Conveyor Belt Misalignment and Pulley Bearing Wear Detection

Conveyor belt misalignment is usually treated as a housekeeping problem: the belt tracks off-centre, the edge frays, a trainer idler gets replaced, production continues. The failure mode that matters is quieter. A persistent tracking offset applies a lateral and axial load path into the pulley shaft, through the bearing races, and into the housing — and it does so at load levels that a condition-monitoring wear sensor may not register as abnormal until the damage is already established.

This article dismantles that specific mechanism and frames the response in IEC 61511 terms: diagnostic coverage, proof-test interval, and the difference between detecting a degraded bearing and detecting the cause of the degradation.

The mechanism: misalignment loads the bearing, not the belt

A belt running true distributes tension symmetrically across the pulley face. The resultant radial load on the shaft is largely vertical and predictable. When the belt tracks to one side, two things change.

First, the effective contact patch shifts toward one edge of the pulley. Tension on the loaded side rises; tension on the slack side falls. The shaft now carries a bending moment it was not designed to carry continuously.

Second, the belt edge contacts the pulley flange or the structure. That contact introduces an axial thrust component along the shaft axis. Deep-groove ball bearings and spherical roller bearings tolerate some axial load, but not a sustained one superimposed on an already elevated radial load.

The consequence is a load zone that migrates around the raceway. Instead of a stationary loaded arc — the condition bearing life calculations assume — the load path moves as the belt wanders. Raceway fatigue accelerates, lubricant film breaks down at the edges of the contact ellipse, and spalling initiates earlier than the L10 life model predicts. The bearing does not fail because it is worn out. It fails because it has been loaded in a way the design case excluded.

Why the wear sensor arrives late

Vibration and temperature sensors on pulley bearings measure the symptom, not the cause. A bearing in early spalling produces broadband vibration energy that may sit inside the normal operating envelope for that asset — particularly on a large conveyor where structural resonance and belt-induced excitation already dominate the spectrum.

Temperature is worse. A bearing can run at stable temperature while subsurface fatigue propagates. Temperature rise becomes measurable when friction increases, which is after the raceway has already degraded. By the time a threshold alarm annunciates, the maintenance decision has shifted from “correct the tracking” to “replace the pulley assembly.”

This is the diagnostic-coverage gap. The sensor loop may have high coverage against gross bearing failure and effectively zero coverage against the initiating cause — belt misalignment — because misalignment is not a bearing variable. It is a geometric relationship between belt, pulley, and structure.

Framing it in IEC 61511 terms

IEC 61511-1 requires that the performance of each safety instrumented function be assessed against its required SIL, and that diagnostic coverage and proof-test intervals be justified for the detected failure modes. Two points follow for conveyor drive and pulley monitoring.

**Separate the failure modes.** “Pulley bearing failure” and “belt misalignment” are distinct failure modes with distinct detection methods. A safety instrumented function credited against bearing failure does not automatically credit misalignment detection. If the hazard analysis identifies belt misalignment as an initiating cause of a hazardous event — belt damage, material spillage into a walkway, pulley seizure on a loaded conveyor — then misalignment needs its own detection path and its own coverage claim.

**State diagnostic coverage honestly.** Diagnostic coverage is the ratio of detected dangerous failures to total dangerous failures for the diagnosed component. A wear sensor that only responds after spalling has initiated contributes little to coverage of the misalignment-initiated failure path. Claiming otherwise inflates the SIL verification result.

A defensible monitoring and proof-test strategy

The engineering approach is to instrument the cause and proof-test the detection chain.

For misalignment, the measurable variables are belt edge position relative to the pulley centreline, and the differential load between left and right sides of the structure. Belt position sensing — optical, ultrasonic, or mechanical edge rollers with position feedback — gives a direct indication of tracking offset. Load-pin or strain instrumentation on the pulley support structure gives an indirect indication of asymmetric loading. Neither is a safety device in itself; both are inputs to a function whose integrity must be assessed.

For the bearing, the useful proof test is not a vibration spot check. It is a periodic verification that the detection chain responds to a known input. Options include:

• Injecting a known tracking offset during a planned shutdown and confirming the detection function annunciates within its stated response time.
• Verifying edge-sensor alignment and calibration against a reference straightedge on the pulley face.
• Trending bearing envelope acceleration against a baseline captured at commissioning, with alarm thresholds set from that baseline rather than generic ISO 10816 zones.
• Inspecting pulley flange wear and shaft end-play at each major shutdown, since axial clearance growth is a direct indicator that thrust loading has occurred.

The proof-test interval should be derived from the failure rate you are claiming, not from the plant’s general maintenance calendar. If the misalignment detection function is credited with a dangerous undetected failure rate, the interval must be short enough to meet the target risk reduction factor.

Getting started

Pick one conveyor with a history of belt tracking problems and a pulley bearing replacement in the last 24 months. Map the failure sequence from first tracking offset to bearing replacement, using maintenance records and any available vibration or temperature history. Identify which detection method would have caught the initiating misalignment, and at what point. Then write the proof test for that method, including the input you will inject and the response you will verify. That exercise produces the diagnostic-coverage argument you need — and it usually shows the existing wear sensor was never covering the failure mode that caused the downtime.

Source: internal-backlog — grounded in S5 engineering design notes, not a current market event.

Talk to S5 System about your site

S5 System designs and manufactures ultra-low-power industrial IoT sensors for mining, agriculture, infrastructure and utilities.

Discuss your IoT monitoring project — book a call.