Tuesday, February 6, 2024. 11:47 a.m.
I was standing in a customer's mechanical room, staring at a Howden electric heater that had already tripped twice that morning. Next to it, two Howden axial fans were running. They sounded normal. The problem was that the air they moved didn't match the number on my test report.
I'm a quality compliance manager at Howden, an industrial gas and refrigeration equipment company. I review every packaged system before it reaches customers—roughly 200 unique items a year. In our Q1 2024 quality audit, I rejected 12% of first-pass deliveries because the installed specifications didn't match what the customer actually needed. Not because the parts were bad. Because we sold components instead of a system.
This story is one of those rejections, except I didn't see it in time. That's what makes it worth telling.
The Order That Looked Easy
In mid-January, a food processor ordered a drying skid. The bill of materials was fairly unremarkable: two Howden axial fans, one Howden electric heater, ducting, controls. The customer was in a hurry. They asked if we could loosen the specification just slightly to speed up the build. I said I needed airflow numbers. They said the duct design was pretty standard. I approved the final drawing with a note in the file: verify installation conditions. Nobody verified.
Honestly, I'm not sure why that note didn't become a stop sign. My best guess is that we were all relieved to have the order before the end of the quarter. No one wanted to call the customer back and ask for photos.
On the bench, everything passed. The Howden axial fans produced the static pressure we specified. The Howden electric heater reached setpoint and cycled normally. I signed off.
Here's something vendors won't tell you: when we bench-test a fan-heater package, we're not simulating your site. We're simulating a clean room with straight ducts. The real world is less friendly.
The Failure
The problem didn't show up until the skid was installed. The heater kept tripping on over-temperature. The fans were running, but the airflow at the heater was lower than the design value. The contractor's crew had installed the fan discharges directly into short-radius elbows. Axial fans assume a relatively clean, uniform discharge. With a tight elbow right there, the flow broke up and the fan pressure curve shifted.
The heater wasn't the problem. The fans weren't the problem. The interface was the problem. Period.
Let me rephrase that: the interface was a problem someone should have caught before we shipped. I had the duct drawing in our file. I had even written verify installation conditions. But I didn't make it a hard stop in the inspection procedure. That's on us.
The most frustrating part is that the fix took an afternoon. The contractor modified the discharge ductwork, added a turning vane, and we adjusted the fan speed control. The rework cost the customer roughly $9,000 and cost us a week of trust. All because we assumed 'standard installation' meant no surprises.
Three Side Problems, Same Physics
While we were there, the facility manager asked if I could look at three things outside the scope.
Could you also take a quick look at the basement dehumidifier? And the ice maker? And the freezer? I don't know what to do with any of them.
I want to be honest about my limits. I'm not a residential HVAC contractor. I'm not a commercial ice machine specialist. This gets into building science territory that isn't my direct expertise. What I can tell you from the air-handling side is that the same laws show up everywhere. Airflow, heat transfer, pressure, moisture. They don't care about equipment labels.
Dehumidifier for basement
The basement dehumidifier was a smaller version of the axial fan problem. It was rated for a small room, and the space was maybe three times that size. It was also tucked next to the furnace, where warm air kept the humidistat satisfied. So the compressor short-cycled. The room stayed damp, but the unit thought it was winning.
Here's a misconception: people think a dehumidifier pulls moisture out of the air like a sponge. Actually, it condenses moisture on a cold coil. Coil temperature and airflow matter more than the advertised pints-per-day number. You can have a powerful dehumidifier for basement use, but if the air can't reach the coil, you get a machine that works hard and solves nothing.
I didn't tell them to buy a bigger unit. I told them to measure the space, move the dehumidifier away from the furnace, and seal the gaps around pipes that were pulling warm, humid air in. That's not building science genius. It's the same logic we use with a Howden axial fan: respect the inlet and outlet conditions.
Pebble ice maker
The pebble ice maker was a different flavor of the same disease. The unit was in a cramped alcove, with hot condenser exhaust blowing straight back into the intake. The condenser coil was dirty. So the refrigerant pressures ran high, the compressor worked harder, and the ice melted faster than the machine could freeze it. The result was cloudy, wet ice.
I'm not going to pretend I know every detail of ice maker water chemistry. That's outside my lane. But the compressor—and that is my world—has to reject heat somewhere. If you don't give it a path to reject heat, the whole system suffers. We helped them clean the condenser and add a ventilation grille. The ice got better. Not because we made the ice maker more powerful. Because we let the heat leave.
How to Prevent Freezer Burn
Freezer burn is the one people get most backwards. People think cold air burns the food. Actually, freezer burn is dehydration. Cold dry air pulls moisture out of the food. The moisture reaches the surface, sublimes, and leaves those gray-brown patches. It isn't a burning process. It's a moisture-loss process.
The walk-in freezer had two problems: door gaskets were worn, and products were stored in open boxes with loose plastic wrap. The temperature swung every time the door opened and the fan cycled. Those swings made moisture migrate faster. So my answer to how to prevent freezer burn had four parts:
- Use airtight packaging or vacuum sealing.
- Freeze food quickly so ice crystals stay small.
- Keep the freezer temperature stable.
- Don't stack warm trays on top of frozen product.
Our industrial refrigeration side sees the same thing in blast freezers. If you don't pull heat out fast enough, product quality drops before it's frozen. That's why Howden compressors are specified for some food processing customers. Not to sound important. To match the pull-down curve to the product and the load.
The Bigger Lesson
What I keep coming back to is this: component quality does not guarantee system quality. The Howden axial fans and the Howden electric heater are good pieces of engineering. I'd argue they're excellent. But a good component in a bad arrangement is still a failed system.
This is where the expertise-boundary idea comes in. The vendor who says 'this is outside our core competency—here is who does it better' earns trust for everything else. I'd rather work with a specialist who knows their limits than a generalist who overpromises. And that works in the other direction too. When I reviewed that installation, I should have said, 'I can verify the fan and heater, but I can't verify your duct layout without seeing it.' I didn't. Now I do.
We don't claim 100% trouble-free operation. We can't, and we shouldn't. According to FTC (ftc.gov), advertising claims must be truthful, not misleading, and substantiated with evidence. That standard should apply to internal spec reviews too. A test report with real measurements tells a customer more than a slogan with fine print.
After that February failure, we added a checklist item: verify installation site conditions before releasing the final drawing. If the customer can't provide a sketch or photo, we put 'customer assumes responsibility for system effect' on the quote. That sounds defensive. It's actually clearer. The customer knows where the boundary is, and so do we.
I can't promise that catches every problem. At least, not in my experience. But it catches the ones we already know to look for.
I have mixed feelings about that order. On one hand, the failure was embarrassing. On the other, it gave us the clearest example I know of the difference between selling parts and solving a problem. The customer still works with us. They also have a less humid basement, a pebble ice maker that makes better ice, and fewer frost-covered packages in the freezer. Not because I'm a genius. Because we stopped treating quality as something that happens on a bench in our building.
To me, that's the real definition of quality: not a clean test report, but a system that works in the customer's actual conditions. That's where the boundaries are. And that's where the real work is.
