Probabilistic Risk Assessment (PRA)

Probabilistic Risk Assessment is the comprehensive, quantitative method for analyzing risk in complex engineered systems by answering three questions: what can go wrong, how likely is it, and how bad would it be. Kaplan and Garrick's 1981 paper gave the field its enduring definition of risk as a set of triplets — scenario, frequency, and consequence — and showed how to extend that definition to incorporate uncertainty through probability distributions. The NASA Probabilistic Risk Assessment Procedures Guide (NASA/SP-2011-3421) operationalizes this framework for high-consequence aerospace systems, combining initiating-event analysis, event trees and fault trees, consequence modeling, and formal uncertainty propagation into an integrated assessment. Unlike qualitative hazard identification, PRA produces a quantified risk picture — typically a frequency-of-exceedance curve with explicit uncertainty bounds — that supports decisions about where scarce safety resources will reduce risk most.