Fitness for Service of Degraded Grade 91 Pipe

Abstract

The use of creep strength enhanced ferritic alloys such as Grade 91 in fossil power plants has become popular for high temperature piping applications. Since Grade 91 has higher stress allowables than Grade 22, a designer can specify thinner component wall thicknesses, resulting in lower through-wall thermal stresses during transient events and lower material and piping support costs. During the past two decades, Grade 91 has been used successfully in fossil power plants. However, this alloy has had some incidents of non-optimal weldment microstructure. In this case study, Brinell hardness tests of an ASME A182 Grade F91 (F91) wye block, including upstream and downstream F91 spools, revealed several readings of soft material, as low as 168HB. A study of creep rupture tests of degraded Grade 91 specimens revealed that the lower bound creep rupture curve of the degraded Grade 91 material is above the average creep rupture curve of Grade 22 material for the range of the specific piping operating stresses. Based on the empirical evidence that the average Grade 22 material creep rupture curve is conservative for the creep rupture of degraded Grade 91 material, a life consumption evaluation was performed for the degraded Grade 91 weldments using Grade 22 creep rupture properties. A life fraction analysis was performed considering the redistributed maximum principal stresses, based on simulation of piping displacements obtained from the hot and cold walkdowns. This study also considered the recent history of the specific piping system operating pressures and temperatures. This study also considered dissimilar metal welds, from ASME A182 Grade F91 (F91) to ASME A335 Grade P22 (P22) materials. It was determined that the Grades F91-to-F91 weldments had about 30% life consumption and the remaining lives were at least 7 years. The Grades F91-to-P22 weldments had less than 40% life consumption and the remaining lives were at least 15 years.

Experimental and Modeling Studies of Creep Performance of Modified 9Cr-1Mo Steels with or without Oxidation-Resistant Coating

“…)] +̇ (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14) where is the material parameter, is the work hardening coefficient of GBS. Appendix B gives details about how to derive and [ 57 ].…”

“…Combining Eqs. (4)(5)(6)(7)(8)(9)(10)(11) and (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14), only the transient part of Eq. (4-14) should be taken, since the steady-state part is already included in Eq.…”

“…To take oxidation effect into consideration in creep modeling, the parabolic oxidation law can be used to describe the growth of oxide scale, , as [ 122 , 123 , 124 ] = √2 (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18) where is the oxidation rate coefficient, which is an Arrhenius-type constant; is exposure time.…”

“…Since now ′ and include the oxidation-modified term for each mechanism, Eqs. (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15) natually describes the creep strain-time response with the involved mechanism contributions as well as the oxidation effect, and it natually reduces to material-intrinsic behavior when oxidation effect is absent.…”

“…[4][5][6][7][8][9][10][11] basically describes the secondary stage of creep when ′ is small or time is relatively short, and the tertiary stage when ′ is large or time is relatively long.For transient creep by GBS, the creep strain function has been derived to be[7] …”