What information do you need before starting an FMEA?
Before starting an FMEA, you need a clearly defined scope, a cross-functional team, and complete documentation of the system or process being analyzed. Without these foundations, the analysis will be incomplete, inconsistent, or too broad to be actionable.
Preparation is the most underestimated phase of the FMEA process. Teams that skip it often produce worksheets that look thorough but miss critical failure paths because the boundaries of the analysis were never properly established. The following inputs are essential before the first failure mode is written down:
- Scope definition: Specify exactly what is being analyzed. Is this a single component, a subsystem, a full production line, or a process? Define where the analysis starts and stops.
- System documentation: Gather engineering drawings, process flow diagrams, P&IDs, operating procedures, and maintenance records. For rotating equipment, this includes equipment specifications, duty cycles, and historical fault data.
- Performance requirements: Define what the system is supposed to do under normal operating conditions. A failure mode is only meaningful relative to an expected function.
- Historical failure data: Maintenance logs, work orders, and incident reports reveal which failure modes have already occurred and how often. This grounds the analysis in real operational experience.
- Cross-functional team: Include design engineers, maintenance technicians, process operators, and safety specialists. No single person holds all the knowledge needed to identify every failure mode accurately.
- FMEA worksheet or software: Agree on the format before starting. Consistency in how severity, occurrence, and detection are scored is critical for meaningful RPN calculations.
In industrial maintenance contexts, it also helps to align on which FMEA type you are conducting. A Design FMEA (DFMEA) focuses on component design; a Process FMEA (PFMEA) focuses on manufacturing or operational steps; a System FMEA takes a higher-level view across interconnected subsystems. Clarifying this upfront prevents the team from mixing methodologies mid-analysis.
How do you identify failure modes in an FMEA?
In an FMEA, failure modes are identified by systematically asking "In what way could this component or process step fail to perform its intended function?" Each answer is a failure mode. The goal is to be exhaustive, not selective, covering every plausible way something can go wrong, not just the most obvious ones.
Start by listing every function of the item being analyzed. A pump, for example, has functions including moving fluid at a specified flow rate, maintaining pressure, and containing the fluid within the system. For each function, the team then asks what could prevent that function from being fulfilled, fulfilled only partially, or fulfilled at the wrong time.
Common categories of failure modes include:
- Complete loss of function: The component stops working entirely (a pump fails to start).
- Partial or degraded function: The component works but below specification (reduced flow due to impeller wear).
- Intermittent function: The component works sometimes but not reliably (vibration-induced intermittent seal failure).
- Unintended function: The component does something it should not (a valve that leaks in the closed position).
- Function at the wrong time: The component activates or deactivates outside the intended window.
For each failure mode, the team then identifies the failure effect (what happens to the system or end user when this failure occurs) and the failure cause (the underlying mechanism or root cause that leads to the failure mode). Causes might include material fatigue, misalignment, contamination, operator error, or design deficiency.
Using structured brainstorming techniques, fault trees, or reference to industry failure databases helps ensure the team does not overlook less obvious failure modes. In rotating equipment applications, common failure causes such as shaft misalignment, bearing wear, seal degradation, and vibration-induced fatigue should always be explicitly considered, as they account for a significant proportion of unplanned downtime across industrial installations.
How is the risk priority number (RPN) calculated in FMEA?
The Risk Priority Number (RPN) in an FMEA is calculated by multiplying three scores: Severity (S), Occurrence (O), and Detection (D). The formula is RPN = S x O x D. Each factor is rated on a scale of 1 to 10, producing an RPN between 1 and 1,000. Higher RPNs indicate failure modes that require the most urgent corrective action.
Understanding what each factor measures is essential to scoring accurately:
- Severity (S): Rates the seriousness of the failure effect on the end user, the process, or safety. A score of 1 means no noticeable effect; a score of 10 means a catastrophic failure with safety implications or regulatory consequences. Severity is assessed based on the worst-case credible outcome.
- Occurrence (O): Rates how frequently the failure cause is likely to occur. A score of 1 means the failure is extremely unlikely; a score of 10 means it is almost certain to occur repeatedly. Historical maintenance data and engineering judgment both inform this rating.
- Detection (D): Rates how likely it is that the failure will be detected before it reaches the end user or causes a consequence. A score of 1 means the failure will almost certainly be detected; a score of 10 means it is virtually undetectable. This reflects the effectiveness of current controls, inspections, or monitoring systems.
One important limitation of the RPN method is that it treats all three factors as equally weighted. In practice, a failure mode with a Severity score of 10 (catastrophic) but low occurrence and detection scores may produce a moderate RPN, yet still warrant immediate action. For this reason, many organizations use the RPN as a prioritization guide rather than an absolute threshold, and apply separate rules that automatically escalate any failure mode with a Severity score of 9 or 10 regardless of its overall RPN.
Scoring consistency across the team is critical. Using agreed-upon rating scales and anchoring each score to specific, observable criteria prevents the subjective drift that can make RPN comparisons meaningless.
What actions should follow a completed FMEA?
After completing an FMEA, the team should prioritize failure modes by RPN, assign corrective actions to address the highest-risk items, implement those actions, and then re-evaluate the RPN to confirm the risk has been reduced. An FMEA without follow-through actions has no practical value.
The action phase is where the FMEA delivers its return on investment. The process typically follows these steps:
- Rank failure modes by RPN: Focus initial attention on the highest-scoring items. Also flag any failure modes with a Severity score of 9 or 10 for immediate review, regardless of their total RPN.
- Define recommended actions: For each high-priority failure mode, specify what action will reduce the risk. Actions generally target one of three areas: redesigning to reduce severity, improving reliability to reduce occurrence, or adding monitoring or inspection to improve detection.
- Assign ownership and deadlines: Each action needs a named responsible person and a target completion date. Unassigned actions rarely get implemented.
- Implement actions: Execute the engineering changes, maintenance procedure updates, monitoring additions, or design modifications identified in the analysis.
- Recalculate the RPN: After implementation, re-score the affected failure modes to verify that the action achieved the intended risk reduction. This closes the loop and documents the improvement.
- Update the FMEA worksheet: Record the actions taken, the revised scores, and the responsible parties. The FMEA becomes a living document that reflects the current risk profile of the system.
In industrial maintenance environments, FMEA actions often translate directly into updated preventive maintenance schedules, new condition monitoring requirements, changes to spare parts stocking policies, or modifications to operating procedures. Connecting FMEA outputs to the maintenance management system ensures that recommended actions are tracked and not lost between review cycles.
How often should an FMEA be reviewed and updated?
An FMEA should be reviewed whenever there is a significant change to the system, process, or operating environment, and at a minimum on a defined periodic cycle, typically every one to three years depending on the criticality of the asset. A static FMEA quickly becomes inaccurate and loses its value as a risk management tool.
Specific triggers that should prompt an immediate FMEA review include:
- A design change, process modification, or equipment upgrade
- A new or recurring failure that was not previously identified in the analysis
- A near-miss event or safety incident
- Changes in operating conditions, such as increased throughput, new materials, or different duty cycles
- Regulatory or standards updates that affect safety or compliance requirements
- Changes in the maintenance strategy, such as moving from time-based to condition-based maintenance
For high-criticality assets in sectors such as oil and gas, chemical processing, or water and wastewater treatment, annual reviews are common practice. For lower-criticality systems with stable operating histories, a two to three year review cycle may be sufficient. The review should involve the same cross-functional team that produced the original analysis, supplemented by anyone who has gained relevant operational experience since the last update.
Treating the FMEA as a living document rather than a one-time exercise is what separates organizations that genuinely reduce failure risk from those that simply complete the paperwork.
What is the difference between FMEA and FMECA?
FMECA (Failure Mode, Effects, and Criticality Analysis) is an extension of FMEA that adds a formal criticality analysis step. While FMEA identifies failure modes and their effects, FMECA goes further by quantifying the criticality of each failure mode using probability data and consequence severity, producing a more rigorous risk ranking. FMECA is the preferred method in high-reliability sectors where quantitative risk assessment is required.
The core difference lies in how risk is ranked:
- FMEA uses the qualitative or semi-quantitative RPN (Severity x Occurrence x Detection) to prioritize failure modes. It is practical, widely applicable, and well-suited to manufacturing, process, and maintenance contexts where historical data may be incomplete.
- FMECA replaces or supplements the RPN with a criticality number derived from the probability of the failure mode occurring, the probability that it leads to a specific consequence, and the severity of that consequence. This approach is grounded in reliability engineering standards such as MIL-STD-1629A and IEC 60812.
FMECA also typically produces a criticality matrix, plotting failure modes by severity category against their probability of occurrence. This visual output makes it easier to communicate risk priorities to decision-makers and supports formal asset management frameworks such as ISO 55000.
In practice, FMEA is the more common choice for general industrial maintenance and product development because it is faster to execute and requires less quantitative failure rate data. FMECA is more appropriate when regulatory requirements demand quantitative risk justification, when designing safety-critical systems, or when detailed reliability data is available to support the calculation. Many organizations begin with FMEA and escalate to FMECA for the assets identified as most critical.
How Zytec Helps Reduce the Failure Modes You Find in an FMEA
One of the most consistent findings in an FMEA for rotating equipment is that mechanical couplings are a significant source of failure modes. Misalignment, wear, lubrication breakdown, seal degradation, and vibration-induced fatigue appear repeatedly across pumps, fans, compressors, and conveyors, each generating its own chain of effects and corrective actions. Zytec's non-contact magnetic drives are specifically engineered to eliminate the root causes behind many of these failure modes at the design level, rather than managing their consequences through maintenance.
Key ways Zytec technology reduces FMEA-identified risks in rotating equipment:
- No mechanical contact: Torque is transmitted through magnetic fields, removing friction, wear components, and the failure modes they generate, including elastomer degradation, metal-to-metal fatigue, and lubrication failure.
- Inherent misalignment tolerance: The F-TORQ accepts up to 12 mm of radial dynamic misalignment, structurally eliminating the misalignment-induced bearing and seal overload that accounts for a significant share of rotating equipment failures.
- Built-in overload protection: If the driven load exceeds the torque limit, the drive decouples automatically, preventing overload propagation into the drivetrain, a failure mode that frequently cascades into motor and gearbox damage.
- No lubrication or wear consumables: Removes entire categories of failure causes related to lubricant contamination, depletion, and scheduled replacement intervals being missed.
- Vibration reduction up to 80%: Lower vibration levels directly reduce occurrence scores for vibration-related failure modes across connected components.
- ATEX-certified for hazardous areas: The F-TORQ.EX series eliminates ignition-risk failure modes that conventional couplings introduce in explosive atmospheres through frictional heating and unpredictable surface temperatures.
When you conduct an FMEA and identify coupling-related failure modes as high-priority items, Zytec's non-contact drives offer a proven engineering solution that reduces Severity, Occurrence, and Detection scores simultaneously. Contact Zytec to discuss how non-contact drive technology can address the specific failure modes identified in your next FMEA.
Frequently Asked Questions
Can FMEA be used effectively on aging equipment with limited historical documentation?
Yes, FMEA can still deliver strong results on older equipment even when documentation is incomplete. In these cases, the cross-functional team's direct operational experience becomes the primary input — maintenance technicians and operators who have worked with the equipment for years often carry tacit knowledge that compensates for missing records. Supplementing team knowledge with generic failure mode databases for the equipment type (such as OREDA for offshore systems or IEEE 493 for electrical equipment) and any available work order history from your CMMS can help fill documentation gaps. The key is to flag assumptions explicitly in the worksheet so future reviewers understand where scored values were estimated rather than data-driven.
What is a good RPN threshold to use for deciding which failure modes require immediate action?
There is no universal RPN threshold that applies to every organization or industry — the right cutoff depends on your risk tolerance, asset criticality, and available resources. A commonly used starting point is to prioritize all failure modes with an RPN above 100 or 125, but this number should be calibrated to your specific context rather than adopted blindly. More importantly, never rely on RPN alone: any failure mode with a Severity score of 9 or 10 should be escalated for immediate review regardless of its total RPN, since a catastrophic but infrequent and detectable failure can still produce an unacceptably low RPN under the standard formula. Many mature FMEA programs use a tiered approach — immediate action above a high threshold, scheduled action in a mid-range band, and monitored-only below that.
How do you prevent scoring inconsistency when multiple team members rate Severity, Occurrence, and Detection differently?
The most effective way to control scoring inconsistency is to anchor each rating scale to specific, observable criteria before the team begins scoring — rather than leaving members to interpret a generic 1–10 scale independently. For example, a Severity score of 8 should be defined as something concrete, such as u0022causes unplanned production shutdown exceeding four hours,u0022 not just u0022high severity.u0022 Holding a short calibration exercise at the start of the FMEA, where the team scores two or three example failure modes together and discusses any disagreements, significantly aligns individual judgment before it affects the full worksheet. Using a standardized FMEA template with pre-defined scale descriptors, ideally drawn from industry references like AIAG-VDA for automotive or IEC 60812 for general applications, provides a consistent baseline across sessions and team changes.
What is the most common mistake teams make when conducting their first FMEA?
The most common mistake is confusing failure modes with failure causes or failure effects — and mixing all three into the same column of the worksheet. A failure mode is the specific way a component fails to perform its function (e.g., u0022pump delivers insufficient flowu0022); the cause is the underlying mechanism driving that failure (e.g., u0022impeller wear due to abrasive mediau0022); and the effect is what happens to the system or user as a result (e.g., u0022downstream process starved, production lossu0022). Keeping these three categories clearly separated is what makes the FMEA actionable, because corrective actions target causes, not modes or effects. Teams that conflate them end up with worksheets that are difficult to prioritize and produce vague corrective actions that don't address root causes.
How long does it typically take to complete an FMEA, and how many sessions should teams plan for?
A focused FMEA on a single piece of equipment or a well-defined process step typically requires between two and five working sessions of two to four hours each, spread over one to three weeks. The wide range reflects the complexity of the system, the completeness of available documentation, and the experience level of the team. Trying to complete an FMEA in a single marathon session is a common planning mistake — fatigue degrades the quality of failure mode identification and scoring, and critical items get missed late in the session. A more effective approach is to complete the function and failure mode identification in early sessions, then score and prioritize in a separate session once the team has had time to verify assumptions and gather any missing data.
Is FMEA software necessary, or can the analysis be done effectively in a spreadsheet?
A well-structured spreadsheet is entirely sufficient for most FMEA applications, particularly for teams conducting analyses on a limited number of assets or running their first few FMEAs. Dedicated FMEA software adds value primarily when managing large numbers of linked analyses, when integration with a CMMS or PLM system is needed, or when regulatory documentation requirements demand audit trails and version control. Tools like ReliaSoft XFMEA, Siemens Teamcenter, or PTC Windchill offer these capabilities, but they also introduce training overhead and licensing costs that may not be justified for smaller programs. The more important factor than the tool is the consistency of the template — whatever format you choose, ensure the worksheet structure, rating scales, and column definitions remain standardized across all analyses in your program.
How does FMEA connect to a broader Reliability-Centered Maintenance (RCM) program?
FMEA is one of the core analytical tools within a full Reliability-Centered Maintenance (RCM) program — specifically, it provides the structured failure mode and effects analysis that RCM uses to determine the most appropriate maintenance strategy for each failure mode. In an RCM framework, the outputs of the FMEA feed directly into maintenance task selection: failure modes with high occurrence scores may warrant condition monitoring or time-based replacement, while those with high severity but low occurrence may be better managed through redesign or redundancy. Organizations that have already conducted FMEAs on their critical assets have completed a significant portion of the analytical groundwork required for a full RCM implementation, making FMEA a practical entry point into a more comprehensive reliability program.