A centrifugal blower runs quietly for months. Then it starts banging — a loud, rhythmic thump every second or two, sometimes with a visible shudder in the ductwork.
Operators often assume a bearing failure or a loose impeller and shut the line down. In most cases, the machine is mechanically fine.
What they’re hearing is surge — an aerodynamic instability, not a mechanical fault. Misdiagnosing it means the real problem never gets fixed.
Surge is one of the most disruptive, least understood failure modes in centrifugal fan operation. It doesn’t look like a typical vibration problem.
It shows up suddenly and loudly, usually right after something changes upstream or downstream: a damper throttles further, a VFD turns down for a low-load period, or a second fan starts in parallel.
This article covers what’s physically happening, why it happens, and how EPC engineers can design it out before commissioning — instead of fighting it in the field.
What Surge Actually Is, Physically
Every centrifugal fan has a pressure-flow performance curve. On the left side of that curve, at low flow, the curve typically flattens or reverses slope.
Operating left of that point puts the fan in an unstable region. Airflow across the impeller blades separates intermittently and static pressure oscillates.
In severe cases, flow through the fan can momentarily reverse before snapping forward again. That cycle repeats at a low frequency — usually well under the shaft’s rotational speed.
That’s exactly why it feels and sounds different from ordinary 1x or 2x rpm vibration. It’s an aerodynamic pulsation, not a rotor imbalance.
Left uncorrected, the repeated pressure and flow reversal cyclically loads the impeller, bearings, and shaft — loading they were never designed to absorb continuously. Ignored surge tends to end in a real mechanical failure, even though it didn’t start as one.
Surge vs. Ordinary Mechanical Vibration
| Characteristic | Surge (Aerodynamic) | Ordinary Mechanical Vibration |
| Frequency | Low — typically well below shaft rpm, felt as a repeating pulse | Synchronous with shaft speed (1x rpm) or a clear harmonic (2x rpm) |
| Correlates with | Damper position, VFD speed, or another fan starting/stopping | Temperature change, bearing wear, or rotor imbalance |
| Sound signature | Rhythmic banging or “chugging” | Steady hum or whine, amplitude-modulated only if a fault is progressing |
| Fix | Change the operating point or add anti-surge control | Re-balance, re-align, or replace worn components |
What Actually Pushes a Fan Into Surge
- An oversized fan running at low load. A blower selected with excess margin “just in case” spends most of its life throttled well below design flow — often right where the unstable region begins.
- Aggressive VFD turndown without a documented stable range. Every speed reduction moves the whole performance curve, and the minimum stable flow point moves with it. Turning down further than the stable range allows is a common, avoidable trigger.
- Rising system resistance over time. Duct fouling, filter loading, or a misadjusted damper all raise resistance, pushing the operating point left — often gradually enough that nobody notices until surge appears.
- Mismatched fans running in parallel. When two fans with different curves or wear states share a duct, the weaker unit can be pushed backward into its own unstable region, even while overall system flow looks normal.
Diagnosing It in the Field
- Check the timing, not just the noise. Confirm the pulsation frequency is well below running speed. A quick vibration spectrum separates surge from a shaft-speed-related fault immediately.
- Correlate the onset with a specific event. Look for a damper change, a VFD setpoint change, or another fan starting or stopping in the same timeframe.
- Plot the actual operating point against the fan’s published curve. If flow is at or below the stated minimum stable flow, that confirms the diagnosis without further guesswork.
Engineering the Problem Out at the Selection Stage
The most reliable fix is never letting the operating point reach the unstable region in the first place.
That starts with selection: size the fan against the actual system resistance curve, not an oversized “safety factor.” The normal operating range should sit clear of the stall point across the full turndown the process requires.
Where a wide turndown ratio is unavoidable — common in kiln and boiler duty with seasonal or load-following operation — a documented minimum stable flow point matters. Pair it with a recirculation or bypass line sized to hold the fan above that point, giving operators a hard engineering limit instead of a guess.
This kind of problem is far cheaper to solve on paper than in the field. A properly matched curve and a stated stable operating envelope cost nothing extra at the quotation stage. Retrofitting anti-surge control onto an installed fan is a real engineering project.
If your process involves a wide turndown ratio, parallel fan operation, or a duct system with resistance that changes seasonally, our engineering team can confirm your minimum stable flow point and recommend a control strategy before the fan is built. Submit your operating parameters here and we’ll size the curve against your actual system — not a generic catalog rating.
