GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Insulate crawl space floor?

FiremanJoyal | Posted in Energy Efficiency and Durability on

We are building a single story home in zone 6B (West Central mountains of Idaho) and the framers are scheduled to arrive Monday 9/14. We are using a sealed, conditioned crawl space, but I haven’t seen any details that included insulating the floor of the crawl (bare dirt in this case). Our stem walls (30″ tall on top of 10″ tall footings) and rim joists will be insulated, likely with spray closed cell foam. One of the framers I interviewed thought I should really consider insulating the floor as well, as the ground temperature will be stealing heat from the home.

Is there any info anyone might point me to to help me decide if this would be worth the money, time, and effort? Your help and input is appreciated.

Thank you,

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. GBA Editor
    Martin Holladay | | #1

    For those who want to insulate a crawl space floor, the most durable approach would be to install a layer of rigid foam (EPS or XPS), followed by a layer of 6 mil polyethylene and a rat slab. A less durable approach would be to install the rigid foam with taped seams, followed by loosely-laid OSB sheets where necessary to provide an access corridor for maintenance workers.

    Most energy modeling programs show that the cost of this work can't be justified by the projected energy savings. If you really want to know how many BTUs per year would be saved in your case, you would need to model it. If I were you, I would spend my extra dollars elsewhere.

    Once more thing: regardless of how you resolve the question about insulating your crawl space floor, your current situation -- "the floor of the crawl [is] bare dirt in this case" -- is unacceptable. You always need a good vapor barrier (at a minimum, 6 mil polyethylene) above such a dirt floor in any sealed crawl space.

    For more information on crawl space details, see Building an Unvented Crawl Space.

  2. Dana1 | | #2

    According to a piece of analysis by the Building Science Corp folks a handful of years ago, a middle-of the road full-lifecycle estimate of what make financial sense for a crawlspace floor or basement slab in zone 6 is about R10, give or take, depending on local construction & energy costs. See Table 2, p10 of BA-1005 (the link to the PDF is in the lower right corner.)

    Note the sub-slab recommendation is ~R10 for zone 6.

    But if the other whole-assembly R values aren't up to the levels in that table it's probably better to boost performance elsewhere. To hit the recommended R20 on the stemwalls would take 3" of closed cell foam. To hit the recommended R35 whole-wall would require 2x6/R20-R23 with R20 of insulating sheathing. If your walls are an IRC 2006 code min 2x6 / R20 with no insulating sheathing, you get more bang/buck putting that foam on the exterior of the wall sheathing than putting it on the crawlspace floor .

  3. user-4524083 | | #3

    I'd consider foam sheets for the stem walls. Foil covered polyiso probably would be cheaper and more environmentally friendly than spray foam. Assuming you are building a well insulated house, I think the crawlspace floor should be insulated. The recommendations in the document Dana suggests are "minimums".

  4. Expert Member
    Dana Dorsett | | #4

    Kevin, those are nominal whole-assembly values in Table 2,not minimums. As explained in the paragraph preceding the table:

    "Recommendations for high R performance targets must, necessarily, include factors such as climate, available enclosure and space-conditioning technologies, desired lifespan, comfort expectations, lifestyle, as well as the costs of construction, energy, environmental damage and borrowed money. Some of these factors can be well defined, but many cannot. However, based on currently available technology, experience with costs & expectations with production and custom builders from coast to coast, and the desire to provide new homes that will be powered by renewable energy sources immediately or in the future, BSC has developed a set of recommendations (Table 0.2). These recommendations will therefore vary for locations with high construction costs, or low energy costs, or clients who value environmental impact highly."

    Going Net Zero Energy today takes somewhat higher R than those recommendations unless it's very carefully designed, but that's the rough mindset behind them. When solar PV is 1/4 the price that is today and 2x the efficiency (more watts per square foot of roof) those will be somewhat generous recommendations, but still not insane.

    (edited to remove line-breaks in the quoted BSC section, for readability.)

  5. user-4524083 | | #5

    Dana - Thanks for the clarification. The title of the table in question is labeled: "Table 2: Current Recommended "True" Minimum R-values including thermal bridging", so I figured they were minimums.

  6. Expert Member
    Dana Dorsett | | #6

    Those are minimums in "whole-assembly-R" for approximately what it takes to get to Net Zero Energy using Y2009 vintage solar PV technology, and Y2009 vintage heating technology, with a PV array that actually fits on the house.

    Since 2009 the efficiency of PV has increased incrementally by only a few percent, but the cost of PV has dropped dramatically. In 2009 the average cost of residential rooftop PV in the US was about $8.50/watt, today it's about $3.50/watt (before subsidies), and falling. In more mature markets like Germany & Australia it's about $2/watt. But until panel efficiency improve dramatically it still takes about the same amount of roof real-estate to produce the same amount of power. But it doesn't quite take the same amount of power:

    Also in that time average efficiency of both lighting and state of the art air-source heat pumps have improved by about 15% (HSPF 12-ish to HSPF 14+ ), and the state of the art of heat pump water heaters has improved by about 50% (EF 2.2 to EF 3.1). So in fact it takes a bit less rooftop real-estate now than it did then.

    The economic "learning curve" on the installed price of PV is that the price drops something like 20-25% every time production doubles, and the recent years trends are that PV volumes are doubling about every 2 years, but that trend is accelerating. This has been a 40 year learning curve, and there are no technical barriers to altering that trend. The prices in Germany & Australia are good indicators that in the short term that cost trend can be rapidly accelerated in the US, since both of those countries have comparable costs of business & labor as the US.

    As these trends continue highly probable (nearly inevitable?) that by 2030 PV is likely to be the cheapest form of energy available (by a good margin), which changes the lifecycle economics of how much more efficiency is rational to put into the building envelope of a house. If you build to the table 2 numbers in BA-2005, you'll almost certainly be able to economically & affordably heat, cool, & light the house with 2030 with low cost, rooftop sourced solar energy, to the point that it's Net Energy Zero (or net-positive).

    Beginning in 2020 all new residential home construction in California (a large market spanning US climate zones 2 through 6) will be required to be Net Zero Energy. Building practices then will then show just how close to reality the BA-1005 prescriptive approximations were. My expectation is that initially many houses will be overbuilt on R-value, but production tract home builders will hone in on the economic balance points within a very few years. And those whole-assembly-R values will evolve with incremental changes in PV cost & efficiency, as well as the evolution of heat pump efficiency.

    So, the minimum performance of the building envelope necessary to get to Net Zero will likely be a bit lower than in Table 2 in another 15 years, but not a LOT lower. If the house is too efficient, the "excess" power available on the rooftop can be economically applied to things like charging the electric car. Those are still reasonable efficiency targets to shoot for in new construction in 2015, but going much beyond that level needs more careful analysis of costs and goals.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |