In flexible operation with increased number of startup, shutdown, and load fluctuations, thermal fatigue damage is exacerbated along with existing creep damage in power plant pipe and pressure vessels. Recently, cracks were found in the start-up vent pipe branching from the reheat steam pipe within a heat recovery steam generator(HRSG) of J-class gas turbine, occurring in the P92 base material and repair welds. This pipe has been used at the power plant for about 10 years. Microstructural analysis of the cross-section indicated that the cracks were primarily due to thermal fatigue, growing within the grains without changing direction along the grain boundaries. To identify the damage mechanism and evaluate the remaining life, temperature and strain monitoring were taken from the damaged piping during startup and normal operation.

The microstructure of MoSiBTiC alloys is very complex, with three to four constituent phases and characteristic structures such as fine precipitates and lamellar structures. To perform the microstructural analysis efficiently, image segmentation was first performed for each phase of the microstructural images. Utilizing the Trainable Weka Segmentation method based on machine learning, the required segmentation time was dramatically reduced. Furthermore, by pre-adjusting the contrast of the images, the segmentation could be performed accurately for gray phases with different shades of gray. In addition, the U-Net method, based on deep learning, could perform highly accurate segmentation of characteristic microstructures consisting of multiple phases. The correlations between microstructural features and hardness were investigated using the segmented images in this study. The findings revealed that the volume fraction of each phase and the number of TiC clusters within the field of view significantly influenced hardness. This suggests that the hardness of MoSiBTiC alloys may be controlled by controlling the amount of TiC precipitates.

This study conducted creep tests, microstructural, and hardness analyses on SA213T23-TP347H dissimilar weld joints of long-term serviced coal-fired boiler final superheater tube. The welded joint (SA213 T23-TP347H) of the superheater tube, after approximately 105,000 hours of service, was sampled for creep life assessment and maintenance planning. Creep tests were conducted at 600°C under three stress conditions: 100, 140, and 160MPa. Most cracks were observed in the heat-affected zone of T23, and compared to unused tubes, the creep life consumption rate was approximately 90%. All dissimilar weld joints used welding rods similar in chemical composition to T23, and significant hardness reduction occurred in the flame-affected zone.

The time-dependent behavior of 9Cr creep strength enhanced ferritic (CSEF) steels has long fixated on the creep life recorded in uniaxial constant load creep tests. This focus is a consequence of the need to develop stress allowable values for use in the design by formulae approach of rules for new construction. The use of simple Design by Formula rules is justified in part by the assumption that the alloys used will invariably demonstrate high creep ductility. There appears to be little awareness regarding the implication(s) that creep ductility has on structural performance when mechanical or metallurgical notches (e.g., welds) are present in the component design or fabricated component. This reduced awareness regarding the role of ductility is largely because low alloy CrMo steels used for very many years typically were creep ductile. This paper focuses on the structural response from selected tests that have been commissioned or executed by EPRI over the last decade. The results of these tests demonstrate unambiguously the importance that creep ductility has on long-term, time-dependent behavior. The metallurgical findings from the selected tests are the focus of the Part II paper. The association of performance with notch geometry, weld strength, and other potential contributing factors will be highlighted with a primary objective of informing the reader of the variability, and heat-specific behavior that is observed among this class of alloys widely used in modern thermal fleet components and systems.

According to ASME Case N-888-3, Similar and Dissimilar Metal Welding Using Ambient Temperature SMAW or Machine GTAW Temper Bead Technique, a 48 hr waiting period before conducting the final nondestructive examination (NDE) is required when ferritic filler weld metal is used. The purpose of the 48 hr hold is to confirm the absence of hydrogen-induced cracking in the temper bead heat-affected zone. In previous research, the effect of post-weld heat treatment (PWHT) and temper bead welding (TBW) on the hydrogen-induced cracking (HIC) susceptibility in the coarse-grained heat-affected zone (CGHAZ) in welds of SA-508, P-No. 3 Group 3, pressure vessel steel was investigated using the Delayed Hydrogen Cracking Test (DHCT). In that previous study, the Gleeble thermomechanical simulator was used to generate six CGHAZ microstructural conditions: as-welded (AW), PWHT, and AW with single a TBW reheat at 675, 700, 725, and 735°C. Hydrogen was introduced to the specimen through cathodic charging under in situ constant tensile stress. The HIC susceptibility for these microstructures was ranked by the DHCT at a diffusible hydrogen level significantly exceeding typical GTAW and SMAW processes. The work described in this paper investigates the susceptibility to HIC of these same CGHAZ microstructures with DHCT at variable current densities, further ranking each condition. Test results were analyzed by fracture surface examination of failed tests, and cross-section microstructural analysis under a scanning electron microscope (SEM). Future steps include evaluating critical hydrogen content levels using gas chromatography for each condition. The results from this study will be used to consider potential elimination of the NDE hold time requirement in Case N-888-3 when ferritic weld metal is used.

The time-dependent behavior of 9Cr creep strength enhanced ferritic (CSEF) steels has long fixated on the creep life recorded in uniaxial constant load creep tests. This focus is a consequence of the need to develop stress allowable values for use in the design by formulae approach of rules for new construction. The use of these simple rules is justified in part by the assumption that the alloys used will invariably demonstrate high creep ductility. There appears to be little awareness regarding the implication(s) that creep ductility has on structural performance when mechanical or metallurgical notches (e.g., welds) are present in the component design or fabricated component. This reduced awareness regarding the role of ductility is largely because low alloy CrMo steels used for very many years typically were creep ductile. This paper focuses on the structural response from selected tests that have been commissioned or executed by EPRI over the last decade. The results of these tests demonstrate unambiguously the importance that creep ductility has on long-term, time-dependent behavior. This is the second part of a two-part paper; Part I reviewed the selected tests and discussed them from a mechanical perspective. The association of performance with specific microstructural features is briefly reviewed in this paper and the remaining gaps are highlighted for consideration among the international community.

Creep strength enhanced ferritic (CSEF) steels have shown the potential for creep failure in the weld metal, heat affected zone (HAZ) or fusion line. Details for this behavior have been frequently linked to metallurgical risk factors present in each of these locations which may drive the evolution of damage and subsequent failure. This work is focused on three weld samples fabricated from a commercially sourced Grade 92 steel pipe section. These weld samples were extracted from the same welded section but were reported to exhibit failure in different time frames and failure locations (i.e., HAZ of parent, fusion-line, and weld metal). The only variables that contribute to this observed behavior are the post weld heat treatment (PWHT) cycle and the applied stress (all tests performed at 650 °C). In this work detailed microstructural analysis was undertaken to precisely define the locations of creep damage accumulation and relate them to microstructural features. As part of this an automated inclusion mapping process was developed to quantify the characteristics of the BN particles and other inclusions in the parent material of the samples. It was found that BN particles were only found in the sample that had been subjected to the subcritical PWHT, not those that had received a re-normalizing heat treatment. Such micron sized inclusions are a known potential nucleation site for creep cavities, and this is consistent with the observed failure location in the HAZ of the parent in the sample where these were present. In the absence of BN inclusions, the next most susceptible region to creep cavitation is the weld metal. This has an intrinsically high density of sub-micron sized spherical weld inclusions and this is where most of the creep damage was located, in all the renormalized samples.

Extensive research and development has been undertaken in the UK on MarBN steels. These were first proposed by Professor Fujio Abe from NIMS in Japan. Within the UK, progress has been made towards commercialisation of MarBN-type steel through a series of Government funded industrial collaborative projects (IMPACT, IMPEL, INMAP and IMPULSE). As part of the IMPACT project, which was led by Uniper Technologies, boiler tubes were manufactured from the MarBN steel developed within the project, IBN1, and installed on the reheater drums of Units 2 and 3 of Ratcliffe-on-Soar Power Station. The trial tubes were constructed with small sections of Grade 91 tubing on either side of the IBN1 to allow direct comparison after the service exposure. This is the world’s first use of a MarBN steel on a full-scale operational power plant. In September 2018 the first tube was removed having accumulated 11,727 hours operation and 397 starts. This paper reports microstructural and oxidation analysis, that has been undertaken by Loughborough University as part of IMPULSE project, and outlines future work to be carried out.

A newly developed ferritic heat-resistant steel; 9Cr-3W-3Co-Nd-B steel has higher creep rupture strength both in the base metal and welded joints than the conventional high-Cr ferritic heat-resistant steels. The creep rupture strengths of 9Cr-3W-3Co-Nd-B steel welded joints were below the lower limit of the base metal in long-term creep stage more than 20,000 hours. The creep rupture position was heat-affected zone (HAZ) from 1.0 to 1.5 mm apart from the fusion line on the welded joint specimen ruptured at 34,966 hours. The equiaxed subgrains and coarsened precipitates were observed in HAZ of the ruptured specimen. In order to clarify the creep fracture mechanism of the welded joints, the microstructures of HAZ were simulated by heat cycle of weld, then observed by EBSD analysis. Fine austenite grains formed along the prior austenite grain boundaries in the material heated just above A C3 transformation temperature, however there were no fine grains such as conventional steel welded joints. The prior austenite grain boundaries were unclear in the material heated at 1050 °C. The creep rupture life of the material heated at just above A C3 transformation temperature exceeded the lower limit of base metal and there was no remarkable degradation, although it was shorter than the other simulated materials. It is, therefore, concluded that the creep fracture of 9Cr-3W-3Co-Nd-B steel welded joint in long-term stage occurred at HAZ heated at from just above A C3 transformation temperature to 1050 °C. It is speculated that the fine austenite grains formed along the prior austenite grain boundaries and inhomogeneous microstructures cause the coarsening precipitates and recovery of lath structure during long-term creep deformation.

A new ferritic steel branded as Thor 115 has been developed to enhance high-temperature resistance. The steel design combines an improved oxidation resistance with long-term microstructural stability. The new alloy was extensively tested to assess the high-temperature time- dependent mechanical behavior (creep). The main strengthening mechanism is precipitation hardening by finely dispersed carbide (M 23 C 6 ) and nitride phases (MX). Information on the evolution of secondary phases and time-temperature-precipitation behavior of the alloy, essential to ensure long-term stability, was obtained by scanning transmission electron microscopy with energy dispersive spectroscopy, and by X-ray powder diffraction on specimens aged up to 50,000 hours. The material behavior was also tested in service conditions, to validate the laboratory results: Thor 115 tubing was installed in a HRSG power plant, directly exposed to turbine flue gasses. Tubing samples were progressively extracted, analyzed and compared with laboratory specimens in similar condition. This research shows the performance of Thor 115 regarding steam oxidation and microstructure evolution up to 25,000 exposure hours in the field. So far, no oxide microstructure difference is found between the laboratory and on field tubing: in both cases, the oxide structure is magnetite/hematite and Cr-spinel layers and the oxide thickness values lay within the same scatter band. The evolution of precipitates in the new alloy confirms the retention of the strengthening by secondary phases, even after long-term exposure at high temperature. The deleterious conversion of nitrides into Z phase is shown to be in line with, or even slower than that of the comparable ASME grade 91 steel.

Inconel 740H is one of the most promising candidate Ni-base superalloys for the main steam pipe of 700 °C advanced ultra-supercritical (A-USC) coal-fired power plants. After processing and welding in manufacturing plant in solution-annealed state, large components was commonly suggested to have an extra aging treatment at 800 °C for 16 h, in order to obtain homogeneous γ′ precipitates. In this present work, creep tests and microstructure analyses were conducted on Inconel 740H pipe specimens under two different heat treatments to verify the necessity of aging process. Here we show that aging treatment has limited effect on the creep rupture life of Inconel 740H pipe. Both in grain interiors and along grain boundaries, crept specimens under two different heat treatments have the same precipitates. But the shape and distribution of γ′ in solution annealed sample is not as regular as the aged ones. Our results provide the underlying insight that aging treatment is not so necessary for the straight pipes if the on-site condition was hard to control. But for both groups of specimens, a small amount of h particles and some banded like M 23 C 6 were emerged during creep, which would be harmful to mechanical properties for the long run.

Welded joints of Ni-base alloys are often the critical part of components operated under high temperature service conditions. Especially welds in thick-walled structures are susceptible to various crack phenomena. Creep rupture and deformation behavior of different similar welds of Alloy 617B, both circumferential and longitudinal, were determined in many research German projects with the aim to qualify the nickel alloys and its welded joints for the use in highly efficient Advanced Ultra Supercritical (AUSC) power plants. Damage mechanisms and failure behavior have also been investigated within these projects. In order to reduce the welding residual stresses in thick-walled components a post weld heat treatment (PWHT) for Alloy 617B is recommended after welding. This PHWT reduces not only residual stresses but causes changes in the damage mechanisms and failure behavior of welded joints of Alloy 617B. Improving effects of PWHT have been investigated in this study and results of microstructural investigations were correlated with the material behavior.

Components such as tubes, pipes and headers used in power generation plants are operated in a creep regime and have a finite life. During partial replacement, creep exhausted materials are often welded to virgin materials with superior properties. The aim of this study was to identify a suitable weld filler material to join creep aged X20CrMoV12-1 to a virgin P91 (X10CrMoVNbV9-1) steel. Two dissimilar joints were welded using the gas tungsten arc welding (GTAW) process for the root passes, and manual metal arc (MMA) welding for filling and capping. The X20 and the P91 fillers were selected for joining the pipes. The samples were further heat treated at 755°C to stress relief the samples. Microstructural evolution and mechanical properties of the weld metals were evaluated. The average hardness of X20 weld metal (264 HV10) was higher than the hardness measurement of P91 weld metal (206 HV10). The difference in hardness was attributed to the high carbon content in X20 material. The characterisation results revealed that the use of either X20 or P91 weld filler for a butt weld of creep aged X20 and virgin P91 pipes material does not have a distinct effect on the creep life and creep crack propagation mechanism. Both weld fillers (X20 and P91) are deemed to be suitable because limited interdiffusion (<10 μm) of chromium and carbon at the dissimilar weld interface was observed across the fusion line. The presence of a carbon ‘denuded’ zone was limited to <10 μm in width, based on the results from local measurements of the precipitate phase fractions using image analysis and from elemental analysis using EDS. However the nanoindentation hardness measurements across the fusion line could not detect any ‘soft’ zone at the dissimilar weld interface. The effect of the minute denuded zone was also not evident when the samples were subjected to nanoindentation hardness testing, tensile mechanical testing, Small Punch Creep Test (SPCT) and cross weld uniaxial creep testing.

Dissimilar metal welds between T91 ferritic steels and TP347H austenitic alloys are commonly used in fossil power plants in China. Premature failure of such dissimilar welds can occur, resulting in unplanned plant outages that can cause huge economic losses. In this article, microstructural evolution of T91/TP347H dissimilar welds after different service conditions were studied, mechanical properties before and after service were also analyzed, a full investigation into the failure cause was carried out. The results show, the dissimilar metal welds in the as-welded condition consists of a sharp chemical concentration gradient across the fusion line, failure is attributed to the steep microstructural and mechanical properties gradients, formation of interfacial carbides that promote creep cavity formation.

A new method of creep life assessment was developed to consider heat-to-heat variations of welded joints of materials used in power plants. This paper explains a scheme of the assessment method and also describes an actual implementation of the method for Grade 91 steel. In the method, creep properties of the welded joints are related to those of each base metal because the heat-to-heat variations of welded joints strongly depend on the creep properties of the corresponding base metals. To estimate the creep properties of each base metal of the target pipe, microstructure analyses and small punch creep tests were conducted using small samples cut from the base metals in service, and evaluations were done on the basis of material data base obtained using standard test samples of long-term service exposed pipes. It is expected that the precision of the life assessment of pipes will be significantly improved using the developed method because it can consider the heat-to-heat variations of their materials, which are not considered in existing life assessment methods.

The present study presents a detailed investigation on the evolution of the microstructure during welding on virgin and long-term service exposed (creep aged 1 = 535°C; 16.1 MPa; 156 kh and creep aged 2 = 555°C; 17.0 MPa; 130 kh) 12% Cr (X20CrMoV11-1) martensitic steel. This study was carried out in order to understand the impact of welding on prior creep exposed Tempered martensite ferritic (TMF) steel and to explain the preferential failure of weldments in the fine grained heat affected zone (FGHAZ) of the creep aged material side instead of the new material side. Gleeble simulation (Tp = 980°C; heating rate = 200 °C/s; holding time = 4 seconds) of the FGHAZ was performed on the materials to create homogeneous microstructures for the investigation. Quantitative microstructural investigations were conducted on the parent plate and simulated FGHAZ materials using advanced electron microscopy to quantify: a) voids, b) dislocation density, c) sub-grains, and d) precipitates (M 23 C 6 , MX, Laves, Z-phase) in the materials. Semi-automated image analysis was performed using the image analysis software MIPARTM. The pre-existing creep voids in the creep aged parent material and the large M 23 C 6 carbides (Ø > 300 nm) in the FGHAZ after welding are proposed as the main microstructural contributions that could accelerate Type IV failure on the creep aged side of TMF steel weldments.

The creep behavior of a γ-TiAl based alloy at 1073 K was investigated, examining three different microstructures: equiaxed γ (Eγ), γ/γ fully lamellar (FLγ), and equiaxed γ with α 2 phase on grain boundaries (Eγα 2 ). The aim was to understand the influence of lamellar interfaces and grain boundary α 2 phase on creep behavior. Initially, creep rates were consistent across all specimens upon loading. However, Eγ exhibited a gradual decrease in creep rate compared to Eγα 2 and FLγ. Notably, the minimum creep rate of Eγ was one order of magnitude lower than that of Eγα 2 and FLγ. Conversely, Eγα 2 and FLγ displayed a slight acceleration and the longest rupture strain, albeit with the shortest rupture time compared to Eγ. Upon microstructural analysis of of the creep-test specimens, it was observed that numerous dynamic recrystallized grains (DXGs) and sub-grains formed along grain boundaries and interiors in Eγ, whereas they were limited to the region along grain boundaries in FLγ. In contrast, very few DXGs were formed in Eγα 2 . These findings indicate that γ/γ interfaces inhibit the extension of DXGs into grain interiors, suggesting that the grain boundary α 2 phase effectively suppresses the formation of DXGs.

Effects of microstructure constituents of α 2 -Ti 3 Al/γ-TiAl lamellae, β-Ti grains and γ grains, with various volume fractions on room-temperature ductility of γ-TiAl based alloys have been studied. The ductility of the alloys containing β phase of about 20% in volume increases to more than 1% as the volume fraction of γ phase increases to 80%. However, γ single phase alloys show very limited ductility of less than 0.2%. Microstructure analysis have revealed that intragranular fracture along γ/γ grain boundary occurred in γ single phase alloy whereas it does not along β/γ interphase in alloys containing β phase. In addition, local strain accumulations along β/γ interphase have been confirmed. The present results, thus, confirmed the significant contribution of β phase, especially the existence of β/γ interphase to enhancement of the room-temperature ductility in multicomponent TiAl alloys.

Oxide scale formation in the inner bore of steam tubing has been identified as a key metric for determining operational parameters and life expectancy of modern boiler systems. Grade 91 tubing is commonly used for the construction of key components within boiler systems designed for power generation operating in the temperature range of 500 to 650 °C. Standard laboratory test procedures involve grinding the surface of test coupons to homogenise their surface structure and improve experimental consistency, however, data presented here shows a discrepancy between laboratory and industrial practices that has long term implications on scale growth kinetics and morphological development. Microstructural analysis of both virgin and ex-service tubing reveals the presence of a pre-existing oxide structure that is incorporated into the inwardly growing scale and is implicated in the formation of multiple laminar void networks. These void networks influence thermal diffusivity across the scale and may function as regions of spallation initiation and propagation.

A new martensitic steel for power generation applications was developed: Tenaris High Oxidation Resistance (Thor) is an evolution of the popular ASTM grade 91, offering improved steam oxidation resistance and better long-term microstructural stability, with equal or better creep strength. Thanks to its design philosophy, based on consolidated metallurgical knowledge of microstructural evolution mechanisms, and an extensive development performed in the last decade, Thor was engineered to overcome limitations in the use of ASTM grade 91, above 600 °C, particularly related to scale growth and liftoff. After laboratory development, Thor was successfully validated at the industrial level. Several heats up to 80 metric tons were cast at the steel shop, hot rolled to tubes of various dimensions, and heat treated. Trial heats underwent extensive characterization, including deep microstructural examination, mechanical testing in the as-received condition and after ageing, long-term creep and steam oxidation testing. This paper presents an overview of metallurgical characterization performed on laboratory and industrial Thor material, including microstructural examination and mechanical testing in time-independent and time-dependent regimes. Data relevant to the behavior and the performance of Thor steel are also included.