โ† Polyodos ยท Methods wiki ยท Industrial and technical risks
Regulatory method

๐Ÿงช PHA โ€” Safety Report (SEVESO)

The regulatory PHA variant for permitted SEVESO and ICPE sites: hazardous phenomena, risk-control measures, and the PIGC probability scale.

1What is a safety report?

L'safety report is the main regulatory document for sites classified as SEVESO (seuil bas/haut) et de nombreuses ICPE subject to authorisation. It follows the PHA structure (hazard โ†’ event โ†’ consequence) but uses terminology and requirements specific to French regulations: initiating event, top event, hazardous phenomenon, with probability assessed using the regulatory PIGC (Probability, Intensity, Severity, Kinetics) scale.

Unlike general-purpose PHA, probability in a safety report is not rated freely: it uses fixed probability classes (PIGC scale), a condition for the authorities to accept the file.

2Method steps

  1. Identify the system and its hazard (hazardous substance, stored energy, etc.).
  2. Describe the initiating event โ€” the failure or error that starts the accident sequence.
  3. Characterise the top event โ€” the resulting loss of containment or control.
  4. Describe the hazardous phenomenon that follows (fire, explosion, toxic release, etc.) and its potential consequences.
  5. Rate severity and probability using the PIGC regulatory scale (fixed, non-editable classes).
  6. List the risk-control measures โ€” passives (containment, separation distance) and actives (detection, automatic extinguishing)โ€”then position the scenario in the regulatory risk matrix (MMR cells).
Clearly distinguish a passive measure (works without action, e.g. a firewall) from an active measure (requires detection and action, e.g. automatic extinguishing): their expected reliability and justification differ.

3Example completed in Polyodos

Section to be completed โ€” step-by-step example (e.g. loss of containment from flammable-product storage), with Polyodos screenshots and positioning in the PIGC matrix.

4Common mistakes

MistakeWhy it is a problem
Customising the PIGC scaleThis regulatory scale is fixed. Changing it within a study removes the file's comparability and may lead the authorities to reject it.
Confusing the top event with the hazardous phenomenonThe top event is the loss of containment or control; the hazardous phenomenon is what follows (fire, explosion). One leads to the other, and they are not interchangeable.
Using only one risk-control measure per line of defenceA single barrier, even a reliable one, is generally insufficient to justify lowering the probability class. Defence in depth, with several independent measures, is expected.

5Further reading

FR