A review of probabilistic techniques: towards developing a probabilistic lifetime methodology in the creep regime

Lifetime assessment procedures for high temperature components incorporate conservatism to account for various uncertainties. These assessments are by nature deterministic, and do not provide advice on assessing their in-built conservatism. Building on conventional deterministic approaches for assessing plant components, this paper outlines the initial stages of developing a probabilistic methodology for high temperature assessments associated with creep and creep-fatigue. The principal aim is to identify the main sources of uncertainty through probabilistic analyses and suggest systematic approaches for estimating conservatism. Such a need is not currently addressed by conventional deterministic assessments. Selected probabilistic techniques are highlighted, and proposed as the building blocks of a prospective probabilistic creep methodology. These help build confidence in the underlying procedures, while guiding future work and areas of further investigation. A case-study is presented to demonstrate the utility of sensitivity analysis techniques in identifying and quantifying parameter uncertainties in the assessment results.