Risk-based inspection
API 580 and 581: risk-based inspection, explained
API 580 is the American Petroleum Institute's recommended practice that defines what a risk-based inspection program must contain, from data collection through risk ranking, inspection planning and reassessment. API 581 is the companion methodology that supplies the quantitative procedures for probability of failure, consequence of failure and risk. Together they are the basis the inspection codes cite when they allow RBI to set intervals. They do not replace API 510, 570 or 653, and API 581's numbers are only as good as the damage mechanism and inspection effectiveness inputs behind them.
What the standard requires
Program elements under API 580
A compliant program has a documented scope, data requirements, a defined method for probability and consequence, a risk ranking, an inspection plan tied to the ranking, a reassessment cycle, and management of change. Each element has to be written down and reviewable.
Probability of failure
Probability is built from the damage mechanisms credible for the equipment, their rates or susceptibility, the equipment's age and the effectiveness of past inspections. API 581 provides damage factors per mechanism; a qualitative or semi-quantitative program scores the same inputs on a defined scale.
Consequence of failure
Consequence covers the safety, environmental and business outcome of a release, driven by fluid, inventory, pressure, temperature, detection and isolation. API 581 gives a calculated area or financial consequence; simpler programs rank it on a defined scale.
Risk ranking and the matrix
Risk is probability multiplied by consequence, placed on a matrix so the equipment can be ranked and the inspection effort concentrated where the risk is. The matrix and its thresholds must be defined before the ranking is done, not fitted afterward.
Inspection planning
The plan sets inspection method, coverage and date to reduce the probability of failure through better knowledge of the damage state. Inspection effectiveness has a defined grade, and the interval must still respect the caps the inspection codes set.
Reassessment
Risk is reassessed after each inspection, after a process or design change, and at a maximum interval the program sets. A ranking that has not been revisited since it was first produced is not a defensible basis for an extended interval.
What the record must carry
- Equipment identification, service, fluid and operating conditions
- Credible damage mechanisms with the basis for each
- Probability of failure inputs, score or damage factor, and method used
- Consequence of failure inputs and score or calculated value
- Risk rank and the matrix position with the thresholds in force
- Inspection plan: method, coverage, effectiveness grade and due date
- Reassessment date and trigger, and the change history behind each rerank
- Reviewer and approver signatures on the ranking and the plan
How intelliSPEC handles it
intelliINTEGRITY carries API 580/581 semi-quantitative RBI scoring, probability times consequence, with a rank from 1 to 5 that drives the inspection plan. Damage mechanisms are first-class records, so the credible mechanisms behind a probability score are linked to the equipment rather than typed into a comment, and the corrosion summary supplies the rates. Interval outcomes still respect the API 510/570 cap at half remaining life, and every rerank, approval, rejection, hold and owner sign-off is a workflow transition recorded in the hash-chained audit trail.
Answers
API 580 and 581 questions
What is the difference between API 580 and API 581?
API 580 is the recommended practice that says what an RBI program has to contain and how it is governed. API 581 is the methodology that supplies the calculations for probability of failure, consequence of failure and risk. A program can follow API 580 with a qualitative or semi-quantitative method; API 581 is the fully quantitative route.
Can RBI extend the inspection interval beyond the API 510 or 570 limit?
The inspection codes allow an RBI assessment to set an interval different from the half-life rule, provided the assessment is documented, reviewed and reassessed at the frequency the program sets. The codes still place an upper limit on how far the interval can be extended, so RBI is a basis for the interval, not a way around the cap.
What makes an RBI ranking defensible in an audit?
The auditor will look for the inputs behind the score, the method that turned them into a rank, the matrix and thresholds that were in force, the inspection plan the rank produced, and evidence the rank was reassessed after the last inspection. A rank with no traceable inputs or no reassessment history is the usual finding.
How often does an RBI assessment have to be updated?
After every inspection that changes what is known about the damage state, after any process or design change that alters probability or consequence, and at a maximum interval the program defines. API 580 requires that reassessment cycle to be written into the program.