# Probability of failure vs consequence of failure

> Probability of failure (PoF) is how likely equipment is to lose containment; consequence of failure (CoF) is how bad the loss would be. Risk-based inspection combines the two into risk, with PoF stated for a set period and CoF covering people, the environment and the business. Inspection mainly sharpens the PoF estimate; isolation, detection and inventory shape CoF.

- Label: Fact
- Type: Explainer
- Published: 2026-11-14T12:02:06+09:00
- By: Jun Zhong
- Researched and drafted by: Beacon's AI newsroom
- Checked by: Jun Zhong on 2026-11-14
- Topics: LNG, Refining, Upstream and offshore
- Canonical: https://beacon.winda.ai/library/pof-vs-cof/

## Why does it matter?

On 16 February 2007, at a refinery near Sunray, Texas, water that had settled in a control station idle for about 15 years froze and cracked an elbow in high-pressure propane piping [2]. That was the probability side: a hazard missed because no management of change review was done when the station was taken out of service [2]. Once propane escaped and ignited, operators could not stop the flow because no remotely operable shut-off valves were installed, and flame on unfireproofed support steel brought down a pipe rack [2]. Four people were injured, the refinery was evacuated and shut for two months, and reported direct losses topped $50 million: the consequence side, set by isolation and fireproofing rather than by the pipe wall [2].

## PoF vs CoF at a glance

In the American Petroleum Institute (API) method, risk-based inspection (RBI) treats failure as loss of containment: a leak to the atmosphere or a rupture [1]. The two halves of risk are estimated separately and then combined [5][6].

| | Probability of failure (PoF) | Consequence of failure (CoF) |
|---|---|---|
| Question it answers | How likely is a loss of containment over a set period? [1][5] | If it happens, what does it do to people, property and the environment? [6] |
| How it is expressed | A frequency rating, or a probability from fault trees, reliability models or Bayesian networks [6] | A severity rating, or modeled outcomes such as fire, explosion and toxic exposure, sometimes turned into money [6] |
| Main inputs | Damage mechanisms, material and coating, protection systems, surroundings and actual operating conditions [5][9] | The fluid, how much of it can escape, how fast the release can be stopped, and who is nearby [7][2][5] |
| How it changes over time | Rises as damage builds up toward a [risk target](/glossary/risk-target/) [1] | Changes when inventory, isolation or the population around the equipment changes [7][5] |
| What inspection does to it | Narrows the uncertainty about the damage present, so the estimate improves; the inspection itself does not lower risk [1] | Little: what a release does is set by what escapes and how fast it is stopped [2][3] |
| What lowers it | Removing or controlling the damage source, for example draining, isolating or removing an idle dead leg [2]; a more resistant material [10] | Remote isolation, leak detection, water curtains, smaller isolatable inventories [2][8] |
| A typical error | Rating it from design data instead of the conditions the unit really runs at [9] | Leaving out how quickly the flow can be stopped [5] |
| In the Library | [Probability of failure](/glossary/probability-of-failure/) | [Consequence of failure](/glossary/consequence-of-failure/) |

Sources: API 2025 catalog, RP 581 entry [1]; US Chemical Safety and Hazard Investigation Board (CSB) reports 2007-05-I-TX [2], 2010-08-I-WA [9] and 2012-03-I-CA [10]; NTSB/SS-15/01 [5]; Collina et al., 2026 [6]; 40 CFR 68.25 and 68.28 [7][8].

## When does probability of failure decide the action?

- **Timing and method of inspection.** API describes risk rising as damage accumulates toward a risk target, with an inspection recommended before it is reached [1]. In US pipeline rules the same logic sets the schedule: liquid lines that could affect high consequence areas are assessed at intervals based on the risk each line poses [12].
- **When the inputs are wrong.** At Anacortes, Washington, in 2010, the probability calculation for high temperature hydrogen attack, run on design conditions, rated as low the exchangers that later ruptured and killed seven workers [9]. CSB also judged inspection for that damage not reliable enough to depend on, and looked instead to inherently safer materials and to checking the conditions the equipment really sees [9][11].
- **When corrosion varies from part to part.** At Richmond, California, in 2012, thickness readings taken mostly on fittings high in silicon missed a low-silicon component that thinned much faster [10]. CSB pointed to two ways out: inspect every component, or replace the line with a steel higher in chromium, which it called more effective [10].

## When does consequence of failure decide the action?

- **Isolation.** After Sunray, the CSB found that the lack of remote isolation greatly increased the size and duration of the fire [2]. At a Philadelphia, Pennsylvania, refinery in 2019 it concluded that remotely operated emergency isolation valves could have shortened the release and prevented the later explosions [3].
- **Inventory.** US accident prevention rules for chemical plants start the worst case from the largest amount held in one vessel or one pipe, counting administrative limits on that amount [7]. The alternative scenario each plant must also report is a release more likely than the worst case, and it may take credit for active and passive mitigation that would survive the event [8]. Smaller isolatable inventories and working mitigation shrink the consequence whatever the probability.
- **Where people are.** Federal gas pipeline integrity programs exist because a failure in a populated area, a high consequence area, carries far more risk; operators have had to run them since 2004 [5]. In NTSB's 2015 study, operators' consequence models leaned mainly on population density, and none counted emergency response factors such as how far away the shut-off valves were or whether they could be closed remotely [5].

## A common mix-up

Risk is a combination, not an average. A 1992 paper by the American Society of Mechanical Engineers (ASME) risk-based inspection task force expressed risk as rupture frequency multiplied by consequence [4]. Probabilistic models can give risk in absolute units, such as deaths per mile per year; relative models have no common unit, so they rely on weights to add different threats and consequences together [5]. Averaging can hide the item that matters: NTSB warned that a length-weighted risk figure over a long stretch of pipeline may bury a short segment with high risk [5]. Plotting the two on a [risk matrix](/glossary/risk-matrix/), consequence on one axis and likelihood on the other, keeps them visible side by side [6]. PoF is also not the same as [probability of detection (POD)](/glossary/probability-of-detection/), the chance that an inspection finds a flaw.

## Where integrity problems show up

| Decision | Data it depends on | What goes wrong without it | In the Library |
|---|---|---|---|
| Rating likelihood for idle piping | Damage mechanisms of out-of-service lines, such as water collecting and freezing | Sunray, Texas, 2007: a control station idle for about 15 years was left connected as a dead leg; trapped water froze and cracked an elbow [2] | [Dead leg](/glossary/dead-leg/) |
| Rating likelihood in hydrogen service | Measured temperatures and hydrogen pressures | Anacortes, 2010: design conditions gave a low result for exchangers that ruptured [9] | [What is high temperature hydrogen attack (HTHA)?](/library/high-temperature-hydrogen-attack/) |
| Rating consequence for large flammable inventories | Isolatable inventory and remote isolation | Sunray, 2007: no remotely operable shut-off valves, so the propane could not be stopped [2]; Philadelphia, 2019: no remote isolation of large hydrocarbon sources near the failed elbow [3] | [Consequence of failure](/glossary/consequence-of-failure/) |
| Weighting threats along a pipeline | Threats specific to each segment | NTSB: weights applied across a whole system, favoring a few threats, can understate a segment exposed to another, such as a pipeline crossing a fault line [5] | [What is pipeline integrity management?](/library/pipeline-integrity-management/) |

Sources: CSB reports 2007-05-I-TX, 2008 [2], 2010-08-I-WA, 2014 [9], and 2019-04-I-PA, 2022 [3]; NTSB/SS-15/01, 2015 [5].

## Sources

1. [2025 Publications Catalog (entry for RP 581) (RP 581, 3rd edition, April 2016), American Petroleum Institute (API)](https://www.api.org/-/media/files/publications/2025-catalog/2025_pubs%20catalog.pdf)
2. [Investigation Report: LPG Fire at Valero-McKee Refinery, Sunray, Texas, February 16, 2007, Report No. 2007-05-I-TX, US Chemical Safety and Hazard Investigation Board (CSB)](https://www.csb.gov/assets/1/20/CSBFinalReportValeroSunray.pdf?13739)
3. [Fire and Explosions at Philadelphia Energy Solutions Refinery Hydrofluoric Acid Alkylation Unit, Investigation Report No. 2019-04-I-PA, US Chemical Safety and Hazard Investigation Board (CSB)](https://www.csb.gov/assets/1/20/PES_Final_Report_Published_October_2022_r1.pdf?16845), 2022-10-11
4. [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](https://www.osti.gov/servlets/purl/7129942)
5. [Integrity Management of Gas Transmission Pipelines in High Consequence Areas, Safety Study NTSB/SS-15/01, US National Transportation Safety Board (NTSB)](https://www.ntsb.gov/safety/safety-studies/Documents/SS1501.pdf), 2015-01-27
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)](https://www.osti.gov/pages/servlets/purl/3364031)
7. [40 CFR 68.25, Worst-case release scenario analysis, eCFR, US Office of the Federal Register](https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-68/subpart-B/section-68.25)
8. [40 CFR 68.28, Alternative release scenario analysis, eCFR, US Office of the Federal Register](https://www.ecfr.gov/current/title-40/chapter-I/subchapter-C/part-68/subpart-B/section-68.28)
9. [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)](https://www.csb.gov/assets/1/7/tesoro_anacortes_2014-may-01.pdf), 2014-05-01
10. [Final Investigation Report: Chevron Richmond Refinery Pipe Rupture and Fire, Report No. 2012-03-I-CA, US Chemical Safety and Hazard Investigation Board (CSB)](https://www.csb.gov/assets/1/17/Chevron_Final_Investigation_Report_2015-01-28.pdf?15397), 2015-01-28
11. [Recommendations Status Change Summary, Recommendation 2010-8-I-WA-R11 to the American Petroleum Institute (API RP 581), US Chemical Safety and Hazard Investigation Board (CSB)](https://www.csb.gov/assets/recommendation/Status_Change_Summary_Tesoro_API_R11_C-UA.pdf), 2016-10-04
12. [49 CFR 195.452, Pipeline integrity management in high consequence areas, eCFR, US Office of the Federal Register](https://www.ecfr.gov/current/title-49/subtitle-B/chapter-I/subchapter-D/part-195/subpart-F/section-195.452)

Published by Winda Beacon (https://beacon.winda.ai).
