How to compare wear measurements between shutdowns

A measurement programme should answer a maintenance question: whether a zone remains stable, whether a repair is progressing, where access should be prioritised or what must be prepared for the next shutdown. Collecting many readings without that purpose can create a dense file that still does not support a decision. Identify the equipment boundary, credible failure modes, high-consequence zones and information needed to distinguish continued monitoring from intervention. Then select locations and methods. The comparison becomes clearer because every point has a reason to exist, and missing data can be assessed by its effect on the intended decision.

A label such as lower chute wall is not precise enough for a trend. Use a stable datum, drawing coordinate, permanent marker or documented grid that can be recovered after cleaning and repair. Include orientation, distance from transitions and photographs showing both context and close detail. If the reference surface may itself move or wear, state that limitation and choose an alternative where possible. When a tile or section is replaced, preserve the history under a new component identifier rather than continuing the old trend as if nothing changed. Traceability depends on knowing that two readings describe the same physical location and construction.

Readings obtained with different instruments, coupling conditions, calibration states, operators or surface preparation may not be directly comparable. Record the method, equipment identification, verification status, accessible surface and any cleaning or grinding performed. Keep raw readings rather than only a rounded summary. If a coating, buildup, curved surface, joint or rough repair affects the method, describe the limitation at that point. Repeating an uncertain value does not automatically make it reliable. A strong record allows a reviewer to separate actual material change from measurement scatter and to decide when a second method or direct examination is justified.

Two shutdowns may be separated by similar calendar periods but very different service. Throughput, operating hours, feed source, particle distribution, moisture, upset events, cleaning and idle conditions can all change the exposure. Link each inspection campaign to the best available operating record and mark where data are incomplete. Normalisation can help only when the chosen exposure measure is relevant to the wear mechanism and consistently recorded. Do not hide variability behind one adjusted rate. Present the raw change, operating context and any derived comparison separately so that reviewers can understand both the observation and the assumptions used to interpret it.

A patch, replaced tile, weld build-up, backing repair or changed joint creates a new local condition. Mark its extent, date, material record and reason, then establish a fresh baseline for the repaired zone. Nearby points may still retain their earlier history, but their exposure can change if the repair alters geometry or flow. Photographs taken before removal, during hidden stages and after completion help explain later measurements. A database that silently overwrites repaired points can produce a smooth but false trend. Preserve superseded records and make the discontinuity visible to anyone planning the next inspection.

Summarise each comparison with location, observed change, operating exposure, data quality, plausible mechanism and consequence if the interpretation is wrong. A consistent pattern across neighbouring points and inspection methods may support more confidence than one isolated value. Where consequence is high and confidence is weak, the response may be better access, verification or contingency preparation rather than an unsupported life estimate. Where evidence is stable, the same framework supports continued monitoring. The value of trending lies in making the next decision more defensible, while preserving enough context for a future reviewer to challenge and improve the interpretation.

Assign responsibility for point definitions, revisions, instrument records and the approved comparison. Store the controlled dataset where maintenance and engineering can retrieve it before planning the next shutdown. Clear ownership reduces duplicate grids, lost photographs and undocumented changes. It also makes assumptions reviewable instead of leaving the trend dependent on one person’s memory.