RI-ISI program design that survives operator turnover
RI-ISI programs are designed for engineers who will not be in the role when the next inspection cycle comes around. We design for the next engineer, not the current one.
The risk-informed in-service inspection methodology produces an inspection schedule that is technically defensible against the plant's PRA at the time the schedule is created. The schedule then has to be executed by inspection engineers, often years later, who were not part of the original design team. If the schedule is not transparent enough that the next engineer can understand why each inspection was chosen, the next engineer has to either re-derive the schedule (expensive) or operate it as a black-box checklist (loses the risk-informed character of the program).
The pattern that produces the un-survivable program is to document the schedule by its outputs (the list of welds and frequencies) without documenting the inputs (the PRA importance rankings, the consequence categories, the inspection-history considerations) that produced those outputs. The next engineer sees the list and trusts it; when the next plant change happens, they cannot tell which inspections need to be re-evaluated and which are unaffected.
The pattern that survives is to document each inspection's selection rationale alongside the inspection itself: this weld is on this segment, the segment's importance rank is X, the segment's consequence category is Y, the inspection frequency was set against Z, the assumptions about plant configuration that supported Z are documented separately. When the plant changes, the new engineer can walk down the rationale and update only the inspections affected by the change.
The work to produce this rationale documentation at design time is modest; the work to recreate it years later, when the plant has changed and the original engineer has moved on, is substantially larger. We design for the recreation cost, not the design cost. The licensee that hires us five years later thanks us for it.