Frequency Distribution Curves for Main Steam Piping Multiaxial Stresses

Abstract

A high energy piping (HEP) asset integrity management program is important for the safety of plant personnel and reliability of the generating unit. HEP weldment failures have resulted in serious injuries, fatalities, extensive damage of components, and significant lost generation. The main steam (MS) piping system is one of the most critical HEP systems. Creep damage assessment in MS piping systems should include the evaluation of multiaxial stresses associated with field conditions and significant anomalies, such as malfunctioning supports and significant displacement interferences. This paper presents empirical data illustrating that lead-the-fleet girth welds of MS piping systems have creep failures which can be successfully ranked by a multiaxial stress parameter, such as maximum principal stress. Both the as-found elastic (initial) stress and inelastic (redistributed) stress at the piping outside diameter surface are evaluated for the base metal of three MS piping systems. Frequency distribution curves are then developed for the initial and redistributed piping stresses. The frequency distribution curves are subsequently included on a Larson Miller Parameter (LMP) plot for the applicable material, revealing the few critical (lead-the-fleet) girth welds selected for nondestructive examination (NDE). By including an evaluation of significant field anomalies, multiaxial operating stress on the outside surface, and weldment performance, it is shown that there is a good correlation of calculated creep stress versus the operating time of observed creep damage. This process also reveals the large number of MS piping girth welds that have insufficient applied stress to have substantial creep damage within the expected unit life time (assuming no major fabrication defects). API 579 recommends an effective stress to compute the creep rupture life using the LMP. This constitutive stress equation includes a combination of the maximum principal, von Mises, and hydrostatic stresses. Considering the stresses in these three MS piping systems, this paper reveals that when the axial and hoop stresses are nearly the same values, the API 579 effective stress may be 10% greater than the maximum principal stress. However, the maximum principal stresses are greater than the API 579 effective stresses at the maximum stress locations in the three MS piping systems, because the axial stresses are significantly greater than the hoop stresses. This study also provides a comparison of the results of a conventional American Society of Mechanical Engineers (ASME) B31.1 Code as-designed sustained stress analysis versus the redistributed maximum principal stresses for a complete set of MS piping system nodes. A comparison of Code-sustained load versus redistributed maximum principal stress results are illustrated on frequency distribution curves.