Diminishing Returns on Insulation

What is the climate zone and is the school one floor?  Is an energy model not being used to evaluate where the diminishing returns are?

There are multiple parameters, helpful to understand the whole life carbon.   If the grid is really dirty, the pay off of insulation can be better, but do we believe US will have a renewable grid by 2045?

It's a good question and one that we are thinking a lot more about. Particularly that many (warranteed) roof systems have specific insulation requirements (often of the higher GWP variety), it very well may be that the lowest life-cycle (and almost definitely the lowest 5-10 year) holistic carbon solution (that addresses both operational and embodied emissions) is not the maximum insulation solution. Issues of climate zone, utility grid, energy sourcing, the rest of the envelope, and HVAC system, as well as choices about the right time-frame to analyze carbon, can all impact optimization. I'd argue that this holistic carbon approach is the definitive way to get to the best solution. While there are a lot of factors, the math is not that complex, and getting incremental embodied carbon factors on roof systems is very straightforward. In searching for an article I read in the last 6 months about this trade-off for you, I found this other article, which also provides some context: Taking a Holistic Approach to Embodied Carbon | Architectural Record I'll also note that BuildingGreen woke up a lot of us about this a decade ago, when they reported on the insane GWP (driven by blowing agents) of foam insulation, XPS in particular, that estimated that it could take centuries of energy savings to "pay back" this embodied carbon debt. Fortunately, b/c of growing awareness, the most egregious blowing agents were phased out, but rigid foam still tends to be on the upper end of insulation GWP. For those not yet integrating LCA into their practice, this type of focused effort seems to me to be a great gateway (as well as focusing on concrete). -C

I'm not sure the articles your engineer shared really back up the point that going beyond code (R-30 in CZ5 if I remember right?) isn't worth it. They're based on single family residential construction areas and values. You can picture how much bigger those "small" returns shown in the Energy Vanguard graphs would get once the area factor is increased to the size of a K-12 school roof (and it's a simple enough equation that you might test that out in Excel yourself). The progress from R-30 to R-45 here specifically seems like it would be pretty significant if scaled up and adjusted for CZ-5. 

Two items: 1. UW Integrated Design Lab studied the total carbon footprint of projects with an ideal envelope U value v the IECC minimum. While it will surprise no one, more insulation lowers total carbon. I believe someone on this group earlier shared this: https://www.sciencedirect.com/science/article/abs/pii/S0378778822007605 2. LMN is about to publish our Envelope installment in the Path To Zero Carbon series. In it, we make a (not original) case that it is peak load reductions that often pay for envelope improvements, since energy is relatively inexpensive and pays back only over time. Peak load reductions can sometimes reduce first cost of mechanical systems. One of the difficult-to-quantify aspects of this is that envelope insulation levels impact thermal comfort to make spaces more thermally uniform. This reduces the likelihood that a facility manager will resort to cranking up the heat in winter and down the thermostat in warmer periods due to thermal complaints…wasting energy. But this thermal comfort issue and associated wasted energy is not a thing I have not seen quantified very well even though the whole idea of mechanical systems is to provide fresh air and thermal comfort (and reduce odors). Good luck! -Kjell F

How do you have an energy model too rudimentary to model differences in roof insulation?  That is one of the most basic things that doesn't require much detail.  I call shenanigans on that to begin with.  Because I can't help it, I built a quick model using the DOE prototype primary school in Boston. I used a TMY weather file made from the most recent 15 years worth of data from Boston Logan. I'm seeing 2.2% savings from R-30 to R-40, 0.6% more savings from R-40 to R-50, and 0.4% more savings from R-50 to R-60.  This is a very simple model I spent 20 minutes on.   I agree the emodied carbon is an import factor to include in these comparisons as well.

This is a tangent, promotional post to the comment Kjell provided. One of the research authors in the link Kjell provided is Tomas Mendez Echenagucia from the University of Washington department of Architecture. 

If you are in the Portland, Oregon area next week on Friday May 5th, Tomas will be the keynote speaker at the single day Passive House Northwest mini conference on embodied carbon. https://passivehousenorthwest.com/ Great chance to hear directly from Tomas and ask questions. There will also be two panel sessions during the day plus a happy hour panel.

Panel 1 focuses on materials with Terry Campbell from Sustainable Northwest Wood, Sadie Carlson from Green Canopy NODE, Michael Bernert from Wilsonville Concrete Products, and Yukari Kubo from Brightworks Sustainability.

Panel 2 focuses on design with Tessa Bradley from Artisans Group, Mia Kalatzes from Studio.e Architecture, and Jesse Elliott from Voussoir Architecture. 

The happy hour panel will be Skyler Swinford from Energy Systems Consultants and Dan Whitmore formerly with RDH and now with Indicator. Skyler and Dan will dive deep into embodied and operational calculations.

Highly recommended going if you can. I wish I could attend, but will be in Washington, DC for Living Future. Hope to see some of you there as well.

Hi all!  This is a fun one.  We often evaluate R-30 vs. R-40 for K12 projects but I don't believe we have ever recommended R-20 outside of southern climates.  I don't think it has been mentioned yet but where diminshing returns set in (in terms of EUI at least) is highly dependent on HVAC efficiency and also outdoor air flow.  Diminishing returns will usually occur quickly with a ground source heat pump system.  Same for a lab building with very high OA/exhaust.  Other factors are emissions intensity of the local grid and whether we are using mineral wool or XPS.  Since this conversation is on roof insulation, polyiso is not quite as bad a XPS but it still has a fairly heavy EC footprint.  So the question is how deep of an EC hole did we dig and what other ECM's could we have purchased instead? It's easy to run an energy model and show that the difference between an extra inch or two of insulation is in the noise for a building with efficient systems but there are other important criteria which Kjell points out.  How much HVAC tonnage did it save (although diminishing returns are certainly a factor here as well)?  How does it affect occupant comfort?  Noise?  In the future, I think one of the most important metrics will be the impact on winter time electrical energy usage and seasonal grid harmonization.

Hi Lara - We've worked on a lot of cold climate (CZ 5, 6, and 7) schools. Since you asked for case studies featuring increased envelope design, I would point to the Sierra Grande pK-12 project we worked on with Cuningham Group Architects in Colorado. Here are a couple features on it:

https://video.rmpbs.org/video/all-electric-near-passive-house-school-hpdfkn/
https://cuningham.com/2020/11/13/future-proof-your-school-with-the-passive-house-framework

The project has ~R-48 walls and ~R-65 roof with triple paned glazing. Fundamentally, one of the approaches we developed was trading off complex mechanical systems to pay for a robust envelope that used passive house standards to set targets. The project inherited the allocated budget the state had for it based off of a previous master plan (done by another firm) and program, so it had to work within that budget. Trading costs between MEP and envelope was a necessity to get the higher envelope performance. 

While I'm very familiar with the graphs plotting EUI vs R-values that show diminishing returns, there becomes a point where you can get so aggressive with the envelope in a cold climate that it unlocks the potential to radically rethink your systems and simplify them. In this project's case we went with predominately electric resistance heating. It's a rural school, and we wanted their facilities person (literally just one person) to be able to get most of the parts they need for the systems within an hour, rather than needing techs to deliver them from 4+ hours away in Denver. At the time of the design, we also didn't have faith in some of the heat pump systems to perform well in the climate (regularly getting to -20F), but that market seems to have changed in the last few years.  

You can run a pretty simple heat balance equation, but with these super-insulated schools in cold climates, you find out the 30 kids in a classroom are essentially heating your school. Early in the project, we had cold temps in the classroom which caused a panic. The contractor immediately started saying the heating plant was undersized because it was <20F outside and we were around 62F in the classroom. It turned out the heating system in that wing wasn't even energized, so they had no heat supply. Once that was discovered, it actually became a fun story about resilience.