You are on page 1of 1522

Advances in Materials

Technology
for Fossil Power Plants
Proceedings from the Seventh International Conference
October 2225, 2013
Waikoloa, Hawaii, USA
Editors
D. Gandy
J. Shingledecker
Sponsored By

EPRI Report Number 3002002375

Published By
ASM International
Materials Park, Ohio 44073-0002
www.asminternational.org

Copyright 2014
by
Electric Power Research Institute, Inc.
All rights reserved
No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form
or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the
written permission of the copyright owner.
First printing, January 2014
Great care is taken in the compilation and production of this book, but it should be made clear
that NO WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION,
WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE,
ARE GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is
believed to be accurate by the Electric Power Research Institute, Inc. (EPRI), EPRI cannot
guarantee that favorable results will be obtained from the use of this publication alone. This
publication is intended for use by persons having technical skill, at their sole discretion and risk.
Since the conditions of product or material use are outside of EPRI's control, EPRI assumes no
liability or obligation in connection with any use of this information. No claim of any kind,
whether as to products or information in this publication, and whether or not based on
negligence, shall be greater in amount than the purchase price of this product or publication in
respect of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE
EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER
PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES
WHETHER OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH
PARTY. As with any material, evaluation of the material under end-use conditions prior to
specification is essential. Therefore, specific testing under actual conditions is recommended.
Nothing contained in this book shall be construed as a grant of any right of manufacture, sale,
use, or reproduction, in connection with any method, process, apparatus, product, composition,
or system, whether or not covered by letters, patent, copyright, or trademark, and nothing
contained in this book shall be construed as a defense against any alleged infringement of letters,
patent, copyright, or trademark, or as a defense against liability for such infringement.
Comments, criticisms, and suggestions are invited, and should be forwarded to EPRI.
ISBN-13: 978-1-62708-060-6
ISBN-10: 1-62708-060-0
SAN: 204-7586
Distributed by ASM International
Materials Park, OH 44073-0002
www.asminternational.org
Printed in the United States of America

Conference Organizing Committee


John Shingledecker
EPRI, United States

Elizabeth Owensby
EPRI, United States

David Gandy
EPRI, United States

Jonathan Parker
EPRI, United States

Jeff Breznak
GE Power & Water, United States

Bruce Pint
Oak Ridge National Laboratory,
United States

Regis Conrad
U.S. Department of Energy, United
States

Dan Purdy
EPRI, United States

Horst Hack
Foster Wheeler, United States

Patricia Rawls
NETL, United States

Jeff Hawk
NETL, United States

Deepak Saha
GE Power & Water, United States

Paul Jablonski
NETL, United States

John Siefert
EPRI, United States

John Marion
Alstom Power, United States

Jim Tanzosh
Babcock & Wilcox Co., United
States

Phil Maziasz
Oak Ridge National Laboratory,
United States

Brian Vitalis
Babcock Power, United States

iii

International Advisory Board


Fusheng Lin
SPERI, China

Fujio Abe
National Institute for Materials
Science, Japan
Giuseppe Cumino
Tenaris-Dalmine, Italy

Fujimitsu Masuyama
Kyushu Institute of Technology,
Japan

Augusto Di Gianfrancesco
Centro Sviluppo Materiali S.p.A, Italy

Alok Mathur
BHEL, India

Jude Foulds
Clarus Consulting, LLC., USA

Peter Mayr
Chemnitz University of Technology,
Germany

Tony Fry
National Physical Laboratory,
United Kingdom

John Oakey
Cranfield University, United
Kingdom

Gerhard Fuchs
University of Florida, USA

Thobeka Pete
ESKOM, South Africa

Masafumi Fukuda
Research Institute for Advanced
Thermal Power Systems, Japan

Ashok Saxena
University of Arkansas, USA

Herbert Heuser
Bhler, Germany

Arthur Stam
DNV Kema, Netherlands

Staf Huysmans
Laborelec GDF Suez, Belgium

Swami Swaminathan
Turbo-Met, USA

Kazuhiro Kimura
National Institute for Materials
Science, Japan

Yasuhiko Tanaka
Japan Steel Works, Japan
Yukio Takahashi
CRIEPI, Japan

Andreas Klenk
MPA Stuttgart, Germany

iv

Rachel Thomson
University of Loughborough, United
Kingdom

Ian Wright
Wright H.T., USA
Xishan Xie
University of Science and Technology
Beijing, China

Christian Ulrich
VGB, Germany

Contents
Preface .................................................................................................................................xix
Section 1: Technology and Program Overviews (Plenary Session)
Progress of China 700C USC Development Program ......................................................... 1
Rui Sun, Zhanzhong Cui, and Ye Tao
ENCIO Project: An European Approach to 700C Power Plant .......................................... 9
A. Di Gianfrancesco, A. Tizzanini, M. Jedamzik, and C. Stolzenberger
Advanced USC Technology Development in Japan ............................................................ 24
Masafumi Fukuda, Eiji Saito, Hiroyuki Semba, Jun Iwasaki, Sakae Izumi, Shinichi
Takano, Takeo Takahashi, and Yasuo Sumiyoshi
Current Status of the U.S. DOE/OCDO A-USC Materials Technology
Research and Development Program .................................................................................. 41
J. Shingledecker, R. Purgert, and P. Rawls
India's National A-USC Mission - Plan and Progress ........................................................ 53
Alok Mathur, O.P Bhutani, T. Jayakumar, D.K. Dubey, and S.C. Chetal
Advantages of A-USC for CO2 Capture in Pulverized Coal Units ...................................... 60
H.L. Hendrix
Section 2: Nickel-Based Alloys for Advanced Ultrasupercritical Power Plants
Nextgenpower Demonstration and Component Fabrication of Nickel Alloys and
Protective Coatings for Steam Temperatures of 750C ...................................................... 74
Arthur F. Stam
A Steam Generator for 700C to 760C Advanced Ultra-Supercritical Design and
Plant Arrangement: What Stays the Same and What Needs to Change .............................. 86
Paul S. Weitzel
An Investigation on Structure Stability of Advanced Austenitic
Heat-Resisting Steels and Ni-Base Superalloys for
600700 C A-USC Power Plant Application ...................................................................... 98
Xishan Xie, Chengyu Chi, Qiuying Yu, Zhihao Yao, Maichang Zhang,
Yaohe Hu, Jianxin Dong, Hongyao Yu, Shuangqun Zhao, Fusheng Lin, Xia Liu,
Linbo Mei, Huachun Yang, and Mingyang Li

vii

Recent Developments in the Characteristics of Haynes 282 Alloy for


Use in A-USC Applications ............................................................................................... 120
S.K. Srivastava, J.L. Caron, and L.M. Pike
Creep-Rupture Behavior of Precipitation-Strengthened Ni-Based Alloys under
Advanced Ultrasupercritical Steam Conditions ................................................................ 131
P.F. Tortorelli, K.A. Unocic, H. Wang, M.L. Santella, and J.P Shingledecker
Microstructural Evolution in Cast Haynes 282 for
Application in Advanced Power Plants ............................................................................. 143
Y. Yang, R.C. Thomson, R.M. Leese, and S. Roberts
Investigations on Nickel Based Alloys and Welds for A-USC Applications ...................... 155
Andreas Klenk and Karl Maile
Qualifications and Experience on A 617 and C263 Boiler Tubes ..................................... 167
Patrik Schraven
Material Advancement Used for 700C A-USC-PP in China ........................................... 171
Zhengdong Liu, Hansheng Bao, Gang Yang, Songqian Xu,
Qijiang Wang, and Yujun Yang
Selection of Super Alloys for A-USC Power Plants under
Consideration of Creep Crack Growth Behavior .............................................................. 180
Falk Mueller, Alfred Scholz, and Matthias Oechsner
Fabrication Trials of Ni-Based Alloys for Advanced USC Boiler Application ................. 190
Nobuhiko Saito, Nobuyoshi Komai, and Yasuhiro Takei
Development and Trial Manufacturing of Ni-Base Alloys for
Coal Fired Power Plant with Temperature Capability 800C .......................................... 202
Shinya Imano, Naoya Sato, and Hironori Kamoshida
Manufacturing Demonstration of Inconel Alloy 740H for A-USC Boilers ................. 215
B.A. Baker, R.D. Gollihue, J.J. deBarbadillo, S.J. Patel, and D. Maitra
Creep-Rupture Performance of Inconel Alloy 740 and Welds .......................................... 230
J.P. Shingledecker
Evaluation of High Temperature Strength of a Ni-Base Alloy 740H for
Advanced Ultra-Supercritical Power Plant ...................................................................... 242
Shengde Zhang and Yukio Takahashi
Cyclic Properties of 50Ni-24Cr-20Co-0.6Mo-1Al-1.6Ti-2Nb Alloy at
Advanced USC Steam Temperature .................................................................................. 254
U.V. Gururaja, A. Patra, P. Mukhopadhyay, and M. Narayana Rao

viii

Microstructure Evolution and Precipitates Stability in


Inconel Alloy 740H during Creep ...................................................................................... 265
Shuangqun Zhao, Fusheng Lin, Rui, Chengyu Chi, and Xishan Xie
Discussion of Delivered Condition Specified in ASME Code Case 2702 on
Inconel 740H Used for A-USC Boiler ............................................................................... 276
Hua-Chun Yang, Lin-Sen Wang, Dong-Ping Wang, Shi-Yong Mao,
Xiao-Chuan Yang, Xian-Hong Lai, Jin-Bing Yang, and Wen Yang
Section 3: Materials for Turbines
Gas and Steam Turbine Forgings for High Efficient Fossil Power Plants ....................... 281
Guenter Zeiler and Albert Putschoegl
Microstructural Change after Long-Term Creep Exposure in
High Cr Steel Forgings for Ultrasupercritical Steam Turbine Rotors .............................. 293
Masato Mikami, Kota Sawada, Satoru Kobayashi, Toru Hara, and Kazuhiro Kimura
Experience in Manufacture of High Chromium Forged Rotor Steels ............................... 304
A. Di Gianfrancesco, S. Budano, P. Lombardi, M. Paura, S. Neri,
M. Calderini, and N. Longari
Manufacturing of Trial Rotor Forging of 9%Cr Steel Containing Co and B
(X13CrMoCoVNbNB9-2-1) for Ultrasupercritical Steam Turbines .................................. 321
M. Nakano, K. Kawano, and M. Mikami
Development and Production of Monoblock Low-Pressure
Turbine Rotor Shaft Made from 670 Ton Ingot ................................................................. 333
Takafumi Yamauchi, Hidenao Kudo, Yasuhiro Kishi Sou Ueda, Hajime Yoshida,
Kimitoshi Kimura, Koji Kajikawa, and Shigeru Suzuki
High Cycle Fatigue Properties of Steam Turbine Materials at
High Temperature under Superheated Steam Conditions ................................................. 344
Jan Dzugan, Tom Mek, and Josef Jurenka
Creep and Creep-Fatigue Crack Growth Behaviors of
30Cr1Mo1V Rotor Steel after Long Term Service ............................................................. 351
Kexian Shi, Fusheng Lin, Shuangqun Zhao, and Yanfeng Wang
The Effect of Water Vapor Content and CO2 on TBC Lifetime ......................................... 360
B.A. Pint, K.A. Unocic, and J.A. Haynes

ix

Formation of Diffusion Zones in Coated Ni-Al-X Ternary Alloys and


Ni-Based Superalloys ........................................................................................................ 371
A.S. Suzuki, G.D. West, and R.C. Thomson
High-Temperature Solid Particle Erosion Testing Standard for
Advanced Power Plant Materials and Coatings ............................................................... 382
V.P. "Swami" Swaminathan, Jeffery S. Smith, and David W. Gandy
Metrology to Enable High Temperature Erosion Testing
A New European Initiative ................................................................................................ 400
A.T. Fry, M.G. Gee, S. Clausen, U. Neuschaefer-Rube, M. Bartscher, D. Spaltmann,
M. Woydt, S. Radek, F. Cernuschi, J.R. Nicholls, and T.W. Rose
NiCoCrAlYHf Coating Evolution through Multiple Refurbishment Processing on a
Single Crystal Nickel Superalloy ....................................................................................... 412
A. Rowe, M. Karunaratne, and R.C. Thomson
Microstructural Evolution in a Ni- Based Superalloy for
Power Plant Applications as a Consequence of
High Temperature Degradation and Rejuvenation Heat Treatments ............................... 424
Z. Yao, M.A.E. Jepson, R.C. Thomson, and C.C. Degnan
Development and Evaluation of Large-Scale Rotor Forging for
Over 700 C-Class A-USC Steam Turbine ........................................................................ 436
Shigekazu Miyashita, Yomei Yoshioka, and Takahiro Kubo
Rotor Forgings for Steam Turbines with High Efficiency ................................................. 448
N. Blaes, B. Donth, A. Diwo, and D. Bokelmann
Similar and Dissimilar Welding of Nickel-Based Superalloys for
A-USC Steam Turbine Rotors in Nextgenpower Project ................................................... 459
Stanislav Hreben, Petr Vitek, Vaclav Zednik, and Lubos Prchlik
Development and Trial Manufacturing of Ni-Based Superalloy LTES700R for
Advanced 700C Class Steam Turbines .............................................................................. 468
Ryuichi Yamamoto, Yoshikuni Kadoya, Shin Nishimoto, Yoshinori Tanaka,
Takuma Okajima, Kouichi Ishikawa, and Kouichi Uno
Tribolayer Formation by Strain-Induced Transformations in Hardfacing Alloys ............ 482
Ryan Smith, J. Siefert, D. Gandy, and S.S. Babu
Advances in Materials Technology for Fossil Power Plants ............................................. 491
S. Birks, S. Roberts, and R. Leese

Trial Production of Alloy 625 and Alloy 617 Casting Component for
Advanced 700C Class Steam Turbines ............................................................................ 504
Yuichi Hirakawa, Yoshikuni Kadoya, Shin Nishimoto, and Yoshinori Tanaka

Section 4: Alloys T23/T24


Material Behavior of T23 and T24 .................................................................................... 513
Eric Dupont
Supercritical Unit Experience with Grade T23 Evaporator Tube Failures ...................... 525
Masafumi Fukuda, Eiji Saito, Yoshinori Tanaka, Takeo Takahashi,
Shinji Nakamura, Jun Iwasaki, Shinichi Takano, and Sakae Izumi
Research on Water Wall Tubes and Welded Joints of
1000MW USC Tower Boiler .............................................................................................. 537
B.C. Wang, Xu Xueyuan, Zhu Yu Feng, Qi Anfang, Lu Zhengran, Zhang Bo,
Jin Yongqiang, and Lu Xiaoli
T/P24 (7CRMOVTIB10-10) a Bainitic-Martensitic Steel Grade for
Super Heater and Water Wall Applications in
Modern Ultra Super Critical Power Plants: Lessons Learned ......................................... 549
M. Jarrar, G. Gevelmann, H. Heuser, and K.G. Tak
Potential Effects of HAZ Hardness on Use of
T24 Tubing for Waterwall Applications ............................................................................ 565
P. Moody and P. Barnard
Phase Transformations and Microstructure in Gas Tungsten Arc Welds of
Grade 23 Steel Tubing ....................................................................................................... 573
B.T. Alexandrov, J.M. Steiner, K.C. Strader, X. Feng, J.C. Lippold,
J.E. Dierksheide, and J.M. Tanzosh

Section 5: Grade 91 and 92


Misorientation Change Caused by the Precipitation Strengthening through
Several MX Type Precipitates in High Cr Ferritic Creep Resistant Steels ....................... 586
Ryosuke Yamagata and Yasushi Hasegawa
Investigation of Creep Damage and Cavitation Mechanisms in P92 Steels ..................... 596
Y. Gu, G.D. West, R.C. Thomson, and J. Parker
Effect of Stress State on Microstructural Change during Creep in
Grade 92 Steel Welded Joint ............................................................................................. 607
Kota Sawada, Kazuhiro Kimura, and Masaaki Tabuchi

xi

The Effect of Post Weld Heat Treatment on the Creep Behaviour and
Microstructural Evolution in Grade 92 Steel Welds for
Steam Pipe Applications .................................................................................................... 615
X. Xu, G.D. West, R.C. Thomson, and J. Parker
Effects of Cr and W Content in High Cr Ferritic Heat-Resistant Steels on
Long-Term Creep Rupture Strength .................................................................................. 627
Yusuke Mito, Kazuhiro Miki, Tsukasa Azuma, and Tohru Ishiguro
Heat-to-Heat Variation in Creep Rupture Ductility of ASME Gr.91 Steels in the
Long-Term-Investigation into Recovery of Microstructure and Void Formation ............. 637
Satoru Kobayashi, Kota Sawada, Toru Hara, Hideaki Kushima, and
Kazuhiro Kimura
Experiences in the Production and Welding of 9-12 % Cr Steels Explained on
Grades 92 and VM12-SHC for USC Power Plants ........................................................... 648
Vida Knezevic, Marko Subanovic, Olivier Hamart, H. Heuser, and K.G. Tak
Creep-Fatigue Interaction in Grade 92 Steel and Its Predictability ................................. 667
Yukio Takahashi, Jonathan Parker, and David Gandy
Creep-Fatigue Properties of Grade 91 Steel ..................................................................... 679
Rami Pohja, Asta Nurmela, Pekka Moilanen, and Stefan Holmstrm
Evaluation of Remaining Life of Gr.91 Welded Elbow Served at
USC Plant for Long-Term ................................................................................................. 690
Seiji Nagai, Masatsugu Yaguchi, Shan Lin, and Souichi Yamazaki
Fracture of Gr. 91 Steel Longitudinal Welded Pipe under
Internal Pressure Creep Condition ................................................................................... 702
Masatsugu Yaguchi, Takayuki Sakai, Takashi Ogata, and Takuaki Matsumura
Creep Cavitation in CSEF Steels ...................................................................................... 714
J. Parker
Evaluation of Long-Term Creep Rupture Life of
Strength Enhanced High Cr Ferritic Steel on the
Basis of Its Temperature Dependence ............................................................................... 732
K. Maruyama, J. Nakamura, and K. Yoshimi
Creep Damage Evaluation of High Cr Ferritic Steel Based on
Change in Hydrogen Desorption Characteristics ............................................................. 744
Shin-ichi Komazaki, Hayato Yamashita, Mitsuharu Yonemura, and Masaaki Igarashi

xii

Section 6: Oxidation and Corrosion


High Pressure Steam Oxidation: Extents and Influences .................................................. 753
S.R. Holdsworth, A.T. Fry, M. Seraffon, and J. Banks
Effect of Temperature, Alloy Composition and Surface Treatment on the
Steamside Oxidation / Oxide Exfoliation Behavior of
Candidate A-USC Boiler Materials ................................................................................... 765
J. M. Sarver and J. M. Tanzosh
Effect of Oxygen Content of Steam on the Steam Oxidation Behavior of
Boiler Tube Materials ........................................................................................................ 777
K. Nakagawa, Y. Matsunaga, T. Tanoue, T. Hayasaka, K. Matsumura, and
Y. Takagi
Evaluation of Supercritical Oxidation Resistance of Boiler Tube Materials .................... 791
Sanni Yli-Olli, Satu Tuurna, Sami Penttil, Pertti Auerkari,
Edgardo Coda Zabetta, and Kysti Vnsk
Effects of Alloy Composition and Surface Engineering on
Steam Oxidation Resistance .............................................................................................. 803
B.A. Pint, S. Dryepondt
Steam Oxidation of the Novel Austenitic Steel of
Fe-20Cr-30Ni-2Nb (at.%) at 973 K ................................................................................... 815
Mitsutoshi Ueda, Masakazu Yamashita, Kenichi Kawamura,
Masao Takeyama, and Toshio Maruyama
Managing Oxide Scale Exfoliation in Boilers with TP347H Superheater Tubes .............. 821
Adrian S. Sabau, Ian G. Wright, John P. Shingledecker, and Peter F. Tortorelli
Steam Loop Testing of A-USC Materials for Oxidation and
Fireside Corrosion Alstoms Experience to Date ........................................................... 832
Reddy Ganta, John Marion, Bruce Wilhelm, Jim Pschirer, Steve Goodstine, and
Charles Boohaker
Inconel Filler Metal 72M Provides Corrosion and Wear Resistance and
Low Delta T Through Walls of Tubing in Fossil-Fired Boilers .................................... 847
Samuel D. Kiser, Martin Caruso, Rengang Zhang, and Brian Gaal
Comparison of Coal-Ash Corrosion Resistance of Alloys Exposed to
Advanced Air-Coal and Oxy-Coal Combustion Environments ......................................... 863
Steven C. Kung

xiii

Fireside Corrosion and Carburization of Superheater Materials in


Simulated Oxyfuel Combustion Conditions ....................................................................... 881
Satu Tuurna, Pekka Pohjanne, Sanni Yli-Olli, Edgardo Coda, and Kysti Vnsk
Effect of Alloy Composition on Fireside Corrosion Rates in
Air- and Oxy-Fired Systems .............................................................................................. 892
B.A. Pint
Section 7: Welding and Weld Performance
Evaluation of Creep Rupture Strength in Ni-Based Alloy Weldments for an
Advanced USC Boiler ........................................................................................................ 903
Nobuyoshi Komai, Nobuhiko Saito, and Yasuhiro Takei
Study of Creep Damage in Creep Exposed Martensitic
High-Chromium Steel Weldments ..................................................................................... 914
C. Schlacher, C. Bal, C. Sommitsch, H. Toda, and P. Mayr
Creep Properties of Heat Affected Zone in Heterogeneous Welded Rotor ....................... 924
Jan Dzugan, Zbysek Novy, Pavel Konopik, Pavel Podany, and Eva Folkov
Flux Cored Wires for Welding Advanced 9-10% Cr Steels ............................................... 936
Susanne Baumgartner, Ronald Schnitzer, Monika Schuler,
Edina Schmidtne-Kelity, and Claus Lochbichler
Investigations on the Degradation Mechanism of Welded Joints of
Advanced 9%Cr-Mo-Co-B Steel Used for 620C USC Steam Turbine ............................. 948
Xiufang Gong, Zhenhuan Gao, Liping Nie, Bo Zhang, Lei Jiang,
Bangqiang Zhang, Gongxian Yang, and Congping Zhang
Creep Degradation and Life Assessment of High Temperature Welds ............................. 960
Fujimitsu Masuyama and Tomiko Yamaguchi
Qualifying Welding Procedures for Repair of
High Temperature Components ........................................................................................ 973
William F. Newell
Influence of Trace RE Element on Properties and Microstructures of
SA335P91 Weld Metal ....................................................................................................... 982
Haoyang Du, Mingliang Du, Jian Yang, Laichang Hou, and Yongmei Yang
Application of EPRI P87 in Dissimilar Austenitic-Martensitic Welded Joints of
Tempaloy AA-1 and T92 Steel Grades .............................................................................. 992
Michael Urzynicok, Robert Jachym, and Krzysztof Kwieciski

xiv

Verification of Long Term Creep Rupture Strength and


Component Fabricability of Candidate Ni-Based Materials for
A-USC Boilers ................................................................................................................. 1006
T. Tokairin, T. Sato, A. Shimada, and R. Nakagawa
Effect of Non-Standard Heat Treatments on Creep Performance of
Creep-Strength Enhanced Ferritic (CSEF) Steel Weldments .......................................... 1016
Yukinori Yamamoto, Michael L. Santella, Xinghua Yu, and
Sudarsanam Suresh Babu
Practical Guide to Welding Inconel Alloy 740H ....................................................... 1025
Ronald D. Gollihue, Brian A. Baker, Joseph E. Dierksheide, and Jim M. Tanzosh
Modeling a Stress Relaxation Cracking Test for
Advanced Ultra Supercritical Alloys ............................................................................... 1038
David C. Tung and John C. Lippold
Development of Welding and Fabrication Technologies in
Advanced USC Boiler ...................................................................................................... 1047
Keiji Kubushirio, Takaaki Matsuoka, Yoshitomo Ohkuma,
Hirokatsu Nakagawa, and Hiroshi Aoki
Section 8: New Alloy Concepts
Characterization of an Extruded Austenitic Stainless Steel for
Advanced Fossil Power Plant Applications .................................................................... 1059
D. Purdy, D. Gandy, and J. Shingledecker
Precipitation Strengthening by the Nitrides in High Cr Containing
Ferrtic Creep Resistant Steels ......................................................................................... 1071
Ippei Shinozaki and Yasushi Hasegawa
Development of High Chromium Ferritic Steels Strengthened by
Intermetallic Phases ........................................................................................................ 1081
B. Kuhn, M. Talik, J. Zurek, T. Beck, W.J. Quadakkers,
L.S. Singheiser, and H. Hattendorf
Long-Term Evolution of Microstructure in VM12-SHC .................................................. 1093
Andr Schneider, Marko Subanovi, and Javier Pirn Abelln
Precipitation Process of Z-Phase in 9-12%Cr Steels ...................................................... 1104
Hilmar Kjartansson Danielsen

xv

Effect of Copper Addition on the Toughness of


New Z-Phase Strengthened 12% Chromium Steels ......................................................... 1116
Fang Liu and Hans-Olof Andrn
Alloy Design of Tempered Martensitic 9Cr-Boron Steel for A-USC Boilers .................. 1127
F. Abe, M. Tabuchi, S. Tsukamoto, and Y. Liu
Section 9: Creep and General Topics
Role of Half Yield on Creep Life Prediction of
Creep Strength Enhanced Ferritic Steels ........................................................................ 1139
Kazuhiro Kimura
Applications of a Phase Analysis Technology to
Advanced Heat Resistant Steels and Nickel-Base Superalloys ........................................ 1151
Peng Zhifang, Peng Fangfang, and Chen Fangyu
Modelling and Optimizing Precipitation in
Creep Resistant Austenitic Steel 25Cr-20Ni-Nb-N .......................................................... 1163
Vujic Stojan, Beal Coline, Sommitsch Christof, Muhammad Farooq,
Sandstrm Rolf, and Zurek Joanna
The Practical Application of Small Scale Sampling and
Impression Creep Testing to Grade 91 Components ....................................................... 1173
S.J. Brett
Development of Boiler Risk Management and Life Prediction System ........................... 1182
Li Yaojun, Wang Dapeng, and Li Shuxue
Harmonizing of Creep-Fatigue Test Methods through
Development of ASTM Standards .................................................................................... 1190
Ashok Saxena, Valliappa Kalyanasundaram, Santosh B. Narasimhachary, and
Stuart R. Holdsworth
Defect Tolerant Design Concepts Applied to Remaining Life Assessments of
Steam Turbines and Weld Repairs of Power Generation Equipment .............................. 1206
Phillip Dowson and David Dowson
Microstructure and Properties of 12Cr2MoWVTiB Steel for
Membrane Walls .............................................................................................................. 1220
Krzysztof Cieszyski, Wadysaw Osuch, Maciej Kaczorowski,
Stanisaw Fudali, and Aleksandra Czyrska-Filemonowicz

xvi

Application of New GMAW Welding Methods Used in


Prefabrication of P92 (X10CrWMoVNb9-2) Pipe Butt Welds ........................................ 1232
Michael Urzynicok, Krzysztof Kwieciski, Jacek Sania,
Paola Mariani, and Marian Szubryt
An Extended Small Punch Test Method for
Providing Measured Displacements Across a Test Specimen ......................................... 1244
L.E. Crocker, A.T. Fry, and J. Banks
Non-Destructive and Optical Thickness Measurements of
Steam Grown Oxide on Contacting Surfaces of Power Plant ......................................... 1256
E.M. Piedra, M.J. Lodeiro, and A.T. Fry
IN718: Higher Temperature Application Range for an Old Superalloys ........................ 1268
A. Di Gianfrancesco, P. Lombardi, D. Venditti, R. Montani, and L. Foroni
Trial Production and Evaluation of 10-Ton Class A-USC Turbine Rotor of
Ni-Fe Base Superalloy FENIX-700 ................................................................................. 1283
Koichi Takasawa, Tatsuya Takahashi, Ryoji Tanaka, Terutaka Kure,
Shin-ya Imano, and Eiji Saito
The New Metallurgical Precipitation Strengthening Model of
W Containing Advanced High Cr Ferrtic Creep Resistant Steels ................................... 1292
Yasushi Hasegawa, Tomohiro Nishiura, and Tetsuya Sato
Evaluation of Reduction in Creep Strength Based on
Fracture Energy in CSEF Steels ...................................................................................... 1304
Y. Nagae and T. Asayama
The Cross-Weld Performance of 9%Cr Creep-Resistant Steels, and the
Influence of Welding Parameters .................................................................................... 1313
John Rothwell and Peter Mayr
A New Developed 9%Cr Steel with High Boron Content Achieving the
Long-Term Microstructural Stability and the
Optimized Mechanical Properties for 625C USC Power Plant Boiler .......................... 1329
Mitsuharu Yonemura, Yasushi Hasegawa, Masaaki Igarashi,
Masaaki Tabuchi, and Fujio Abe
Development and Evaluation of Fe/Ni Dissimilar-Weld Turbine Rotor for
Advanced USC ................................................................................................................. 1341
Shun Oinuma, Shigekazu Miyashita, Yoshihiro Fujita, Yomei Yoshioka, and
Takahiro Kubo

xvii

Creep of the Novel Austenitic Heat Resistant Steel of


Fe-20Cr-30Ni-2Nb under Steam Atmosphere at 1073 K ................................................ 1352
Yu Misosaku, Imanuel Tarigan, Naoki Takata, Mitsutoshi Ueda,
Toshio Maruyama, and Masao Takeyama
Evaluation of Long Term Creep Strength of 9Cr Heat Resistant
Ferritic Steel Containing Boron with the Aid of System Free Energy Concept .............. 1363
Yoshiki Shioda, Shohei Fujii, Yoshinori Murata, and Yasushi Hasegawa
Creep Crack Growth in T23 Weldments .......................................................................... 1372
J.A. Siefert and J.R. Foulds
Corrosion Characteristics of Alloy622 Weld Overlay for
Waterwall Tubes in Coal Fired Boilers ........................................................................... 1388
Masakazu Matsui, Hirotoshi Matsumoto, and Takeshi Miyazawa
Detection of Incipient Stress Corrosion Cracking Damage in
Primary Loop Piping Using Fiber Optic Strain Gages ................................................... 1397
Benjamin K. Jackson, David Bosko, Michael T. Cronin, and
Jonnathan L.W. Warwick
Creep Rupture Properties of Ni-Base Superalloy USC141 as
Solution Treated for 700C Class A-USC Boiler ............................................................ 1407
Toshihiro Uehara, Chuya Aoki, Takehiro Ohno, Patrik Schraven,
Hironori Kamoshida, and Shinya Imano
High Temperature Oxidation Behavior of Fe-9Cr Steel In CO2-O2 Gas Mixture ........... 1417
Shigenari Hayashi, Kietaro Kaya, and Shigeharu Ukai
Hot Corrosion Properties of Ni-Based Alloys Used in an Advanced-USC Boiler ........... 1422
Yasuhiro Tanaka, Nobuyoshi Komai, and Yasuhiro Takei
In-Situ Full Field Creep Deformation Study of Creep Resistant Materials Welds ......... 1432
Xinghua Yu, Zhili Feng, and Yukinori Yamamoto
A Computational Design Study of Novel Creep Resistant Steels for
Fossil Fuel Power ............................................................................................................ 1441
Qi Lu, Wei Xu, and Sybrand van der Zwaag

Section 10: Reference Information


Author Index ..................................................................................................................... 1453
Subject Index .................................................................................................................... 1457

xviii

Preface
There were 185 participants representing 18 countries traveling from around the world
to participate in the EPRI 7th International Conference on Advances in Materials
Technology for Fossil Power Plants on the Big Island of Hawaii, USA. This conference
built on the success of the previous 6 conferences originally started by Prof. R. (Vis)
Viswanathan, EPRI (FASM) in 1995 in London, England, and held every three years
since that time in San Sebastian (Spain), Swansea (Wales), Hilton Head Island (USA),
Marco Island (USA), and Santa Fe (USA). This proceeding contains the largest number
of papers ever collected for this conference. Overwhelming response to the 2013
conference (over a 20% increase in number of papers and technical talks) prompted the
organizers to hold the first ever poster session. EPRI and ASM have again partnered to
publish this proceeding which is organized into 9 topical areas: technology overviews,
nickel-based alloys for advanced ultrasupercritical power plants, materials for turbines,
alloys T23/24, Grades 91/92, oxidation and corrosion, welding and weld performance,
new alloys concepts, and creep and general topics.
The high level of interest in the topic of advanced materials for fossil power plants is
being primarily driven by the desire for higher-efficiency advanced ultrasupercritical
(A-USC) steam cycles. The conference brought together, for the first time,
representatives from all the national A-USC projects including the Europe, USA,
Japan, China, and India. Each program is in a different stage of development from
materials fabrication and property studies, operation of test loops and component test
facilities, and planning for a demonstration plant. China appears poised to build and
operate the worlds first coal-fired 700C+ A-USC steam cycle around the year 2020.
Materials, specifically nickel-based alloys and advanced steels and stainless steels, are
the key enabling technology to realize these high efficiency cycles. This proceeding
provides a wealth of information on these alloys as well as new understanding on
advanced steel alloys being used in todays fossil power plants. New alloy concepts
and fundamental understanding on degradation mechanisms (creep, fatigue, oxidation,
corrosion, etc.) were also highlighted in the conference program. Finally, innovative
research into alloys for combustion turbines, a growing area of interest (especially in
countries like the USA where low cost natural gas is changing the focus from coal to
gas), was presented.
The conference award dinner recognized the lifetime achievements of Prof. Fujimitsu
Masuyama, Kyushu Institute of Technology, for his contributions to implementation of
new alloys in fossil power generation. Prof. Masuyama joined past recipients of this
award: Rudolph Blum and Dr. Ian Wright. Attendees were given the opportunity to
rank the 20 posters in the poster session. The best poster award was granted to Shun
Oinuma, Toshiba Corporation, for his poster on Development and Evaluation of Fe/Ni
Dissimilar Metal Weld Turbine Rotor for Advanced USC.

xix

The 7th conference had broadest participation yet for such an event. The chairmen are
indebted to the EPRI conference support staff, members of EPRIs Fossil Materials and
Repair Program (P87), the organizing committee, the international advisory board, and
all the participants who made the conference a huge success.
John Shingledecker and David Gandy
EPRI (2013)

xx

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

PROGRESS OF CHINA 700 USC DEVELOPMENT PROGRAM


Rui SUN, Zhanzhong CUI, Ye TAO
Electric Power Planning & Engineering Institute, Beijing, China

ABSTRACT
This paper presents an overview of Chinas electric power development and the National 700
Ultra-Supercritical (USC) Coal-Fired Power Generation Technology Innovation Consortium.
Besides, the R&D plan and latest progress of China 700 USC coal-fired power generation
technology is also introduced in this paper.
INTRODUCTION
The Status quo of Chinas electric power industry
Chinas power industry has made remarkable achievement with the sustainable development. The
power generation technologies have been improved significantly and the installed capacity has
been growing (Fig. 1). By the end of 2012, the installed capacity of power generation has reached
1145 GW, among which, fossil-fuel power is 796 GW (69.5%), hydropower is 249 GW (21.7%),
nuclear power is 13 GW (1.1%), and wind & other renewable energy power is 64 GW (5.6%),
respectively (Fig. 2).

Figure 1: Chinas total installed capacity from 2003 to 2012

Figure 2: Compose of Chinas installed capacity in 2005 and 2012


The power generation technology has been rapidly developed in recent 10 years. The first 600
MW domestic manufactured supercritical unit with the initial parameters of 24.2MPa/566
/566, located in Huaneng Qinbei Power Plant, was put into operation in 2004. And the first
1000 MW/26.25MPa/600/600 domestic manufactured USC unit in Huaneng Yuhuan Power
Plant was put into operation in Nov. 2006. There are approximately 60 units with the capacity of
1000 MW operating in China. The thermal efficiency of a 1000 MW USC unit has achieved
45.4% (LHV), and its coal consumption for power generation is about 271 g/kWh. The technical
performance has already approached the advanced level of similar coal-fired units in the
worldwide.
There are two units in Guodian Taizhou Power Plant Phase project, which are double-reheat
1000 MW USC units with the initial parameters of 31 MPa /600/610/610, is now carrying
out the preliminary work. The project are planned to achieve the technical target as following: the
coal consumption for power generation is 256.7 g/kWh, thermal efficiency is 47.92% (LHV), coal
consumption for power supply is 267.1 g/kWh and auxiliary power rate is 3.91%, which will
approach the highest level in the world.
The initial parameters of Chinas coal-fired units have been gradually rising during the past years
(Fig. 3).

Figure 3: Raised initial parameters of Chinas coal-fired units

Due to the efforts of raising the initial parameters and closing down the condensing units smaller
than 100MW, the average coal consumption for power supply by coal-fired units nationwide has
been declining year by year in China. By the end of 2012, the average coal consumption has
decreased to 326 grams/kWh, which is about 20% lower than 2003 (Fig. 4).

Figure 4: Average coal consumption for power supply by coal-fired units nationwide per year
Forecast of Chinas power Generation development
For the future trend of Chinas power generation industry, Fig. 5 indicates the prediction made by
our institute, comparing the capacity in 2012, 2015 and 2020. We can see that Chinas installed
generation capacity will hit 1490 GW and 1800 GW respectively in 2015 and 2020; among
which, the capacity of fossil-fuel power plants will be 1040 GW and 1200 GW respectively.
Additionally, for proportion of fossil fuel units among the total power generation units, fossil fuel
powers share will slowly decline; on the other hand, the renewable energy will be developed
vigorously.

Figure 5: Forecast of Chinas installed power generation capacity in 2015 and 2020
BRIEF INTRODUCTION OF NATIONAL 700 USC CONSORTIUM
Members of the Consortium
On July 23, 2010, the launching ceremony of the National 700 USC Coal-Fired Power
Generation Technology Innovation Consortium was held in the Great Hall of the people by
National Energy Administration (NEA), shown in Fig. 6. The Consortium was lead by NEA, and
3

its 18 members are as follows: Xian Thermal Power Research Institute, Shanghai Power
Equipment Research Institute, Institute of Metal Research (Chinese Academy of Science), China
Iron & Steel Research Institute Group, Electric Power Planning & Engineering Institute, China
Power Engineering Consulting Group, Shanghai Electric Group, Dongfang Electric Corporation,
Harbin Electric Corporation, China First Heavy Industries, China National Erzhong Group,
Baoshan Iron & Steel Co.,LTD, Dongbei Special Steel Group Co.LTD, China Huaneng Group,
China Datang Corporation, China Huadian Corporation, China Guodian Corporation, China
Power Investment Corporation.

Figure 6: Launching ceremony of the Consortium


Organizational Structure of the Consortium
The organization structure of National 700 Consortium is shown in Fig. 7. The Council is the
highest decision-making body, leading the Secretariat and the Technical Committee. As the
executive section, the Secretariat is responsible for dealing with daily affairs. The Technical
Committee is the decision-making body on technical issues, composed of domestic well-known
experts. To facilitate the work, the Technical Committee is divided into four special working
groups, which are system & engineering solutions group, boiler group, turbine group and material
group.

Figure 7: Organizational Structure of the Consortium


Operation of the Consortium
The Council holds annual session, to decide on key issues of National 700 Consortium. The
Secretariat practices the Secretariat working meeting system, by holding quarterly meetings,
managing the routine work. The Technical Committee holds meetings in due time to make
decisions on major technical issues, responsible for the R&D of significant technical issues,
4

formulating the technical roadmap and R&D plans, carrying out project studies, supervising
project execution, and reviewing achievements of the research project at different stages. The
Technical Committee leads the four special working groups.
The research projects related to China 700 USC development are organized and supported by
the Consortium. The Consortium organized members making applications for the national
scientific research funds to the National Energy Administration, the Ministry of Science &
Technology (MOST) and other departments. After approved, the research projects are carried out
according to relevant management regulations or contract. During this process, the Consortium
coordinates the impediment and provide technical support to the projects (Fig. 8).
NEA

Fund

Apply

MOF

R&D Project

Approve
Lead

Consortium

Application
&
Approval

organize

Coordinate Support

Fund

Implementation

Review

Acceptance

MOST
Apply
Approve

Inspect

Appraise

* NEA=National Energy Administration;


MOF=Ministry of Finance;
MOST=Ministry of Science and Technology

Figure 8: Administration of Chinas 700 USC R&D projects


R&D PLAN OF CHINA 700 USC POWER GENERATION TECHNOLOGY
China 700 USC Technology R&D Roadmap
On the base of the R&D and application achievements in China 600 USC units, constructing
the R&D platform of 700 USC power generation technology aiming to master the 700 USC
technology and achieve the technical ability, taking the experience of other countries on
developing 700 USC unit for reference:
Proposing the overall technical plan and key equipment design plan of 700 USC unit;
Selecting, developing, assessing and optimizing the metal materials used in high
temperature components, to determine the series of heat-resistant materials for China
700 USC units;
Developing the production technology of key heat-resistant materials and the
manufacturing technology of key components;
Constructing the verification test platform, to take performance test for key high
temperature components;
Constructing the 700 USC demonstration projects, and mastering the technology of
700 USC.

Plan of China 700 USC R&D


Based on the actual situation of electrical, mechanical, and metallurgical industry in China,
overall R&D plan of China 700 USC has been made, intending to spend about 10 years to
achieve the 700 USC demonstration project.
R&D plan includes 5 sections (Table 1): 1 Overall design (2011~2013); 2 development of heatresistant materials, including selecting, developing, optimizing and assessing (2011~2017); 3
development of key components of main equipments and high temperature pipes (2012~2018); 4
Construction of test platform for key components and valves (2011~2017); 5 Demonstration
project (2015~2021).
Table 1: Overall plan of China 700 USC R&D
NO.

CONTENT

Overall design
Development of heat resistant materials (selecting,
developing, optimizing and assessing )
Development of key components of main equipments and
high temperature pipes

2
3

11

12

13

14

15

16

17

18

19

20

21

1)Boiler tubes
2)Boiler key components
3)Turbine Large Forgings
4)Turbine key Components
5)High temperature pipes and fittings
6)High temperature and high pressure valves
4

Construction of test platform and Conduction of test


1)Design and construction of test platform
2)Verification test of key components and valves of boiler

Demonstration power plant


1)Preparation of construction
2)Construction
3)Operation and experiences feedback

THE R&D PROJECTS IN PROGRESS


NEA-Initiated National Research Projects in Progress
National 700 USC Coal-fired Power Generation Technology and Its Engineering
Demonstration, an energy application technology research and engineering demonstration
project, has been initiated by National Energy Administration, including 6 sub-topics. 50 million
(RMB) will be provided from national research funds.
1) Sub-topic 1: Research on Overall Design
In this sub-topic, the overall design scheme of the 700 USC coal-fired units will be proposed,
including the initial parameters, capacity, thermal system and layout scheme.
Preliminary proposals on the initial parameter (35MPa/700/720) and the single-unit capacity
(~600MW) have been confirmed. Research and design work is carried out in accordance with two
proposed scheme, i.e. single reheat and double reheats. The power efficiency is expected to be
higher than 50%.
6

2) Sub-topic 2: Research on key materials


Heat-resistant material is the bottleneck problem of 700 USC. This sub-topic will carry out the
material characteristics tests, so as to master the related welding and non-destructive testing
technologies, in order to select, develop, optimize and assess the material in the 700 plan.
Currently the characteristics tests and assessment work on foreign materials, such as 617M
740740H high temperature alloys and T24 ferrite steel, etc., are in progress.
3) Sub-topic 3: Research on boiler key technology
This sub-topic will develop the fundamental engineering and manufacturing solutions for the
domestic-manufacturing of 700 USC boiler, as well as its ancillary equipments.
700 boiler general design and performance calculation has been completed, including boiler
frame design, arrangement of column, process design of steam/water/flue gas/air, heating surface
arrangement, De-NOx system selection, preheater selection and flue gas/air duct design, etc.
Preliminary proposals on materials for different boiler components have been made. Research on
material and its welding performance is in progress.
4) Sub-topic 4: Research on turbine key technology
This sub-topic is aimed at proposing the overall scheme for 700 turbine, mastering the design
of 700 turbine, as well as the key technologies for turbine materials and production, developing
700 USC turbine with better economics and high reliability.
700turbine general design and main technical specifications, as well as preliminary research on
heat-resistant materials have been done. Further R&D work, i.e. thermal system optimization,
flow path design, strength calculation and analysis of HP/IP welding rotor, research on
manufacturing cylinder castings, assessment test on heat-resistant materials, are in progress.
5) Sub-topic 5: Construction and operation of test platform for key components
A test platform with main steam temperature of 700 will be constructed based on a on a unit
already in operation; field tests on the heat-resistant alloys and key high temperature components
for 700 USC units will be carried out. The test platform is primarily built to meet the
requirements of doing verification tests on heat-resistant alloys and high temperature components.
The host unit, a 320MW SC unit, has been confirmed. The overall design of test platform has
been finished. Related R&D work is in progress, i.e. thermal calculation and checking, detail
design, material selection, drawing up of test plan, impact analysis of test platform on host unit,
etc.
6) Sub-topic 6: Feasibility study of demonstration power plant
This sub-topic will carry out the research on the proposed capacity scheme, and propose the
technical conditions of the main equipments, including an overall design plan for the
demonstration power plant, process systems design (such as thermal system and combustion

system), and layout proposal of each machine shop. Besides, it is also aimed at estimating the
project investment, comparing the cost, and conducting economic evaluation, etc.
This sub-topic is closely related to the construction of the demonstration projects. Currently,
China 700 USC technology R&D focus on some fundamental issues, especially the
development, test, and manufacturing of heat-resistant materials and high temperature
components, so it is premature for the demonstration projects. This sub-topic is still in the
preparation status.
MOST-Initiated National Research Projects in Progress
The Ministry of Science and Technology of china has set up a national program, to develop key
boiler tubes and pipes used for 700 advanced USC power plant. China Iron & Steel Research
Institute Group (CISRI) is the leading member of the project, along with 9 major partners as
follows: Baosteel Group Corporation, Fushun Special Steel Shares Co., Ltd, Yangzhou Chengde
Steel Pipe Co., Ltd, Institute of Mental Research of Chinese Academy of Sciences(IMR CAS),
Xian Thermal Power Research Institute Co., Ltd, Inner Mongolia North Heavy Industries Group
Corporation, Ltd, University of Science and Technology Beijing, Jiangsu University, and Jiangsu
Yinhuan Precision Steel Tube Co., Ltd. The government budget for this project was 29.5 million
RMB and the aforementioned companies, participated in this project, matched up their funding as
required in the course of this great adventure. The duration of the project is from 2012 to 2015.
The research and development of major candidate materials used for 700 USC power plants
were included in this project, such as G115, CCA617 equivalent, Inconel740H equivalent, and
GH2984G. To date, all candidate materials are under industrial trial to make tubes and pipes
according to their proposed application. Hopefully, full-size industrially manufactured tubes and
pipes will be available in china before the end of 2013.
PENDING R&D PROJECTS
China 700 USC R&D is in progress orderly, and there are nine more topics prepared to apply
for support from national scientific research funds, i.e.:

Topic 1: Research on the boiler water wall for 700 USC units

Topic 2: Research on the components of boiler superheater and reheater for 700 USC
units

Topic 3: Research on boiler header of 700 USC units

Topic 4: Research on high temperature steam pipes and fittings for 700 USC units

Topic 5: Research on manufacturing of turbine HP-IP rotor for 700 USC units

Topic 6: Research on manufacturing of high temperature cylinder valve housing for 700
USC units

Topic 7: Research on turbine high temperature blades and fasteners, valve cores wearresisting parts for 700 USC units

Topic 8: Research on turbine high temperature forgings for 700 USC units

Topic 9: Research on turbine high temperature castings for 700 USC units

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

ENCIO PROJECT: AN EUROPEAN APPROACH TO 700C POWER


PLANT.
A. Di Gianfrancesco: Centro Sviluppo Materiali S.p.A., Rome Italy
A. Tizzanini: ENEL Ingegneria e Innovazione, Pisa Italy
M. Jedamzik: Hitachi Power Europe, Duisburg, Germany
C. Stolzenberger: VGB, Essen Germany

Abstract
ENCIO (European Network for Component Integration and Optimization) is an European
project aiming at qualifying materials, components, manufacturing processes, as well as
erection and repair concepts, as follow-up of COMTES700 activities and by means of
erecting and operating a new Test Facility.
The 700C technology is a key factor for the increasing efficiency of coal fired power
plants, improving environmental and economic sustainability of coal fired power plants
and achieving successful deployment of carbon capture and storage technologies.
The ENCIO-project is financed by industrial and public funds. The project receives
funding from the European Community's Research Fund for Coal and Steel (RFCS)
under grant agreement n RFCPCT-2011-00003. The ENCIO started on 1 July 2011.
The overall project duration is six years (72 months), to allow enough operating hours,
as well as related data collection, investigations and evaluation of results.
The ENCIO Test Facility will be installed in the Andrea Palladio Power Station which is
owned and operated by ENEL, located in Fusina, very close to Venice (Italy). The Unit 4
was selected for the installation of the Test Facility and the loops are planned for 20.000
hours of operation at 700C.
The present paper summarizes the current status of the overall process design of the
thick-walled components, the test loops and the scheduled operating conditions, the
characterizations program for the base materials and the welded joints, like creep and
microstructural analysis also after service exposure.
Key words: Advanced Ultra Supercritical power plants, Nickel superalloys, welded
joints, test loop, creep behaviour, microstructural stability.
1. Introduction
In the ENCIO-project scientific and technological efforts aim to the successful
deployment of 700C technology in coal fired power plants. The key elements of ENCIO
are the installation and operation of a test facility in Fusina, at an ENEL power plant in
Italy. The project focus is on practical investigations, aiming at proving manufacturing,
welding, repair and life-time concepts for thick-walled components.
Europe is still in the global lead for the 700C technology in the field of fossil-fuel based
power generation. In order to continue with this successful technology development
ENCIO is the next important step to deploy this technology and to remain competitive.
Therefore a group representing the leading European power generators and
equipment suppliers has increased its efforts to speed up the evaluation and
assessment process out of COMTES700.
The 700C technology is a key factor:
9

Increasing efficiency of coal fired power plants,


Improving environmental and economic sustainability of coal fired power plants,
Achieving successful deployment of carbon capture and storage technologies.

The strategic importance of these R&D activities is evident for the electricity sector. The
integration of the renewables (RES) is one of the top priorities of European policy. The
consequence for the electricity supply system is that an appropriate balance between
intermitting and dispatchable generation is absolutely necessary. The fossil fuel based
power generation provides the back-up power and grid control function which are
necessary to ensure that the targeted contribution of renewable resources to the
electricity generation portfolio can be achieved.
It is expected that the price of electricity will increase further as soon as the RES share,
the demand for reliable back-up power as well as the global demand for coal will
increase. Different investigations in forecasting the evolution of the electricity sector - in
Europe and worldwide - show that beyond 2025 the commercial viability will be given
which would fit into the time line of the technical maturity of both 700C technology and
carbon capture & storage technologies.
The ENCIO project meets the goal of the RFCS program for coal in terms of efficiency
enhancements and in terms of carbon capture. The scientific and technical approach of
ENCIO consists of the connection in use of new and essentially improved Ni-based
alloys, the development and selection of new manufacturing processes and
consequently to test (under realistic operation condition) new design features of
components needed for the boiler as well as for the turbine.
ENCIO is the perfect transition from pilot towards demonstration, because it contains
elements of both with the goal to have a mature technology available at the end of the
project. In ENCIO the whole European expertise and competence generators and
suppliers are concentrated. This is a clear signal that the proposed work can be done
in a proper manner ensuring that the project objectives can be fully achieved.
Out of the structure and content of the work it is secured that the outcome of ENCIO will
deliver significant progress for the 700C technology, a progress which will be unique in
a global view.
The innovative part can be derived from the fact that new material concepts, new
manufacturing issues and new design features for the components will be demonstrated,
an absolute condition for the successful environmental, technical and sustainable
deployment of the 700C technology and the basis for an efficient carbon capture
process and for saving our coal resources.
The key elements of ENCIO are consequently the use of new and essentially improved
Ni-based alloys, the development and selection of new manufacturing processes and to
test under realistic operation condition new design features of components needed for
the boiler as well as for the turbine. The manufacturing is the key hurdle for a technically
feasible and commercially viable deployment of the 700C technology and consists of
more precise specifications for the ingot, forging, casting, welding including heat
treatment etc. These new process steps will be tested by a smart selection of
component elements required for a successful future operation.
The consequence of the significant progress beyond state of the art is a significant
improvement for the coming generation of fossil fired power plants by the application
and demonstration of their maturity of a complete new type of materials.
The industrial benefit is twofold and can be described as follows: for the European
industry active in the electricity sector the 700C technology provides an urgently
10

needed technology and a clear opportunity for world-wide application of this technology
as a technology leader.
The strategic relevance is given as pointed out before through its role as technology
provider in order to meet the European Councils decisions in mastering the climate
change and to sell this technology worldwide. The competitiveness of both electricity
sector and equipment supplier will be improved. Concerning the generators the
improvement is given by the strategic contribution to a balanced portfolio in lowering the
cost increase due combatting climate change.
For the equipment suppliers their market position will be strengthened in a global sense.
The credibility of the envisaged goals can be derived by the consistency of the working
packages and their deliverables as well as by the integrated responsible experts. The
ENCIO project is targeted for field testing but includes also modelling combined with
simulation. The industrial participation will cover at minimum level about 60% of all
efforts and is relevant in terms of numbers and competence.
The thematic area is an European one: it helps to make the European Council decisions
for decarbonizing the electricity sector reality. The direct link between highly efficient
thermodynamic processes and carbon capture technology is the key for a successful
deployment. The activities of the European Union (EU) in the field of CCS justify this in
addition. The clear need for a European approach is the fact that the partners constitute
themselves European wide and fulfill the specific request for high level skills and
expertise. The successful demonstration of all features of the 700C technology will be
applied all over Europe and globally as a follow-up.
As described previously the strategic importance is evident for the electricity sector. Only
by the back-up and grid control function of the dispatchable generation as nuclear and
fossil the strongly intended major contribution of renewable resources to the electricity
generation portfolio can be achieved. A reduction of the emissions in terms of CO2,
SOx, NOx and dust as well as the saving of resources by less coal consumption will
deliver an essential improvement to the health and safety.
The ENCIO project will contribute to the preservation of natural sources and
environment. This is possible by more efficient processing technologies, reducing
materials and energy consumption.
At the end it will support the use of renewable resources in the electricity sector.
2. PROJECT OBJECTIVES
The key goal of ENCIO is to concentrate all scientific and technological efforts in order to
make the 700C technology ready for deployment. ENCIO is an important step before
the erection of a 700C power plant can start. The ENCIO project consists of the
installation of a Test Facility in the ENEL coal fired power station Andrea Palladio,
based in Fusina (north-east Italy). The rationale behind is that the 700C technology is
the pre-requisite for a successful deployment of the significant technology paths for
carbon capture (i.e. post-combustion and oxyfuel requesting). Both technology lines high efficiency and carbon capture - will converge in a commercially viable power plant:
700C technology combined with carbon capture technology, i.e.
< 100g CO2/kWh and > 40% net efficiency by 2020
The experimental, demonstrational and investigational activities of ENCIO pursue the
following targets:
11

Provide proof of design and material behaviour of thick-walled components


under real operating conditions,
Close main technical open items derived out of the comprehensive analysis of
COMTES700, [1]
Test new developed materials and manufacturing options (e.g. post weld heat
treatments) to improve the reliability of weldments made out of Ni-based alloys,
Develop a life-time monitoring concept for pipes made out of Ni-based alloys,
Explore materials and manufacturing options having the potential to reduce the
investment cost of 700C technology and improve the load change behavior,
Verify the technical conditions for achieving high efficiency and better
environmental figures (lower emissions).
The ENCIO Test Facility (TF) will be installed in the Andrea Palladio Power
Station which is owned and operated by ENEL, located in Fusina (Italy). The
Unit 4 was selected for the installation of the Test Facility [2-4]. This unit has the
following characteristics:
o Boiler type: two pass boiler
o Burner configuration: tangential
o Steam capacity: 1050 t/h
o Production capacity: 320 MWe
o Fuel: hard coal + RDF (Refuse Derived Fuel)
o Superheater steam temperature: 540C
o Superheater steam pressure: 177 bar

Fusina

Figure 1: Enels power station Andrea Palladio Fusina (Venice Laguna) (Courtesy
of ENEL)
The Fusina power plant is perfectly qualified for installing the TF, as it has previous
experiences in the frame of innovation projects and it was already involved as host
power plant in other EU-funded projects, such as DEBCO (Demonstration of large scale
Biomass Co-firing) and H2-IGCC (Low emission gas turbine for hydrogen rich syngas).
Andrea Palladio station also already hosts a hydrogen fuelled demonstration plant,
including a demonstrative 16 MW GTCC (Gas Turbine Combined Cycle), where hot
component materials (combustor and gas turbine) are tested under long term operation
with hydrogen rich gas mixtures. All DCS (Distributed Control System) data of such a
facility are remotely monitored in real time by ENEL IIN offices based in Pisa. The
12

control, maintenance and operation of the hydrogen fuelled power plant are performed
with the support of personnel of the power station.
3. Structure, timeline, organization and funding
The ENCIO-project is financed by industrial and public funds. The project receives
funding from the European Community's Research Fund for Coal and Steel (RFCS)
under grant agreement n RFCPCT-2011-00003. The ENCIO project started on 1 July
2011. The overall project duration is 6 years (72 months), to allow enough operating
hours, as well as related data collection, investigations and evaluation of results.
The industrial fund consists of contributions of generating companies and of equipment
and materials suppliers. Five partners are contractors to RFCS:

VGB (VGB PowerTech e.V.) as co-ordinator


CSM (Centro Sviluppo Materiali S.p.A.)
ENEL IIN (ENEL Ingegneria e Innovazione S.p.A)
ENEL GEM (ENEL Produzione S.p.A)
HPE (Hitachi Power Europe GmbH)

VGB will be responsible for the overall project co-ordination and management including
the relations towards the EC RFCS. VGB is also the coordinator of the COMTES700+
program, that is the overall umbrella also for the other project running in Germany: the
GKM HWT II hosted in Mannheim GKM power plant [5,6]. The structures and the main
targets are summarized in figure 2.

Figure 2: COMTES700+ umbrella with GKM HWT II and ENCIO targets


The generating companies behind this proposal are CEZ a.s. (Czech Republic),
EDF/Electricit de France, EnBW Kraftwerke AG (Germany), ENEL (Italy), E.ON New
Building & Technology GmbH (Germany), STEAG GmbH (Germany), EVN AG (Austria),
GDF Suez (France), GKM/Grosskraftwerk Mannheim AG (Germany), RWE Power AG
13

(Germany), Vattenfall Europe Generation AG (Germany) and Vattenfall A/S (Sweden) as


well as Eskom Generation Business Engineering (South Africa). Those companies
represent a large share of the generated electricity in Europe. ESKOM is an important
international partner in clean coal technologies and in abating climate change via CDM
(Clean Development Mechanism) and other instruments.
Their high overall contributions to the project emphasize their strong interest and their
commitment to clean coal technologies with very high efficiencies. On the other side the
generators regard the funding from RFCS as very important as it renews the European
Union's interest in supporting 700C technology and underlines the continuity of
European energy research.
Manufacturers and suppliers are also involved in ENCIO - either as a contractor (HPE)
or as an associated partner (HORA/Holter Regelarmaturen GmbH &Co. KG, Sandvik
AB, Sempell AG, Outokumpu (formerly ThyssenKrupp VDM GmbH), V&M/Vallourec &
Mannesmann Tubes, Voestalpine Giessen Traisen GmbH). Nippon Steel Sumitomo
Metals Co. (formerly Sumitomo Metals) is also an associated partner supplying HR6W
pipes and the testing has been allocated at CSM under separate contract.
Associated partners have agreed to contribute to the project in relation to the added
value of their order volume. They will have access to the know-how related to the
provided materials and components. This partnership approach will ensure a close cooperation, share of expertise and commitment of all parties involved.

Figure 3: ENCIO project structure

14

4. Project Schedule
The main time milestones are:
late February 2014 start of erection activities outside the boiler
Summer 2014 Fusinas outage for erection activities inside the furnace (i.e.
additional
superheater (SH) bundles, etc.)
early Fall 2014 start of experimental activity
Fall 2017 stop of experimental activity and start of investigation activities after
dismantling
The project structure of the ENCIO project is shown in the chart.
2010

2011

2012

2013

2014

2015

2016

2017

Application to
RFCS Negotiation
Basic Design
EPC
Integrity
Document
Erection &
Commitment
Test Rig
Operations
Investigations

Figure 4: Project structure


5. Description of the tests loops
The Test loop description is shown in figure 5 as well as the 3D schematic view:

Steam is taken from existing boiler (Host Plant) at 540C


Steam is additionally superheated up to 705C by additional SH bundles
installed into the furnace
Additionally superheated steam is sent to experimental components
Experimental components are installed in four Test Loops having the following
scopes:
o Test Loop 1: Development of pipe repair concept
o Test Loop 2: Test of Hot Isostatic Pressing (HIP) parts and weldments
as well as life-time monitoring
o Test Loop 3: Test of different Ni-based alloys and elements
o Test Loop 4: Test of turbine cast material and weldments

15

Figure 5: P&I Diagramm and 3D-View of the Test Facility


5.1 TEST LOOP 1: Development of pipe repair concept
Take advantage of additional investigations carried out in COMTES700 and in ENCIO to
get optimized WPS. Prior to start any weld testing basic investigations have been done
to characterize the aged material.
Complimentary to COMTES700 different heat treatments (pre-/post-) will be applied and
a trial weld of each material combination and/or welding method will be tested
destructively. Trial welds will be carried out and all those welding methods without any

16

crack indication will then be selected to produce components to be installed in the test
loop.
Different NDT (Non Destructive Test) methods will be applied and tested to assure high
quality standards.
Non Destructive Surface Tests will be performed frequently during operation for all test
welds. A final repair with orbital TIG (Tungsten Inert Gas) narrow gap and electrode
welding will be executed on additionally aged pieces of A617B (having been in operation
in COMTES700).
5.2 TEST LOOP 2: Hot Isostatic Pressing (HIP) parts + Lifetime monitoring
The HIP technology is interesting for fabrication of T-pieces, valve bodies and turbine
parts. Although this technology is already commercially applied in other fields, it has not
yet been adopted for boiler and turbine pressure parts in power plants (alloy 617B or
alloy 625). The HIP-technology is promising for substituting expensive castings and
consequently a cost reduction may arise from its application to 700C technologies.
NDTs will be performed frequently during operation for all test welds. After the end of
operation final repair welds will be executed on all dismantled pipes of Test Loop 2.
The creep behaviour of alloy 617B is monitored by running tests under respective load
and temperature, as well as by using a thin wall piece designed for ~30000 h (measuring
and monitoring of the creep online).
5.3 TEST LOOP 3: A617 OCC / A263 / HR6W / A625 cast base material and
dissimilar welded joints
The optimized chemical composition of A617B, called A617OCC is applied to explore
possible improvements in weldability, which is due to less formation of chromium
carbides. Additionally, an optimized melting process is implemented to reduce the
amount of impurities in the ingot. Such an optimization has the potential to make welds
more reliable. This is also expected to be an option to reduce relaxation-cracking and
hot-cracking occurrences, which is one of the objectives of ENCIOs tests and
investigations. It can be expected that due to the new melting process the improved
weldability may lead to reduced pre- and post-weld heat treatment requirements. Thus,
A617B OCC has also to be considered as a possible option for reduction of investment
costs of 700C technology.
Furthermore, other Ni-based alloys like A263, HR6W and A625 will be tested. A263
shows a high potential in cost reduction but has not yet been in test operation.
It is also planned to test other material like HR6W for the temperature range 620C to
700C from suppliers such as Sumitomo. These materials will be installed additionally to
the ENCIO test program.
The weldments of the material combinations A617B OCC HR6W and A263 - A625
cast are tested and compared with each other. This is necessary as the combinations
A263 - A625 cast (neither with nor without heat treatment) and A617B OCC HR6W
have not yet been tested and investigated with the required heat treatment.
This aims at an optimized design for the overall plant with regards to competiveness,
costs and performance.
NDTs will be performed frequently during operation for test welds. After the end of
operation final repair welds will be executed on dismantled pipes of Test Loop 3 which
are long enough for this purpose, followed by microstructural investigations, mechanical
and creep testing.
17

5.4 TEST LOOP 4: Turbine cast material


Need to prove thick walled welds in the range of the real pipe dimensions of a
demonstration plant with material combination alloy 617 OCC - alloy 625 cast. The
material combinations alloy 617 OCC/alloy 625 is not tested in COMTES700 (real
dimensions).
Even if there was no negative indication about the behaviour of the welded material in
the combination alloy A617B forged - A625 cast material, the dimensions of the test
weld in COMTES700 had been at relatively thin wall thickness (80 mm). So there is the
need to prove thick-walled welds in the range of the real cast dimensions of a
demonstration plant with material combinations A617B OCC - A625 cast. The
weldments A625 cast - A625 cast with wall thickness in the range of 150 mm was not
foreseen in earlier designs for this component but with the post-weld heat treatment
needed for A617B OCC - A625 cast this design issue is necessary to be introduced and
tested.
6. Erection concept
Welding technologies that are suitable for on-site applications (i.e. orbital TIG narrow
gap welding and electrode welding), are applied both for new material and aged
material, because it will be mandatory needed to develop a repair concept for the future
construction of commercial size power stations based on 700C technology. Since these
welding technologies make it possible to perform welding activity not only in the
workshop, ENCIO will provide the proof of feasibility of such a welding approach.
Different non-destructive methods will be applied to secure the safety of the TF. Among
others, PTs (Penetration Tests) will be applied to check the surface of weld seams for
flaws.
UTs (Ultrasonic Tests) are decisive to identify flaws in the weld seam.
The overall assembly details of the 4 test loops are summarized in figure 6.

Figure 6: details of the test loops

18

7. Project work plan


The ENCIO project is structured in six WPs (Work Packages) with clearly defined
responsibilities and deliverables, meeting the envisaged objectives.
WP1 Engineering (Work Package leader: Hitachi Power Europe GmbH, HPE)
This WP covers all activities necessary to design the TF and to integrate the equipment
into the host plant. It also includes the selection and design of components, such as
superheaters, steam pipes, valves and test loops, as well as of the C&I (Control &
Instrumentation) system and the according measuring devices and the permitting
process.
WP 2 Materials and Manufacturing (Work Package leader: HPE)
This WP includes all procurement of materials and components, as well as
manufacturing of trials to prove weldability and mechanical properties prior to
installation. Components to be installed will then be properly manufactured. NDT
methods will also be applied. It also includes the qualification through inspection bodies.
WP 3 Erection and Commissioning (Work Package leader: ENEL Ingegneria e
Innovazione S.p.A., ENEL IIN)
This WP includes all activities needed to install the TF in the host plant, the related cold
and hot commissioning, as well as the dismantling of the TF after completion of the
demonstration program.
WP 4 Investigation (Work Package leader: Centro Sviluppo Materiali S.p.A., CSM)
This WP includes all investigations necessary to qualify components and to characterize
material properties. This comprises the assessment of base material (before and after
operation), welded joints (before and after operation) and repair welds (after
dismantling).
It also includes a creep test program, which is essential especially for materials (e.g.
A617B aged, A617B OCC, A263) and manufacturing processes (e.g. forging, extrusion,
casting, HIP), where knowledge of material behaviour is very poor. Evaluation of results
is another important item addressed in the WP in order to provide a basis for the design,
operation and repair methods for future 700C power plants.
WP 5 Operation (Work Package leader: ENEL Produzione S.p.A., ENEL GEM)
This WP includes the start-up, operation and monitoring of relevant TF data. Operating
data will be properly stored on a server and periodic inspections and non-destructive
testing will be carried out.
WP 6 Co-ordination and dissemination (Work Package leader: VGB PowerTech e.V.,
VGB)
This WP includes the project management and reporting to assure meeting the project
objectives, the budget, time line and compliance to RFCS rules. Furthermore, the
management of the collaboration between the project partners and the protection of IPR
(Intellectual Property Right) will be part of the Work Package. Dissemination actions
(e.g. events, publications) are an essential part of the scope of this WP.
The inter-dependencies between the different WPs are consisting of well-defined
interfaces as input and output features - a pre-requisite for an efficient and successful
performance.

19

Due to the ENCIO project objectives comprising the implementation of new materials,
new manufacturing processes and new welding and testing features a comprehensive
material qualification procedure is required to comply with obligations of the certified
bodies. As this certification process follows defined rules the preparatory work before
erection cannot be expedited. Nevertheless the feasibility to start qualification tests prior
to the RFCS commencement date has to be explored.
8. Main facts
TF installed - tie-ins installed end of Aug. 2012 (figure 7) [4]

all time critical orders already put in place


test loop 1: material available, trial welds started
test loop 2: per-investigation trials accomplished and pipe manufacturing completed
test loop 3: pipe produced (figure 8) and weld manufacturing started. NDE started.
Pipe cut and sampling started (PMA, WPQR, R&D purposes)
test loop 4: material manufacturing completed
Investigations on all the base materials are started
Investigations on HR6W welded are started

End of erection scheduled end of August 2014


Commissioning planned to start on September 2014
Experimental operation to be started by end 2014

Figure 7: tie-in installation [2]

20

Figure 8: HR6W pipe experimental component (courtesy of NSSM co.) [2]


9.

Work package Investigation

This work package is leaded by Centro Sviluppo Materiali S.p.A. (CSM) with the
cooperation of Technische Universitt Darmstadt (TUD) and VGB PowerTech e.V.
(VGB).
The WP4 main objectives are:
The definition of the ranking of manufacturing processes and repair concepts to
reduce the risks in future long-term plant operations,
The identification of life time determinants in particular including main damage
mechanisms in long-term service exposed components (through a
comprehensive characterisation of materials and components), in order to
provide a basis for design, operations and repair methods on future 700C
power plants.
The materials scheduled to be installed in the five test loop are: A617B, A617B OCC,
A625, A263, HR6W. The production processes scheduled for component production
are: forging, casting, extrusion, pilgrim and HIP.
The welding procedures scheduled for the components production are: Orbital TIG
narrow gap welding (cold wire), Electrode welding and TIG narrow gap welding (hot
wire).
This WP will also support WP2, by carrying out laboratory analyses on trial welded joints
devoted to destructive testing (DT) for achieving the proof of weldability and fabricability.
Depending on the case, welded joints will be tested with or without pre weld heat
treatment. Post weld heat treatment will always be applied.
Task 4.1 Base materials: A comprehensive mechanical assessment (hardness, tensile,
indentation, impact tests) and microstructural characterisation (light optical microscopy
(LOM), scanning electron microscopy (SEM), transmission electron microscopy (TEM),
and X-ray diffraction (XRD) analysis) of base components (made by A617B,
A617B OCC, A625, A263), as new material batches or priory exposed in the
COMTES700 loop (without overlapping with the COMTES700 tests), manufactured by

21

different process routes (forging, casting, extrusion, pilgrim, HIP) and with different sizes
and/or wall thicknesses.
Task 4.1 Welded joints: A comprehensive mechanical assessment (hardness, tensile,
impact tests) and microstructural characterisation (LOM, SEM, TEM, XRD analysis) of
welded joints (similar and dissimilar), produced by different welding procedures (Orbital
TIG welding - cold wire, Electrode welding, TIG narrow gap welding - hot wire) and with
or without pre welding heat treatment. Post weld heat treatment will applied be always.
Task 4.3 Simulation of in-service exposure conditions: Specimens from the creep
test program will be adopted for indentation tests to obtain information on mechanical
properties. TEM examinations (head and gauge length) at different times and
temperatures to evaluate on base metals and in weld metals or HAZ (Heat Affected
Zone), for the welded joints, in order to assess and quantify the strain-induced
precipitation effect on overall microstructure evolution (measured in terms of
precipitation evolution and coarsening, formation of new phases).
Task 4.4 Dismantled components: A comprehensive mechanical assessment
(hardness, tensile, indentation, impact tests) and microstructural characterisation (LOM,
SEM, TEM, XRD analysis) of exposed base metals and welded joints (after dismantling)
will be performed.
Task 4.5 Welding repair: A comprehensive mechanical assessment (hardness, tensile,
indentation, impact tests) and microstructural characterisation (LOM, SEM, TEM, XRD
analysis) of the welded joints carried out on in-service exposed pipes taken from
COMTES700, will be performed.
Task 4.6 Creep Assessment program: A creep test program will be carried out to
characterize the long-term behaviour of the different base metals and welded joints
(similar and dissimilar) produced by different WPS (Welding Procedure Specification), as
well as, the effect of pre-/post- weld heat treatments, giving indication on the stress
reduction factor to be applied for the industrial components.
It is foreseen to share the work within all tasks among CSM, VGB and TUD according to
the corresponding resources.
An unique combination of mechanical (instrumented tensile tests or special indentation
procedures from RT to 800C, hardness and micro hardness indentations, etc.) and
metallographic investigation techniques (LOM, SEM with X-ray microanalysis (energydispersive spectroscopy, EDS), EF-TEM incl. scanning transmission electron
microscopy (STEM), XRD) have been selected and applied.
Data generated and metallurgical evidences correlated to the mechanical behaviour of
laboratory samples and full-scale components will be critically analysed and evaluated in
order to identify main damage mechanisms in long-term service exposed components.
This evaluation will be shared with the other partners and results will be made available
to WP 2, WP 5 and WP 6 to provide a basis for design, operations and repair methods
on future 700C power plants.
The key purpose of WP 4 is that all results and findings out of the operation of the TF
will be compiled and evaluated. Based on this assessment the partners will have the
necessary conclusions for engineering, designing, manufacturing and commissioning in
order to start with the erection of a 700C power plant. Therefore, VGB will integrate all

22

expertise familiar with the 700C technology in Europe as Laborelec (Belgium),


Fraunhofer institutes (Germany) or others.
Acknowledgements
The ENCIO project receives funding from the European Unions Research Fund for Coal
and Steel (RFCS) research programme under grant agreement n RFCP-CT-201100003.
Acknowledgements to all the other colleagues of the team:
-

VGB: Sabine Polenz;


CSM: Paolo Lombardi, Arianna Gotti, Susanna Matera, Pietro Gimondo, Luigi
Russo, Gennaro Inserra, Silvia Tiberi Vipraio;
ENEL Innovazione e Ricerca: Silvia Olivotto, Nicola Rossi, Marco Gazzino;
ENEL Produzione: Vittorio Bellini,
Hitachi Power Europe: Ralf Udo Husemann, Friedrich Klauke, Martin Becker;

Bibliography
1) http://www.vgb.org/en/research_project261.html
2) M. Gazzino: ENCIO project: overview, status up-date and planned fireside
corrosion tests Enel Engineering and Research Division VGB/IEA Workshop
Advanced USC coal-fired power plants; 19-20 September 2012 Vienna, Austria
3) http://www.encio.eu/contact.html
4) http://www.vgb.org/en/research_project355.html
5) http://www.vgb.org/fue_projekt321.html

23

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

ADVANCED USC TECHNOLOGY DEVELOPMENT IN JAPAN


Masafumi Fukuda
Research Institute for Advanced Thermal Power Systems, Tokyo, Japan
Eiji Saito
Hitachi Ltd., Hitachi, Japan
Hiroyuki Semba
Nippon Steel & Sumitomo Metal Corporation, Amagasaki, Japan
Jun Iwasaki
Babcock Hitachi K. K., Tokyo, Japan
Sakae Izumi
Fuji Electric Systems Co. Ltd, Kawasaki, Japan
Shinichi Takano
IHI Corporation, Tokyo, Japan
Takeo Takahashi
Toshiba Corporation, Tokyo, Japan
Yasuo Sumiyoshi
Mitsubishi Heavy Industries, Ltd., Yokohama, Japan
ABSTRACT
We have reported on the effort being done to develop the A-USC technology in Japan, which
features the 700 deg-C steam condition, since the 2007 EPRI conference. Our 9 year project
began in 2008. There have been some major changes in the electricity power market in the world
recently. At first, the earthquake changed the power system violently in Japan. Almost all nuclear
power plants have been shut down and natural gas, oil and coal power plants are working fully to
satisfy the market's demands. In the USA, the so called 'Shale gas revolution' is going on. In
Europe, they are working toward the target of reducing CO2 emissions by the significant use of
renewables with the backup of the fossil fuel power systems and enhancing power grids. A very
rapid increase in power generation by coal is being observed in some countries. Despite some
major changes in the electric sector in the world and the CO2 problem, the global need for coal
power generation is still high. We can reconfirm that the improvement of the thermal efficiency
of coal power plants should be the most fundamental and important measure for the issues we are
confronting today, and that continuous effort should be put towards it. Based on the study we
showed at the 2007 conference, we developed 700 deg-C class technology mainly focusing on the
material and manufacturing technology development and verification tests for key components
such as boilers, turbines and valves. Fundamental technology developments have been done
during the first half of the project term. Long term material tests such as creep rupture of base
materials and welds will be conducted for 100,000hrs continuing after the end of the project with
the joint effort of each participating company. Today, we are preparing the plan for the second
half of the project, which is made up of boiler components test and the turbine rotating tests.
Some boiler superheater panels, large diameter pipes and valves will be tested in a commercially

24

operating boiler from 2015 to 2017. The turbine rotor materials which have the same diameter as
commercial rotors will be tested at 700 deg-C and at actual speed.
INTRODUCTION
Figure 1 shows the power generation trend and its fuel share in Japan1). The amount of power
generation increased steadily during 1980s and 1990s. In 2000s, it was maintained around 100
TWh. After the Oil shock in the 70s, coal fired power plants had been steadily replacing oil
fired power plants until the earth quake in 2011, which struck northern Japan and caused tsunami.
Fukushima #1 nuclear power plant was severely damaged by the tsunami. Today, almost all
nuclear power plants have been shut down and natural gas, oil and coal power plants are working
fully to satisfy the market's demand. Natural gas dominates almost half of the power generation
and old oil power plants have returned to the market. Consequently, the cost of fuel which is
imported from abroad increased significantly and turned the countrys international trade into
deficit from surplus.
Coal is expected to become more favorable choice for the newly built power plants in addition to
replacing old coal power plants because of its price and availability. In this situation, the
reduction of CO2 emission from coal power plants is the crucial issue for the country.

Figure 1 Power generation and fuel share trend in Japan1)


The improvement in the efficiency of the coal fired power plants has been mainly achieved by
raising steam conditions as shown in Figure 2. The steam temperature was raised from 538deg-C
to 566deg-C at the end of the 50s, and remained at this temperature until 1993. Steam power
plants that have been built recently usually have a steam temperature of around 600deg-C and a
steam pressure of 25MPa. We usually call such a steam condition USC (Ultra Super Critical
steam condition).

25

Figure 2 Trend of steam conditions in Japan


Electric Power Development Company (J-POWER) started a comprehensive development
program of USC technology in 1981 to develop the USC technology, subsidized by the Japanese
government. Materials to be used for 600 to 650deg-C systems were developed through the year
2000. The materials which contain 9 to 12 Cr steels and were developed at that time are being
used for the USC plants in Japan today.
CONCEPT OF A-USC
Following J-Powers programs, we did some case studies on a system which has a 700deg-C
class steam temperature. The result of this study was reported at EPRI conference in 2007 and
contributed to establish the concept of A-USC. The 700deg-C class A-USC technology will be
developed based on todays latest 600deg-C class USC technology by raising the steam
temperature 100deg-C (Figure 3). The target net thermal efficiency for the higher heating value
base is 46 to 48%. This is more than 10% higher than that of the 600deg-C class USC. That
means more than a 10% drop in CO2 emissions.
In addition to the efficiency improvement, biomass co-firing and CCS (Carbon Capture and
Storage) can be coupled to reduce CO2 emissions further. If a CCS system is added on to a coal
power plant, a considerable amount of energy is consumed and there is a large reduction in the
efficiency of the plant. It is necessary to use a high efficiency system such as A-USC for power
generation to make CCS feasible.
35MPa, 700

A-USC
Net Thermal Efficiency 4648%HHV

Boiler

Biomass Co-Firing
CO2 Recovery
Oxyfuel
Exhaust Gas

Steam Turbine

25MPa, 600

USC
Net Thermal Efficiency 42%HHV

Boiler
Steam Turbine

Figure 3 700deg-C class advanced USC (A-USC)


26

In the case study, we checked existing coal power plants in Japan. There are many older coal
plants in Japan, built in the 70s and early 80s, which will reach the point where they will need
to be rebuilt or retofitted in the near future. Almost all Japanese coal power plants already have
environmental protection equipment like DeNOx, DeSOx, and EP which make up a large part of
the plants. If we can retrofit the old plants with the higher steam temperature technology it is
possible to reduce CO2 at a much lower cost than if we used the other technologies because we
can reuse the expensive environmental protection equipment and some other cold parts from the
existing plants.

Figure 4 Number of Existing Plants and Steam Conditions

Figure 5 Case Studies


Then, we selected model plants with the parameters shown on Figure 5
Case A had a 700 deg-C class double reheat condition. It was expected to have the highest level
of thermal efficiency improvement. But, it would be necessary to change a single reheat plant to a
double reheat plant. Case B had a 700 deg-C class single reheat condition. It was expected to
have a fairly high level of thermal efficiency improvement and not require extreme remodeling.
Case C had a 700 deg-C class temperature only in the reheat system. The temperature of the main
steam system is kept at 610 deg-C. This configuration would enable us to choose ferritic
materials for the main steam system. It was expected to have a good level of thermal efficiency
improvement and require only light remodeling without the heavy use of Ni-based alloys.
27

We considered Ni-based alloys for the 700 deg-C class turbine rotors. It is necessary to use Ni or
Ni-Fe-based wrought material for rotors of 700 deg-C class steam turbines unless advanced
cooling technology is used. However, it is difficult to make large Ni or Ni-Fe-based wrought
parts and it is assumed that the largest possible size of a part is around 10 tons. Turbine rotors for
large capacity power stations usually weigh 30 to 40 tons. The Welded Rotor concept was
proposed to make large rotors by welding Ni or Ni-Fe-based wrought material and steel. Figure 6
shows a typical welded rotor.

Figure 6 Welded Rotor (Courtesy of MHI)


By introducing the gas turbine cooling technology, we can reduce the use of Ni or Ni-Fe-based
materials. Figure 7 shows a plan of an intermediate pressure turbine (IPT) that is cooled. Ni-based
alloy is used only for the blades and some stationary parts around the steam inlet.
Inlet Pipe, Nozzle Casing
Ni-Based-Alloy
Outer Casing CrMoV
Inner Casing 12% Cr
Rotor 12% Cr

Blades Ni-Based-Alloy

Figure 7 Cooled Turbine (Courtesy of Toshiba)


A typical example of material selection for A-USC is shown in Figure 8. The blue color
represents conventional materials; green means gas turbine materials, pink means materials under
development. And solid dark pink means Ni-based alloys under development. Ni-based alloys,
which have not been used for USC, were chosen for a part of the superheaters and reheaters, the
large steam pipes and the valves going from the boiler to the turbines, and a part of the turbine
rotors and casings. The turbine rotors consist of Ni-based alloy and 12Cr steel, which are welded
together. The turbine nozzles and blades for the high temperature stages use Ni-based materials
that are being used for gas turbines.
28

Figure 9 shows the estimated plant thermal efficiency (Net, HHV) of each case after
refurbishment. The thermal efficiency of the original plant is about 40%. Case D is a reference
case, which uses the current technology.

Figure 8 Selected Materials for case A


Net Thermal EfficiencyHHV

48

46

44

42

40
A

Case

Figure 9 Estimated Thermal Efficiency


Main and reheat steam temperatures are 600 deg-C and the system configuration is a single reheat
cycle. Compared to Case D, Case A has a 3.8 point advantage. Case A has the best thermal
efficiency, 46%, and requires relatively heavy remodeling. The thermal efficiency of Case B is
44.3% and 2.1 points better than that of Case D and 1.7 points less than that of Case A. As Case
B has a single reheat cycle, the remodeling is not as extensive as in Case A. The thermal
efficiency of Case C is 43.4%. Because Case C employs a 610 deg-C main steam temperature and
a 720 deg-C reheat steam temperature, the thermal efficiency advantage of Case C is less than
Case A and B. However, it doesnt need any Ni based alloy parts in the main steam system.

29

A-USC TECHNOLOGY DEVELOPMENT PROJECT


A long term A-USC technology development project began in 2008. In the first half of the project,
boiler, turbine and valve materials are being developed and verified. In the second half, boiler
components and small turbine tests will be done to verify the reliability of each component
(Figure 10). Throughout the project, long term creep rupture tests will be done on each candidate
material and welded joint.
On EPRI conference in 2010, we reported our project status of the beginning part as shown on
Figure 10. We would like to show brief summary of previous report and the latest status of the
project. We began the project with eight members. Today, twelve companies and institutes, ABB
Bailey Japan, Babcock-Hitachi, Central Research Institute of Electric Power Industry (CRIEPI),
Fuji, Hitachi, IHI, Mitsubishi heavy Industries (MHI), National Institute for Materials Science
(NIMS), Okano Valve, Nippon Steel & Sumitomo Metal, Toa Valve, and Toshiba are
participating (Figure 11). ABB Bailey Japan, CRIEPI, IHI, MHI, NIMS, and Nippon Steel &
Sumitomo Metal are working together on the boiler technology development. ABB Bailey Japan
is developing the valve technology for turbine bypass valves. The boiler materials for testing are
provided mainly by Nippon Steel & Sumitomo Metal. Fuji is in charge of the valve technology
for steam turbines. Hitachi, MHI and Toshiba are supporting Fujis research. The steam turbine
technology is being developed by Hitachi, MHI and Toshiba. Okano Valve and Toa Valve are
developing the technology for general-purpose valves and safety valves.

Figure 10 Master Schedule

30

Figure 11 Project Structure


BOILER TECHNOLOGY DEVELOPMENT
Once we decide to develop and use a new material, we need to thoroughly develop related
technologies such as maintenance, fabrication, component, and system design technology (Figure
12). Long term reliability, high temperature corrosion resistance, steam oxidation resistance, and
fatigue resistance are crucial characteristics for the material. Life assessment and welding
technology for aged materials are necessary to repair used components. Welding, bending, and
NDT technologies must be developed to fabricate components. Valves, desuperheaters, and
headers are essential to build a boiler system. In addition to the hardware technologies, system
design technologies, such as piping design, structural design, and thermal analysis are the key to
the successful construction and operation of the boiler system. The companies participating in the
project are working together to develop each technology.
Figure 13 shows the candidate materials for boilers which were prepared by Nippon Steel &
Sumitomo Metal. HR6W, HR35, Alloy 617, Alloy263, Alloy740, and Alloy141 are Ni based alloys
for use at temperatures higher than 650deg-C. High boron 9Cr steel, low carbon 9Cr steel and
SAVE 12AD are ferritic steels for use at temperatures lower than 650deg-C. These materials are
being tested to verify the characteristics regarding creep rupture, fatigue, oxidation and corrosion.
Welding and bending tests have been conducted to check the manufacturability of the materials.
Figure 14 shows an example of creep rupture strength which was derived from HR6W creep
rupture tests. HR6W is being developed by Nippon Steel & Sumitomo Metal for pipes and tubes
in the A-USC system, and designed to have good corrosion resistance to combustion gas in the
boiler and good creep strength for pressurized pipes and tubes at 700deg-C. 100,000hr creep
rupture stress at 700deg-C is expected to be about 90MPa which is the target for large steam pipe
materials
The creep rupture strength of HR6W weld joint is shown in Figure 15. Creep tensile specimens
which contain welded portions were cut out from welded HR6W plates. Weld material was
Alloy617. When tested, all specimens broke at base material. We can see that the strength of the
welded material is equivalent to or higher than that of HR6W without a weld joint.
31

Improvement of creep strength

Long

term reliability
(including Weldment)
Temp. Corrosion
Steam Oxidation
Fatigue

Materials

High

Maintenance

System Design
Piping

(Thermal Exp.)
design
(Thermal Stress)
Basic Engineering
(Boiler Concept)
System

Boiler
Technology
Development

Components
Valves

(SV/CV etc)
DeSH
(including Turbine bypass Spray )
Header
Boiler

Life
Assessment
Welding for
Aged materials

Fabrication
Welding

(including Dissmilar)

Bending
NDT

Figure 12 Boiler development tasks

Figure 13 Boiler Candidate Materials (Courtesy of Nippon Steel & Sumitomo Metal)

Figure 14 HR6W creep rupture strength (Courtesy of Nippon Steel & Sumitomo Metal)

32

Figure 15 Creep rupture strength of HR6W weld (Courtesy of Nippon Steel & Sumitomo Metal
and Babcock-Hitachi)
We have made pipes of HR6W, HR35 and Alloy617 using the Ehrhardt push bench method
(Figure 16). These pipes were prepared in order to conduct material tests, such as welding,
bending and the long term creep test. A large diameter pipe was made for a reheater header
mock-up (Figure 17). The diameter and the wall thickness of the pipe are 675mm and 72mm.

Figure 16 Pipes prepared for material test (Courtesy of Nippon Steel & Sumitomo Metal)

Figure 17 Large diameter pipe prepared for header mock-up (Courtesy of Nippon Steel &
Sumitomo Metal)
Figure 18 shows bended pipes which are made of Alloy617 and HR6W. These pipes were cut into
pieces to check the shape of the cross section and to make test specimens. Pipe welding tests were
carried out. A sample of welded pipes is shown on Figure 19. We put some specimens which
were cut out from the welded portion of pipes into long term creep rupture test.
Some advanced 9Cr steels were also tested. Figure 20 shows the cross section of welded
SAVE12AD pipe and a bended pipe. In the HAZ section of welds, we didnt find any small grain
structure.

33

Some header mock-ups were made to check manufacturability of actual parts of A-USC boiler.
Figure 21 shows a reheater mock-up which uses a pipe shown on Figure 17.

Figure 18 Pipe trial bending (Courtesy of Babcock-Hitachi and Mitsubishi Heavy Industries)

Figure 19 Pipe welding test (Courtesy of IHI)

Figure 20 SAVE12AD pipe welding and trial bending (Courtesy of Babcock-Hitachi)

Figure 21 Reheater Header Mock-up (Courtesy of Mitsubishi Heavy Industries)


TURBINE TECHNOLOGY DEVELOPMENT
There are three candidate materials for steam turbine rotors (Figure 22). Ni based alloys, FENIX700, LTES and TOS1X, are being developed and tested for use at temperatures higher than
700deg-C. The aim of the development of FENIX-700 is to build a rotor heavier than 10 tons
without segregation in the material. The purpose of LTES and TOS1X is to target a weight of
around 10 tons which will be welded to steel parts to make a 30 to 40-ton rotor.

34

Materials

Temperature
Level

Weight

Development Target

FENIX-700

700

>10ton

Ni-base material heavier than 10 tons


without segregation

LTES700R

>700

3040ton

>720

Ni:10ton
+
Steel:20~30ton
Welding

10 ton Ni-base material with good


weldability to steel

TOS1X

10 ton Ni-base material with good


weldability to steel

Figure 22 Steam turbine rotor materials


FENIX-700 which has superior long-term stability at 700deg-C was developed from Alloy706 by
reducing Nb content and increasing Ti and Al content2). The 100,000 hour creep rupture strength
at 700deg-C is expected to be higher than 100MPa.
LTES700R is a Ni-based alloy that has been developed by MHI. This alloy was developed to have
a thermal expansion coefficient similar to 12Cr steel, so it conforms well to conventional steels
(Figure 23). In addition, the creep rupture strength of LTES700R is higher than the target for 700
deg-C class rotor material (Fig24). Originally, LTES700 was developed for small parts, such as
casing bolts. LTES700R was developed from LTES700 for large steam turbine rotors. Fig. 25
shows a large LTES700 material. Welding technology is crucial for this material. Welding tests
including the welding of dissimilar materials were carried out. Figure 26 shows a similar and a
dissimilar welding of LTES700R.

Figure 23 Coefficient of thermal expansion of LTES700R (Courtesy of Mitsubishi Heavy


Industries)

35

Figure 24 Creep rupture strength of LTES700R (Courtesy of Mitsubishi Heavy Industries)

Figure 25 Large LTES700R material before forging (Courtesy of Mitsubishi Heavy Industries)

Figure 26 Welding test of LTES700R


(Courtesy of Mitsubishi Heavy Industries)

36

Figure 27 TOS1X creep rupture strength (Courtesy of Toshiba)


TOS1X was developed from Alloy617. The earlier version of TOS1X, which is now called
TOS1X-I, is expected to have around 200MPa of 100,000 hour creep rupture strength at 700deg-C
(Figure 27). A piece of forged material, 1000 mm in diameter and weighing 7tons, has been made
successfully using TOS1X-I. TOS1X-II was developed from TOS1X-I by increasing the Al and Ti
content. TOS1X-II is expected to have around 200MPa of 100,000 hour creep rupture strength at
700deg-C. A 13 tons forged material of TOS1X-II has been made successfully (Figure 28). Figure
29 shows a similar welding of TOS1X-II.

Figure 28 TOS1X rotor material (Courtesy of Toshiba)

Figure 29 TOS1X rotor material (Courtesy of Toshiba)


Figure 30 shows materials casted for testing. At the beginning of the project three alloys, which
are Alloy740, Alloy 625, and Alloy 617, were casted into step blocks. The thickness of the casting
varies like steps. The maximum thickness resembles the thickness of the flanges of steam turbine
casings. We cut them into pieces to make specimens for material tests, such as the tensile test,
creep test, macro and micro observation and so on. Last year, we made castings of an inner casing,
valve and nozzle boxes to check the castability of actual parts. These are also cut into specimens
and tested.

37

Figure 30 Casted materials (Courtesy of Toshiba and Mitsubishi Heavy Industries)


The concept of 1000MW class steam turbine is shown in Figure 31. Five turbine casings and
rotors are aligned in tandem. From the left, there is a very high pressure turbine (VHPT), a high
pressure turbine (HPT), an intermediate pressure turbine (IPT) and two low pressure turbines
(LPT). Part of the rotors of VHPT, HPT and IPT consist of a newly developed Ni based material.

Figure 31 Concept of 1000MW class double reheat steam turbine (Courtesy of Toshiba)
BOILER COMPONENT TEST AND TURBINE ROTOR TEST
Today, we are preparing for the boiler component test and the turbine rotor test as shown on Fig.
32. In 2015 and 2016 boiler components such as super heaters, pipes, valves, and turbine casing
will be tested by using an actual boiler (Figure 33). Three rotors made of the three candidate rotor
materials will be tested in 700 deg-C atmosphere and at actual speed from 2014 to 2016 (Figure
34). The rotors will be heated by electric heaters in a vacuum chamber, and driven by an electric
motor.

38

2012

2013

2014

2015

2016

Boiler Component Test


Basic Design
Components Design
Components Prodution & Installation
Test

Turbine Rotor Test


Test Facility Design
Test Facility Production & Installtion
Test Rotors Production
Rotating Tests

Fig. 32 Schedule for boiler component test and turbine rotor test

Figure 33 Boiler component test facility

Figure 34 Turbine rotor test facility

39

CONCLUSION
The development of the USC technology began in Japan in the 80s, aiming at a cleaner and
more economical use of coal. Today, almost half of the coal power plants have a 600deg-C class
USC steam condition.
A-USC is one of the remarkable technologies being developed to reduce CO2 emissions from
fossil fuel power plants A large scale 9-year project began in 2008 to develop A-USC technology
thoroughly. Major Japanese manufacturers of boilers, steam turbines and valves and some
institutes are cooperating in the project to develop the technology. Five years from the beginning
of the project, we have made some boiler materials and carried out some fundamental tests, such
as tensile, creep, welding, bending etc. Turbine rotor and casing materials have been forged and
casted and are being tested as well.
Today, we are preparing for the boiler component test and turbine rotor test. In 2015 and 2016
boiler components such as superheaters, pipes, and valves will be tested in an actual boiler. Three
rotors made of the three candidate rotor materials will be tested in 700 deg-C atmosphere and at
actual speed from 2014 to 2016.
REFERENCES
[1] http://www.fepc.or.jp/about_us/pr/sonota/__icsFiles/afieldfile/2013/05/17/kouseihi_2012.pdf
[2] Shinya Imano, Jun Sato, Koji Kajikawa and Tatsuya Takahashi,'Mechanical Properties and
Manufacturability of Ni-Fe base Superalloy(FENIX-700) for A-USC Steam Turbine Rotor Large
Forgings', 5th international conference on advances in materials technology for fossilpower plants,
EPRI, 2007
Product names mentioned herein may be trademarks of their respective companies.

40

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

CURRENT STATUS OF THE U.S. DOE/OCDO A-USC MATERIALS


TECHNOLOGY RESEARCH AND DEVELOPMENT PROGRAM
J. Shingledecker
Electric Power Research Institute, Charlotte, NC USA
R. Purgert
Energy Industries of Ohio, Independence, OH USA
P. Rawls
National Energy Technology Laboratory, Pittsburgh, PA USA
ABSTRACT
The United States Department of Energy (U.S. DOE) Office of Fossil Energy and the Ohio Coal
Development Office (OCDO) have been the primary supporters of a U.S. effort to develop the
materials technology necessary to build and operate an advanced-ultrasupercritical (A-USC)
steam boiler and turbine with steam temperatures up to 760C (1400F). The program is made-up
of two consortia representing the U.S. boiler and steam turbine manufacturers (Alstom, Babcock
& Wilcox, Foster Wheeler, Riley Power, and GE Energy) and national laboratories (Oak Ridge
National Laboratory and the National Energy Technology Laboratory) led by the Energy
Industries of Ohio with the Electric Power Research Institute (EPRI) serving as the program
technical lead. Over 10 years, the program has conducted extensive laboratory testing, shop
fabrication studies, field corrosion tests, and design studies. Based on the successful development
and deployment of materials as part of this program, the Coal Utilization Research Council
(CURC) and EPRI roadmap has identified the need for further development of A-USC technology
as the cornerstone of a host of fossil energy systems and CO2 reduction strategies. This paper will
present some of the key consortium successes and ongoing materials research in light of the next
steps being developed to realize A-USC technology in the U.S. Key results include ASME Boiler
and Pressure Vessel Code acceptance of Inconel 740/740H (CC2702), the operation of the
worlds first 760C (1400F) steam corrosion test loop, and significant strides in turbine casting
and forging activities. An example of how utilization of materials designed for 760C (1400F)
can have advantages at 700C (1300F) will also be highlighted.
INTRODUCTION
Need for A-USC
In 2010, coal provided 48% of the electric generation in the United States [1] and 40.6% of the
electric generation in the world [2]. Coal is abundant with some studies estimating 150 years of
world-wide reserve at current consumption rates [3]. Despite its use as major source of electricity,
coal faces strong regulatory and economic challenges as the world adopts policies for reducing
carbon consumption. As world-wide demand for electricity grows, it is clear that a robust
portfolio of power generation options is needed to ensure reliable and environmentally
responsible electricity. To meet the challenges of the 21st century, advancements in coal
technology are needed. The Coal Utilization Research Council (CURC)-EPRI Coal Technology
Combustion roadmap, Figure 1, identifies key technologies needed to ensure that coal remains an
option for power generation in the future. A key aspect of this roadmap is the deployment of
41

higher efficiency pulverized coal combustion using Advanced Ultrasupercritical (A-USC)


technology. For the world, the International Energy Agency (IEA) has proposed, in its High
Efficiency Low Emission (HELE) roadmap for coal technology that coal generation from
inefficient subcritical plants be replaced by higher efficiency USC and A-USC plants as a first
step to carbon reduction, prior to commercial deployment of Carbon Capture and Storage (CCS)
technologies [4]. Thus, there are strong environmental and economic drivers for the development
and deployment of A-USC technology in the U.S. and worldwide.

Figure 1: CURC-EPRI Roadmap Combustion Timeline and Impact


National Programs
The world-wide development of A-USC technology aimed at steam temperatures of 700C
(1300F) and greater started initially around 1998 with a variety of European Projects [5]. In
2001, the U.S. Department of Energy in conjunction with the Ohio Coal Development Office
(OCDO) and cost share from all the major U.S. boiler and turbine original equipment
manufacturers (Alstom, B&W, Foster Wheel, Riley Power, GE), the Energy Industries of Ohio
(EIO), and the Electric Power Research Institute (EPRI) with support from Oak Ridge National
Laboratory (ORNL) and managed through the National Energy Technology Laboratory (NETL)
began an ambitious pre-competitive research and development project that would lead to higher
efficiency coal-fired power plants with reduced CO2 emissions [6,7]. Achieving major increases
in coal-fired power plant efficiency requires an increase in steam conditions (temperature and
pressure) and therefore demands the utilization of new materials and technologies to implement
these new materials. Hence, the R&D programs goal was to develop the materials technology
necessary to achieve Advanced Ultrasupercritical steam (A-USC) power plant steam conditions
up to temperatures and pressures of 760C (1400F) and 35 MPa (5000 psi), respectively which
can reduce all emissions, including CO2, by 20% or greater compared to todays U.S. fleet [8,9].
Beyond Europe and the United States, national programs now exist in Japan (initiated in 2008),
China (2011), and India (2012) with announced plans for demonstration in both China and India
around 2020 [10, 11, 12]. In the following sections, some of the successful developments of the
U.S. Program are highlighted which shows that materials and fabrication technologies now exist
to construct such a system.

42

ALLOY SELECTION
The consortium first defined the conditions required for the materials to operate at A-USC
conditions up to 760C (1400F). Based on these preliminary design studies, nickel-based alloys
were selected for detailed evaluation because only these alloys had the requisite strength required
to operate at A-USC conditions. Current USC boilers operate at temperature of approximately
600C which is the limit of the most advanced creep-strength enhanced ferritic (CSEF) steels, but
for temperatures above 700C (1300F), nickel-based alloys are clearly required. Figure 1 shows
the temperature required to produce creep-rupture in 100,000 hours for various alloys. With the
aim of 760C (1400F), age-hardenable alloys (Inconel alloy 740 and Haynes 282) were
considered beyond solution strengthened alloys such as 617. Using the criteria of 100,000 hour
strength at 100MPa (14.5ksi), these temperature limits can be observed from the plot. Although
strength is an important consideration, a number of other critical material properties, must be
considered for materials operating at A-USC conditions. Much of the work highlighted in the
following sections is related to the other important properties needed for boilers and turbines
including: fabricability, weldability, weld performance, tensile and fatigue properties, notch
sensitivity, and steam side oxidation and fireside corrosion resistance. Table 1 provides a list of
some alloys which were studied initially by the consortium and eliminated from future testing.
Table 2 provides a partial list of the key nickel-based alloys selected by the consortia for study
along with comments on their applicability and limitations.

Figure 2: 100,000 hour creep-rupture strength as a function of temperature for alloy/alloy class.

43

Table 1. Ferritic and Iron-based Austentic Alloys Selected for Early Evaluation and Some
Important Findings (Status)
Alloy
Component
Comments
SAVE12
Pipe
Unstable microstructure, welding was challenging [13]
(eliminated from testing)
Super 304H
SH/RH
Met strength projections, may need coatings in some
(CC2328)
environments and shot-peening for oxidation resistance [14]
(only for tubing)
HR6W
SH/RH,
Did not meet strength projections [15] (stopped research, new
Pipe
chemistry now available)
*SH/RH: Superheater and Reheater Tubing

Alloy
Haynes 230
CCA617

Alloy 263
Inconel
740/740H

Haynes 282

Waspalloy
Nimonic
105

Table 2. Nickel-Based Alloys Under Evaluation


Component
Comments
SH/RH, Pipe Successful welding trials, maximum size limitations for pipe may
limit applicability
SH/RH, Pipe
Higher strength than 617 but not enough data to change ASME
code stress values, not suitable for high sulfur coals, only
successful SMAW welds in nickel-based alloys, strain-age
cracking concerns, low strength limits applicability for turbine
rotor
Castings,
Back-up cast alloy to 282, good castability and weldability, lower
Rotor
strength but good ductility
SH/RH, Pipe Highest strength alloy in ASME B&PV code to enable A-USC up
to 760C (1400F), excellent fireside corrosion resistance,
successful fabrication and welding, prime candidate for boiler
components, cannot be air cast for valves and shells
Castings,
Higher creep strength than 740, relatively insensitive to starting
Rotor
microstructural condition, good forging window for rotor, can
be cast for valves and casings
Rotor, Bolts,
Back-up alloy with good turbine history Cannot be welded
Blades
reliability. Poor ductility
Bolts, Blades
Highest creep strength alloy. Only considered for bolting and
blading (non-welded components).

PROGRAM SUCESSES
Inconel Alloy 740/740H Development and Code Case Acceptance
Inconel Alloy 740 is an age-hardenable nickel-based alloy developed by Special Metals
Corporation (Huntington, WV USA) for use as superheater and reheater tubing in A-USC power
plants [16]. Due to its excellent high-temperature (creep) strength, see figure 2, and corrosion
resistance, the consortium also evaluated its use for thicker components such as boiler piping and
headers. This involved numerous welding and fabrication trials combined with long-term testing.
Initial challenges were encountered when welding the alloy in sections up to 75mm (3) in
thickness. The consortium worked closely with the alloy designers and other research institutions
to refine the alloy and weld metal composition. Revolutionary progress was made, and the alloy,
Inconel 740H, is now considered weldable as shown in Figure 3 with numerous successful welded
joints produced [17, 18].
44

Figure 3. 75mm (3) weldment in Inconel 740H Pipe (left) and Header mock-up (right) [17]
Materials used in the construction of fired pressure vessels must be designed to ASME B&PV
Code Section I. While there are some materials which are allowed for Section I construction at
760C (1400F), they do not have the requisite strength (allowable stresses) needed to design and
build an A-USC boiler. Thus, the consortium developed a comprehensive data package which
included test data on multiple material heats and product forms containing long-term data [19].
Babcock & Wilcox championed the case with supporting data from the other project members.
ORNL conducted the long-term testing and EPRI conducted the stress analysis. The approved
Section I code case, Code Case 2702, contains fabrication rules, welding specifications (including
weld strength reduction factors), stress allowables, and other key requirements. Users can now
design and specify the alloy for use in Section I construction to a maximum use temperature of
800C (1472F). The alloy has also been adopted by ASME Section B31.1 for power piping.
Creep-rupture testing has surpassed 45,000 hours with no drop-off in rupture strength or ductility.
Research has included studies on notch sensitivity and microstructural development which suggest
the alloy is suitable for long-term service at A-USC conditions [20, 21].
Fireside Corrosion Tests & Oxidation Behavior
Concerns with fireside corrosion at A-USC temperatures and the wide variety of coals burned in
the U.S., necessitated the consortium to do detailed evaluation of all the alloys fireside corrosion
resistance. The aim was to determine the suitability of the alloys over the range of temperatures
expected in superheaters and reheaters (SH./RH) in an A-USC plant and determine if higher
chromium claddings or coatings would be required. Extensive laboratory tests have been
conducted on a wide range of alloys, overlays, and coatings with varying sulfur contents [22], air
cooled probes have been fabricated and testing for up to 16,000 hours in actual boiler
environments [23], a steam-cooled corrosion test loop was operated in a high sulfur coal
environment [24], and the worlds first 760C (1400F) steam cooled loop is now operating in a
U.S. boiler. Figure 3 shows some results from one of the air-cooled probes. In this test, Inconel
740 performed as well as some of the high chromium weld overlays. Figure 4, shows the A-USC
programs steam loop which operates with a steam outlet of 760C (1400F) and contains a
variety of materials and overlays. Current plans are to remove the loop in 2014 for destructive
evaluation and comparison with other field and laboratory data. Work is also ongoing to evaluate
oxycombustion environments. The results show there are alloys and weld overlays which can be
used with confidence to enable an A-USC boiler with acceptable corrosion performance.

45

Figure 4. Cleaned surface of ring samples (identification at the top) from an air cooled probes
after 16,000 hours operation in a utility boiler. Note alloy 740 shows virtually no corrosion
wastage with similar performance to high chromium weld overlays (52 and 72WO). Other
alloys, including HR6W show moderate attack.

Figure 5. Steam-cooled A-USC corrosion test loop after installation.


The steam cooled-corrosion loops provide limited information on steam-side oxidation and
exfoliation. An extensive laboratory test program has been studying steam oxidation resistance
for all A-USC alloys including the use of shot-peening on stainless steels. Parabolic rate
constants, effect of pressure, weld oxidation, exfoliation and long-term (10,000 hour) exposures
are all being examined in a comprehensive test program involving multiple laboratories [14].
Boiler Fabrication Trials and Heavy-Section Welding
Extensive boiler fabrication studies have been completed which show it is possible to use typical
boiler fabrication processes to construct an A-USC boiler using nickel-based alloys [25, 26, 27].
In some cases, flux processes were not possible for heavy section nickel-based piping and gas
tungsten arc (GTAW) and gas metal arc (GMAW) welding was utilized. Hot wire GTAW (also
known as hot wire TIG) to enable higher deposition rates has been successfully demonstrated on
alloys including Inconel 740H in section thickness up to 75mm (3) including narrow groove
configurations to improve productivity (see Figure 3). Long-term weldments creep tests [30] are
being conducted to establish weld strength reduction factors (WSRF) and ongoing work includes
strain-age cracking studies. Figure 6 shows demonstration articles which include: forming,
46

bending, welding (pipe and tube), machining, application of overlays, and dissimilar metal
welding applied to A-USC alloys.

Figure 6. A-USC Boiler Demonstration Articles


Casting Developments and Scale-Up
Steam turbine casing and valve bodies are traditionally made via casting. While previous
European efforts had had some success with casting solid solution strengthened nickel-based
alloys, the use of higher strength precipitation hardened alloys with high Ti and Al contents
(which are very reactive in air) needed to be explored to see if conventional air/protective
atmosphere casting techniques could be used. A research team from NETL-Albany and ORNL
cast a large number of trial alloys and subject them to destructive evaluation and testing [28].
Based on successful results, scale up of the most promising alloys, Haynes 282 and Nimonic 263
was started with two different vendors. To date, static sand cast test blocks, step blocks, and
centrifugal castings have been made and are being evaluated. Planning is now underway for
larger pours ~2700kg (6,000lb) in valve geometries. Figure 7 shows a ~450kg (1000lb) finished
weight step casting of cast Haynes 282. Modeling work includes casting simulations to determine
suitability for pouring larger casing geometries and thermodynamic/diffusion modeling to develop
appropriate homogenization heat-treatment cycles [29]. Weld repair studies, microstructural
evaluation and mechanical property determination are all ongoing.

Figure 7. ~450kg (1,000 lb) finished weight step casting of Haynes 282
47

Haynes 282 Forging


Scoping studies by the A-USC Turbine Consortium identified Haynes 282 as the most promising
turbine rotor/disc alloy. The studies showed that creep behavior of Haynes 282 was relatively
insensitive to starting microstructural condition and had adequate tensile and fatigue behavior for
a 760C (1400F) rotor [30]. The current HP and IP A-USC turbine design being considered
calls for a bolted rotor similar to an industrial gas turbine [31]. In such a design, the highesttemperature component is a forged disk. Prior to forging, ingots of Haynes 282, Figure 8, were
produced via triple melting (VIM/ESR/VAR). A first ingot was cast with a weight of ~4500kg
(10,000lbs) and sectioned to evaluate defects and effects of processing variations. A similarly
sized second ingot was also cast and will be forged into a rotor disc for full property evaluation.
Extensive mechanical property studies are planned for this full-scale component.

Figure 8. Upset and draw operation (left) and worlds first Haynes 282 triple melt ingot (right)
Economy of 760C Materials at 700C
The U.S. programs temperature aim of 760C (1400F) steam is greater than the other programs
worldwide focused on A-USC at 700C (1300F). This has lead to the extensive study of alloys
Inconel 740H and Haynes 282. A critical advantage of these alloys is that they can be used more
cost effectively than solid solution strengthened nickel-based alloys at 700C (1300F) due to
their higher strength at lower temperatures with similar alloy cost. One specific study [32]
compared alloy 617 to 740H for a prototypical A-USC main steam and hot reheat piping system
and showed that by using 740H in lieu of 617, the number of pipes required would decrease, the
length of each individual extruded pipe section would increase, the number of welds for the
piping system would decrease, the pipe thickness would decrease, and the amount of welding
(both filler material and welding time) would decrease. The combined effect of these results
demonstrated that the utilization of 740H over 617 would result in a significant cost savings with
regard to both construction and life management of an A-USC piping system at 700C (1300F).
Figure 9 is a comparison of the piping system in that study. Of interest is that for the main steam
pipe, two 617 pipes were required because the wall thickness of the 617 in a single pipe
arrangement exceeded typical boiler fabrication sizes (thickness was greater than 100mm).
Additionally, the processing characteristics for extrusion (the basis of this study) showed 740H
could be extruded in large diameter thus only one reheat pipe was needed compared to two for the
617 system. Depending on alloy cost, the materials savings alone for using 740H in lieu of 617
was 5 to 20 million USD.

48

Figure 9. Comparison of a 617 and 740H piping system for a 700C (1300F) steam plant. The
617 system requires two main steam and two hot-reheat pipes to avoid wall thickness in excess of
75mm and to allow for extrusion. In contrast only one main steam pipe and one reheat pipe is
needed for the 740H system due to its excellent strength and good fabrication characteristics.
CONCLUSIONS AND NEXT STEPS
Revolutionary progress has been made in advancing the materials technology to enable a 760C
(1400F) A-USC power plant. These successes have only been made possible through a unique
cost-sharing U.S. Consortium of manufacturers and research organizations dedicated achieving to
the aggressive goals set forth by the U.S. DOE. The U.S. DOE/OCDO A-USC Consortium has
had many successes including: demonstration of welding and fabrication of nickel-based alloys
for an A-USC plant, extensive successful fireside corrosion testing including in-plant operation of
the worlds first 760C (1400F) steam cooled corrosion test loop, development of casting
techniques for high-alloy age-hardneable alloys, new materials for rotor forgings, and code
acceptance of Inconel 740H which gives manufacturers, for the first time, an alloy with the
requisite strength for a 760C (1400F) plant. Current research includes scale-up, demonstration,
testing, and weld repair of castings, testing and evaluation of a Haynes 282 forged disc, long-term
creep testing in excess of 45,000 hours to ensure long-term reliability, field corrosion studies, and
economic evaluations. A new project to build a 760C (1400F) component test facility is now
underway. This facility will be unique in that it will test heavy section components 60C (100F)
hotter than any other test facility in the world, will incorporate cyclic operation of welds and
valves, and will include a turbine test. Figure 9 shows a schematic of this planned facility. The
facility plan is an outcome of workshops held with utility stakeholders who desire to see a
demonstration before construction of an A-USC plant. In summary, the materials technology to
design, build, and operate an A-USC power plant now exists or the research is nearing
completion. Current research and development is ongoing and planned to improve the economy of
building such a plant and to ensure its reliable operation.

49

Figure 10. General arrangement of the planned 760C (1400F) component test facility
ACKNOWLEDGMENTS
The authors are greatly indebted the members of the A-USC consortium who have undertaken this
endeavor. First and foremost appreciation is given to Prof. R (Vis) Viswanathan for his
leadership of the project prior to his retirement. Special thanks are given to the many members of
the consortium past and present including but not limited to: H. Hendrix (EPRI), R. Ganta, J.
Pschirer, and J. Marrion (ALSTOM Power), J. Tanzosh (B&W), H. Hack (Foster Wheeler), D.
Saha and J. Breznak (GE), P. Torterelli (ORNL), J. Hawk and P. Jablonski (NETL-Albany), and
B. Vitalis (Riley). The support and guidance of our sponsors is also greatly acknowledged: B.
Romanosky (NETL), R. Conrad (DOE), and Ohio Office of Development.
REFERENCES
[1] The CURC-EPRI Coal Technology Roadmap. August 2012: Update. Available at:
www.coal.org
[2] International Energy Agency OECD/IEA, 2012. Available at: www.iea.org
[3] B. Diczfalusy. HELE Coal Technology Roadmap. IEA Clean Coal Centre Workshop: on
Advanced ultrasupercritical coal-fired Power Plants, Vienna, Austria 12-20 September
2012.
[4] K. Burnard, O. Ito. HELE Coal Technology Roadmap. IEA Clean Coal Centre Workshop:
on Advanced ultrasupercritical coal-fired Power Plants, Vienna, Austria 12-20 September
2012.
[5] R. Blum, R. Vanstone. Materials Development for Boilers and Steam Turbines Operating at
700C, Proceedings to the 6th International Charles Parsons Materials Conference, Dublin,
Ireland (2003).
[6] R. Viswanathan, R. Purgert, U. Rao: Materials for Advanced Power Engineering 2002,
Proceedings Part II, Forschungszentrum Julich GmbH, 2002. 1109-1129
50

[7] U.S Department of Energy and Ohio Coal Development Office Advanced Ultra-Supercritical
Materials Project for Boiler and Steam Turbines. EPRI, Palo Alto, CA: March 2011.
1022770.Available: http://my.epri.com/portal/server.pt?Abstract_id=000000000001022770
[8] R. Viswananthan, J.F. Henry, J. Tanzosh, G. Stanko, J. Shingledecker, B. Vitalis, R. Purgert.
U.S. Program on Materials Technology for Ultra-Supercritical Coal Power Plants. Journal
of Materials Engineering and Performance. Vol. 14 (3) June 2005. 281-292.
[9] J.P. Shingledecker, I.G. Wright. Evaluation of the Materials Technology Required for a
760C Power Steam Boiler. Proceedings to the 8th Liege Conference on Materials for
Advanced Power Engineering 2006. Forschungszentrum Jlich GmbH (2006) pp. 107-120.
[10] F. Masuyama. R&D Program for A-USC Material Development with Creep
Strength/Degradation Assessment Studies. Proceedings to the 6th International Conference
on Advances in Materials Technology for Fossil Power Plants, Santa Fe, New Mexico,
August 30-September 4, 2010. EPRI, March 2011: 1022300. Distributed by ASM
International. 11-29.
[11] R. Sun. Development of 700C USC Coal-Fired Power Generation Technology in China. IEA
Clean Coal Centre Workshop: on Advanced ultrasupercritical coal-fired Power Plants,
Vienna, Austria 12-20 September 2012.
[12] Personal communication: A. Mathur, O.P. Bhutani, BHEL February 2013.
[13] M. Santella et al. Weld Development for Heavy-Section SAVE12. Presentation at the 5th
International Conference on Advances in Materials Technology for Fossil Power Plants,
Marco Island, FL USA Oct. 3-5, 2007.
[14] R. Viswanathan, J. Sarver, J. Tanzosh. Boiler Materials for Ultra-Supercrtical Coal Power
Plants Steamside Oxidation. Journal of Materials Engineering and Performance. Vol 15
(3), June 2006. 255-274
[15] J.P. Shingledecker. Creep rupture behaviour and recrystallisation in HR6W and Haynes
alloy 230 cold bent boiler tubing for ultrasupercritical steam boiler applications. Energy
Materials, Vol. 2, No. 4, 2008. 235-240.
[16] G.D. Smith, H.W. Sizek. Introduction of an Advanced Superheater Alloy for Coal-Fired
Boilers. Corrosion 2000. Paper 00256 2000 NACE International, TX.
[17] B.A. Baker, et al. Welding and Repair Technology for Power Plants, Tenth International
EPRI Conference. June 26-29, 2012 Marco Island, FL USA
[18] J.A. Siefert, J.M. Tanzosh, J.E. Ramirez. Proceedings to the 6th International Conference on
Advances in Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August
30-September 4, 2010. EPRI, March 2011: 1022300. Distributed by ASM International.
1045-1066.
[19] J.P. Shingledecker, N.D. Evans, G.M. Pharr. Influences of Composition and Grain Size on
Creep-Rupture Behavior of Inconel Alloy 740. Materials Science and Engineering A. 578
(2013) 277-286. 10.1016/j.msea.2013.04.087
[20] J.P. Shingledecker, G.M. Pharr. The Role of Eta Phase Formation on the Creep Strength and
Ductility of INCONEL Alloy 740 at 1023 K (750 C). Metallurgical and Materials
Transactions A., Vol. 43, Issue 6 (2012). 1902-1910. 10.1007/s11661-011-1013-4
[21] J.P. Shingledecker, G.M. Pharr. Testing and Analysis of Full-Scale Creep-Rupture
Experiments on Inconel Alloy 740 Cold-Formed Tubing.Journal of Materials Engineering
and Performance. Volume 22, Issue 2 (2013), Page 454-462. 10.1007/s11665-012-0274-4
[22] H. Hack, G. S. Stanko, Update on Fireside Corrosion Resistance of Advanced Materials for
Ultra-Supercritical Coal-Fired Power Plants, The 31st International Technical Conference
on Coal Utilization & Fuel Systems, May 21-25, 2006, Clearwater, Florida, USA.
[23] M. Gagliano, H. Hack, G. Stanko. Update on the Fireside Corrosion Resistance of Proposed
Advanced Ultrasupercritical Superheater and Reheater Materials: Laboratory and Field Test

51

Results. The 2009 Clearwater Coal Conference, 34th International Technical Conference on
Coal Utilitzation & Fuel Systems. Clearwater, FL May 31-June 4, 2009.
[24] E. Robitz, J. Tanzosh. In Situ Corrosion Testing of Ultrasupercritical Tube and Weld
Overlay Materials. Proceedings to the 6th International Conference on Advances in
Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August 30-September
4, 2010. EPRI, March 2011: 1022300. Distributed by ASM International. 288-302.
[25] W.R. Mohn, J.M. Tanzosh. Considerations in Fabricating USC Boiler Components from
Advanced High Temperature Materials. Proceedings to the Fourth International
Conference on Advances in Materials Technology for Fossil Power Plants (Hilton Head, SC,
Oct. 25-28, 2004). ASM-International, Materials Park, OH, 2005. 165-176.
[26] B.A. Baker, R.D. Gollihue. Proceedings to the 6th International Conference on Advances in
Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August 30-September
4, 2010. EPRI, March 2011: 1022300. Distributed by ASM International. 96-109.
[27] M. Borden. Weldability fo Materials for Ultrasupercritical Boiler Applications.
Proceedings to the Fourth International Conference on Advances in Materials Technology
for Fossil Power Plants (Hilton Head, SC, Oct. 25-28, 2004). ASM-International, Materials
Park, OH, 2005. 837-854.
[28] P.J. Maziasz et al. High Temperature Mechanical Properties and Microstructure of Cast NiBased Superalloys for Steam Turbine Casting Applications. 6th International Conference
on Advances in Materials Technology for Fossil Power Plants, Santa Fe, New Mexico,
August 30-September 4, 2010.
[29] P. D. Jablonski, C. J. Cowen, "Homogenizing a Nickel-Based Superalloy: Thermodynamic
and Kinetic Simulation and Experimental Results", Met Trans 40B 182 (2009).
[30] R. Viswanathan, J. Shingledecker, J. Hawk, S. Goodstine. Effect of Creep in Advanced
Materials for Use in Ultrasupercritical Power Plants. Proceedings: Creep & Fracture in
High Temperature Components, 2nd ECCC Creep Conference, April 21-23, 2009, Zurich,
Switzerland. 2009 DEStech Publications, Inc. 31-43.
[31] R Schwant, C. Shen, M. Soare. New Materials Enable Unprecedented Improvement in
Turbine Performance. Advanced Materials & Processes, January 2013. ASM
International. 18-22.
[32] J. Shingledecker et al. U.S. Program on Advanced Ultrasupercritical Power Plant Materials
The Economy of Using Advanced Alloys. Proceedings to the IEA Clean Coal Centre
Workshop: Advanced ultrasupercritical coal-fired power plants. Vienna, Austria, 19-20
Sept. 2012.
LEGAL NOTICE/DISCLAIMER
This report was prepared by J. Shingledecker (EPRI) pursuant to a Grant partially funded by the
U.S. Department of Energy (DOE) under Instrument Number DE-FG26-0 1 NT 41175 and the
Ohio Coal Development Office/Ohio Department of Development (OCDO) under Grant
Agreement Number CDO/D-OO-20 (now D-05-02A). NO WARRANTY OR
REPRESENTATION, EXPRESS OR IMPLIED, IS MADE WITH RESPECT TO THE
ACCURACY, COMPLETENESS, AND/OR USEFULNESS OF INFORMATION
CONTAINED IN THIS REPORT. FURTHER, NO WARRANTY OR REPRESENTATION,
EXPRESS OR IMPLIED, IS MADE THAT THE USE OF ANY INFORMATION,
APPARATUS, METHOD, OR PROCESS DISCLOSED IN THIS REPORT WILL NOT
INFRINGE UPON PRIVATELY OWNED RIGHTS. FINALLY, NO LIABILITY IS
ASSUMED WITH RESPECT TO THE USE OF, OR FOR DAMAGES RESULTING FROM
THE USE OF, ANY INFORMATION, APPARATUS, METHOD OR PROCESS DISCLOSED
IN THIS REPORT

52

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

INDIA'S NATIONAL A-USC MISSION - PLAN AND PROGRESS


Alok Mathur
Bharat Heavy Electricals Limited, New Delhi, India
O.P Bhutani
Bharat Heavy Electricals Limited, New Delhi, India
Dr. T. Jayakumar
Indira Gandhi Centre for Atomic Research, Kalpakkam, India
D.K. Dubey
National Thermal Power Corporation, New Delhi, India
S.C. Chetal
Office of Principal Scientific Adviser to Govt. of India, New Delhi, India
ABSTRACT
India's current installed power generating capacity is about 225,000 MW, of which about
59% is coal based. It is projected that India would require an installed capacity of over
800,000 MW by 2032. Coal is likely to remain the predominant source of energy in India
till the middle of the century. India is also committed to reducing the CO2 emission
intensity of its economy and has drawn up a National Action Plan for Climate Change,
which, inter alia, lays emphasis on the deployment of clean coal technologies. With this
backdrop, a National Mission for the Development of Advanced Ultra Supercritical
Technology has been initiated. The Mission objectives include development of advanced
high temperature materials, manufacturing technologies and design of equipment. A
corrosion test loop in an existing plant is also proposed. Based on the technology
developed, an 800 MW Demonstration A-USC plant will be established. Steam
2
parameters of 310 kg/cm , 710 C / 720 C have been selected. Work on selection of
materials, manufacture of tubes, welding trials and design of components has been
initiated. The paper gives details of India's A-USC program and the progress achieved.
INTRODUCTION
Indias current power generating capacity is about 225,000 MW, which is predominantly
coal based. About 59% of generation capacity is coal fired, 18% hydro, 2% nuclear, 9%
natural gas and 12% renewables. Although the installed capacity has been increasing
rapidly in the last few years, there is an overall shortage of electricity in relation to the
demand. With expected economic growth of over 6% per annum in the long term, the
demand for electric power is likely to shoot up at a much faster rate. According to the
Integrated Energy Policy adopted by the Government of India in 2008, the demand for
electric power is projected to be about 800,000 MW in 2032. Coal is likely to remain the
major source of energy for power generation.
.
India has also adopted a National Action Plan for Climate Change (NAPCC) which
mandates a reduction in the CO2 intensity of the Indian economy. This implies that CO2
emissions resulting from coal combustion in power plants have to be substantially
lowered. The bulk of Indias utility power generating capacity comprises pulverized coal
fired subcritical thermal power plants of 600 MW, 500 MW, 250 MW and 210/ 200 MW
53

ratings, in addition to older plants of lower ratings. In recent years, supercritical power
plants of 660 MW and 800 MW have been established and are in operation. The steam
2
parameters of the supercritical plants being set up currently in India are 247 kg/cm ,
565 C/ 593 C. Future capacity additions will be mainly in the supercritical range. Ultra
supercritical plants, with steam temperatures over 600 C, are also on the anvil.
In the medium to long term, however, India needs to adopt clean coal technologies that
promise higher efficiencies, reduced coal consumption per unit of power produced and
lower emissions of CO2 and pollutants such as SOx, NOx, etc. Advanced Ultra
Supercritical (A-USC) is one of the most promising of such clean coal technologies,
currently under development in several countries. Among major economies, the drivers
for adopting A-USC technology are perhaps the strongest in India.
With the above backdrop, the Government of India has initiated a National Mission for
Development of Clean Coal (Carbon) Technologies, under which there is a Sub-Mission
for the Development of Advanced Ultra Supercritical Technologies for Thermal Power
Plants. The National Mission is being executed under the overall leadership of the
Principal Scientific Adviser (PSA) to the Government of India. The Mission has a special
focus on achieving a high degree of self reliance in technology apart from economic and
low cost production through indigenization.
The objective of Indias National Mission for the Development of Advanced Ultra
Supercritical Technology for Thermal Power Plants is to develop all the technologies
required for an A-USC plant and to design, manufacture and establish an 800 MW AUSC Demonstration Power Plant. After successful development and demonstration,
there is enormous potential to set up a large number of coal fired power plants based on
the A-USC technology developed.
The development work is being executed by a consortium of three organisations: Bharat
Heavy Electricals Limited (BHEL), the leading power equipment manufacturer of India;
Indira Gandhi Centre for Atomic Research (IGCAR), a premier R&D institution
specializing in the development of high temperature materials; and NTPC Limited
(NTPC), Indias largest power generation utility. Together, the three organizations have
the capabilities required to develop and establish an A-USC power plant. In addition,
wherever necessary, the help of R&D institutions and other organizations is sought.
The time frame for the development is seven years from the date of approval of the
project by the Government of India - comprising 2 years for development and 4
years for setting up the demonstration plant. A comprehensive and detailed roadmap of
the development activities across the entire spectrum, covering materials selection and
development, manufacturing technologies, corrosion test loop, equipment design,
thermal cycle design and overall plant design, has been prepared. Well before the
formal start of the R&D project, several proactive R&D initiatives have been taken to
gain a head start in critical areas of technology and accelerate the pace of development
work and significant progress has been achieved.
DEVELOPMENT AREAS TAKEN UP AND PROGRESS ACHIEVED
The following sections give details of the major developmental activities that have been
completed or are in progress.

54

Materials Selection and Development


The materials selected for use in the high temperature zones of the boiler are SS
304HCu and Alloy 617M. The selection is based on the following criteria:

The materials are included in the ASME code/ code cases


Fairly good commercial availability
Experience of use in various applications
Availability of material properties for design

The specific composition of the materials to be used in the Indian program has been
selected to lie in a narrow zone within the overall material specifications of ASME, in
order to achieve the desired properties and reduce variability.
Development of Processing Technologies
In order to gain experience in processing and using these materials, forged ingots of SS
304HCu and Alloy 617M, starting from the basic ingredients, have been manufactured
on a pilot scale. The materials were characterized at each stage of the production
process and the process was optimized to achieve the specifications and the desired
properties. The manufacturing process has been stabilized and is capable of being
scaled up.
Boiler tubes of SS 304HCu and Alloy 617M have also been manufactured on a pilot
scale, starting from the forged ingots developed earlier. A detailed stage wise
characterisation approach was adopted for the development of tubes of SS 304HCu and
Alloy 617M in order to ensure that the process of manufacture was optimized to obtain
the specified characteristics. This approach helped to overcome problems faced during
the initial trials. The series of heat treatments and characterizations at various stages of
tube manufacture helped in understanding the materials behavior towards mechanical
working and heat treatment. Based on these metallurgical inputs, the process flow
sheets were evolved and optimized.
Materials Testing
A comprehensive materials testing program covering long term creep tests and hot
corrosion for the parent metal, tubes and welded joints, both similar and dis-similar, has
been put in place. The tests being carried out are briefly described below:
Creep Tests
Long term creep tests on Alloy 617M at 650, 700, 750, 800 C, with stress 100 MPa to
360 MPa, and test durations ranging from 100 hours to 10,000 hours are in progress.
Creep tests on Alloy 617M welded with Alloy 617 filler material are being carried out at
650, 700, 750, 800 C, with stress 100 MPa to 360 MPa, and test durations ranging from
100 hours to 1200 hours. Tensile and hot tensile (600-800 C) impact tests are also
being carried out.
Creep tests on Super 304HCu tubes welded with Alloy 617M and Alloy 625 filler material
at 600, 700, 750 C, and stress 125 MPa to 275 MPa, with test durations ranging from
100 hours to 1,500 hours are in progress. Creep tests on parent metal at 600, 650,

55

700C, with stress 100 MPa to 360 MPa, and test durations ranging from 100 hours to
2,000 hours are being carried out.
Creep tests are also being carried out on dissimilar welds of Alloy 617M with SS
304HCu using Alloy 617 filler material welds at 600, 650 and 700 C, and stress 180
MPa to 280 MPa, with test durations ranging from 100 hours to 3,000 hours.
Hot Corrosion Tests
Hot corrosion tests in a simulated flue gas environment have been carried out on Alloy
617M and SS 304HCu tubes in the laboratory at 600, 700, 800 C for up to 1000 hours
duration.
Hot Corrosion Test Loop
The materials chosen have been extensively tested elsewhere for hot corrosion. To
check for hot corrosion with firing of Indian coal, a corrosion test loop is planned to be
installed at an existing power plant in India.
The test loop comprises tubes of T91, SS 304HCu and Alloy 617M. The steam flow is
taken from the main steam line at a temperature of about 540 C. In the test loop, the
steam is heated to a temperature of 720 C. After the test loop, the steam pressure is
reduced to hot reheat pressure in a multistage steam regulation valve, which regulates
the steam quantity to achieve the steam temperature of 720. The steam from the
regulating valve is mixed in a mixing piece with relatively cold steam at about 390 C,
taken from the Low Temperature Superheater (LTSH) outlet link, to get steam at a
temperature of approximately 540 C, in order to return the steam flow into the hot
reheat pipeline. The test loop is likely to be installed in 2014 and is planned to be in
operation for at least two years.
Manufacturing Technologies
To gain hands-on experience in manufacturing with high temperature materials, trials for
all the operations required during the manufacture of boilers, steam turbines and other
power plant equipment, are being carried out. Some of these are briefly mentioned
below.
Welding of Tubes and Plates
Welding was carried out on tubes and plates using various processes like Gas Tungsten
Arc Welding (GTAW), Hot Wire GTAW, semi-automated GTAW on Alloy 617M. The
trials were successful and it was possible to get consistent quality of the butt welds in all
the cases. The process parameters have been finalized for the welding with the
available materials. Some dissimilar welding combinations were also tried out and the
outcome was satisfactory. Figure 1 shows a sample tube to tube butt weld joint.

56

Figure 1: Tube to Tube Butt Weld with Hot Wire GTAW


Forming
Tube bending is another area of interest, especially for the fabrication of high
temperature superheater coils. The forming trials on the tubes involved bending the
tubes to different radii. The bends gave good results, with a controlled level of ovality
and thinning. There were no cracks or any significant indications on the entire bend. The
bending trials were conducted on a Pine 150 CNC tube bending machine. Figure 2
shows Alloy 617M tubes cold bent to R/D 1,2 and 3.

Figure 2: Alloy 617M Tubes Cold Bent to R/D 1, 2 and 3


Studies on Critical Aspects of Technology
Studies on various critical aspects of the technology have been initiated. These include
the following:
Thick Walled Superheater Header
Manufacture of thick walled headers, such as superheater headers, involves challenges
in manufacturing, especially in welding, as failures in welding of thick sections have
been reported in the literature. It is, therefore, important to study in detail and optimize
the welding technologies for thick walled headers.
A mock up superheater header made of Alloy 617M is being fabricated. The header will
be similar to the final superheater header assembly in terms of diameter and thickness
but of shorter length. Several developmental tasks will be undertaken, including
development of end covers, establishment of circumferential seam welding procedures
between end cover and pipe, establishment of fillet welding process for stub to pipe
57

welding and establishment of advanced ultrasonic testing techniques for the detection of
discontinuities in butt and fillet welds.
HP Bypass Valve
Development of valves to suit the pressures and temperatures encountered in A-USC
boilers has been identified as a critical area, due to the complexity of valve design. The
High Pressure Bypass valve for high pressure steam turbine is one such valve, which is
subjected to A-USC steam parameters at the inlet to the HP turbine. The design and
manufacture of an HP Bypass valve for A-USC parameters is therefore being taken up.
Several developmental tasks will be undertaken, including development of materials for
body and other components, stem, hard facing, fasteners, and packings of the valve,
establishment of body spherical and other machining, stem milling, deep hole drilling,
destructive and non-destructive testing procedures, establishment of body-nozzle
welding, seat hard facing, body assembly-seat welding procedures, and establishment
of advanced ultrasonic techniques for detection of discontinuities in welds.
Dis-similar Weld Joint of Turbine Rotor
Development of steam turbine rotor forgings of advanced high temperature nickel base
materials is an important area to be addressed while considering the 710C / 720C
steam cycle. Considering that a monolithic rotor of Alloy 617M would be expensive, a
welded rotor is being considered. The rotor would have an Alloy 617M forging for the
high temperature portion of the HP/IP rotor, with a dis-similar metal weld joint between
the cylindrical forgings of Alloy 617M and 10% Cr steel. Operating experience on dissimilar metal weld joints is not very satisfactory. The joining technique therefore needs to
be developed and tested thoroughly. To address the main issues related to welded
turbine rotor manufacture, laboratory scale studies to demonstrate the weldability of
large size forgings of Alloy 617M to 10% Cr steel and optimization of weld procedures
and welding consumables, while maintaining the properties of the forgings, are being
undertaken.
Technical specifications for the Alloy 617M rotor forgings suitable for use in the 800 MW
A-USC plant, developing the supply base for manufacture of scaled down Alloy 617M
cylindrical forgings, procedure development for narrow-gap tungsten inert gas (NG-TIG)
welding of dissimilar metal weld joints, finalizing the associated non-destructive testing
methodologies, and characterization of the short-term and long-term mechanical
properties of the Alloy 617M/ 10%Cr steel joint are some of the outcomes of the work.
Blade Profiles for HP Turbine
Design of steam flow path and blade profiles for A-USC parameters are important
aspects of steam turbine design as blades using advanced high temperature alloys are
being designed for the first time. A project to develop the blade profiles for A-USC
parameters is under way and is nearing completion. Trial manufacture of steam turbine
blades with Alloy 617M is also being undertaken as part of the project, to ensure that the
manufactured blades satisfy all the specifications. The profile design is also being
verified through Computational Fluid Dynamic (CFD) analysis and also through physical
cascade testing in the laboratory.

58

Thermal Cycle
The target efficiency of the A-USC plant has been set as 46% on HHV basis, under
Indian environmental conditions (cooling water temperature of 33 C), and with Indian
coal firing. This gives a reduction in CO2 emissions of about 11% compared with a
supercritical plant of the same rating. The thermal cycle of the plant has been designed
after a series of techno economic studies to achieve the most economical and efficient
configuration.
Conceptual Design of Boiler
Several techno-economic studies were carried out to decide the optimum configuration
of the boiler. A two pass, single reheat boiler configuration with tilting tangential firing
has been selected.
Tubes of SS304HCu and Alloy 617M are used in the high temperature zones in the
superheater and reheater. Some new features to maximize the utilization of heat are
also included.
CONCLUSION
A National Mission for the development of A-USC technology for thermal power plants
has been initiated by the Government of India, with BHEL, IGCAR and NTPC
constituting the core team. The objective is to establish an 800 MW A-USC
demonstration power plant in a time frame of seven years. A comprehensive plan of
action has been prepared for developing the entire range of technologies necessary.
Several proactive initiatives have been taken to gain a lead time and accelerate the work
before the formal start of the project. There is a strong focus on achieving self reliance in
technology and low cost manufacturing through indigenization.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the help provided by the Office of the Principal
Scientific Adviser to the Government of India and the managements of BHEL, IGCAR
and NTPC in carrying out the work described in this paper and for permitting them to
present this paper. The authors also acknowledge the enormous efforts of the teams in
each organization who carried out the work.

59

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

ADVANTAGES OF A-USC FOR CO2 CAPTURE IN PULVERIZED


COAL UNITS
H.L. Hendrix
Electric Power Research Institute, Charlotte, NC USA
ABSTRACT
Increasing the steam temperature of a coal-fired pulverized coal (PC) power plant increases its
efficiency, which decreases the amount of coal required per MW of electrical output and
therefore decreases the emissions from the plant, including CO2. However, increasing the steam
temperature requires that the materials for the boiler pressure parts and steam turbine be upgraded
to high-nickel alloys that are more expensive than alloys typically used in existing PC units. This
paper explores the economics of A-USC units operating between 595oC and 760oC (1100oF to
1400oF) with no CO2 removal and with partial capture of CO2 at an emission limit of 454 kg
CO2/MW-hr (1000 lb CO2/MW-hr) on a gross power basis. The goal of the paper is to
understand if the improved efficiency of A-USC would reduce the cost of electricity compared to
conventional ultra-supercritical units, and estimate the economically optimal steam temperature
with and without CO2 removal.
INTRODUCTION & BACKGROUND
Over the history of power generation, the thermal efficiency of PC power plants has improved as
steam temperatures and pressures have increased. There are many benefits to increased thermal
efficiency including the decrease in operating costs associated with purchasing coal, limestone for
the flue gas desulfurization (FGD) unit, ammonia for the selective catalytic reduction (SCR) unit,
reduced CO2 emissions, and reduced water consumption. The higher efficiency will decrease the
expense of all of these variable operating costs and allow the power plant to be dispatched earlier,
which will allow it to be operated primarily as a base-loaded plant.
Ultra supercritical steam (USC) conditions are roughly defined at having temperatures in excess
of 593oC (1100oF). Advanced ultra-supercritical (A-USC) steam conditions are at temperatures
above that of USC, typically in the range of 705-760oC (13001400oF). The maximum steam
temperature achievable using currently available ferritic steels is 620oC (1148oF). Utilizing
higher steam temperatures requires a transition to high-nickel alloys. The U.S. Department of
Energy (DOE) and the Ohio Coal Development Office (OCDO) began a research program in
2001 for the development and certification of these alloys. The research program formed a
consortium of companies and organizations which currently includes DOE/NETL, OCDO,
Energy Industries of Ohio (EIO), EPRI, ALSTOM Power, Babcock and Wilcox, Foster Wheeler,
General Electric, Oak Ridge National Laboratory and Riley Power, Inc. One of the recent
achievements for the consortium was the successful approval of Inconel Alloy 740 by the ASME
B&PV Code for Section 1 and B31.1 in 2011 and 2012. This material is rated for continuous
operation at steam conditions of up to 800oC (1472oF).
The business driver for operating at the higher efficiency of an A-USC power plant is largely
dependent on the geographic region. For example, as fuel costs are higher in Asia than the USA,
improved efficiency to lower fuel usage results in greater cost savings and a correspondingly
lower cost of electricity.
In the U.S. and Europe, one of the primary drivers is the likelihood that there will be limitations
on the emissions of CO2. In March 2012, the U.S. Environmental Protection Agency issued a
Standard of Performance that would limit CO2 emissions on new fossil power plants to 454 kg
60

CO2/MW-hr (1000 lb CO2/MW-hr) on a gross power basis. This Standard of Performance is


currently in a period of comment and review, but if it becomes an emission regulation then all
new coal fired power plants in the U.S. would require some form of CO2 capture. CO2 emission
regulations or a tax on CO2 emissions are already in place or have been proposed in Canada, the
United Kingdom, South Korea, Australia, Japan, and India.
A key approach to lowering all emissions from a PC power plant is increasing the efficiency of
the generating unit. Raising the efficiency decreases the amount of coal required to generate a
given output, which decreases the CO2 emissions from the unit. With less CO2 being generated,
the energy penalty associated with adding CO2 removal will be less.
EPRI has been evaluating the performance and economics of USC and A-USC power plants with
and without PCC for several years [1-5]. These projects have evaluated CO2 removal at 90% and
at a level equivalent to a natural gas fired combined cycle power plant. One conclusion from
these reports is that as the steam cycle temperature increases, the cost of the coal system, the Air
Quality Control Systems (AQCS) (particulate removal, flue gas desulfurization (FGD). NOx
control) and the PCC system decreases as the overall efficiency increases. However, as the steam
temperature increases, the quantity of high-nickel alloys required in the boiler, steam piping, and
steam turbine increases. These alloys are substantially more expensive than carbon steel or
advanced ferritic alloys and increase the cost of the boiler pressure parts and the steam turbine.
These studies have lead to several important questions what are the optimum steam
conditions for an A-USC power plant? As the steam temperature increases and fuel requirements
decrease, does the cost increase of the boiler and turbine offset the balance-of-plant cost savings?
Does the efficiency increase of operating at higher steam cycle pressures justify the extra
thickness of the nickel alloy pressure parts? Does the efficiency increase associated with double
reheat justify the additional material and equipment cost? Using in-house performance and cost
estimation tools, EPRI has embarked on a study to evaluate the relative cost of single reheat AUSC power plants operating at 3 pressures: 275, 345, and 415 bar (4000, 5000 and 6000 psia)
and four main steam temperatures: 595oC, 650oC, 705oC, and 760oC (1100oF, 1200oF, 1300oF,
and 1400oF). Double reheat cycles operating at 345 bar (5000 psia) and the four steam
temperatures will also be evaluated. Performance and economics will be determined for the units
with no CO2 capture system, and with a 30 wt% monoethanolamine (MEA) Post Combustion
Carbon Capture (PCC) system designed to meet the U.S. EPA proposed Standard of Performance
at 1000 lb CO2/MWhr gross.
This paper discusses the initial results of the 345 bar (5000 psia) single reheat cases at 595oC,
650oC, 705oC, and 760oC (1100oF, 1200oF, 1300oF, and 1400oF) main steam temperatures with
and without the PCC system.
STUDY BASIS
The base units evaluated in this study have a gross turbine output of 850 MW, are fired with subbituminous coal and achieve emissions performance for criteria pollutants lower than currently
permitted values. In the cases with the PCC system the 30% MEA system is described elsewhere
[2]. A portion of the flue gas leaving the FGD system flows through the MEA absorber and 90%
of the CO2 entering is removed. The remaining flue gas is bypassed around the absorber and the
two streams combine before flowing out of the stack. The CO2 removed by the PCC system is
compressed to pipeline pressure and this cost and power consumption is included in the
calculations.
The plant of is located in Kenosha, Wisconsin. For weather protection the boiler and steam
turbine are enclosed, and the site is clear and level in a Seismic Zero zone requiring 30-m (10061

feet) deep pile foundations. Available at the site boundary are rail and transmission access, raw
water supplied from Lake Michigan, and natural gas. The fuel delivered by rail is Wyoming
Powder River Basin (PRB) sub-bituminous coal, with characteristics as detailed in Table 1.
The A-USC PC plant is designed with an annual capacity factor of 80%. Annual capacity factor is
defined as the actual annual production divided by the plant rated capacity times 8,760 hours.
The plant design is based on using components suitable for a 30 year life, with provision for
periodic maintenance and replacement of critical parts.
Steam Generator
The boiler island scope and general design basis are summarized below.

Greenfield, balanced-draft unit designed for base-load operation.

The unit is sized to have a gross output of 850 MW before the addition of the PCC system.

Low-NOX axial-swirl burners with over-fired air and SCR is used to achieve emission limits
of 0.03 lb/MBtu (0.013 kg/GJ).

Separator, recirculation pump and start-up system, and economizer.

A bottom ash system (submerged chain conveyor) to remove ash from the hopper throat
feeding it into a water-filled trough.

Soot-blowing system and mechanical draft cooling tower.

Single fans for forced-draft (FD) and primary-air (PA) duty.


The design stack gas emission limits for the unit are listed in Table 2. To meet these emissions,
the AQCS consists of the following components:

NOX Control: low-NOX burners with SCR.

Electrostatic Precipitator (ESP) for particulate control.

Wet FGD for sulfur control.

Mercury Removal Halogen injection into the boiler promoting mercury oxidation over SCR
catalyst with co-capture in the FGD. Possible supplemental capture using activated carbon
injection ahead of the ESP, if required.
The PCC technology for this study includes a conventional two column absorber/regenerator
scheme designed for 90% CO2 removal and utilizing 30 wt% MEA as the solvent.

62

Table 1
Coal Analysis for Wyoming Sub-Bituminous Coal
Proximate Analysis Weight Percent As Received
Moisture

30.24

Ash

5.32

Volatile

31.39

Fixed Carbon

33.05

Ultimate Analysis Weight Percent As Received


Carbon

48.18

Hydrogen

3.31

Nitrogen

0.70

Chlorine

0.01

Sulfur

0.37

Oxygen

11.87

Ash

5.32

Moisture

30.24
Heating Value As Received

HHV, kJ/kg (Btu/lb)

19,400 (8,340)

LHV, kJ/kg (Btu/lb)

17,900 (7,710)

Table 2
Emissions Limitation HHV Basis
Pollutants

Emission Limits
3

PM10

0.01 lb/MBtu

~10 mg/m

PM2.5

0.013 lb/MBtu

~13 mg/m

SO2

0.03 lb/MBtu

~30 mg/m

NOX

0.03 lb/MBtu

~30 mg/m

VOC
Mercury

3
3
3

0.0025 lb/MMBtu (0.0011 kg/GJ)


90-percent capture

Carbon Dioxide

Varies

63

The portion of the flue gas exiting the FGD unit and entering the PCC system first enters a flue
gas scrubber where it is contacted with circulating, cooled water. This scrubber cools the flue
gas, which by decreasing volumetric flow and condensing water improves absorber performance.
By adding a dilute caustic solution to the circulating water, residual SO2 in the flue gas can be
removed reducing degradation product formation
From the scrubber, the cooled flue gas enters a blower that provides the head to overcome the
pressure drop of the absorber and piping without increasing the back-pressure on the FGD
system.
The flue gas enters at the bottom of the column and flows upward, and the CO2-lean solvent
enters at the top of the column and flows downward. The CO2-depleted flue gas enters the upper
wash section of the column and is brought in contact with circulating water that cools the flue
gas and scrubs out any amine present. The cleaned flue gas then flows to the stack for discharge
to the atmosphere.
The CO2-rich solvent exits the bottom of the absorber where it is pumped through the rich/lean
solvent exchanger and into the regenerator. The hot rich solvent enters the top of regenerator
where the absorbed CO2 is released by the addition of heat. The heat breaks the chemical bonds
between the CO2 and the solvent, liberating the CO2 and regenerating the solvent so that it can be
returned to the absorber for CO2 removal. The heat is provided by the condensation of low
pressure (~4 bar (60 psi) steam) in the regenerators kettle-type reboilers, the steam being
extracted from the crossover between the IP and LP steam turbines. Heating the rich solvent
before it enters the regenerator decreases the extraction steam required.
The gas leaving the regenerator is cooled before being sent to the reflux drum. The gas exiting
the reflux drum is the product CO2 which is sent to the CO2 compressor system and compressed
to approximately 150 bar (~2200 psia) before entering the pipeline.
The lean solvent exits the bottom of the regenerator and flows through the rich/lean solvent
exchanger where it is cooled. The exiting lean amine is cooled in a trim cooler before returning
to the absorber.
MODELING
PC Cost is a Microsoft Excel based costing tool developed by EPRI that has evolved over a
24 year period of time. Its purpose is to allow engineers and planners to estimate the conceptual
and preliminary costs of PC-fired, subcritical and supercritical power plants [6]. It calculates the
heat and material balance for the unit and the overall plant performance based on the fuel
specification and then determines the cost of the major equipment subsystems by scaling from
reference costs. These reference costs are based on budgetary quotes and/or in-house developed
estimates which have been updated periodically over the life of the costing tool. The goal is to
provide a +/- 30% estimate for the costs for the equipment and materials within the plant
boundary, which includes all the direct and indirect costs for: site preparation, earthwork,
concrete and structural steel, building construction, major equipment, auxiliary equipment,
piping, electrical/instrumentation/control equipment as well as construction labor, bulks, and
subcontractors.
Beginning in 2011, PC Cost was modified to allow it to estimate the performance and cost of AUSC units. As part of that modification, it was decided to use AspenTechs AspenPlus to
perform the majority of the heat and material balance calculations. This would also allow for the
use of the physical properties generated by AspenPlus to be used for heat transfer calculations.
Additional capability was added to allow PC Cost to be able to choose the most cost effective
material for each row of tubes in the boiler backpass, depending on the design pressure and
64

calculated metal temperature. The possible material selections include carbon steel, T91, S304H,
HR3C, HR6W, IN617, HR230 and IN740.
PC Cost assumes a boiler configuration to perform its heat transfer calculations. This
configuration and the corresponding AspenPlus model are shown in Figure 1. Sizing of the PC
furnace is performed using proprietary methods. PC Cost sizes the backpass of the boiler using
conventional heat transfer methods using the flowrates, temperatures, and physical properties
from AspenPlus to determine the heat transfer coefficients and the surface area. The
temperature and pressure of the steam/water are used to determine the wall thickness of the tubes,
tube weights, and then the cost is determined by using unit costs ($/kg) for the boiler components.
One of the main objectives of this project was to produce these studies on a consistent basis as the
pressure and temperature changed. However, one early conclusion from this study is that using
the same boiler geometry for all temperature cases significantly penalizes the higher temperature
cases. PC Cost inherently assumes a long piping run from the top of the boiler to the steam
turbine. As the temperature increases, the material of piping quickly transitions to Inconel 740,
which is roughly 10 times the cost (per kg or lb) than T91. Boiler and turbine manufacturers are
designing plant layouts that minimize the amount of high-nickel alloy required. One option is a
horizontal boiler design proposed by Siemens, [7] which would minimize the length of piping
between the boiler and the turbine. An alternate design would separate the HP and IP turbines
into multiple components [8]. A high temperature HP section would be located near the top of
the boiler, and this would exhaust at a lower temperature and pressure into a conventional HP
steam turbine. After the steam is reheated, the high temperature, hot reheat steam would enter a
similar IP section near the top of the boiler which would exhaust into a conventional IP turbine.
For the remaining equipment, the costs are scaled from reference plant costs using industryaccepted algorithms whose exponents and constants are tailored to each type of equipment.
Additional changes were made to the cost estimate calculation of the steam turbine. Since there
is no commercial A-USC turbine at this time, a method was devised to estimate its cost relative to
a USC turbine. The temperature limit for the steam turbine estimated in PC Cost is 595oC
(1100oF). After discussions with steam turbine consultants, it was decided to apply a
multiplication factor to this estimated steam turbine cost to account for the change in materials
required for the higher operating temperature. At 595oC (1100oF), this multiplication factor was
1.0, and it linearly increased to 1.2 at 760oC (1400oF). Several organizations are currently
working to provide a preliminary design of an A-USC turbine, and a more accurate cost. When
this information becomes available, it will be incorporated into PC Cost.
The steam system heat balance was modeled in Gate/Cycle using Spencer-Cotton-Cannon
algorithms to predict the efficiency of the turbine under the various operating conditions. A
simulation was created for the base plant (no-PCC) and converged on a steam turbine gross
power of 850 MW. This information was used by PC Cost/AspenPlus to calculate the coal flow
rate and the heat and material balance for the boiler systems. The power consumption within the
steam cycle (HP BFW pumps, condensate pumps, etc) was calculated by Gate/Cycle, and the
auxiliary loads for the boiler and AQCS systems were calculated by AspenPlus based on the
calculated flowrates. Values not calculated by AspenPlus for certain plant equipment (mill
power, for example) were estimated from previous EPRI reports.
For the PCC cases, the coal feed and the steam/water flows calculated for the non-PCC cases
were fixed. The LP steam extraction for the PCC system was removed from the IP/LP crossover.
This steam was condensed in the PCC regenerator reboilers and returned to the steam cycle.
However, as expected, the steam extraction resulted in a significant decrease in gross power.
Within the PCC system, there are opportunities to recover useful heat and use this energy to heat
the boiler feed water leaving the condenser. This allows the steam that would normally be
65

extracted from the steam turbine for boiler feedwater heating to be used to generate power. A
heat integration study was carried out for all cases based on the experience gained in previous
reports (References 2-5). This allowed for the increase of gross power and improved the plants
performance as well as reducing load on the cooling tower. Typically 30% of the heat in PCC
system was transferred to the boiler/steam turbine system.

SPRAY2

SPRAY1

MSPIPE
MAINSTM

STM7

FINSH
(HEATER)

STM6

ATT2
(MIXER)

PLATSH
(HEATER)

STM5

ATT1
(MIXER)

STM4

STM3

PSH
(HEATER)

STM2

FGSPLIT
(FSPLIT)
PLATSHQ

FINSHQ

SCR1Q

ROOFWLQ

SCR2Q
RWSTOTQ

PLATSHFG
(HEATER)

FG2

FINSHFG
(HEATER)

FG3

SCR1
(HEATER)

FG4

ROOFWALL
(HEATER)

FG5

SCR2
(HEATER)
PSHQ
FG6

FG7A

FGSPLIT
(FSPLIT)

HRH2TUR

RWSCR
(HEATER)

HRHPIPE

FG7B

HRHOUT
FG1
PSHFG
(HEATER)

REHEATFG
(HEATER)

REHEATQ

REHEAT
(HEATER)
STM1

CRH
FG8A

FG8B

MXFG
(MIXER)
BOILERWA
(HEATER)

BOILERQ
BOILERFG
(HEATER)

BFW2
FG9
FG0
FWECONQ
FWECON
(HEATX)

FWECONW
(HEATER)

BFW1

LOSS
(HEATER

SPL-SPRA
(FSPLIT)

TOSCR

TSPRAY

BFWIN
DRY-COAL
BURNER
(RGIBBS)

HOPPER
(SEP)

AMMONIA

SCR
(RSTOIC)

INGRESS

MXINGRESS
(MIXER)

FROMSCR
BTMASH

AIR
TOAIRHTR
ESPDP
AIRHEAT
(HEATER)

AIRHTOUT

TOESP

FLYASH

AIRHTQ

PASAHEAT
(HEATER)

Figure 1
PC-Cost Boiler Configuration

66

ESP
(SEP)

AMBAIR

TOID

IDFAN
(COMPR)

TOFGD

In addition to the loss of gross power, there is a significant increase in the auxiliary power with
the addition of the PCC system, primarily due to the large CO2 compressors. The decrease in
gross power and increase in auxiliary power creates a substantial decrease in net power,
especially for the 90% CO2 removal cases.
It is intuitive that reducing the percentage of the CO2 removed from the flue gas would have a
significant improvement on the plants performance and economics. To meet the EPAs
Standard of Performance approximately half of the CO2 would have to be removed from the
flue gas compared to 90% CO2 removal. This reduces the LP steam extracted which reduces the
effect on gross power, as well as the auxiliary power load.
BASE PLANT RESULTS (NO PCC CASES)
The results of the Base Plant cases are shown in Table 3. As steam temperature increased, the
overall plant efficiency increased and CO2 emissions decreased. Since the coal flow decreased
with increasing temperature, the power requirements associated with coal feeding and the AQCS
systems decreased. As steam flowrate per MW decreased, the HP BFW pump power also
decreased. These improvements decreased the auxiliary load and increased the net power.
As coal flow decreases, the capital costs for the coal feed system, ash removal, and the AQCS
systems all decrease. As the steam temperature increases, the cost of the boiler increases
primarily for three reasons:

In general, alloy mechanical strength decreases with increasing temperature. Therefore,


the tubing thickness increases with increasing steam temperature.

Beyond a certain temperature, the strength is too low and/or corrosion rates too high for
one material and another material must be substituted for the application. Typically,
these upgraded materials are more expensive on a per pound or per kilogram basis.

Even though the steam temperature is increasing, the temperature profile of the flue gas
through the boiler is approximately the same. Therefore, the driving force for heat
transfer is decreasing which requires a larger surface area for a given heat duty.

This can be evidenced in Table 3 which shows the capital cost increasing by 15% between the
595oC (1100oF) case and the 760oC (1400oF) case.
Although the coal feed decreases with increasing steam temperature, the cost of electricity (COE)
increases due to the increasing capital cost. For example in going from 595oC to 760oC (1100 F
to 1400F) the CO2 emissions fall 7.3% but the COE increases 9.1%. One important observation
is that the results in Table 3 reflect conditions in the U.S. market. EPRI has been working with
members in Asia to understand the differences between projects in the U.S. and Asia, and work is
ongoing in this area. The COE in Table 4 shows a comparison between the U.S. market and Asia
based on the following assumptions:

67

Table 3
Preliminary Performance and Economics of 345 bar (5000 psia) units with no PCC.
o

Description

595 C/615 C
1100F/1140F

650 C/671 C 705 C/727 C 760 C/760 C


1200F/1240F 1300F/1340F 1400F/1400F

Gross plant
output, kW

850,000

850,000

850,000 kW

850,000 kW

Auxiliary
load, kW

87,716

82,894

78,925

76,055

Net plant
output, kW

762,284

767,101

771,075

773,945

Net plant
heat rate

8,534 Btu/kWh
9,056 kJ/(kWh)

8,282 Btu/kWh 8,054 Btu/kWh 7,885 Btu/kWh


8,738 kJ/(kWh) 8,497 kJ/(kWh) 8,319 kJ/(kWh)

Net plant
efficiency
(HHV)

39.7%

41.2%

42.4%

43.3%

Plant fuel
consumption

784,500 lb/hr
355,900 kg/hr

761,800 lb/hr
345,500 kg/hr

744,600 lb/hr
337,700 kg/hr

731,700 lb/hr
331,900 kg/hr

CO2
Emission

1,382,000 lb/hr
627,400 kg/hr

1,342,000 lb/hr 1,312,000 lb/hr 1,287,000 lb/hr


609,200 kg/hr 595,500 kg/hr 584,100 kg/hr

CO2
Emission

1,625 lb/MW (g)


738 kg/ MW (g)

1,579 lb/MW (g) 1,543 lb/MW (g) 1,514 lb/MW (g)


717 kg/ MW (g) 700 kg/ MW (g) 687 kg/ MW (g)

Capital
Cost, $1000

$1,732,000

$1,766,000

$1,834,000

$2,001,000

Capital
Cost, $/kW

$2,270

$2,300

$2,380

$2,590

$363,000

$365,000

$378,000

$402,000

$67.90

$68.00

$70.00

$74.10

Total
Annual
Costs,
$1000/yr
Cost of
Electricity,
$/MWh

68

EPRI has learned anecdotally that capital costs on a $/kW basis in Asia are 50-70% of
the cost based on U.S. estimates. For this comparison it is assumed that the capital cost
is 60% of what is estimated by PC Cost.

Richardson International Cost Factor Manual [9] estimates the labor index relative to the
U.S. market. The average labor index for Asia given in the manual (Taiwan, Korea, and
India) is 0.4. This was used to adjust the operating and administrative labor rates.

Platts Coal Trader International [10] predicts a value of approximately $3.70 to


$4.55/GJ ($3.90 to $4.80/MMBTU) for comparable coal shipped to Asia from Indonesia
or South Africa in the 2013-2016 timeframe. The midpoint of this range, $4.12/GJ
($4.35/MMBTU) is used instead of the U.S. coal price of $1.71/GJ ($1.80/MMBTU) for
PRB to Kenosha, WI.

Table 4
Comparison of the COE between the U.S. and Asia
Steam Cycle

U.S. COE

Asia COE

595 C/615 C
(1100F/1140F)

$67.90

$68.70

650oC/671oC
(1200F/1240F)

$68.00

$67.80

705oC/727oC
(1300F/1340F)

$70.00

$68.20

760oC/760oC
(1400F/1400F)

$74.10

$70.20

In the U.S. the COE is largely determined by the capital cost and the coal component is relatively
small. In Asia the reverse is true; the coal cost is a significant portion of the COE. This is
illustrated in Figure 2, which shows the major components of the COE for the 595oC (1100oF)
case. The coal component is about 22% of the COE in the U.S. and about 50% in Asia.
Therefore, operating at higher efficiency is economically important.
As Table 4 shows, the COE is less for the 650oC and 705oC cycles (1200oF and 1300oF) than for
the 595oC (1100oF) cycle. If coal prices increase, this difference becomes more significant.
Because these units all have the same boiler/turbine configuration, the higher temperature cases
are penalized because of the long runs of Inconel 740 main/reheat steam piping. EPRI will
evaluate the cost savings of alternate A-USC plant designs in future work.

69

U.S.

Asia
Capital
Coal
FOM
VOM

Figure 2
1
Cost of Electricity Components in the U.S. and Asia
A-USC AND CO2 CAPTURE
The most significant economic driver for A-USC in the U.S., Europe, and other countries may be
CO2 removal from PC power plants. As previously mentioned, the EPA has released a Standard
of Performance for the removal of CO2 from fossil plants exhaust gas. The Standard proposed a
limit of 1000 lb CO2/MW on a gross basis. Since its release the Standard has been in a comment
and review period, but it is expected that a final limit will be released by the EPA by the end of
2013.
To meet the partial capture limit, slightly less than half of the flue gas enters the PCC absorber.
Because of this, less steam is extracted from the steam turbine and the reduction in gross power
decreases. To meet the EPAs Standard of Performance for the partial CO2 capture cases the
decrease in efficiency was about 4 percentage points as shown in Table 5.
Because of the loss of net power, there is an amplification of the capital cost on a $/kW basis.
For example, in the 595oC (1100oF) case, the overall capital cost increases 14% over the unit with
no PCC. However, on a $/kW basis, the capital cost increased 27% over the case with no PCC
system. On average, the COE increased approximately 30% over the non-PCC cases.

Note: FOM Fixed Operating & Maintenance Costs, VOM Variable Operating &
Maintenance Costs.
70

Table 5
Preliminary Performance and Economics of 345 bar (5000 psia) units with 1000 lb
CO2/MW(g) Emissions Target
Description

Gross plant output, kW

815,348

816,484

818,526

821,256

Auxiliary load, kW

130,018

122,143

115,705

110,821

Net plant output, kW

685,330

694,341

702,821

710,435

Net plant heat rate


Net plant efficiency
(HHV)
CO2 Emission
CO2 Emission

595 C/615 C 650 C/671 C 705 C/727 C 760 C/760 C


1100F/1140F 1200F/1240F 1300F/1340F 1400F/1400F

9,547 Btu/kWh 9,150 Btu/kWh 8,836 Btu/kWh 8,590 Btu/kWh


10,072 kJ/(kWh) 9,653 kJ/(kWh) 9,322 kJ/(kWh) 9,062 kJ/(kWh)
35.7%

37.3%

38.6%

39.7%

816,000 lb/hr
370,000 kg/hr

817,000 lb/hr
370,000 kg/hr

819,000 lb/hr
371,000 kg/hr

821,000 lb/hr
373,000 kg/hr

1,001 lb/MW 1000 lb/MW (g) 1,000 lb/MW 1,000 lb/MW


(g)
(g)
(g)
454 kg/ MW (g)
454 kg/ MW (g)
454 kg/ MW (g) 454 kg/ MW (g)

Capital Cost, $1000

$1,975,000

$2,002,000

$2,063,000

$2,225,000

Capital Cost, $/kW

$2,881

$2,883

$2,936

$3,133

Total Annual Costs,


$1000/yr

$435,000

$434,000

$440,000

$462,000

Cost of Electricity,
$/MWh

$90.50

$89.20

$89.50

$92.80

CONCLUSIONS AND FUTURE WORK


As with any engineering performance and economic study, the conclusions are largely dependent
on the assumptions made. From this work it is apparent that the economic advantage of using AUSC will depend on:

The boiler/turbine configuration. One significant conclusion from this study has been that
the cost of the long main and reheat steam lines between the boiler and the steam turbine
significantly increases the capital cost at the higher steam temperatures. Alternate
boiler/steam turbine configurations will be necessary for A-USC designs to be economically
competitive at temperatures higher than 700oC (1300oF).

The alloy costs. Currently Inconel 740 pipe and tubing is estimated to cost 25 times more on
a per pound or kilogram basis than carbon steel. However, Inconel 740 pipe and tubing have
71

not been manufactured in commercial quantities, so this cost is largely speculation. As the
first A-USC units are built, the price for these materials is expected to decrease.

The cost of the turbine. The cost for an A-USC turbine is unknown. As the market
develops, the price for an A-USC turbine will become better understood.

The applicability of A-USC in Asia will largely be driven by coal price. A-USC power
plants will become more economic at higher coal prices.

The plant availability. The Cost of Electricity for these studies is based on a plant
availability of 80%, which is typical of U.S. units in todays market due to increased
renewable energy usage for power generation. Coal fired units in Asia may be operating
closer to 90%, which would decrease the COE.

CO2 capture requirements. If coal is burned in future U.S. power plants, some amount of
CO2 capture will be required. The final regulations for CO2 removal in new fossil power
plants are expected by the end of 2013. The limit set by this regulation will determine to a
large extent the economics of future coal fired units in the U.S.

Finally, significant work is ongoing in many areas of the world related to A-USC:

Both China and India have announced plans to build 700oC A-USC units within the next 5-7
years.

Turbine manufacturers are working to develop A-USC steam turbine designs and estimate
the cost to build an A-USC turbine.

Due to the high cost of the nickel alloys, the cost of the main and reheat steam lines from the
boiler to the turbine will be a significant percentage for the overall power plant capital cost.
Boiler, turbine manufacturers, and EPCs are evaluating plant layout configurations to
minimize the length of these lines, which could change the economic conclusions reached by
this paper

The DOE/OCDO A-USC Materials Consortium continues to evaluate materials performance


at high temperatures. As part of this work, the consortium is working to identify a host for a
760oC (1400oF) Component Test (COMTEST 1400). This test loop would take steam from a
host utilitys main steam line, heat it to 760oC and then flow the steam through a small (~4
MW) A-USC steam turbine.

The Materials Consortium continues to work with materials suppliers to understand alloy
prices in commercial quantities.

Because the EPA regulation is based on gross output, there may be advantages to using CO2
removal technologies that do not decrease gross power. EPRI is currently researching the
use of membranes for the removal of CO2. The results of this work are planned for release
later in 2013.

72

REFERENCES
[1] Engineering and Economic Evaluation of 1300F Series Ultra- Supercritical Pulverized
Coal Power Plants: Phase 1, EPRI, Palo Alto, CA 2008, 1015699.
[2] An Engineering and Economic Assessment of Post-Combustion CO2 Capture for 1100oF
Ultra-supercritical Pulverized Coal Power Plant Applications: Phase II Task 3 Final
Report, EPRI, Palo Alto, CA 2010, 1017515.
[3] Engineering and Economic Evaluation of 1300oF Series Ultrasupercritical Plant with PostCombustion Capture: Engineering and Economics, EPRI, Palo Alto, CA 2012, 1026645.
[4] Engineering and Economic Analysis of a 1300F Series USC Demonstration Plant with
Natural Gas Equivalency Post-Combustion Capture. EPRI, Palo Alto, CA, 2013. 1026644.
[5] Minimizing the Derate of Adding Post Combustion Capture to Pulverized Coal Units
EPRI, Palo Alto, CA: 2012. 1023864.
[6] Hoskins, Bill (URS), and Booras, George (EPRI), Assessing the Cost of New Coal-Fired
Power Plants, Power Magazine, October 2005, pg 24-28.
[7] Smith, David, Horizontal Boilers make 700oC Steam Economic, Modern Power Systems,
May 2000, pg 36-41.
[8] Feng, Weizhong, Technology Innovation of Most Efficient Coal Power Plant and Plan for
Future, presented at EPRIs CoalFleet Technical Meeting, Mobile, AL, December 2012.
[9] Richardson International Construction Factor Manual, Cost Data On Line, Inc., Prescott, AZ,
2011.
[10] Coal Trader International, Platts, Volume 13, Issue 135, July 13, 2013

73

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

NEXTGENPOWER DEMONSTRATION AND COMPONENT


FABRICATION OF NICKEL ALLOYS AND PROTECTIVE COATINGS
FOR STEAM TEMPERATURES OF 750C
Arthur F. Stam
DNV KEMA Energy & Sustainability, Arnhem, The Netherlands
ABSTRACT
The EU NextGenPower-project aims at demonstrating Ni-alloys and coatings for application in
high-efficiency power plants. Fireside corrosion lab and plants trials show that A263 and A617
perform similar while A740H outperforms them. Lab tests showed promising results for NiCr,
Diamalloy3006 and SHS9172 coatings. Probe trials in six plants are ongoing. A617, A740H and
A263 performed equally in steamside oxidation lab test 750C while A617 and A740H
outperformed A263 at 800C; high pressure tests are planned. Slow strain rate testing confirmed
relaxation cracking of A263. A creep-fatigue interaction test program for A263 includes LCF
tests. Negative creep of A263 is researched with gleeble tests. A263 80 - 500mm trial rotors are
forged with optimized composition. Studies for designing and optimizing the forging process
were done. Segregation free 300 and 1,000mm rotors have been forged. A263 A263 and
A293 COST F rotor welding show promising results (A263 in precipitation hardened
condition). Cast step blocks of A282, A263 and A740H showed volumetric cracking after heat
treatment. New as cast blocks of optimized composition are without cracks. A 750C steam
cycle has been designed with integrated CO2 capture at 45% efficiency (LHV). Superheater life at
750C and co-firing is modeled.
Keywords: USC, Ni-alloys, coatings, coal, biomass
INTRODUCTION
NextGenPower (www.nextgenpower.eu) is an EU-funded project aiming at fabrication and
demonstration of Ni-based alloy components and protective coatings for application in power
plants to increase their efficiency. A higher efficiency results in lower CO2-emissions and
reduced fossil fuels consumption. The project aims at plants combusting coal where biomass cofiring decreases the need for carbon capture. To reach 45% plant efficiency (LHV) with
application of carbon capture, steam temperatures of 750C and biomass co-firing is necessary.
The combination of co-firing and elevated temperatures creates a more aggressive fireside
environment requiring the use of coatings.
The state-of-the-art in steam temperature applied in coal-fired power plants is 620C. Decades of
research precede this project aiming for increase in efficiency and improvement of material
properties. In Europe, the Thermie / AD700 project started in 1998 and included material
development and component testing for steam temperatures up to 700C [1]. Even higher
temperatures of 760C were aimed at in the US [2]. Advanced-USC projects run worldwide, e.g.
in Japan, China and India.
The NextGenPower-project is led by DNV KEMA and includes three technical sub-projects:
Boiler (subdivided into work packages (WPs) fireside corrosion, steamside oxidation and
74

mechanical testing),Steam Turbine (WPs rotor forging, rotor welding and casting) and
Integration (WPs operating conditions and environments, steam cycle modeling integrating
carbon capture and superheater life modeling).
The partners in the project are DNV KEMA, Doosan Babcock, Skoda Power, E.ON, Cranfield
University, Goodwin Steel Castings, Monitor Coatings, Saarschmiede, Aubert & Duval, VTT,
VZ, TU Darmstadt and Special Metals. The project started May 1st, 2010 and has a duration of
48 months. The budget is 10.3 million (6 million EU contribution).
BOILER
Fireside corrosion
Co-firing biomass and higher metal temperatures increase the risk on high-temperature corrosion
and its rate. Co-firing biomass causes a more severe oxidation due to the higher alkali and
chlorine content in biomass. Alkalis can form low-melting eutectics causing Hot Corrosion. The
oxidation rate in chlorinated atmospheres is more severe compared to sulphidation. Generally,
salts containing chlorides have a lower melting point compared to salts including only sulfates.
Advanced materials contain, in general, Ni and Cr. It was generally found that a higher Crcontent and (Ni+Cr)-content will increase the fireside corrosion resistance under coal firing
conditions [3], [4], although test results for Ni-alloys in [5] do not confirm this findings. Ni
alloys in general suffer from fireside corrosion due to Hot Corrosion and internal sulphidation,
depending on the temperature [6].
A test program was set up including lab screening trials and rig trials for investigation of
materials and coatings at A-USC temperatures, allowing systematic research in well-defined
conditions. Plant trials are performed to demonstrate the material behavior in service for a longer
time.
Fireside corrosion screening tests
Cranfield University executes a screening program with base materials T24, T92, A316, Sanicro
25, S304H, TP347HFG, A617, A263 and A740H, coated and uncoated, with and without
deposits and under different gas environments in a temperature range of 600 800C. The tests
were carried out in a furnace (Fig. 1). Different coatings are applied: NiCr, FeCrAl, NiCrAlY,
NiCrWMoCu, NiCrMoWFe, CrNiC, and NiCrMo(Nb+Ta).
Results after 1,000hrs [7] show that Ni-alloys A617, A263 and A740H have a broad front type
corrosion attack at 650C and 700C without internal damage. The corrosion mechanisms in all
three alloys changed at higher temperatures (750C and 800C) and the alloys suffered from
significant internal corrosion. A263, A617 and A740H all suffered from similar metal losses at
the lower temperatures (650 and 700C); however, at the higher temperatures (750C and 800C)
A740H outperformed A263 and A617.

75

Figure 1: Schematic diagram of a controlled atmosphere furnace setup for fireside corrosion in
simulated air-firing combustion gases at Cranfield University
Coatings A625, NiCr, FeCrAl and NiCrAlY were deposited on T91 using HVOF (High Velocity
Oxy Fuel) and plasma spraying. The thermally sprayed coatings suffered significant corrosion
attack. The plasma-sprayed coatings showed less corrosion than the HVOF-sprayed coatings due
to a lower porosity. NiCr showed the lowest thickness reduction of ~87m (HVOF-sprayed) and
~13m (plasma-sprayed). In general, the median degradation for coatings had the following
ranking (in the descending order: most to the least damage): NiCrAlY > A625 > FeCrAl > NiCr
[8]. Based on the screening tests, a selection of the base materials and coatings will be tested in a
100kW burner test rig at Cranfield University.
VTT tested substrates X20, Sanicro25 and A263 with coatings NiCr, A625, Diamalloy 4006 and
Nanosteel SHS9172 (DJ, Diamond Jet Spray by Sulzer Metco and CJS, Carbide Jet Spray by
Thermico) [9]. The materials were covered with 6.5wt%NaCl-59wt%Na2SO4-34.5wt%KCl salt
and exposed to a air-0.1%HCl-1%SO2-10%H2O atmosphere for 168hrs at 575C and 625C.
The Diamalloy 3006 DJ and SHS9172 CJS coatings had the best corrosion resistance against
molten salt attack. SHS9172 DJ showed good corrosion behavior, but vertical cracking should be
prevented with sufficient coating optimization. NiCr DJ, A625 CJS and A625 DJ formed
protective but thick corrosion layers especially at 575C. High melting state coatings NiCr, A625
and Diam4006 showed sufficient corrosion protection.
Fireside corrosion plant demonstrations
VTT performed corrosion tests [10] at the 550MWth CFB Alholmens Kraft power plant (Finland)
firing 30% peat, 10% coal, 50% biomass (mainly wood) and 10% SRF (Solid Recovered Fuel).
During the measurements (~1,300 hrs) an air/water-cooled superheater probe contained two sets
of specimen rings, one controlled at 550C, and one at 750C. T92 was used as substrate for the
550C set and exposed both uncoated and with five different coatings: NiCr, A625, Diam4006,
NiCrTi and SHS9172. An identical test was carried with A263 substrate at 750C. As a result the
coated T92 showed a decreased corrosion compared to the uncoated martensitic steel. For A263
the uncoated material as well as A263 coated with NiCr and NiCrTi showed the lowest
degradation while A625 showed severe corrosion. DIAM4006 was nearly completely oxidized.
Two additional measurement campaigns will be carried out.

76

To investigate membrane wall corrosion, DNV KEMA installed KEMCOPs i.e. M8 probes
protruding the strip of the membrane. The materials 15Mo3 (450 500C), 10Cr9-10 (450
550C), T24 (500 600C), T92 (450 500C), A617 (550 750C) and A263 (500 750C),
are exposed for 5,000 15,000 hours. The probes are installed at three Dutch coal/biomass fired
power plants of GDF Suez, RWE/Essent and E.ON. The length reduction rate was determined
after 5,000 hours for materials T24, A617 and A263. Figure 2 compares the results from RWE
(C) and E.ON (B) plants with results from an earlier corrosion test (coal plant (A)). Both Nialloys showed similar length reduction which was lower compared to T24. Superheater probe
tests are currently running for 15Mo3, A263 and Alloy 740H at 540 750C at the above
mentioned plants including electrochemical noise measurements.
120

length reduction rate (m/year)

100
T24 Coal boiler A

80

T24 Coal + MBM-mix boiler B


T24 Coal + wood boiler C

60

Inconel 617 Coal boiler A


Inconel 617 Coal + MBM-mix boiler B
Inconel 617 Coal + wood boiler C

40

Nimonic 263 Coal boiler A


Nimonic 263 Coal + MBM-mix boiler B

20

Nimonic 263 Coal + wood boiler C

0
400

500

600

700

800

temperature (C)

Figure 2: Time averaged length reduction rates of KEMCOP probes


Doosan Babcock has installed four exposure probes in the E.ON Ratcliffe coal plant (UK), unit 2;
Two probes are installed near superheaters (600-700C and 700-800C) and two probes near
reheaters (600-700C and 680-760C). Each probe includes 22 test sections. Applied base
materials are T91, X20 and TP347HFG; applied coatings are Ni50Cr, A625, FeCrAl, and
NiCrAlY. Another test program is planned at Tilbury (UK) firing biomass. Two probes will be
exposed near superheater and one near reheates. The materials that are tested include Esshete
1250, TP 347HFG, Super 304H, Sanicro 25, A617 and A740.
Steamside oxidation tests
The spallation of oxide from the steamside of a once-through boiler circuit, leads to excessive
levels of oxide flakes causing blockage leading to overheating and causing erosion in the steam
turbine. Ferritic and austenitic steels form Fe and Cr-containing oxides that become less
protective >600C. In many high temperature systems >700oC, Al is added either to the alloy or
in a coating to promote the formation of alumina which is thermodynamically more stable at
higher temperatures than chromia. Diffusion aluminide coatings are among the best performers
and offer an economic benefit [11], [12]. While the use of Ni alloys will perform better than their
ferritic and austenitic counterparts in steam, there remain doubts over the long term steam
oxidation resistance.

77

Cranfield University conducts lab-scale exposure tests with small ring sections of test materials,
either coated or uncoated, in a flowing steam atmosphere at 600 800C. The steam is supplied
in a closed system, with de-ionized water supplied to the tube furnace where is it converted into
steam directed over the samples. The materials include T24, T92, A316, Sanicro 25, S304H,
TP347HFG, A617, A263 and A740H. Two aluminum coatings are tested, a slurry coating and an
ionic liquid application coating.
The results show that mostly thin oxides are formed on the highly alloyed materials - as expected
for samples with ground surface finishes. Nodular growths increased with time and temperature.
Sanicro 25 outperformed other stainless steels (304H, 316, 347 HFG) at 700- 800 C and all
stainless steels, except for Sanicro 25, suffered from scale spallation at 800 C. Ni-based alloys
showed small mass gain at 700C and 750C (~0.2 and ~0.4 mg/cm2 respectively). At 800 C,
A617 and A740H outperformed A263. After the screening tests, exposure tests are performed in a
pressurized rig at Cranfield University (up to ~40bar, 800C) and at VTT (~40bar, 700C).
Relaxation Cracking
Relaxation cracking occurs in zones with residual stresses that are relieved by time dependent
inelastic deformation during PWHT or service. Materials with low (creep) ductility cannot
withstand the inelastic strains due to relaxation, and exhibit cracking within 3 years of operation.
Precipitations of second phases in the material matrix hinder relaxation of the residual stresses in
the matrix. As a result, the stresses preferentially relax in a small precipitate-free zone along the
grain boundaries. This may result into inter-crystalline cracking and brittle rupture.
A hot tensile test serves as a screening test to observe if A263 can accommodate plastic
deformation in the temperature range of 700 750C, where the lower the ductility, the higher
the susceptibility to relaxation cracking. The test is performed by VZ and is strain controlled.
Material before and after PWHT is tested, base metal and cross weld specimens are compared. A
notched stress relaxation tests is performed to show the effect of heat treatment at 980C for
180min on stress relaxation (according to standard EN 10319-1:2003).
TU Darmstadt performs a cyclic notch stress relaxation test which is considered to be similar to
service conditions. A263 samples specimens are tested in the as welded condition, after PWHT
and after service simulation (i.e. strain aged at temperature to 1% creep strain, over 3,000hrs).
The load controlled tests are performed at 700 750C, with dwell times (1h and 10h) to allow
for relaxation, and without dwell time for comparison. Strain is locally measured with an
extensometer, in and outside the notch. Stress in the notch is calculated with Neuber parabola and
the re-distribution of stress in the notch is studied.
Gas Phase Embrittlement
Ni-based alloys contain the hardening precipitates NbC. Oxidation of NbC can form CO2 gas
bubbles, forming intergranular cavities. Cracking follows on grain boundaries leading to brittle
failure. NbC can also oxidize to Nb2O5 but it is still investigated whether Nb is a crucial factor in
the cracking system. Gas phase embrittlement is identified with a Slow Strain Rate Test (SSRT).
Tensile and strain are controlled while a strain rate of 0.006%/min is applied. Specimens include
parent material and cross welds, and in all cases pre- and post-service material is tested.
Microscopic investigations of the base material (both pre and post-service simulation) show
disaggregation indicating gas phase embrittlement, see Fig. 3.

78

Figure 3: Base material (left), base material after SSRT at 750C (middle) and base material
with service simulation after SSRT at 750C (right)
Creep-Fatigue Interaction
Creep, fatigue and its interaction of A263 is investigated in the region 700 750C. Doosan
Babcock performs creep rupture tests (load controlled) on parent material and longitudinal welds,
both at virgin material and pre-strained (service simulation) material. LCF tests (R=0) are
performed on parent material and longitudinal welds, only at virgin material. Dwell times of 168h
and 1,000h are applied. VTT and TU Darmstadt perform LCF tests with R = -1 that are strain
controlled. Tests without dwell time and with dwell times of 3min, 20min, 1h and 10h are
performed. More precisely, virgin parent material and with longitudinal weld (dwell times of 0, 3,
20min. and 10h) and post-service parent material and with longitudinal weld (dwell times 0 and
10hrs).
Negative Creep
Negative creep describes the time dependent contraction of a material due to microstructural
changes as opposed to the elongation seen for a material experiencing normal creep behavior.
Negative creep is an ordering effect explained by existence of two phases occurring in a Ni-base
super alloy: a disordered matrix (r) and small ordered precipitates (r'). Al and Cr atoms diffuse
through r-r' interface and part of the Ni3Al transforms into Ni3Cr resulting in lattice contractions
[13]. The effect is researched by VZ for A263 where a conventional creep specimen is loaded in
a gleeble at 700 750C with a stress close to 0MPa and strain is monitored using laser
dilatometry to measure any contraction as a function of time (max. 20,000hrs) and temperature.
STEAM TURBINE
Rotor forging and welding alloy selection
Creep strength and the ability to fabricate full-scale forgings (e.g. segregation sensitivity,
forgeability, large-scale ingot manufacture) and weldability are the main parameters for alloy
79

selection. A230, A263, A282, A625, A706, A718, A901, A740, A740H, Waspaloy and A105
have highest creep strength at 750C. While A105 was discarded due to its low toughness and
expected difficulties in the forging process, the pre-selected alloys based on creep strength were
A263, A282, A740, A740H and Waspalloy. Weldability of A740 and Waspalloy was considered
problematic and were discarded. A263 was selected as the best option due to its commercial
availability, low sensitivity for segregation, experience of the partners with the material, and
ability to weld the material to 10Cr-steels.
Rotor Forging Alloy Optimization
After long time ageing eta () phase precipitates arise in A263 [14] where is an intermetallic
phase: Ni3Ti. As the phase is formed at the expense of the hardening gamma prime phase and the
precipitates are large platelets extending across grains, the phase generally has an effect on
long-term creep behavior in superalloys. However, no effect on creep behavior for A263 was
observed by [15] while [14] state that further research on creep behavior would be needed.
Aubert & Duval optimized the composition of A263 in order to remove phase. Five trial melts /
forgings (80mm) have been produced, one reference alloy 263 and four with optimized
composition. The standard heat treatment for A263 (1150C/2hours/water quench and
800C/8hours/air cooled) was applied. After ageing (870C/100hours) the reference alloy 263
showed phase whereas the optimized 263 alloys with identical heat treatment and ageing did
not show phase. Tensile tests, creep tests (Fig.6 (left)) and toughness tests were performed,
showing that for the short term, the optimized alloy had better creep properties. Based on the
results, one modified alloy 263 composition was selected and a new trial melt / forging was made
(160mm). Longer ageing was performed (750C/3,000hours) and the optimized alloy showed
improved creep properties and toughness compared to the reference alloy 263. The mechanical
test program is not yet completed and final conclusions can be drawn on basis of long-term test
results. Meanwhile, two large-scale rotors are produced (300 and 500mm) which are
subjected to a mechanical test-program and welding trials. The rotors were free of cracks and the
rotor produced from the optimized alloy showed good workability.

Figure 4: Creep properties of reference A263 and optimized compositions of alloy 263; trial
forgings produced by Aubert & Duval (left). Creep properties of C263; 300mm trial forging
produced by Saarschmiede (right). Measurements performed by TU Darmstadt.

80

Rotor Forging Forging studies and production of full-scale rotors


The forging process starts with producing ingots by melting electrodes using Vacuum Induction
Melting (VIM), followed by Electro Slag Re-melting (ESR) and Vacuum Arc Re-melting (VAR)
to remove impurities and to reduce inclusions. Computer simulations (MeltFlow, see e.g. [16]) of
the re-melting processes were performed in order to optimize re-melting parameters and therefore
to avoid or suppress segregations. The forging process consists of upsetting and stretching. The
forging process was modeled using SIMUFACT, see e.g. [17] which is a finite element model
calculating distribution of temperature, deformation degree and grain size at each position in the
forging and at every moment during the forging operation. The calculation results are used to
design and optimize the forging process. The description of micro-structural changes (e.g.
recrystallization) during the forging process depending of the forging parameters (temperature,
deformation degree, sequence of forging steps) is of key importance in the modeling. A material
dataset for C263 was therefore experimentally determined, including measurement of flow
curves, static and dynamic recrystallization and grain growth, all as function of temperature,
deformation degree, deformation speed and initial grain size.
A bar (300mm) was produced and the forging process was simulated. The metallographic
determined grain size of the forged bar and the simulation results were in quite good agreement.
The measured results were also used to calibrate and optimize the forging model. This model was
then used to perform several calculations with variations in temperature, deformation degree,
deformation speed and forging passes, resulting in the detailed forging procedure for the fullscale rotor. The material model for A263 showed a fast development of static recrystallization at
high temperatures. Therefore several calculations were carried out in order to minimize holding
times at high temperatures without plastic deformation and to reduce temperatures where
possible. The full-scale rotor (1,000mmx3,000mm) now has been produced and is free of
segregations. The 300mm and 1,000mm bar are now subject to mechanical testing (see Fig. 6
for first results) and welding trials.
Rotor Welding
A welding process is designed full scale rotor welding of A263 with COST F (a 10%Cr steel) is
demonstrated. The A263 has to be welded in precipitation hardened condition for the dissimilar
metal joint. A263 - A263 welding was performed at Polysoude (France) on the trial melt / forging
(160mm) produced by Aubert & Duval. Hot-TIG welding was applied using Ni-base alloy
filler. Three different PWHTs were executed and mechanical testing was performed and the
results were compared with material without PWHT. All welds met the requirements of quality
degree B (highest quality) according to EN ISO 5817. The test results (Fig. 5) show that a longer
heat treatment increases the material strength which was attributed to carbide formation and
precipitation of -phase while the toughness decreased with longer PWHT.

Figure 5: Mechanical test results. Measurements performed by TU Darmstadt.


81

For the dissimilar joint of A263 and COST F, the PWHT temperature is very important with
respect to the tempering temperature of COST F. PWHT #2 was chosen as the best option (good
combination of strength and toughness). The hot tensile test results (see figure 10) show that the
tensile strength is in the range of 650 750 MPa for temperatures of 700 - 750C. All samples
were broken in the weld metal.
Based on the experience of welding A263 to A263, a dissimilar weld of A263 to COST F was
made. These were made on 300mmx160mm discs produced by Saarschmiede. Two approaches
were taken: with and without buttering on the COST F side. A nickel based alloy was used both
for filler and buttering material. The COST F material was preheated during buttering. Both
welds were mechanically tested. Tensile tests at room temperature showed that the Rm of cap,
filler and root lies around 800MPa where the value for buttered material is slightly lower than for
material without buttering. The Rp0.2 values lie around 600MPa. Both the Rm and Rp0.2 are
reasonable values for future rotor applications. Charpy impact tests show that the most critical
point is the HAZ between A263 filler metal at the root and cap (impact energies 31 - 72J while
20J is considered a minimum allowable value based on experience). The macrostructure was
investigated at three different locations for both welds and showed no cracks. Also the
microstructure was investigated and both welds meet the requirement of grade B according to EN
ISO 5817. Welding of the 500mm and 1000mm discs is performed at the time of writing of
this paper.
Casting
Materials with sufficient creep strength for cast heavy section components at 750C are A105,
A282, A740, A740H, and A263. A105 has low ductility at room temperature which is a concern
for casting applications and the Al+Ti content raises concerns for their affinity to oxygen causing
oxygen defects with casting in air. A105 was therefore discarded as well as A740 that has more
susceptibility for formation of and G phase than its technically updated variant 740H. A740H,
A263 and A282 were considered materials with at least sufficient strength at 750C, sufficient
room temperature ductility (although values reported in [18] for 740H raise some concerns), good
oxidation performance, reasonable Al + Ti content, and a good perceived weldability (repair and
fabrication) and castability.
The chemistry of the three materials was optimized for casting in conjunction with Special
Metals, Haynes and Loughborough University. Casting was simulated using the finite element
software MagmasoftTM which is a computer program calculating flow, heat exchange, etc., for a
meshed 3D model, on basis of the casting parameters and the material properties. Solidification
simulations carried out to calculate the feed demand of the step blocks (see further) so that the
correct size feeder can be used for the production of each block. The feeder pipe in Ni alloys is
much longer than in conventional materials and it is important that allowance is made for this
phenomenon. Results from simulation and practice are shown in Fig. 6.
Step blocks were cast from the selected alloys (A263, A740H and A282). The mold is produced
with an induction furnace. The aluminum and titanium are added as late additions. As Al and Ti
can form oxides that can get trapped especially in thinner sections causing weak points, the whole
mold and pouring are protected by an inert atmosphere. However, all step blocks castings
produced, to varying degrees, suffered from microstructural cracking during the heat treatment
process, which resulted in volumetric cracks that were detected upon radiographic examination.
The volumetric cracking, although different in each alloy, was always associated with cracked
carbides dispersed along the grain boundaries and often worse in the heaviest sections of the
material. Therefore, the material chemistry and heat treatment of the three alloys was optimized
82

and in cooperation with the FP7 project MACPLUS, new step blocks are cast. Full scale castings
will be made in collaboration with MACPLUS after mechanical testing of the material obtained
from the step blocks.

Figure 6: X-ray of feeders (carbon steel and A625) compared with feeder pipe prediction using
MagmasoftTM for solidification simulation. Courtesy of Goodwin Steel Casting Ltd.
INTEGRATION
Cycle Calculations
An optimal boiler/steam system configuration has to be defined that has a CO2 emission
comparable to a coal fired power plant with 90% CO2 capture and 45% net efficiency requiring
an optimal steam turbine and boiler design. To achieve the required efficiency, the system is
equipped with a double reheat and 750C steam temperatures for all pressure levels. Integration
with the CCS unit is optimized. Design a high efficient power plant and integrate a CCS unit later
on is not optimal and not realistic because optimal integration of a CCS unit will result in a
different boiler and steam turbine design. An integrated approach has therefore been taken. All
other approaches will result in a lower efficiency of the system. Three different candidate steam
turbine designs were studied each with their own advantage. Firstly, the normal cycle (close to
existing cycles) is studied. For optimization of investments, the low pressure (LP) turbine is
reduced in size because the necessary steam for regenerating the solvent is taken just before the
LP-turbine. Disadvantage is that operation without CCS is not possible. Secondly, the master
cycle is taken into consideration. Steam for regenerating the solvent is supplied via the T-turbine,
so that the T-turbine will be larger compared to a design without CCS. With this system, the
steam turbine efficiency increases but it turned out that this system is not able to operate without
CCS and start-up of this system could be difficult. Thirdly, a combination of a normal cycle and a
master cycle is calculated where the necessary steam for the CCS unit is extracted from the cold
reheat and expanded through a separate turbine. This system is able to operate with and without
CCS and the resulting efficiency is in between the normal and master cycle. This cycle is selected
to be applied for the boiler design. As CO2 capture processes the Econamine FG Plus is selected
as state-of-the-art (and only commercial available) technology and is compared with the Cansolv
process which is the most promising technology with respect to low energy consumption in the
future.
The steam turbine and boiler are designed based on the capture techniques. As the steam demand
for the capture processes differs, so does the reheating steam flow and therewith the sizing of heat
83

exchanger surface in the boiler. All calculated metal temperatures (SH 777C, RH 752C,
membrane wall 578C) are within the operating range of the materials with a margin of 20 - 30K
that allows for actual higher metal temperatures in practice. Only 1% biomass (based on thermal
heat input) has to be co-fired to achieve the outlined prerequisites.
Superheater Life Modeling
The objective of the superheat lifetime modeling is to explore the optimization of component
lives while balancing steam conditions and any limits on percentage of co-firing/oxy-firing (and
while maintaining CO2 emissions in the target range (equivalent to >90% reduction for a coalonly plant)). The effects of steam-side degradation and any enhancement due to oxy-combustion
will also be included. The life model includes corrosion models that are derived from extensive
laboratory testing and are a function of gas composition, deposition composition and flux, metal
temperature, time and alloy composition. The model has been validated against plant data,
recognizing the differences between laboratory and plant, e.g. isothermal conditions in the lab vs.
a temperature difference between gas and surface and that deposition in plants is governed by
fuels, design and operating conditions rather than simulated deposits.
CONCLUSIONS
Fireside corrosion testing includes lab screening, lab rig tests and plant trials which are ongoing.
740H outperforms A263 and A617, the latter two perform equally. NiCr coatings performance is
amongst the best. Steamside oxidation screening tests show equal performance of A617, A263
and A740H 750C while A617 and A740H outperform A263 at 800C. LCF-testing of A263 is
ongoing, as well as investigation of relaxation cracking, gas phase embrittlement (indication of it
is found) and negative creep. Full scale rotors have been forged successfully of A263 in reference
and optimized composition. Cast step blocks showed volumetric cracking but new blocks with
optimized composition dont show cracks in the as cast condition. Casting of a full-scale
component is foreseen. An optimal 750C steam cycle with integrated capture has been designed
showing that for a 45% (LHV) efficiency only little biomass has to be co-fired to meet the
required CO2-emission levels. Superheater life modeling for 750C steam and co-firing is
planned work.
ACKNOWLEDGMENTS
The European Union is acknowledged for their financial contribution. Alhomens Kraft, RWE,
and GDF Suez are acknowledged for making their boilers available for corrosion testing.
REFERENCES
[1] Blum, R., Burgge, J., Kjaer, S. USC 700C Power Technology A European Success
Story, VGB Power Tech, Vol. 4 (2009), pp. 26 32.
[2] Viswanathan, R., Henry, J.F., Tanzosh, J., Stanko, G., Shingledecker, J., Vitalis, B., Purgert,
R., U.S. program on materials technology for ultra-supercritical coal power plants,
JMEPEG, Vol. 14 (2005), pp. 281 292.
[3] McDonald, D.K., Coal ash corrosion resistant materials testing program evaluation of the
first section removed in November 2001, Proc 28th Clean Coal Conference on Coal
Utilization and Fuel Systems, Clearwater, FL, USA. 2003.
84

[4] Stein-Brzozowska, G., Flrez, D.M., Maier, J., Scheffknecht, G., Fireside corrosion of
dedicated austenitic steels in ultra-supercritical coal-fired power plants, Proc 37th
International Technical Conference on Clean Coal and Fuels Systems, Clearwater, FL, USA.
2012.
[5] Stein-Brzozowska, G., Flrez, D.M.; Maier, J., Scheffknecht, G., Nickel-base superalloys in
ultra-supercritical coal-fired power plants: fireside corrosion. Laboratory studies and power
plant exposures, Fuel, Vol. 108, No. 62 (2012), pp. 521 - 533.
[6] Spiegel,
M.,
Krein,
R.
Werkstoffentwicklung
fr
Kohlekraftwerke

Hochtemperaturkorrosion und Festigkeit. Proc Dampferzeugerkorrosion 2011, Freiberg,


Germany. 2011.
[7] Hussain, T., Simms, N.J., Oakey, J.E, Fireside corrosion of Ni-based alloys in simulated cofiring combustion environment, Proc Baltica IX, Helsinki Stockholm Helsinki, 11 13
June. 2013.
[8] Hussain, T.; Dudziak, T.; Simms, N.J.; Nicholls, J.R, Fireside corrosion behavior of HVOF
and plasma-sprayed coatings in advanced coal/biomass co-fired power plants, J. Therm.
Spray Techn., Vol. 22, No.5 (2013), pp. 797 807.
[9] Oksa, M., Tuurna, S., Varis, T., Increased lifetime for biomass and waste to energy power
plant boilers with HVOF coatings: High temperature corrosion testing under chlorinecontaining molten salt, J. Therm. Spray Techn., Vol. 22 (2013), pp. 783 796.
[10] Oksa, M., Krki, J., Metsjoki, J., Coating solutions against high temperature corrosion
performance validation and feasibility at biomass fired boilers, Proc Baltica IX, Helsinki
Stockholm Helsinki, 11 13 June. 2013.
[11] Agero, A., Muelas, R., Gutirrez, M., Van Vulpen, R., Osgerby S., Banks, J.P., Cyclic
oxidation and mechanical behaviour of slurry aluminide coatings for steam turbine
components, Surf. Coat. Tech. Vol 201 (2007), pp. 6253 - 6260.
[12] Prez, F.J., Hierro, M.A., Trilleros, J.A., Carpintero, M.C., Snchez, L., Brossard J.M.,
Bolvar, F.J., Iron aluminide coatings on ferritic steels by CVD-FBR technology,
Intermetallics, Vol. 14 (2006), pp. 811-817.
[13] Ren, D.G., Atom probe and field ion microscope investigation of the negative creep
mechanism in nickel-base superalloys, J.Mat. Proc., Vol. 73 (1998), pp. 74 77.
[14] Smith, S.A., West, G.D., Chi, K., Gamble, W., Thomson, R.C., Microstructural evolution in
Nimonic 263 for high temperature power plant, Proc Sixth International Conference
Advances in Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, USA.
2010.
[15] Zhang, Y.H., Knowles, D.M., Prestraining effect on creep behaviour of nickel base C263
superalloy, Mater. Sci. Technol., Vol. 18 (2002), pp. 917-923.
[16] Patel, A.D., Kelkar, K.M., New insights into the electro-slag remelting process using
mathematical modeling, In: Modeling of Casting, Welding, and Advanced Solidification
Processes XII. Ed. Steve L. Cockcroft and Daan M. Maijer. TMS (The Minerals, Metals &
Materials Society). 2009.
[17] Schafstall, H., McBain, G., Barth, C., Terhaar, J., Jarolimeck, J., Automatisierte Simulations
der Prozesskette vom Blockguss zum Wrmebehandelten Stahl, Stahl und Eisen, Vol. 10
(2010), pp. 57-62.
[18] P.J. Maziaz, N.D. Evans, J.D. Jablonski, High-temperature mechanical properties and
microstructure of cast Ni-based superalloys for steam turbine casing applications, Proc
Sixth International Conference Advances in Materials Technology for Fossil Power Plants,
Santa Fe, New Mexico, USA. 2010.

85

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

A STEAM GENERATOR FOR 700C TO 760C


ADVANCED ULTRA-SUPERCRITICAL
DESIGN AND PLANT ARRANGEMENT:
WHAT STAYS THE SAME AND WHAT NEEDS TO CHANGE
Paul S. Weitzel
Babcock & Wilcox Power Generation Group, Inc., Barberton, Ohio, U.S.A.

ABSTRACT
Increasing the efficiency of the Rankine regenerative-reheat steam cycle to improve the
economics of electric power generation and to achieve lower cost of electricity has been a long
sought after goal. Advanced ultra-supercritical (A-USC) development for materials to reach 760C
(1400F) is a goal of the U.S. Program on Materials Technology for Ultrasupercritical Coal-Fired
Boilers sponsored by the United States (U.S.) Department of Energy and the Ohio Coal
Development Office (OCDO). As part of the development of advanced ultra-supercritical power
plants in this program and internally funded programs, a succession of design studies have been
undertaken to determine the scope and quantity of materials required to meet 700 to 760C (1292
to 1400F) performance levels. At the beginning of the program in 2002, the current design
convention was to use a two pass steam generator with a pendant and horizontal tube bank
arrangement as the starting point for the economic analysis of the technology. The efficiency
improvement achieved with 700C (1292F) plus operation over a 600C (1112F) power plant
results in about a 12% reduction in fuel consumption and carbon dioxide emissions. The reduced
flue gas weight per MW generated reduces clean up costs for the lower sulfur dioxide, nitrogen
oxides and particulate emissions.
The operation and start up of the 700C (1292F) plant will be similar in control methods and
techniques to a 600C (1112F) plant. Due to arrangement features, the steam temperature control
range and the once through minimum circulation flow will be slightly different. The expense of
nickel alloy components will be a strong economic incentive for changes in how the steam
generator is configured and arranged in the plant relative to the steam turbine. To offer a view
into the new plant concepts this paper will discuss what would stay the same and what needs to
change when moving up from a 600C (1112F) current state-of-the-art design to a plant design
with a 700C (1292F) steam generator and turbine layout.

INTRODUCTION
Much of the previous work published on the development of A-USC steam generators has
covered the materials developments needed to make operation possible at greater than 700C
(1292F) [1, 2]. The development effort has also included a task for conceptual design of the
steam generator in order to provide information on material temperature conditions and an
estimate of size and quantity of the components [3]. The arrangement configurations started with
current two-pass designs used for 600C (1112F) ultra-supercritical (USC), as this meets current
acceptance and expectations of the industry. The two pass style has pendant surface over the
furnace and arch and horizontal heating surface in a down pass that is divided for gas biasing to
86

control reheat steam temperature. With A-USC there will be about 150C (300F) increase in the
outlet mainsteam temperature and the furnace wall enclosure will deliver about 66C (150F)
higher fluid outlet temperature over the present supercritical design, which has been in U.S.
utility service for more than 55 years. Selection of higher alloy materials for the lower and upper
furnace enclosure and the convection pass enclosure and tube banks is a primary activity of the
sponsored A-USC boiler materials development program and Babcock & Wilcox Power
Generation Group, Inc. (B&W PGG) research and development activities [1-5]. The program
includes participation of the U.S. boiler suppliers on the basis of being limited to pre-competitive
information needed for ASME data and the understanding of material/mechanical properties
needed industry wide. The power industry needs solutions which demonstrate that operation at
greater than 700C (1292F) will be economic, sustainable and will deliver the higher efficiency
expected.
Due to the high cost of nickel steam piping, new concepts have been proposed for the
arrangement of the steam generator and the turbine in the plant. Siemens AG has proposed a
horizontal boiler. China has some interest in the two-turbine concept with the HP and IP sections
at high elevation by the steam outlet headers and the LP section and condenser on the ground
level. Horizontal tube banks could provide a means to better handle the exfoliation of steam side
oxidation where distribution and removal can be improved. The higher feedwater temperature
from A-USC turbine cycles makes it difficult to achieve low boiler outlet gas temperature
entering the air heater. The soot blowing steam requirements could be met by a lower pressure
steam generator, similar in function like a two pressure heat recovery steam generator (HRSG).
Soot blowing with compressed air would save high pressure water. Flue gas cooling using
condensate may be needed. Some new revisited designs will be presented later.
The expectation that some steam generator design changes will occur is valid. These must be
understood and gain acceptance through proof of performance. The U.S power generation
industry is particularly conservative and functions in a risk-averse commodity market with
constraining forces that seek to achieve lower electricity prices. Still the industry needs to remain
economically vital by advancing technical development. This paper provides a discussion of what
will certainly change, what will probably change and what should remain unchanged in the plant
and steam generator design.
A-USC DESIGN TECHNOLOGY DEVELOPMENT
Supercritical Furnace Enclosure Design for 540C (1005F)
B&W supercritical 540C (1005F) once-through steam generators use SA213 T-2 for the entire
vertical tube furnace enclosure. The enclosure is called UP-UP, having a first and second pass of
alternating tubes from the lower inlet headers so the fluid moves up twice through the same
elevation, then a third pass continues up to the roof [6]. Exit fluid temperature reaches about
385C (725F). Enclosure circulation requires increasing pressure to 24.1 MPa (3500 psi)
supercritical operation (called constant pressure) after the convection pass outlet reaches about
232C (450F) at 6.9 MPa (1000 psi). This prevents two-phase flow overheating due to boiling
crisis (dryout). On start up, a division valve separates the enclosure from the superheater until the
fluid conditions are high enough to produce steam by throttling the fluid into the start up flash
tank and the superheater. When the superheater is pressurized to match the upstream furnace
enclosure, the division valve is full open and the flash tank system is taken out of service.

87

Ultra Supercritical (USC) Furnace Enclosure Design for 600C (1112F)


The present state-of-the-art design is the USC steam generator [6]. The USC furnace enclosure
and superheater operate on a variable pressure ramp through the upper load range. The evaporator
end point location is variable with load so the upper outlet portion provides superheating duty at
higher load. The full arc turbine inlet valves are nearly wide open (VWO) or with some throttle
reserve, on a control program from about 30% or 35% up to 95100%. The purpose of this
method of control is to not throttle and reduce the turbine inlet steam temperature at partial load
operation. The higher steam temperature 600C (1112F) USC design is more efficient and more
suitable for load cycling. The furnace enclosure wall may be spiral or vertical tube construction if
the flow to perimeter ratio is high enough. The enclosure operates with two phase flow conditions
at lower loads and the water/steam vertical separators run dry after reaching the Benson point
where all steam enters the separator, at about 30% load. The enclosure tube materials are SA213
T-12 for the entire B&W USC furnace. Post weld heat treating (PWHT) is not required after
construction with this material. This enclosure design is used at American Electric Power,
Southwestern Electric Power Companys John W. Turk, Jr. plant with 600C (1112F) steam
conditions. The average fluid outlet temperature of USC is about 414C (775F) to less than 427C
(800F). Subcritical pressure operation during start up is permitted due to the spiral design which
better accommodates the heat upset and unbalances. The furnace enclosure design must be able to
handle appropriately located boiling crisis. Multi-lead ribbed tubing is used to allow lower
minimum circulation flows at low load. This permits a wider mainsteam temperature control
range.

Figure 1 B&W State-of-the-Art 600C (1112F) USC Steam Generator


The USC start up system uses vertical steam separators (VS) and a water collection tank (WCT)
along with a boiler circulation pump (BCP) to return VS water drains back to the economizer
inlet. The system operates up to the Benson point much like a pump assisted natural circulation
boiler. There is no division valve between the enclosure and the superheater. Above the Benson
point to full load, the VS remains in service at pressure.

88

Advanced Ultra Supercritical (A-USC) Furnace Enclosure Design for +700C


(1292F)
The B&W 700 760C (1292 1400F) A-USC steam generator is developed with features and
characteristics that are the same as the 600C (1112F) USC steam generator. With higher steam
temperature there is more superheating than steam generation (evaporation) duty, and the gas-tosteam temperature difference is lower so more convection bank heating surface is needed and the
superheating duty of the enclosure is more. The design requirements of the fuel are setting the
constraints on flue gas side which are the same for supercritical, USC and A-USC furnaces. The
furnace enclosure material is currently selected to allow operation to 454C (850F) to less than
510C (950F) average outlet fluid temperature. This is about +66C (150F) for the enclosure and
+93 C to 149C (200F to 300F) for the final superheater over the temperature for USC. Achieving
up to 760C (1400F) final superheat will require T-92 in the enclosure and 740H nickel alloy tubes
and headers for the superheater, reheater and steam piping. A-USC produces a lower heat rate and
lower carbon dioxide emission starting at about a 12% improvement over 600C (1112F) USC.
The price of nickel is high enough that new arrangements for the steam generator and the position
of the steam turbine must be a consideration.
A-USC upper enclosure walls will need to use creep strength enhanced ferritic (CSEF) steels, T23, T-91, T-92 requiring post weld heat treatment. Coatings on these materials have been
included in fireside corrosion and steam oxidation testing programs [2]. B&W has made large T23 and T-92 test panels, Figure 2, for lab weld trials to further develop fabrication practices with
shop burner inserts and smaller openings, and construction methods and field repairs [4]. This
included the development of PWHT procedures and requirements. T-91 division walls, floor
panels, and convection pass enclosure walls have already been used in eight retrofit projects at
existing B&W and competitor units. T-23 walls on competitor units are already in service so
further experience with this material will be become available. There are 600C (1112F) double
reheat units in progress in China that will use these CSEF steels (T-23, T-91) because the
additional second reheat absorption requirements push up the required outlet temperature of the
furnace enclosure.

Figure 2 T-92 Wall Panel for Welding Fabrication, Installation and Repair Testing
89

Operating Concept for Variable Pressure (A-USC) Steam Generator +700C (1292F)
New steam generator designs are expected to be capable of variable pressure operation. The
purpose of variable pressure, also known as sliding pressure, is to not change the temperature of
the steam turbine high pressure components with load cycling. Figure 3 shows the enthalpypressure operating domain results for an A-USC steam generator with suitable partial load
characteristics. A constant pressure steam generator can maintain the turbine inlet temperature.
However, the turbine control valve will throttle the steam at partial loads and thereby lower the
metal temperature during load cycling. Variable pressure operation means the turbine inlet valves
operate nearly wide open and do not throttle the steam resulting in maintaining high full load
steam temperature and metal temperatures of the turbine and superheater outlet headers and
piping at partial loads. Lower temperature intermediate components of the steam generator will
cross some isotherms during load cycling while the higher temperature outlet components do not
cross very many isotherms until shut down.

Figure 3 B&W A-USC 700C (1292F) Steam Generator Operating Domain h-P

90

Variable pressure is an important operating mode capability due to the increase to higher steam
temperature with A-USC and the situation with electric grids needing units to load cycle. The
load change rate of response will be slower than a constant pressure with some throttle pressure
reserve. Variable pressure needs to use additional methods to improve load change response. A
technique called frequency control or condensate throttling is used to quickly start the load
change without serious over-firing from the delayed control action. Condensate throttling works
by using fast valves on the turbine steam extraction lines to the LP feedwater heaters so the
turbine quickly retains more steam to produce the power increase. The slow coal pulverizers can
then meet the firing demand for the load change without the catch up delay creating severe
temperature overshoot and damage. Firing rate control should not be too fast. Condensate
throttling control is fast. Water storage design of the hotwell and deareator are increased to
handle the condensate flow transient.
First B&W Conceptual A-USC Design Air Fired with Gas Biasing
The first B&W conceptual design was developed to provide the materials quantity and service
temperature requirements for early cost analysis in the materials development program [3]. The
steam conditions produced 750 MW net at 34.5 MPa (5003 psi) 735C/760C (1355/1400F). The
design fuel was an Ohio, U.S. bituminous coal. This once through steam generator arrangement is
the most typical in U.S. utility service and is called a two pass or B&W Carolina type, Figure 4.
The top supported pendant heating surface may be set on any spacing intervals and is not
governed by stringer supports. The down pass is baffled to provide for gas biasing to control
reheat steam temperature through a wider load range.
The gas side operation is set by the fuel design limitations. For the design fuel, the gas velocity
limits are high so the convection heat transfer is more effective. The lower differential
temperature between the gas and steam increases the heating surface requirements. The steam
outlet terminals are at the very top of the structure.

Figure 4 B&W A-USC Conceptual Study-Two Pass Air Fired Steam Generator
91

B&W Conceptual A-USC Design Oxygen Fired Series Back Pass


A follow-on B&W conceptual design was developed using oxygen combustion to provide the
conditions and materials quantity and service temperature requirements for a comparison to the
earlier air fired design. Again the steam conditions produced 750 MW net at 34.5 MPa (5003
psi), 735C/760C (1355/1400F). The design fuel was an Ohio bituminous coal. The once through
steam generator arrangement selected was a two pass type with an in-series down pass, Figure 5.
Oxygen combustion uses 95% oxygen to replace air and recycles flue gas to increase the carbon
dioxide concentration. Cleaning and compression of the flue gas will prepare the carbon dioxide
for pipeline transportation and deep well sequestration. The fuel dictates the type of gas
recirculation system employed: cold, cool and warm. Hot gas recirculation is not a consideration
because of the concentration of the fuel sulfur and chloride. Gas from the economizer outlet was
used in the early once through supercritical units. Gas recirculation (GR) dilutes the furnace gas
temperature and maintains similar gas flow rates that provide good convection heating surface
effectiveness. The reheat steam temperature control with GR is also effective and within the
operating parameters of the two constraints. The required heating surface in the down pass can be
less than the two-pass, parallel-path, gas-biasing design. The furnace absorption and convection
pass absorption is enhanced by the higher carbon dioxide due to the higher density and specific
heat.

Figure 5 B&W A-USC Conceptual Study-Two Pass Oxy Fired Steam Generator
B&W A-USC Conceptual design- 898 MW Modified Tower
Figure 6 is a modified tower design that was developed for a Toshiba steam turbine 30 MPa
(4351 psi), 700C/730C (1292F/1346F). The design fuel is an Indian coal [5]. Featured is the use
of stringer supports for horizontal surface. All bank heating surface in the gas upflow pass must
be on a multiple of the interval spacing of the stringers. The material of the stringers is T-92
92

supporting nickel superheater and reheater tubing. The down pass is also stringer supported and is
typical of current practice. A series down pass may be used when the reheat temperature control
range can be limited. A parallel down pass is used when a wider temperature control range is
required. The high ash coal requires much lower gas velocity limits so the convective heat
transfer degrades quickly with reduced load. Gas recirculation and gas biasing may be employed
to meet performance requirements. A low pressure steam generator is placed after the economizer
to reduce exit gas temperature and provide auxiliary steam instead of extracting very high
pressure and high temperature steam.
This modified tower arrangement puts the outlet headers closer to the steam turbine although, due
to the high ash fuel the height, of the furnace is much higher. By locating the base of the steam
generator below grade the nickel steam leads are shorter and offer significant savings. The
feedwater heaters and deareator building bay may also be moved from between the steam turbine
and steam generator to along one side of the steam generator.

Figure 6 B&W A-USC Conceptual design- 809 MW modified tower


Other Changes to the A-USC Configuration

Steam generator configurations will most likely evolve because of the high nickel steam lead
cost. The first Babcock & Wilcox once through steam generator was a low profile, nearly
horizontal configuration, Figure 7 (UP-1 for American Electric Powers 125 MW Philo 6, 31.4
MPa /621C/565C/538C (4550 psi /1150F/1050F/1000F) [6]. Some of these new forms have
already started appearing for A-USC alternatives to the Figure 1 convention. The horizontal
93

boiler was proposed by Siemens for the planned Thermie 700C European demonstration plant. A
conventional tower arrangement was also proposed. A plant arrangement might change with the
entire steam turbine partially elevated relative to the boiler. Another solution has been suggested
to divide the steam turbine where the HP and IP sections are placed at high elevation closer to the
superheater and reheat outlet headers, and the second part, the LP sections and condenser remain
on the ground. For A-USC to become economical, the steam generator and steam turbine
configuration and relative placement is most likely going to change.

Figure 7 Babcock & Wilcox UP-1 AEP Philo 6 125 MW 31.4 MPa /621C/565C/538C
Comparison of Supercritical, Ultra Supercritical, Advanced Ultra Supercritical
Table 1 shows a comparison of the key features of supercritical, USC and A-USC. Most A-USC
features are the same as for USC. Particular exceptions are:
1. Final superheater and reheater tube banks will use materials like 740H and 230 nickel.
2. The furnace enclosure material in the upper enclosure is creep strength enhanced ferritic
steel requiring field PWHT of the tube to tube joints and possibly the membrane panel
seams.
3. Minimum circulation flow load is more likely 5% to 10% higher than USC which limits
temperature control load range.
4. Steam piping is 740H nickel or better.
5. The arrangement will continue to evolve.

94

Table 1 Technology Comparison of B&W Supercritical, Ultra-Supercritical and


Advanced Ultra-Supercritical Steam Generators

Feature

Supercritical

Ultra Supercritical

Advanced Ultra Supercritical

Steam Pressure &


Temperature
Pressure Control

24MPa (3500psi) 540C to


565C (1005F to 1050F)
Constant Pressure

~25.5MPa (3700psi)
600C to 650C(1112F to 1200F)
Variable Pressure

~30MPa (4350psi)
700C to 760C (1292F to 1400F)
Variable Pressure

Load Control

Unit Load Master


Feed forward to Firing and
Feedwater Flow
Feedwater: Firing Rate
Ratio, nominal single stage
spray attemperation
Feed forward with Unit
Load Master
All Volatile Treatment
(AVT) with full condensate
polishing
Some use OWT
5 to 10% on mainsteam

Firing Rate Demand


Some use of Frequency Control
(Condensate Throttling)
Multiple Stage Spray
Attemperation vs. Load
Program
Furnace Enthalpy Differential
Pick Up vs. Load & Trim with
First Stage Attemperator
Differential Temperature
Oxygenated Water Treatment
(OWT) with full condensate
polishing
AVT used in early operation
40 to 100% HP & LP

Firing Rate Demand


Some use of Frequency Control
(Condensate Throttling)
Multiple Stage Spray
Attemperation vs. Load
Program
Furnace Enthalpy Differential
Pick Up vs. Load & Trim with
First Stage Attemperator
Differential Temperature
Oxygenated Water Treatment
(OWT) with full condensate
polishing
AVT used in early operation
40 to 100% HP & LP

Division & Throttling


Valves between furnace &
superheater,
1500psi Flash Tank with
steam & water drain
recovery system [6]
Pump Minimum Furnace
Circulation Flow 25% to
33%. Superheater shut off
and throttling valves
closed. Bypass valves
throttle water to flash tank
which returns steam to SH.
When upstream enthalpy
high enough to pressurize
the SH, Throttling division
valve opened on a program
to raise superheater to 24.1
MPa (3500 psi) (Once
through operation).

Vertical Separator (VS) &


Water Collection Tank (WCT)
with Boiler Circulation Pump
(BCP) [6]. WCT level control
valve to condenser for flow up to
7%.
Pump Minimum Furnace
Circulation Flow (30% to 35%)
with a minimum feedwater flow
of 7% and use of the BCP
controlling WCT level. V is in
service up to full pressure.
Feedwater flow meets minimum
as BCP handles less and less
water drains.
BCP is shut off above Benson
Load point, about 30 to 35%
load. Once through operation
feedwater control listed above
becomes highest demand.

Vertical Separator (VS) &


Water Collection Tank (WCT)
with Boiler Circulation Pump
(BCP). WCT level control valve
to condenser for flow up to 7%.

Arrangement
Configuration
Piping Material

Two Pass & Tower

Two Pass

P22

P92

Two Pass, Tower & Modified


Tower
740H nickel

Furnace Enclosure
Material

T-2 smooth & multi-lead


ribbed, vertical multi-pass
UP-UP furnace with 1st to
2nd pass full or partial mix
and 2nd to 3rd pass full or
partial mix
T-22, 304H

T-12 smooth & multi-lead


ribbed spiral lower furnace
(vertical tube lower furnace
needs T-23 or high flow per foot
perimeter), vertical upper
furnace above transition
Previous Column Plus
T-91, T-92, 347HFG, 310HCbN

T-12 T-22 T-23 T-91 T-92


smooth and multi-lead ribbed
lower furnace (spiral or vertical
based on flow per foot of
perimeter), vertical upper
furnace above transition
Previous Column Plus
Super 304H, 230, 740H

Steam Temperature
Control
Feedwater Control

Water Treatment
Chemistry
Turbine Bypass
Start Up System

Start Up Operation

Superheater Material

95

Pump Minimum Furnace


Circulation Flow (40% to 45%)
with a minimum feedwater flow
of 7% and use of the BCP
controlling WCT level. V is in
service up to full pressure.
Feedwater flow meets minimum
as BCP handles less and less
water drains.
BCP is shut off above Benson
Load point, about 40 to 45%
load. Once through operation
feedwater control listed above
becomes highest demand.

CONCLUSIONS
An important next step in A-USC development is to build at significant scale and demonstrate the
features that extend beyond present USC experience. The first demonstration will test the
capability of suppliers to support the new materials required. Confidence in meeting the quantity
and schedule commitments for plant projects is needed by power plant suppliers. Material
suppliers will need to make investments based on increased certainty of the timing when the AUSC market demand will form. The first lead plants will establish procurement standards and
quality specifications for A-USC components. It is also important to place the technology into the
hands of utility operations and maintenance personnel to gain industry acceptance. The most
important step is to provide a plant design that is high performing and cost acceptable.
ACKNOWLEDGMENTS
The U.S. Department of Energy and the Ohio Coal Development Office support for the A-USC
Materials Development project is greatly appreciated. The efforts of industry-wide organizations
have fostered an environment of cooperation in working toward the common pre-competitive
needs for ASME Code materials development for A-USC. The goal of these efforts is to advance
the state-of-the-art for higher efficiency power generation using abundant low cost coal while
reducing carbon emissions of a major energy source of electric power.
REFERENCES
[1] Viswanathan, R., et.al., U.S. Program on Materials Technology for Ultrasupercritical CoalFired Boilers, in Proceedings of the 5th International Conference on Advances in Materials
Technology for Fossil Power Plants, ASM International, (2008).
[2] Rao, K.R., (ed.), Energy and Power Generation Handbook, ASME, (New York, 2011),
Chapter 17.
[3] Bennett, A.J., Weitzel P.S., Boiler Materials for Ultrasupercritical Coal Power Plants Task 1B,
Conceptual Design, Babcock & Wilcox Approach, USC T-3, Topical Report, DOE DE-FG2601NT41175 & OCDO D-0020, (February 2003).
[4] Weitzel, P.S., Steam Generator for Advanced Ultra-Supercritical Power Plants 700 to
760C, ASME Power 2011, Denver, CO, 2011.
[5] Weitzel, P.S., et al., Advanced Ultra-Supercritical Power Plant (700 to 760C) Design for
Indian Coal, Power Gen Asia, Bangkok, Thailand, October 2011.
[6] Kitto, J.B, Stultz, S.C., Steam/its generation and use, Edition 41, The Babcock & Wilcox
Company, (Barberton, OH, 2005).
Benson is a registered trademark of Siemens AG
B&W and Babcock & Wilcox are registered trademarks of Babcock & Wilcox Power Generation
Group, Inc.
Copyright 2013 Babcock & Wilcox Power Generation Group, Inc. All rights reserved.

96

DISCLAIMER
Although the information presented in this work is believed to be reliable, this work is published
with the understanding that the Babcock & Wilcox Power Generation Group, Inc. (B&W PGG)
and the authors are supplying general information and are not attempting to render or provide
engineering or professional services. Neither B&W PGG nor any of its employees make any
warranty, guarantee, or representation, whether expressed or implied, with respect to the
accuracy, completeness or usefulness of any information, product, process, method, or
apparatus discussed in this work, including warranties of merchantability and fitness for a
particular or intended purpose. Neither B&W PGG nor any of its officers, directors, or
employees shall be liable for any losses or damages with respect to or resulting from the use
of, or the inability to use, any information, product, process, method, or apparatus
discussed in this work.

97

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

AN INVESTIGATION ON STRUCTURE STABILITY OF ADVANCED


AUSTENITIC HEAT-RESISTING STEELS AND NI-BASE
SUPERALLOYS FOR 600-700 A-USC POWER PLANT
APPLICATION
Xishan Xie, Chengyu Chi, Qiuying Yu, Zhihao Yao, Maichang Zhang, Yaohe Hu,
Jianxin Dong
University of Science & Technology Beijing, Beijing 100083, China
Hongyao Yu
Central Iron & Steel Research Institute, Beijing 100081, China
Shuangqun Zhao, Fusheng Lin
Shanghai Power Equipment Research Institute, Shanghai 200240, China
Xia Liu, Linbo Mei
Shanghai Electric Power Generation Equipment Co., Ltd, Shanghai 200240, China
Huachun Yang
Dongfang Boiler(Group) Co. Ltd, Zigong, Sichuan 643001, China
Mingyang Li
Beijing Beiye Functional Materials Corporation, Beijing 100192, China
ABSTRACT
This overview paper contains a part of structure stability study on advanced austenitic heatresisting steels (TP347H, Super304H and HR3C) and Ni-base superalloys (Nimonic80A,
Waspaloy and Inconel740/740H) for 600-700 A-USC fossil power plant application from a
long-term joint project among companies, research institutes and university in China. The long
time structure stability of these advanced austenitic steel TP347H, Super304H, HR3C in the
temperature range of 650-700 and Ni-base superalloys Nimonic80A, Waspaloy and
Inconel740/740H in the temperature range of 600-800 till 10,000h have been detailed studied
in this paper.
INTRODUCTION
Ultra-Super-Critical(USC) fossil power plants are developed world-widely for raising thermal
efficiency and reduction of CO2 emission. China had put her first USC power plant in service with
steam parameter of 600 and 25MPa in the year of 2006. Up to now China has established more
than 100 USC electric power units with 600 steam temperature[1]. Advanced austenitic steels
TP347H, Super304H and HR3C are the most important high temperature materials for China
600 USC power plants.
China has also initiated a 700 advanced ultra-super-critical(A-USC) technology project in the
year 2011. The highest temperature components in this project are superheater and reheater
tubings. Their metal temperature can reach 750 even higher. The Ni-Cr-Co-Mo-Nb-Ti-Al type
superalloy Inconel740H characterizes with better structure stability at high temperatures in
comparison with Inconel740.
Inconel is trademark of the Special Metals Corporation group of companies
98

The highest temperature components of 600-700 A-USC steam turbines are different stage
blades. These blade materials can be Nimonic80A and Waspaloy.
The long time structure stability of these advanced austenitic steels TP347H, Super304H, HR3C
in the temperature range of 650-700 and Ni-base superalloy Inconel740/740H, Nimonic80A
and Waspaloy in the temperature range of 600-800 till 10,000h have been detailed studied in
this paper.
MATERIALS AND EXPERIMENTS
Materials
The nominal chemical compositions of 7 investigated advanced heat-resistant steels and
superalloys are listed in Table1. They are namely in two groups of austenitic steels TP347H,
Super304H, HR3C and Ni-base superalloys Nimonic80A, Waspaloy and Inconel740/740H.
Table 1: Nominal chemical compositions of 7 investigated austenitic heat-resistant steels and Nibase superalloys(in Mass%)
Alloy
C
Cr
Ni
Co
Fe
Mo
Nb
Ti
Al
Cu
N
TP347H
0.08 18
10
Bal. 0.8
Super304H
0.08 18.5 9.5
Bal. 0.5
3
0.1
HR3C
0.06 25
20
Bal. 0.5
0.2
Nimonic80A 0.06 20
Bal. 2.4
1.7
Waspaloy
0.04 20
Bal. 14
4.5
3
1.5
Inconel740
0.03 25
Bal. 20
0.5
2
1.8
0.9
Inconel740H 0.03 25
Bal. 20
0.5
1.5
1.35 1.35 Heat treatments of investigated materials are as follows:
o
1. TP347H: 1150 C 30/WC
o
2. Super304H: 1150 C /30/WC
o
3. HR3C: 1150 C /30/WC
o
o
o
4. Nimonic80A: 1080 C /8h/AC+845 C /24h/AC+700 C /16h/AC
o
o
o
5. Waspaloy: 1080 C /4h/AC+845 C /24h/AC+760 C /16h/AC
o
o
6. Inconel740: 1150 C /30/WC+800 C /16h/AC
o
o
7. Inconel740H: 1150 C /30/WC+800 C /16h/AC
Experiments
All advanced austenitic steels TP347H, Super304H and HR3C tubes were received from steel
plants after mill annealing and followed by long time thermal aging at 650 C till 8,000-10,000hrs
for TP347H and Super304H and aging at 700 C till 5,000hrs for HR3C.
Ni-base superalloy Nimonic80A for 600C USC steam turbine blade application was conducted
long time aging at 550C to 750C for 10,000hrs.
Ni-base superalloy Waspaloy is selected for 700C A-USC steam turbine blade application was
conducted long time aging at 550C to 800C for 10,000hrs.
Ni-base superalloy Inconel740/740H for 700C A-USC boiler superheater/reheater application
was conducted long time aging at 700,750 and 800C for 10,000hrs.
Structure characterization of long time aging samples of investigated materials were analyzed by
SEM, TEM, HRTEM, XRD and EDAX. For quantitative determination of precipitates in several
investigated materials by electrolytic isolation of precipitates and followed by micro-chemistry
99

analyses. The precipitation behavior of Cu-rich phase in Super304H was analyzed by 3


dimensional atomic probe(3DAP) in detail. Mechanical properties of hardness, impact toughness
and stress rupture lives were also tested in some investigated materials.
RESULTS AND DISCUSSIONS
Austenitic steels
Austenitic heat-resistant steels TP347H, Super304H and HR3C are wildly used for 600 USC
power plant boiler superheater/reheater tubing application not only in China but also all over the
world. Fig.1 shows the precipitation hardening effect of TP347H and Super304H at 650 long
time aging. Fig.2 shows the precipitation effect of HR3C at 700 long time aging. The
precipitation hardening of these 3 austenitic steels are quite similar. These austenitic steels can
quickly reach the peak hardness in 1,000h aging and still keep high level hardness at 600 and
700 long time aging for TP347, Super304H and HR3C respectively. However, Super304H
characterizes with the highest stress rupture strength and allowable stress among these 3 austenitic
steels(see Table 2).

Figure 1: Precipitation hardening curves of TP347H and Super304H at 650C long time aging
till 8,000-10,000h

Figure 2: Precipitation hardening curves of HR3C at 700C long time aging till 5,000h
100

Table 2: 650C stress rupture strengths of TP347H, Super304H and HR3C


Steel
TP347H
Super304H
Minimum 105h stress rupture strength, MPa
86.9
104
Allowable stress/0.8, MPa
76.3
98.0

HR3C
98.1
84.0

Typical microstructure of these 3 austenitic steels are nano-size strengthening phases precipitated
in austenitic matrix and M23C6 carbide distributed at grain boundaries. The main precipitation
strengthening phase for TP347H in austenitic matrix is NbC phase as shown in Fig.3 by high
resolution transmission electron microscope(HRTEM). The precipitation strengthening phases in
Super304H are Cu-rich phase, Nb(CN) type MX phase in austenitic matrix and M23C6 carbide
mainly at grain boundaries and occasionally in grains as shown in Fig.4. It can be seen that Curich phase particles with the finest sizes homogeneously distribute in austenitic matrix, and Curich phase occupies the largest amount of the total precipitates. The high resolution TEM image
of Cu-rich phase can be seen in Fig.5. Except MX the main precipitation strengthening phase in matrix phase of HR3C is nano-size NbCrN. The HRTEM image of NbCrN is clearly shown in
Fig.6.

Figure 3: HRTEM image of nano-size NbC in -matrix of TP347H

500nm

Cu-rich phase: 3%
MX: 0.38%
M23C6: 0.99%
Figure 4: The main precipitates M23C6, Cu-rich phase and MX in Super304H

101

Figure 5: HRTEM image of Cu-rich phase in Super304H

Figure 6: HRTEM image of nano-size NbCrN in -matrix of HR3C


Fig.7 shows the precipitation hardening behavior of Super304H at 650C aging from 1h till
1,000h. TEM image and EDAX analyses have confirmed the precipitation of Cu-rich phase with
average size of about 10nm in Super304H after 650C aging for 1,000h(see the middle picture of
Fig.4)[2]. At the very early stage of 650C aging for 1-5h TEM images can not clearly show the
existence of Cu-rich phase because of its extremely fine particle sizes.

102

Figure 7: Typical precipitation hardening curve of Super304H at 650 aging till 1,000h
The precipitation procedure of very fine Cu-rich phase can be described by 3DAP technology[3].
o
The Cu atoms have quickly segregated to form Cu-rich clusters just after 1h aging at 650 C as
o
shown in Fig.8. Fig.9 shows the growth of a Cu-rich particle after 100h aging at 650 C. The Cu
atoms continuously concentrate to Cu-atom segregated regions(Cu-rich clusters). As Fig.9 and
o
Fig.10 show that the Cu-content(at%) has almost reached 90at% in a Cu-rich phase at 650 C
aging for 500h. The Cu-rich phase formation from Cu atom segregated regions(Cu-rich clusters)
characterizes with an evolution process, because there is Cu atom concentration process only but
no crystallographic structure change. Cu-rich phase and Fe-Cr-Ni -matrix both characterize with
FCC structure and they are only in a small difference of crystallographic parameters(small misfit
between Cu-rich phase and -matrix) and keep in coherent relationship between Cu-rich phase
o
and -matrix. The growth rate of Cu-rich phase at 650 C aging is quite slow as shown in Fig.11.
o
The Cu-rich phase can still keep nano-size(34nm) even at 650 C aging for 10,000h. It is an
important reason for excellent strengthening effect of Cu-rich phase precipitation in Super304H
heat-resistant steel[2,4].

Solution treatment

650C /1h
Figure 8: The concentration of Cu atoms to form Cu-rich clusters in Super304H just after 1h
o
aging at 650 C (size of selected box is10nm10nm70nm)

103

Figure 9: The Cu atom concentration procedure during 650 aging from 5h to 500h: (a)5h,
(b)100h and (c)500h
(a)

(b)

(c)

Figure 10: The precipitation procedure and formation of Cu-rich phase in Super304H: (a) after
solid solution treatment (10nm10nm1nm ), (b)aged at 650 for 1h(2nm2nm1nm ) and
(c) aged at 650 for 100h(12nm12nm1nm )

Figure 11: The growth rate of Cu-rich phase at 650 long time aging till 10,000h in Super304H
Experimental results show that MX type NbC or Nb(C,N) nano-size can be precipitated in
TP347H, Super304H and HR3C steel because of Nb addition in these 3 steels. Except MX phase
the Cu-rich phase and NbCrN nano-size precipitates can be formed in Super304H and HR3C
respectively. Among these 3 steels Super304H characterizes with the highest stress rupture
strength because of the Cu-rich phase precipitation strengthening effect. In fact of these
experimental results a new idea is developed to design an advanced Ni-Cr austenitic heat-resistant
104

steel with multi-phase precipitation strengthening(MX+Cu-rich phase+NbCrN in -matrix and


M23C6 at grain boundaries) to reach high stress-rupture strength and high corrosion/oxidation
resistance performance both for superheater and reheater tubing application at higher
temperatures[5].
Ni-base superalloys
Nimonic 80A
A Ni-base superalloy Nimonic 80A mainly strengthened by Ti and Al to form '-Ni3(Al, Ti)
precipitation in Ni-Cr solid solution strengthened austenite matrix had been adopted as 600
USC steam turbine blades by Shanghai Turbine Company (STC) since 2006[6]. Up to now more
than forty 600 USC steam turbines have been safely and successfully put in service.
Typical microstructure of Nimonic 80A at standard heat treatment condition is shown in Fig.12.
The fractions of precipitates were quantitatively determined as shown in Table. 3. It can be seen
that the uniform grain structure (Fig.12a), and a certain amount of Cr-rich carbides M7C3 and
M23C6 (0.636%) mainly distributed at grain boundaries (Fig.12b) and high fraction (17.713%) '
strengthening phase dispersively distributed in grains (Fig.12c). A small amount of MC carbide
(0.036%) formed during solidification randomly distributed in the alloy.
(b)

(a)

(c)

Figure 12: Typical microstructure of Nimonic80A alloy at standard heat treatment condition (a)
Grain structure; (b) Carbides at grain boundaries; (c) precipitates in grains

105

Table 3: Factions of different precipitates in Nimonic 80A at standard heat treatment condition
Phase
matrix
phase
M7C3+M23C6
MC
Fraction (wt. %)
81.615
17.713
0.636
0.036
precipitation

The effect of long-time aging on phase is illustrated in Fig.13. It can be seen that
morphology has no apparent change below 700C, almost the same as that at standard heat
treatment condition. However, phase becomes unstable and its morphology changes after
750C long time aging (see Fig.13h and 13i). Fig.13 clearly shows the coarsening process
at high temperature long time aging. The radius of phase increases with aging
temperatures and times, especially more rapidly after 750C, 1,000h aging. Through these
observations, it can be concluded that aging temperature more intensively contributes to
coarsening than ageing time. Quantitative phase analyses (see Table. 4) show that the
fraction of phase for 10,000h at 700C (18.038%) is lower than those at 600 (19.382%)
and 650C (19.468%).

Figure 13: Effect of aging on behavior in Nimonic80A alloy (a) 550 C, 500h; (b) 550 C,
o
o
o
o
o
5000h; (c) 600 C, 1000h; (d) 600 C, 10,000h; (e) 650 C 1000h; (f) 650 C, 10,000h; (g) 700 C,
o
o
10,000h; (h) 750 C, 2000h; (i) 750 C, 5000h
o

Table 4: Factions of different precipitates in Nimonic80A aged for 10,000h at 600, 650 and
700C
10,000h aging
600C
650C
700C
phase (wt. %)
19.382
19.468
18.038
M7C3+M23C6 (wt. %)
0.641
0.676
0.755
106

growth

Fig.14 represents the growth of with aging times at 550~750C. It is clearly seen that
radius almost has no change at 550C long-time aging till 5,000h. And phase grows more
slowly at 600~650C than at 700C with aging times. These average radiuses at 600, 650
and 700C long-time ageing till 10,000h by quantitative phase analyses are 38.7, 37.2 and
42.1nm, respectively. From Fig.14(a), it can be seen that growth at 750C is the most rapid
among 550~750C and its average radius has reached about 87.5nm with aging times up to
5,000h, which would be harmful to the mechanical properties of the alloy for hightemperature long-term service.
3

The correlation of effective radius r of precipitates with aging times is plotted in Fig.14(b). It
can be seen that the results show linear relationships at 550~750C (Fig.14(b)). It is suggested
3

that the coarsening of precipitates obeys the standard r t kinetics of diffusion-controlled


particle growth. The slope of each fitted curve in Fig. 14(b) corresponds to a temperaturedependent coarsening rate constant, k. It is evident that k in Nimonic80A alloy increases much
rapidly with long-time ageing temperatures. It means that aging temperature is a more important
parameter for the coarsening of precipitates.

Figure 14: The relationships between size and aging time for Nimonic80A at different
temperatures (a) r -t; (b)

r -t

Grain boundary carbides


The morphology of grain boundary carbides at different aging conditions is shown in Fig.15. The
quantity of grain boundary carbides M7C3+M23C6 at 600, 650 and 700C for 10,000h is shown in
table. 4. No other precipitates after 10,000h aging have been found through detail phase analyses.
The amount of M7C3+M23C6 increases with aging temperatures and times in comparison with
table.3 and 4, and their quantities at 600, 650 and 700C for 10,000h are 0.641%, 0.676% and
0.755%, respectively. Grain boundary carbides produce a discontinuous distribution at early stage
of aging (see Fig.15a) and near continuous distribution at grain boundaries has formed with
prolonged aging times and temperatures (see Fig.15d~15g). At 750C, some grain boundary
carbides coarsen and aggregate (see Fig.15i), which may lead to the weakening of mechanical
properties.

107

Figure 15: Effect of aging on grain boundary carbides behavior in Nimonic 80A alloy (a) 550C,
500h; (b) 550C, 5000h; (c) 600C, 1000h; (d) 600C, 10,000h; (e) 650C, 1000h; (f) 650C,
10,000h; (g) 700C, 10,000h; (h) 750C, 2000h; (i) 750C, 5000h
Stress-rupture strength after long time aging.
Fig. 16 shows a very positive result that the stress-rupture test (750C, 310MPa) lives of Nimonic
80A alloy specimens after 10,000h long time aging at 600, 650 and 700C can still meet the
specification requirement ( 100h).

Figure 16: Stress rupture tests (750C, 310MPa) of Nimonic80A samples after 10,000h long time
aging at 600, 650 and 700C

108

Above mentioned results show that Nimonic 80A characterizes with good structure stability at
high temperatures[7]. This alloy can be not only used for 600C USC turbine blades, also can be
possibly used for 700C A-USC steam turbine blades.
Waspaloy
WASPALOY is a Ni-Cr-Co-Mo base superalloy with Ti and Al to form -Ni3(Al,Ti)
precipitation strengthening for creep strength at high temperatures. WASPALOY has been wildly
used as blades, disks, fasteners and other forgings for aero-engine application in the temperature
o
range 700-750 C even above. WASPALOY has been also adopted for gas expander blades and
o
disks application in petro-chemical industry for long time service at temperature 650-700 C ever
higher. Now WASPALOY is considered to be used as blade material for 700 A-USC steam
turbine application.
Typical microstructure and fraction of precipitated phases after standard heat treatment is shown
in Fig.16. It can be clearly seen that about 22% wt homogeneously distributed in Ni-Cr-Co-Mo
-matrix and about 0.2% M23C6 carbide discontinuously precipitated at grain boundaries. A small
amount (0.1%) of MC carbide formed during solidification and randomly distributes in the alloy.

Precipitate

MC
M23C6

Composition
(Ni0.883Fe0.03Cr0.048Co0.039)3.28(Al0.38Ti0.62Mo)
Ti(CN), (TiN)
(Cr0.746Mo0.094Ni0.084Co0.041Fe 0.023Ti0.012)23C6

Wt%
~22
Small amount
~0.5

Figure 17: Typical microstructure and phase fraction of WASPALOY at standard heat treatment
condition (1080/4hrs/AC+ 845/24hrs/AC+760/16hrs/AC.)
The microstructure of WASPALOY looks very similar to Nimonic80A (see Fig.17 and Fig.12).
However, WASPALOY contains more solid solution strengthening elements such as Co and Mo
and higher fraction (~22%) of than that of Nimonic80A (~17%). In results of that WASPALOY
characterizes with higher creep strength and to be used at higher temperature than Nimonic 80A.
The main strengthening effect of WASPALOY is contributed by precipitation in Ni-Cr-Co-Mo
-matrix. So the stability of phase is the key issue for guarantee high creep strength at high
temperatures. The structure stability study of WASPALOY has been conducted by long time
thermal aging in the temperature range from 550-850[8]. Fig.18 shows the morphology after
long time aging at 700,5000h and 750,4500h respectively. The long time aging at 700-750
is still continuing until 10,000h. The growth rate of in WASPALOY is shown in Fig.19. It can
be seen that strengthening phase is quite stable in the temperature range of 550-750 .
109

However its growth rates in the temperature range of 800-850 are quite high. It can be
concluded that from view-point of structure stability WASPALOY is suitable to be used as 700
blade application for A-USC steam turbine.

Figure 18: precipitation behavior in WASPALOY after long-time aging at 700 for 5000h(a)
and 750 for 4500h(b)

Figure 19: growth rates in WASPALOY at temperatures 550-850


The long time stress rupture tests are just going. Fig.20 shows the remaining lives (WASPALOY
disk after 60,000h service) calculated stress-rupture curves in the temperature range from 600850[9]. The estimated stress rupture strength of 105hrs should be higher than 100MPa at 700
to fulfill the blade material requirement for 700 A-USC steam turbine.

110

Figure 20: The calculated stress rupture curves for WASPALOY in the temperature range of 600850
Inconel740
Inconel740, a new Ni-Cr-Co-Mo-Nb-Ti-Al superalloy, is recently developed by Special Metal
Corp.(SMC, Huntington) USA for European THERMIE AD700 A-USC project with steam
parameters of 700 , 35MPa. At this condition the superheater and reheater fire-side metal
temperature can be 750-760 even higher. The superheater and reheater require stress rupture
strength(100MPa, 105h) at temperatures 750-760 , together with the high corrosion
resistance(2mm cross-section loss in 2105h). Inconel740 characterizes with the highest stress
rupture strength and corrosion/oxidation resistance among todays available candidate
materials(such as Haynes230, HR6W, Inconel617B and Nimonic263) and can meet above
mentioned strict requirements.
Typical microstructure and phase fraction of Inconel740 at standard heat treatment are shown in
Fig.21. The main strengthening phase (12.98%) homogeneously distributes in Ni-Cr-Co matrix, M23C6 carbide(0.115%) and a very small amount(0.054%) of high Si-containing G-phase
mainly precipitate at grain boundaries and MC-(Nb.Ti)C(0.183%) carbide formed at solidification
process randomly distributes in the alloy.

phase

MC
M23C6
G
Wt%
12.980
0.183
0.115
0.054
Figure 21: Typical microstructure and phase fractions of Inconel740 after standard heat
treatment
111

The microstructure of Inconel740 is quite stable at 700 long time aging as shown in Fig.22.
However, it is quite different at 760 long time aging. There are a lot of plate-like -phase
nearby grain boundaries and rapidly grow to the grains. Moreover the large globular high Sicontaining G phase distributed at grain boundaries as determined by EDAX(see Fig.23).
(b)

(a)

Figure 22: Stable micro-structure of Inconel740 at 704 long time aging: (a) 704 /1000h and
(b) 704 /2000h
Si (at%):
19.25%
A6B16Si7:
24.13%

Figure 23: The large amount of -phase and G-phase in Inconel740 after 760 long time aging
Detail quantitative determination of brittle high Si-containing G-phase after 704-760 long time
aging is shown in Fig.24. It is very harmful for Inconel740 that the amount of G phase(0.471%)
after 2000h aging at 760 is almost 10 times than the amount of G-phase(0.054%) as at standard
heat treatment condition. Moreover the unstability of strengthening phase and the formation of
-phase have been also detaily studied as shown in Fig.25.

112

standard heat treatment

aged at 760 for 2000h


704
760

G phase percent/%

0.5

0.4

0.3

0.2

0.1

0.0

500

1000

1500

2000

Aging time/h

Figure 24: High Si-containing G-phase formation and plate-like -phase precipitation in
Inconel740 at 760 long time aging

(a)

(b)

Figure 25: Coarsening of phase(a) and the precipitation of -phase(b) in Inconel740


It can be recognized that the structure stability is good for Inconel740 at long time aging at the
temperatures under 725. The unstability of Inconel740 at 760 has been confirmed by 3
important factors: high coarsening rate of phase in grains, plate-like -phase and high Sicontaining brittle G-phase formation at grain boundaries. It will clearly develop degradation of
strength and ductility both. Just for these reasons the structure stability improvement and
modification of Inconel740 for developing Inconel740H should be done to fulfill the
superheater/reheater requirements at the temperature 750 and above[10].
113

Inconel740H
Inconel740H is a modification of Inconel740 for improvement of structure stability by the
adjustment of Nb, Ti, Al and Si[11]. Chemical composition and phase fraction comparison between
Inconel740H and Inconel740 is shown in Table.5 and 6 respectively. Fig.26 shows the
comparison of calculated phase diagrams of Inconel740 and Inconel740H. It can be seen from
these results that the strengthening phase is more stable in Inconel740H than Inconel740. The
-phase is eliminated in Inconel740H and there is not existence of brittle high Si-containing G
phase in Inconel740H.
Table 5: Chemical composition (in wt%) comparison between Inconel740H and Inconel740
Alloy C
Cr
Ni
Co
Mo
Nb
Ti
Al
Mn
Fe
Si
740
0.03
25
bal
20
0.5
2
1.8
0.9
0.3
0.7
0.5
740H 0.03
25
bal
20
0.5
1.5
1.35
1.35
0.3
0.7
0.15
Table 6: Phase fraction(wt%) comparison between Inconel740H and 740
Alloy

MC
M23C6
740
12.980
0.183
0.115
740H
14.623
0.220
0.202

(a)

(b)

Figure 26: Phase diagram comparision between Inconel740(a) and Inconel740H(b)

500nm
Figure 27: Typical microstructure of Inconel740H after standard heat treatment

114

G
0.054
0

Typical microstructure of Inconel740H after standard heat treatment as shown in Fig.27 is very
similar to the structure of Inconel740 in comparison with Fig.21. The main strengthening phase
(14.623%) homogeneously distributes in Ni-Cr-Co -matrix, M23C6 carbide(0.202%) and also
MC-(Nb,Ti)C carbide(0.220%) formed at solidification process randomly distributes in this alloy.
However, the comparison of structure stability of Inconel740H and Inconel740 at 750 and
800 long time aging is quite different(see Fig.28 and Fig.29)[12]. Inconel740H keeps stable
strengthening in Ni-Cr-Co -matrix and M23C6 carbide at grain boundaries. However, plate-like phase formation and high Si-containing G-phase precipitation obviously happen in Inconel740.
Moreover the growth rate comparison between Inconel740H and Inconel740(see Fig.30) clearly
shows that Inconel740H characterizes very good structure stability at 750 long time aging.

Figure 28: Long time structure stability comparison between Inconel740H and 740 at 750
aging

115

Figure 29: Long time structure stabiltiy comparison between Inconel740H and 740 at 800
aging

(b)

(a)

Figure 30: growth rate comparison between Inconel740H(a) and 740(b) at high temperatures
The structure stability study has been also conducted after high temperature stress rupture tests.[12]
The structure of Inconel740 sample tested at 775, 170MPa, after 2779hrs shows a large
amount of -phase near grain boundaries and G phase formation at grain boundaries(see Fig.31a).
However, the structure of Inconel740H sample tested at 750, 280MPa, after 1087hrs still keeps
stable precipitation in -matrix and M23C6 carbide distributed at grain boundaries and there are
no existence of plate-like -phase and brittle G-phase at grain boundaries(see Fig.31b).

116

(b)

(a)

1m
Figure 31: Microstructure comparision of Inconel740(a) and Inconel740H(b) stress rupture
tested samples
The impact toughness of Inconel740H after high temperature long time aging is much higher than
Inconel740 as shown in Table.6.
Table 6: Impact toughness comparison of Inconel740H and Inconel740 after high temperature
long time aging
CVN J/cm2
Alloy
Exposure
740

740H

0h

45

750-2000h

13.6

0h

90.6

700-1050h

29.5

750-1000h

35.7

800-1000h

50.5

700-3042h

23.3

750-3000h

29.8

800-3000h

46.2

700-5000h

34.7

750-5000h

23.8

800-5000h

45.3

solution treatment at 1121

93.7

750-1000h(solution treatment at 1121)

51.2

750-3000h(solution treatment at 1121)

44.4

750-5000h(solution treatment at 1121)

45.2

117

CONCLUSIONS
1. Almost 80% of total electricity is generated by coal-fired power plants in China. It will be
still kept around 70% of total electricity till the year of 2020.
2. China started her 1st 600 1000MW USC power unit in November 2006. To the end of
2012 China already put more than hundred 600 USC units in service. These are about 80%
of 600 USC power units in the world.
3. The structure stability of advanced austenitic steels(TP347H, Super304H and HR3C) for
600 USC power units have been studied in detail and all these heat-resisting steels can be
made in China.
4. China has started 700 A-USC Project for coal-fired power plants in June, 2011. A series of
Ni-base superalloys (such as Nimonic80A, Waspaloy and Inconel740/740H) have been
studied for Chinese 700 A-USC Project.
5. Nimonic80A shows good structure stability at 600 and up to 700. This alloy has been
used as blade material for 600 USC steam turbines in routine production. From the
viewpoint of structure stability Nimonic80A is possible to be used as blade material for USC
steam turbines under 700 .
6. From creep rupture strength and structure stability view point Waspaloy should be a good
candidate blade material for 700 A-USC steam turbine application.
7. Inconel740H developed from Incoenl740 shows good structure stability especially no phase
nor G phase formation. Impact toughness of Inconel740H is obviously improved after longterm high temperature aging in comparison with Inconel740.
8. Inconel740H can meet the requirements of 105h creep strength 100MPa and
oxidation/corrosion resistance 2 105h 2mm for 700 A-USC technology. So,
Inconel740H is one of the excellent candidate materials of tube and pipe application for
700 A-USC power plant technology.
9. For world-wide promotion of 700 A-USC power plant project China is seeking the
international cooperation in this field.
ACKNOWLEDGMENTS
University authors appreciate China Natural Science Foundation(NO. 50931003), Central Iron &
Steel Research Institute, China. Special Metals Corporation(Huntington) WV, USA and CBMM
Brazil for partial funding on heat resisting steels and superalloys joint projects and appreciation
also to Shanghai Power Equipment Research Institute for providing long time aged heat resistant
steel samples.
REFERENCES
[1] Fusheng Lin, Shichang Cheng and Xishan Xie., Ultrasupercritical power plant development
and high temperature materials applications in China, Energy Materials, Vol. 3(2008), pp.
201-207.
[2] Hongyao Yu, Jianxin Dong and Xishan Xie., 650 Long-Term Structure Stability Study on
18Cr-9Ni-3CuNbN Heat-Resistant Steel, Materials Science Forum, Vol. 654-656(2010),
pp. 118-121.

118

[3] Chengyu Chi, Jianxin Dong, Wenqing Liu, et al., An Investigation on Precipitation Behavior
of Cu-Rich Phase in Super304H Heat-Resistant Steel by Three Dimensional Atom Probe,
Material Science Forum, Vol. 654-656(2010), pp. 110-113.
[4] Chengyu Chi, Hongyao Yu, Jianxin Dong, et al., The precipitation strengthening behavior of
Cu-rich phase in Nb contained advanced Fe-Cr-Ni type austenitic heat resistant steel for USC
power plant application, Progress in Natural Science: Materials International, Vol. 22, No.
3 (2012), pp: 175-185.
[5] Hongyao Yu., Ph. D Dissertation, University of Science & Technology Beijing, (2013).
[6] Hongwei Shen, Lihong Zhang, Xishan Xie, et al., The Application of Ni-Base Alloy
Nimonic80A for Buckets of USC Steam Turbine in China, Proc of Advances in Materials
Technology for Fossil Power Plants, ASM, 2008, pp. 402-411.
[7] Qiuying Yu, Zhi Shen, Maichang Zhang, et al., Long time Thermal Structure Stability
Study on NiCr20TiAl Alloy, Advanced Materials Research, Vol. 399-401(2012), pp: 71-75.
[8] Zhihao Yao, Jianxin Dong, Xu Chen, et al., Gamma Prime Phase Evolution during Long
Time Exposure for GH738 Superalloy, Transactions of Materials and Heat Treatment, Vol.
34(2013), pp: 3084-3098.
[9] Zhihao Yao, Maichang Zhang, Jianxin Dong, Stress Rupture Fraction Model and
Microstructure Evolution for WASPALOY, Metall. Trans. A, Vol. 44(2013), pp: 30843098.
[10] Xishan Xie, Shuangqun Zhao, Jianxin Dong, et al., An Investigation of Structure Stability
and Its Improvement on New Developed Ni-Cr-Co-Mo-Nb-Ti-Al Superalloy, Materials
Science Forum, Vol. 475(2004), pp: 613-516.
[11] Xishan Xie, Shuangqun Zhao, Jianxin Dong, et al., A New Improvement of Inconel
Alloy740 for USC Power Plants, Proc of Advances in Materials Technology for Fossil
Power Plants, ASM, 2008, pp. 220-224.
[12] Chengyu Chi, Shuangqun Zhao, Fusheng Lin and Xishan Xie., High Temperature Longterm Structure Stability Study on Inconel740/740H Ni-Base Superalloy, Keynote Lecture at
Proc of The 5th Symposium on Heat Resistant Steels and Alloys for High Efficiency USC/AUSC Power Plants 2013, May 26-29, Seoul, Korea,2013.

119

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

RECENT DEVELOPMENTS IN THE CHARACTERISTICS OF


HAYNES 282 ALLOY FOR USE IN A-USC APPLICATIONS
S.K. Srivastava, J.L. Caron, and L.M. Pike
Haynes International, Inc., Kokomo, IN 46904-9013

ABSTRACT (#106)
In an earlier paper, preliminary data for HAYNES 282 alloy was presented for potential advanced
steam power plant applications [1]. Since then, 282 alloy has continued to be evaluated for a
variety of A-USC applications: superheater boiler tubing, large header piping, rotors, casings, etc.
Per current practice the alloy achieves its strengthening by a two-step age hardening heat
treatment. Given the difficulty of such a procedure, particularly for larger components in the
power plant, interest has focused on the development of a single step age hardening treatment.
While considerable work on 282 alloy is still going on by a number of investigators, during the
preceding years a large amount of data was generated in characterizing the alloy at Haynes
International. This paper will briefly review the behavior of 282 alloy in air and water vapor
oxidation (10% H2O) at 760C (1400F), low cycle fatigue properties at 649C to 871C (1200F
to 1600F) and long-term thermal stability at 649C to 871C (1200F to 1600F). Special focus
of the paper will be mechanical behavior: tensile and creep; microstructural analysis, and
weldability of 282 alloy as a result of single step age hardening heat treatment: 800C
(1475F)/8hr/AC.
HAYNES, 230, and 282 are registered trademarks of Haynes International, Inc.

INTRODUCTION
It is recognized that for long-term service at high-temperature high-pressure conditions operative
in A-USC power plants, nickel based alloys would be required. In contrast to solid solution
strengthened alloys, -strengthened alloys offer considerable strength advantages to many
components at 760C (1400F) at 24 MPa (3.5 ksi). HAYNES 282 alloy is a -strengthened
nickel-based alloy combining superior creep strength, stability, and fabricability vis--vis other
age-hardened alloys.
Since being introduced in the market place in 2005 [2], the alloy has been a commercial success
and has been produced and sold in a variety of forms such as sheet, plate, bar, and forgings of
various configurations. The alloy has been specified for a number of applications for flying as
well as land-based gas turbines. A preliminary account of the alloy was presented during 5th Intl.
Conference on Advances in Materials Technology for Fossil Power Plants [1]. Since then, the
alloy has continued to be evaluated for a variety of A-USC applications: superheater boiler
tubing, large header piping, rotors, casings, etc. This paper will briefly update on environmental
characteristics, low cycle fatigue properties at 649C to 870C (1200F to 1600F), and long-term
thermal stability at 649C to 871C (1200 to 1600F). In subsequent sections, the paper will
120

focus on mechanical behavior and weldability of 282 alloy as a result of the one step age
hardening heat treatment (1475F/8hr/AC).
Physical Metallurgy
The nominal composition of 282 alloy along with alloys mentioned in this study are given In
Table 1.
Table 1: Nominal compositions of the alloys in this study (wt.%)
Alloy

Ni

Cr

Co

Mo

Ti

Al

Fe

Mn

282

57a

20

10

8.5

--

2.1

1.5

1.5*

0.3*

R-41

52

19

11

10

--

3.1

1.5

5*

0.1*

0.5*

0.09

0.006

--

Waspaloy

58a

19

13.5

4.3

--

1.5

2*

0.1*

0.15*

0.08

0.006

Zr-0.05

263

52a

20

20

--

2.4*

0.6*

0.7*

0.4

0.2

0.06

0.005

Al+Ti-2.6

22

5*

14

--

0.3

3*

0.5

0.4

0.10

0.015*

0.2La

230
a

57

As Balance

Si
0.15*

Other

0.06

0.005

--

*Maximum

The development of 282 alloy was described in Ref [3]. The main goal in its development was to
achieve exceptional creep strength concomitant with superior fabricability. These were achieved by
optimizing the composition of the alloy to form significantly less amount of than R-41 and
Waspaloy yet achieving comparable creep strength. The alloy achieves its strengthening by a twostep age hardening heat treatment to form phase, Ni3 (Al,Ti) with L12 structure. The phase in
282 alloy forms, presumably due to a lower - mismatch, in a spherical morphology as would be
seen later in the paper. The phase fraction of in 282 alloy is estimated to be 19% versus 27% in
R-41 alloy. A much lower phase fraction of along with the sluggish kinetics of its formation, is
the key to its superior fabricability especially for thicker parts.
Environmental Properties
Cyclic oxidation studies were carried out in flowing air [4] and in air + 10% H2O at 760C
(1400F) [5], cycled weekly for 1008 hours. The results are tabulated below.
Table 2: Oxidation Testing at 1400F, in both air [4] and water vapor [5] environments
Environment

Alloy

Metal Loss, m
(mils)

Air

Avg Met. Aff. m


(mils)

Max. Met. Aff. m


(mils)

282
2.5 (0.1)
10.2 (0.4)
263
2.5 (0.1)
12.7 (0.5)
Waspaloy
2.5 (0.1)
15.2 (0.6)
R-41
2.5 (0.1)
10.2 (0.4)
Air + 10% H2O
282
2 (0.08)
13 (0.5)
263
2 (0.08)
10 (0.4)
R-41
2 (0.08)
15 (0.6)
230
1 (0.04)
5 (0.2)
Avg. (Max) Metal Affected = Metal Loss + Avg. (Max) Internal Attack; 1 mil = 0.001 inch = 25.4 m

12.7 (0.5)
15.2 (0.6)
17.8 (0.7)
12.7 (0.5)
18 (0.7)
13 (0.5)
18 (0.7)
8 (0.3)

Details of the experimental procedures can be found in Ref [5]. Many combustion environments
are likely to contain significant amounts of water vapor. It would suffice to state that at 760C
(1400F) 282 alloy suffered little to no corrosion attack in air + 10% water vapor. Comparative
121

static oxidation data for 282 alloy at 871C, 927C, and 982C (1600F, 1700, and 1800F), as
also the dynamic oxidation data at 871C (1600F) were reported in Ref [6]. In static oxidation
test at 871C (1600F), 282 alloy suffered 41 m (1.6 mils) of corrosion attack similar to the
other alloys in the test. Dynamic oxidation testing is much more aggressive wherein samples
are subjected to severe thermal cycling every half hour. At 871C (1600F), 282 alloy suffered
corrosion attack of 117 m (4.6 mils) which was similar to other alloys but more than the solid
solution strengthened 230 alloy which suffered 63 m (2.5 mils) of corrosion attack.
Low Cycle Fatigue
Low cycle fatigue behavior of 282 alloy was reported by Pike [7]. Fully reversed (R = -1.0) axial
low cycle fatigue testing was carried out at 649, 760, 816, and 871C (1200, 1400, 1500, and
1600F) on 282 alloy, and at 1500F (816C) for the Waspaloy, R-41, and 263 alloys. The LCF
resistance was found to gradually decrease with temperature up to 816C (1500F). Above this
temperature, the LCF resistance decreased more rapidly. It was found that 263 alloy had the
lowest LCF resistance at all strain ranges of the four alloys in the study. The 282 alloy, Waspaloy
alloy, and R-41 alloy were found to have very similar LCF resistance for higher total strain range
conditions. The stress amplitude was virtually constant in the elastic regime, while cyclic
softening was observed in tests at a higher total strain range. The paper did not include
examination of hold-time effects where the interaction of creep and fatigue can be determined.
Also, most of tests lasted less than 100 hours as such negative effects of environmental exposure,
as is likely in service, were not a factor.
Thermal Stability
Thermal stability of 282 alloy has been studied in exposures at 649C, 760C, and 871C (1200,
1400 and 1600F) up to 16000 hours [8]. Thermal stability is a key feature of alloys intended for
long-term service at elevated temperatures. For gamma-prime strengthened alloys in particular,
there are two main issues which can arise due to microstructural instability: loss of ductility and
loss of strength. Formation of deleterious phases such as the TCP-type sigma, Laves, or mu
phases can lead to a loss of ductility and embrittlement. The embrittlement is more often seen at
room temperature (RT) than at elevated temperatures (ET). Loss of strength can arise in these
types of alloys due to a number of reasons, including gamma-prime coarsening and/or dissolution
as well as formation of TCP-type phases which can deprive the matrix of critical elements such as
Mo. The 282 alloy did not suffer embrittlement and it did not form deleterious phases within the
1,000 hour exposures at any temperature. Optical micrographs of 282 alloy exposed at various
temperatures for 1,000 hours are shown in Fig 1. The phases present comprise primary TiN and
MC, and secondary M23C6 and M6C in addition to the matrix , being non-observable. The
alloy was found to increase in strength after the 1,000 hour exposure at 760C (1400F). Based on
RT and ET tensile tests of materials exposed for up to 16,000 hours at temperatures between 649
and 871C (1200 and 1600F), the results were very favorable for use in that temperature range in
terms of both elongation and yield strength.

122

a)

b)

c)

d)

Figure 1: Optical micrographs of thermally exposed 282 alloy (1000-hour exposure).


a) 1200F (649C), b) 1400F (760C), c) 1500F (816C), d) 1600F (871C). [8]
HEAT TREATMENT STUDIES
The 282 alloy was originally developed for gas turbine applications. On account of its strength,
thermal stability and fabricability, the alloy has also been found eminently suitable for
applications in A-USC power plants. However, it was found that the two-step heat treatment
would pose difficulty especially for larger components in the power plant. Therefore, considerable
efforts were directed toward characterizing a one-step aging heat treatment. In this section
mechanical behavior of 282 alloy as a result of the one-step age hardening heat treatment in
contrast to that of two-step age hardening heat treatment is examined.
Tensile Properties
The alloy is mill annealed in the range of 1107 to 1149C (2025 to 2100F) to obtain a typical
grain size of ASTM 4-4.5 for optimum creep strength and fatigue life. Per the standard practice, a
two-step heat treatment consisting of 1010C (1850F)/2hr/AC + 788C (1450F)/8hr/AC is
imparted to form the (L12) strengthening phase. After an initial screening, a one-step aging heat
treatment of 800C (1475F) was selected and a set of mechanical property data was generated.
In Tables 3A and 3B comparative tensile data for the two different aging heat treatments are
given.

123

Table 3A: Effect of Different Age-Hardening Treatments: 0.062 Sheet


Age-Hardening Treatment

1010C (1850F)/2h/AC
+788C (1450F)/8h/AC

800C (1475F)/8h/AC

Temp
C (F)

0.2% YS
MPa (ksi)

UTS
MPa (ksi)

Elong.
%

RT

676.6 (98.1)

1115.2 (161.7)

27.3

649 (1200)

608.3 (88.2)

1006.9 (146.0)

31.9

760 (1400)

597.9 (86.7)

838.6 (121.6)

20.4

871 (1600)

496.6 (72.0)

554.0 (80.4)

24.2

RT

705.5 (102.3)

1150.3 (166.8)

30.8

649 (1200)

608.3 (88.2)

1005.5 (145.8)

26.9

760 (1400)

604.8 (87.7)

832.4 (120.7)

12.9

871 (1600)

500.0 (72.5)

559.3 (81.1)

20.0

Table 3B: Effect of Different Age-Hardening Treatments: 0.500 Plate


Age-Hardening Treatment

Temp
C (F)

0.2% YS
MPa (ksi)

UTS
MPa (ksi)

Elong
%

R.A.
%

None

RT

394 (57.1)

831 (120.5)

57

N/A

1010C (1850F)/2h/AC
+788C (1450F)/8h/AC

RT

706 (102.3)

1152 (167.1)

29.9

33.1

760 (1400)

632 (91.7)

866 (125.6)

21.1

22.6

RT

741 (107.4)

1183 (171.5)

31.3

35.4

760 (1400)

658 (95.4)

857 (124.2)

18.3

18.3

800C (1475F)/8h/AC

The strength properties as a result of the both aging heat treatments were similar. The most
significant difference was a discernible drop in ductility at 760C (1400F), more so in the sheet
product than in the plate. These observations can be interpreted with the help of microstructural
observations as shown in Figures 2 and 3.
The images on the left correspond to the two step heat treatment: 1010C (1850F)/2hr/AC +
788C (1450F)/8hr/AC; on the right correspond to the single step heat treatment: 800C
(1475F)/8hr/AC. The two step heat treatment leads to a discontinuous blocky morphology of the
carbides precipitating (presumably M23C6 type) in the grain boundaries. At the higher
magnification, the carbide particles show a stone wall construction. In contrast, the carbides
forming at 800C (1475F) during the one step aging heat treatment show a nearly continuous
network and ledges protruding into the matrix. The apparent difference in the width of grain
boundaries for the two conditions is attributed to the predominance of the growth morphology at
1010C/2hr/AC.

124

Figure 2. SEM micrographs of 282 alloy grain boundaries; Left Two step heat treatment; Right
One step heat treatment. Bottom images are at higher magnification.

The strength properties are interpreted with the help of SEM images shown in Fig. 3. In the case
of the two step age hardening heat treatments, particle size was estimated to be in the range of
21-33 nm; for the single step in the range of 38-71 nm. Though in both instances, the spherical
shape of particle indicative of a low mismatch strain is observed.
The equilibrium mole fractions after either heat treatment would not be expected to be
significantly different. However, the difference in the particle sizes as a result of the two heat
treatments does not lead to a substantive difference in strength properties.

125

Figure 3. Gamma prime particle distribution in 282 alloy; Top two step heat treatment;
Bottom single step heat treatment.
Creep Strength
Effect of age-hardening treatments on creep-rupture properties for sheet is shown in Table 4.
Table 4: Effect of Different Age-Hardening Treatments: Creep-Rupture at 927C (1700F)/
48.3MPa (7 ksi); Sheet, 0.062
1% Creep,
Rupture Life,
Age-Hardening Treatment
hours
hours
1850F/2h/AC +1450F/8h/AC
318.5
773.0
1475F/8h/AC
237.5
465.2

126

Clearly there is some degradation in 1% creep and rupture lives as a result of the single step heat
treatment. Comparative creep rupture strength data derived over the range of 649C (1200F) to
816 (1500F) from a 0.500 plate showing the effect of different age-hardening treatments is
plotted in Fig 4. The figure clearly shows that the one-step heat treatment resulted in a consistent,
but a very small decrease in creep-rupture strength relative to the two-step treatment. The loss is
attributed to a slight coarsening of the .

100

Larson-Miller Parameter (K x 10-3, C = 20)


22
24

20

26

90

600

80

500

70
60

400

50

Stress (ksi)

30

200

Stress (MPa)

300

40

20
282 alloy - 1850F(1010C)/2h/AC + 1450F(788C)/8h/AC
282 alloy - 1475F(800C)/8h/AC
263 alloy - 1475F(800C)/8h/AC

100
90
80

10

36

44
42
38
40
Larson-Miller Parameter (R x 10-3, C = 20)

70

46

Figure 4: Creep-Rupture strength over 650 to 816C.

In addition to the creep-rupture strength data for 282 alloy, data for 263 alloy is also plotted in the
figure. Both aging heat treatments resulted in considerably better creep-rupture strength than 263
alloy. Note that 263 alloy also is strengthened by a single step aging heat treatment at 800C
(1475F)/8h/AC.
FABRICATION
As stated in the Introduction, in addition to its creep strength, the defining characteristic of 282
alloy is its fabricability, which is derived not only from an optimum volume fraction of the
phase, but also equally from the sluggish kinetics of its formation. In R-41 alloy, the formation of
cannot be suppressed by rapid cooling from the mill annealing temperature. In contrast, 282
alloy requires almost two hours of isothermal exposure at ~ 788C (1450F) which is at the nose
of TT Hardening plot. This feature has important implications in processing and fabrication, as it
mitigates the strain-age cracking phenomenon frequently associated with strengthened alloys.
The hardening kinetics of the 282 alloy are similar to those of the 263 alloy. The latter is known

127

for its superior weldability and formability which accounts for its widespread popularity and
applications in the aerospace industry.
During controlled heating rate tensile (CHRT) testing, anomalously large elongations were
exhibited by 282 alloy compared to R-41 and Waspaloy alloys, which indicates better strain-age
cracking resistance than those alloys [9]. This finding is attributed to its relatively sluggish precipitation kinetics. Based on Varestraint test results, 282 alloy also exhibits excellent weld
solidification cracking resistance, with crack lengths comparable to 263 alloy [10]. All
experimental testing and fabrication experience gathered to date indicate that 282 alloy has
excellent overall weldability when welded in the mill-annealed condition. Enough is not yet
known to determine the overall weldability of 282 alloy in the single-step aged condition.
Effect of Heat Treatment on Welds
In order to further investigate the weldability of 282 alloy, we have studied the effect of different
aging heat treatments on all-weld-metal (AWM) samples. The following three tables show tensile
properties of AWM plate samples subjected to various postweld heat treatments. Comparing the
results of the single-step (Table 5) to the two-step age-hardened (Table 6) AWM samples, it can
be seen that the single-step condition provided higher strength and generally lower ductility at
elevated temperatures up to 760C (1400F). The strength of the single-step condition fell below
that of the two-step condition at 871C (1600F) while its ductility continued to decrease and
remained lower. When compared to samples given a solution anneal treatment prior to the 788C
(1450F) aging (Table 7), the same trends in strength and ductility were observed. Samples given
a solution anneal treatment exhibited higher and more consistent ductility compared to samples
directly aged at 788C (1450F). Compared to the tensile properties of similarly aged base metal,
AWM samples exhibited higher YS, lower UTS, and lower ductility at RT and 760C (1400F).
While the precise microstructural reasons for these differences are not yet known, it is common
for weld metal to exhibit lower ductility than carefully controlled base metal microstructures.
The as-solidified weld microstructure of 282 alloy consists of discrete (Ti,Mo)C MC-type carbide
eutectic formed along the interdendritic boundaries. In addition, Cr-rich M23C6-type carbides
have also been identified in the as-welded condition. The formation of / eutectic in the
interdendritic regions, which is associated with solidification cracking in -strengthened Ni-base
alloys, has not been observed in 282 weld metal. Previous experiments have shown that 282 weld
metal achieves only partial recrystallization during the 1135C (2075F)/30min/AC solution
anneal as some initial solidification grain boundaries are left intact. The higher ductility of
solution annealed samples could be attributed to grain refinement via recrystallization and
dissolution of secondary phases in the weld microstructure. According to thermodynamic
calculations, MC carbides are not stable at 788C (1450F) and there is likely a propensity for
them to transform to M23C6 and upon aging. At 1010C (1850F), both MC and M23C6 are
stable whereas the dissolution temperature is exceeded. While microstructural analyses of
these particular samples have not been conducted, characterization of similar samples revealed
that Cr-rich M23C6 carbides are present along the initial weld solidification grain boundaries upon
postweld aging at 1850F. However, there appeared to be significantly more precipitation of
blocky M23C6 along the recrystallized grain boundaries. Continuous carbide precipitation as
a result of aging at 788C (1450F), similar to that observed in the base metal, may account for
the decreased ductility of single-step aged samples. Microstructural analyses are being conducted
to investigate this possibility. Through both qualitative SEM observations and semi-quantitative
XRD measurements, comparatively fewer carbides were observed in samples aged at 788C
(1450F) only.
128

Table 5: Tensile Properties AWM 0.375 Plate: GMAW + 788C (1450F)/8h/AC


Test Temp.
C (F)
RT
538 (1000)
649 (1200)
760 (1400)
871 (1600)

0.2% YS
MPa (ksi)
857 (124.2)
739 (107.2)
730 (105.9)
714 (103.5)
449 (65.1)

UTS
MPa (ksi)
1109 (160.9)
917 (133.0)
905 (131.3)
847 (122.9)
543 (78.8)

Elong.
%
21.0
20.5
18.5
9.0
5.5

R.A.
%
21.0
24.2
27.6
12.0
7.4

Table 6: Tensile Properties AWM 0.375 Plate: GMAW + 1010C (1850F)/2h/AC + 788C
(1450F)/8h/AC
Test Temp.
C (F)
RT
538 (1000)
649 (1200)
760 (1400)
871 (1600)

0.2% YS
MPa (ksi)
759 (110.1)
646 (93.6)
632 (91.6)
640 (92.8)
488 (70.7)

UTS
MPa (ksi)
1077 (156.1)
887 (128.6)
897 (130.0)
841 (122.0)
563 (81.6)

Elong.
%
19.6
22.5
20.8
12.8
14.8

R.A.
%
17.8
25.8
23.2
19.0
25.7

Table 7: Tensile Properties AWM 0.375 Plate: GMAW + 1135C (2075F)/30Min./AC +


788C (1450F)/8h/AC
Test Temp.
C (F)
RT
538 (1000)
649 (1200)
760 (1400)
871 (1600)

0.2% YS
MPa (ksi)
730 (105.9)
603 (87.5)
616 (89.3)
621 (90.1)
512 (74.2)

UTS
MPa (ksi)
1079 (156.4)
832 (120.6)
866 (125.6)
836 (121.2)
575 (83.4)

Elong.
%
26.4
25.7
23.7
24.5
25.8

R.A.
%
23.5
30.6
27.3
21.8
29.2

Additionally, previous transverse weld tensile tests indicate that the inclusion of the 1010C
(1850F) aging step does indeed provide for enhanced elevated temperature ductility, which may
be the result of carbide precipitation in the preferred morphology [10].
SUMMARY
1. Since the introduction of 282 alloy a large amount of data has been generated to characterize
the alloy, which includes physical metallurgy, environmental resistance, thermal stability,
weldability, etc.
2. As a result of the single step hardening heat treatment, the material showed similar strength
properties and some loss of ductility at 760C (1400F), especially in sheet product. The one-step
heat treatment resulted in a consistent, but a very small decrease in creep-rupture strength relative
to the two-step treatment. The loss is attributed, at least in part, to a slight coarsening of the .
3. Microstructural analysis showed that the single step treatment led to a continuous network of
carbide morphology as opposed to a discontinuous blocky morphology, the former resulting in
lower ductility at elevated temperatures.
4. While enough is not yet known to determine the overall weldability of 282 alloy in the singlestep aged condition, all experimental testing and fabrication experience gathered to date indicate
129

that 282 alloy has excellent overall weldability. The inclusion of either a post-weld solution
anneal at 2075F or aging step at 1850F provided for enhanced elevated temperature ductility of
AWM samples compared to those in the 1450F single-step aged condition.
5. While it may be necessary to generate even more data for specific applications, it is evident that
the available data reinforces the potential of 282 alloy for A-USC applications.
ACKNOWLEDGMENTS
Authors would like to thank John Cotner, Mark Richeson and John Ryan for help with
metallography, the Weld lab for all-weld-metal samples, and the Test lab for tensile and creep
testing.
REFERENCES
[1] D.L. Klarstrom and L.M. Pike, Materials Solutions for Advanced Steam Power Plants,
Advances in Materials Technology for Fossil Power Plants, Proc. Fifth Int. Conf., Oct. 2007, pp.
107-118, EPRI 2008.
[2] L.M. Pike, U.S. Patent 8,066,938.
[3] L.M. Pike, Development of a Fabricable Gamma-Prime (') Strengthened Superalloy, (Paper
presented at Superalloys 2008, Champion, PA, September 14-18, 2008) 191-200.
[4] V.P. Deodeshmukh, Unpublished Research, Haynes International, Inc., 2013.
[5] V.P. Deodeshmukh and N.S. Meck, Paper No. 11195, NACE Int., Corrosion 2011.
[6] L.M. Pike and S.K. Srivastava, Materials Science Forum, Vol. 595-598 (2008) pp. 661-671.
[7] L.M. Pike, Paper No. GT 2007-28267, ASME Turbo Expo 2007: Power for Land, Sea and
Air, May 14-17, 2007, Montreal, Canada.
[8] L.M. Pike, Long-Term Thermal Exposure of HAYNES 282 Alloy, (Paper presented at
Superalloy 718 and Derivatives, Pittsburgh, PA, October 10-13, 2010) 644-660.
[9] Metzler, D.A., Welding Journal, vol. 91, no. 6, June 2012, pp. 163-s-168-s.
[10] Caron, J.L., unpublished research, Haynes International, Inc. 2013.

130

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

CREEP-RUPTURE BEHAVIOR OF PRECIPITATION-STRENGTHENED


NI-BASED ALLOYS UNDER ADVANCED
ULTRASUPERCRITICAL STEAM CONDITIONS
P.F. Tortorelli, K.A. Unocic, H. Wang, M.L. Santella,*
Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831 USAt
J.P Shingledecker
Electric Power Research Institute, Charlotte, North Carolina 28262 USA
(*Retired)

ABSTRACT (107)
To achieve the necessary creep-rupture lifetimes at the temperatures and pressures associated with
advanced ultrasupercritical (A-USC) steam conditions (100,000 h at 100 MPa and 760C),
precipitation-strengthened nickel-based alloys are required for the superheater and reheater tubing
in A-USC boilers. Two alloys were considered to have potential for this application:
Inconel 740 and Haynes 282 alloy. In support of this application, creep-rupture testing of
several heats of Inconel 740 was conducted over a range of temperatures and stresses to develop
confidence in qualitatively predicting creep lifetimes under conditions relevant to A-USC steam
conditions, with the longest rupture times exceeding 30,000 h. For comparison, the creep-rupture
behavior of Haynes 282 alloy was mapped as a function of temperature and stress, but with a
significantly smaller dataset. Only a small difference in creep-rupture results between Inconel
740 and Inconel 740H was found although the latter alloy showed significantly greater resistance
to phase formation during testing. Little effect of prior aging treatments (for setting the
precipitate structure) on creep-rupture behavior was observed. Results from a modified power law
analysis showed that, while both Inconel 740 and Haynes 282 are projected to meet the A-USC
lifetime requirements, the latter offered the potential for better long-term creep resistance.
INTRODUCTION
There are ongoing efforts worldwide to develop higher-efficiency coal- and nuclear-based
electrical power generation systems utilizing advanced (ultrasupercritical) steam.[1-5] For coalfired plants, the increased energy conversion efficiency from the use of ultrasupercritical (USC)
steam (gains of up to 35%) also, of course, translates to a reduction in CO 2 emissions for a given
plant power output. To achieve such efficiency gains using advanced USC steam, boiler metal
temperatures in excess of 700C and pressures greater than ~30 MPa will be necessary for the
superheater/reheater components. To withstand these steam conditions, non-traditional boiler
alloys will be required in the hottest parts of the boiler in order to achieve adequate strength and
corrosion resistance at these temperatures and stress levels.[6] These higher strength alloys must
also be formable and weldable during fabrication as well as for repair. Any alloy for this
application will have to meet the requirements of the ASME Boiler and Pressure Vessel Code for
creep strength and show acceptable fireside corrosion and steamside oxidation resistance.

131

Table 1. Nominal compositions for several Ni-based alloys under consideration for use in
parts of USC coal-fired boilers (wt%, balance Ni)
Designation

Cr

Co

Mo

Al

Ti

Fe

Mn

Si

Other

Alloy 617
Haynes 282
Inconel 740
Nimonic 105
* maximum

22
20
25
15

12
10
20
20

9
8
0.5
5

1.2
1.5
0.9
4.7

0.3
2.1
1.8
1.2

3.0*
1.5*
0.7
1.0*

1.0*
0.3*
0.3
1.0*

1.0*
0.15*
0.5
1.0*

0.1
0.06
0.03
0.17*

0.006 B*
0.005 B
2.0 Nb
0.007 B

Worldwide, several materials development/qualification programs have identified nickel-based


alloys as being necessary for meeting the requirements for advanced steam (ultrasupercritical)
operations of coal-fired boilers.[1-4,7] Depending on the target temperatures and pressures, these
Ni-based alloys include those that use solid-solution strengthening (for example, the original
Inconel alloy 617, cf. Table 1) or when temperatures exceed ~700C, compositions that include
sufficient concentrations of Al and Ti to form (Ni 3 Al or Ti 3 Al) precipitates upon appropriate
aging. Examples of -strengthened alloys include Inconel alloy 740, Nimonic alloy 105, and
Haynes 282 alloy (Table 1).
This paper presents creep and microstructural results on two specific -strengthened Ni-based
alloys being investigated as part of the U.S. Department of Energy/Ohio Economic Development
Corporation project on materials for advanced USC steam systems.[1,6,8] This program has
targeted 760C (1400F), 35 MPa (5000 psi) for the final steam conditions to gain high efficiency
steam generation and energy conversion. This long running, comprehensive effort involving
boilermakers, turbine manufacturers, other engineering firms, federal laboratories, and the
Electric Power Research Institute, has produced creep data for an ASME pressure vessel and
piping code case for Inconel 740 [9], creep-rupture results for other high-temperature alloys, and
an extensive database on fireside corrosion and steam oxidation. In this paper, analysis of Inconel
740 creep data, including that used for an ASME Boiler and Pressure Vessel Code case as well as
more recent results, are presented and accompanied by observations of microstructural
development after exposures to elevated temperatures in the presence and absence of stress.
These Inconel 740 results then serve as a benchmark against which a smaller set of creep-rupture
and microstructural data for Haynes 282 is described. It is concluded that, based on creep
resistance, Inconel 740 can meet a lifetime goal of 100,000 h at 760C and 100 MPa, with the
initial Haynes 282 data indicating better creep resistance under such conditions.
EXPERIMENTAL APPROACH
Solution annealed or aged smooth-bar specimens having a 6.4-mm diameter with a 31.8-mm gage
length were used for the creep-rupture testing, which was conducted in accordance with
procedures described in ASTM E139. More details of the specific approach have recently been
described by Shingledecker and Pharr.[10] All results contained herein were for specimens tested
to rupture. Although, for most tests, creep deformation was measured as a function of time using
extensometers, only times to rupture as a function of stress and temperatures are reported in this
paper. After testing, a subset of the creep specimens were selected for microstructural analysis,
which included imaging of polished and etched cross sections from the gage and grip parts of the
samples by optical and scanning electron microscopy (OM or SEM, respectively). In cases
requiring higher spatial resolution images, transmission electron microscopy (TEM) or scanning
132

Table 2. Compositions of the various heats of Inconel 740 and Haynes 282 used in this study
(wt%, balance Ni)
Alloy*
Cr
Co
740 (t)
24.4 19.9
740 (p)
24.4 20.0
740 (b)
24.4 19.9
740 (b)
24.9 19.9
740H (p)
24.2 20.3
282 (p)
19.7 10.1
282 (p)
19.6 10.6
* b: bar; p: plate; t: tube
nr: not reported

Mo
0.54
0.50
0.50
0.53
0.49
8.36
8.60

Al
0.94
0.98
0.98
1.20
1.20
1.46
1.48

Ti
1.80
1.78
1.78
1.41
1.51
2.08
2.21

Nb
1.98
1.98
2.00
2.05
1.58
nr
<0.1

Mn
0.28
0.26
0.26
0.30
0.27
0.04
0.03

Fe
0.42
0.46
0.46
0.69
1.09
0.20
0.15

Si
0.54
0.55
0.51
0.48
0.25
<0.05
<0.05

C
0.03
0.03
0.03
0.06
0.03
0.06
0.06

B
0.002
0.004
0.004
0.001
<0.001
0.003
0.005

transmission electron microscopy (STEM) was used to examine thinned sections of specimens
prepared by electropolishing.
The creep-rupture specimens were fabricated from plates or bars of either Inconel alloy 740
(produced by Special Metals Corporation) or Haynes 282 alloy (from Haynes International,
Inc.). Several different heats of each alloy were used (Table 2). With the exception of the H
refinement of the Inconel 740 compositional specification, the results reported in this paper are
not analyzed in terms of the concentration variations amongst the heats of the same alloy. With
respect to the Inconel 740 alloy type, 740H represents a compositional modification to improve
fabricability as described by Baker et al.[11] This was accomplished by refinement of the Ti, Al,
Nb, and Si concentrations. (Table 2).
RESULTS
Creep
Creep-rupture results for the heats of Inconel 740 (Table 2) are summarized in Fig. 1, which
includes, as open symbols, data from specimens that have not yet ruptured at the time of
reporting. The data in Fig. 1 are from specimens that were thermally aged (8 or 16 h at 800C) to
form the precipitate structure (see below) prior to creep testing. The curves for each
temperature are approximate fits to the data, yet it is apparent from simple extrapolation that
lifetimes of the order of 105 h at 100 MPa are achievable for temperatures of 750C and below.
This, of course, assumes no change in failure mechanisms at times beyond the range of those
reached in the current experimental program (~45,000 h at 750C).
In addition to the Inconel 740 data shown in Fig. 1, creep-rupture data for one heat of Inconel
740H were also generated. The data for aged (8h, 800C) Inconel 740H specimens (half-filled
circles) are compared to those for Inconel 740 (filled circles) in Fig. 2, which presents the creeprupture results as applied stress versus a Larsen-Miller parameter (LMP), T{log(t)+C} where T is
the test temperature in K, t is the time at load, and C=19.392.[9] The solid curve represents the
fitted stress vs. LMP curve for Inconel 740. It is clear from Fig. 2 that there is no significant
difference in creep-rupture lives between Inconel 740 and its H refinement at the highest test
stresses. While the H rupture times at intermediate stresses may be lower than the corresponding
Inconel 740, the situation at the lowest stresses may be the reverse in view of the results from the
ongoing Inconel 740H tests (open circles).
133

Figure 1. Applied stress versus rupture time for aged Inconel 740 specimens. Much of
data was used in ASME code case 2702[9]. Open symbols and arrows indicate ongoing
tests.
Some specimens of the Inconel
740H heat were tested in the
solution
annealed
(SA)
condition. The creep-rupture
data for these are shown in Fig. 3
as the square symbols. (For ease
of comparison, the individual
data points for Inconel 740
shown in Fig. 2 have been
removed, but the master curve
remains.) Except for the tests at
the highest stress levels, there
appeared to be little difference in
creep-rupture lives between the
SA specimens and those that
were SA then aged. The shorter
rupture lives of the SA material
at high stresses could reflect the
Figure. 2. Applied stress versus Larson-Miller parameter
fact that the yield stress was
(LMP) for Inconel 740 and 740H specimens (including,
exceeded during the creep
but not limited to, data used in code case 2702). Open
loading; the high-temperature
symbols represent ongoing creep tests
yield strength of Inconel 740H in
this condition was about 60% of
what it was in the aged condition.[12]
A smaller number of creep-rupture tests of Haynes 282 have been completed to date. These
results are shown in Fig. 4, where the solid circles represent data from specimens aged before the
start of the creep exposures. A two-step aging treatment of 1010C for 2 h followed by 8 h at
134

788C is recommend by the alloy


producer[13] and most of the data
shown in the Fig. 4 data were for
specimens heat treated in this
way. However, two of the
specimens with a single aging
time and temperature (788C/8 h
or 800C/4 h) showed rupture
times that followed the general
trend of Fig. 4. The arrows in
Fig. 4 indicate these data points
with the one at the higher test
stress being the 4-h treatment.
Separately, a study of single-step
aging times at 800C for Haynes
282 indicated that 4 h was
Figure. 3. Applied stress versus LMP for aged and
sufficient
to
set
the

solution annealed (SA) Inconel 740H specimens. Open


structure[14] and a series of
symbols represent ongoing creep tests. Line is fitted
creep-rupture tests over a range of
curve for aged Inconel 740, same as in Fig. 2.
stresses and temperatures is
underway to determine if such
aging will produce creep-rupture lifetimes that match those of Haynes 282 with the traditional
two-step aging treatment. However, the results for solution annealed Haynes 282, (the squares in
Fig. 4) show no significant difference in creep-rupture behavior from the aged specimens over the
range of temperatures (750, 800C) and lifetimes (~300 to ~13,500 h) included in this data set.
Such observations for both Haynes 282 and Inconel 740 (see above) suggest that for application
conditions (such as boilers) where long lifetimes are desired and temperatures are relatively high,

Figure. 4. Applied stress versus LMP for aged and solution annealed (SA) Inconel 740H
specimens. Arrows indicate specimens with a single step aging treatment.
self aging to precipitate might be occurring.
135

Microstructure
Initial microstructural analysis focused on a subset of Inconel 740 and 740H specimens exposed
at 750C at the same three stress levels. Both OM and SEM were used to generally assess
microstructural evolution as a function of time and stress. Thermal aging processes were
distinguished from stress-induced effects by comparing the grip and gage sections from the same
rupture specimen.
Figure 5 shows SEM
micrographs of the grip and
gage sections of crept
Inconel 740 specimens that
were held at 750C at 220,
265, and 370 MPa, with
rupture times as shown. At
this magnification, the most
dramatic
change
in
microstructure as a function
of time at temperature was
the increase in the amount
of eta () phase (Nb 3 Ti)
compare Fig. 5a, c, and e.
The phase is often
recognized by its acicular
shape at certain grain
boundaries.[4,15]
This
morphology can be clearly
seen in Fig. 6, which is a
higher magnification view
of one such grain boundary.
Fig. 5. Secondary electron images of Inconel 740 after
Figure 5 also shows that the
exposures at 750C for indicated times and applied stress
presence of stress promotes
levels.
development. The images
on the right hand side of the figure represent the gage (stressed) section of the same specimen and
there was always a greater amount of phase observed at this location than in the grip, where the
applied stress was essentially nil. Grain boundary carbides are also present but are distinctly
different morphologically (cf. Fig. 6) and were not observed to undergo much change in their
characteristics with time or stress. Figure 7 shows examples of the microstructures of Inconel
740H after testing at 750C. It reveals that, in contrast to Inconel 740, much less phase formed
during thermal aging at the same stresses and temperature (compare Fig. 7a, c, e). Indeed, it was
difficult to observe in these specimens; even at the longest exposure time shown (4864 h,
Fig. 7e), a higher magnification was needed to clearly identify the morphological evidence of this
phase. As with the Inconel 740, development was greater in the presence of stress, with this
being most clearly observed for Inconel 740H at 370 MPa (compare Fig. 7a to b).
The creep resistance of these alloys relates directly to the periodic distribution and shape of the
precipitates (see Fig. 6). The TEM micrographs in Fig. 8 are used to compare the size and shape
of the after short and extended times at 750C in the absence of stress (that is, these images
were taken from grip sections). This comparison shows that there was a 2-3X increase in size
136

Figure 6. Secondary electron


image of Inconel 740 that
failed after 2185 h under a
stress of 265 MPa at 750C.
Two different magnifications
of same grain boundary area

Figure 7. Secondary electron images of Inconel 740H after


exposures at 750C for indicated times and applied stress
levels.

over several thousand hours for both Inconel 740 and 740H.
However, there was no change in the cuboidal shape of the . Furthermore, while preliminary,
there was no evidence of depletion near the grain boundaries for either Inconel 740 (2185 h,
265 MPa, 750C, Fig. 6) or Inconel 740H (4864 h, 220 MPa, 750C, Fig. 9)

Figure 8. TEM images of grip (zero stress)


sections of creep specimens of Inconel 740 (top)
and 740H (bottom) exposed at 750C for the
times indicated.
137

Figure 9. STEM image of a grain


boundary in Inconel 740 after
4864 h under a stress of 220 MPa.

Only preliminary microstructural


analysis of as-tested Haynes 282 has
been conducted to date. As with the
Inconel 740, grain boundaries were
decorated with carbides with typical
arrays within the grains (Fig. 10). In
this particular case, (2180 h, 300 MPa,
750C), depletion at the grain
boundaries was noted. Eta phase was
not observed under these conditions,
but, with only small concentrations of
Ti and no purposefully added Nb
(Table 2), Haynes 282 would not be
expected to form such precipitates.

Figure 10. Secondary electron image of Haynes


282 after 2180 h at 220 MPa and 750C.

DISCUSSION
As with any alloy being used for long-term service at elevated temperatures, microstructural
stability over a defined period of operation is key to attaining desired properties and lifetimes.
With precipitation-strengthened alloys such as Inconel 740 and Haynes 282, particular concerns
are with (i) the coarsening of and concomitant disruption of the periodic structure needed for
effective obstruction of dislocation movement and/or (ii) the complete dissolution of near grain
boundaries.[4,10] For Nb- and Ti-containing alloys, there is a further issue of phase formation
and its direct and indirect effect on creep.[4,10] The microstructural data presented in this paper
clearly shows the development of in Inconel 740 at the temperatures of interest for A-USC
conditions (750C) and that stress exacerbates its formation. This influence of stress has been
reported previously (see, for example, refs. 4 and 10), may involve a reduction in the kinetic
barrier to precipitation,[10] and is considered to be relatively minor compared to the thermal
driving forces for formation and growth.[10] Qualitatively, the present observations also
indicated that phase development was certainly more strongly dependent on time at temperature
than on applied stress.
In contrast to Inconel 740, less formed during creep exposures of Inconel 740H, presumably
due to its lower concentration of Ti.[7,12] Despite the differences in susceptibility to phase
development between Inconel 740 and 740H, the creep-rupture lives of these two compositional
variants were similar at low-to-moderate stresses (Fig. 3). This is consistent with the conclusion of
Shingledecker and Pharr regarding a minimal effect of on creep resistance of Inconel 740
despite their observation of a few volume percent of this phase in crept specimens.[10]
In principle, coarsening of the phase during the long-term exposures should decrease creep
lifetimes.[16] However, as moderate coarsening of was observed for both Inconel 740 and
740H, no conclusion about the effect can be made because, taken as a whole, the test results from
the present study indicate comparable creep resistance of the two alloys. Xie et al. reported little
long-term coarsening for Inconel 740.[4] The depletion of near grain boundaries is also a
potential detriment to creep resistance, but no definitive effect can be deduced from the creep data
and the limited amount of microscopy performed to date. As shown above, depletion has not been
observed in the present study for the Inconel 740 and 740H specimens, but it has been reported
for one of these heats of Inconel 740 creep tested under the same conditions.[10] Interestingly,
while depletion was detected in Haynes 282 (Fig. 10), greater creep lifetimes are projected for
138

this latter alloy than that for aged Inconel 740 based on the analysis of the current data (see
below).
As shown in Fig. 3, Inconel 740H specimens that were only solution annealed before creep testing
had comparable rupture lives to those which were aged (4 or 8 h at 800C), except at the highest
applied stresses (> 300 MPa). Additionally, initial creep-rupture data for Haynes 282 (Fig. 4)
showed no difference in lifetimes between solution annealed and aged specimens across the entire
ranges of test stresses (120 to 420 MPa). Any conclusions about the implications of this set of
observations await detailed microstructural analysis and further creep-rupture data. Nevertheless,
these results would seem to indicate that creep testing at or close to the typical aging temperature
provides the necessary heat treatment to precipitate the phase responsible for long-term
strengthening. This self-aging phenomenon, of course, is only possible for alloys, such as those
studied here, that have sufficient intrinsic strength at these temperatures to avoid substantial
deformation damage during the time at which the precipitate structure is being developed. Indeed,
as mentioned above, this may explain why, at the highest stress levels (only), ruptures times for
Inconel 740H in the solution annealed condition were significantly shorter than those were aged
prior to testing.
The ability to predict longer-term rupture lives of a particular alloy based on relatively short-term
creep tests has been a longstanding problem in materials science and engineering that is made
more challenging for power-plant boiler applications where extremely long lifetimes (100,000s
of hours) are required.[3,17] This is even more problematical for boilers operating under A-USC
steam conditions, which require operation at higher temperatures and stresses that can more
readily promote processes that destabilize the creep-resistant microstructures. Indeed, in his recent
review, Evans[17] noted that while progress has been made in creep lifetime prediction for
traditional power plant materials (for example, 9-12Cr steels and austenitic stainless steel), there
is a need to extend such models to nickel-based alloys such as (specifically) Inconel 740 and
Haynes 282 and examine their effectiveness. While such a task is well beyond the scope of this
work, an attempt has been made to apply one of the current models to the present data to gain a
rough approximation of the time to rupture for Inconel 740 and Haynes 282 for A-USC steam
service.
Referring to Fig. 1, a simple extrapolation of the power law curve fit to the data shows that the
A-USC Steam Program goal of a 100,000 h lifetime at 100 MPa and 750C would be met by aged
Inconel 740. Furthermore, an analysis based on the best-fit curve to the stress versus LMP data for
aged Inconel 740 (Fig. 2) predicts, for 750C, a 100,000-h rupture life at 127 MPa. However, as a
first attempt to more rigorously predict long-term rupture time (t f ) based on short-time data, an
analytical treatment proposed by Wilshire et al.[17,18] was used. This approach goes beyond a

simple power law description by relating the applied stress, , to the rupture time, t f , through
where TS is the tensile strength, Q c is an activation energy, T the absolute temperature, R the gas
constant, and k 1 and u best-fit parameters. This analysis constructs, obviously, the right boundary
conditions at t f = 0 and = and has the form of a probability distribution function. It should be
noted that there are certainly other approaches to predicting long-term lifetimes that may be more
sophisticated in terms of physically based mechanics and microstructural models or accuracy.
Indeed, the present findings that, under many conditions, SA Inconel 740 and Haynes 282 show
similar creep-rupture lifetimes to those of their respective aged versions (cf. Figs. 3 and 4)
indicate that the Wilshire relationship between initial tensile strength and rupture time does not
139

hold, in this case at least, because TS in


the SA condition for both of these alloys
is 60-75% lower than after aging.[12]
Nevertheless, because the intent here was
simply to produce an estimate of creep
lifetime for both Inconel 740 and Haynes
282 from the present datasets for aged
material, and to compare the predicted t f
for the two alloys, a Wilshire type of
approach was used.
The Wilshire relation was used to
calculate the stress at which a 100,000 h
creep lifetime is at 750C (*) is
achieved. For Inconel 740, the dataset
used in the analysis was that for aged
(800C) specimens with only one 750C
data point for Inconel 740H included.
(For the purpose of this analysis, the data
points in Fig. 1 representing ongoing
creep tests were not used.) The data
fitting produced values of k 1 and u and a
reasonable activation energy (Q c ) for
nickel-based alloys. A plot of the 750C
data and the fitted line for as a function
of t f is shown in Fig. 11a. The predicted
* is 111 MPa. (Note that this is a more
conservative prediction than that coming
from the LMP analysis described above
where * was 127 MPa.) The Wilshire
analysis applied to the aged Haynes 282
750C creep-rupture results produced
similar agreement between the fitted line
and the cluster of data points (Fig. 11b).
For this alloy, a * of 119 MPa was
predicted.

(a)

(b)

Figure. 11. Wilshire analysis of creep-rupture


data at 750C. Solid curve is fitted Wilshire
function with dashed lines representing 10%.
(a) aged Inconel 740; (b) aged Haynes 282

Based on the Wilshire approach, both Inconel 740 and Haynes 282 are found to meet the A-USC
Steam Program goal of 100,000 h at 750C at an applied stress of 100 MPa. A similar analysis of
the aged Inconel 740H data yielded values of Q c and * (750C) that were physically unrealistic
given the data shown in Figs. 2 and 3. This was thought to be due to the lack of an adequate
number of data points, particularly at low stress values (longer t f s). Completion of the tests
indicated as still running in Fig. 2 should help in this regard. The 740H data shown in Figs. 2 and
3 suggest a similar * as Inconel 740.
SUMMARY
Creep-rupture testing of several heats of Inconel 740 was conducted over a range of
temperatures and stresses to develop confidence in qualitatively predicting creep lifetimes over a
140

range of conditions relevant to advanced ultrasupercritical steam conditions, with the longest
rupture times exceeding 30,000 h. For comparison, the creep-rupture behavior of Haynes 282
alloy was mapped as a function of temperature and stress, albeit not as extensively. Little
difference in creep-rupture results between Inconel 740 and Inconel 740H was found although the
latter alloy showed significantly greater resistance to phase formation during testing. For both
Inconel 740H and Haynes 282, little effect of prior aging treatments (for setting the precipitate
structure) on creep-rupture behavior was found; solution annealed specimens produced results
comparable to aged ones except for high stress tests with Inconel 740H. Results from a modified
power law analysis proposed by Wilshire et al. showed that, while both Inconel 740 and Haynes
282 are projected to meet the A-USC lifetime requirements, the latter offered better long-term
creep resistance.
ACKNOWLEDGMENTS
This research was sponsored by the U.S. Department of Energy (DOE), Office of Fossil Energy,
Advanced Research Materials Program. The authors thank the members of the DOE/Ohio
Economic Development Corporation A-USC consortium for comments on this work and
S.N. Dryepondt (Oak Ridge National Laboratory) and I.G. Wright (retired, Oak Ridge National
Laboratory) for review of the manuscript.
REFERENCES
[1] Viswanathan, R., Henry, J. F., Tanzosh, J, Stanko, G., Shingledecker, J., Vitalis, B., and
Purgert, R., U.S. Program on Materials Technology for Ultra-Supercritical Coal Power
Plants, J. Mater. Eng. Perf., 14 (2005), pp. 281-292.
[2] Blum, R. and Bugge, J., The European Perspective and Advancements for Advanced USC
Steam Power Plants, Proc. Sixth Intl. Conf. on Advances in Materials Technology for
Fossil Power Plants, Gandy, D., Shingledecker, J. and Viswanathan, R. (eds.), ASM
International, Materials Park, Ohio (2011), pp. 1-10.
[3] Masuyama, F., R&D Program for A-USC Material Development with Creep
Strength/Degradation Assessment Studies, Proc. Sixth Intl. Conf. on Advances in
Materials Technology for Fossil Power Plants, Gandy, D., Shingledecker, J., and
Viswanathanm, R. (eds.), ASM International, Materials Park, Ohio (2011), pp. 11-29.
[4] Xie, X., Chi, C., Yu, H., Yu, Q., Dong, J., Chen, M., and Zhao, S., Structure stability study
on fossil power plant advanced heat-resistant steels and alloys in China, Proc. Sixth Intl.
Conf. on Advances in Materials Technology for Fossil Power Plants, Gandy, D.,
Shingledecker, J., and Viswanathan, R. (eds.), ASM International, Materials Park, Ohio
(2011), pp. 30-52.
[5] Duffey, R. B., Pioro, I. L., and Kuran, S., Advanced Concepts for Pressure-Channel
Reactors: Modularity, Performance and Safety J. Power Energy Sys. 2 (2008), pp. 112221.
[6] Shingledecker, J. P. and Wright, I. G., Evaluation of the Materials Technology Required
for a 760C Power Steam Boiler, Materials for Advanced Power Engineering 2006,
Lecomte-Beckers, J., Carton, M., Schubert, F., and Ennis, P. J. (eds.), Schriften des
Forschungszentrums Jlich, Jlich (2006), pp. 107-119.

141

[7] Patel, S. J., deBarbadillo, J. J., Baker, B. A., and Gollihue, R. D., Nickel Base Superalloys
for Next Generation Coal Fired AUSC Power Plants, Procedia Eng. 55 (2013), pp. 246252.
[8] Viswanathan, R., Shingledecker, J., and Purgert, R., Evaluating Materials Technology for
Advanced Ultrasupercritical Coal-Fired Plants, Power 154 (2010), pp. 41-45.
[9] ASME Code Case 2702, ASME Boiler and Pressure Vessel Code, ASME International,
New York, 2012.
[10] Shingledecker, J. P. and Pharr, J. M., The Role of Eta Phase Formation on the Creep
Strength and Ductility of INCONEL Alloy 740 at 1023 K (750C) Mater. Metall. Trans. A
43A (2012), pp. 1902-1910.
[11] Baker, B. A. and Gollihue, R. D. Optimization of INCONEL Alloy 740 for Advanced Ultra
Supercritical Boilers, in Proc. Sixth Intl. Conf. on Advances in Materials Technology for
Fossil Power Plants, Gandy, D., Shingledecker, J. and Viswanathan, R. (eds.), ASM
International, Materials Park, Ohio (2011), pp. 96-109.
[12] Santella, M. L. and Stevens, C. O., Oak Ridge National Laboratory, unpublished results
(2011).
[13] Haynes International, Inc., Haynes 282 Alloy, technical bulletin H-3173 (2008).
[14] Unocic, K. A. and Tortorelli, P. F., Oak Ridge National Laboratory, unpublished results
(2013).
[15] Evans, N. D., Maziasz, P. J., Swindeman, R. W., and Smith, G. D., Microstructure and
Phase Stability in INCONEL Alloy 740 During Creep, Scripta Mater. 51 (2004), pp. 503507.
[16] Tien, J. K. and Copley, S. M. The Effect of Orientation and Sense of Applied Uniaxial
Stress on the Morphology of Coherent Gamma Prime Precipitates in Stress Annealed
Nickel-Base Superalloy Crystals, Metall. Trans. 2 (1971), pp. 543-553.
[17] Evans, M., The Potential United Kingdom Energy Gap and Creep Life Prediction
Methodologies, Mater. Metall. Trans. A 44A (2013), pp. S109-127.
[18] Wilshire, B. and Scharning, P., A New Methodology for Analysis of Creep and Creep
Fracture Data for 9-12% Chromium Steels J. Int. Mater. Rev. 53 (2008), pp. 91-104.

142

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

MICROSTRUCTURAL EVOLUTION IN CAST HAYNES 282 FOR


APPLICATION IN ADVANCED POWER PLANTS
Y. Yang1 and R. C. Thomson1
1

Department of Materials, Loughborough University, Loughborough, Leicestershire, LE11 3TU,


UK
R. M. Leese2 and S. Roberts2

Goodwin Steel Castings Limited, Ivy House Foundry, Ivy House Road, Hanley, Stoke-on-Trent,
Staffordshire, ST1 3NR, UK

ABSTRACT
There is a worldwide drive to increase the efficiency of power plants in order to reduce the
amount of fossil fuel consumed and associated CO2 emissions. Raising the operating temperature
and pressure can improve the thermal efficiency, however, this necessitate the use of materials
which have high temperature performance. Steels are currently used at temperature up to 600C
with the efficiency of 38-40 %. Advanced Ultra Supercritical (A-USC) design plans power plants
to operate at steam temperatures of 700C and pressure up to 35 MPa with a lifetime of at least
100 000 hours. Ni-base superalloys are leading materials due to their significant strength and
creep resistance.
Haynes 282 is one possible candidate to meet the A-USC conditions for turbine engines. This
alloy is a precipitation strengthened material and exhibits very good creep properties and
thermal stability. The alloy examined in this research was produced by sand casting, and
therefore the aim of this research is to investigate the microstructural evolution in large scale cast
components.
The alloy has been examined in both the as-cast condition and as a function of a range of
different pre-service heat treatments. The microstructural changes during different heat treatments
have been fully identified and quantified. The results have also been compared with predictions
from thermodynamic equilibrium calculations using a Ni alloy database. It has been found that
variations in the heat treatment conditions can have a significant effect on microstructural
development and hence, potentially, the mechanical properties of Haynes 282 alloy.
1 INTRODUCTION
Haynes 282 was first introduced in 2005 as an advanced wrought gamma prime strengthened
superalloy for high temperature applications such as aircraft and land based gas turbines [1]. It is
notable for its exceptional high temperature creep strength, good weldability and fabricability and
its resistance to strain-age cracking [1].
After fabrication, wrought Haynes 282 materials often undergo a standard heat treatment: a
solution annealing treatment followed by quenching and then a two-stage precipitation treatment
with air cooling to achieve optimum mechanical properties [1, 2]. The solution annealing
temperature is designed to be above the solvus of M23C6 to make the material soft and ductile [2].
The first precipitation treatment temperature is above the solvus of gamma-prime and it is
designed to precipitate out M23C6 both intergranularly and intragranularly to give optimum creep
strength and high temperature ductility [2]. The second precipitation treatment is designed to
form to strengthen the material [2].

143

It is well known that there are many phases present in Ni-based superalloys, including different
carbides and TCP phases [3] and that these phases can have different effects on mechanical
properties. It has been reported that the microstructure of the heat treated Haynes 282 contains
MC carbide, Cr rich M23C6 both intergranularly and intragranularly, and also a uniform
distribution of spherical particles with a typical size of 20 nm [2].
There are a number of studies in the literature concerned with the microstructure of Haynes 282,
particularly focused on welds [4-8]. One study of the fusion zone in a laser weld of Haynes 282
observed that Co, Cr and Al were homogeneously distributed between the dendrite core and
interdendritic regions, whereas the Ti and Mo were rejected to the interdendritic region and
formed Ti-Mo rich, MC type carbides [6]. Three types of particles have also been identified along
the grain boundaries in Haynes 282 weld metal which were Cr-rich, Mo-rich and Ti-rich [8]. It
was also observed that there was B peak in the Mo-rich particle, and it was suggested that the
Mo-rich phases are M5B3 with lattice parameter of a=0.56 nm and c=1.01 nm [8]. There are
reports of other borides present in the alloy [5], including Nirich, Cr-rich and Mo-rich borides of
the form M3B2. Additionally, it has been reported the grain boundaries were covered by both
M23C6 and M5B3 [4, 7]. However, other authors [9] have also found a Mo rich phase before
precipitation treatment which was also associated with C and Si peaks, and they postulated that
the Mo rich phase was M6C. Therefore, it is still necessary to undertake a detailed microstructural
study of Haynes 282, and in particular most of the literature is focussed on the wrought alloy.
This research therefore is concerned with a full microstructural investigation of cast Haynes 282
in various heat treated conditions with possible application for heavy sections in the power sector.
2 MATERIALS AND EXPERIMENTAL PROCEDURES
The material studied in this research was produced by Goodwin Steel Castings Limited using
sand casting. The material examined was cut from the feeder part of a large cast section. The
nominal composition of Haynes 282 is given in Table 1. The initial solution heat treatment was
slightly modified to allow for the fact that this was a large casting rather than a wrought alloy,
with a two-stage solution treatment being applied comprising of 1100C for 6 hrs and then a ramp
up to 1150C for 15 hrs followed by water quenching. After the solution treatment, a two-stage
ageing treatment was applied which was the same as the standard Haynes precipitation treatment
but with a longer holding time at 1010C for 5 hrs followed by quenching, and then a second
ageing treatment at 788C for 15 hrs followed by air cooling.
Eight samples cut from the cast 282 material were examined in this paper and their condition and
code names are provided in Table 2. The material was examined in the as cast condition (AC) for
comparison purposes, then after solution treatment (ST), after the first stage aged condition (STFA), and after solution treatment and both aged conditions for different ageing times of the
second ageing treatment: 15 hrs, 100 hrs, 1000 hrs, 2000 hrs and 3000 hrs (ST-FA-SA** where
** indicates the length of the heat treatment). Samples were cut to a suitable size and then
mounted in conducting Bakelite. Mounted samples were ground on SiC papers from 220 to 1200
grit, and then polished with 9, 3, 1 m diamond solution and finally polished with colloidal silica
(0.5 m) for 30 min. To quantify gamma prime, the etchant Kallings no.2 (CuCl2-2g, HCL-40 ml
and ethanol-40-80 ml) was used for 10 s to highlight the . The Vickers hardness of all samples
were measured by a Mitutoyo HM-124 machine using a 10 kg load and a 15 s dwell time. For
each sample, ten measurements were performed and the average value was taken.
Table 1: The nominal composition in wt. % of Haynes 282
C
0.06

Cr
20

Mo
8.5

Si
0.15

Ti
2.1

144

Co
10

B
0.005

Al
1.5

Ni
Bal.

Table 2: Conditions and code names of all the samples in this paper. The solution treatment was
carried out at 1100oC for 6 hrs, followed by a ramp to 1150oC for 15 hrs and water quenched.
First ageing was carried out at 1010oC for 5 hrs followed by water quenching, and the second
ageing was at 788oC for various times followed by air cooling.
Sample name
AC
ST
ST-FA
ST-FA-SA15
ST-FA-SA100
ST-FA-SA1000
ST-FA-SA2000
ST-FA-SA3000

As Cast

Heat treatment condition


Solution Treated
First Aged

Second Aged

15 hrs
100 hrs
1000 hrs
2000 hrs
3000 hrs

Thermodynamic calculations were performed using the MTDATA [10] software developed by
the National Physical Laboratory with an appropriate Ni-database [11]. The alloy composition
was used to predict the equilibrium phases between 300-1600C and the chemical composition of
each phase as a function of temperature. Solidification and chemical segregation profiles were
also predicted using the Scheil methodology (Tscheil).
Different microstructural features were assessed using a Carl Zeiss (Leo) 1530 VP field emission
fun scanning electron microscope (FEG-SEM). The backscatter mode was used to show different
phases due to its atomic number contrast. Energy dispersive X-ray (EDS) analysis was also used
for chemical analysis, with an operating voltage of 20 kV. Transmission electron microscopy
(TEM) was used to study small precipitates. A JEOL 2000FX TEM was used to obtain the
diffraction patterns of different precipitates and an fei F20 Tecnai FEG-TEM was used to obtain
EDS maps and spectra of different small precipitates. The site specific TEM samples were
prepared using an fei Nova Nanolab 600 dual beam which consists of a field emission electron
column and a gallium source ion column.
3 RESULTS AND DISCUSSION
3.1 Thermodynamic calculations
The calculated equilibrium phase diagram is illustrated in Figure 1. The phases considered in the
calculation were liquid, gamma, gamma prime, MX, M23C6, M6C, M3B2 and Mu. From Figure 1
(a), it can be seen that the predicted melting range is from 1300C to 1380C. Gamma prime
forms below 1000C and nearly 20 wt. % is precipitated at 788C, which is the standard second
ageing treatment temperature. The magnified phase diagram of the precipitates present in smaller
quantities, including MX, M6C, M23C6 and M3B2, is illustrated in Figure 1 (b). It can be seen that
MX is present until melting, and M3B2 and M23C6 precipitate out below 1250C and 800C
respectively. It should be noticed that M6C is present between 780C to 1080C. This is because
below 780C, the M23C6 is more stable [3].
The equilibrium composition of the different precipitates including gamma prime, MX, M6C,
M23C6, M3B2, and Mu were also predicted as illustrated in Figure 2. It can be seen that MX is a Ti
rich carbon nitride and M23C6 is Cr rich. It should be noted that the M6C, M3B2 and Mu phases
are all Mo rich, however, M6C is Mo and Ni rich, M3B2 is Mo and Cr rich and the Mu phase is
Mo and Co rich.
145

b
Mass percentage of different phases

Mass percentage of different phases

a
Liquid

100
80

LIQUID
GAMMA_PRIME
Gamma
MX
M23C6
M6C
M3B2_TETR
MU_PHASE

Gamma
60
40
Gamma_prime

20

Mu

0
400

600

800

1000

1200

1400

1600

1.2
1.0

M6C

LIQUID
GAMMA_PRIME
Gamma
MX
M23C6
M6C
M3B2_TETR
MU_PHASE

0.8
0.6

M23C6

0.4
0.2

MX
M3B2

0.0
400

600

800

Temperature (C)

1000

1200

1400

1600

Temperature (C)

Figure 1: Plots of the equilibrium phase diagram of the Haynes 282 (a) and at a magnified scale
for the minor precipitate phases M6C, M23C6, M3B2 and MX (b).
a

b
MX

Gamma prime

0.85

Ni

0.80
0.75

0.9
0.8

0.60
0.15

Ti

0.10
0.05
0.00
400

Co

Al

Cr
500

600

700

800

900

Mo

0.6

0.4

0.2

0.0
400

1000

600

800

Temperature (C)

0.1

1200

1400

0.0
400

1600

Co
840

880

550

920

960

Temperature (C)

1000

1040

1080

0.3

Cr

0.2

400

700

750

800

800

850

0.35

0.4

0.0

650

Mo

0.40

0.5

0.1

600

Mu phase

0.45

Mass fraction

Mass fraction

Mass fraction

Ni
Cr

800

500

0.50

Mo

0.7

0.4

0.1

450

Co

0.6

0.2

Ni

C
Temperature (C)

M3B2 phase

0.8

Mo

0.3

Mo

0.2

M6C

0.5

1000

0.6

Temperature (C)

d
0.6

Cr

0.7

Mass fraction

Mass fraction

0.65

0.0

M23C6

1.0

Ti

0.8

0.70
Mass fraction

0.30

Co

0.25
0.20

Cr

0.15
0.10

Ni

0.05

600

800

1000

Temperature (C)

1200

0.00
400

450

500

550

600

650

700

750

Temperature (C)

Figure 2: Plots of the calculated equilibrium composition of different phases: (a) gamma prime;
(b) MX; (c) M23C6; (d) M6C; (e) M3B2; and (f) Mu
Since segregation is a normal phenomenon during casting, Scheil calculations have also been
carried out and the results are illustrated in Figure 3. From the phase diagram in Figure 3 it can be
seen that the solidification sequence in the Haynes 282 is gamma, MX, M6C, M3B2, Mu and
gamma prime. M6C is present from 1250C to 1175C, then M3B2 forms and followed by Mu
phase. From Figure 3 (b), it can be seen that there is a relatively high amount of MX which forms
together with gamma at the start of solidification, and Figure 3 (c) shows that the initial MX is
rich in nitrogen, a Ti rich carbonitride whereas later it becomes a Ti, Nb rich carbide. This is very
similar to the equilibrium prediction but with much more Nb and Mo due to segregation. Figure 4
146

illustrates the segregation ratio of elements in the liquid during solidification and it is obvious
that Nb and B are most heavily segregated, together with Ti and Mo.
b

Liquid
Gamma_prime
Gamma
MX
Liquid
M6C
M3B2
Mu-phase

80
60
40
20

Gamma

0
1100

1150

1200

1250

1300

1350

1400

0.06
0.05

c
Liquid
Gamma_prime
Gamma
MX
M6C
M3B2
Mu-phase

Gamma
Gamma_prime

0.04
0.03
0.02

M3B2

0.00
1100

1150

0.80

M6C

Mu

0.01

0.85

Mass fraction

100
Mass percentage of different phases

Mass percentage of different phases

MX
1200

1250

1300

0.70

1350

0.10

Cr

Nb

Mo

0.00
1100

1400

0.15

0.05

Temperature (C)

Temperature (C)

Ti

0.75

1150

1200

1250

1300

1350

C
Ni
Cr
Mo
Nb
N
Ti
Co
B
Al
1400

Temperature (C)

Figure 3: Predictions from Tscheil calculations: (a) phase diagram; (b) enlarged phase diagram
for MX, M6C, M3B2, Mu and gamma prime; (c) predicted composition of MX formed during
solidification
b

70

C
Ni
Cr
Mo
Nb
N
Ti
Co
B
Al

60
50

40
30

Nb

20

10
0
1100

1150

1200

1250

1300

1350

1400

Temperature (C)

Segregation ratio of different elements

Segregation ratio of different elements

C
Ni
Cr
Mo
Nb
N
Ti
Co
B
Al

5.0
4.5
4.0
3.5

Ti

3.0
2.5

Mo

2.0

Co

1.5

Cr

1.0
0.5
0.0
1100

Ni

Al
1150

1200

1250

1300

1350

1400

Temperature (C)

Figure 4: Tscheil calculation predicted plots of segregation ratio of elements in liquid during
solidification (a) and the enlarged plot for C, Ti, Mo, Ni, Co, Cr and Al (b)
3.2 Hardness
The results from hardness testing of all samples are shown in Figure 5. It can be seen that the AC
sample has a higher hardness than the ST sample, which in turn has the lowest overall hardness as
a result of the dissolution of all of the strengthening precipitates. The AC sample was observed to
contain some gamma prime, as indicated in Figure 3 (b). After first and second ageing, the
hardness increased by up to ~160 HV compared to the solution treated sample. However, it was
noted that the ST-FA sample was ~40 HV higher than the ST sample in hardness, implying that
this increase in hardness is a result of precipitation during the first ageing. Therefore, in
comparing the hardness of the FA and SA samples, it can be inferred that gamma prime results in
~120 HV strengthening. Comparing the second ageing samples, it can be seen that the ST-FASA1000 sample and the ST-FA-SA2000 sample approach the highest hardness with very similar
values. The absolute values of hardness are a function of both the amount and size distribution of
147

gamma prime. It was, however, observed that the hardness of the ST-FA-SA3000 sample was
lower than the other aged samples, which could be caused by the coarsening of the gamma prime
distribution, as discussed further in section 3.5.
400

315

350

Hardness (HV10kg)

300

250

243

250

281

314
264

189

200

151

150
100
50
0

AC 282

ST 282 ST-FA
282

ST-FA- ST-FA- ST-FA- ST-FA- ST-FASA15 SA100 SA1000 SA2000 SA3000


282
282
282
282
282

Figure 5: A plot showing the hardness of all of the samples examined in this research.
3.3 MX precipitates
Two types of MX precipitates were found in the AC condition: one was a Ti rich nitride with a
cubic morphology as shown in Figure 6 (a) and the other was a Ti, Mo rich carbide with a blocky
morphology along grain boundaries (Figure 6b). These results are in agreement with the predicted
composition of MX in the segregated state, as presented in Figure 3 (c). The diffraction pattern of
the cubic TiN illustrated in Figure 6 (a) is consistent with an FCC structure with a lattice
parameter of 0.425 nm, and the diffraction pattern of the blocky boundary (Ti,Mo)C illustrated in
Figure 6 (b) indicates that the (Ti,Mo)C phase also has an FCC structure with a lattice parameter
of 0.432 nm. Compared to TiN, the increase in lattice parameter of the (Ti,Mo)C is caused by the
larger substitutional atom, Mo, substituting for Ti.
After the solution treatment, the blocky (Ti,Mo)C particles disappeared and instead complex MX
particles formed comprising TiN surrounded by (Ti,Mo)C, as illustrated in Figure 7, which shows
the morphology of the complex MX in different aged conditions. It can be seen that the complex
MX were stable during subsequent ageing treatments. However, it should be noticed that after the
first ageing treatment, white precipitates were observed around the complex MX. From the
equilibrium phase diagram, the white precipitates could be M6C or M3B2. However, the solvus of
M3B2 is above the solution treatment temperature and it was not observed in the ST sample.
Therefore the white precipitates are more likely to be M6C, as discussed in section 3.4, and were
observed to coarsen slightly during the second ageing treatment with increasing ageing time.

148

Figure 6: Micrographs of the microstructure of the AC sample: (a) cubic TiN surrounded by
(Ti,Mo)C on a grain boundary and the diffraction pattern of the TiN, zone axis [-2-11]; (b)
blocky (Ti,Mo)C along a grain boundary and its diffraction pattern, zone axis [0-1-3]; (c) a EDS
spectrum of the TiN; and (d) a EDS spectrum of the (Ti,Mo)C
a

TiN

(Ti,Mo)C

Figure 7: Micrographs of the heat treated samples: (a) the ST sample; (b) the ST-FA sample; (c)
the ST-FA-SA15 sample (d) the ST-FA-SA100 sample; (d) the ST-FA-SA1000 sample and (f) the
ST-FA-SA2000 sample.

149

3.4 Boundary precipitates


Two types of grain boundary precipitates were found in the AC condition as shown in Figure 8
(a). The EDS spectra in Figure 8 (b and c) indicated that the white precipitates were Mo, Si rich
carbide and the dark precipitates were Cr rich carbides. The diffraction patterns of the dark
precipitate in Figure 8 (f and g) show an FCC structure with a lattice parameter of 1.06 nm.
Therefore, it can be concluded that the Cr rich dark precipitates are Cr rich M23C6. From the
equilibrium composition diagram in Figure 2, it can be seen that the Mo containing phases are
M6C, M3B2 or Mu. The diffraction patterns of the white precipitates in Figure 8 (d and e) show a
diamond cubic structure with a lattice parameter of 1.10 nm. However, M3B2 has a primitive
tetragonal structure, where lattice parameters a and c are 0.59 and 0.31 nm respectively [12] and
Mu has a rhombohedral structure where lattice parameters a and c are 0.9 and 3.08 nm
respectively [13]. Therefore, the boundary white precipitates are consistent with M6C.
a

b
1

1
c
2

Figure 8: A micrograph of the grain boundary in the AC sample (a); (b) a spectrum from the
white precipitate; (c) a spectrum from the dark precipitate; (d and e) the diffraction patterns of
the white precipitate with zone axis [-112] and [-111] respectively and (f and g) the diffraction
patterns of the dark precipitate with zone axis [1-1-2] and [-123] respectively
After the solution treatment both boundary precipitates were dissolved, as illustrated in Figure 9
(a). It can be seen that there are only complex MX particles in the grain boundaries in the ST
sample. This is because the solution treatment temperatures are above the solvus of both M6C and
M23C6. The disappearance of the M6C after the solution treatment also provides further evidence
that the white precipitates are not M3B2 since the solvus of the M3B2 is above the solution
treatment temperature. During the first ageing treatment, M6C re-precipitated and covered the
grain boundaries again as seen in Figure 9 (b). From the equilibrium phase diagram in Figure 2
(b), it can be seen that at 1010C, the microstructure is predicted to comprise gamma, MX and
M6C, which is consistent with those observed in the microstructure of the ST-FA sample. After
the second ageing treatment, M23C6 re-precipitated and covered the grain boundaries, together
with M6C, as shown in Figure 9 (c). A specific lift out sample of the grain boundary in the ST150

FA-SA15 was made and EDS maps from the FEG-TEM also proved the grain boundaries were
covered by alternating Mo and Cr rich carbides. Figure 10 shows micrographs of the ST-FA-SA
samples after different ageing times. It can be seen that with the increasing ageing time at 788C,
both M6C and M23C6 coarsened.
a

Figure 9: Micrographs of the grain boundary structure of different samples: (a) the ST sample;
(b) the ST-FA sample; (c) the ST-FA-SA15 sample; (d) A TEM micrograph of a grain boundary
in the ST-FA-SA15 sample and the corresponding EDS maps of Mo, Cr and C.
a

Figure 10: Micrographs of the grain boundary structure of different aged samples: (a) the STFA-SA100 sample; (b) the ST-FA-SA1000 sample and (c) the ST-FA-SA2000 sample

151

3.5 Gamma prime


Gamma prime has been found in the AC sample, in agreement with the predicted solidification
sequence. It should be noticed that the gamma prime in the AC sample appeared as clusters of
gamma prime particles in a star morphology, consistent with previous results [14]. The gamma
prime particles were then observed to re-precipitate out during the second ageing treatment in a
spherical morphology, as shown in Figure 11.
The coarsening behaviour of the gamma prime particles has been quantified by investigating their
average size and distribution in each aged sample. Image analysis software was used to analyse
15 micrographs for each sample condition and a significant number of the particles were
measured (over 5000) to ensure good statistical significance. The average size of the gamma
prime in the ST-FA-SA100 sample, the ST-FA-SA1000 sample, the ST-FA-SA2000 sample and
the ST-FA-SA3000 sample were 42 nm, 102 nm, 119 nm and 231 nm respectively. Figure 12 (a)
illustrates the distribution of the gamma prime in the different aged samples, and it is obvious that
the distribution of gamma prime broadened with increasing time. It should be noted that the
average size in the ST-FA-SA1000 sample and the ST-FA-SA2000 sample are nearly the same,
which may explain why the hardness of these two samples was very similar. However, the size of
the gamma prime in the ST-FA-SA3000 sample is about twice that of the ST-FA-SA2000 sample,
consistent with the drop in hardness observed. Figure 12 (b) illustrates the coarsening rate of the
gamma prime. It can be seen that the result agrees with the growth rate r3t [15].
It was also noted that the morphology of the gamma prime was still spherical up to 3000 hrs
ageing. It is known that spherical gamma prime has the smallest misfit with gamma, and this may
explain why Haynes 282 has good creep properties.
a

Figure 11: Micrographs of the gamma prime in different samples: (a) the AC sample; (b) the STFA-SA100 sample; (c) the ST-FA-SA1000 sample; (d) the ST-FA-SA2000 sample and (e) the STFA-SA3000 sample.

152

b
250

ST-FA-SA100 282
Average size (nm)

Thousands

ST-FA-SA1000 282

Count

ST-FA-SA2000 282

ST-FA-SA3000 282

200
150
100
50
0

0
0

500
Feret diameter (nm)

1000

5
10
15
Ageing time t1/3 (h)

20

Figure 12: Plots showing the distribution of the gamma prime in the different aged samples (a)
and the coarsening of the gamma prime (b).
4 CONCLUSIONS
Microstructural evolution of cast Haynes 282 has been fully studied using a number of analytical
techniques, focusing on the different precipitates present. The observed phases present within the
microstructures agreed well with the thermodynamic equilibrium and Scheil predictions.
In the AC condition, two types of MX were found, which were TiN and (Ti,Mo)C. M23C6, which
is Cr rich and M6C, which is Mo rich, were found to cover the grain boundaries. Gamma prime
particles present in star morphology were also found in the AC sample, which had a relatively
high hardness value of 243 HV.
After the solution treatment, the (Ti,Mo)C particles dissolved and formed a complex MX
comprising TiN surrounded by (Ti,Mo)C. The other precipitates were all dissolved and thus the
material was measure to have a significantly lower hardness value (151 HV).
M6C was precipitated out during the first ageing treatment and covered the grain boundaries. This
precipitation provided the material a hardness increase of 40 HV. During the second ageing
treatment, M23C6 was precipitated and covered the grain boundaries, alternating with M6C. With
increasing ageing time, both M23C6 and M6C coarsened. Spherical gamma prime particles were
found during the second ageing treatment, and gave the material a maximum hardness increase of
120 HV. The coarsening of the gamma prime particles was observed to be proportional to time to
the power one third. It should be noted that spherical gamma prime particles were found up to
3000 hrs, which could contribute to the alloys relatively good creep properties.
5 ACKNOWLEDGEMENTS
The authors would like to acknowledge the support of Goodwin Steel Castings and
Loughborough University for this study.
6 REFERENCES
[1] Haynes 282 alloy Brochure, Online Literature, Haynes International, Inc. [Online] Access at:
www.haynesintl.com/literature.htm
[2] Pike, L.M., Development of Fabricable Gamma-prime (c9) Strengthened Superalloy, Proc.
11th Int. Symp. Superalloy 2008, Champion, PA, USA, September 2008, TMS, pp.191200.
[3] Sims, C.T., Stoloff, N.S. and Hagel, W.C., Superalloy II, John Wiley & Sons (New York,
1987), pp. 101.

153

[4] Buckson, R.A. and Ojo, O.A., Cyclic Deformation Characteristics and Fatigue Crack Growth
Behaviour of a Newly Developed Aerospace Superalloy Haynes 282, Materials Science and
Engineering A, Vol. 555, (2012), pp. 63-70.
[5] Ghoneim, A and Ojo, O.A., Microstructure and Mechanical Response of Transient Liquid
Phase Joint in Haynes 282 Superalloy, Materials Characterization, Vol. 62, (2012), pp.1-7.
[6] Osoba, L.O., Ding, R.G. and Ojo, O.A., Microstructural Analysis of Laser Weld Fusion
Zone in Haynes 282 Superalloy, Materials Characterization, Vol. 65, (2012), pp. 93-99.
[7] Osoba, L.O. and Ojo, O.A., Influence of Laser Welding Heat Input on HAZ Cracking in
Newly Developed Haynes 282, Materials Science and Technology, Vol. 28, No. 4, (2012), pp.
431-436.
[8] Osoba, L.O. and Ojo, O.A., Improved Resistance to Laser Weld Heat-Affected Zone
Microfissuring in a Newly Developed Superalloy Haynes 282, Metallurgical and Materials
Transactions A, Vol. 43A, (2012), pp. 4281-4295.
[9] Boehlert, C.J. and Longanbach, S.C., A Comparison of the Microstructure and Creep
Behaviour of Cold Rolled Haynes 230 Alloy and Haynes 282 Alloy, Materials Science and
Engineering A, Vol. 528, (2011), pp. 48888-4898.
[10] Davies, R.H., Dinsdale, A.T., Gisby, J.A., Robinson, J.A.J. and Martin, S.M., MTDATAThermodynamics and Phase Equilibrium Software from the National Physical Laboratory, (2002),
pp. 229-271.
[11] Saunders, N., Proc. 8th Int. Symp. On Superalloys (Superalloys 1996), (ed R.D. Kissinger
et. al), 101-110; 1996, Warrendale, P.A., The Minerals, Metals & Materials Society, Seven
Springs.
[12] Zhang, H.R., Ghoneim, A. and Ojo, O.A., TEM Analysis of Diffusion Brazement
Microstructure in a Ni3Al-based Intermetallic Alloy, Materials Science, Vol. 46, (2011), pp.
429-437.
[13] Darolia, R., Labrman, D.F. and Field, R.D., Formation of Topologically Closed Packed
Phases in Nickel Base Single Crystal Superalloys, Superalloys 1988, (ed. S. Reichman, D.N.
Duhl, G. Maurer, S. Anolovich and C. Lund), pp. 255-264, 1988, The metallurgical Society.
[14] Durand-Charre, M., The Microstructure of Superalloys, Gordon and Breach Science
Publishers (1997), pp.71-72.
[15] Zhao, S., Xie, X., Smith, G.D. and Patel, S.J., Research and Improvement on Structure
Stability and Corrosion Resistance of Nickel-Base Superalloy Inconel Alloy 740, Vol. 27,
(2006), pp.1120-1127.

154

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

INVESTIGATIONS ON NICKEL BASED ALLOYS AND WELDS


FOR A-USC APPLICATIONS
Andreas Klenk, Karl Maile,
Materials Testing Institute University of Stuttgart (MPA)

ABSTRACT
In several material qualification programs tubes and thickwalled components mainly from Alloy
617 and Alloy 263 were investigated. Results as low cycle fatigue and long term creep behavior
of base materials and welds are presented. Numerical models to describe the material behavior
have been developed and verified by multiaxial tests. In order to ensure the feasibility of A-USC
plants two test loops have been installed in GKM Mannheim one for tube materials and a new
one for thick-walled piping and components. The latter consists of a part with static loading and a
part subjected to thermal cycles and is in operation since November 2012. First results of
measurements and numerical calculations for a pipe bend (static loading) as well as pipes and a
header (thermal cycles) are presented.
INTRODUCTION
At the beginning of this millennium the discussions on CO 2-reduction lead to strong efforts to
develop materials and components for high efficiency fossil fired steam power plants. In this
context world-wide research initiatives were established, Figure 1. Nowadays several new
challenges for power generation arise from the changes in political restrictions and demands.
There are still strong demands on reduction of CO2-emissions and subsequently raising efficiency
of fossil fired plants. All conventional plants are facing the demand of more flexible operation
due to the growing production using renewable energy sources, which are not constantly
available. This means that Nickel based alloys are required to ensure sufficient high-temperature
creep resistance [1],[2]. Research work on qualification and lifetime assessments for all materials
used in a A-USC plant, i.e. new nickel based materials, martensitic steels as VM12 and P92 and
ferritic steels, e.g. 7CrMoVTiB10-10 (T24) and their welds have been performed [3] and the
investigations on cyclic behavior of these materials have been enhanced recently.
For tubes, piping and headers in the boiler of A-USC-plants the Nickel alloys 617, 263 and 740
are under investigation. For these materials which are partly known from other application (e.g.
for gas turbines) first problems of manufacturing of tubes and mainly heavy-section components
had to be solved. Since the material properties which are needed for design and life assessment
are dependent on the manufacturing and heat treatment processes applied, basic material
qualification of material taken from components and weldments were necessary. Therefore the
first sections will focus on properties of Alloy 617, Alloy 263 and their welds.
For the application in a real plant tests under conditions as close to the service as possible are
necessary. This was the reason for installing test rigs in plants as shown in Figure 2. Two test rigs
in the GKM-plant in Mannheim are currently in operation.

155

Figure 1: World-wide research and development for high efficiency plants.


EU-FUNDED PROJECTS

NATIONAL FUNDED PROJECTS

700 FIELD TEST RIGS

Figure 2: Test loops to complement material and component qualification by basic research
Up to now, components are being designed with the help of codes and standards or by empirical
knowledge. In order to reduce the costs related to the new nickel based alloys and an economic
and efficient use of other materials, a Taylor-made design of components is required. Extended
steam conditions would also yield high wall thickness in components and piping if they are
designed according to currently existing codes and standards. However, in some cases these
requirements are contrary to the possibilities of manufacturing. Furthermore high thermal stresses
during start-up and shut-down must be taken into account in such components due to high wall
thickness. Actual codes and standards do not take into account the inelastic material response (for
example the relaxation of secondary stresses). One way to improve the design of new high
temperature components is to use modern constitutive equations for design calculations. The
European Pressure Equipment Directive (PED) allows for example the use of Design By Analysis
(DBA). Consequently, Finite Element Method (FEM) can be used to predict the creep and
relaxation behavior of boiler components for 700C fossil power plants. Since these models are
mostly phenomenological, a detailed experimental database is necessary to fit simulation
parameters. Therefore the development of such design and life assessment tools and their
application to components are of growing importance in the recent research projects [4].

156

EXPERIMENTAL DETERMINATION AND NUMERICAL DESCRIPTION OF THE


BEHAVIOR OF NI-BASED ALLOYS
Overview on Research Projects
Within the German Cooretec initiative a couple of materials qualification projects have been
already finished, Table 1. These projects solved manufacturing problems and provided a first
basis of knowledge on material properties. Furthermore advanced descriptions of material
behavior and parameters of constitutive equations have been provided. In order to enhance this
data base in a way that it is sufficient for approvals according to existing design rules a
comprehensive investigation program was established in the research project Cooretec-DE4 [5],
Fig. 2. Interfaces to other Cooretec projects were given by supplying material for investigations
on oxidation behavior (WP9) and nondestructive testing (WP10) and applying results from basic
investigations e.g. to allow fracture mechanics assessments using NDE-findings.
Table 1: Overview on research project for A-USC material qualification
Boiler application
MARCK0 700 Material qualification for the 700/720 C power plant
DE-1
Fireside corrosion and steam side oxidation behavior of materials for
700 C
power plant
DE-2
Characterization of superheater materials after cold deformation
FDBR02
Qualification of pipes with longitudinal welds made of Alloy 617
DE-4
Characterization of strength and deformation of pipes and forgings made of Nibased alloys

Status
Finished
Finished

725HWT (1)
Phase 2

Investigation of the long term service behavior of tubes for the future highefficiency power plant

Finished
Ongoing

HWT 2

Investigation of the long term service behavior pipes, bends and headers under
static and cyclic loading

Ongoing

Finished
Ongoing
Finished
06/2013

Turbine application
DT-3
Qualification of dissimilar welds between 10%Cr-steels and Ni-based alloys

Finished

DT-4

Procedures and Fracture mechanics approaches for life assessment of


components operating in high temperature regime

Finished

WK2

Advanced material concepts for forged and cast turbine components

Ongoing

Turbine and pipe work application


TD-1
Optimization of non-destructive testing methods for thick walled components
made of Ni-base alloys

Workpackage 1:
Sampling and Analysis of
existing data
Workpackage 4 : Creep
and Fatigue Strength and
Ductility of base materials

Workpackage 5:
Microstructural
characterisation

Workpackage 6:
Creep and Fatigue
Strength and Ductility of
Welded Joints

Workpackage 2:
Manufacturing and
Processing

Strength and
Ductility of Alloy 263
Piping and Forgings
and Alloy 617
Forgings in the
longterm range for
use in High
Efficiency Power
Plants
Workpackage 3:
Welding Processes,
Optimisation of
welding consumables,
Welding

Finished

Workpackage 7:
Evaluation of material and
component behaviour
Workpackage 8:
Fracture Mechanics
Characterisation and
Prevention from Brittle
Fracture
Workpackage 9: High
Temperature
Oxydation and
Corrosion

Workpackage 10:
Quality assurance

Figure 3: Work packages of the Cooretec-DE 4 project.


In order to assess the material behavior for power plant operation up to 725 C and 350/200 bar
pressure two test rigs have been installed in the Grosskraftwerk Mannheim AG (GKM). Test rig
one (HWT 1 high temperature materials test loop 1) is focused on tubes [6]. The operation of
HWT 1 delivered experiences with the oxidation and corrosion behavior of superheater tubes,
157

creep deformation and rupture behavior of tubes, oxidation behavior of coated turbine materials,
functionality of high temperature control and shut down valves. In the first phase almost 9000 h
of operation under more than 700 C have been accomplished with the operation of HWT 1.
Currently a second phase is in progress, tests on some samples are continued and new samples
have been included into the test rig. The test rig and results so far are described in more detail in
[7] and [8].
Test rig two (HWT 2) is aimed towards material behavior of thick walled pipes under static and
cyclic load conditions with additional superimposed secondary stresses [9],[10]. By injection of
cooling steam and water more than six thermal load cycles per day have been achieved. Duration
of one cycle is approximately 105 minutes, approximately 2000 cycles have been accumulated up
to now. The main focus of this test rig lays in the component behavior under primary and
additional secondary stresses, thermal fatigue behavior with temperature cycles between 400 C
and 700 C, stress-strain situation in pipes, bends, headers and T-pieces for lifetime analyses,
operation of high temperature control and shut off valves.
feedwater
feedwater
feedwater

B 3 steam

da = 219,1 x 50 mm

Pipe + bend + weld


Prim. Stresses from
inner pressure
Sec. cycl. Stresses from
hangers
Combustion chamber
1200C
Superheater Pipes

Pipe + weld
Prim. Stress from inner pressure
Therm. gradient from simul. start-up and shut-down
(injection T 300 C steam)

Static load

725C

Thick walled pipe


Cyclic Load

Bypass

hangers, insulation

PRDS

HT-conditioning valves
Design, reliable
longtime operation
Mass flow control
Set to 725 C/162 bar
KZ steam
< 400 C

Bypass with PRDS


PRDS functionality
530C, 170 bar

Flow Chart HWT II


Steam extraction
530 C

2,5 kg/s
from B 4

Figure 4: HWT 2 test loop with thickwalled pipes and bends, T-piece and header made from
Alloy 617 and Alloy 263.
Manufacturing and Qualification
Within the MARCKO 700 and DE4 projects manufacturing processes for tubes, pipes, T-pieces
and headers have been established. In DE 4 the products and weldments shown in Table 2 have
been manufactured and tested. In HWT 2 project numerous welds for the test rig and for
accompanying tests including dissimilar welds (A617/A263) have been produced.
Beside basic pedigree tests, Charpy-V tests in virgin and aged (up to 30,000 h) state have been
performed, Fig. 4. In the virgin state of the pipe specimen taken from the weldmetal show the
lowest values, for all specimens an increase with temperature can be shown, ageing results in a
reduction of impact energy, however this is not significant for higher temperatures. For the tubes
impact energy is highest in hardened state, further reduction after 30,000 h ageing is observed.
Additional results from qualification tests, e.g. tensile tests for A617 and A263 are given in [11]
and [12].

158

Table 2: Products and welds manufactured in DE4 and HWT 2

100

State: @800C/4h
State: @800C/4h+700C/3000h
State: @800C/4h+725C/3000h
State: LW@800C/4h+725C/30000h

90
80

140

70
60
50
40
30
20
10
0

200

300

400

Alloy263,
263.Rohr
Pipe380x60mm,
TIG weld
Alloy
380x60mm,autom.
Rundnaht,
vWIG-Orbital

80
60
40
20 Grundwerkstoff,

Remark: Mini-specimen 3x4x27 mm according to DIN50115-1991

100

500

HWT 2

100

Alloy 263, Tube 38 x 6,3 mm

Project
DE 4

120

Kerbschlagarbeit / J

Charpy-V-notch toughness J/cm

Material
Product
Welds
Weld material
Alloy 617
Forging
TIG (SA)
Similar A617
Alloy 263
Forging
TIG (A) and SMAW (A)
Similar A263
Alloy 263
Seamless Pipe
TIG (SA, A), SMAW (SA, A)
Similar A263
Alloy 263
Tubes
TIG (SA and A)
Similar A263
Alloy 263
Centrifugal Cast
TIG (SA)
Similar A263
Alloy 263
Induction Bend
Alloy 263
Cold Bend (Tube)
Alloy 263
Forging
TIG, SMAW
A617
Alloy 617
Forging
TIG, SMAW
A617
A617/A263
Forging
Dissimilar Welds, TIG
A617
A617/A263
Induction Bends
A617/A263
Header
Tubes: A617, A263, A740
A617
Welded in state: SA.solution annealed, A solution annealed and aged (800 C)

600

700

800

AZ;
700C/3kh;
Schweiverbindung, Kerbe WEZ,
AZ,
Schweiverbindung, Kerbe SG,
AZ,

50

100

150

725C/3kh;
700C/10kh;
700C/10kh;

600

700

700C/10kh;

800

Temperatur / C

Temperature / C

a)
b)
Figure 5: Results from Charpy V tests for tubes (a), with miniaturized specimens) and pipes (b) in
the virgin state and after 3000 and 30,000 h ageing.
Creep Behavior of Base Materials and Welds
One aim of the investigations in DE4 is to provide design data. For Alloy 617 within the project a
reevaluation of existing and new data has been made [13]. The results are shown in Table 3. The
values are in agreement with findings from previous projects and are much higher than the valued
given in the recent standard (DIN-EN-10302). So far this evaluation could be confirmed with all
tests. Assessment of weld strength of similar welds (Alloy 617) result in an overall weld strength
factor of 0.85. These results are confirmed by tests carried out within the recent projects. In Fig. 5
results of creep rupture test for different Alloy 617 melts and various weldments are shown. Most
base material results are in the upper part of the scatterband of the new evaluation.
So far the test results and running tests of TIG-welds with 980C postweld heat treatment are also
in the upper scatterband of the base material evaluation.
Figure 6 shows a summary of results at 700C for Alloy 263. Results are well above the Alloy
617 mean line and higher than the meanlines given by manufacturers. The scatterband of the
standard DIN-EN-10302 which is based on data for thin products is reached by the base metal
results. Only small reduction is visible for the investigated weldments.
159

Rupture stress / MPa

Table 3:New evaluation of creep rupture properties for Alloy 617 [13]

500
450 Alloy 617 mod.; T=700C
400
350
300
1

250
200
150
100

Location of
fracture:
1 = BM
3 = WM
4 = HAZ

33

33

3
3

3 2x3
1
Weld metal and welds:
3
pure WM, SMAW
pure WM, TIG
3 3
pure WM, TIG, running Forged Material Test (Pipe):
weld, TIG orbital (+980C/3h)
Header Material Test (Pipe):
weld, TIG orbital (+980C/3h), running
weld, SMAW
Base Material Test (Forged) (IfWD):
weld, SMAW, running
BM (Pipe 360 x 80)
weld, TIG orbital
BM (Pipe 360 x 80), running
Tube Material Test:
weld TIG, Tube
Forged Material Test (Pipe):
weld TIG, Tube, running
weld, TIG orbital (+980C/3h)
Base Material Test (SZMF):
weld, TIG orbital (+980C/3h), running
BM (Tube)
Base Material Test (Forged) (IfWD):
BM (Pipe 460 x 80)
BM (Pipe 360 x 80)
BM (Pipe 460 x 80), running
BM (Pipe 360 x 80), running

10

10

Scatterband (A130) :
Average
+-20 %
New scatterband
COORETEC DE4:
Average
+-20 %

10

10

10

Rupture time / h
Figure 6: Creep rupture strength of Alloy617(mod) and its weldments.
600

Scatter band (800C/8h):


DIN EN 10302: 2008
Average
+-20 %
Special Metals
Average
VDM
Average

500

Rupture stress / MPa

400
300

Alloy 263; T=700C

Location of
fracture:
1 = BM
3 = WM
4 = HAZ

1
1
1
1
1

200
Base material (IfWD) 800C/4h:
BM (forged, Pipe 360 x 60)
BM (forged, Pipe 360 x 60), running
BM (Tube 38x6,3)
BM (Tube 38x6,3), running

100

Welds (MPAS) 800C/4h:


weld TIG (Tube 38x6,3: L)
weld TIG (Tube 38x6,3: L), running
weld TIG (Tube 38x6,3: LW)
weld TIG (Tube 38x6,3: LW), running
weld TIG orbital (seamless pipe 360 x 60: L)
weld TIG orbital (seamless pipe 360 x 60: L), running

10

10

10

Scatter band Alloy 617:


COORETEC DE4 2010:
Average
+-20 %

10

10

Rupture time / h

Figure 7: Creep rupture strength of Alloy 263 and its weldments


160

Creep curves are obtained for the materials in order to describe numerical models as shown in
Fig. 7 for the seamless pipe made of Alloy 263. Using such data it was possible to approximate
parameters for creep models, e.g. the incremental Graham Walles creep law and a Chaboche type
model (CNOW) [14]. In order to qualify the pipe weldments introduced into the test rigs, same
additional welds (TIG and SMAW) were manufactured. From these welds large scale specimens
were taken (cross section appr. 40x40 mm2). Creep strain measured by high temperature strain
gauges over the complete weld zone are shown in Fig. 8. All welds were manufactured using
Alloy 617 weld material. It is visible that similar Alloy 263 welds and dissimilar welds show
smaller creep strains than Alloy 617 welds.
SA: welded in solution annealed state, afterwards hardened (800C/4h)

Alloy 263, seamless pipe 360 x 60 mm, T=700C


1

Strain p / %

10

BM(forged), 350 MPa


BM(forged), 310 MPa
BM(forged), 270 MPa
BM(forged), 230 MPa

10

-1

10

-2

10

Weld (seamless pipe, SA), 390 MPa


Weld (seamless pipe, SA), 340 MPa
Weld (seamless pipe, SA), 280 MPa
Weld (seamless pipe, SA), 234 MPa

-3

10

10

100

1000

10000

100000

Time / h

Creep strain / %

Figure 8: Creep curve tests on welded samples of a seamless pipe compared to Alloy 263 base
material

Dissimilar weld

time / h

Figure 9: Measured Creep creep strain of large scale specimens with a cross section of
40x40mm2 taken from pipe welds used in HWT2 project

161

Fatigue and Creep Fatigue Behavior


Fatigue and creep fatigue tests are performed for thickwalled pipes and forgings as well as for
tubes within the DE4 and HWT projects for Alloy 617 and Alloy 263. Exemplarily S-N curves
for Alloy 617 are shown in Fig. 9. The influence of hold time on the crack initiation time is
visible. Tests from different melts are in good agreement. The influence of temperature is shown
in Fig. 9b. A comparison of the both Nickel base alloys under investigation is shown in [12].

Figure 10: S-N-curves for Alloy 617


Crack Initiation and Crack Resistance
Since the ductility of the Nickel based alloys is less than so far used ferritic and martensitic steels
the question of sufficient crack resistance arises. Exclusion of catastrophic failure and the
respective assessments need fracture mechanics characteristics for base and weld metals. Figure
10 shows static crack resistance curves for Alloy 263 after hardening and for comparison after
ageing at 700C and 725C for 3000 h. Crack resistance at room temperature is somewhat
reduced after ageing. At 700C, the material after hardening has the lowest crack resistance
which is significantly increased after ageing at 700C and more after ageing at 725C. It has to be
noted, that even the smalles crack resistance curve demonstrates sufficiently high static crack
initiation.
J-R Curve, Alloy 263
1000

J-integral (N/mm)

900
800

R1, T=23C (State: 800C/4h)


R3, T=23C (State: 800C/4h + 700C/3000 h)
R5, T=23C (State: 800C/4h + 725C/3000 h)
R2, T=700C, T=23C (State: 800C/4h)
R4, T=700C (State: 800C/4h + 700C/3000 h
R6, T=700C (State: 800C/4h + 725C/3000 h

700
600
500
400
300
200
100
0
0,00

0,50

1,00

1,50

2,00

2,50

3,00

Crack extension Da (mm)

Figure 11: J-R curves for Alloy 263 at different temperatures and after ageing

162

Cyclic crack growth behavior is shown in Fig. 11. The comparison of both Nickel alloys gives
somewhat higher crack growth rate for Alloy 617. Creep fatigue crack growth data and creep
crack growth data are discussed and compared in more detail in [12].
1,E-01

ASME XI, ferritic steels, T=23C


Alloy 263, threshold, notch: HAZ, T=23C
Alloy 263, threshold, BM, T=23C
Alloy 263, notch: HAZ, T=23C
Alloy 263, BM, T=23C
Alloy 263, notch: HAZ, T=700C
Alloy 617, BM, T=23C
Alloy 617, BM, T=700C
Alloy 617, notch: HAZ, T=700C
Alloy 617, notch: WM, T=23C

da/dN / mm/LW

1,E-02
1,E-03

R=0,1

1,E-04
1,E-05
1,E-06
1,E-07

10

DKI / MPam

100

Figure 12: Cyclic crack growth for Alloy 617 and Alloy 263 at RT and 700C.
COMPONENT TESTING IN TEST LOOPS INSTALLED IN A POWER PLANT
Tube Test Loop (725 HWT 1)
In the tube test loops superheater tubes of various materials (Austenite and Nickel based alloys
where tested. Oxidation and corrosion behavior in the boiler are compared to findings in the
project Cooretec DE1 where tube samples and oxidation plates are investigated as well. Such
plates are also used in the turbine material section of the loop. In the outside loop ferritic and
austenitic materials are subjected to a temperature of about 630C and high Chrome austentitics
and Nickel based alloys are subjected to 725C. The test samples of the materials have a reduced
cross section in order to provoke damage within shorter durations. The investigations on the
oxidation behavior are now ongoing after finishing the first phase of the project. Some new
materials and some of the previously tested samples are now included in the test loops which is
under further operation now.
Test Loop for thick-walled components (HWT 2)
In HWT II a thick walled test rig with components made from the nickel based alloys A617 and
A263 is operated at a maximum temperature of 725C in Mannheim power plant GKM. The test
rig consists of a stationary and a transient part. With the exception of the operational startups and
shutdowns the stationary part is operated with steam at a temperature of 725C and an inner
pressure of 170 bar to investigate the creep behavior. In the transient part of the test rig the steam
at 725C is cooled down to 520C with cooler steam and after this to 400C with water. After
cooling the temperature increases from 400C to 725C, where a short hold time of almost 75
minutes takes place before the cooling starts again. One cycle has a duration of almost 105
minutes. Operation reached approximately 2000 cycles up to now which is close to the aim
formulated in the project. The intention of this operation mode is to simulate the fast and many
startups and shutdowns in future power plants and to investigate the thermal fatigue of the used
nickel base alloys. In the stationary part the main focus lies on a pipe bend made from Alloy 617
that is loaded with internal pressure and loads resulting from the weight and the hanger
construction of the pipework. Furthermore the bend is pre-tensioned with a bar to produce
stresses.

163

Numerical Analyses for thickwalled components


To investigate the creep behavior in the pipe bend different numerical simulations were perfomed
using a model of the full pipe work (left) and the pipe bend only (right). The detailed model of
the full pipework including all applied hanger constructions is created with the aim to determine
the creep behavior in the pipe bend. On the other hand a model of only a small area of the
pipework with an artificial fixation point is created to investigate the influences of the boundary
conditions. For both models an inner pressure of 170 bar is applied. For the reduced model the
boundary conditions (applied forces and moments at the cuts) are a result of an elastic calculation
of the pipework. Furthermore two simulations for each models under strain controlled conditions
(except internal pressure) and stress controlled conditions were carried out. Figure 13 shows a
comparison of Mises stress and equivalent strain for the different methods. The results show
significant differenced between both the variation of model and variation of boundary conditions.
It has to be expected that the strain controlled case and the use of the full model is most realistic.

Figure 13: Different models to simulate the pipe bend behavior under static creep load

Figure 14: Mises stress and equivalent strain prediction obtained by an incremental creep law
From the section of the test loop which is subjected to thermocycles an analyses for the header is
shown in the following. The header and its model with different bore hole radii is shown in Fig.
14. The plots in Fig. 15 show the transient of steam temperature and mass flow dependent heat
transfer coefficient. The inner pressure is steady during one cycle and amounts to 170 bar.
Three simulations with an elastic-plastic material model, an incremental creep model (GrahamWalles type) and a complex Chaboche model were performed. In the elastic-plastic simulation
the cyclic flow curves from low-cycle fatigue (LCF) tests were used. With this material model it
is only possible to describe the isotropic hardening of the material. For the second simulation
164

using an incremental creep model the Graham-Walles creep law was used. This model is fitted
with uniaxial creep tests and it does not consider deformation due to cyclic plasticity. The
Chaboche Nouailhas Ohno Wang (CNOW) model was used for the third calculation. This model
is able to describe the isotropic and kinematic hardening of the material. The model contains two
different flow rules. The first one is used to describe the flow at high stresses and second one at
low stresses. The results are shown in Fig. 15.

2 capacitive strain
gauges

Figure 15 Header included in the part of the test loop which is subjected to thermocyclic loading
and geometrical model representing different radii for the boreholes of the header

Figure 16: Mises stress, first principal stress and principal strain obtained by a simulation of
the thermal cycle using the Chaboche type CNOW material model
From this simulation a strain range can be derived. Using a Manson-Coffin approximation of the
S-N-curve of the materials as a result for different radii and different material models the number
of cycles to crack initiation can be estimated, Table 4.
Table 4: Estimation of the number of cycles to crack initiation in the header for different radii
and applying different material models
Manson-Coffin
a
N

%
-

elastic-plastic
R3
R10
0,33
5435

0,26
14731

165

GWD
R3

R10

0,33
5435

0,28
10443

CNOW
R3
R10
0,38
3243

0,31
6995

SUMMARY AND CONCLUSIONS


Investigations aimed to provide data and tools for design and lifetime assessments for
components of the A-USC plant made from Nickel based alloys have been described. After
testing within a couple of projects a reasonable data base to characterize the materials under
investigation could be established. This database comprises pedigree data, creep strain and creep
strength data, information on cyclic behavior and failure as well as fracture mechanics data.
Constitutive equations for the simulation of static creep and thermal cyclic load could be
established and the parameters have been approximated. Test loops with tubes and thickwalled
boiler components have been operated successfully giving experience on the material behavior in
the plant and the functionality of components. Using the constitutive equations established
numerical simulations have been carried out with the aim to predict stresses and strain and
lifetime of the component. A prediction for strain accumulation in the statically loaded pipe bend
as well as the prediction of cycles to failure was obtained. These results are basis for future
comparisons with experimental findings and damage models e.g. the DTMF approach used in
[15]. As a conclusion it could be stated, that with the described projects valuable data and tools
for the assessment of components made of Nickel based alloys could be established.
ACKNOWLEDGEMENTS
The projects are funded by the Federal Ministry of Economy (BMWi) under grant No 0327705Y
(DE4) and 03ET2017 (HWT2) and an industrial consortium of material and plant manufacturers
and utilities. The authors wish to thank the research institutes involved (IfW Darmstadt
Fraunhofer IWM Freiburg, the steering committees and companies for good co-operation.
REFERENCES
[1]

[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]

R. U. Husemann, A. Helmrich, J. Heinemann, A. Klenk and K. Maile. Applicability of Ni-based welding


consumables for boiler tubes and piping in the temperature range up to 720 C. 4th Int. Conf. On Advances in
Materials Technology for Fossil Power Plants, Hilton Head, 2004.
Quirong Chen et al. Materials Qualification for 700 C Power Plants. Fifth International Conference on Advances in
Materials Technology for Fossil Power Power Plants, October 3 - 5, 2007, Marco Island, USA
Schmidt, K. and E. Roos: Qualifying Materials for the 700/720 C Power Plant (Part I: Materials for the 700/720 C
Power Plant, VGB Power Tech, VGB Power Tech 1/2, 2012, pp. 74-84.
Schmidt, K., A. Klenk and E. Roos: Qualifying Materials for the 700/720 C Power Plant Part II: Oxidation
behaviour, Component Tests and Simulation, VGB Power Tech 3, 2012, pp. 97-103.
Cooretec-DE4 Research project Investigations on the longterm strength and deformation behavior of Nickel
based alloys, BMWi project , Final Report in preparation
K. Metzger, K. H. Czychon, K. Maile, A. Klenk, A. Helmrich, Q. Chen. GKM Test Rig: Investigation of the long
term operation behavior of tubes and forgings made of alloys for future high efficient power plants. 6th International
EPRI Conference on Advances in Materials Technology for Fossil Power Plants, 31.8-3.9.2007, Santa Fe
K. Metzger, S. Zickler and K. Maile. Experience with the Operation of a test-rig for 700 C power plant material.
6th International Conference on Creep, Fatigue and Creep-Fatigue Interaction, January 22 - 24, 2012, Chennai, India
Zickler, S., S.Wagner, K. Maile, K. Metzger and K.H. Czychon: High-Temperature Materials Test Rig: Materials
for Future High-Efficient Power Plants, 38th MPA Seminar, 7-8 October 2013, Stuttgart
Maile, K., S. Zickler and K. Metzger: Qualification of Structural materials for the Advanced 700C Fossil Fired
Power Plant experience gained in the GKM Field Test Loops, BALTICA IX, 2013, Helsinki-Stockholm
K. Metzger: Commissioning of the Thick walled Pipe Test Rig for 725C Power Plants, 39 th MPA Seminar, 8-9
October 2013, Stuttgart
Schraven, P.: Qualifcation and Experience on A-617(mod) and C263 Boiler Tubes, 7th Int. Conf. on Advances in
Materials Technology for Fossil Power Plants, 22-25 Oct 2013, Waikoloa Village, HI
Mller, F., A. Scholz and M. Oechsner: Selection of Super Alloys for A-USC Power Plants under consideration of
creep crack growth behavior, 7th Int. Conf. on Advances in Materials Technology for Fossil Power Plants, 22-25
Oct 2013, Waikoloa Village, HI
Schubert, J.: Neuauswertung des Zeitstandverhaltens von lsungsgeglhtem Alloy 617 samt seiner
Schweiverbindung im Temperaturbereich 550 bis 1000C, 33. Vortragsveranstaltung des VDEH, Dsseldorf,
November 2010
Hggenberg, D. and A. Klenk: Numerical Assessment of Secondary Stresses in Pipe Bends and Thermal Fatigue in
Headers, 39th MPA Seminar, 8-9 October 2013, Stuttgart.
Oesterlin, H and G. Maier: Numerical Assessment of Power Plant Components by Use of Mechanism-Based
Material Models for Deformation and Lifetime, 39th MPA Seminar, 8-9 October 2013, Stuttgart

166

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

QUALIFICATION AND EXPERIENCE ON A 617 AND C 263 BOILER


TUBES
Dipl.- Ing. Patrik Schraven, Muelheim an der Ruhr, Germany

ABSTRACT #

19

SMST is producing Ni alloy Boiler tubes since more than 10 years with application in
several test loops and R&D programs. This paper will give an overview about the
experience with the common grades A617 as well as C263 plus some additional
information on the new developed austenitic material Power Austenite MoW.

INTRODUCTION
As fuel resources are shrinking and CO 2 emissions are getting more and more
important, global energy policy is adapting to these new challenges. The trade with CO 2
certificates is one of the driving forces to investigate new technologies in PC power plant
technology.
In 2008, European utilities decided to perform a study on a 500 MW coal fired power
plant with net efficiencies of > 50%. The design studies for this project showed that live
steam temperatures of 1290 F are necessary to reach this high efficiency. Latest
austenitic stainless steels have proven to be capable up to 1150F steam temperature
but for the highest temperature range, Nickel Base alloys are necessary to withstand the
tough thermal and mechanical conditions.

EXPERIENCE
During the last 10 years Salzgitter Mannesmann Stainless Tubes has been fully
committed to this technology development and has gained comprehensive experience in
alloys which include DMV617 mod. & DMV263. The most promising candidate is alloy
DMV617mod.

167

In scientific terms the following major steps were taken:


DMV 617 mod
-

Modification and qualification of alloy 617 for the application as boiler tube
Results of several test loops running at 1200 1337 F (650 -725C)
Significant improvement of creep rupture strength of A 617
Prove of creep rupture strength with test durations of up to 200.000h (25
years)

DMV 263
-

Long term creep rupture testing of C 263 in aged and solution annealed
condition
Experience in C-263 boiler tubes fabrication

POWER AUSTENITE MoW


-

New developed austenitic stainless steel with a unique chemical composition


(Table 1) and strengthening concept
Excellent creep rupture properties around 1300 F due to a combination of
small sigma phase formation within the grains and a grain boundary
strengthening by M 23 C 6 and Nb(C,N) (Figure 1)
Superior fire side corrosion and oxidation properties due to increased Cr
content (Figure 2) compared to 22Cr25Ni materials

C
Si
Cr
Ni
Mo
Nb
B
0,020,1-2,0
25-33
22-38
1-6
0,4-1,5
10-100
0,15
ppm
Table 1: Chemical composition of POWER AUSTENITE MoW [weight %]

168

N
0,01-0,2

M23C6
Nb(C,N)
phase
grain boundary
Figure 1: Structure of POWER AUSTENITE MoW

POWER AUSTENITE MoW

22Cr25Ni alloy

Figure 2: Microstructure of samples after 3,000h flue gas corrosion test under salt
In parallel to the scientific development, Salzgitter Mannesmann Stainless Tubes works
in partnership with the European utilities as well as boiler works to prove our capability of
producing more than 1000t of Ni base boiler tubes in accordance with a rigid project
schedule.

169

Highlights of this development process include:


-

Fully qualified prematerial sources


Experience with ESR / VAR billets
Deformation and heat treatment design tool to optimise production processes
Two completely redundant process routes in case of mayor problems during
project execution
Advanced time schedule management

CONCLUSIONS
Salzgitter Mannesmann Stainless Tubes is able today to supply the requested stainless
steel and Nickel alloy boiler tubes for a 1300F PC power boiler in accordance with the
latest and greatest industry standards.
A new tube material with good fire side corrosion resistance and excellent creep
properties was developed and will be further qualified.

170

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

Material advancement used for 700C A-USC-PP in


China
Zhengdong LIUHansheng BAOGang YANG
Central Iron and Steel Research Institute (CISRI), Beijing, China, 100081
Songqian XU, Qijiang WANG
BaoSteel Co., Ltd., Shanghai, China, 201900
Yujun YANG
Fushun Special Steel Co., Ltd., Liaoning, China, 113001
ABSTRACT
This paper briefly introduces the state-of-the-art of the research and development of candidate
heat resistant materials used for the manufacturing of 700C advanced ultra-super-critical (AUSC) fossil fuel power plants (PP) in china, especially, focus on the impressive progress in the
past three years. The detailed advancements (technical exploration and industrial investigation) of
candidate materials spectra for the boiler system of A-USC PP will be presented in the current
paper, including novel ferritic heat resistant steels, advanced austenitic heat resistant steels, FeNi-based alloys and Ni-based alloys, which serve and cover the steam temperature scope from
600C to 720C. Some newly available data associated with above materials will be released.
INTRODUCTION
In recent years, the construction of advanced ultra-supercritical (A-USC) fossil-fired power
generation plant with higher efficiency has been accelerated by the demand of reducing CO2
emission for the protection of global environment. The story is especially true in China in the past
years. Since the first 600C USC power plant was successfully launched in the November of
2006, more than 80GW electric capacity of 600C USC units has been commercially installed in
china mainland by the end of 2011 and more 600C USC units will be constructed in the near
future. The steam parameter evolution of Chinese USC power plants was schematically plotted in
Figure 1.

Figure 1 Steam parameter evolution of Chinese USC power plants


On July 23, 2010, Chinese government officially announced a national program to build a
demo 700C A-USC power plant in the near future, probably starting from the year of 2018,
171

depending the availability and maturity of heat resistant steels and alloys. To fulfill such a goal, a
national consortium has been directly organized by the National Energy Administration (NEA) of
China, including seventeen research facilities, manufacturers and power plant operators, which all
are the leading player in their own sectors. Well known, the critical challenge of 700C A-USC
technology lies in the R&D of high-temperature materials, not only for boiler system but also for
turbine system. Due to the space limitation, the current paper will mainly address on the recent
boiler materials advancement in china, which may be hopefully used for Chinese 700C A-USC
demo plant.
BOILER MATERIALS SPECTRA FOR 700 A-USC PP
Basically, those mature materials successfully used in 600C USC power plant will still be
applied in 700C A-USC power plant, in where the working temperature is below 600C. When
the steam temperature increases from 600C to 700C, different novel materials are necessarily
required for various temperature scopes. According to published literature and our knowledge,
the materials R&D sub-committee of Chinese consortium for 700C A-USC power plant has
discussed, evaluated and summarized the candidate materials spectra used for 700C boiler of AUSC power plant in China, as schematically shown in Figure 2.

Figure 2 Candidate materials spectra used for 700C boiler of A-USC PP in china
For water cooling wall tubing, T23 and T91 are considered to be base-steel, on which some
modification for the improvement of on-site welding performance during plant manufacturing
and following maintain has to be carried out. G102 may be the alternative choice for the
replacement of T23 and T91, upon certain modification and improvement which are under
investigation at CISRI and BaoSteel. Most recently, T92 and G115 were selected as one of
candidate materials by Chinese boiler designers to make tube at superior temperature scope of the
water cooling wall tubing system, due to its outstanding creep strength, which make it possible
that all tubes from lower temperature to higher temperature of the system are of similar outside
diameter and wall thickness. Alloy 617 is not necessary to be a candidate material for the water
cooling wall system, unless the hidden weakness of T92 and G115 for such an application occurs
during manufacturing and in service in the future. Welding is a key concern for the making of
water cooling wall tubing.
For the tube-making of super-heat (SH) and reheat (RH), the candidate materials spectra are
T91, T92, NF709R, Sanicro25 (GH2984G), and Inconel740H equivalent (or an Inconel740HM
alloy), with the increase of steam temperature from 580C to 700C. GH2984 was developed by
Institute for Metals Research of Chinese Academy of Sciences some years ago, and 2984G will
be an upgraded iron-nickel based super-alloy on the basis of GH2984, which is supposed to be

172

used to make tube or pipe up to 680-700C. Thus, GH2984G and Sanicro25 are competitive
materials.
For the pipe/header-making, the candidate materials spectra are P91, P92, G115/G112,
(GH2984G), and CCA617 equivalent alloy, with the increase of steam temperature from 580C
to 700C. G115 was originally developed and patented by the Central Iron and Steel Research
Institute (CISRI) of China in the past years, which is proposed to be used to make pipe or header
up to 650C,although it was selected to be a candidate for water cooling wall tube.
Table 1 The time-table of R&D of materials used for 700 A-USC PP in china

The milestone and time-table of research and development of candidate materials used for
700C A-USC power plant in china were included in Table 2. Hopefully, the final products of
aforementioned candidate materials will be commercially available within five more years from
now on, subject to the determination and certainty of designing requirements. The typical
chemistry of candidate materials used for 700C A-USC power plant was listed in Table 2 and
the typical mechanical property of candidate materials used for 700C A-USC power plant was
listed in Table 3. The desirable creep rupture strength of candidate materials is around 100MPa
under its service temperature for 100,000 hours.
Table 2 Typical chemistry of candidate materials used for 700 A-USC PP

Alloy

Cr

Mo

G102
G115
NF709R
Sanicro25
GH2984
CCA617
In740H

0.12
0.08
0.08
0.08
0.06
0.06
0.03

2
9
21
22.5
19
22
25

0.6

0.4
3.0

Co
3.0

3.6

1.5

1.5
2.0
9.0
0.5

12
20

Nb
0.05
0.25
0.50
1.00

Al

0.35
0.20

Ti

Cu

0.13

0.005
0.015

1.0
3.0

0.4
1.0
1.35

1.50

1.0
0.4
1.35

0.003

Ni

Others

25
25
43
55
48

Fe
Others
Others
Others
33Fe
Others
Others

Table 3 Typical mechanical property of candidate materials used for 700 A-USC PP

Alloy
G102
G115
NF709R
Sanicro25
GH2984*
CCA617
In740H

Room temperature
Rp0.2 Rm
A
Z
345
623
330
366
686
330
705

630
763
700
760
1107
750
1110

>18
23
44.5
48
26.6
66
36

>40
72
74.5
68
46.6
70
42

Rp0.2
325
171
209
539
188
630

* Data based on published literature


173

Elevated temperature
Rm
A
Z
T,
375
454
513
745
555
850

30
40
44
34.4
80
23

83
65
57
52.3
56
28

650
700
700
700
700
750

G115 FERRITIC HEAT RESISTANT STEEL


In recent years, the Central Iron and Steel Research Institute (CISRI) of China successfully
developed G115 and G112 ferritic heat resistant steels, which have been industrially made and
qualified, showing the imposing potential to manufacture pipe and steam container in the
temperature scope from 620 to 650C.[1] Among them, G115 is of higher creep rupture strength at
650C and G112 is of better steam oxide resistance at 620-650C.
G115 tube and pipes have been industrially made at BaoSteel of China since 2008 by various
processing routines. The specimens investigated in the current paper were from hot-extruded
G115 pipe, with the dimension of 254253500 mm.
Table 4 Mechanical properties of G115 steel pipe
Temp Rm, Rp0.2, A,% Z,%
Akv2,J
HB
MPa
MPa
R.T.
763
623
23
72.3 115 112 210
338
26.5
85
650 380
The mechanical properties at room temperature and 650 of the G115 pipe were listed in
Table 4. Obviously, the steel is of reasonable impact energy, implying that the steel is of potential
to make heavy thickness pipes. The impact energy is higher than 112J at room temperature. The
photo-graphics of G115 pipe after aging at 650 up to 8000 hours were shown in Figure 3, in
which, it is clearly that the martensitic lath width keeps stable, thinner than 425nm. However,
when the aging temperature is 700, the martensitic lath width grows much faster, as indicated
in Figure 4. The 650 should be the upper limitation of the steel.

Figure 3 TEM photos of G115 steel after aging at 650

174

Figure 4 Evolution of G115 martensitic lath width during aging

Figure 5 Evolution of M23C6 and Laves phases of G115 steel during aging

Figure 6 High temperature strength of G115 steel during aging


The major precipitation of G115 steel during 650-700 aging includes MX, Cu-rich
precipitate, M23C6 and Laves phase. Among them, MX and Cu-rich particles are very fine,
basically in nanometer-scale, and very stable during aging. From the viewpoint of precipitation
hardening, the deterioration of creep strength of the steel results from the coarsening and
aggregation of M23C6 and Laves phase particles. The evolution of M23C6 and Laves phase during
aging was shown in Figure 5. The velocity of coarsening of M23C6 and Laves phase under 650
175

aging are much lower than that of 700 aging. The size of M23C6 is less than 170 nm and Laves
phase is about 200 nm under 650 aging up to 8000 hours. Correspondingly, the yield strength
and tensile strength of G115 at 650 and 700 after 650 aging were drawn in Figure 6. The
mechanical properties of G115 under 650 aging are very stable.

Figure 7 Creep rupture strength of G115 steel pipe

Figure 8 Oxide resistance of G115 steel pipe


The curves of current available creep test results of the G115 steel pipe and other G115 trial
samples under 650 were plotted in Figure 7. Longer time creep tests of G115 are undergoing at
CISRI. In fact, the G115 is of excellent creep strength among all commercially available ferritic
heat resistant steels worldwide so far, much higher than that of P92 steel. Meanwhile, G115 has a
reasonable oxidation resistance as shown in Figure 8, comparing to P92 and T122 steels. Various
welding trial and verification of G115 steel are also under the way at CISRI and Chinese boiler
makers.
CN617 NICKEL-BASED ALLOY
Although alloy 617 was developed as early as in 1960s, it was frequently addressed being a
candidate material for 700 A-USC power plant in Europe since 1990s. Alloy 617 was not
suitable to make superheat and reheat tubes due to the poor ash-corrosion resistance resulting
from the addition of about 9% molybdenum, however, the alloy was considered to be an
important candidate for the manufacturing of pipe and header. Being a solid solution hardening
dominated alloy, alloy 617 is more difficult to be hot-deformed, comparing to alloy 740. The
176

operational temperature window is very narrow and the flow stress is higher during pipe-making.
Further, relatively lower creep strength is another weakness of alloy 617. Modified 617 (or socalled 617B) alloy has been developed at ThyssenKrupp VDM, mainly by the boron addition of
10-50ppm, through which the creep strength is evidently increased. Big 617B alloy pipes were
successfully made and had been qualified in various European test loops. Cracks in the base
metal of an injection cooler, made of 617B alloy, in Scholven power plant were found and the
cracks was analyzed to be stress relaxation cracks. Proposed solution to eliminate the cracks is to
apply stress-relief annealing at 980for 3 hours.[2] In spite of aforementioned problem, alloy
617B is still one of most possible candidate materials for the pipe and header making of 700 AUSC power plant, because of its accumulated creep data and rich engineering experiences. Alloy
263 and 740H are of higher creep strength, but both alloys are in the early stage in the course of
feasibility verification used to make pipes and headers of 700 A-USC power plant.

Figure 9 CN617 alloy ingot after VIM+VAR


(Courtesy of CISRI and Fushun Special Steel, 2013)
In recent years, CISRI systematically and experimentally investigated the optimization of
chemistry, hot-deformation and heat treatment of 617B alloy. Optimized chemical composition
and processing parameters were proposed against the available results, which led to an industrial
trial of the novel alloy, tentatively termed as CN617, at the FuShun Special Steel Company in
China. A 6-ton CN617 alloy ingot was made by VIM and VAR melting routines with the
dimension of 660 mm (outside diameter) and 1940 mm in length, as shown in Figure 9. The ingot
had been transferred to Inner Mongolia Heavy Industries Company of China, in where the ingot
will be hot-extruded into a big pipe with the dimension of 460804000 mm in the coming
July 2013, with the help of 360MN press.
INCONEL740HM ALLOY
It was well known that the brittle eta (, Ni3Ti-type) phase precipitates along the grain
boundaries of Inconel740 during 750aging. After aging at 750 for 1000 hours, small amount
of eta phase was observed near the grain boundaries. After 5000 hours, not only eta phase
initiates at grain boundaries but also a significant amount of eta phase forms inside the grains as
Widmanstatten pattern structure.
The eta phase is rich Ti-containing, which partly attributes to the degradation of creep
strength. To improve the microstructure of Inconel740, the researchers from University of
Sciences and Technology Beijing (USTB) and Special Metals Corporation (SMC) modified the
content of critical elements (i.e. Nb, Al, Ti, Si, etc) and named the modified alloy as

177

Inconel740H. After the modification, it was reported that there is no acicular eta phase in the
microstructure of Inconel740H after 700-750 aging.
With the light of the modification of USTB and SMC, CISRI melted some heats of
Inconel740H in Laboratory for verification. Among them, one heat of Inconel740H contained
1.98%Nb, 1.08%Al, 1.64%Ti and 0.050%Si. As seen in Figure 10, acicular eta phase clearly
exists along the grain boundaries of the specimens after 5000-8000 hours under both 750 and
800, which implies that the microstructure of Inconel740H should be further modified and
improved.[3]

Figure 10 TEM photograph of Inconel740H during aging up to 8000hours

Figure 11 TEM photograph of Inconel740HM during 750 aging up to 8000hours


When further decreasing the content of Titanium to 1.38%, meanwhile with 1.48% Niobium
and 1.47% Aluminum, together with some other modification, the microstructure of the alloy
after aging at 750 up to 8000 hours was shown in Figure 11. Clearly, no eta phase precipitates,
the carbides along the grain boundaries of the specimens keep in very stable, the size of gamma
prime increases. This indicates a much better microstructure (Figure 11) comparing to that in
Figure 10. The improved alloy was marked as Inconel740HM. More experimental findings on
this interested and important alloy will be published soon.

178

CONCLUSIONS
(1) Candidate materials spectra used for 700 boiler of A-USC demo Power Plant in china was
carefully discussed and summarized. The research and development in laboratory and
industrial trials of all tubes and/or pipes associated with the selected materials spectra are
under their way now.
(2) G115 has been successfully developed at CISRI and BaoSteel and it is an imposing material
for the making of boiler pipes at the temperature scope from 620 to 650 and the water
wall tubing of 700 A-USC power plants, due to its excellent creep rupture strength and
reasonable oxide resistance as well as its economic competitiveness.
(3) Although Inconel740H has been recognized as an excellent candidate alloy for 700 boiler,
the further optimization of chemistry and processing techniques is still necessary and
possible, based on existed and characterized weakness of the imposing alloy.
ACKNOWLEDGMENTS
The authors are very grateful for the continuously financial support from the Ministry of
Sciences and Technology of China under the grant of 2010CB630804 and 2012AA03A501. The
thankfulness of the authors also goes to all collaborative professionals and graduate students
participated in the project.
REFERENCES
[1] Z.D.Liu, P.Yan,H.S.Bao, S.Q.Xu, Q.J.Wang, Development of advanced ferritic heat resistant
steel G115 for A-USC power plants, The 5th Symposium on heat resistant steels and alloys
for high efficiency USC power plants, May,23-26, 2013, Seoul, Korea
[2] J.Klower, Development of nickel alloys for USC boilers, CISRI Heat Resistant Steels and
Alloys Lecture, Oct.,31,2012, Beijing, China
[3] Z.D.Liu, The research and development of heat resistant steels and alloys for 700 A-USCPP in china, IEA-Vienna Workshop-2012, Vienna, Austria, Sept., 19-20, 2012

179

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

SELECTION OF SUPER ALLOYS FOR A-USC POWER PLANTS


UNDER CONSIDERATION OF CREEP CRACK GROWTH BEHAVIOR
Falk Mueller, Alfred Scholz, Matthias Oechsner
Institut fr Werkstoffkunde, Technische Universitt Darmstadt
Grafenstr. 2, 64283 Darmstadt, Germany

ABSTRACT
To improve efficiency and flexibility and reduce CO2 emissions, advanced ultra super critical (AUSC) power plants are under development, worldwide. Material development and its selection are
critical to the success of these efforts. In several research and development programs / projects the
selection of materials is based on stress rupture, oxidation and corrosion tests. Without doubt,
these criteria are important. To improve the operational flexibility of modern power plants the
fatigue properties are of increased importance. Furthermore, for a safe operation and integrity
issues the knowledge about the crack behavior is essential. Crack initiation and crack growth may
be caused by natural flaws or cracks induced by component operation.
In order to develop new materials, properties like tensile strength and creep strength are an
important part of qualification and subsequent approval by notified bodies. Consequently short
term properties as well as time-temperature dependent properties are generated and taken into
considerations. In the case of high strength '-strengthening nickel-base alloys investigating the
creep crack behavior is also strongly recommended.
This article shows results of currently investigated nickel-based alloys for newly developed
headers, pipes and other high temperature boiler applications and their critical creep crack
propagation behavior.
INTRODUCTION
The nickel-based super alloys Alloy 617 and Alloy 263 are candidate materials for A-USC power
plants. Both alloys are known as materials with good mechanical properties and creep values at
elevated temperatures [1, 2]. Due to the precipitation hardening mechanisms Alloy 263 shows a
higher strength in comparison to Alloy 617.
Future requirements for power plants, e. g. improved efficiency and flexibility, raise the demand
regarding the mechanical properties of the deployed materials in order to avoid dimensional
changes of power plant components, e. g. wall thickness of pipes, which would cause increased
demands on processing and manufacturing of these parts.
Important properties of Alloy 617 and Alloy 263 were investigated and compared in a project
DE-4 [3] founded by the German government. Material was provided by the industry. The main
focus of the investigations has been the determination of creep properties and fatigue behavior.
Especially, the creep crack growth behavior of both candidate materials is of great interest.
Additionally, notch impact energy and fatigue crack growth data were determined. In the
following the results of tensile, creep, fatigue and creep crack growth tests are summarized.

180

MATERIALS
In this study a single melt of Alloy 617 and a single melt of Alloy 263 were investigated and high
temperature mechanical properties were compared. The product forms of the materials are thick
walled pipes with a wall thickness of 80 mm for Alloy 617 and 60 mm for Alloy 263 and an
outside diameter of 360 mm each. The heat treatment, chemical composition and microstructure
are conforming to current specifications [1, 2]. Alloy 617 was investigated in solution treated
state. Alloy 263 was age hardened at 4 h / 800 C / AC (AC - air cooling).
RESULTS AND DISCUSSION
In the following the results of the investigations of mechanical properties on Alloy 617 and
Alloy 263 are presented.
Tensile properties
For determination of the tensile properties at 700 C smooth cylindrical specimens with 8.4 mm
diameter were used. The tensile tests were performed in accordance to DIN EN ISO 6892-2 [4].
Their respective stress-strain-curves are shown in Figure 1.
As expected, Alloy 263 shows higher yield and tensile strength values compared to Alloy 617.
The rupture elongation A for Alloy 263 is much lower than for Alloy 617, but in an expected
range. The values of the reduction of area Z were determined to be 23 % for Alloy 263 and 47 %
for Alloy 617, respectively. In general, both materials show typical tensile properties at 700 C.

*800 C / 4 h / AC

T = 700 C

(MPa)
Alloy 263* / dET
500

Alloy 617 / dEO

50

(%)

Figure 1: Stress-strain-curves of Alloy 617 and Alloy 263 at 700 C

181

Creep rupture behavior


On both materials Alloy 263 and Alloy 617, creep rupture tests were conducted on smooth
cylindrical specimens in accordance to DIN EN ISO 204 [5] at 700 C with different stress levels.
In Figure 2 the creep rupture strength Ru/t/T and the creep rupture elongation Au are compared for
the materials tested. It could be observed that the creep rupture strength of Alloy 263 is much
higher than that of Alloy 617. According to DIN EN 10302 [6], the values for Alloy 263 are
within the specified scatter band. For Alloy 617 the measured creep rupture strength is clearly
above the specified values. This is not unusual, as modern melts of Alloy 617 often exceed the
standard requirements [6].
The creep rupture elongation values Au of Alloy 617 and Alloy 263 are found to be above 10 %,
with Alloy 617 showing higher values. The same can be observed for the reduction of area Zu.
Both investigated materials show high creep rupture strength and acceptable values of creep
rupture elongation and reduction of area.

*800 C / 4 h / AC

T = 700 C

Ru/t/T
(MPa)

1)

Alloy 263* / dET

Alloy 617 / dEO


2)

100
1)
2)

Alloy 263, DIN EN 10302:2008-6


Alloy 617, DIN EN 10302:2008-6

open symbols: ongoing test


40 (%)
A
u
30
20
10
10000

time (h)

Figure 2: Creep rupture strength Ru/t/T and creep rupture elongation Au of Alloy 617 and
Alloy 263 at 700 C
Low cycle fatigue behavior
The low cycle fatigue tests were performed using smooth cylinder specimens with a diameter of
7.9 mm. The tests were conducted under strain controlled loading conditions at 700 C according
to ISO 12106 [7].
Figure 3 shows the relationship between total strain range and fatigue life. The fatigue life is
determined as the number of cycles corresponding to a 5 % load drop. The fatigue life of Alloy
617 is found to be slightly lower compared to the fatigue life of Alloy 263.
Tests with hold times of 10 min in tension and pressure (Figure 4), respectively, show shorter
fatigue lives compared to tests conducted without hold times (Figure 3). The reduction of fatigue
lives is caused by the interaction of fatigue and creep [8]. Here, too, the fatigue life of Alloy 263
is slightly higher than the fatigue life of Alloy 617.

182

(%)

T = 700 C

R = -1
d/dt = 6 %/min

Alloy 263* / dET

Alloy 617 / dEO

* 800 C / 4 h / AC
1000

10000

Ni 5 (cycle)

Figure 3: Fatigue life as a function of total strain range of Alloy 617 and Alloy 263 at 700 C,
tests without hold time

(%)

T = 700 C

R = -1
d/dt = 6 %/min
tht = thp = 10 min

Alloy 263* / dET


Alloy 617 / dEO

*800 C / 4 h / AC
1000

10000

Ni 5 (cycle)

Figure 4: Fatigue life as a function of total strain range of Alloy 617 and Alloy 263 at 700 C,
tests with hold time in tension and pressure tht = thp = 10 min

183

Creep crack growth


Creep crack growth data for both materials Alloy 263 and Alloy 617 were generated using
standard side-grooved compact-tension specimens [9] (Figure 5). The Cs25-specimens have a
thickness of B = 25 mm, a thickness between side grooves of Bn = 20 mm, a width of W = 50 mm
and an initial crack length of a0 = 28.5 mm (this corresponds to a0/W = 0.57). The initial crack
length was pre-fatigued at room temperature.
Based on results from creep crack tests conducted on Cs25-specimens essential assessment
methods (e. g. two criteria diagram, Nikbin-Smith-Webster model, time dependent failure
assessment diagram) for crack initiation and crack growth were developed for steels [10, 11] and
austenitic materials [12].
The creep crack growth tests were performed according to ASTM 1457-07 [9] at 700 C. During
each test the creep crack length was measured by an AC-potential drop system. Additionally, the
load line displacement was measured. Both signals are used in further creep fracture mechanics
assessment.
In Figure 6 and Figure 7 the potential drop signal and the related load line displacement is shown
for a total of four creep crack growth tests, two for each material.
The first compact-tension specimen made of Alloy 617 was tested at 700 C with a constant load
of 18.5 kN (this corresponds to an initial stress intensity factor of KI0 = 45.1 MPa m1/2). After
2,600 h, the test was stopped and the specimen was brittle fractured to reveal the creep crack
surface. The creep crack growth length of this specimen was determined to be 3 mm.
At a load level of 13.5 kN (which corresponds to an initial stress intensity factor of
KI0 = 32.9 MPa m1/2) the second compact-tension specimen made of Alloy 617 is currently being
tested and shows a slow creep crack growth after a loading time of more than 9,000 h.
The first compact-tension specimen made of Alloy 263 was tested at 700 C with a constant load
of 18.5 kN. This specimen already failed after 14 h. The creep crack growth length on this
specimen was observed to be about 8 mm. At a load of 13.5 kN the test on the second compacttension specimen made of Alloy 263 was stopped after 54 h, which corresponds to a creep crack
growth length of 7 mm.
In a next step the tests were analyzed by a fracture mechanics assessment according to
ASTM 1457-07. The stress intensity factor KI and the parameter C* were determined. For side
grooved compact specimens the stress intensity factor can be computed as:
KI = F/(BBnW)1/2(2+a/W)f(a/W) ,

(1)

f(a/W) = [0.886+4.64(a/W)-13.32(a/W)2+14.72(a/W)3-5.6(a/W)4]/(1-a/W)3/2 ,

(2)

with the crack length a, the applied load F, the specimen width W, the specimen thickness B and
the specimen thickness between side grooves Bn.
The parameter C* can be determined by [9]:
C* = (dvLLD/dt)cF/[Bn(W-a)]n/(n+1)

(3)

with = 2.2 for compact-tension specimens, the creep portion of measured load-line displacement
rate (dvLLD/dt)c and the Norton-exponent n.
For the creep crack growth tests on Alloy 263 and Alloy 617 the corresponding results are shown
in Figure 7 and Figure 8. In Figure 7, the creep crack growth rate da/dt is plotted versus the stress
intensity factor KI and in Figure 8 versus the parameter C*.

184

0.71W

1.2W 0.01W

Utilizing the stress intensity factor KI (Figure 7), the difference between the measured creep crack
growth rates in Alloy 263 and Alloy 617 is obvious. At the same loading parameter (e. g.
KI = 50 MPa m1/2) the creep crack growth rate in Alloy 263 is more than two decades higher than
in Alloy 617.
In the C*-diagram (Figure 8), the difference appears to be smaller. However, it needs to be
considered how similar the C*-values for Alloy 263 and Alloy 617 are attained. The high load
line displacements rates in Alloy 263 compensate for the reduced loads.
Metallographic sections were prepared for the Alloys 263 and 617 (Figure 9). Even though the
two materials showed vastly different creep crack growth behavior, no differences can be seen in
the intergranular crack propagation. The fracture surfaces (Figure 10) show no substantial
differences, as well.
The creep crack behavior of Alloy 617 correlates with published results [13, 14]. In contrast to
the results of tensile, creep and fatigue tests on smooth specimens, the creep crack growth
behavior of Alloy 263 was unexpected. Published results of creep crack growth behavior of Alloy
263 do not exist.
It is assumed that the reason for the differences in creep crack growth behavior is the deformation
and damage behavior ahead of the crack tip (Figure 11). In the higher strength Alloy 263,
extremely high strains concentrate on a relatively small volume ahead of the crack tip. Further
investigations to reveal this phenomenon are ongoing.

Bn

W 0.005W

1.25W 0.01W

Figure 5: Drawing of a standard side grooved C(T)-specimen

185

0.47

0.850

Alloy 617 / dEO


T = 700 C
Cs25-specimen
Fmax = 18.5 kN

U0.46
(V)
0.45

Alloy 263 / dET*


T = 700 C
Cs25-specimen
Fmax = 18.5 kN

U0.825
(V)
0.800

* 800 C / 4 h / AC
0.44

0.775

0.43

0.750

0.42

0.725

0.41

0.700

vLLD
-250
(m)

vLLD
-250
(m)

-500

-500

-750

-750

-1000

-1000

-1250

-1250

specimen: dEO7c32
-1500

500

1000

1500

2000

2500(h)
time

specimen: dET7c22
3000

-1500

10

15

20

25(h)
time

30

Figure 6: Measured potential drop signal U and load line displacement vLLD at 700 C on creep
loaded Cs25-specimens, Fmax = 18.5 kN, left: Alloy 617/dEO, right: Alloy 263/dET
0.480

0.61

Alloy 617 / dEO


T = 700 C
Cs25-specimen
Fmax = 13.5 kN

U0.60
(V)
0.59

Alloy 263 / dET*


T = 700 C
Cs25-specimen
Fmax = 13.5 kN

U0.475
(V)
0.470

0.58

0.465

0.57

0.460

0.56

0.455

0.55

0.450

-1100

-800

vLLD
-1200
(m)

vLLD
-900
(m)

-1300

-1000

-1400

-1100

-1500

-1200

-1600

* 800 C / 4 h / AC

-1300

specimen: dET7c21

specimen: dEO7c33
-1700

2000

4000

6000

8000

10000
time
(h)

12000

-1400

20

40

60

80

100(h)
time

120

Figure 7: Measured potential drop signal U and load line displacement vLLD at 700 C on creep
loaded Cs25-specimens, Fmax = 13.5 kN, left: Alloy 617/dEO, right: Alloy 263/dET

186

da/dt
(mm/h)

T = 700 C

Alloy 263* / dET7c21


Alloy 263* / dET7c22
* 800C / 4h / AC

e
atur
liter

C
700
and
er b 617 @
t
t
a
sc Alloy
data

Alloy 617 / dEO7c32


50

100

KI (MPa m1/2)

Figure 8: Creep crack growth rate da/dt vs. stress intensity factor KI of Alloy 617 and Alloy 263,
T = 700 C

da/dt
(mm/h)

Alloy 263* / dET7c21


Alloy 263* / dET7c22

T = 700 C

* 800C / 4h / AC

Alloy 617 / dEO7c32


1

1000
C* (N/mmh)

Figure 9: Creep crack growth rate da/dt vs. parameter C* of Alloy 617 and Alloy 263,
T = 700 C

187

Figure 10: Details of creep crack growth, T = 700 C, Fmax = 18.5 kN, left: Alloy 617/dEO
(tmax = 2 600 h), right: Alloy 263/dET (tmax = 14 h), creep crack growth direction from left to
right

Figure 11: Details of creep crack growth surfaces, T = 700 C, Fmax = 18.5 kN, left:
Alloy 617/dEO (tmax = 2 600 h), right: Alloy 263/dET (tmax = 14 h), creep crack growth direction
from below to top

Rp0.2*

Rp0.2**
crack tip

rpl*
rpl**

Figure 12: Crack tip stress field (* Alloy 263, ** Alloy 617), schematically

188

CONCLUSIONS
Based on the presented results on two candidate materials (Alloy 617 and Alloy 263) for thick
walled pipes in A-USC power plants, it is strongly recommended to consider short term as well as
time-temperature dependent properties. These properties should be determined in standard tensile,
creep and fatigue tests on smooth specimens. A lack of information on creep crack growth
behavior can lead to an overestimation of the material's properties. Especially Alloy 263 shows a
significant lower resistance against creep crack growth in comparison to Alloy 617. These
findings are unexpected, as the tensile and creep strength values as well as fatigue life times of
Alloy 263 are higher compared to Alloy 617.
Consequently critical crack lengths are to be determined and non-destructive testing methods to
detect cracks in components are necessary in terms of integrity issues.
ACKNOWLEDGMENTS
The investigations were conducted as part of the joint research program COORETEC-DE4. The
work was supported by the Bundesministerium fr Wirtschaft und Technologie (BMWi) as per
resolution of the German Federal Parliament under grant number PTJ 0327705Y. The authors
gratefully acknowledge FDBR-Forschungsstiftung and the members of the project group for their
support. The responsibility for the content lies solely with its authors.
REFERENCES
[1] Alloy 617, "Material Data Sheet No. 4119", ThyssenKrupp VDM, (2005); "SMC-029",
Special Metals Corporation, (2005).
[2] Alloy 263, "Material Data Sheet No. 4020", ThyssenKrupp VDM, (1993); "SMC-054",
Special Metals Corporation, (2004).
[3] COORETEC-DE4: "Untersuchung zum langzeitigen Festigkeits- und Verformungsverhalten
von Rohren und Schmiedeteilen aus Nickel-Basislegierungen", PTJ 0327705Y, (2007-2013).
[4] DIN EN ISO 6892-2, Metallic materials Tensile testing Part 2: Method of test at elevated
temperature (2011).
[5] DIN EN ISO 204, Metallic materials Uniaxial creep testing in tension Method of test
(2009).
[6] DIN EN 10302, Creep resisting steels, nickel and cobalt alloys (2008).
[7] ISO 12106, Metallic materials - Fatigue testing - Axial-strain-controlled method (2003).
[8] Kobayashi, K. et al., "Creep-fatigue interaction properties of nickel-based superalloy 617",
Acta Metall. Sin., Vol. 24, No. 2 (2011), pp. 125-131.
[9] ASTM 1457, "Standard Test Method for Measurement of Creep Crack Growth Times in
Metals" (2007).
[10] Abe, F. et al., Creep-resistant steels, Woodhead Publishing Limited (Cambridge, 2008), pp.
446-471.
[11] Berger, C. et al., "Perspectives on improved life assessment methods for new plants", Proc.
4th Intern. Conf. on Advanced in Materials Technology for Fossil Power Plants, South
Carolina, 2004, pp. 653-671.
[12] Klenk, A. et al., "Developments in the use of creep crack initiation for design and
performance assessment", ECCC Information Day, 2003, pp. 45-57.
[13] Ren, W. et al., "A Review on Current Status of Alloys 617 and 230 for Gen IV Nuclear
Reactor Internals and Heat Exchangers", Journal of Pressure Vessel Technology, Vol. 131
(2009).
[14] Roesler, J. et al., "Wrougth Ni-Base Superalloys for Steam Turbine Applications beyond
700 C", Advanced Engineering Materials, 5, No. 7 (2003), pp. 469-483
189

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

FABRICATION TRIALS OF NI-BASED ALLOYS FOR ADVANCED


USC BOILER APPLICATION
Nobuhiko Saito*, Nobuyoshi Komai* and Yasuhiro Takei**
*Nagasaki R&D Center, Mitsubishi Heavy Industries, Ltd.
5-717-1 Fukahori-Machi, Nagasaki 851-0392, Japan
**Boiler Engineering Department, Mitsubishi Heavy Industries, Ltd.
3-3-1 Minatomirai, Nishi-ku, Yokohama, 220-8401, Japan

ABSTRACT
In order to reduce CO2 emissions and improve power generation efficiency, a development
project involving an advanced USC (A-USC) plant has been carried out in Japan since 2008. Nibased alloys are candidate materials for boiler components with high temperature steam
conditions, which are much stronger than conventional heat resistant steel. However, Ni-based
alloys have never been applied with respect to the high pressure parts and thick walled
components of USC coal-fired power plants. In this study, therefore, fabrication and characteristic
properties, such as weldability, the weld joint and bent part properties, and weld cracking
susceptibilities of Ni-based alloys such as HR6W, HR35 and two types of Alloy617 (High B and
Low B) pipes were evaluated. Additionally, two types of HR6W header mock-ups and a HR6W
tube element mock-up were fabricated. With the exception of Alloy617 (High B), the fabrication
trials of Ni-based alloy pipes were conducted successfully, and the long-term creep strength of
weldments and bends of Ni-based alloy pipes were found to be nearly equivalent to those of base
metal. In the case of Alloy617 (High B), hot cracking was observed.
INTRODUCTION
From the standpoint of environmental protection, reduction of CO2 emissions is urgently required.
In the case of power plants, improvement of power generation efficiency with higher steam
temperature and pressure conditions is the most effective set of countermeasures. Research and
development work on A-USC power plants with steam conditions of 700C and higher has been
initiated in Europe, the USA and Japan[1-4]. In Europe, field testing has also begun[2,3]. An AUSC power plant development project has been carried out in Japan since 2008[4]. From 2008 to
2012, boiler, turbine and valve materials were developed and verified. From 2013 to 2016, boiler
components and small turbine tests will be undertaken to verify component level reliability.
To achieve steam conditions of 700C and higher, the application of Ni-based alloy materials
becomes necessary for high temperature components. To date, however, there has been no actual
experience in the application of Ni-based alloy for thick walled components in boilers. Therefore,
the establishment of optimized and practical fabrication procedures such as welding and bending
technology is needed. Furthermore, long-term reliability is also highly important. In this study, in
190

order to establish welding and bending technology for large diameter and thick walled pipes of
Ni-based alloys, weldability, welded joint properties, induction bending properties, as well as
long-term creep rupture properties were evaluated in comparison with those of the base metal. In
addition, mock-up tests were conducted for the purpose of applying the developed technologies to
components.
TEST MATERIALS AND EXPERIMENTAL METHOD
HR6W, HR35, Alloy617, Alloy263, Alloy740 and Alloy141 were candidate materials for A-USC
boiler components in the Japanese project. To evaluate the manufacturability of large diameter
and thick walled pipe, HR6W, HR35 and two types of Alloy617 (High B and Low B) pipes were
tested. The large diameter and thick walled pipes of these alloys were 350mm in outside diameter
and 50mm in wall thickness, and their chemical compositions are shown in Table 1. In the case of
Alloy617, two types of materials differing in the amount of B content were tested. The
appearances of the test materials are shown in Fig.1. Six types of candidate plate materials with
thickness of 25mm were also tested to evaluate weld cracking susceptibilities. In mock-up testing,
large diameter and thick walled HR6W pipes with sizes of 558mm OD x 138mm WT and 635mm
OD x 72mm WT were used, together with small diameter tubes having sizes of 45mm OD x
12mm WT and 63.5mm OD x 12mm WT.
The welding conditions and groove configurations for the Ni-based alloy pipes are shown in
Table 2. Welded joints were made by auto gas tungsten arc welding (Auto-GTAW) with
narrow groove using two types of matching wires, and additional HR6W weldments were made
Table 1 Chemical compositions of tested Ni-based alloy
C

Si

Mn

Ni

Cr

Mo

Co

Al

Ti

(mass%)
Nb

Fe

HR6W

0.070 0.15 1.02 0.002 <0.001 45.19 23.37

7.54

0.09 0.18 Bal. 0.0019

HR35

0.051 0.07 0.10 0.004 0.001 49.71 29.90

4.44

0.73

Alloy617
0.061 0.07 0.04 0.002 <0.001 Bal. 22.22 9.12
(High B)

11.92 1.01 0.41

Alloy617
0.062 0.15 0.04 <0.01 <0.001 Bal. 21.83 9.22
(Low B)

12.09 0.90 0.49

Figure 1 Appearance of large diameter Ni-based alloy pipes

191

Bal. 0.0033
-

0.0041

0.32 0.0023

by shielded metal arc welding (SMAW) using matching filler metals. The inter-pass temperature
was controlled to reduce hot cracking susceptibilities. Post weld heat treatment was not
performed. After welding, side bend testing, metallography examination, tensile testing, hardness
measurement, Charpy impact testing, and creep rupture testing were conducted.
Ni-based alloy materials are characterized by high hot cracking susceptibilities in comparison
with conventional heat resistant steel[5]. In addition, the re-heat cracking susceptibilities of Nibased alloy are higher than that of the conventional heat resistant steel due to strengthening of
phase precipitates in the grain[6]. Therefore, in the case of large diameter and thick walled pipe,
there is concern that re-heat cracking may occur in the grain boundary during relaxation of stress
in high temperature operations. Re-heat cracking was indeed reported in Alloy617 pipe
weldments in European A-USC field testing[6,7]. Given this background, hot cracking
susceptibilities and re-heat cracking susceptibilities were evaluated by trans-varestrain testing[5]
and slow strain rate tensile testing, respectively[8,9]. The trans-varestrain test was conducted with
3% loading strain. Hot cracking susceptibilities were evaluated by comparing maximum crack
length and total crack length. Slow strain rate tensile testing was conducted at 700C, 750C and
800C using weld thermal simulated HAZ specimens. The specimens were heated to 1300C in
about 70 seconds by means of high frequency induction heating, held for 5 seconds, and then
cooled down to 300C in about 80 seconds. The crosshead speed was 0.5mm/min (gauge length:
25mm). The re-heat cracking susceptibilities were evaluated by observing the reduction of area of
the specimens.
The induction bending properties of HR6W pipe were evaluated. Before induction bending,
support piping of carbon steel was welded with the HR6W pipe. Circumferential zone heating was
undertaken using a high frequency induction heater, after which the heated zone was bent. The
heating temperature was controlled such that the temperature of the outer surface could be
maintained between 1050C and 1150C, while the pipe feeding speed was approximately
0.1mm/sec. The pipe was bent at an angle of 50 and the bending radius was 1400mm (4 x OD).
The bent pipe was re-solution heat treated after bending under the condition of 1220C X 1hWQ,
which was the same condition of the heat treatment for base metal. After bending, metallography
examination, tensile testing, hardness measurement, Charpy impact testing, and creep rupture
testing were conducted.
Table 2 Welding conditions for large diameter Ni-based alloy pipes

192

For the purpose of component testing, applying the developed technologies, HR6W superheater
and reheater header mock-ups, as well as a HR6W tube element mock-up, were made to confirm
the manufacturability in the context of actual application for A-USC boiler compornents.
RESULTS AND DISCUSSION
Weldability and Weldment Properties
Figure 2 shows the appearance of GTAW and SMAW applied for HR6W pipes. The results of
side bend testing are shown in Fig.3, with all results being acceptable except for Alloy617 (High
B) were found. In the case of Alloy617 (High B) weldment cracking was observed after side bend
testing. Figure 4 shows the macrostructure and microstructure with respect to the longitudinal
sections of weldment. No defects were found except for Alloy617 (High B). There, hot cracking
in the heat affected zone (HAZ) near the fusion line could be observed. It was considered that the
damage seen in the Alloy617 (High B) weldment after side bend testing was due to hot cracking
in the HAZ. On the other hand, no such cracking in the HAZ of Alloy617 (Low B) could be
observed, suggesting that reduction of the amount of B content may be effective in lowering hot
cracking susceptibilities. However, the grain size of Alloy617 (Low B) was finer than that of
Alloy617 (High B), and thus grain size may also be effective in this regard.
Figure 5 shows the results of the elevated temperature tensile strengths of Ni-based alloy
weldments when compared with base metal properties. Although the tensile strength portions of
HR6W (SMAW) and Alloy617 (High B) weldments were lower than those of base metal due to
weld metal fracturing, other all tensile strengths were higher than for base metal.
Figure 6 shows the hardness distributions in the weldments. Hardness was measured at the inner
and outer sides of the pipes, as well as at midwall. In all Ni-based alloys, a hardening region was
seen in the HAZ. The hardening at the inside of the pipe was the greatest. The higher hardness in
the HAZ was thought to result from work hardening induced by tensile stress during shrinkage
due to solidification, and precipitation induced by the heat input of the welding thermal cycle.
Figure 7 shows the 0C Charpy impact test results of the weldments. The notch positions were
located at the center of the weld metal, the fusion line, the HAZ and the base metal. Although the
portion of absorbed energy at the fusion line and the HAZ of Alloy617 (High B) weldments

(a) GTAW

(a) HR6W

(b) SMAW

(d) Alloy617 (Low B) (e) HR6W (SMAW)

Figure 2 Welding of Ni-based alloy pipes


(HR6W)

Figure 3 Appearance of specimens after side


bend testing

193

(b) HR35

(c) Alloy617 (High B)

were lower than for other specimens due to slight hot cracking, other all specimens showed good
toughness of 100J and over.

Figure 4 Macrostructures and microstructures of Ni-based alloy weldments

Figure 5 Elevated temperature tensile strengths of Ni-based alloy weldments

194

HR6W

Weld metal

Weld metal

HR35

Alloy 617

Weld metal

250

200

B
A

A
A
1

150

100

Open : GTAW
Solid : SMAW

40

A
B

Vicker's hardness (HV)

300

Open : High B
Solid : Low B

Distance

Distance

Distance

(a) HR6W

(b) HR35

(c) Alloy617

1mm

Figure 6 Hardness distributions in Ni-based alloy weldments

Figure 7 Charpy impact properties in Ni-based alloy weldments

Figures 8, 9 and 10 show the creep rupture properties of HR6W, HR35 and Alloy617 weldments,
respectively, compared with base metal properties. The creep rupture tests exceeding 20,000h
under low stress conditions are still ongoing. The creep rupture strengths of HR6W weldments
made by both of GTAW and SMAW were higher than those of base metal, and no difference of
strength relating to welding method could be observed. The creep rupture strengths of the HR35
weldments were also higher than those of base metal, although the long creep rupture strengths
exhibit a slightly declining trend. It is necessary to verify the long-term properties, i.e., 30,000h
and more. In the Alloy617 weldments, while the creep rupture strength of the weldment at 700C
was slightly lower than that of base metal (caused by weld metal fracturing), the test results at
750C and 800C were nearly equivalent to base metal. The higher creep rupture strength of the
weldments was assumed to result from increased weldment hardness.

195

Figure 8 Creep rupture properties of HR6W weldments

Figure 9 Creep rupture properties of HR35 weldments

Figure 10 Creep rupture properties of Alloy617 weldments

196

Weld Cracking Susceptibility


Figure 11 shows the trans-varestrain test results for six types of Ni-based plates. Hot cracking
susceptibilities of Alloy617 (High B) were the highest; followed by HR35, Alloy740 and
Alloy141, which were high; and then by HR6W and Alloy263, which were relatively low. In
order to prevent of hot cracking, control should be exerted to reduce the heat input and the interpass temperature during the welding process, with optimization of the amount of B content and
the grain size in the base metal.

12
10
8
6
4
2
0

Total crack length

Alloy 141

Alloy 740

Alloy 263

Alloy 617
(High B)

HR35

Maximum crack length

HR6W

Crack length (mm)

Figure 12 shows the results of slow strain rate tensile testing of simulated HAZ specimens. The
re-heat cracking susceptibilities of Alloy617 (High B), Alloy263, Alloy740 and Alloy141 were
somewhat higher than those of HR6W and HR35. In the case of Alloy617, recovery of reduction
of area at 800C due to solid solution of the phase was observed[10]. It was considered that
HR6W and HR35 have good creep ductility due to the absence of phase precipitates in HR6W
and the presence of -Cr precipitates instead of phase in HR35. Accordingly, it is better to
perform post weld heat treatment such as re-solution heat treatment or stress relief heat treatment
so as to prevent re-heat cracking in thick walled components of Ni-based alloy, especially
Alloy617, Alloy263, Alloy740 and Alloy141.

Figure 11 Trans-varestrain tests results of Ni-based plates


80

Alloy617
(High B)

Reduction of area (%)

70

HR35
HR6W

60
50
40
30

Alloy263

20

Alloy740
Alloy141

10

Crosshead speed:0.5mm/min
0
650
700
750
800
Test temperature ()

850

Figure 12 Slow strain rate tensile test results of Ni-based alloy plates

197

Induction Bending Properties of HR6W Pipe


Figure 13 shows the induction bending process and the appearance of bent pipe made of HR6W.
The results of pipe wall thickness measurements and the wall thickness reduction rates are shown
in Figure 14. Decreased thickness was seen in the extrados (tension side), with increased thickness
in the intrados (compression side). A small difference in the wall thickness was seen between the
beginning portion of the bend and the center of bending, with very slight change in wall thickness
observed in the finishing portion of the bend. Such changes normally occur when pipe is bent, and
the reduction in wall thickness was about 7% at maximum.
The microstructure of the bent pipe was observed to be (gamma) phase across all sections. The
tensile properties of the bent pipe were nearly the same as for base metal, and no variation was
seen along the circumferential locations. The impact properties of the bent pipe were superior to
those of base metal.

Wall thickness (mm)

Figure 13 Induction bending process of HR6W pipe and appearance of bent pipe

45

40
Start of bending
Center of bending
End of bending
Straight part

35

Reduction ratio of
wall thickness (%)

-15
-10
-5

a Extrados
b

h
g

c
f

d
e Intrados

0
5
10
15
a

c
d
e
f
g
Location in circumference

Figure 14 Dimensional changes in HR6W bent pipe circumference

198

Figure 15 shows the creep rupture test results of bent pipe, as compared with base metal
properties. Test specimens were taken from the extrados portion, neutral portion and intrados
portion of the bent pipe. The creep rupture strengths were nearly equivalent to those of the
respective base metals, and it was thus recognized that there had been no strength reduction due to
bending. Also, no differences in strength related to sampling location could be seen.
Header and Tube Element Mock-up Testing
Figure 16 shows the appearance of the superheater and reheater header mock-ups, and the tube
element mock-up. These mock-ups test program included butt welding, stub tube welding, stub
welding, tube bending and spacer welding; no particular problems were experienced with these
manufacturabilities. According to the mock-ups test results, it was confirmed that fabrication of
HR6W pipe can be practically applied for A-USC boiler components.
500

Stress (MPa)

300
200

700
750

100

800

70
50
30 1
10

Base metal
Tension side
Neutral
Compression side

102

103
Time to rupture (h)

104

105

Figure 15 Creep rupture properties of HR6W bent pipe after solution heat-treatment
at 1220 for 1h

(a) Superheater header

(b) Reheater header

(c) Tube element

Figure 16 Mock-up test results of HR6W headers and tube element


199

CONCLUSIONS
1. In order to evaluate manufacturability with respect to welding and bending technology of Nibased alloys for thick walled components in A-USC boilers, fabrication trials of HR6W, HR35
and two types of Alloy617 (High B and Low B) pipes were conducted. The long-term creep
rupture strength of weldments and bent pipes made of these alloy were also discussed.
2. The welding trials of Ni-based alloy pipe materials were conducted successfully, with the
exception of Alloy617 (High B). Although the hot cracking was observed in HAZ near the
fusion line of Alloy617 (High B) weldments, no hot cracking was observed in Alloy617 (Low
B). Also, the mechanical properties and creep rupture properties of Ni-based weldments were
found to be nearly equivalent to those of base metal.
3. In order to prevent of hot cracking, control should be exerted to reduce the heat input and the
inter-pass temperature during welding process, and to optimize the amount of B content and the
grain size in the base metal. Moreover, it is better to perform post weld heat treatment such as
re-solution heat treatment or stress relief heat treatment so as to prevent re-heat cracking in
thick walled components of Ni-based alloy, especially Alloy617, Alloy263, Alloy740 and
Alloy141.
4. The bending trial of HR6W pipe was conducted successfully. The metallic structure,
mechanical properties and creep rupture properties of HR6W bent pipe were nearly the same as
those of base metal.
5. Two types of HR6W header mock-ups and a HR6W tube element mock-up were manufactured
without any particular problems. According to these test results, it was confirmed that
fabrication of HR6W pipe can be practically applied for A-USC boiler components.
REFERENCES
[1] R. Viswanathan et al., U.S. Program on Materials Technology for Ultra Supercreitical CoalFired Steam Power plants, Paper No.Creep2007-26826, Proceeding of 8th International
Conference on Creep and Fatigue at Elevated Temperatures, San Antonio, Texas, July.
2008.
[2] R. Blum, The European Perspective and Advancements for Advanced USC Steam Power
Plants, Proceedings of Sixth International Conference on Advances in Materials
Technology for Fossil Power Plants, Santa Fe, New Mexico, August. 2010, pp. 1-10.
[3] K. Metzger et al., GKM Test Rig: Investigation of the Long Term Operation Behavior of
Tubes and Forgings Made of Alloys for Future High Efficient Power Plants, ibid, pp. 86-95.
[4] M. Fukuda, Advanced USC technology Development in Japan, ibid, pp. 325-341.
[5] M. Suzuki et al., Research for the Quantitative Evaluation of Solidification Cracking
Susceptibilities in Ni-base Alloy Weld Metals, KOBE STEEL ENGINEERING REPORTS,
Vol. 54, No. 2 (2004), pp. 43-46.
[6] G. Schmidt et al., Component Test Facility for a 700C Power Plant, COMTES700,
(2012).

200

[7] H. Schaffon, 700C Power Plant Technology Status and Challenges, Proceedings of 9th
Liege Conference: Materials for Advanced Power Engineering 2010, September. 2010, pp.
20-28.
[8] A. Dhoog, Survey on reheat cracking in austenitic stainless steels and Ni base alloys,
Welding in the World, Vol. 41, No. 3 (1998), pp. 30-43.
[9] K. Okada et al., Sensitivity of Reheat Cracking on Austenitic Stainless Steel Containing
Nb, CAMP-ISIJ, Vol. 17, (2004), pp. 928.
[10] H. Semba et al., Creep properties and strengthening mechanisms in 23Cr-45Ni-7W(HR6W)
alloy and Ni-base superalloys for 700C A-USC boiler, Proceedings of fifth international
conference on advances in materials technology for fossil power plants, Florida, USA,
March. 2008, pp. 168-184.
[11] N. Saito et al., Creep Rupture Properties of HR6W for Advanced-USC Boilers,
Proceedings of Sixth International Conference on Advances in Materials Technology for
Fossil Power Plants, Santa Fe, New Mexico, August. 2010, pp. 962-971.

201

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

DEVELOPMENT AND TRIAL MANUFACTURING OF NIBASE ALLOYS FOR COAL FIRED POWER PLANT WITH
TEMPERATURE CAPABILITY 800C
Shinya Imano, Naoya Sato, Hironori Kamoshida
Hitachi Research Laboratory, Hitachi, Ltd.
7-1-1 Omika-cho, Hitachi 319-1292, Japan
Abstract
Large scale components of the conventional 600C class steam turbine were
made of the ferritic steel, but the steam turbine plants with main steam
temperatures of 700C or above (A-USC) using the Ni-base superalloys are now
being developed in order to further improve the thermal efficiency. The weight of
the turbine rotor for the A-USC exceeds 10ton. A lot of high strength superalloys
for aircraft engines or industrial gas turbines have been developed up to now. But
it is difficult to manufacture the large-scale parts for the steam turbine plants
using these conventional high strength superalloys because of their poor
manufacturability. To improve high temperature strength without losing
manufacturability of the large scale components for the A-USC steam turbine
plants, we developed Ni-base superalloy USC800(Ni-23Co-18Cr-8W-4Al-0.1C
[mass %]) which has temperature capability of 800C with high
manufacturability achieved by controlling microstructure stability and
segregation property. The 700C class A-USC materials are the mainstream of
current development, and trial production of 10 ton-class forged parts has been
reported. However, there have been no reports on the development and trial
manufacturing of the A-USC materials with temperature capability of 800C. In
this report, results of trial manufacturing and its microstructure of the developed
superalloy which has both temperature capability 800C and good
manufacturability are presented. The trial manufacturing of the large forging,
boiler tubes and turbine blades using developed material were successfully
achieved. According to short term creep tests of the large forging and the tube
approximate 100,000h creep strength of developed material was estimated to be
270MPa at 700 C and 100MPa at 800C.

202

1. Introduction
To reduce CO2 emissions, further improvements of the coal-fired power
generation plants are required. Research and development on the Advanced Ultra
Super Critical (A-USC) power generation is underway in Europe and Japan. The
double-reheat system for steam temperatures of 700C/720C/720C with an
electric power transmission efficiency of 46% is expected to improve by 10%
compared to the conventional USC power plants (electric power transmission
efficiency of 42%). For further improvement of the efficiency, increasing the
main steam temperature to 750C or higher is being discussed (U.S.
ULTRAGEN III). Among the candidate materials for 700C class A-USC plants,
IN617, DT706, Waspaloy, DT750, etc., are being studied in Europe1). In Japan,
the development of the Ni-base alloys such as FENIX-7002), LTES-7003), and
USC1414) have been reported. At 750C, Waspaloy1) exhibits a creep rupture
strength of approximately 100 MPa, which is the highest among the candidate
materials, but manufacturability of the ingot and hot workability are not sufficient
for the large scale components such as steam turbine rotors. With regard to the
USC141 alloy, the manufacturability of large forgings superior to that of the
Waspaloy, whereas the temperature at which the 100,000h creep rupture strength
is 100 MPa (the temperature that is satisfied with the strength for turbine rotors,
hereinafter, creep resistant temperature) is below 750C. The creep resistant
temperature of the other candidate materials is 720C or below. Because the
boiler tube is exposed to a temperature of 750C or higher even when the main
steam temperature is 720C, it is necessary to develop new materials to raise the
main steam temperature and reheat steam temperature to more than 720C and
750C, respectively. The large forgings such as steam turbine rotors and boiler
pipes are formed by hot forging at approximately 1000C. However, it is
generally difficult to satisfy both hot workability at 1000C and high temperature
strength in the range of 750 - 800C, so it is a major obstacle to enhancing the AUSC efficiency. The purpose of this research is to develop the Ni-base alloys
which can be used at 750C or higher, and that have superior large ingot
manufacturability and forgiability. In this study, using a thermodynamic
calculation5), a promising composition of the Ni-base alloys (USC800) was
determined and fundamental high temperature properties are investigated6) using
specimens from 20 kg ingots. Fig. 1 shows the feature of the USC800.
In this report, results of trial manufacturing of the large forging and the tubes for
boiler using the USC800 according to evaluation of the large ingot
manufacturability were reported. And results of short time creep tests using
specimens from large forging and 10,000h class creep tests using specimens from
20kg ingot were also reported.

203

phase fraction (at.%)

50
40

Service
temperature

30

Forging
temperature

USC800

USC141
20

IN740
10 Nimonic263
IN617
0
600
700

800
900
Temperature (C)

1000

1100

Fig.1 Metallurgical Feature of USC800

2. Experimental Procedure
2.1 Evaluation of large ingot manufacturability
Fig. 2 shows the dimension of the specimen for the horizontal directionally
solidification furnace shown in Fig. 3. The chemical composition of the specimen
is shown in Table 1. The horizontal directionally solidification furnace enables
the changing of the cooling rate () and solidification rate (R) by controlling the
heater and the water cooled chill. The cooling rate () and solidification rate (R)
were evaluated by the temperature datas from thermo couples. The horizontal
directionally solidification furnace was placed in the chamber filled with high
purity argon. Segregation index (R1.1) 6) was controlled between 7 and 2([K/min]
[mm/min] 1.1). Macro segregation was examined by using the horizontal slice of
the ingot.
For the forging tests, the ingot was sliced into 4 parts which has different and
R respectively and 4 round bars with 15mm were made by high temperature
forging followed by swaging. Temperatures during these processes were kept in
the range from 1000C to 1150C. Each round bar was heat treated at 1160C/4h,
204

1000C/6h, 900C/24h, and 700C/16h. After the swaging, microstructures of


these round bars were investigated, also tensile and creep tests were performed.
2.2 Trial Manufacturing of large ingot and tube for boiler
An electrode was made by VIM. Weight of the VIM electrode was 6tons. An
ingot was made by VAR using the VIM electrode. Diameter of the ingot was
800mm. A 3ton forging was made by high temperature forging using the VAR
ingot. The 3ton forging was heat-treated at 1100C/4h, 1000C/6h, 900C/24h,
and 700C/16h after forging.
Tubes with an outer diameter of 50mm and an inner diameter of 34mm were
made by piercing, hot drawing and cold roll in a commercial production line.
Maximum area reduction ratio of cold working was 50%. Visual test and ultra
sonic inspection were performed after final cold working. To evaluate the hightemperature properties, short time creep tests were performed using the specimen
from 3ton forging and tubes.

50

Bolt Hole for Handling


M162
70
Hole for Thermo Couple

120

70

50

25

50
200

50

50

25

25

25
50

Fig.2 Dimension of ingot for Remelting

205

Table1 Chemical composition of the ingot for remelting(mass%)

Bal

4.05

23.20

17.00

5.95

<0.01

2.03

0.0027

0.012

0.041

Vacuum or Argon Chamber

Thermo Couples

Chill

Diffusion Pump

Rotary Pump

Carbon Heater
Ingot
Fig.3 Appearance of Furnace Chamber and schematic of the furnace

3. Results and Discussion


3.1 Evaluation of large forging manufacturability
Fig. 4 shows the variation of segregation index (SI) in the ingot. Fig. 5 shows
the cross section of the ingot. As shown in Fig. 5, there was no macro segregation
all over the cross section. And this result represents that the critical value for
segregation index of USC800 is lower than 2([K/min] [mm/min] 1.1). Critical
value for segregation index is defined as lower limit of segregation index to make
an ingot without macro segregation. It is reported that critical value for
segregation index of Alloy706 is 2.0([K/min] [mm/min] 1.1), so the segregation
property of USC800 is as good as or much better than that of Alloy706. Since
diameter limit of Alloy706 is about 800mm by ESR, solidification condition of
the ingot shown in Fig. 5 is equivalent to that of 800mm-dia ESR ingot.
Using the ingot shown in Fig. 5, 4 round bars were made from the 4 parts (A to
D) of the ingot by high temperature forging and swaging. Fig. 6 shows the
206

optical micrographs of the round bars after hot forging and heat treatment.
Microstructure of the bar from the part D is coarser than that of the bar from the
part A. There are precipitation rich areas (dark contrast) in all the specimens. Fig.
7 shows the SEM micrographs of the bar from the part C. There were lots of
carbides in the precipitation rich areas. Most of the precipitates on grain
boundaries were Cr rich carbides. The precipitates in the grain were Cr rich
carbides and Mo rich carbides. No other phases were found in the precipitation
rich areas of the all bars.
Fig. 8 through 11 show the tensile properties at both room temperature and
800C as a function of the segregation index. These results indicated that the
effect of solidification condition on tensile properties was negligible. And the
ductility is favorable level at both room temperature and 800C. It is concluded
that mass effect due to the difference of solidification condition is negligible.

1.1
SI
(/min)(mm/min)1.1
SI(K/min)(mm/min)

7.0
6.0
5.0
4.0
3.0
Segregation index
SI=R1.1
:Cooling Rate
R:Solidification Rate

2.0
1.0
0.0
0

50

100

150

Distance from chill(mm)

(mm)
Fig.4 Variation of segregation index in the ingot

207

200

Chill Side

Fig.5 Horizontal Cross Section of the Ingot

Fig.6 Microstructure after Forging and Heat Treatment


208

Cr

Mo

Cr

Fig.7 SEM Micrographs after Forging and Heat Treatment

209

1400
1200

Stress(MPa)

1000
800
600
0.2% Yield Stress
0.2%

400

Tensile Strength)
(

200

Room Temperature

0
0

4
6
8
SI (K/min)(mm/min)1.1
Fig.8 Effect of Solidification Condition on Tensile Properties(1)
45.0
40.0

Ductility (%)

35.0
30.0
25.0
20.0
15.0

Elongation

10.0

Reduction of area

5.0

Room Temperature

0.0
0

4
6
8
SI (K/min)(mm/min)1.1
Fig.9 Effect of Solidification Condition on Tensile Properties(2)

210

800
700

500
400
300

0.2% Yield Stress


0.2%800

200

(800
Tensile Strength )

100

800

0
0

4
6
8
1.1
SI (K/min)(mm/min)
Fig.10 Effect of Solidification Condition on Tensile Properties(3)

60.0
50.0

Ductility (%)

Stress(MPa)

600

40.0
30.0
20.0

800
Elongation
800
Reduction of area

10.0

800

0.0
0

4
6
8
SI (K/min)(mm/min)1.1
Fig.11 Effect of Solidification Condition on Tensile Properties(4)

211

3.2 Trial manufacturing of 3ton forging and tube


Fig. 12 shows the appearance of the 800mm-dia VAR ingot weighing more than
6 tons. It was forged by 4,000 ton class open die press. Macro segregation was
not observed in the cross section of the ingot. This ingot was forged into
octagonal bar with a 450mm diameter and 2,500mm length as shown in the right
side of Fig. 12. According to the ultrasonic testing, any defects were not detected.
As shown in the lower left of the Fig. 12, cylindrical machined specimens
exhibited good weldability. Fig. 13 shows the appearance of the tubes using
USC800. Any defects were not detected by ultrasonic testing after final cold
working.
Fig. 14 shows the results of creep rupture tests. Approximate creep rupture
strength of the specimens from 20kg ingot at 800C is about 90 to 100MPa
according to 10,000h class creep tests. Short time creep strength of the specimens
from 3ton forging and tubes was as high as that of specimens from 20kg ingot.
VAR Ingot800mm, 6 tons)
3 ton

Forging

forging

t = 8mm(38mm)
450mm
Machining
Welding
Fig.12 Appearance of VAR ingot, 3 ton forging and welding test specimen

212

50mm dia. 8mmt


Fig.13 Appearance of the tube for boiler

4. Summary
1. Segregation property of USC800 is as good as or much better than that of
Alloy706.
2. It is possible to manufacture 800mm-dia VAR ingot without segregation
defects by using USC800.
3. It is possible to manufacture 3ton forging from 6ton ingot of USC800 by
4,000 ton class open die forging press.
4. Creep rupture strength of the specimens from 3ton forging and tubes is as
high as that of the specimens from 20kg ingot.
Acknowledgments
A part of this work was supported by the New Energy and Industrial
Technology Development Organization (NEDO) under the Ministry of Economy
Trade and Industry of Japan. The authors are grateful to Dr. Katsunari Oikawa
from Tohoku University for the designing the directionally solidification furnace
in this report.
References
1. F. Schubert, H. J. Penkalla, and J. Rosler: Wrought Ni-Base Alloys For Rotor
Shafts in Advanced USC Power Plants. Proceedings from the Fourth
International Conference Advances in Materials Technology for Fossil
Power Plants, (2004), 587.
2. S. Imano, H. Doi, T. Takahashi, and K. Kajikawa, Proc. Conf. Superalloys
213

3.

4.
5.

6.

7.

718, 625, 706 and Various Derivatives, (2005), 77.


R. Yamamoto, Y. Kadoya, S. Ueta, T. Noda, R. Magoshi, S. Nishimoto, and
T. Nakano: Development of Wrought Ni-Based Superalloy with Low
Thermal Expansion for 700C Steam Turbines. Proceeding from the Fourth
International Conference Advances in Materials Technology for Fossil
Power Plants, (2004), 623.
S. Imano and H. Doi: CAMP-ISIJ Vol. 17, (2004), 926.
N. Saunders, X. Li, A. P. Miodownik: J-Ph. Proc. Symp. Materials Design
Approaches and Experiences, eds. J.-C. Shao et al., 185-197, (2001),
Warrendale, PA, TMS
Jun Sato, Hironori Kamoshida amd Shinya Imano: Proceedings from the
Fourth International Conference on Advances in Materials Technology for
Fossil Power Plants, (2010), 386.
H. Yamada, K. Kajikawa, S. Suzuki, and F. Takahashi: JSPS 123rd
Committee Research Report, vol. 47, No. 3, (2006), 293.

214

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

MANUFACTURING DEMONSTRATION OF INCONEL ALLOY 740H


FOR A-USC BOILERS
B. A. Baker, R. D. Gollihue, J. J. deBarbadillo, S. J. Patel, and D. Maitra
Special Metals, a PCC Company
3200 Riverside Drive
Huntington, WV 2570
ABSTRACT
INCONEL alloy 740H was specifically developed for use in coal-fired AUSC boilers. This
alloy displays a unique combination of steam and coal-ash corrosion resistance, microstructure
stability, creep strength and heavy section weldability. During the past two years Special Metals
and Wyman-Gordon have undertaken an intense effort to demonstrate their capability to
manufacture full-size boiler components, characterize their properties and simulate field assembly
welds. This work was performed according to the requirements of ASME Boiler Code Case 2702
that was recently issued. This paper covers manufacturing of tube and pipe products and property
characterization including recent data on the effect of long time exposure on impact toughness of
base and weld metal. New data will also be reported on coal ash corrosion of base metal and weld
metal. An overview of welding studies focused on integrity of circumferential pipe joints and a
discussion of remaining technical issues will be presented.
INTRODUCTION
Impetus for utilization of nickel-base alloys for construction of coal-fired power plant
components has been provided by ambitious goals for operating parameters and material
properties established by various consortia around the world.1-3 The standard for strength
applied for superheater tubing capable of withstanding 700C steam at 35 MPa, originally
established by the European AD700 project, is a rupture life of 100 MPa at 750C.4 Operation
with such high temperatures and pressures enables increases efficiency, thereby reducing fuel
consumption, and emissions (Figure 1).5 Fabrication requirements are demanding, with
component diameters and weights pushing the limits of nickel-base alloy fabrication capabilities.
Add to this the corrosion allowance for the superheater, set forth by the European AD700 Project
as less than 2mm of corrosion loss in 200,000 hours .4 In addition, end-users desire long-term
thermal stability in materials of construction to ensure predictable long-term strength,
repairability, etc.
While worldwide interest continues in the furtherance of advanced ultra-supercritical coal-fired
boiler technology, as evidenced by the continued efforts of consortia in the United States and
Europe, Japan, India and China, application of nickel-base alloys to high-efficiency powergeneration technology is expanding beyond the utilization of coal as a fuel, and steam as the
working fluid.1-3 Utilization of carbon dioxide to transfer thermal energy via the Brayton cycle
under consideration for concentrated thermal power systems6, and gas turbine systems for power
generation. Operating temperatures under consideration for such systems, like A-USC coal
215

combustion, could extend into the 700C-750C range, with similarly high operating pressures
(25-35 MPa), necessitating the utilization of nickel-base alloys to sustain such operating
conditions. Such systems could be utilized for thermal transfer in generation IV nuclear systems
as well.7
Currently, the most efficient commercial ultra supercritical designs operate only up to 605C with
steam pressures in the 25-30 MPa range, not requiring the utilization of nickel-base alloys.
However, the creep strength enhanced ferritic grades (Gr. 91/92) in use are satisfactory only up to
about 620C. Advanced austenitic stainless steel grades can be used up to about 680C.
However, their high coefficient of thermal expansion, low thermal conductivity and strength
capability limit impose limitations upon their utility. To cope with the conditions described for
A-USC plants, nickel-base alloys must be considered.
Alloy 740H has been developed to meet the myriad demands of high temperature/high pressure
service with imposed corrosion requirements, in the required section thicknesses and forms.

Figure 1. Effect of efficiency upon CO2 emission for steam-generating pulverized coal.5
Figure 2 illustrates the rupture strength capability of the various classes of materials that are being
considered for A-USC conditions.1 Alloys 740H and 617 are primary candidates in most current
A-USC designs.

Figure 2. Stress to produce 100,000 hour rupture life for 9-12 Cr ferritic steels, austentic steels,
and selected nickel-base alloy1.
216

Work by the U.S. A-USC consortium has culminated in issuance of code case 2702 for UNS
N07740 in September of 2011. This action represents significant progress in the A-USC
community as this is a first instance of use of an age-hardenable nickel-base alloy being
approved by a very conservative code body. The code case provides rules for the utilization of
wrought sheet, plate, rod, seamless pipe and tube, fittings and forgings under the auspices of
section I of the ASME code. The code case stipulates allowed ranges for annealing and aging of
supplied material. In addition, welding qualification guidelines, and post-weld heat treatment
requirements are provided. Allowable stresses are specified; see Figure 3 for a comparison with
allowable stresses for alloy 617. Alloy 740/740H possesses an 80% advantage over 617 at
1300F (704C), 74% advantage at 1350F (732C) and 62% advantage at 1400F (760C).
PRODUCTION OF COMMERCIAL-SCALE INCONEL ALLOY 740H PRODUCTS
Steam Header Piping Fabrication, Properties and Weldability
Three full-scale commercial heats of the INCONEL alloy 740H composition have been produced
by PCC Energy. Given that at least one ingot from each of these melting campaigns has been at
30 inches (760mm) in diameter or greater, the melting sequence of following vacuum induction
melting (VIM) with vacuum arc re-melting (VAR) was implemented, in the interest of ensuring
optimal ingot macro/microstructure. Table 1 shows individual electrode and ingot sizes and
weights for the three melts of INCONEL alloy 740H completed to date. Ingot enumeration will
consist of 1-A, 1-B, etc.

Figure 3. ASME maximum allowable stress comparison between alloy 617 (UNS N06617) and
alloy 740/740H (UNS N07740). The advantage of alloy 740/740H is 80% at 1300F (704C),
74% at 1350F (732C), and 62% at 1400F (760C).

217

Table 1. Electrode and Ingot Sizes and Weights for Three Melts of INCONEL alloy 740H
Melt
1
1
2
2
3
3

Ingot
A
B
A
B
A
B

VIM Electrode Diameter, In. (mm) VAR Ingot Diameter, In. (mm) Ingot Weight, lb (kg)
27 (690)
30 (760)
21000 (9520)
18 (460)
20 (508)
10000 (4540)
30 (760)
33 (840)
27000 (12240)
18 (460)
20 (508)
10000 (4540)
27 (690)
30 (760)
19500 (8840)
27 (690)
30 (760)
19500 (8840)

Figure 4 shows a photograph of ingot 1-A after VAR. Table 2 shows the chemical analyses
obtained for the ingots. Values shown are the average of head and toe analyses obtained from
drillings at head and toe of the VAR re-melted ingot.
Figure 5 shows chemical analysis results for a top, or head, end ingot slice acquired after
homogenization and controlled cooling, as a function of location for ingot 1-A. Consistency of
composition across the diameter was excellent. No evidence of intermittent solidification or
macro-segregation was detected. The ingot microstructure was also evaluated, after
homogenization and controlled cooling, with results displayed in Figure 6. A coarse, equiaxed
grain structure was observed. Carbides were observed on the grain boundaries, and a dispersion
of coarse gamma prime was observed.
Figure 7 shows chemical analysis results for a head end ingot slice acquired after homogenization
for ingot 2-A. As with slices evaluated for other ingots, no evidence of intermittent solidification
or macro-segregation was detected.
The homogenized. cropped and conditioned ingot 1-A was transferred to Wyman Gordons
Houston facility for extrusion. The 16660 pound (7570 kg) ingot was pierced and blocked using
a 14 kT press, followed by light machining of OD and ID. Final pipe extrusion was performed at
2175F (1190C) using Wyman Gordon Houstons 35 kT vertical press. The as-extruded pipe
size was 14.9 inches (378mm) x 3.47 inches (88mm) x 34.5 feet (10.5m). Figure 8 shows a
photograph of the as-extruded pipe. The pipe was solution annealed at 2050F (1120C) and
water quenched, and then aged at 1472F (800C) for 5 hours and air cooled. Table 3 shows
room-temperature tensile data for solution annealed and aged pipe. Elevated-temperature tensile
data from the front end at mid-wall are shown in Table 4.
Samples from the solution annealed and aged pipe were subjected to long-term exposures at 700,
750 and 800C. Results are shown in Figure 9. Figure 10 shows SEM photomicrographs of
samples exposed for 10000 hours at each temperature, revealing intra-granular gamma prime
phase and globular grain boundary carbides.

218

Figure 4. INCONEL alloy 740H ingot 1-A, having diameter of 30 inches (760mm) and weighing
21000 pounds (9500 kg), and melted via VIM+VAR.
Table 2: Chemical Analysis Results for INCONEL alloy 740H Ingots
Ingot
1-A
1-B
2-A
2-B
3-A
3-B

C
0.045
0.046
0.024
0.023
0.043
0.043

Cr
24.6
24.7
24.6
24.6
24.6
24.6

Mo
0.01
0.01
0.35
0.32
0.35
0.35

Co
20.2
20.2
20.1
20.1
20.1
20.1

Al
1.34
1.33
1.34
1.34
1.33
1.34

Ti
1.36
1.37
1.35
1.34
1.32
1.33

Nb
1.51
1.52
1.47
1.50
1.46
1.46

Figure 5. chemical analysis results for a top end (head end) 30 inch (760mm) diameter ingot
slice from ingot 1-A acquired after homogenization, as a function of location.

219

Figure 6. Photomicrographs from homogenized and controlled cooled 30 inch (760mm)


INCONEL alloy 740H ingot 1-A. Light micrographs from edge, center and mid-radius show
carbide precipitation. Photomacrograph at right shows equiaxed grain structure. SEM
photomicrograph from center shows coarse gamma prime distribution.

Figure 7. Chemical analysis results for a top end 33 inch (840mm) diameter ingot slice from
ingot 2-A acquired after homogenization, as a function of location.
Table 3: Room-Temperature Tensile Data for 14.9 Inch (378mm) OD x 3.47 Inch (88mm) Wall
Pipe Extruded at 2175F (1190C), Solution Annealed at 2050F (1120C) and Aged at 1472F
(800C) for 5 Hours.
Location
in Pipe
Front, Mid-Wall
Front, Mid-Wall
Back, Mid-Wall
Back, Mid-Wall

ksi
105.2
102.3
106.9
108.3

0.2% YS

MPa
725.0
705.0
737.0
747.0

220

ksi
157.3
155.9
160.9
160.6

UTS

MPa
1080
1075
1109
1107

Elong., % R of A, %
31.9
33.4
31.4
29.9
31.5
31.9
32.4
34.3

Table 4: Elevated Temperature Tensile Data for 14.9 Inch (378mm) OD x 3.47 Inch (88mm) Wall
Pipe Extruded at 2175F (1190C), Solution Annealed at 2050F (1120C) and Aged at 1472F
(800C) for 5 Hours.
Test Temperature
F
1112
1202
1292
1382
1472

C
600
650
700
750
800

0.2% YS
ksi
81.3
87.0
83.0
83.1
74.0

UTS

MPa
560
600
572
573
510

ksi
128.3
133.5
129.1
115.9
97.1

MPa
885
920
890
799
670

Elong., % R of A, %
36.1
46.3
29.8
36.0
25.9
21.8
22.5
25.3
22.2
20.4

Figure 8. As-extruded 740H steam header pipe measuring 14.9 inches (378mm) x 3.47 inches
(88mm) x 34.5 feet (10.5m).

Figure 9. Charpy impact results (10mm x 10mm samples) for solution annealed and aged pipe
produced from ingot 1-A, after exposure at the indicated temperatures and times.

221

Figure 10. SEM photomicrographs showing microstructure of pipe samples exposed for 10000
hours (see Figure 9).
Table 5: Chemical Composition of INCONEL alloy 740H Filler Metal
Material

Ni

Cr

Mo

Co

Al

Ti

Nb

Mn

Fe

Si

FM1

0.031

50.81

24.13

0.580

19.76

1.26

1.31

0.93

0.37

0.55

0.19

0.001

Two 24 (610mm) long sections of the steam header pipe produced from ingot 1-A, with final
dimensions consisting of 8.071 (205mm) ID with wall nominally at 3 (76.2mm), were prepared
for a full section welded joint. A full-section girth weld was completed using hot-wire GTAW
using matching filler metal heat FM1 (composition shown in Table 5). The sample was postweld heat treated using ceramic pads with ramping and cooling rates intended to simulate field
heat treatment. A previous publication describes the welding process and heat treatment in
greater detail.8
The macro-etched full cross section can be viewed in Figure 11 as well as the etched
microstructure of a mounted and polished cross section. No micro-fissuring was observed in the
heat-affected-zone (HAZ) or in the weld metal. Side bend testing (4T, Figure 12) and transverse
tensile testing (Table 6) were successfully performed as per ASME section IX.

222

Figure 11. Photograph Showing Hot Wire GTAW Bead Sequence in the 3 (76.2mm) Thick
Steam Header Pipe (left) and microstructure at the fusion line (right).
Samples from the discussed pipe weld have been stress rupture tested, in the post-weld heat
treated condition. Results are shown in Figure 13. So far, to rupture times approaching 10000
hours, a weld strength reduction factor (WSF) of 0.78 has been maintained. ASME code case
2702 calls for a WSF of 0.70 for longitudinal seam welds.

Figure 12. Photograph Illustrating Acceptable Full Section Direct Aged 4T Side bends Sectioned
from the HWGTAW Steam Header Pipe Weldment and a Fractured (outside the weld) Room
Temperature Tensile Specimen

223

Table 6. Room Temperature Tensile Properties for ASME Section IX Qualifications from the Full
Section HWGTAW Steam Header Pipe Weldment
0.2% YS

UTS

Sample
Location

ksi

OD

118.2

815

164.6

1134.9

21.3

22.8

Weld

OD

116.4

802.6

162.8

1122.5

21.2

24.9

Weld

Mid-Wall

110.6

762.6

159.3

1098.4

24.6

22.3

Base

Mid-Wall

109.9

757.8

158

1089.4

20.6

21

Base

ID

108.8

750.2

156.8

1081.1

21.4

22.9

Base

ID

109.7

756.4

157.7

1087.3

24.4

24.5

Base

MPa

ksi

MPa

Elong. %

R of A %

Failure
Location

Samples from the fabricated pipe weld were also exposed at 700, 750 and 800C for up to
10000 samples followed by impact testing and evaluation of microstructure. Impact test results
are shown in Figure 14 and are observed to exceed the values displayed by the base material.
Figure 15 shows weld microstructures after aging for 10000 hours.

Figure 13. Stress rupture data for transverse INCONEL alloy 740H pipe weld samples
compared with data for transverse INCONEL alloy 740 tube weld samples. All samples were
tested after post-weld aging.

Figure 14. Results from impact testing transverse impacts at weld center from alloy 740H header
pipe joined using hot-wire GTAW.
224

Superheater Tubing
Superheater tubing production began with casting a VIM electrode, weighing 10,900 lb (4955 kg)
and measuring 18 inches (457mm) in diameter. This electrode was re-melted via VAR to a
diameter of 20 inches (508mm) ingot 1-B. The resulting ingot was homogenized and open-die
forged at 2175F (1190C) to an 11.375 inch diameter bar. After peeling, billets were cut and
prepared for extrusion by center boring and applying a radius to the front end. Billets were
extruded at 2175F (1190C) to tube shells. Figure 8 shows a shell after extrusion. de-glass,
pickled and solution annealed with water quenching, followed by cold-pilgering. The cycle of
annealing, pickling and pickling was then repeated until the final size of 2 inch (50.8mm) OD x
0.315 inch (8mm) wall. The final annealing temperature was 2125F with water quenching.
Table 7 shows room-temperature tensile properties for three lots of superheater tubing, before and
after aging at 1472F (800C)/4h/air cool; elevated-temperature tensile data for lot 1 is shown in
Table 8.
Samples from lot number 2 (see table 4) were exposed for up to 10000 hours at 700, 750C and
800C and impact tested (samples were exposed in the solution annealed and aged condition).
Figure 16 shows the resulting impact test results (5mm x 10mm sample section). Sample
microstructures are shown in Figure 17.

Figure 15. Alloy 740H header pipe weld microstructure after exposure for 10000 hours at the
indicated temperature.

225

Table 7: Room-Temperature Tensile Data for 2 Inch (50.8mm) Diameter x 0.315 Inch (8mm)
Wall Cold-Pilgered Tubing Solution Annealed at 2125F (1163C) and Aged at 1472F (800C)
for 4 Hours.
Lot #
1
1
2
2
3
3
1
1
2
2
3
3

Condition
SA
SA
SA
SA
SA
SA
SA + Age
SA + Age
SA + Age
SA + Age
SA + Age
SA + Age

0.2% YS
ksi MPa
52 358.5
51.2 353.0
51.2 353.0
51.6 355.8
76.8 529.5
69.9 482.0
105.6 728.1
105.6 728.1
103.9 716.4
106.7 735.7
121.3 836.4
124.0 855.0

UTS
ksi
MPa Elong., % R of A, %
114.3 788.1
57.5
68.0
113.9 785.3
58.5
71.5
113.9 785.3
59.5
66.0
113.9 785.3
54.0
67.0
136.0 937.7
56.0
61.0
125.2 863.3
51.7
65.4
168.7 1163.2
36.0
41.0
168.2 1159.7
36.5
37.5
166.1 1145.3
35.0
31.5
166.3 1146.6
37.0
38.0
175.0 1206.6
34.1
41.1
177.3 1222.5
36.5
40.5

Table 8: Elevated Temperature Tensile Data for 2 Inch (50.8mm) Diameter x 0.315 Inch (8mm)
Wall Cold-Pilgered Tubing Solution Annealed at 2125F (1163C) and Aged at 1472F (800C)
for 4 Hours (Lot 1).
Test Temperature
F
C
1202
650
1202
650
1292
700
1292
700
1382
750
1382
750
1472
800
1472
800

0.2% YS
ksi
MPa
87.6
604.0
89.2
615.0
94
648.1
90.8
626.1
88.5
610.2
88.5
610.2
78.2
539.2
78.2
539.2

UTS
ksi
MPa Elong., % R of A, %
139.8 963.9
28.5
34.5
135.7 935.7
25.5
29.0
131.7 908.1
15.5
19.0
130.9 902.6
17.0
20.5
116.3 801.9
13.0
25.0
113.9 785.3
12.0
19.0
97.2
670.2
10.5
14.5
99.2
684.0
10.5
16.5

Figure 16. Impact strength of solution annealed and aged 2 (50.8mm) OD x 0.315 (8mm) wall
superheater tubing samples (produced from ingot 1-B) exposed for 10000 hours at the indicated
temperature.
226

Figure 17. SEM photomicrographs showing microstructure of superheater tubing samples


exposed for 10000 hours (see Figure 15).
COAL ASH CORROSION TESTING
Simulated coal ash corrosion testing has been performed at 750C, exposing coupons having a
painted-on slurry comprised of a simulated ash containing 6% Fe O -29% CaSO -1.9% Na SO 2

1.9% K SO -39% SiO -22% Al O -0.05% KCl-0.05% NaCl.


2

Ash components were pulverized thoroughly with a mortar and pestle and applied to sample
surfaces as a slurry with acetone. Average mass gain via application of the slurry was 73
mg/cm2. Sample dimesions were approximately 0.75 (19mm) x 0.5 (12.7mm) x 0.25 (6.4mm)
A simulated flue gas comprised of N 15% CO 3.5% O 5% H O 0.1% SO
2

was flowed over the samples at 500 sccm (with platinum catalyst preceding the samples.
Samples of both base metal and weld metal were tested (from GTAW welds made in matching
base plates; alloy 740H samples were fabricated using welding wire made from ingot 1-A).
Cross sections of tested samples were made and evaluated, yielding corrosion depth results
shown in Figure 18 for alloys 740H, 263 and 617. For alloys 740H and 263, both base metal and
weld samples were tested after post-weld aging at 800C (1472F); the alloy 617 samples were
tested in the solution annealed condition (2150F/1177C). Samples were exposed for 1000
hours, with the ash coating re-applied at weekly intervals. Figure 19 shows light
photomicrographs for base metal and weld samples, and SEM photomicrographs for the base
metal samples. Corrosion depths and corrosion product morphologies were similar between base
metal and weld metal samples. External oxide scales were chromium-rich with low levels of
titanium present in all cases. The external oxide scale was thicker in the case of alloys 263 and
617; the alloy 740H samples possessed a thinner, more compact chromia scale. All alloys
developed internal oxides along grain boundaries rich in Al/Ti. Chromium sulfides (verified with
SEM-EDS and backscatter visualization), evident in the micrographs as medium-gray globular
internal precipitates tending to reside on grain boundaries, were much more pronounced in the
227

alloy 617 and 263 samples than in the alloy 740H samples. While the alloy 617 sample exhibited
very significant pitting, the alloy 263 samples exhibited incipient shallow pit development.

Figure 18. Depth of corrosion attack after exposure under simulated coal ash corrosion
conditions at 750C for 1000 hours.

Figure 19. Photomicrographs acquired after exposure under simulated coal ash corrosion
conditions at 750C for 1000 hours; light micrographs of base metal samples are at the top, light
photomicrographs of weld metal samples are at the center, and SEM photomicrographs of base
metal samples are shown at the bottom.

228

CONTINUING CHALLENGES
Significant challenges remain for the commercialization of age-hardenable nickel-base alloys for
applications such as A-USC. Feasibility of fabrication of numerous components must be
demonstrated; wyes, flanges and fittings, and pipe bends, for example. The need for larger forged
components will continue to push the capabilities of nickel-base manufacturers.
The question of relaxation cracking remains a ponderous and pertinent issue, although activity is
beginning to explore this phenomenon for materials of interest. Devising a testing protocol
which is relevant to actual service is still a significant obstacle.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge Ian Dempster and Tom Armstrong of Wyman Gordon who
have contributed greatly to the success of the alloy 740H pipe extrusion. Also, the authors would
like to thank Jim Tanzosh, Joe Dierkscheide and Doug Ziegler of Babcock and Wilcox for their
significant contribution in the completion of circumferential pipe welds and tube-to-pipe welds.
REFERENCES
1. R. Viswanathan, J. P. Shingledecker, J. Hawk and M. Santella, Proceedings of the 34th
International Technical Conference on Clean Coal and Fuel Systems, Coal Technology
Association, Clearwater, FL, May 31 to June 4 (2009).
2. R. Blum and R. W. Vanstone, Materials Development for Boilers and Steam Turbines
Operating at 700C, Proceedings of the Sixth International Charles Parson
Turbine Conference, Institute of Materials, Minerals and Mining, London, pp 489-510
(2003).
3. M. Fukuda, et al., Advanced USC Technology Development in Japan, Proceedings of the
Sixth International Conference in Advances in Materials Technology for Fossil Power
Plants, Santa Fe, NM, USA, August 31-September 3, 2010, EPRI Report Number 1022300
(2010).
4. R. Blum, S. Kjaer, and J. Bugge.Development of a PF Fired High Efficiency Power Plant
(AD700). Proceedings of the Riso International Energy Conference Energy Solutions for
Sustainable Develoment, Denmark, 2007 Riso-R-1608, EN (2007).
5. R. Viswanathan, A. F. Armor, and G. Booras, Power, April 2004, pp 42-49.
6. Concentrating Solar Power Program Review 2013, National Renewable Energy
Laboratory, Phoenix, Arizona, April 23-25, 2013.
7. V. Dostall, M. J. Driscoll, and P. Hejzlar, A Supercritical Carbon Dioxide Cycle for Next
Generation Nuclear Reactors, Report # MIT-ANP-MR-100, Thesis for Doctor of Science in
Nuclear Engineering at MIT, March 10, 2004.
8. Baker, B. A., Gollihue, R. D., deBarbadillo, J. J., Welding and Repair Technology for Power
Plants, Tenth International EPRI Conference, EPRI, Marco Island, FL, June 26-29 (2012).

229

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

CREEP-RUPTURE PERFORMANCE OF INCONEL ALLOY 740 AND


WELDS
J.P. Shingledecker
Electric Power Research Institute, Charlotte, NC USA

ABSTRACT
Inconel alloy 740/740H (ASME Code Case 2702) is an age-hardenable nickel-based alloy
designed for advanced ultrasupercritical (A-USC) steam boiler components (superheaters,
reheaters, piping, etc.). In this work, creep testing, beyond 40,000 hours was conducted a series
of alloy 740 heats of varying product form, chemistry, and grain size. Long-term creep-rupture
strength was found to be weakly dependent on grain size. Analysis of the time-to-rupture data
was conducted to ensure long-term strength projections and development of ASME stressallowables. Testing was also conducted on welded joints in alloy 740 with different filler metal
and heat-treatment combinations. This analysis shows the current weld strength reduction factor
of 30% (Weld Strength Factor of 0.70) mandated by ASME Code Case 2702 is appropriate for
740 filler metal but other options exist to improve strength. Based on these results, it was found
that alloy 740 has the highest strength and temperature capability of all the potential A-USC
alloys available today.
INTRODUCTION & BACKGROUND
INCONEL alloy 740/740H (ASME Boiler & Pressure Vessel Code Section I Code Case 2702
[1]), hereafter referred to as alloy 740, is an age-hardenable nickel-based superalloy which is
purposely designed for advanced ultrasupercritical (A-USC) steam boiler applications at highstress and temperatures where creep will be the dominate deformation mode [2]. Figure 1 shows
that alloy 740 has the highest allowable stress (highest strength) of any section I alloy at 700 to
800C. Code Case 2702 limits alloy 740 to a maximum use temperature of 800C, but additional
testing is underway to extend this limit to 850C.
A-USC steam boilers with operating temperatures up to 760C will help to increase efficiency and
decrease emissions of all effluents, including CO2, in coal-fired power plants by up to 25% in
comparison to current technology [3,4]. The typical microstructure of alloy 740 is a gamma
matrix containing gamma prime precipitates and very little precipitation at the grain boundaries.
After high-temperature aging, the gamma prime coarsens, and grain boundary M23C6 carbides, Gphase (a complex silicide), and the plate-like eta phase forms [5,6]. Studies on alloy 740 have
shown that this microstructure, and in particular the formation of eta phase, is sensitive to alloy
chemistry [7].
Early research on alloy 740 showed that the formation of a small amount of eta phase, which
grows by consuming gamma prime precipitates, is not detrimental to the creep strength or
ductility at 750C [8].Later work suggested the alloys creep strength and ductility over a larger
range of temperatures was relatively insensitive to chemical compositional differences, but grain
size did have some notable effect on creep-rupture strength [9]. Successful welding of the alloy in
thin and thick sections has been accomplished [10, 11], but the current code case applies a weld
230

strength reduction factor (WSRF) or weld strength factor (WSF) of 0.70 [1] for seam-welded
components. A reduction in creep strength of nickel-based alloys has been observed for other AUSC nickel-based alloys including Haynes 230 [12] and alloy 617 [13], but a thorough analysis of
the available cross-weld creep-rupture data on alloy 740 is required. This paper provides an upto-date analysis of some of the research first reported in [9] with additional analysis of cross-weld
creep-rupture data for multiple filler metals and welding processes.

Figure 1. ASME Section I Allowable Stress Values for Code Case 2702 (Inconel 740/740H)
Compared to Other Nickel-based Alloys (617 and 230), 347H stainless steel, and grade 91
EXPERIMENTAL PROCEDURE
Various heats of alloy 740 were evaluated representing a range of chemistries and product forms
as shown in Table 1. In some cases, the solution annealing (SA) heat-treatment temperature was
varied to modify the grain size of the alloy. All base materials were given the standard aging
heat-treatment of 760 to 800C for 4 to 16 hours per code case 2702. Welds were made and
supplied by Babcock & Wilcox, Barberton, OH USA, using gas tungsten arc welding (GTAW),
gas metal arc welding (GMAW), and hot-wire narrow groove GTAW (hot-wire TIG). Table 2
provides the relevant welding details and post-weld heat-treatments given to the weldments.
Smooth bar creep rupture specimens with a 6.35mm diameter and a 31.75mm gauge length were
utilized for the non-welded base material tests, and specimens with a 6.35mm diameter and a
57.15mm gauge length with the weld centered in the specimen gauge length were utilized for
cross-weld creep rupture testing. Testing was performed at temperatures from 600 to 875C in
accordance with ASTM E139 in lever-arm type creep machines with most test run until failure.
At the time of writing this paper, the longest failure durations exceeded 20,000 hours for base
metal and 15,000 hours for cross-weldments with base metal specimens still running beyond
45,000 hours.
Microstructural analysis was conducted using light optical microscopy (OM) on etched samples
using a solution of 40 ml H20, 40 ml HN03, and 20 ml HF. Grain size was determined by the
manual mean lineal intercept method in ASTM E112. Failure location was evaluated by
specimen observation and in some cases through OM of polished metallurgical mounts of
specimens after testing.
231

Table 1. Alloy 740 Heat Descriptions


Composition wt% (Ni balance)
Heat (SA Grain Size
C
Temp, C)
A (1120)

82.4

A (1190)

Fe

Si

Cr

Al

Ti

Co

Mo

Nb

0.03 0.28 0.42 0.0010 0.54 24.43 0.94 1.81 20.00 0.55 1.98 0.005 0.0030

165

B (1120)

0.03 0.26 0.46 0.0010 0.53 24.38 0.98 1.77 19.90 0.50 1.97 0.005 0.0043

188

C (1120)

0.03 0.26 0.46 0.0010 0.54 24.34 0.97 1.78 19.80 0.50 1.99 0.005 0.0037

127

D (1200)

0.03 0.27 1.02 0.0002 0.45 24.31 0.75 1.58 19.63 0.52 1.83 0.003 0.0006

169

E* (1190)
E (1120)

Mn

(m)

89.6

0.06 0.30 0.69 0.0060 0.48 24.86 1.20 1.41 19.90 0.53 2.05 0.004 0.0010

**113

F (1121)

0.04 0.31 1.05 0.0100 0.30 24.28 1.30 1.50 19.88 0.53 1.57 0.002 0.0007
-

0.30 1.07

0.20 24.35 1.28 1.45 20.08 0.53 1.53 0.002

*Material furnished in hot-rolled condition


** Bimodal grain size distribution, average grain size reported (center region grain size = 92.4
and outer sample region grain size = 145.1)

ID
740GMAW

Base
Metal
B

263GMAW

740GTAW
740GTAWSA
282GTAW

A
(1120)
G

Table 2. Alloy 740 Weld Descriptions


Product
Welding
Filler
Post Weld Heat
Form
Process
Metal
Treatment
15.9mm
GMA
740
4hrs-800C
plate
15.9mm
GMA
263
4hrs-800C
plate
50.8mmOD
GTA
740
4hrs-800C
10mm WT
1hr-1120C air cool,
tube
4hrs-800C
38.1mm
Hot-Wire
Haynes
4hrs-800C
plate
TIG (GTA)
282

232

RESULTS
Base metal rupture data and analysis
To evaluate the large range of times and temperatures, and to provide an estimate of long-term
rupture life, the base metal creep-rupture data were analyzed using the Larson Miller Parameter
(LMP). To find a best fit to the data, a regression to minimize the error in time to rupture (tr) was
performed using the following:

equation 1

Stress (MPa)

where C is the LMP constant, Ax are the regression coefficients, is stress (MPa), and T is
temperature in K. Figure 1 is a Larson-Miller (LMP) plot for all the base metal creep-rupture data
including ongoing tests (open symbols) using the best fit LMP constant of 19.392. Table 3
provides the regression coefficients. Inspection of the figure shows the data are tightly grouped
and fall within a +/- 20% scatterband on stress. The only exceptions are two data points which
are short-term tests at 875C. To analyze the goodness of the fit and its ability to accurately
predict long-term behavior, the rupture data are plotted as actual versus predicted life (equation 1)
in Figure 2. The figure shows excellent agreement and all the data fall within a factor of +/-2 on
life. Inspection of the ongoing test data (open symbols) show the prediction of long-term
performance is very good with all tests exceeding 30,000 hours meeting or exceeding predicted
life. A fit to the rupture data gives a slope of 0.93 which is close to unity (perfect fit) but is
slightly conservative for long-term strength predictions. Based on these results, the equation was
found to be accurate in the assessment of long-term material performance. Earlier research [8,9]
found that alloy 740 was insensitive to microstructural variation due to chemistry, but some effect
of grain size was noted. Figure 3 segments the heats of tested material by grain size. The figure
clearly confirms the trend that the lower scatterband of creep strength is occupied by heats with
finer grain size while slightly coarser grain sizes results in average or above average life with
long-term tests exceeding the average prediction.
A (1120)
A (1190)
B (1120)
C (1120)
D (1200)
E (HR1190)
E (1120)
F (1121)
Average
- 20%

100

Open Symbols = Ongoing Test


10
20000 21000 22000 23000 24000 25000 26000 27000 28000

LMP C=19.392

Figure 2. Optimized Larson-Miller-Parameter (LMP) plot for Inconel 740/740H creep-rupture


test data segmented by heat. Open symbols on ongoing (not ruptured) tests.
233

Table 3. Regression coefficients determined for equation 1


Regression
740 BM
740 WM (GTAW & 740WM (WM
Constants
GMAW Aged)
failures only)
C
19.392
16.513
17.0245
A1
2.336e-4
6.044e4
4.478e4
A2
5.532e3
-4.612e4
-2.457e4
A3
-2.0.65e-3
1.9481e4
1.005e4
A4
-1.027e2
-3.066e3
-1.709e3

Predicted Rupture Life (hrs) - eq. 1

1000000

A (1120)
A (1190)
B (1120)
C (1120)
D (1200)
E (HR1190)
E (1120)
F (1121)

Open Symbols = Ongoing Test

100000

10000

1000
Slope = 0.93
100

+/- factor of 2
10

10

100

1000

10000

100000

1000000

Rupture Life (hrs)

Figure 3. Predicted (equation 1) versus actual rupture life for alloy 740. Note, the predicted life
is based off an earlier assessment of the data. All non-ruptured specimens (open symbols) are
within a +/-2 scatterband on expected life with most ongoing tests (including those over 30,000
hours) exceeding predicted life.

Predicted Rupture Life (hrs) - eq. 1

1000000

Grain Size < 100m


Grain Size > 100m

Open Symbols = Ongoing Test

100000

10000

1000
Slope = 0.93
100

+/- factor of 2
10

10

100

1000

10000

100000

1000000

Rupture Life (hrs)

Figure 4. Predicted versus actual failure life segmented by grain size

234

Cross-weld testing results and analysis

Stress (MPa)

All the cross-weld time to rupture results for the welds in Table 2 are plotted against the average
alloy 740 base metal (BM) behavior in Figure 4. All but one data point (black square at LMP =
~23,500) for a variety of weld filler metals and heat-treatment conditions fall between the average
expected behavior and -30% on stress. A 30% reduction in stress is equivalent to a WSF of 0.70.
740GMAW
263GMAW
740GTAW
740GTAW-SA
282GTAW
740BM Ave.
-30%

100

20000 21000 22000 23000 24000 25000 26000 27000 28000

LMP C=19.392

Figure 5. Alloy 740 Cross-Weld Rupture Results


Using the LMP of 19.392 and the base metal fit (Table 3), the stress for rupture (r) data in Figure
4 were analyzed to determine the specific WSF for each data point. These data are plotted as a
function of rupture time in Figure 5 and Table 4 provides the average of the measured WSFs for
each weldments. The data show that the 740GMAW and 740GTAW specimens had a WSF of
slightly greater than 0.70, but that the application of a solution annealing heat-treatment after
welding and prior to aging improved this to close to 0.90. The use of alternate filler metals, alloy
263 and Haynes 282, also improved the strength of the welded joint to 0.82 and 0.85,
respectively.
WSF = Weldment-Actual
/ BM-Average
[LMP=19.392]
tr
tr

1.0

740GMAW
263GMAW
740GTAW
740GTAW-SA
282GTAW

0.9

0.8

0.7

0.6

0.5

10

100

1000

10000

100000

Weldment Rupture Life (hrs)

Figure 6. Inconel 740 measured cross-weld Weld Strength Factors (WSF) as a function of test
time using average Inconel 740 base metal

235

To examine the behavior of the 740 filler metal welds in more detail, the data are plotted in Figure
6 along with information on the failure location. The weld given the solution annealing heattreatment failed in the base metal outside the weld. Because these solution-anneal weldments
were made in heat A with the finer grain size, a WSF of ~0.90 is equivalent to base material heats
in the lower scatterband of the data. In contrast, the GMAW and GTAW welds all failed in the
weld metal with the exception of one test which had a reported heat affected zone (HAZ) failure.
This location was reported based on visual observation and detailed OM was not complete at the
time of writing this paper. Based on these findings, the 740GMAW and GTAW data (note these
data were aged following welding) were combined and analyzed in same manner as the base metal
to determine optimized LMP fits. Table 3 provides the results of this analysis for all the data and
with the HAZ data removed because a different failure mechanism or failure due to a welding
defect may be influencing results. Tables 5 and 6 compare the predicted stress to produce a
rupture in the base metal and a 740 weld metal for various times up to 100,000 hours based on the
LMP constants in Table 3. The optimized fits of the data show, in some cases, a slightly more
pessimistic value for the WSFs compared to the average data in Table 4 since the LMP constants
and equations vary between the base metal and the weldments data. However, the values do not
show any consistent trend with temperature or test time. The values of the WSF calculated from
optimized fits and reported in Table 6 are 0.70+/-0.3 which suggests the value of 0.70 is
appropriate.
Table 4. Average of measured WSFs (Figure 5) for Inconel 740/740H
ID
WSF
740GMAW
0.71
263GMAW
0.82
740GTAW
0.74
740GTAW0.88
SA
282GTAW
0.85

740GMAW

Stress (MPa)

740GTAW
740GTAW-SA
740BM Ave.
-30%

100

HAZ Reported
Failure
Solid symbol = weld metal failure
Open symbol = base metal failure

20000 21000 22000 23000 24000 25000 26000 27000 28000

LMP C=19.392

Figure 7. Inconel 740 Cross-Weld Rupture Results for 740 Filler Metals and Failure Locations

236

Table 5. Comparisons between average Inconel 740 data and an optimized fit to a combined
dataset of cross-weld data on 740GMAW and 740GTAW (aged condition, all failure modes)
Time to
Rupture
Temp.
(C)
700
750
800

10,000 hours (MPa)


BM
307.5
197.9
114.6

Weldment
211.6
132.1
82.3

30,000hours (MPa)
WSF
0.69
0.67
0.72

BM
260.5
160.3
84.8

Weldment
171.8
105
66.3

100,000 hours (MPa)


WSF
0.66
0.66
0.78

BM
214.1
123.7
84.8

Weldment
134.4
82.4
-

WSF
0.63
0.67
-

Table 6. Comparisons between average Inconel 740 data and an optimized fit to a combined
dataset on 740GMAW and 740GTAW weldments (aged condition, only weld metal failures)
Time to
Rupture
Temp.
(C)
700
750
800

10,000 hours (MPa)


BM
307.5
197.9
114.6

Weldment
215.5
136.1
83.2

30,000hours (MPa)
WSF
0.70
0.69
0.73

BM
260.5
160.3
84.8

Weldment
177.1
108.5
65.5

100,000 hours (MPa)


WSF
0.68
0.68
0.77

BM
214.1
123.7
84.8

Weldment
141.1
84.3

WSF
0.66
0.68

b.

c.
Figure 8. OM of failure (a, b) and fusion line (c) in 740GTAW after1668.9hours at 750C

237

To examine failure modes in more detail, OM was conducted on cross-sections of failed samples.
Figure 7 provides an overview of the typical failure mode observed in the 740 weld metal tests
where creep failure occurred in the weld metal (a) with no evidence of distress in the HAZ (c).
Higher magnification images suggest the interdendritic regions (final region to solidify) had a
large population of creep cavities and microcracks (b). This is consistent with more detailed
ongoing research which shows cavitation initiates in the interdendritic regions having large areas
denuded in gamma prime [14] which is similar to reported initiation sites in base metal [9].

Solution annealing after welding restored base metal strength and moved the failure location from
the weld metal to the base metal. Figure 8 shows the change in microstructure in the GTAW
weldments after creep testing. In this case, it appears the large dendritic weld microstructure was
completely recrystallized resulting in a wrought weld structure after solution annealing. This
suggests very high residual stresses in this tube to tube butt weld. Creep damage was not
observed in the recrystallized weld metal which exhibited a coarser grain size than the base metal.
This is again consistent with the base metal analysis which showed larger grains are beneficial for
creep strength. This data also suggests that the welding processes utilized did not have significant
defects and that they were controlled to minimize loss of specific strengthening elements. If weld
quality were poor, strength restoration via solution annealing would not be possible.

Figure 9. Change in microstructure of 740GTAW weld (left) with application of a solution


annealing heat-treatment (right) after testing at 800C and 180MPa for ~300-600hrs. Fusion
line is indicated by dashed white line.
Alternative weld filler metals offer a second approach (compared to heat-treatment) to improve
weldment strength because failure (see Figure 7) of 740 weldments are due to poor filler metal
strength. The alloy filler 263 failures were all reported in the weld metal similar to the 740 filler
metals, but overall strength was slightly better. Figure 9 is an OM of one of the 263 failures.
Similar to the other 740 weldments, the base metal did not exhibit creep damage and cavitation
was observed in the weld metal. However, a significant difference was noted. In the weld metal
adjacent to the fusion line a region of finer grains/refinement compared to the rest of the weld
metal (Figure 9 left) was intermittently observed. Creep damage was concentrated in these weld
regions (Figure 9 right) suggesting the failure mechanism may be different for the examined filler
materials and process improvements may increase the strength of the filler metal.

238

Figure 10. Cross-sectional micrographs for failure in 740 welded with 263 after 250.2hrs at
750C. Cavitation is observed in a refined region of the weld metal near the fusion zone.
DISCUSSION
Analysis of the current base metal testing shows an optimized Larson Miller Parameter analysis
can be used to accurately represent the data and to make extrapolations for longer life. Long-term
ongoing tests, beyond 40,000 hours, suggest the equation provided in this paper will be slightly
conservative for such purposes. At 750C, the analysis shows the 100,000 hour rupture life on
alloy 740 is greater than 120MPa which is in excess of the generally accepted criteria of 100MPa100,000 hour rupture life needed for power steam boiler components [2]. At 700C, the value is
greater than 210MPa which confirms the stress allowable values (Figure 1). In the interest of AUSC powerplant construction, alloy 740 has the highest strength of any ASME-approved alloy for
this application.
Cross-weld creep-rupture testing shows that 740 weldments generally have inferior creep strength
to the base metal. For 740 matching filler metals given a standard aging heat-treatment after
welding, a WSF of 0.7 was found, through multiple analysis methods, to be an accurate measure
of long-term creep performance. No clear trend with test time or temperature was observed
suggesting a single factor can be applied in the creep regime. Microscopy shows the interdendritic
regions are prone to creep cavitation and microcracking and at failure no creep damage was
observed in heat affected zone (HAZ) or base metal. More detailed microscopy is required to
understand the failure mechanism in the weld metal and ultimately improve the strength of the 740
weld metal. Solution annealing offers one path for improved weld metal strength. This research
shows that base metal strength can be achieved through application of a standard solution
annealing heat-treatment which results in a stronger weld metal and moves the failure location to
the base metal. A second option to improve the cross-weld strength of alloy 740 weldments is to
use an alternative filler metal. In this work, both alloy 263 and Haynes 282 filler metals improved
the strength of the welded joint to within an acceptable -20% scatterband of the base material.
However, neither alloy 263 nor Haynes 282 restored cross-weld strength in alloy 740 to that of
the base metal. Examination of the 263 joints revealed that creep cavitation was observed in fine
grained regions (associated with weld metal refinement) in the weld metal suggesting process
improvements may be warranted for this combination.
Currently, ASME only applies WSRF (WSF) to longitudinally seam welded components which
are not envisaged for A-USC designs. A WSF less than 1.0 is standard for many power plant
materials including extensively used ferritic steels such as grade 91 which has a WSF between 0.5
and 0.9 (due to the Type IV failure mechanism). Other studies of nickel-based alloys such as
239

Haynes 230 [12] and alloy 617 [13] show WSF around 0.8. Therefore, from a practical
standpoint, a WSF of 0.7 can be successfully utilized in design by avoiding seam welded
components. If seam welds are required, the data clearly indicate alternative filler metals or
solution heat-treatment are viable options for improving weldments strength.
CONCLUSIONS
Analysis of a large set of creep-rupture data on Inconel alloy 740/740H of varying compositions
showed the alloy has sufficient strength to achieve desired lifetimes for A-USC boiler
components. Optimized time-temperature parameter analysis was shown to give accurate results
for the base metal dataset and ongoing tests are meeting predictions. Extensive cross-weld
rupture tests confirm the use of a 0.70 WSRF in the current ASME code case for matching filler
metal made with various processes. Although not needed in most designs, if higher weldments
strength is required, the data show alternate filler metals and solution heat-treatment are both
options available to the end user and manufacturer. Microstructural analysis of failed cross-weld
specimens showed different behavior for different filler metals and heat-treatments which
illuminate the needs for further study. Future work should include microstructural analysis of
weldments samples to understand the failure mechanisms, continued long-term base metal testing,
and interrupted tests to improve understanding of damage development.
ACKNOWLEDGMENTS
The author wishes to acknowledge the significant contributions of Oak Ridge National Laboratory
(ORNL) in performing the long-term creep testing in this work including the leadership of R.
Swindeman (who taught me all I know about creep testing), P. Torterelli and M. Santella and the
experimental work of B. Sparks, J. Moser, F. Adames, and T. Geer. The guidance and support of
the U.S. DOE/OCDO A-USC Steam Boiler consortium R. Purgert (EIO), P. Rawls (NELT), B.
Romanosky (NETL), and R. Conrad (DOE) is appreciated. Special thanks goes to J. Sanders, J.
Siefert, and J. Tanzosh (B&W) who provided the welds. Finally, thank you to J. Siefert for
helpful discussions on the preparation of this manuscript.
REFERENCES.
[1] Case 2702 Seamless Ni-25Cr-20Co Material Section I. Cases of the ASME Boiler and
Pressure Vessel Code. BPV-Supp.7. 2011 American Society of Mechanical Engineers
[2] J.P. Shingledecker, I.G. Wright: Proceedings to the 8th Liege Conference on Materials for
Advanced Power Engineering 2006, Forschungszentrum Jlich GmbH, 2006.107-120.
[3] R. Viswanathan, R. Purgert, U. Rao: Materials for Advanced Power Engineering 2002,
Proceedings Part II, Forschungszentrum Julich GmbH, 2002. 1109-1129
[4] R. Viswananthan, J.F. Henry, J. Tanzosh, G. Stanko, J. Shingledecker, B. Vitalis, R. Purgert:
J. .Mater. Eng. Perform., 2005, 3., Vol. 14. 281-292
[5] N.D. Evans, P.J. Maziasz, R.W. Swindeman, G.D. Smith: Scr. Mater., 2004, Vol. 51. 503507.
[6] S. Zhao, X. Xie, G.D. Smith, S.J. Patel: Mater. Sci. and Eng., 2003, Vol. A355, 96-105.
[7] X. Xie, S. Zhao, J. Dong, G.D. Smith, B.A. Baker, S.J. Patel: Mater. Sci. Forum, 2007, Vols.
561-565. 471-476.

240

[8] J.P. Shingledecker, G.M. Pharr. Met. Mat. Trans. A, Vol. 43 (6) 2012. 1902-1910.
10.1007/s11661-011-1013-4
[9] J.P. Shingledecker, N.D. Evans, G.M. Pharr. Influences of Composition and Grain Size on
Creep-Rupture Behavior of Inconel Alloy 740. Materials Science and Engineering A. 578
(2013) 277-286. 10.1016/j.msea.2013.04.087
[10] J.A. Siefert, J.M. Tanzosh, J.E. Ramirez. Proceedings to the 6th International Conference on
Advances in Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August
30-September 4, 2010. EPRI, March 2011: 1022300. Distributed by ASM International.
1045-1066.
[11] B.A. Baker, R.D. Gollihue. Proceedings to the 6th International Conference on Advances in
Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August 30-September
4, 2010. EPRI, March 2011: 1022300. Distributed by ASM International. 96-109.
[12] R. Viswanathan, J. Shingledecker, J. Hawk, and S. Goodstine, Proceedings Creep & Fracture
in High Temperature Components, 2nd ECCC Creep Conference, April 21-23, 2009 ( Zurich,
Switzerland), DEStech Publications, Inc., 2009. 31-43
[13] A. Klenk, M. Speicher, K. Maile. Weld Behavior of Martensitic and Ni-based Alloys for
High Temperature Components. Procedia Engineering 5 (2013) 414-420.
[14] D. Bechetti, J. Dupont, Lehigh University, unpublished research, June 2013.
LEGAL NOTICE/DISCLAIMER
This report was prepared by J. Shingledecker (EPRI) pursuant to a Grant partially funded by the
U.S. Department of Energy (DOE) under Instrument Number DE-FG26-0 1 NT 41175 and the
Ohio Coal Development Office/Ohio Department of Development (OCDO) under Grant
Agreement Number CDO/D-OO-20 (now D-05-02A). NO WARRANTY OR
REPRESENTATION, EXPRESS OR IMPLIED, IS MADE WITH RESPECT TO THE
ACCURACY, COMPLETENESS, AND/OR USEFULNESS OF INFORMATION
CONTAINED IN THIS REPORT. FURTHER, NO WARRANTY OR REPRESENTATION,
EXPRESS OR IMPLIED, IS MADE THAT THE USE OF ANY INFORMATION,
APPARATUS, METHOD, OR PROCESS DISCLOSED IN THIS REPORT WILL NOT
INFRINGE UPON PRIVATELY OWNED RIGHTS. FINALLY, NO LIABILITY IS
ASSUMED WITH RESPECT TO THE USE OF, OR FOR DAMAGES RESULTING FROM
THE USE OF, ANY INFORMATION, APPARATUS, METHOD OR PROCESS DISCLOSED
IN THIS REPORT

241

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

EVALUATION OF HIGH TEMPERATURE STRENGTH OF A NI-BASE


ALLOY 740H FOR ADVANCED ULTRA-SUPERCRITICAL POWER
PLANT
Shengde Zhang and Yukio Takahashi
Central Research Institute of Electric Power Industry, 2-6-1 Nagasaka, Yokosuka-shi,
Kanagawa-ken 240-0196 Japan

ABSTRACT
High temperature strength of a nickel-based superalloy, Alloy 740H, was investigated to evaluate
its applicability to advanced ultrasupercritical (A-USC) power plants. A series of tensile, creep
and fatigue tests were performed at 700C, and the high temperature mechanical properties of
Alloy 740H was compared with those of other candidate materials such as Alloy 617 and Alloy
263. Although the effect of the strain rate on the 0.2% proof stress was negligible, the ultimate
tensile strength and the rupture elongation significantly decreased with decreasing strain rate, and
the transgranular fracture at higher strain rate changed to intergranular fracture at lower strain
rate. The time to creep rupture of Alloy 740H was longer than those of Alloy 617 and Alloy 263.
The fatigue limit of Alloy 740H was about half of the ultimate tensile strength. Further, Alloy
740H showed greater fatigue strength than Alloy 617 and Alloy 263, especially at low strain
range.
INTRODUCTION
For enhancement of the thermal efficiency and reduction of CO 2 emission in coal-fired power
plants, an advanced ultra-supercritical (A-USC) power plant in which steam condition is raised
from 600C in the conventional USC power plant to 700C or higher, is under development as a
worldwide activity [1-3]. Conventional ferritic and austenitic steels cannot fulfill the requirement
of creep strength at these temperatures and Ni-based alloys with higher creep strength and greater
corrosion resistance are needed to attain the above target.
Alloy 740 is a relatively new Ni-based superalloy developed on the basis of Alloy 263 at Special
Metals Corporation as a candidate material for boiler tubing and piping in A-USC power plants
[4]. Compared with Alloy 263, the alloy increases the fraction of Cr and decreases the fraction of
Mo for improving corrosion resistance, and adds the strengthening element Nb for high
temperature strength. This alloy exhibits the stress rupture strength over 100MPa at 750C and a
good oxidation and hot corrosion resistance at elevated temperatures [4]. However,
microstructure of the alloy changes during prolonged aging or creep loading, such as gamma
prime () coarsening, to eta () phase transformation and formation of G-phase [5, 6]. The
structure stability of this alloy can be improved by the adjustment of Al, Ti and Si level reported
by Zhao et al. [7], and a new modified alloy was developed as Alloy 740H. Overview of available
literature indicates that the microstructure characteristic of Alloy 740 has been investigated in
detail. However, limited efforts have been made on the topic of high temperature strength for
Alloy 740 or Alloy 740H.
242

The objective of this paper is to study the high temperature strength of Alloy 740H, for evaluating
its applicability to A-USC power plants. Tensile and creep tests were performed at 700C and
effect of strain rate on the tensile properties and creep fracture behaviors was discussed, followed
by the fracture surface observation. Both strain- and load-controlled fatigue tests were also carried
out at 700C and the fatigue strength was discussed in relation to the tensile properties. In
addition, the high temperature strength of Alloy 740H was compared with the data of other
candidate materials such as Alloy 617 and Alloy 263.
EXPERIMENTAL PROCEDURE
The material tested was Alloy 740H piping (205mmID72.5mmt150mmL) of which the
chemical composition and heat treatment are listed in Table 1. The material was received in
solution annealed and aged condition, following the ASME Code Case 2702. Microstructure of
the material is shown in Fig. 1. Almost equiaxed grain structures were observed in both axial and
circumferential sections, indicating that there is no effect of rolling process on the microstructure.
The average grain size was found to be about 150m by a cutting method according to the
JIS0551 [8].
Figures 2(a) and 2(b) show the geometry and dimensions of tensile, creep and fatigue specimens.
Solid bar specimens with a gage diameter of 6mm and 30mm gage length shown in Fig. 2(a) were
used in the tensile and creep tests. The fatigue specimen was a round bar whose diameter and
length at the test section were 8mm and 16mm as shown in Fig. 2(b). All the specimens were
machined in a way that the loading direction is parallel to axial direction of the material.
Table 1: Chemical composition of Alloy 740H (wt%).
C
Cu
0.045
0.031
Ti
Nb
1.36
1.52
Solution annealed and aged

Cr
24.7
Mn
0.246

Mo
0.006
Fe
0.238

(a) Axial section

Co
20.2
Si
0.137

(b) Angular section

Figure 1: Microstructure of the material tested.

243

Al
1.33
Ni
Bal.

22 (Tensile)
15 (Creep)

30

10

10

22 (Tensile)
15 (Creep)

21.5

O8

O 14

O 6.3

O6

M10 (Tensile)
M12 (Creep)

16

M18P2.5

103

(a) Tensile and creep specimen

(b) Fatigue specimen

Figure 2: Shape and dimensions of specimen tested.

10 %/s
-2

-3

10 %/s
10 %/s

10 %/s

-1

10 %/s

-4

T=700C

(a) up to failure

(b) up to 1% strain

Figure 3: Stress-strain curves at 700C.


A universal testing machine (SHIMADZU Autograph/AG-I, 100kN) and a creep testing machine
with a level system (lever ratio of 1:10) were used in tensile and creep tests, respectively. The
experimental apparatus used in the fatigue tests was an electric-mechanical uniaxial fatigue
machine (INSTRON 8861). Temperatures of all the specimens were raised by an electric
resistance furnace and the temperature variation in the gage part was within 5C.
Tensile tests were performed at five strain rates (100, 10-1, 10-2, 10-3 and 10-4%/s) at 700C. Creep
and fatigue tests were also conducted at 700C. A fully reversed symmetrical triangular strain
wave with a strain rate of 0.1%/s was applied in the fatigue tests, where the gage length for strain
measurement was 12.5mm. The number of cycles to failure (N f ) was defined as the cycle of 25%
drop of tensile stress amplitude from the stress amplitude at half life. Load-controlled fatigue tests
with a nominal stress rate of 500MPa/s were also performed to determine fatigue limit of the
material.
RESULTS AND DISCUSSION
Tensile Properties
Figure 3 shows stress-strain curves of Alloy 740H at five strain rates at 700C. The alloy showed
clear strain hardening behavior and strain softening presumably caused by a local necking. There
was no large difference in stress-strain curves at different strain rates until yielding as shown in
244

Fig. 3(b). Comparing Figs. 3(a) and 3(b), the strain rate mainly affected the strain range from
yielding to necking (strain hardening range) in stress-strain curve. That is, the strain hardening
range decreased significantly with decreasing the strain rate.
Figure 4 presents variation of Youngs modulus (E), 0.2% proof stress ( 0.2 ) and ultimate tensile
strength ( B ) with strain rate at 700C. There was a little difference in 0.2% proof stress at the
five strain rates but Youngs modulus slightly decreased with decreasing the strain rate. Ultimate
tensile strength, on the other hand, decreased monotonically as the strain rate decreased. These
results indicate that ultimate tensile strength was more sensitive to the strain rate than Youngs
modulus and 0.2% proof stress, due to a large stress dependency of recovery rate.
In order to discuss the effects of strain rate on the ductility of Alloy 740H, rupture elongation ()
and reduction of area () are plotted against the strain rate in Fig. 5. Elongation as well as
reduction of area decreased as the strain rate decreased. Generally, increasing the strain rate
increases the strength, but decreases the ductility at elevated temperatures in most of carbon steels
and low-alloy steels [9]. However, Alloy 740H showed the positive strain rate dependency of both
ultimate tensile strength and elongation, which may have resulted from the change in fracture
mode due to the strain rate.

B
0.2
E
T=700C

Figure 4: Variation of E, 0.2 and B with strain rate at 700C.


T=700C

Figure 5: Variation of and with strain rate at 700C.

245

Figure 6 shows fracture surfaces observed at three strain rates (100%/s, 10-2%/s and 10-4%/s) at
700C. At strain rate of 100%/s, transgranular fracture with many dimples was dominant and
intergranular fracture was partly found as seen in Fig. 6(a). When the strain rate decreased to 102
%/s as shown in Fig. 6(b), both transgranular and intergranular fracture surfaces were observed
with the ratio of intergranular fracture being larger than that at the strain rate of 100%/s. In
contrast, almost the whole fracture surface was found to be intergranular at the strain rate of 104
%/s. These results mean that the fracture mode was changed from transgranular to intergranular
with decreasing the strain rate, which caused the decrease in ductility with decreasing the strain
rate shown in Fig. 5.

500m

100m
(a) 100%/s

(b) 10-2%/s

(c) 10-4%/s

Figure 6: Fracture surfaces observed after tensile test.


Creep Deformation and Rupture
Figure 7 correlates times to creep rupture with applied stress at 700C, where the data of Alloy
740 are quoted from the literature [4] for comparison of their rupture time at the same
temperature. Solid line in the figure is drawn based on the data of Alloy 740H. The rupture data
of Alloy 740H were almost the same as those of Alloy 740, which indicates that modification of
alloy has a little effect on time to creep rupture. It should be noted that Alloy 740H is a modified
alloy based on Alloy 740 to improve its structure stability. Perhaps, there may be no significant
microstructure evolution in the short creep region at 700C, which results in similar creep rupture
life of both alloys.
Figures 8 (a) and (b) illustrate creep curves and variation of creep strain rate with time at three
stresses obtained in this study. The creep strains at rupture were smaller than 5% in all the tests,
which means that creep rupture ductility of the alloy is similar to the ductility in the tensile test at
slow strain rate, Fig. 8 (a). Clear primary, secondary and tertiary creep stages were found but the
primary and secondary stages were relatively shorter than the tertiary stage, irrespective of the
applied stress. Further, this alloy exhibited relatively small change of the strain rate in the primary
stage, indicating that strain hardening occurring in the primary stage is insignificant, Fig. 8 (b).

246

= 855.3t r 0.11
T=700C

Figure 7: Correlation of times to creep rupture with stress.

T=700C

T=700C

520MPa
430MPa

430MPa
520MPa

395MPa

Primary

Steady

Tertiary

395MPa

(a) Creep curves

(b) Variation of strain rate with time

Figure 8: Creep deformation behavior at 700C.


Figure 9 shows fracture surfaces observed after creep test. Almost the whole fracture was found to
be intergranular in all the tests, indicating that intergranular fracture is dominant at 700C even
though the testing time is comparatively short.
Figure 10 plots the relationship between the time to rupture and true fracture strain in tensile and
creep tests. The true fracture strain in tensile test decreased with increasing the time to rupture
because the fracture mode was changed from transgranular fracture at high strain rate to
intergranular fracture at low strain rate as shown in Fig. 6. On the other hand, the true fracture
strain in both tests took almost a constant value of about 10% in the intergranular fracture region.

247

500m

100m
(a) 520MPa

(b) 430MPa

(c) 395MPa

Figure 9: Fracture surfaces observed after creep test.

T+I
I
T: Transgranular fracture
I: Intergranular fracture

T=700C

Figure 10: Relationship between true fracture strain and time to rupture.
Fatigue Life
Figure 11(a) illustrates the variation of maximum and minimum stresses with number of cycles in
strain-controlled fatigue tests at 1.2% and 1.0% strain ranges. Slight cyclic hardening was
observed in the early cycles but the material exhibited a tendency to cyclically soften beyond
about 10% of the life. Similarly, slight cyclic hardening and softening behavior was observed in
the load-controlled tests at stress amplitudes of 500MPa and 450MPa as shown in Fig. 11(b). In
the test at stress amplitude of 400MPa without plastic strain, however, strain range was almost
constant until the end of life, indicating that there is no effect of the cyclic deformation on the
stress-strain relationship in the predominantly elastic region, which is easily understood from the
fact that cyclic hardening or softening is mainly caused by the change of dislocation structure.

248

Figure 12 correlates the fatigue lives in both the strain- and load-controlled tests with stress
amplitude, where the stress amplitudes in the strain-controlled tests are obtained from hysteresis
loops at half life. The solid line is drawn based on all the data at 700C. Fatigue lives in the strain
controlled tests were comparable to those in the load controlled tests, which means that the
control mode and loading frequency have a little effect on fatigue life. Furthermore, it can be seen
that the fatigue limit was about 400MPa. Since ultimate tensile strength at the same strain rate
(0.1%/s) was 870MPa as shown in Fig. 4, the fatigue limit was approximately a half of ultimate
tensile strength.

T=700C

(a) Strain controlled test

(b) Load controlled test

Figure 11: Cyclic deformation behaviors in strain and load controlled fatigue tests.

1/2B
T=700C

Figure 12: Correlation of numbers of cycles to failure with stress amplitude.

249

Figure 13 correlates the fatigue lives in all the tests with elastic ( e ), plastic ( p ) and total strain
range ( t ), where these strain ranges are obtained from the hysteresis loops at half life. In all the
tests, the elastic strain range was larger than the plastic strain range. The fatigue lives were
linearly correlated with the elastic strain and plastic strain range and can be expressed by the
following equations, so the Manson-Coffin relationship [10] holds for this alloy at 700C.

e = 1.25 N f 0.08 , p = 7.65 N f 0.44 , e , p : %

(1)

t = 1.25 N f 0.08 + 7.65 N f 0.44 , t : %

(2)

Figure 14 shows the photograph of fracture surface after fatigue test at strain range of 0.7% as a
representative case. Crack initiated at surface and propagated macroscopically in the normal
direction to the specimen axis. Transgranular fracture with clear striations morphology was
observed.
T=700C

e = 1.25N f 0.08

p = 7.65N f 0.44

Figure 13: Correlation of numbers of cycles to failure with strain range.


Crack initiation

200m

2mm

200m

Figure 14: Fracture surfaces observed after fatigue test at strain range of 0.7%.

250

Comparison with Other Materials


Figure 15 compares ultimate tensile strength and rupture elongation of Alloy 740H to those of
Alloy 617 and Alloy 263 [11]. Ultimate tensile strength and elongation of Alloy 617 decreased
with decreasing the strain rate similar to the results of Alloy 740H. Among the three alloys, the
largest ultimate tensile strength was obtained in Alloy 740H and the smallest in Alloy 617.
Conversely, Alloy 617 exhibited the largest elongation and elongation of Alloy 263 was larger
than that of Alloy 740H.
Comparison of time to creep rupture of the three alloys is shown in Fig. 16, in term of LarsonMiller Parameter. The same trend was found in the time to creep rupture as ultimate tensile
strength shown in Fig. 15. That is, the alloy with larger ultimate tensile strength in Fig. 15 had the
longer time to creep rupture in Fig. 16. This may result from that the contents of phase is
different in three alloys. In addition, creep rupture strengths of three alloys for 105h at 700C were
above 100MPa predicted by Larson-Miller method, indicating that the three alloys have creep
strength required as boiler material used in A-USC power plants.

T=700C

Alloy 740H
Alloy 617
[11]
Alloy 263

Figure 15: Comparison of B and of three kinds of Ni-based alloys at 700C.

700C, 10 h

[12]

Figure 16: Comparison of creep strength of three kinds of Ni-based alloys.


251

Figure 17 compares fatigue life of Alloy 740H to those of Alloy 617 [13] and Alloy 263 [11] at
700C. The largest fatigue life was found in Alloy 740H and fatigue lives of Alloy 617 were
almost same as those of Alloy 263.
The above discussion has revealed that Alloy 740H exhibits better tensile, creep and fatigue
strength than Alloy 617 and Alloy 263 and is a leading material for constructing A-USC power
plants from a viewpoint of the high temperature strength. Ductility of Alloy 740H tends to be
smaller than those of other two alloys, but the thermal stress/strain can be reduced by decreasing
the thickness of components due to its enhanced strength and the possibility of creep-fatigue
failure is not necessarily larger.

T=700C

Figure 17: Comparison of fatigue life of three kinds of Ni-based alloys at 700C.
CONCLUSIONS
There was a little effect of strain rate on the 0.2% proof stress but the Youngs modulus slightly
decreased with decreasing strain rate in the tensile tests. The ultimate tensile strength and the
elongation significantly decreased with decreasing strain rate with a transition from the
transgranular fracture at higher strain rate to intergranular fracture at lower strain rate.
The time to creep rupture of Alloy 740H was similar to that of Alloy 740. Almost the whole
intergranular fracture was found in all the tests, indicating that intergranular fracture is dominant
at 700C even though the creep loading time is short. Creep rupture ductility of the alloy was
similar to that in the tensile test at slow strain rate.
Alloy 740H exhibited a slight cyclic hardening initially and softening behavior subsequently in
the fatigue tests and the control mode had a little effect on fatigue life. Fatigue limit of this alloy
was approximately a half of the ultimate tensile strength.
Ultimate tensile strength of Alloy 740H as well as creep strength was larger than those of Alloy
617 and Alloy 263 but elongation was the smallest among three alloys. In addition, Alloy 740H
showed greater fatigue strength compared with Alloy 617 and Alloy 263 especially at low strain
range.

252

ACKNOWLEDGMENTS
The authors gratefully acknowledge Daido Steel Co., Ltd. for providing the testing material and
Mr. Katsuaki Hosino for the supports in conducting the tensile and fatigue tests.
REFERENCES
[1] Fukuda, M., Advanced USC Power Generation Technology, Journal of the Japan Society
of Mechanical Engineers, Vol. 114, No. 1109 (2011), pp. 22-25.
[2] Kjaer, S., Klauke, F., Vanstone, R., Zeijseink, A., Weissinger, G., Kristensen, P., Meier, J.,
Blum, R and Wieghardt, K., The advanced Supercritical 700 Pulverised Coal-Fired
Power Plant, VGB Power Technology, Vol. 82, No. 7 (2002), pp. 46-49.
[3] Palkes, M., Boiler Materials for Ultra Supercritical Coal Power Plants Conceptual Design
ALSTOM Approach, NETL-DOE, USC T-1 (2003).
[4] Patel, S.J., Introduction to Inconel Alloy 740: an Alloy Designed for Superheater Tubing in
Coal-Fired Ultra Supercritical Boilers, Acta Metallurgica Sinica (English Letters), Vol. 18,
No. 4 (2005), pp. 479-488.
[5] Zhao, S., Xie, X., Smith, G.D and Patel, S.J., Microstructural Stability and Mechanical
Properties of a New Nickel-Based Superalloy, Materials Science and Engineering A, Vol.
355 (2003), pp.96-243.
[6] Evans, N.D., Maziasz, P.J., Swindeman, R.W and Smith, G.D., Microstructure and Phase
Stability in INCONEL Alloy 740 during Creep, Scripta Materialia, Vol. 51 (2004), pp. 503507.
[7] Zhao, S., Xie, X., Smith, G.D and Patel, S.J., Research and Improvement on Structure
Stability and Corrosion Resistance of Nickel-Base Superalloy INCONEL Alloy 740,
Materials and Desigh, Vol.27 (2006), pp.1120-1127.
[8] Japanese Industrial Standards Committee ed., JIS G 0551: Steel-Micrographic Determination
of the Apparent Grain Size, Japanese Standards Association (2005), p. 12.
[9] JSMS Committee on High Temperature Strength of Materials ed., High Temperature
Strength of Materials, Japan Society of Materials Science (2008), p. 38.
[10] Manson, S.S., Fatigue: a Complex Subject: Some Simple Approximations, Experimental
Mechanics, Vol. 5 (1965), pp. 193-226.
[11] Yamamoto, M., Shingledecker, J., Ogata, T and Santella, M., High Temperature Strength
Evaluation of an Advanced USC Candidate Material Alloy 263, CRIEPI report, Q08003
(2009).
[12] http://www.specialmetals.com/products/inconelalloy617.php
[13] Noguchi, Y., Miyahara, M., Okada, H., Igarashi, M and Ogawa, K.,Creep-Fatigue
Properties of Fe-Ni Base Alloy HR6W for Piping in 700 USC Power Plants, Journal of
the Japan Society of Materials Science, Japan, Vol. 57, No. 6 (2008), pp. 569-575.

253

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

CYCLIC PROPERTIES OF 50Ni-24Cr-20Co-0.6Mo-1Al-1.6Ti-2Nb


ALLOY AT ADVANCED USC STEAM TEMPERATURE
U V Gururaja, A Patra, P Mukhopadhyay, M Narayana Rao
Mishra Dhatu Nigam Ltd., Hyderabad-500058, India

ABSTRACT
Significant development is being carried out worldwide for establishing advanced ultra
supercritical power plant technology which aims enhancement of plant efficiency and reduction of
emissions, through increased inlet steam temperature of 750C and pressure of 350 bar. Nickel
base superalloy, 50Ni-24Cr-20Co-0.6Mo-1Al-1.6Ti-2Nb alloy, is being considered as a
promising material for superheater tubes and turbine rotors operating at ultra supercritical steam
conditions. Thermal fluctuations impose low cycle fatigue loading in creep regime of this material
and there is limited published fatigue and creep-fatigue characteristics data available. Scope of
present study includes behavior of the alloy under cyclic loading at operating temperature. Strain
controlled low cycle fatigue tests, carried out within the strain range of 0.2%-1%, indicate
substantial hardening at all temperatures. It becomes more evident with increasing strain
amplitude which is attributed to the cumulative effects of increased dislocation density and
immobilization of dislocation by ' precipitates. Deformation mechanism which influences fatigue
life at 750C as a function of strain rate is identified. Hold times upto 500 seconds are introduced
at 750C to evaluate the effect of creep fatigue interaction on fatigue crack growth, considered as
one of the primary damage mode. The macroscopic performance is correlated with microscopic
deformation characteristics.
Keywords: Ultra supercritical, Low cycle fatigue, Alloy 740
INTRODUCTION
Establishment of state of the art advanced ultra supercritical (AUSC) power plant is being pursued
worldwide to meet ever increasing demand of electricity due to increased population and global
warming due to CO2, SOX and NOX emissions. A key aspect of this strategy of development of
AUSC technology is to have an operating inlet steam temperature of 750C and pressure of
350bar which will in turn increase the efficiency of Rankine cycle. Henceforth performance of
superheater tubing, headers and HP rotors becomes crucial in terms of creep strength, low cycle
fatigue strength. Because of the structural stability of newly developed nickel base superalloy
(50Ni-24Cr-20Co-0.6Mo-1Al-1.6Ti-2Nb (alloy 740) primarily strengthened by gamma prime
and additionally solid solution strengthened by Co, Cr at 750C, it is considered as the potential
candidate material for A-USC power plant [1]. Low cycle fatigue damage caused by cyclic
thermal stress and strain generated during daily operation, start-up and shut-down of a thermal
power plant is a critical issue to be addressed. As limited published data is available about fatigue
and creep fatigue properties of alloy 740 and thus deformation behavior under cyclic loading and
in creep fatigue interaction condition need to be investigated.

254

The scope of present study includes preliminary tensile and creep property characterization,
assessment of low cycle fatigue life at room temperature and operating temperature range of an
industrial scale heat of alloy 740. The effect of reducing strain rate and introduction of hold time
during cyclic loading at 750C is also investigated. In order to generate complete understanding
of mechanisms of failure, macroscopic fracture mode is correlated with microscopic damage
mechanisms.
EXPERIMENTAL
Material
50Ni-24Cr-20Co-0.6Mo-1Al-1.6Ti-2Nb alloy was manufactured at industrial scale through
vacuum induction melting followed by consumable electrode vacuum arc remelting. Composition
of product sample is given in Table 1.
Table 1. Chemical composition of alloy
Element
C
S
P
Si
Mn
Cr
Co
Mo
<0.002 <0.004 0.02
0.03
24.7
20.8
0.56
Weight% 0.02
Element
Ti
Al
Nb
Fe
B
N2
O2
Ni
1.70
1.98
0.26 50ppm 30ppm 10ppm Balance
Weight% 1.16
This ingot was subjected to long term homogenization cycle followed by various
thermomechanical processing steps to obtain 100mm forged bar and was subsequently hot
rolled to 28mm . Heat treatment at 28mm involved solution annealing at 1150C for 30
minutes minimum, quenching in water and aging at 800C for 16hrs followed by air cooling.
Microstructural observation (Fig. 1) revealed that the material is having austenitic grains with
annealing twins of average grain size of 100 microns, a small quantity of un-dissolved primary
MC carbides and fine discontinuous M23C6 along the grain and twin boundaries. On aging fine
particles precipitate in the matrix.

100X
Fig. 1 Microstructure of solution treated alloy
Tensile properties obtained at room temperature and application temperature range in the alloy
under study in comparison with international standard [2] are given in Table 2.
Table 2. Tensile properties of alloy
Tensile properties
0.2% YS (MPa) UTS (MPa)
620
1035
Specified at RT (Code case 2702)
890
1250
Obtained values at RT
621
1023
Typical values at 650C
716
985
Obtained values at 650C
608
766
Typical values at 750C
650
907
Obtained values at 750C
255

%El
20
30
38
36
32
40

As application of this high temperature material demands primarily creep strength, creep rupture
tests were carried out to qualify the creep characteristics of the material before evaluating the
creep fatigue interaction behavior at application temperature. Creep rupture strength of 280MPa
for 1000 hours was obtained at 750C.
Methods
The specimens used in low cycle fatigue test were 4.5mm and 14mm in gauge length. A
digitally controlled servo hydraulic fatigue testing machine equipped with an electrical resistance
furnace was used for the test. Fully reversed (R= -1) strain controlled isothermal low cycle fatigue
tests were conducted by using an extensometer with a gauge length of 12 mm. A symmetrical
triangular waveform with 3X10-3 s-1 strain rate was applied for tests with axial strain range of
0.2% to1% at room temperature and 750C. The test temperature of 750C was chosen to
reproduce the condition of AUSC. Also fatigue tests with lower strain rate i.e. 3X10-5 s-1 were
carried out at 750C. Tensile hold time upto 500 seconds were introduced at 750C by applying
trapezoidal waveform. Fracture surfaces were examined by optical and scanning electron
microscope to determine the fatigue crack initiation and propagation modes. Studies were also
conducted on longitudinal section of the specimen to characterize the microstructural changes that
occur during fatigue and creep-fatigue damage.
RESULTS
The relationship of total strain range with fatigue life at room temperature and operating
temperature i.e 750C was shown in Fig. 2. Although relatively low strain range ( = +1%) led
to endurances of around 500 cycles, the gradients of endurance curves were shallow and yielded
endurance of 5000 cycles at = +0.3%. Effect of decreasing strain rate to 3X10-5 s-1 on fatigue
life was observed at 750C. There was a trend towards lowering of fatigue life by 1/4th with
decreasing strain rate (Fig. 3).The hysteresis loop of stable cycle at 750C with different strain
rate were shown in Fig. 4, where it was observed that, a significant drop in peak tensile stress
occurred at lower strain rate test. It indicated that there was considerable amount of increase in
inelastic strain per cycle and henceforth identified as the possible reason for lower fatigue life
with reducing strain rate.

Fig. 2 Relationship between total strain range and low cycle fatigue life at RT and 750C

256

Fig. 3 Effect of strain rate on fatigue life at 750C

Fig. 4 Hysteresis loop of stable cycle at different strain rate at 750C, = +0.4%
Change in peak stress with number of cycles in all strain ranges at 750C is shown in Fig. 5. Alloy
740 does not show typical cyclic hardening or softening behavior instead cyclic softening follows
a rapid hardening similar to precipitation strengthened nickel base superalloys [3]. This general
behaviour of hardening does not change with increase in strain range, but the magnitudes of
hardening have been found to be dependent on strain range. At lower strain range a stable phase is
observed in between cyclic hardening and softening. Also at all temperatures cyclic hardening
behaviour is prominent followed by a stable phase and cyclic softening (Fig. 6).

257

900
800

Peak stress(MPa)

700
600
500
400
300
200

= + 0.8%
= + 0.6%
= + 0.4%

100
0

500

1000

1500

2000

No of cycles

2500

Fig. 5 Cyclic hardening behaviour at 750C with increasing strain range

Peak stress at(MPa) =+0.4%

800
700

RT

600
500
400

750 deg_C

300

650 deg_C

200
100
0

2000

4000

6000

8000

10000

No of cycles
Fig. 6 Cyclic hardening behaviour at different temperature, = +0.4%
Figure 7 showed the relationship between the stress ranges obtained at half of fatigue life. The
lives in terms of stress range were shorter in lower strain rate test. Stress range of alloy 740 was
notably higher than those of other alloys [4]. Masing behaviour was observed in alloy 740 at room
temperature upto total strain range under study as indicated in Fig. 8. By bringing the compressive
tips of all stable hysteresis loops belonging to various strain levels to a common origin, it was
observed that all the loading branches overlap and form a common envelope curve. Hold times at
the tensile peak strain led to pronounced reduction in fatigue life. A tensile hold of 100s seconds
resulted in shorter fatigue life by 60%. It was noted that fatigue life reduction increased
258

continuously, but at lower rate with an increase in hold time above 100 seconds (Fig. 9). Load
drop was similar with hold time above 200 seconds as observed in hysteresis loops of stable
cycles (Fig. 10).
2000

Stress range, (MPa)

At 3X10-3/s
At 3X10-5/s

1000
900
800
700
600
500

10

100

1000

10000

100000

No of cycles to failure,Nf

Fig. 7 Total stress range vs no of cycles to failure in different strain rate at 750C
1000

Stress(MPa)

500

-500

-1000
-0.8

0.8%
0.6%
0.4%
-0.6

-0.4

-0.2

0.0

0.2

Strain(%)

0.4

0.6

0.8

1.0

Fig. 8 Masing behaviour at room temperature

No of cycles to failure,Nf

10000

1000

100

10

100

200

300

400

500

Hold time(sec)

Fig. 9 Effect of hold time on fatigue life at 750C, = +0.4%


259

Fig. 10 Load drop during hold at 750C and =+0.4%


DISCUSSION
Cyclic stress response depends on the internal resistance to deformation of a material. The
resistance comes from pinning of dislocation movements by precipitation. Creep rupture strength
of 280MPa at 1000 hrs life indicates the suitability of the alloy for designed strength greater than
100MPa at 100000 hours of AUSC application. Hence significance of low cycle fatigue studies
becomes more important from the point of view of application condition.
Fractographic study
The effect of loading condition on the macro crack initiation and propagation behaviour is
interpreted through fractographic study (Fig. 11, Fig. 12). At room temperature crack initiation
and propagation occurred in transgranular cleavage mode. With the increase of temperature to
650C, mode of crack initiation and propagation remains same but a more wavy fracture surface
was observed. Whereas at 750C crack initiation is transgranular and it propagated in mixed
mode. With decreasing strain rate and introducing hold time at 750C, fracture surface of the
specimen shows intergranular mode of fracture.

(a)

(b)

Fig. 11 Fracture surfaces of fatigue tested samples (a) at room temperature, (b) at 650C
260

(a)

(b)

(c)

Fig. 12 Fracture surfaces of fatigue tested samples (a)) at 750C, (b) at 750C with 3X10-5 strain
rate and (c) at 750C with tensile hold of 200seconds.
Deformation behaviour study
Total strain range of around 0.3% for endurance of 5000 cycles as required by the steam turbine
application is achieved by alloy 740 at 750C in air [5]. Masing behavior is observed when a
material has very stable microstructure under cyclic deformation. Specifically matrix containing
finely dispersed non-shearable precipitates and low stacking fault energy [6]. Cyclically deformed
substructure consists of planar arrays of dislocation, instead of dislocation cell, in this case. At
room temperature slip was predominantly planar and concentrated in parallel slip band which
have highest resolved shear stress and there is no change in precipitate morphology even within
the slip band. The tendency of planar slip is attributed to characteristic ordering and coherency of
precipitates, relatively difficult cross-slip of dislocation around the particles due to sufficient
low stacking fault energy of Ni-Cr-Co matrix [7]. This phenomenon is strongly supported by
occurrences of slip band step in SEM microphotograph (Fig. 13) and masing behavior too.

261

Fig. 13 Slip band step observed at broken fatigue sample tested at room temperature
Alloy 740 does not show typical hardening or softening behavior. It hardens till it reaches a
cyclically stable structure then continuous softening follows until failure similar to other
precipitation strengthened superalloy. The extreme hardening and softening behaviour reflects
precipitates are in peak aged condition. The peak aged particles are very strong, but once slip is
initiated on a few relatively weak planes, they soften drastically, concentrating further strain in
these few slip bands. When dislocation structure reaches towards a saturated state with increasing
strain range, resistance to deformation is reduced leading to lower fatigue life due to decreased
precipitation contribution, owing to the reduced precipitation cross section. With increasing strain
range, the dislocation density increases. Severe dislocation interaction leading to dislocation
tangles gives rise to increased peak stress as well as faster rate in cyclic hardening at higher strain
ranges. Cyclic hardening behavior is predominant even at 750C because of more slip system in
operation with increasing temperature, leading to frequent slip band interaction and dislocation
tangle formation.
The drop in number of cycles to failure with the introduction of hold time and decreased strain
rate is associated with higher dwell period at 750C. It induces grain boundary oxidation as
indicated in Fig. 14 by Cr2O3 scale on crack surface which results into accelerated intergranular
crack propagation. Once initiated crack will open exposing new material surface to gas phase,
oxidation of the crack edge would occur eventually with growth of Cr rich oxide. Figure 15 shows
the features of a typical grain boundary failure.

C r 2O 3

Fig. 14 Oxidation of crack path


262

Grain boundary crack

Fig. 15 Typical grain boundary failure


The shift in crack propagation mode from transgranular to intergranular is a result of grain
boundary damage from grain boundary oxidation is responsible for lower fatigue strength.
CONCLUSION
1. Low cycle fatigue life at room temperature and AUSC steam temperature is obtained
at
various strain ranges. Endurance of 5000 cycles of designed turbine life can be withstood by
alloy 740 with a strain range of 0.3% in air. To evaluate the suitability of the alloy in real steam
environment AUSC test loop study need to be carried out.
2. Reducing strain rate at 750C introduces considerable inelastic deformation and hence reduces
fatigue life.
3. Cyclic hardening followed by softening till failure is observed at all temperatures similar to
precipitation strengthened nickel base superalloy.
4. Masing behaviour is observed in the strain range under study i.e =+0.8% at room
temperature for this alloy which is attributed to fine phase precipitation and low stacking
fault energy due to higher Co.
5. Total stress range that the material can bear at operating temperature at different strain ranges is
much higher than other materials, like austenitic steels, solution strengthened nickel base
superalloys intended to be used for this application.
6. Change in crack propagation mode from transgranular to intergranular with reducing strain rate
or introducing hold time at 750C is attributed to grain boundary oxidation.
7. The phenomenon of tolerance of higher peak stress at 750C in comparison to 650C, although
the fatigue life is comparatively lower gives further scope of study on change in mechanism of
precipitate behaviour under cyclic deformation at 650C and 750C.
REFERENCES
[1] Shingledecker, P. et al, Strengthening issues for high-temperature Ni-based alloys for use in
usc steam cycles, (2006).
[2] Inconel alloy 740, Publication Number SMC-090, Copyright Special Metals Corporation,
(2004).
[3] Bressers, J., Creep and fatigue in high temperature alloys, Applied science publishers
(London, 1981), pp. 73-81.
263

[4] Kobayashi, K. et al, Creep Fatigue interaction properties of Nickel-based superalloy 617,
Acta Metallurgica, Vol. 24, No. 2(2011), pp. 125-131.
[5] Blum, R. et al, Materials Development for Boilers and Steam Turbines Operating at 700C,
Proceedings of the Fourth International Conference on Advances in Materials Technology
for Fossil Power Plants 2004 (ASM International), South Carolina, USA,June. 2005, pp.
116-136.
[6] Plumtree, A. et al, Cyclic stress strain response and substructure, International journal of
fatigue, Vol. 23 (2001), pp. 799-805.
[7] Sims, C. T. et. al., Superalloys II, Wiley-Interscience publication, (New York, 1987), pp. 263267.

264

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

MICROSTRUCTURE EVOLUTION AND PRECIPITATES STABILITY IN


INCONEL ALLOY 740H DURING CREEP
Shuangqun Zhao, Fusheng Lin, Rui Fu
Shanghai Power Equipment Research Institute, Shanghai 200240, China
Chengyu Chi, Xishan Xie
School of Materials Science and Engineering, University of Science and Technology
Beijing, Beijing 100083, China

ABSTRACT
Inconel alloy 740H is designated for boiler sueprheater/reheater tubes and main steam/header
pipes application of advanced ultra-supercritical (A-USC) power plant at operating temperatures
above 750. Microstructure evolution and precipitates stability in the samples of alloy 740H
after creep-rupture test at 750, 800 and 850 were characterized in this paper by scanning
electron microscopy, transmission electron microscopy and chemical phase analysis in details.
The phase compositions of alloy 740H were also calculated by thermodynamic calculation. The
research results indicate that the microstructure of this alloy keeps good thermal stability during
creep-rupture test at 750, 800 and 850. The precipitates are MC, M23C6 and during
creep-rupture test. The temperature of creep test has an important effect on the growth rate of
phase. No harmful and brittle phase was found and also no to transformation happened
during creep. Thermodynamic calculations reveal almost all the major phases and their stable
temperatures, fractions and compositions in the alloy. The calculated results of phase
compositions are consistent with the results of chemical phase analysis. In brief, except of
coarsening of , Inconel alloy 740H maintains the very good structure stability at temperatures
between 750 and 850.
INTRODUCTION
The challenge and pressure on environmental protection and energy saving have been more and
more serious since the last decade of the twentieth century in the world wide. In one hand, new
energy, such as solar energy, wind energy, biomass energy, etc, are paying more and more
attentions and their developmental speed and increasing rate are very high in the last ten years, in
other hand, the great efforts to improve the thermal efficiency of the traditional fossil power
plants by increasing the steam temperature and pressure are also carried out in this century. The
600-class ultra-supercritical (USC) coal-fired power technology has been the major choice of
the newly-built power plant. There are more than 110 units (600~660MW & 1000MW) of
600-class USC coal-fired power plant put into operation in China from 2006 to 2012, which
accounts for about more than 80% of USC units in the whole world.
Simultaneously, the 700 advanced ultra-supercritical (A-USC) coal-fired power technology
has moved forward into the researchers insight. Since 1998, European Union initiated the
Thermie AD700 project to develop 700A-USC coal-fired power technology and several
265

successive projects were carried out to develop or demonstrate the new materials and key
components in Europe in the following years [1, 2]. USA, Japan and China also started their
research project to develop the 700A-USC pulverized coal-fired power technology in the last
decade [3, 4].
Inconel alloy 740 is a precipitate-strengthened Ni-Cr-Co superalloy and was invented by Special
Metals Corporation, Huntington from 1998 to meet the requirement (750/105h creep rupture
strength larger than 100MPa and metal loss of flue gas/coal ash corrosion/oxidation less than
2mm in 2105h) as boiler superheater/reheater tubing materials in Europe 700A-USC project
[5,6]. The patent of this Ni-Cr-Co alloy was issued in 2001. However, according to the
experimental research results, the microstructure instability of Inconel alloy 740 in the elevated
temperatures are as follows: (1) high growth rate at temperature >750; (2) large amount of
plate-shape phase formation nearby the grain boundaries and in the grains and (3) big block of
Si-containing G-phase precipitation at grain boundaries [7-10]. Base on the consideration of
improving the microstructure stability and the elimination of liquid cracks in thick-wall welded
joint of Inconel alloy 740, chemical composition modification of Inconel alloy 740 was carried
out through adjustment of the precipitation-strengthened elements of Al, Ti, Nb and Si, B as well
[11-15]. The new alloy, Inconel alloy 740H, was specifically developed in 2009 for
superheater/reheater tubes and main steam/header pipes applications and has been the main
candidate materials in 700 A-USC power plants.
Over the past decade, the corrosion resistance, microstructure evolution and mechanical
properties of the alloy 740 have been extensively investigated in the world. From 2009, the
researchers interests concentrate on the microstructure stability, creep-rupture property,
workability and weldability of Inconel alloy 740H. As one of the most important candidate
materials for Chinese 700 A-USC power plant, Inconel alloy 740H is still being evaluated in
several research institutes, boiler companies and universities in China.
In this paper, several creep-ruptured samples of Inconel alloy 740H were investigated through
SEM and TEM observation, chemical phase analysis and Thermodynamic calculation as well to
detail the microstructure evolution and precipitates stability during creep at different high
temperatures. Creep-rupture tests are still ongoing and creep-rupture property, mechanism and
the influencing factors will be discussed in the future.
EXPERIMENTAL MATERIALS AND METHODS
The chemical compositions of alloy 740H used in the present investigation and the ASME
specification are given in Table 1. The chemical compositions of alloy 740H are in the range of
specification of ASME Code Case 2702. The sample material was solution annealed and aged
according to the specification of ASME Code Case 2702. The actual aging conditions are carried
out at 800 for 5hrs for the sample of alloy 740H. The optical microscopy was performed on
mechanically polished and chemical etched specimen of alloy 740H to get its ASTM grain size
grade. The kinds and weight fractions of the precipitates were ascertained through microchemical phase analyses method. Creep-rupture tests were carried out at 750, 800 and 850
in several stresses. The longest rupture-life samples of alloy 740H were selected for
microstructure characterization, which were electro-polished and electro-etched first and then
were carried out using scanning electron microscopy (SEM). The samples of longest rupture-life
of alloy 740H were mechanically ground to about 80m-thickness and then were double-jet
266

electro-polished or argon-ion milled to produce better thin regions for transmission electron
microscopy (TEM) observation. Then, TEM was employed on these samples to get the
morphology of phases and their characteristic patterns of the selected area diffraction (SAD).
Thermodynamic calculations were performed to predict the precipitate stability of alloy 740H by
using Thermo-calc software and a corresponding Ni-database developed by Thermo-Tech. The
calculation results were also compared with the results of chemical phase analysis of the alloy.

Alloy 740H
ASME CC
2702
(Continued)
Alloy 740H
ASME CC
2702

Table 1: Chemical compositions of Inconel alloy 740H (wt.%)


C
Si
Mn
S
P
Cr
Co
0.031
0.11
0.11
0.0003
0.005
25.1
20.1
0.005
23.515.01.0
1.0
0.03
0.03
-0.08
25.5
22.0
Fe
0.45

Nb
1.55

Ti
1.61

Al
1.60

Cu
0.06

3.0

0.5-2.5

0.5-2.5

0.2-2.0

0.5

B
0.002
0.00060.006

Zr
0.03
-

Mo
0.53
2.0
Ni
Bal
Bal

EXPERIMENTAL RESULTS AND DISCUSSION


Microstructure and precipitates of alloy 740H
Figure 1 shows the OM image of the grain of alloy 740H at the solution-annealing state. The
average grain size of the sample is about 73m and its ASTM grade is 4.5. SEM image of the
sample in the state of solution-annealed and aged treatment is shown in Figure 2. The
morphologies of in the grain and M23C6 at grain boundary of alloy 740H are shown in Figure
3. Microstructure of the alloy indicates that carbides precipitated in the matrix and at grain
boundaries. The particles distributed in the matrix randomly and at grain boundary are MC and
M23C6 carbides respectively. As shown in Figure 2 and Figure 3, the globally and uniformly
dispersive fine particles are observed in the grain. The shape of M23C6 carbide at grain
boundary usually is granular and the size mostly is less than 100nm. The SAD pattern in Figure
3a corresponds to disordered fcc with ordered superlattice reflections in a [110] beam
direction. The SAD pattern of grain boundary carbide M23C6 in Figure 3b shows that it has a
cube-on-cube orientation relationship with the matrix in [110] zone. The average dimension of
particles in the sample of alloy 740H at standard heat-treatment state is about 33nm. The
exacted quantity of the precipitates shown in Table 2, which were analyzed by a physical and
chemical phase analyses technology, indicates that the precipitates of the alloy include , MC
and M23C6. The fractions of , MC and M23C6 in the sample of alloy 740H are 17.246%, 0.157%
and 0.131% respectively. The MC is (Nb,Ti)C-type carbide and M23C6 is Cr23C6-type carbides.
According to the above-mentioned experimental results, there are three kinds of precipitates, ,
MC and M23C6 in alloy 740H at standard heat-treatment conditions. G phase wasnt found at
grain boundary, which usually forms at grain boundary due to the high silicon fraction in original
alloy 740. The harmful TCP phase, such as phase, also was not found in alloy 740H. In
addition, the result of chemical phase analysis indicates that phase fraction of alloy 740H is
larger than that of original alloy 740 at standard heat-treatment state, in which the fraction of
phase is less than 13wt%. The differences of these results can be attributed to the variation of
267

chemical compositions between alloy 740H and alloy 740. The total atomic fractions of
precipitation-strengthening elements Ti, Al and Nb of alloy740H are larger than that of alloy 740
especially for Al fraction.

Figure 1: OM image of alloy 740H

Figure 2: SEM image of alloy 740H


(a)

(b)

[110]

[110]

Figure 3: TEM images and SAD patterns of (a) and (b) M23C6 of alloy 740H
Table 2: Fractions of the precipitates of alloy 740H (wt.%)
Precipitate
MC
M23C6

Weight Fraction 17.246


0.157
0.131
Microstructure evolution of alloy 740H during creep
Figure 4 shows microstructures of alloy 740H after creep-rupture test at 750/265MPa for
3147h. Compared with the image of the sample at standard heat-treatment state (Figure 2), the
growth of is obvious after creep-rupture test at 750 and the carbides at grain boundary also
become coarse and no other phases were found in TEM image as well.
TEM images of alloy 740H after creep-rupture test shown in Figure 5 give the morphologies of
particles in the grain and M23C6 carbide at grain boundary of alloy 740H. The SAD pattern shows
that phase and M23C6 carbide have same orientation relationship with the matrix in [211]
crystal zone axis. The average dimension of particle is about 70nm. Moreover, during
268

coarsening, the shape of particle is also changed from global to cubic. The cubic shape
indicates that there is a large mismatch degree of crystal lattice between matrix and phase. It
is worth mentioning that there isnt phase formation nearby the grain boundary and also no free zone nearby the grain boundary. On the contrary, phase was found to accumulate at or
nearby the grain boundary. This is different from the alloy 740, in which phase forms very fast
nearby the grain boundary and then a -depletion zone happens at and above 750 during longterm aging or creep-rupture test.

Figure4 : SEM image of alloy 740H after creep-rupture test at 750(265MPa/3147h)


(a)

(b)

M23C6

M23C6

[211]

Figure 5: TEM images of (a) and (b) M23C6 of alloy 740H after creep-rupture test at 750
SEM image in Figure 6 shows microstructure of alloy 740H after creep-rupture test at 800
/160MPa for 2807h. The growth of particles in the grain and carbides at grain boundary is
remarkable and also no other phase can be found in this sample.
The morphologies of particles and grain boundary carbides of the sample after creep-rupture
test at 800 are shown in Figure 7. The phase was also found to form at or nearby the grain
boundary. The average dimension of phase is about 130nm. Same as in Figure 5, the phase,

269

grain boundary carbide also have same orientation relationship with the matrix in [110] and
[211] zone.

Figure 6: SEM image of alloy 740H after creep-rupture test at 800(160MPa/2807h)


(a)

(b)

M23C6

M23C6
[110]

[211]

Figure 7: TEM images of (a) and (b) M23C6 of alloy 740H after creep-rupture test at 800
SEM microstructure of the alloy 740H after creep-rupture test at 850/100MPa for 2943h is
shown in Figure 8. There are many big size particles form at and nearby the grain boundary.
This is same as the samples crept at 750 and 800. Figure 9 shows the morphologies of
phase and carbide of the sample after creep-rupture test at 850 and their distributions are same
as in former figures. The average size of phase is about 270nm. The size homogeneity of
particles of the sample creep-rupture test at 850 is worse than the former samples after creeprupture test at 750 and 800.
From the above results of microstructure of samples creep-ruptured at 750, 800 and 850,
except for the phases of , MC and M23C6, no other phase was found in the samples after creeprupture test. The coarsening of particle is the main style of structure evolution. Even the
rupture life is different, the creep test temperature also can be considered as the major influencing
factor on the growth rate of phase. With different from the previous alloy 740, although the
270

creep test temperature is relatively high, the formation of phase nearby the grain boundary and
the transformation of -to- phase t didnt happen in the crept samples of alloy 740H.

Figure 8: SEM image of alloy 740H after creep-rupture test at 850(100MPa/2943h)


(a)

(b)
M23C6

Figure 9: TEM images of (a) and (b) M23C6 of alloy 740H after creep-rupture test at 850
On the formation and accumulation of at and nearby grain boundary, the reason can be
explained as follows. The surrounding area of matrix around the carbide particle has a different
composition from the matrix composition. Both chemical phase analyses and Thermo-calc
predictions, shown in the following Figure 11, Table 3 and Table 4, reveal that M23C6 carbide
contains no Nb, Al and Ti and only small amounts of Ni and Co. As a consequence, the matrix
region surrounding the carbide is rich in the rejected -forming elements, and Cr as a M23C6forming element is depleted in this region. Thus phase is easy to nucleate and finally
precipitate at or nearby grain boundary.
Calculation of precipitate stability of alloy 740H
The calculated equilibrium phase diagram of the alloy 740H is shown in Figure 10 by Thermocalc software using exact chemical compositions in Table 1, in which all the predicted
equilibrium phases and their weight fractions are given. Thermodynamic calculation result
271

reveals that the equilibrium phases above 600 include , , , , MC and M23C6. The stable
temperature scopes of precipitate of the alloy, , and , are 766, 920~1100, and
1021, respectively. The predicted solidification temperature range of this alloy is about 65
and the solid phase and liquid phase coexist at temperatures between 1304 and 1369.
Compared with the calculation results of alloy 740 [7, 9, 10], the fraction and dissolution
temperature of phase increase dramatically and the stable temperature scope of phase also
becomes narrow and especially no phase forms in the potential operating temperature scope of
750-800 in A-USC units.
1304
1369

Weight fraction

Liquid

1021
1100

766
920

M23C6

MC

Temperature /
Figure 10: Calculated phase diagram of alloy 740H showing equilibrium phase and their weight
fractions
The compositions of precipitates were also calculated by Thermo-calc and the change tendencies
of major elements are presented in Figure 11. For the composition of phase, with the increase
of equilibrium temperature, the fraction of Ni decreases and the fractions of Co and Cr increase
slightly. However, the fractions of precipitation-strengthening Al, Nb and Ti almost keep same
with the temperature increasing and the fraction of Ti is larger than those of Nb and Al. For the
composition of M23C6 carbide, with the increase of equilibrium temperature, the fractions of Cr,
Co and Ni increase slightly and the fraction of Mo decrease obviously. M23C6 carbide is rich in
Cr and the fraction of Cr accounts for about 75%. For the composition of MC carbide, with the
increase of equilibrium temperature, the fraction of Nb decreases and the fraction of Ti increase.
The fraction of Nb is much more than the fraction of Ti in the composition of MC carbide.
The calculated phase compositions of precipitates at 800 are shown in Table 3 and the
measured phase compositions of precipitates at standard heat-treatment state by chemical phase
analysis are shown in Table 4. Thermodynamic calculation is on the basis of complete balance of
the system. Although there is deviation between the results of calculated and measured, the
calculated results have good agreement with the experimental results and the major composition
and their variation trend of the precipitates are almost same. The big difference is the calculated
composition and measured composition of MC carbide. The results of calculation and
measurement show that M23C6 carbide is very rich in Cr, and it is based on Cr23C6 with some Mo,
272

Co and Ni replacing the Cr, and that it is obvious that is (Ni,Co,Cr)3(Ti,Nb,Al) and MC carbide
is (Nb,Ti)C with small amount of Mo and Cr in it.
0.9

0.8

(a)

(b)

0.7

Ni
Co
Ti
Nb
Al
Cr

0.6

Ni

Co
Ti
Nb
Al
Cr

0.1

0.0

Weight Fraction / %

Weight Fraction / %

0.8

600

700

800

900

Cr
0.7

Cr
Mo
C
Co
Ni

0.6

Mo
0.1

0.0

1000

C
Co
Ni
600

Temperature /

700

800

900

1000

Temperature /

0.9

(c)
Weight Fraction / %

0.8

0.7

Nb

Nb
Ti
C

0.6

0.2

Ti
C

0.1

0.0

600

700

800

900

1000

Temperature /

Figure 11: Calculated phase compositions of alloy 740H ((a) , (b) M23C6 and (c)MC)
Table 3: Calculated compositions of precipitates of alloy 740H by Thermo-calc(wt.%)

M23C6
MC

Cr

Co

Mo

Nb

Ti

Al

Ni

5.390
11.460

1.751
77.747
0.130

9.023
3.620

0.029
10.669
0.069

6.874

78.450

7.272

9.891

5.451

69.601
3.261

Table 4: Measured compositions of precipitates of alloy 740H by chemical phase analyses(wt.%)

M23C6
MC

Cr

Co

Mo

Nb

Ti

Al

Ni

5.582
14.997

6.811
84.266
0.455

4.818
3.688

0.191
1.425
0.408

5.175

52.418

6.591

31.722

5.142

71.272
5.040

From Figure 10, the fractions of precipitates also can be predicted and the precipitates at 800
are 21.7% weight percent and 0.6% weight percent M23C6. These results are higher that the
experimental results shown in Table 2, which is from the samples aged at 800 for 5h and
therefore it is not in a complete equilibrium state theoretically. Generally, thermodynamic
calculations correctly predicted almost all the phases and their compositions in this alloy, except
harmful phase, which was not founded in the sample of creep-rupture test at 750. This is
273

probably due to the relatively poor assessment of the thermo-dynamic parameters for phase.
Either thermodynamic calculations probably overpredicted the stability of phase, or it is
possible that phase precipitation is some very sluggish and the exposure time was not so long
enough to form this phase. In addition, it is noted that the thermodynamic calculation predicts the
formation of phase between 920 and 1100. The results of calculation and measurement
didnt reveal the existence of phase in alloy 740H crept at temperature between 750 and
850, which forms very fast above 750 in original alloy 740. However, the temperature scope
of experiment carried out doesnt cover the range of 920-1100and the existence of phase
cannot be confirmed in the present stage.
CONCLUSIONS
Inconel alloy 740H after creep-rupture test at elevated temperatures was investigated on its
microstructural evolution and precipitate stability by means of SEM, TEM, chemical phase
analysis and Thermodynamic calculation. The precipitates observed in this alloy are , M23C6
and MC. The phases of , M23C6 and MC are mainly composed of Ni3(Ti,Nb,Al), Cr23C6 and
(Nb,Ti)C respectively. After creep-rupture test at 750, 800 and 850, the category of phase
doesnt change and the major difference is the growth of particle. The particle dimension of
phase increases fast with the increasing of creep temperature. No , G and phase were found in
all samples.
Thermodynamic calculation predicts that the structure stability of this alloy including the stable
temperature, fraction and composition of , M23C6 and MC. The calculated phase compositions
of , M23C6 and MC are in good agreement with experimental results of chemical phase analysis.
REFERENCES
[1] R. Blum, Advanced (700oC) PF Power Plant, EC Contact No SF/1001/97/DK(1997).
[2] R. Blum, Vanstone R.W., and Messelier-Gouze C., Materials Development for Boilers and
Steam Turbines Operating at 700, Proceedings from the Fourth International Conference
on Advance in Materials Technology for Fossil Power Plants, Hilton Head Island, South
Carolina, October 25-28, 2004, pp.116-136.
[3] R. Viswanathan, J. F. Henry, J. Tanzosh, G. Stanko, J. Shingledecker, and B. Vitalis, U.S.
Program on Materials Technology for USC Power Plants, Proceedings from the Fourth
International Conference on Advance in Materials Technology for Fossil Power Plants,
Hilton Head Island, South Carolina, October 25-28, 2004, pp.3-19.
[4] M. Fukuda, E. Saito, Y. Tanaka, T. Takahashi, S, Nakamura, J. Iwasaki, S. Takano, S.
Izumi, Advanced USC Technology Development in Japan, Proceedings from the Sixth
International Conference on Advance in Materials Technology for Fossil Power Plants,
Santa Fe, New Mexico, August 31-September 3, 2010, pp. 325-341.
[5] G. D. Smith, S. J. Patel, N. C. Farr, The corrosion resistance of nickel-containing alloys in
coal-fired boiler environments Paper presented at Corrosion 99, San Antonio, Texas,
April, 25-30, 1999, pp.12.
[6] G. D. Smith, H. W. Sizek, Introduction of an advanced superheater alloy for coal-fired
boilers Paper presented at Corrosion 2000, Orlando, Florida, March, 26-31, 2000,
pp.00256.1.

274

[7] S. Zhao, X. Xie, G. D. Smith, S. J. Patel, Microstructural stability and mechanical


properties of a new nickel-based superalloy, Materials Science and Engineering, Vol.
A355, (2003), pp. 96-105.
[8] S. Zhao, X. Xie, G. D. Smith, S. J. Patel, Gamma prime coarsening and age hardening
behaviors in a new nickel base superalloy, Materials Letters, Vol. 58, No. 11(2004), pp.
1784-1787.
[9] S. Zhao, J. Dong, X. Xie, G. D. Smith, S. J. Patel, Thermal stability study on a new Ni-CrCo-Mo-Nb-Ti-Al superalloy, Proceedings from the 10th International Symposium on
Superalloy , Seven Springs, PA, September 19-23, 2004, pp.63-72.
[10] N.D. Evans, P.J. Maziasz, R.W. Swindeman, G.D. Smith, Microstructure and phase
stability in INCONEL alloy 740 during creep, Scripta Materialia, Vol. 51, (2004), pp.503507.
[11] X. Xie, S. Zhao, J. Dong, G. D. Smith, S. J. Patel, An Investigation of Structure Stability
and Its Improvement on New Developed Ni-Cr-Co-Mo-Nb-Ti-Al superalloy, Materials
Science Forum, Vol. 475-479, (2005), pp. 613-618.
[12] S. Zhao, X. Xie, G. D. Smith, S. J. Patel, Research and improvement on structure stability
and corrosion resistance of nickel base superalloy INCONEL alloy 740, Materials and
Design, Vol. 27, No. 10, (2006), pp. 1120-1127.
[13] X. Xie, S. Zhao, J. Dong, G. D. Smith, B. A. Baker and S. J. Patel, A New Improvement of
INCONEL Alloy 740 for USC Power Plants, Proceedings from the Fifth International
Conference on Advance in Materials Technology for Fossil Power Plants, October 3-5,
Marco Island, Florida, 2007, pp. 220-230.
[14] J. M. Sanders, J. E. Ramirez and B. A. Baker, Weldability Investigation of Inconel Alloy
740for Ultrasupercritical Boiler Applications, Proceedings from the Fifth International
Conference on Advance in Materials Technology for Fossil Power Plants, October 3-5,
Marco Island, Florida, 2007, pp. 818-829.
[15] B. A. Baker, R. D. Gollihue, Optimization of INCONEL alloy 740 for Advanced Ultra
Supercritical Boilers, Proceedings from the Sixth International Conference on Advance in
Materials Technology for Fossil Power Plants, Santa Fe, New Mexico, August 31September 3, 2010, pp. 96-109.

275

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

DISCUSSION OF DELIVERED CONDITION SPECIFIED IN ASME


CODE CASE 2702 ON INCONEL 740H USED FOR A-USC BOILER
Hua-Chun Yang, Lin-Sen Wang, Dong-Ping Wang, Shi-Yong Mao,
Xiao-Chuan Yang, Xian-Hong Lai, Jin-Bing Yang, Wen Yang
Dongfang Boiler (Group) Co.,Ltd., Zigong, Sichuan 643001,P.R.China

ABSTRACT
Delivered condition of Inconel740H specified in ASME Code Case 2702 is solution heat treated
and aged condition, fabricating performances are also based on the condition, and a re-annealing
and aging shall be performed if cold forming to strains is over 5%. These almost harsh requirements bring great inconvenience for its engineering practice and utilization. The comparative
bending tests on 740H tubes in solution heat treated + aged condition and solution heat treated
condition were performed, and the rules reasonability of the specification on delivered condition
was discussed and analyzed from point view of deformability and weldability in the paper. The
bending test results showed that tube bent was difficult because of its high strength and limited
deforming capacity in solution heat treated + aged condition. Therefore, the material supplied in
the solution condition may be better from fabricating points. Whether re-solution for the bent
tube is performed after bending depends on its bending radius, followed by welding and post
weld heat treatment of component (this treatment can also be the aging treatment for annealed
sector at the same time), this treating manner can meet regulatory requirements. For solution
tubes, however, there are some inconveniences to its engineering application because fewer researches were carried out on its properties up to now, and no regulations on them were given for
the material in the specification. Suggestions are: 1) deeply investigating the properties of tubes
in solution condition, including mechanical and fabricating performances, 2) adding the mechanical properties, maximum allowable cold forming to stain without performing re-solution and
weld strength reduction factor of solution material to the code case.
1. INTRODUCTION
Inconel740H, a Ni-based super-alloy, is further modified by Special Metals Corporation on the
basis of Inconel740 developed from alloy Nimonic263. The alloy has better oxidation and fireside corrosion resistance because of being of higher Cr and lower Mo content, very good creep
rupture strength because of more dispersed precipitation of and its stability after adding of Al,
Ti and Nb. Now, Inconel740H was approved by ASME code case 2702 in September, 2011, and
is as an important candidate tube for A-USC boiler parts with about a temperature of 750[1-2].
The delivered condition specified in ASME Code Case 2702 is material shall be supplied in the
solution heat treated and aged condition. Following regulations such as allowable stress, fabricating performances and control of cold strains without re-annealing are also specified based on
the aged condition.
Of course, the tube shall be used in solution annealed and aged condition because the tube in this
condition will have better strength and good structural stability. However, aged hardenable alloys

276

are of poor formability, compared with those of annealed condition, especially for parts needing
bending or welding.
The comparative bending tests on 740H tubes in solution heat treated + aged condition and solution heat treated condition were performed and the rules reasonability of the specification on
delivered condition was discussed and analyzed from point view of deformability and weldability
required in boiler manufacturing processes in this paper, it is expected that this may be helpful for
the perfect of the specification and easy to be used in the future.
2. MATERIALS AND EXPERIMENTAL PROCEDURES
2.1. Test Materials
The materials used in this experiment were 740H tubes, with the dimension of 50.88 and
chemical compositions shown in Table 1.
Tubes in solution annealed (referred to as simply SA): Provided by Special Metals Corporation.
Tubes in solution annealed + aged (referred to as simply "SG"): Cut from tubes in SA, and
aged as per ASME Code case 2702, the same as aged by SMC: at 800oC for 4 hours, followed by
air cooling. The mechanical properties at room temperature were shown in Table 2.
Table 1: Chemical compositions
Element
Composition
(wt)
Element
Composition
(wt)

C
0.005
0.08
Ti
0.5
2.5

Mn

Fe

Si

Cu

1.0

3.0

0.03

1.0

0.50

Co
15.0
22.0

Mo

Cb+Ta
0.5
2.5

B
0.0006
0.006

2.0

0.03

Cr
23.5
25.5
Zr

Al
0.2
2.0

0.02

Table 2: Mechanical properties


Condition RP0.2 (MPa)
Rm (MPa)
A50mm (%)
Z (%)
HBW Hardness
SA
353.0/355.8
785.3/785.3
57.5/58.5 68.0/71.5
229/224/215
SG
728.1/728.1 1163.2/1159.7 36.0/36.5 41.0/37.5
313/328/321
NoteBecause of limitation of samples, tensile properties were courtesy of Special Metals
Corporation, and hardness values measured by Dongfang boiler.
2.2. Bending Tests
Bend forming tests were performed on 740H tubes in SA and SG conditions.
The flowchart of bending was as follows: Pre-cold bending Induction heating Small R hotextrusion.
Firstly, the tubes were pre-bent by boost bending machine to 76 of radius, about 1.5 times nominal diameter of the tube; the pre-cold elbow followed by heating in induction furnace to 1150,
then hot-extruded to 51 of radius in hot-extrusion equipment.
3. RESULTS AND DISCUSSIONS
The pre-cold elbows of the tubes in two conditions were given in Figure 1, and small R elbow
hot-extruded were shown in Figure 2.

277

As shown in Figure1, solution annealed tube exhibited excellent cold deformation capacity, metal
deformed uniformly, there was no wrinkle, deflated and appearance quality defects on the elbow
in Figure1-a.
Compared to elbow made by solution annealed tube, elbow from aged tube had relatively poor
cold deformation ability, uneven deformation, obvious transition zone, obvious rebound and apparently deflated uniform deformation region on the elbow in Figure1-b.
This phenomenon was closely related to the different strengths in two conditions, because of the
stronger strength, the greater resistance deformation, the harder deformation. Aged tube had
higher strength, hardness and lower ductility than annealed tube, as stated in Table 2. Therefore,
annealed 740H tube had better cold deforming ability than the aged tube.

a) SA

b) SG

Figure 1: Pre-cold elbows of 740H in two conditions

a) SA

b) SG

Figure 2: Small R hot-extruded elbow of 740H in two conditions


As for hot formability, small R elbow obtained by hot extruded on pre-cold elbow from annealed
tube was shown in Figure2-a, its profile was good and met the requirements of specification.
However, small R elbow manufactured from aged tube bent cracked in the transition region as
shown in Figure2-b, this showed that aged tube has poor hot formability.
In fact, there were several factors affecting hot formability of tubes such as hot ductility, deformation resistance and deforming rate. During hot extrusion of small R elbow, because of quick
forming requirements in short time and high deforming rate, it is necessary for tube to be lesser
deformation resistance and better hot ductility. In order to be softened and easily bent, the precold elbow was heated in induction furnace to 1150 very quickly and there was no keeping
time before hot extrusion. However, because uneven geometry of aged elbow after cold defor278

mation resulted in stress concentration during hot extrusion deformation, the plastic flow of partial location was affected and defects prone to be produced in pre-cold elbow. On the other hand,
because of very high heating rate and no keeping time, strengthening phases in the alloy cannot
completely be re-solution annealed into solid solution; these residual phases would block dislocation movement in plastic deformation. Therefore, pre-cold elbow in aged condition appeared
lesser hot ductility and larger deformation resistance than those of annealed, and was worse hot
bend ability.
The results from tests mentioned above showed that annealed tubes has much better formability
than that of aged tubes, whether it is cold forming or hot forming.
On the other hand, alloy microstructures markedly influenced its weldability. Annealed alloy,
phases mainly solved in solution, was in softened condition with lower strength and better ductility; In contrast, aged alloy with precipitated and dispersed phases, and higher strength and lesser ductility, enhanced the restraint degree during welding and increased sensitivity of the welding
crack[3]. At the same time, since there were some low melting of eutectic phases at grain boundaries for the aged alloy, liquation crack was easy to form while welded, so welding for precipitation strengthening alloy is better to be performed in annealed condition as far as possible.
4. DISCUSSION OF CONDITION SPECIFIED IN ASME CODE CASE 2702
Reply (a) in ASME Code Case 2702 specified that Material shall be supplied in the solution heat
treated and aged condition. Solution heat treatment shall be performed at 2010F (1100) minimum for 1 hour per inch (25 mm) of thickness but not less than hr. Aging shall be performed at
1400F to 1500F (760 to 816) for 4 hours minimum up to 2 inches (50 mm) of thickness,
plus an additional hour per additional inch (25 mm) of thickness. Aging shall be followed by
air cooling. This regulation for the alloy means that alloy delivered condition by manufacturer of
tube or pipe is solution annealed and aged. Allowable stress in Reply (b) and fabricating performances such as Reply (c), (d) and (e) are also based on the aged condition.
In Reply (e), the regulation is After cold forming to strains in excess of 5%; after any swages,
upsets, or flares; or after any hot forming of this material, the component shall be heat treated in
accordance with the requirements specified in (a) above. No local solution annealing may be performed. The entire affected component or part that includes the cold strained area and transition
to unstrained material must be included in both heat treatments.
However, deformation strain of all elbows will exceed 5% in the manufacturing process of actual
boiler parts; these almost harsh requirements bring great inconvenience for its engineering practice and utilization. From point view of fabricating, as mentioned above in bending tests results
and discussions, firstly, boiler parts with elbows have difficulty manufacturing by aged alloy
with poor bending performances and weldability; Second, even if they are able to be fabricated
into component, followed complete solution annealing of the parts would have huge work; Third,
because of solution annealing temperature very high, fabricated parts re-annealed would face the
problem of difficulty correcting deformation caused by high annealing temperatures; Finally, if
fabricator is to re-anneal in-work for bending after purchased from manufacturer of aged tube and
then bending, this will cause the waste of resources and increase of costs.
Of course, the tube shall be used in solution annealed and aged condition because tube in this
condition will have better strength and good structural stability. However, material in solution
heat treated will be better from engineering practice as solution tube has good forming performance from the test results mentioned above.
There may be a way out of the dilemma. Actually, whether elbows from solution tubes were resolution annealed or not depends on the bending radius (namely the strains), welding and post
weld heat treatment (PWHT) followed. This PWHT may also be as aging treatment for parent
solution tube, elbow (re-solution or not) and weldments. This can meet the requirements of appli-

279

cation condition specified in the case, and the treating manner may be an access to possible engineering application for the tube, too.
However, all properties are specified in aged condition and there are no regulations on annealed
condition in the specification, it is necessary to deeply study the properties in solution annealed
later, such as mechanical properties, welding, cold and thermal forming performance, so that requirements on solution tube can put forward into the case. For example, 1) mechanical properties
for annealed tube at room temperature may be added in Table 2; 2) are the other welding methods
allowed to be used? Reply (c) may give answer; and, 3) Reply (e) may answer what strain for the
elbow needing re-solution annealing, because annealed tube seems may not be same as strains in
excess of 5% of requirements for aged tube.
5. CONCLUSIONS
From comparative bending tests on 740H tubes in solution heat treated + aged condition and solution heat treated condition and mentioned analysis above, the conclusions and suggestions are
as follows:
(1)Annealed tubes has much better formability than that of aged tubes, whether it is cold forming
or hot forming.
(2)Suggestions: a) deeply investigating the properties of tubes in solution annealing condition,
including mechanical and fabricating performances; b) adding the regulation on the solution annealing tubes, such as mechanical properties, maximum allowable cold forming to stain without
performing re-solution and weld strength reduction factor of solution material to the code case.
ACKNOWLEDGMENTS
The financial support from the National Energy Utilization Technology Research and Engineering Demonstration Program of China under grant No. NY20110103A-1 and materials support
from Special Metals Corporation are greatly acknowledged.
REFERENCES
[1] S.J. PATEL. Introduction to Inconel alloy 740: an alloy designed for superheat tubing coal
fired ultra supercritical boilers, Acta Metallurgica Sinica (English Letters), Vol.18, No.4
(2005) , pp.479-488.
[2] S.Q. Zhao, Y. Jiang and J.X. Dong. Experimental Investigation and Thermodynamic Calculation on Phase Precipitation of Inconel 740, Acta Metallurgica Sinica (English Letters),
Vol.19, No.6 (2006), pp.425-431.
[3] J.T. Guo. Materials science and engineering for super-alloys (Part II). Beijing: Science Press,
(2009), pp.425-427 (in Chinese).

280

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

GAS AND STEAM TURBINE FORGINGS FOR HIGH EFFICIENT


FOSSIL POWER PLANTS
Authors: Guenter Zeiler
Albert Putschoegl
Boehler Edelstahl GmbH & Co KG, Kapfenberg, Austria

ABSTRACT
Sufficient available energy in combination with lowest environmental pollution is a basic
necessity for a high standard of living in every country. In order to guarantee power supply for
future generations it is necessary to use fossil fuels as efficient as possible. This fact calls for the
need of power plants with improved technologies to achieve higher efficiency combined with
reduced environmental impact. In order to realize this goal it is not only a challenge for power
station manufacturers, but also for manufacturers of special steels and forgings, who have to
produce improved components with more advanced materials and more complex manufacturing
processes.
This paper reports about experiences in the fabrication of forged components for gas and steam
turbines followed by achievable mechanical properties and ultrasonic detectability results. The
materials are the creep resistant martensitic Cr steels developed in the frame of the European Cost
research programme.
Whereas Boron containing 10% Cr steels are suitable for steam temperatures of 625C and
slightly higher, Ni-based alloys shall be used for temperatures of 700C and above. One pilot
rotor forging, representing a HP-rotor for welded construction, has been manufactured out of
alloy Inconel 625 within the frame of the European Thermie project AD700.
KEYWORDS
Forgings for power generation, melting processes, Cost programme, 10% Cr steels, Boron
containing 10% Cr, Cost E, F and FB2, properties, AD700 project, Inconel 625
INTRODUCTION
The continuous trend towards more economic electricity production in parallel with reduced
environmental pollution can only be sustained by improving the thermal efficiency of power
generation plants. The efficiency is increased by raising the temperature and also the pressure of
the steam which finally results in the need for improved materials for the boiler and turbine
design.
Since many years Boehler Edelstahl GmbH & Co KG is a premium supplier of forged
components for the power generation industry, e.g. discs, centre shafts, turbine shafts, shaft
components and accessories for gas and steam turbines. Based on a complete range of special
steelmaking equipment, consisting of an electric steel plant (electric arc furnace, AOD converter,
secondary metallurgy equipment), a special melting shop with vacuum and re-melting facilities
281

(VIM-VAR, ESR, PESR), a 52 MN hydraulic forging press and the R&D support of FEM
modelling to achieve a uniform deformation, Boehler is in a good position for the manufacture of
high-quality forgings, which are further processed in downstream facilities for heat treatment,
machining and testing.
Our customers are counted among the most important manufacturers of steam and gas turbines
such as Siemens, Alstom and licensees.
Seen from the material side Boehler manufactures all the typical steels introduced in energy
engines application so far, but more or less focussed on higher alloyed steels (Figure 1).
Materials

Analysis [%]

DIN

Bhler

Si

Mn

Cr

Mo

Ni

V116

0,30

0,06

0,30

1,50

0,40

2,90

0,10

<0,010

0,30 <0,03 <0,04 1,70

0,40

3,70

0,10

<0,010

26NiCrMoV 11 5
26NiCrMoV 14 5mod
(Super Clean)

V128SA

Al

Nb

30CrMoNiV 5 11

D102

0,32

0,10

0,80

1,10

0,65

0,30

<0,010

X20CrMoV 12 1

T550

0,22

0,30

0,70 11,50 1,00

0,50

0,30

<0,015

T505

0,12

0,10

0,45 10,30 1,06

0,75

0,18

0,90 <0,010 0,050 0,050

T507

0,12

0,10

0,50 10,30 1,50

0,60

0,18

<0,010 0,050 0,050

T559

0,13

0,10

0,35

0,15

0,25

X12CrMoWVNbN 10 1
(COST E)
X12CrMoVNbN 10 1
(COST F)
X12CrMoCoVNbB 9 1
(COST FB2)

1,30

Co

9,50

1,50

1,30

0,030

<0,010 0,060 0,020

0,010

Figure 1: Most common materials for gas and steam turbine components in Europe
Those higher alloyed steels are specifically the 9-10% Cr class, developed in the frame of the
European Cost research programme, where Boehler Edelstahl participated since 1987 (Cost 501),
manufacturing experimental heats as well as trial rotor shafts. Today Boehler is also active in the
current KMM-VIN action, the continuation of Cost.
During the last years Boehler Edelstahl Open Die Forge has gained a lot of experience in the
manufacturing of forgings in highly alloyed steels. Many rotor forgings, discs and hollow shafts
have been manufactured for high temperature applications primary in the creep resistant
chromium steels COST grade E (10%CrMoWVNbN), grade F (10%CrMoVNbN) and also FB2
(9%CrMoVNbNB). Figure 2 gives the scope of the forgings produced so far, covering all turbine
relevant components.
Quantity
Product

Cost E

Cost F

Cost FB2

GT-Discs

924

---

---

Hollow Shafts

414

---

---

HP-Shafts

101

60

Other Shaft Components

113

195

---

Figure 2: Turbine components made of COST grade E, F, FB2

282

1. FORGINGS FOR GAS TURBINES


There is a relatively wide range of forgings for gas turbine components, starting with the main
product as discs for the compressor and the turbines, turbine rings, followed by front, centre and
rear end rotor shafts, tie rod and nuts. Materials used so far are the 3 - 3,5% NiCrMoV steels
which have been improved substantially (super clean version to avoid or minimise long term
embrittlement) [1, 2] over the last two decades and the meanwhile well established and qualified
10% Cr Cost E.
Cost E is the more challenging material in manufacturing turbine discs and centre rotor shafts. It
has originally been developed under COST 501 [3-4] for steam turbine applications as a rotor
material, but the results in mechanical properties and long term behaviour illustrated, that the
material is also suitable for gas turbine applications. The background and requirements, which are
very different to those for steam engine applications, have already been reported [5].
To date, mainly disc forgings with a diameter of up to 2100 mm and thicknesses from 280 to 680
mm in the as-forged condition are manufactured by Boehler. Due to the high requirements on gas
turbine discs the steel is re-melted by P-ESR process.
After heating the ingots up to forging temperature, deformation is carried out on a 52 MN forging
press by upsetting, cogging and disc-forging to a final shape. Especially for cogging and discforging the manufacturing parameters, such as the forging temperature, soaking time and
deformation rate, are very important to ensure a defect free forging with uniform microstructure
and grain size distribution to achieve the high requirements on mechanical properties and
ultrasonic detectability (MDDS). In order to optimise the forging sequence and improve the
reproducibility of the forging process, the development of the forging technology has been
supported by the use of FEM modelling.

Fig. 3: FEM model for optimised forging steps to achieve uniform deformation
and reproducibility
When forged, the preliminary heat treatment is of great importance in order to provide the
necessary preconditions. After pre-machining, the QHT is performed by austenitising at 1050C
followed by double tempering to a yield strength 800 N/mm2, machining for ultrasonic testing
(UT), mechanical testing and final machining. The achieved ultrasonic detectability as well as
FATT test results are shown in Figure 4.

283

Fig. 4: MDDS and FATT properties of Gas Turbine discs in


X12CrMoWVNbN10-1-1 (COST E steel)
Permanent developments in the power generation plant business have led not only to increased
requirements for highly stressed components and their materials, but also to improved and
increased efforts in ultrasonic testing, which can be fulfilled by two automated ultrasonic
inspection facilities invested in 1999 and 2008. Conventional as well as phased array technique is
available.
2. FORGINGS FOR STEAM TURBINES
2.1. Cost E and Cost F materials
To date, many rotor forgings in different sizes and weights have been manufactured by Boehler
Edelstahl Open Die Forge, with diameters from 700 to 1180 mm and shaft ends with flange
diameters of up to 1800 mm. In total there are more than 450 forgings now in Cost E respectively
Cost F material (Fig. 2). At the beginning of industrial production of these steels the focus was
more on Cost E; meanwhile this has changed and there is a higher demand in Cost F. Both
materials are qualified and allow the new generation of fossil-fired ultra-super critical thermal
power stations operating at live steam and reheat steam temperatures of 600C and supercritical
live steam pressures up to approximately 300 bar [6]. The applied melting route used so far is the
already mentioned P-ESR process for smaller shaft parts for welded rotor constructions as well as
the BEST process for larger rotors with ingot weights up to approximately 45 to. The principle of
the process has already been reported [7, 10].
Again, as already described in the chapter for gas turbines, the ingots are heated up to forging
temperature and forged on the 52 MN forging press by mainly multiple upsetting (in case of
BEST ingots for a sufficient consolidation) followed by forging to the final shape.
When forged, the rotors are preliminary heat treated, pre-machined and quality heat treated. QHT
is performed by austenitising at 1050C 1070C followed by double tempering to a yield
strength 700 N/mm2. Final machining and testing have to be carried out prior to shipment.
Typical test results concerning US detectability are summarised in Figure 5.

284

Fig. 5: Relationship of sound path and final ultrasonic detectability (COST F steel)
Figures 6-7 show some examples of forgings produced for the power generation industry.

Fig. 6: Shaft component made of


COST steel F, Boehler grade T507

Fig. 7: Shaft component made of


COST steel E, Boehler grade T505

2.2. Cost FB2 material


Specifically, the 9 to 10% Cr grades steels have the highest potential to meet the required creep
resistance level for the critical components in steam power plants.
Under COST 501 (1983-1997) the 9-10% Cr steels E, F and B2 for use up to 600C have been
developed, trial rotor forgings with body diameters up to 1200 mm have been manufactured and a
lot of testing work has been performed on the material [8, 9]. Cost E and F have meanwhile
become standard materials for ultra-supercritical (USC) power plants with 580-600C steam
temperature. The trend to even higher steam conditions was the subject of the COST 522
programme (1998-2003), where the very promising properties of FB2 test material, produced by
Boehler and based on B2 from Cost 501 with the addition of Co, led to an upscale to industrial
heat to manufacture a trial rotor forging (Fig. 8).
Steel

Si

Mn

Cr

Mo

Ni

Co

Nb

B2

0,17

0,07

0,06

9,3

1,55

0,12

0,27

0,064

0,015

0,010

FB2 test melt

0,13

0,05

0,82

9,32

1,47

0,16

0,96

0,20

0,05

0,019 0,0085

FB2 trial rotor

0,13

0,09

0,33

9,08

1,43

0,16

1,26

0,22

0,054

0,022 0,0076

Fig. 8: Chemical compositions of steels B2 and FB 2

285

The first full size FB2 rotor forging with a final weight of 17 to was manufactured by Boehler in
2000-2001 via the Best melting route; then forgings from Societa delle Fucine Terni / Italy
(conventional steel making method) and Saarschmiede (ESR process) followed. Manufacturing
the test melts, trial rotor forgings and test results have already been reported in several papers
[10-13].
The results of creep tests gained so far show that all data lies in one narrow scatter band
confirming the trial melt behaviour of FB2. In Figure 9 [11] the values are all above the rotor B2
line, a Boron containing steel from COST 501 and the basis for the development of FB2.

Fig. 10: Microstructure of FB2:


after QHT (a); gauge section (b)
(600C, 100 MPa) after 56.500
hours [14]

Fig. 9: Creep rupture strength of Cost FB2


test rotors in comparison to other Cost rotor
alloys (with arrow: running specimens) [11]

During creep, under the influence of temperature and stress, usually dislocation densitiy
decreases, sub-grains form und grow, M 23 C 6 particles coarsen, the martensite laths become much
wider and also new phases as the Laves phase, appear. Not so in FB2; the good high temperature
stability is caused by:
- a martensitic structure with narrow martensite laths; which are decorated with and
therefore stabilised by M 23 C 6 carbides, stabilised itself by Boron [15]
- a high dislocation density, which is preserved for longer times
- and Nb, V carbonitrides (up to about 1 m); which avoids grain growth during
austenitisation and are stable during creep.
A representative TEM image of steel FB2 is shown in Figure 10, [14]
Under the influence of temperature, Laves phase also appears in FB2, but very small (0-6-0,8m)
and homogeneously distributed and therefore has no negative effect on creep.
The excellent creep resistance of FB2 and the successful transformation of trial melt behaviour to
large components led to the start of industrial production and to orders for new USC power plant
projects in Germany, USA and Asia, which have been placed within the last years.
2.2.1. Experiences in manufacturing FB2 rotor forgings
Based on the experience with the Cost FB2 trial rotor forging, the manufacturing technology had
to be fitted to the appropriate dimensions and the melting route has been changed from Best to
conventional casted ingots with total weights of 22 to 28 to due to the smaller dimensions of the

286

shafts. Till today seven forgings have been manufactured with delivery weights from 7 to 9,2
metric to.
The basic manufacturing steps for the forgings were set as follows:
Melting the steel in an EAF, than ladle furnace followed by AOD and LF with VD, casting the
ingot by bottom pouring process, homogenising, hot-forming, preliminary heat treatment, premachining, ultrasonic testing, quality heat treatment (QHT), machining and final testing (Figure
11).

Fig. 11: Flow chart of FB2 rotor manufacturing route


Forging was carried out on the 52 MN forging press, consisting of cogging, upset forging,
followed by final forging (shaping) and preliminary heat treatment (PHT) by martensitic
transformation and tempering.
After pre-machining the rotors to achieve a defect-free surface, ultrasonic tests were performed to
confirm the internal quality of the forgings and then the quality heat treatment for adjusting the
mechanical properties was carried out by autenitising at 1100C / spray quenching and double
tempering to the target 0,2 % yield strength of 650 N/mm. Double annealing was performed in
order to ensure a totally annealed martensitic microstructure.
The rotors were then machined and checked (Figure 12) using ultrasonic testing to determine the
minimum detectable defect size (MDDS). No defects could be found; the measured MDDs
ranged from 1,5 mm for a smaller part with 860 mm up to 2,2 mm for parts with a diameter of
1120 mm respectively.
There was an improvement of 0,5-0,8 mm when compared with the preliminary heat treated
condition.

Fig. 12: Quality heat-treated and machined FB2 Forging; final weight 9,2 to
287

2.2.2. Test Results


All turbine shaft parts are subjected to mechanical technological testing in order to ensure their
suitability for use. The properties are checked at different test positions; specimens in tangential
direction from the outer segments on both ends of the forging (top and bottom) and in some cases
specimens were taken from a near centre test ring of 350 mm in diameter and 300 mm from the
face end. The results of basic strength and toughness properties are summarised in Figure 13.

80

Charpy Impact +40C[J]

Edge, 1st Shaft


Centre, 1st Shaft

60

Edge, 7th Shaft


Centre, 7th Shaft

40

20

670

680

690

700

710

0,2% Yield Strength [N/mm]

Fig. 13: Yield strength and Charpy impact properties of FB2 rotors
In addition, short-term creep tests at 600C and 160 MPa load were performed on one shaft part
at t = 1300h in order to compare the creep behaviour with the results achieved on the FB2 trial
melt and the FB2 trial rotor forging from the Cost programme. Four test specimens in total, one
from each test location, were taken in tangential direction, two from the edge and two further
from the near centre position. The achieved results, plotted as a creep strain versus time curve, are
shown in Figure 14.
Up to 1200 hours testing time the creep strain is on an average on the same level as the FB2 trial
rotor.
0,45

F41, FB2 test melt


F44, FB2 test melt
F1, FB2 trial rotor
A1, FB2 production rotor
B1, FB2 production rotor
C1, FB2 production rotor
D1, FB2 production rotor

0,40

Creep strain [%]

0,35
0,30
0,25

Temp. = 600C
Stress = 160 MPa

0,20
0,15
0,10
0,05
0,00
0

200

400

600

800

1000

1200

1400

1600

Time [hrs]

Fig. 14: Results of short term creep testing in comparison to FB2 trial melt
as well as trial rotor forging
Metallographic examinations showed, that the microstructure in the edge as well as the near
centre position are tempered martensite with a grain size of 0-1 acc. to ASTM E112 in the edge
and about 00 in the near centre area. Furthermore, in all examined sample locations the
microstructure was free of -ferrite, but contained some isolated Boron nitrides.
288

In order to check the homogeneity of the chemical composition and in particular the distribution
of Boron, analysis checks were carried out again at the end positions and the near centre area.
Figure 15 represents 5 investigated forgings so far. It can be seen, that the chemical composition
is very homogenous from top to bottom as well as from edge to centre related to the ingot. The
scatter of max 8 ppm Boron over the length / cross section of the ingot demonstrates an excellent
Boron distribution.

Fig. 15: Check analysis at 4 test positions of a FB2 rotor forging


One important issue has to be taken into consideration. Adjusting the correct chemical
composition is in fact important, but not enough. Additionally, the manner of Boron in the steel is
essential. Boron has to be in solid solution in order to stabilise the M 23 C 6 carbides for a high
level of creep resistance. Also Nitrogen is necessary for the formation of the MX precipitates
which are basically Nb, V carbonitrides. If there are too many Boron nitrides (BN) in the
microstructure, the creep properties will not achieve the expected level due to a lack of B and N
in the matrix. B and N have therefore to be balanced very carefully (Fig. 8).

Fig. 16: Influence of Boron and Nitrogen on BN generation


All these results are achieved by industrial melts with at least 3,5 to and based on 9-12% Cr
steels. Larger BN were observed among and above the red line; smaller isolated ones among the
blue line. Having N contents below the blue dotted line, BNs could never be observed. Our
production rotors were in the field of some isolated and small BN, which are not avoidable and in
this size and content not detrimental to the creep properties.
289

In order to improve the quality of the forgings especially MDDS, development work is carried
out mainly in respect of further improvement the grain size for a better ultrasonic detectability.

Fig. 17: Results of grain refinement trial in laboratory


Furthermore, with respect to the fabrication of larger forgings, the intention of Boehler is more
and more focussed on the use of re-melted (PESR) materials due to a better consolidation of the
ingots compared with less segregations and a better homogeneity of the product. Test material
manufactured out of ingots up to 540 mm in diameter has already been produced successfully and
an up-scaling to a 32 to ingot is currently in progress.
2.3. Ni-Base Alloy 625
While Boron containing steels are suitable to steam temperatures of 625C and slightly higher, Ni
based alloys shall be used for temperatures of 700C and possibly above. The efficiency can be
raised from 47,5% (modern plants with 600C, 300 bar) to 55% (projected, 700/720C, 375 bar)
[16].
The metallurgical challenge hereby is the manufacturing of larger ingots, which are needed for
the hot middle part component for the HP turbine. In contrast to steel ingots, Ni Base alloys have
several limitations such as ingot sizes caused by homogeneity and segregations or the forgeability
of such alloys.
Within the frame of the European Thermie project AD700, a pilot rotor forging in alloy 625 was
forged, heat treated and tested. The rotor was manufactured by VIM-VAR process to an ingot of
diameter 810 mm and 10 to in weight. Forging was carried out on our 5200 metric tons press
using multiple upset and cogging operations to get the final machined dimension of 960 mm in
diameter. After a solution treatment of 1100C / 7,5hrs followed by water quenching, the trial
forging was ultrasonic tested, sectioned for macro and microstructure analysis, chemical analysis
and mechanical properties as well.
2.3.1. Test results
The macrostructure testing showed no segregations; the grain size was determined acc. to ASTM
E112 with values of 2-3/some 1 (surface) respectively 0-1/some 2 (centre). Mechanical properties
were examined at the centre and edge position; specimens were taken in longitudinal direction at
one end of the forging piece. The achieved results are summarized in Fig. 18.

290

Fig. 18: Results of tensile test data and microstructure in surface


and centre area of Alloy 625 trial rotor
3. CONCLUSIONS
Boehler Edelstahl Open Die Forge is an active partner for manufacturing components for gas and
steam turbine applications in higher alloyed materials up to the latest generation of high
chromium steels Cost E and F. The development work within the Cost programme by
contributing trial melts as well as full size trial rotors resulted in the fabrication of meanwhile
more than 1800 forgings made of COST steel E and F. Steel FB2, containing Boron, is now the
most promising candidate material for the next turbine generation. During the last years Boehler
has started with industrial production of components ranging from 860 mm in diameter up to
1090 mm and length dimensions from 1300 up to 2700 mm with delivery weights from 7 to 9,2
to.
In terms of 700C steam temperature it has been shown, that rotor components for the hot part of
a welded rotor construction can be manufactured out of Inc. 625 up to a diameter of 1000 mm
with sufficient mechanical properties. Concerning ultrasonic detectability an improvement in
grain size is necessary. Due to the knowledge of today a large rotor forging made of Alloy 617 is
planned to be forged in near future, using the 80 MN press at our sister company in Germany.

REFERENCE

(1) Berger C., Mayer K.H., Jaffe R.I.; Superclean 3,5% NiCrMoV Rotor Steel for

advanced Steam and Gas Turbines; 3rd Int. Conference on Improved Coal-Fired
Power Plants EPRI-Meeting, San Francisco, April 1991
(2) W. Meyer, R. Bauer, G. Zeiler: Development of Production Technology and
Manufacturing Experiences with Super Clean 31/ 2 NiCrMoV Steels, Conf. on
Superclean Steel, London, UK, March 6-7, 1995, Proc. p. 89-100
(3) Berger C., Scarlin R.B., Mayer K.H., Thornton D.V., Beech S.M.; Steam turbine
materials: high temperature forgings October 3-6, 1994, Liege, Belgium.
(4) C. Berger, S.M. Beech, K.H. Mayer, R.B. Scarlin, D.V. Thornton; High
Temperature Rotor Forgings of High Strength 10% CrMoV Steel; Chicago, Illinois,
September 11-17, 1994

291

(5) T.U. Kern, K.H. Schoenfeld; Improved 10% CrMoWVNbN Steel for Gas Turbine

Application Manufacturing and Properties; The 13th International


Forgemasters Meeting, Pusan, Korea, October 12-16, 1997
(6) K.H. Mayer, R. Blum, P. Hillenbrand, T.U. Kern, M. Staubli; Development Steps
of new Steels for advanced steam Power Plants, 7th Liege COST Conference, 29
Sep.-Oct. 2002, Liege, Belgium.

(7) Meyer W., Zeiler G., Bauer R.; Recent Developments on 9 to 12% Chromium

Steel Open Die Forgings for Steam and Gas Turbine Applications at Boehler;
14th International Forgemasters Meeting, 03-08. September 2000, Wiesbaden.

(8) T.U. Kern, M. Staubli, K.H. Mayer, K. Escher, G. Zeiler; The European Effort
in Development of new High Temperature Rotor Materials up to 650C COST
522, 7th Liege COST Conference, 29. Sep.-Oct. 2002, Liege, Belgium

(9) K.H. Mayer, T.U. Kern, K.H. Schoenfeld, M. Staubli, E. Tolksdorf; Long-term
Investigations of Specimen of 8 Production Rotors manufactured of the Advanced
Martensitic 10% Cr-Steels X12CrMo(W)VNbN 10 1 1, 14th International
Forgemasters Meeting, 03-08. September 2000, Wiesbaden, Germany.

(10)

G. Zeiler, W. Meyer, K. Spiradek-Hahn, J. Wosik: "Experiences in


Manufacturing and Long-Term Mechanical & Microstructural Testing on 9-12%
Chromium Steel Forgings for Power Generation Plants"; 4th International
Conference on Advances in Materials Technology for Fossil Power Plants (EPRI),
Conference CD, page 222-236,Hilton Head Island, South Carolina, USA, October
2528,2004

(11)

T.U. Kern, B. Scarlin, B. Donth, G. Zeiler, A. DiGianfrancesco; The European


Cost 536 Project for the Development of New High Temperature Rotor Materials
17th International Forgemasters Meeting, 03-07. November 2008, Santander

(12)

A. Di Gianfrancesco, L. Cipolla, D. Venditti, S. Neri, M. Calderini; High


Temperature Properties and Creep Behaviour of a CrMoCoB (FB2) Steel Trial
Rotor, 17th International Forgemasters Meeting, 03-07. November 2008, Spain

(13)

T.U. Kern, M. Staubli, K.H. Mayer, B. Donth, G. Zeiler, A. DiGianfrancesco;


The European Effort in Development of new High Temperature Rotor Materials
COST 536, 8th Liege COST Conference, 18. -20. Sep. 2006, Liege, Belgium

(14)

F. Kager, N. Boeck, K. Spiradek-Hahn, S. Hoefinger, M. Brabetz, G. Zeiler;


Superior Long-term Creep Behaviour and Microstructural Evolution of 9% CrSteels with Boron, 8th Liege COST Conference, 18. -20. Sep. 2006, Liege;

(15)

Lundin L.; High Resolution Microanalysis of Creep Resistant 9-12% Chromium


Steels, 1995, Goeteborg, Chalmers University, PhD Thesis

(16)

R. Blum, R.W. Vanstone; Materials Development for Boilers and Steam


Turbines Operating at 700C, 2003, Parsons Conference 2003, Proceedings page
489-510

292

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

MICROSTRUCTURAL CHANGE AFTER LONG-TERM CREEP


EXPOSURE IN HIGH CR STEEL FORGINGS FOR
ULTRASUPERCRITICAL STEAM TURBINE ROTORS
Masato Mikami
Japan Casting & Forging Corporation, Kitakyushu, Fukuoka Japan
Kota Sawada, Satoru Kobayashi, Toru Hara and Kazuhiro Kimura
National Institute for Materials Science, Tsukuba, Ibaraki Japan

ABSTRACT
Microstructural change of 10 % Cr steel trial forgings subjected to different heat treatment
conditions which aim to improve the creep rupture strength and microstructural stability during
creep was investigated. Creep rupture strength of the forging subjected to the quality heat
treatment with the austenitizing temperature of 1090 oC is higher than that of the forging solutiontreated at 1050 oC, however, the difference of creep rupture strength is reduced in the long-term
region around 40,000 h. Decrease in creep rupture ductility of the forging until 43,300 h is not
observed. Progress of the martensite lath recovery in the forging solution-treated at 1090 oC is
slower than that in the forging austenitized at 1050 oC. Higher temperature solution treatment
suppresses the recovery of lath structures. Formations of Z-phase are found in the specimens
creep-ruptured at 37,300 h in the forging solution-treated at 1050 oC and at 43,400 h in the
forging austenitized at 1090 oC. Z-phase precipitation behavior in this steel is delayed in
comparison with the boiler materials, regardless of austenitizing temperature.
KEYWORDS: 9-12 % Cr steels; turbine rotors; forgings; heat treatment; microstructure
INTRODUCTION
Higher creep strength is required for turbine components for fossil fuel power generations since
the generating efficiency increases with increase in steam temperature. Alloy design is one of
important factors to improve the creep rupture strength of 9-12%Cr steel turbine rotor forgings as
well as the boiler materials and a lot of research work on alloy design has been done [1-4].
However, the effect of heat treatment conditions, particularly solution temperature in the quality
heat treatments on creep rupture strength and microsturctural stability during creep exposure
remains unclear, whereas they control the microstructures which affect the creep deformations.
For instance, in CrMoV steel turbine rotors, it is well known that the long-term creep strength
with notched specimens is lower than that with smooth specimens in the case of forgings
austenitized at 1010 oC, so that the solution temperature is restricted to 954 oC at the highest [5].
Based on these experiences, the selection of solution temperature is very important in the case of
low alloy steel rotor forgings, but there are not so many reports regarding the effect of
austenitizing temperature on creep strength. Thus, 10 % Cr steel trial forgings have been
subjected to different heat treatment conditions in order to investigate a possibility of
293

improvement in the creep rupture strength and microstructural stability of them by means of
higher temperature solution treatment, and their creep strength and microstructural change during
creep exposure have been investigated [6, 7]. The results of previous work have revealed that the
creep strength of the forging solution-treated at higher temperature was higher than that of the
forging austenitized at lower temperature until 30,000 h testing time [7]. This paper describes the
results on creep rupture tests and investigations of the microstrucural evolution after 43,400 h
creep exposure.
EXPERIMENTAL PROCEDURES
Two trial rotor forgings examined were made from a single ingot with the weight of 69.5 metric
ton. These two forgings were manufactured through same process from the steel and ingot making
to the preliminary heat treatments [6]. In the quality heat treatments, forging-A was solution
treated at 1050 oC, and another forging-B was done at 1090 oC as shown in Figure 1. After that,
each forging was quenched in oil followed by double temperings (T T1 , T T2 ) and additional
tempering (T A ) treatments. Materials for creep tests and microstructural investigations were cut
from center of both forgings. There is no large difference in chemical composition between the
both materials as shown in Table 1. Detail information of the manufacturing sequence of forgings
are described in the previous work [6, 7]. The specimen dimensions for creep tests are 5.5 mm
7.1 mm in diameter and 25 mm in gage length, the testing temperature is 600 oC. Observations of
microstructure were performed by optical microscope, transmission electron microscope (TEM)
and scanning electron microscope (SEM) equipped with electron back scattered diffraction
(EBSD) system. Analyses of precipitates with carbon extracted replicas were examined by
scanning transmission electron microscope (STEM) with energy dispersion X-ray (EDX)
analyzer.

Quality Heat Treatment

Additional
Tempering

TT1

TT2

TA

Forging-A

1050C
(1922F)

570C
(1058F)

680C
(1256F)

680C
(1256F)

Forging-B

1090C
(1994F)

570C
(1058F)

685C
(1265F)

690C
(1274F)

Oil
Quenching

TT1

TT2

Furnace
Cooling

Furnace
Cooling

TA

Additional
Tempering

Quality Heat Treatment

Figure 1: Quality heat treatment of the trial forgings examined.


294

Furnace
Cooling

Table 1: Chemical composition of materials tested. (mass %)


C

Si

Mn

Ni

Cr

Mo

Nb

Forging-A 0.12 0.06 0.43 0.008 0.001 0.76 10.45 1.04 1.02 0.20 0.047 0.053
Forging-B 0.13 0.07 0.44 0.009 0.001 0.78 10.52 1.07 1.03 0.20 0.052 0.054
RESULTS AND DISCUSSIONS
Creep Rupture Strength of Trial Forgings
Figure 2 shows the creep rupture strength (a) and creep rupture ductility (b) of two trial forgings
as a function of time. Creep tests under 100 MPa continue beyond 60,000 h. Creep rupture
strength of the forging-B solution treated at 1090 oC is higher than that of the forging-A
austenitized at 1050 oC, however, the difference of both curves is reduced in the low stress and
long-term region around 40,000 h as shown in Fig.2 (a). On the other hand, the creep rupture
ductility of specimen ruptured at 43,300 h in the forging-B still keeps higher level, and a decrease
in ductility at long-term exposure is not observed in spite of higher solution temperature. Further
long-term tests are required to evaluate the efficacy of higher temperature solution treatment on
the creep strength.

Testing temperature: 600oC


Center

200

El (%)

Stress (MPa)

300

100

(a)

RA (%)

400

Ongoing

100
90
80
70
60
100

Forging-A (T =1050oC)
Forging-B (T =1090oC)

1000

(b)

Center

80
60
40

Forging-A, T =1050oC
Forging-B, T =1090oC

20
o

10

10

Testing temperature: 600 C


0
100
1000

10

10

Time to rupture (h)

Time to rupture (h)

Figure 2: Creep rupture strength and ductility of the forgings tested.


Microstructural change after long-term creep exposures
Microstructure of the specimens creep-ruptured at 37,300 h in the forging-A and at 43,400 h in
forging-B under the test condition of 600 oC and 120 MPa was investigated to assess the influence
of solution temperature on microstructural change during long-term creep exposure. Figure 3
shows the optical micrographs of the specimens after the additional tempering heat treatment in
the forging-A (a) and forging-B (b), together with those of creep ruptured specimens after 37,300
h in forging-A (c) and 43,400 h in forging-B (d). Tempered martensitic microstructure is observed
on each specimen, but the large difference in structures between them is not found by optical
microscope.

295

(a)

(b)

200um
(c)

200um
(d)

200um

200um

Figure 3: Optical micrographs after additional tempering heat treatment of forging-A (a),
forging-B (b), after creep rupturing at 37,300 h in forging-A (c) and at 43,400 h in forging-B (d).

TEM images of specimens after the additional tempering heat treatments and creep exposures are
shown in Figure 4. Whereas fine martensitic lath structures containing high-density dislocations
are observed after the additional tempering heat treatments in Figs. 4 (a) and 4 (b), coarsened
laths with low-density dislocations and cell-like sub-grain structures are found after creep rupture
in Figs. 4 (c) and 4 (d). Martensitic lath widths after the additional tempering heat treatment of the
both forgings are estimated to be about 0.5 m, and those width after creep rupturing of the both
forgings are estimated to be around 1 m. Figure 5 shows some preferential recovery of the lath
structure observed in the creep-exposed specimens of the both forgings at a point distant from
Figs. 4 (c) and (d), respectively. Coarsened sub-grains with about 2 m in diameter are observed
in the fine sub-micro size martensitic lath structure as shown with arrows in Figure 5. It is found
that they are long and thin shape inherited from the initial martensitic lath structures in many
cases, but specific laths in the vicinity of prior austenite grain boundary are coarsened in some
cases. The recovery of martensitic lath microstructure accompanying the preferential recovery
during creep exposure is found from the results of TEM observations. However, the influence of
the solution temperature on martensitic lath width after the additional tempering heat treatment
together with long-term creep exposure is not clear by observations on small area of thin foils.
Then, microstructural observation was performed by SEM equipped with EBSD to investigate the
preferential recovery of lath structure in the large areas.

296

(a)

(b)

2um
(c)

2um
(d)

2um

2um

Figure 4: TEM images after additional tempering heat treatment of forging-A (a), forging-B (b),
and creep ruptured specimens after 37,300 h in forging-A (c) and 43,400 h in forging-B (d).

(a)

(b)

2um

2um

Figure 5: Preferential recovery of lath structures observed in forging-A and forging-B creep
ruptured after 37,300 h in forging-A (a) and 43,400 h in forging-B (b).

297

Figure 6 shows the backscattered electron images of creep ruptured specimens on the both
forgings. There are cell-like structures displayed with gray contrast caused by the difference in
crystal orientations between martensitic laths or blocks as well as bright white particles on the
surface of those specimens. Cell-like sub-grain sizes of the creep ruptured specimen after 37,300
h in the forging-A seemed to be larger than those after 43,400 h in the forging-B. Then, crystal
orientation analyses were examined by EBSD to estimate the degree of lath recovery found in
backscattered electron images clearly. The boundary mapping and the crystal orientation image
around prior austenite grain boundaries on creep ruptured specimen of the forging-A at 37,300 h
are shown in Figure 7, together with the back scattered image. The comparison of the boundary
mapping, Fig. 7 (b), with the crystal orientation image, Fig. 7 (c) indicates that there is a close
correspondence between the prior austenite grain boundaries or the martensitic block boundaries
and the high angle boundary shown by yellow lines in Fig. 7 (c). The martensitic lath boundaries
are unclearly detected and represented as the areas where contain densely low angle boundary in
the case of fine lath structure. On the other hand, they are represented as low angle boundaries in
the case of coarsened lath structure. Therefore, the identifications of the recovered areas in the
microstructures are possible by the observations of grain boundary mapping as shown in Fig. 7
(c). Figure 8 shows the grain boundary maps on creep ruptured specimens of the both forgings.
The progression of recovery towards grain interior of the forging-A is faster than that of the
forging-B during creep exposure since sub-grain sizes expressed by black areas in the forging-A
(Fig. 8 (a)) are larger than those in the forging-B, which is consistent with the observation results
of the back scattered images as described above (Fig. 6). On the other hand, there is the
progression of preferential recovery near the prior austenite grain boundaries in the both forgings
because the areas where the densities of lath boundaries or sub-boundaries are low shown by
white lines are found in the vicinity of prior austenite grain boundary of the both specimens. It can
be said from the results of microstructural investigations that higher temperature solution
treatment seems to suppress the progression of the martensitic lath recovery in the prior austenite
grain.

(a)

(b)

10um

10um

Figure 6: SEM back scatter images of creep-ruptured specimens in forging-A at 37,300 h (a) and
forging-B at 43,400 h (b).

298

(a)

(b)

(b)

15um

15um

Figure 7: SEM back scatter image (a), boundary mapping (b) and orientation image micrograph
(c) after creep rupturing at 37,300 h in forging-A.

(b)

(a)

15um

15um

Figure 8: Boundary mapping after creep exposure at 37,300 h in forging-A (a) and at 43,400 h
in forging-B (b).

299

Change in Precipitates during Creep Exposure


Phase analyses were performed to reveal the reasons of why the higher temperature solution
treatment suppresses the recovery of martensitic structure in this steel. TEM images of carbon
extracted replicas on the specimens of both forgings after the additional tempering heat treatments
and creep exposure are shown in Figure 9. Most of particles after creep rupturing coarsen in
comparison with those after the heat treatments in both forgings. Figure 10 displays the
distribution of precipitates for each phase by composite elemental mappings. The phases found in
the specimens after the heat treatments are (Fe, Cr) 23 C 6 labeled by red color, (Cr, V) 2 X pink,
NbX green and VX blue as shown in Figs. 10 (a) and (b). Besides those precipitates, the particle
type detected after the long-term creep exposures are Fe 2 (W, Mo) Laves phases, yellow, and
other pale blue phases containing Cr, V, Nb shown with white arrows in Figs. 10 (c) and (d).
Figure 11 is the energy spectrum of the particle indicated with white arrow in Fig. 10 (c). This
particle was identified as Z-phase by comparing the intensity ratios of Nb, V and Cr [8, 9].

(a)

(b)

1um

1um
(d)

(c)

1um

1um

Figure 9: TEM images with extract replicas after heat treatment in forging-A (a), forging-B (b),
after creep rupturing at 37,300 h in forging-A (c) and at 43,400 h in forging-B (d)

300

(a)

(b)

1um
(c)

1um
(d)

1um

1um

Figure 10: Elementary mapping after additional tempering heat treatment in forging-A (a),
forging-B (b), after creep exposure at 37,300 h in forging-A (c) and at 43,400 h in forging-B (d).
White arrows showing Z-phase particles.

(a)

(c)

1um
(b)

1um

Figure 11: EDX analysis showing Z-phase particle after creep exposure at 37,300 h in forging-A.
301

Table 2 is a summary of the phases found in this research and the previous work [7] during creep
at 600 oC. Several particles of M 23 C 6 , M 2 X, NbX and few VX exist after the additional tempering
heat treatments, Laves phase appears during short-term creep exposure around 1,000 h but Z
phase is not observed after long-term creep exposure, up to approximately 28,000 h at least in this
steel. It is known that formation behavior of Z-phase in the 9-12 % Cr steels boiler materials is
affected by Cr content and Z-phase appears after 10,000 h in Grade T91 and after 2,500 h in
Grade T122 (11%Cr) at 600 oC [10]. Precipitations of Z-phase in the both forgings are delayed in
comparison with those boiler materials whereas the steel investigated contains Cr of 10.5 %.
Sawada et al. have mentioned that tempering at lower temperature (953 K) provides M 2 X instead
of VX, the presence of such M 2 X does not promote Z-phase formation and no Z-phase is
observed after creep even after 20,000 h in the case of 9Cr-1Mo-V-Nb-N steel [11]. According to
the paper, the reason of retard precipitation of Z-phase in the both forgings is assumed to have a
relationship with the tempering at lower temperature. Also, there is a possibility that the reason of
why the creep rupture strength of the forging-B decreases and close to that of forging-A in the
long-term creep region is connected with Z-phase formation. It may means that solution
temperature does not affects nucleation of Z-phase. However, it is still unclear that reason why the
higher temperature solution treatment suppresses the recovery of the martensitic lath structures
from the results of the phase analyses described above. Further creep rupture data and quantitative
evaluations of the particles in detail are required to assess the efficacy of higher temperature
austenitizing on long-term creep strength.
Table 2: Summary of particle observed in previous and this work.
Phase detected
Specimen
M 23 C 6
M2X
VX
NbX
Laves Z-phase
As HT
y
y
y
y
950h
y
y
y
y
y
Forging-A
T=1050C
15,700h
y
y
y
y
y
37,300h
y
y
y
y
y
y
As HT
y
y
y
y
1,850h
y
y
y
y
y
Forging-B
14,800h
y
y
y
y
y
T=1090C
28,500h
y
y
y
y
y
43,300h
y
y
y
y
y
y

CONCLUSIONS
Microstructural change of 10 % Cr steel trial forgings subjected to different solution tempeature
treatments which aim to improve the creep strength and microstructural stability during creep was
investigated. The results obtained were as follows:
Creep rupture strength of the forging subjected to the quality heat treatment with the
austenitizing temperature of 1090 oC is higher than that of the forging solution-treated at
1050 oC, however, the difference of creep rupture strength is reduced in the long-term

302

region around 40,000 h. Decrease in creep rupture ductility of the forging until 43,300 h is
not observed.
Progress of the martensite lath recovery in the forging solution-treated at 1090 oC is slower
than that in the forging austenitized at 1050 oC. Higher temperature solution treatment
suppresses the recovery of lath structures.
Formations of Z-phase are found in the specimens creep-ruptured at 37,300 h in the forging
solution-treated at 1050 oC and at 43,400 h in the forging austenitized at 1090 oC. Z-phase
precipitation behavior in this steel is delayed in comparison with the boiler materials,
regardless of austenitizing temperature.
REFERENCES
[1] Y. Kadoya, T. Kitai, A. Matsuo, I. Tsuji, M. Kishimoto, T. Tsuchiyama and M. Okamura,
Practical Application of an Advanced 12Cr Rotor to 593oC-700MW, Tetsu-to-Hagane,
Vol. 78 (1992), pp. 910-917.
[2] M. Yamada, O. Watanabe, Y. Yoshioka and M. Miyazaki, Development of Advanced
12Cr Steel Rotor Forgings, Tetsu-to-Hagane, Vol. 76 (1990), pp. 1084-1091.
[3] M. Shiga, Y. Fukui, S. Kirihara, R. Kaneko, F. Ito and S. Sugai, Improved 12Cr Rotor
Forging for Ultra-Super-Critical Steam Turbine, Tetsu-to-Hagane, Vol. 76 (1990), pp.
1092-1099.
[4] C. Berger, S. M. Beech, K. H. Mayer, R. B. Scarlin and D. V. Thornton, High Temperature
Rotor Forgings of High Strength 10% CrMoV Steels, 12th International Forgemasters
Meeting, Chicago Illinois (September 1994)
[5] J. D. Conrad and N. L. Mochel, Operating Experience With High-Temperature SteamTurbine Rotors and Design Improvements in Rotor-Blade Fastening, Trans. ASME, Vol. 80
(1958), pp. 1210-1224.
[6] M. Mikami, Y. Wakeshima, T. Miyata and K. Kawano, Manufacturing Experiences and
Investigation of Properties of 12%Cr Steel Forgings for Steam Turbines, Proceedings of 5th
International Conference on Advances in Materials Technology for Fossil Power Plants,
Marco Island Florida USA (2007).
[7] M. Mikami, Y. Wakeshima and T.Miyata, Creep Rupture Strength and Microstructural
Investigation of 12 % Cr Steel Large Forgings for Ultra-Supercritical Steam Turbine Rotors,
Advances in Materials Technology for Fossil Power Plants, Proceedings of 6th International
Conference, Santa Fe, New Mexico, USA (2010), pp. 408-422.
[8] Strang and V. Vodarek, Z Phase Formation in Martensitic 12CrMoVNb Steel, Materials
Science and Technology, 1996, pp. 552-556.
[9] K. Sawada, H. Kushima and K. Kimura, Z-phase Formation during Creep and Ageing in 912% Cr Heat Resistant Steels, ISIJ International, Vol. 46 (2006), pp. 769-775.
[10] K. Sawada, H. Kushima, K. Kimura and M. Tabuchi, TTP Diagrams of Z Phase in 9-12%
Cr Heat-Resistant Steels, ISIJ International, Vol. 47 (2007), pp. 733-739.
[11] K. Sawada, K. Suzuki, H. Kushima, M. Tabuchi and K. Kimura, Effect of tempering
temperature on Z-phase formation and creep strength in 9Cr-1Mo-V-Nb-N steel, Materials
Science and Engineering A 480 (2008), pp. 558-563.

303

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

EXPERIENCE IN MANUFACTURE OF HIGH CHROMIUM FORGED ROTOR STEELS


A. Di Gianfrancesco, S. Budano, P. Lombardi, M. Paura: Centro Sviluppo Materiali,
Roma, Italy
S. Neri, M. Calderini, N. Longari: Societ delle Fucine, Terni, Italy
Abstract
The target to increase the steam parameters, temperature and pressure, in the new
power generation plants for reduction of CO2 emission generated the development of
enhanced high Chromium steels with to temperature with improved creep behaviours
and stability at service temperature.
After several years of research activities in Europe currently the composition Rotor E,
developed in the frame of COST programs (501, 522 and 536), become a commercial
product.
This forged steel is mainly produced by remelting of ingot by Electro Slag Remelting
(ESR), but this paper demonstrate that also the conventional route, without ESR, can be
also used to produce successfully large rotor components using tailored process control
of the production steps.
These paper describe the current production of Societ delle Fucine for Rotor E steel
grade by conventional process route, based on ladle furnace and vacuum degassing,
and the mechanical and creep behaviours of the forged products.
Societ delle Fucine (SdF) produced also a FB2 prototype rotor using a conventional
process route. In fact in the COST 522 the best candidate, coded FB2 - a 10%Cr steel
with additions of Co and B, without W - was selected for scale-up from laboratory trial to
full industrial component. In fact the addition of Boron was successfully adopted to
increase the stability of the microstructure and as consequence to improve the creep
behaviour of the new enhanced martensitic steels.
Also the update of the long term characterization program of trial rotor made by FB2
steel grade will be also presented.
Keywords: COST 501-522-536, 10%Cr Steels, High Temperature Application, Creep
strength, microstructural evolution, Boron effect.
1. Introduction
The energy production is faced with the introduction of more and more stringent
emission regulations to safeguard health and to preserve the environment for the future
generations.
The thermal efficiency is influenced by several factors, but the adoption of
ultrasupercritical conditions by increasing steam temperatures and pressures plays a
key role. On the other hand, the increase of steam parameters from 600C up to
650C/300bar will generate an efficiency improvement of 8-10% with a corresponding
CO2 reduction [1]. These very high temperatures and pressures make mandatory the
use of steels suitable for these severe conditions [2]. In fact these advanced steam
parameters require materials with adequate creep strength and resistance to oxidation.
Experience with austenitic materials was unsatisfactory showing considerably
restrictions in the operational flexibility of the plants due to the difference in the thermal
expansion between austenitic and ferritic components and the consequent stresses. The

304

class of the 9-12% Cr steels offers the highest potential to meet the required creep
resistance level for the critical components in steam power plants.
In Europe the main efforts to improve the 9-12%CrMoV steels were concentrated in the
COST (CO-operation in Science and Technology) Programmes: COST501 (1986-1997),
COST 522 (1997-2003) [1-6]. In these programmes new ferritic steels for forging,
casting and pipework were developed and characterised to increase the operating
steam temperatures:
- from the subcritical plant: with live steam pressure (PLS) of 166bar, live steam
temperature (TLS) of 540C and reheat steam temperature (TRH) of 540C,
- to the supercritical plant: PLS 250bar, TSL 540C, TRH 560C,
- to the first generation of Ultra Super Critical (USC) plants PLS 270bar, TLS 580C,
TRH 600C,
- up to the second generation USC plants: PLS 300bar, TLS 600C, TRH 620C.
The qualifications of these materials are still on-going after the end of COST 536
Programme (2004-2009), in to the KMM-VIN, WG2 EMEP.
2. Materials development for 600C turbine rotor applications
In COST 501 forged and cast 9 to 10%Cr steels were developed with additions of
1.5%Mo (type E) or a combination of 1%Mo and 1%W. They showed much improved
creep strength, resistance to embrittlement in operation and weldability [1]. In addition
samples from production components were subjected to low cycle fatigue and long term
creep testing permitting a statistical evaluation of the results. They are in use at
temperatures up to 600C.
The trend to even higher steam conditions was the subject of the COST 522 programme
which explored the possibilities of stabilising the tempered martensitic microstructure
through addition of small quantities of Boron. These are the steels (forged steel FB2 and
cast steel CB2) now being employed in orders currently being executed in Germany and
the USA.
Within COST 501 a series of advanced steels for forgings, castings and pipe/tube
application as given in Table 1 was qualified and currently the Type E composition is
utilized for rotor manufacture.
After the good results obtained in the COST 501 on composition E, F and B a new
modified steel called FB2 was produced as trial melt. Furthermore the promising
properties of the trial melt, at the beginning of the new COST 522; it was decided to
scale up FB2 steel to industrial heat in order to manufacture a trial forged rotor.
Boehler/Austria has manufactured a full-size rotor forging with a final weight of
17,000kg. The steel making process was Boehler-BEST, consisting in a pouring process
with special measures to improve the homogeneity of the ingot [7].
A second proposal for a FB2 trial rotor manufacturing was brought into COST522 by the
Italian Societ delle Fucine (SdF) in Terni; it produced a 52,000kg ingot by conventional
steel making (Ladle Furnace and Vacuum Degassing) with a final rotor weight of
28,000kg [2].
One more trial rotor has been produced by ESR remelting of 57,000kg ingot in
Saarschmiede [8].

305

Forged
Steels
1CrMoV

Cr

Mo

0.25

1.0

1.0

0.25

12CrMoV

0.23

11.5

1.0

0.25

501

Type F

0.1

10

1.0

501

Type E

0.1

10

501
522

Type B
0.2
Type FB2 0.13
(SdF)

9.0
9.32

COST

Ni

Nb

0.2

0.7

0.05

0.05

1.5

0.2

0.6

0.05

0.05

1.5
1.47

0.2
0.2

0.1
0.16

0.05
0.05

0.02
0.019

0.01
0.085

100MPa Status
100.000h
550C
Long
term
operating
570C
Long
term
operating
597C
Operating
in
plant
597C
Operating
in
plant
620C
Trial
rotor
manufactured

Table 1: Compositions of improved ferritic steels developed in COST 501 and operating
temperature for 100MPa/100.000h. [2]
3. The manufacture of FB2 trial rotor at Societ delle Fucine (SdF)
The aim of the investigations is to qualify the different steelmaking processes for this
class of Boron containing 10CrMoCoVB alloys. One of the main tasks is how to increase
and optimise the composition homogeneity and the properties of the final rotor forging
are. Figure 1 shows the lay-out of the SdF manufacture route: from the melting shop to
the final machining. Due to the characterisation work on experimental rotor the
component machining was stopped at the step of NDT control after quality heat
treatment.

Figure 1: flow chart of rotor manufacturing route at SdF


The chemical composition of the SdF trial rotor is based on the FB2 trial melt from COST
501. Table 2 shows the chemical analysis of the cast product. Very good agreement with
the aimed composition has been obtained with very low content of residual elements.

306

FB2
min
max
cast

C
0.12
0.14
0.14

Si
0.006
0.032

Mn
0.3
0.4
0.32

P
0.01
0.007

S
0.005
0.003

Cr
9.0
9.5
9.1

Mo
1.45
1.55
1.5

Ni
0.1
0.2
0.14

Al
0.008
0.001

B
0.006
0.009
0.009

Co
1.2
1.4
1.23

N
0.015
0.030
0.015

Nb
0.04
0.06
0.046

V
0.18
0.22
0.2

Table 2: Required and obtained chemical composition of FB2 trial forged rotor
(residual elements: H2= 1,2ppm; Sb= 0,001%; Sn=0,001%; As=0,006%; Cu=0,035%; W <0.01%)

The final dimensions of the trial component are shown in Figure 2. After forging the trial
rotor has been treated as follows:
- Austenitizing: 1100C/ 17h/oil quenched;
st

- 1 Tempering: 570C/ 24h/ air cooled;


nd

- 2 Tempering: 700C/ 24h/ air cooled.


The trial component in as treated condition and after preliminary machining has been
subject to NDT ultrasonic inspection (US) with a 2MHz source. The maximum defects
discovered (flat bottom hole equivalent) in the different positions are summarised in
Table 3 [9]. Figure 2a show the dimension of the trial component after final forging and
heat treatment and Figure 2b during the NDT control tests.
FB2 SdF Trial rotor
Dimensions (mm)

Total (L1+L2+L3)
4380

D1
D2 D3
925 1110 790

Max Defect discovered (mm)

1.0

1.5

Table 3: results of the US inspection on SdF trial rotor

Top
D1

L1

D
E

D2

L2

Bottom
D3

L3

307

1.5

a
b
Figure 2. COST trial rotor forgings (type FB2) dimensions [9]; SdF rotor during US NDT
control.
3.1 Product characterisations
In order to verify the homogeneity of the properties of the trial rotor, the specimens for
chemical analysis and mechanical tests have been obtained from different parts of the
component (A,B,D,E and C) positions from the core (see Figure 2a). The chemical
analysis (Table 4) shows a very good homogeneity in the composition of the main alloy
elements as well as for the Boron and Nitrogen content that could be critical for their
distribution in a large component processed without remelting.
Position/Element
A
B
D
E
FB2 trial melt
C
0,12
0,13
0,13
0,15
0,13
Cr
9,08
9,13
9,11
9,02
9,32
Mo
1,57
1,59
1,59
1,62
1,47
Ni
0,14
0,14
0,14
0,14
0,16
V
0,21
0,21
0,21
0,22
0,20
Nb
0,054
0,054
0,054
0,060
0,05
B
0,0095
0,0096
0,010
0,012
0,0085
N
0,015
0,016
0,015
0,014
0,019
Co
1,28
1,27
1,27
1,29
1.23
Table 4: chemical composition (wt%) of forged trial rotor in different positions compared
with the nominal composition.
The chemical composition of FB2 steel guarantees a fully martensitic structure. Also with
very low cooling rate, corresponding to claim air, no traces of ferritic transformation
appear. In fact the microstructural analysis performed after the final heat treatment
(normalizing + tempering) (Figure 3a) shows a typical tempered martensitic structure
with a 0-2 ASTM grain size. In Figures 3b is showed the FB2 microstructure as obtained
by Transmission Electron Microscopy on thin foils. The prior austenite and the
martensitic lath boundaries are decorated by a typical precipitation of M23C6 carbides.
The microstructure contains a high dislocation density.
Conventional mechanical tests have been performed in different positions and
orientation. The results, just discussed in previous papers [10-11], shown a quite good
homogeneity of the obtained values in term of YS, UTS and ductility: elongation and
reduction of area.

308

400 m

2 m

Figure 3: fully tempered martensitic microstructure of trial rotor after quality heat
treatment: a) by Light Microscopy; b) by Transmission Electron Microscopy on thin foil
3.2 Microstructural characterisations
In the COST 536 Alloy Design Group discussion it is still open on the effect of B addition
not only as M23C6 and grain boundary stabiliser, but also on the correct amounts of B
and N to avoid the formation of BN particles as defined by Abe & Co-workers [12] in
figure 4. The formation of these particles have the double negative effect to reduce the
amount of B available in solution and able to diffuse into the M23C6 and otherwise to form
particles reducing the ductility in creep as demonstrate for the FB2 trial melt.
The FB2 material from SdF it is on the border line of the equilibrium (red point in the
figure 4) and therefore BN formation could be present.

Figure 4: BN equilibrium diagram [12]: circle point is the FB2 from SdF
Then investigations have been made by FEG-SEM +EDS on RT tensile test specimens
fracture surface: no BN pure particles have been identified, but some dimples appears
sometime associated with particles identified as complex inclusions that contain also BN
particles:
- Figure 5a shows a BN particles jointly with a MnO inclusion,
- Figure 5b shows a BN associated with other complex oxide
309

Figure 6 shows a sulphide inclusion with BN particle.

As conclusion of these analyses the following consideration can be made: SdF FB2
seems have a quite good balancing of B and N, because pure particles have not been
discovered and the dimensions of the discovered BN particles are very small. The SdF
FB2, conventionally processed without remelting, shows small sulphide and oxide
inclusion where sometime BN can nucleate.

Figure 5: a) BN + MnO on FB2 tensile specimen; b) BN plus MnO, Al and Ti on FB2


tensile specimen

Figure 6: particle of BN plus MnS on FB2 tensile specimen


3.3 Creep properties
A large creep test programme was defined in the Turbine Working Group of COST 536
to qualify the trail rotor and the test have been carried out by CSM. Notched and unnotched specimens have been machined from positions A, B and C and tested in the
temperature range 600-650C with continuous strain measurements. Some tests have
been planned to reach the rupture in 100.000 hours. The creep results obtained on the
SdF trial rotor confirm with the behaviour of the 500kg trial melt [9]. Figure 7 summarises
the current status of the tests compared with the master curves (Larson-Miller) of COST
501 rotors E, F and B2. Some specimens are still running: 3 at 625C currently in the
range of 25.000 hours and 1 notched specimen at 600C 130MPa that is close to the
80.000 hours.
It possible to observe that the results of FB2 SdF trial rotor are in the upper band of the
previous trial rotors and the ductility of the creep specimens broken up to now it is very
310

good both in term of elongation and reduction of area, as shown in figure 8 without
evidence of ductility drop for the longer tests at the different temperatures.

Figure 7: Creep master curve of tests of SdF FB2 trial rotor (open point = specimens
running) compared with B2 and E + F rotors curves [15]
90

Reduction of Area (%)

30

Elongation (%)

25
20
15
10
5
0
19,5

20

20,5

21

21,5

22

22,5

PLM (C=20)

80
70
60
50
40
30
20
10
0
19,5

20

20,5

21

21,5

22

22,5

PLM (C=20)

Figure 8 elongation and reduction of area of broken specimens


3.4 Microstructural evolution
In parallel with the mechanical and creep tests a programme to investigate the
microstructural evolution of the FB2 steel is started. The thermodynamic tools predict, at
equilibrium the main presence of M23C6 carbides and M(C,N), Laves phase and a little
amount of Z-phase. As well known the Z-phase appearance has been identified as the
main reason of the drop of the creep properties of all the recently developed 10-12%Cr
steels [12].
The microstructural investigations have been carried out by TEM quantitative
investigation with EDS analysis on the as treated material and on specimens after aging
at different temperatures.
The forged material after heat treatment is characterised by a fine precipitation of M23C6
and, in the meantime, by a low amount of larger particles of MX: their dimensions and
mean chemical compositions is reported in table 5 and can be summarised as:
- No fine MX are present after heat treatment,
- The large MX are bigger then MX in 9%Cr steels,
311

M23C6 are 20-40% smaller then M23C6 in 9% grades,


M23C6 frequency is much higher than the in 9% grades.

Precipitates/
Dimensions
(nm)
MX

FB2
SdF

Grade 92 Grade 911


1070+780 1060+760

180

45

M23C6

98

125

Grade 91
1070+780

Frequency
(%)

Cr
(%)

Mo
(%)

V
(%)

Nb
(%)

Fe
(%)

50

35

17

68

130

139

96

56

14

1,5

22

Table 5: mean dimension of precipitates present in N+T forged FB2 compared with the
other and quantitative analysis (% in weight) of FB2 precipitates [13-14]
3.5 Microstructural analysis on aged specimens
Microstructural analysis by SEM+ EDS and STEM+EDS started on crept aged
specimens in order to investigate the evolution of the precipitates: the first 2 specimens
have been selected after 23.576 hours at 625C and after 10.026 hours at 650C (figure
9).
It was possible to observe that in aging condition there are:
- still a relevant amount of small M23C6 particles without appearance of
coarsening process,
- the precipitation of some small MX particles, not present in the as treated
condition,
- the appearance of Laves phase in the specimen aged at 625C,
- the start of transformation of the bigger MX in to Z-phase (only 1 particle
analysed) (Figure 10).

Figure 9: Microstructure after 23576 hours at 625C and after 10026 hours at 650C

312

Figure 10: Precipitation after 23576 hours at 625C; a) general view; b) Z-phase particle
(extraction replica)
The M23C6 dimensions in FB2 aged specimens have been compared with the dimension
of the precipitates in the grade 91: it was observed that the coarsening phenomena are
strongly reduced [14-18]. This effect could be related to the introduction of B in the
chemical composition, that seems have a stabilising effect on the M23C6 and on grain
boundary [19-21], but this item is still under discussion worldwide.
4. Type E Rotor
After this quite good results on this experimental steel in year 2001 Societ delle Fucine
decided to made some other forged rotors in Steel X12CrMoVWNb10-1-1 (COST Type
E), following Technical Specification suggested by COST program. The ingots have
been melted in ThyssenKrupp Acciai Speciali Terni.
An example of the chemical composition (%wt) obtained in one delivered rotor is
reported in Table 6. It is possible to observe that all the elements in the component are
in the range the standard. The other residual elements analyses are: Cu<0.06%; Sn
<0.003%; As <0.004%.

Table 6: Chemical composition of a rotor produced with Steel X12CrMoVWNb10-1-1


(COST Type E)
Manufacturing process for Rotor E components is the same of FB2, described in the
scheme just illustrated in Figure 1. The CCT diagram of the type Steel is reported in
figure 11.
After completion of the first rough machining, the forging rotors were subject to quality
heat treatment, for instance, as follow:

Austenitizing at T=1070C; oil quenching


st

1 Tempering at T=520C, and air cooling;


nd

2 tempering at T=710C, and air cooling.


313

Figure 11: CCT curve of Type E steel


Tempered martensite microstructure of one rotor component has been obtained in all the
different positions of the components, as shown in figures 12.

Figure 12: Tempered martensite microstructure in different positions of the rotor (x100;
Vilella etching)
An example of the basic mechanical characteristics of Rotor E components are shown in
table 7. The scattering of strength values are quite low: these results confirm the good
heat treatment practises and the homogeneous mechanical properties. [23].
The FATT value has been evaluated as 38C. Tensile tests have been carried also in
the temperature up to 620C. Figure 13 shows the tensile tests results which are
summarized in in table 8.
The material strength values are in the range of nominal values of the Rotor E [17]. No
recordable indications were found in all the forging components.

314

Typical requested values


min
max
MPa
800
900
MPa
700
%
15
%
Joule
30

UTS
YS
El.
RoA
KCV

SdF
nim
824
706
17
54
35

max
880
740
19
63
96

Table 7: Tensile tests report


1000

1000

900

900

800

800
700

600

T.Stress [MPa]

T.Stress [MPa]

700

500
400
300

Rotor E R.T.
Rotor E 300C
Rotor E 550C
Rotor E 620C

200
100
0
0.00

0.04

0.08

600
500
400
300

Rotor E R.T.
Rotor E 300C
Rotor E 550C
Rotor E 620C

200
100

0.12

0.16

0.20

T.Strain [mm/mm]

0
0.000

0.004

0.008

0.012

0.016

0.020

T.Strain [mm/mm]

Figure 13: Rotor E Tensile test results at different temperatures


Temperature
C
RT
300
300
550
600
620

UTS(MPa)

YS (MPa)

Elongation
(%)
882
775
16
760
665
15
745
658
15
513
484
16
409
383
18
378
346
18
Table 8: Rotor E tensile tests report

Red. Of
Area (%)
59
58
60
83
89
89

In order to have a full qualification of the material Rotor E a large and comprehensive
testing program including tests on smooth and notched specimens, cracked and uncracked specimens were planned as follows:
Creep program: 550-600-650C,
high cyclic fatigue tests (HCF),
Low cycle fatigue tests at high temperature (HCF),
Thermo-mechanical fatigue tests (TMF),
Crack growth tests at high temperatures and static loads and under isothermal
conditions (CG),
Fatigue and creep/fatigue crack growth (CCG/FCG) characterisation of crack
propagation on CT specimens under TMF condition.
315

4.1 Fatigue tests


The high cyclic fatigue tests are designed with the scope to determine the maximum
cycling load without promoting a catastrophic failure of the component. The HCF tests
were conducted under load control. The cyclic loading was applied in tensioncompression with sinusoidal waveform. The value of the relationship between the loads
R=Pmin/Pmax is equal to: R=-1. The cyclic loading frequency was f=121Hz.
Figure 14 shows the HCF test results at room temperature, the y-axis shows the cyclic
load applied, it is expressed as:
The x-axis indicates the number of cycles at which the specimen fails at an imposed .
Each point, therefore, is characterised by an imposed load and the cycle number at
which the specimen has failed [22, 23].
Figure 14 shows also the results of the HCF tests performed at high temperature using
an induction furnace. They were later subjected to a sinusoidal waveform tensilecompression cyclic load, with a load factor of R= -1, and a cyclic loading frequency of
113Hz.
7
Under these conditions, the f fatigue limit (the f at which the Nf = 10 cycles) was
estimated at: f= 380MPa with a confidence level of 95%. The red round symbol
represents an HCF test with imposed load of =375MPa which was stopped after 40
million cycles.
950

900

620C
Linear Fit
Confidence bands (95%)

400

850

[MPa]

[MPa]

420

RT
Linear Fit
Confidence bands (95%)

800
Equation
Adj. R-Square

380

y = a + b*x
-0.08046
Value

750

Ds

Intercept

Ds

Slope

3.03557

0.1175

-0.01448

0.01828

1000000

Equation
Adj. R-Square

Standard Error

y = a + b*x
0.78187
Value

1E7

Intercept

549.74927

32.34484

Ds

Slope

-23.30839

4.91309

1000000

Cycles to failure

Standard Error

Ds

1E7

1E8

Cycles to failure [n]

Figure 14: Rotor E HCF tests at RT HCF tests at 620C

LCF tests were carried out in agreement with ASTM E606. The LCF tests were
performed at three temperature levels: room temperature (RT), 300C and 620C. The
test temperatures were achieved by an induction furnace. The actual temperature test
was measured and monitored by thermocouples welded directly onto the specimen. The
test has been conducted in strain control with imposed strain rate of 0.01 s-1 with the
triangle waveform, and a load ratio R= -1. Figure 15 summarize the LCF results in
comparison with the HCF results.

316

Figure 15: Comparison between LCF and TMF results


4.2 Creep Tests
The creep test program has been planned at 550-600-650C. Figures 16 shown the
creep result at three temperatures; the creep results at all the temperatures are in
agreement with the Cost E mean lines, the red symbols with arrows represent the still
running tests: all the tests have been performed with strain measurements.
400

400
350

350

300

300

250

Stress [MPa]

200

200

150

100
100

150

100

Rotor E 600C
Cost E 600C

Cost E 550C
Rotor E 550C
1000

10000

100000

50

50

100

1000

Time to rupture [h]

Time to rupture [h]


200

150

Stress [MPa]

Stress [MPa]

250

100

Rotor E 650C
Cost E 650C
50
100

1000

10000

100000

Time to rupture [h]

Figure 16: Creep tests at 550C, 600C, 650C [15]

317

10000

100000

Rotor E creep data have been elaborated to evaluate nmin and Amin .The minimum strain
rate,
has been plotted against stress to find nmin, Amin for each temperature. Table
9 shows the nA and AA Rotor E found parameters at the different temperatures.
Rotor E

nmin

Amin

550C
9.82
3.51E-30
600C
6.78
3.39E-21
650C
3.18
2.11E-12
Table 9: Summary of the Rotor E creep parameters
5. Summary and Conclusions
The result obtained up to now on FB2 trial rotor and on Steel X12CrMoVWNb10-1-1
(COST Type E) produced at Societ delle Fucine, without ingot remelting, demonstrate
that it is possible to realise full scale forged components by conventional process route
with very good homogeneity of chemical analysis, mechanical behaviour and creep
properties.
The FB2 from SdF has a quite good balancing of the B and N content, therefore a very
low amount of BN that are present and there are always in connection with MnS or
inclusion oxides particles.
In term of creep behaviour FB2 seems to offer a significant improvement over the
parent E-F-B type steels and it could be suitable for 625C applications. In fact the
experience gained during the last few years seems guarantee that the 9%Cr are more
stable materials from the microstructural evolution point of view, and suggests that the
formation of some Z-phase precipitates may not significantly affect by mean of the long
term creep behaviour of this steel.
Obviously these results have to be confirmed with longer creep tests in order to
generate a more consistent database, assessment and extrapolation. Tests still running
are planned to reach 100.000 hours life.
The TEM investigation on long term aged specimens shown that the microstructural
evolution is significantly different respect to the other 9%Cr steels: the dimension of the
M23C6 seems not slightly affected by time and temperature precipitates and few Z-phase
particles have been found in FB2 after 23576 hours at 625C.
If the slower coarsening of the M23C6 particles will confirmed by the analysis of the more
aged specimens, these will be an experimental confirmation of the positive effect the
Boron on the microstructural stability of this steel and therefore on the creep behaviours.
Also the parent Steel X12CrMoVWNb10-1-1 (COST Type E) has been successfully
produced in Societ delle Fucine and the microstructural analysis, mechanical and creep
tests shown properties in agreement with the literature values.
Currently about 20 rotor components have been produced in Societ delle Fucine
starting from a maximum ingot of 75tons: the dimension of the components have a
maximum delivered weight of 27 tons and a maximum diameter of 1200mm.
Acknowledgement
The authors are grateful to their colleagues and partners in the programme COST522
and 536 for their contributions and many discussions during the course of this
endeavour. Many thanks are also extended to the COST Management Committee for
their guidance. A special thanks to Dr. Silvia Tiberi Vipraio, Dario Venditti, Marcello
nd
Ballone (CSM) for the TEM investigations and Dr. Riccardo Polini (2 University of
Rome, Tor Vergata) for the support in FEG-SEM analysis.
318

REFERENCES
1. T.U. Kern, K. Wieghardt, H. Kirchner: Material and design solutions for advanced
steam power plants, EPRI Fourth International Conference on Advanced in
Materials Technology for Fossil Power Plants; October 25-28, 2004, Hilton Head
Island, South Carolina USA
2. R.W. Vanstone, Alloy design and microstructural control for improved 9-12%Cr
power plant steels, Annex A, COST 522 Steam Power Plant, Final Report, 19982003
3. B. Scarlin, R. Vanstone, R. Gerdes: Materials developments for Ultrasupercritical
Steam Turbines, EPRI Fourth International Conference on Advanced in Materials
Technology for Fossil Power Plants; October 25-28, 2004, Hilton Head Island,
South Carolina USA
4. B. Scarlin, T-U. Kern, M. Staubli: The European efforts in material development for
650C USC power plants cost522, Ibidem
5. Y. Tanaka, T. Azuma, K. Miki: Development of steam turbine rotor forging for high
temperature application high temperature steel forgings for power generation,
Ibidem
6. P. Peel, B. Scarlin, R.Vanstone: From materials development to advanced steam
turbines: PARSONS 2007: 7th International Charles Parsons Turbine Conference,
Power Generation in an Era of Climate Change, 11-13 September 2007, Univ. of
Strathclyde, Glasgow, UK.
7. G. Zeiler, W. Meyer, K. Spiradek, J. Wosik: Experiences in manufacturing and longterm mechanical & microstructural testing on 9-12 % chromium steel forgings for
power generation plants, EPRI Fourth International Conference on Advanced in
Materials Technology for Fossil Power Plants; October 25-28, 2004, Hilton Head
Island, South Carolina USA
8. N. Blaes, B. Donth, K.H. Schnfeld, D. Bokelmann: High temperature steel forgings
for power generation, Ibidem
9. T.U. Kern, M. Stabli, K.H. Mayer, B. Donth, G. Zeiler, A. Di Gianfrancesco: The
European effort in development of new high temperature rotor materials COST
536: Int. Conf. Materials for Advanced Power Engineering, 19-21 September 2006
Liege, Belgium
10. T.-U. Kern, B. Scarlin, B. Donth, G. Zeiler, A. Di Gianfrancesco: The European
COST536 project for the development of new high temperature rotor materials: 18
International Forgemaster Meeting: 3-6 November 2008 Santander, Spain
11. A. Di Gianfrancesco, L. Cipolla, D. Venditti, S.Neri, M. Calderini: High Temperature
Properties and Creep Behaviour of a CrMoCoB (FB2) Steel Trial Rotor, Ibidem
12. F. Abe: Alloy Design of Creep and Oxidation Resistant 9Cr Steels for Thick Section
Boiler Components Operating at 650C: EPRI Fourth International Conference on
Advanced in Materials Technology for Fossil Power Plants; October 25-28, 2004,
Hilton Head Island, South Carolina USA
13. A. Di Gianfrancesco, L. Cipolla, D. Venditti, S.Neri, M. Calderini: Creep Behaviour
and microstructural stability of FB2 (CrMoCoB) Steel Trial Rotor Sixth International
Conference on Advances in Materials Technology for Fossil Power Plants August
31September 3, 2010 Santa Fe, New Mexico, USA
14. Microstructural evolution of ASTM P91 after 100,000 hours exposure at 550C AND
600C: L. Cipolla, S. Caminada, D. Venditti, H. K. Danielsen, A. Di Gianfrancesco:
9th Lige Conference on Materials, for Advanced Power Engineering: September
27th 29th, 2010, Lige Belgium

319

15. The European efforts in development of new high temperature rotor materials
COST536: T.-U. Kern, K.H. Mayer, B. Donth, G. Zeiler, A. Di Gianfrancesco: ibidem
16. Long term creep properties and microstructural evolution of IN718 prototype forged
disk: M. Calderini, S. Neri, A. Di Gianfrancesco, P. Lombardi, L. Foroni, R. Montani:
18th International Forgemasters Meeting: September 12-16, 2011 Pittsburgh, PA,
USA
17. J. Hald, H. Danielsen: Z-phase in 9-12%cr steels: EPRI Fourth International
Conference on Advanced in Materials Technology for Fossil Power Plants; October
25-28, 2004, Hilton Head Island, South Carolina USA
18. S. Caminada, G. Cumino L. Cipolla, A. Di Gianfrancesco: Long term creep
behaviour and microstructural evolution of ASTM grade 91 steel; ibidem
19. Fujio Abe: Effect of boron on creep deformation behaviour of 9cr steel for USC
boilers at 650C: PARSONS 2007: 7th International Charles Parsons Turbine
Conference, Power Generation in an Era of Climate Change, 11-13 September
2007, Univ. of Strathclyde, Glasgow, UK.
20. H.O. Andren, A. Golpayegani: Microstructure of a high boron chromium steel for
steam turbine applications: ibidem
21. H.O. Andren, A. Golpayegani: Creep resistant high boron 9-12% chromium steels
for steam power plants: Int. Conf. New Developments on Metallurgy and
Applications of High Stregth Steels: Buenos Aires, Argentina 26-28 May 2008
22. P. Lombardi, S. Budano, A. Di Gianfrancesco, K. Nikbin, F. Biglari, M. A. Gomes, A.
Sanguineti, E. Poggio, S. Neri: Development of methods for the characterisation,
fracture assessment and life prediction of new high strength steels under variable
temperature operating conditions: DELOC: RFSR CT 2007-00021; 01/01/2007
30/06/2010, Final Report
23. Y. W AKESHIMA, M. MIKAMI, MANUFACTURING OF TRIAL ROTOR FORGINGS OF COST E
STEEL (X12CRMOWVNBN10-1-1), IFM 2006, 16TH INTERNATIONAL FORGEMASTERS
MEETING,15 19 OCTOBER 2006, CUTLERS HALL, SHEFFIELD, UK

320

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

MANUFACTURING OF TRIAL ROTOR FORGING OF 9%CR


STEEL CONTAINING CO AND B (X13CRMOCOVNBNB9-2-1)
FOR ULTRASUPERCRITICAL STEAM TURBINES
M.Nakano
K.Kawano
M.Mikami
Japan Casting & Forging Corporation
46-59, Sakinohama Nakabaru, Tobata-ku, Kitakyusyu-city 804-8555, Japan

ABSTRACT
9%Cr steel containing Co and B, X13CrMoCoVNbNB9-2-1, has been manufactured by electro
slag remelting to confirm the qualities and compare creep strength and microstructure of that
forging with those of a forging made from electro slag hottopping ingot. It is confirmed to
manufacture rotor forgings with homogeneous composition and good properties by electro slag
remelting process. Creep strength of the forging made from electro slag remelting ingot is similar
to that of the forging applied electro slag hottopping process from the results of creep rupture tests
up to 5000h. Martensitic lath microstructures with high density dislocations and the precipitations
of M 23 C 6 , VX, NbX and M 2 X are observed after the quality heat treatments at the center portion
of the both forgings. There is no large difference in the martensitic lath widths, distributions and
sizes of those particles between the both trial forgings.
KEYWORDS: FB2 material; rotor forgings; ingot making process; creep strength; microstructure
INTRODUCTION
To increase thermal efficiency and reduce CO2 emissions, advanced 9-12%Cr steels containing
Co and B have been actively developed in Japan and Europe [1-4]. Recently, FB2 material,
X13CrMoCoVNbN9-2-1, which was developed by COST program in Europe has been utilized
for USC plants and its demands have been increased [4]. Quality and reliability of the high Cr
steel rotor forgings are required because those rotor forgings turn at 3000 or 3600 revolutions per
minute. Therefore homogeneous large ingots which are obtained with special process such as
electro slag remelting (ESR) and electro slag hottopping (ESHT) have been needed to
manufacture the high quality turbine rotors [5,6]. Figure 1 shows a schematic illustration of ingot
making process, conventional, ESHT and ESR. In the case of high Cr steel ingot by conventional
ingot making process, macroscopic segregation arises at center and top portion of the ingot [6].
For this reason, it is difficult to manufacture the homogeneous large ingot of high Cr steel. ESHT
process controls the segregation by melting dilution electrode heated with graphite electrode and
slag. On the other hand, ESR process controls the segregation by continuously melting and
solidifying electrodes manufactured with conventional ingot making process in the water cooling
mold. Japan Casting & Forging Corporation has manufactured a lot of 9-12% Cr steel rotor
forgings since 1991, and ESHT trial rotor forging of FB2 material made from 80 metric ton ingot
in 2007 [7,8]. However, the workability and productivity such as yield ratio in ESHT process are
321

not so good for large size ingots. Thus, 145 ton ESR facility has been installed in 2012 to improve
the workability and productivity. This paper describes the results of FB2 trial rotor forging made
from ESR ingot to confirm the qualities and compare the creep rupture strength and the
microstructure of that forging to those of a forging manufactured with ESHT process.
COMPOUND
HOTTOP MOLD
REFRACTORY

Conventional process

UNSOLIDIFIDE ZONE
SOLID
MOLD

CONSUMABLE ELECTRODE
GRAPHITE ELECTRODE
SLAG
HOTTOP MOLD
REFRACTORY

ESHT process

UNSOLIDIFIDE ZONE
SOLID
MOLD

ELECTRODE
SLAG
WATER COOLING MOLD
ESR process
UNSOLIDIFIED ZONE
SOLID

Figure 1: Ingot making process of Conventional, ESHT and ESR process [5, 6].

322

MATERIALS EXAMINED AND EXPERIMENTAL PROCEDURES


Manufacturing Sequence of Trial Rotor Forging Made from ESR Ingot
Figure 2 shows manufacturing sequence of the trial rotor forging made from ESR ingot.
Electrodes were manufactured by conventional ingot making process. The steel was melted and
refined in electro arc furnace and ladle furnace. The molten steel was casted into the molds by
bottom poring. The as-cast electrodes were remelted in ESR facility, and 66 metric ton ESR ingot
with the diameter of 1900mm was produced. Figure 3 shows the forging process of the trial
forging. Upsetting and stretching were performed by 130MN press. Maximum forging diameter of
the forging was 1370mm. In the preliminary heat treatment, pearlitic transformation was carried
out to refine the grain size. Figure 4 shows the heating cycle of the quality heat treatment after the
first rough machining. The trial rotor forging was austenitized at 1100 degrees C and quenched in
oil. And double temperings were subsequently performed. Ultrasonic test was examined after
machining, and it was confirmed that there was not harmful defect internally. The manufacturing
sequence described above is similar to that of trial forging made from ESHT ingot. The detail
manufacturing process is refered in the paper [8].
Steel & ingot making

Electro Slag Remelting (ESR)

Forging
Preliminary heat treatment

Isothermal transformation

Pre-machining
Ultrasonic test
Quality heat treatment

Oil quenching and Double tempering

Machining
Ultrasonic test
Mechanical tests

Tensile, Impact, Creep Test

Figure 2: Manufacturing sequence of the trial rotor forging.

323

Dia. 2350mm (92.5)

Dia. 1900mm (74.8)

Dia. 1350mm (53.1)

ESR ingot 66ton

Dia. 1370mm (53.9)

Dia. 1370mm (53.9)

715mm (28.1)
3040mm (119.7)
ESR Trial forging 21ton
Figure 3: forging process of ESR trial rotor forging.
1100C
(2012F)
Oil
Quenching

700C
(1292F)

570C
(1058F)

Figure 4: Schematic the heating cycle for quality heat treatment of ESR trial rotor forging.
Mechanical Tests and Microstructural Investigations
Materials for metallurgical investigations and mechanical tests were cut from each location of the
trial forging after ultrasonic inspection. Chemical analyses, tensile, Charpy impact, creep and
creep rupture tests were carried out. microstructural observations were performed by optical
microscope (OM) and transmission electron microscope (TEM). Observations and Analyses of
precipitations were examined by scanning transmission electron microscope (STEM) and energy
dispersion X-ray (EDX) analyzer with carbon extracted replicas. Comparison of creep strength
and microstructure between the trial forging manufactured with ESR process and the forging
made from ESHT ingot was done [5,6].

324

RESULTS AND DISCUSSIONS


Results of Trial Rotor Forging Manufactured with ESR Process
Table 1 shows the results of chemical analysis at each location of the trial rotor forging. Boron
content at each portion is in the range from 90 to 101ppm and similar level as that of electrode.
Homogeneous ingot can be obtained by ESR process in spite of the large diameter of 1900mm
since there is no large difference in the amount of each alloy element besides boron between the
electrode and the trial forging. Table 2 shows the results of tensile and Charpy impact tests at the
surface and center portions. 0.2% yield strength at surface is about 670MPa and that at center is
around 700MPa. The impact energy at 20 degrees C is in the range from 14 to 20J at both surface
and center portions. It is confirmed to manufacture high Cr steel rotor forgings with homogeneous
composition and good properties by ESR process.
Table 1: Results of chemical analysis on ESR trial rotor forging (mass %).
Top Side

709mm
(27.9)

1309mm
(51.5)

1109mm
(43.7)

S2

S1

C1
Locations

809mm
(31.9)

C2

Si

Mn

Ni

Cr

Mo

Nb

Co

Electrode

0.13

0.09

0.34

0.005

0.001

0.12

9.13

1.52

0.20

0.050

1.28

0.0099

0.020

S1

0.13

0.09

0.34

0.005

0.001

0.12

9.14

1.52

0.20

0.050

1.28

0.0100

0.020

S2

0.13

0.08

0.34

0.005

0.001

0.12

9.16

1.53

0.20

0.050

1.29

0.0097

0.020

C1

0.14

0.09

0.34

0.005

0.001

0.12

9.20

1.53

0.20

0.051

1.28

0.0090

0.019

C2

0.13

0.09

0.34

0.006

0.001

0.12

9.15

1.55

0.20

0.052

1.30

0.0101

0.020

Table 2: Results of tensile and Charpy impact tests on ESR trial rotor forging.

0.02%Y.S.
[MPa]
(ksi)

0.2%Y.S.
[MPa]
(ksi)

T.S.
[MPa]
(ksi)

El.
[%]

R.A.
[%]

S1

574
(83)

677
(98)

811
(118)

19

S2

571
(83)

668
(97)

800
(116)

C2

596
(86)

695
(101)

832
(121)

Locations

325

Impact Energy
[J]
(ft-lb)

FATT
[C]
(F)

at 20C
(68F)

at 40C
(104F)

63

20
(15)

35
(26)

67
(153)

19

63

20
(15)

37
(27)

74
(165)

17

60

14
(10)

31
(23)

68
(154)

Comparison of Creep Strength between Forgings Made from ESR and ESHT Ingot
Figure 5 shows creep rupture strength at the center portions of the forgings made from ESR and
ESHT ingot. Creep testing time of the ESR forging is around 5000h and that of ESHT beyond
27000h, and these tests still continue. There is no large difference in creep rupture strength
between the both forgings since creep rupture curve of the ESR forging is in good agreement with
that of the ESHT. Figure 6 displays change in creep rate at the center portions of the both
forgings. Minimum creep rate of the forging manufactured with ESR process under the testing
condition of 650 degrees C and 150MPa is 8.0x10-6/h and that of the forging manufactured with
ESHT process 7.0x10-6/h, furthermore both curve are similar. It can be said from the results of
creep tests up to 5000h that there is no large difference in the creep strength between both
forgings. However, long-term creep strength is important for high temperature materials, thus
long-term creep test has to be continued.

400
Center Portion

40
30

200

20
100 Ongoing:
90
ESR Trial Forging in 2012
80
ESHT Trial Forging in 2007
70
60
22
23
24
25

L.M.P.=T(25+log(t))

Stress (ksi)

Stress (MPa)

300

50

10
26

27

Figure 5: Creep rupture strength of trial rotor forgings made from ESR and ESHT ingot.

326

-1

10

ESR Trial Forging in 2012


ESHT Trial Forging in 2007

-2

Creep rate (/h)

10

Center Portion
650C (1202F)
150MPa (22ksi)

-3

10

-4

10

-5

10

-6

10

10

100
Time (h)

1000

10

Figure 6: Change in creep rate of trial forgings manufactured with ESR and ESHT process.
Comparison of microstructure between Forgings Made from ESR and ESHT Ingot
It was found from the results described above that the difference in creep strength between the
forgings made from ESR ingot and ESHT is not so large, which suggested that the both forgings
assume to be similar in microstructure since the creep strength is strongly affected by the
microstructure. Then, the microstructures of both forgings were investigated. Figure 7 shows the
optical micrographs at the center portion of trial forgings manufactured with ESR ingot (a) and
ESHT ingot (b). Tempered martensitic structures are observed and no delta ferrite is found at the
both forgings, but the difference in structures between them is not found by optical microscope.
TEM images of the both forgings with thin foils are shown in Figure 8. Martensitic lath structures
with high dislocation density are observed on the both forgings. The martensitic lath widths of the
both forgings are around 0.6m. A lot of ellipsoidal particles with the average diameter of about
120nm are found on the lath boundaries and about 30nm fine particles are observed in and on the
lath boundaries. Analyses with the carbon extracted replicas were carried out by STEM and EDX.
Figure 9 shows the STEM image (a) and phase mapping (b) of the replica at the center portion of
the forging made from ESR ingot. Figure 10 shows the STEM image (a) and phase mapping (b)
of the forging made from ESHT ingot. Each particle was identified by the chemical analysis
results of individual phases from the papers [9,10]. In Figures 9 (b) and 10 (b), M 23 C 6 is labeled
by red, MX including a lot of V (VX) blue, MX including a lot of Nb (NbX) green and M 2 X
cyan. Mainly M 23 C 6 and coarsened NbX with the average diameter of about 80nm are observed.
However, precipitation densities of VX, NbX and M 2 X in the both forgings are extremely low in
comparison with 9Cr-1Mo-V-Nb-N steel [10]. Figure 11 shows the result of extracted residue. Fe
and Cr are thought to be mainly included in M 23 C 6 . Amounts of V and Nb in the both forgings are
327

around 0.04% and 0.03% respectively. These amounts are in good agreement with the distribution
density of fine M 2 X, VX and NbX found in Figure 9 (b) and Figure 10 (b). However, the amount
of extracted residue of the both forgings is not so large difference. It can be said from the results
of these microstructural investigation that there is no large difference in martensitic lath widths,
phase distribution and sizes of the particles observed between the both forgings.
(b)

(a)

100m

100m

Figure 7: Optical micrographs at center portion of the forging manufactured from ESR ingot (a)
and ESHT ingot (b).
(a)

(b)

Figure 8: TEM images at center portion of the forging manufactured from ESR ingot (a) and
ESHT ingot (b)

328

(a)

0.5m

(b)

M23C6

Red

VX

Blue

NbX

Green

M2X

Cyan

Figure 9: STEM image (a) and phase mapping (b) with carbon extracted replica at center
portion of forging made from ESR ingot.

329

(a)

0.5m

(b)

M23C6

Red

VX

Blue

NbX

Green

M2X

Cyan

Figure 10: STEM image (a) and phase mapping (b) with carbon extracted replica at center
portion of forging made from ESHT ingot.

330

Fe
Cr
Mo
Co
V
Nb

Residue(%)

ESR

1
ESHT

Figure 11: Results of residue on trial rotor forgings at center portion.


Conclusions
9%Cr steel containing Co and B, X13CrMoCoVNbNB9-2-1, trial rotor forging has been
manufactured by electro slag remelting process to confirm the qualities and compare creep
strength and the microstructure of that forging with those of a forging made from electro slag
hottopping ingot. The results obtained were as follows:
1.

It is confirmed to manufacture rotor forging with homogeneous composition and good


properties by electro slag remelting process.

2.

Creep rupture strength of the forging made from electro slag remelting ingot is almost similar
to that of the forging applied electro slag hottopping process from the results of creep rupture
tests up to 5000h.

3.

Martensitic lath microstructures with high density dislocations and the precipitations of
M 23 C 6 , VX, NbX and M 2 X are observed after the quality heat treatments at the center
portion of the both forgings. There is no large difference in the martensitic lath widths,
distributions and sizes of those particles between the both trial forgings.

Further long-term investigations on the creep strength and microstructural evolution in the both
forgings will be performed.

331

REFERENCES
[1] Y. Kagawa, F. Tamura, O. Ishiyama, O. Matsumoto, T. Honjo, T. Tsuchiyama, Y. Manabe
and H. Kawai, Development and Manufacturing of The Next Generation of Advanced 12%Cr
Steel Rotor for 630C Steam Temperature, 14th International Forgemasters Meeting, (2000), pp.
301-308.
[2] M. Arai, H. Doi, T. Azuma and T. Fujita, Development of High WCoB-Containing 12Cr
Rotor Steels for Use at 650C in USC Power Plants, 15th International Forgemasters Meeting,
(2003), pp. 261-268.
[3] Y. Tsuda, M. Yamada, R. Ishii, Y. Tanaka, T. Azuma and Y. Ikeda, Development of High
Strength 12%Cr Ferritic Steel for Turbine Rotor Operation above 600C, 13th International
Forgemasters Meeting, (1997).
[4] T. -U. Kern, H. Cerjak, K.H. Mayer, M. Staubli, R. Vanstone, G. Zeiler and K. -H.
Schoenfeld, Development and Qualification of Rotor Steels for 600 to 650C Application in
COST501/COST522, 14th International Forgemasters Meeting, (2000), pp. 270-276.
[5] K. Morinaka, Y. Futamura, I. Kitagawa and S. Watanabe, The Manufacture of the Large
ESHT-J Ingot, I&SM (1989), pp. 9-15.
[6] D. L. Newhouse, C. J. Boyle and R. M. Curran, A Modified 12-Percent Chromium Steel for
Large High Temperature Steam Turbine Rotors, Preprint of the ASTM Annual Meeting, (1965).
[7] M. Mikami, K. Morinaka, Y. Okamura, S. Tanimoto, Y. Wakeshima, R. Magoshi, Y. Kadoya
and A. Matsuo, Manufacturing of 12%Cr Rotor Forgings for High Temperature Steam Turbines
from Electro Slag Hot Topped Ingot, 14th International Forgemasters Meeting Proceedings,
(2000), pp. 295-300.
[8] K. Kawano, Y. Wakeshima, T. Miyata, Y. Okamura, F. Gatazka and S. L. Breitenbach,
Manufacturing of COST-FB2 Trial Rotor Forgings, 17th International Forgemasters Meeting,
(2008), pp. 303-308.
[9] M. Mikami, Y. Wakeshima and T. Miyata, Creep Rupture Strength and Microstructural
Investigation of 12 % Cr Steel Large Forgings for Ultra-Supercritical Steam Turbine Rotors,
Advances in Materials Technology for Fossil Power Plants, Proceedings of 6th International
Conference, Santa Fe, New Mexico, USA (2010), pp 408-422.
[10] K. Sawada, K. Suzuki, H. Kushima, M. Tabuchi and K. Kimura, Effect of Tempering
Temperature on Z-phase Formation and Creep Strength in 9Cr-1Mo-V-Nb-N Steel, Materials
Science and Engineering A 480 (2008), pp. 558-563.

332

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

DEVELOPMENT AND PRODUCTION OF MONOBLOCK LOWPRESSURE TURBINE ROTOR SHAFT MADE FROM 670 TON INGOT

Takafumi Yamauchi, Hidenao Kudo, Yasuhiro Kishi


Sou Ueda, Hajime Yoshida, Kimitoshi Kimura,
Koji Kajikawa, Shigeru Suzuki
The Japan Steel Works Ltd., 4 Chatsu-Machi, Muroran, 051-8505, Japan

ABSTRACT
Monoblock low-pressure (LP) turbine rotor shaft forgings for nuclear power plants have been
produced from up to 600 ton ingots. However, ingots greater than 600 tons are necessary to
increase the generator capacity. Segregation, non-metallic inclusions, and micro porosities
inevitably increase with the increase in ingot size. Manufacturing such massive ingots with high
soundness is a quite difficult. Thus, the development of 650 ton ingot production was carried out
in 2010. The 650 ton ingot was dissected and investigated to verify its internal quality. The
internal quality of the 650 ton ingot was found to be equal to that of 600 ton ingots. Subsequently,
in 2011, we produced a 670 ton ingot, the worlds largest, to produce a trial LP rotor shaft
forging with a diameter of 3,200 mm. Results show that the internal quality, mechanical
properties, and heat stability are the same as LP rotor shaft forgings made from 600 ton ingots.
INTRODUCTION
To increase the unit capacity of power plants, the power generation efficiency is being improved
with the maximum unit capacity reaching 1,700 MW. Rotor shafts for power generation, nuclear
reactor components, and petroleum refining reactors used at an advanced power plants have
become larger and heavier. Since 1985, The Japan Steel Works, Ltd. has been manufacturing
monoblock low-pressure turbine rotor shafts forgings using 600 ton ingots[1]. To date, more than
280 turbine rotor shaft forgings have been produced using 600 ton ingots. However, with the
increasing unit capacity of power plants, the next-generation designs of nuclear turbine rotor
shafts calls for ingots larger than 600 tons. Segregation, non-metallic inclusions, and
microporosities inevitably increase with the increase in ingot size. Manufacturing such massive
ingots with high soundness is quite difficult. Thus, we developed 650 ton ingot production in
2010. The 650 ton ingot was dissected and investigated to verify its internal quality. Factors
related to internal quality, such as segregation, non-metallic inclusions, and microporosities of
the 650 ton ingot were found to be equal to those of 600 ton ingots[2][3]. We subsequently
333

produced a 670 ton ingot, the worlds largest, to produce a trial LP rotor shaft (hereinafter
referred to as trial LP rotor), which has a body diameter of 3,200 mm. Such a large LP rotor shaft
has never been produced as a mono-block forging. Enlarging the drum diameter will decrease the
toughness and ultrasonic permeability at the center of the rotor. Therefore, it is an enormous
challenge to obtain good mechanical properties and ultrasonic permeability in such a massive
forging. In this study, we investigated the internal quality, mechanical properties, and heat
stability of the trial LP rotor.
PRODUCTION OF TRIAL LP ROTOR
Production plan of trial LP rotor
A body diameter of 3,200 mm for the trial LP rotor was chosen based on design information
from heavy electric machinery companies to evaluate the ultrasonic permeability and mechanical
properties at the center of the body section. A target tensile strength of 870 MPa was chosen as
the highest strength demanded. Furthermore, we planned to adjust the tensile strength of the trial
LP rotor from 870 MPa to 780 MPa by second tempering to evaluate the mechanical properties
at low-level strength.
Figure 1 shows the manufacturing process and quality control of the trial LP rotor. Ultrasonic
and mechanical tests were conducted to investigate internal quality and mechanical properties.
Furthermore, a heat stability test was conducted after gashing to evaluate the deflection-related
base metal.

Press

Basic Electric Arc Furnace


Ladle Analysis

V.C.D.

Ladle Refining
Periphery ultrasonic Test
Dimensional Test
Visual Test

Vac.
Vac.

Horizontal Furnace

Ar

Melting & Ingot Making

Forging

(Before Quality Heat Treatment)

Preliminary Heat Treatment

Cleaning

Rough Machining

Inspection

Vertical Electric Furnace

A.C.

W.Q.

Periphery ultrasonic Test


Mechanical Test

Tempering

Tensile Test
Impact Test
Product Analysis etc.

Tensile Test
Impact Test
Product Analysis
etc.

F.C.

Quality Heat Treatment

Rough Machining

Inspection

Center Core

Trepanning

10

11

Mechanical Test And


Sampling For Macro Structure

Horizontal Electric Furnace

Bore Sonic Test


Bore Visual Test
Bore Magnetic Particle Test

12

Tempering &
Mechanical Test

13

Axial Bore
Finishing

No7

Inspection

14

15

Gashing

16

No1 3

Macrostructure
observation
Mechanical Test

No3

Heat Stability Test

17

Dissection &
Investigation

Figure 1. Manufacturing process and quality control plan of the trial LP rotor

334

12

Production of 670 ton ingot


Segregation, non-metallic inclusions, and micro porosities inevitably increase with increasing
ingot size. Although a large number of studies have been conducted on ingot making, prediction
accuracy is not enough. Therefore, manufacturing such massive ingots with high soundness is
quite difficult. The 670 ton ingot was produced using an optimized manufacturing method based
on knowledge gained from investigation of a 650 ton ingot[2][3]. The 670 ton ingot is shown in
Figure 2. To obtain more than 670 tons of molten steel using an electric arc furnace (EAF), raw
materials were melted five times with the EAF. At the ladle refining furnace (LRF), degassing
was performed by agitation with argon and by vacuum treatment to reduce gas contents such as
oxygen, hydrogen, and nitrogen. At the same time, the chemical composition was adjusted
separately in each ladle. Although using five ladles has been carried out in the past, this was the
first multiple-ladle casting using 670 tons of steel. The chemical composition and molten steel
temperature of each ladle were carefully controlled. Subsequently, the refined molten steel was
poured into an ingot case set up in a vacuum tank under vacuum conditions.
The chemical composition of the 670 ton ingot is shown in Table 1. The material was
3.5%NiCrMoV steel, which is widely used in low-pressure steam turbine rotor shafts.

Table 1. Chemical composition of


the 670 ton ingot (mass %)
C
Si
Mn
P
S
Ni
Cr
Mo
V
0.24 0.05 0.33 0.0025 0.0014 3.68 1.74 0.43 0.07

Figure 2. Appearance of the 670 ton ingot


Forging
Solidification contraction of the ingot inevitably generates microporosities in the ingot. In the
forging process, the solidification structure of the ingot must be eliminated using hightemperature diffusion.
The microporosities formed in the ingot
must also be consolidated to maintain
the integrity of the inner properties.
Therefore, the forging process was
optimized using the finite element
method. Figure 3 shows the appearance
of the trial LP rotor with a diameter of
Figure 3. Appearance of the trial LP rotor with
3,270 mm at its body finishing stage.
3,270 mm at its body finishing stage.
335

The trial LP rotor was forged using a 14,000 ton hydraulic press.
Heat treatment
Heat treatment comprises two processes. One is preliminary heat treatment for normalizing and
tempering to attain inspectability. Obtaining a fine grain structure from repeat austenitization is
important so that there is good penetration of ultrasonic waves. The other is quality heat
treatment, consisting of quenching and tempering, to obtain the optimum mechanical properties.
For the trial LP rotor, preliminary heat treatment was carried out in a horizontal furnace. Similar
to the LP rotors made from 600 ton ingots, normalization was implemented three times for the
trial LP rotor. After finishing the preliminary heat treatment, the trial LP rotor was machined to a
diameter of 3,200 mm and an ultrasonic test was performed to confirm the internal quality. The
quality heat treatment was then conducted. In the quality heat treatment process, heating was
carried out in the vertical furnace and spray quenching was performed in the vertical position to
obtain a homogeneous microstructure. However, for the trial LP rotor, which has a body
diameter of 3,200 mm, the cooling rate decreases, which in turn decreases the toughness at the
center of the rotor. Consequently, a fourth normalization stage was added in the quality heat
treatment process.
RESULTS
Ultrasonic testing
After the quality heat treatment, oxide scale was removed from the trial LP rotor by machining,
and the internal quality was inspected via an ultrasonic test in accordance with ASTM A418.
Figure 4 shows the minimum detectable flaw size (MDFS) measured at a frequency of 2.25 MHz.
The MDFS at the center of the body section was 1.0 - 1.2 mm, which is equal to or greater than
that of the LP rotors made from the 600 ton ingots.
4941

3260

Bottom of ingot

894

3182

895

1528

1176

3234

(Unitmm) Top of ingot

Applicable spec.ASTM A418Frequency2.25MHzMethodLongitudinal wave method


Shaft
Body
Shaft
Position
Item

Surface MDFS(,mm)
0.3
0.2
0.4
0.4
0.4
0.2
0.3
Center MDFS(,mm)
1.0
0.7
1.0
1.0
1.2
0.7
1.0
MDFSMinimum detectable flaw size
Sensitivity multiplication factor such that a 10% indication at the forging center will be
equivalent to 1.6mm diameter flat bottom hole.

Figure 4. Result of ultrasonic test after quality heat treatment

336

This MDFS value is a good value for the ultra-large forging with a diameter of 3,200 mm.
Furthermore, no indication exceeding MDFS, such as non-metallic inclusions, was detected in
the trial LP rotor.
Mechanical properties
A tensile test and Charpy impact test were performed in accordance with JIS Z 2241 and 2242.
Figure 5 shows the distribution of tensile properties and impact properties of the trial LP rotor at
a tensile strength of 870 MPa class. It was confirmed that a aim of properties, i.e., the tensile
strength, yield strength, and elongation and reduction in area at the surface and center, could be
fully achieved and that these are equal to the date for an LP rotors made from 600 ton ingots.
Regarding the Charpy impact test, the absorbed energy is rather low and the fracture appearance
transition temperature (FATT) is rather high at the center of the body section compared with the
date for the LP rotors made from 600 ton ingots. One of the reasons for the deteriorating
toughness is the change in cooling rate during quenching with the increase in the body diameter.
Furthermore, in the body center portion, the absorbed energy and the FATT gradually
deteriorated toward the top of the ingot. The main reason is that the element increased toward the
top of the ingot as a segregation. Figure 6 shows the distribution of the FATT from the body
surface to the center of the trial LP rotor. The FATT increased toward the center of the rotor,
with a maximum value of 22 , because the cooling rate decreased during quenching. This
result is similar to the data for the LP rotors made from 600 ton ingots at a tensile strength of 870
MPa class.
Trial LP rotor
3200mm

LP rotor made from 600 ton ingot


2720mm

Body surfaceradial

Body centeraxial

3200

Body centerradial

Bottom of ingot
0.2%YS , TS Carbon content
2
at center(wt,%)
(N/mm )

B ottom of i ng ot

Top of i ng ot

0.25
0.20
950
900
850

TS

40

0.2%YS

20

800
750

Sampling position of test specimens

Body diameter
2650mm
2720mm
3200mm

0.30

Top of ingot


Remarks
- LP rotor made from
600
ton
ingot
Trial LP rotor

700

740
700

FATT ()

Absorbed energy
(J)

EL.,RA. (%)

660
100
80
60
40
20
0

RA.

FATT ()

0.02%YS
2
(N/mm )

780

EL.

-20
-40
-60
After quality heat treatment
TS870MPa class

300

-80

200
100

-100

0
40

-120

0
-40

-80
-120

Figure 5. Distribution of tensile properties


and impact properties

200
400
600
800 1000 1200 1400
Depth of test specimen from body surface (mm)

1600

Figure 6. Distribution of FATT from


body surface to center
337

To evaluate the mechanical properties at a tensile strength of 780 MPa class, the trial LP rotor
underwent tempering and mechanical testing. In addition, test specimens were taken from bottom
of a gash to the center. This area has never been investigated because it is equivalent to the
delivery contour. The distribution of the FATT is shown in Figure 7. The FATT was -20
which is equal to that at the center of the body section of an LP rotors made from a 600 ton ingot.

Bottom of ingot

Top of ingot

Sampling position of test specimens

Body diameter

2850mm

3200mm

Remarks
LP rotor made from

600 ton ingot


Trial LP rotor

40
After re-tempering
(TS870MPa780MPa

20

FATT ()

0
-20
-40
-60
-80
-100
-120
0

200
400
600
800 1000 1200 1400
Depth of test specimen from body surface (mm)

1600

Figure 7. Distribution of FATT from body surface to center after re-tempering


The fracture toughness (KIC and JIC), low-cycle fatigue, and fatigue crack growth testing were
performed using specimens taken from the top, middle, and bottom of the ingot, which is the
equivalent to the center of the body section in the trial LP rotor. Figure 8 shows the relationship
between the temperature and fracture toughness. The fracture toughness is similar for the top,
middle, and bottom sections. Moreover, the test
result is nearly equal to the LP rotors made from
350
Trial LP rotor
600 ton ingots. Compared with the past record,
Center of body
300
Middle Top Bottom
this result falls on and near the lower bound of
K
250
K
the scatter band of a conventional LP rotors made
K
200
from the 600 ton ingots. There is a possibility of
150
influence for decreasing the cooling rate during
LP rotor made from
100
600 ton ingot
quenching because of the large diameter of the
K
50
K
body. Figure 9 shows the results of low-cycle
K
0
-250
-200
-150
-100
-50
0
50
100
fatigue testing, and the relationship between the
Temperature
fatigue crack growth rate and stress intensity
Figure 8. Fracture toughness properties
factor is shown in Figure 10. It was confirmed
K IC (MPam1/2)

IC

JIC

JC

IC

JIC

JC

338

that these test results fall on and in the scatter band of a conventional LP rotors made from 600
ton ingots
10-6
Trial LP rotor
Middle of body
Top of body
Bottom of body

Trial LP rotor
Middle of body
Top of body
Bottom of body

10-7

da /dN (m/cycle)

Total strain range, t (%)

10

10-8

10-9

LP rotor made from


600 ton ingot
0.1
102

103

104

105

10-10

Number of cycles to failure, N f

LP rotor made from


600 ton ingot
1

10

100

K MPam1/2

Figure 9. Low cycle fatigue properties

Figure 10. Fatigue crack growth rate


properties

Investigation using center core bar


Macrostructure observation, chemical analysis, cleanliness, and non-metallic inclusions were
investigated using a core bar taken from the center of the trial LP rotor.

135mm

The macrostructure observation was performed for a cross section based on the JIS G0553
method using the core bar taken from the center of the trial LP rotor. The macrostructure of the
cross section of the center core bar is shown in Figure 11. Compared with the body section, the
shaft section had a dense dendritic structure. The main reason is that the forging ratio is defferent
between the body and shafts. However, it was confirmed that a sound microstructure was
obtained at all positions.

G1
Bottom of ingot

G2
G1G3Sampling position(135mm10mm)

G3
Top of ingot

Figure 11. Macrostructure of the cross section of center core

339

Chemical analysis was performed using the center core bar. To compare the trial LP rotor and
650 ton ingot dissected in 2010, the positions of the test pieces sampled from trial LP rotor was
restored to their original ingot state. Figure 12 shows the distribution of the carbon content at the
center of the ingot as an example of the chemical analysis. Carbon has a relatively severe
segregation profile in the ingot. The carbon content gradually increased toward the top of the
ingot. This behavior is inevitable in the progress of the solidification process for the ingot.
Although the ingot is large, the chemical profile in the trial LP rotor was almost the same as the
650 ton ingot and LP rotors made from the 600 ton ingots.

Carbon content at the center (mass%)

0.50
650 ton ingot
Trial LP rotor
LP rotor made from 600 ton ingot

0.45
0.40
0.35
0.30
0.25
0.20
0.15
0

1000

2000

3000

4000

5000

6000

Distance from ingot bottom (mm)

Figure 12. Chemical analysis result of carbon distribution

Furthermore, the area fraction of non-metallic inclusion in the trial LP rotor made from the 670
ton ingot ranged from 0.0024 % to 0.0064 %, and the inclusions consist of oxides and sulfides.
This result is almost the same as the that of 650 ton ingot, which ranged from 0.0002 % to
0.0082 %.
Heat stability test
After the mechanical test, the trial LP rotor was gashed to investigate the heat stability. The
contour of the trial LP rotor was decided based on design information acquired from heavy
electric machinery companies. The heat stability test was performed in accordance with ASTM
A472 as the international standard, and the test temperature was 400 , the highest demanded.
Moreover, a new heat furnace was produced because the trial LP rotor has a large body with a
diameter of 3,200 mm. Figure 13 shows the results of the heat stability test. The deflection was
evaluated as the deference of the movement of center between in 400 and room temperature
were within the 2.5 / 100 mm limit stipulated by ASTM A472. The deflection of the trial LP
rotor was equal to the LP rotors made from the 600 ton ingots.

340

Measurement points

Top of ingot

Bottom of ingot

Measurement points
Movement of shaft center
(1/100mm)
Specified range
ASTM A472 (1/100mm)

0.35

0.80

0.50

2.5

40
Trial LP rotor
LP rotor made from 600 ton ingot
Test temparature 400
Number of rotor 22

35

Number of data

30
25
20
15

ASTM A472
Specified range2.5 / 100mm

10
5

4.54.9

4.04.4

3.53.9

3.03.4

2.52.9

2.02.4

1.51.9

1.01.4

0.50.9

00.4

Movement of shaft center (1/100mm)

Figure 13. Heat stability of trial LP rotor

Macrostructure observation
The macrostructure observation was performed for the entire longitudinal section of the trial LP
rotor based on the JIS G0553 method using flat polished specimens for macrostructure
observation. The macrostructure is shown in Figure 14. Inverse V segregation (A segregation)
streaks are observed at the top side of the ingot in the journal near the body. Observed V
segregation streaks in the top side of the ingot in the journal are shown in Figure 15. Although
inverse V segregation has a tendency to become severe with increasing ingot size, not much
inverse V segregation was found in the trial LP rotor. And the inverse V segregation was slightly.

500mm

Bottom side of ingot

Figure 14. Macrostructure of the longitudinal section of trial LP rotor

341

Top side of ingot

Bottom side of ingot

Observed position

Top side of ingot

Surface side

Center side
Bottom side of ingot

Top side of ingot

Figure 15. Observed Inverse V segregation


SUMMARY
To establish the manufacturing technology for rotor shafts made from a large ingot, The Japan
Steel Works Ltd. Developed the production of a trial LP rotor with a diameter of 3,200 mm made
from a 670 ton ingot, the largest ingot in the world. The trial LP rotor was investigated and
dissected to evaluate the internal quality. The results are summarized as follows.
(1) The MDFS at the center of the body section was 1.01.2mm, which is equal to or
greater than the LP rotors made from a 600 ton ingots, and no indication exceeding MDFS
was detected in the trial LP rotor.
(2) The mechanical properties after quality heat treatment and re-temper heat treatment were
equal to those of the LP rotors made from 600 ton ingots.
(3) At the center of the ingot, the carbon content gradually increased toward the top of the ingot.
However, the chemical profile in the trial LP rotor was almost the same as those of the 650
ton ingot and LP rotors made from 600 ton ingots.
(4) Results from heat stability test show that the heat stability was satisfied ASTM A472 and
was equal to that of the LP rotors made from a 600 ton ingots.
(5) Inverse V segregation streaks were observed at the journal of the top side of the ingot, but
they were slightly.

We have established the technology for manufacturing rotor shafts made from a 670 ton ingot.
Consequently, we can contribute to increase the unit capacity at the power plants.
342

REFERENCES
[1]

I. Sato, E. Murai, O. Tsumura, I. Kurihara, Y. Saitoh, T. Karaushi, Y. Tanaka Histry and


future prospects of technologies and products of the Muroran Plant focusing on energy
and preservation of environment , JSW Technical Review, Vol. 58, (2007), p.15

[2]

K. Kajikawa, S. Suzuki, F. Takahashi, S. Yamamoto, T. Suzuki, S, Ueda, T. Shibata, H.


Yoshida, Development of 650-ton-class ingot production technology Investigation on
the internal quality of the 650-ton trial ingot JSW Technical Review, Vol. 63, (2012),
p.48

[3]

K. Kajikawa, S. Suzuki, F. Takahashi, S. Yamamoto, T. Suzuki, S, Ueda, T. Shibata,


Yoshida, Development of the World Largest 650ton Ingot for Rotor Shaft Application
The Thermal and Nuclear Power Generation Convention: Collected Works, (2011), p.93

343

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

HIGH CYCLE FATIGUE PROPERTIES OF STEAM TURBINE


MATERIALS AT HIGH TEMPERATURE UNDER SUPERHEATED
STEAM CONDITIONS
Jan Dzugan
COMTES FHT Inc., Dobrany, Czech Republic
Tom Mek
Doosan koda Power, Pilsen, Czech Republic
Josef Jurenka
Faculty of Mechanical Engineering, Czech Technical University in Prague,
Czech Republic
ABSTRACT
Increasing demand for reliable design of all kinds of structures requires materials properties
evaluated under the conditions as close to real service conditions as possible. Presently resolved
project dealing with development of new turbine blades geometry requires better understanding of
the material behavior under service conditions. Service conditions of turbine blades are cyclic
loading at high temperatures under superheated steam conditions and complex mechanical
loading.
There are not commercially available testing systems providing such functionality and thus the
system allowing samples testing under considered conditions was developed. The system allows
cyclic loading at temperatures up to 650C under superheated steam conditions. Typical blade
steel is investigated here and experimental approach considering complex mechanical loading as
well as thermal and corrosion is shown here. The results of high cycle fatigue tests in superheated
steam corrosive environment are shown here.
INTRODUCTION
Lightweight and at the same time safe design requires more accurate input data thus development
of a new testing procedures is necessary. In order to increase the steam turbines efficiency,
higher operation temperatures are required leading to new materials exploitation as the current
ones are not able to fulfill new requirements. The new materials shall be tested and failure free
safe service design demands data obtained at appropriate testing conditions as close to service
ones as possible. In the case of steam turbine blade materials is current demand on materials
operating up to 650C. When realistic conditions are to be considered not only high temperature
has to be applied, but also, superheated steam has to be taken into account as well as complex
mechanical loading.
A superheated steam is used in turbines at temperatures reaching and soon possibly exceeding
650C. The superheated steam is present in the most parts of the turbine, except the last few blade
rows. Saturated steam can be found at the last blades rows, where a temperature drop of the steam
due to its movement through the turbine creates conditions for its formation. The highest
temperatures are present in the turbine first stage, which is the most critical part from materials
temperature resistance point of view. The higher steam temperature, the higher turbine efficiency,
therefore temperatures increase is an issue nowadays together with new materials development
and proposed system has to assure such conditions.
344

The current work is dealing with presentation of the experimental approach for blade material
testing taking into account multiaxial loading conditions and development of testing equipment
for high cycle fatigue tests at temperatures up to 700C under superheated steam condition.
EXPERIMENTAL APPROACH
In order to assess appropriately fatigue life of the components considered, analyze of the in
service loading conditions was performed. A turbine blade is loaded by a complex loading
consisting of mechanical, thermal and corrosive components as main loading factors. All of these
factors are considered in the current investigations.

Figure 1: Testing fixture for multiaxial loading of a blade model


The analyze of the turbine blade mechanical loading during the service consists of the centrifugal
forces in the axial direction of the blade, torsion moment induced by locking blades together and
finally there are service vibrations. These conditions are considered in the current investigation.
Based on pre-mentioned analyze, testing fixture simulating these multiaxial loading conditions
was designed for a fatigue testing, Fig. 1. The scheme explaining the loading is shown in the Fig.
2. The original aim of the fixture was to allow testing of turbine blade models under almost in
service mechanical loading condition at high temperature under superheated steam conditions. In
the course of the experiments it was found to be experimentally very difficult to obtain suitable
temperature field over the blade within the designed fixture, due to its complexity. It was decided
to run experiments with multiaxial loading fixture at room temperature only and fit the fatigue
behavior simulation model under these conditions and subsequently use the same model for high
temperature and superheated steam with cylindrical specimens. The model at room temperature is
based on fatigue tests under multiaxial loading of the blade model and also with the use of round
samples of the same geometry that is to be tested at high temperatures. The notched cylindrical
samples had to be used with similar fatigue index to blade as it is described below.

345

Figure 2: Scheme of multiaxial blade loading (The blade is red).


The simulation model is based on fatigue experiments performed under multiaxial loading as well
as using uniaxial fatigue tests. As it was above mentioned, it is very difficult to run tests with
multiaxial loading at high temperatures, but testing of standard smooth samples also doesnt
provide appropriate information. Notched round samples with similar fatigue behavior were then
considered for the experimental testing.

Figure 3: FEM ananlyze results with the blade critical spot detection.
FEM simulation of the blade loading condition was performed and the critical location was
detected on the blade, Fig. 3. For this critical spot sensitivity analyzes using Fatigue Index (FI)
was performed on the basis of fatigue criterions according to Dang Van (DV) and MansonMcKnight (MMK). The simulation was performed for tilt angle of the blade 45 and appropriate
axial load and torque. The analyze shown, that round sample with a notch radius 1,5mm should
provide similar fatigue behavior. There was performed verification of the notched sample
comparability of the fatigue behavior with the blade. There were applied three loading cyclic
forces for the FEM simulation within the expected range of service loads: 1, 2 and 3 kN. Results
of this analyze are shown in Fig. 4 together with results obtained for the blade loaded by the same
forces. There can be clearly seen very good agreement between the results for the blade and
proposed notched sample for all considered load levels and both applied approaches. Based on
these results notch with radius 1,5mm was further used for the uniaxially loaded round samples
that can be tested under high temperature and superheated steam.

346

Figure 4: Fatigue index calculation for represnetative notch proposal


TESTING SYSTEM FOR SUPERHEATED STEAM CONDITIONS
The first step of the testing system development was analyze of demands placed on the
experimental system. The system must allow high cycle testing under high temperature and
superheated steam conditions. Complex required functionality of the system can be basically
divided into three basic parts:

Mechanical loading unit


Heating unit
Superheated steam production unit

The mechanical loading unit should provide cyclic loading to the testing sample at desired load
level. In the current case, magnetoresonant testing system with the load capacity +/- 150kN
allowing testing frequencies up to 250Hz was employed.
The heating unit goal in the testing system is to provide a stable testing temperature of the testing
chamber with sample. A three zone split furnace was utilized here. The operation range of the
used furnace was up to 800C. Internal space of the furnace had to be sufficient for the steam
chamber accommodation
The last and the most difficult part was the system for superheated steam production. There was
not found any commercially available system applicable for considered case. Therefore, it was to
be developed so, that it fulfills all demands on a steam parameters and it fits with the other system
parts and provides desired functionality

347

Figure 5: Scheme of the testing system


The proposed system, shown in Fig. 5., consists of following main parts: demineralized water
reservoir, steam generator, steam superheater, testing chamber, steam condensate cooler, steam
condensate tank, split furnace and magnetoresonant testing machine. All piping and fittings are
made of stainless steel prohibiting steam contamination from corrosion products on its way
towards sample.
A key part of the steam chain is testing steam chamber. The chamber must enable sample
attachment to the testing system and also movement of the sample during high cycle fatigue
testing. At the same time it must be sufficiently sealed in order to keep superheated steam within
the chamber around the testing sample, thus appropriate sealing is crucial here. A sample
temperature must be precisely monitored and controlled in order to provide tests at desired
temperature levels. This is assured by thermocouple attached directly to the sample acting as a
control signal feedback. Set up of the testing chamber together with testing sample and furnace
can be seen in Fig. 6.
Designed system was assembled and on the basis of preliminary performance tests there were
added some additional control and safety features. The used demineralized water reservoir was
fitted with level gauges assuring pumping of additional demineralized water into the reservoir
when the low level is reached. In the case of empty tank, testing is automatically interrupted
preventing testing without steam. At the same time the furnace and steam superheater are also
switched off preventing an idle long term operations at high temperature. The end of tests defined
either by certain number of cycles or sample rupture acted also as a triggering signal for switching
of the system in case of long term tests without operators presence. The most difficult part of
testing is insertion of the sample with welded thermocouple into testing fixtures and chamber and
subsequently into the testing system. Any failure of the thermocouple lead to test invalid test,
therefore a special attention has to be paid to this point. Successful tests series were run with the
system describer here.

348

Figure 6: Detail of testing chamber with sample and furnace


HIGH CYCLE FATIGUE TESTING UNDER SUPERHEATED STEAM CONDITIONS
High cycle fatigue tests were carried out on smooth and notched round samples at room
temperature and at 560C under superheated steam environment. Smooth samples active part
diameter was 8mm. The notched samples had the same active part geometry as the smooth ones,
but in the middle there was introduced notch with the radius of 1,5mm resulting in minimal
diameter of 5mm. Tests were performed with stress ratio R= -1. Samples high temperature tests
were conditioned for at least 2 hours at testing temperature prior to tests. Results at tests at room
temperature and at 560 for smooth and notched samples can be found in Fig. 7. Difference in
fatigue strength between room temperature and 560C evaluated for 107 cycles is about 100MPa.
Difference between notched and smooth samples is slightly bigger in the case of testing at 560C
in comparison to room temperature results.

349

Figure 7: Example of the high cycle fatigue results obtained for smooth and notched samples at
room temperature as well as at 560C.
CONCLUSIONS
The presented work summarizes results of fatigue tests of steam turbine blades material under
superheated steam environment at high temperatures. There were explained blade service
mechanical loading conditions together with developed experimental set up enabling laboratory
simulation of these conditions on the blade model at first. The proposed experimental fixture for
multiaxial loading did not allow tests at high temperatures. Secondly, a simplification of high
temperature tests had to be done. On the basis of the FEM simulation of the blade critical point
and the fatigue index analyze was executed for this point and notched round fatigue samples with
similar fatigue conditions were proposed. Finally, fatigue testing set up for high temperature
testing under superheated steam environment up to 700C was developed and successfully applied
to tests of the smooth and notched samples. Results obtained at high temperatures were compared
with results obtained at room temperature for samples of the same geometries.
The work presented here is initial part of the project dealing with development and modernization
of turbine blades and there are expected further activities leading to procedure proposal for
fatigue life optimization from fatigue point of view for steam turbine blades.
ACKNOWLEDGMENTS
The paper was realized due to the support of project TA01020985 - Development and
modernization of turbine blades from the perspective of the reliability and service life increase
sponsored by Technology Agency of the Czech Republic.
REFERENCES
[1] Jan Dzugan et al., 2012, Advanced Materials Research, 538-541, 1630

350

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

CREEP AND CREEP-FATIGUE CRACK GROWTH BEHAVIORS OF


30Cr1Mo1V ROTOR STEEL AFTER LONG TERM SERVICE
Kexian Shi, Fusheng Lin, Shuangqun Zhao, Yanfeng Wang
Shanghai Power Equipment Research Institute, Shanghai 200240, China

ABSTRACT
This paper presents the creep and creep-fatigue crack growth behaviors of 30Cr1Mo1V turbine
rotor steel which had been in service for 16 years. Two typical sections of the rotor, i.e. high and
low temperature sections, are examined at 538, with crack initiation and propagation monitored
by D.C. potential drop method in a compact tension (CT) specimen. The material of the high
temperature section has the lower resistance to creep and creep-fatigue crack growths than the low
temperature section. The creep crack initiation (CCI) time decreases with the increase of initial
stress intensity factor. The creep-fatigue crack growth (CFCG) is dominated by the cycledependent fatigue process when the hold time at the maximum load is shorter, but it becomes
dominated by the time-dependent creep process when the hold time becomes longer. The high
temperature section shows a larger influence of time-dependent creep behavior on CFCG than the
low temperature section.
1. INTRODUCTION
The turbine rotors in power plants are often subjected to variable loading conditions during
service such as creep, fatigue and creep-fatigue, which leads to the initiation of cracks from the
pre-existing defects in manufacturing or the damages in deformation. The creep arises from the
high pressure on the rotor at high temperature, and the fatigue is due to the alternating thermal
stresses developed during turbine start and stop. Such multiple loading conditions can induce the
appearance of a macro-crack evolved from some inherent micro-defects or deformation damages
such as inclusions and voids. The crack initiation, from the viewpoint of fracture mechanics, does
not mean the failure of component since the unstable crack propagation happens only when the
initiated crack grows to a certain size. There is generally a long period of time for the crack to
initiate and grow to the critical size under creep-fatigue loading condition. To ensure the safety
and reliability of operation, it is necessary to study the creep-fatigue lifetime of turbine rotor in
service conditions.
Cr-Mo-V steel is the typical material for turbine rotors in supercritical power plants. Many
experimental and theoretical studies have been made to predict the service lifetime, focusing on
the effects of stress level, hold time and temperature on the creep-fatigue crack initiation and
growth [1-2]. However, most of them are based on the as-received material, and fewer are
reported for the performance of Cr-Mo-V steel which has been in service for a long time.
Studying the changes of the creep-fatigue crack growth behaviors of rotor steel after long-term
service is helpful to not only the better understanding of the effect of service conditions on the
351

material property, but also the reliable analysis of material degradation and life expenditure that
serve basis for the management and assessment of rotor lifetime.
In this paper, an experimental investigation on the crack growth behavior of 30Cr1Mo1V (1Cr1Mo-0.25V) rotor steel that has been operated for 16 years is carried out in both creep and creepfatigue conditions. Tests are performed in the compact tension (CT) specimens at 538 for two
typical sections of the rotor, i.e. high and low temperature sections. The effects of initial stress
level and hold time on the crack initiation and propagation are studied via D.C. potential drop
method.
2. EXPERIMENTAL PROCEDURES
2.1 Materials and specimens
The material used is 30Cr1Mo1V turbine HP-IP rotor steel, which had been in service for 16
years in subcritical plant. The chemical composition of the steel is shown in Tab.1. Two typical
sections of the rotor, i.e. high and low temperature sections, are investigated as shown in Fig.1.
The mechanical properties of materials at these two sections are shown in Tab.2. The property of
high temperature section has degraded after long-term service, but the low temperature section
still satisfies the performance requirement of ASTM A470-03[3]. The service temperature and
stress calculated by finite element analysis according to the rotors service condition are shown in
Tab.3. The high temperature section worked at a temperature more than 500 and the other at
the room temperature, but the stresses in both of the sections are very small.
The specimen used is a standardized CT specimen for creep crack growth tests based on ASTM
E1457-00[4]. The height, width and thickness are 24, 25, and 10mm, respectively. All specimens
were notched by Electrical Discharge Machining (EDM).
Tab.1 Chemical composition of 30Cr1Mo1V steel (wt.%)

Si

Mn

Ni

Cr

Cu

Mo

Al

0.29

0.22

0.75

0.007

0.002

0.43

1.12

0.06

1.20

0.27

0.005

Tab.2 Mechanical properties of examined materials


Sampling location

RP0.2/MPa

Rm/MPa

A/%

Z/%

High temperature section

608

761

19.0

60.3

Low temperature section

640

810

19.1

54.2

Original material

630

790

22.4

65.2

ASTM A470-03

620

725-860

17

43

352

Tab.3 Service temperature and stress of specimen


Section

Low temperature section

high temperature section

Temperature()

25

525

Von Mises Stress (MPa)

67.4

22.6

Fig.1 Sketch map of steam turbine rotor


2.2 Test method and conditions
The creep crack growth (CCG) tests are conducted based on ASTM E1457-00. The test
temperature is kept at 538 with the precision of 1. Dead weight loading is applied with
initial stress intensity factor KI in the range of 25 to 45 MPam1/2. The load waveform of the
creep-fatigue tests is shown in Fig.2. A stress ratio of 0.1 is kept constant, but the hold time (th) at
the maximum load varies from 1 to 120 minutes in different tests. The crack length is measured
by DCPD method.

Fig.2 Load waveform adopted in creep-fatigue tests

353

3. EXPERIMENTAL RESULTS AND DISCUSSION


3.1Creep crack growth behavior
Similar to the creep curve, the creep crack growth (CCG) curve of both high and low temperature
sections can be divided into 3 stages, as shown in Fig.3. The crack growth first experiences a
relative long stage with the slow increasing rate, and then reaches essentially a steady stage in
which the rate changes little with time, and finally the crack rapidly grows with time until fracture
occurs. The first stage can be considered as the incubation of creep crack. It generally defines the
time for the crack to grow to a size of 0.2 mm as the crack initiation time [5]. As shown in Fig.4,
the creep crack initiation (CCI) time decreases with the increase of initial stress intensity factor,
yet following an approximately linear logarithmic relation. In addition, the CCI time of high
temperature section is smaller than that of the low temperature section, which is an indication of
material degradation of resistance to crack growth at the high temperature section.
Fig.5 shows the CCG rate da/dt as a function of stress intensity factor K. The curve of each
condition is a typical CCG rate curve [6]. The CCG rate increases with the increase of K at the
first stage, then comes to a flat stage in which the rate change little with K, and finally it grows
rapidly with the stress intensity factor. Again, material at the high temperature section has a
degraded resistance to creep crack growth, because the CCG rate of the high temperature section
was much higher than that of the low temperature section under the same loading.

Fig.3 Creep crack growth length versus time for different initial stress intensity factor KI of (a)
the high temperature section and (b) the low temperature section.

354

Fig.4 Creep crack initiation time t0 versus initial stress intensity factor KI

Fig.5 Creep crack growth rate da/dt versus stress intensity factor K
3.2Creep-fatigue crack growth behavior
Different from the CCG curves, the creep-fatigue crack growth (CFCG) curves have marginal
incubation stages, especially when the hold time becomes short, as shown in Fig.6. The
introduction of fatigue condition effectively reduces the creep-fatigue crack initiation (CFCI)
time, which is found to decrease with the reduction of hold time at the maximum load. When the
hold time is reduced to 1 min, no visible incubation period can be observed from the CFCG
355

curves for both high and low temperature sections. Similar to the situation in pure creep
condition, the high temperature section has the lower resistance to CFCG than the low
temperature section.
The CFCG rate can be defined as crack growth length per time (da/dt) or crack growth length per
cycle (da/dN). They are often used together to differentiate the effects of creep and fatigue on the
crack growth, as to be discussed below. The relations of da/dt and da/dN with the stress intensity
factor range K are shown respectively in Figs.7a-b for the high temperature section, and
respectively in Figs. 8a-b for the low temperature section. The CFCG rate da/dt is generally larger
than the CCG rate, but the difference in between decrease with the increase of hold time (Fig. 7a
and 8a). . As for the rate da/dN, it shows weak dependence on the hold time when the hold time
itself is shorter (e.g. 1-10 min in Figs. 7b and 8b). Only when the hold time becomes longer
(e.g. >30 min) da/dN increases with the increase of hold time..
To examine the influences of creep and fatigue on the crack growth behavior, a relationship
between the CFCG rate at a given stress intensity factor of K=36MPam1/2 and the frequency is
constructed, as shown in Fig.9. The frequencies are transferred from the hold times. The CFCG
rate da/dt depends greatly on the frequency when the frequency itself is higher, so the crack
growth behavior in this case is dominated by the cycle-dependent fatigue process. With the
decrease of frequency, i.e. the increase of hold time, the dependence becomes weaker due to the
contribution of the time-dependent creep process. Similar effects were reported by O Yokota et al.
[7] for the creep-fatigue crack growth of 12Cr steel.
The influence of time-dependent creep behavior on CFCG at the high temperature section is
larger than that at the low temperature section. This is supported by the facts that the slope of
da/dt-f curve of the high temperature section is smaller than the low temperature section (Fig. 9a),
and the scope of hold time in which the CFCG rate da/dN changes little is narrower in the former
than in the latter (Fig.9b). The reason could be that the creep property of the high temperature
section has degraded after long term service compared to the low temperature section.

Fig.6 Creep-fatigue crack growth rates of (a) da/dt and (b) da/dN versus stress intensity factor
range K of the high temperature section.

356

Fig.7 Creep-fatigue crack growth rates of (a) da/dt and (b) da/dN versus stress intensity factor
range K of the high temperature section.

Fig.8 Creep-fatigue crack growth rates of (a) da/dt and (b) da/dN versus stress intensity factor
range K of the low temperature section

357

Fig.9 Creep-fatigue crack growth rate of (a) da/dt and (b) da/dN with range of stress intensity
factor K

4. CONCLUSIONS
An investigation has been carried out on the creep and creep-fatigue crack growth behaviors of
30Cr1Mo1V steel which had been operated for 16 years. The following conclusions can be
drawn from the current investigation:
(1)The material resistance to creep and creep-fatigue crack growth of the high temperature section
has decreased, comparing to the low temperature section.
(2) Introduction of fatigue condition promotes the crack initiation and growth. The creep-fatigue
crack growth behavior is dominated by the cycle-dependent fatigue process when the hold time is
shorter, but it turns to be dominated by the time-dependent creep process when the hold time is
longer.
(3) The time-dependent creep behavior affects the CFCG at the high temperature section more
greatly than at the low temperature section.

REFERENCES
[1] Henrik Andersson, Rolf Sandstrom. Creep Crack growth in service-exposed weld metal of
2.25Cr1Mo [J].International Journal of Pressure Vessels and Piping, 2001, 78: 749-755.
[2] Takeo Yokobori, Toshimitsu Yokobori. High Temperature Creep, Fatigue and Creep-fatigue
Interaction in Engineering Materials [J].International Journal of Pressure Vessels and Piping,
2001, 78: 903-908.
[3] ASTM A 470-03, American Society for Testing and Materials, Standard Specification for
Vacuum-Treated Carbon and Alloy Steel Forgings for Turbine Rotors and Shafts [S]New
York: ASTM International, 2003.
358

[4] ASTM E 1457-00, American Society for Testing and Materials, Standard Test Method for
Measurement of Creep Crack Rates in Metals [S]New York: ASTM International, 2000
[5] Ellisonet E.G. International Conference on Creep Fatigue in Elevated Temperature
Application [C], Philadelphia 1973 and Sheffield UK 1974: Paper 173/73
[6] Duo Wang. Fracture Mechanics [M]. NanningGuangxi People Press, 1982
[7] Yokota O, Sugiura R, Yoda M ,et al. Crack Growth Characteristics and Damage in 12Cr Steel
Under High Temperature Creep and Creep-fatigue Conditions [J]. Strength Fracture and
Complexity, 2006, 4: 41-46.

359

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

THE EFFECT OF WATER VAPOR CONTENT AND


CO2 ON TBC LIFETIME
B. A. Pint, K. A. Unocic and J. A. Haynes
Oak Ridge National Laboratory, Oak Ridge, TN USA

ABSTRACT
While the water vapor content of the combustion gas in natural gas-fired land based turbines is
~10%, it can be 20-85% with coal-derived (syngas or H2) fuels or innovative turbine concepts for
more efficient carbon capture. Additional concepts envisage working fluids with high CO2
contents to facilitate carbon capture and sequestration. To investigate the effects of changes in the
gas composition on thermal barrier coating (TBC) lifetime, furnace cycling tests (1h cycles) were
performed in air with 10, 50 and 90 vol.% water vapor and in CO2-10%H2O and compared to prior
results in dry air or O2. Two types of TBCs were investigated: (1) diffusion bond coatings (Pt
diffusion or simple or Pt-modified aluminide) with commercially vapor-deposited yttria-stabilized
zirconia (YSZ) top coatings on second-generation superalloy N5 and N515 substrates and (2) high
velocity oxygen fuel (HVOF) sprayed MCrAlYHfSi bond coatings with air-plasma sprayed YSZ
top coatings on superalloy X4 or 1483 substrates. In both cases, a 20-50% decrease in coating
lifetime was observed with the addition of water vapor for all but the Pt diffusion coatings which
were unaffected by the environment. However, the higher water vapor contents in air did not
further decrease the coating lifetime. Initial results for similar diffusion bond coatings in CO210%H2O do not show a significant decrease in lifetime due to the addition of CO2.
Characterization of the failed coating microstructures showed only minor effects of water vapor
and CO2 additions that do not appear to account for the observed changes in lifetime. The current
50-100C de-rating of syngas-fired turbines is unlikely to be related to the presence of higher
water vapor in the exhaust.
INTRODUCTION
While the exhaust in natural gas-fired land based turbines contains ~10 vol.% water vapor, the
exhaust gas with coal-derived fuels or innovative turbine concepts for more efficient carbon
capture may contain 20-85% H2O [1,2]. Turbines designed for coal-derived synthesis gas (i.e.
syngas) or hydrogen fuels are expected to have higher water vapor contents than natural gas or jet
fuel due to the combustion reaction or the use of steam diluents to assist with combustion [3,4].
Syngas-fired turbines, such as those being installed in two gasification plants in the US (one began
operation in June 2013, the other is under construction), are de-rated (i.e. fired at a lower maximum
temperature). This is a significant concern as the de-rating will reduce the turbine efficiency
resulting in lower plant electricity output for the same amount of fuel. Improved
materials/coatings may enable a reduction in the de-rating; however, it is critical to understand the
syngas/H2 turbine environment. Sulfur or ash contamination from imperfect syngas cleanup (or
Notice: This manuscript has been authored by UT-Battelle, LLC, under Contract
360No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States
Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up,
irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes.

upset conditions) could certainly degrade materials performance as could the higher water vapor
content [3]. While sulfur or ash contamination could be mitigated, especially by carbon capture
processes applied to the syngas prior to combustion, the higher water vapor content is determined
by the fuel and combustion conditions. The effects of water vapor on high temperature oxidation
have been widely studied in the past 15 years [5-17], including alumina-forming alloys and
thermal barrier coating (TBC) bond coatings. There is a general need to understand the role of
water vapor and, in this case, to understand the effect of higher water vapor levels on coating
performance.
The first phase of this work studied Pt-containing bond coatings, including single phase and two
phase (+) Pt diffusion coatings [9,15,18-22] in furnace cyclic testing at 1150C because of the
extensive prior experience and in-depth characterization database available for these coatings. The
second phase examined the 1100C performance of MCrAlY bond coatings , which are typically
used in land-based gas turbines [23,24]. Included in this study is the current change in superalloy
compositions in land-based gas turbines to lower cost, low Re 2nd generation compositions [25]
or returning to 1st generation superalloys without Re, such as alloy 1483 [26].
EXPERIMENTAL PROCEDURE
For fabricating the Pt-containing coatings, superalloy Ren N5 or N515 (Table 1) coupons 1.5mm
thick, ~16mm diameter with chamfered edges were first grit blasted with alumina and then
electroplated with 71m of Pt in a laboratory scale bath at Tennessee Technological University.
For the Pt diffusion coatings, the plated coupons were then annealed in a vacuum of 10-4 Pa
(10-6 Torr) for 2h at 1175C resulting in a ~25m thick coating [19]. For Pt-modified aluminide
coatings, the plated specimens were aluminized for 6h at 1100C to form a ~40m thick layer in
a laboratory-scale chemical vapor deposition (CVD) reactor, described elsewhere [27]. A ~125m
thick ceramic YSZ top coating was deposited on one side of the coupons by electron-beam
physical vapor deposition (EB-PVD) using a standard commercial process that included a light grit
blasting prior to YSZ deposition. For the thermally sprayed coatings, 2mm thick CMSX-4 and
1483 coupons, ~16mm diameter were high velocity oxygen fuel (HVOF) coated using a
commercial-type process and commercial NiCoCrAlYHfSi [28,29] (Table 1) powders at
Stonybrook Univ. The HVOF-coated specimens were annealed in a vacuum of 10-4 Pa (10-6 Torr)
for 4h at 1080C and had a bond coating thickness of ~100m. Coupons were then returned to
Stonybrook University for air-plasma spraying (APS) a ~200m-thick YSZ layer on one side using
standard conditions.
For both types of coatings, four specimens were evaluated at each condition: three were coated
with YSZ and the fourth was tested without YSZ (and without grit blasting for both diffusion
coatings) to study the surface morphology evolution. The specimens were hung by a Pt-Rh wire
in an automated cyclic rig and were exposed for 1h at 1100C (MCrAlY) or 1150C (PtTable 1. Chemical compositions (atomic% or ppma) determined by inductively coupled plasma
analysis and combustion analysis.
Material

Ni

Cr

Al

Re Co

N5
N515
X4
1483
MCrAlYHf

64.7
66.5
62.8
60.8
41.2

8.0
6.6
7.5
13.6
16.2

13.3 0.9 7.6


13.2 0.5 7.3
13.0 0.9 9.7
7.3 < 8.8
22.9
18.4

Ta

Mo Ti

Other (ppma)

1.6
2.0
2.1
1.3
0.01

2.2
2.1
2.2
1.7
<

0.9
1.2
0.4
1.2
<

0.29
0.02
0.36
0.05

26C,132Zr,70Y,540Hf,17S
66Zr,104Y,1993Hf,13Ce,<1S
<3Y,270Hf,17S
<3Y,7Hf,80Si,<1S
3860Y,710Hf,6500Si,3S

361

0.01
<
1.2
4.9
<

containing) in (1) dry air or O2, (2) air with 101%, 502% or 902% vol.%H2O or (3) buffered
90%(CO2-0.15%O2)+10%H2O and cooled in laboratory air for 10min for each cycle. Oxidation
in wet air was conducted by flowing air at 850 ml/min with distilled water atomized into the gas
stream above its condensation temperature. The injected water was measured to calibrate its
concentration. Mass change was measured using a Mettler-Toledo model XP205 balance. Every
200 cycles, the specimens without YSZ were characterized by scanning electron microscopy
(SEM) equipped with an energy dispersive x-ray spectrometer (EDS) to examine the surface scale
adhesion and roughness [15-17]. After completing testing, the specimens were metallographically
sectioned and examined by light microscopy and electron probe microanalysis (EPMA) using a
JEOL model 8200.
RESULTS & DISCUSSION
Pt diffusion bond coatings
Figure 1 shows the average lifetime data for the Pt diffusion bond coatings in various environments
[15,16,22,30,31]. Massive YSZ spallation defined the coating lifetime and the average and a
standard deviation of the three specimens is shown. The lifetimes in dry O2 are higher than other
studies [32,33] but numerous experimental differences make comparisons difficult. For the Ptdiffusion bond coatings, the addition of water vapor appeared to have little effect on YSZ lifetime.
With 10%H2O, a similar lifetime was noted for the N5 and N515 substrates and the average
lifetime was highest in CO2-10%H2O. In contrast, water vapor had a significant effect on the
lifetime of Pt-modified CVD aluminide bond coatings on N5. The reduction in average lifetime
was inversely proportional to the water content, 53% with 10%H2O, 37% with 50%H2O and 16%
with 90%H2O (the 90%H2O life is within the standard deviation of the dry O2 lifetime). A similar
drop in lifetime was observed for coatings on X4 substrates [34]. For the same bond coating
on the N515 substrate in 10%H2O, the lifetime was more than double that observed on N5 and
slightly higher than for N5 in dry O2, Figure 1. No comparable dry O2 data has been generated

Figure 1. Average lifetimes (number of 1 h cycles to failure) for EB-PVD YSZ-coated superalloy
specimens with two different Pt-containing diffusion bond coatings exposed at 1150C in several
environments. The bars note one standard deviation for 3 specimens of each type. [15,16]
362

for coated N515. Replacing air with buffered CO2 in the experiment further increased the lifetime.
For the environments containing H2O, the coating lifetime appeared to increase with decreasing
pO2. However, the oxidation of alumina-forming alloys is thought to be independent of pO2 [35]
even in steam [36] and changes in environment had only minor effects on the thermally grown
alumina scale thickness [16]. Previously, the coating results were explained by an observed
effect of water vapor on the rumpling of the bond coating [15,30], however, more recent
experiments could not confirm that hypothesis [34]. For the N515 substrates, the increased
coating lifetime is attributed to the higher Hf content in this alloy (Table 1) and the well-known
benefit of Hf on improving the oxidation resistance of NiAl and aluminide coatings on Hf-doped
substrates [37-39], including improving the creep strength of the alumina scale [40]. The Pt
diffusion coating lifetime did not increase with the N515 substrate because of the difference in
coating phases. For -, the higher Hf content increased the amount of internal oxidation, which
has been shown previously in model alloys and coated superalloys [41,42]. The difference in
reaction products between the two types of Pt diffusion coatings on N515 is shown in Figure 2.
EPMA x-ray maps from a similar - coated specimen show that the internal oxidation contains
numerous bright particles (Z-contrast in the backscattered image) that are rich in Hf, Figure 3.
Both coatings appear relatively flat with failure at the alumina-coated metal substrate interface.
Limited rumpling of the coating, which is widely observed on similar coatings without a ceramic
top coating [43,44], was likely suppressed by the adjacent ceramic top coating [30,45]. There was
no evidence of carbide formation in the CO2-rich environment.
NiCoCrAlYHfSi bond coatings
While diffusion coatings are widely used in aero-engines, these turbines are not likely to operate
with coal-derived fuels. Larger land-based power generation turbines typically use sprayed
MCrAlY-type bond coatings and sprayed YSZ top coatings. Figure 4 shows current results for
HVOF (widely used in land-based turbines [23]) NiCoCrAlYHfSi-coated X4 and 1483 substrates
but with a rougher bond coating (Ra~8) than prior X4 results with a lower (Ra~5) roughness
[17,30]. In this case, the exposure temperature was reduced to 1100C because the higher
coefficient of thermal expansion (CTE) of these coatings [46] induces more rapid spallation of the
low CTE alumina scale [45]. As for the bond coatings in Figure 1, the addition of 10% water
vapor decreased the lifetime for coatings on both X4 and 1483 substrates. Comparing the results
for these two substrates, in both dry and wet environments, the lifetimes were reduced for the 1483
substrates compared to X4. This reduction is attributed to the lower Al content in the 1483
substrates, Table 1. Exposure at 1100C greatly accelerates the Al loss from the coating (as will
be discussed later), likely leading to early failures. The roughness was increased in the 2nd set of

YSZ

YSZ

alumina scale

epoxy

epoxy

-(Ni,Pt)Al bond coating

N515

50 m

internal oxidation

- coating
N515

Figure 2. Light microscopy of polished cross-sections of failed N515 substrates with EB-PVD
YSZ top coated specimens at 1150C in buffered CO2-0.15%O2+10%H2O (a) -NiPtAl bond
coating after 1320 1-h cycles and (b) - bond coating after 2680 cycles.
363

10 m

Al

Hf

Figure 3. (a) Backscattered electron image of a polished cross-section of - bond coated N515
after 2080 1-h cycles at 1150C in buffered CO2-0.15%O2+10%H2O with corresponding x-ray
maps (b) O, (c) Al and (d) Hf.

specimens to be more similar to commercial HVOF coatings. With the same X4 substrate, the
higher roughness did not increase coating lifetime as expected. With the observed scatter in the
data, the average lifetimes were very similar. However, the bond coating was thinner in the second
batch, ~100m, compared to 125m in the first batch which may have affected the results. Finally,
one set of coated 1483 specimens were exposed in air with 50%H2O. The lifetime was slightly
lower than for 10%H2O and all three specimens failed after 240 cycles.
Figure 5 shows the affect of coating roughness on the failure. Rougher interfaces lead to mixed
failure near the YSZ-bond coating interface compared to the smoother diffusion bond coatings,
Figure 2. When the top coat spalls (usually in one piece), interfacial cracks proceed through the

Figure 4. Average coating lifetimes (number of 1 h cycles to failure) for APS YSZ coated X4 and
1483 16mm diameter superalloy coupons with HVOF NiCoCrAlYHfSi type bond coatings with
two different average roughness (Ra) values exposed at 1100C in dry and wet environments.
Bars note one standard deviation for three specimens of each type. [17,46]
364

epoxy

epoxy

YSZ
Al-rich denuded zone

YSZ
alumina scale

20 m

alumina scale

NiCoCrAlYHf bond coating b

Figure 5. Light microscopy of polished cross-sections of failed X4-coated specimens exposed in


10% water vapor (b) Ra=5 after 440 cycles and (c) Ra=8 after 380 cycles.

alumina scale and the YSZ, leaving some metal exposed but also some YSZ on the substrate
surface. Thus, higher roughness is thought to inhibit this failure. However, the large asperities in
the higher roughness bond coating appeared to be undercut by oxidation, arrow in Figure 5b. In
some cases, the asperities sheared off leaving small pieces of metal in the spalled YSZ coating.
This behavior may be due to the high temperature and short 1h cycles used to induce rapid failures
in laboratory testing. (Previous work with 100-h cycles at 1100C resulted in an average YSZ
lifetime of 3733h.) Figure 5a shows the region of the coating near the surface was depleted in the
Al-rich phase in the coating. The phase depletion occurs both due to formation of the alumina
scale (and reformation when spallation occurs) on the gas side and interdiffusion with the substrate
on the substrate side. An example line profile from a coated 1483 specimen exposed at 1100C in
air with 50%H2O is shown in Figure 6. After only 240 1-h cycles, the Al content had dropped to
~6%, about 50% of the starting powder composition, and there is no evidence of Al-rich phase
remaining in the ~100m thick coating after this exposure. In addition, the Ti content in the

Figure 6. Electron microprobe line profiles across failed a 1483 HVOF (Ra=8)/APS-coated
specimen at 1100C after 240 1-h cycles in air with 50% H2O. The gas interface is on the left
and the substrate is on the right in both cases.
365

YSZ

YSZ

epoxy

epoxy

YSZ

alumina scale

50m

1483

X4
b
alumina
Figure 7. Light microscopy of polished cross-sections of failed HVOF (Ra=8)/APS-coated
specimens exposed in 10% water vapor at 1100C (a) X4 substrate after 340 1-h cycles
(b) 1483 substrate after 280 cycles.

coating is 1.4% throughout the coating and both Mo and Ta have diffused from the substrate to
similar levels in the coating. At this level, Ti is thought to be detrimental to alumina scale adhesion
[22] but this has not been conclusively established. However, Ti has been minimized or excluded
from most 2nd generation superalloys like N5 and N515, Table 1.
Figure 7 compares the failed coatings on X4 and 1483 substrates after cycling in air with 10%H2O.
Both failures appear similar. Chemical analysis has not been completed on these specimens but
the coating on 1483 appears to be fully depleted in the Al-rich phase after 280 cycles, consistent
with the profile in Figure 6, and some phase is apparent in the coating on X4 after 340 cycles.
Slower interdiffusion on X4, due to its higher Al content, likely accounts for this difference. Also,
the higher Ti content in 1483 may cause an additional effect on interdiffusion, explaining the
decreased coating lifetimes at 1100C for 1483. In Figure 7a, the larger dark particles at the
coating substrate interface are alumina, presumably embedded grit used to clean the surface prior
to coating. The particles appear slightly larger for the less oxidation-resistant 1483 substrates,
Figure 7b. This difference needs to be further investigated. Comparing Figure 7 to Figure 2b,
there is an obvious difference in internal oxidation. The lower Hf content in the NiCoCrAlYHfSi
coating reduced the amount of internal oxidation expected. In addition, some or all of the Hf (and
Y) may be oxidized during deposition. In both Figures 5 and 7, the small dark precipitates in the
coating are oxides present after deposition. When tied up as oxides, the Hf is less mobile and
unlikely to form large internal oxides as shown in Figure 3. Furthermore, recent results on model
NiCrAlY, YHf and YHfTi alloys suggested that co-doping (i.e. adding both Y and Hf to the
coating) reduces the depth of internal oxidation [47].
While water vapor is well-known to have a detrimental effect on oxidation behavior, including
TBC life, these results suggest that higher levels of water vapor do not appear to further affect
lifetime under these conditions. For alumina-forming alloys and coatings, water vapor has been
shown to increase internal oxidation and the amount of scale spallation observed [7]. While some
studies on FeCrAl alloys suggest that water vapor could affect the growth rate of the scale, the
observations are that it reduces the growth rate [48,49]. However, measurements of the scale
beneath YSZ have seen little effect or a slight increase in the scale thickness [16,17]. Evidence of
366

evaporation of Al(OH)3 at 1100C also has been observed [9] but this should not be a factor
beneath a YSZ top coating. Scale adhesion appears to be the most likely mechanism for a water
vapor effect, perhaps accelerating cracking at the metal-oxide interface [10]. While ideal model
alloys, such as Hf-doped NiAl or NiCrAlYHf, do not spall more with the addition of water vapor,
these coatings include a variety of substrate and coating impurities and may be more susceptible.
A few recent studies have considered the effect of CO2 on the oxidation behavior of aluminaforming alloys [50,51]. As with the current results, these studies suggest that the effect of CO2 is
minimal. Additional experiments are being conducted with HVOF bond coatings, but using 100h
cycles to better simulate a base load turbine duty cycle [17].
Actual metal temperatures in turbines are not likely to exceed 900C in order to achieve 25kh
coating lifetimes. Water vapor effects may need to be examined over a broader range of
temperatures. For syngas fired turbines, the higher water vapor contents are not likely to affect
coating performance. Thus, other environmental factors such as S and ash should be evaluated in
order to better understand coating performance in these systems.
CONCLUSIONS
Laboratory furnace cycling at 1100 and 1150C was used to compare the YSZ coating lifetime in
two different types of TBC systems with and without the presence of H2O and CO2. At 1100C,
NiCoCrAlYHfSi bond coatings with APS YSZ top coatings on X4 and 1483 substrates showed a
~15-30% reduction in average TBC lifetime in air with 10%H2O compared to dry O2 or air. When
the H2O content was increased to 50%, only a slight further decrease in lifetime was observed.
Compared to 2nd generation superalloy X4, the average coating lifetimes on 1483 were lower and
this was attributed to the lower Al and higher Ti in this alloy and the accelerated interdiffusion
occurring at 1100C, compared to the application temperature of 900C. At 1150C, the average
EB-PVD YSZ lifetime with Pt diffusion bond coatings on N5 substrates were relatively unaffected
by 10%-90% H2O additions and increased in buffered CO2-10%H2O with a N515 substrate. In
contrast, Pt-modified aluminide coatings on N5 showed a 50% lifetime reduction with the addition
of 10% H2O. Higher water vapor contents of 50 and 90% reduced the lifetime only 37 and 16%,
respectively. Longer YSZ coating lifetimes were observed for phase coatings on the higher Hf,
low Re N515 substrates in air with 10%H2O and buffered CO2-10%H2O. This was attributed to
the beneficial role of Hf on alumina scale growth and adhesion. Higher water vapor contents do
not appear detrimental to YSZ coating lifetime and other factors unique to coal derived H2 and
syngas-fired turbines should be investigated.
ACKNOWLEDGMENTS
The authors would like to thank G. W. Garner, T. M. Lowe, K. M. Cooley, H. Longmire, T. Jordan
and D. Leonard for assistance with the experimental work. Plating of Pt was conducted at
Tennessee Tech. University by Prof. Y. Zhang. B. Hazel and B. Nagaraj at General Electric
Aircraft Engines provided the N5 and N515 substrate materials and coated the specimens with EBPVD YSZ and Stonybrook Univ. applied the HVOF and APS coatings. The X4 substrates were
provided by K. Murphy at Alcoa Howmet and the 1483 substrates by A. Kulkarni at Siemens. P.
F. Tortorelli provided helpful comments on the manuscript. This research was sponsored by the
U.S. Department of Energy, Office of Coal and Power R&D, Office of Fossil Energy, (R. Dennis
program manager).

367

REFERENCES
[1] Chiesa, P., Lozza, G. and Mazzocchi, L. Using Hydrogen as Gas Turbine Fuel, Journal of
Engineering for Gas Turbines & Power, Vol. 127 (2005), pp. 73-80 .
[2] Ruth, L. A. Advanced clean coal technology in the USA, Mater. High Temp., Vol. 20 (2003),
pp. 7-14.
[3] Wright, I. G. and Gibbons, T. B. Recent developments in gas turbine materials & technology
& their implications for syngas firing, Int. J. Hydrogen Energy, Vol.32 (2007), pp.3610-3621.
[4] White, B. M., Ames, R. W. and Burke, P. Conditions in Advanced Turbines for IGCC Power
Plants with Carbon Capture, ASME Paper #GT2013-94609, presented at the International
Gas Turbine & Aeroengine Congress & Exhibition, San Antonio, TX, June, 3-7, 2013.
[5] Leyens, C., Fritscher, K., Gehrling, R., Peters, M. and Kaysser, W. A. Oxide Scale Formation
on an MCrAlY Coating in Various H2-H2O Atmospheres, Surf. Coat. Technol., Vol. 82
(1996), pp. 133-144.
[6] Jnsson B. and Svedberg, C. "Limiting Factors for Fe-Cr-Al and NiCr in Controlled Industrial
Atmospheres," Mater. Sci. Forum, Vol. 251-254 (1997), pp. 551-558.
[7] Onal, K., Maris-Sida, M. C., Meier, G. H. and Pettit, F. S. Water Vapor Effects on the Cyclic
Oxidation Resistance of Alumina-Forming Alloys, Mater. High Temp., Vol.20 (2003), pp.
327-337.
[8] Opila, E. J. Volatility of Common Protective Oxides in High-Temperature Water Vapor:
Current Understanding and Unanswered Questions, Mater. Sci. Forum, Vol. 461-464 (2004),
pp. 765-774.
[9] Pint, B. A., Haynes, J. A., Zhang, Y., More, K. L. and Wright, I. G. The Effect of Water Vapor
on the Oxidation Behavior of Ni-Pt-Al Coatings and Alloys, Surf. Coat. Technol., Vol. 201
(2006), pp. 3852-3856.
[10] Smialek, J. L.Enigmatic Moisture Effects on Al2O3 Scale and TBC Adhesion, Mater. Sci.
Forum, Vol. 595-598 (2008), pp. 191-198.
[11] Quadakkers, W. J., urek, J. and Hnsel, M. Effect of water vapor on high temperature
oxidation of FeCr alloys, JOM, Vol. 61, No.7 (2009), pp. 44-50.
[12] Dneux, V., Cadoret, Y., Hervier, S. and Monceau, D. Effect of Water Vapor on the Spallation
of Thermal Barrier Coating Systems During Laboratory Cyclic Oxidation Testing, Oxid.
Met., Vol. 73 (2010), pp. 83-93.
[13] Pint, B. A. and Zhang, Y. Performance of Al-rich Oxidation Resistant Coatings for Fe-Base
Alloys, Mater. Corros., Vol. 62 (2011), pp. 549-560.
[14] Pint, B. A., Brady, M. P., Yamamoto, Y., Santella, M. L., Maziasz, P. J. and Matthews, W. J.
Evaluation of Alumina-Forming Austenitic Foil for Advanced Recuperators, J. Eng. Gas
Turb. Power, Vol. 133 No. 10 (2011), Art. No. 102302.
[15] Pint, B. A., Garner, G. W., Lowe, T. M., Haynes, J. A. and Zhang, Y. Effect of increased water
vapor levels on TBC lifetime with Pt-containing bond coatings, Surf. Coat. Technol., Vol.
206 (2011), pp. 1566-1570.
[16] Unocic, K. A. and Pint, B. A. Effect of Water Vapor on Thermally Grown Alumina Scales on
Bond Coatings, Surf. Coat. Technol., Vol. 215 (2013), pp. 30-38.
[17] Haynes, J. A., Unocic, K. A. and Pint, B. A. Effect of Water Vapor on the 1100C Oxidation
Behavior of Plasma-Sprayed TBCs with HVOF NiCoCrAlX Bond Coats, Surf. Coat.
Technol., Vol. 215 (2013), pp. 39-45.
[18] Bouhanek, K., Adesanya, O. A., Stott, F. H., Skeldon, P., Lees, D. G. and Wood, G. C. "High
Temperature Oxidation of Thermal Barrier Coating Systems on RR3000 Substrates: Pt
Aluminide Bond Coats," Mater. Sci. Forum, Vol. 369-372 (2001), pp. 615-622.
[19] Zhang, Y., Haynes, J. A., Pint, B. A. and Wright, I. G. "A Platinum-Enriched + Two-Phase
Bond Coat on Ni-Base Superalloys," Surf. Coat. Technol., Vol. 200 (2005), pp. 1259-63.
[20] Gleeson, B. Thermal Barrier Coatings for Aeroengine Applications, J. Propul. Power, Vol.
368

2 (2006), pp. 375-383.


[21] Haynes, J. A., Pint, B. A., Zhang, Y. and Wright, I. G. Comparison of the Cyclic Oxidation
Behavior of -NiAl, -NiPtAl and + NiPtAl Coatings on Various Superalloys, Surf. Coat.
Technol., Vol. 202 (2007), pp. 730-734
[22] Pint, B. A., Haynes, J. A. and Zhang, Y. Effect of Superalloy Substrate and Bond Coating on
TBC Lifetime, Surf. Coat. Technol., Vol. 205 (2010), pp. 1236-1240.
[23] Itoh, Y., Saitoh, M. and Tamura, M. Characteristics of MCrAlY Coatings Sprayed by High
Velocity Oxygen-Fuel Spraying System, Journal of Engineering for Gas Turbines and
Power, Vol. 122 (2000), pp. 43-49.
[24] Nicholls, J. R. Advances in Coating Design for High-Performance Gas Turbines, Mater.
Res. Soc. Bull., Vol. 28 (2003), pp. 659-670.
[25] Fink, P. J., Miller, J. L. and Konitzer, D. G. Rhenium reductionalloy design using an
economically strategic element, JOM, Vol. 62, No. 1 (2010), pp. 55-57.
26] Shah, D. M. and Cetel, A. Evaluation of PWA1483 for Large Single Crystal IGT Blade
Applications, in T. M. Pollack, R. D. Kissinger, R. R. Bowman, K. A. Green, M. McLean, S.
L. Olson and J. J. Schirra eds., Superalloys 2000, TMS, Warrendale, PA, 2000, p.295-304.
[27] Lee, W. Y., Zhang, Y., Wright, I. G., Pint, B. A. and Liaw, P. K. Effects of Sulfur Impurity on
the Scale Adhesion Behavior of a Desulfurized Ni-based Superalloy Aluminized by Chemical
Vapor Deposition, Metal. Trans. A, Vol. 29 (1998), pp. 833-841.
[28] DeMasi-Marcin, J. T. and Gupta, D. K. Protective coatings in the gas turbine engine, Surf.
Coat. Technol., Vol. 68-69 (1994), pp. 19.
[29] Unocic, K. A. and Pint, B. A. Characterization of the Alumina Scale Formed on a
Commercial MCrAlYHfSi Coating, Surf. Coat. Technol., Vol. 205 (2010), pp. 1178-1182.
[30] Pint, B. A. High-Temperature Corrosion in Fossil Fuel Power Generation: Present and
Future, JOM, Vol. 65 (2013), pp. 1024-1032.
[31] Pint, B. A. and Haynes, J. A. "The Effect of Water Vapor Content on TBC Lifetime, Mater.
Sci. Technol., Vol. 29 (2013), pp. 828-834.
[32] Tawancy, H. M., Mohamed, A. I., Abbas, N. M., Jones, R. E. and Rickerby, D. S. Effect of
Superalloy Substrate Composition on the Performance of a Thermal Barrier Coating System,
J. Mater. Sci., Vol. 38 (2003), pp. 3797-3807.
[33] Wu, R. T., Kawagishi, K., Harada, H. and Reed, R. C. The Retention of Thermal Barrier
Coating Systems on Single-Crystal Superalloys: Effects of Substrate Composition, Acta
Mater., Vol. 56 (2008), pp. 3622-3639.
[34] Lance, M. J., Unocic, K. A., Haynes, J. A. and Pint, B. A. Effect of Water Vapor on Thermally
Grown Alumina Scales on Simple and Pt-modified Aluminide Coatings, Surf. Coat. Technol.,
(2013) in press.
[35] Ramanarayanan, T. A., Raghavan, M. and Petkovic-Luton, R. "Al2O3 Scales on ODS Alloys,"
in Proc. JIMIS-3, High Temperature Corrosion Transactions Supplement, v. 24, Japan Inst.
Met., Tokyo, Japan, 1983, pp.199-206.
[36] Pint, B. A., Terrani. K. A., Brady, M. P., Cheng, T. and Keiser, J. R. "High Temperature
Oxidation of Fuel Cladding Candidate Materials in Steam-Hydrogen Environments," J. Nucl.
Mater., Vol. 440 (2013), pp. 420-427.
[37] NGandu Muamba, J. M., Streiff, R. and Boone, D. H. L'Influence du Hafnium du Substrat
sur la Rsistance l'Oxydation des Revtements d'Aluminiures sur les Superalliages Base
Nickel, Mater. Sci. Eng., Vol. 88 (1987), pp. 111-121.
[38] Pint, B. A., Wright, I. G., Lee, W. Y., Zhang, Y., Prner, K. and Alexander, K. B. Substrate
and Bond Coat Compositions: Factors Affecting Alumina Scale Adhesion, Mater. Sci. Eng.
A, Vol. 245, (1998), pp. 201-211.
[39] Tolpygo, V. K., Murphy, K. S. and Clarke, D. R. Effect of Hf, Y, C in the Underlying
Superalloy on the Rumpling of Diffusion Aluminide Coatings, Acta Mater., Vol. 56 (2008),
pp. 489-499.
369

[40] Veal, B. W., Paulikas, A. P., Gleeson, B. and Hou, P. Y. Creep in -Al2O3 Thermally Grown
on -NiAl and NiAlPt Alloys, Surf. Coat. Technol., Vol. 202 (2007), pp. 608-612.
[41] Izumi, T. and Gleeson, B. Oxidation Behavior of Pt plus Hf-modified gamma-Ni plus Ni3Al Alloys, Mater. Sci, Forum, Vol. 522-523 (2005), pp. 221-228.
[42] Haynes, J. A., Pint, B. A., Zhang, Y. and Wright, I. G. The Effect of Pt Content on +
Diffusion Coatings, Surf. Coat. Technol., Vol. 203 (2008), pp. 413-416.
[43] Deb, P., Boone, D. H. and Manley, T. F. "Surface Instability of Platinum Modified Aluminide
Coatings During 1100C Cyclic Testing," J. Vac. Sci. Technol. A, Vol. 5 (1987), pp.3366-3372.
[44] Tolpygo, V. K. and Clarke, D. R. Rumpling of CVD (Ni,Pt)Al diffusion coatings under
intermediate temperature cycling, Surf. Coat. Technol., Vol. 203 (2009), pp. 3278-3285.
[45] Evans, H. E. Oxidation failure of TBC systems: An assessment of mechanisms, Surf. Coat.
Technol., Vol. 206 (2011), pp. 1512-1521.
[46] Haynes, J. A., Pint, B. A., Porter, W. D. and Wright, I. G. "Comparison of Thermal Expansion
and Oxidation Behavior of Various High-Temperature Coating Materials and Superalloys,"
Mater. High Temp., Vol. 21 (2004), pp. 87-94.
[47] Unocic, K. A. and Pint, B. A. Oxidation Behavior of Co-Doped NiCrAl Alloys in Dry and
Wet Air, Surf. Coat. Technol., in press.
[48] Young, D. J., Naumenko, D., Niewolak, L., Wessel, E., Singheiser, L. and Quadakkers, W. J.
Oxidation kinetics of Y-doped FeCrAl-alloys in low and high pO2 gases, Mater. Corr., Vol.
61 (2010), pp. 838-844.
[49] Unocic, K. A., Essuman, E. K., Dryepondt, S. and Pint, B. A. Effect of Environment on the
Scale Formed on ODS FeCrAl at 1100C, Mater. High Temp., Vol. 29 (2012), pp. 171-180.
[50] Pint, B. A., Dryepondt, S., Rouaix-Vande Put, A. and Zhang, Y. Mechanistic-Based Lifetime
Predictions for High Temperature Alloys and Coatings, JOM, Vol. 64 (2012), pp. 1454-1460.
[51] Dryepondt, S., Rouaix-Vande Put, A. and Pint, B. A. Effect of H2O and CO2 on the Oxidation
Behavior and Durability at High Temperature of ODS-FeCrAl," Oxid. Met., Vol. 79 (2013),
pp. 627-638.

370

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

FORMATION OF DIFFUSION ZONES IN COATED NI-AL-X TERNARY


ALLOYS AND NI-BASED SUPERALLOYS
A. S. Suzuki, G. D. West and R. C. Thomson
Loughborough University, Loughborough, LE11 3TU, United Kingdom

ABSTRACT
Coatings are an essential part of the materials system to protect the turbine blades from oxidation
and corrosive attack during service. Inter-diffusion of alloying elements between a turbine blade
substrate and their coatings is a potential concern for coated turbine blades at ever increasing
operating temperatures because this can cause the formation of undesirable Secondary Reaction
Zones (SRZs), which may degrade the mechanical properties of coated Ni-based superalloys.
Understanding the effects of each element on the SRZ formation is essential in order to
understand both the mechanism and inter-diffusion behaviour between coatings and substrates. In
this research, a number of simpler aluminized ternary Ni-Al-X (where X is Co, Cr, Re, Ru or Ta)
alloys were investigated in order to elucidate the separate effects of each element on the
microstructural evolution, especially at the coating/substrate interface. The aluminized ternary
alloys developed distinctive diffusion zones, depending on the third alloy element, X.
Specifically, it has been found that both Ni-Al-Re and Ni-Al-Ta alloys developed a continuous
SRZ-like diffusion layer. This diffusion zone persisted in the Ni-Al-Re alloys after high
temperature exposure, indicating that Re has a stronger effect on SRZ formation than Ta.
INTRODUCTION
Higher operating temperatures of power plant improve thermodynamic efficiency, which in turn
leads to a reduction in CO2 emissions and lower fuel consumption. Ni-based superalloys, used as
substrate materials for turbine blades in gas turbines, have excellent high temperature capability.
Coatings also have an important role to play: these are frequently applied to the surfaces of the
turbine blades to provide additional protection against oxidative and corrosive environments
during service.
As the operating temperature increases and the environment becomes more aggressive, the
oxidation and corrosion resistance of turbine blade materials becomes more crucial than before.
Furthermore, inter-diffusion of alloy elements between a coating and a substrate becomes more
important, as this will be very significant at higher temperature, and will accommodate
microstructure changes during service, such as the formation of Secondary Reaction Zones
(SRZs) [1-6] and Topologically Close-packed (TCP) phases [7]. The term SRZ was suggested
by Walston [1], and they were first observed in coated third generation Ni-based superalloys. The
SRZ [1] which can form between a coating and a Ni-based superalloy substrate is generally
considered to be a detrimental layer for coated turbine blades because it can degrade the
mechanical properties of the blades. Sato et al. compared the mechanical properties of various
aluminized Ni-based superalloys and uncoated Ni-based superalloys [8], and concluded that the
SRZ formation in aluminized Ni-based superalloys potentially degrades creep properties. These
SRZs typically have been observed in the case of aluminized or Pt-aluminized coated turbine
blades, however, SRZs can also appear in other coating systems. In general, SRZs consist of , '
371

and TCP phases growing into substrate by the migration of high-angle grain boundaries. The
propensity to form such a layer is a sensitive function of composition and surface finish prior to
coating. Coatings are an essential part of the materials system to protect the turbine blades from
oxidation and corrosion attack during service, and thus it is essential to fully understand the
microstructural evolution of SRZs.
Previous studies have shown both positive and negative effects of different alloying elements in
the case of Pt-aluminized fourth generation Ni-based superalloys, e.g. Ru encourages SRZ
nucleation [5], and Mo is an effective element in impeding SRZ formation [4]. However, these
effects have been observed from various multicomponent Ni-based superalloys, with complex
alloy compositions, and are not always consistent across alloy systems. Some alloying elements,
such as Co shows clear effects in controlling SRZ formation in Pt-aluminized third-generation Nibased superlaloys, whereas this is not the case for Pt-aluminized fourth-generation Ni-based
superalloys [6]. This suggests that the effect of some alloy elements on SRZ formation can be
overwhelmed by the other elements within Ni-based superalloys.
For turbine blade applications, both the alloy and coating compositions can vary significantly, and
therefore how each element affects the formation of SRZs in coated turbine blades is key to
understanding their nucleation mechanism and predicting their growth behavior. In this research,
Ni-Al-X (where X is Co, Cr, Re, Ru or Ta) ternary alloys have been used as substrates in order to
simplify the effect of alloying elements on SRZ formation. This paper presents a possible
mechanism of diffusion zone formation in aluminized ternary alloys, especially in terms of their
microstructural changes during coating preparation and high temperature exposure. It is
anticipated that a greater understanding of the formation of SRZs will facilitate the development
of materials systems with superior mechanical properties and greater microstructural stability. The
results of this study will also act as an indicator of possible problems in terms of microstructural
stability when the alloy compositions are varied for future generations of Ni-based superalloys.
EXPERIMENTAL
The ternary alloy compositions used in this study are presented in Table 1. The Ni-Al-Co, Ni-AlCr, Ni-Al-Re and Ni-Al-Ru alloys were designed to contain both the ' and phases at 1100 C
from ternary diagrams [9-15], and the Ni-Al-Ta alloy was designed to have ' phase at elevated
temperature using the thermodynamic equilibrium software package, MTDATA [16]. The alloys
were prepared via arc melting under an argon atmosphere, followed by an appropriate
homogenization heat treatment. Ternary alloys were cut into coupons with a 3 mm thickness. The
alloys were grit-blasted and then an aluminizing coating process was applied by a Low
Temperature High Activity (LTHA) process, followed by post-production heat treatment (PPHT)
at 1100C for 1 hour under flowing Ar. CMSX-4 was also used as a reference material, and its
composition is also listed in Table 1. Isothermal oxidation tests were subsequently conducted at
880C for up to 500 hours, followed by cooling in air. The resulting microstructures were
analysed using scanning electron microscopy (SEM) with a back-scattered detector (BSE), and
scanning transmission electron microscopy (STEM) and Transmission Electron Microscopy
(TEM) with energy-dispersive X-ray analysis (EDX). Cross-sectional TEM samples were
prepared using a dual-beam focused ion beam scanning electron microscope (FIBSEM).

372

Alloy
Ni-Al-Co
Ni-Al-Cr
Ni-Al-Re
Ni-Al-Ta
Ni-Al-Ru
CMSX-4

Ni
Bal.
Bal.
Bal.
Bal.
Bal.
Bal.

Al
20.3
20.0
18.0
20.7
20.0
12.6

Table 1. Alloy Compositions (at.%)


Cr
Re
Ta
Co
Mo
---4.4
-10.0
-----1.0
-----3.1
-------7.6
1.0
2.2
9.3
0.4

W
-----2.0

Ti
-----1.3

Hf
-----0.03

Ru
----2.1
--

RESULTS
As Aluminized Microstructures
Figure 1 shows cross-sectional microstructures of all of the aluminized ternary alloys. All alloys
formed a coating layer within a thickness range of 35 to 70 m, and consisted of a -NiAl phase.
The Ni-Al-Cr system consisted of two distinctive coating layers, which are indicated as layers 1
and 2 in Fig. 1. Both layers consisted of the phase, however, layer 1 was found to contain less
than 1.5 at.% of Cr, whereas layer 2 contained approximately 6 at.% of Cr. However, the coating
layer in the Ni-Al-Re specimens did not contain Re, suggesting that Re is not incorporated into NiAl. The coating layer in Ni-Al-Ta developed a number of discrete particles. EDX analyses also
indicated that there was a small amount of Ta content in the coating layer, and therefore it is likely
that the discrete particles observed are Ta-rich precipitates. CMSX-4 exhibited typical crosssectional microstructures of as aluminized Ni-based superalloys [17]. The thicknesses of the
coating layer in the specimens were in the order: Cr>Co, Ru> Ta>Re>CMSX-4.
Ni-Al-Co and Ni-Al-Ru alloy did not form any clear diffusion zones after aluminizing at the
interface between coating and substrate, however, the other ternary alloys developed continuous
diffusion layers with distinct morphologies which depended on the third alloy element, X. The
diffusion layer in Ni-Al-Cr in Fig. 1 was approximately 20 m thick. In the diffusion zone in NiAl-Cr, dark phases were observed which were composed of 30.8Al-10.6Cr-58.7Ni (at.%),
whereas bright phases contained 14.9Al-23.1Cr-62.1Ni (at.%). These phase compositions are
comparable with the phase composition of the and phases reported in Ni-Al-Cr ternary
diagram by Dupin [12], and this means that the diffusion zone in Ni-Al-Cr consists of the phase
surrounded by a matrix phase, which was probably formed due to the outward diffusion of Ni
during post-production heat treatment. CMSX-4 formed a typical inter-diffusion zone (IDZ)
microstructure, consisting of -NiAl with topologically close-packed (TCP) phases. Except for
the TCP phase precipitation in the IDZ, the interfacial morphology between -NiAl and substrate
('/ phases) formed in CMSX-4 was similar to that in Ni-Al-Cr, and therefore was classified as an
inter-diffusion zone.
However, both the Ni-Al-Ta and Ni-Al-Re alloys formed a diffusion zone, which was more
similar to a SRZ, as in Fig. 1. Ni-Al-Ta formed the narrowest diffusion zone of all the specimens,
which was approximately 6 m in extent, and consisted of a number of grains with a phase
matrix and Ta-rich precipitates, which grew perpendicular to the substrate with a rod shape and
bright contrast in Fig. 1. Earlier work [18] has shown using EDX mapping that the particles
precipitated in the diffusion zone of Ni-Al-Ta were rich in Ta, but with an absence of Al.

373

Whole microstructure

Diffusion zone

Whole microstructure

Diffusion zone

Figure 1: Cross-sectional microstructures of as aluminized Ni-Al-X ternary alloys and CMSX-4.

Figure 2: Bright field images with the selected area diffraction patterns (SADPs) of Ta-rich
precipitates in the diffusion zone of as aluminized Ni-Al-Ta: (a) Ni2AlTa and (b) Ni2Ta.
Table 2: TEM-EDX analysis results of the
precipitates formed within the diffusion
zone of aluminized Ni-Al-Ta. (Numbers
corresponds to Fig. 2.)

Cross-sectional TEM analysis was carried out for


the diffusion zone in Ni-Al-Ta. Figure 2 illustrates
the bright field images with the corresponding
selected area diffraction patterns (SADPs) of Tarich precipitates and TEM-EDX point analysis
Al
Ni
Ta
Phase
results (Table 2) respectively. It should be noted
(1) 15.8 62.1 22.0
Ni2AlTa
that some stacking faults were observed within the
precipitates. Fig. 2 indicates that there were
(2)
3.0 68.0 29.0
Ni2Ta
differences in the morphologies and size of the
precipitates present: some precipitates exhibited a plate-like morphology, Fig. 2 (a), and others had a
more blocky morphology, Fig. 2 (b). According to TEM-EDX analysis and SADPs of the
precipitates, the majority were identified as the FCC Ni2AlTa phase, whereas some of the other
precipitates in Fig. 2(b), were identified as the tetragonal structured Ni2Ta. This suggests that
several different phases can coexist within the diffusion layer of as received Ni-Al-Ta.
Microstructural Evolution at 880C
Aluminized Ni-Al-X ternary alloys and CMSX-4 were exposed at 880C, and their representative
microstructures are shown in Fig. 3 for 20 hours (diffusion zone only) and in Fig. 4 for 500 hours
respectively. Due to the relatively low exposure temperature, the microstructural changes in the
coating layer after 20 hours were relatively small, and all samples remained in the -NiAl phase.
374

However, the diffusion zones started to show some microstructural changes after 20 hours, as
shown in Fig. 4. The phase in the diffusion zone in Ni-Al-Cr showed decreases in both the
number and the size of particles. With respect to Ni-Al-Ta, very fine precipitates, having a needle
morphology, appeared above the diffusion zone, whereas the size of the Ni2Ta and Ni2AlTa
precipitated in the diffusion zone decreased. Ni-Al-Ru started to form a layer consisting of very
fine precipitates, with a brighter contrast and located just above the substrate. In terms of Ni-AlCo, Ni-Al-Re and CMSX-4, the microstructural changes in their diffusion zone after 20 hours
exposure were less significant than the other alloys.

Figure 3: Cross-sectional microstructures of diffusion zones formed in all ternary alloys at


880C for 20 hrs.
Whole microstructure

Diffusion zone

Whole microstructure

Diffusion zone

Figure 4: Microstructures of representative aluminized alloys at 880C for 500 hrs.


After 500 hours exposure in Fig. 4, the phase transformation observed in the coating layer was
very striking, depending on the third element X. Ni-Al-Co showed the fastest coating layer
degradation during the exposure, and the coating layer consisted of the '-Ni3Al and phases after
375

500 hours. Most of the other alloys retained the phase as matrix phase, however, some
transformation from to ' phase was observed in Ni-Al-Ru.

Figure 5: STEM-EDX mapping of the diffusion zone in Ni-Al-Ru at 880oC after 500 hours.
(Numbers correspond to Table 3.)
Table 3: STEM-EDX point analysis of diffusion
zones of aluminized Ni-Al-Ru at 880oC after
500 hours (at.%).
Al
Ni
Ru
(1)
27.3
72.7
0.0
(2)
42.4
41.5
16.1
(3)
45.9
35.1
19.0

In terms of the diffusion zone, all alloys showed


microstructural changes, with the differences
between alloys being more significant than after
only 20 hours exposure. The phase
precipitates in Ni-Al-Cr disappeared, and the
interface between the coating layer and the
substrates became more planar, similar to the
interfacial morphology in Ni-Al-Co.

Ni-Al-Re retained the SRZ-like diffusion zone, and some matrix transformation from to ' was
observed, as shown in Fig. 4. However, the thickness of the diffusion zone in Ni-Al-Re did not
change significantly. The very fine needle-like precipitates formed in the diffusion zone of Ni-AlTa coarsened, and started to form discrete precipitates along the grain boundaries. Also, under the
diffusion zone, some needle-shaped small precipitates, appeared after 500 hours. CMSX-4
demonstrated some TCP phase precipitation underneath the IDZ, which are frequently observed
after thermal exposure [17].
In the case of Ni-Al-Ru, the diffusion zones increased in extent. The brightly imaging precipitates
shown in Fig.4 became more noticeable, and increased in number with exposure time. The
precipitates formed in the diffusion zone of Ni-Al-Ru after thermal exposure were investigated
further with STEM. Fig. 5 shows the STEM-EDX mapping results for diffusion zones in Ni-AlRu at 880C for 500 hours, and Table 3 shows the EDX point analysis results. Needle-like
precipitates were distributed within the matrix phase ('- Ni3Al). The width of the precipitates was
less than 500 nm, but their length varied. The EDX data presented in Fig. 5 demonstrated that the
precipitates contained higher concentration of Al and Ru, compared to the matrix phase. Point
analysis in Table 3 revealed that the precipitates consist of approximately 35~41 at.% Ni, 43~46
at.% Al, and 16~19 at.% Ru; these compositions are comparable to the NiAl based 1 phase,
which was reported by Chakravory [13,19].
DISCUSSION
In this study, aluminized Ni-Al-X ternary alloys showed a distinctive morphology within the
diffusion zone before and after thermal exposure, depending on the third element present. In the
LTHA process, Al is the predominant diffusing species, and Al diffuses inwardly to form a
coating layer. The coating layer after the aluminizing process typically consists of -Ni2Al3, and
the -Ni2Al3 in the coating layer transforms to -NiAl during the post-production heat treatment at
376

1100C for 1h. During this process, a diffusion zone is also formed at the interface between the
coating layer and the Ni-based superalloy substrate [20, 21]. During the LTHA process, Al
diffuses into /' phases to form -Ni2Al3 [22], and Ni moves towards the surface to form -NiAl
during the post-production heat treatment. In Ni-Al-X ternary systems, the formation of diffusion
zones after aluminizing depends on the solubility of X elements in the -Ni2Al3, -NiAl and '
Ni3Al phases. For example, Co and Cr have a large solubility in the -NiAl phase [9,11,12]. Co
can form a complete solid solution in Ni-Al and Ni, and this resulted in no diffusion zone
formation after the aluminizing process, as shown in Fig. 2. In terms of Cr, the binary -Ni2Al3
phase also has approximately 12 at.% of Cr solubility [23]. However, the relatively small
solubility of Cr in -NiAl is responsible for the formation of phase diffusion zones.
The solubility of both Re and Ta into -Ni2Al3 and -NiAl phase is much smaller than that of Cr
[10, 24-28]. Specifically, the Re solubility in the -NiAl phase is 0.53 at.% in the as cast
condition [24], and 0.5 at.% in -Ni2Al3 [28]. Therefore, Re rejected from the -NiAl and Ni2Al3 phases and forms Re-rich phase precipitates, as observed in Fig. 2. Nrita et al. [29]
reported that Re has limited diffusion in Ni aluminides. This also suggests that Re in the substrate
was rejected from the -NiAl phase, with segregation occurring at the interface between the
coating and substrate.
With respect to Ta, the low solubility in the phase, which is less than 1 at.% [28], contributed to
the formation of discrete particles in the coating layer and diffusion zone. In particular, a variety
of precipitates and intermetallic compounds have been reported in the Ni-Al-Ta system [25,30]. It
can be considered that the small solubility and variety of possible precipitates in the Ni-Al-Ta
system contributed to the formation of an SRZ-like diffusion zone. Ni-Al-Ta contained a number
of discrete particles within the coating layer in both as coated and as received conditions.
Therefore, it is expected that the lower solubility of a particular X element in -NiAl and Ni2Al3 is strongly associated with diffusion zone formation, as observed in the aluminized
condition in this study.
The morphology of the diffusion zones observed in these systems were particularly distinctive
(see Fig. 2). Although the matrix phase of the diffusion zone in both Ni-Al-Re and Ni-Al-Ta is NiAl, the morphology and the interface between the diffusion zone and substrate is almost
identical to the typical SRZ formed in coated Ni-based superalloys [1,2]. Many of the X-rich (X:
Re or Ta) precipitates within the diffusion zone also grow perpendicular to the substrate in Fig. 2,
and this is a consistent propensity of TCPs growing in SRZ formed in coated Ni-based superalloys
[1-4]. However, the interfacial morphology between -NiAl and the substrate (/' phases) formed
in Ni-Al-Cr was very similar to that in CMSX-4, and therefore was classified as an inter-diffusion
zone. It has been noted that the interface morphology between the diffusion zones and the
substrates was very different, and therefore the diffusion zone in Ni-Al-Cr is defined as an interdiffusion zone, whereas the one in Ni-Al-Re and Ni-Al-Ta has been defined to be an SRZ.
The X-rich precipitates (X: Re or Ta) in the diffusion zones were observed to preferentially
segregate at grain boundaries [2]. X-rich precipitates may cause the formation of many grains in
the diffusion zone during the aluminizing process or post-production heat treatment. A SRZ is
generally formed via a phase transformation from -NiAl to '-Ni3Al in the IDZ during the postproduction heat treatment, however, the results from the aluminized ternary alloys have indicated
that the SRZ can nucleate before the phase transformation occurs. It is believed that the formation
of X-rich precipitates can also act as a nucleation site of a SRZ, and may accelerate the formation
of SRZ nuclei during the post-production heat treatment.

377

The ternary Ni-Al-X alloys demonstrated microstructural changes in both the coating layers and
diffusion zones during the exposure at 880C, however the changes were slower than at 1100C
as expected [18]. In Ni-Al-Co, the coating layer changed to and ' phase, suggesting that all the
-NiAl phase was consumed. This might be due to the inferior oxidation resistance of Co at
elevated temperatures [31]. All of the other alloys still contained the phase after 500 hours
exposure, however, Ni-Al-Ru exhibited some transformation from -NiAl to '-Ni3Al phase near
the specimen surface. The effect of Ru on the oxidation resistance of coatings and Ni-based
superalloys is the subject of debate [32-35], however, it is clear that Ni-Al-Ru showed slightly
faster coating layer degradation than the other samples, except for the Ni-Al-Co.
With respect to the diffusion zones, it must be noted that Ni-Al-Ru exhibited slightly different
diffusion zone formation from the other ternary alloy systems. All alloys except for Ni-Al-Ru
formed the diffusion zones before exposure, however Ni-Al-Ru developed the diffusion zones
consisting of very fine Ru-rich precipitates during exposure. This is likely to be associated with
the solubility of X elements into the various Ni-Al phases. In the case of Ru, there is complete
solid solubility between the NiAl and RuAl phases at 800oC and a high amount of Ru solubility in
NiAl at 1000oC [9]. However the solubility reduced as Al receded. This suggests that Ru
remained within the NiAl phase during the aluminizing process, and it was gradually rejected and
appeared as precipitates during the exposure, due to the Al loss during exposure. The matrix
phase around the Ru-rich precipitates was the Ni3Al phase, with a very limited solubility for Re,
which explains the formation of Ru-rich precipitates during the exposure. Although the formation
of the diffusion zones was observed in Ni-Al-Ru during the exposure, all of the other ternary
alloys exhibited microstructural degradation of diffusion zones in respect of di diffusion zone
thickness and precipitate morphology.
The matrix phase of the diffusion zones transformed into the '-Ni3Al phase, in particular in NiAl-Re and Ni-Al-Ta, and the some precipitates were reduced their numbers, and integrated into
other precipitates in both Ni-Al-Re alloys and Ni-Al-Ta, with the progress of interdiffusion of
alloying elements. Also, Ni-Al-Ta exhibited a greater reduction in precipitates than those in NiAl-Re. At 900C, the diffusion coefficient of Ta in Ni is approximately 8x10-17-2x10-16 m2/s and
that of Re is 4-5x10-17 m2/s respectively [36], with Ta clearly having a faster diffusion rate than
Re. It is expected that part of the Ni2Ta/Ni2Al2Ta phase was consumed into the '-Ni3Al phase
during the phase transformation of matrix phase in the diffusion zone, resulting in a reduced
density of the precipitates in Ni-Al-Ta after the exposure.
In this study, both as aluminized Ni-Al-Ta and Ni-Al-Re alloys formed distinctive diffusion zones
which were defined to be SRZs, indicating that both Ta and Re are prone to SRZ formation at an
early stage. Comparing the microstructural degradation of the diffusion zone between Ni-Al-Re
and Ni-Al-Ta, Re maintained a clear layer and the effect of Re for SRZ formation is persistent
after the exposure. It should therefore be concluded that both Re and Ta promote SRZ formation,
and the effect of Re is stronger than that of Ta. Ni-Al-Ru did not show any diffusion zone
formation after aluminizing, however 1 precipitates appeared after the exposure. Due to the
formation of precipitates, it is rationalized that Ru might have similar effect on diffusion zone
formation to Re and Ta. On the other hand, it can be concluded that Ni-Al-Co and Ni-Al-Cr have
no clear negative effect on the diffusion zone formation. Ni-Al-Cr formed a precipitate-free
diffusion zone, and Ni-Al-Co did not show any diffusion zones. The solubility of X elements
into - Ni2Al3, -NiAl, '-Ni3Al and phases, and the diffusion coefficients of X element in Ni
are contributing factors to the change in diffusion zone thickness and their morphology.

378

CONCLUSIONS
Ni-Al-X ternary alloys designed to contain and ' phase were aluminized, and their
microstructural evolution, especially in terms of diffusion zones, was investigated. The following
conclusions can be drawn from this research:

Aluminized ternary alloys developed distinctive diffusion layers depending on the third
elements: Ni-Al-Cr alloys formed a layer consisting of both and phases, which is more
similar to a traditional inter-diffusion zone in coated Ni-based superalloys, whereas Ni-AlRe and Ni-Al-Ta formed a continuous layer consisting of phase and Re-rich/Ni2Ta or
Ni2AlTa precipitates with high-angle grain boundaries present.

Ni-Al-Ru did not form a diffusion zone after aluminizing, however, the microstructure
developed to contain '-Ni3Al phase and NiAl based 1 precipitates.

The matrix phase of the diffusion layer formed in Ni-Al-Re and Ni-Al-Ta alloys comprised
the phase, however, this layer can be defined to be a secondary reaction zone, because the
layer also consisted of many grains and Re or Ni2Ta/Ni2AlTa precipitates with high-angle
grain boundaries between the layer and substrate.

After exposure at 880C for 500 hours, the diffusion layer disappeared in the Ni-Al-Cr
system, and the diffusion layer in Ni-Al-Re and Ni-Al-Ta demonstrated a loss of Rerich/Ni2Ta and Ni2AlTa precipitates and phase transformations in the matrix phase.

Both Re and Ta can promote the formation of SRZs. The effect of Re is persistent after
thermal exposure, and it is concluded that Re has the stronger effect than Ta.
ACKNOWLEDGEMENTS
This research was supported by the Japan Society for the Promotion of Science (JSPS)
Postdoctoral Fellowship for Research Abroad scheme and the Department of Materials,
Loughborough University. The authors would like to acknowledge Mr K. Roberts (University of
Cambridge), Mr S. Pickering and Dr Z. Zhou (Loughborough University) and Dr R. Jones (RollsRoyce plc) for their technical assistance.
REFERENCES
[1]

[2]

[3]

[4]

[5]

[6]

Walston, W. S., Schaeffer, J. C., and Murphy, W. H., A New Type od


Microstructural Instability Superalloys SRZ-, Superalloys 1996, eds. R.D.
Kissinger et al., (TMS, Pittsburgh, 1996) pp. 9-18.
Suzuki, A., Rae, C. M. F., Yoshida, M., et al., Secondary Reaction Zones in Coated
4th Generation Ni-based Blade Alloys, Superalloys 2008, R. C. Reed, et al. eds,
(TMS, Pittsburgh, 2008) pp. 777-786.
Das, D. K., Murphy, K. S., Ma, S., et al. Formation of Secondary Reaction Zones in
Diffusion Aluminide-Coated Ni-base Single Crystal Superalloys Cotaining
Ruthenium. Metal. Mater. Trans. A, Vol. 39, No. 7 (2008), pp. 1647-1657. .
Suzuki, A. S., Rae, C. M. F., Hobbs, R. A., et al., Secondary Reaction Zone
Formations in Pt-aluminized fourth generation Ni-base single crystal superalloys, Adv.
Mater. Res., Vol. 278, (2011), pp. 78-83.
Suzuki, A. S, and Rae, C. M. F., Comparison of Microstructural Evolution in Pt
aluminized Ni Based Superalloys with and Without Ru, Mater. Sci. Tech., Vol. 29,
No. 6 (2013), pp. 723-732.
Suzuki, A., Microstructure Evolution of Coated Fourth Generation Single Crystal
Superalloys, PhD Thesis, University of Cambridge, UK (2009)
379

[7]

[8]

[9]
[10]

[11]

[12]
[13]
[14]
[15]

[16]

[17]

[18]

[19]
[20]

[21]
[22]
[23]
[24]

[25]
[26]

Rae, C. M. F., and Reed R. C., The Precipitation of Topologically Close-packed


Phases in Rhenium -containing Superalloys, Acta Mater., Vol. 49, (2001), pp. 41134125.
Sato, A., Aoki, Y., Arai, M., and Harada, H., Effect of Aluminide Coating on Creep
Properties of Ni-base Single Crystal Superalloys, J. Japan Inst. Met., Vol. 71, No. 3
(2007), pp. 320-325.
Villars, P., Prince, A., Okamoto, H., Handbook of ternary alloy phase diagrams, Ohio,
USA, ASM International, 1995
Miyazaki, S., Murata, Y., and Morinaga, M. Site Occupancy of Re Atoms in Ni3Al (')
and -' Phase Equilibria in Ni-Al-Re Ternary Alloy, Tetsu-to-Hagane, Vol. 80, No. 2
(1994), pp. 73-77.
Dupin, N., PhD Thesis, Contribution Lvaluation Thermodynamique des Alliages
Polyconstituts a Base de Nickel Institut National Polytechnique de Grenoble, France
(1992)
Dupin, N., Ansara, I., and Sundman, B., Thermodynamic Re-assessment of the
Ternary System Al-Cr-Ni, Calphad, Vol. 25, No. 2 (2001), pp. 279-298.
Chakravorty, S., and West, D. R. F., The Constitution of the Ni-Al-Ru System, J.
Mater. Sci., Vol. 21, (1986), pp. 2721-2730.
Zhu, J., Zhang, C., Cao, W., et al,, Experimental Investigation and Thermodynamic
Modeling of Ni-Al-Ru Ternary System, Acta Mater., Vol. 57, (2009), pp. 202-212.
Tin, S., Yeh, A. C., Ofori, A. P., et al., Atomic Partitioning of Ruthenium in Ni-Based
Superalloys, Superalloys 2004, K. A. Green, et al. eds, (TMS, Pittsburgh, 2004), pp.
735-742.
R.H. Davies, A.T. Dinsdale, J.A. Gisby, et al., MTDATA-thermodynamic and phase
equilibrium software from the national physical laboratory, CALPHAD, Vol. 26, No. 2
(2002), pp. 229-271.
Rae, C M F, M S Hook, and R C Reed, The Effect of TCP Morphology on the
Development of Aluminide Coated Superalloys, Mat Sci & Eng A, Vol. 396, (2005),
pp. 231-239.
Suzuki, A. S., and Thomson, R. C., Inter-diffusion behaviour of coated Ni-Al-X
ternary alloys and coated Ni-based superalloys, Proc 8th High Temperature Corrosion
and Protection of Materials, Les Embiez, France, May 2012, Oxi. Met. In press.
Chakravorty, S., and West, D. R. F., Phase Equilibria Between NiAl and RuAl in the
Ni-Al-Ru System, Scripta Metal., Vol. 19 (1985), pp. 1355-1360.
Das, D. K., Joshi, S. V., and Singh, V,. Evolution of Aluminide Coating
Microstructure on Nickel-base Cast Superalloy CM-247 in a Single-step High-activity
Aluminizing Process Metal. Mater. Trans. A, Vol. 29, No. 8 (1998), pp. 2173-2188.
Goward, G. W., and Boone, D. H., Mechanisms of Formation of Diffusion Aluminide
Coatings on Nickel-Based Superalloys, Oxi. Met., Vol. 3, No. 5 (1971), pp. 475-495.
Reed, R.C., The Superalloys Fundamentals and Applications, Cambridge, Cambridge
University Press, 2006.
Grushko, B., Kowalski, W., Pavlyuchkov, D. et al., A Contribution to the Al-Ni-Cr
Phase Diagram, J. Alloys Comp., Vol. 460, No. 1-2 (2008), 299-304.
Saito, S., Kurokawa, K., Hayashi, S., et al., Phase Equilibria and Tie-Lined
Compositions in a Ternary Ni-Al-Re System at 1423 K, J. Japan Inst. Met., Vol. 71,
No. 9 (2007), pp. 793-800.
Raghavan, V., Al- Ni- Ta (Aluminum-Nickel-Tantalum), J. Phase Equi. Diff., Vol.
27, No. 4 (2006), pp. 405-407.
Huang, A, and Chang, Y. A., A Thermodynamic Description of the Ni-Al-Cr-Re
System, Mater. Sci. Eng. A, Vol. 259 (1999), pp. 110-119.
380

[27] Raghavan, V., Al-Ni-Re (Aluminum-Nickel-Rhenium), J. Phase Equi. Diff., Vol. 31,
No. 1 (2010), pp. 53-54.
[28] Povarova, K. B., Kazanskaya, N. K., Drozdov, A. A., et al., Physicochemical Laws of
the Interaction of Nickel Aluminides with Alloying Elements: I. Formation of Nickel
Aluminide-based Solid Solutions, Russian Metallurgy (Metally), No. 5 (2006), pp.
415-426.
[29] Narita, T., Thosin, K. Z., Fengqun, L., et al., Development of Re-based Diffusion
Barrier Coatings on Nickel Based Superalloys, Mater. Corr., Vol. 56, No. 12 (2005),
pp. 923-929.
[30] Palm, M., Sanders, W., and Sauthoff, G., Phase Equilibria in the Ni-Al-Ta System, Z.
Metallkd, Vol. 87, No. 5 (1996), pp. 390-398.
[31] Bose, S., High Temperature Coatings, Burlington, USA, Butterworth-Heinemann.
[32] Kawagishi, K., Sato, A., Sato, A., et al., Oxidation Behavior of Ru-cotaining Ni-base
Single Crystal Superalloys, Mater. Sci. For., Vol. 522-523, No. 8 (2006), pp. 317322.
[33] Kawagishi, K., Aki Sato, T Kobayashi, and H Harada. "Oxidation Properties for 2-5th
Generation Ni-base Single Crystal Superalloys at 1023, 1073, 1373K." Journal of
Japan Institute of Metals 71, no. 3 (2007): 313-319.
[34] Edmonds, I. M., Evans, H. E., Jones, C. N., et al., Intermediate Temperature Internal
Oxidation in Fourth Generation Ru-bearing Single Crystal Ni-base Superalloys, Oxi.
Met., Vol. 69 (2008), pp. 95-108.
[35] Edmonds, I. M., Evans, H. E., Jones, C. N., et al., The Kinetics of Oxidation of Rubearing Nickel-based Superalloys, Mater. Sci. For., Vol. 595-598 (2008), pp. 59-67.
[36] Karunaratne, M. S. A., Carter, P., and Reed, R. C., Interdiffusion in the Face-centred
Cubic Phase of the Ni-Re, Ni-Ta and Ni-W Systems Between 900 and 1300C Mater.
Sci. Eng. A, Vol. 281, (2000), pp. 229-233.

381

Advances in Materials Technology for Fossil Power Plants


Proceedings from the Seventh International Conference
October 2225, 2013, Waikoloa, Hawaii, USA

Copyright 2014 Electric Power Research Institute, Inc.


Distributed by ASM International. All rights reserved.
D. Gandy, J. Shingledecker, editors

HIGH-TEMPERATURE SOLID PARTICLE EROSION TESTING


STANDARD FOR ADVANCED POWER PLANT MATERIALS AND
COATINGS

V.P. "Swami" Swaminathan, TurboMet International, 8026 Winter Park, San Antonio,
TX 78250
Jeffery S. Smith, Material Processing Technology, LLC, 1965 Forest Park Road,
Norton Shores,
MI 49441
David W. Gandy, Electric Power Research Institute
P.O. Box 217097, Charlotte, NC 28221

Abstract
Solid Particle Erosion (SPE) of hardware remains an ongoing concern with the operation of Steam
Turbine power plants as well as Land Based Gas Turbines. SPE of steam piping, rotating and
stationary components in turbines leads to loss of efficiency, higher cost of operation and
maintenance downtime and cost. Ultra Supercritical (USC) and advanced USC programs
underway in North America, Europe and Asia have created renewed interest in the understanding
of the effects of SPE on the advanced alloys at high temperature. The Gas Turbine industry has
also been conducting studies for improving the elevated temperature erosion resistance of
compressor and hot section components and coatings. ASTMs G76 Standard Test Method for
Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets defines a standard test
method for conducting room temperature SPE erosion testing but has many limitations that
restrict its usefulness to evaluate alloys and coatings used in turbines.
The objective of the current Electric Power Research Institute (EPRI) sponsored program is to
develop an elevated temperature SPE standard that will provide more appropriate reference
parameters for SPE conditions encountered in current and next generation steam turbines, gas
turbines as well as in the selection and qualification of alloys used in other steam path
components. Currently such test standard is not available from any of the standards organizations.
Various laboratories around the world have developed their own equipment and procedures to
conduct their elevated temperature (ET) SPE tests. This makes it difficult to compare these interlaboratory test results for the purpose of screening and selecting alloys and coatings for erosion
mitigation. Organizations from the United States, United Kingdom, China, Germany, Italy and
India have been participating in the EPRI inter-laboratory Round Robin test program to develop
a new ASTM ET erosion test standard. Standard test coupons from Type 410 stainless steel have
been tested using a standard alumina erodent powder at specified test conditions at room
temperature and 600C. Impingement angles of 30 and 90 degree were used. An international
conference on solid particle and liquid droplet erosion was organized in 2012 to gather the world
experts in erosion testing, modeling and application [1]. Valuable information was obtained at
this conference which helped in the preparation of the first draft standard which has been
382

submitted to ASTM for balloting by the committee membership. This presentation will describe
the program structure, test conditions used, test results and the development of the draft ASTM
ET SPE test standard.
Key Words: Solid particle erosion; high-temperature; round-robin testing; ASTM Standard;
steam turbines; gas turbines; particle velocity measurement
INTRODUCTION
Solid particle erosion has been a pervasive generic problem in fossil power generation equipment.
Exfoliation of oxide particles from the steam side surfaces of high-temperature steam path
components in boilers and stem pipes lead to erosion of turbine blades, nozzles and control
valves. Electric Power Research Institute (EPRI) conducted studies on the extent of this problem
and costs associated with this on utility steam turbines [2,3,4]. While thermal spray coatings
have been effective at increasing the SPE resistance of the uncoated component, component
erosion damage is still observed in service.[5] Some examples of the effects of SPE on hightemperature steam turbine parts are shown in Figure 1.

Figure 1. SPE Damage on High Pressure Steam Turbine Blades (courtesy of Encotech)
This SPE problem is expected to be more pronounced in USC turbines which operate at much
higher temperatures and pressures. Several high-temperature steels and exotic alloys in
combination with protective coatings are being developed and evaluated for the USC application.
Several papers on this subject were presented at the EPRI conference in 2010[6].
However, their erosion properties and resistance to SPE of these alloys under plant operating
conditions have not been reported.
It is important to understand the SPE behavior of the alloys used in high-temperature application
under various conditions in order to develop effective erosion mitigation options which may
include coatings and other surface modification techniques. The variables which control the SPE
behavior are the temperature, particle size, velocity impact angle, hardness of the particles as well
as the substrates, morphology of the particles (sharp, blunt, angular etc). The particle size
distribution found in a typical boiler scale of a fossil power plant is shown in Figure 2 [7] The
composition of the scale is mainly magnetite (Fe3O4). The particle size distribution varies from
about 5 microns to 100 microns. Depending on the volume fraction, velocity, angle of the
particles, the erosion characteristics will vary.

383

Figure 2. Particle size Distribution of Boiler Scale Recovered from Boiler Water and Superheater Steam Drains[7]
SPE in aero engines is also a problem for engines operating in dusty environment. Damage to
the compressor blades as well as the hot section components reduce the efficiency and life of the
engines resulting in costly change outs and repairs. The effect of SPE on an aero engine
compressor blade is shown in Figure 3. Increasing the SPE resistance of aero engine compressor
components with protective coating continues to be an active area of development.[8,9]

Figure 3 Comparison of compressor blade damaged by SPE (left) with a virgin blade (right) [10]
A brief overview of the SPE behavior of some of the materials under different conditions is
presented below.
Solid Particle Erosion Characteristics
The topic of solid particle erosion has been studied by numerous researchers over the years.
Several good reviews of the particle erosion literature can be found in articles by Wright [11],
Finnie [12], and Mathews.[13] One of the key concepts that has been identified is the difference
in the erosion behavior of ductile materials such as metals and that of brittle materials such as
most ceramics. With ductile materials the erosion response as a function of particle impact angle
384

has been show to approach zero at very low angles of attack, increases to a maximum as the angle
of incidence is between 15-20 degrees and then drops to 1/2 to 1/3 of the maximum erosion rate
as the particles impacting the surface approach 90 degrees. The erosion rate of brittle materials is
at a maximum at 90 degrees with the rate decreasing continually to a negligible mass loss at very
low angles of impact. This response reflects fracture induced mass loss where the extent of the
erosion is dependent on the vertical component of the particle impact energy. This difference in
ductile and brittle material behavior is plotted in Figure 4.
~ 1.7 mg/g

~ 10 mg/g

Figure 4. Erosion Behavior of Ductile (Al) and Brittle (Al2O3) Materials [12]
In ductile erosion, the metal is indented by the particles impacting the surface and material is
extruded around the indentation. At high angles the energy of the particle is dissipated through
ductile deformation and is more resistant to erosive wear than at low angles were the metal
indentation proceeds by a plowing or micromachining action. At high angles the material removal
mechanism is thought to proceed by work hardening of the extruded material by repeated impacts,
leading to local fatigue or fracture based loss of material. With brittle materials the particle
impact generates brittle fracture within the near surface zone of the material, with cracks radiating
outward and downward from the point of impact. These mechanisms are illustrated in Figure 5.

Figure 5. Particle Erosion Mechanisms. A) Plowing and Extrusion Mechanism for Ductile
Materials B) Brittle material Erosion Proceeds through Lateral and Radial Cracking around the
Crush Zone. (From DAlessio 1994) [14]

385

EPRI Program Objectives and Approach


The main objective of this EPRI project is to promote and facilitate high-temperature solid
particle erosion test standard development for application in steam turbines and land based & aero
gas turbines engines. The technical approach to accomplish the objective is as follows:
1.

Conduct a literature search and survey of high-temperature erosion test facilities and
capabilities around the world

2.

Define the various testing parameters and develop a test matrix for elevated temperature
erosion testing

3.

Conduct round robin tests and perform statistical analysis of the results

4.

Organize an international workshop on solid and liquid particle erosion

5.

Develop a high-temperature solid particle erosion test standard which could be used by
the international community to conduct SPE testing and compare the results on a
common basis

The standard testing method or guideline developed under this EPRI project is used to develop an
ASTM standard for high-temperature SPE testing. The SPE test methods used for the evaluation
and characterization of various alloys and coatings vary greatly among the various laboratories.
Several test equipments, techniques and procedures were used by the participating laboratories
and provided results to develop this test standard.
Elevated Temperature Erosion Testing
Under this EPRI program, a detailed survey of the high-temperature erosion testing capabilities
around the world was conducted. All of the steam and gas turbine manufacturers, independent
testing laboratories, universities, coating manufacturers and turbine repair companies were
surveyed to gather information. The data collected include temperature limits, velocity, methods
used to obtain high-velocities (compressed air jets vs. combustion tunnels), methods used in
velocity measurements, types of erodent and sizes used. Details of the test equipment and
methods used by the laboratories were presented in Reference [15].
In general this test method utilizes a repeated impact erosion approach involving a small nozzle
delivering a stream of gas containing abrasive particles which impacts the surface of a test
specimen. A standard set of test conditions for the elevated temperature testing were provided to
the test labs. However, deviations from some of the standard conditions are permitted if described
thoroughly. This allows for laboratory scale erosion measurements under a range of conditions.
Test methods are described for preparing the specimens, conducting the erosion exposure, and
reporting the results. Among the key variables are the type of erodent (i.e. Alumina, Silica,
Arizona Road Dust, Iron Chromite, Magnetite, etc.), particle size range (10 micron 200 micron
or larger), particle velocity (30 m/s to 215 m/s or higher) and angle of impingement (15 to 90
degrees). In reporting erosion test results it is important for these parameters to be specified. The
ASTM G76 test method for conducting solid particle erosion via gas jets is often referenced for
erosion test studies conducted at room temperature. It was developed for erosion characterization
of structural materials and is not completely suited for conducting studies of turbine components
or coatings. This has resulted in a number of different approaches being taken for testing these
386

materials. Initial room temperature screening is generally carried out with some form of a
modified G76 test since it is the simplest test to set up [16].
A schematic diagram of the elevated temperatures testing system used by the participating
laboratories around the world is shown in Figure 6.
Insulated Mixing Chamber
and Supply Tube

Figure 6. Schematic drawing of test system used for elevated temperature SPE testing and
erosion scar geometry achieved for two angles of impingement (30 and 90 degrees) on Type 410
stainless steel test coupons
Typically researchers will modify the test conditions to be more representative of field conditions
and to be better at monitoring erosion of coatings especially those that are in the 10 -20 micron
thickness range for compressor airfoil applications [9]. Typical modifications are use of silica,
instead of alumina, to use higher particle velocities to better simulate turbine conditions, a larger
diameter nozzle to increase the area tested and evaluating weight loss instead of volume loss.
Erosion results are influenced by the hardness, friability and angularity of the particles. Silica
provides results that are closer to conditions in the field than alumina does for aero engine
applications [6]. Several particle size ranges may be used depending on the SPE test objective.
The larger test area helps to screen for coating defects that might be present and to improve the
resolution of the test. Erosion rates are usually defined as the number of grams of material or
coating eroded per gram of erodent impacting the sample (mg/g) rather than as a volume loss
(mm3/g) as specified in G76. For coatings it is difficult to establish a reliable density to calculate
volume loss from the weight loss.

387

A critical feature in all erosion testing is the use of a witness coupon for comparison of erosion
results taken at different times. Often it will be the substrate material being coated (e.g., Ti-6Al4V, IN-718, 17-4 PH etc.) rather than 1020 steel called out in G76. This also provides some
measure of erosion performance compared of the substrate to guide the development of any
surface modifications needed. It is difficult to make comparisons erosion data generated in
different labs since there are often differences in the actual practice of the test that lead to
different erosion rates. Two keys areas of variation are the actual particle velocity and geometry
factors with the amount of erodent hitting the coupon. In the case of particle velocity it is
difficult to measure accurately without specialized equipment (e.g., laser doppler velocimeter
(LDV) or particle image velocimeter (PIV) etc). Some of the laboratories use double rotating disc
(DRD) method to determine the particle velocity [17]. Another factor is the amount of erodent
actually hitting the sample. At low angles this can be especially significant since the erosion
footprint may be larger than the coupon being tested.
The following organizations have high-temperature erosion test facilities and provided details of
their testing capabilities and participated in the EPRI sponsored round-robin test program to
develop the ASTM test standard..

Air Force Materials Lab (AFML) / University of Dayton Research Institute (UDRI),
Dayton, USA (room temp testing only and baseline data); LDV used for real-time
velocity measurements
General Electric Company, Schenectady, USA (test facility at GE Global Research Ctr,
Bangalore, India) - room temperature and 600C tests; DRD for velocity measurements
Cranfield University, Cranfield, UK Room temperature and 600C tests; no direct
velocity measurements but fluid dynamics principles used to compute particle stream
velocity
ERSE SpA, Milan, Italy Room and 600C tests; DRD used to measure the velocity
Institute of Turbomachinery, Xian Jiaotong University, Xian, P.R. China Room and
600C tests; PIV used in real time for velocity measurements
DUCOM, Bangalore, India Room temp and 600C tests; DRD used for velocity
measurements

Configurations and photos of the test systems, their maximum velocity (at RT) and temperature
capabilities from these test facilities are shown below in Figures 7 through 12.

Figure 7. SPE Test facility at AFML/UDRI, Dayton, OH (330 m/s; RT Only)

388

sample
holder

screw feeder

heating system

acceleration tube

compressor
exhaust
pressure vessel

Figure 8. Test system schematic (left) and photo at Cranfield University, UK, (200 m/sec; 850C
max. )

Figure 9. Schematics of the test system at the Turbomachinery Labs, Xian University, China
(450 m/s; 650C)

Figure 10. High-temperature erosion test rig at GE Global Research, Bangalore, India (305 m/s;
982C)
389

Figure 12. Test system used at DUCOM,


(200 m/s; 800C max.)

Figure 11. Test system at ERSE SpA, Mila,


Italy, (200 m/s; 800C max.)

As can be seen in these figures the architecture of the test systems and their capabilities are
diverse. The nozzle, specimen design, mounting methods, stand off distances, erodent feeding,
mixing, temperature control, etc., vary among these test facilities. Two of the test rigs employ
combustion gases and the others use compressed air with external heat. Velocity measurements
methods are also unique to each of these systems.
Round Robin Test Matrix
The material selection was based on the alloys used in high-temperature steam turbine application.
Type 410 stainless steel or similar steels are used in the steam inlet regions of most of the
turbines. The erosion test coupons size is fixed at 75 mm x 25 mm x 4.5 mm which is suitable for
all of the laboratories. Some of the labs need to make minor modification of their specimen
fixture arrangements to accommodate this sample size. The test matrix is shown in Table 1.

390

Table 1. EPRI Round Robin Test Matrix for room temp. and 600C in comparison to
ASTM G76 specification
EPRI ROUND ROBIN TESTS
(Room Temp. & 600C)
200 m/s particle velocity
Adjust nozzle to test coupon standoff
distance to create 14 mm diameter erosion
scar
1.5 - 9 mm (0.060 0.360) Nozzle Diameter
(dependent on lab)
50 Alumina erodent (from a single batch)
410 Stainless Steel Substrates
2 grams/minute powder feed
5 - 10 minute min. test intervals per sample
RT & 600C Test Temperatures
30 & 90 degree impingement angles
mg/gram of erodent to be reported

ASTM G76 SPECIFICATION


30m/s particle velocity
10 mm stand off distance; results in relatively
small erosion scar
1.5 mm (0.060) Nozzle Diameter
50 Alumina erodent (multiple sources)
1020 Steel Substrates
2 grams/minute powder feed
10 minute min. test time
Room Temperature Test Only
90 degree impingement angle
mm3/gram of erodent (need to know
substrate/coating density to calculate weight
loss)

The chemistry and mechanical properties of the Type 410 stainless steel coupons used in this
program are given below.
Table 2. Chemistry of Type 410 Stainless Steel Test Coupons
Grade 410 SS

Fe

Cr

Ni

Mn

Si

min.

bal

11.5

Coupon Lot

bal

12.1

0.13 0.31 0.49

max.

bal

13.5

0.75

Ti

0.13 0.014 0.021 0.002 0.0074


-

0.15 0.04 0.03

Dimensions: 25mm x 75 mm x 4.5 mm


Yield strength: 42.5 KSI
Microstructure of the 410SS Coupons

Tensile Strength: 64.5 KSI


Hardness: 74 - 76 Rb
Surface Finish: < 0.2 microns Ra

The erodent selected was 50 alumina for this round robin test program. This powder lot was
evaluated for particle size distribution and particle characteristics. It is highly desirable to keep
the size distribution tightly around 50 . A single master lot of white alumina powder was
procured from Japan and distributed to the test labs. (JIS 6001-320: D10 = 33.8; D50 = 50.3;
D90 = 74.6). Additional details on the particle size distribution are covered in Reference [18].

391

TEST RESULTS AND DISCUSSIONS


Room Temperature (24C) Data
Initial baseline tests were conducted at room temperature at AFML/UDRI labs. The erosion rate
data obtained showed linear trend as shown in Fig. 14. This baseline data was used to compare
the results from the other labs. Other feed rates up to 5 g/min were also used to evaluate the effect
of feed rate on the erosion rare. No significant differences were obtained in the erosion rate
expressed as mg/g. However, all of the tests in this program were conducted at a feed rate of 2
g/min.

Figure 13. Tight clustering of baseline erosion test data on Type 410 Stainless Steel at Room
Temperature (24 C), 200 m/s at 90 degrees with 50 micron alumina at 2 g/min feed rate
Table 3 summarizes the results of room temperature tests conducted by the various labs. Five
readings were taken on each coupon at 20 g intervals. Thus a total dose of 100 g erodent was
applied for each coupon. Five coupons were used for repeat tests and the average of the five
reading from each coupon was reported along with the individual readings. Thus a total of 25
readings were obtained for each angle of incidence at a given temperature. The baseline data by
Lab 9 with known measured velocity (by LDV) of 200 m/s was used to compare the results from
the other labs. As mentioned previously, each of the labs used different methods to measure or to
estimate the particle velocities. The reported velocities are 200m/s +/- 10m/s. However, it can be
seen from this table that the results reported by Lab 3 are much lower than that of the other labs.
This raised the question about the particle velocity measurement accuracy at these various labs.
Additional tests by selected few laboratories were conducted at different velocities to understand
the effect of velocity on the erosion rates and arrive at a common basis for cross comparison of
the results among these labs.

392

Table 3.

Results of the solid particle erosion tests conducted at room temperature by the
laboratories at 90 and 30 degree impingement angles at 200m/s (Note: the lab ID
numbers were assigned at random)

Since the Type 410 stainless steel is a ductile material, the erosion rates follow the trend similar to
that shown in Fig.4 for ductile materials. The erosion rates at 30 degree impingement are higher
by approximately a factor of 2x compared to the 90 degree data in Table 3.
Test Data at 600C
Out of the seven labs that conducted room temperature testing only five of the labs were able to
conduct high-temperature tests at 600C at 200 m/s particle velocity due to equipment limitations.
Labs 1 employed direct in-situ velocity measurements by PIV whereas three labs used DRD and
the one of the labs estimated the velocity using fluid dynamics principles and prior calibration
data. When DRD is used, the particle velocity is measured prior to the erosion tests on the
coupons. Then the DRD is removed from the test chamber and the test coupon is positioned for
the erosion tests. These velocity measurements are done at room temperature since DRD cannot
be used in high temperature environments. The hopper which supplies the erodent powder is kept
at room temperature. The particle stream is exposed to high temperature as it enters the career hot
gas stream or the heated test chamber depending on the test system.
The results from the 600C tests from five labs are summarized in Table 4. For comparison, data
from room temp from Lab 9 is also included in this table.

393

Table 4. Summary of results of erosion tests at 600C by five laboratories at 200 m/s along with
the baseline room temperature data from Lab 9.

The rests fall with an acceptable scatter band. The data at 90 impingement angle at both
temperatures (RT and 600C) show very small difference. This infers that the erosion mechanism
is similar at both temperatures. The 30 degree data at 600C show higher erosion rates with more
scatter than the room temperature data. It should be noted that there was no noticeable oxidation
on the Type 410 stainless steel test coupons at 600C. This is expected since the cumulative
exposure time to high temperature is less than an hour for any given sample during these tests.
The interlaboratory study (ILS) group at ASTM International conducted a precision and bias
study on all of the test data. Both the repeatability and reproducibility were determined for the
data sets and they ascertained that all of the data are within acceptable limits. A detailed treatment
and discussions will be presented in a research report prepared by ASTM in support of the
proposed test standard.
Erosion Tests at Additional Velocities
Additional tests were conducted by two of the laboratories at velocities ranging from 60 m/s to
232 m/s at room temperature and 90 degree impact angle. The velocities were measured by
different techniques at each of these labs using PIV, LDV or a DRD system. One of the labs
conducted a series of tests at predetermined velocities. The test data summary and a plot are
presented in Fig. 14.

394

Figure 14. Summary of average erosion rate data and power-law plot at additional velocities at
room temperature and 90 degree impact angle
The relationship of erosion rate (E) and impact velocity () is shown below in Equation 1.

E=kvm

(Equation 1)

where k is a constant and m is the velocity exponent.


The data falls in a tight scatter band and describes the erosion behavior very well in this velocity
range. There does not appear to be a mechanism change in this velocity range. The pre
exponential constant and the velocity exponent are included in the plot in Fig.14. The exponent
value of 2.2 indicates the parabolic nature of the erosion behavior in this velocity range. This
equation could be used to cross check the velocities from tests conducted by any other lab and
could be considered as a calibration curve. One of the participating labs conducted tests at
additional velocities at 600C at 90 deg. and 30 deg. impact angles. Particle velocity was measured
using a DRD apparatus. Figure 15 shows the summary plot of the data at 90 deg. The velocity
exponent of 2.24 indicates, similar erosion mechanism at room temperature is also occurring at
600C.

Figure 15. Power Law data fit for the erosion rate as a function of particle velocity at 600 C and
90 degree impingement angle
395

A summary plot of erosion test data obtained at 600C/30 deg. angle is shown in Fig. 16. Again,
the velocity exponent is close to 2 which indicate the parabolic nature of the erosion phenomenon
with velocity and the erosion mechanism seems to be independent of temperature under the
testing conditions used in this program. Of course, the elative positions of the curves, i.e., the
actual erosion rates, are determined by the pre-exponential constants.

Figure 16. Power-law erosion rate plot at additional velocities at 600C and 30 degree
impingement angle
Erosion Scar Analysis
Erosion scars formed on some of the randomly selected test coupons were analyzed using optical
microscope and contact profilometry techniques. The diameter of the circular scar formed at 90
degree impingement angle, major and minor axes of the elliptical scar formed at 30 degrees and
the maximum depth of the scar are important parameters. Scar analysis is more important when
the erosion tests are conducted on coated samples. Volume and the depth of the coating removed
are critical since the weight may vary depending on the composition of the coating. Also,
breaching of the coating is undesirable to have valid data on the coatings. Typical surface
profilometry of a scar for 90 degree impingement is shown in Fig. 17 and that for 30 deg. is
shown in Fig. 18.

Figure 17. Contact profilometry data of RT test scar at 90 deg. impingement; 5mm nozzle dia,;
14 mm stand-off distance; 25 g dose; Scar diameter = 10mm; maximum depth =
0.42mm
396

Figure 18. Contact profilometry of RT test


scar at 30 deg; 5mm nozzle dia, 18 mm SOD;
50 g dose; major axis = 20 mm; minor axis =
11 mm; Maximum scar depth = 0.23mm

Figure 19. Optical profilometry of erosion scar


on test specimens eroded 30 degrees at room
temperature. 2mm nozzle dia; 16 mm SOD;
100g dose.

Figure 19 shows the results obtained using an optical profilometry system on a scar of a specimen
tested at room temperature at 30 degree impingement. The difference in the measured parameters
between Fig. 18 and 19 are due to the differences in the nozzle diameter, SOD and the total dose
applied. These results illustrate methods of scar analysis and the variability in the scar geometry
due to test system set-up parameters and test conditions. Such scar analysis will help refine the
test coupon set-up and standoff distance to obtain desirable scar sizes.

Summary
Currently high-temperature solid particle erosion testing is conducted by many organizations
around the world using their own equipment and procedures developed in-house. This EPRI
sponsored round-robin testing and ASTM test standard development program was directed at
developing a high-temperature solid particle erosion test standard which will help the industry to
perform such tests and compare the results on a common basis. Such common test method and
data analysis procedure is essential to screen materials and coatings which may be selected for use
in USC and other high-temperature steam power plants as well as gas turbine applications. This
program was very successful in organizing an international conference on this critical subject to
assemble many well known experts in the field of solid particle and liquid droplet erosion to share
the current technology, ideas and information developed over many years. There was a consensus
in this forum for the development of a common test standard which could be used by any
organization to conduct these erosion tests. Many laboratories around the world voluntarily
participated in a round robin test program to perform tests on standard coupons provided by
EPRI. Type 410 stainless steel coupons from a single heat and 50 micron alumina powder from a
single batch was distributed to these laboratories.
Tests were conducted at room temperature and at 600C at impingement angels of 90 and 30
degrees. The test results
reported by these labs show good general agreement at these different test conditions. The erosion
rates show a linear relationship with the applied erodent dose at all test conditions. However,
noticeable difference was found in the initial reported erosion rate results from some of the labs
for the same reported velocity. Additional tests were conducted at different known particle
velocities by selected few labs. A power law with an exponent of 2 applies to the erosion rate as a
function of particle velocity. Due to this power law nature of the erosion rate, it is highly critical
397

to accurately measure the particle velocities. A master calibration curve relating the erosion rate
and velocity is proposed. Rigorous statistical analysis of all the data conducted by the ASTM
interlaboratory study group show good repeatability and reproducibility of the results. Limited
erosion scar analysis was also conducted to provide some guidance on the use of two different
methods. A draft test standard has been developed and submitted to the ASTM subcommittee
G02-10 for balloting. At the time of this writing, first ball