“Apples and oranges aren’t that different really. I mean they’re both fruit. … I could understand if you said, ‘That’s like comparing apples with hermaphroditic ground sloths.” — Chuck Klosterman
A client recently asked us to “outline the difference in method and objective for FMEA vs. PHA/HAZOP and LOPA.” They were interested in evaluating system resiliency/reliability of an integrated power and use application. We thought others would be interested in the answer, so here goes.
PHAs
A process hazard analysis—a PHA—is a study to “identify, evaluate, and control hazards” in a process, typically a manufacturing process and in the case of the readers of this blog, a chemical manufacturing process. OSHA, in its Process Safety Management (PSM) standard (and by extension, the USEPA in its Risk Management Planning (RMP) rule) lists six methodologies that may be appropriate for PHAs, depending on “the complexity of the process”. In the order listed by the regulations, they are
- What-If
- Checklist
- What-If/Checklist
- Hazard and Operability Study (HazOp)
- Failure Mode and Effects Analysis (FMEA)
- Fault Tree Analysis
The regulations also allow for “an appropriate equivalent methodology.”
Notice that the regulations list both HazOp and FMEA as methodologies that may be appropriate for conducting PHAs. Also notice that Layer of Protection Analysis (LOPA) is not listed.
Layer of Protection Analysis
The Center for Chemical Process Safety introduced the LOPA methodology to the world in 2001 with its purple book, Layer of Protection Analysis – Simplified Process Risk Assessment. Coming nine years after OSHA promulgated the PSM standard, the LOPA methodology wouldn’t have been included on the list.
Is LOPA “an appropriate equivalent methodology.”
Risk assessment consists of estimating the consequences and severity of a potential incident, which the book touches on, and estimating the likelihood of the potential incident, which is the primary focus of the book and the LOPA methodology. The LOPA methodology applies after the hazard has been identified. So, while helpful in a PHA insofar as LOPA can address evaluating and controlling hazards, it is not a complete PHA tool in itself because it does not identify hazards.
But it is still very useful in completing a PHA.
Hazard and Operability Study
The HazOp methodology was already a very mature tool when OSHA promulgated the PSM Standard in 1992. Originally developed by engineers from ICI in 1963, the HazOp methodology was introduced as a class in 1974. The first class happened to come shortly after the Flixborough disaster, and the method quickly gained wide acceptance. Trevor Kletz, of ICI, first introduced the term “HazOp” in 1983.
The HazOp methodology looks at the process from the perspective of individual nodes within the process (lines or vessels in the case of chemical processes) and questions what effect a change in process parameters will have on the node or the process. The parameters studied include flow, temperature, composition, pressure, and the variations include too high, too low, in the wrong direction, or not present. Some variations may not apply to the given parameter, and some additional variations may be applicable to others.
The HazOp relies on the team having sufficient knowledge of the process to adequately assess how the deviations will affect the process, and what safeguards will protect against such effects. Based on the findings of the HazOp, an action list of safeguards and scenarios for further study can be created. LOPA is often the methodology of choice for the additional study.
Failure Modes and Effects Analysis
The FMEA methodology is even older than the HazOp methodology. The U.S. military began using FMEA in the late 1940s, and first issued a military standard, MIL-P-1629, in 1949 to evaluate and reduce risks in military systems and equipment. It was adopted by the aerospace industry in the 1960s, then the automotive industry in the 1970s.
The FMEA methodology looks at components within a system. It analyzes what would cause each component to fail and the effect of that failure on the overall system. The methodology requires the FMEA team to assign a severity, likelihood, and detection ability rating to each failure. These factors are then combined to generate a risk priority number which the team uses to rank which failure mode will have the most severe effect. The goal of FMEA is to create a prioritized list which can then be addressed according to the magnitude of the risk priority number.
Given the origin and purpose of the FMEA methodology, it has never been clear to us why OSHA included it as one of the possible methodologies for PHAs in a chemical process. It has its champions, though, and OSHA does allow its use.
Evaluating System Resiliency/Reliability
The query we received was about evaluating the system resiliency and reliability of an integrated power and use application. When it comes to evaluating systems, rather than processes, there is probably no better tool than FMEA. As a MIL standard, the Department of Defense cancelled the last edition of FMEA standard, MIL-STD-1629A (24-Nov-1980), in August 1998. The DoD did not replace it. Instead, many turn to the Society of Automative Engineers standard, SAE J1739_202605, Potential Failure Modes and Effects Analysis (FMEA) Including Design FMEA, Supplemental FMEA-MSR, and Process FMEA. It is worth noting that when automotive engineers refer to “process”, they are referring to assembly processes.
Use the Appropriate PHA Methodology
The difference between HazOp and FMEA is not as extreme as the difference between apples and hermaphroditic ground sloths. It really is more like the difference between apples and oranges. Or apples and pears. Or granny smith apples and golden delicious apples. There are differences, and every situation there is a methodology that is the best. HazOp is best in chemical processes where the deviations include pressure, temperature, flow, and level excursions—the things that worry chemical engineers. FMEA is best in systems where component failure is the concern—the things that worry mechanical, automotive, and aerospace engineers. While other methodologies can be forced to work (although we don’t recommend trying to make an apple pie with oranges) the effort will be most beneficial when the right methods are used.
