Three failures, one condenser unit, and a question I didn't want to ask
I'm the office administrator for a 240-person manufacturing company. I manage facilities and MRO purchasing—roughly $380,000 a year across 14 vendors—and I report to both operations and finance. When I took over facilities purchasing in 2021, I assumed the hard part would be the paperwork.
It turned out to be the fans.
Specifically, the axial fan on a condenser unit at the north end of our production floor. It failed in June 2022, again in November 2022, and a third time in April 2023. Same unit, same mounting plate, same failure mode: bearing noise first, then heat, then the shutdown alarm at 2 a.m.
The first time, I assumed bad luck. The second time, I blamed the vendor and switched suppliers. The third time, I sat down with the maintenance lead and admitted something uncomfortable: I had no idea what I was actually buying.
The surface problem was never the fan
Here's what I knew about fans before 2023. They blow air. That was roughly the depth of my expertise on a component that could stop a production line.
And to be fair to myself, I'd bought plenty of fans before. Desk fans for the summer, a neck fan for the woman in accounting who runs hot, a countertop ice maker for the break room that I cleaned twice and then forgot about. None of that prepared me for specifying an axial fan for a condenser unit, but I didn't know that yet. I thought a fan was a fan.
The uncomfortable truth is that we had been buying industrial fans the same way I buy a neck fan online: match the approximate size, pick a reasonable price, hope it works. That's fine when the stakes are thirty dollars and a warm office. It is not fine when the stakes are a shutdown at 2 a.m. and a maintenance crew on overtime.
The problem underneath the problem
When I finally pulled the purchase history, three things stood out. None of them were about the fan itself.
We had no spec review process
The original workflow was this: maintenance foreman sends me a photo or a part number, I find something with matching dimensions online, I compare prices, I order. Nobody checked airflow, static pressure, operating temperature, IP rating, or control signal. We were matching bolt patterns and calling it procurement.
That came back to bite us when a replacement arrived with the right frame size and the wrong speed curve. It fit. It ran. It also ran hot and loud, and it was back on the work order four months later. We didn't have a formal verification step, and it cost us a full day of downtime plus the price of a second unit.
We treated a fan as a part number, not a system component
A fan sits inside a system—ducting, filters, a control board, an ambient temperature that changes with the season. Every one of those variables affects which fan is correct. We were treating it like a light bulb: same socket, same job.
The datasheets were on the vendor sites the whole time. I just never opened them. I didn't know what I was looking at, so I skipped it and ordered the one that looked closest.
We didn't understand AC versus EC
Every replacement we bought was an AC fan, because that's what had been there before. EC fans existed, but I assumed they were expensive and fussy—something for the engineers to worry about.
What I didn't know is that EC motors run cooler and draw less power at partial load than the AC units we were replacing, and they handle speed control without extra hardware. For a condenser unit that doesn't need to run full tilt in January, that matters. Not enough to make me an expert. Enough to make me ask better questions.
What three failures actually cost
I don't have hard data on how much the old AC motors cost us in energy. Our utility bills are bundled by building, not by equipment, and I never tracked it at the unit level. What I can say anecdotally is that two of the three failures happened during peak summer load, and both times the unit was already running near its limit.
The costs I can count:
- Roughly 11 hours of unplanned downtime across the three incidents, with the April failure landing on a Thursday we did not have to spare
- Three fan purchases where one properly specified unit should have lasted longer
- Overtime for the maintenance crew on two of the three callouts
- Enough awkward conversations with operations to make me want to avoid a fourth
And then there's the cost I almost missed. Last summer I cleaned the condenser coils on the break room countertop ice maker for the first time in two years. The change in how fast it made ice was obvious within an hour. Nobody had ever put that machine on a maintenance schedule—it just quietly got worse until someone mentioned it.
It occurred to me later that we had no filter-change or coil-cleaning schedule for the industrial units, either. We were replacing fans on equipment that was breathing through clogged coils. Some of those failures were probably self-inflicted.
Where the “we do everything” pitch fell apart
One of the vendors we tried during this mess advertised themselves as a one-stop shop for motors, fans, and controls. They quoted every request we sent. Fast, too.
They also shipped us a motor with a wiring diagram that didn't match our control board, and when we called, the answer was “it should work.” It didn't.
Compare that to the supplier we eventually went with. Their first response to our inquiry wasn't a quote. It was a question: what's the ambient temperature at the fan location, and what control voltage is available? They hadn't sold us anything yet. They just wanted to know whether they could actually solve the problem.
That conversation changed how I buy. Not because the first vendor was dishonest—I don't think they were. They just didn't know the edges of what they were selling. The second vendor did, and said so.
I've started asking vendors a blunt question now: what's something you don't do? The ones who answer without flinching are the ones I trust with everything else.
What we changed
The fix was less dramatic than the problem.
Before any fan order now, we need three documents on file: the datasheet, the wiring diagram, and the manual. No documents, no purchase order. It sounds bureaucratic. It has already caught two wrong-speed units before they shipped.
We also stopped treating every fan position as interchangeable. The condenser unit got a properly specified axial fan from ebm-papst—an EC unit, chosen against the actual operating conditions rather than a photo and a guess. It wasn't the cheapest quote we received. It was the one that came with a datasheet I could actually read and a technical contact who answered questions about temperature derating without deflecting.
Not every position needs that level of spec. Some of our exhaust fans are fine with a commodity unit. The point is that we now know which is which, because we asked someone who knew the difference instead of assuming nobody did.
What I still get wrong
I'm not an engineer. I still can't read a fan curve without help, and I still call our maintenance lead before I finalize anything that touches the production floor. That's fine. The job isn't to become the expert—it's to know when I'm not the expert and get the right one on the phone.
The third failure taught me that. Expensive lesson, but a lasting one.
If your facility is burning through fans and the replacements keep failing, look at your buying process before you blame the product. In our case, the fans were doing exactly what we asked. We just kept asking for the wrong thing.