Contact
Home/Insights/

API Inspection

API Inspection

Brief

Risk-based inspection: spending inspection budget where the risk is

How risk-based inspection under API 580 and API 581 directs inspection effort to the equipment that matters, and what it needs to stay credible.

September 5, 2026

5 min read

KARR technical team
Risk-based inspection: spending inspection budget where the risk is
At a glance
  • API 580 sets the principles of an RBI programme, while API 581 provides a specific quantitative methodology.
  • Risk combines probability and consequence of failure, and keeping them separate leads to better inspection and mitigation choices.
  • Credible damage mechanisms, with API 571 as a common reference, must be matched to techniques that can detect them.
  • RBI only works as a living programme fed by good data, and it can sharpen turnaround scope considerably.

Every inspection budget is finite, and most plants have more equipment than they can inspect thoroughly in a single cycle. A purely time-based programme spreads effort evenly, so low-risk items get the same attention as equipment where a failure would be serious. The result is often money spent confirming that benign equipment is still benign, while the items that matter are inspected with the wrong technique or not often enough.

Risk-based inspection (RBI) addresses this by ranking equipment according to the risk it carries and directing inspection effort accordingly. Done well, it gives a defensible reason for every line on the inspection plan. Done badly, it becomes a software exercise that produces confident numbers from weak data.

API 580 and API 581

The two key API documents do different jobs. API 580, Risk-Based Inspection, sets out the principles and minimum elements of an RBI programme. It is deliberately methodology-neutral: it explains what a credible programme should contain without dictating how risk must be calculated.

API 581, Risk-Based Inspection Methodology, provides a specific quantitative approach. It defines how probability and consequence of failure are calculated, including damage factors, a management systems factor and consequence models. Many commercial RBI software tools are built on it. A plant can run a qualitative or semi-quantitative programme consistent with API 580 without adopting API 581 in full, but it should be able to explain and justify the method it has chosen.

Probability and consequence

Risk in RBI is the combination of how likely a failure is and how bad it would be. Probability of failure depends on which damage mechanisms are active, how fast they progress and how effective past inspections have been at detecting them. Consequence of failure considers flammable and toxic releases, injury, environmental harm and business interruption.

Keeping the two separate is useful. Equipment with high consequence but low probability may need modest inspection and strong mitigation, such as isolation or detection systems. Equipment with high probability and modest consequence may need more frequent or better-targeted inspection. Collapsing both into a single colour on a risk matrix hides that distinction.

Damage mechanisms drive everything

An RBI assessment is only as good as its understanding of how equipment degrades. API 571 describes damage mechanisms affecting fixed equipment in the refining industry, from general and localised corrosion to environmental cracking and high-temperature damage, and is widely used for screening. Each credible mechanism should be linked to an inspection technique that can actually find it. Spot thickness readings will not reliably detect most cracking, and an external visual inspection will not size internal wall loss.

Inspection effectiveness matters as much as inspection frequency.

Data quality and the living programme

RBI depends on data that many plants hold only in fragments. The inputs that most often weaken an assessment include:

  • Process data: operating conditions, fluid composition and contaminants that drive damage mechanisms.
  • Inspection history: past findings, techniques used and coverage achieved, not just inspection dates.
  • Corrosion rates: measured rates at fixed locations rather than assumed or default values.
  • Design and materials: construction details, materials of construction and any modifications since.
  • Integrity operating windows: limits that keep damage rates within the assumptions of the assessment.

Where data is missing, conservative assumptions are reasonable, but they should be recorded so the gap can be closed later. RBI is also not a one-off study. Process changes, new inspection findings, changes in feed and management of change events should all trigger a review. A programme that is never updated slowly drifts back into a time-based plan with a risk label attached.

From RBI to turnaround scope

One of the clearest benefits of RBI is a better turnaround worklist. Equipment whose risk can be managed by on-stream inspection with suitable NDT can come out of the shutdown scope, while high-risk items that genuinely need internal access are prioritised. A straightforward way to use RBI in scope development:

  1. Confirm the assessment is current and reflects recent operating history and findings.
  2. Identify equipment whose risk will exceed the acceptance threshold before the next planned opportunity.
  3. For each item, decide whether on-stream inspection can manage the risk or internal access is needed.
  4. Specify the technique and coverage required for the credible damage mechanisms.
  5. Feed the results into the turnaround worklist and record the reasoning for each decision.

That record matters. When scope is challenged on cost grounds, a documented risk basis makes it far easier to explain why an item stays in or comes out.

A practical first step

Take your ten highest-ranked items and check three things: are the damage mechanisms credible, can the planned technique detect them, and when was the assessment last updated? The answers will tell you how much confidence the programme deserves. If you would like a second opinion, KARR's inspection engineers can review the assessment or help rebuild the data behind it.

Topics

RBI, API 580, Turnaround planning

From insight to site
Want this applied to your assets?
Talk to our team
Keep reading

More insights.

View all insights →