What Makes a Good Outage Inspection?

The Problem With Poor Inspections

Poor inspections usually share a common pattern: they are organised around activities rather than decisions. The inspector works through a list of things to look at, records what is seen, and produces a report that describes observations. The question of what to do with those observations — which ones require action, which ones don't, in which order and why — is left unanswered.

This creates a second phase of work after the inspection, where someone else tries to interpret the findings and make decisions under time pressure without having been in the machine. In many cases, the person making the decision was not the person who made the observation. Important context is lost. The result is either over-maintenance (replacing things that didn't need replacing) or under-maintenance (missing things that did).

Preparation Before Opening

A good inspection starts before the machine is opened. The following preparation directly improves inspection quality:

• Review operating history: vibration trends, alarm history, bearing temperature trends and any operational events since the last outage. This tells you where to look first and what anomalies to expect or watch for.
• Review previous outage findings: what was found last time, what was repaired, and what was left for monitoring. Any "monitor until next outage" findings become priority inspection items.
• Define scope before entry: which components must be inspected, which are conditional on findings, and which are definitely not in scope. A clear scope means that decisions about additional work can be made consciously, not by default.
• Prepare documentation in advance: measurement sheets, sketch formats, and photo labelling convention. Documentation prepared in advance is more complete than documentation improvised during the inspection.

During the Inspection: What Good Practice Looks Like

Sequence and priority

The most critical components should be inspected first, not last. In a time-constrained outage, if the outage window closes unexpectedly, the critical items should already be done. Bearings, rotor surfaces at steam paths and coupling condition are typically the highest priority; casing external hardware and instrumentation checks are typically lower.

Measurement discipline

Clearance measurements, alignment readings and dimensional measurements must be taken carefully, recorded immediately (not later from memory) and documented in consistent formats. The value of a measurement is only as good as the recording of it. A bearing clearance measured but not written down is the same as a bearing clearance not measured.

Photography

Every significant finding should be photographed before any cleaning or remediation is performed. A cleaned bearing surface contains less information than the same surface before cleaning. Label photographs consistently with component name, location and orientation — a photograph without clear labelling is difficult to use weeks later when a decision is being made.

On-site finding assessment

The inspection should not only record findings — it should assess them in the field, while the inspector still has direct access to the component. For each significant finding:

• What is it? (Description and location)
• How severe is it, based on what is visible?
• What does it indicate about the cause?
• What action is recommended?

This assessment may be refined later when laboratory results are available, but the initial field assessment made with the component in front of you is often the most valuable data point.

Prioritising Findings

Every significant finding should be assessed against two dimensions: severity and urgency.

• Severity — what happens if this finding is left as-is? Will it cause a failure before the next planned outage? Is the failure mode potentially catastrophic or would it give warning?
• Urgency — does this need to be addressed in the current outage window, or can it be scheduled for the next? Can it be addressed during operation with corrective maintenance?

A finding that is severe and urgent requires repair in the current outage. A finding that is low severity and low urgency can be documented and included in the next outage scope. Findings between these extremes require judgement — which is exactly why the field assessment by someone with experience is valuable.

What a Good Handover Looks Like

The inspection is only complete when the findings have been properly handed over to the decision-makers. A good handover includes:

• A summary of what was inspected and what was not
• A prioritised finding list — not an unstructured list of observations
• For each significant finding: description, location, severity assessment, recommended action and basis for the recommendation
• Photographs referenced to the finding descriptions
• A clear recommendation on what to repair in this outage, what to monitor and what can wait
• Open questions — what is not yet resolved and what additional information or action would resolve it