COMAH Regulation 7: internal emergency planning in practice

COMAH 2015 implements the Seveso III Directive in Great Britain and applies to establishments holding dangerous substances above the thresholds in Schedule 1. Regulation 5 requires a major-accident prevention policy for all operators; Regulation 7 adds the duty, for upper-tier establishments, to prepare an adequate internal emergency plan before operations begin.

Regulation 7 does not stand alone. The internal plan dovetails with the local authority's external emergency plan under Regulation 12, and with the safety report an upper-tier operator must submit. It is also tied to Regulation 8, which requires the plan to be reviewed and tested at suitable intervals not exceeding three years. Read together, these regulations expect a plan that is written, resourced, rehearsed and demonstrably effective, not a document that sits in a drawer.

HSE guidance frames an adequate plan around a set of objectives: containing and controlling incidents to minimise harm, protecting people and the environment, communicating with staff, emergency services and the public, and restoring and cleaning up afterwards. To meet them, the plan has to name the people who take charge, define how the alarm is raised, set out evacuation and mustering arrangements, and list the resources those actions rely on.

Personnel accountability runs through all of it. You cannot confirm an area is clear, direct fire and rescue to the right building, or decide whether re-entry is safe, unless you know who was on site and where they are. That includes contractors and visitors, not just direct employees, which is precisely where paper registers tend to fail on a busy day.

The plan must also be kept current. New process units, changed shift patterns, additional contractors and altered muster arrangements all age a plan quickly, so Regulation 7 expects it to be revised whenever there is a significant change and reviewed at least every three years.

A plan is only as good as the last time it was exercised. Regulation 8 requires operators to test the internal emergency plan at suitable intervals and at least every three years, and to review it in light of what the test reveals. Regulators increasingly want to see that lessons from an exercise fed back into the plan.

This is where many operators struggle to produce evidence. A drill happens, wardens report back verbally, and the debrief is a few handwritten notes. There is rarely a defensible record of how long the muster took, whether anyone was unaccounted for, or how the count compared with the number of people who were actually on site.

An automated muster turns each exercise into hard data: the time the alarm was raised, the moment each person reached safety, the maximum time to a full head count, and any tags that never appeared. That is the objective evidence a competent authority looks for when it asks whether your Regulation 7 plan has been properly tested.

The Salvus Network places Safety Detection Points at gates, muster points and zone thresholds, and issues every worker, contractor and visitor a safety tag. The moment an alarm sounds, the operator sees a live SAFE, EVACUATING and MISSING rollup, and where each missing person was last detected, so a search can be directed rather than guessed.

At the assembly point a single Safety Detection Point reads up to 500 tags at once and marks everyone safe automatically, with nothing to sign and no name-call, so the count keeps pace with people arriving in a rush. Every exercise and every real activation produces a timestamped drill report automatically, giving you the Regulation 8 test evidence and the after-incident record in the same system.

None of this replaces the human structure of your plan. Incident controllers, wardens and the works emergency team still run the response; the platform removes the part people are worst at under pressure, counting accurately and fast, and leaves behind the auditable proof that your Regulation 7 arrangements do what the plan says they will.