Process Safety Hazards and Risk Assessment
Hazard types and assessment methods
Candidates must recognise the main hazard families: fire, explosion (including vapour cloud, dust and boiling liquid expanding vapour explosion), toxic release, reactive chemistry, and asphyxiation. Understand the fire triangle and flammability envelope - flash point, autoignition temperature, lower and upper flammability limits - and how inerting, ventilation and elimination of ignition sources control them.
Risk assessment methods must be matched to purpose. Qualitative tools include HAZOP (systematic deviation study using guide words), What-If analysis, checklists and FMEA. Semi-quantitative tools include risk matrices and layers of protection analysis (LOPA), which compares scenario frequency with independent protection layers against a target. Quantitative risk assessment (QRA) estimates individual and societal risk. Consequence modelling (dispersion, fire, explosion) supports all of these.
Traps in questions
- Confusing HAZOP (deviation-based, team-based) with FMEA (component failure-based).
- Counting the same safeguard twice in LOPA - protection layers must be independent.
- Treating a risk matrix score as absolute rather than a decision aid.
- Forgetting human error and domino effects as credible initiating events.
Worked example
A flammable solvent drum store: identify release scenario, estimate frequency from historical data, list independent protection layers (bunding, ventilation, ignition control, detection), and compare residual risk against the site's tolerability criteria. If a layer depends on the same utility or operator action as another, it is not independent.
Rule of thumb
A protection layer must be independent, auditable and capable of preventing the scenario on its own.