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

How Much Air Leakage in Your Home Is Too Much?

Whatever your air tightness goal, ACH50 isn’t the best unit to measure it

Image 1 of 3
To determine a home's air leakage rate, energy auditors run a blower-door test.
Image Credit: Energy Vanguard
To determine a home's air leakage rate, energy auditors run a blower-door test.
Image Credit: Energy Vanguard
The blower door is all set up and ready to operate.
Image Credit: Energy Vanguard
Eliminate the big holes in the building envelope: this 700-square-inch hole was put in by a contractor to vent the refrigerator to the attic.
Image Credit: Energy Vanguard

Whether you want to build a new home or fix an old one, the way to ensure that you get the best performance is to do the building envelope right. That means installing the right amount of insulation and installing it well, and it means having an air barrier with minimal leakage. But how do you know when you’ve done enough air sealing? How tight is tight enough?

I get asked this question every once in a while, and I love to talk about the measures for air leakage anyway, so let’s dive in. First, of course, you have to be able to measure how much air leakage the house has. That’s what a blower door is for.

About air leakage units

First, you have to choose how you want to specify the air leakage. The most common unit used by blower door operators is ACH50, which stands for Air Changes per Hour at 50 Pascals. I prefer cfm50 per square foot of building envelope, or better, cfm50 per hundred square feet of building envelope (sfbe). (One cfm50 is a cubic foot per minute with a pressure difference of 50 Pascals between inside and outside.)

The two reasons for that choice are that (i) air leakage happens at the surface, not in the volume, and (ii) it’s the best unit, in my opinion, to express what a Blower Door is really good at — measuring the amount of air moving across the building envelope at elevated pressure.

Please don’t talk to me about ACHnat (‘Natural’ Air Changes per Hour). I loathe that measure! If you’re using a blower door, you can’t measure it; the only ones who can are researchers using tracer gas analysis.

Now we’re ready to discuss the actual question: How much air leakage should you aim for? OK, we’re not really there yet. I lied. First, we have to know about your house. Are you building new or fixing an existing home? If the latter, what’s your budget and how complex is the building envelope?

New homes

Let’s start with the easier one: new construction. The first rule here is that a house can never be too tight. The Passive House program takes houses about as far as you can go with air tightness, and their threshold is 0.6 ACH50. I tested a net-zero house a couple of weeks ago that had less than 200 cfm50 of air leakage, or about 0.5 ACH50. That’s really tight!

A target that’s more achievable for anyone — and which the 2012 International Energy Conservation Code (IECC) will require for most climate zones (CZ 3 and higher) — is 3 ACH50. That’s also the level that Joe Lstiburek identified as a good target in his great article on blower door testing new homes, Just Right and Airtight.

That number — 3 ACH50 — translates to about 0.25 cfm50 per square foot of envelope, or 25 cfm50 per hundred square feet of envelope. Since roofers have already abbreviated 100 square feet as 1 square, I like the latter form the best. It gets it into a whole number form and is easy to remember. Get your blower door number down to 25 cfm50 per square (or below), and you’ve got a tight house. The house I built ten years ago came in at 14 cfm50 per square (1.7 ACH50).

But wait. I can hear you wondering, “Is he saying that this one number, 25 cfm50 per square, is good for every climate zone?” My friend David Butler says that we should consider cost effectiveness and not just apply a “one-size-fits-all” threshold. “We should always ask ourselves which combination of efficiency measures will provide the lowest energy usage within a given budget,” he wrote recently. “The cost-benefit of achieving 3.0 ACH50 would look very different in Buffalo than in Tucson, on multiple levels.”

He does have a point, He went on to explain that “in hot climates, I’d rather fight to bring ducts inside (always challenge in non-basement homes) than to arbitrarily specify a 3-ACH50, especially in dry climates where latent loads are negligible.”

Lstiburek, a Canadian who lives in New England, thinks the roughly equivalent 3 ACH50 is good because it’s achievable, and it solves a lot of comfort and efficiency problems associated with air leakage. I understand David’s point, but here in Georgia, home builders now have to pass a blower-door test and get their houses 7 ACH50 or below. What we’re finding is that a significant number of builders are getting down to 3 ACH50 in their attempt to meet the code target. Once builders get their crews trained, 3 ACH50 should cost them about the same as 5 or 7 ACH50. Now, if we can just get them to change the target to 25 cfm50 per square!

Existing homes

This can of worms is really too big for the little article I’m writing here, but let me try to shed at least a little light on it. The amount of air sealing you’re going to be able to do in an existing home is limited — unless your budget isn’t. Generally, with a good attack on the holes, you can get about a 20% to 30% reduction in your Blower Door number (whether you specify it in cfm50, ACH50, or cfm50/sfbe). Sometimes you can get much more when you add surface area or volume by moving the building envelope.

The first thing you want to know here, though, is: How leaky is your home to start with? If you’re already at 25 cfm50 per square, it’s going to be really hard to get a 20% reduction. If you’re at 100 cfm50 per square, it should be a snap to reduce it to 75 cfm50 per square or even lower. The higher that starting number is, the more big holes you probably have in your building envelope.

If you’re starting at 100 cfm50 per square, however, don’t count on getting down to 25 cfm50 per square unless you’re doing a deep-energy retrofit. Sometimes deep-energy retrofits are called deep pockets retrofits, for obvious reasons. They cost a fortune! Check out Martin Holladay’s recent post called The High Cost of Deep Energy Retrofits here at GBA.

I’ve done a number of air-sealing jobs when I was in the contracting business, and the results varied. Sometimes it was a frustrating 5-10% reduction, and other times it was an easy 20%. The most reduction I ever achieved was about 40% in a house that started with about 120 cfm50 per square. We encapsulated the crawl space and did a lot of work on the kneewalls and can lights on the upper floor.


If you’re looking to make your home tight and you really want some numbers to go after, here are my recommendations:

New Homes:  25 cfm50 per hundred square feet of building envelope (or 3 ACH50)

Existing Homes:  Get a blower door test, and try to reduce the number by 20% or more.

Allison Bailes of Decatur, Georgia, is a RESNET-accredited energy consultant, trainer, and the author of the Energy Vanguard blog.


  1. davidmeiland | | #1

    Not a bad plan
    but of course it calls for measuring completely enough to determine surface area. That can be a bit time consuming in houses with lots of cathedral ceilings or other detail. The Washington State Energy Code is using a formula based on square footage that roughly equates to 5ACH50 (houses with high ceilings are penalized by this approach), and I assume it's to avoid making testers measure surface area. It's quick to go through most houses and get square footage, and then get an average ceiling height so that you have approximate volume. In an average house I think that a number that's somewhat under 1CFM50 per square foot is pretty good.

  2. ewoodhouse | | #2

    Maybe I'm missing something...
    but it seems like measuring based on surface area (or volume, for that matter) makes less sense than measuring based on living area. A building with a large surface-area to floor-area ratio is going to use more energy than one with a simpler shape and an equal cfm50 per square, won't it? I guess that's fine if what you're aiming to measure is quality of construction alone - but it seems like the design certainly has an impact.

  3. davidmeiland | | #3

    Depends what you're measuring
    A test result based on surface area tends to measure construction quality, both that of the builder on site, and of the manufactured products like doors and windows, dampers, etc. A test based on square footage is a dumbed-down version of the same thing. A test based on volume is useful if you want to estimate air changes for determining ventilation requirements, or heat loss via air leakage.

  4. albertrooks | | #4

    It's really not that hard to get well below 3ACH 50
    And I mean for new construction. Retrofit is a "whole 'nother kettle of fish"

    After countless (seriously) new construction passive house projects that we've helped with (in a small way). It can be done without question.

    The first (of yours at less than 0.6ach) is the hardest.

    The third (of yours at less than 0.6ach) passes by without notice or fanfare: It has become a matter of course.

    To be fair, those projects go at it from a different design perspective: A really good lid, no can lights, Floor connected to wall, wall connected to lid: The unbroken line (air barrier wise). The term "obsessiveness" might come up in some circles, but it's really just a "goal" and good workmanship towards the goal.

    If these excessive Passive House folks can do it, it can't be that tough. Often it's really not. It's in the design & planning stage that you can get rid of most of the big "holes" and "disconnections". After that, it's just seeking out and keeping the small "holes" filled and making sure that new holes aren't created down stream.

    There are a lot of good materials available to accomplish this with: Foams, Tapes, Gaskets, Sheet Goods, Membranes, Sealants. They are all reasonable ways to accomplish the goal.

    Beating 3 ACH 50 is about as easy as reaching it. If you really care to...

  5. GBA Editor
    Allison A. Bailes III, PhD | | #5

    Response to David Mieland
    Yes, calculating the surface area is difficult in more complex houses, but so is volume when you have insulated rooflines. If you're doing energy modeling on the home for a load calculation or HERS rating anyway, then you'll have all the info. For existing homes where you just want to get in and out as quickly as possible, ACH50 is adequate.

    In your response to Eric, you wrote:

    A test based on volume is useful if you want to estimate air changes for determining ventilation requirements...

    See the article by Joe Lstiburek that I mentioned in the section on New Homes or my linked article on what a Blower Door is good at. The amount of air moving through the Blower Door at elevated pressure doesn't extrapolate very well to "natural" pressures, so a Blower Door doesn't help a whole lot with determining ventilation requirements, in my opinion.

  6. GBA Editor
    Allison A. Bailes III, PhD | | #6

    Response to Eric Woodhouse
    We're talking about air leakage here, not general energy usage, and a Blower Door is only good at telling us how much air leakage a house has. The reason for bringing the size of the house into it is so you can compare different sizes houses to each other and to a standard that's independent of size (e.g., a code requirement for 3 ACH50). I'm not sure exactly what you mean about the design being important, but a Blower Door test is all about measuring the quality of construction in new homes.

    Am I missing your point?

  7. GBA Editor
    Allison A. Bailes III, PhD | | #7

    Response to albert rooks
    I'm with you, Albert, but the only house I ever built was a custom home that took me two years. I hit 14 cfm50 per square (1.7 ACH50) without ever having seen a Blower Door test. I think any custom builder who takes the time to figure it out can easily do better than 3 ACH50. Joe Lstiburek, who's been doing this work for decades and has worked with production builders, makes a pretty convincing case that 3 ACH50 is a good target, however, and that's why I settled on that number.

  8. michael anschel | | #8

    Nice job
    Nice job with this post. A good 101 for foks just getting their feet wet with the subject. It would be great if the folks at BPI and ResNet would heed your advice and make the switch over to CFM per SF. (I get the idea behind the 'square' but I'm not sure if representing the home in numbers like 23.578 is really valuable. Anyone not involved with roofing or carpet is much more comfortable with SF.)

    HOWEVER, I think that what is missing in this whole discussion is a conversation about where that leaking takes place. A buiding forensics expert (the work that provides Joe with enough dough for that Porche) will quickly find the point where air is leaking into the enclosure and causing damage. Without being specific about location we need to be cautious about making blanket statements about the acceptable volume of air that passes through the enclosure (not connected to mechanical or intentional passive components)

    One additional point to consider. The push for air tightness in the South creates a building environment that demands the use of AC. Promoting a policy that relies entirely on mechanical delivery of conditioned air in lieu of smart building architecture that can effectively keep a home in the hotest of climates comfortable seems slightly irresponsible.

    I think the incesent desire to oversimplify a structure and set of systems for the border line literate trades needs to stop. Keep it complicated and variable.

    Rock on-

  9. GBA Editor
    Allison A. Bailes III, PhD | | #9

    Response to Michael Anschel
    Good points, Michael! Yes, the location of leakage is critical, and a wall between the house and an attached garage, for example, needs to be as tight as a vacuum chamber. But Blower Door operators are measuring the leakage from the whole house. That tool can't separate out the various components.

    Regarding the South, I'd say you're ascribing the demand for air conditioning to the wrong cause. People here want AC in leaky houses, too, and comfort is the driver, not how tight the house is. I do agree that we should build houses so they can be comfortable without AC, but that's a design issue, getting windows in place for cross ventilation and ceiling fans and porches.

    It'd be great if every budget included enough money so we could engineer houses the way we do commercial buildings and cars and computers. Then we'd have to "keep it complicated and variable."

  10. michael anschel | | #10

    Keeping it Complex
    I think at GBA you should take up that challenge to go beyond simple measurements. You are in a great position (and have enough brain power) to take up the challenge. (You also have enough perspective, I think, to understand and explain the nuanced complexities involved)

    I hear what you are saying about the leaky v non leaky home demand for AC. My point was that point to to 3 ACH as the metric we are not doing the home design justice, and that when we are discussing appropriate levels of tightness there is an argument to be made for spending $$ on smart design in lieu of mechnical systems. (the $12-15K spent on mechanicals will more than cover the design and modeling fees for smart ventilation and cooling systems)

    Funny that we will pay hundreds of thousands of dollars for a poorly engineered product, but will spend hours pouring over the details of a $800 computer or $30K car. Sad sad sad.

  11. GBA Editor
    Allison A. Bailes III, PhD | | #11

    Nuanced complexities - Yes!
    You're right, Michael. (Wow, did I just say that?) I'll take up this challenge and see if maybe we can get the subject of air leakage in homes up to where it needs to be. And yes, it is sad, sad, sad about misplaced priorities. Since we do design (both architecture & HVAC), we deal with this issue all the time. Some people do really get it, but they're in the minority.

  12. ewoodhouse | | #12

    Response to Allison A. Bailes III, PhD
    We may be missing each other's point, Allison. Part of the issue, I guess, is what you call "size" when you're comparing one house to another. If I have a 2,000 s.f. house with 9' ceilings, and another house of identical design but with 15' ceilings, the latter house will obviously have far more surface area, and more volume. But the two houses will appear equally tight when looking at ACH or cfm per square... and to my way of thinking, the 9' ceiling house is the "tighter" of the two.

  13. GBA Editor
    Allison A. Bailes III, PhD | | #13

    Yes, size matters.
    Eric, I think I see what you're getting at now, and it's basically the difference between absolute and normalized leakage. Yes, in the example you gave, the house with 9' ceilings will have less absolute leakage (cfm50) if they have equal normalized leakages (cfm50/sf).

    Energy codes and green building programs, though, specify normalized leakage to take the size of the house out of the equation (though LEED for Homes and now ENERGY STAR Version 3 add it back in somewhere else). If they specified absolute leakage thresholds, say 1000 cfm50 (to make up a number), builders of larger houses would have to do much more air-sealing to hit the target, and the small homes might not have to do much at all.

    So, yes, size matters.

  14. user-1005777 | | #14

    Did we do the right things?
    We have a mini home just over 1000 square feet. We had a blower door test done in November of 2007. It showed 6.9 ACH50. We have taken all the trim off the inside of the house and foamed or caulked every joint in the structure that we can get to and even sealed light fixtures. The blower door test done this last December showed 3.1 ACH50. Should we have attacked the envelope from the outside? That probably would have been much more difficult, and costly. Does it matter?
    Wouldn't the energy auditor have to know the volume of the house to figure the ACH figure?

  15. GBA Editor
    Allison A. Bailes III, PhD | | #15

    Response to Roger Williams
    If you cut the infiltration rate by more than half on an existing home, I'd say that you did the right things. That kind of reduction isn't easy for existing homes. And getting down to 3.1 ACH50 is also good, so be happy with where you're at.

    Yes, the volume figures into the ACH50 result. It's the cfm50 (the amount of air flowing through the Blower Door fan while the house is depressurized to 50 Pa) times 60 (to convert minutes to hours) and then divided by the volume (to convert cubic feet to air changes).

  16. davidmeiland | | #16

    On an existing home, you do what you can. It would be more effective to remove all your siding and trim and seal leaks at the sheathing, but in most cases you aren't going to do that. I agree with Allison that getting to 3ACH50 with interior retrofit measures is quite an accomplishment. I assume you had some very large holes that you were able to seal.

  17. HhsHwZFSCs | | #17

    1984 Blower Door Test results
    I have about 3100 sq feet and completed my first blower door test. The result was 3300CFM50 or 8 ACH50 so I was told by the vendor. This test was completed in 2011. I have sealed all the kneewalls and floor baseboard areas. Also sealed the attic, Garage, resided two outside walls and resheathed. In 2012 the test result was 2864cfm50 (estimated number) This equates to 6.8 ACH50 I thought the numbers would be lower than these. Any other areas to focus to lower this number. I have addressed the fireplace with rigid foam and sealed all the sill plate areas in the basement. Any thoughts would be greatly appreciated.

  18. USEuG9CHFL | | #18

    nobody mentioned this
    One of the most important things you can use a blower door for is "zonals". Using the manometer, you can isolate rooms, light fixtures, the wall between the garage and the living space, the attic(the most important). I work on older homes and making sure the attic is as unattached to the home as possible is our main goal. Older homes usually don't have enough insulation in the attic or venting. This tends to keep the attic warm enough to inhibit mold growth. Throw down insulation without air sealing enough and during the Fall or Spring when you have some humidity in the air compared to the dead of Winter and now those air leaks are producing mold. This winter was the worst here in PA because it never got cold enough out to really dry out the attics. I have to throw this out there because a little bit of knowledge can get you in a lot of trouble.
    Anyway the manometer will tell you how connected the attic is to the house and you shouldn't be done airsealing until that manometer reads the attic at +49 to +50pascals with relation to the house.

  19. heidner | | #19

    RE:1984 Blower Door Test results.
    GBA has a number of articles and blog postings on how to find and go after the air leaks. If you are having a blower test done, make sure you have a thermograph imaging done while the testing is ongoing... and have some smoke pens handy. You should be able to find large leaks quickly.

    You didn't mention if you have recessed lights. If so, you may wish to consider alternative lights and/or verify the recessed cans can be sealed up. (NOT ALL recessed lights are certified for insulation and air sealing)!!

    Check behind sinks, toilets, dishwashers, and appliances. Sometimes there are gaping holes where the plumbing enters the wall.

    Look for leaks in sofits, between floor wire/pipe chases, etc. If you have forced air -- is the ducting outside of the heated space and is it leaking?

    If you are doing this your self and trying to tighten the house -- I would consider buying a simple dwyer 460 air meter, some used furnace blowers and build a crude blowr assembly to depressurize the house while you use a smoke pen. Then when you've found and sealed more leaks -- have it retested by a certified blower/energy audit professional.

  20. user-648002 | | #20

    Skin Calculations
    We have always looked at skin tightness this way. The size of the hole per 100 sqft of wall = leakage ratio or LR. We would get results down in the 2" per 100 sqft of wall or an LR of 2. I think this is the best way to talk about how tight your building is.

  21. user-1135248 | | #21

    alphabet soup
    [Is it better to follow up on an older thread, or start a new
    Q&A for this? Let's try this first...]

    Before my retrofit I did a quick-n-dirty volume and surface area
    calculation on my relatively simple Cape, and have just redone it
    a bit more accurately. I'm now reading up on blower-door testing
    because I expect that eventually the place will get one, and the
    confusion over which numbers are which, better, meaningful, BS,
    etc is *profoundly* irritating. Anybody who comes along talking
    about CFM leakage per square foot of *living area* needs to be
    shown the door and by that I mean as a prompt exit assisted by my
    foot, not the door to mount the blower frame in ... then you've got
    the problem of Equivalent Leakage Area over the entire envelope vs.
    MLR which only considers above-grade envelope surface, and for the
    latter I probably want to add in the below-grade area of my basement
    bulkhead door to make things more realistically accurate.

    By breaking the house into sections -- attic, living floors, basement --
    I know all my volumes and areas, and derived an easy reference point
    that 1 ACH for the whole enclosed space, just shy of 15,000 cubic feet,
    would be 248 CFM. That's regardless of pressure, of course, and that
    volume includes interior walls, closets, the volume of the remaining
    part of the chimney inside, etc. But what is the "geek with the blower
    door" going to say about what any CFM reading actually means in terms
    of CFM/sf? How many "square feet" should be considered, and where are
    they located?? How does anyone know how much "living" I'd do in my
    unfinished basement before deciding whether to include that in any
    such figures?

    And then in a heavily-insulated house one has to consider where the
    measurements were taken -- inside or out? I started with my interior
    dimensions because that's effectively the volume where air will be
    changed, measuring heights floor-surface to floor-surface to include
    all of those thicknesses. But if I did the same measurement outside
    now, with the exterior bulked up 5 or 6 inches more with polyiso and
    furring and siding, everything would be different.

    So really, "CFM50/sf" has to be *thoroughly* qualified and explained
    in any test results if it's going to mean jack. And it likely won't
    account for *where* leakage sites actually exist, i.e. would they
    contribute to stack effect more or less. At least "ACH50" makes more
    sense because it keeps everything in terms of volume, as long as you
    know how that volume was determined.

    The poor sod who eventually shows up here to test the place is probably
    going to get a real earful about this, as he's given an envelope with
    *no* penetrations to test. I mean, why should I leave my HRV vents
    open if we're testing the *envelope*?? I know how many CFM that is


Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |