Cycle frequency affects both high-temperature oxidation behavior and the method in which the cyclic test is conducted. Several issues are discussed using examples taken from results for Ni-base and Fe-base, alumina-forming alloys. For alloys that form adherent scales, cycle frequency has little effect on results over extended test times ( ≥500 hr). When an alloy forms a less adherent scale, reducing the cycle time often has the expected effect of increasing the mass loss per unit exposure time; however, the opposite effect is observed in other cases. Low-frequency cycle experiments can be conducted with specimens contained in alumina crucibles. This has the important benefit of collecting the spalled oxide and measuring the “total” mass gain, equivalent to the metal wastage. However, higher-frequency-cyclic tests cannot be performed with crucibles because of the large thermal mass and thermal-shock problems of alumina crucibles. The test method and cycle frequency ultimately have a strong effect on lifetime predictions.

Oxide dispersion-strengthened (ODS) FeCrAl ferritic steels constitute a class of alloys with a combination of mechanical strength capabilities and environmental resistance that permits operation at temperatures significantly higher than available classes of wrought alloy. A significant issue for these ODS alloys is that, because of the extreme operating temperatures, unexpected failure of the protective scale in service could lead to catastrophic degradation of high-temperature, high-pressure components. In the absence of handbook data for quantification of their high-temperature environmental degradation, significant efforts have been made to develop approaches for predictive oxidation lifetime modeling. The main basis for such models is an aspect of oxidation behavior peculiar to these alloys at very high temperatures, specifically, their ability to maintain a protective alumina scale until practically all available Al in the alloy has been consumed. At the high temperatures involved the concentration profile for Al throughout the component remains flat while the alumina scale is intact, so that knowledge of the oxidation kinetics provides a basis for calculating the time to exhaustion of the alloy’s Al reservoir, hence service life. Difficulties arise when unanticipated gradients in the Al concentration profile are introduced following, for instance, premature local mechanical failure of the protective scale. The intent of this work was to consider if and how scale failures change with differences in surface shape, and how related mechanistic understanding of the actual oxidation behavior may be used to extend the utility of the modeling approaches. Experimental results for coupons of different shapes were very largely in agreement with the basic tenets of the lifetime modeling approaches, but a notable finding was that cylindrical shapes reproducibly exhibited a reduction in unexpected early scale failures, leading to more consistent oxidation-limited lifetimes.

Alloys containing a dispersion of fine, stable oxides can develop useful creep strength to 90 percent of their melting temperature, providing a capability of 100 to 150 degrees C or more over conventional high-temperature alloys. Such oxide dispersion-strengthened alloys allow other alloying additions to be optimized to provide superior oxidation resistance. Activities to develop such alloys for severe high-temperature applications are described, together with opportunities for further understanding of the development and growth of protective oxide scales of chromia and alumina. In particular, the use of oxides such as ThO2 and Y2O3 as dispersd phases resulted in the generation of important observations of the 'reactive element effect' in high-temperature oxidation. By analogy to the separation of the requirements for alloy strengthening and environmental resistance by insertion of a dispersion of oxide particles, examples are provided of other issues that can be addressed by various dispersed phases, without compromising existing alloy properties.

The influence of frequency of thermal cycling on the oxidation behaviour of alloys TP446, TP310, and Inconel 617 in dry air at very high temperatures (871 degrees C and 982 degrees C) is reported. The effects on the iron-basedalloys were complex, whereas that of the Ni-based alloy was not greatly affected by cycling frequency. Unexpected trends were increased oxidation rates of TP310 and Inconel 617 with increased cycle frequency at 871 degrees C, whereas those of TP446 decreased. At 982 degrees C, the oxidation rates of TP310 and TP446 decreased with increasing cycle frequency, though less so for TP310. While the scales formed on the iron-based alloys consisted of a basal layer corresponding to chromia containing Mn, and an outer layer corresponding to manganese chromate, differences in the rates of transport of Cr, Mn and Si in the alloys at the different temperatures appeared to account in large measure for the differences in response to thermal cycling.

The influence of frequency of thermal cycling on the oxidation behaviour of alloys TP446, TP310, and Inconel 617 in dry air at very high temperatures (871°C and 982°C) is reported. The effects on the iron-basedalloys were complex, whereas that of the Ni-based alloy was not greatly affected by cycling frequency. Unexpected trends were increased oxidation rates of TP310 and Inconel 617 with increased cycle frequency at 871°C, whereas those of TP446 decreased. At 982°C, the oxidation rates of TP310 and TP446 decreased with increasing cycle frequency, though less so for TP310. While the scales formed on the iron-based alloys consisted of a basal layer corresponding to chromia containing Mn, and an outer layer corresponding to manganese chromate, differences in the rates of transport of Cr, Mn and Si in the alloys at the different temperatures appeared to account in large measure for the differences in response to thermal cycling.

Over the last 40 years oxide growth and exfoliation (OGE) in superheater and reheater tubing have been responsible for a number of power plant problems which seriously have affected reliability. In the same time period, the authors have collected a data base of scale morphologies that has been used to describe in detail the progression of oxide scale development to the point where failure can occur. The concomitant evolution of knowledge of the factors that determine the mode of scale growth and failure in steam has provided the foundation for defining the specific stages in that progression, understanding differences among ferritic and austenitic alloys, and for categorizing the influence of plant operating characteristics. In particular, while tube/steam temperature and the maximum temperature drop at plant shutdown are major variables, the specific cycle chemistry used for the plant feedwater has very little influence. Key stages in the progression of scale growth to the point of failure are identified as OGE indices that are specific for ferritic and austenitic steels. These indices are intended to be used proactively to determine the current condition of a superheater or reheater on the path to exfoliation and possible plant damage. Also, by analyzing samples of exfoliant or oxide deposits responsible for damage, the origin of the oxide, and thus a possible superheater or reheater problem, can be identified retroactively.

In an ongoing effort to predict corrosion/oxidation rates of metallic heat exchangers for open-fired supercritical CO2 (sCO2) power cycles, candidate metallic materials were subjected to a series of isothermal laboratory tests in a simulated sCO2 working fluid at high temperatures and 200-bar pressure. The extent of oxide growth and morphologies of the formed oxide scales were studied using a scanning electron microscope (SEM). Recent advances in SEM detector technology allow for ‘closer’ examination of the oxide scales using ultra-high resolution imaging capabilities without limitations to the field-of-view, unlike other characterization techniques like TEM. Variations in morphological features of the oxide layers were investigated for five alloys (ferritic-Grade 91, austenitic-304H and TP347H and nickel based 617 and 740H) using such high-resolution imaging techniques. Oxide scale thickness measurements were performed and compared for the five alloys after test durations of 1000, 3500, and 5000 hours, at test temperatures of 700°C. It was found that ferritic and austenitic materials exhibit similar oxide scale features as typically seen when exposed to high temperature steam. In contrast, nickel based alloys show a bimodal scale structure consisting predominately of a sub-micron size uniform single layer with dispersed formations of surface nodules that exhibit a duplex scale structure.

Recent experience of rapid corrosion of superheater tubes in coal-fired boilers has led to the application of weld-overlay coatings, which have proved effective but are expensive and can degrade mechanical tube life. Lack of understanding of how such corrosion differs from the established mechanism involving low-melting, complex, alkali-iron trisulphates is a major obstacle to devising alternative protective measures. Suggested scenarios for such accelerated corrosion are considered, including (1) increased tube surface temperatures allowing other sulphate mixtures to melt, analogous to Hot Corrosion mechanisms in gas turbines; (2) changes in the corrosivity of tube deposits from carryover of reduced species; and (3) increased release of corrosive species from the coal under substoichiometric combustion conditions. The potential for such scenarios to change the corrosive environment is assessed on the basis of current mechanistic understanding, though the general lack of phase diagrams for relevant ternary and quaternary sulphate systems is a major drawback to specifying the processes involved.

This project was the first research to address oxidation of alloys under supercritical CO2 conditions relevant to a semi-open Allam Cycle system. The levels of impurities expected in the CO2 for typical operation were determined by thermodynamic and mass balance calculations, and a test rig was assembled and used to run corrosion tests at temperatures from 650 to 750°C in CO2 at 200 bar for up to 5,000h, with and without impurities. Oxidation rates were measured for seven alloys representing high-strength ferritic steels, standard austenitic steels, and Ni-based alloys with higher-temperature capabilities. The very thin, protective scales formed on the high-temperature alloys provided significant challenges in characterization and thickness measurement. The rates of mass gain and scale thickening were possibly slower when oxidizing impurities were present in the sCO2, and the scale morphologies formed on the ferritic and austenitic steels were consistent with expectations, and similar to those formed in high-pressure steam, with some potential influences of C. Some surface hardening (possibly due to carbon uptake) was identified in ferritic steels Grade 91 and VM12, and appeared more severe in commercially-pure CO2. Hardening was also observed in austenitic steel TP304H, but that in HR3C appeared anomalous, probably the result of work-hardening from specimen preparation. No hardening was found in Ni-base alloys IN617 and IN740H. An existing EPRI Oxide Exfoliation Model was modified for this application and used to evaluate the potential impact of the scales grown in sCO2 on service lifetimes in compact heat exchanger designs. Results suggested that reduction in flow area by simple oxide growth as well as by accumulation of exfoliated scale may have a major effect on the design of small-channel heat exchangers. In addition, the specific oxidation behavior of each alloy strongly influences the relationship of channel wall thickness to service lifetime.

A laboratory study was performed by exposing seven candidate heat-exchanger alloys to simulated conditions of advanced open supercritical CO2 (sCO2) Brayton power cycles. The alloys, consisting of ferritic steels, austenitic stainless steels, and nickel-base alloys, were exposed to impure CO2 containing initially 3.6% O2 and 5.3% H2O at a constant pressure of 200 bar. The test temperatures varied from 650 to 750°C. The total exposure time of each test was 1000 hours, with alloy coupons removed from the reaction retort after approximately 300, 600, and 1000 hours. Metallurgical analyses were performed on the exposed samples using optical microscopy, SEM/EDS, and a micro-hardness tester to characterize the scale morphologies and the extent of carburization in the alloys. Results indicate that the oxidation kinetics were significantly influenced by the test temperature and alloy composition. In general, a change of 50°C in temperature can lead to a change in oxide thickness by a factor of up to 4. By comparison, the ferritic steels suffered the highest oxidation and carburization attack, followed by austenitic stainless steels and nickel-base alloys. Alloys containing a high combined Cr and Ni content offered the best resistance to oxidation and carburization. However, only the nickel-base alloys exhibited adequate carburization resistance to the laboratory conditions after the exposure time of 1000 hours.

Because of the problems experienced with steam-side oxidation in commercial power plants, there has been continuing interest in better understanding the steam oxidation behavior of creep strength enhanced ferritic steels such as grades 23, 24 and 91 as well as 300-series stainless steels such as 347H and 304H. Analysis of field-exposed tubes has provided information on the oxidation reaction products but relatively few specimens are available and there is limited information about the kinetics. Specimens have included tube sections with a shot peened surface, a treatment that is now widely used for austenitic boiler tubes. To complement this information, additional laboratory studies have been conducted in 1bar steam at 600 degrees-650 degrees C on coupons cut from conventional and shot-peened tubing. Exposures of 1-15 kh provide some information on the steam oxidation kinetics for the various alloys classes. While shot-peened type 304H retained its beneficial effect on oxidation resistance past 10,000 h at 600 degrees and 625 degrees C, the benefit appeared to decline after similar exposures at 650 degrees C.

Interpretation and use of a body of field data for corrosion of superheater and reheater (SH/RH) tubes in coal fired boilers was found to be unsatisfactory without some indication of how the corrosive environment encountered differed when different coals were burned. A new factor intended to represent the relative corrosion potential (RCP) of the coal burned is suggested, based on the accepted mechanism of accelerated corrosion of SH/RH tubes by the development of low melting complex sulphates beneath ash deposits. Initial testing of the RCP concept was encouraging, given the difficulty in some cases of ensuring accurate representation of the actual coal burned during the corrosion exposures. However, some data suggested that new boiler operating modes, such as various approaches to emissions reduction through staging of the combustion process, appeared to be capable of causing significantly more rapid corrosion of SH/RH tubes than would be expected from the RCP value for the coal. Although increased corrosion with some emissions control systems has been recognised in practice, understanding of the particular corrosion process involved is lacking. Some degree of mechanistic understanding of the key features of this form of corrosion is needed to provide a firm basis for application of concepts such as RCP or development of improved versions.

Research into improved materials systems and associated manufacturing and reliability issues is a major part of initiatives to produce cleaner and cheaper energy systems in the UK and the USA. Under the auspices of a Memorandum of Understanding on Energy R&D, a work programme concerned with steam oxidation has been conducted. The focus was on the generation of definitive information regarding the oxidation behaviour in steam of current and developmental ferritic steels, austenitic steels, and nickel-based alloys required to enable advanced steam power cycles. The results were intended to provide a basis for quantifying the rate of metal loss expected under advanced steam cycle conditions, as well as understanding of the evolution of oxide scale morphologies with time and temperature to identify features that could influence scale exfoliation characteristics. This understanding and acquired data were used to develop and validate models of oxide growth and loss by exfoliation. This paper provides an overview of the activity and highlights a selection of the results coming from the programme.

Abstract A model based on a concept of “fraction of exfoliated area” as a function of oxide scale strain energy was developed to predict the extent of exfoliation of steam-side scale from boiler tube superheater loops. As compared with the Armitt diagram, which can be used to predict when scale damage and exfoliation would be likely to occur, a “fraction of exfoliated area” approach provides an estimation of mass of scale released and the fraction of tube likely to be blocked by the exfoliation. This paper gives results for the extent of blockage expected in a single bend of a superheater loop was predicted as a function of operating time, bend geometry, and outlet steam temperature under realistic service conditions that include outages. The deposits of exfoliated scale were assumed to be distributed horizontally the tubes bends. Three types of bends were considered: regular bends, short bends, and hairpin bends. The progressive increase in steam and tube temperatures along a single loop of superheater tubing and the ensuing variation of oxide scale thickness are considered. Numerical simulation results for a superheater loop made of TP347H austenitic steel indicated that tube blockage fractions larger than 50% are likely to occur within the first two years of boiler operation (with regularly scheduled outages) for outlet tube temperatures of 540-570°C, which is consistent with practical experience. Higher blockage fractions were predicted for tubes with hairpin bends than for tubes with regular bends, of length that are larger than five internal tube diameters. Finally, the blockage model presented can be used with some confidence to devise operating schedules for managing the consequences of oxide scale exfoliation based on projections of time to some critical blockage fraction for specific boiler operating conditions.

Materials are the key enabling technology driving development of high-efficiency power plants, which requires better fundamental and practical understanding of the effects of operating environments.

AbstractObservations of steam-formed scales on type T22 ferritic steel reheater tubes have provided some insights into the scale morphologies associated with exfoliation. Measurements indicated that as the total thickness of the adherent scales increased, the ratio of the thicknesses of the main inner (Fe–Cr spinel) and outer (magnetite) layers increased continuously to a maximum of approximately 2.5. The often-assumed thickness ratio of 1.0 did not persist after a laminated structure of pairs of spinel and magnetite layers was developed in the main inner layer. The development of such multi-laminations appeared to be associated with a decrease in the rate of growth of the main outer, magnetite layer. In this particular set of samples, exfoliation occurred in scales of total thickness greater than approximately 270 μm, and involved separation of a single pair of oxide layers of similar thickness that consisted of the original outer magnetite layer and part of the original spinel layer. The scale lost in subsequent exfoliation events at essentially the same site consisted of a further pair of layers comprising the new magnetite layer re-grown after the initial event, and a further part of the original spinel layer. Exfoliation events did not appear to involve the whole of the spinel layer existing at that time, only the equiaxed region immediately subjacent to the magnetite and a portion of the laminated structure.Keywords: T22high-temperature oxidationsteamexfoliationscale morphology

The sections in this article are Introduction Oxidation Fundamentals Protective Oxide Scale Formation Reactive Elements Overview Effects on Oxidation Production of Alloys with RE Additions Applicability and Limitations Pesting NiAl Based Intermetallics Overview NiAl Transient Oxidation Steady-State Oxidation Effect of Al Content Alloying Additions Other Corrosive Environments Ni3Al Transient Oxidation Steady-State Oxidation Alloying Effects FeAl Based Intermetallics Overview Historical Perspective Thermodynamic and Kinetic Considerations in the FeAlO System Fe3Al FeAl Other Corrosive Environments Sulfur-Containing Gases Chlorine-Containing Gases Carbon-Containing Gases Molten Salts and Condensed Deposits TiAl Based Intermetallics Overview TiAl2 TiAl3 τ Phase γ-TiAl Thermodynamic Considerations The Nitrogen Effect Effects of Alloying Additions Engineering Considerations Other Corrosive Environments α2-Ti3Al and Orthorhombic Ti2AlNb α2 Alloys Orthorhombic Alloys NbAl Based Intermetallics Overview NbAl3 Nb2Al Nb3Al NbTiAl Precious Metal, Exotic, and Miscellaneous Aluminides PtAl Intermetallics IrAl Intermetallics RuAl Intermetallics CoAl Intermetallics VAl Intermetallics Laves Phases and In-Situ Composites Overview Single-Phase Laves Alloys NbNbCr2 In-Situ Composites CrXCr2 Type In-Situ Composites NiAl-Based In-Situ Composites γ-TiAl + Ti(Cr,Al)2 and NbAl3 + Nb(Cr,Al)2 In-Situ Composites Silicides Overview and General Considerations MoSi2 Overview High-Temperature Oxidation Accelerated Oxidation and Pesting Effects of Alloying Additions Composites Other Corrosive Environments Mo5Si3 TiSi2 and Ti5Si3 TiSi2 Ti5Si3 V5Si3 Cr3Si Fe and Ni Silicides Other Silicides Beryllides Overview Complex Beryllides–MBe13, MBe12, and M2Be17 MBe2 MBe Phases Acknowledgements

Proposed open and closed Brayton-type cycle systems employing supercritical CO2 (sCO(2)) as the working fluid develop hot gas path environments that present challenging requirements of strength and environmental resistance for the materials of construction. The likely materials properties required by different cycle configurations are examined and compared with the capabilities of available materials. In some instances where peak operating conditions in the turbines in closed-cycle systems approach 500 - 700 degrees C at 200 bar, or 1150 degrees C at 300 bar for open cycles, the range of available alloys with the requisite temperature-strength and capabilities is limited.The cycles are highly recuperated: hot, lower-pressure turbine exhaust is used to pre-heat high-pressure. CO2 entering the external heater or combustor just upstream of the turbine. Obviously, there is a need to match alloy capabilities with the performance requirements of individual components. Where alloys are employed under conditions beyond current experience, or in unusual configurations (the unique properties of sCO(2) may impact the design of some components), practical acceptance will depend on successful qualification testing involving close collaboration among equipment manufacturers, materials suppliers, and materials research and development groups.Examination of the range of scenarios published for sCO(2) turbine systems suggests that approaches used in the U.S. Advanced, Ultra-Supercritical (A-USC) Steam Boiler program, and some of the data generated for fabrication, joining, mechanical properties, and performance in simulated service environments in that program, will be applicable to some of the components in these systems. An attempt to elaborate and prioritize the materials property and qualification activities required for the successful realization of these sCO(2) cycles is presented as a guide to activities needed to facilitate materials selection.

This paper provides an introduction to a comprehensive model being developed to predict and control oxide scale exfoliation from the steam-side of superheater and reheater tubes in steam boilers. The model deals with the main phenomena involved in scale growth and failure in steam, and incorporates major variables related to boiler design and operation. The considerations used to calculate oxide growth under the specific constrains of small diameter tubes carrying high-pressure steam and operating with large temperature gradients under temperature and pressure cycling conditions, as well as the evolution of stresses and strains in the scales, are indicated but only a cursory description is given of the details of the analytical treatments. An example is presented of calculations made with the model to predict the extent of blockage expected in a single superheater loop as a function of time and outlet steam temperature under several realistic service conditions. The results suggest that problems due to scale exfoliation would be expected early in the operating life of superheater tubes made from austenitic steel TP347H.