Minimum Reasonable wall width for good U values?

[joe90]

2 hours ago, Oz07 said:

@joe90 what's your u value? Are you at around 425mm thickness total?

I have to confess I don’t know!!! I didn’t use any calculations but others here may be able to help with that. My wall build up is 100mm brick, 200mm rockwall batts, 100mm 7n block, render then plaster, so yes close on to 425mm.

[SteamyTea]

Vacuum panels, 0.004 W/m².K

Can't get better.

Except cellulose is probably more practical in s building.

[Ed Davies]

20 hours ago, Temp said:

A combination of Rockwool between and PIR outside would be reasonable. 

 

Perhaps Rockwool between and PIR inside would be better from the giving-any-water-a-path-to-evaporate point of view.

[puntloos]

3 hours ago, scottishjohn said:

why is wall thickness the factor you seem to be concentrating on ?

 

Frankly I've just been very scattershot with my 'what to focus on'.. 

3 hours ago, scottishjohn said:

 small plot and room sizes ?

But yes, this is a factor, I was struggling to achieve a certain size of room, seating area, kitchen next to it.. and the distances started to add up so to speak. Of course say 30cm ((440-300)*2) isn't going to make a night-and-day but I was basically able to achieve all my target distances between things if I could.. just.... get.. 40cm.. somewhere..

 

But yes, 250sqm house on a 420sqm plot. Fairly cramped.

 

3 hours ago, scottishjohn said:

 you can get passiv levels on any type of build

Is this a self build project ?

Yes

3 hours ago, scottishjohn said:

how much of it do you intend to do yourself

I'll make tea occasionally.

(and am deeply involved with design/architecture, probably project management bits too)

 

But I will not be doing anything 'physical' other than perhaps some minor electrics.

 

3 hours ago, scottishjohn said:

 do you have plans ? 

laid by the best mice and men.

 

But also: 

 

Note that while the basics haven't dramatically changed we have recently received a major update from the architect, I will update the topic soon.. probably today

 

3 hours ago, scottishjohn said:

all the above will guide you intype of construction you choose probably

If you have plans --then send  toa TF company 

an ICF company  .SIPS company 

 

Good idea.. when is a good moment though? Now-ish? Or if/when we have the external shell nailed down & submitted for planning?

 

3 hours ago, scottishjohn said:

your exterior finish  you desire will also guide you possibly

type of ground you are on 

 

Clay. We're just north of London

 

3 hours ago, scottishjohn said:

can it be slab ,or strip founds ,or piles --

Indeed, I haven't at all made my mind up on this.. 

 

3 hours ago, scottishjohn said:

the choice is very complicated until you tie down some real site conditions and choices  and your desires first  

and get some quotes -?

 

Yup probably true

[puntloos]

4 hours ago, Jeremy Harris said:

 

 

There are two considerations, really, as U value is only part of it, decrement delay is at least, if not more, important than U value, as that's largely what determines how comfortable the house will be, in terms of not suffering from rapid temperature fluctuations. 

 

Decrement delay is described here pretty well: http://www.greenspec.co.uk/building-design/decrement-delay/ and, although related to insulation performance, it's equally closely related to the heat capacity of the fabric. 

 

One problem with making very thin walls, apart from decreasing the U value to a sensible level, is in increasing the decrement delay to a sensible level, too.  In general, lightweight foam materials tend to have a pretty short decrement delay, as they have a very low heat capacity, whereas some other insulation materials, like some fibre materials, wood fibre or blown cellulose tend to have a reasonably long decrement delay.  It's quite possible to combine material layers so as to achieve both a low U value and high decrement delay, as another option. 

 

Clearly I want all-the-things at none of the thickness.

 

4 hours ago, Jeremy Harris said:

Alternatively, you can accept having a short decrement delay and just fit a heating and cooling system with a very rapid response time, so it can cool the house as fast as it heats up, and vice versa.

 

Do you mean more active cooling (e.g. aircon over ASHP(alone))?

 

4 hours ago, Jeremy Harris said:

Location and orientation also play a big part in this decision.  If the fabric of the house isn't exposed to bright sunshine for long periods during the day, then decrement delay becomes less of a concern.  It's primarily an over-heating issue, as in bright sunshine it's pretty easy for an external surface to get to well over 40°C (I've seen outside surface temperatures of around 50°C at times here).  That creates a pretty high temperature differential across any wall or roof, and will tend to drive heat more rapidly through the fabric of the house unless the heat capacity of that fabric can absorb some of that heat and so slow down the rate of heat transfer.  If it can slow it down enough for the heat not to reach the inside face, before the sun has moved away, then generally that's good enough, as as soon as the outer temperature drops heat will start to flow back out.

 

Yup seen this too, and indeed my inside garden face is exposed to full sun between 10 and 4 (we're SE facing). 

 

So I guess I have to strike some 'fair balance'. Would you e.g. do the layer approach with one PIR and one cellulose layer stapled together or somesuch?

[Jeremy Harris]

39 minutes ago, puntloos said:

 

Do you mean more active cooling (e.g. aircon over ASHP(alone))?

 

Yes, probably the fastest way to cool/heat a house is to use some form of cooled/heated air system.  With a low decrement delay the system needs to be able to respond fairly quickly, in order to keep the temperature reasonably stable.  Clearly, using a fair bit of power to rapidly control temperature isn't really in keeping with the PassivHaus concept, but if the cooling power is being self-generated by a PV system than I can't see that it's really an issue.

 

 

39 minutes ago, puntloos said:

So I guess I have to strike some 'fair balance'. Would you e.g. do the layer approach with one PIR and one cellulose layer stapled together or somesuch?

 

If going for layers, then something like ICF makes sense.  That has a very short decrement delay insulation layer (often EPS) coupled with a moderately high heat capacity structural core.  The combination of the two produces a structure with a decent U value as well as a reasonably long decrement delay.

[puntloos]

48 minutes ago, Jeremy Harris said:

 

Yes, probably the fastest way to cool/heat a house is to use some form of cooled/heated air system.  With a low decrement delay the system needs to be able to respond fairly quickly, in order to keep the temperature reasonably stable.  Clearly, using a fair bit of power to rapidly control temperature isn't really in keeping with the PassivHaus concept, but if the cooling power is being self-generated by a PV system than I can't see that it's really an issue.

 

 

 

If going for layers, then something like ICF makes sense.  That has a very short decrement delay insulation layer (often EPS) coupled with a moderately high heat capacity structural core.  The combination of the two produces a structure with a decent U value as well as a reasonably long decrement delay.

 

Would ICF work well together with timberframe? ICF has great audio dampening qualities too, so I'd be very very tempted to pick ICF as one of the main components...

 

[Jeremy Harris]

29 minutes ago, puntloos said:

 

Would ICF work well together with timberframe? ICF has great audio dampening qualities too, so I'd be very very tempted to pick ICF as one of the main components...

 

 

 

If going for ICF, then it doesn't really make sense to also use timber frame as a structural element, as the ICF will do both.

[puntloos]

1 minute ago, Jeremy Harris said:

 

 

If going for ICF, then it doesn't really make sense to also use timber frame as a structural element, as the ICF will do both.

 

Hmm, while I like insulation (both audio and thermal) to be good, I prefer the off-site construction, accuracy and speed of standing up the house when onsite. I guess the audio is secondary... 

[Dan F]

22 hours ago, puntloos said:

 

Sorry, I appreciate I'm a newbie here, but combining the above 2 statements, can I say safely-ish that:

 

- With MBC's 0.14,

- making sure the render is on carrier board, vent cavity behind (I assume this is not part of the default MBC necessarily, so needs to be separately checked)

-> and therefore staying under 300mm

 

I would be 'okay' - mortgageable (perhaps not the industry-best rates but nothing crazy), warranty, insurable etc?

 

 

IMO yes, no issue at all.  We're doing the same, but with twin-wall 300mm system and haven't had an issue with mortgage or warranty. 

 

- Timber frame without block skin does limit lenders slightly.

- Timber frame without BBA or BOPAS accreditation limits lenders further.

- But, there are still a fair number of lenders, including Ecology (which a number of people on buildhub have used). We went with someone else in the end as they could lend us more.

 

Of course, it'll be wise to check with a mortgage broker before signing off on anything.

 

 

[Temp]

12 hours ago, Ed Davies said:

 

Perhaps Rockwool between and PIR inside would be better from the giving-any-water-a-path-to-evaporate point of view.

 

Thats ok but If the structural timbers are on the cold side of insulation they need to be well ventilated. 

[DavidFrancis]

I want to see someone build a bungalow using sandwich panels (Kingspan, Tata etc) on a lightweight steel frame. Looks like you can get a U value of 0.15 from a panel of a bit over 120mm thickness, if I'm reading the Kingspan details correctly.

[scottishjohn]

lightweight don,t mean cheap!!

[Jeremy Harris]

15 minutes ago, DavidFrancis said:

I want to see someone build a bungalow using sandwich panels (Kingspan, Tata etc) on a lightweight steel frame. Looks like you can get a U value of 0.15 from a panel of a bit over 120mm thickness, if I'm reading the Kingspan details correctly.

 

 

Why bother with the steel frame, though?  SIPs are structural, at least up to two or three storeys, so there's no need for any steel.

[DavidFrancis]

The ones I'm thinking of are the metal-clad panels, as used in large commercial/industrial buildings. They're always used with a heavyweight frame, but with a true bungalow I imagine you could get away with something very lightweight. Some of the panels have quite respectable finishes both inside and outside, so you'd only have to clad them if you didn't like the aesthetics or wanted to have a service void for cables/pipework etc. 

[kxi]

4 hours ago, DavidFrancis said:

I want to see someone build a bungalow using sandwich panels (Kingspan, Tata etc) on a lightweight steel frame. Looks like you can get a U value of 0.15 from a panel of a bit over 120mm thickness, if I'm reading the Kingspan details correctly.

 

We are using the KS1000RW Kingspan Quad-Core panels as the outer shell (walls and roof) in our build.

 

A single component that provides insulation, water-tight shell, and a reasonable degree of airtightness (though this will need some additional taping). It is on a heavy steel frame. We are using the 150mm version giving 0.12 u-value.

 

Thermal breaks in the column floor junctions handled with armatherm FFR structural thermal breaks. With a lightweight steel frame you probably wouldn't have such an issue since it could all sit on an insulated slab.

 

The internal face of the panels we couldn't stomach so will be covered to give a deep service void.

 

There are one or two examples of it used on domestic buildings, but mainly modern barn conversions, i.e. they were a panelled shed that someone turned into a house.

 

@Simon Brooke has used the KS1000RW for the roof of their dairy conversion, which seemed to be successful.

 

Edited January 13, 2020 by kxi
added link to other relevant post

[kxi]

N.b. the kingspan panels have to be supported on purlins or rails, which may then need further support, so the 150mm doesn't give you the full required thickness.

 

Though you can leave this open to some extent if you don't mind the internal face.

 

The panels provide little sound insulation (or decrement delay...), so you may want to cover them internally for other reasons.

[DavidFrancis]

Thanks for that @kxi. Hadn't seen the insulated roofing sheet thread. And didn't realise the sound insulation was poor. Have you posted more details about your build somewhere? Looked at your recent posts, but only found something about your floor.

[Moonshine]

Can i throw my hat / question in the ring, as i am also thinking about exterior wall constructions for a proposed build which is going to be rendered mostly and timber clad in some areas.

 

I have had quotes from timber framer for just a single stud exterior wall, which i don't really like the idea of, and i'm not sure i really want to be doing a full cavity block masonry wall with PIR insulation in the middle (300 - 325mm)

 

However i have been thinking the hybrid construction of timber with a masonry exterior wall as below (out site to inside);

• Render / timber cladding
• 100mm lightweight / dense block
• 50mm cavity
• 50mm PIR boards
• 9mm OSB (needed?)
• 90-100mm timber stud filled with mineral / glass wool (e.g. Superglass 032)
• internal vapour control membrane
• 25mm battens for services void
• 12.5mm plasterboard.

What do you think that the U-value of this would be?

 

Also the construction could ditch the PIR insulation boards, and could have thicker insulation filled studs, in this case what do you think the U value would be and also how close could this go to the exterior blockworks.

 

Can anyone share their thoughts on this and in general of a timber with a masonry exterior wall, and what is the typical way to do it

[Dreadnaught]

@Moonshine, I am no expert, but how would any moisture in the timber frame exit, on one side is a VCL and on the other PIR, both vapour impermeable?

 

Rule-of-thumb I learnt was to have increasing vapour permeability of materials from the inside to the outside (or to the ventilation void in your case).

[ProDave]

4 minutes ago, Moonshine said:

Can i throw my hat / question in the ring, as i am also thinking about exterior wall constructions for a proposed build which is going to be rendered mostly and timber clad in some areas.

 

I have had quotes from timber framer for just a single stud exterior wall, which i don't really like the idea of, and i'm not sure i really want to be doing a full cavity block masonry wall with PIR insulation in the middle (300 - 325mm)

 

However i have been thinking the hybrid construction of timber with a masonry exterior wall as below (out site to inside);

• Render / timber cladding
• 100mm lightweight / dense block
• 50mm cavity
• 50mm PIR boards
• 9mm OSB (needed?)
• 90-100mm timber stud filled with mineral / glass wool (e.g. Superglass 032)
• internal vapour control membrane
• 25mm battens for services void
• 12.5mm plasterboard.

What do you think that the U-value of this would be?

 

Also the construction could ditch the PIR insulation boards, and could have thicker insulation filled studs, in this case what do you think the U value would be and also how close could this go to the exterior blockworks.

 

Can anyone share their thoughts on this and in general of a timber with a masonry exterior wall, and what is the typical way to do it

That is almost the standard timber frame method up here.

 

Except use a 150mm timber frame and no insulation in the cavity.  Usually you also over sheet with PIR on the inside of the frame.

[Carrerahill]

10 minutes ago, Moonshine said:

• 50mm cavity
• 50mm PIR boards
• 9mm OSB (needed?)

I am not too keen on that, over your OSB you would have a membrane like Protect TF200 or Dupont Reflex and if you stuff PIR board of the same thickness as the depth of the cavity then you have removed the ventilation over the face of the membrane and this would as I see it create a perfect moisture trap at the OSB layer. 

 

We went for render, block, 50mm cavity, breather reflective membrane over 11mm OSB (I wouldn't use anything less than 11 for the cost of it - even 11mm is fairly flimsy), onto 6x2 studs, 100mm of QuinnTherm PIR in-between studs, then plaster and skim - we could have gone for the same makeup but a layer of QuinnTherm over the studs too but we didn't need it, we did however do that on the ceiling. We could have also use 125mm QuinnTherm but that would have reduced our service void too much and to be honest, as it stands the rooms take only the warmth of Henry the hoover running to take the chill of it! 

 

 

[ADLIan]

I wouldn't use 032 mineral wool between timber studs. Will give little to U-value but at much more cost, 035 product will be OK.

[tonyshouse]

Decide on the U-value first then design to make it work, less than 0.1 in my book

 

good idea to do a thermal model too. 

[Moonshine]

On 16/01/2020 at 11:47, ProDave said:

That is almost the standard timber frame method up here.

 

Except use a 150mm timber frame and no insulation in the cavity.  Usually you also over sheet with PIR on the inside of the frame.

 

Do you know what a typical U value of that build up would be with 50mm pir on the inside of the timber.

 

How do you fix the pir to the timber frame? I would have thought any screws for battens will need to go all the way through the pir and create a cold bridge?