Thinner and better alternative to 150mm PIR in cold roof insulation

[Adsibob]

I found this thread, which discussed frametherm vs PIR very helpful in educating me about the benefits of greater decrement delay in insulation materials. The article that @Jeremy Harris posted in particular was enlightening. 

 

I need to insulate a pitched cold roof of my ground floor extension. My architect has specified 150mm PIR insulation. However, discussing this with him he explained that one of the issues we have is reduced ceiling height in that room due to a structural issue that has arisen. If possible, I would like to reduce the thickness of the insulation to 100mm or even 80mm, without compromising the insulation. It would be nice to go with something with more density to also help with decrement delay.

 

I'm considering Pavaflex or SteicoFlex wood fibre boards which both have a lamda value of 0.038 W/(mK) and a specific heat capacity of 2100 J/(kgK). Further details attached. As far as I can tell, these products perform similarly, the main difference being cost. For some reason SteicoFlex is quite a bit cheaper, working out at about £11 plus VAT per square metre for the 120mm thickness and less for the 80mm thickness.

 

Although the lamda value is slightly worse than PIR's 0.022 W/(mK), the specific heat capacity of wood fibre board is so much better that it should be equivalent or possibly better even using a thinner amount. Does anybody know how thick the wood fibre boards would need to be to replace the 150mm PIR and still give me a similar U-value?

 

If I can make this work, I may also consider wood fibre boards for my loft conversion, where decrement delay is going to be key, but I'm not there yet.

 

PAVAFLEX_engl.pdf Steico_Flex_technical_data_sheet.pdf

[Ferdinand]

It is proportional to the U-value, everything else being equal.

 

That will be 3/4 thicker for the same insulation value.

 

To out perform PIR you will be into aerogel afaik.

 

Calculator: https://www.vesma.com/tutorial/uvalue01/uvalue01.htm

[Russell griffiths]

I’m not understanding 

if it’s a cold roof then your insulation is fitted between the joists, so how does reducing the insulation thickness gain you headroom. 

Surely the ceiling height is dictated by the joist thickness, not the thickness of the insulation in between the joists. 

 

You wont find much to better pir for insulation value versus thickness. 

[Adsibob]

37 minutes ago, Russell griffiths said:

I’m not understanding 

if it’s a cold roof then your insulation is fitted between the joists, so how does reducing the insulation thickness gain you headroom. 

Surely the ceiling height is dictated by the joist thickness, not the thickness of the insulation in between the joists. 

 

You wont find much to better pir for insulation value versus thickness. 

Yes, you are right, but we are also looking at having thinner joists. 

[Russell griffiths]

Just now, Adsibob said:

Yes, you are right, but we are also looking at having thinner joists. 

Whatever roof buildup, I don’t think your joist size will go below 125mm 

[Olf]

8 hours ago, Adsibob said:

Although the lamda value is slightly worse than PIR's 0.022 W/(mK), the specific heat capacity of wood fibre board is so much better that it should be equivalent or possibly better even using a thinner amount

No. 

Wood fibre boards in your own numbers are nearly twice worse than PIR - and as a result you need nearly twice the thickness for equivalent conductive heat transfer resistance.

Decrement alone does little, it is how it then translates into phase shift - and that only helps if >10h (so the morning heat only starts flowing in the evening, peak midday heat would only spread in the night and is being cancelled), requiring larger thickness than 100mm. On top of that it stabilises heat changes during daily cycle, but does not affect long term heat loss - here low U is still what matters.

My advice:

- stick to PIR, if desperate go phenolic (0.018) - but that comes at a price

- improve air tightness

- if there is any void between ceiling boards and insulation you can use reflective film better than the face of PIR boards - but this, like phenolic, gives you diminishing return

Edited February 3, 2021 by Olf

[JamesP]

I fitted Pavaflex above the rafters, a warm roof and would recommend as much as possible.

As @Russell griffiths says, rafter size limitation as well as your maximum external heights.

[Adsibob]

33 minutes ago, JamesP said:

I fitted Pavaflex above the rafters, a warm roof and would recommend as much as possible.

As @Russell griffiths says, rafter size limitation as well as your maximum external heights.

Thanks @JamesP. I don't think i have the space for a warm roof for my extension, but I could do this in my loft conversion where space is not such an issue. Did you just have Pavaflex or did you combine with PIR? Would be interested to know your build up and the thickness of each layer.

[JamesP]

@Adsibob I used 40mm Pavatherm to cover all the walls and roof. Then 265mm and 275mm of Rockwool in 3 layers. Drawing enclosed.

 

Floor, Wall, Roof Build up Jan 2018..pdf

[Temp]

10 hours ago, Adsibob said:

If possible, I would like to reduce the thickness of the insulation to 100mm or even 80mm, without compromising the insulation.

 

I think only Aerogel is a better material than PIR but I don't think its available in thick enough sheets to give you the same insulation.

 

 The lambda value of Spacetherm (an aerogel derived insulation) is 0.014 W/mk according to..

 

https://www.phstore.co.uk/aerogel-spacetherm-blanket

 

That compares with around 0.022 to 0.032 W/mk for PIR

 

So you would need between..

 

150*14/32 = 65mm

And

150*14/22 = 95mm

 

Of Spacetherm to equal 150mm of PIR.

 

Two layers of 40mm making 80mm in total would probably do it.... but at a cost of  £375 per square meter !

 

 

[Temp]

An alternative  might be a combination of a multifoil type insulation and PIR but opinions are divided on how good multifoil is in practice. Your BCO might have an opinion.

[PeterW]

11 hours ago, Adsibob said:

If possible, I would like to reduce the thickness of the insulation to 100mm or even 80mm, without compromising the insulation.

That isn’t possible unless you go to Aerogel. Kooltherm is slightly better than the standard PIR but you won’t get the insulation value halved. 

[Olf]

Actually it is possible: https://www.kingspan.com/gb/en-gb/products/insulation-boards/insulation-boards/optim-r/optim-r-roofing-system

0.007 W/mK, totally airtight and well reflective.

Oh, and £££ that makes aerogel a bargain - but sometimes there is economy that makes it the right choice, otherwise they would not market it.

[Adsibob]

52 minutes ago, Olf said:

Actually it is possible: https://www.kingspan.com/gb/en-gb/products/insulation-boards/insulation-boards/optim-r/optim-r-roofing-system

0.007 W/mK, totally airtight and well reflective.

Oh, and £££ that makes aerogel a bargain - but sometimes there is economy that makes it the right choice, otherwise they would not market it.

@Olf this really is very interested and exciting! (Is it a reflection of the times, or my obsession with my build, that results in me getting so excited about insulation?)

 

Having just searched for it online, I found about 5 retailers who won't commit to a price online, it's all "POA" or "call to enquire". So I guess it's very expensive. But might be worth it given the area of roof we are talking about is actually quite small.

Given the Calculator  @Ferdinandlinked to above doesn't have this Optim-R system as an option, does anybody know how I would calculate how much I would need to replicate the U value provided by 150mm of PIR with this?

 

[Russell griffiths]

That looks very similar to the insulation product from kingspan that I used for a different problem, if it is it was about £40 a metre. 

[Olf]

3 hours ago, Adsibob said:

how I would calculate how much I would need to replicate the U value provided by 150mm of PIR

~ U-0.16 (

The thinnest (65mm) option is a match

 

3 hours ago, Adsibob said:

won't commit to a price online, it's all "POA" or "call to enquire"

They are made to order, to the exact dimensions - call to enquire seems to be sensible option.

 

[Adsibob]

14 minutes ago, Olf said:

~ U-0.16 (

The thinnest (65mm) option is a match

 

They are made to order, to the exact dimensions - call to enquire seems to be sensible option.

 

@Olf, that is really quite incredible that I can save almost 100mm by going for these two Kingspan products or for the aerogel that @Tempsuggested.

 

Now the limiting factor is the rafter. @Russell griffiths you suggested 125mm was the shortest rafter I could use, is that assuming I stick to timber? What about a steel rafter, anyone know how thin short that could be?

[PeterW]

@Adsibob These are VIPs - Vacuum Insulated Panels. 
 

https://www.kingspan.com/gb/en-gb/products/insulation-boards/insulation-boards/optim-r

 

They are bespoke and made to exact size

They cannot be cut, drilled or altered on site 

They have to be protected from damage by another product (Kooltherm In this case)

They are expensive - about 3 times the price of the same uValue per square metre of PIR. 
 

They are an insulation of last resort. Why do you need to reduce ceiling height ..? You are gaining possibly 15mm as the overlay panels go below the joists and the Optim-R goes between. If you have 150mm joists, you can fill to 125mm with PIR and then use 30mm of Kooltherm over the top to get the same uValue for about 35-40% of the cost of VIPs. 

[Gus Potter]

21 hours ago, Adsibob said:

I found this thread, which discussed frametherm vs PIR very helpful in educating me about the benefits of greater decrement delay in insulation materials. The article that @Jeremy Harris posted in particular was enlightening. 

 

I need to insulate a pitched cold roof of my ground floor extension. My architect has specified 150mm PIR insulation. However, discussing this with him he explained that one of the issues we have is reduced ceiling height in that room due to a structural issue that has arisen. If possible, I would like to reduce the thickness of the insulation to 100mm or even 80mm, without compromising the insulation. It would be nice to go with something with more density to also help with decrement delay.

 

I'm considering Pavaflex or SteicoFlex wood fibre boards which both have a lamda value of 0.038 W/(mK) and a specific heat capacity of 2100 J/(kgK). Further details attached. As far as I can tell, these products perform similarly, the main difference being cost. For some reason SteicoFlex is quite a bit cheaper, working out at about £11 plus VAT per square metre for the 120mm thickness and less for the 80mm thickness.

 

Although the lamda value is slightly worse than PIR's 0.022 W/(mK), the specific heat capacity of wood fibre board is so much better that it should be equivalent or possibly better even using a thinner amount. Does anybody know how thick the wood fibre boards would need to be to replace the 150mm PIR and still give me a similar U-value?

 

If I can make this work, I may also consider wood fibre boards for my loft conversion, where decrement delay is going to be key, but I'm not there yet.

 

PAVAFLEX_engl.pdf 361.42 kB · 1 download Steico_Flex_technical_data_sheet.pdf 828.75 kB · 2 downloads

Just a thought. I wonder what the structural issues are. I would start here. Often with an extension you need to get the pitch so you can get the tiles to work at low angle and if two storey get the flashings / cavity trays to function. One way of making a roof thinner is to close up the spacing of the rafters, however you have more wood that is more conductive than say PIR insulation, law of return and so on. 

 

I did a dormer attic conversion where I used cold formed steel rafters and stuck a warm roof on top, managed to squeeeezzzzze it in by about 10mm so it did not stick above the main ridgeline of the house. There are standard sizes available and brackets too so it may be worth a look if you are tight for space.

 

A cold formed steel rafter is just like what you see on a B & Q roof but can carry more load than a timber rafter of similar depth  I'll leave out the width for now. You can stick a rubber say EPDM membrane on a shallow pitched roof.

 

 

 

Edited February 3, 2021 by Gus Potter
typo

[Temp]

22 hours ago, Adsibob said:

 

I need to insulate a pitched cold roof of my ground floor extension. My architect has specified 150mm PIR insulation. However, discussing this with him he explained that one of the issues we have is reduced ceiling height in that room due to a structural issue that has arisen.

 

Can you tell us more about issue? Photos?

 

Perhaps there is a better solution than expensive insulation.

[Adsibob]

33 minutes ago, Temp said:

 

Can you tell us more about issue? Photos?

 

Perhaps there is a better solution than expensive insulation.

 

@Temp you may well be right. The issue is mainly about increasing the ventilation gap at the top of the cold roof. I'll try and explain what you are seeing below:

 

I have a green roof (the shallow type for very basic plantings). This slopes at approximately 20 degrees down towards a concealed gutter. Then there is a wall which is right on the border with my neighbour's garden. Below the green roof is a cold roof. We are concerned that there is insufficient ventilation because the entry to the ventilation area is itself concealed because it is so close to the concealed gutter. Another issue is that the gutter is so close to the entry that any foliage falling on the gutter might block the ventilation. One way of mitigating these two issues would be to  reduce the thickness of the insulation, allowing the ventilation gap (which is currently a channel that is 50mm wide, but with a narrower entry point) to be wider and the entry point to be wider. If we did that, we would probably also drop the positioning of the gutter by 25mm or so.

 

The reason the gutter is concealed is both for aesthetic reasons and because if it was on the other side of the wall it would pretty much overhang the neighbour's property.

 

The design also includes 70mm of resilient bars (the green things between the plasterboard and the 150mm of PIR) but we are doing away with those anyway as I'm not sure sound proofing is essential under this part of the roof, whereas head height is an issue and so we are looking at other ways to increase head height. 

 

 

 

 

 

 

[Temp]

Perhaps I've missed it (it is nearly 1am) Im still not sure why you can't just raise the roof say 150mm higher.? Is the roof height limited by Permitted Development rules for the eaves height near a boundary?  A window on the side of the house?

 

 

[Adsibob]

41 minutes ago, Temp said:

Perhaps I've missed it (it is nearly 1am) Im still not sure why you can't just raise the roof say 150mm higher.? Is the roof height limited by Permitted Development rules for the eaves height near a boundary?  A window on the side of the house?

 

 

It's not PD, it's planning permission. I have planning permission for the height shown, not higher.  The reason we are in this position is that we had planned to sink the extension into the ground and drop the floor of our existing house, but we cannot do that to the extent we planned on doing because on exposing the foundations they are shallower than we had expected. So as expensive as aerogel or this special kingspan product is, it's going to be a lot cheaper than underpinning my house! 

[Russell griffiths]

You seam a bit mixed up with this, how can you consider making the roof rafters smaller with the high load from a green roof, the roof loading will be higher than normal. 

Im afraid I don’t like the design at all, green roof is pointless, gutter details are a disaster waiting to happen, leaf debris building up will lead to the gutter no working correctly and water getting into the structure. 

 

Nope, not for me I’m afraid. 

[Cpd]

You had a plan - lowering the floor and building a green roof. 
that plan is now shot as you don’t want to lower the floor due to cost. 

you need to go back to the drawing board and start again as your now trying to get a square peg into a round hole. 
I think by starting to replan with the new restraints you could come up with a better option. 
Also take on board what Russell has said about the existing design not being best practice. 
I had a similar situation and had to go down the underpinning route.....