The EHS Audit Checklist for Seveso Sites: Auditing the VBS

An external Seveso audit is not the same exercise as an internal EHS review. When DCMR or the Nederlandse Arbeidsinspectie schedules an inspection at a BRZO-classified site, they are not there to help you find gaps - they are there to verify that your veiligheidsbeheersysteem (VBS) is real, current, and actually followed on the floor. The difference between a smooth inspection and a difficult one usually comes down to preparation logistics, not the underlying safety program itself.

This checklist covers what to do in the days and weeks before an external inspection: who needs to be available, which documents to have staged, what inspectors typically ask, and how to run an internal walk-through so nothing surprises you on the day. For the full breakdown of the seven Annex III elements your VBS must cover, see our compliance audit checklist - this post assumes that structure is already in place and focuses on getting ready to defend it in front of a regulator.

How External Seveso Inspections Differ From Internal Audits

Internal audits are exploratory: you're looking for weak spots before they become findings. External inspections are evaluative: the inspector already has a scope, and they are checking whether your documented VBS matches what actually happens on site. Three practical differences follow from this:

Inspectors sample, they don't review everything. They will pick a handful of installations, procedures, or incidents and trace them end to end - document, training record, maintenance log, and a conversation with the operator who does the work.
Inconsistency is the finding, not the gap itself. A procedure that's slightly out of date is less damaging than a procedure that doesn't match what an operator describes when asked directly.
You can't prepare the content in the week before. You can only prepare access, logistics, and the people who need to speak to it. If the VBS itself has holes, an inspection week is too late to close them - that's a year-round program, not a pre-audit task.

This checklist is about that second category: logistics and readiness, not rewriting your safety management system on short notice.

Two to Four Weeks Out: Confirm Scope and Assign Owners

Most DCMR and Arbeidsinspectie inspections come with some advance notice, even if it's short. Use it.

Confirm the scope with the inspecting authority where possible - which installations, which Annex III elements, or which follow-up items from a previous inspection are in focus. Don't assume it's a full-site review if the invitation suggests otherwise.
Assign a single point of contact who will accompany inspectors for the duration of the visit and coordinate document requests. Inspectors lose confidence quickly when they have to chase down who owns what.
Identify backup owners for each Annex III element (hazard identification, operational control, change management, emergency planning, training, monitoring, audit and review) in case the primary owner is unavailable on the day.
Check open items from the last inspection. Regulators routinely follow up on prior findings before moving to new ground. If a corrective action was promised with a deadline, know its current status before you're asked.

Who Needs to Be Available on Inspection Day

Document review only tells half the story - inspectors interview people to confirm the documents reflect reality. Have these roles confirmed and reachable, not just "somewhere on site":

The VBS/HSE manager or Seveso coordinator, who can speak to the safety management system as a whole and explain how its elements connect.
Site or installation managers for any process, unit, or installation named in the inspection scope.
Operators and shift personnel who actually run the installations being reviewed - inspectors frequently ask floor-level staff to describe a procedure in their own words, not just point to the document.
Maintenance and technical staff who can produce inspection, testing, and maintenance records for safety-critical equipment on request.
Emergency response coordinator, able to walk through the internal emergency plan and how it links to external/municipal response plans.
Someone with authority to make commitments - a director or site manager who can agree to corrective action timelines during the closing meeting, rather than needing to check with someone off-site.

Documents to Stage Before the Inspection

Don't wait to be asked. Have current, correct versions of the following pulled and ready - not buried across shared drives, binders, and someone's inbox:

Current MAPP (major-accident prevention policy) and safety report, plus the version history showing when each was last reviewed
VBS documentation mapped to the Annex III elements, with clear ownership per element
Hazard identification and risk assessment records for the installations in scope
Management of change (MOC) records for any recent modifications to process, equipment, or procedures
Maintenance, inspection, and testing records for safety-critical equipment (SIL-rated systems, pressure relief, containment, gas detection)
Training records and competency evidence for personnel connected to the installations under review
Internal audit reports and their corrective action tracking - including whether past actions were actually closed, not just marked closed
Incident and near-miss logs, with evidence of root-cause investigation and follow-up
Internal and external emergency plans, and records of the most recent emergency exercise or drill
Permit-to-work records for any hot work, confined space entry, or other high-risk activity in the review period

If any of these live in Capptions or another digital system, confirm in advance that the right people can pull filtered exports (by installation, by date range, by element) without needing IT support mid-inspection.

Questions Inspectors Commonly Ask

Preparing your team for the type of question - not scripting specific answers - is what makes interviews go well. Inspectors tend to probe in a few consistent directions:

"Walk me through what happens when..." - a process deviation, an alarm, a near-miss. They're checking whether the answer matches the documented procedure.
"Who is responsible for this?" - testing whether roles and responsibilities in the VBS are actually known by the people who hold them, not just written down.
"When was this last reviewed or tested?" - for procedures, risk assessments, and equipment. Vague answers ("recently," "I think last year") read worse than an honest "I'll confirm and follow up."
"Can you show me the record for that?" - the follow-up to almost every claim. If a procedure says an inspection happens monthly, expect a request for the last several months of records.
"What changed here, and how was that change assessed?" - for any visible modification to equipment or layout, checking whether it went through management of change.
"How does this connect to the emergency plan?" - for any hazard or scenario discussed, checking that risk assessment and emergency response are linked, not separate exercises.

If a staff member doesn't know an answer, "I don't know, but I'll find out" is a better response than guessing. Inspectors are used to this and generally prefer it to a confident wrong answer.

Run an Internal Walk-Through Before the Real One

The most useful preparation step is a dry run that mimics the actual inspection format, not just a document review.

Pick a sample the way an inspector would - one or two installations or procedures, ideally ones you haven't scrutinized recently rather than your best-documented example.
Trace it end to end: pull the procedure, pull the supporting records, then ask the operator or technician to describe it without looking at the document. Compare the three.
Time the document retrieval. If it takes twenty minutes to locate a maintenance record during your own walk-through, it will take at least that long - and look worse - during the real inspection.
Debrief gaps immediately and assign owners with dates, the same way you'd track a real corrective action. A walk-through that surfaces a gap and then loses track of it hasn't accomplished anything.
Repeat on a rotation. A single dry run before a scheduled inspection is useful; doing this periodically, on installations chosen at random, is what actually keeps the VBS audit-ready year-round rather than only in inspection week.

During and After the Inspection

Stick to what's asked. Answer the question in front of you; don't volunteer unrelated information or speculate about causes you haven't confirmed.
Take your own notes on what was reviewed, what was asked, and any verbal feedback given during the walk-through - the closing meeting will move faster if your team already has a shared record.
Clarify findings before the inspector leaves, if a closing meeting is offered. It's easier to ask "can you point to the specific requirement" in the room than to interpret an ambiguous written finding afterward.
Assign corrective actions with real owners and dates as soon as the formal findings arrive, and track them the same way you track internal audit actions - so the next inspection doesn't reopen the same item.

The Underlying Point

Readiness for an external Seveso inspection isn't a separate program from your day-to-day VBS - it's what a well-run VBS looks like when someone from outside the organization is watching. Sites that struggle with inspections are usually not sites with worse safety programs; they're sites where the people, documents, and records aren't organized well enough to demonstrate the program under time pressure. Fixing that is mostly a logistics and access problem, which is exactly what a pre-inspection checklist like this one is for.