Foam Insulation and Global Warming, Part 3

24 Comments Read/write comments

Well, I really stirred things up with my last article on insulation and global warming. My intention was to explain why Alex Wilson's results could be doing a disservice to the green building community. In the end, I was rightly accused of have done a disservice myself. So, here goes with part three of my take on the global warming impact of insulation. Let's see if I can get closer to the truth this time.

drainage plane wrb rigid foamboard sheathing

A public apology to Alex Wilson

First, I need to apologize to Alex Wilson. I apologized privately at the time of my last article. Now I do so publicly. I wrote things that went too far, saying his results were bogus and he was engaging in pseudoscience. I regret those comments and have removed them from that article.

As I've said before, Alex Wilson has done great work in his career. He's taken green building further than just about any other person in the field. His Environmental Building News has set the standard for green building news and analysis for decades. He also has a background in science and takes science seriously.

Wilson’s insulation and global warming study

Since that last article, I've done more reading of Wilson's 2010 article, Avoiding the Global Warming Impact of Insulation, and related works, as well as discussing it with others. Briefly, what he did was to calculate the amount of time a highly-insulated wall assembly would have to be in service to "pay back" in reduced carbon emissions the amount of global warming created by the insulation itself. That insulation impact, he stated, comes from two things: the embodied global warming potential (GWP) and the GWP of the blowing agents used in foam insulation. His conclusion was that extruded polystyrene (XPS) and closed-cell spray foam had long payback periods and should be avoided.

To do the calculations, he had to assume certain things and set the parameters for his model. As I've written before, some of his assumptions were:

  • The manufacturers used high GWP blowing agents.
  • The offgassing profile is uniform.
  • The lifetime of the product is somewhere between 50 and 500 years.

Those were the main things I focused on in my two previous articles. After doing more reading, discussing, and thinking, I now see I was looking at the wrong things. Here are the assumptions of his model that are more relevant:

  • He looked only at walls
  • He calculated GWP payback from various insulation materials after the 2x6 cavities were insulated with cellulose
  • He didn't give any credit to the air-sealing properties of any insulation
  • Energy savings and carbon reductions were based on heating with a 90% efficient gas furnace

In the next two sections, I'll discuss these other assumptions.

Looking at the whole system rather than individual materials

Wilson's study looked at the contribution to global warming from additional insulation on a 2x6 wall after the stud cavities were filled with cellulose or fiberglass. So the starting point was a wall with a whole-wall R-value of 14, which results from putting R-20 in the cavities and factoring in the effects of the framing.

As I discussed in my article on the diminishing returns of adding more insulation, most of the energy savings are due to the first few inches of insulation. As you add more and more insulation, the amount of energy you save keeps diminishing. If you're calculating the global warming impact of insulation based on the amount of energy saved after you've already got a wall with R-20 in the cavities, the results are guaranteed to be worse than if you take the initial energy savings from the insulation in question.

Since Wilson was looking at highly-insulated walls with R-values up to 60, the results don't tell you what the payback would be if you used closed cell spray foam insulation or extruded polystyrene (XPS) by themselves in a wall that just meets code. In that case, it wouldn't look nearly as bad for those two insulation materials because they'd get credited with all of the energy savings.

Also, Wilson's study excluded any air-sealing benefits of the insulation materials they looked at. But if you're using closed cell spray foam, airtightess is one of the biggest benefits you get. If you're going to compare the global warming impact of fiberglass, mineral wool, or cellulose to closed cell spray foam, you need to account for the global warming impact of the air sealing materials used for the fibrous insulation materials. And you'd need to give credit for the savings due to airtightness.

The bottom line here is that I don't think you can look only at the global warming impact of specific materials here. You have to look at the complete system. If we want to get serious about understanding the global warming impact of buildings, we need to model all the impacts together and see how those results compare.

Consider the source

Another really important component lacking from this study is a consideration of the source of the energy being saved. In their calculations, Wilson used a 90% efficient natural gas furnace. That's an easy one to do. Natural gas doesn't vary much in its carbon emisions from one area to another.

But what if the house you're building is going to be all-electric? And what if your electricity comes from one of the utilities with the highest carbon emissions? My electricity here in the Atlanta area comes from the Southern Company, which is reported to have the highest carbon emissions in the US. That makes anything I do to save energy very good at reducing carbon emissions.

Or let's say you're going to build a net-zero house in the Pacific Northwest? You've got lots of photovoltaics on the roof, and when you draw from the grid, you're getting some of the lowest-carbon electricity in the country. In that case, you're going to have a hard time reducing carbon emissions at all. (Keep in mind, though, that insulating homes does a lot more than just helping with our global warming problem.)

The fact is that you've got to be able to account for the source of the energy being saved if you want to know something about the real global warming impact of your insulation.

The real takeaway from Wilson's study

After spending so much time thinking about this recently, I've discovered that what bothered me the most about Wilson's 2010 article. He simplified a complex calculation and drew a conclusion that seemed to define a static property for various insulation materials. He wrote:

These differences are dramatic enough that, even if our assumptions are off by a significant factor, we can draw some general conclusions about sensible choices. If we're building highly insulated buildings and doing so in part to mitigate global warming, we should use insulation materials other than XPS or SPF...

I believe he underestimated that "significant factor" of uncertainty. That all-electric, code-minimum house using high-carbon electricity would have payback that is perhaps only a tenth of what Wilson found. Would you feel the need to avoid using XPS or closed-cell spray foam if the payback were only 10 years instead of 100?

The real takeaway from Wilson's study, however, is something most of us seem to have missed. What he showed is that we can calculate the global warming impact of buildings. It just needs to include more inputs than he used. It needs to be based on assemblies at a minimum, although it really should be done for a whole enclosure. That way you can take into account the airtightness. You can also look at different combinations of insulation in the walls, floors, and ceilings. The calculations also need to include the actual carbon data about the energy being saved.

The good news is that this process has already begun. David White put together a spreadsheet in 2011 that is a bit more nuanced than Wilson's calculations. Martin Holladay wrote about it and included a link so you can download it and use it. It still lacks inputs for carbon emissions from the source of electricity and for modeling different assemblies and enclosures. It's a good start, though.

I've been ranting about Wilson's article since it first came out, but now I see its true value. He took L.D. Danny Harvey's academic paper on the subject, which is inaccessible to most, and started the process of creating a useful tool. Eventually, designers and builders will be able to run what-if scenarios to minimize the carbon emissions of a particular building. Kudos to Alex Wilson for that.


Related Articles

Foam Insulation, Global Warming Potential, and BS

Don't Forget the Science in Building Science

The Diminishing Returns of Adding More Insulation


NOTE: Comments are moderated. Your comment will not appear below until approved.


Thomas Dugan

Very good article. I tell my customers that because any specific vendor is selling their part of a system, they come across as having the total solution. Certainly not the case. I advise them that we spend our first energy dollar on the convection issue and stopping air movement both in and out of the building envelope. The second dollar goes to conduction and the third dollar to radiation. After that, additional dollars can then go to beefing up the conduction issue with whatever works best for specific areas of the building.
When you add in the structural benefits of closed cell spray foam, it is a natural for number one and stopping convection. I use it under lower floor framing with a 1" flash to completely seal that area and then supplement with fiberglass batts. I also put several inches under the roof decking to "glue" the roof deck to the structure and seal in the attic space.
My walls are already sealed and insulated as I use precast insulated concrete panels from Superior Walls. I pretty much end up with a boat when completed.

Shawna Henderson

Kudos to you, too, Allison, for digging deeper and getting to the meat of the good work that Alex did. I did some minor studies on insulation materials for some work for NRCan and CMHC, nothing as focussed as AW's work or yours, but came up with the exact same issue: the material choice, when it comes to GWP can't be made without putting the material into context. Here in Nova Scotia, we're looking at onsite oil or coal-fired electricity as primary heating solutions. Our electricity produces GHG/carbon equivalent at about 1.1kg/kWh, and the system efficiency means that even a high-performance heat pump has an overarching efficiency that is no better than a code-required new oil furnace or boiler.

ANYTHING we can do to minimize energy consumption and leapfrog past code requirements is going to have a major offset to the GWP of a material, whether it's on retrofit or new construction.

Thanks for this!

John Nicholas

How does all this change if you are using recycled rigid foam? I recently found a local source of foam that is recycled from a major roofing contractor.


The payback is even shorter (possibly immediate?), because the product has already been 'in service' over a period of time.

Matt Sharpe

Great subject of discussion. I'd love to see this take monumental steps forward to cover embodied energy of heating systems options, PV, transportation, batteries, etc. so tools can be developed that will output the carbon impact of retro-fit options and NC scenarios that help us get closer to the ultimate goal of reduced global carbon emissions.

Andrew Kosick

I agree application and assembly really do matter. Where as I probably could (and would personally) avoid both of these materials in new construction, Ive found that the labor savings of closed cell SPF for retrofitting crawlspaces (something I do a lot of) to be so profound as to make the difference between someone having the work done or not. Not to mention that a cut ''n cobble never achieves the level of air tightness possible with SPF. When considering the difference in GWP between these crawlspaces not being insulated at all, or even not being insulated for some years, and using closed cell foam, I've concluded that the foam is a good option.

Alex Wilson

John, by re-using rigid foam that has a high GWP you are keeping that foam out of a landfill, where its blowing agent would likely be released much more quickly. Thus, the high-GWP blowing agent is being sequestered longer, and that's a good thing. I would argue that if using recycled foam, it would be reasonable to ignore the GWP impact of the foam. This is the same argument I use with salvaged materials for ignoring the embodied energy of the original material.

BTW, the issues that Allison raised in this blog are exactly the sort of issues that I had hoped my original article would raise. Nice job, Allison!

Allison Bailes

Thanks, Alex. Sorry it took me so long to figure it out.


Allison your original thoughts I believe were largely correct. The SPFA actually tried to address specific concerns or outright inaccuracies in Alex's article but I do not see that the original EBN article was ever edited to reflect these factual inaccuracies that helped to form the basis of the erroneous conclusions. These concerns or outright inaccuracies outlined by SPFA are as follows, and SOME of them you are addressing:
1. Amount of blowing agent in HFC-245fa is overstated (by about 100%!!!) - in fact my company Eco-Panels uses a newer more advanced blowing agent than 245fa with a GWP in effect about 25-30% of what Alex is discussing.
2. Embodied GWP assumptions may be incomplete or inaccurate
3. Sole consideration of SPF as a secondary or additional insulation over a baseline R‐value unfairly biases the conclusions of the study
4. GWP impact of fourth‐generation blowing agents not fully reported
5. Air sealing properties of foam plastics ignored
6. Inaccurate statements regarding quality, chemical safety and off‐gassing of SPF

Our contention is that stick frame construction remains the weakest and least energy efficient method of construction allowed by law. The single best use of dimensional lumber is as a tree in the forest, containing CO2, and if as a tree in your yard (if developers would not clear-cut and sell lumber to the market) could reduce your HVAC bills by up to 30 or 40% (according to DOE).

Allison Bailes

Charles, we agree on many of your points. I think my understanding of it now is a lot more complete than when I wrote my original article in 2010.

Allison Bailes

Having spent a lot of time in crawl spaces fastening plastic liners and rigid foam insulation, I have a great appreciation for your labor savings!

Allison Bailes

That definitely should be the ultimate goal here.

Allison Bailes

Thanks, Shawna. As you well know, the dirtier the fuel source, the more valuable the savings.

Allison Bailes

Sounds like you have a good system, Thomas. But shouldn't you be doing more than $3 of energy efficiency work on your homes? ;~)

Rachel Kingsley

I don't know much about WUFI or any of the really good energy modeling programs (we use REM Design) but as Ed Mazria from Architecture 2030 says, all architects (and home improvement contractors) should have a window in the bottom corner of their CAD screens that keeps track of embodied energy, projected energy usage, life cycle analysis, etc. When these professionals change the plans in any way, the data gets updated. In this way we will be able to assess what types of energy improvements are preferable in each unique situation. Right now we have a tool offered on the EPAs website called "Target Finder"
If you haven't gone on Architecture 2030, here is my favorite page:
Thanks Allison, keep up the good work!

Allison Bailes

I like Mazria's idea! I first heard him describe it 7 years ago when he spoke at the RESNET conference.

Ted Kidd

Allison, I LOVE that you are taking that next big step back and thinking systemically using first principles logic!

It's painful, because it exposes previous thinking disconnects, but it also takes you on the path to being a great designer who doesn't fixate on the unimportant.

We need homes that are controllable so occupants experience long term comfort, health, safety, which means getting things as tight as possible so you can manage air quality, temperature, humidity, and surface dew points all within a very narrow band all while driving EUI to 30--50% below average.

Keep going!!


Agreed. His article was based on mis-information and misunderstanding and actually cost us business. If he actually used good data like shown in the information I have sent you he would not have had an article to write in the first place. They have had this information since 2010 correcting their mistakes, yet they keep the article live as proven by your link. Why?

David White

A couple of notes that might help move this forward:

- the GWP calculator can be unlocked (password: "unlock") so that the CO2 emissions for electricity can be adjusted for a specific location.

- The tool unfortunately does not model assemblies with more than one type of insulation. However you can fury rig as follows: 1) model with one of the types included in the "reference R value" calculation, then track the other type in the graph. This gives you embodied GWP of the 2nd type plus heating energy GWP. 2) set HDD to zero and track the 1st type in the chart. This gives you embodied GWP of the 1st type only. 3) add the two together. Email me with questions.

- Regarding the total GWP of XPS/ccSPF as the only insulation in an assembly, I want to make clear that even at low R-values it's hard to get XPS/ccSPF to come close to the rest. E.g. in my NYC 5000 HDD climate at a whole-assembly R-15, you'll see about 50% higher total GWP (embodied GWP plus heating energy) using XPS as you would using dense pack cellulose in a wood frame. It's not that it's fine at low R values; it just gets worse at high R values. By the way, in my opinion it clears up a lot of this to jettison the concept of payback in favor of total GWP, as used in the calculator.

- Regarding accounting for the benefits of air tightness, in my work most air barriers are composed of tape, caulk, and/or sealant applied to materials that are already in the construction for other reasons. I don't expect the embodied GWP of the tape/caulk/sealant will add up to much.

I request that any responses to the two points above use real calculations. I am allergic to public speculation. I worked hard on that GWP calculator so that you all could download it for free, and use it, thereby sparing me a rash.

Tom Bassett-Dilley

Hi David and Allison,
I was glad to hear Allison present on this at the PH conference a few days ago, and would love to have the calculator to use in the office (also would like to know what was in that brown bag, Allison! ;-) But the links I've tried so far haven't worked--can you help with this? Thanks
Tom B-D

David White

Hi Tom, I think it's posted on GBA but not sure exactly where. It was on EBN but the link seems to no longer work. If you email me at david[at]rightenvironments[dot]com, I can send you a copy.

Erich Riesenberg

I have used expanded polystyrene throughout my house, with a topping of mineral wool for fire safety.

EPS seems exemplary. Air barrier, vapor retarder, environmentally less damaging, maintains performance due to the absence of offgassing. Much cheaper at least when bought at Menards.

The only downside is a few extra inches to reach the same r value, which to a non builder seems like an easy adjustment.

Allison Bailes

Erich, EPS is a great product. If you get the type with graphite mixed in, you can get the same R-value as XPS (5 per inch).