In order to understand these differences it is important to understand what FMEA and FTA are.
FMEA stands for Failure Modes and Effect Analysis, where ‘failure’ means something stops functioning correctly and ‘failure mode’ relates to how the failure happened. The versatility of FMEA means that it is widely used for root cause analysis.
To help understand the difference between the failure and a failure mode, we could use the example of a printer that isn’t working properly. The failure is the fact that the machine is not printing, but the potential failure modes could be that the paper is jammed in the tray or that the printer ink has run out. So, one failure with two failure modes.
FMEA not only highlights potential failure modes, but also assigns a value to the effect and impact of each failure. This can be combined with a criticality analysis which demonstrates the different outcomes as a result of a failure, this is known as FMECA (Failure Modes, Effects and Criticality Analysis).
Other FMEA alternatives include DFMEA (Design Failure Mode and Effect Analysis), which looks at failures in product design, and PFMEA (Process Failure Mode and Effect Analysis), which investigates process-related failure.
FMEA works by thoroughly detailing a list of failure modes, causes and impacts. This information is usually presented as a table with seven columns that each detail a different step, as follows:
1. List Failure Modes
This is quite simply listing the potential causes of a given failure. For example, with a printer not working, the failure modes could include a paper jam, running out of paper, running out of ink, the power not being on, and so forth. It is important not to forget any failure modes as this could compromise your risk assessments. This step time-consuming step can be sped up by cross-referencing failure modes for similar components or systems.
2. Describe Failure Effect
The second step is to describe the effect of the failure, which can then be used to determine the severity of the failure in step three. Our broken printer may not have a big effect, depending on how often it needs to be used and for what tasks, or it could be deemed a critical item of equipment.
3. Determine Severity
This third step involves determining the severity rating of your failure, based on the description in step two. The severity rating is scored on a scale of 1-10, as follows:
- Rating 1: Minor Risk – faults are almost unnoticeable
- Rating 2-3: Low Risk – faults can be noticed but only have a minor impact
- Rating 4-6: Moderate Risk – faults may impact asset performance and are noticeable
- Rating 7-8: High Risk – the fault compromises your operation, disrupting schedules
- Rating 9-10: Critical/Very High Risk – the fault has totally compromised your asset with high security, safety or other risks
4. List Root Causes
Each failure mode could have a number of root causes that need to be listed in step four. For example, if there is no power to our printer it could be that the power has not been turned on, there could be a power cut, or a fuse may have blown. Listing these causes will make it easier to check each one to find the reason for your failure.
5. Determine Occurrence
The occurrence shows how often a particular failure is likely to happen. This is given a rating from 1-10, with one being ‘extremely unlikely’ and ten being ‘extremely likely’ or even ‘inevitable.’ Running out of printer ink or paper is pretty much inevitable, so you may mark that as a ten.
6. Determine Detection Rating
This step involves setting a detection rating from 1-10, with one being ‘extremely likely’ and ten being ‘extremely unlikely.’ A lack of paper may get a detection rating of one, as our printer will probably alert you as to what the problem is, but a blown fuse may be harder to detect.
7. Calculate the Risk Priority Level (RPN)
The RPN is a final number based on the ratings you gave for severity, occurrence and detection. The formula for this is severity x occurrence x detection = RPN. The higher this number, the more critical the failure and the more need there is for monitoring and improvements.
FMEA can be used to improve productivity, reliability and safety, as well as providing a number of additional advantages:
- Improved Work Methods: FMEA will create working methods that are more efficient, safer and more reliable
- Failure Prioritisation: By assessing and evaluating failure modes and their impacts it is easy to prioritise them according to likelihood and criticality
- Maintenance Planning: Understanding when a failure may occur allows you to improve your inspection and preventive maintenance planning
- Improved Security: Identifying potential failure points allows you to address and improve any security measures that may be compromised as a result
- Product Design: An understanding of failure modes and effects allows you to adjust product designs to reduce the chance of a failure occurring
- Faster Troubleshooting: By checking against the potential failure modes, it is easier to troubleshoot the cause of a problem
Despite the advantages of FMEA, there are a number of weaknesses with the method:
- Multiple Failures: FMEA is not able to take account of situations where multiple failures may occur at once since there is no linking between different failures in the method
- Time Consuming: As mentioned above, FMEA can be time-consuming as it requires all of the potential failure modes to be listed. As a result, it is also reliant upon the expertise of staff to identify them all
- Frequent Updates Required: Even the best FMEA may miss some failure modes or you may discover new ones with knowledge and experience of your asset. This means that your FMEA will need frequent assessment and updating
- Underestimating Risk: If you fail to consider a failure mode or modes, you may end up underestimating the associated risk
- Potential Waste of Resources: Conversely, taking too much time over your FMEA can be a waste of time and resources
First developed by the United States Army in the 1950s, FMEA was soon adopted by the aerospace industry and NASA, who used it for the Apollo missions, the Viking programme, and their Voyager missions.
FMEA continues to be used in aero and astrospace applications as well as in the automotive and oil and gas industries.
Because of its ability to analyse and assess functions, manufacturing processes and design, FMEA is used for a wide range of products and processes, including when there are design updates, new processes and regulations, or even problems highlighted by your customers.
FTA, or fault tree analysis, offers a system for problem-solving, troubleshooting and finding the root cause of a failure. It can either deliver an analysis of a single failure or assess several components together.
While FTA is often compared to FMEA, there are notable differences between the two techniques. In order to understand these differences, it is important to understand how FTA works.
A fault tree analysis diagram begins with the failure itself, offering a top-down approach that applies Boolean logic to assess each potential cause of the failure. This is assisted by using a series of symbols that represent input and output events, as follows:
Reliability software can also be used to create an FTA diagram, integrating the information with statistical probabilities to create a quantitative method.
So, how does it work?
An FTA is used to assess the cause of a failure (or near failure) by ruling out potential causes until the real reason is found. For example, if a system to prevent a fire fails it could be caused by either the system failing to detect the fire or the sprinkler system failed once the fire was detected. Of course, both of these could have failed, but an FTA allows you to systematically rule out different causes to find the final root cause of the failure. Once the cause has been found the reason behind that cause can be investigated, such as a lack of maintenance or wear on components.
Finding the root cause of a failure allows you to fix the problem and alter maintenance plans and safety procedures to reduce the chance of it happening again, increasing the availability and reliability of your assets over time.
FTA can be used for simple or more complex failures and also allows you to highlight several causes at the same time.
FTA offers a range of advantages, including:
- Safety: Improved compliance with safety regulations
- Assess Several Systems: FTA allows you to assess and correlate the relationship between failures and different systems
- Minimised Risk: FTA can reduce risk by highlighting where product or system design changes are needed
- Prioritisation: FTA allows you to establish overall priorities for your systems and assets
- Probabilistic Risk Assessments (PRA): The characteristics of FTA mean that it can be used to carry out safety and probabilistic risk assessments. A PRA is a systematic approach for analysing risk and reliability to determine the chances and the consequences of a failure.
While FTA offers some advantages, there are also limitations:
- Rigidity: Since FTA works on a binary system whereby something has either caused the problem or not, it can be too rigid for assets that may fail due to certain conditions, such as high or low temperatures. FTA is also not very good at determining partial failures
- Probability: FTA is not good at determining the probability of a failure, which means it is no use for quantitative purposes
- Lifetime Planning: FTA doesn’t account for the amount of time an asset has been in use or its probable lifecycle. This means that FTA is not ideal for product development purposes
FTA is used to diagnose the root cause of a failure and understand what can cause a system to break. This helps determine the inherent risks and identify methods to mitigate against them. Because of this, FTA is ideal for Probabilistic Risk Assessments, which are used in a range of potentially hazardous industries such as aerospace, nuclear, chemical, oil and gas and even the pharmaceutical industry. FTA is also used in software engineering to eliminate the causes of bugs and viruses.
The most evident difference between FMEA and FTA is how they approach failure. FMEA takes a ‘bottom up’ approach, looking at each component in turn and creating a list of potential failure modes. By contrast, FTA takes a ‘top down’ approach, beginning with the failure and then diagnosing what could have caused the problem through a series of questions or checks.
These two approaches mean that FTA is good at finding the cause for a particular fault while FMEA is better at exhaustively checking and cataloguing initiating faults and their effects. However, despite this intrinsic difference in approach, both FMEA and FTA require experts with a strong knowledge of the asset being analysed.
A successful FMEA analysis requires the prediction of all possible failure modes, which can be separated according to how critical they are – from a full breakdown to almost imperceptible problems. However, unlike FTA, FMEA will not account for conditional events or establish a relationship between multiple failures.
However, unlike FMEA, FTA does not account for partial failures as something is either broken or not. This binary system of ‘yes’ or ‘no’ makes it difficult to assess nuances or scale. On the other hand, this lack of scale means that FTA is efficient for design and analysis that requires potential failures to be identified ready for safety improvements.
Because the risk priority numbers associated with FMEA are given subjectively (by deciding how ‘risky’ something is on a scale of 1-10), it is a qualitative method of analysis. This means it can be unreliable and makes it potentially dangerous when used in hazardous situations.
FTA, meanwhile, is a quantitative tool, whose binary nature makes it ideal for probabilistic risk assessments. This is because it works to a set rule or ‘0’ or ‘1’ with no space for conjecture. FTA is also better at taking account of external events (i.e. a flood) and conditioning events (i.e. low temperatures) that may cause problems. FMEA, on the other hand, only offers isolated failure modes without looking at other factors.
Another key difference is in how easy FTAs are to keep updated compared to FMEAs. FMEA requires a great deal of detail and expertise to keep updated while FTA can easily be carried out using software to automate the process while also including statistical data.
The following table shows the main differences between FMEA and FTA:
Top-down, deductive approach
Bottom-up, inductive approach
Can show the correlation between multiple failures
Lists component failures without analysing the system as a whole
Considers external events
Does not consider external events
Does not consider partial failures
Does not consider unexpected failures
Easily updated with correct software
Often difficult, time-consuming, and resource-heavy to update
Given the differences, it may be difficult to work out whether to use FMEA or FTA.
FMEA should be used when:
- You can’t locate a top-down failure event to begin an FTA
- You need to identify every possible failure mode, including those that are not critical or dangerous, for example, when producing a product manual
- Your system is highly automated and predictable, making it easier to predict all failure modes
FTA should be used when:
- A ‘top-down’ failure can be identified
- You need to perform a safety assessment on a system or asset
- You need to perform a probability risk assessment
- Your system is complex with several interacting elements
- Your system is liable to human error or other external factors that could cause issues
Many organisations and industries opt to use FMEA and FTA together in order to attain the benefits of both. Because risk analysis can be performed quantitatively or qualitatively, FMEA and FTA can work alongside each other.
This has led to the creation of hybrid solutions that blend the different approaches, such as Variation Modes and Effect Analysis (VMEA). These type of tools are widely used for predictive maintenance of critical assets.
While both are tools that can be used for root cause analysis, there are notable differences in the approaches of FMEA and FTA. These differences lend each method its own strengths and weaknesses that can be used for different applications.
Where FMEA can be too subjective, FTA can prove to be too rigid as an approach. Because of this, many industries use a combination of the two.
Related Frequently Asked Questions (FAQs)