[Strawbale] Heat loss culprits (was re: Decrement factor of SB walls) + "True tested " R-value

[RT]

On Wed, 07 Dec 2011, in GSBN Digest, Vol 9, Issue 3

carolatkn at aol... wrote:
Re: Decrement factor of SB walls.

> basing heat loss calculations on conduction (r values or u-values) is  
> wholly inadequate because most heat loss is lost through convection
[snip]
(The entirety of the above message and the thread from which it came may  
be viewed at the accessible-to-the-public-at-large GSBN List archives at:   
http://sustainablesources.com/pipermail/gsbn/ )

I find the above assertion to be curious and contrary to my experience.

Heat loss as a result of convection within an envelope component would be  
significant only if the insulation were poorly fitted and/or of  
insufficient density and/or the air barrier strategy is ineffective or in  
the case of insulating window glass, if the space between the panes is too  
wide.

When a building is insulated to super-insulation levels so as to minimise  
heat loss via conduction and the requisite effective air barrier strategy  
is implemented, then the biggest potential heat loss factor for the  
building becomes that which is the result of the ventilation air changes  
necessary to ensure a steady supply of fresh air to the building's  
occupants. I use the term "potential" because that heat loss is typically  
minimised with the use of heat recovery (minimum 75% efficiency per 1980's  
Canadian R-2000 performance standard, don't know nor care what the  
PassivHaus standard is because I think that it is very confused) on the  
exhaust air stream.

and Bohdan Dorniak <bdco at adam....au> wrote:
>
> Are there any tests on the R-value of rendered strawbales?
> I?ve never been able to find a true tested value.

I've never had a problem with the thermal resistivity values for straw  
that were provided by Joe McCabe in his Master's thesis on the subject so  
many years ago.

They are in perfect agreement with the values predicted by the following  
correlation that is used to estimate thermal conductivity (k) of materials  
in the absence of test data:


	k = 0.281 * Exp(0.0268 * density ) (expressed in Btu* in /h* ft^2 * degF,  
density in lb/ft^3)

(courtesy of Douglas M. Burch, Heat Transfer Group, Building Environment  
Division, (US) National Institute of Standards and Technology (NIST), July  
14, 1995)

OTOH, the thermal resistance of SB wall assemblies is understandably all  
over the board due to the variability of bales, building methods, test  
methods, skill level of the builders, locations of the test subjects etc.

I view the thermal resistance values yielded by tests on SB wall  
assemblies as being like what one might get if one took a brush loaded  
with paint and then flung that paint at a wall.

The resultant paint splatter would represent the data points.

And since people like to be able to cite a discrete "true tested R-value"  
it would be a matter of picking a spot somewhere in the midst of where the  
the paint splatter is most dense.

As ridiculous as the above may sound, the reality of the derivation of  
many of the R-values for SB that are in widespread use today is arguably  
moreso.

Pretty much the only "true tested R-value" that one is going to get (IMO)  
is if one builds a SB house while maintaining as much uniformity  
throughout the process as one is capable of doing, documenting every  
detail and variance throughout the course of construction and then after a  
year or three (depending upon locale) after the materials have reached  
something resembling equilibrium, use a reasonably accurate device to take  
readings of interior and exterior surface temperatures of the wall at 3:00  
am and then use those readings and the detailing info to do the arithmetic  
necessary to yield the R-value for _that_  particular construction and  
quite likely not representative of 75% (a WAG pulled out of my hat) or  
more of other existing SB constructions.


-- 
=== * ===
Rob Tom
Kanata, Ontario, Canada

< A r c h i L o g i c  at  Y a h o o  dot  c a  >
(manually winnow the chaff from my edress if you hit "reply")