Oxidation kinetics of steels for improved life assessment for steam generating systems
2017-02-23T23:58:08Z (GMT) by
A detailed understanding of the steam oxidation kinetics of boiler steels is required for creep rupture based remaining life assessment of high temperature components. This has significant industrial importance, particularly in the case of boiler tubing in thermal power stations. Although the oxidation behaviour of traditionally employed steels is quite well understood, there is significantly less known about the more advanced boiler alloys designed to meet the creep strength requirements of higher operating temperatures and pressures utilised in modern plants with reduced greenhouse gas emissions. This research was carried out to improve understanding of the oxidation kinetics of these steels through direct comparison against traditional alloys using a novel approach designed to facilitate reliable laboratory based steam oxidation testing. Materials investigated represent both traditional and advanced variants from several of the major boiler alloy classes, including a 2.25% Cr ferritic steel (T23), 9% Cr martensitic steels (T9, T91, T92), 18-25% Cr austenitic stainless steels (347H, 347HFG, Super304H, HR3C) and an alumina-forming austenitic steel (AFA: Fe-25% Ni-14% Cr-3.5% Al-2.5% Nb). Alloys were subjected to long term high temperature oxidation at two specifically developed test facilities that represent considerably different approaches for studying steam oxidation. A simulated steam environment consisting of 50% water vapour in argon carrier gas was generated in the first experimental setup, where oxidation exposures were performed at atmospheric pressure for up to 7,510 h in the temperature range of 650 to 760°C for all alloys. The second setup utilised pressurised steam from the hot reheat line at Loy Yang B Power Station (Victoria, Australia). Alloys were exposed to oxidation in plant steam at 4.2 MPa for up to 12,170 h at metal temperatures ranging from 540°C for T23 to 810°C for AFA. Oxide scales were assessed using optical microscopy, non-destructive ultrasonic testing, Raman spectroscopy, SEM-EDS and mass-change measurements. Correlations describing oxide growth kinetics were determined for T23, each of the 9% Cr alloys and both Grade 347 stainless steel variants investigated in the study. In each case the growth rates well represented by parabolic kinetics. Distinct differences in steam oxidation resistance were observed between traditional and advanced alloys. This lends support to the argument that oxidation kinetics used for life assessment should be determined for individual alloys, rather than applied across an entire class of boiler steels. The oxide scale morphologies and growth rates observed for alloys in 50% H₂O/Ar and plant steam were generally consistent. The exception to this was more ready breakdown of the protective Cr₂O₃ layer developed on austenitic steels and one of the 9% Cr alloys (T92) in plant steam, which was attributed to the higher steam pressure and moisture content. The thin and irregular oxide scales that developed on austenitic alloys could not be measured by ultrasonic thickness testing and were better suited to mass based assessment of oxide scale growth. The findings are promising for future studies involving the two steam oxidation laboratories, however further testing of additional alloy batches is required to validate the correlations for accurate life assessment.