One of the most frustrating and costly patterns in process safety is the repeat incident — the same or very similar event happening again despite previous investigations and recommendations.
When incidents recur, it is rarely due to a single failure. More often, it results from weaknesses in the learning cycle: investigation, recommendation development, implementation, and sustainment. Below are five common systemic reasons why repeat incidents occur.
1. Incomplete or Incorrect Root Cause Identification
Many investigations stop at the immediate or obvious causes (equipment failure, human error) without digging deep enough into system root causes.
Common problems include:
- Using the wrong investigation methodology or applying it poorly
- Failure to identify all physical, human, and especially organizational/system causes
- Missing the right participants on the investigation team
- Bias, politics, production pressure, or flawed assumptions (“we know what happened”)
When root causes are missed or misidentified, the resulting recommendations cannot prevent recurrence.
2. Weak or Incomplete Recommendations
Even when causes are identified, recommendations often fall short. They may:
- Not address all identified causes
- Be vaguely worded (“improve training,” “review the procedure”)
- Fail to directly fix the underlying problem
- Focus on symptoms rather than system fixes
Poor recommendations give the illusion of action while leaving the real vulnerabilities in place.
3. Recommendations Not Completed (or Completed Poorly)
This is one of the most common failure modes:
- Recommendations are turned into MOCs or new action items that are never executed (“Closing on a promise”)
- Actions are only partially completed
- Fixes are installed incorrectly or without proper verification
- Actions taken don’t reflect the original intent of the recommendation
The original recommendation is marked “complete,” but the hazard remains.
4. Recommendations Not Sustained Over Time
Even when initially completed, improvements can erode:
- One-time fixes instead of permanent changes to procedures, training, or standards
- New changes (personnel, technology, workload) that undermine the fix
- Lack of ongoing operational discipline to maintain the new standard
- Drift back to old practices because the change was never fully embedded in the management system (in some cases, done once to close the recommendation, but then never done again as needed)
5. Failure to Apply Lessons Broadly
Lessons learned from one incident are not shared or applied to similar equipment, processes, or areas across the facility (or company). A problem fixed in one unit reappears in another because the organization treated it as an isolated event rather than a systemic vulnerability.
Example: BP Texas City Refinery (2005)
The CSB investigation into the March 23, 2005 explosion (15 killed, 180 injured) found that BP had experienced multiple previous serious incidents and near-misses on the same isomerization unit. As the CSB report noted:
- Many of the safety problems that led to the March 23, 2005, disaster were recurring problems that had been previously identified in audits and investigations.
- Incidents were often ineffectively investigated and appropriate corrective actions were not taken.
Breaking the Cycle of Repeat Incidents
Preventing repeat events requires disciplined execution of the entire learning loop:
- Thorough, unbiased investigations that reach true system root causes
- High-quality, specific, actionable recommendations
- Rigorous verification that fixes are properly implemented
- Strong sustainment through updated procedures, training, and operational discipline
- Broad sharing and application of lessons across the organization
- A quality review process of the initial recommendations and a later (e.g., 6 or 12 months) review that they are appropriately completed and being sustained
- If a repeat incident occurs, the investigation should evaluate the earlier investigation to determine why it was ineffective and to identify improvements in the investigation process
The key question for every facility: Are we truly learning from our incidents — or are we just documenting them?
References
- Process Safety Recommendations: Turning Insights into Lasting Change
- Klein, J.A. and Vaughen, B.K., Process Safety: Key Concepts and Practical Approaches, CRC Press, 2017
- Klein, J.A., “The ChE as Sherlock Holmes: Investigating Process Incidents,” CEP, 2016
- U.S. Chemical Safety Board, BP Texas City Refinery Explosion, 2007
What repeat incidents or near-misses have you seen in your career, and what systemic gaps allowed them to recur? Share your experiences in the comments or contact me at jim@psmnews.com.