Risk Assessment and Hazard Identification
Core concepts
Hazard identification asks what could cause harm; risk assessment estimates how likely and how severe that harm is. Identification methods include job hazard analysis, inspections, checklists, what-if analysis, HAZOP, FMEA, fault tree analysis, task analysis, change review, and review of incident and near-miss data. Risk is commonly expressed as likelihood × severity, but matrices are ordinal scales, not precise arithmetic. Assessment may be qualitative, semi-quantitative, or quantitative; exposure assessment may require sampling and comparison with occupational exposure limits. Consider routine and non-routine tasks, emergency situations, maintenance, and foreseeable misuse. Controls follow the hierarchy, and residual risk must be evaluated.
Traps
- Multiplying ordinal matrix scores and treating the product as a precise risk value.
- Calling an inspection a risk assessment when no severity or likelihood judgment is made.
- Assessing only normal operations and missing startup, shutdown, and maintenance.
- Treating “acceptable” as risk-free; acceptable or tolerable risk still needs monitoring.
- Ignoring change: new chemicals, equipment, staffing, or work methods alter risk.
Worked example
For a solvent parts-washing task, a team lists hazards (skin contact, vapor inhalation, fire), rates exposure as likely and severity as moderate without controls, then applies local exhaust ventilation, closed containers, and gloves. After controls, likelihood drops and residual risk is low but not zero. Rule of thumb: identify hazards first, then assess risk, then select controls, then reassess.