火力ボイラ伝熱管のクリープ破断時間に対する確率論的評価の試み

抄録

A probabilistic evaluation for the creep rupture time of heat transfer tubes made from Super304H austenitic stainless steel in ultra-supercritical power plants is presented. A conventional assessment of creep life is based on deterministic evaluation method by using creep life prediction equation with constant values of stress and temperature. In the traditional assessment, estimated temperature is used due to the difficulty of temperature measurement of tubes exposed to combustion, and creep rupture time is evaluated conservatively due to variations of rupture time caused by heat-to-heat differences. In this study, assuming that the temperature uncertainty follows a normal distribution and the variation of creep rupture time follows a lognormal distribution, a Monte Carlo method and an approximate evaluation method are proposed to evaluate a creep rupture probability quantitatively. Then, the effect of temperature uncertainty on creep rupture probability was evaluated. Consequently, creep rupture probability at the rupture time obtained by the deterministic evaluation increases as the standard deviation of temperature increases. Furthermore, the result obtained by the Monte Carlo method is in good agreement with that of the approximate evaluation method.