I'm the procurement manager at a 140-person mechanical services contractor. I've managed our air-movement budget for six years, negotiated with more than 30 vendors, and logged every order in a cost-tracking system I built after getting burned twice by hidden freight charges. Our spending in this category ran about $640,000 in 2024.
Every couple of months, a project manager forwards me a spec that says "industrial type fan" and asks for three prices by Friday. Every time, I have to ask the same two questions first: what's the system static pressure, and how long is the duct run? Those two numbers decide whether you're buying an axial or a centrifugal fan, and the gap between the two options on the same job has run anywhere from 8% to 40% on our projects.
Here's how I compare them.
Sticker price is the least useful number on the quote. I compare on five lines:
Run the comparison that way and the answer flips more often than the spec suggests.
This is the physics that decides most of it, so I'll be blunt about it.
Axial fans move a lot of air against very little resistance. The types of axial fans we buy most are propeller fans (cheap, no duct, wall or roof mounted), tube axial fans (impeller in a cylinder, fine for straight duct runs), and vane axial fans, which add straightening vanes behind the impeller and pick up a real efficiency gain — usually 5 to 10 points of static efficiency in the submittals I've compared, though it varies by selection.
Centrifugal fans trade volume for pressure. Forward-curved wheels are compact and cheap but not especially efficient. Backward-inclined and airfoil wheels are the ones worth paying for when the fan runs thousands of hours a year. Radial and radial-tip wheels show up in dirty process exhaust, where you accept lower efficiency in exchange for a wheel that won't clog.
Where the split happens on our jobs is around 1 to 1.5 in. w.g. of total system resistance. Below that, an industrial axial fan is typically smaller, lighter and cheaper to install. Above it, you're asking an axial fan to do something it's bad at, and the noise and stall margin get ugly fast.
One thing I check before anything else: whether the published curve comes from a certified test. For package fans, that means airflow and pressure tested under AMCA 210 / ASHRAE 51 and published through the AMCA certified ratings program (AMCA 211). For larger industrial machines, ISO 5801 is the reference standard. Sound data is a separate question — ask for sound power tested to AMCA 300, because "85 dBA at five feet" measured in a vendor's own room isn't comparable to anything.
Verdict on this dimension: short straight run, low resistance — industrial axial fan, and it isn't close. Filters, long runs, or a layout full of elbows — centrifugal fan exhaust selection, airfoil or backward-inclined.
This is where I've saved the most money, and where I've made my worst call.
Two metrics matter. First, motor efficiency: IEC 60034-30-1 sets the IE classes, and stepping from IE3 to IE4 usually adds somewhere in the 10-20% range to the motor's price. Second, and bigger, the fan's own efficiency at your operating point.
ASHRAE 90.1-2016 was the edition that moved the industry metric to the Fan Energy Index, calculated per AMCA 208. Roughly speaking, an FEI of 1.0 means the fan draws about as much power as the reference baseline fan at that duty point, and higher is better — an FEI of 1.25 works out to roughly 20% less power. Check the current edition and the exceptions before you write it into a spec, because coverage isn't universal.
Here's the math we actually ran in 2023 on a 15 kW duty. Two fans, same airflow, same pressure. One had a certified FEI of 1.0, the other 1.25. At our duty point the difference came to about 8,800 kWh a year at 4,000 run hours — roughly $970 annually at our 2024 blended rate. Over ten years, before any rate escalation, that's close to $9,700. The premium on the better fan was about $2,800.
Simple payback under three years, and then it just keeps paying. Your hours and your tariff will be different, so run your own numbers rather than borrowing mine.
The upside was $2,800 in capital savings on the cheap option. The risk was that the published curve wouldn't hold up in the field, because those ratings weren't certified and I'd have no way to prove anything when the balancing report came in 15% short. I kept asking myself whether $2,800 was worth potentially eating a rebalance and a re-selection on a job with a penalty clause. It wasn't.
What I'm less sure about is whether FEI alone is the right thing to buy on. I have mixed feelings about it. On one hand, it's the first number that lets me compare two vendors' fans on the same basis. On the other, a fan can post a great FEI at peak efficiency and still spend most of its life at 60% flow, where it's mediocre. So I ask for FEI at the actual operating point, not just at the peak. If you ask me, that's the number that belongs in the spec.
Axial fans are inline. That's basically the whole story. They sit in the duct, need a support or two, and go in during a normal working day. A comparable centrifugal unit weighs more, needs a base, spring isolators, a flexible connection, and enough room for transitions and a belt guard.
On one 2021 job we swapped a centrifugal selection for a vane axial and took about 6,400 pounds off the roof structure. That's not a rounding error in the steel package.
Now the part that catches people. If you're controlling flow, you have two options: throttle with a volume control damper, or slow the fan with a VFD.
Throttling works and it's cheap to install. It's also the most expensive way to control flow over the life of a fan, because you're paying for the pressure drop you just created. On our numbers a VFD added roughly $1,100 to a 7.5 kW fan package and paid for itself in under two years on any fan running more than 2,000 hours a year. Dampers still belong in the design — for balancing between branches, for isolation, for zoning. Just not as the primary flow control on a fan that runs all day.
If you do specify one, ask for certified leakage data. AMCA 500D is the test standard and the leakage class comes out of that testing; the phrase "low leakage" in a submittal means nothing until it's a class tied to a standard. I still kick myself for not putting that requirement in our 2019 spec template. We ended up with dampers on two roof exhausts leaking enough that the fans ran near full speed year-round, and it took me 18 months to get the language into every subsequent package.
Verdict on this dimension: axial wins on installation and footprint in nearly every case. The damper-versus-VFD decision moves lifetime cost more than the choice of fan does.
Everything above assumes ducted ventilation. Car parks break that assumption.
A ducted centrifugal exhaust system for a basement car park means a lot of duct, a lot of hangers, and a lot of ceiling height consumed. The alternative is a car park jet fan layout: a set of small impulse fans near the ceiling that push air toward extract points, so most of the ductwork disappears.
On capital cost, we've seen jet fan layouts come in below ducted systems on simple, well-shaped basements. Not because the fans are cheap — they aren't — but because you're deleting duct, hangers and installation labor, which is usually the larger line. The counterweight is design effort. Jet fan layouts need real engineering: thrust calculations and, in most cases, CFD. The codes are specific too. BS 7346-7 covers smoke and heat control for covered car parks in the UK, AS 1668.2 covers car park ventilation including impulse systems in Australia, and where a garage ventilation system doubles as a smoke control system in the US, NFPA 92 governs the approach.
Operation is where jet fan systems either earn their keep or don't. With CO sensors staging the fans — setpoints are the mechanical engineer's call under local code, and ours have landed in the 25-50 ppm range — the fans sit idle most of the day and run hard during the morning and evening peaks. Without that control, you're running a bank of 1 kW fans continuously for no reason at all.
Verdict on this dimension: simple basement geometry plus a competent designer — jet fans usually win on installed cost. Complicated smoke strategy or a ceiling that's already a maze of services — ducted centrifugal is the lower-risk answer, and I'd rather pay for it than argue about it in year two.
I'd rather spend ten minutes explaining why I'm asking for a certified fan curve than deal with a mismatch three months later. An informed buyer asks better questions and gets to a decision faster — that's not a slogan, it's just what the last six years of invoices taught me.
When a quote comes in with certified airflow and pressure data, FEI at the duty point, motor efficiency class, and the damper's leakage class, I know I'm dealing with someone who's done this before. When it's a price and a lead time, I'm going to spend a week building the comparison myself. The vendor who made me do that work rarely gets the order.