CWT (combined water treatment) was introduced in Japan in 1990 and over 50 power generation boilers are now in operation. However, the effect of oxygenated treatment on the steam oxidation of the ferritic-martensitic steels and austenitic stainless steels that are used for superheaters and reheaters is currently far from clear. In this study, laboratory tests were used to examine the effect of the oxygen level of the feed water on the scale growth and the scale exfoliation propensity of T91 ferritic-martensitic steel and 300-series austenitic stainless steels, as represented by TP316H and TP347H (coarse- and fine-grained, respectively). The oxygen level of the feed water had little effect on the steam oxidation rates of all the steels tested. Hematite (Fe 2 O 3 ) formed in the outer layer of the oxide scales on both the ferritic and austenitic steels and is considered to have been encouraged in the simulated CWT atmosphere. The adhesion strength of the oxide scale formed on T91 in the simulated CWT atmosphere, that is, scale in which hematite was present, was lower than that of the oxide scale formed in the simulated AVT (all volatile treatment) atmosphere. The oxidation rate of fine-grained TP347H was confirmed to be slower than that of coarse-grained TP316H. Hematite significantly influenced the scale exfoliation of the austenitic steels and the critical oxide thickness for exfoliation decreased with increasing proportion of hematite in the outer scale.

The high-temperature oxidation of Fe-9Cr-1Mo steel in a CO 2 environment, with varying oxygen content (0.6-3%), was investigated at 700°C. While the steel heavily oxidized in pure CO 2 , the oxidation mass gain decreased significantly with increasing oxygen content. Microscopic analysis revealed the formation of Fe-rich nodules with an internal Cr-carbide layer beneath them. Notably, the number of nodules decreased with increasing oxygen content but remained independent of the oxidation time. To explain these observations, the authors propose that “intrinsic” defects within the initially formed protective Cr 2 O 3 scale facilitated gas permeation. This mechanism is believed to be responsible for the observed dependence of nodule formation on the oxygen content in the gas mixtures.

The acceptance of materials for long-term, safety-critical power generation applications requires multiple testing stages and data generation. Initial screening involves short-term exposures under simplified, constant atmospheres and temperatures, which can eliminate unsuitable materials but fail to distinguish between those with broadly acceptable properties. Subsequent pilot plant testing, costing over £100K for month-long exposures, is typically required. An intermediate laboratory testing step that better replicates in-service conditions would offer a cost-effective approach to material selection and lifetime prediction. For steam oxidation degradation, key experimental parameters—such as water chemistry, pressure, steam delivery, and flow rate—must be tailored to produce oxide scale morphologies similar to those observed in actual plant conditions. This study examines the effects of these parameters through steam exposure tests on ferritic (P92), austenitic (Esshete 1250), and superalloy (IN740) materials. Results indicate that oxidation rates vary with dissolved oxygen levels in feed water, increasing for austenitic materials and decreasing for ferritic materials, while also influencing spallation tendencies. Additionally, steam pressure and delivery methods impact oxidation rates and scale morphology. A comparison with service-exposed materials revealed that traditional oxide scale morphologies were not adequately replicated, whereas cyclic oxidation tests provided a closer match to service-grown scales.