What compression actually costs a flex runout

Method and provenance

1. Friction loss of fully extended flex from the Darcy-Weisbach equation Δp = f · (12L/D) · pv, with f from Colebrook-White and pv = ρ(V/1097)² at standard air, ρ = 0.075 lbm/ft³.
2. Absolute roughness ε = 0.003 ft. ASHRAE Fundamentals Table 1 classes "flexible duct, nonmetallic and wire, fully extended" as medium rough at 0.003 ft, citing Abushakra et al. 2004 and Culp 2011. Older editions of the same table put all fabric-and-wire flex in the rough category at 0.01 ft; that classification has been superseded, which is why design software still carrying 0.01 for runouts disagrees with this tool.
3. Compression enters as a pressure drop correction factor multiplying the result of step 1 — the application the ASHRAE Duct System Design Guide specifies. Slopes in PDCF = 1 + a·rc are a = 25.35 at 6 in., 21.61 at 8 in. and 16.18 at 10 in., from Abushakra, Walker and Sherman, "Compression Effects on Pressure Loss in Flexible HVAC Ducts" (LBNL, 2004).
4. Carrying capacity is solved by bisection on airflow until the friction rate meets the allowable rate — 100 × ASP ÷ TEL, the standard Manual D relationship.

What this tool refuses to do. It does not model compression by inflating roughness. Compression is a measured multiplier on pressure drop, and an expression like ε = 0.003 × (1 + k·compression) hides an invented coefficient inside a physical-looking constant. It does not use ACCA Manual D Appendix 17 correction figures, which are paywalled and reach the public only through secondary characterization. It does not supply fitting or boot losses from a private table; those enter only as a total effective length you read off your own report.

Limits. Straight horizontal runs only — no sag, no bends, no fittings, which means real installations lose more than this predicts, not less. The correction factors are measured at three diameters; 7 and 9 in. are interpolated and nothing is extrapolated past 10 in. or below 6 in. The linear PDCF fit tracks the measured points closely up to about 15 percent compression; at 30 percent it runs above the measured 8 in. value, so treat the high end of the slider as the published fit rather than a measured point. The velocity ceilings and the 0.10 in. w.c./100 ft design rate are conventional practice and also the defaults in the software that produced the reports behind them.

This is deliberately a conservative calculation. When the tool shows a problem, the real run is probably worse, not better, for three reasons:

1. It models one straight, fully supported run. Sag, bends, kinks, crowded plenum takeoffs, boots and fittings all add resistance that the tool does not count.
2. It starts from the most favorable roughness for fully extended flex, the current ASHRAE value of ε = 0.003 ft, not the older 0.01 ft that some design software still carries for runouts.
3. It ignores leakage. Air that escapes through unsealed collars, wyes and boots never reaches the room, and none of it is subtracted here.

One choice runs the other way, and it is stated above: at the high end of the slider the published linear fit gives more loss than LBNL measured for 8 in. duct. Up to about 15 percent compression, fit and measurement agree closely.

So the tool is good for showing that a run is starving a room. It is not a tool for proving an existing system is adequate. That takes measured static pressure and measured room airflow.

Written and maintained by Jeff Haag, who installs insulation and designs duct systems in New Orleans for a living. Default values were calibrated against four ACCA-approved Manual J/D reports for Gulf Coast homes; those reports are private and no figure from any individual house appears here. If you want this run against your own measured static pressures and effective lengths rather than against defaults, that is a job for a consult, not for a web form.