Phoenix Journal · Ductwork
A busy kitchen extract can look right on paper and still run hot, loud and greasy - because the ductwork was designed to fight itself. Here is how to design that resistance out from the first drawing.
A cautionary tale
The canopy looked the part, the fan was rated for the job on paper, yet six months in the kitchen was hot, the extract was noisy and the grease was already backing up.
Picture a busy production kitchen fitted out to a tight budget. The extract fan was correctly sized for the airflow the canopy needed, and on the day of commissioning the numbers just about held up. Within a season, though, the chefs were complaining that the space never cleared, the fan was howling at full speed, and the first grease clean turned up heavy deposits far sooner than anyone expected. Nothing had broken. No motor had failed. The system was simply fighting itself, and it had been designed to do so from the first drawing.
The culprit was resistance - the sum of every bend, squeeze and rough transition the air had to shoulder its way through between the hood and the roof. When you design that resistance out from the start, you get a quieter, cheaper, cleaner system that keeps its performance for years. When you design it in, no fan on the market will save you.
Good extract ductwork is a balancing act, and the sweet spot is narrower than most people think. Move the air too slowly and grease-laden vapour drops out onto the duct walls; push it too fast and you pay in noise, energy and wear. The recognised UK reference here is BESA’s DW/172, the Specification for Kitchen Ventilation Systems, which environmental health officers, insurers and consultants all treat as the benchmark even though it is guidance rather than statute.
DW/172 points designers toward roughly 6 to 8 metres per second in the main runs, easing back a little through branches and spigots. That range is not arbitrary. Grease-carrying ducts need to hold a transport velocity - broadly a minimum around 7.6 metres per second, or 1,500 feet per minute in the older units the research still uses - to keep particles airborne rather than settling on the metal. Drop much below about 7 metres per second and deposition climbs sharply; push past roughly 12 to 14 metres per second and the noise and the fan energy climb with it. Design your duct sizes so the air naturally lands inside that window at the airflow the canopy actually demands, and half the battle is already won.
Trace the failed kitchen back to its drawings and the problems were all in the geometry, not the kit.
First, the route was tortured. To dodge a steel beam the installer had thrown in two tight, mitred 90-degree bends within a metre of each other. A sharp elbow like that can add the resistance of five to ten metres of straight duct, and stacking them so close together makes it worse than the sum of the parts - the air leaving the first bend is still churning when it hits the second. Long-radius bends or turning vanes would have cut that penalty by around sixty per cent, but they cost a little more and took a little more thought, so they were skipped.
Second, the duct was the wrong shape. Where it crossed the ceiling void it had been flattened into a wide, shallow rectangle to gain headroom - an aspect ratio well past 3:1. The trouble is that a squashed duct has far more wall in contact with the air than a round or square one of the same area, and friction rises with it. ASHRAE guidance is to keep the aspect ratio as close to 1:1 as you can and never above 4:1; round is best of all. This run had quietly become the biggest single source of drag in the system.
Third, the transitions were abrupt. The connection from the canopy plenum into the duct, and from the duct into the fan, were blunt step-changes rather than tapered. Fitting-into-fitting transitions like these are often the largest hidden losses in a whole system, throwing the airflow into turbulence exactly where you least want it. Every one of those losses landed on the fan, which had to spin faster and draw more power to hold the design airflow - hence the noise, the heat and the energy bill.
And because the velocity sagged in the flattened, over-resisted sections, the grease had time to fall out of the air and stick. The system had effectively been designed to soil itself, which is why the first TR/19 clean found so much so soon.
Designing for minimal resistance is not exotic engineering. It is a handful of disciplined choices made early, while the route is still lines on a drawing and cheap to change. If you are specifying or reviewing a system, work through these in order.
Get those six right and the fan can run slower, quieter and cooler for the same airflow, the grease stays airborne to the point of extract, and the whole system is measurably cheaper to run and to keep compliant. Specific Fan Power - the yardstick Approved Document L uses to judge how hard a fan works for the air it moves - falls as a direct result, which increasingly matters as the regulations tighten.
A system built for minimal resistance is also a system built to be cleaned, and the two ambitions pull in the same direction. Straight runs, gentle bends and generous access panels are exactly what a hygiene technician needs to reach every internal surface, and a duct that holds its transport velocity is one that soils more slowly between visits. Even a well-designed heavy-use kitchen still needs a TR/19 grease clean around every three months, but good geometry means each clean is faster, more thorough and easier to verify. The same care that shapes efficient ductwork tends to run through a whole kitchen; the way you plan the cooking line and its workflow and service times often decides where the canopy sits and how the duct has to travel in the first place.
Questions
BESA's DW/172 points to roughly 6 to 8 metres per second in the main extract runs, easing back slightly through branches and spigots. Grease-carrying ducts also need a transport velocity of at least around 7.6 metres per second (1,500 feet per minute) to keep particles airborne. Below about 7 metres per second grease drops out and settles; above roughly 12 to 14 metres per second noise and fan energy climb sharply, so the aim is to size ducts so the air naturally lands inside that window.
A single sharp 90-degree elbow can add the resistance of five to ten metres of straight duct, and a flattened rectangular duct has far more wall touching the air than a round or square one, so friction rises with it. Using long-radius bends or turning vanes can recover around sixty per cent of the loss a sharp turn would cost, and keeping the aspect ratio near 1:1 (and never beyond 4:1) keeps friction down. Every bit of resistance you remove is fan energy, noise and grease deposition avoided.
A low-resistance duct holds its transport velocity, so grease stays airborne longer and the system soils more slowly between cleans. Straight runs, gentle bends and TR/19 access panels at roughly two-metre intervals also let a technician reach every internal surface. Even a well-designed heavy-use kitchen typically still needs a grease clean around every three months, but good geometry makes each visit faster, more thorough and easier to verify.
Phoenix Duct Clean · by the numbers
Phoenix surveys and cleans kitchen and building ductwork to the TR19 standard - measured, cleaned and certificated, UK-wide.