QA-spotlightheader image
1 Helpful?

Air Leaks or Thermal Loss: What’s Worse?

Have we been brainwashed to consider only R-values?

Posted on Dec 20 2010 by Scott Gibson

Beefing up R-values and reducing air leaks are the twin rallying cries of builders focusing on energy efficiency. Regardless of the particulars of the house design, more insulation and fewer air leaks make houses more comfortable, more durable, and less expensive to heat and cool.

No one seems to argue that point. But Al Cobb wonders which is more significant.

"My real goal is to find the tipping point when a leaky building loses more energy via air changes then via the insulated envelope," he writes in GBA's Q&A forum. "I've had many answers where the losses from air leakage have been as low as 10% or as high as 50%."

Cobb believes home buyers have been "brainwashed" into thinking only about R-values, as energy codes give short shrift to the importance of airtightness. Energy modeling is especially frustrating, he says, because it asks for highly specific information on R-values but only broad generalizations when it comes to airtightness.

"Therefore, I'm looking for a study or analysis of homes (real or not) that have been modeled to the extent that heat loss from conductive and air infiltration losses are clearly defined," Cobb adds. "It only makes sense that as leakage rates increase, the decision to ignore air-sealing can be shown as a critical mistake."

Ain't no such animal

Good luck and God speed, suggests Robert Riversong: "Your question is similar to, 'What's the difference between an apple?'," he writes. "The answer could range from near zero to near 100%, and is entirely dependent on whole-house R-value and whole-house air exchange rate during normal operation (not under blower door testing). If you're asking about 'average' existing housing, there is some data on that. If you're asking about a particular new construction project, you have to do the heat loss analysis for that specific building including design or actual air exchange losses."

And as Riversong suggests, it might be useful to pin down exactly what we mean when we refer to "the code," as if there was a single, universal book of rules accepted by building inspectors everywhere.

"There is no 'The Code,' " he says. "There are international model codes which are becoming more comprehensive by the year and which each jurisdiction can choose to adopt or modify. And all building codes are minimum standards. You're free to exceed them as much as you'd like."

That said, there is always the HERS rating system, an approach to quantifying energy efficiency established by the Residential Energy Services Network. Its 0 to 100 scale compares the performance of a tested home to that of a house built to standards of the 2006 International Energy Conservation Code.

HERS inspections include both a blower-door testTest used to determine a home’s airtightness: a powerful fan is mounted in an exterior door opening and used to pressurize or depressurize the house. By measuring the force needed to maintain a certain pressure difference, a measure of the home’s airtightness can be determined. Operating the blower door also exaggerates air leakage and permits a weatherization contractor to find and seal those leakage areas. to measure airtightness and a test to measure duct leakage.

Some truth to the argument

As to Cobb's basic premise that too much attention is focused on R-values, AJ Builder is right on board.

"R-value — My biggest pet peeve, especially when discussed here on this supposedly leading source of green building advice site!" he says. "You all love to quote rules and 'the codes,' yet two homes can be built with the same approved R-value insulation and have 100% different energy needs. The codes should spec insulation in more regard as to how continuous it is, in how it actually performs at all temperatures, when it is needed most and other aspects of the entire assembly. We all know that fiberglass batts in an attic that is at 0°F is not giving the same true R-value as, say, the R-value of other installs — but where [does] the beloved code address this fact?

"The code and discussions here would make much more sense if R-value [measured the R-value of the] actual whole assembly, rated at worst design temperatureReasonably expected minimum (or maximum) temperature for a particular area; used to size heating and cooling equipment. Often, design temperatures are further defined as the X% temperature, meaning that it is the temperature that is exceeded X% of the time (for example, the 1% design temperature is that temperature that is exceeded, on average, 1% of the time, or 87.6 hours of the year). of stated climate," AJ suggests. "Otherwise the info is garbage."

That's exactly what I'm talking about, Cobb says. "The average consumer, building official, and architect is consumed with the perception of the importance of R-value," says Cobb. "That misperception is the basis of my question that started this thread. If modeling identifies the cost of ignoring good air-sealing, it can be used as a tool that educates all people about building better buildings by using products and systems that perform properly. The suggestion that the code has comprehensive standards is laughable."

Moreover, Cobb bristles at Riversong's tone: "I'm looking for data to help us build better homes and help the consumer make better choices," Cobb says. "What I'm not looking for is criticism of my professional capabilities."

Other solutions suggested

While Riversong suggests there is more of value in the model energy code on air sealing than Cobb or AJ acknowledge, J Chesnut has another idea, providing you're a math nut. You could, he says, play with heat loss formulas developed by ASHRAEAmerican Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE). International organization dedicated to the advancement of heating, ventilation, air conditioning, and refrigeration through research, standards writing, publishing, and continuing education. Membership is open to anyone in the HVAC&R field; the organization has about 50,000 members. . J Chestnut points to some specifics in a 2005 publication which offer a way of calculating infiltration rates for load calculations.

Hunter Dendy adds this idea to the mix: "Maybe the simplest approach would be to model a house and then adjust those variables to see how it changes the performance/heating loadRate at which heat must be added to a space to maintain a desired temperature. See cooling load. of the model," Dendy writes. "You probably won't find an across-the-board answer to your question since there are too many variables involved in house design and site conditions. Better to look at it on a project-by-project basis, but after doing this for a while you will get a feel for it."

And in the end, that's where the discussion seems to put us. Thermal insulation and air sealing are both essential for high-performance houses, but there may be no cut-and-dried answer to Cobb's original question that fits all circumstances.

Expert's opinion

Here's how GBA Technical Director Peter Yost sees it:

Mathematically, you can constrain the two equations for convective and conductive heat loss and force them to be equal. And then in some part the relationship will be driven by the geometry of the building and the relationship between surface area and volume. But problems quickly arise:

1. ConductionMovement of heat through a material as kinetic energy is transferred from molecule to molecule; the handle of an iron skillet on the stove gets hot due to heat conduction. R-value is a measure of resistance to conductive heat flow. is a field phenomenon and convection is a point effect. In the equations we just sum each effect and yet we know it is the sizes and locations of all the holes that are really important.

2. If we set the equations equal to each other, we are assuming that one does not affect the other. And yet we know that where and how the air leakage occurs can erode R-values.

3. Many building materials 'care' a lot more about convective than they do conductive heat loss because moisture moves with the air and can condense. So, we typically worry a lot more about convective than conductive heat loss in terms of durability.

And then there are the occupants. Their concern about the relationship between the two phenomena is complicated as much by the quality as it is the quantity of air exchange. While ASHRAE 62.2A standard for residential mechanical ventilation systems established by the American Society of Heating, Refrigerating, and Air-Conditioning Engineers. Among other requirements, the standard requires a home to have a mechanical ventilation system capable of ventilating at a rate of 1 cfm for every 100 square feet of occupiable space plus 7.5 cfm per occupant. gives us some guidance on how much fresh air — or more exactly, outside air — we should have in terms of occupant health and safety, how much outside air occupants need is a far from settled question and dependent on many more factors than the ones that architects and builders control.

There is no question that most folks, lay people and too many in our industry, are “R-centric” and give convective heat transfer short shrift. But a ton of great work has gone in to EPA’s Thermal Bypass Checklist to guide our priorities in cost effectively and practically tackling the big and typical holes. And both in terms of overall energy efficiency and building durability, the continuity of the air or convective barrier is just as important as and even codependent with the continuity of the thermal or conductive barrier.

Tags: , , ,

Dec 20, 2010 8:18 AM ET

I think air leaks is worse
by Buy Advair

I think air leaks is worse thing..

Dec 20, 2010 12:56 PM ET

Other factors too
by James Morgan

Our own experience in a mixed climate suggests the simplistic conclusion that air sealing improvements over base-level standards tend to have the greater impact on cooling load, while increasing R-value has the greater impact on heating load. This may be particularly true when the heating input is largely radiant: Michael Chandler made the very good point in a recent post however that air sealing in a heating climate is also important for maintaining indoor humidity levels.

The meta-discussion might also go to the many other aspects of maintaining thermal comfort while lowering energy inputs. Top of the list for me would be the effects of factors like surface emissivity and thermal capacitance of building materials on localized comfort and discomfort zones within the total energy economy of the structure, particularly in renovation situations where substantial R-value upgrades can be very difficult to achieve at reasonable cost.

Dec 20, 2010 3:41 PM ET

It's a dual approach
by John Zito

This is a similar discussion I have with spray foam contractors. They want to spray 4" of cc in the attic and call it good, that's it's better than r-49 (our code prescriptive in Michigan). And at roughly $1/in/sqft, I can see why. They also have a bit of oppositional defiant disorder when they are informed that there other ways to air seal that don't involve their product.

I think in the last few years, the building industry has done a decent job of getting the message out that there is more to reducing energy loss than r-value alone. But I also see a shift where those involved in reducing energy loss have ignored or forgotten that r-value is the other part of the equation.

Dec 20, 2010 3:47 PM ET

Response to Zito
by Martin Holladay

I love your diagnosis -- "oppositional defiant disorder."

Dec 20, 2010 5:34 PM ET

Edited Dec 20, 2010 5:36 PM ET.

R value is a distraction
by aj builder, Upstate NY Zone 6a

We should quickly change the way we talk about homes. When a home uses higher levels of insulation, that is good. But R value of the insulation product is a small aspect. The larger issue is the gross and net energy consumed per inhabitant. That is where we should be today when designing and building and discussing better building enclosures.

Passive House is leading the way in many aspects. It is time for others to do the same in their own way so as we can get this race to sustainability moving. The natural build community is another group worth mentioning.

And lastly... Google is setting one of the best examples of how we can all get there alive... I.E.-Open Sourcing. Google is applying the merits of open sourcing more than any other large entity. Thank you Google, Thank you open source communities of all nature.

Open Sourcing what we know, what we do, what we design and talk about, and where we are and what are goals are. If you have participated in a worthy green project, don't just tell us your R values, post your entire process, plans, details and all, including what you would change in the future. If you have made the best spreadsheets you have ever punched numbers into, post them. Tell us the good, bad and the ugly of it all including the damn "R values."

Capitalism raises all. The top too high and the bottom, well to just fat enough to need pharmaceuticals. Working more together with a Ben and Jerry's attitude and a Google need to be open source... we can get someplace better than where "has been" capitalism, has been... taking us of late.

Dec 20, 2010 5:40 PM ET

rule of thumb
by Michael Blasnik

If you just want to ballpark the energy impacts, then you can take the blower door measured CFM50 and divide by 20 to get the equivalent UA for the air leakage. Put another way, each CFM50 of measured air leakage creates about the same heat loss as 1 square foot of R-20 envelope.

Dec 20, 2010 11:16 PM ET

Air Leaks and More
by Doug McEvers

This is a wonderful document penned in part by energy whiz, Larry Kinney. It is worth a read for anyone interested in energy efficiency improvements.

Dec 22, 2010 10:09 AM ET

bang for the buck
by Mark Marcoplos

I do low-tech, "common sense energy audits" that cost the customer about $125. I can almost write my report before I even inspect the house. I will find multiple opportunities for air-sealing, as well as generic duct leakage, and sloppy insulation. Since most homeowners have a tight budget these days, I usually offer four hours of air-sealing & insulation tidying for about $200. This small investment has a big pay-off in lower energy costs & comfort.

They could get a blower door test and infra-red camera imagery with impressive computer print-outs and photos for $400-600. While this scientific verifiability is accurate and useful information, my approach identifies a huge percentage of the same problems and gets them fixed quickly and inexpensively.

Dec 22, 2010 10:54 AM ET

What About?
by R S Mills

How do breathable walls fit in to the equation?

Dec 22, 2010 10:59 AM ET

Response to R. S. Mills
by Martin Holladay

R. S. Mills,
Your question is unclear.

1. Are you talking about walls that are vapor-permeable or walls with air leaks?

2. Once you have explained what type of wall you are talking about, please tell us what your question is.

Dec 22, 2010 11:20 AM ET

Convective looping
by Allan Bullis CEM, LEED AP

One area not discussed is convective looping and its effect on R-value. An example is a building with a tight drywall inteior and rigid foam on the exterior. If the drywall is tighter than the rigid insulation, that will be the determining factor in the blower door test. The effective R-value of the system will be determined by how much air looping there is around the edges of the rigid insulaiton. I would like somebody smarter than me to discuss convective effect on R-value.

Dec 22, 2010 11:45 AM ET

by Darin

It is interesting to see these articles being posted today yet. Anyone that truly knows the industry knows that the way R-Values are required to be tested do not give the actual value to resist heat and cold at maximum ambient or conditioned temperature differencesfor their home. There is a FTC report 16 CFR Part 460 that was published over 5 years ago that will get a lot of people upset at the gov. Furthore more the majority of entities that claim they have the answer on green, energy efficiency, and affordible comfort for the homeowner just create confusion.
Dear M. H. in responce to your comment on 7/16/2009 we are still retrofitting hundreds of homes with 100% satisfaction. My advice to you is not to offer advice.

Dec 22, 2010 12:21 PM ET

Response to Darin
by Martin Holladay

R-value testing is performed at temperature differences established by ASTM. These tests are consistent for a reason -- to allow comparison of materials.

Of course, actual temperatures experienced by houses will differ from laboratory conditions. That doesn't mean, however, that R-values aren't useful.

More information here: Understanding R-Value

Dec 22, 2010 12:32 PM ET

Type of construction matters
by Mike Thomas, Michael Alan Thomas, Architect

If you're trying to make a seat-of-the-pants judgment about whether to correct air leakage or low insulation levels, it helps to consider the type of construction you're dealing with. Here in the City of Saint Louis most homes were built before 1920 and are three wythes of brick with a plaster interior finish. The R value is awful but through-wall air leakage is minimal. One problem is air leakage around double-hung windows (both between sash and frame and through the weight boxes concealed in the jambs), as are both air leakage and conduction through the roof (which may be flat or steeply pitched with 'garret' bedrooms with sloped ceilings but in any case with little or no insulation). This is a very different set of problems than found in a balloon-framed Victorian or a post-war ranch house in the suburbs.

Dec 22, 2010 1:08 PM ET

Progress in the Pacific Northwest
by Albert Rooks, small planet workshop

I'm happy to say that the Washington State Energy Code requires a blower door test starting January 1, 2011. Here come performance based requirements! The initial requirement is low: 5.0ACH 50, and it's actually presented as surface leakage area (.0003SLA = 5.0 ACH 50).

I think that it's important to elevate the industries understanding of the importance of air sealing in the context of building envelop energy efficiency. While the number is probably a moving target due to the number, and types of infiltration potentials, it's quoted by the Swiss Airsealing company Siga, as being in the 30 to 45% range of heat or cooling loss. Even though it's in their interest as a material vendor to post a higher number, I'm going to give them credit that the Western European Building industry has 20 years of experience over us on: how to, and the positive effects of air- sealing on light frame residential construction. In their markets we are just talking about their code requirements and not volunteer programs like Passivhaus. They have built up a wealth experience that we can tap. How about some investigatory pieces on how they do it?

I'm with Al Cobb on the "what does it all mean and how do we quantify it". Answering this question takes us a long way in being able to quantify the direct effects and savings of simple air sealing in conjunction with the other building characteristics of existing homes. I live in Thurston County in Washington State where our local Utilities have funded a home print (energy assessment) program. At this point we now have a data base of 400 existing homes, their building characteristics complete with blower door test. With this data now available, and if the effects of increasing airsealing were reliably known, Then it would be possible to start an assessment program with the existing home data to quantify both the cost and benefit of airsealing and thermal upgrades to these homes. When this is viewed on a Thurston County wide perspective, we can begin to see how systematic deep green retrofits would affect regional energy consumption and future generation needs.

Yes... I've gone beyond the scope of this column, but the point is it starts with a knowledge base within our building industry of the value and effects of airsealing.

Keep up the good work!

Dec 22, 2010 1:22 PM ET

by Dennis E. Hayes

The R value of a home is misleading. The coefficient of heat loss, which is a small decimal point number describing the heat flow in BTU’s per hour per inch of thickness, is the "U" factor and the reciprocal of that decimal number, or 1 divided by the U factor, is called the "R" factor, which is a whole number. The Insulation’s "R" value is only one small part of the formula used in calculating a building’s total heat loss or gain. The single most important factor is the shape of the building,of all possible geometric forms the pyramid has the least amount of exposed exterior surface area per square foot of heated or conditioned floor space. Heat loss or gain is a direct mathematical function of surface area. has some articles I've written on this subject.

Dec 22, 2010 1:33 PM ET

Air leaks and Thermal loss
by Jason Scheurer

Some of the things that we need to consider is not only the thickness of the R value, but how good it was installed. As certain protocols go, we rate the install as good, fair and poor. Believe me, most of the time, even in brand new construction, we see poor installs. As Building Science (BS) is concerned, the Air Barrier and Thermal Barrier need to be continuous and aligned at all times. Most of the time there is some type of break of these two components, which highly affect our (BS) evaluations. It could be wall penetrations, gaps in the Batts, etc. Also, we need to have all trades do what we call "Drill and Fill" their own holes as they go about their daily work. Building Science has been around for along time and we need to start implementing the techniques again. As for air leaks, I come from a mild climate, but as long as we all start doing air sealing techniques properly, we will see that our "R" values will work more realistically in any type of climate. Air sealing is the most important meassurement we must do to make our insulation perform at its best and hold that R value evrybody is talking about.

Dec 22, 2010 1:39 PM ET

Response to Albert Rooks
by Martin Holladay

You suggested, "How about some investigatory pieces on how they [Europeans] do it [air sealing]?"

In fact, GBA reports regularly on the topic. In addition to our many articles on the Passivhaus standard, GBA has published an article on the SIGA method of air sealing:
One Air Barrier or Two?

Dec 22, 2010 2:21 PM ET

Insulation and Right Sizing
by Trish Holder

Here's my two cents, as a keen observer of how consumers react to what they hear and learn about green building. I think R-Value and insulation has surpassed other strategies in terms of consumer awareness because it's something that consumers can relate to. We all wear coats, right? So consumers GET the value of insulation in their homes. To them it is very simple (even though it really isn't). But right-sizing equipment is something that is rarely discussed with homeowners--and an oversized system can cause serious problems for a house that is exceptionally well-insulated. I think it is crazy (and a little scary) that some homeowners will research the heck out of how to get the best R-Value and never give the sizing of their HVAC a second thought. But then again, homeowners are rarely invited into that conversation. Why, I do not know.

Dec 22, 2010 2:39 PM ET

Air Barrier or Insulation
by John

I am convinced in my area of New England, air sealing is the most important thing to do and must be done before insulation. In new construction this is pretty straight forward if you know how to do it. In existing housing its a lot more work. The blower door test and ACH 50 are good indications of leakage, but not perfect and are done differently from person to person. However if the test is conducted first, then air sealing is done and tested again, improvements can be measured and quantified. Basements and attics are usually the biggest problems in this area of the country. Understanding stack effect and bypasses is important to air sealing and it takes some time to figure out what is easily and less costly to fix. I have seen homes with a CFM50 of 8500 before air sealing and CFM50 of 1600 after proper air sealing. What this means in energy savings is enormous.

Insulation is a different beast and is very regional, as is the discussion of vapor barriers. It is always better after air sealing to insulate rim joist in the basement, and all penetrations into the attic area. Between floor penetrations and rim joist insulation makes a big difference too. Adding attic insulation and installing a good sealing attic hatch is always worth the cost. One or two inches of foam added to the sheeting or blown in while new siding is needed is a good way to get to the walls.

Dec 22, 2010 5:16 PM ET

by Darin

R-values as given or applied to insulation and other materials that are classified as insulation but do not insulate is not an accurate value because the testing in itself to determine such value is faulty.

Dec 22, 2010 6:42 PM ET

Response to Darin
by Martin Holladay

And what aspect of the testing are you calling "faulty"?

Dec 22, 2010 8:55 PM ET

r-value vs aier leakage, what's more important.????
by Steve McCarthy

Impossible question to answer!!!!!!!!!!!!!!!!
I do know that air leakage has a significant effect on insulation performance. Whenever I do an IR scan on a typically insulated fiberglass wall and some low density cellulose blows, it is embarassing to the customer and me, and I did not do the work. I just show how bad it performs.
I want to hear more from Blasnic on this one. One cfm 50 = heat loss of an R-20 sq. ft. so, for a climate with an average winter temp of 30 deg. f. and indoor @ 70. 40 delta t times .05 =2 BTU's for every sq. ft. per hour during the heating season. Michael! I' dumb. what are you trying to tell us? when I do the math, It looks like a BIG number. Does this big number just account for air in air out or does it also include the impact of R-value degradation from air flow thru the insulation????

Dec 23, 2010 1:15 AM ET

Re: R Value vs. Air Leakage/Sealing
by Peggy Deras, CKD, CID

All I know is that our home is comfortable for the first time in 25 years after sealing our attic only. We had plenty of attic insulation and crawl space insulation - no sidewall insulation. San Francisco Bay Area ranch home. Our heating bills are reduced quite dramatically as well.

I vote for the sealing.

Dec 23, 2010 12:48 PM ET

response to Steve M
by Michael Blasnik


I hope you realize that a Btu is a tiny amount of energy. Here's a sample calculation that might put things in perespective. Let's use a house that leaks 3,000 CFM50 and is located in a climate with about 5000 HDD (I would use HDD60, so that's equal to maybe 6500 HDD65)

3000 CFM50 /20 = 150 Btu/hr/F heat loss rate
150 Btu/hr/F * 24 hr/day * 5000 HDD (F*day/yr) = 18,000,000 Btu/yr in heat loss

That's might seem like a big number, but Btus are really really small. If you have a 90% efficient gas furnace, then you'd use about 200 therms/yr of natural gas (180/0.9). The typical heating use in such a climate might be about 700 therms/yr.

This number is only meant as a rough estimate of the energy loss associated with the natural infiltration. It doesn't attempt to estimate any interactive effects between air leakage and insulation which may lead to larger or smaller heat loss.

Dec 23, 2010 4:21 PM ET

on anectdotal evidence; & R-values vs leakage in CD's
by John (Gus) Gramling, architect

I'll second Peggy Deras' observation. Mine is also a small Bay Area 1950's rancher, but exposed to the Pacific (I'd call it a view, but it's foggy so often...). Replacing sash a few years back with good quality, better-sealing IG units made a big improvement in comfort. I've recently increased attic insulation, which has not had as noticeable an effect on comfort.
So why the focus on R-values as the standard? Because traditionally, R-values can be specified in construction documents in a way that's readily enforceable during construction. With a quick job-site look, the architect can observe (or the building official can inspect) that yes indeed, R-xx insulation is there, and in the quantities and locations it should be. Unlike sealant, which largely comes under "workmanship" (trust the contractor), and would require a more continuious (read : expensive) inspection to enforce, than most residential or commercial clients & projects care to bear. Many important seal locations need get covered up by sequence pretty quick to stay on schedule, and can't wait around another week for the arch or BI (niether couldn really tell if it was effective just by looking at it anyway) to make their site visit. Code-prescribed blower-door tests will be a start, good ammo for withholding final payment from the contractor if workmanship wasn't tight.

Dec 23, 2010 8:08 PM ET

An R is an R
by Riversong

Those who criticize the value of R-value simply don't understand how it's measured, what it measures, and why standardized measurements are necessary for objective comparisons of insulation materials. We choose insulation like we choose cars - the EPA mileage figures give a very accurate comparison between different makes and models, but have no necessary correlation to the mileage we'll get when we drive.

A large number of those who criticize the value of R-value are using fiberglass batts - the worst insulation on the market - as the baseline. Fiberglass, more than almost any other insulation material, is highly dependent on the quality of the installation and of the assembly which encloses it.

And a growing number of those who criticize the value of R-value are either hucksters of, or true believers in, spray foam insulation - and use the deceptive marketing hype that less R-value is OK because of the better air tightness. Such claims are not only scientifically wrong, but also violations of code minimum standards (and ignore the fact that very little air leakage happens in the stud cavities or rafter bays).

The current IECC is specifically written to make it easier for designers and builders, who don't know how to calculate as-built whole assembly R-values, to meet the prescriptive standards by giving different R-value requirements for continuous vs cavity (i.e. interrupted) insulation. And it also gives credit for the dynamic thermal mass advantage of heavy assemblies.

R-value remains the single most important standard for energy efficient envelope construction. Air tightness is secondary to that in new construction.

Dec 23, 2010 8:19 PM ET

Edited Dec 23, 2010 8:21 PM ET.

More Deceptive Hype
by Riversong

Dennis E. Hayes, a purveyor of pyramid homes, claims above that "the single most important factor is the shape of the building,of all possible geometric forms the pyramid has the least amount of exposed exterior surface area per square foot of heated or conditioned floor space."

This is wrong on two counts. While it's true that, all other things (such as R-value and air leakage) being equal, geometry does have an impact on the relative energy efficiency of a home, the primary determinant is the insulative capacity (R-value) of the envelope and the tightness of the structure. It's also wrong, however, in its geometrical comparison: a sphere (or hemisphere, like a dome) has the least surface area ratio to either volume or floor area. But even a cube has a lower surface area to floor area ratio than a pyramid of the same volume. A cube also has a lower surface area to volume ratio than a pyramid (counting all sides, including the base which is also a heat loss surface). And a cube of the same volume or surface area has more usable floor space than a pyramid because of the vertical walls.

Dennis also claims, on his website, that a pyramid is "250% more energy efficient than a box home." This statement is more absurd than even the claims of spray foam installers. All things being equal, a pyramid would be slightly less energy efficient than a box home and has a lot more roof to maintain (which also means more solar radiant heat gain in the summer).

His website reminds us that "figures can lie and liars can figure". Go figure!

Dec 23, 2010 8:27 PM ET

Michael Blasnik question
by Riversong


If you're still lurking, why do you use HDD60 when HDD65 is the industry standard and most homeowners set their winter thermostats to 70° or higher?

If this is to compensate for internal heat gains, they become significant only in very well insulated homes. So why would you use a different standard than everyone else?

Dec 23, 2010 8:29 PM ET

Scott Gibson's "tone"
by Riversong


While I appreciate that you included a few, though not the most relevant, of my Q&A comments on this topic, why in the world would you choose to focus on "Cobb bristles at Riversong's tone" rather than the substance of my remarks?


Dec 23, 2010 9:11 PM ET

Air Leakage vs Insulation
by Rob

Which wing is more important, the right one or the left? In the very cold climate where I live, air leaks lead to moisture and mold problems. Air leaks are also a significant source of heat loss. It is important to continuously improve air sealing in the house. There are good and better practices to use, but the homes we build get better every month.
We use insulation at a practical level, but we are constantly looking for ways to improve the insulation on a house.
Educating installers continuously makes standard practices work better.

Dec 23, 2010 9:18 PM ET

Strong Persistent Winds - A vote for Infiltration
by Rick Gowen

I am a homeowner with a home exposed to gusty and persistent winds. With a lot of work, energy audits, some Infrared analysis, I've reduced heat consumption and done a pretty credible job in cooling costs. But the winter winds have taught me that if air can infiltrate the insulation cavity, rated R values are pure fantasy. A simple low tech test exposed the problem. I drilled a hole through the drywall into the fiberglass insulation cavity and held my hand over the whole. The cold blast of air was confirmed by a simple infrared thermometer. I am fairly certain that the builder of the home felt that using rigid foam as sheathing eliminated the need for Tyvek and/or systematic sealing of the envelope.

I am facing having to remove all the siding from the house and sealing up the envelope to some acceptable level.

I am glad that there is some movement towards certifying the air tightness of construction. But there's a ton of research and standards and education of both trades and consumers required. And with tools like a blower door, it has to stay around until everything gets sealed....or you may be working with the wind as your blower door down the road.

Dec 24, 2010 11:21 AM ET

Response to Robert Riversong
by Michael Blasnik

If you're still lurking, why do you use HDD60 when HDD65 is the industry standard and most homeowners set their winter thermostats to 70° or higher?

If this is to compensate for internal heat gains, they become significant only in very well insulated homes. So why would you use a different standard than everyone else?

The short answer is that base 60 works better -- both in terms of physics and in terms of actual energy use patterns.

The use of base 65 for heating degree days was developed decades ago when internal gains were much lower and buildings were less insulated than they are today. It has only stayed in existence due to momentum. Internal gains (solar, plug/appliance, and occupants) typically provide about 10F in "free heat" (aka temperature float). You can estimate this value by dividing the Btu/hr of usable internal gains by the UA of the envelope (Btu/hr/F) -- resulting in a degrees F of float. Back when they developed HDD, appliance/plug loads were much smaller and envelope R values generally lower than they are today. In the 1940s there might have been about 5-7F of temperature float while today it's probably more like 10F.

The problems with HDD65 were recognized quite a while ago -- my 1989 ASHRAE Fundamentals includes a graph of correction factors (Cd) to apply when using HDD65 (in Chapter 28) and explains that these adjustments are needed because HDD65 don't work very well anymore.

You can use HDD65 but then you really should multiply by some sort of correction factor that might be around 0.7. I'd rather just use HDD60 and make other fixes to the simple UAdT-type equations.

But my main reason for using HDD60 is that I've found that this provide the best correlation between actual energy use and degree days in large groups of homes I've analyzed. My analysis has employed a variable-base degree day regression (similar to the PRISM model) finds the HDD base that provides the best fit between energy use and HDD. I've run this analysis for millions of homes. The average HDD balance point from this analysis was 59.9 for a gas utility in Ohio (based on 600,000 homes), 59.6F for an Illinois gas utility (634,000 homes), 59.8 from an analysis of 51 homes in upstate NY. In low income homes, I've found higher balance point temperature -- probably related to the poorly insulated and leaky housing stock and lower gain -- in an Ohio low income analysis of 1,400 homes I found 61.6F, in a program targeted at high use low income homes in Pennsylvania I found 63.4F. On the other hand, in 265 Energy Star homes in Washington State the average was 56.8F.

I'm a little surprised that you would wonder why someone wouldn't just follow along with what everyone else does ;)

Dec 24, 2010 11:25 AM ET

sorry for the typos
by Michael Blasnik

apologies for my lack of proofreading in my last post -- sure wish they had an edit button. Now to some holiday fun and cheer -- although what could be more fun than balance point temperature discussions?

Dec 24, 2010 11:29 AM ET

one more thing
by Michael Blasnik

Robert -- I also wanted to say that I agree with you 100% on your comments about R value and also for dispelling that nonsense about pyramids -- Bucky Fuller is probably turning over in his geodesic dome ;)

Dec 24, 2010 5:20 PM ET

Air Leaks vs. R-value
by Cameron Ware

Leaks are many times more important than R-value. Why? Because the R-value test does not contemplate convection in any way -- it's a hot plate conduction test that does not consider convection in any way shape or form. Any leak and you have R-0.

Try building an R-50 structure and leaving a window open. A very low R-value building with no leaks can easily trump a high value building with leaks. Why? This is not because R-value is not important -- it is because in the presense of leaks you have no R-value at all.

Dec 25, 2010 9:43 PM ET

Edited Dec 25, 2010 9:44 PM ET.

Michael Blasnik
by Riversong

You say "I'm a little surprised that you would wonder why someone wouldn't just follow along with what everyone else does."

Because without standardization of data, there's no way to compare data from difference sources.

A complete heat loss analysis must start with an assumption about set-point temperature, which determines local HDD. The industry standard, I believe, is a conservative one at 65°. Then it must separately take into account internal occupancy gains, which are part of the design program and will require an adjustment to a balance point HDD. The passive solar gains should be considered separately from balance point design heat load, since they don't effect the heat load of the occupied envelope but only reduce the required annual heat consumption.

These are simple calculations which require only a spreadsheet and which result in a much more accurate analysis. And I don't see 10° balance point adjustments until a house achieves superinsulation levels.

There's no way to compare the efficiency of different homes if we don't all talk the same language. HDD65 is the accepted common language.

P.S. If you're logged in, you get to edit your comments.

Dec 25, 2010 9:53 PM ET

Cameron Ware
by Riversong

You clearly don't understand R-values.

ASTM R-value tests are simply a measure of complete heat transfer from one side of a material to the other, and hence includes all modes of internal heat transfer: conduction, radiation and internal convection.

Unless you open all the windows of a house and the wind's blowing, you will not have the heat loss equivalent of zero R.

Dec 25, 2010 9:56 PM ET

Rick Gowen
by Riversong

When you say, "if air can infiltrate the insulation cavity, rated R values are pure fantasy" you're generalizing from your experience with fiberglass insulation - arguably the worst insulation on the market.

No other insulation is as vulnerable to internal convection as fiberglass, either batts or loose fill. But it's the cheapest, and you get exactly what you pay for.

Dec 26, 2010 1:52 PM ET

response to Robert Riversong
by Michael Blasnik

There's no way to compare the efficiency of different homes if we don't all talk the same language. HDD65 is the accepted common language.

But we aren't discussing comparisons between homes -- we are discussing energy use calculation methods. In the case of standard comparisons such as using Btu/sqft/HDD65 I think they usually aren't very useful and often misused but if you want to do something like that it's fine to use HDD65 if you want -- the conclusions aren't really affected as long as you use the same HDD base.

A complete heat loss analysis must start with an assumption about set-point temperature, which determines local HDD. The industry standard, I believe, is a conservative one at 65°.

This is false -- the set point does not determine HDD and HDD65 has not been used to imply a 65F conservative setpoint.

Then it must separately take into account internal occupancy gains, which are part of the design program and will require an adjustment to a balance point HDD. The passive solar gains should be considered separately from balance point design heat load, since they don't effect the heat load of the occupied envelope but only reduce the required annual heat consumption.

This is incorrect. The balance point is the setpoint minus the temperature float. The temperature float may be thought of as the increase in the indoor temperature over the outdoor when the heating system isn't used. That temperature would depend on the internal gains as well as solar gains. I agree that you can take a different approach and calculate a solar heating fraction and come up with similar energy use estimates but the reality is that the solar gains affect the balance point. That is why the energy use data I analyzed indicates that the average balance point is about 60F -- averaged across more than 1 million homes.

These are simple calculations which require only a spreadsheet and which result in a much more accurate analysis. And I don't see 10° balance point adjustments until a house achieves superinsulation levels.

Your decision to treat solar gains separately will lead to lower estimates of temperature float, but the actual energy use patterns of homes disagrees with you.

But getting back to the start of this discussion. You took exception to my use of HDD60 in an energy loss calculation. You have also disagreed with someone else who used HDD65 but multiplied by 0.75. Apparently you also disagree with ASHRAE Fundamentals (1989, which still included HDD energy estimating methods) which applied a correction factor to HDD65 to reflect the lower true balance point. But if you use HDD65 without any correction then you will dramatically over-estimate the energy use of most homes.

Dec 26, 2010 6:10 PM ET

by Cameron Ware

You said: "ASTM R-value tests are simply a measure of complete heat transfer from one side of a material to the other, and hence includes all modes of internal heat transfer: conduction, radiation and internal convection."

I don't buy it. R-value is only part of what's happening in the real world. R-value is a conduction hot plate test. Where do you get the idea that the R-value calculation/test somehow includes convection? Internal convection as you call it means very little in the absense of anything initiating convection like a pressure differential or wind etc. Convective loops between fibers etc. in a laboratory do not cover it -- particulary inside a machine in steady state. Please send some laboratory data that says a test of conduction somehow includes convection.

Where do you get the idea that the conduction R-value test somehow includes radiation? Just exactly where is the radiation coming from? If you are saying the hot plate is modeling the suns radiation? The test does not include either a convective or radiative source.

Dec 27, 2010 12:21 AM ET

Michael Blasnik sign up for an account, then sign in to edit
by aj builder, Upstate NY Zone 6a

Once you have an account and sign in each time, you will have the ability to edit and post attachments.

Please join the discusion here and post often. Learn to post sometimes and leave it at that. You will stay younger longer.


PS... would be nice to have you start a thread in Q&A and teach us all the basics of what you are discusing here on this thread.

Dec 27, 2010 6:14 AM ET

Response to Cameron
by Martin Holladay

You wrote, "R-value is a conduction hot plate test. Where do you get the idea that the R-value calculation/test somehow includes convection?"

The ASTM R-value tests measure heat flow via all three heat transfer mechanisms: conduction, convection, and radiation. To learn more, see Understanding R-Value.

Dec 27, 2010 1:00 PM ET

Fascinating Discusion, But What About Radiant Heat?
by Anonymous

I've been wondering about R values for years, mostly because I have seen them not work: rooms staying too cold or too hot, etc. Once I noted that a room felt cold even though air temperatures were fine and saw that this room was adjacent to a massive, very cold, masonry structure off of its North wall. In winter folks avoided the room; In summer they congregated there.

A second area of concern is air stratification by temperature. In winter, the hot air is pressing hard against the ceiling where conduction and convection work hard on cooling it, but in summer quite the other way: the cool air is down where we want it. If we eliminate leaks from the bottom 2/3 of a room structure, the cool, dry air stays put.

Lastly, conduction loss/gain rate depends upon the square of the difference in temperature. In winter in the North, 70 degrees of delta is common. In the summer, 30 degrees of delta T is common at ground level, and with proper attic ventilation and radiant barrier, also common at the ceiling. It would seem, then, that R-value is often over-rated, and that other tools should be considered with equal weight. I have often seen the simple application of extra R-value have no effect on comfort or cost.

Dec 27, 2010 1:52 PM ET

Edited Dec 27, 2010 1:53 PM ET.

Response to Anonymous
by Martin Holladay

Whether or not R-value is "overrated" obviously depends on the person assessing the importance of insulation R-value.

If we are talking about a person with a basic understanding of the principles of building science, they probably have a fairly good assessment of the importance of R-value.

However, if there is a person who doesn't understand air leakage or pressure dynamics, they may mistakenly think that thick fiberglass batts over a very leaky ceiling will keep them warm.

You can't just choose one idea and say, "this is all that matters." It doesn't make any sense to buy a $2,000 window for your living room if there are big holes in the floor and ceiling.

You have to think about all the elements of a house. Ideally, your house will be as airtight as you can make it; it will be equipped with above-code levels of insulation; it will have excellent windows and doors; and it will have a mechanical ventilation system.

Dec 27, 2010 6:22 PM ET

by Cameron Ware

Martin, I appreciate your directing me to your article. I've read it. I would appreciate it if you could share a link or article from a testing organization or something like ORNL etc. that states that R-value is more than conduction based. If I am wrong about this I would really like to know. thanks.

Dec 28, 2010 12:04 AM ET

Thanks Scott for starting this debate
by Thinkitthrough

Very interesting debate. It highlights the need for more research into the effects of air infiltration into and through an insulated wall. With the right software (or even using a spreadsheet) a comparison could be made between a perfectly insulated R20, R40, R60, .... wall or attic and one that had a known amount of air infiltration. This could be stepped from a little to a lot and the effect on the R value calculated. This is research that is well worth doing. We all have our "opinions" about the importance of air sealing, but what's needed is solid data. If there are any Architects or Engineers out there looking for a juicy project, this one should top the list.

Dec 28, 2010 6:37 AM ET

More information for Cameron Ware
by Martin Holladay

Among my sources is David Yarbrough, a research engineer at R&D Services in Cookeville, TN, and a nationally recognized expert in heat flow and R-value measurements. Yarbrough is chairman of the Chemical Engineering Department at Tennessee Technological Institute and on the research staff at Oak Ridge National Laboratory. He also served for ten years as Chairman of the Reflective Insulation ASTM Committee.

Yarbrough has written, "The R-value used to describe thermal insulating products includes heat being transferred by all three mechanisms. The term used in the thermal insulation community is ‘apparent thermal conductivity.’ A formal definition for apparent thermal conductivity is contained in document C168 published by the American Society for Testing and Materials. The term is applied to situations involving the simultaneous flow of heat by all three transport mechanisms. The statement ‘R-values are measures of conductive thermal resistance’ is incorrect, if it implies a limitation, since ‘R-value’ includes radiation and convection when they are present."

Dec 28, 2010 1:47 PM ET

Simple question. Simple answer.
by Kelly Merke

Without question, 'air leaks' (convective heat loss) trumps 'thermal loss' (conductive heat loss). Picture yourself inside a perfect box built from rigid styrofoam in the middle of a frozen lake somewhere. Now you could imagine that box with 50% of it riddled with air holes and see the difference that would make... given equal portions of air and insulation... it's no competition. But you don't have to go that far, starting from a perfect box again... after just a few holes, and you will fully appreciate the impact. Air leakage is more important.

Dec 28, 2010 2:48 PM ET

by Cameron Ware

Martin, I've read quite a bit of David Yarbrough's material. Even more now that you've suggested it. In doing some other research, I did find the following out there which may lead one to conclude changes may be under way:

A definition of R-value based on apparent thermal conductivity has been proposed in document C168 published by the American Society for Testing and Materials. This describes heat being transferred by all three mechanisms—conduction, radiation, and convection.

I fully agree that R-value is important and that convection and radiation are real and present when one enters the real world. Perhaps this proposed new definition (which is somewhat mathematically rigorous) is based on the fact that today's R-value does not appropriately describe all that it should. Kelly's post somewhat summarizes my belief -- if we take Kelly's styrofoam box and give it twice the thickness - the lab will say it has twice the R-value and yet given a few small holes one would still freeze to death.

Register for a free account and join the conversation

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