Home / Guides / Bowtie vs. fault tree
Bowtie vs. fault-tree analysis
Both techniques analyse hazards. They answer different questions. The short answer: fault trees give you logic and probability; bowties give you communication and barrier visibility. Many serious risk studies use both.
Side by side
| Bowtie | Fault tree | |
|---|---|---|
| Primary question | What controls stand between threats and consequences? | What combinations of basic events lead to a top event? |
| Direction | Threats → top event → consequences (left to right) | Top event → causes (top down) |
| Logic | Implicit (parallel paths) | Explicit (AND, OR gates) |
| Quantification | Qualitative by default | Quantitative (probability of failure) |
| Audience | Engineers, operators, auditors, regulators, executives | Reliability engineers, safety analysts |
| Best at | Communicating risk, identifying barrier gaps and shared failures | Quantifying probability of complex causal chains |
When to use a bowtie
- You need to communicate a hazard to a non-specialist audience.
- You want a single picture of all the controls that mitigate a top event.
- You want to identify shared failures across multiple barriers (escalation factors do this well).
- You are running a workshop with a mixed audience.
When to use a fault tree
- You need a probability for the top event.
- The causal chain is logically complex (multiple AND conditions, voting logic).
- You are building a SIL determination, a PFD calculation, or a quantitative safety case.
When to use both
In a mature safety case, the bowtie is the front page. Each preventive barrier and each escalation factor on the bowtie is supported by a fault tree (or a LOPA) underneath it that quantifies its contribution. The bowtie tells the story; the fault tree shows the working.