The rating condition nobody meets
Flexible duct is tested and rated fully extended. Not mostly extended, not extended and then eased back a little to get it round a joist: pulled taut and supported so the inner liner is a smooth tube. Every pressure drop figure published for a flex product is measured in that state, and the state is achievable, but it takes deliberate effort and it wastes duct, because a run of eighteen feet installed properly consumes about eighteen feet of duct and not the twenty five feet that came in the bag.
Compressed flex is a different duct. The corrugations that make it flexible turn into a series of restrictions when the duct is shortened, and the effect on friction is not proportional to the compression. A few percent of slack is measurable. Fifteen percent, which looks like a gentle wave from below, can put the friction rate at several times the stretched figure. This is why the multiplier on this page is an input with a range rather than a constant: the effect is large, it is real, and its exact size depends on the product and on how the slack is distributed along the run.
Reading the multiplier field honestly
| What it looks like | Rough multiplier over stretched | How it happens |
|---|---|---|
| Taut, supported every few feet | 1.0 | Cut to length, stretched, strapped with wide saddles |
| Small sag between supports | Around 1.4 | Supports too far apart, duct cut a touch long |
| Visibly wavy | Around 1.9 | Slack never taken up, one or two dips |
| Clearly compressed | Around 2.6 | Full bag length used on a shorter run |
| Concertina between two joists | 4 or worse | Excess duct pushed into a bay to get it out of the way |
These are a bracket, not a specification. Published measurements on compressed flex vary between products and between test setups, and the honest summary is that the direction and the order of magnitude are well established while any single number is not. The value of putting it in as an input is that you can run the same duct at 1.0 and at 2.6 and see what the installation quality is worth in inches of water gauge. That difference is almost always larger than anything else on the page.
Bends are a second, separate penalty
A bend in flexible duct has no fixed geometry. A metal elbow has a radius the manufacturer chose; a flex bend has whatever radius the installer happened to leave, and a tight one collapses the inner liner on the inside of the turn into exactly the shape that causes separation. That is why the equivalent length per bend on this page defaults high and is a field you can raise. A bend supported on a wide radius saddle is a modest fitting. The same bend pulled round a joist with a strap is not.
Sharp bends near a takeoff or immediately at a boot are the worst placement, because the flow arriving there is already disturbed. Two gentle bends spread along a run cost less than one hard one at the end.
What it changes downstream
A branch that runs three times the intended friction rate does not only deliver less air to its own room. It changes the balance of every other branch on the trunk, because the split between parallel paths depends on their relative resistance and one of them just got much harder. A house with metal trunks and flex branches, where the flex quality varies room to room, is unbalanceable in any stable way, because the resistances are not what the layout assumed.
It also raises total system static, which pushes the blower up its curve and reduces total airflow. Feed the pressure drop from this page into the duct static pressure calculator to see where the whole path lands against the blower rating, and use the duct branch balancing calculator to see how a single bad flex run redistributes air away from every other room.
Where flex earns its place
None of this is an argument against flexible duct. It is fast, it is cheap, it seals easily at the collars, it absorbs sound at moderate velocity, and a short well-stretched run from a metal takeoff to a boot is a perfectly good piece of ductwork. The failure mode is specific: long runs, excess length, unsupported spans and bends made by pushing rather than by forming.
The practical rules that follow are short. Cut it to the length of the run and no longer. Support it at close spacing with wide saddles that do not pinch. Stretch it before you strap it. Keep the bends wide and away from the ends. Every one of those is free at install time and impossible afterwards, which is the real reason the penalty is worth understanding before the ceiling goes up rather than after.
Questions people ask
How much airflow does compressed flex duct actually lose?
Enough that it is usually the largest single defect in a residential duct system, though the exact figure depends on the product and on how the slack sits. Work the default case on this page: eight inch duct, four hundred CFM, twenty five feet. Rigid metal at that size and flow runs about 0.28 inches of water gauge per hundred feet. Fully stretched flex is somewhat rougher, so call it 0.34. The same duct clearly compressed lands near 0.87, which over the run turns a small loss into a substantial one. Expressed the way it matters, the run behaves like roughly three times its own length of rigid duct. Because flow through a resistance falls as the resistance rises, and because that branch is competing with others, the room at the end sees a meaningful reduction, and the neighbouring rooms see an increase they did not ask for.
Is flex duct always worse than rigid?
Fully stretched and properly supported, flex is somewhat rougher than galvanised steel of the same diameter, typically in the region of ten to thirty percent higher friction depending on the product. That is a real penalty but a manageable one, and it buys you speed, cost, easy sealing at the collars and some sound absorption. The gap becomes large only when the flex is compressed or bent tightly, and at that point it is not really a comparison between two materials any more. The comparison is between a duct installed as the manufacturer intended and one that was not, and the answer would be similar for any duct that had its cross section obstructed.
Why does the calculator ask for the multiplier instead of just knowing it?
Because there is no single correct value and pretending otherwise would be the wrong kind of confidence. Published measurements on compressed flexible duct come from different products, different compression patterns and different test rigs, and they do not agree on a constant. What they agree on is the shape of the effect: the penalty rises steeply and disproportionately with compression, and even small amounts of slack are measurable. Making it an input lets you bracket the answer, run the same duct at both ends of the range, and see whether the conclusion changes. If it does not change, you have your answer regardless. If it does, you have learned that this run needs to be installed carefully or replaced with metal.
Can I fix a compressed run without replacing it?
Often yes, and it is the cheapest improvement available in most duct systems. If the run was cut long, the fix is to disconnect one end, pull the duct taut, cut off the excess, and reconnect. If the run is the right length but sagging, the fix is more supports at closer spacing using wide saddle straps rather than narrow ties that pinch the liner. Both are attic or crawlspace work on the duct itself rather than on any appliance. Anything involving the air handler cabinet, the coil, the refrigerant circuit or a gas appliance is licensed work and a different job entirely, and so is anything that requires the equipment to be opened.
Should I go up a duct size to compensate for flex?
It works, and this page calculates the size that would do it, but treat it as the second choice. Friction falls very steeply with diameter, so one nominal size up can absorb a substantial multiplier, and where a run genuinely cannot be installed tight it is a reasonable engineering response. The reasons it is second choice are that a larger duct costs more, is harder to fit in a joist bay, and lowers velocity in a system that may have been counting on that velocity for throw at the register. Stretching the duct you already have costs nothing and gives back more. Upsize when you have a real constraint, not as a substitute for installation quality.