Hazard Identification, Analysis, and Tracking
This domain tests your ability to select a hazard analysis technique, interpret its output, and maintain a defensible hazard tracking record from identification through closure.
Technique selection
- PHA is early, qualitative, and broad; it screens hazards before detailed design.
- HAZOP uses process nodes and guide words such as no, more, less, reverse, as well as, part of, and other than to find deviations.
- FMEA works bottom-up from component failure modes to effects; FMECA adds criticality.
- FTA works top-down from a defined top event through Boolean gates to minimal cut sets, exposing single points and combinations.
- LOPA evaluates scenario frequency against independent protection layers and their probability of failure on demand.
- STPA examines unsafe control actions and causal scenarios in complex control structures.
Tracking discipline
Each hazard needs a unique ID, description, causal factors, risk, controls, verification method, owner, due date, status, residual risk, and acceptance. Closure requires objective evidence, not a statement that a procedure was revised.
Worked example
A fault tree for “loss of cooling” has the logic (pump failure AND backup failure) OR operator fails to start backup. The minimal cut sets are {pump failure, backup failure} and {operator fails}. The second is a single-point failure. The correct response is an independent automatic start with proof testing, not simply more operator training. Reasoning: the cut set shows that one human action defeats the whole barrier.
Traps that cost marks
- Using an RPN threshold that lets a high-severity, low-occurrence failure pass.
- FMEA that ignores common-cause and interaction failures.
- HAZOP nodes drawn so broadly that deviations are meaningless.
- Closing a hazard after a document change without verifying field implementation.