How one bottling plant and KCF Technologies turned a routine vibration alert into a root-cause discovery and quieted five cooling tower fans.
Walk up to almost any condition monitoring dashboard and you’ll eventually see it: Imbalance detected — cooling tower fan.
Most systems would print that alert, generate a work order, and call it a day. The maintenance team would head out, check the fan, find nothing obviously wrong, and start wondering whether their monitoring system actually knows what it’s talking about.
At a beverage bottling facility in Pennsylvania, that’s exactly the moment where the story could have ended. Instead, it’s where it got interesting.
A Critical Asset with a Hidden Problem
Cooling tower fans don’t get much glory, but at a bottling plant they’re process-critical. If enough of them go down, production stops. And these aren’t quick swaps. They’re large, expensive assets that take real time and money to repair or replace.
During a pilot deployment of KCF’s SMARTdiagnostics® platform, continuous vibration monitoring on the plant’s cooling towers revealed something unusual. For the vast majority of operation, vibration levels were stable and healthy. But for roughly 7.5% of runtime, vibration spiked fivefold, jumping from around 0.4 inches per second to 2 inches per second.
That’s serious shaking. And it wasn’t isolated. The same pattern was showing up on every cooling tower at the site.
A conventional system would have flagged it as imbalance and moved on. The FFT data even supported that conclusion at first glance, with a high 1x run speed peak, the classic imbalance signature. But something didn’t add up. Rather than simply prescribing a fix, the KCF team brought the question to the customer: Why is this happening?
The Test: 75%, 90%, 30%
Here’s where continuous monitoring became a genuine partnership.
The initial theory was load-related: on hot days, the fans ramp up to cool more water, and higher speed amplifies an underlying imbalance. It’s a logical story. So together with the plant team, KCF designed a simple controlled experiment. On a mild day, operators ran the fans at three distinct speeds — 75%, 90%, and 30% — while SMARTdiagnostics captured the vibration response at each step.
If the imbalance theory were right, vibration should have climbed steadily with speed: low at 30%, higher at 75%, highest at 90%.
That’s not what happened.
The worst vibration on the machine showed up at 75% speed. At 90%, it actually went down.
The imbalance theory was wrong — and if the team had simply acted on the original alert, the “fix” would have missed the real problem entirely.
Finding the Root Cause
With the speed test data in hand, KCF’s analysts dug deeper. By adjusting the sensor sampling strategy and using SMARTdiagnostics’ run-speed tracking, they could overlay the motor’s VFD speed directly against fan vibration over time.
The picture snapped into focus. Within one specific VFD speed range, vibration shot through the roof. Outside that range, the fans ran smoothly. This wasn’t imbalance at all. It was structural resonance. At certain speeds, the fans were exciting a natural frequency of the structure, shaking the equipment violently every time operations happened to land in that band.
KCF shared the analysis with the plant team, not as a prescription but as evidence. The customer’s own reliability and controls engineers immediately saw the answer: program the drives to skip the problematic speed range entirely.
They didn’t need to be told what to do. The data made the conclusion obvious.
The Result
The plant’s controls team programmed the resonant band out of the VFDs across the cooling tower fans. In the weeks since, time in alarm on those assets has dropped dramatically, and the ongoing monitoring is now helping the team fine-tune the exclusion range and evaluate whether structural stiffening could eliminate the issue for good.
That’s the difference between a fault-finding tool and a machine health partnership:
- A conventional system says: Imbalance. Go check the fan.
- SMARTdiagnostics and the KCF team said: Here’s what’s happening, here’s when it happens, here’s the experiment to prove why — and here’s the data your team needs to solve it at the root.
Five cooling tower fans, one root cause, zero unnecessary teardowns.
Insights, Not Just Alerts
Resonance problems are exactly the kind of issue that gets missed by monitoring solutions built to spit out automated diagnoses. They don’t fit the standard fault categories. They require correlating operating conditions with vibration behavior over time. And solving them means working with the people who know the plant best, not talking past them.
That’s the KCF approach: continuous monitoring, advanced vibration analytics, and human expertise working together to move plants from reactive firefighting to proactive problem elimination. Because the goal isn’t just to detect failures earlier. It’s to make entire categories of failure never happen at all.