Which assets should you maintain, and how? Rank by criticality, then choose by failure consequence

Start from what the asset must do

The international standard on asset management frames the job as coordinating asset-related decisions and activities so that the portfolio delivers the organisation's objectives over its lifecycle, and its 2024 revision stresses outcomes over a fixed set of maintenance tasks. Begin from criticality and business value, not from a generic task list.

Rank on consequence and likelihood

One published approach ranks asset criticality from two criteria: the consequences of failure and the probability of failure, in a risk-matrix framework. It weights the sub-criteria under each with the analytic hierarchy process and combines them with a fuzzy inference system. The authors tested it on a gas refinery and report that it was effective for ranking assets.

That is one application, and the weights rest on expert judgment, so the value to take from it is the structure: two axes, explicit written criteria, and a ranking other people can inspect and challenge. A simpler scoring on the same two axes is a sensible start; that suggestion is Portolan's, not the study's.

Then choose the task by consequence

Reliability-centred maintenance selects tasks by asking what the asset must do, how it can fail, and what each failure would cause. Where no preventive task is justified it uses a deliberate default (run to failure, redesign, or duplicate). Applying one interval to every asset wastes effort on low-consequence failures and misses hidden ones, so the consequence, not habit, should pick the task.

The method depends on reliable failure history, which means the same definitions of a failure and a repair across teams. If those differ, fix them first, or the ranking will rest on data that cannot tell assets apart.

Tie the ranking back to production

A case study linking maintenance to the quality, efficiency and effectiveness of operations in a paper mill makes the same point from the production side: judge a maintenance decision by the production outcome it protects, not only by its own cost.

The path, step by step

These steps come from the Maintenance and asset reliability playbook. Each has a checkpoint you can observe.

1. 01

   Classify assets by function and criticality

   List the assets that carry operational, safety or environmental consequence. For each, state the function it must perform and what happens when it fails.

   CheckpointA criticality ranking tied to operational and safety consequences, with the top assets named.

   Common failureTreating every asset as equally critical and spreading maintenance thin.

2. 02

   Capture failure history with consistent definitions

   Record failure mode, cause, detection, downtime and repair for critical assets. Use the same definitions across shifts and sites so the history can be compared.

   CheckpointA failure log with consistent categories and enough history to rank modes.

   Common failureChoosing maintenance strategy from vendor claims because the failure history is unusable.

3. 03

   Choose tasks by failure consequence

   For each significant failure mode, decide whether a scheduled task, condition-based check, redesign or run-to-failure is justified. Delete tasks that cannot address a credible mode.

   CheckpointA task decision per major failure mode, including the explicit run-to-failure decisions.

   Common failureAdding more preventive tasks to every asset and calling it reliability.

What this guide does not settle

The criticality method is one application at one refinery, and the standard and the method are not evidence that ranking assets improves reliability. The library has no comparison of criticality methods. Treat your ranking as a hypothesis to revise after the first review of real failures.