An airport terminal can fail at air conditioning in two ways. It can fail where passengers notice it, in drafts, grille noise and hot and cold spots, or it can fail silently, conditioning an upper volume where nobody stands and paying for that mistake in kilowatts every hour the building is open. Both failures start in the same place, and it is not the mechanical room but the distribution.
A building that resists being conditioned
Few commercial buildings are harder to condition than a terminal. Ceilings run from roughly 33 to 130 ft (10 to 40 m). Glazed curtain walls pick up solar gain all day long. Occupancy arrives in pulses: a thousand people at a gate that was empty twenty minutes earlier, then empty again.
ASHRAE states the consequence directly in its guidance for transportation centers. In an airport, the single largest problem is often thermal drafts created by large entranceways, high ceilings and long passageways with openings to the outdoors. Not installed capacity. Not the chiller. Air movement.
That diagnosis moves the argument from tonnage to distribution.
Why cold air does not fall from a 100 ft ceiling
Supply air leaves an outlet as a jet, and a jet does two things on the way down.
First, it slows. In the fully developed zone of the jet, the ASHRAE Handbook — Fundamentals chapter on space air diffusion gives the centerline velocity as Vx = Kc3 · Vo · √Ao / X: velocity falls inversely with throw distance. The constant is not a generic figure. It comes from outlet testing to ASHRAE Standard 70, and the tabulated values are published as examples only.
Second, it entrains room air. That is the working principle of mixing ventilation, and it has a cost: as the jet advances, both the supply velocity and the temperature difference between supply air and room air get smaller and smaller.
From 65 to 100 ft (20 to 30 m) that leaves two outcomes. Throw hard enough to reach the floor and the air arrives in the occupied zone as a draft, which is the visible failure. Throw less and it never arrives: the jet stalls in the upper volume, mixes into the warm stratified air already sitting there and loses the capacity to cool anything. The cooling was produced and paid for. It simply never reached a passenger.
The leakage that never enters the comparison
Then there is the metal itself: tons of galvanized sheet suspended over the heads of the public, assembled from transverse joints sealed by hand in the field.
Field work at Lawrence Berkeley National Laboratory measured ten large commercial duct systems at operating conditions. Three leaked less than 5 percent of duct inlet flow. The other seven leaked between 9 and 26 percent. That spread is not a product property. It is what field assembly does to a drawing.
The consequence is not only wasted fan power. ASHRAE recommends that a terminal building be designed to maintain a substantial positive pressure, precisely because air balance is hard to control across many outdoor openings, high ceilings and long, low passageways that are frequently not conditioned. ASHRAE does not attach a figure to the word substantial, and neither should a specification that quotes it. And air balance is what leakage disturbs first: air released into a ceiling plenum instead of the space it was sized for is still inside the building, but it has left the balance the designer calculated.
The duct is the diffuser
Fabric duct changes the geometry rather than the material list. There is no grille at the end of the run, because the fabric surface is itself the diffusion element: air is released gently along the entire length instead of being concentrated at a handful of outlets. Where it has to come down from height, the design uses high-throw or low-throw nozzles, sized and aimed to deliver into the occupied zone, which in a terminal means gates, check-in lines and seating.
That is consistent with the design intent ASHRAE describes for high-ceiling transportation spaces: concentrate conditioning where it is wanted and avoid disturbing the stratified air above. ASHRAE describes sidewall distribution as the means; what transfers to a textile system is the intent, not the hardware. Less air mass treated for the same conditions in the occupied zone means fewer kilowatts.
Sound follows the same logic. Diffusers and grilles typically generate their highest noise levels in the octave bands centered at 1000 Hz or above, which is the band that carries speech intelligibility. That is why a noisy concourse buries its own announcements. A textile surface diffusing at low face velocity does not whistle, drum or vibrate.
Weight is the last piece of it. A fabric run weighs a fraction of the equivalent metal and hangs from a single steel cable, so suspended load on the structure drops and installation can proceed at night, section by section, without taking a concourse out of service for a single day.
Nothing is guessed
Before a foot of duct is cut, the terminal is modeled in computational fluid dynamics: the air in the building divided into millions of cells, with the physics solved in each one for direction, temperature and velocity. It is the same class of numerical method used in aerospace and motorsport, and it takes high-performance computing to run.
It earns its keep because large volumes do not mix uniformly. Microclimates form. Flow descends in some regions while convective plumes rise in others, and a design that ignores the plume fights it. The plume needs room to develop, and extraction belongs where it rises. A terminal does not behave like a supermarket, a hospital or an office.
The output is numeric. Air speeds in the occupied zone are checked against ANSI/ASHRAE Standard 55. It is worth being exact about what that standard actually sets. What it limits is the mean speed the occupant is exposed to within the occupied zone, not the velocity in a duct or at an outlet face, and it publishes no single blanket maximum: the allowance moves with operative temperature, moves again where people can adjust the air themselves, and is judged separately at ankle height. Those are inputs to a comfort model for spaces designed for human occupancy, not duct design limits, and the standard does not specify setpoints.
From that velocity field comes the perforation pattern: diameters, rows and orientation, laser-cut and sewn at the plant.
Aesthetics and maintenance are specified, not inherited
Color is a decision, not a default. A duct can contrast with the architecture and read as part of it, or be specified in a low-visibility tone that lets it recede into the ceiling structure. Internal hoops are equally optional: with them the duct holds its shape when the system cycles off, without them it breathes with the equipment.
Maintenance never shuts the terminal down. Sections come down run by run, go out for industrial laundering and come back with a service report for each one.
Frequently asked questions
How do you get air down 100 ft without creating a draft?
Not by increasing throw. The jet decays inversely with distance and entrains room air the whole way, so more supply velocity buys arrival at the cost of a draft. The workable route is directed delivery from nozzles sized for the occupied zone, with the resulting air speeds verified against ASHRAE Standard 55 rather than against a catalog throw figure.
Does conditioning only the occupied zone violate any ventilation requirement?
No. It separates two questions that get confused. Outdoor air rates and air changes are set by the applicable ventilation standard and are unaffected by the diffusion element. What changes is where the conditioned air is placed, and ASHRAE explicitly describes concentrating conditioning where it is wanted in high-ceiling transportation spaces.
How does fabric duct compare to metal on leakage?
Metal leakage in large commercial buildings is a field variable, not a catalog value: LBNL measured 9 to 26 percent of inlet flow in seven of ten systems, and less than 5 percent in the other three. A textile system has no hand-sealed transverse joints at all, so the field variable that produces that spread does not exist in it.
Is CFD necessary, or is it a sales exercise?
For a terminal it is necessary, because the geometry has no precedent in smaller buildings. Ceiling height, glazing load and pulsed occupancy produce microclimates and convective plumes that no rule of thumb predicts, and the perforation pattern is derived from the resulting velocity field rather than chosen from a standard product.
The decision in a terminal is not made by comparing the installed price of a linear foot of duct. It is made by asking three questions the traditional comparison usually leaves out: what fraction of the design airflow actually reaches the occupied zone after distribution losses, what air speed the passenger standing at a gate is exposed to at that flow, and what sound the diffusion element contributes in the octave bands where announcements have to be understood. A system that answers all three with measured or computed numbers is specifiable. One that answers them with a catalog throw figure quoted to a standard terminal velocity is a hypothesis about a building that has no equivalent at small scale.
Request a CFD-based design and sizing study for your terminal at fabricduct.co.
