Application of probabilistic modelling to the lifetime management of nuclear boilers in the creep regime: Part 2

•Flow restrictions and metal losses adversely affect boiler creep-fatigue lives.•Rupture and crack initiation assessments employ the R5 methodology.•Probabilistic assessment provides a quantitative guide to the commercial risk.•Probabilistic assessment identifies the most significant parameters.

Monte Carlo probabilistic simulation has been applied to a large population of nominally identical components in an AGR boiler operating in the creep regime. Some of the components have a history of partial steam flow restrictions which can cause an elevation of their operating temperature, potentially raising the rate of creep life usage. Metal losses due to steam-side and gas-side oxidation and chemical cleaning operations can also exacerbate the rate of creep life usage. The R5 procedure has been used within a probabilistic program to calculate the expected frequency of both creep rupture and creep-fatigue crack initiation. The probabilistic approach is shown to provide a better quantitative guide to the commercial threat than traditional deterministic methodologies based on bounding data. In particular, probabilistic assessments identify the parameters which most significantly influence plant life.