Incident Report Checklist for Seveso and BRZO Sites

Most incident report checklists are written for general workplace safety: slips, minor injuries, near-hits on a factory floor. At a Seveso-classified site, that's not the checklist you need. Incident and near-miss reporting isn't just an HR or general-safety task - it's a mandated input to your veiligheidsbeheersysteem (VBS), specifically the "monitoring performance" (prestatiebewaking) element required under Seveso III Annex III.

That distinction changes what you capture, why you capture it, and where the data has to go afterward. This checklist is built around that requirement, not a generic incident form.

Why Near-Miss Reporting Matters More at a Seveso Site

At a non-Seveso worksite, a near miss is usually a lesson about individual behavior or a local hazard: someone almost slipped, a forklift almost clipped a pallet rack. At a major-hazard site, a near miss can be a precursor signal for a major accident - a small deviation in a process that, under slightly different conditions, could have led to a loss of containment, fire, explosion, or toxic release.

This is the core reason Seveso III treats performance monitoring as a formal management-system element rather than a housekeeping task. A near miss involving a safety-critical valve, an interlock bypass, a pressure excursion, or a deviation from a safe operating limit isn't a minor event to log and forget - it's evidence about whether your barriers and safeguards are actually working as designed. If that evidence isn't captured and analyzed, you lose the early-warning function that near-miss reporting exists to provide.

In practice this means your reporting threshold at a Seveso site has to be lower and more specific than "did someone get hurt." You need to capture:

Deviations from safe operating procedures, even when no harm resulted
Failures or near-failures of safety-critical equipment (interlocks, relief systems, alarms, containment)
Situations where a control measure identified in your MAPP or safety report activated, was bypassed, or didn't perform as expected
Loss-of-containment events of any scale, including those fully contained by secondary measures

What Needs to Be Captured for a Major-Hazard-Relevant Incident or Near Miss

A general incident form asks for names, dates, and a description. That's necessary but not sufficient here. For an incident or near miss with major-hazard relevance, the report needs enough detail to support a real investigation into whether a safety barrier failed, degraded, or was never adequate in the first place.

Identification and classification

Date, time, and precise location (unit, installation, process step)
Classification: was this a loss of containment, a near miss, a process deviation, or an equipment/safety-system failure
Whether a hazardous substance covered by your Seveso inventory was involved, and in what quantity or concentration if known

What happened, in sequence

The sequence of events leading up to and following the incident, in chronological order
The process conditions at the time (pressure, temperature, flow, equipment status) where relevant
Which safety-critical systems were involved - interlocks, relief valves, gas detection, fire suppression, containment - and how each performed

Barrier and control performance

Which control measures from the MAPP or safety report were expected to prevent or mitigate this scenario
Whether those measures functioned, partially functioned, or failed
Any manual interventions that substituted for an automated safeguard

People and consequences

Individuals involved, including operators, contractors, and anyone in the affected zone
Injuries or exposure, including psychological effects, even if minor or ultimately unconfirmed
Actual or potential offsite consequences (release beyond the fence line, community impact) even if none occurred - the potential matters as much as the actual outcome for major-hazard reporting

Evidence

Witness statements, taken separately and factually
Photo or video evidence of the area, equipment, and any damage
Relevant instrumentation data or alarm logs, where available

Immediate response

Who was notified and when (shift supervisor, HSE manager, emergency services if applicable)
Immediate containment or mitigation actions taken
Whether the emergency plan was activated, even partially

What the Investigation and Root-Cause Process Should Cover

Capturing the incident is the input; the investigation is where it becomes useful. For a major-hazard-relevant incident or near miss, the investigation needs to go beyond "what happened" to "why the barriers didn't fully do their job."

At minimum, the investigation should establish:

Direct cause - the immediate technical or human action that triggered the event
Underlying causes - equipment condition, procedural gaps, competency or training gaps, or organizational factors that allowed the direct cause to occur
Root cause(s) - the systemic issue in the management system itself: a maintenance interval that was too long, a procedure that didn't reflect actual operating practice, a change that was made without a management-of-change review
Barrier analysis - for each safety-critical system involved, whether it was present, functional, and adequate for the scenario, or whether it was missing, degraded, or bypassed
Comparison to the safety report/MAPP scenario set - whether this event maps to a scenario already identified and controlled for, or reveals a gap in the hazard identification

The investigation should be proportionate to the potential severity, not just the actual outcome. A near miss that could have escalated into a major accident under slightly different conditions warrants the same rigor as an incident that did cause harm - this is the point of near-miss reporting at a Seveso site, and it's where organizations most often under-invest.

Feeding the Data Back Into the VBS and MAPP

None of this has value if it stops at a closed investigation file. Seveso III requires that performance monitoring data actively inform the management system, and that means a defined path from "incident closed" to "system updated."

Into performance monitoring. Incident and near-miss data, including trends across multiple minor events, should be part of the metrics your VBS tracks under the monitoring performance element - not just lagging indicators like injury counts, but leading indicators like near-miss frequency by type, barrier failure rate, and time-to-close on corrective actions.

Into procedure updates. Where an investigation identifies a procedural gap - a safe operating limit that was unclear, a permit-to-work step that was skipped, a maintenance procedure that didn't catch a degrading component - the corrective action needs to result in an actual, version-controlled update to that procedure, communicated to the people who use it.

Into the MAPP, where relevant. If an investigation reveals that a hazard scenario wasn't adequately identified or controlled in the safety report or MAPP - a barrier assumed to be effective that wasn't, a scenario likelihood that was underestimated - that finding needs to feed into the next review of those documents, not wait for the scheduled multi-year revision cycle if the gap is significant.

Into management review. Incident and near-miss trends, root causes, and the status of corrective actions should be visible at the management review level, not just held at the site or department level. This is what closes the loop between individual incidents and the continuous improvement Seveso III expects from the VBS as a whole.

The Practical Takeaway

A generic incident report checklist will get you a record of what happened. A Seveso-fit incident and near-miss reporting process gets you evidence about whether your major-accident-prevention barriers are actually working - and a defined route for that evidence to reach performance monitoring, procedures, and the MAPP. If your current reporting process can't tell you that, it's worth reviewing against the Annex III requirements before your next inspection does it for you.