Development of Reliability Assessment Method for Piping Containing Local Metal Loss at Structure Discontinuity

<jats:p>In chemical plants, Corrosion Under Insulation (CUI) is a typical maintenance issue of internal pressed pipes. This form of localized corrosion leads to metal loss at the outside surface of pipes, reduces the pipe’s resistance to internal pressure [1]. Once the local metal loss is inspected, it is important to apply a Fitness-For-Service (FFS) assessment to evaluate the remaining safety margin against working internal pressure. However, if the local metal loss occurs at the piping discontinuity, it is impossible to calculate the critical collapsing pressure without any existing theoretical expression. Moreover, evaluation based upon standards may produce too much conservative evaluation and lead to unnecessary repairing-replacement cost. In this study, both deterministic analysis and probabilistic reliability analysis are performed to evaluate the safety margin of piping containing local metal loss at piping branch connection, and the obtained results are compared with those evaluated based upon API579-1/ASME FFS-1 standard. In addition, sensitivity analysis is conducted to show the dependence of metal loss geometries regarding both width and depth, material strength and working internal pressure on the safety margin. Finally, a proper reliability assessment procedure by using advanced Response Surface Method (RSM) is proposed. In the deterministic analysis part, a FEM structural analysis is performed to calculate the critical collapsing internal pressures (CIP) for several cases of various metal loss geometries regarding width and depth and material tensile strength. These results are compared with maximum allowable working internal pressure (MAWP) calculated by the procedures which are provided in API579-1/ASME FFS-1 standard. (The safety margin is defined as CIP/ MAWP). It is shown that the safety margins are greater than 6 in most of the cases. Since there are uncertainties in the geometrical dimensions of local metal loss, material tensile strength and working internal pressure, it is necessary to perform a reliability analysis to quantify the probability of failure (Pf) caused by these uncertainties. In the probabilistic analysis part, the metal loss geometries, material strength and internal pressure are considered as random variable with distributions. The reliability index of each group of geometries-strength-pressure combinations is calculated by first order reliability method (FORM). The reliability index of the case working under MAWP is also investigated. In order to define the limit state function, response surface method is used to predict the response which relates the critical collapsing internal pressure to the metal loss geometries and material strength. Finally, the sensitivity analysis is performed to clarify the dependence of metal loss geometries, material strength, and internal pressure on the Pf, and to specify the dominant variables and negligible variables.</jats:p>