roof insulation question

chillit

Hi

I am looking for some advice about insulating a sloping roof. This ceiling is open to rafter level so the insulation needs to be between and below the rafters. i.e it is not a flat attic where I can stack up lots of rockwool.

The u-value target is 0.16 for the sloping ceiling but I would like to try to target 0.12 to make it the same as the flat ceiling target.

I have 7x2 inch rafters (so 175mm of depth) and it is a cold roof construction.

I seem to have 2 options. Either

A. metac/rockwool between the rafters and pir board below the rafters
or
B. Pir in both locations - between and below the rafters.

From my initial research metac and pir boards are priced more or less the same but pir boards at 0.022 W/mK are much better than metac at 0.034 W/mK

Rockwool is by far the cheapest product but has the lowest performance at 0.043 W/mK. However it still seem to have the best price versus performance metric.

My engineer and roofer say that everyone uses metac these days but I don't understand this based on the fact that the performance of PIR is much better and the price similar. Infact if you have the space (i.e. in a flat ceiling attic) it seems like standard rockwool is by far the most cost effective approach.

Anyway, given 175mm deep rafters what would people recommend. I want to leave the recommended 50mm air gap above the insulation so this means max 125mm between the rafters and whatever else below.

My planned makeup from the outside is

Slates
Battens
Breathable felt
Rafters - with some type of insulation in between
PIR board
Airtightness and moisture membrane
Counter battens for services - probably rockwool/metac in the service channel
Plasterboard

One possible option which gets a 0.12 u-value would be

100mm rockwool (0.043 W/mK) between the rafters
100mm PIR (0,022 W/mK) below the rafters
50mm rockwool in the service channel

Another option is
120mm pir between the rafters
50mm pir below the rafters
50mm rockwool in the service channel

The second option gives me 50mm of extra ceiling height which in certain parts of the ceiling would be useful but probably not essential.

What do people think of these options or do you have any better suggestions?

Am i crazy to target 0.12 u-value on a sloping roof ?

sydthebeat

is the roof finished?

have you considered putting insulation OVER the rafters?

Wegian

I have the same situation (I think) and will need to deal with it in the next few months so I would be interested in what is advised

Anecdotally I believe metac is preferred to PIR between the rafters as it involves less work to fit - I am open to correction on this one

kieran.

is this a new build or a renovation project

chillit

this is a new roof and the rafters and felt are in place so a warm roof (over the rafters) is not really an option.

I do understand that warm roof is probably better but unfortunately I am where I am.

BryanF

Wegian wrote: »

I have the same situation (I think) and will need to deal with it in the next few months so I would be interested in what is advised

Anecdotally I believe metac is preferred to PIR between the rafters as it involves less work to fit - I am open to correction on this one

glass/mineral wool allow for moisture to pass through it, whereas PUR/ PIR doesn't.

air

I did this in a retrofit situation and used 100mm PIR between the rafters and 50mm across them underneath.
If I was doing it again I'd try making up the 100mm between the rafters with 2 layers of 50mm. The 100mm sheets need to be cut with a saw but the 50mm sheets can be cut with a blade, much less mess & a squarer cut overall on average.
I don't think rock wool or anything else similar to it makes any sense off the horizontal plane as it's bound to slump over time.

kieran.

From a cost point of view what you could do is splice another 175 mm timber onto the the underside of the 175mm rafter using field splice plates and place 300mm of a (cheap) 0.04 /m²k Wool between the rafters and then PIR board to the underside of the timber with the joints taped and plasterboard out.

This is all on the assumption that space is not a problem

air

kieran. wrote: »

From a cost point of view what you could do is splice another 175 mm timber onto the the underside of the 175mm rafter using field splice plates and place 300mm of a (cheap) 0.04 /m²k Wool between the rafters and then PIR board to the underside of the timber with the joints taped and plasterboard out.

This is all on the assumption that space is not a problem

I can't see this being economic by the time you buy and install the additional timber and the end result is unlikley to be very effective due to slumping.
Added to that the extra 175mm timbers will add a not insignificant dead load to the roof.
Also it will have a huge impact on available space within the room, had to imagine it would be practical in very many attics.

kieran.

300mm Isover G3 is approx €5.60m², 150mm Kingpspan for the equivalent value is around €26m²

I would not be concerned about slumping on the slope if the insulation is installed properly intially.

air

You also need to allow for about 2.5m of your 175mm wood per m2 and the cost of the spicing plates and the labour to install it all.
On paper it might be the same U value, but in practice the wool on the outside is exposed to wind and won't have any barrier over it which has been shown to radically disimprove it's insulation performance. The Kingspan won't suffer from the same problem.

Out of interest, what's the correct method of installing that stuff on a slope to avoid slumping and maintain the correct (low) installed density?

kieran.

air wrote: »

has been shown to radically disimprove it's insulation performance

Point me to these studies please for modern mineral wool not fibreglass insulation.

air wrote: »

Out of interest, what's the correct method of installing that stuff on a slope to avoid slumping and maintain the correct (low) installed density?

Friction fitted (in layers if necessary) to fully fill the space between the rafters. All layers should be closely butt jointed, and are usually installed from below once the roof is weathertight. If installed in layers, ensure joints are staggered. The roof insulation should be cut to shape to link with the wall insulation to avoid a thermal bridge at the eaves. A continuous vapour control layer should be fixed to the underside of the rafters with all joints, tears and perforations sealed.

chillit

hi kieran

Thanks for the comments

You mentioned to fully fill the rafters but are you not supposed to leave a 50mm air gap between the felt and the insulation?

I agree re your point that standard rockwool is the best value for money IF you have the space to put in enough of it.

I also think that BryanF's point about rockwool allow moisture through is also an important consideration

Wegian

BryanF wrote: »

glass/mineral wool allow for moisture to pass through it, whereas PUR/ PIR doesn't.

Good or bad thing?

Good I assume - just afraid I am missing something

chillit

Wegian wrote: »

Good or bad thing?

Good I assume - just afraid I am missing something

yes this would be good and in general is important.

although with a proper moisture membrane on the warm side of the insulation vapour in the roof should not be a major concern

kieran.

Sorry I was referring to fully fill the horizontal space between the rafters !!

As your felt and I assume battens are in place full fill is not an option for you and you will need to main the 50mm air gap, that's why suggested 300mm in the increased 350mm space plus PIR (taped) + p/board.

Wegian

To clarify,you are suggesting PIR between the rafters but allowing for a 50mm gap between PIR and felt?

air

kieran. wrote: »

Point me to these studies please for modern mineral wool not fibreglass insulation.

What's the difference between the two exactly? They are both open fibrous materials that insulate by trapping air. If there is significant air pressure and movement on one side, it's going to move cold air through the insulation reducing it's effectiveness. Does "modern mineral wool" possess some kind of air tightness properties that I'm not aware of?

I can't quote any studies as I haven't researched it, the principal as explained to me is pretty logical.
Have you any studies that support your own position?

kieran. wrote: »

Friction fitted (in layers if necessary) to fully fill the space between the rafters. All layers should be closely butt jointed, and are usually installed from below once the roof is weathertight. If installed in layers, ensure joints are staggered. The roof insulation should be cut to shape to link with the wall insulation to avoid a thermal bridge at the eaves.

Not convinced by that to be honest, wouldn't have it myself but each to their own.
I have seen claims that a 4% void in fibreglass insulation can result in a 50% reduction in effectiveness.
To my mind it needs a wind tight layer on the cold side and some kind of mechanical retention mechanism to have any hope of doing it's job in the real world.

kieran.

air wrote: »

What's the difference between the two exactly?

In fibreglass the fibers are spun together to make the fiberglass quilt, generally the fibers are long, and therefore hang together easily to make a cohesive quilt.

With mineral wool, the fibers are much shorter, so in order to keep the quilt in one piece, they have to be packed to a higher level of density and are there less prone to slumping and voids . [/QUOTE]

In real world situations PIR is incredibility difficult to fix effectively between rafters especially is you want to fix circa 150mm of it.

But as you say each to there own

air

Have to agree on the difficulty of fixing PIR having done it myself, hence why I recommended multiple thinner layers.
Having said that if I was doing a new build I'd fix the boards under the rafters first and do the infill from above, be far easier to foam the boards into place.

Surely the mineral wool has an inferior U-value if it's packed so densely? It tears very easily too in my experience.

chillit

thanks for all the feedback on this. I'm not sure I am any closer to a decision as there seems to be strongly differing opinions but the input is useful anyway