Fault Tree Analysis
Fault tree analyses (FTAs) study specific system, process, or product failures using a tree diagram. The process can be used to study a failure that actually occurred, or it could study a potential failure. The technique starts with the failure and works backward toward increasingly fundamental events or root causes.
To predict failure events and potential causes.
To analyze a complex system containing numerous interconnected causes of failure.
To identify causes of a failure after it has happened.
To identify causes of a potential system failure during the design process.
1. IDENTIFY THE SYSTEM OR PROCESS THAT WILL BE STUDIED. A flow diagram may be particularly useful.
2. IDENTIFY THE TYPE OF FAILURE to be analyzed. Be as specific as possible in the definition. This is called the top event.
3. IDENTIFY EVENTS that are potential immediate causes of the top event. Record these events at the next level, under the event they would cause.
4. For each event, DETERMINE THE NATURE OF THE EVENT. Evaluate whether it is a basic failure or can be further analyzed for immediate causes. Circle the event if it is a basic failure and enscribe it within a rectangle if it can be analyzed for its own causes.
5. ASK how the events you have identified are related to the event they cause. Show relationships via gate symbols. Lower-level events are called input events and the event they cause is denoted as the output event.
6. CLASSIFY CAUSES.
7. REPEAT STEPS 3, 4, 5, and 6 for each event that is not considered basic, continuing the process until every branch of the tree terminates in either a basic or undeveloped event.
8. (Optional) ASSIGN PROBABILITIES to each of the basic events.
9. ANALYZE THE TREE to understand how the causes are interrelated and think of how to prevent failures. Pay closest attention to failures that are the most likely to happen.
Defines possible failure events along with the associated causes.
Useful for modeling team-based failures.
Can be used in either a predictive or retrospective analysis.
Diagrams can become large and complex.
Training needed to use the method quantitatively.
Stanton N, Salmon P, Walker G, et al. Process charting methods. Human factors methods: a practical guide for engineering and design. Great Britain: Ashgate; 2005. p. 109-37.
Tague N. The tools. In: O'Mara P, editor. The quality toolbox. 2nd ed. Milwaukee, WI: ASQ Quality Press; 2005. p. 93-521.