Hi Mark

Need to read the stuff properly (always a good idea!!!!). Don't want this to turn into a silly one because I believe the issue to be important. ON first proper (quick) look beyond the executive summary though the issues with these properties look scary bad.

Particularly for my lovely Bikes!

Plastering ceilings for the first time in my life at the moment so bear with me.

S.

Well I think you should give up.

You challenge a post I made and then cite a reference that I suspect you didn't read in depth which actually supports the hardly confrontational statement contained in my post that it's the construction techniques "wot dun it", i.e. sabotage insulations best attempts to insulate.

But have a nice day, anyway.

I'm coming over to the UK tomorrow...and bringing a nice thick polar fleece to keep me warm! Should I bring my foil blanket too?

As a self builder, I salute your efforts, especially plastering a ceiling.

First of all my apologies for boring the arse of anyone not really wanting to read all of this

Okay marktime, since you persist – lets pick this apart. The post to which I originally responded:
……………………………………………
CommentAuthor marktime CommentTime2 days ago edited
quote
I can't speak for others but I've laid glass wool in lofts under a ventilated roof and packed it in well to the rafter height, (or better said, truss tie height), layered the next layer at right angles etc. and after 300 mm of wool I reckon I must have got close to the calculated u-value. The gale blowing over the top, provided it isn't allowed to lift the wool, is factored in as a boundary condition.
If you use PUR or other rigid board, I reckon you could do even better. However, I agree that where poorly supervised installation takes place, you'd be lucky to get anywhere near the design figure and having had a shufty at a Barret estate during building, that's probably the norm.

Long term? What would cause settling of glass wool in a loft? Dust deposits? And by how much would that change the thermal conductivity? I don't have a real world answer to that but I would think, not significantly.
…………………………………………………

This is what I wanted to say, but didn’t at the time
This is purely speculative on your part. Words like ‘I reckon’ for example. Based on what exactly? Have you measured the u-value- No? Did you use a thermal imaging camera? No? Have you read around the subject? Do you have a commercial bias/interest?
You also demonstrate here that you know little regarding the way air entrained insulation works – which is largely by entraining air. If you squash the air out of it..or.. it settles for some mysterious reason then you will indeed change the thermal conductivity. Please feel free to fault my logic here.
Instead of responding with such an aggressive post I said:
…………………………………………………

CommentAuthor Mike George CommentTime2 days ago edited edit
quote
Regarding u-values v real life performance in general, have a look at this http://www.lmu.ac.uk/as/cebe/projects/stamford/index.htm

Regarding settlement of some compressive insulation it can and does occur. I have seen it on site many times and I have also seen research which highlights it as a problem. It can and does affect the conductivity significantly. I'll try and remember where the report on this is.

Marktime, can you point me to a citation which discusses the boundary conditions you mention that are factored into the u-value of mineral wool. The only hot plate test I am aware of which factors in any level of air movement is one recently conducted by the NPL in their investigation of multifoil http://publications.npl.co.uk/dbtw-wpd/exec/dbtwpub.dll?&QB0=AND&QF0=ID&QI0=%204835%20&TN=NPLPUBS&RF=WFullRecordDetails&DL=0&RL=0&NP=4&AC=QBE_QUERY certainly not the norm
………………………………………………

Your response was rather selective
CommentAuthormarktime CommentTime1 day ago edited
quote
@ MikeG

Your link: "The Stamford Brook PII project is a unique record of the achievements, successes, failures, problems and solutions that can occur during the implementation of an advanced energy standard on a large scale housing development."

Problems encountered: "Party Wall Thermal Bypass/Airtightness Issues/The Gap between Designed Performance and As-Built Performance".

Nothing about settling that I could see but you could point me in the right direction if I have missed something. What would make my glass wool settle to such a point that it would impact the thermal conductivity? You say you have seen it on site. In a roof or in a cavity?

The coefficient of thermal transmission for a simple enclosure comprising two parallel planes, (air -air) is given by: 1/k = 1/hi + L/thermal conductivity (lambda) + 1/he

where hi and he represent the heat transfer at the internal and external faces exposed to air respectively i.e the boundary conditions. They have been found empirically and are given in tables for heat flows horizontally and vertically, including in the latter, for flow up or down.

Thus the fabric u-value takes these boundary conditions into consideration in their calculation. We could of course go into the formula for sandwich construction but I think you get the drift.
…………………………………………………………………………………………

I note you concentrated the post above on the Stanford Link. Perhaps I should have said this
Thermal bypass = something not allowed for in u-value calculation.
Airtightness issues = something not allowed for in u-value calculation
Which is why I posted the link, and which was AND I QUOTE “Regarding u-values v real life performance in general”
I did not say that the Stanford link contained any information about settlement
I said [in a new paragraph!!] AND I QUOTE “I'll try and remember where the report on this is.”
You either didn’t read this or you ignored it :(
You also totally ignored the relevance of the NPL report :( though it is heavy reading :)
The relevance of the NPL report here is
Air infiltration = something not allowed for in u-value calculation.
Solar gain = something not allowed for in u-value calculation.
Heat transfer via ratiation = something not allowed for in u-value calculation
Lots more stuff worth reading
Again instead of responding with such an aggressive post I said
…………………………………………………………………………………………

CommentAuthor Mike George CommentTime1 day ago edited
edit quote
Marktime,

The stamford link concludes I believe that there is a difference between predicted steady state u-values and real life performance.

There is nothing about settling in the Stamford report - that is in a different publication, the one I am going to try to find. Hopefully someone else here has seen it and can point us to it. I do not know what makes the insulation settle for sure - gravity? But I do know that it does - I have seen it in lofts and I have also seen it in the cheek walls of dormer walls. The dormer walls were full of flies by the way. Other lodgers in the loft included many wasp nests and at least a pair of bats - who knows if they have something to do with it.

Presumably the boundary conditions you refer to are what I would call Ri and Ro: Resistance of the surface inside and outside respectively. In my opinion these do not accurately take account of the passage of air in a cold loft, or any other such construction where air can be funneled through part or all of a construction, such as above insulation fitted between rafters; or in a ventilated cavity. I believe it is for this reason that NPL introduced a new methodology to their hot box testing. If you read their report you will see that this is indicated as being the case. Indeed there is an acknowledgement that u-values do not take account of either air infiltration or solar gain.
…………………………………………………………………………………………

and
CommentAuthorMike George CommentTime1 day ago
edit quote
Here's someone else who doesn't think much of hot box u-values http://www.limetechnology.co.uk/upload/documents/1201787723_the_thermal_performance_of_tradical_hemcrete.pdf. Claims seem to be accepted by Building Control by the way - LABC certification here: http://www.limetechnology.co.uk/upload/documents/1210172463_labccertificateweb.pdf. You need rather a lot of hemcrete to obtain the current breg requirement by the method set out in BR443 Ie. hot plate derived conductivity. An example upgrade to a 600mm stone wall requires approx 1500mm hemcrete at a hot box tested k-value of 0.679W/mK This gives a u-value of approximately 0.35W/m2K. Conclusion: Either all jobs using Hemcrete leave a limited floorspace or Building Control are ignoring the guidance set out in BR443.
…………………………………………………………………………………………

Once again you chose to fall back on the Stamford report and ignored the very relevant references above. Here’s a part of a report I was asked to prepare which précis the pertinent issues:
The contentions are [and I mostly agree with them]

• A hot plate test is unable to accurately measure the real life performance of Hemcrete due to the nature of the test which results in samples being thoroughly dried out prior to steady state conditions being reached.
• Since buildings have various levels of Relative Humidity, the fabric is never in a completely dry state. It is a contention that a small amount of moisture present in the Hemcrete is important and that it results in a phase change property. ‘this is where the change from liquid to vapour, or back, stores or releases energy’
• Under dynamic conditions, only a portion of heat is conducted through a wall from one side to the other. Walls very rarely reach steady state conditions.
• Research from Switzerland [1] uses ‘lab test figures’ input into WUFI software ,[2] which simulates time taken for materials to reach steady state. Of relevance is the comparison between Hemcrete and mineral wool 72hrs [Hemcrete]:12hrs [mineral wool] Average heat loss of Hemcrete is equated to 0.11 W/m2K [theoretical u-value 0.29 W/m2K].
• Further research by Evrard [3] states ‘ thermal diffusivity is very low for hemcrete compared to other materials, which means it will take longer to change temperature’
• A significant amount of heat is lost from buildings through air leakage. Tests on Hemcrete have achieved air leakage figures of around 2m3/m2@ 50Pa. This is 5 times better than the current Building Regulation minimum requirement of 10m3/m2@ 50Pa.
• The thermal effusivity of Hemcrete is low. This is ‘why it feels warm to the touch’ ‘This means you feel warm even though the heating is turned down, saving hundreds of pounds per year in heating costs’.
• Hemcrete is normally used as an infill between timber frame constructions. Since the thermal capacity, diffusivity and effusivity are similar to wood; timber framing does not behave as a thermal bridge in the same way as it does in other forms of construction.

[1]Herde, A and Evrard, A Bioclimatic envelopes made of hemp and lime. Proceedings from CISBAT Conference 2005
[2]Available at: 2 http://www.eere.energy.gov/buildings/tools_directory/software.cfm/ID=362/pagename=alpha_list
[3]Evrard, A Transient hygrothermal behaviour of lime-hemp materials PhD in Applied Science, Universite catholique de Louvain 200

Given this and my u-value calculations in my last post above, some might conclude that even Building Control themselves have recognised the limitations of the u-value

:( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :( :(

Your next post AGAIN fell back on the Stamford link [rather old hat by this stage] :)

CommentAuthormarktime CommentTime1 day ago
quote
This is why the tests demonstrated that these developers were not delivering what was written on the tin and confirms what I posted for goodness sake.

Quote <In addition to losses attributed to ventilated party walls ....

35. Other possible explanations for the remaining unaccounted for difference between measured and predicted heat loss are as follows:

a) The U-values of the various construction elements may not have been achieved in practice due to quality or technical problems during construction.

b) The values for linear thermal bridging obtained by thermal modelling and used in parametric SAP may have been unrealistic due to either quality or technical issues during construction.

c) The air permeability of the test dwellings may have increased significantly during the test due to the high thermal stresses imposed on the building fabric, joints and junctions.

d) There may be other, as yet unidentified heat loss mechanisms, which are not included in the parametric SAP model. > End quote.

The next time you cross a bridge you want to hope that the construction has met the design criteria and that no waivers on quality have been allowed so that at least you, like the proverbial chicken, will get to the other side.

The science is out there, why do you need to second guess it?
…………………………………………………………………………………………

LOL at this one especially your very first sentence

Quote <In addition to losses attributed to ventilated party walls ....

Reminder:
Air infiltration = something not allowed for in u-value calculation.

Also:
Linear thermal bridging = something not allowed for in u-value calculation.
Non Linear thermal bridging = something not allowed for in u-value calculation.
As yet unidentified heat loss mechanisms = something not allowed for in u-value calculation.
SAP = LOL all around - I won’t go there

And Yes, of course quality of wormanship affects performance as may 'technical problems' [whatever they were]

…………………………………………………………………………………………

Some examples to think about.

How accurate is the u-vale for a South facing window in the middle of winter on a nice sunny day? Is it the same as, or different from an identical North facing window? Now is the heat loss through each the same?

How accurate is the u-value for a solid floor given that the temperature difference to the ground is nowhere near as predictable as the temperature difference to the outside, say, through a wall? Is the heat lost at the perimeter the same as that lost from the centre of the floor? Does this make a difference as floors get bigger? Say 2000m2?

…………………………………………………………………………………………

Finally I have been second guessing u-values for some years now. Science is there to be questioned for goodness sake :)

One final link from the AECB regarding the complexity of what they call “Real u-values” [and how they differ from the recognised ones] http://www.aecb.net/forum/index.php?topic=602.0

…………………………………………………………………………………………

Sorry about the length of the post but I had to get my goat back. [shouldn’t have let it stray across the bridge ]

Now I really would like to retire from this one for a while. I have more constructive things to do

You have a nice day too


Once again, my apologies for boring the arse of anyone not really wanting to read all of this

Not as long as mine but very to the point

From the BRE

The results of the present work, together with findings from previous research,
show that the level of agreement between measured and calculated U-values
varies widely. Results also suggest that as-built U-values of walls are typically
around 20% (or 0.1 W/m²K) higher than U-values predicted by BS EN ISO 6946
(although the degree of difference depends very much upon the type of
construction being examined). It cannot at this stage be established whether this
difference arises from workmanship, ageing processes, weathering or defective
insulation materials but the level of difference appears to depend upon the type of
construction used.


http://projects.bre.co.uk/uvalues/U-values.pdf

S.

And add this little nugget (p.28):

"It cannot at this stage be established whether this
difference arises from workmanship, ageing processes, weathering or defective
insulation materials but the level of difference appears to depend upon the type of
construction used."