Two meanings. In process safety, QRA means estimating individual and societal risk of harm from hazards, with F-N curves and risk contours. This entry covers project QRA: the cost and schedule risk of delivering a capital project.
Qualitative versus quantitative
| Qualitative assessment | Quantitative assessment |
|---|
| Output | Ratings on a 5x5 matrix | P50, P80 and P90 costs and dates |
| Combines risks? | No, each risk is scored alone | Yes, all risks and uncertainty together |
| Answers | Which risks look worst | How much contingency and time you need, and why |
| Best for | Early screening | Sanction, funding, baselines and contingency |
How project QRA works
Quantify each risk
Give each risk a probability and a cost or time impact range, written as a clear cause, event and effect.
Add estimating uncertainty
Apply three-point ranges to cost items and activity durations.
Build the model
Link risks to the cost estimate (QCRA) or the schedule (QSRA), with correlation where items move together.
Simulate
Run a Monte Carlo simulation to produce the distribution of outcomes.
Decide
Set contingency, commit dates and rank mitigations by the value they add.
A worked number
A risk with a 30% chance of a $4m to $10m impact (most likely $6m) has an expected value of about $2m. A heat map would call it "medium". A QRA puts it in the model alongside every other risk, and the simulation shows how much it adds at P80, which is usually more than its expected value.
Common mistakes
- Multiplying probability by impact and adding it up. Summed expected values ignore the spread, so they understate the contingency needed at P80.
- Converting heat-map scores into numbers. A "4 x 3" is not a quantity.
- Leaving out estimating uncertainty. Discrete risks alone miss much of the real spread.
Related terms
