Restarting after a failure: why ‘fixed’ does not mean ‘understood’

After a failure or incident, the plant is under pressure. Customers are waiting, deadlines are slipping, and everyone wants the equipment running again. So the symptom is often fixed: a part replaced, a parameter adjusted, a sensor bypassed. The equipment runs again. Whether the cause has been established remains open, and with it the risk that the incident will happen again.

Typical traps after an incident

• Restarting too quickly: the equipment is released before anyone checks whether consequential damage has occurred.
• Only the obvious: the defective component is replaced, but nobody asks why it failed.
• Lost evidence: data, parts and observations are not secured and are no longer available for later clarification.
• Unclear responsibility: nobody is responsible for clarifying the cause once production is running again.

A structured procedure

1. Assess and secure the situation. First secure people and equipment. Then record what happened: time, condition, people involved, control system messages.
2. Secure the evidence. Keep defective parts, measurements, photos and statements before they are lost.
3. Investigate causes and consequences. Look not only at the failed component, but also at conditions, routines and possible consequential damage elsewhere.
4. Agree the restart. Define checks, a trial run, release and responsibilities. The timing follows the findings and safety requirements, not just deadline pressure.
5. Capture the lessons. What has been learnt goes into maintenance, work instructions and training so that the incident does not recur.

Statutory reporting obligations and requirements from authorities or insurers take precedence over any internal procedure.

Why an outside view can help

After an incident, those involved are often close to events, sometimes affected themselves. An outside view helps separate facts from assumptions, set priorities and document the restart traceably.

Frequently asked questions