Incident Investigation Beyond Five Whys: Reconstruct Controls and Decisions
Five Whys can start a conversation, but a reliable investigation reconstructs the event, examines failed barriers and verifies corrective action.
Incident Investigation Beyond Five Whys: Reconstruct Controls and Decisions
“Why did the operator do that?” is rarely a complete investigation. A process safety event develops through interactions among equipment, procedures, people, supervision, planning and safeguards. Five Whys can help a team explore one line of inquiry, but it cannot guarantee that the team has found all contributing conditions.
OSHA’s incident investigation guidance recommends investigating close calls as well as harmful events and looking beyond a single immediate cause. It points investigators toward equipment, procedures, training and safety-program weaknesses. OSHA guidance is not South African law, but the systems approach is useful for any facility that needs to learn from an event.
Reconstruct the event before naming causes
Secure the scene and preserve records as soon as it is safe to do so. Build a timeline from alarms, control-system data, permits, shift logs, maintenance records, interviews and physical evidence. Separate confirmed facts from estimates and recollections. Record what was known at the time of each decision rather than judging it with information that became available later.
Map the prevention and mitigation barriers for the scenario. Ask which barriers were required, which were available, which were challenged, and how the event moved through them. Include detection, isolation, shutdown, containment and emergency response where relevant. Check whether the protection depended on shared utilities, common procedures or the same maintenance system.
Ask questions that reveal conditions
- What changed before the event: process conditions, equipment, staffing, workload, materials or procedures?
- Did the procedure reflect the actual task, and could the team follow it with the tools and time provided?
- Were alarms useful and understood, or had repeated nuisance alarms weakened response?
- Did maintenance, inspection or management-of-change decisions affect a safeguard?
- Which assumptions shaped the decisions, and what information was available to the people making them?
- Could a similar event occur elsewhere in the facility?
Avoid assigning “human error” as the root cause. A person’s action may be part of the sequence, but investigators still need to examine the conditions that shaped it. Worker participation in the review helps surface work-as-done and can improve the corrective actions.
Make corrective action testable
Prefer changes that remove the hazard or strengthen a barrier over reminders to “be more careful.” Give each action an accountable owner, due date, completion evidence and effectiveness check. If a temporary control is needed, document its limits and expiry. Revisit the event after implementation to confirm that the change works under real operating conditions.
For a structured view of process hazards and safeguards, see MMRisk’s Bowtie analysis and PHA services and our safety report guidance. Contact MMRisk for investigation support or a barrier review.