May 16, 2013

Reason Foam Fails #3 – Degrading Thermal Insulation Values

Polyisocyanurate (polyiso), XPS, and spray closed-cell foams substantially degrade over time as the insulating blowing agent gases diffuse out and plain air diffuses in.  How much does it degrade and at what rate?  To answer, industry has come up with something called Long Term Thermal Resistance (LTTR) testing.  LTTR testing takes into account degradation by providing R-values measured on a time weighted average over the “long term”.

What do you suppose is meant by long term?   100 years?  75 years?  50 years? 25 years?  No.  According to industry, “long term” is 15 years.

How could this be?  To answer that, we need to understand that foam is viewed, by-and-large, by industry, as flat roofing material.  Foam started on flat roofs and its testing and evaluations have been set based on those industry parameters.  As foam use has grown to cover entire buildings – the parameters for testing and evaluation have largely remained stuck at roofs.

So again, why 15 years?  Because industry says that 15 years is the average life of a commercial roof.  At which time, presumably, it all gets replaced.  However, with closed-cell foam now buried in facades and framing it may be more reasonable to assume that – in all these non-flat commercial roof situations – the foam won’t be removed and replaced.

So let’s talk crazy and assume you’re planning on having the closed-cell foam stay in place on the building indefinitely – for 25, 50, 75 years or more, what then?  What’s the weighted average over the actual functioning life of the foam installation?  There’s no data.  No one can tell you.  The only thing we know for sure is that it is going down, down, down.

The chemical companies that make foam are tight-lipped about leakage rates.  When Alex Wilson and John Straube looked to leaking gases to examine the global warming impact of some of these foams for Alex’s article Avoiding the Global Warming Impact of Insulation – they were left having to make educated guesses.  They explained their guess of 50% depletion over its useful lifetime as being conservative.

This general lack of understanding has been well know and documented for many years including an authoritative January 1988 article, Aging of Cellular Plastics: A Comprehensive Bibliography, in The Journal of Thermal Insulation by Ronald P. Tye.  And still today, 25 years later, architects, consultants and builders have no solid idea what is the true insulating value of that closed-cell foam that they are wrapping their “high-performance” buildings in.  Astonishing, no?

The onus is on the chemical companies to prove their products’ effectiveness through clear and useful information on truly long-term performance.  If you are putting foam in a place that is intended to stay there for 50 years or more – then the test data should reflect that.  A 15-year average is not long term.

To add insult to injury, the data that the chemical companies have provided on LTTR is substantially misleading, overstating the R-values by 6% for polyiso and 10-25% for XPS, according to a study by Sachchida N. Singh and Paul D. Coleman of the Huntsman Advanced Technology Center.

As the rate of diffusion generally decreases at a slower rate over time, the LTTR 15 year time weighted average is typically represented by the insulation value measured at five years of aging.  The National Roofing Contractors Association have also completed independent LTTR testing and conclude in an article by Mark Graham in Professional Roofing, in January 2006:

Seventeen of the 20 samples tested exhibited R-values less than their established LTTR values.  All these samples were less than five years old, the relative age the LTTR methodology is intended to represent.  Four of the samples with R-values less than the established LTTR values were less than one year old.  On the basis of this data, a positive bias in the LTTR methodology clearly is apparent—that is, the LTTR methodology appears to overstate a product’s actual R-value at five years of relative aging.

So if independent testing has demonstrated that the R values are consistently being overstated by manufacturers – geez, I wonder, industry wide, how much lost energy and money to consumers these overstatements might represent?

But for the moment it seems we are stuck with these “long term” tests – so what are some of the parameters?  A key parameter is that the foam is tested with an exterior mean reference temperature of 75 F.  This may make  perfectly good sense if you are in the roofing industry – roofs get darn hot after all.  But what if the external temperature is lower?  What if you dare use foam insulation on a wall?  Or soffit?  With a vented rainscreen in front of it?  Or a north facing one at that?  What happens if the exterior temperature is freezing?  The foam performs substantially still worse.  The Building Science Corporation recently issued a paper identifying some of these temperature-testing anomalies, extensively referencing the work – from 2003 through 2010 – of the NRCA.

Although polyisocyanurate in heating climates is typically sold by the chemical companies as R6 or 6.5 based on it’s testing of non-aged foam, at colder, more realistic temperatures, the NRCA concluded in 2010 that in heating climates specifiers and designers should use an in-place design value of R5.

Mark Graham of NRCA wrote in Professional Roofing Magazine:

Although the LTTR method of R-value determination and reporting may be appropriate for laboratory analysis, research comparison and procurement purposes, NRCA does not consider LTTR use to be appropriate for roof system design purposes when actual in-service R-value can be an important aspect of roof system performance.

One wonders what the recommendation should be for foam walls with vented siding over 50 years in a cold climate?  Can we reasonably speculate that the polyiso foam should be valued at less than R5?  How low should we go?  We don’t dare speculate.

The chemical companies are not going to change unless they are forced to change.  As long as you keep specifying and installing foam, you will never know what you have 25 years from now.

So as opposed to foam where you don’t know what you’ll have 25 years from now, look at dense-pack cellulose and fiberglass, mineral wool, wood fiberboard, cork and cellular glass – insulations that will, when properly installed, maintain their R-value indefinitely; and in their reliability, providing true high performance.

This post is the third in a series called Foam Fails.


Aging of Cellular Plastics: A Comprehensive Bibliography, by Ronald P. Tye, Dynatech Scientific Inc, Journal of Thermal Insulation, January 1988

Avoiding the Global Warming Impact of Insulation, by Alex Wilson,  Environmental Building News, June 2010.

Accelerated Aging Test Methods for Predicting the Long Term Thermal Resistance of Closed-Cell Foam Insulation, by Sachchida N. Singh and Paul D. Coleman, Huntsman Advanced Technology Center

Info-502: Temperature Dependence of R-values in Polyisocyanurate Roof Insulation,  By Building Science Corporation   Created: 2013/04/11

Revised R-values NRCA has revised its longstanding design R-value recommendation for polyisocyanurate insulation, by Mark S. Graham, Professional Roofing Magazine, December 2010

R-value concerns, by Mark S. Graham, Professional Roofing Magazine, May 2010

Testing LTTR, Research reveals the LTTR method may be over-reporting results, By Mark s. Graham,  Professional Roofing Magazine, January 2006

Testing polystyrenes’ R-values: R-value tests for polystyrene insulation produce mixed results, by Mark s. Graham,  Professional Roofing Magazine, May 2011

, , , , , , , , , ,

Tweet about this on TwitterShare on FacebookShare on LinkedInPin on PinterestEmail to someone

4 Responses to Reason Foam Fails #3 – Degrading Thermal Insulation Values

  1. MushroomSam May 20, 2013 at 10:15 am #

    Thanks for another excellent post. We’re developing a new insulation material that joins the club of cellulose, cork, wood fiberboard and other materials that don’t lose R value over time. I just posted a link to this article on

  2. Martin Holladay May 22, 2013 at 3:35 pm #

    Ken, The bottom of the range of R-values per inch is not as mysterious as you maintain. It is fairly easy to figure; if you want to be
    pessimistic, just assume that 100% of the gas blowing agent is replaced
    by air. I think you will end up with a worst-case scenario of about R-5
    per inch. This is a rather shallow critique of useful building
    materials. It’s a little like saying, “You should never install
    cellulose insulation on your attic floor, because it settles over time.”
    So what? Just install a little more to account for the settling.

    • Ken Levenson May 22, 2013 at 5:26 pm #

      Martin, Thanks for the visit, always a pleasure. It’s pathetic that after decades the chemical companies still don’t give straight answers regarding a fundamental quality of their product. The testing done by the NRCA and others raise important questions and point to more questions. But I’m glad your confident that R5 is the bottom – I guess we can all take that to the bank.

  3. W. Blake Talbott July 11, 2014 at 2:48 pm #

    Very good article with respected resources and references. I am an XPS foam panel proponent for roofs and walls. About four years ago Duke University had a 30 year old ballasted XPS roof, removed and tested revealing 98% +/- of original R Value, reinstalled with new roof. This is not a 100 year old insulation yet and I do not know of any insulation that claims to be or has aging or advanced aging testing performance records of a 100 years or more.

    If you have or know of an independent comparison chart based on aging for insulations, I would appreciate you sharing?

Leave a Reply