How to Convert Blower Door CFM50 to Manual J Infiltration

ACCA's Technical Reference Note 2016-1 converts a blower door CFM50 into an effective leakage area of 0.055 sq in per CFM50 at sea level, then runs that area through Manual J's stack and wind equation once for heating and once for cooling. On the two-story house worked below, that comes out to CFM50 divided by about 10 for the heating load and by about 22 for cooling.

The divide-by-20 shortcut would have put half the design heating infiltration into that load. For one, two and three-story homes, see the divisor table further down. The free Manual J calculator runs the whole-house heating and cooling load from a tight, average or leaky class. It doesn't take your ICFM directly, as covered at the end of this page.

Why doesn't ACH50 divided by 20 work for Manual J?

Because 20 estimates an average infiltration rate, and Manual J needs the airflow on the design day. ACCA's method sizes infiltration at the full heating temperature difference and a 15 mph wind, so on the example below the winter number comes out about twice the divide-by-20 figure, 289.6 against 145 CFM. Jones, Persily and Sherman traced the rule of 20 to early-1980s field studies and found no physical basis for it.

Their 2016 paper describes the ratio as a way to infer "the mean heating season, or other typical, infiltration rate" from a 50 Pa test. For the heating load on the example house below, you need the leakage at a 76°F indoor-outdoor difference with a 15 mph wind.

Jones, Persily and Sherman, ASHRAE and AIVC IAQ 2016 § Abstract

"In practice, the value of 20 is not applicable to all buildings and climates, and should be adjusted according to a number of factors, such as building height, shielding, air leakage path size, and climate." - Benjamin Jones, Andrew Persily and Max Sherman, nist.gov

The LBL n-factor tables are the usual refinement. In the BPA Journal, Randy Williams reads the n-factor table and gets an n of 18.5 for a one-story house with normal exposure in zone 2 of the LBL map, and 9.8 for a three-story house in an exposed spot in zone 1, which nearly doubles the estimate for the same 2.75 ACH50. He still writes that the result "isn't very reliable." None of these divisors is tied to the design-day temperature difference a Manual J load runs on.

ACCA starts from the test pressure itself. A blower door holds the house at 50 Pa (0.2 in. w.c.), while weather pushes on it far more gently.

ACCA Technical Reference Note 2016-1 § 3.0 Leakage for a 4 Pascal Test

"Per industry research, the pressure difference value for natural infiltration circumstances is about 4 pascal, so the envelope leakage test should be a CFM4 test." - ACCA, acca.org

So the note converts the 50 Pa reading to a leakage area at 4 Pa and lets Manual J's own stack and wind coefficients do the rest.

How does ACCA turn CFM50 into infiltration CFM?

Five steps. Get CFM50 from the test, or from ACH50 times volume divided by 60. Multiply by 0.055 at sea level for the effective leakage area at 4 Pa. ICFM equals that area times the square root of Cs times TD plus Cw times wind speed squared, with Cs and Cw from Manual J Table 5D or 5E. Run it once for heating and once for cooling, then turn each ICFM into Btu/h.

Here's the sequence from the note, written for ANSI/ACCA 2 Manual J-2016, version 2.50:

1. Volume. You need it only when you start from an ACH50 target. For heating, the volume includes all heated rooms. For cooling, it includes all air-conditioned rooms.
2. CFM50. On a tested house, use the blower door reading. For a code target, CFM50 = ACH50 x volume / 60.
3. Leakage area, ELA4. A single-point test can't give the flow exponent, so ACCA assumes n = 0.65. That makes C = 0.079 x CFM50 and ELA4 = 0.055 x CFM50 in square inches at sea level. Figure 2 of the note drops the factor to 0.054 at 1,000 ft and 0.050 at 5,000 ft.
4. Infiltration CFM. ICFM = ELA4 x (Cs x TD + Cw x V^2)^0.5. TD is the Manual J heating or cooling design temperature difference, and V is wind speed in mph, 15 for heating and 7.5 for cooling by default.
5. Loads. Sensible heating = 1.1 x ACF x ICFM heating x HTD. Sensible cooling = 1.1 x ACF x ICFM cooling x CTD. Latent cooling = 0.68 x ACF x ICFM cooling x grains difference. ACF is the altitude correction, 1.0 - 0.0000308 x elevation in feet.

Cs is the stack coefficient and grows with height. Cw is the wind coefficient and shrinks as the house gets more shelter. These are the Table 5D values for houses with three or four exposures, as printed in the note:

ACCA calls class 4 "typical suburban shielding" and class 2 "a few nearby obstructions: sheds, trees." Table 5E covers houses with one or two exposures, for example an interior townhouse with neighbors on both sides.

A worked example with a 2,900 CFM50 test result

Take ACCA's own sample house, 24,896 cu ft over two stories at 1,000 ft, and give it a hypothetical test result of 2,900 CFM50, about 7.0 ACH50. The ACCA method gives 289.6 CFM of heating infiltration and 134.3 CFM for cooling, or 23,484 Btu/h of sensible heating load. Divide-by-20 gives 145 CFM for both seasons and half that heating load.

The note's house has a 56 x 32 ft main level with 8 ft ceilings (14,336 cu ft) and a 33 x 32 ft lower level with 10 ft ceilings (10,560 cu ft). It sits at 955 ft, rounded to 1,000, with shielding class 4. The Manual J Worksheet A values in the note are a 76°F heating TD, a 15°F cooling TD and a 38 grain moisture difference. It hasn't been built yet, so ACCA sizes it at a 3.0 ACH50 limit set by the AHJ. We kept every input and swapped in a hypothetical test result of 2,900 CFM50.

1. ACH50 = 2,900 x 60 / 24,896 = 6.99. This is for reference only. The leakage-area path never uses it.
2. ELA4 = 0.054 x 2,900 = 156.6 sq in.
3. Heating. (0.0299 x 76 + 0.0051 x 15^2)^0.5 = (2.272 + 1.148)^0.5 = 1.849. ICFM = 156.6 x 1.849 = 289.6 CFM.
4. Cooling. (0.0299 x 15 + 0.0051 x 7.5^2)^0.5 = (0.449 + 0.287)^0.5 = 0.858. ICFM = 156.6 x 0.858 = 134.3 CFM.
5. Loads with ACF 0.97. Heating 1.1 x 0.97 x 289.6 x 76 = 23,484 Btu/h. Sensible cooling 1.1 x 0.97 x 134.3 x 15 = 2,149 Btu/h. Latent cooling 0.68 x 0.97 x 134.3 x 38 = 3,366 Btu/h.

Now run the same house through the shortcut. 2,900 / 20 = 145 CFM, used for both seasons.

Divide-by-20 lands within 8% of the ACCA cooling numbers and at half the ACCA heating infiltration. In this house the shortcut drops 11,726 Btu/h of heating infiltration, about the output of a 3.4 kW strip heater.

To check the arithmetic against ACCA, run the same steps at the note's 3.0 ACH50, which is 1,244.8 CFM50. You get ELA4 of 67.2 sq in, 124.3 CFM and 10,080 Btu/h of heating, the same as the note. Our cooling ICFM comes out 57.6 CFM where the note prints 57.8, a difference of 3 Btu/h in sensible cooling and 5 Btu/h in latent.

What divisor does the ACCA method imply?

Divide CFM50 by ICFM and you get the divisor the ACCA method implies. At sea level with suburban shielding and default winds, it runs about 8 to 15 for heating and 15 to 27 for cooling, depending on stories and design temperature difference. Taller houses and bigger temperature differences push it lower. A flat 20 only lands inside the cooling range.

The table is our arithmetic from Table 5D, shielding class 4, 15 mph heating and 7.5 mph cooling wind, and the 0.055 sea level factor. The TD columns are sample values. Use your own heating and cooling TD from Manual J Worksheet A, or pull the outdoor design temperatures for the ZIP code from our design temperature lookup.

Every heating cell is under 20. Dividing by 20 understates design heating infiltration by 25% on a one-story house at a 40°F TD and by 59% on a three-story house at 80°F. On the cooling side the error swings both ways, from 37% high on a one-story house at a 15°F TD to 25% low on a three-story house at 25°F.

You can't fold this into one better rule of thumb either. The heating and cooling divisors for the same house differ by a factor of about two, so a single natural ACH number can't be right for both loads.

Which inputs change the answer the most?

Stories, temperature difference, shielding and altitude. On the example house, moving from shielding class 4 to class 2 raises heating infiltration from 289.6 to 350.0 CFM, about 21% more. Altitude cuts it through the leakage-area factor and the air density correction, by about 5% of the infiltration load at 1,000 ft and about 23% at 5,000 ft. Volume matters only when you start from an ACH50 target.

Go through these before you trust the number:

• Shielding class. Class 4 is the Manual J default when the site plan doesn't tell you more. A house on an open lot with a few trees is class 2, and the example house then needs 350.0 CFM for heating and 166.4 CFM for cooling.
• Wind speed. The defaults are 15 mph for heating and 7.5 mph for cooling. The note allows higher values "for locations that have a reputation for wind velocities that consistently exceed the default values."
• Stories. Cs doubles from one story to two (0.0150 to 0.0299), so the stack term drives most of the winter number on a two-story house.
• Altitude. The Figure 2 factor goes from 0.055 at sea level to 0.050 at 5,000 ft, and ACF drops to 0.85 at 5,000 ft. Together that's 0.050 / 0.055 x 0.85 = 0.77, so leave altitude out at 5,000 ft and the infiltration load comes out about 29% high.
• Multipoint test. If the blower door software reports C and n, use the note's Section 5.0 equation, ELA4 = 0.2835 x C x 4^n at sea level, in place of the 0.055 factor.
• Volume. A measured CFM50 goes straight to ELA4, so the old argument about whether the basement belongs in the volume doesn't change the infiltration CFM. It only changes the ACH50 figure on the report.

Putting the number into a load calculation

If your software asks for air changes rather than CFM, convert back with ICFM x 60 / volume, using the volume your software multiplies by. On the full 24,896 cu ft, the example house comes out at 0.70 ACH for heating and 0.32 ACH for cooling. Enter 0.70 in the heating infiltration field and 0.32 in the cooling field, and note the shielding class and wind speed you assumed in the job file.

BuildSolver's free whole-house Manual J in the chat doesn't take a CFM50. Its heating and cooling tools each take a tightness class, tight, average or leaky (average if you don't give one), and assume 0.10, 0.18 and 0.30 ACH for cooling and 0.30, 0.55 and 1.0 ACH for heating. Those are BuildSolver's calibrated assumptions, not code values. The tool also figures volume as floor area times 8 ft, which gives 22,784 cu ft on the example house against the actual 24,896.

Run the example through it and no class matches the ACCA numbers. For cooling, leaky gives 113.9 CFM, 15% under the 134.3 CFM from the ACCA method. For heating, average gives 208.9 CFM, 28% under 289.6, and leaky gives 379.7 CFM, 31% over. Pick leaky for heating if you'd rather the load err high, and treat the result as a preliminary number.

The room-by-room Manual J on the Pro and Team plans accepts your own heating and cooling ACH, so you can enter 0.70 and 0.32 directly. To check the blower door result against the 2021 or 2024 IECC limit, use the air changes per hour calculator.

This is preliminary sizing for a quote. For a permit, the AHJ may want a load report from ACCA-approved software, covered in Manual J for permit. Start with the whole-house load in the Manual J calculator and keep the ICFM math from this page in your notes.

Sources

• ACCA, Technical Reference Note 2016-1, Computing Manual J Infiltration Loads Based Upon a Target Envelope Leakage Requirement, October 2016, sections 1.0 to 7.0, Figure 2, Tables 5D, 5E and 10A: acca.org
• Benjamin Jones, Andrew Persily and Max Sherman, The Origin and Application of Leakage-Infiltration Ratios, ASHRAE and AIVC IAQ 2016: nist.gov
• Randy Williams, Measuring Natural Air Leakage, BPA Journal, May 15, 2023: building-performance.org
• ACCA, ANSI/ACCA 2 Manual J-2016, version 2.50, Section 21 and Tables 5D and 5E, as quoted in Note 2016-1

For estimation purposes only. Not a substitute for a licensed engineer, and not ACCA-approved for permit submission.