Patching every shutdown or relining once: how to tell which one you are actually doing

A position that is touched up at every shutdown rarely appears in a review, because no single intervention is large enough to trigger one. It is absorbed into general maintenance, carried out by people already on site, and closed without a record that connects it to the previous time. Three years later nobody can say how many times it has been done.

The first useful step costs nothing: tag repeat positions by location rather than by work order, so the count accumulates somewhere visible. A position that appears four times in a row is making an argument, and it cannot make it while each appearance is filed separately.

Reaching the position usually costs more than filling it. Scaffolding or a platform, confined-space entry, isolation and lock-out, gas testing, a standby attendant, surface preparation, the cure wait, and the outage minutes the whole sequence consumes — all of these recur every time, and none of them scale down because the repair itself is small.

Expressing the comparison in material price alone therefore compares the two smallest numbers available. Express it per intervention instead, including access and outage, and the arithmetic often reverses without anything else changing.

Patching wins outright in several cases, and treating relining as automatically superior is its own error. Equipment with a defined end of life, where a durable construction would outlive the asset. Positions expected to change when a process is modified, so that a permanent installation would be removed before it paid for itself. Assets awaiting a capital decision, where committing a large sum now would prejudge it. And unstable or damaged shells where a rigid construction cannot be retained until the structure is repaired.

In each of those, the patch is a deliberate decision with a stated horizon — which is exactly what distinguishes it from the same patch applied because nobody decided anything.

The choice is rarely all-or-nothing on a whole asset. Converting the two or three positions that account for most of the repeat work, and continuing to patch the rest, is usually both cheaper and lower-risk than a single large project — and it produces evidence. If the converted zone survives a campaign that the patched zones did not, the case for the next conversion writes itself from your own plant.

Pick the positions by repeat count and access cost, not by visible severity. The worst-looking wear is often in an accessible place that is cheap to maintain, while the expensive one is a modest amount of loss somewhere that takes a day to reach.

Five fields are enough: position, date, what was done, how long the access took, and a photograph from a repeatable viewpoint. Kept per position rather than per work order, that record answers the conversion question within two or three outages, without any modelling and without anyone having to estimate a service life.

Bring that record to any supplier discussion. A proposal built on your intervention history addresses the position that is actually costing you; one built on a general claim addresses whichever position the claim was written about.

Repeat repairs persist mainly because no one owns the choice. The work is authorised at a level where each instance is routine, and the cumulative pattern sits above that level where nobody sees it. Naming an owner for repeat positions, and reviewing the list once a year, is a governance change rather than a technical one — and it is usually the change that unlocks the rest.

The decision that follows may well be to keep patching. That is a legitimate outcome, and a documented one is worth considerably more than the same practice continued by default.