Can Sound Diagnose Mining Pump Faults Early?

2026-08-04

Upgrade traditional manual listening experience into an intelligent monitoring system for centrifugal pumps. Acoustic signals enable early‑stage fault warning and cut unplanned downtime losses for mining pump, mine drainage pump and products from multistage pump manufacturer.

1. Two Types of Acoustic Signals

Centrifugal pump acoustic signals cover audible sound and Acoustic Emission (AE), each for specific scenarios.

Audible sound: Identified on‑site by maintenance staff with low cost and easy operation. Cavitation generates gravel‑like or popping noises; bearing failure brings continuous metal friction and high‑pitched squealing, serving as quick on‑site preliminary judgement.
Acoustic Emission (AE): High‑frequency elastic waves with high signal‑to‑noise ratio for incipient fault capture. AE signals occur in bearing lattice wear, impeller crack propagation and cavitation bubble collapse. Sensors support precise online monitoring for predictive maintenance.

2. Acoustic Fingerprints of Four Typical Faults

Cavitation: Common hydraulic fault marked by persistent crackling noise from bubble implosion. AE parameters such as ring‑down count and RMS voltage quantify cavitation severity with fitting goodness above 0.96. Proven in marine and overseas cases, AE detects early cavitation before obvious head drop and helps locate the Best Efficiency Point (BEP).
Bearing failure: Major cause of rotor breakdown, featured by persistent rustling or sharp buzzing. VMD decomposition and envelope spectrum extract fault features; time‑frequency diagrams pinpoint bearing wear positions to prevent cascading damage from rotor‑stator rubbing.
Wear‑ring rubbing: Friction between impeller and wear ring creates distinct kurtosis peaks in frequency spectrum. Acoustic‑signal‑based modelling delivers advance warning and avoids severe wear and sharp flow reduction.
Rotor component defects: Impeller looseness and shaft bending produce periodic impact noise. AE event count rate directly reflects rotor damage progression.

3. Complete Intelligent Acoustic Diagnosis Workflow

Signal acquisition: Microphones collect audible noise; dedicated AE sensors capture high‑frequency acoustic emission. LabVIEW builds integrated online monitoring systems.
Pre‑processing & denoising: Offset superposition amplifies fault features while distance evaluation filters irrelevant industrial background noise.
Feature extraction: Multi‑dimensional fault markers are obtained via time‑domain statistics, Fourier spectrum, VMD and wavelet time‑frequency decomposition.
Intelligent identification: Machine‑learning models including SVM and probabilistic neural networks classify faults automatically for unattended early warning.

4. Industry Applications & Existing Challenges

Applications

Acoustic monitoring is deployed across oil & gas, marine, power and water sectors. It monitors bearing conditions for oil‑field equipment, identifies cavitation via IVMD algorithm for ship pumps, combines manual stethoscopy and spectrum inspection for water supply stations, and eliminates assembly‑related abnormal noise for power‑station oil pumps.

Challenges

Strong on‑site background noise masks weak fault signals; mixed acoustic features appear under coupled faults such as cavitation plus bearing damage; signal models lack universality across different pump types and operating conditions; conventional feature extraction relies heavily on human experience, and automatic mining algorithms are still under optimization.

5. Conclusion

Abnormal noise acts as one of the most sensitive early warning indicators for centrifugal pumps. Combining traditional manual listening with acoustic emission and AI algorithms forms a cost‑effective diagnosis solution. With wider adoption of Industrial IoT and edge computing, online acoustic monitoring will gain popularity. It helps avoid severe failures including impeller pitting and bearing seizure, and improves the full‑life predictive‑maintenance framework for mining pump, mine drainage pump and products from multistage pump manufacturer.

Previous: How to Diagnose & Prevent Cavitation in Boiler Feed Pump?

Next: How to choose the model of a mixed flow pump?

RELATED INFORMATION

How to Diagnose & Prevent Cavitation in Boiler Feed Pump?