SIP wall thickness: diminishing returns?

[joshwk]

Hi all.

 

We will be constructing our self-build in SIPs and I was curious to understand if there is a school of thought on 'diminishing returns' when it comes to wall thickness and U-value effectiveness.

I realise the black and white answer is 'the lowest u-value is the best' and that there is a significant variable in the form of cost, but what I am curious to understand is if there is any general thinking on what a 'sweet spot' might be when it comes to investing into additional PIR vs the incremental gains made from improving u-value? Currently, I am approaching our build on the basis that walls would be 142mm + 25mm PIR to achieve a u-value of 0.15 and roof would be 142mm + 75m PIR to achieve a u-value of 0.11 but am wondering if I should be looking to improve the wall insulation further? Our property is not excessively large (~200sq/m total) so I am also conscious of trying to maximise room sizes.

We are not going for Passivhaus but as this will be our 'forever home', I'm keen to ensure I'm building out house with the long term in mind. I have included a table below that we received from our SIPs supplier for visibility in case helpful.

 

Thanks in advance.

 

Walls U-Value Table

Roof U-Value Table

 

 

Walls + Roof Summary

 

 

 

Edited June 9 by joshwk

[JohnMo]

People will say this is quite black and white, but it's not. One thing often missed is house form factor. So a house that is a cube is great, you need very little insulation to heat a low W/m2 heating required. However my house is rubbish it's long and thin, single storey, with vaulted ceilings in every room. So although all surfaces are insulated to better than advertised PH requirements it's energy inputs are way higher.

 

Your house as far as form factor goes is about as bad as it gets. Lots of external walls, roof and floor for a given m2. So cram as much insulation in as you can get away with. Aim for 0.1 U values or better everywhere.

[joshwk]

Thanks  @JohnMo, that's a great callout. Very helpful.

[SteamyTea]

Another thing to add in is the local weather regime.

 

There is little difference in mean temperatures across the UK.

There is quite a difference in how they are distributed though.

So West Cornwall may have a similar mean temperature to Hull, but the extremes are closer to the mean.

Sunshine hours are also different, which can make a difference to cooling costs.

 

3 years ago when people where paying 2.5p/kWh for gas and 12p/kWh for electricity, there was a case to be made for just meeting building regulation levels.

But different now that gas is 7.5p and electric 30p.

 

And then there is airtightness.

More energy can be lost through this than through walls, floors and roofs.

 

The much mentioned Jeremy Harris spreadsheet was originally designed to compare price comparisons to help establish best value for money.

Easy enough to use it as such, just fill in different numbers and see what it churns out.

 

One word if caution, if anyone says that "it meets building regs', or 'airtight houses overheat', walk away from them, they will do a bad job.

[Iceverge]

Do the best you can is the only thing I can say. 

 

I used PHPP and my own spreadsheet to refine the insulation levels.

 

In terms of money spent to improve energy efficiency it's a little further down the list than you might expect. 

 

From memory the best bang for your buck was 

 

1.Airtightness

2.MVHR

3. 3g Windows 

4. Roof insulation 

5. Wall insulation 

6. Floor insulation. 

 

The metric I used was payback years. EG an extra 100mm of attic insulation might take 30 years to payback and and extra 150mm might take 50 years. 

 

I started at bRegs and worked my way up year by year until everything was at 25 years payback. I stopped there as I was at passivhaus performance and a bit extra. At least I knew then I was spending our cash in as balanced a manner as possible. 

 

In the end I had 450mm cellulose in the attic, 250mm EPS bonded beads in the wall, 200mm EPS70 in the floor. 

 

 

I could have probably put 300mm everywhere but in our case it would have been dearer for the same result as the floor insulation was much dearer than the cellulose. 

 

[SBMS]

Have you looked at a thicker SIPS panel to do away with the 25mm insulation?

Edited June 9 by SBMS

[joshwk]

Good tips,  thank you @SteamyTea & @Iceverge.

 

@SBMS: 142mm appears to be the standard manufacturing for TEK Kingspan. 172mm is also possible but requires additional work in cutting transition materials down to suit the non standard wall panels width (quoting the images shown above). Per the table, 172mm with no additional insulation is estimated to achieve a 0.16 U/value.

To achieve close as close to a 0.1 U/value as possible based on @JohnMo's suggestion, in my personal situation, I think the best route is to use 142mm but upgrade from +25mm to +75/80/90mm to achieve a 0.11 on the walls (in line with the current estimated U-Value for the roof).

Edited June 10 by joshwk

[Iceverge]

I'm not a massive fan of SIPs in general as they have issues with thermal bridging at junctions. 

 

If I was going to upgrade them I would probably opt for a layer of something like mineral wool to reduce this. 

 

How is your external skin going to be constructed? Brick/blocks or cladding hung from the SIPS.

[SBMS]

4 hours ago, joshwk said:

Good tips,  thank you @SteamyTea & @Iceverge.

 

@SBMS: 142mm appears to be the standard manufacturing for TEK Kingspan. 172mm is also possible but requires additional work in cutting transition materials down to suit the non standard wall panels width (quoting the images shown above). Per the table, 172mm with no additional insulation is estimated to achieve a 0.16 U/value.

To achieve close as close to a 0.1 U/value as possible based on @JohnMo's suggestion, in my personal situation, I think the best route is to use 142mm but upgrade from +25mm to +75/80/90mm to achieve a 0.11 on the walls (in line with the current estimated U-Value for the roof).

There's a local company near us that do a 195mm SIPS panel that achieves 0.13 U value. Or a 140mm that achieves 0.16.

 

https://flitcraft.co.uk/injectawall/

 

Out of interest, have you modelled the difference between a 0.16 and a 0.1 u value?  Its about a £80 in additional energy costs per annum when modelled for our build (YMMV, but its worth checking).

[MikeSharp01]

On 09/06/2024 at 17:35, Iceverge said:

 

From memory the best bang for your buck was 

 

1.Airtightness

2.MVHR

3. 3g Windows 

4. Roof insulation 

5. Wall insulation 

Do you recall your methodology for working this out? Did you use just savings in energy input or this and / or other factors?

[G and J]

I took best bang for your buck to mean final heating requirement vs total build cost.

 

Sounds easy when it’s in a list like that but I’m finding out that there’s soooooo many factors.

[Iceverge]

9 hours ago, MikeSharp01 said:

Do you recall your methodology for working this out? Did you use just savings in energy input or this and / or other factors?

 

I started with a bRegs basic house. Double glazing, natural ventilation, minimum insulation and 5ACH. 

 

I assigned a realistic cash value to every upgrade of the fabric. Obviously some were either there or not like the 3g and triple glazing buy others like airtighess and insulation I stepped up gradually. 

 

I then fed them into PHPP and looked at the annual heat demand reduction for each upgrade. I costed energy at the time at 10c/kWh and noted the payback of the upgrade.

 

For example if MVHR cost €3000 but saved 1500kWh/year @ 10c/kWh = €150. €3000/€150 = 20 year payback. 

 

Anything that paid back in less than 25 years at the time got the thumbs up. 

 

It was a bit of an arbitrary target but I felt that it was at least a balanced and economic method of spending money. 

 

 

[Kelvin]

I did something similar but being a bit older the payback period matters slightly less. Therefore I considered it both in terms of the payback period and a much longer term window. MVHR self selects though as you make the building more airtight. 

[G and J]

51 minutes ago, Kelvin said:

I did something similar but being a bit older the payback period matters slightly less. Therefore I considered it both in terms of the payback period and a much longer term window. MVHR self selects though as you make the building more airtight. 

Some things like MVHR just get in due to how nice it is to live with it there, like nice sanitary wear or cladding that looks right for the location and design.  The latter two do not have a payback financially, except perhaps sale value and I don’t much care how much our executors sell for.  MVHR requires airtight therefore that’s in too.  
 

[joshwk]

On 10/06/2024 at 13:44, Iceverge said:

I'm not a massive fan of SIPs in general as they have issues with thermal bridging at junctions. 

 

If I was going to upgrade them I would probably opt for a layer of something like mineral wool to reduce this. 

 

How is your external skin going to be constructed? Brick/blocks or cladding hung from the SIPS.

 

Thanks for taking the time to respond, everyone. I'm coming back to this thread after trying to move our project along coupled with a busy work/family life.

@Iceverge The "North" and "West" building in the below will have a skin of brickwork, whereas the centre of the building are large glass openings and this area alongside the "East" building will be timber clad.

 

[joshwk]

On 10/06/2024 at 13:46, SBMS said:

Out of interest, have you modelled the difference between a 0.16 and a 0.1 u value?  Its about a £80 in additional energy costs per annum when modelled for our build (YMMV, but its worth checking).

@SBMS Thanks for taking the time to respond. I haven't! Is there a simple way to do this? Thank you.

[G and J]

6 minutes ago, joshwk said:

@SBMS Thanks for taking the time to respond. I haven't! Is there a simple way to do this? Thank you.

Yes.  Jeremy’s spreadsheet.  It’s really easy to use and informative.  

[joshwk]

On 17/06/2024 at 03:34, Iceverge said:

 

I started with a bRegs basic house. Double glazing, natural ventilation, minimum insulation and 5ACH. 

 

I assigned a realistic cash value to every upgrade of the fabric. Obviously some were either there or not like the 3g and triple glazing buy others like airtighess and insulation I stepped up gradually. 

 

I then fed them into PHPP and looked at the annual heat demand reduction for each upgrade. I costed energy at the time at 10c/kWh and noted the payback of the upgrade.

 

For example if MVHR cost €3000 but saved 1500kWh/year @ 10c/kWh = €150. €3000/€150 = 20 year payback. 

 

Anything that paid back in less than 25 years at the time got the thumbs up. 

 

It was a bit of an arbitrary target but I felt that it was at least a balanced and economic method of spending money. 

 

 

This sounds very sensible. Is there a tool you used to do this?

• I've so far committed to the fabric (walls currently 0.15 and roof 0.11 but in trying to determine if I should upgrade the walls, based on the responses so far the answer is yes)
• We have committed to MVHR (rigid ducting, Zehnder Q350 system)
• Appreciate the importance of air tightness; we will be working with the SIPs company and our respective contractors to ensure they understand the importance of this (i.e.. not puncturing holes everywhere).
• RE: thermal bridging, this is acknowledged in my quote ("Under new 2022 Part L Building Regulations, Thermal Bridging needs to be calculated and measures should be taken to reduce thermal bridging where possible. In the case of TEK (SIPS), where required, this involves the simple application of a secondary layer of insulation, (typically) positioned in the room side of the external wall or underside of TEK (SIPS) roof. This detail reduces our already excellent U Values further, whilst improving our psi values at the same time. A further benefit of this detail is an improvement in air permeability, due to the addition of a secondary taped layer of insulation TW55 (walls) - TP10 (roofs).".
  • Am I right in my understanding that the extra +25mm PIR to walls and +75mm PIR to roof speaks to this?
• I've noticed the quote from my SIPs supplier hasn't allowed for supply/fix of VCL to walls and roof. Is this something I need?

Edited November 10 by joshwk

[SBMS]

10 hours ago, joshwk said:

based on the responses so far the answer is yes

When I modelled this it was a diminishing return past 0.15 in the walls. I decided to go 200mm cavity with EPS (traditional masonry) and increasing cavity to 250mm and 300mm made very little difference to overall energy loss per annum. Nowhere near worth the additional cost of the insulation, larger (non standard) lintels etc. Your mileage may vary if using SIPS.

 

Interesting also the comment from Kingspan that their fabric makeup also isn’t SAP compliant which I think from memory was what I discovered. I love the ‘simple application of internal insulation’ - not particularly simple to skin the entire inside of your build with another layer of PIR.. Good wall u values but failing thermal bridging elements hence their recommendation of internal insulation over the timber bridging elements. It was a contributing reason to why I decided traditional construction. 
 

That said there are SIPS suppliers that don’t use timber connections and thus avoid that issue. 

[SBMS]

10 hours ago, joshwk said:

rigid ducting

Don’t do rigid if it’s a branch system. Was a right pain in our last build. Next build we will use a radial semi system. 

[SBMS]

10 hours ago, joshwk said:

I've noticed the quote from my SIPs supplier hasn't allowed for supply/fix of VCL to walls and roof. Is this something I need?

I think the internal 25mm and 75mm insulation will serve as your vapour check layer. Your initial post references this in the Kingspan notes. 

[Nick Laslett]

4 minutes ago, SBMS said:

Don’t do rigid if it’s a branch system. Was a right pain in our last build. Next build we will use a radial semi system. 

Not sure which approach is the best or the trade offs. I did my own install, so rigid intake and exhaust. Then rigid to the supply & extract manifold boxes, then radial semi rigid to all the rooms. 
 

I could not have got my head around a rigid branch system for the whole house. All of the runs from the plant room to the rest of the house are already very congested. To much pre-planning, with too many unknowns. 

[Iceverge]

We used 

Stuff like this from outside to the unit. 

 

These from the unit to the manifolds and

 

 

To the rooms. 

 

If doing it again the only thing I would change would be to include some of these. 

 

[JohnMo]

8 hours ago, Iceverge said:

We used 

Stuff like this from outside to the unit. 

 

These from the unit to the manifolds and

 

 

To the rooms. 

 

If doing it again the only thing I would change would be to include some of these. 

 

Pretty much the same materials use by me also. We used 90mm semi flexible, as you only need a single run of it to most rooms. Then all you need to add is the plenum/manifold. No room cross talk to worry about, a branch system need attenuation between rooms.

 

Semi flexible is just so easy to install, the shopping list is pretty simple also - why would you bother going rigid?

[SBMS]

12 hours ago, JohnMo said:

Pretty much the same materials use by me also. We used 90mm semi flexible, as you only need a single run of it to most rooms. Then all you need to add is the plenum/manifold. No room cross talk to worry about, a branch system need attenuation between rooms.

 

Semi flexible is just so easy to install, the shopping list is pretty simple also - why would you bother going rigid?

As someone who went rigid (ahem) with the association attenuators etc, couldn’t agree more. Ridiculous amount of extra kit and complexity for no real benefit in my opinion.