Blower Door Report: CFM50 Is Measured, ACH50 Is Not

Two auditors can leave the same house holding the same fan reading and write down different ACH50 figures, without either of them having made a mistake. That is not a flaw in the blower door. ACH50 is a calculation performed on a measurement, and the inputs to that calculation are conditions of the test, recorded on the report, rather than facts about your walls.

The measurement is CFM50 — cubic feet per minute of air the fan has to move to hold the house at a 50 pascal pressure difference against outdoors. Everything else on the report is downstream of it.

The rules below come from ANSI/RESNET/ICC 380, the standard a rating-grade blower door test is run under. RESNET publishes the 2019 edition in full as a free PDF, and every passage quoted below was read in that copy on 23 August 2026. The edition now referenced for ratings is 380-2022, which is sold rather than published; RESNET's own technical bulletin of 4 August 2026 walks through the same chapter of the 2022 text and cites the same section numbers where the two overlap. Equation numbers are given as the 2019 PDF prints them.

The denominator nobody quotes you

The conversion is one line — Equation 7a, in section 4.5.2:

ACH50 = CFM50 x 60 / Infiltration Volume in cubic feet

Infiltration volume is not floor area times ceiling height. The standard defines it as the sum of the conditioned space volume and adjacent volumes — a crawlspace, an attic, a basement — which count only "when the access doors or hatches between the crawlspace and Conditioned Space Volume are open during the enclosure airtightness test," and in equivalent words for the other two.

So the denominator depends on which doors stood open, and that is not left to taste. Section 4.2.5 says all doors between the conditioned space and a basement shall be opened, with an exception at 4.2.5.1.1: where the floor above the basement is air sealed and insulated, those doors shall be closed and the basement drops out of the volume altogether. Sections 4.2.3 and 4.2.4 do the same job for crawlspaces and attics, and turn on the same question of where the air barrier and the insulation actually sit.

Watch what that does. Take a two-storey house with 2,000 sq ft of conditioned floor area on a 1,000 sq ft footprint, 8 ft ceilings, over a basement with 7.5 ft of headroom:

The same fan reading, a third off the headline figure.

Holding CFM50 fixed across those two rows overstates the effect on purpose: with the basement inside, its own leaks are being measured too, so the numerator would rise as well. But the denominator moves 47% and the numerator moves far less, so the direction holds. Which is why section 4.3.1.6 requires the infiltration volume to be recorded whenever a result will be reported as ACH50 — 4.3.2.14 repeats it for an attached unit — and 4.2.3 through 4.2.5 require the position of every crawlspace, attic and basement door and hatch to be recorded alongside it. A report that prints ACH50 with no volume beside it has left out the half you would need to check it.

CFM50 has the opposite property. Nobody can redefine it. If you are going to hold a contractor to a before-and-after improvement, hold them to CFM50.

The 1.1 that only appears on the short test

RESNET 380 offers two procedures. A one-point test brings the house to 50 pascals (within ±3 Pa) once and records the flow. A multi-point test sweeps a series of pressures and fits a curve, producing the coefficient C and exponent n, from which CFM50 is calculated.

Then section 4.5.1 does something unexpected:

If the results of the building or Dwelling Unit enclosure air leakage test are to be used for conducting an energy rating or assessing compliance with a building or Dwelling Unit enclosure leakage limit, then the corrected airflow determined using a one-point test shall be adjusted using Equation 5a or 5b.

Adjusted CFM50 = 1.1 x Corrected CFM50

The same clause closes: "Other applications of building or Dwelling Unit enclosure air leakage testing and the results of multi-point testing do not require the corrections in this section." RESNET's 2026 bulletin puts the current rule in plainer words — a single-point result "must also be adjusted using a 10 percent single-point adjustment factor after the correction for temperature and elevation has been made (4.5.1)" — so this is live practice rather than a 2019 leftover.

The one-point test therefore carries a built-in 10% penalty whenever the result is used for a rating or checked against a leakage limit. The 2,640 CFM50 in the table above is what a one-point measurement of 2,400 becomes. Run multi-point on the same house and 2,400 stands: 9.0 ACH50 rather than 9.9 on the smaller volume. That 0.9 gap is larger than the entire tightening the 2024 IECC asked of climate zones 6 through 8, which moved from 3.0 to 2.5. If a contractor's guarantee is written in ACH50 — and some are — the procedure matters as much as the sealing.

Two more clauses worth knowing. "Corrected" CFM50 is not a figure of speech: section 4.4.1.5 sends the calculation to section 9 of ASTM E779 for the density and viscosity adjustments, so the ASTM method sits inside the RESNET path rather than beside it as an alternative.

The other clause covers the house that beats the fan. Where 50 pascals cannot be reached, 4.4.1.2 lets the tester record the highest pressure the equipment did achieve, against a hard floor: "A minimum of 15 Pa (0.06 in. H2O) must be induced across the enclosure for the test to be valid." Section 4.4.1.4 then scales that reading up to a nominal CFM50 with Equation 1a, which raises the ratio of 50 to the measured pressure to the power 0.65. On a very leaky house with one fan, that floor is reachable. Ask what pressure was achieved.

A note on the standards themselves, because audit reports cite them loosely. ASTM E779 is Determining Air Leakage Rate by Fan Pressurization, the multi-point method. E1827 is Determining Airtightness of Buildings Using an Orifice Blower Door, which is where the two-point procedure lives. E3158 is Measuring the Air Leakage Rate of a Large or Multizone Building, written for commercial work and not the residential test. The code clause that requires the test names ANSI/RESNET/ICC 380, ASTM E779 and ASTM E1827 — that is the wording at N1102.4.1.2 (R402.4.1.2) in the widely adopted text carried on UpCodes, read 23 August 2026.

Setup decisions that move the number before anyone reads it

Section 4.2 of RESNET 380 is a long list of positions things must be in, and several items pull against what people assume.

Chimney dampers and combustion-air inlets on solid fuel appliances are closed (4.2.7). Combustion appliance flue vents are not: they are left in their as-found position (4.2.8), and undampered combustion air and make-up air openings stay open (4.2.11.5). Sealing them would flatter the result, and the point is to measure the house as it stands, holes included.

• Interior doors between rooms inside the conditioned space: open (4.2.6).
• Anything able to move air across the enclosure — clothes dryer, attic and crawlspace fans, bath and kitchen exhaust, air handler, ventilation fan — off (4.2.9).
• Plumbing drains with empty p-traps: sealed or filled with water (4.2.16). A dry floor drain is an open pipe to the sewer.
• Motorized dampers to outside or to unconditioned space: closed, and not sealed any further (4.2.10.2). Non-motorized ones are left as found (4.2.10.1).

Ventilation openings get split by how the fan is controlled rather than by what room it is in. Openings serving continuously operating exhaust or continuous dwelling-unit ventilation are sealed at the exterior of the enclosure where conditions allow (4.2.11.3, 4.2.11.4). Openings serving intermittent local exhaust are left open, and intermittent dwelling-unit ventilation openings are not sealed at all (4.2.11.1, 4.2.11.2). Two bath fans in one house can end up treated differently for that reason alone.

For an attached unit — section 4.3.2 is the compartmentalization procedure — there is a penalty clause with a number attached. The door the blower door is mounted in has to be inspected for a seal, and where one is absent or improperly installed, section 4.3.2.4 requires 140 CFM50 to be added to the measured airflow, with the adjustment and the condition of the seal documented in the final report. RESNET's standing interpretation of that clause (IR 380-2019-009 / 380-2022-002, read 23 August 2026) settles the argument about the bottom of the door: the seal must be present on all four sides, and a door shoe or brush sweep counts.

Put that in the units the report uses. An 1,100 sq ft unit with 8 ft ceilings holds 8,800 cu ft, so 140 x 60 / 8,800 is roughly 0.95 ACH50 added before anything else is computed.

None of this is about catching a dishonest tester. It is about knowing that "as found" is a defined state, and that a test run with the fireplace damper open measured something other than your envelope.

Which line you are being compared against depends on the zone and the code year

Here the sloppy version of this article would give a single number. There isn't one.

The Department of Energy's own analysis of the 2024 IECC lays out the change, in tables D.6 and D.7 of Energy Savings Analysis: 2024 IECC for Residential Buildings (energycodes.gov, read 23 August 2026). These are the code's prescriptive limits rather than assumptions the analysts chose: the paragraphs above each table name the approved code changes the values came from, REPI-063 and REPI-064, both headed "Prescriptive Air Leakage."

Two editions, two different answers for the same house in Minnesota. And a third possibility, since plenty of states still enforce the 2018 IECC or an amended version of something older — DOE keeps the adoption picture by state at energycodes.gov/status.

There is an alternative metric too, and it exists because a volume-based limit is hard on small units: an apartment has a lot of shared and exterior surface for the cubic feet behind it. RESNET's Equation 10 produces the alternative — CFM50 divided by the compartmentalization boundary area, giving cfm per square foot of enclosure at the same 50 pascals — and section 4.3.2.16 requires that boundary area to be recorded when a result will be reported that way. Whether your jurisdiction lets a dwelling unit comply on that basis instead of ACH50, and at what figure, is an exception written into the adopted code text rather than into the test standard. Read R402.4.1.2 in the edition your building department actually enforces before assuming either.

Now the part that matters more than any of it: these limits govern new construction and permitted work. Your existing house is not out of compliance for reading 9.9, and no inspector is coming. The zone figure is a yardstick for judging distance, and that distance is what a sealing quote is offering to close. Which is a different question from whether the quote is worth the money — the ranking method for that is in the audit that sets the order.

Getting from CFM50 to dollars is modelling, not measurement

Somewhere on many audit reports, ACH50 gets divided by 20 and reappears as a "natural" air change rate, which then gets multiplied into therms and dollars. That shortcut has an actual pedigree, and it also comes with an explicit warning from the person who published it.

Max Sherman's The Use of Blower-Door Data (Lawrence Berkeley Laboratory, LBL-35173, 1995, read 23 August 2026) sets out the chain. Equation 12 gives, for a typical single-storey house, normalised leakage as ACH50 divided by 20. The equivalent ventilation from infiltration is then 1.44 × w × NL, where w is an annual weather factor, and Sherman notes that "the factor 1.44w ranges between two-thirds and unity for most climates in the U.S." — which, combined with the first approximation, "comes close to approximating the 'divide by 20' rule."

Then the sentence immediately after:

This air change rate is a good estimate of the equivalent amount of ventilation produced by infiltration, but it is not a good estimate of the average air change rate or an air change rate suitable for making energy estimates.

The paper dates from 1995, and w depends on height, leakage distribution and wind sheltering, none of which a divide-by-20 captures. But the caveat is the point. If the dollars on your report trace back through ACH50/20, they rest on a conversion its own author disclaimed for exactly that use.

What you want instead is a modelled therm and kWh figure — the reduction in fuel use the software predicts for that measure, in this house, with local weather — priced at your own all-in rates, computed the way the four numbers on your bill get pulled out. Payback is that quotient and nothing else. An ACH50 figure is an input to the model, not a substitute for running it.

The fan the sealing quote did not mention

Sherman's Table 1 sorts houses into the leakage classes of ASHRAE Standard 119 and says, for each, whether infiltration can be counted on for ventilation. Reading across the typical ACH50 column: around 1, no credit at all, everything mechanical and balanced; around 2 to 3, still "quite tight construction" where the contribution "will be too small to count on and there is still a ventilation system requirement"; around 5 to 7, infiltration "may become a significant part of the ventilation requirement"; around 20, purpose-provided ventilation is not expected at all.

So a house going from 9.9 to 4 ACH50 has crossed a line. In new construction, code answers with a whole-house mechanical ventilation system, and IRC section M1505.4.3 sizes it — Equation 15-1, in the 2024 IRC text on UpCodes, read 23 August 2026:

cfm = (0.01 x total square foot area of house) + [7.5 x (number of bedrooms + 1)]

For the 2,000 sq ft, three-bedroom house above: 20 + 30 = 50 cfm, continuous. That is a fan, an install, a hole in the envelope you just sealed, and a small permanent electricity draw. Some programs size ventilation with ASHRAE 62.2 instead. It is a different calculation and will not hand back the same number, so ask which one a report used before comparing it with another.

The failure mode here is quiet: air sealing gets priced, ventilation does not, the payback is computed on the sealing alone, and the fan turns up as a separate job the following spring. The size of that error is just the ventilation cost over the sealing cost. If the fan, its install and its wiring land at half the sealing price, the measure costs 50% more than the figure you compared and the payback is half again as long. Ask for both on the same page, with the fan's wattage, so its running cost goes into the same sum.

Federal help with either half is gone for 2026 work — the §25C credit that covered air sealing materials and part of the cost of a home energy audit expired at the end of 2025 — so state and utility programs are the only subsidy left to check.

What the number will not tell you

An airtightness result describes the total size of the holes. It says nothing about where they are, and location decides whether sealing them is cheap or expensive. Two houses at 9.9 ACH50 can be an afternoon of foam and caulk at accessible top plates, or a job that needs a finished cathedral ceiling opened up, and the quotes will differ by an order of magnitude for that reason alone. The diagnostic work behind the number — zonal pressure readings, infrared while the house is depressurised, duct leakage measured separately — is what you are actually buying.

Which gives you something concrete to do with the report already on your desk. Find four things on it: the CFM50, the infiltration volume in cubic feet, whether the test was one-point or multi-point, and which spaces sat inside the boundary. Redo the division yourself. If any of the four is missing, the report is not complete under the standard it claims to follow, and the auditor can supply them from the same test file in a couple of minutes.