Sub-Slab Mineral Wool

musingsheader image

Sub-Slab Mineral Wool

Some builders have begun installing a continuous horizontal layer of mineral wool insulation under concrete slabs

Posted on May 22 2015 by Martin Holladay

UPDATED on April 5, 2016

Most green builders who need a layer of horizontal insulation under a concrete slab specify expanded polystyrene (EPSExpanded polystyrene. Type of rigid foam insulation that, unlike extruded polystyrene (XPS), does not contain ozone-depleting HCFCs. EPS frequently has a high recycled content. Its vapor permeability is higher and its R-value lower than XPS insulation. EPS insulation is classified by type: Type I is lowest in density and strength and Type X is highest.), an affordable product that performs well in this application. If a builder specifies high-density EPS rated for below-grade use, the product is very durable.

That said, many green builders don’t like EPS. Some object to the fact that polystyrene is made from petroleum, while others worry about possible health problems associated with the brominated flame retardants that polystyrene manufacturers add to EPS.

These builders have always wished there was a good, affordable way to insulate concrete slabs without the use of foam.

What about mineral wool?

Mineral wool insulation is available in several different forms. Some builders have looked at the densest available mineral wool panels and wondered, “Are these insulation panels dense enough to install under a slab?” When asked that question, mineral wool manufacturers have usually answered “no” — at least until the last few months.

A couple of years ago, I asked Iain Stuart, a technical solutions coordinator at Roxul, about using Roxul under concrete slabs. Stuart told me flatly that Roxul insulation products were “not for horizontal use under slabs.”

After several years of dithering, however, the engineers at Roxul, a manufacturer of mineral wool insulation, recently gave builders the OK to use horizontal mineral wool insulation under concrete slabs.

The announcement was oral

The first sign that Roxul was ready to change its public position came last fall. In the November 2014 issue of Environmental Building News, former editor Alex Wilson shared the news that a Roxul representative at a Passive HouseA residential building construction standard requiring very low levels of air leakage, very high levels of insulation, and windows with a very low U-factor. Developed in the early 1990s by Bo Adamson and Wolfgang Feist, the standard is now promoted by the Passivhaus Institut in Darmstadt, Germany. To meet the standard, a home must have an infiltration rate no greater than 0.60 AC/H @ 50 pascals, a maximum annual heating energy use of 15 kWh per square meter (4,755 Btu per square foot), a maximum annual cooling energy use of 15 kWh per square meter (1.39 kWh per square foot), and maximum source energy use for all purposes of 120 kWh per square meter (11.1 kWh per square foot). The standard recommends, but does not require, a maximum design heating load of 10 W per square meter and windows with a maximum U-factor of 0.14. The Passivhaus standard was developed for buildings in central and northern Europe; efforts are underway to clarify the best techniques to achieve the standard for buildings in hot climates. conference in Portland, Maine, announced that the company would “stand behind” the use of Roxul mineral wool insulation installed horizontally under slabs.

Ever since Wilson shared that news, I’ve been waiting to see whether the oral announcement by a Roxul representative would be confirmed on the Roxul web site — for example, by the publication of installation guidelines for this application. As of this week, no such confirmation had occurred.

So I telephoned Roxul to ask, “What’s up?” I discussed the situation with Paraic Lally, the North American manager for specifications at Roxul. “There was a little bit of tug-of-war internally,” Lally told me. “Some people here at Roxul were not so sure about this.”

It turned out, however, that my phone call was well-timed. Lally told me that Roxul is about to publish a technical bulletin providing recommendations on the use of mineral wool insulation under concrete slabs. [Author's note: Since this article was published, Roxul has published the technical bulletin. Here is the link: Roxul Under-Slab Insulation.]

Does mineral wool have sufficient compressive strength?

In the months leading up to Roxul’s decision to support the horizontal use of mineral wool insulation under slabs, green builders discussed three different Roxul products that might be used for this purpose: RockBoard 80, ComfortBoard IS, and ComfortBoard CIS.

  • RockBoard 80 has a compressive strength of 353 psf at 10% compression;
  • ComfortBoard IS has a compressive strength of 745 psf at 10% compression;
  • ComfortBoard CIS has a compressive strength of 1,220 psf at 10% compression.

All of these products are rated at either R-4 or R-4.1 per inch.

Any of these three products should be able to handle the required loads, although the two denser products provide more of a safety factor. In its November 2014 article, “Move Over, Foam: Sub-Slab Mineral Wool Is Here,” Environmental Building News quoted engineer Bruce King. “‘Roxul and every other rigid insulation I know of is more than strong enough to support a ground slab,’ according to Bruce King, P.E. King explains that the weight of slabs varies from 50 psf for a 4-inch-thick slab to 150 psf for a 12-inch-thick slab, while code-imposed live loads vary from 40 psf (residential) to 50 psf (office) to 300 psf (heavy storage). ‘At the extreme, a fully loaded 12-inch storage slab is still far less weight than would appreciably compress the insulation.’”

Sub-slab mineral wool in Maryland

Before Roxul made its public announcement about the sub-slab use of its products, at least two North American construction teams went ahead and poured slabs over horizontal Roxul insulation. One of these projects was in Maryland; the other was in Wakefield, Quebec.

The Maryland project is nicknamed “Passive Birdhouse.” The home was built by Gosnell Builders of Oalkand, Maryland. The house has a basement foundation; mineral wool insulation was installed under the basement slab.

The architect, Carri Beer of Brennan+Company Architects in Ellicott City, Maryland, reported in a blog posting, “Two 5-inch staggered seam layers of Roxul RockBoard 80 have been laid as our under slab insulation. The total 10 inches of insulation was decided on because it simplified foundation construction and ordering. Only 8 inches was necessary per dynamic modeling.” (See Image #1 at the top of the page and Image #2 below.)

Although Beer specified the use of mineral wool insulation under the slab, she specified a denser insulation — Foamglas — under the footings. (For more information on Foamglas, see On the Jobsite with Foamglas.)

An architectural drawing showing these foundation details is reproduced as Image #7, below.

The project’s Passive House Consultant, Michael Hindle, is Beer’s husband. In his own blog posting, Hindle wrote, “Last Thursday we poured the slab on a bed of RockBoard 80. We had two layers with staggered seams. The vapor barrier is placed over the mineral wool and taped prior to pouring the slab. … We will have 8 lb. density mineral wool from Roxul as exterior insulation of the foundation and below the slab. I am particularly proud of this achievement given the resistance we have faced trying to get this material approved for this application, despite its adequate compression strength.”

RockBoard 80 is somewhat fragile

I telephoned Hindle to get more details. “The builder has been fantastic, but like most builders, they are averse to liability,” Hindle told me. “They said, ‘This is terrific, but we need an engineer to sign off on this.’ Roxul said that their insulation has the compressive strength to work under a non-load-bearing slab. We explained that to our engineer, but the engineer said, ‘No, I’m not signing off on it. It hasn’t been done before.’ So I pushed the people at Roxul. I really did push. I said, ‘We want to create a whole new market for your product.’ So eventually we got Roxul to approve the use of Roxul 80 or their CIS board for this application. When our engineer heard that, he said, ‘OK. I’ll do it. I’m just saying it’s rigid insulation.’”

The material was easy to work with. “The guys in the field had no problems,” said Hindle. “They cut it cleanly with a bread knife. They chinked with it. They loved it.”

Beer and Hindle were brave pioneers, but the decision to use RockBoard 80 — a product with only 29% of the compressive strength of ComfortBoard CIS — led to a few problems that could probably have been avoided if a denser product had been specified.

Hindle reports that the RockBoard 80 was fragile. “I showed up at the job site the day before the concrete pour, when they were laying down the vapor barrier. Some areas of the insulation were a little soft, as if they had been banged or pounded on or stepped on repeatedly. I don’t see any inherent flaw. I’m just expressing an excess of caution. We poured the slab. The concrete filled in any slight depressions, and then it hardened.”

Because the insulation had soft spots and depressions, Hindle told me, “Workers should be advised to take special care not to bruise the material, and it should not be saturated with water. I would advise workers to lay it down, then put down plywood strips or OSB strips for walking on it, so you don’t bruise it.”

He also said, “My feeling is, if you damage it, it could get soft and the fibers could be pulled apart. That’s just conjecture on my part. … It’s probably better to delay pouring the slab until the roof is on, so the slab insulation won’t get wet.”

Although Hindle worries that rain might make RockBoard 80 more fragile, I think that there is little evidence to support this worry. Mineral wool insulation is hydrophobic and fast-draining. As long as there is a well-drained layer of crushed stone under the mineral wool, water shouldn't hurt it.

Summing up his experience with RockBoard 80, Hindle said, “I would go with the CIS next time.”

It cost more than EPS

Hindle admitted that the decision to insulate the foundation with Roxul and Foamglas added cost to the project compared to EPS.

“We did a rough cost breakdown, and we figured that the incremental cost for using mineral wool and Foamglas was about $6,000,” said Hindle. “We could probably have eliminated some of the Foamglas, and we could have used thinner mineral wool, so at the end of the day, my guess is that the incremental cost could be brought down to $3,000 to $5,000.”

A demonstration project in Quebec

The house in Quebec was a demonstration home built by Mike Reynolds for EcoHome, the English-language arm of Écohabitation, a Quebec-based green building web site.

“I am a home builder and project manager, and I’m always trying green building stuff,” Reynolds told me. “We were looking for something to do. I thought, ‘Let’s build a house and put everything in it. Let’s really go for it.’”

Reynolds was ready to experiment with the use of horizontal mineral wool insulation under the slab-on-grade foundation. “The engineer at Roxul suggested that we could use the CIS board under the slab but not the footing,” Reynolds told me. “I spoke to a few engineers. They were worried that there had been no independent testing. They said, ‘You can use it under the slab, no problem — but under the footing, or under a load-bearing wall, I can’t sign off on that.’ So we used EPS under the footing.”

In a phone conversation, Reynolds said, “The engineers will tell you, ‘It will compress when you put concrete on it.’” But Reynolds feels confident that after this initial compression, the insulated slab won’t settle further.

Reynolds installed a thickened-edge slab on a pad of compacted gravel. In a blog posting about the project, Reynolds wrote, “In this case we have gone with 8 inches of insulation for a total of approximately R-32. … The insulation we used here is a mix of Roxul ComfortBoard CIS under the main slab, and ComfortBoard IS on the exterior of the slab and high-density Type III EPS foam under the footing, as Roxul has not been tested yet for the weight of a footing and load bearing wall. Following that will be a [horizontal] skirt of R-8 insulation extending 4 feet out from the slab just below final grade, to ensure the ground below and surrounding the slab stays above freezing.” (See Images #3, #4, and #5, below.)

Elsewhere, Reynolds wrote, “We placed 8 inches of Roxul ComfortBoard vertically to protect the exterior of the slab, and we included a cement board outside the Roxul but inside the form. We later attached the cement board with plastic tie straps that passed through the insulation and vapor barrier into where the future footing will be; a 1.5-inch screw was put into the end of the tie strap to act as an anchor inside the concrete after it sets.”

Sure, you can walk on it

Since Reynolds used ComfortBoard CIS for the horizontal insulation (rather than the RockBoard 80 used by Beer and Hindle), they didn’t have any problems due to foot traffic during construction. “The CIS is more dense than the IS,” Reynolds told me. “You can walk on it.”

I asked, were there any signs of damage — perhaps at the corners of the panels? Was it really OK to step on it? “Step on it? Sure,” Reynolds answered. “You can walk all over it, and you won’t even leave any footprints.”

Reynolds continued, “This stuff stays where you put it. Unlike rigid foam, there is no slippage. We put down four layers of mineral wool and it stayed put. And it’s easy to cut with a bread knife. I was very happy with it. I have laid down enough foam to say that I loved the Roxul. It stays where you put it, it cuts easy, you can compress it if you have to. If you are cutting around plumbing pipes, it cuts easily compared to foam, so you can do a better job.”

The video below shows how the mineral wool insulation was installed at the EcoHome project in Quebec.

A sneak peek at Roxul’s technical bulletin

Roxul’s technical bulletin has not yet been published, but Thomas Hackett and Paraic Lally from Roxul graciously allowed me to see a draft version of the bulletin. The document I saw is titled, “Roxul Under Slab Insulation.”

Somewhat surprisingly, the bulletin recommends the use of ComfortBoard IS (which has a compressive strength of 745 psf at 10% compression) rather than ComfortBoard CIS (which has a compressive strength of 1,220 psf at 10% compression) for this application. According to the bulletin, a well-drained layer of crushed stone is required under the mineral wool, to make sure that ground water never saturates the insulation.

The draft version of the bulletin advises that ComfortBoard IS can be used “under poured concrete slab and above crushed stone” as well as “between slab edge and foundation wall, acting as thermal break.” It goes on to advise, “Do not use under footings and load bearing walls. High density EPS or cellular glass should be used for this application. Do not use in applications with a slab thickness less than four inches thick. … Vapor barriers should be used per building code or common local practices. In most climate zones the vapor barrier will be used above the insulation (between the insulation and the concrete).”

The draft document also notes, “Roxul ComfortBoard IS will compress slightly under the weight of the slab. It is important to ensure the design thickness takes into account the compression, causing a small reduction in R-valueMeasure of resistance to heat flow; the higher the R-value, the lower the heat loss. The inverse of U-factor. . Ensure not to step on board corners during installation to prevent damage to the product. If using 2 layers of ComfortBoard IS, overlap joints for added strength. Crushed stone below insulation will provide drainage and when coupled with perforated radonColorless, odorless, short-lived radioactive gas that can seep into homes and result in lung cancer risk. Radon and its decay products emit cancer-causing alpha, beta, and gamma particles. pipe will facilitate soil gas control. Do not use mineral wool sub slab in areas of high ground water levels.”

Image #6, below, is a reproduction of the architectural drawing in Roxul's draft bulletin.

Because it's possible that the final published version of this bulletin may differ somewhat from the draft I was shown, architects or builders interested in this method of insulation should consult Roxul before finalizing their plans.

Martin Holladay’s previous blog: “The Return of the Vapor Diffusion Bogeyman.”

Click here to follow Martin Holladay on Twitter.

Tags: , , , ,

Image Credits:

  1. Image #1 and #2: Michael Hindle
  2. Image #3, #4, and #5: EcoHome
  3. Image #6: Roxul
  4. Image #7: Brennan + Company Architects

May 22, 2015 11:17 PM ET

Sub-Slab Roxul

Martin - Nice ,informative job. Does anyone have info. on prices for any of the 3 types of Roxul boards for Maine or New England area?

May 23, 2015 6:05 AM ET

Edited May 23, 2015 6:08 AM ET.

Response to Kevin Zorski
by Martin Holladay

Contractors who contact local distributors and ask about the price of Roxul products get a wide variety of answers, depending on where they live in North America and whether other contractors in the area are requesting the material. In some areas, contractors are still being told that they must order an entire truckload of insulation; when that's the case, residential builders are in a quandary.

The bottom line: if you want to know about availability and prices for any building materials, you have to talk to distributors (lumberyards) in your area.

In his 2013 article on the topic, Alex Wilson shared what he learned about the price of Roxul insulation in southern Vermont:

"I was pleasantly surprised recently when I asked Leader Home Center in Brattleboro, Vermont, to price a number of insulation materials for a BuildingGreen report we’re revising. The contractor pricing for ComfortBoard IS came to $0.64 per board foot, compared to $0.48 per board foot for standard polyiso, $0.75 for fire-rated polyiso (Thermax), and $1.07 for XPS. While pricing will doubtless differ in other regions and for different quantities, the fact that ComfortBoard is in the same ballpark as these other materials is great. Even after correcting for the lower insulating value (you need more thickness of ComfortBoard to achieve R-10 than with the foam plastics), Comfortboard IS locally was more affordable than XPS: roughly $1.59 per square foot at R-10 for ComfortBoard vs. $2.14/sf @ R-10 for XPS."

May 24, 2015 8:10 PM ET

Basement walls
by Malcolm Taylor

For those of us who are a bit foam-wary this is very good news.
Martin, any comment on the basement detail Roxul provides for wall insulation?

May 25, 2015 5:09 AM ET

Response to Malcolm Taylor
by Martin Holladay

As some GBA readers know, I usually don't recommend the use of air-permeable insulation materials like fiberglass or mineral wool to insulate the interior side of a basement wall, because of the likelihood that condensation will form between the concrete and the insulation.

In its detail, Roxul appears to be admitting that condensation is indeed possible. The text in the illustration notes that the mineral wool layer on the interior side of the basement wall is a "drainage layer to footing." Since a layer of mineral wool connects the wall insulation with the crushed stone layer (which is drained by a perforated pipe), anticipated condensation can flow downward to the crushed stone relatively harmlessly.

It's an interesting solution to the condensation problem.

May 25, 2015 8:21 PM ET

by Malcolm Taylor

I bet there will be a lot of condensation on the poly.

May 26, 2015 3:57 AM ET

Response to Malcolm Taylor
by Martin Holladay

The Roxul detail shows just 12 inches of polyethylene at the bottom of the wall -- not a wide enough strip of poly to cause problems, in my viewpoint. There is a "vapour control layer" on the interior side of the wall -- presumably this could be gypsum wallboard painted with vapor-retarder paint, I suppose, although the vagueness of the text isn't helpful.

Strangely, the layer of material directly under the concrete slab is also called a "vapour control layer," not polyethylene, even though the words "6 mil poly" appear elsewhere (describing the 12-inch-wide strip of material at the base of the wall).

So the detail raises several questions. Remember, it's a draft detail.

May 26, 2015 12:45 PM ET

10" of RockBoard 80 seems risky with just a 4" slab
by Jonathan Rupp


Assuming a linear compression coefficients (of the RockBoard 80), given the code min live loads of 40-50 psf, and the 10" of sub-slab insulation, that would imply a dynamic +/- 0.1" + of compression and de-compression of the insulation from changing the live loads around on the finished slab (unless i am calculating something wrong). In my basement, i definately have a few loads much closer to the 300 psf under the shelves in my basement which get cycled moving from summer to winter seasons (which would be close to 1" of compression movement...

Granted concrete has rigidity, but I think you would need to add significant amounts of steel to reinforce the slab to prevent cracking and floor settling -- especially if a future home owner would decide to repurpose a room.

May 26, 2015 12:50 PM ET

Response to Jonathan Rupp
by Martin Holladay

First of all, this article recommends that anyone interested in trying this technique should use Roxul IS or Roxul CIS, not Rockboard 80.

Second, if this approach to insulating under a slab seems risky to you, I would never urge you to go this route. Just use EPS.

May 26, 2015 9:24 PM ET

by Malcolm Taylor

Another way to look at it might be that 10" of any type of sub-slab insulation is a bit nuts.

May 26, 2015 10:37 PM ET

Price and results
by michael hindle

Kevin, In my pricing I have found that mineral wool is less than XPS and competing with Poly-iso. It is a bit higher than high density EPS though.

Martin, thanks for writing on this and shedding more light on this product. I want to be clear that our experience with the RockBoard 80 was really very good. I would and in fact am using it again on another Passive House project. I am using CIS on yet another project and will have comparative experience to report on. My expressions of caution stem in part from my experience on affordable housing job sites. So, while the guys at Gosnell Buildiers handled the Rock Board 80 extremely well and got terrific results, I am always looking to find the design solution that will be applicable whether working with high end craftsmen like Gosnell, or crews on my affordable housing projects. Hence my excess of caution, and if you can get a little more margin of confidence, why not?

I am delighted that Roxul will release their technical Bulletin and bring this product more into the mainstream.

May 27, 2015 6:54 AM ET

Response to Michael Hindle
by Martin Holladay

Thanks for your comments. I'm glad to hear that you are using mineral wool under slabs, because job-site experience is essential to understanding how this technique works. I look forward to any future reports comparing Rockboard 80 to Roxul CIS.

May 27, 2015 10:20 AM ET

RockBoard 80 approval history
by carri beer

Just a little history of why we selected RockBoard July of 2014, prior to the Technical Bulletin release and after much pursuing, Roxul provided us with an official letter and multiple emails from their technical department stating that RockBoard 80 was among their products approved for sub-slab installation. The material was then approved by our structural engineer and installed a month later. The recent Technical Bulletin does not, however, list RockBoard 80. When you are a pioneer of something you tend to be overly cautious and critical of small issues and maybe want to change something the next go around, but there has been no problem in overall performance with the RockBoard 80 and we appreciate the support of Roxul in this endeavor.

Jun 2, 2015 9:47 AM ET

Response to Malcolm Taylor
by Mike Reynolds

"Another way to look at it might be that 10" of any type of sub-slab insulation is a bit nuts."

We likely live in different climates so this is possibly and apples vs oranges thing, but where we built the Ecohome house in Canada, the 8 inches of insulation we installed under our slab was arrived at after lengthy consideration. Air and ground temperatures, the cost of insulation, the cost of heating along with the type of ground we were building on were all factored. So the 'optimum' amount of insulation will vary depending on where you build and even the ground on which you build (rock conducts heat much quicker than soil for example).

When you separate concrete from the ground with enough insulation, it stops acting as a conductor and instead assumes the temperature of the conditioned space. This will for sure reduce and possibly eliminate all together the need to heat a concrete floor.

So if you direct your design and budget more towards heat retention rather than heat generation, you are then comparing the price of insulation against the price of heating infrastructure. If I add a couple of thousand to my sub-slab insulation budget and in turn I save even more than that on heating infrastructure (meaning if I no longer need a radiant floor but can get by with a heat pump and still have warm feet), I may save money on the initial build, I'll definitely save monthly on heating and I'll be paying off what effectively amounts to being a lifetime supply of heat.

In a super-insulated house like this you are effectively 'purchasing' heat at the bank through your mortgage, rather than 'renting' it monthly from utility companies, so a bigger bill is offset by a smaller one. The actual cost you have to put out monthly can be surprisingly similar, until the mortgage is paid off then it will be much different, in a good way.

Jun 2, 2015 8:27 PM ET

by Malcolm Taylor

With ten inches of Comfortboard IS you are looking at R-40 sub slab insulation. Surely you are well beyond the point of diminishing returns where the increase over, say, half that amount makes no appreciable difference.
While I agree entirely with your characterization of purchasing heat through the increase in efficiency of the building envelope as being a good thing, there is a point where more makes no difference, and I think R-40 well exceeds that point.

Jun 3, 2015 9:56 AM ET

by Mike Reynolds

Ours was 8 inches of Comfortboard CIS, so somewhere in the R32-R33 area. Where are you located Malcolm? I could see how it would seem like overkill for someone in a more hospitable region. Our engineer figures under most circumstances in this climate at least R20 -25 would be justified under slabs, more in cases of building on rock. Using 2 inch sheets our options were either R24 or 32, so we preferred being slightly over instead of slightly under. Finding the true 'sweet spot' of insulation is at best guess work anyway, and impossible to pin down stretched over different climate zones. Along with climate and ground considerations you would also have to factor the estimated and useful lifespan of a building to be accurate, another impossible feat. I would say though, that under-insulated slabs are more common that over-insulated ones.

Register for a free account and join the conversation

Get a free account and join the conversation!
Become a GBA PRO!