vapour open plywood with a vapour control layer on top. The insulation on top needs to be rigid to support the waterproof covering, so typically it will be non-vapour open rigid board insulation. So, as you say, this is in effect equivalent to d)[ii]. Ventilation is not required because it's easy to make airtight, the rate of vapour transfer is extremely low & there's nothing on the cold side of the insulation that's vulnerable to decay.

You could in theory use vapour open woodfibre board insulation giving a construction equivalent to d)[i]. However, this would require the creation of a ventilated cavity above the insulation & a "cold deck" to support the waterproof covering. I would have concerns about the ability to get effective ventilation at the centre of the roof leaving the battens, deck & insulation vulnerable to decay. So, if I had to have a flat roof, I would use d)[ii].

David

All that didn't answer the question about why walls dont need to ventilated did it.

David: Putting all the insulation on the outside makes for a very fat roof construction and it seems a shame not to put any insulation in the 150-200mm provided by the roof timbers (typically taller than rafters in a pitched roof). The problem there is that we need vapour to be able to get out of this area and the ply/PUR/waterproof membrane has very high vapour resistance, and normally we want highest vapour resitance on the warm side, lowest on the cold side, so I can see why putting all the insulation on top of the ply is recommended. I'm working my way through BS5250 to get a proper understanding of this. It would be nice to have a construction without 200mm of uninsulated space in it - can we really not think of one that would be robust?

djh: The problem I see with the inverted roof is that letting water down behind your insulation seems a great way to remove all the heat. This is a bad enough problem with air and water shifts hundreds of times more heat. The design no doubts attempts to minimise water movement behind the insulation, but some seems almost inevitable.

Posted By: tonyAll that didn't answer the question about why walls dont need to ventilated did it.

Timber framed walls have typically been built in the same way as warm roofs. The insulation has typically followed the line of the timber structure, the external sheathing board has been continuous & it has not been replaced by tiling battens.

Where the sheathing board is not vapour open a vapour barrier has been used on the inside as in d)[ii]. Where the sheathing board is vapour open, e.g. woodfibre board, then typically an plywood or OSB lining board provides sufficient vapour control. This is equivalent to d)[i] & is how timber framed homes are typically built in Alpine regions.

Both will typically have a ventilated batten cavity, but this is sometimes ommitted in d)[ii], where a shingle or hung tile finish is used.

David