What duct sizing actually decides
A duct has one job: deliver a defined airflow to a defined place without spending more pressure than the blower has. Every foot of duct converts some of the fan's static pressure into friction against the wall of the pipe, and every fitting converts some more into turbulence. Size the duct too small and the pressure runs out before the air reaches the far register. Size it too large and you spend metal, space and heat loss for nothing.
The rate at which straight duct spends pressure is called the friction rate, quoted in inches of water column per 100 feet of duct. It rises steeply with airflow and falls steeply with diameter, which is the whole reason sizing works: the exponents are 1.9 on flow and 5.02 on diameter. Increasing a duct from 6 in to 7 in, a 17% change in diameter, cuts its friction to (6/7)^5.02 = 46% of what it was, a 54% reduction at the same airflow. That asymmetry is why the last inch of duct size is usually the cheapest performance you can buy.
Two methods dominate. Equal friction fixes the pressure loss per 100 ft and lets velocity fall out of it; every section of the system is sized at the same friction rate, so the pressure drop of any run is simply its total effective length times that rate. Constant velocity fixes the speed and lets friction fall out; it is used where noise sets the limit, where duct depth is constrained by a joist bay, or in the return where velocity governs grille sizing.
Equal friction is what ACCA Manual D uses for residential systems, because it turns the whole design into one number: find the friction rate from the blower's spare pressure and the longest run, then size everything at it.
The friction equation and where its constants come from
The relation used here is the standard power-law fit to the Darcy-Weisbach and Colebrook solution for galvanised steel duct with an absolute roughness of 0.0003 ft:
Δp100 = 0.109136 × Q1.9 ÷ D5.02
with Q in cfm, D in inches, and Δp in inches of water column per 100 ft. The exponent 1.9 on flow rather than a clean 2.0 reflects that the friction factor itself falls slowly as Reynolds number rises. The exponent 5.02 on diameter rather than 5.00 is the same effect seen through the hydraulic diameter. Both come out of curve-fitting the exact solution over the range of sizes and flows real duct occupies, and the fit reproduces the published ASHRAE friction chart closely: 1,000 cfm at 0.10 in.wg per 100 ft returns 13.90 in against a chart reading of about 13.9.
Rearranged for design, the equation gives the diameter directly, and the one-fifth power on the right is what makes the answer so insensitive to the friction rate you choose. Tripling the allowed friction from 0.06 to 0.18 shrinks the duct by only about 20%.
Velocity follows from geometry alone. The free area of a round duct in square feet is πD² ÷ 576 with D in inches, so V = 576Q ÷ (πD²), or 183.35Q/D².
For rectangular duct the calculator uses the Huebscher relation, De = 1.30(ab)0.625÷(a+b)0.25, which gives the round duct that has the same friction loss at the same airflow. A 12 × 12 duct has an equivalent diameter of 13.1 in, not the 13.5 in that equal cross-sectional area would suggest, because the corners of a rectangle carry slow-moving air that contributes area without contributing flow.
Worked example: a 1,000 cfm trunk at 0.10 in.wg per 100 ft
A furnace moves 1,000 cfm and the design friction rate, taken from the available static pressure and the total effective length, is 0.10 in.wg per 100 ft. The trunk has to fit between joists, one dimension fixed at 12 in.
- Raise the flow. 1,0001.9 = 501,187.
- Apply the constant. 0.109136 × 501,187 = 54,698.
- Divide by the friction rate. 54,698 ÷ 0.10 = 546,976.
- Take the root. 546,9761/5.02 = 13.90 in.
- Round up. The next stocked size is 14 in.
- Check the velocity. 183.35 × 1,000 ÷ 14² = 935 fpm. That is at the top of the comfortable residential range, which is expected: the equal-friction method lets velocity land where it lands.
- Check the friction actually delivered. 0.109136 × 501,187 ÷ 145.02 = 54,698 ÷ 566,974 = 0.0965 in.wg per 100 ft, a little under the target because you rounded the diameter up.
- Rectangular equivalent. Solving 1.30(12b)0.625÷(12+b)0.25 = 13.90 gives b = 13.5 in, so a 12 × 14 duct is the practical choice and is very slightly generous.
Run the same 1,000 cfm at 0.06 instead and the diameter goes to 15.39 in with a velocity of 774 fpm; at 0.18 it falls to 12.36 in at 1,199 fpm. The duct changes by about three inches across the entire usable range of friction rates, while the velocity rises by 55%. Duct size is insensitive to the friction rate; duct noise is not.
Reading velocity, and when to override the friction answer
Velocity is the output to check after every sizing, because the friction method never looks at it. In residential work, supply trunks are comfortable up to roughly 900 fpm and branch runs lower; above that the register becomes audible in a quiet bedroom, and above about 1,200 fpm the duct itself starts to rumble. Commercial mains routinely run at 1,500 to 2,500 fpm, but they are lined, they serve occupied spaces through terminal boxes, and they are designed with a sound calculation.
Where velocity comes out too high at your friction rate, do not lower the friction rate for the whole system. Lowering it globally enlarges every duct and, more importantly, it means you were not going to spend the blower's available pressure, so the fan will simply move more air than designed and the velocity will climb back. The correct fix is to upsize the specific noisy section, accept the slightly lower friction there, and keep the rest of the system at the design rate.
The friction-at-standard-size output shows you what rounding up cost. It is always at or below the target, because a larger duct loses less. Over a whole system those small surpluses accumulate into real margin, which is one reason equal-friction designs generally deliver a little more airflow than nominal.
Aspect ratio is the constraint on rectangular sections. A 4:1 duct has far more sheet metal, more surface area for heat loss, and more joints per unit of airflow than the round duct it replaces. Where geometry forces a high ratio, prefer two smaller round ducts side by side.
All of this presumes you have a friction rate to design at. Get it from the available static pressure calculator, which subtracts the filter, coil and accessory losses from the blower's rated pressure and divides what is left across the total effective length. Get the airflow per room from the room CFM calculator and the system total from the CFM per ton calculator.
Round duct capacity at 0.10 in.wg per 100 ft
| Diameter (in) | Airflow (cfm) | Velocity (fpm) |
|---|---|---|
| 6 | 109 | 553 |
| 8 | 232 | 665 |
| 10 | 419 | 769 |
| 12 | 679 | 864 |
| 14 | 1,019 | 953 |
| 16 | 1,451 | 1,039 |
| 18 | 1,980 | 1,120 |
| 20 | 2,617 | 1,200 |
Each airflow is the friction equation inverted: Q = (0.10 x D^5.02 / 0.109136)^(1/1.9). Note that velocity climbs steadily with size at constant friction, which is why the largest trunks in an equal-friction design are also the fastest.
What this calculation does not include
- Fittings. Elbows, takeoffs, boots and transitions can easily exceed the loss of the straight duct they connect. Manual D handles them as equivalent lengths, which is why the design length is called total effective length rather than measured length.
- Flexible duct. Flex is rougher than galvanised steel even when fully stretched, and compressed flex is dramatically worse. Sizing flex from this equation and then installing it with slack is one of the most common causes of a system that will not deliver its airflow.
- Duct liner and duct board. Internally lined duct and fibrous glass board both have higher roughness than bare metal. Size them from the manufacturer's own friction data.
- Air density. The correlation assumes standard air at 0.075 lb per cubic foot. At 5,000 ft the density is about 83% of standard, and friction losses fall in proportion; the same duct passes more air for the same pressure.
- Return-side grille and filter losses. Those are component pressure drops, not duct friction, and they belong in the available static pressure calculation instead.
- Leakage. A duct sized for 400 cfm at the takeoff delivers less than that at the register if the run leaks. Seal to the applicable duct-tightness requirement and verify by test, not by assumption.
Which standard governs your job
The friction correlation and the equivalent-diameter relation both appear in the ASHRAE Handbook of Fundamentals duct design chapter. For residential systems, ACCA Manual D is the design standard that most codes and utility programmes reference; it consumes room-by-room loads from Manual J, applies the available static pressure to a total effective length to get a friction rate, and then sizes every section at that rate. For sheet metal construction, pressure classes and reinforcement, SMACNA's HVAC Duct Construction Standards are the reference. This calculator does the sizing arithmetic those documents specify; it does not replace the design procedure.
Where duct sizing sits in the design sequence
Duct sizing is the last quantitative step, and it depends on everything before it. The load calculation gives the room loads, which set the room airflows. The equipment selection gives the blower and its rated external static pressure at that airflow, which the furnace size calculator ties back to the heating load. The available static pressure and the total effective length together give the friction rate. Only then does this calculator have both of its inputs.
Design in that order and the numbers come out consistent. Skip a step and the failure is usually invisible until commissioning: the system is built, the blower runs at full speed, and the measured airflow is 30% below design because the ducts were sized from a friction rate nobody derived. Measuring total external static pressure at that point tells you immediately whether the duct system or the equipment is the problem.
A last practical note. Round duct outperforms rectangular of the same free area, keeps its shape, seals more easily and costs less to install in most residential geometries. Rectangular exists because of the spaces we build in, not because it moves air better. Where you have the depth, use round, and use this calculator's rectangular output only where the joist bay or the soffit leaves you no choice. Once the sizes are set, check the resulting air change rate in the rooms with the air changes per hour calculator to confirm the distribution is sensible.