Skip to content
Beaconby Winda
FactAsset Integrity Management · Risk-based inspection

Qualitative vs semi-quantitative vs quantitative RBI

·6 sources

Short answer

Qualitative risk-based inspection (RBI) ranks probability and consequence of failure by expert judgement, usually as low, medium or high on a risk matrix. Quantitative RBI calculates them from models, so it can predict when a risk limit will be reached. Semi-quantitative RBI mixes the two. API RP 580 allows all three; API RP 581 is quantitative.

By , Editor·Researched and drafted by ·Checked by , Editor·Last checked ·Next review

Why does it matter?

Numbers are only as good as their inputs: after the 2010 heat exchanger rupture at Anacortes, Washington, the US Chemical Safety and Hazard Investigation Board (CSB) showed that the 2008 edition of American Petroleum Institute (API) RP 581, fed the design conditions used in the unit's damage reviews, gave the failed exchangers a low susceptibility to high temperature hydrogen attack (HTHA) 2. Seven workers died 2.

Judgement-based methods have their own blind spots. In a 2015 study of US gas pipeline programs, the National Transportation Safety Board (NTSB) found risk models with hundreds of factors, each with its own scale and weight, that state inspectors felt unequipped to challenge; it concluded that nobody knew whether the different approaches allowed by the rules give the same safety benefit 4. Choosing a method is also choosing what you will later be able to check.

Qualitative vs semi-quantitative vs quantitative at a glance

The labels are used loosely. US gas pipeline rules let operators choose among four kinds of risk model: judgement by subject matter experts, relative models that score and weight factors, scenario models built from event or fault trees, and probabilistic models 4. A 2018 report by the Pipeline and Hazardous Materials Safety Administration (PHMSA) grouped them as qualitative, relative or index, quantitative system, and probabilistic 5. In the table, expert judgement stands for qualitative, scored and weighted models for semi-quantitative, and probabilistic models for quantitative.

QualitativeSemi-quantitativeQuantitative
How likelihood and consequence are setBy the collective judgement of people who know the system 4Rating scales turn facts such as coating type or soil into scores, which are weighted and added 4Probabilities from fault trees, reliability models or Bayesian networks, and modeled release outcomes 6
What comes outA cell on a risk matrix, or a priority number from frequency and severity ratings 6A relative score for ranking, with no physical unit 4An absolute risk, such as expected deaths or dollars per mile per year 4
What it can supportRanking and first priorities 6Ranking segments or items for assessment 4A target risk level, cost-benefit comparisons and inspection timing 4,1
What it needsExperienced people, with time to write down the reasoning behind each estimate 4,3Weights chosen and checked against operating history 4Failure data that many operators do not yet have 4
Its weak pointResults checked mostly against the team's own sense of where risk lies 4Weights that suit some threats can mute others 4Wrong inputs give confident wrong answers 2

Sources: NTSB/SS-15/01, 2015 4; Collina et al., 2026 6; API 2025 catalog, RP 580 and RP 581 entries 1; CSB report 2010-08-I-WA 2.

API RP 580 (4th edition, August 2023) sets the elements of an RBI program and allows a spectrum from qualitative to fully quantitative; API RP 581 (3rd edition, April 2016) supplies quantitative procedures 1. See API 580 vs API 581: what is the difference?.

When is qualitative RBI used?

  • To find the hazards in the first place. Structured judgement tools such as hazard and operability studies and failure mode and effects analysis are the usual way to list how each component could fail, according to a 2026 review 6. See What is FMEA?.
  • Where data are thin. The same review notes that studies of electrolyzers, a new component in hydrogen plants, have stayed qualitative because operating and reliability data are limited 6. NTSB found pipeline operators held back from probabilistic models for the same reason 4.
  • As the everyday method. Six of the seven operators NTSB interviewed used relative models and the seventh relied on experts, and NTSB saw little difference between the two, since the relative models drew heavily on expert input 4.
  • To check other models. Operators validated their rankings by review: engineers looked at the highest- and lowest-ranked segments each year and adjusted weights until the results matched experience 4.

When is semi-quantitative RBI used?

  • To rank many items on one scale. Relative models give each threat on each segment a score, combine the threats with weights, multiply by a consequence score and rank the result 4.
  • To mix tools. The 2026 review joins failure mode and effects analysis, which identifies scenarios, with Bayesian networks, which estimate how likely each is and can be revised as data arrive 6.
  • With care over weights. Weights let unlike threats be added without a common unit. NTSB found operators starting from defaults and tuning them by experience, and warned that one set of weights for a whole system can understate a segment with an unusual threat 4.

When is quantitative RBI used?

  • To time inspections against a risk target. Quantitative results show when the risk acceptance limit will be crossed as damage grows 1. Where the likelihood estimate is good enough, the plan can be worked backward from a target probability of failure 6. See Risk target.
  • To compare options in common units. Operators moving to probabilistic models told NTSB they wanted risk in units managers understand, a target risk level and cost-benefit comparison of measures and schedules 4. PHMSA's report weighed each model type against decisions such as setting assessment intervals and benefit-cost analysis 5.
  • Outside oil and gas. In 1992 an American Society of Mechanical Engineers (ASME) task force described using probabilistic risk assessment to quantify rupture risk and set inspection priorities for nuclear plant components 3. It started from the view that inspection programs then rested on experience and engineering judgement, with risk considered at best implicitly 3.

A common mix-up

Quantitative is not the same as objective. Even the ASME task force's probabilistic method drew rupture probabilities from a panel of experts, because ruptures are rare and history gives little data 3. At Anacortes, design-condition inputs gave a low susceptibility result for equipment that had in fact failed 2. NTSB added that the safety benefit of probabilistic models over the other types had not been studied much 4. A number carries the judgement that went into it.

Where integrity problems show up

DecisionData it depends onWhat goes wrong without itIn the Library
Trusting a quantitative outputActual operating conditions as inputsAnacortes, 2010: the 2008 RP 581 HTHA calculation with design conditions gave the failed exchangers low susceptibility; the CSB judged it unreliable for preventing HTHA failure in carbon steel 2What is high temperature hydrogen attack (HTHA)?
Setting weights in a scored modelThreats specific to each item or segmentNTSB: system-wide weights that favor a few threats can understate a pipeline crossing a fault line 4Risk matrix
Moving to a probabilistic modelFrequency data for each failure mechanismNTSB: many operators lack the data to run probabilistic models well 4Probability of failure
Reviewing an operator's modelTraining and a defined, documented logicNTSB: inspectors lacked the training to verify whether a risk assessment was technically valid 4What is risk-based inspection (RBI)?

Sources: CSB report 2010-08-I-WA, 2014 2; NTSB/SS-15/01, 2015 4.

Common questions

Is API 581 the only quantitative RBI method?
No. Quantitative risk methods for inspection are older and wider than one document. In 1992 an ASME task force applied probabilistic risk assessment to rank nuclear plant components for inspection, and US gas pipeline rules allow probabilistic models that give risk in absolute units.
Which approach do pipeline operators use in practice?
In NTSB's 2015 study, six of the seven gas transmission operators interviewed used relative index models and the seventh relied on subject matter experts. Several were moving toward probabilistic models but said they lacked the failure data to do it.
Is semi-quantitative RBI less rigorous?
Not by definition. A 2026 review pairs qualitative tools such as failure mode and effects analysis with probabilistic models, using each where it fits. What decides rigor is whether the inputs and the weights have been checked.

Beacon explains these concepts in its own words for education. It does not reproduce or replace the standards cited, and Winda is not affiliated with or endorsed by the organizations that publish them. Engineering decisions need the current edition of the standard and a qualified engineer.

Sources

  1. 1.2025 Publications Catalog (entries for RP 580 and RP 581) (RP 580, 4th edition, August 2023; RP 581, 3rd edition, April 2016) · American Petroleum Institute (API)
  2. 2.Investigation Report: Catastrophic Rupture of Heat Exchanger (Seven Fatalities), Tesoro Anacortes Refinery, Anacortes, Washington, April 2, 2010, Report No. 2010-08-I-WA · US Chemical Safety and Hazard Investigation Board (CSB)
  3. 3.Risk Based ASME Code Requirements (B. F. Gore, T. V. Vo, K. R. Balkey), PNL-SA-21211, OSTI 7129942 · Pacific Northwest Laboratory, US Department of Energy, via OSTI.gov
  4. 4.Integrity Management of Gas Transmission Pipelines in High Consequence Areas, Safety Study NTSB/SS-15/01 · US National Transportation Safety Board (NTSB) ·
  5. 5.Pipeline Safety: Gas and Hazardous Liquid Pipeline Risk Models, notice and request for comments, 83 FR 40843 (FR Doc. 2018-17659) · Federal Register, US Pipeline and Hazardous Materials Safety Administration (PHMSA) ·
  6. 6.Toward the application of the Risk-Based Maintenance approach to a hydrogen-fueled manufacturing plant (G. Collina and others), International Journal of Hydrogen Energy 238, 155318, open access under CC BY 4.0 · Elsevier, via OSTI.gov (co-authored by the US National Laboratory of the Rockies)