Field fan vibration diagnostics: four faults on one spectrum

A fan rarely fails without warning — it fails without anyone reading the warning. The signal is a velocity spectrum, and it takes about thirty minutes on site to capture. The useful part is that one spectrum carries four different diagnoses at once: imbalance, misalignment, bearing wear and looseness each sit at a different place on the frequency axis. Learn where they sit and a single reading tells you not just that a fan is unhappy, but why.

This matters most on the fans you cannot casually stop. A boiler induced-draft (ID) or forced-draft (FD) fan runs continuously, and pulling it apart on a hunch is expensive. So instead of guessing, you measure velocity RMS (mm/s) at both bearing housings — horizontal, vertical and axial, on the housing, never the base frame — and you let the spectrum point at the fault.

Four signatures on one axis

Anything rotating out of balance pushes once per revolution, so it shows up at running speed (1×). Misalignment loads the shaft twice per revolution and pushes along it, so it climbs at 2× and in the axial direction. Bearing damage rings at much higher frequencies long before it moves the velocity number. Looseness rattles across a whole comb of harmonics. None of it is mysterious — it is geometry.

Fault	Where it sits on the spectrum	Common cause
Imbalance	1× dominant; horizontal & vertical high, axial low	Dust or process build-up on the impeller, blade damage, a shed balance weight
Misalignment	2× larger than 1×; axial high at the drive end	Shifted coupling, slipped pulley, a foundation that has settled on one side
Bearing wear	High-frequency envelope (gE) band rising; velocity still low but climbing	Spalled races, cage damage, marginal lubrication
Looseness	Raised harmonics — 3×, 4×, 5× — often with an audible rattle	Loose anchor bolts, a cracked base weld, soft foot, or structural resonance

The bearing row is the one that catches people out. A defect shows in the high-frequency envelope band (gE) months before it touches the velocity RMS. The classic progression is a gE reading that creeps up over two-to-four months, then a velocity number that rises sharply over four-to-six weeks, then failure within days. Watch velocity alone and you meet that fan at the very end of its story.

Why the trend beats the limit

Acceptance tables have their place — we assess new fans by the ISO 20816 method and expect a healthy machine to sit in the good zones. But an absolute limit is a blunt instrument between measurements. The sharper signal is the rate of change. A fan that reads a modest number this month and double that next month is telling you something a fan sitting steadily at a higher-but-stable value is not. A doubling in the monthly trend is a slope, and a slope has a destination.

That is the whole argument for a monthly log over a one-off inspection. Motor current works the same way — the absolute amps mean little; the drift over time is the tell. Baseline every fan within its first month in service, then let the trend line, not the limit line, drive the decision.

A real reading: 11.8 → 1.6 mm/s

Early in 2025 we were called to a continuously-run process fan, roughly 75 kW, that had been getting louder by the week for months. Nobody had put a number on it. Our technician's first velocity reading at the motor bearing was 11.8 mm/s RMS — not merely “high” but the kind of value that says a bearing is days from letting go. We stopped it that afternoon; the opened bearing had spalled races and cage fragments loose in the housing.

The instructive part was the root cause. The spectrum and the operator's memory of an earlier axial (2×) rise pointed past the bearing to a foundation that had settled on one side. Replacing the bearing alone would have re-wound the same failure inside a year or two. We jacked and shimmed the foundation, replaced both bearings, laser-aligned the coupling, and returned the fan to service at 1.6 mm/s RMS. Reading the whole spectrum — not just the one alarming number — is what separated the symptom from the cause.

What we bring to the fan

This is our own field practice, not a brochure. We capture the full FFT velocity spectrum on site, so the 1× / 2× / harmonic story is visible rather than inferred; we balance impellers in place with an on-site balancer; we set couplings with laser alignment and keep the report; and we check bearing and lubrication condition directly before condemning either. Being able to diagnose the signature before touching a component is the point — the wrong fix costs twice.

Vibration also has a close cousin: when the frequency that jumps is a structural natural frequency rather than a fault, you are looking at resonance, and the fix is stiffness, not balance. We cover that decision in our companion note on resonance versus imbalance on a boiler ID fan.

If a critical boiler fan or process fan is drifting on you, send us the make and model with a photo of the nameplate and we will walk the spectrum through with you.

Talk to us about field vibration diagnostics →

Jitamitra Electro Engineering · Fan-engineering notes, written for the engineer.