LIFE ASSESSMENT OF CRITICAL BOILER AND TURBINE

COMPONENTS USING EPRI’s Creep-FatiguePro

SOFTWARE

Darryl A. Rosario
Structural Integrity Associates
San Jose, CA

Richard M. Tilley
EPRI
Charlotte, NC

ABSTRACT
The accumulation of creep and fatigue damage over time are two principal degradation mechanisms
which eventually lead to crack initiation in critical high temperature fossil power plant boiler and
turbine components. In extreme cases, the growth of these cracks can lead to leakage and rupture. The
monitoring and prediction of the rate of accumulation of creep-fatigue damage is therefore essential in
developing an effective life assessment strategy to prevent catastrophic failures and to manage the
remaining life of these components. Because the rate of accumulation of creep-fatigue damage is
directly dependent on actual plant operating conditions, the ability to access plant operating data directly
from the main plant computer system and process this data in real time is critical in the design of an
advanced life consumption monitoring system.
In response to this challenge EPRI has funded the development and application of a stand-alone
windows-based personal computer program, Creep-FatiguePro, to access and process real-time plant
operating data to track creep-fatigue damage accumulation and crack growth in critical high-temperature
fossil plant boiler and turbine components. This software system is typically configured to reside on a
single dedicated personal computer located either at the plant or an off-site engineering office. An
interface with the existing plant computer/data acquisition system is established to collect temperature,
pressure, flow rate and other instrument data from the components to be monitored. Creep-FatiguePro
processes this operating data to determine the creep-fatigue stresses, damage accumulation and crack
growth of known or postulated defects over time. By projecting the monitored damage and crack
growth rates into the future, Creep-FatiguePro determines the expected additional operating time
required to reach critical levels of damage accumulation or crack size. This information helps the utility
engineer make accurate and timely decisions regarding component inspection need and schedule,
run/repair/replace decisions for detected damage, and evaluate the impact of alternative plant operating
modes on component life.
This paper summarizes the Creep-FatiguePro software methodology, new software features and presents
a recent utility application.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 1
INTRODUCTION
Thick-section fossil plant components operating at temperatures of 1000ºF or higher are subject to
material degradation over time due to creep and fatigue. This includes boiler superheater and reheater
headers, main steam and hot reheat piping, and turbine steamchests, casings and rotors. The actual rates
of accumulation of creep and fatigue damage and crack growth in these components are strongly
dependent on the specific temperature and pressure operating histories to which they are exposed.
Therefore, significant improvement in predictions of damage accumulation, crack growth, and
remaining useful life can be attained when these calculations are based on actual plant operating
conditions as determined by on-line monitoring of data from plant instrumentation.
The Creep-FatiguePro software was originally developed under EPRI sponsorship in 1992 for
monitoring of critical girth and nozzle weld locations in main steam and hot reheat piping systems [1].
Since then, the Creep-FatiguePro technology has been extended to include monitoring of high-
temperature boiler header body welds or tubebore ligaments, turbine inlet steamchests, inner/outer
turbine casings, and rotor bores [2-4]. The DOS version of software has recently been upgraded to
Windows with built-in data trending and plotting capability [5]. The current Windows software version
provides utilities with a stand-alone comprehensive life assessment package for accurate and economical
monitoring of creep and fatigue damage in all critical fossil plant components.

OVERVIEW OF METHODOLOGY
The monitoring approach employed by Creep-FatiguePro, is illustrated in Figure 1, and can be
summarized as follows:
• Temperature, pressure flow rate and RPM data for the monitored components is collected from the
plant computer at 30-60 second intervals and stored in an appropriate format for processing in Creep-
FatiguePro.
• Component-specific analytical Stress Transfer Functions programmed into Creep-FatiguePro
compute creep and fatigue stresses as a function of time from the collected plant instrument readings.
• Using the monitored temperatures, calculated stresses, and built-in material properties for creep
rupture and fatigue crack initiation, creep and fatigue damage is calculated and accumulated over
time to provide an indication of the time to crack initiation.
• Using the monitored temperatures, calculated stresses, and built-in material creep and fatigue crack
growth properties, crack growth is calculated for an initial known or postulated defect.
• Monitored damage accumulation and crack growth rates are projected into the future to predict the
time required to reach critical levels of damage and crack size. Probabilistic remaining life output,
allowing for variations in the initial crack size and the material creep crack growth properties, is also
provided.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 2
 Figure 1. Flowchart Illustrating Creep-FatiguePro Monitoring Approach.

Key aspects of this monitoring approach are the concept of stress transfer functions, and the
methodologies for calculation of damage accumulation and crack growth. A brief description of each of
these is given below.

Stress Transfer Functions

Stress transfer functions are geometry-dependent functions which are applied to plant instrument
readings to determine component stresses. They are specifically determined for each component
monitor location when configuring Creep-FatiguePro for a given plant installation. For simple
geometries such as straight sections of main steam or hot reheat piping, these transfer functions can be
adequately determined using standard closed-form solutions. For more complex geometries such as
header tube bore ligaments, turbine inlet steamchests, casings and rotors, detailed elastic and creep finite
element analyses are typically performed to derive the transfer functions. The inelastic creep analyses
are required to adequately account for the effects of creep stress redistribution and/or relaxation.
For the majority of components monitored with Creep-FatiguePro, internal pressure and operating
thermal transients are the principal stress sources for which transfer functions are developed. In the case
of high energy piping, transfer functions to address piping deadweight and global thermal expansion
effects are included. For turbine rotor bore locations, rotational stresses are also considered.
With the exception of thermal transients, the magnitude of stresses produced by the above-noted
loadings are dependent only on the instantaneous reading of the applicable plant instrument (i.e.,
pressure transducer reading for pressure stress, thermocouple reading for piping thermal expansion
stress, turbine RPM for rotor rotational stress). Thermal transient stresses are dependent not only on the
instantaneous temperature reading, but also on the prior temperature versus time history of the fluid in
the component. A specialized technique referred to as the Green's Function approach is employed in

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 3
Creep-FatiguePro for the calculation of these thermal transient-related stresses and is described in
reference [1].

Damage Calculation
Creep and fatigue damage is calculated in Creep-FatiguePro using a life fraction approach equivalent to
that employed in ASME Code Case N-47 for high temperature nuclear components. This approach,
described in reference [6], is summarized as follows:
D = Σn/Nf + Σt/tr (1)
where, D is Life fraction or Damage Index, representing the fraction of total creep-fatigue life consumed
at any given time, n is number of applied fatigue stress cycles for each loading condition, Nf is
corresponding number of cycles to failure for each loading condition, reflected in a fatigue S-N curve
for the applicable material, t is Time duration of each temperature-stress loading condition, and, tr is
time to creep rupture for each temperature-stress loading condition reflected in stress rupture or failure
LMP curves for the applicable material.
Under conditions of pure creep or pure fatigue, crack initiation is theoretically predicted when the
damage index sums to unity. Detrimental creep-fatigue interaction effects can be incorporated in Creep-
FatiguePro by specifying reduced allowable values of D. The creep-fatigue damage envelope of Code
Case N47 can serve as a guide for determining this reduction in allowable D; an allowable of
approximately 0.15 is prescribed for an equal mix of creep and fatigue.

Crack Growth Calculation

Fatigue crack growth is calculated in Creep-FatiguePro using the approach outlined in Appendix A of
the ASME Code, Section XI, Rules for Inservice Inspection of Nuclear Power Plant Components. This
approach employs linear elastic fracture mechanics concepts and a Paris Law crack growth equation as
follows:
da/dN = C∆Kn (2)
where, da/dN is fatigue crack growth per cycle, ∆K is the range of fracture mechanics stress intensity
factor, and, C, n are constants describing the fatigue crack growth behavior of the applicable material.
The Ordered Overall Range (OOR) cycle counting routine, described in reference [1], is used to extract
peak and valley stresses from which ∆K values are computed. Material fatigue crack growth constants
from ASME Section XI are used to bound crack growth for both 2¼Cr-1Mo and 1¼Cr-½Mo piping and
header steels. Fatigue crack growth constants for Cr-Mo-V rotor steels are obtained from reference [7].
Creep crack growth is computed using Saxena's transient creep parameter, Ct [8], and is given by the
following equation:
da/dt = H (Ct)q (3)
where, da/dt is creep crack growth rate (inch/hr), Ct is the transient creep crack tip parameter (in-
kip/in2-hr), and, H, q are constants describing the creep crack growth behavior of the applicable
material. The methodology for computing Ct is documented in reference [8], along with the material

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 4
creep crack growth constants H and q for a variety of applicable boiler steels. Creep crack growth
constants for turbine rotor steels are obtained from reference [9].
In general, a database of all necessary material properties required for creep and fatigue damage and
crack growth calculations is maintained within Creep-FatiguePro. This database covers the majority of
alloy steels typically used for high temperature boiler and turbine components and distinguishes
between base material and weld heat-affected-zone properties. The database can be readily updated or
expanded as new data becomes available or to suit a particular application. Similarly, a library of
fracture mechanics flaw models, containing linear elastic (K) and creep (Ct) solutions for a number of
flaw geometries, is available in Creep-FatiguePro for application to a variety of component monitoring
locations.

CREEP-FATIGUEPRO SOFTWARE DESCRIPTION

The latest Version 3.0 of the Creep-FatiguePro software program [5] was written in Microsoft C++ for
operation on a personal computer (PC) in a Windows environment (Figure 2).

Figure 2. Creep-FatiguePro Splash screen and Main Menu Options.

Key main menu options, shown in Figure 2, are as follows.

The “File” Option allows the user to setup a new project, and to access or edit an existing project. Each
project defines which plant data, database, charts and report files are to be used for evaluation.
The “View” Option allows the user to view the contents of the binary databases, analysis log or “.CDT”
data files in text format.
The “Analysis” Option offers the user the choice of processing data files, creating reports, performing a
probabilistic analysis or undoing a previous analysis.
The “Report” Option allows the user to view/print monthly, summary or probabilistic text (tabular)
results reports.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 5
The “Chart” Option contains a powerful set of built-in graphics tools that allows the user to view/print
various charts (plots) of input and calculated data values.
The “Window” Option allows the user to switch between various active project windows and/or report
and chart windows.
The “Help” Option allows the user to access “Help Topics” which cover the entire content of the user’s
manual and to verify the version number of the program. Context-specific help can also be accessed
using the “F1” function key from anywhere in the program.

SOFTWARE IMPROVEMENTS
The latest software release incorporates the following significant improvements over earlier DOS
versions:
(i) Windows GUI: an easy-to-use Windows graphical user interface.
(ii) US and SI Units: Setup/configuration that can support either US customary or SI units.
(iii) Charting Tools: Powerful charting tools are incorporated within the latest software release
eliminating the need for external links to other spreadsheet programs like MS Excel and Lotus,
required by earlier versions. Considerable flexibility is offered under the main menu “Chart”
option allowing users to access and plot pre-defined “Custom Charts”, as shown in Figure 3a or to
create their own chart using any number and combination of plant instruments and calculated
variables (temperature, stress, damage and crack growth), as shown in Figure 3b.

(a) Custom (b) User-Defined

Figure 3. Creep-FatiguePro Built-in “Chart” Options.

Charting tools also allow the choice of two levels of data refinement used in plotting: “Trend” data
which represents hourly average values and “Detail” data which includes minute-by-minute data
resolution. Further data interrogation is facilitated via horizontal/vertical data zooming and
panning within a plot, as shown in Figures 4a and 4b.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 6
 (a) May 1-31,2001 Data (b) 9-Hour Zoom for Startup on May 8, 2001
Figure 4. Creep-FatiguePro Built-in “Chart” Zoom Option for Data Interrogation.
(iv) Database Conversions: A built-in coded-to-binary database converter for site-specific
customization. Database updates and changes can be easily implemented using this feature.

RECENT UTILITY APPLICATION

Since the recent software upgrade to Windows, Creep-FatiguePro has been successfully installed and
tested at a 250 MW fossil plant. This unit is equipped with a boiler and turbine manufactured by
Mitsubishi Heavy Industries with 176 kg/cm2 (2503 psi) and 541°C (1006°F) main steam conditions and
45.2 kg/cm2 (643 psi) and 541°C (1006°F) reheat steam conditions. The unit was placed into
commercial operation in 1982 and had accumulated approximately 100,000 operating hours and 2770
unit starts up to the time Creep-FatiguePro monitoring was initiated. Locations selected for monitoring
included all major plant components, namely: (a) Piping - the highest stressed girth welds in the main
steam and hot reheat piping, (b) Headers - the boiler superheat and reheat outlet headers and the
economizer inlet header, (c) Turbine - the HP turbine inner cylinder and the turbine chest throttle valve
and governor valve bodies.
Typical monitoring locations for header ID surface bore-hole ligament cracking are illustrated in Figure
5 for the secondary superheater outlet header. For the turbine, cracking at the at the HP turbine inner
cylinder flange near the bolt holes which was removed by grinding, prompted monitoring of this
location for recurrence of cracking (Figure 6). Cracking in the original steam chest throttle valve (TV)
and governor valve (GV) bodies prompted replacement of the entire steam chest, and monitoring at the
TV seat fillet radius (Figure 7) and the GV seat fillet and outlet transition radii (Figure 8) locations
known to be susceptible to crack initiation was implemented after steam chest replacement.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 7
 Loc.#1: HPInrCyl
(HP Turbine Inner
Casing Crack)

Loc.#1: SSOH-AX116
Loc.#3: SSOH-CIRC Loc.#2: SSOH-AX91

Figure 5. Superheater Outlet Header Ligament Figure 6. HP Inner Cylinder Flange Cracking
Cracking Monitoring Locations. Monitoring Location.

1 2 1 2

MX
MN

MX
MN
Y
X
Z
MX
MN MN

MX

TV Monitoring
Loc #1 Loc #1 Loc.#2
Y GV Monitoring
Z X

-10.329 6.237 22.802 39.368 55.934 2.841 23.137 43.433 63.728 84.024
-2.046 14.519 31.085 47.651 64.217 12.989 33.285 53.581 73.876 94.172
TV Pressure Evaluation - P=176 kg/cm**2 (2500 psi) GV Pressure Evaluation - P = 176 kg/cm**2 (2500 psi)

Figure 7. Throttle Valve Body ID Surface Cracking Figure 8. Governor Valve Body ID Surface Cracking
Monitoring Location. Monitoring Locations.

Results from the first year of monitoring have confirmed that the system can provide valuable input to
inspection planning, run/repair/replace decisions and changes to operating procedures to optimize the
remaining life of the monitored equipment. Separate monitoring reports are available for each of the
three major systems – piping, headers and turbine. An example summary report for monitoring
locations in the turbine is shown in Figure 9. The first part of the report (Figure 9a) contains a summary
of calculated creep-fatigue damage and crack growth for each of the seven turbine monitoring locations
– one at the HP inner cylinder flange, one each for the A and B TVs and two for each of the A and B
GV potential crack initiation sites. The second part of the report (Figure 9b) lists damage and crack
growth projections into the future based on the recent past (last 3 months) and last one year of operation.

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 8
 (a) Calculated Damage/Crack Growth (b) Projections into the Future
Figure 9. Creep-FatiguePro Summary Report for Critical Turbine Locations.

UTILITY BENEFITS
In the increasingly cost-competitive deregulated utility environment, accurate and timely information on
equipment condition is vital for input to repair/replacement outage planning/scheduling. By automating
the process of computing damage accumulation, crack growth and remaining life, and basing these
calculations on actual plant operating conditions, Creep-FatiguePro offers utilities many benefits:
• By taking periods of low-stress operation into account, Creep-FatiguePro provides more accurate
damage predictions than those based on maximum design or operating conditions, thus enabling
utilities to significantly reduce inspection frequency and extend the useful life of critical components.
• Advanced data trending and analysis tools allow operators to identify and adjust plant operating
procedures to minimize creep and fatigue damage at locations with higher/abnormal damage
accumulation rates, thereby avoiding unnecessary component degradation. The effect of revised
operating procedures can also be evaluated using off-line simulations.
• When a crack or fabrication defect is detected during inspection, Creep-FatiguePro monitor results
provide critical input for run/repair/replace decisions, possibly enabling utilities to avoid or delay
unnecessary repairs and outage extensions.
• Automated tracking of damage and crack growth in critical components eliminates the need for
tedious and less accurate manual calculations of remaining life.

SUMMARY AND FUTURE DEVELOPMENTS

Successful application of the Windows-based personal computer software program, Creep-FatiguePro,
to track creep and fatigue damage and crack growth in critical fossil plant components, has been
demonstrated via recent implementation and testing at a 250 MW fossil plant. Results from the first
year of monitoring indicate that the system can provide valuable input to inspection planning,

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 9
run/repair/replace decisions and changes to operating procedures to optimize the remaining life of the
monitored equipment.
Future developments will expand the current probabilistic evaluation capability to include the treatment
of additional variables which affect both crack initiation and propagation. Other software improvements
will be identified and prioritized based on end-user testing and feedback.
The Creep-FatiguePro software is currently available under license to EPRI.1

ACKNOWLEDGMENTS
The authors would like to thank participating EPRI member utilities who provided valuable feedback
from the initial stages of development through the release of the of the latest Windows version. Special
thanks are also due to Tim Gilman and Curt Carney of Structural Integrity for their efforts in upgrading
the software to Windows.

REFERENCES
(1). J.F. Copeland, C. E. Carney, P. C. Riccardella, T. P. Sherlock, "Creep-FatiguePro: On-line Creep
Fatigue Damage and Crack Growth Monitoring System", Structural Integrity Associates, Inc.,
EPRI Research Project 1893-11, Final Report No. TR-100907, July 1992.
(2). D. A. Rosario, M. L. Berasi, A. F. Deardorff, “Automated Creep-Fatigue Damage and Crack
Growth Monitoring in Power Plants,” Presented at the Joint ASME/IEEE Power Generation
Conference, October 17-22, 1993, Kansas City, Kansas.
(3). D. A. Rosario, P. C. Riccardella, L. H. Bisbee, M. G. Luttrell, “Remaining Life Assessment of a
Secondary Superheater Outlet Header,” 1995 ASME PVP Conference, July 13-27, 1995,
Honolulu, Hawaii.
(4). D. A. Rosario, M. L. Berasi, P.C. Riccardella, “Creep-Fatigue Damage and Crack Growth
Monitoring in Turbine Components using Creep-FatiguePro,” Presented at the Fourth EPRI
Turbine/Generator Workshop, August 1-14, 1995, Milwaukee, Wisconsin.
(5). “Creep-FatiguePro for Windows, Operating Procedures and User’s Manual,” Version 3.0, prepared
by D. A. Rosario, Structural Integrity Associates Report No. SIR-00-076, May 2000.
(6). R. Viswanathan, Damage Mechanisms and Life Assessment of High-Temperature Components,
ASM, Metals Park, Ohio, © 1989.
(7). F.V. Ammirato, et al, "Life Assessment Methodology for Turbogenerator Rotors", Volumes 1
through 4, EPRI CS/EL-5593-CCM, Projects 2481-3, 2785-1, March 1988.
(8). A. Saxena, J. Han, and K. Banerji, "Creep Crack Growth Behavior in Power Plant Boiler and
Steam Pipe Steels", Journal of Pressure Vessel Technology, Vol. 110, ASME, May 1988, pp. 137-
146.
(9). R. Viswanathan, and C.H. Wells, "Life Prediction of Turbine Generator Rotors", Technology for
the Nineties, ASME, 1993.

1
For additional information on Creep-FatiguePro, contact the EPRI program manager Rich Tilley (email: rtilley@epri.com)
or Darryl Rosario (email: drosario@structint.com).

EPRI International Conference on Advances in Power Plant Life Assessment, March 11-13, 2002, Orlando, Florida 10