Remaining Life Sensitivity to Longitudinal Weld Seam Peaking in High-temperature Low Chrome Piping

Introduction

Numerous failures of high-temperature, low chrome piping in the refining and power generation industries have been attributed to “peaking” of longitudinal weld seams. Typically, local weld seam peaking occurs during pipe manufacturing, where the rolled pipe locally deviates from a true circular cross-section at the weld, as shown in Figure 1. Additionally, most welded piping fabrication standards have no specific acceptance criteria for this type of out-of-roundness. Furthermore, some of the high-temperature pipes that have failed conformed to the required original fabrication tolerances, but no documented peaking measurements were available in the equipment inspection files. This makes it difficult to quantify the risk associated with elevated temperature operation of low chrome piping. Depending on original heat treatment, creep damage progression is known to be accelerated by the mismatch in creep properties of the weld deposit, heat affected zone (HAZ), and adjacent base metal. This property mismatch results in stress intensification and triaxial tension that can accelerate creep damage near the weldment. Longitudinal weld seam peaking can induce significant local bending stresses in the pressure boundary, and for piping components that operate in the creep regime, the presence of local peaking can lead to an increased propensity for creep crack initiation, propagation, and eventual gross rupture of the pressure boundary.

This article summarizes a recent finite element analysis (FEA)-based study that employs creep simulation techniques to investigate the elevated temperature response of piping with peaked longitudinal weld seams. The objective of this study is to use analytical methods to estimate the remaining life of specific low chrome piping geometries and to assess the sensitivity in results to variations in key parameters, such as operating temperature, magnitude of longitudinal weld seam peaking, and the effect of pipe heat treatment. This study compares the creep damage progression for multiple examples of 30-inch and 36-inch diameter 1 1/4 Cr - 1/2 Mo pipes with and without local weld seam peaking. Simulation techniques utilizing the Materials Properties Council (MPC) Omega creep methodology, such as the ones discussed herein, are valuable in estimating remaining life of in-service piping and establishing recommended local weld seam peaking fabrication tolerances, appropriate inspection practices, and reasonable non-destructive examination (NDE) intervals for high-temperature low chrome piping systems.

Documented Failures

Early failures of 1 1/4 Cr - 1/2 Mo components, such as super-heater outlet headers and piping components, operating in the creep regime were attributed to cracking [2].  In 1968, the ASME Code reduced the (time-dependent) allowable stresses for 1 1/4 Cr - 1/2 Mo materials, such that the allowable stresses at 1,000°F and 1,050°F were reduced by 16 and 26 percent, respectively [2]. Therefore, low chrome headers and piping components operating in the creep regime designed during the 1950s and 1960s are potentially under-designed. A second decrease in allowable stresses, prompted by industry failures, took place in the 1989 addenda to the ASME Code [3], where the allowable stresses for 1 1/4 Cr - 1/2 Mo decreased from 6.9 ksi to 6.3 ksi at 1,000°F and from 4.6 ksi to 4.2 ksi at 1,050°F. Additionally, ASME B31.3 introduced a weld joint strength reduction factor (W) in the 2004 Edition [4]. The purpose of this factor is to account for long-term behavior of welds at elevated temperatures in the absence of creep tests (above 950°F). ASME B31.1 introduced this parameter in the 2008 Edition [5].