[Equest-users] Questions - Foam Board R-value Variations by Temp

[Aaron Powers via Equest-users]

Chris,


DOE2/eQuest doesn't model temperature-dependent R-Values.  The technical
reason is that DOE2 uses the response factor method for 1-D heat
conduction.  This is a Laplace Transform method which relies on the theory
of residues to numerically solve the heat equation.  This method greatly
speeds up simulation times compared to direct finite-difference methods,
but it relies on the underlying equations being time-invariant and linear.
The problem with a temperature dependent R-Value is that it makes the heat
equation non-linear.  Almost any practical detailed simulation of a whole
building is going to use some form of the response factor method for
conduction, so this isn't just a limitation to DOE2/eQuest.  So to add this
capability to the engine would require a fundamental change to the
underlying mathematics with a huge penalty in computation time.  The best
solution might be to aim for some annual average value based on climate
data suggested by others.


Aaron




One thing one of our suppliers has asked about is how the R-values get
treated in equest, specifically, for foam board insulation and it opened up
some reflection and in-depth thought on how this might be addressed in
equest moving forward.



I decided to post this more to incite discussion (my knowledge on this is
limited).



It appears that NRCA has been publishing R-value testing data separate from
the actual NRCA roofing manual as they get newer data and more accurate
data.



I was emailed a 2016 PDF file called “Insulation Design Predictability in
Alaska” containing testing data and R-value variations by temperature for
different foam board media.  If you email me I can forward it.



The result -  that has been the trend over the last few years - is that
rated R-values have been going down for polyeurethane board and polyiso
board.



The example for polyisocyanurate from 2014 was telling.  At 75 degF a 2”
polyiso board had an average (mean) R-value of 5.55 out of a total of 7
samples.

The sampling for the same 2” polyiso at 25 degF had an average R-value of
4.049 (some say the newer data is 3.9).



One question is how we might be able to address this in equest in the
future as we push for more accurate representations of energy usage and
energy modeling in equest.



A published R-value on a cut sheet is not necessarily the performing
R-value of the material in an Alaskan winter where we can get -40F or -50F
for up to a week before warming up to -10 or -20 (sometimes colder).  And
Alaska is definitely not alone in this.



But it doesn’t appear that even NRCA has testing data for any foam board
under a temperature of 25 deg F (which is warm for Alaska in winter).



The question some people have in Alaska is how this will be addressed in
terms of energy efficiency calculations – which, obviously, isn’t easy to
answer.



There’s a definite bell curve associated with the testing data that shows a
steep dropoff of R-value with temperature – particularly temps colder than
25F – for both polyeurethane board and polyiso.



It would appear that the only way to achieve an accurate representation in
cold climates would be to conduct independent testing and establish an
overall annual average R-value (or at least a calculated number).



But even this would not necessarily be accurate. An equest model with a
single (averaged) input R-value would represent slightly elevated cooling
needed in summer and heating in winter that would likely be reflected as
“too efficient” than what you would see in reality.



Given the lack of data we ended up going with cutsheet standard values for
our previous model in equest but I, obviously, now know that 2”
polyeurethane board is not necessarily one R-value or another.



When we model a wall or roof buildup in equest we are doing so to represent
the thermal performance of the unit throughout a model year – with the foam
being just one piece of the puzzle.



What isn’t clear is how we might be able to move forward with more accurate
data.  Our foam supplier that manufactured all the material for our LEED
project was asking me if there was some way to address this moving
forward.  I can’t say…



To my knowledge, it still remains to be addressed since we don’t have a lot
of accurate testing data applying to these materials in places like Climate
Zone 8 or 8A here in Alaska.



So without accurate testing data we really can’t even calculate the R-value
accurately other than to continue the bell-curve of the current data into
the lower temps.  But then you would be seeing a 2” polyiso foam board with
R2 when you get into your below zero temps which is – wow…  Something is
wrong here.



The manufacturer said this was due to the nature of how the foam changes
with temp – specifically, if the cells within the foam can shrink with cold
temps.



It is even possible that this stressing of the material over extended
periods can cause it to pull in moisture and moisture-laden air (slight
amounts) which could also play a roll in cold temps.  But to what extent
this can occur is unknown.



I’m curious what others have seen or come across or how we might be able to
reflect this more accurately in equest.



Thank you!



Chris Baker
CCI Alliance of Companies

Fort Wainwright, AK
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.onebuilding.org/pipermail/equest-users-onebuilding.org/attachments/20160831/6c8ee471/attachment-0005.htm>