Failure Modes and Effects Analysis

Why Failure Modes and Effects Analysis Matters

Failure Modes and Effects Analysis (FMEA) is one of the most widely used system safety and reliability tools in industry, and it appears regularly on the BCSP Associate Safety Professional (ASP) exam within the Safety Programs and Concepts domain. Safety professionals are expected to understand FMEA because it is a proactive, structured way to find out how things can go wrong before they actually do. Rather than waiting for an incident to reveal a weakness, FMEA systematically examines each component or process step, asks how it could fail, and evaluates what the consequences of that failure would be.

For the ASP candidate, FMEA is important for three reasons. First, it is a core hazard analysis technique that you will be compared against other methods such as Fault Tree Analysis (FTA), Hazard and Operability Studies (HAZOP), Job Hazard Analysis (JHA), and What-If analysis. Second, it introduces the concept of risk ranking through the Risk Priority Number (RPN), which the exam may ask you to calculate or interpret. Third, FMEA illustrates the hierarchy of controls and risk reduction philosophy that underpins modern safety management systems such as ANSI/ASSP Z10 and ISO 45001.

What Is Failure Modes and Effects Analysis?

FMEA is a systematic, inductive, bottom-up analytical technique used to identify potential failure modes of components, subsystems, or process steps; determine the effects of each failure on the larger system; and prioritize those failures so that corrective actions can be targeted where they will reduce risk the most.

Key terminology you must know:

Failure Mode – The specific manner in which a component, assembly, or process step could fail to perform its intended function. Examples include a valve that sticks open, a wire that shorts, a weld that cracks, or an operator step that is skipped. Think of a failure mode as how something fails.

Failure Effect – The consequence of the failure mode on the system, the product, the process, or the people involved. Effects may be described at the local level (what happens to the component itself), the next-higher level (what happens to the subsystem), and the end-effect level (what happens to the overall system or user). Think of the effect as what happens as a result.

Failure Cause – The underlying reason the failure mode occurs, such as corrosion, fatigue, contamination, improper installation, software error, or human error.

Severity (S) – A ranking of how serious the effect of the failure would be, typically on a 1 to 10 scale where 10 is catastrophic (death or total system loss) and 1 is negligible.

Occurrence (O) – A ranking of how likely the failure cause is to happen, typically 1 (extremely remote) to 10 (almost certain).

Detection (D) – A ranking of how likely the current controls are to detect the failure mode or its cause before it reaches the end user or results in harm. Importantly, this scale is reversed in logic: a 1 means detection is almost certain, and a 10 means the failure is almost impossible to detect. The harder it is to detect, the higher the number.

Risk Priority Number (RPN) – The product of Severity, Occurrence, and Detection: RPN = S × O × D. With 1 to 10 scales, the RPN ranges from 1 to 1,000. Higher RPNs indicate higher priority for corrective action.

Criticality – When an FMEA includes a formal quantitative assessment of the probability and severity of failures, it is called a Failure Modes, Effects, and Criticality Analysis (FMECA). FMECA has its roots in military standard MIL-STD-1629A and is common in aerospace, defense, and nuclear applications.

Types of FMEA

System FMEA (SFMEA) – Performed early in the concept phase to analyze interactions between subsystems and the overall system architecture.

Design FMEA (DFMEA) – Focuses on how a product design could fail during its service life. It examines components, materials, geometry, and tolerances. It is performed before the design is released to manufacturing.

Process FMEA (PFMEA) – Focuses on how a manufacturing, assembly, service, or work process could fail. It examines each process step, equipment, methods, and human actions. PFMEA is the type most often relevant to occupational safety because it directly analyzes work tasks.

Other variants include Machinery FMEA, Software FMEA, and Healthcare FMEA (HFMEA), the latter promoted by the Joint Commission and the VA National Center for Patient Safety.

How FMEA Works: The Step-by-Step Process

Step 1: Define the scope and assemble the team. Decide what system, product, or process will be analyzed and establish the boundaries. FMEA is a team activity; it works best with a cross-functional group including design engineers, operators, maintenance personnel, quality staff, and safety professionals. A facilitator who understands the FMEA method keeps the group on track.

Step 2: Break the system down into its elements. Use block diagrams, process flow charts, or bills of material to list every component or process step. Each element is examined one at a time. This is the bottom-up nature of FMEA: you start with individual parts and work upward to system effects.

Step 3: Identify the function of each element. State what each component or step is supposed to do. A failure mode is defined as a departure from intended function, so you must know the function first.

Step 4: Identify potential failure modes. For each function, brainstorm all the ways it could fail. Common generic failure modes include: fails to operate, operates prematurely, operates intermittently, fails to stop, degraded output, and operates at the wrong time. Every element may have multiple failure modes.

Step 5: Determine the effects of each failure mode. Describe the consequences at the local, next-higher, and end levels. Ask what the operator, customer, or system would experience.

Step 6: Assign Severity ratings. Rate each effect on the agreed scale. Severity is tied to the effect, not the cause, and generally cannot be reduced unless the design itself is changed to eliminate or lessen the consequence.

Step 7: Identify potential causes and assign Occurrence ratings. List the root causes or mechanisms that could lead to each failure mode and rate how often each is expected to happen. Historical data, reliability data, and expert judgment inform this rating.

Step 8: Identify current controls and assign Detection ratings. List existing prevention controls (which reduce Occurrence) and detection controls (which improve Detection) such as inspections, alarms, interlocks, tests, and procedures. Rate the likelihood that the failure would be caught.

Step 9: Calculate the RPN. Multiply S × O × D for each failure mode and cause combination. Rank the results from highest to lowest.

Step 10: Develop and assign corrective actions. Prioritize the highest RPNs and any item with a high Severity rating regardless of RPN. Recommended actions should follow the hierarchy of controls: eliminate the failure mode through design, substitute, add engineering safeguards or redundancy, improve detection, then rely on administrative controls and PPE. Assign responsibility and a target completion date.

Step 11: Recalculate the RPN after actions are taken. Once actions are implemented, reassess S, O, and D to produce a revised RPN and verify that risk has actually been reduced. This closes the loop and documents continuous improvement.

Step 12: Maintain the FMEA as a living document. Update it whenever the design, process, materials, or operating conditions change, or after incidents reveal new failure modes.

A Worked RPN Example

Suppose a PFMEA on a hydraulic press identifies the failure mode light curtain fails to detect operator hand. The team rates Severity as 9 (serious injury, amputation), Occurrence as 3 (low, but possible due to misalignment), and Detection as 6 (moderate, since misalignment might not be noticed until the daily check). RPN = 9 × 3 × 6 = 162. After adding a self-checking light curtain with automatic fault indication and a pre-shift verification procedure, Detection drops to 2 and Occurrence drops to 2, producing a revised RPN of 9 × 2 × 2 = 36. Note that Severity stayed at 9 because the potential injury has not changed; only the likelihood and detectability have improved.

Strengths of FMEA

FMEA is systematic and thorough, forcing the team to consider every component. It is relatively easy to learn and does not require advanced mathematics. It produces a documented, auditable record and a prioritized action list. It is proactive and can be applied at any lifecycle stage. It integrates well with quality systems such as IATF 16949 and with reliability-centered maintenance programs.

Limitations of FMEA

FMEA examines failures one at a time and therefore is weak at identifying hazards caused by combinations of simultaneous failures or by interactions between components; Fault Tree Analysis is better suited for that. It can be time consuming and labor intensive for complex systems. RPN ratings are subjective, and because RPN is a product, very different combinations of S, O, and D can yield the same number, which can mask a high-severity item. For this reason, many practitioners treat any high Severity rating as a priority regardless of RPN. FMEA also tends to focus on hardware failure and may under-represent human error and organizational factors unless the team deliberately includes them.

How FMEA Compares to Other Hazard Analysis Methods

FMEA vs. Fault Tree Analysis (FTA): FMEA is inductive and bottom-up, starting with component failures and reasoning forward to effects. FTA is deductive and top-down, starting with an undesired top event and reasoning backward through AND and OR logic gates to find contributing causes. They are complementary; FMEA findings often provide inputs to a fault tree.

FMEA vs. HAZOP: HAZOP uses guide words (No, More, Less, Reverse, Other Than, etc.) applied to process parameters (flow, pressure, temperature) and is especially suited to chemical process plants. FMEA focuses on component and step failure modes and is more common in manufacturing and product design.

FMEA vs. Job Hazard Analysis (JHA): JHA breaks a job into steps and identifies hazards to the worker performing each step, with a focus on human exposure. PFMEA breaks a process into steps and identifies how the process can fail, including equipment faults, with a formal risk ranking. JHA is simpler and task-focused; FMEA is more rigorous and system-focused.

FMEA vs. Preliminary Hazard Analysis (PHA): PHA is a broad, early-stage screening tool used when little design detail exists. FMEA is performed later when components and process steps are defined.

FMEA vs. Event Tree Analysis (ETA): ETA is inductive like FMEA but starts with an initiating event and traces forward through success or failure branches of safeguards to determine possible outcomes.

Exam Tips: Answering Questions on Failure Modes and Effects Analysis

Tip 1: Memorize the classification words. If a question asks which technique is inductive, bottom-up, or examines single component failures, the answer is FMEA. If it asks which is deductive, top-down, or uses logic gates, the answer is Fault Tree Analysis. Exam writers love this distinction.

Tip 2: Know the RPN formula cold. RPN = Severity × Occurrence × Detection. Be ready to compute it from given values and to identify which failure mode has the highest priority. Double-check that you multiplied all three factors; a common distractor is the sum rather than the product.

Tip 3: Remember the Detection scale is inverted. A high Detection number means the failure is hard to detect, which increases risk. If a question says detection is almost certain, the Detection rating is low (near 1). Misreading this scale is a frequent error.

Tip 4: Severity is tied to the effect, and it rarely changes. Adding an inspection or an alarm improves Detection; adding maintenance or better materials reduces Occurrence; only a design change that alters the consequence itself reduces Severity. If a question asks which factor a new inspection program would most directly affect, the answer is Detection.

Tip 5: Distinguish failure mode from failure effect and failure cause. Mode is how it fails (valve sticks closed). Effect is what happens as a result (loss of cooling flow, overheating). Cause is why it fails (corrosion, debris). Questions often present a scenario and ask you to classify a phrase into one of these categories.

Tip 6: Know the difference between FMEA and FMECA. FMECA adds a formal criticality analysis, usually quantitative, and is associated with MIL-STD-1629A. If a question mentions criticality ranking or military or aerospace origins, think FMECA.

Tip 7: Recognize the main limitation. If a question asks about a weakness of FMEA, the best answer is typically that it does not readily identify hazards resulting from multiple simultaneous failures or component interactions. Other acceptable limitations include subjectivity of ratings and the time required for complex systems.

Tip 8: Match the tool to the scenario. Chemical process with guide words means HAZOP. Worker task steps and PPE means JHA. Early design concept with little detail means PHA. Component-by-component failure review with RPN means FMEA. Backward analysis of a specific catastrophic event means FTA.

Tip 9: Apply the hierarchy of controls to corrective actions. When asked for the best recommended action from an FMEA, prefer elimination or engineering redundancy over procedures or PPE. The exam rewards answers that reduce risk at the source.

Tip 10: Remember FMEA is a team-based, living document. Questions about who should perform an FMEA point to a cross-functional team, not a single engineer. Questions about when to update an FMEA point to design changes, process changes, or new incident information.

Tip 11: Watch for the phrase single-point failure. FMEA is particularly good at revealing single-point failures, meaning a single component whose failure alone causes system failure. Identifying and eliminating single-point failures through redundancy is a classic FMEA outcome.

Tip 12: Read numerical questions carefully. If given a table of several failure modes with S, O, and D values, compute each RPN and compare. Also check whether the question asks for the highest priority (highest RPN) or asks which item should be addressed first even if RPN is lower because Severity is 9 or 10. Many modern FMEA standards, including the AIAG-VDA harmonized handbook, replace RPN with an Action Priority table that emphasizes Severity first.

Summary

FMEA is a proactive, inductive, bottom-up hazard analysis technique that identifies how each component or process step can fail, what the effects would be, and how risky each failure is based on Severity, Occurrence, and Detection. The Risk Priority Number (S × O × D) drives corrective action, and actions are verified by recalculating the RPN. FMEA excels at finding single-point failures and documenting risk reduction but is limited in analyzing combined failures, where Fault Tree Analysis is the complementary tool. For the ASP exam, focus on the vocabulary, the formula, the inverted Detection scale, the inductive versus deductive distinction, and the ability to match the right analysis tool to a given scenario.