I need to to insulate and plasterboard a previously unfinished room upstairs in our chalet bungalow.
It will be a room for a baby, so it needs to be comfortable in winter and summer (we get some hard frosts here). Also, it is the first time I have tackled a job like this, so I need something that is reasonably easy to fit. I don't believe I am required to comply with building regs.
The rafers are 100mm deep and are spaced between 350 and 450mm approx. I think this rules out the "raftersqueeze" type insulation, as some of the gaps are too wide.
I've been looking at Celotex which looks ideal - easy to work with and quite cheap. I know I need to keep a 50mm air gap (the sarking is not breathable), and a local builder told me that 50mm of Celotex would be plenty. Reading the Celotex site, they suggest 50mm between rafters + 35mm over the top of them, sealed with foil backed tape, and then the plasterboard on top of that.
I'm wondering if the extra 35mm is worth it for comfort - will the difference be noticeable in temperature terms? It would cost more, and take away from the room which is not large as it is.
50mm is OK..its to regs, and will be an immense improvement. The purpose of the 35mm over the top is because the Celotex is SO good that the rafters themselves will be a significant cold spot..covering the lot removes that problem.
Having said that, I am sitting in a room with exactly what you propose over my head right now, and its more than good enough..
Unless you have a BCO breathing down your neck go for just the between the rafters bit BUT....
Be UTTERLY meticulous about sealing with the tape and cutting the celotex accurately so there are no airgaps. And when you plasterboard up, seal every joint to every frame with decorators caulk. The smallest airgap can have a huge impact in an icy winters gale.
I'm not particulalry au-fait with this, so I'm approaching from physics principles:
[Using
formatting link
source of data]
If you put 50mm Celotex between the rafters, at first sight you get insulation to the effect of:
Thermal-conductivity (K-value) = 0.025 W/mK, which is a contraction of
0.025 Wm/m2K ie: Watts x thickness / (area x temp difference)
So 1m2 at a temp diff of 1K for 50mm Celotex would transmit about:
So, where A = area, d = thickness, P = power, T= temp diff,
K=Pd/AT P=KAT/d
So P (50mm & 1m2 & 1K) = 0.025 x 1 x 1 / 0.05 = 0.5W.
That's apparantly good, as for, say, 50m2 of roof area and 20C difference you need to supply 0.5 x 50 * 20 = 500W to keep the loft warm in winter in a largish house.
Problem is that wood has a much worse K-value (varies, I'm picking 0.35W/mK which I've seen for fir for this example.) Similar to electrical resistors, your rafters are thermally shunting the Celotex by being thermally in parallel.
Repeat above for wood, assuming rafters are 100mm deep:
P=KAT/d = 0.35 * 1 * 1 / 0.1 = 3.5W. Bit meaningless on its own, lets take into account the expected areas of wood and celotex in a 1m x 1m section of your roof, assuming the rafters are 50mm wide and spaced 500mm apart:
Number of rafters in 1m width of section = 2, so area of wood = 1 x 2 x 0.05 = 0.1 m2 thus area of celotex = 1 - 0.1 = 0.9 m2
Redoing the calculations, for 1C difference for 1m x 1m section:
Wood: P=KAT/d = 0.35 x 0.1 x 1 / 0.1 = 0.35W (that's a lot for the tiny area)
Celotex: P = KAT/d = 0.025 x 0.9 x 1 / 0.05 = 0.45W (for 9 times the area of the wood - gives you a feel how good celotex is)
Add together, you now have 0.35 + 0.45 = 0.8W
so for 50m2 roof at 20C diff, total power = 0.8 x 50 x 20 = 800W - quite a bit worse.
So you can see that adding more celotex inside covering the rafters is likely to improve things.
What's wrong with the above calculation:
a) Noone (I assume) does it like that in reality - it's an illustration.
b) The same temperature diff is assumed for both materials - in practice, one might expect the wood to be colder on the outer face as it is abutting the tiles. The celotex has a layer of contained air, so might have a lower temperature difference across it. However the air is moving and sourced from outside (your ventilation) so I'm not even going to try and work that out.
c) I ignored the internal plasterboard and roof tiles completely.
Adding 0.035mm celotex to the inside:
Well, the 0.8W/K for a single 1m2 section represents thermal conductance which (like electrical equivalent) is the inverse of resistance.
So thermal resistance of rafters + 50mm celotex is 1/0.8 = 1.25K/W Without repeating details above, conductance of 0.035mm celotex for 1m2 section is 0.714W/K, so resistance is 1.4K/W. Like electrical resistors in series, you add these giving:
1.4 + 1.25 = 2.65, so final conductance = 0.377 W/K
So 50m2 roof at 20C diff now needs 377W to heat, compared to 800W without the 35mm internal layer of celotex, so adding the extra layer make the insulation twice as effective.
The interesting question is: if headroom were not a problem, would it be better (cost effectivly) to add more depth of celotex inside and not bother packing any between the rafters at all. I suspect it might be so.
=============================
Sorry - not a very good explanation, I was doing it partly for my own benefit too, out of interest.
There might be a hideous error(s) in there - best way to really evaluate it would be to get a copy of Superheat from here:
formatting link
work it out properly.
(Free trial version, limited time and no printing). I believe the author is a regular here...
Not what you wanted to hear - but if you can construct a spreadsheet or grab Superheat you can play with the numbers and reach a compromise perhaps?
I seem to remember that there might be a combined urethene foam/finishing board composite product which might save you a tiny bit of depth. Might have been on either Celotex's or Kingspans websites?...
Since were on the subject - I have no idea about ventilation requirements, other than there has to be some. Perhaps you could help me out here, for future reference:
1) Dealing with a hipped roof - how do you ventilate the hipped bit? Is one vent tile between each pair of rafters enough at the top end?
2) If fitting soffit vents would be a pain, is a vent tile low down (betwen each pair of rafters) sufficient?
3) If using ridge vents for the main part of the roof, I've seen roofs with maybe 3 such vents - not one between each pair of rafters - so do you form an air cavity in the inside of the roof apex to allow air to circulate up from between several sets of rafters to one vent? Something like: _ / \
That is basically what I did, and can confirm that it does give a very good level of insulation.
Since this is a job you will only likely do once I would recommend the extra layer under the rafters since it ought to pay for itself reasonbly quickly, and will make some difference ot the room comfort. It is also much easier to maintain a complete seal (using adhesive backed foil tape).
The loss of headroom is pretty minimal. Have a look at the 2nd photo down here:
formatting link
can see by the roof window reveal the extra 30mm of foam under the PB
Actually Tim I *have* done these calcs myself. And your analysis is totally correct. The best place to put the Celotex is of course inside the rafters not between them. However the recommended 50mm between and
35mm over is a good compromise between space taken up and efficiency. And sells more Celotex :-)
The overall plasterboard does make a significant, but not large difference.
You cannot rely on air trapped between the Celotex and the roof,because under windy conditions, its an icy draught there. In this house, done that way, heat loss in a wind is several times higher due to underfloor and soffit ventilation. Once I've got rid of the BCO I'll cut both down a bit. Way more than is needed.
Ive got hips and part thatched part tiled roof space - all 'cold'.
The BCO made me put soffit vents in everywhere and ridge vents. My ceiling are only taking up part of the roof space..like yours.
Given the 'A' shape of the roof cross section as drawn, you need celotex up inside the rafters till the horizontal part of the A,. then insulation over the flat bit of the ceiling. In my case thats rockwool between 6" joists, with chipboard over to reduce draughts.
Essentially the part of the roof above the insulation should be a freely ventilated 'umbrella' with as little attempt to stop air movement inside it as possible.
This is best for damp build up prevention, however restricting airflow allows the air inside to warm up and reduces heat loss, so its a balance between insulation and ventilation.
In a cold still air day, the ridge vents are used to allow some airflow..any damp warm air that leaks into the roof space will rise, and exhaust from the ridge vents, and be replaced by cold dry air from outside via the soffit vents.
Complete overkill IMHO but thats the regulations.
As far as hips go, its more of the same..run the insulation up to the ceiling only, and arrange soffit vents below the airgap and exhaust into the main ridge space and the ridge vents.
How you arrange soffit ventilation - vent tiles. insect proof slats between the rafters, or actual drilled out soffit vents, is a matter f practical convenience.
The Natural Philosopher wrote in news: snipped-for-privacy@doris.uk.clara.net:
Let's hope he doesn't monitor this group, eh? ;)
It's an interesting question - I asked a builder, and he said no, of course I didn't need to worry about regs. I phoned Celotex today, and they said I did need to comply.
If I remember correctly, the regs refer to a significant alteration, which is a bit vague.. but I could be wrong?
As to whether it's a change of use, I suppose that is a grey area too. Do I need to inform building control? We sought planning to fit Velux windows in that room when we got the extension at the other end of the house, although the drawings make no mention of insulation etc in this room. And yes, indeed, it is currently bare rafters.
I have decided to go for 50+35, since the man from Celotex told me there would be a noticeable difference in comfort levels in the room, and that alone is motivation enough for me. And as someone else said, you only do it once..
That guy also told me that regs changed (this month I think he said) and I would *now* need 50+75 to comply. However, I'm only going to comply with the previous ones.
I'm not particulalry au-fait with this, so I'm approaching from physics principles:
[Using
formatting link
source of data]
If you put 50mm Celotex between the rafters, at first sight you get insulation to the effect of:
Thermal-conductivity (K-value) = 0.025 W/mK, which is a contraction of
0.025 Wm/m2K ie: Watts x thickness / (area x temp difference)
So 1m2 at a temp diff of 1K for 50mm Celotex would transmit about:
So, where A = area, d = thickness, P = power, T= temp diff,
K=Pd/AT P=KAT/d
So P (50mm & 1m2 & 1K) = 0.025 x 1 x 1 / 0.05 = 0.5W.
That's apparantly good, as for, say, 50m2 of roof area and 20C difference you need to supply 0.5 x 50 * 20 = 500W to keep the loft warm in winter in a largish house.
Problem is that wood has a much worse K-value (varies, I'm picking 0.35W/mK which I've seen for fir for this example.) Similar to electrical resistors, your rafters are thermally shunting the Celotex by being thermally in parallel.
Repeat above for wood, assuming rafters are 100mm deep:
P=KAT/d = 0.35 * 1 * 1 / 0.1 = 3.5W. Bit meaningless on its own, lets take into account the expected areas of wood and celotex in a 1m x 1m section of your roof, assuming the rafters are 50mm wide and spaced 500mm apart:
Number of rafters in 1m width of section = 2, so area of wood = 1 x 2 x 0.05 = 0.1 m2 thus area of celotex = 1 - 0.1 = 0.9 m2
Redoing the calculations, for 1C difference for 1m x 1m section:
Wood: P=KAT/d = 0.35 x 0.1 x 1 / 0.1 = 0.35W (that's a lot for the tiny area)
Celotex: P = KAT/d = 0.025 x 0.9 x 1 / 0.05 = 0.45W (for 9 times the area of the wood - gives you a feel how good celotex is)
Add together, you now have 0.35 + 0.45 = 0.8W
so for 50m2 roof at 20C diff, total power = 0.8 x 50 x 20 = 800W - quite a bit worse.
So you can see that adding more celotex inside covering the rafters is likely to improve things.
What's wrong with the above calculation:
a) Noone (I assume) does it like that in reality - it's an illustration.
b) The same temperature diff is assumed for both materials - in practice, one might expect the wood to be colder on the outer face as it is abutting the tiles. The celotex has a layer of contained air, so might have a lower temperature difference across it. However the air is moving and sourced from outside (your ventilation) so I'm not even going to try and work that out.
c) I ignored the internal plasterboard and roof tiles completely.
Adding 0.035mm celotex to the inside:
Well, the 0.8W/K for a single 1m2 section represents thermal conductance which (like electrical equivalent) is the inverse of resistance.
So thermal resistance of rafters + 50mm celotex is 1/0.8 = 1.25K/W Without repeating details above, conductance of 0.035mm celotex for 1m2 section is 0.714W/K, so resistance is 1.4K/W. Like electrical resistors in series, you add these giving:
1.4 + 1.25 = 2.65, so final conductance = 0.377 W/K
So 50m2 roof at 20C diff now needs 377W to heat, compared to 800W without the 35mm internal layer of celotex, so adding the extra layer make the insulation twice as effective.
The interesting question is: if headroom were not a problem, would it be better (cost effectivly) to add more depth of celotex inside and not bother packing any between the rafters at all. I suspect it might be so.
=============================
Sorry - not a very good explanation, I was doing it partly for my own benefit too, out of interest.
There might be a hideous error(s) in there - best way to really evaluate it would be to get a copy of Superheat from here:
formatting link
work it out properly.
(Free trial version, limited time and no printing). I believe the author is a regular here...
Not what you wanted to hear - but if you can construct a spreadsheet or grab Superheat you can play with the numbers and reach a compromise perhaps?
I seem to remember that there might be a combined urethene foam/finishing board composite product which might save you a tiny bit of depth. Might have been on either Celotex's or Kingspans websites?...
Cheers
Tim
Hi Tim,
Thanks for posting this. I had not really grasped the significance of the heat loss due to the timber before. I believe however it may look worse in your calculation than it really is because of the pessimistic figure of 0.35 w/mK that you have used for wood. In your own quoted source, wood is listed as 0.14W/mK which agrees more with another figure of 0.13W/mK I have gleaned from
formatting link
makes the heat loss through wood to be about 40% of the loss that you suggest.
I add this in the spirit of useful info rather than being critical of your posting which I did find useful.
That's good. That means that my understanding is basically correct - which will be useful in the future when I need to do these for real. Thanks for taking the time to vet my half-arsed calculations :)
OK - that's a useful assumption.
Two things have never grokked with me - structure: "how do all the bits of wood in the loft held together with 4" nails not fall down" and ventilation WRT to avoiding rot. I always wondered how stuffing a roof with glass wool under the sarking then plasterboarding over never caused rot, but I suppose it can breath. Dad's house is like that (loft conversion done OK, but not great, and no evidence of rot there after 20 years).
Does seem to lessen the case for worrying about too much celotex under the rafters. Some, to prevent cold spots on the ceiling is probably a good idea
- but as I said, sure I;ve seen some lining board with insulating properties that might be good enough for that. Wish I could remember what it was called.
Not at all critical Bob - it's all very interesting. I don't own a property at the moment, so I'm spending my time gathering knowledge for when the house prices get sensible again (ha!). Now SWMBO has read
formatting link
and has gone off buying for a while, next job is to screw a bit of 8x4 plasterboard to the garage wall and learn to plaster :) I;ve identified that as a big money saving skill worth having.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.