API RP 580 vs 581: Which RBI Document?

Back to Blog

The Decision in One Paragraph

If you are asking what a proper RBI program needs to include, read RP 580 - it sets the requirements. If you need a concrete, quantitative recipe to compute risk numbers, RP 581 gives one. Most programs cite both: 580 to establish they are doing RBI credibly, and 581 (or a comparable method) to show how the numbers were produced. Neither replaces the deterministic inspection codes; both feed them.

The general concept both documents serve is covered under risk-based inspection, and the program-building view is in the elements of an API RP 580 program. This page is specifically about which of the two documents to reach for.

Side-by-Side

The table draws the practical distinction. Both are recommended practices published by the American Petroleum Institute; the difference is scope and specificity.

Aspect	API RP 580	API RP 581
Role	Defines RBI requirements and principles	Provides a quantitative RBI methodology
Answers	What must the program contain?	How do I compute the risk numbers?
Nature	Framework / expectations	Detailed calculation method
Substitutable?	The requirements stand	Other credible methods may satisfy 580

When Each Wins

Reach for RP 580 when you are setting up governance, training an audit response, or deciding whether an existing program is credible. It is the yardstick for whether you are doing RBI at all, versus just calling something RBI. It governs the roles, the reassessment loop, and the documentation that make the program defensible.

Reach for RP 581 when you need to actually calculate. It supplies the damage-mechanism models, probability factors, and consequence models to turn condition and process data into risk numbers. Programs feeding an API 570 piping program or vessel intervals often lean on 581's math to justify the interval changes RP 580 permits.

Pitfalls

The first pitfall is treating RP 581's numbers as compliance by themselves. Running the 581 calculation without the 580 program around it - roles, reassessment, documentation - produces risk figures with no governance, which an auditor will not accept as a credible program.

The second is assuming 581 is the only acceptable method. RP 580 sets requirements that other quantitative or qualitative approaches can also satisfy. 581 is a well-known way to meet them, not the mandatory one. Locking into it without understanding that can make a program brittle when the tooling changes.

How RBI Output Feeds the Inspection Codes

Neither RP 580 nor RP 581 replaces the deterministic inspection codes; their output feeds them. The codes governing vessels and piping allow interval and scope decisions to be informed by an RBI assessment, and that is where the two recommended practices earn their keep: RP 580 makes the program credible enough to rely on, and RP 581-style math produces the risk ranking that justifies inspecting one item sooner and another later. The vessel-side interaction is described under API 510 pressure vessel inspection.

The practical loop looks like this: the RBI assessment ranks equipment by risk, the resulting inspection plan sets what to examine, where, how, and when, the inspection produces findings, and those findings update the damage-rate assumptions that drove the ranking. RP 580 requires that loop to exist and be documented; RP 581 supplies the recalculation each time new data lands.

The Data RP 581 Needs Before the Math Helps

RP 581's models are only as good as their inputs, and a common failure is running the calculation on defaults for everything. Before the quantitative method adds value over a qualitative ranking, a program needs, at minimum:

1. Design and construction data for each item - materials, thicknesses, design conditions.
2. Actual operating conditions, including excursions, not just the design values.
3. Identified damage mechanisms for each item, from a corrosion study or materials review.
4. Inspection history with measured thicknesses and the effectiveness of each past inspection.
5. Consequence inputs: the fluid, its inventory and phase, and what surrounds the equipment.

Where these are missing, RP 581's conservative defaults take over, and the computed risk reflects the defaults more than the plant. That is not a reason to skip the method; it is a reason to treat data gathering as the first project phase and to read the early numbers as provisional.

The Shape of the 581 Calculation, in Symbols

Stripped of its tables, RP 581 computes risk as the product of a probability of failure and a consequence of failure, each item evaluated over time: risk(t) equals POF(t) times COF. The probability side starts from a generic failure frequency for the equipment type and adjusts it with a damage factor, which grows as damage mechanisms progress and shrinks when effective inspection reduces uncertainty, and a management systems factor reflecting the quality of the surrounding programs. The consequence side is evaluated in area-based or financial terms from the fluid and inventory inputs.

Two behaviors of this structure are worth internalizing. First, inspection does not change the physical damage rate; it changes what you know, and the damage factor rewards that knowledge with a lower computed probability. Second, because the damage factor grows with time, every item's computed risk drifts upward between inspections, and that drift is exactly the mechanism that generates a recommended inspection date rather than a fixed calendar interval.

Frequently Asked Questions