[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: GSBN:CalifoRnian values (was re: from the Crest list)


The R-30 is NOT based on the CEC/ATI tests.  We all recognized they were
flawed (though not quite so incompetently as you seem to remember). The R-30
is based on the best straw bale wall test (watched over by David Einsenberg
and Jeff Christian) at Oak Ridge National Lab.  The CEC/ATI test yeilded a
value in the neighborhood of R-20.  And yes, it was due to moisture.  But
no, it was not due to the incompetence of the ATI staff - but rather to my
own.  To test a wall in the type of chambe that ATI has, we had to build the
wall in place - in the hole in the face of the chamber where they typically
insert the windows that they test as their bread and butter.  Obviously,
building a straw bale wall into a hole of an already defined set of
dimensions means that you cannot get the same tight structure top-to-bottom
that you would if you were building a wall for a house and then placing a
ceiling/roof structure on top of it.  Therefore, we thought we were doing
the best we could when we stuffed the space above the wall (and below the
top of the chamber opening) with straw - as tightly as we could.  Turns out
that we did not get it tight enough to represent real world conditions.
Therefore, the water that we had the ATI staff spray onto the stucco finish
for three days (to prevent large cracks), migrated into the less dense space
above the bales - the space that we stuffed relatively less tightly.  It is
unlikely that three weeks or even three months of drying out would have
removed all (most?) of that moisture.  Yet, the lab could not afford to
leave the test chambers (we used both of the chambers that they had) idle
waiting for the walls to get as dry as they would in your house wall.

I am sorry if I disappointed people by not supervising a procedure that
resulted in straw bale wall test R-values that directly correlated with the
straw bale test results.  But at least it provided enough stimulus for Oak
Ridge to run another test, and to take the quality control serious enough to
invite someone from the California Energy Commission and David Einsenberg to
help supervise the construction of the wall.  Their procedure was
appropriate and I doubt that anyone else, running a high-quality test on a
bale wall, would come up with thermal performance values significantly
different from Oak Ridge's second test (~1997).

Joe's test was probably of equal quality.  The fact that the values diagree
should not surprise anyone.  One test was of a bale, the other was of a
straw bale wall system.  It is time to get over the fact that the R-value
results were different.  Besides the fact that they SHOULD be different
since different things were being tested, the difference between R-30 and
R-45 just doesn't matter.  A fifteen point difference when one of the values
is around R-10 DOES matter; but an equal difference at the high R-values we
are discussing doesn't much matter.  Since R-values are the reciprical of
U-factors (the amount of heat that flows through a SF of your system in one
hour given a one degree temperature difference), the difference between R-10
and R-25 means that one lets 2.5 times more heat through than the other
(0.10 Btu/hr/SF/F and 0.040 Btu/hr/SF/F).  The difference between R-30 and
R-45 means the difference between 0.033 Btu/hr/SF/F and 0.022 Btu/hr/SF/F.
Repeat after me, "It just doesn't matter!  It just doesn't matter."  I
suggest we move on to other, more meaningful disagreements.


-----Original Message-----
From: GSBN [mailto:GSBN@...]On Behalf Of Rob Tom
Sent: Thursday, May 08, 2003 9:43 AM
To: GSBN@...
Cc: sb-r-us@...
Subject: GSBN:CalifoRnian values (was re: from the Crest list)

Rocky et al;

I don't recall the specifics of the CEC tests at the ATI labs but I do
recall that at the time,  I was shaking my head at the manner in which the
test panels were prepared before being presented to the ATI people for

I have vague memories of things like foam insulation board being jammed into
cracks at the perimeter by the ATI people *during* the tests, presumably to
reduce the haemorrhaging so that their results wouldn't make them look
incompetent as a testing facility. (ie the GIGO syndrome)

I alos have vague memories of meaningless moisture content readings being
taken (I think that it was of bales that weren't even inside of the test
panel), while ignoring the  wet condition of the bales in the actual test
panel which hadn't been allowed to dry out properly after stucco

It was therefore no surprise that the CEC/ATI tests yielded disappointingly
low R-values for straw but even more disappointing was the fact that the CEC
chose to accept the test results as being representative of the insulating
properties of straw bales.

There is no doubt in my mind as to the accuracy of the thermal resitivity
values that JoE McCabe's testing provided, as they agree perfectly with the
theoretical value that would be yielded by the equation used to estimate
thermal conductivity (k) in the absence of data:

                     k = 0.281*Exp((0.0268*density)

(note that the above equation will yield a value that is dead-nuts in the
middle of the two values that JoE provided in which he accounted for fibre
orientation WRT heat flow... which also makes sense because the equation
does not have a modifier for bale orientation)

The Sandia lab resistivity test results (although of bales of lower density
than that used by JoE in his testing) would seem to further confirm that the
CEC/ATI thermal resistance values are seriously flawed.

ie The ASHRAE values for the thermal resistivity of materials (presumably
derived from testing of the materials under lab conditions) is regularly
used to estimate the nominal thermal resistance of building assemblies for
permit purposes etc. so why should there be a double standard for straw
bales ?

Clearly, it was the process (ie of the  panel preparation, not of the ATI
testing) which was flawed and it would seem to make sense that the tests
should be re-done (making the necessary changes each time of course... they
say that one of the signs of insanity is repeating the same actions over and
over again expecting a different result each time) until the thermal
resistance of the test walls approaches the values that are predicted by
JoE's resistivity values.

Aside from the issue of respectability for strawbale housing (ie why would
anyone go to the trouble of doing plastered SB walls to achieve a mere R-30
when better than R-30 can be easily achieved in a thinner and
readily-accepted-by-Conventionaldumb stick-framed wall ?), one would think
that there would be more pragmatic consequences, such as in the reduction of
the sizing of Code-required auxiliary heating (and cooling ?) systems,
possibly even eliminating them in some instances, ultimately saving the
homeowner's financial resources and perhaps more importantly, the planet's

--- * ---
Robert W. Tom
Kanata, Ontario, Canada

please visit:  http://www.theHungerSite.com daily

>Date: 6 May 2003 23:58:40 -0500
>From: stoneandstraw@...
>Subject: RE: From the Crest list
>In California, the accepted R-value (for compliance with the state building
>standards) is R-30.  That was based on a the results of a number of tests.
>One that I managed while at the California Energy Commission (CEC), two at
>Oak Ridge National Lab, the tests that Joe McCabe did, and testing that
>Sandia Labs did.  Every test had issues, and didn't all test the same
>For example, McCabe and Sandia tested straw bales, while the other tests
>were of straw bale walls.

Add photos to your messages with MSN 8. Get 2 months FREE*.