Defending an RBI-Derived Inspection Interval

Extending an interval is a decision you will be asked to justify years later, by someone who was not in the room.

Risk-based inspection earns its keep by moving effort away from equipment that has been demonstrably stable and toward equipment that is genuinely degrading. That only works if the underlying condition data supports the probability-of-failure input — and when RBI runs on default corrosion rates rather than measured ones, an interval extension is a decision resting on an assumption. Regulators and clients increasingly test exactly that.

What has to be defensible

  • The corrosion rate: measured from a real thickness time series at identified CMLs, with the readings, dates, instruments and technicians behind it recoverable.
  • The damage mechanisms assigned per API RP 571, and why they are the credible ones for that circuit's actual service.
  • The consequence-of-failure basis — inventory, fluid, location, and the assumptions behind them.
  • The RBI methodology and its version, plus any deviations from the standard approach and their justification.
  • The competence of the people who performed the assessment, and the review and approval chain.
  • The integrity operating windows the assessment assumed, and evidence that operation stayed inside them.

The weak link is almost always the condition data

RBI methodology is well documented and rarely the point of challenge. What gets challenged is the input: a corrosion rate derived from two readings whose CML identity cannot be proven, a thickness history missing the instrument and technician provenance, or a default rate quietly substituted where measurements were unavailable and never revisited when they became available.

This is why CML register quality and inspection-evidence provenance are RBI issues rather than administrative ones. The methodology inherits the credibility of its inputs, and no amount of analytical rigour compensates for a broken time series.

Integrity operating windows are part of the argument

An RBI assessment assumes an operating envelope. If the unit subsequently ran outside it — higher temperature, different feedstock sulphur, changed water chemistry — the damage-mechanism susceptibility that underpinned the assessment may no longer hold, and the extended interval is no longer supported. Demonstrating that operation stayed within the assumed windows, or that excursions were assessed, is part of defending the interval and is frequently the part nobody has evidence for.

Building the trail as you go

  • Bind every thickness reading to a persistent CML identity, with instrument, technician and procedure provenance captured automatically.
  • Record damage-mechanism assignment with its rationale and the process-service basis, not just the mechanism name.
  • Version the assessment: inputs, methodology, results and approvals frozen at the point of decision.
  • Re-run the assessment on new data rather than accepting the old interval by default; an interval that has never been re-examined is a weak position.
  • Track excursions against integrity operating windows and link them to the affected assessments.

Frequently Asked Questions

Is RBI accepted as a basis for extending statutory intervals?

It depends entirely on jurisdiction and on the regulator, and the answer ranges from broadly accepted to not permitted for certain equipment classes. What is consistent is that where RBI is accepted, the burden of demonstrating the technical basis sits with the operator — so the quality of the condition data and the audit trail determines whether the position holds under examination.

What happens when a new inspection contradicts the RBI assumption?

The assessment should be re-run rather than the reading treated as an outlier. A measured corrosion rate materially higher than assumed invalidates the interval that was derived from the assumption, and the correct response is to reassess and, if necessary, bring the inspection forward. Documented reassessment is a sign of a functioning programme; unexplained readings left in place are the opposite.

Can RBI be run without a good CML register?

It can be run; it cannot be defended. Probability of failure driven by default corrosion rates rather than measured ones produces a ranking that reflects assumptions rather than condition, and any interval extension derived from it is exposed the moment someone asks how the corrosion rate was obtained.

How often should assessments be refreshed?

Whenever significant new condition data arrives, whenever process service or operating windows change, and on a defined cycle regardless — commonly aligned to the inspection cycle. The practical failure mode is an assessment performed once at programme inception and inherited unchanged for a decade, which is difficult to defend however sound the original work was.

See it running on your own workflow

Thirty minutes, your job types and your reporting formats, co-presented by an ASNT NDT Level III. Affordable, accessible, fully customizable — request a demo and a tailored quote.

Related: Asset integrity management software · RBI programme design consulting · Atlantis Digital Twin platform · API 510 certification guide