Written and maintained by CASRAI Editorial Board
Published
On this page
A process hazard analysis (PHA), also called a hazard evaluation, is the documented, team-based examination of a chemical process that US regulations require for covered processes: OSHA’s Process Safety Management standard (29 CFR 1910.119) and EPA’s Risk Management Program rule (40 CFR Part 68, Program 3). The analysis must identify, evaluate and control the hazards of the process, use at least one of seven named methods, be revalidated at least every five years, and be kept for the life of the process.
Definition and scope
Under 29 CFR 1910.119(e)(1), the employer performs an initial process hazard analysis on processes covered by the standard. The text says the analysis must be “appropriate to the complexity of the process” and must “identify, evaluate, and control the hazards involved in the process.” It is a process-level study: the unit of analysis is the whole process, including its equipment, controls, siting and people, not a single task or a single failure.
The OSHA standard applies to a process that involves a chemical at or above the threshold quantity listed in its Appendix A, and to a process involving a Category 1 flammable gas or a flammable liquid with a flashpoint below 100 °F (37.8 °C) on site in one location in a quantity of 10,000 pounds (4,535.9 kg) or more. Two exceptions sit inside that second trigger: hydrocarbon fuels used solely for workplace consumption (for example, propane for comfort heating) when they are not part of a process with another highly hazardous chemical, and flammable liquids in atmospheric tanks, or transferred, that are kept below their normal boiling point without chilling or refrigeration. Paragraph (a)(2) states that the section does not apply to retail facilities, oil or gas well drilling or servicing operations, or normally unoccupied remote facilities. A laboratory is not named in that exclusion list, so applicability turns on the chemicals and quantities in the process rather than on the word “laboratory.”
For research settings, that quantity logic is the practical point. A bench or core facility that holds laboratory-scale amounts of reagents is judged against the same thresholds as a plant; a PHA becomes a regulatory obligation only when a covered process exists. Where it does not, the task-level approach in job hazard analysis for lab procedures and the hazard-class vocabulary in the lab chemical hazard classes primer are the more usual tools.
What 29 CFR 1910.119(e) requires
Paragraph (e) of the OSHA standard breaks the PHA obligation into separate duties. Each one below is stated in the regulation text.
Setting priorities and doing the initial analysis
The employer must determine and document the priority order for conducting PHAs, using a rationale that includes the extent of the process hazards, the number of potentially affected employees, the age of the process, and its operating history. The original compliance schedule required at least 25 percent of initial PHAs by May 26, 1994, 50 percent by May 26, 1995, 75 percent by May 26, 1996, and all of them by May 26, 1997. Analyses completed after May 26, 1987 that meet the paragraph are acceptable as initial PHAs.
What the analysis must address
Paragraph (e)(3) lists seven subjects: the hazards of the process; any previous incident with a likely potential for catastrophic consequences in the workplace; engineering and administrative controls applicable to the hazards and their interrelationships, such as detection methods that give early warning of releases (the text gives process monitoring and control instrumentation with alarms, and detection hardware such as hydrocarbon sensors, as examples of acceptable detection); the consequences of failure of those controls; facility siting; human factors; and a qualitative evaluation of a range of the possible safety and health effects of failure of controls on employees in the workplace.
Who performs it
Under (e)(4), the PHA is performed by a team with expertise in engineering and process operations. The team must include at least one employee with experience and knowledge specific to the process being evaluated, and one member must be knowledgeable in the specific PHA methodology being used. Separately, paragraph (c) of the standard requires the employer to consult employees and their representatives on the conduct and development of process hazard analyses and to give them access to the PHAs.
Acting on the findings
Under (e)(5), the employer must establish a system to promptly address the team’s findings and recommendations, assure that recommendations are resolved in a timely manner with the resolution documented, document what actions are to be taken, complete actions as soon as possible, develop a written schedule for completion, and communicate the actions to operating, maintenance and other employees whose work is in the process and who may be affected.
Revalidation and retention
At least every five years after the initial PHA, the analysis must be updated and revalidated by a team meeting the (e)(4) requirements, so that it stays consistent with the current process. The PHAs, their updates or revalidations, and the documented resolution of recommendations must be retained for the life of the process. A PHA is therefore a living file, not a one-time report.
The seven methodologies named in the rules
Both rules require the employer to use “one or more” of the following methodologies, whichever are appropriate to determine and evaluate the hazards of the process being analyzed:
- What-If
- Checklist
- What-If/Checklist
- Hazard and Operability Study (HAZOP)
- Failure Mode and Effects Analysis (FMEA)
- Fault Tree Analysis
- An appropriate equivalent methodology
The regulation names these methods but does not define how to run each one, and it does not rank them. Choosing among them is a judgement the rule leaves to the employer, limited by the phrase “appropriate to determine and evaluate the hazards of the process.” Because the same seven options appear in both rules, one methodology choice can serve both programs.
FMEA appears on this list as a process-safety method and also appears in other fields. For how the same technique is used in a clinical setting, and how it differs from root cause analysis, see healthcare failure mode and effects analysis and the RCA vs FMEA comparison.
How EPA’s 40 CFR 68.67 differs from OSHA’s paragraph (e)
EPA’s Program 3 prevention requirement at 40 CFR 68.67 opens with nearly the same language as OSHA’s. It also states that PHAs completed to comply with 29 CFR 1910.119(e) are acceptable as initial PHAs, and that updated and revalidated OSHA-compliant PHAs satisfy the five-year paragraph. The differences are in what the analysis must address. The data table below sets the two rules side by side, paragraph by paragraph.
Three EPA additions deserve attention. First, 68.67(c)(3) requires the owner or operator to ensure that monitoring equipment associated with prevention and detection of accidental releases from covered processes has standby or backup power to provide continuous operation. Second, 68.67(c)(9) adds a safer-technology analysis. For covered processes in NAICS codes 324 and 325, the owner or operator must consider and document, in the following order of preference, inherently safer technology or design, passive measures, active measures, and procedural measures; a combination may be used to reach the desired risk reduction. Third, 68.67(c)(10) requires the analysis to address any gaps between the codes, standards or practices to which the process was designed and built and the most current version of the applicable codes, standards or practices.
For a narrower set of NAICS 324 and 325 processes, 68.67(c)(9)(ii) goes further: processes within 1 mile of another stationary source with a covered process in those NAICS codes, hydrofluoric acid alkylation processes in NAICS 324, and processes that have had one accident meeting the accident-history reporting requirements of 40 CFR 68.42 since the most recent PHA. For those, the owner or operator must also determine and document the practicability of the inherently safer technologies and designs considered, document the methods used, and, for any implemented, submit a description of the technology to EPA. Under (c)(9)(iii) the analysis must be performed by a team that includes members with expertise in the process being evaluated, including at least one member who works in the process, and the team members must be documented.
The ordering in the safer-technology paragraph (inherently safer design, then passive, active and procedural measures) resembles the general logic of the hierarchy of controls, though the regulation sets out its own four-step order and does not cite that framework.
PHA compared with job hazard analysis and FMEA
Three terms are often confused. The distinction below is editorial interpretation based on how the regulation defines the PHA’s scope, not a regulatory definition of the other terms.
- PHA: examines a whole covered process, its controls, siting and human factors, as required by 1910.119(e) and 68.67, and is revalidated on a five-year cycle.
- Job hazard analysis (JHA): works task by task, breaking one job into steps to find the hazards in each. The lab version is covered in job hazard analysis for lab procedures. A JHA is not one of the seven PHA methodologies in the regulation text.
- FMEA: one of the seven named PHA methodologies. It is a method used inside a PHA, not a substitute for the PHA’s required contents.
Where it matters for labs and facilities
Three situations are worth checking against the text. First, a facility that scales a research process up to pilot or production quantities should compare its inventory to the Appendix A thresholds and the 10,000 pound flammables trigger, because applicability is quantity-based. Second, a site covered by both programs can use one set of PHAs for both, because 68.67 accepts OSHA-compliant analyses; the extra EPA topics (natural hazards, backup power for monitoring equipment, public receptors, design-code gaps, and safer technology where NAICS 324 or 325 applies) must still be shown to have been addressed. Third, the findings-resolution system is the part that carries the most day-to-day weight: both rules require documented, scheduled follow-through with the people whose work is in the process, not just a completed report.
Common misunderstandings
“A PHA is the same as a risk assessment of any lab task.”
The regulated PHA is a defined obligation tied to covered processes, with a required team, required subjects, five-year revalidation and life-of-process retention. A routine task assessment may be a good practice but does not meet those elements by itself.
“One method is mandatory.”
The rule requires one or more of the listed methodologies “that are appropriate.” No single method is mandated, and “an appropriate equivalent methodology” is allowed.
“The PHA is done once.”
The initial PHA must be updated and revalidated at least every five years, and the records stay for the life of the process.
“The two rules require identical analyses.”
They share the same core, but EPA’s paragraph adds topics such as natural hazards, standby power for monitoring equipment, proximity to public receptors, design-code gaps and, for NAICS 324 and 325, safer technology.
Frequently asked questions
How often must a process hazard analysis be updated?
At least every five years after the initial PHA, by a team that meets the team requirements. This is stated in 29 CFR 1910.119(e)(6) and 40 CFR 68.67(f).
Who must perform a PHA?
A team with expertise in engineering and process operations, including at least one employee with experience and knowledge specific to the process, and one member knowledgeable in the PHA methodology used (29 CFR 1910.119(e)(4); 40 CFR 68.67(d)).
What methods can be used for a PHA?
What-If, Checklist, What-If/Checklist, HAZOP, FMEA, Fault Tree Analysis, or an appropriate equivalent methodology, using one or more as appropriate to the process.
How long must PHA records be kept?
For the life of the process, including updates, revalidations and the documented resolution of recommendations.
Does the OSHA PSM standard apply to my laboratory?
Applicability depends on the process: a chemical at or above its Appendix A threshold quantity, or 10,000 pounds (4,535.9 kg) or more of a Category 1 flammable gas or flammable liquid with a flashpoint below 100 °F in one location, subject to the exceptions in the text. Retail facilities, oil and gas well drilling or servicing, and normally unoccupied remote facilities are excluded.
Sources
- 29 CFR 1910.119, Process safety management of highly hazardous chemicals, paragraphs (a), (c) and (e). Legal Information Institute, https://www.law.cornell.edu/cfr/text/29/1910.119. Checked 2026-11-15.
- 40 CFR 68.67, Process hazard analysis (Program 3 prevention program). Legal Information Institute, https://www.law.cornell.edu/cfr/text/40/68.67. Checked 2026-11-15.
Data: Process hazard analysis requirements compared: OSHA 29 CFR 1910.119(e) and EPA 40 CFR 68.67
Side-by-side comparison of what the OSHA Process Safety Management standard, paragraph (e), and the EPA Risk Management Program Program 3 rule, 40 CFR 68.67, require of a process hazard analysis. Compiled by reading the regulation text of both paragraphs and recording each requirement row by row; the EPA-only additions are the rows where OSHA's (e)(3) list is silent.
Coverage: United States.
| Requirement | OSHA 29 CFR 1910.119(e) | EPA 40 CFR 68.67 |
|---|---|---|
| Initial analysis of covered processes | Required, appropriate to complexity; priority order documented; compliance schedule ended May 26, 1997 | Required, same wording; to be conducted not later than June 21, 1999; OSHA-compliant PHAs acceptable as initial PHAs |
| Methodologies | One or more of seven: What-If, Checklist, What-If/Checklist, HAZOP, FMEA, Fault Tree, equivalent | Same seven options |
| Team | Engineering and process operations expertise; one employee with process experience; one member knows the methodology | Same; (c)(9)(iii) adds team-member documentation and at least one member who works in the process for the safer-technology analysis |
| Siting | Facility siting | Stationary source siting, including placement of processes, equipment and buildings, proximate stationary sources, and consequences for the public and public receptors |
| Natural hazards | Not listed in (e)(3) | Natural hazards that could cause or exacerbate an accidental release |
| Standby or backup power | Not listed in (e)(3) | Monitoring equipment for release prevention and detection must have standby or backup power for continuous operation |
| Safer technology | Not listed in (e)(3) | NAICS 324 and 325: consider and document inherently safer technology or design, passive, active, then procedural measures |
| Design-code gaps | Not listed in (e)(3) | Gaps between design-basis codes, standards or practices and the most current versions |
| Findings resolution | Prompt system; timely, documented resolution; written schedule; communicate to affected employees | Same elements in 68.67(e) |
| Revalidation | At least every five years | At least every five years; OSHA-compliant revalidations acceptable |
| Records retention | For the life of the process | For the life of the process |
Sources
- 29 CFR 1910.119 Process safety management of highly hazardous chemicals (accessed 2026-11-15)
- 40 CFR 68.67 Process hazard analysis (accessed 2026-11-15)
Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.







