Why Does My Radiator Fan Stay On? A Fleet Buyer's $2,500 Lesson in Root-Cause Repairs

One of our drivers called me on a Tuesday morning. "Check engine light is flashing." I could hear the radiator fan roaring in the background. "And the fan won't shut off." The van—a 2015 Ford Transit—had been parked for 45 minutes, fan still running. By the time I got to the lot, the battery was dead.

I'm the office administrator for a 62-person logistics company. I handle all the fleet maintenance ordering—roughly $120,000 a year across 14 vendors. I report to both operations and finance, which means I'm the one explaining to drivers why they can't just take the van to any repair shop, and the one defending every repair bill to my VP.

My first instinct was simple. The fan is doing something weird, so the fan is the problem. Replace the fan. Done.

But the Fan Is Just the Messenger

It's tempting to think a failing part is the problem. The fan runs when it shouldn't, so the fan must be broken. That logic feels clean. It's also wrong more often than you'd expect.

A radiator fan stays on because something is telling it to stay on. The fan itself is usually fine. The real culprits I've seen in our fleet:

• The engine coolant temperature sensor sending a false "hot" signal to the ECU
• Corrosion in the wiring connector at the fan—surprisingly common on Transits of that generation
• A thermostat stuck closed, making the engine genuinely run hot
• A failing fan control module
• The A/C system triggering the fan due to a refrigerant pressure issue

Five different root causes. One identical symptom. If you just replace the fan, you waste your money and the problem comes back.

That pattern didn't stay in the cooling system for long. Once I started asking "why is this happening?" instead of "what part do I order?", I saw it everywhere.

The Same Lesson, Three Different Parts

One of our vans kept eating tail light bulbs. Every three weeks, another one blown. We kept ordering replacement tail light bulbs—and honestly, I kept picking the cheapest ones. The bulb wasn't the problem. The socket was corroded, and the vibration from a worn strut mount was shaking the filament to death. The fix was an $8 socket repair, not another $14 bulb.

Another vehicle's centrifugal fuel pump failed at eight months. The first shop replaced it with a no-name unit. That pump failed too. The actual diagnosis on the second visit: a weak relay causing a voltage drop that was cooking the pump motor. A $45 relay and a proper centrifugal fuel pump ended the saga. The cheap pump wasn't a deal—it was a deposit on a second repair bill.

This is the part that changed how I think: symptom-based replacement is the most expensive buying strategy you can have.

What the "Simple Fix" Actually Costs

Let me walk you through the math on that radiator fan.

Visit one: the shop replaced the fan relay. $187, and the van was down for the rest of the day.

Visit two: the fan ran again on the next long drive, so they replaced the coolant temperature sensor. $231. The scan tool threw a code, but nobody actually tested the wiring between the sensor and the ECU.

Visit three: a different mechanic actually checked the wiring harness, found corrosion in the connector, cleaned it, and fixed the contact. $315.

That's $733 in shop charges for what was, from the start, a $40 connector and maybe an hour of proper diagnosis. But the cost doesn't stop there.

The van was out of service for parts of three separate days. At roughly $600 a day in lost utilization for a box truck, that's $1,800 in productivity gone. The total tab for that single "simple" problem cleared $2,500.

And that's just the visible cost. Every one of those visits generated a work order, an invoice, an accounting match, and a conversation with a frustrated driver. Three visits meant three POs, three invoice approvals, three expense lines. That's my time, finance's time, and the driver's lost hours—none of which shows up on the repair bill.

There's also the failure cost that never shows up in a ledger. When a van is out of service, deliveries get missed and drivers get reassigned. Customers notice. That's the most expensive cost of all, and it's the one nobody wants to put a number on.

I've seen the exact same math with suspension parts. When a Transit starts wandering at highway speed and clunks over bumps, most shops say "struts are worn, replace them." Then they quote a generic strut built to a "good enough" spec.

Here's the thing: a 2015 Ford Transit is a heavy van that sometimes runs empty and sometimes runs loaded. On paper, a standard strut fits. In real life, it gets beat up. I went back and forth on this for two weeks—the budget option that saved $90 per corner, or the better option. On paper, the budget strut made sense. My gut said we'd be doing this job again in a year.

The upside was $360 saved up front. The risk was a dead damper at 20,000 miles, a full redo on labor, and a driver who thinks I'm cheaping out on safety parts. I kept asking myself: is $360 worth potentially doing the whole job again?

I went with the Koni special active Ford Transit 2015 strut. It solved a problem I hadn't even identified. The Special Active valve adjusts the damping to the load in real time—the van rides smooth when it's light, and it doesn't wallow when it's loaded. That's the difference between a part that fits and a part that's actually right for how the vehicle is used.

I am not here to tell you every brand name is worth paying for. I am here because the math is clear: that budget strut wasn't $90 per corner cheaper in the long run. When you count the second installation, the alignment, the downtime, and the ride complaints, it was the expensive option.

What I Do Differently Now

When I took over purchasing in 2020, I approved whatever the shop recommended. I do not do that anymore. Three changes made the biggest difference.

First, every repair has to answer a question before I sign off: what is the root cause, and how do you know? Not "the fan is broken." Why is it broken? What test confirmed it? A shop that can't answer that doesn't get my PO.

Second, I buy parts that match real operating conditions. For our Transits, that means a Koni strut assembly built for the vehicle's actual weight range. It means a proper-spec centrifugal fuel pump, not the cheapest drop-ship option. And for bulbs, it means a vibration-resistant replacement tail light bulb—a few dollars more—and fixing the socket while we're in there.

Third, I treat the whole vehicle as a system. The burning tail light bulb was a clue about a failing strut mount. The noisy fuel pump was a clue about an electrical issue. The radiator fan that would not shut off was a clue about a corroded connector. Parts don't fail in isolation, and buying decisions shouldn't either.

In our 2024 vendor consolidation project, we cut from fourteen vendors down to seven. The ones that stayed are the ones who can explain a system instead of just shipping a part. Their invoices actually match the work order, too—which sounds trivial until you've had finance reject an expense because the line items don't line up. We cut parts spend around 18% last year, not by buying cheaper parts, but by buying the right ones the first time.

Look, the word "efficiency" gets thrown around a lot. But in my world, efficiency is simple: fewer repeat repairs, fewer emergency orders, fewer phone calls from drivers. The radiator fan that taught me this lesson cost us $2,500. The question I ask before every order since then costs nothing.

That's the best part of the whole story. It's a free fix.