Does More Volume Mean More Heating and Cooling Load?

16 Comments Read/write comments

Attics encapsulated with spray foam insulation have more volume

What happens to the heating and cooling loads when you encapsulate an attic? With the insulation and air barrier at the ceiling below the attic, you're excluding the attic space. That volume of air up there isn't involved in the conditioning of the home. But when you move the enclosure to the roofline (usually by installing spray foam insulation beneath the roof deck), now the attic's volume is included inside the building enclosure. Occasionally I hear people say the loads will be higher because of the extra volume. Does having more air inside really increase the loads?

A load calculation refresher

To understand this issue, let's take a look at what goes into doing a load calculation. To do it right, you have to consider all the ways heat enters or leaves a home. Here they are:

  1. Enclosure loads - This is where most of the heating and cooling load comes from. It's the heat that conducts through walls, windows, doors, ceilings, and floors. It's also the radiant heat that comes (mainly) through the windows, also known as direct solar gain.
  2. Infiltration load - The air that leaks in through the building enclosure is really a sub-category of enclosure loads. But it's worth separating it out if for no other reason than it's entered separately in the calculation.
  3. Ventilation load - The outdoor air you bring in for ventilation adds heat (both sensible and latent) in summer and results in heat loss in winter.
  4. System loads - When you put a heating or cooling system in unconditioned space, there's heat gain in summer and heat loss in winter at the system. Likewise for the distribution system, whether you use air or water. Duct insulation and sealing and pipe insulation help reduce those gains or losses but you still have to include that extra heat gain or loss when you're figuring the total loads.
  5. Dehumidification and humidification loads - All dehumidifiers give off heat. Most of them dump that heat into the space you're dehumidifying. (An exception is the Ultra-Aire SD12,* which is a split-system dehumidifier.) Humidifiers need heat to vaporize the water. Sometimes that heat comes from the heating system itself, in which case it adds to the heating load.
  6. Internal loads - People give off heat. Lights give off heat. Appliances give off heat. You get the idea. Those things all get included, too.

By entering all the relevant details of the house under consideration, you get the loads for each room, each zone, and the whole house.

What's air got to do with this?

So, which one of those loads is related to having a higher volume of air to heat and cool? Well, a couple of them are actually. When air leaks into a house, that unconditioned air has to be heated or cooled. Same with ventilation, except that in this case the "leakage" is intentional.

But the question here is about the effect that increasing the volume has on the load. When you encapsulate an attic and bring it into the conditioned space, the additional air in the attic doesn't add anything to the load. If a sloppy spray foam job resulted in the attic not being airtight, then infiltration into the attic does add to the load, but that's not what we're talking about here.

Now, you may have more heating and cooling load when you encapsulate an attic, but it's not because of the extra volume of air inside the conditioned space. When you move the enclosure to the roofline, you also increase the surface area of the home. That can increase the enclosure load. But if you've moved your HVAC system from unconditioned to conditioned space, you may end up with a lower load. In hot climates that can be significant.

Another factor that increases the loads when you encapsulate an attic is the insulation level. A lot of times, contractors will use a lower R-value for insulation on the roofline than they would on the ceiling below. (Martin Holladay wrote an article about this several years ago on Green Building Advisor. It's called It's OK to Skimp on Insulation, Icynene Says.) For example, most of Georgia requires R-30 for ceilings but you can install spray foam insulation at the roofline as low as R-19.

Correlation ain't causality

The problem here may be that the people who believe a bigger volume means a bigger load are confusing a correlation with causality. Hey, I get it. All my friends in Maine are trying to reduce their consumption of margarine because it'll improve their chances of staying married. See the graph below, and you'll be convinced, too. (See the website Spurious Correlations for a lot more of these interesting connections.)

Eating margarine causes divorce in Maine

Wait, no! That's a correlation between two variables but there's not a shred of evidence that there's a causal relationship between the two. Same conditioned volume and heating and cooling loads. We know what factors contribute to heating and cooling loads. When you increase the size of a home by encapsulating an attic or a crawl space, it's the extra surface area and the insulation levels that affect the loads, not the volume.


* In full disclosure, Therma-Stor, which makes the Ultra-Aire line of dehumidifiers, advertises in the Energy Vanguard Blog.


Related Articles

Humidity in a Spray Foam Attic

HVAC Design Done Right - Manual J, S, T, & D

3 Reasons to Remove Attic Floor Insulation in a Spray Foam Attic

The #1 Question to Ask before Putting Spray Foam in Your Attic

Infiltration Occurs at the Surface, Not in the Volume


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


May 11 2017 - 9:03am

Thanks for this Allison. I still have to think about it some more, but thanks. Agreed that our knee-jerk reaction is to associate sizing with volume, not surface area. And I did not realize that the GA code made exceptions for insulating on the underside of the roof deck as opposed to the attic floor - sounds a little crazy to me. Is there sound science behind that reasoning?

May 11 2017 - 9:15am

Charles, reducing the R-value at the roofline is allowed only through tradeoffs. Since the ceiling insulation is reduced, some other part of the building enclosure has to be better than code to counter the weaker ceiling. I linked to Martin Holladay's article on this issue in the text above, and that's a great place to look for more discussion on this.

May 11 2017 - 9:19pm

@Charles, there's a logic for putting less insulation on the roof. When you look at the heat flux across the ceiling, it becomes obvious that it takes less R-value at the roof of an unvented attic that's not directly conditioned to end up with a similar ceiling load, compared to an insulated ceiling under a vented attic. Whether an R-21 roof equates to an R-30 ceiling as Allison described depends on several factors (roof pitch, ceiling construction, ceiling-to-roof ratio). The final trade-off depends largely on whether HVAC is in the attic. As Allison said, whenever you can incorporate HVAC into conditioned space, it's better to go with a vented attic, assuming ceiling is tight, even ignoring the savings in first-cost. That's always my recommendation to clients.

Code officials will usually accept a letter from the foam manufacturer certifying the equivalency of lower roof R-value (I know that Icynene has issued these letters). And as Allison mentioned, you can always use the Total UA Alternative path to demonstrate compliance via trade-offs (Section R402.1.5).

Also note that the IECC for commercial buildings has a lower prescriptive R-value for insulated (unvented) roofs as compared to insulated ceilings under a vented attic (Table C402.1.3).

May 11 2017 - 9:11am

Another great and timely article. We are simply changing the way we build the envelope with attic and crawlspace encapsulation. This one answers a question that I have received many times from clients and subs. Great job. Achieving a blower door test below 1.5 ACH is getting easier all of the time.

May 11 2017 - 9:23am

You mention two factors which may increase load by insulating at the roofline, the increased area and the common practice of settling for a lower R-value. A third occurred to me in a discussion with a contractor who was arguing for spray foam at the roof deck rather than cellulose at ceiling level, this after I'd taken pains to included raised heel framing, kept all mechanicals and ductwork out of the attic, kept all can lights out of the ceiling and indicated a well sealed scuttle instead of a pull-down stair. He claimed the delta-T would be less because the outside air temperature in summer conditions would be lower than the temperature inside a vented attic. True enough, but I felt the delta-T should be measured not to outside air but to the sheathing temperature which with a dark shingle would probably be higher than either outside air or the air temperature of the attic. Any thoughts/empirical studies that you know of?

May 11 2017 - 9:27am

James, yes, I've heard that, too. If you think in terms of BTUs, though, it's clear that the attic temperature is higher than ambient because of the heat that conducted through the roof. Not all of that heat stays in the attic as some gets carried out through the attic vents (soffit, ridge, gable...). If you put your insulation at the roofline, you have to deal with all the BTUs. At the ceiling level, you have fewer BTUs to stop.

May 11 2017 - 12:40pm

Hello Allison,

Thanks for all the articles and everything that you do to help educate everyone about building science.

Your response to James has me a little confused (I confuse easily). Are your your point simply that you need more insulation at the roof line that at the ceiling line? Do we know if the additional insulation at the roof line (due to greater square footage and to achieve higher R-Value) and the related costs are justified by the energy savings?

While I'm focused on the co$t$, I have not lost sight of how valuable it is to be able to enter (without melting) and storing things in a conditioned attic (additional useful sq/ft).

May 11 2017 - 5:25pm

I just remember, having been to about 3 lectures by Joe Lstiburek, he always said that insulating at the ceiling level is better than at the roof.

May 11 2017 - 3:35pm

Corwin, sorry for the confusion. No, you don't need more insulation at the roofline than you do on the attic floor. You can get good performance with less insulation at the roofline. If you have ducts and air handlers in the attic, R-19 at the roofline usually beats R-30 on the attic floor. But, if you do a really good job with air sealing the attic floor, getting full insulation thickness over the exterior walls, and prevent wind-washing, that method beats the insulated roofline when there are no mechanicals in the attic.

And as you mention, there are other benefits to encapsulated attics beyond energy savings.

If you want to take a deep dive into this issue, read Martin Holladay's article and the comments (311 of them as of 5/11/17) on this topic:

It’s OK to Skimp On Insulation, Icynene Says

This discussion goes a bit beyond the point of my article so I'll probably write about it at some point, too.

May 11 2017 - 9:24am

To some degree, doesn't the same principle apply to a sealed conditioned crawlspace? You add the foundation walls to the surface area of insulated walls. But sealing them for leakage is much easier than a floor system with lots of holes cut and drilled in it. And, you get heated floors in the winter time. You are also "trading" the earth surface for the insulated floor space. If it were a basement, the surface area of insulated walls greatly increases over a crawl space. Is there likely to be a point where an insulated crawl space would increase the load if the crawl space were high with a large surface area of insulated walls?

May 11 2017 - 9:31am

Bob, yes, the same applies for encapsulated crawl spaces. I did mention them briefly at the end of the article. The more above-grade wall area you have in a crawl space, the more heat gain and heat loss you're have to deal with. Whether that increases the load or not depends on how well insulated and air-sealed the enclosure is.

May 11 2017 - 9:48am

Spot on useful information as usual. I do enjoy the effort to explain these complex issues. I wish you would have sent it out last week though; I just sent out an email to some local builders here in Southwest Florida explaining why having built in truss duct chases is better than encapsulating the attic. One of my points was the volume of air to condition. I guess the saving grace is that it is still a much better method, having the duct chase should be easier to seal more completely and we can pile on 12" of insulation.
Also from your other blog; hopefully someone will come with better products for buried ducts here. I wondered about a flexible insulation with vapor barrier that could be draped over the flex ducts and sealed off against the ceiling before insulating the attic.

May 11 2017 - 2:15pm

Q: Does GA code currently(or in future?) allow for a small supply & return duct in these unvented attics?

May 11 2017 - 9:29pm

@JC, returns are generally not allowed in an encapsulated attic. Hopefully someone else can be more specific, but I think you would have to cover the foam with drywall.

Supplies are not required but are sometimes added to manage moisture levels (although I would argue it's better to ensure the attic tight and thus keep moisture out). Without direct conditioning, the attic should stay within about 10F of the house. In the winter, it's important for the attic to stay warm enough to avoid condensation, especially if rafters are not fully insulated. But in that case, moisture would be coming from the house, not from the outside. The ceiling should remain uninsulated. In retrofit encapsulation projects, ceiling insulation is often removed because it's dirty and may contain rat droppings (and worse), but aside from that, it's important to remove at least some of the insulation so the attic can be indirectly heated by the house.

May 12 2017 - 8:37am

Thanks Allison, you should contact Ken Carpenter with Designed Climate. They have a patent for ac units with a dehumidifier installed and integrated that looks at humidity before changing the air temp. It also does correlation between actual and "feel like" temp.

Add new comment