See discussions, stats, and author profiles for this publication at: [Link]
net/publication/37617168
Ramifications Of Cracks At The Fillet Welds Of High Temperature Steam
Headers
Article · January 2004
Source: OAI
CITATION READS
1 1,884
3 authors, including:
Russell Coade
HRL Technology Pty Ltd
11 PUBLICATIONS 119 CITATIONS
SEE PROFILE
All content following this page was uploaded by Russell Coade on 22 May 2014.
The user has requested enhancement of the downloaded file.
SIF2004 Structural Integrity and Fracture. [Link]
Ramifications Of Cracks At The Fillet Welds Of High Temperature Steam
Headers
R W Coade1, J D Butler2 & D Zimmerman2
1
HRL Technology Pty Ltd., e-mail: rcoade@[Link]
2
ATCO Power Canada Ltd., Alberta, Canada
ABSTRACT: During a major statutory inspection of a 150MW steam raising boiler, cracks were found at
the stub tube-to-header fillet weld of both the secondary superheater outlet header and the reheater outlet
header. These cracks extended in some instances up to halfway through the tube wall thickness. The extent
of the cracking precluded immediate and widespread repair and so an engineering solution was devised to
identify the most critical cracks and repair these while leaving less critically cracked stubs in service.
While the unit was returned to service, the long-term life of the boiler was in question. The expected life of
these headers was compared with the required life of the unit and the economic issues of repairing stubs
compared to replacing with a new header and stubs, the result being a new for old replacement.
This paper will describe the extent of the cracking found, the engineering calculations undertaken to ensure
the safe return of these headers to service and the longer-term life issues.
1 INTRODUCTION
The Battle River Generating Station consists of 3 black-coal fired unitized boilers that feed three
steam turbines. Of these units Boiler 3 is the oldest, being commissioned in 1969 and having
operated up to December 2003 for a total of approximately 265,000 hours and undergone more than
600 starts. The station expects to operate Boiler 3 until around 2021, which equates to a life
expectation of almost 400,000 hours.
A statutory inspection of Boiler 3 in 2001 looked at emerging life issues and identified cracking
at the superheater and reheater header stub welds, at dissimilar metal welds (DMWs), and
deterioration of the superheater and reheater banks as issues expected to limit the future reliability
and life of this boiler. This paper describes only the actions taken to manage cracking at the
reheater and secondary superheater outlet header stub welds of Boiler 3.
2 SECONDARY SUPERHEATER AND REHEATER OUTLET HEADERS
2.1 Secondary superheater outlet header, SH4
The superheat circuit for Boiler 3 consists of three banks of primary superheater tubes in the rear
furnace pass and a secondary superheater in the front furnace pass immediately above the furnace.
Steam passes from the primary superheater via cross-over pipes into the secondary superheater inlet
header, travels through the secondary superheater tube banks to the secondary superheater outlet
header, SH4. The steam then passes to the steam turbine via a main steam line.
The secondary superheater outlet header, identified as SH4, has a design metal temperature of
557°C for a pressure of 12.8 MPa and is fabricated from SA 387 D which is a plate steel
specification indicating that this header is seam welded. The central-T is also manufactured from
SA 387 D material. The header is supported by a centrally located steam outlet branch and four
brackets connected to the four downcomers emanating from the steam drum. The SH4 header has
sixty six rows of tubes across the header with two tubes per row. The upper row is labeled row “A”
and the lower row “B”. The SA213 T22 SH4 stub tubes are 50.8mm OD x 8.6mm MWT and these
stub tubes join the SA213 T347H secondary superheater outlet tubes via a transition piece
approximately 457mm from the header.
SIF2004 Structural Integrity and Fracture. [Link]
The creep life of SH4, based on the operating temperature of 541°C (rather than the design
temperature of 557°C), a nominal wall thickness of 65.2 mm and creep-rupture properties reported
in the literature, ranges from a lower bound life of 162,000 hours to a mean life of >500,000 hours.
The header microstructure (Figure 1) shows bainite in a ferrite matrix with no signs of significant
thermal deterioration nor was there creep cavitation on grain boundaries or triple points.
Consequently it was considered reasonable to assume that the header material exhibited at least
mean creep-rupture properties.
Figure 1:Microstructure of the SH4 header parent steel in the vicinity of a seam weld.
2.2 Reheater outlet header, RH2
Exhaust steam from the high pressure turbine travels via the cold reheat line to the reheater inlet
header, through the reheater banks to the reheater outlet header, from where the superheated steam
travels via the hot reheat line to the intermediate pressure turbine.
The reheater outlet header, designated RH2, has a design metal temperature of 557°C for a
pressure of 4.1 MPa. The header is fabricated from SA 335 P22 material, a pipe specification
indicating that this header does not have a seam weld. The central-T is, however, manufactured
from SA 387 D material. The header is supported by a centrally located steam outlet branch and
four brackets connected to the four downcomers.
The RH2 header has sixty six rows of tubes across the header with three tubes per row. The
upper row is labeled row “A”, the middle Row “B” and the lower row “C”. The SA213 T22 stub
tubes are 54.0 mm OD x 4.6 mm MWT and these stub tubes join the SA213 T347H reheater outlet
tubes via a transition piece approximately 457 mm from the header.
The design metal temperature of 557°C is well within the creep range for 2¼Cr1Mo steel. The
creep life for this header was calculated to be > 500,000 hours and so quite adequate for the
maximum proposed life for this boiler.
SIF2004 Structural Integrity and Fracture. [Link]
3 ENGINEERING ASSESSMENT OF THE SUPERHEATER OUTLET HEADER
From a review of the inspection records it was found that in July 1993 all stub to header fillet welds
were intact but several tube weld toes were cracked between the 1:00 to 2:30 o’clock position and
other tubes were cracked between the 9:30 to11:00 o’clock position. All cracked tube welds were
repaired at that stage. The SH4 header had also experienced cracking at the support brackets which
has been managed successfully with remedial grinding to remove these defects.
The inspection in 2001 found some weld profiles to be quite poor as shown in Figure 2.
Inspection of the SH4 stub tube to header weld toes and stub tube weld toes found extensive
cracking at the tube weld toe, however, no cracking was found at the stub tube to header weld toes.
This cracking was removed by pencil grinding as shown in Figure 3. Figures 4 and 5 plot the crack
depths against tube number. Clearly the deepest and largest number of cracks were located at the
ends of the header. Figure 6 plots the total number of cracks against the position of these cracks
around the tube. Again a clear trend emerged with the majority of cracking at between the 9:00 to
3:00 positions, i.e. at the top of the tube. This was similar to that found in the survey undertaken in
July 1993.
The location of cracking suggests an imbalance between the vertical movement of the furnace
and the SH4 header during startup and shutdown. SH4 is supported through attachment to the
downcomers which will expand downwards from the steam drum. The movement of the stub tubes
where they penetrate the furnace wall is governed by the expansion of the furnace. It is therefore
suggested that this cracking may develop as a result of differential expansion between the
downward movement of the header and the movement of the furnace.
To prevent further failures of these stub tube welds this imbalance in movement must be addressed.
If this is not possible, the profile of the welds needs to be significantly improved to provide the
smallest stress concentrations possible.
Figure 2: SH4 header and stubs. Approximately 18 Figure 3: Example of the excavation at the stub weld
inches from the header the transition welds between the toe on the SH4 header. Where the remaining tube wall
2¼CrMo stubs and the 347H superheater tubes can be thickness was equal to or greater than the minimum
seen. design thickness, the excavation was left without the
need for weld repair.
SIF2004 Structural Integrity and Fracture. [Link]
0.5 Row A 0.5
Row B
12-3 o'clock
12-3 o'clock
Crack depth, inches
0.4 3-6 o'clock
3-6 o'clock
Crack depth, inches
0.4
6-9 o'clock 6-9 o'clock
9-12 o'clock 9-12 o'clock
0.3 0.3
0.2 0.2
0.1 0.1
0.0 0.0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Element number Element number
Figure 4: Row A. The cracks at the stub tube, tube to weld Figure 5: Row B. The depth of each crack at the SH4
toes were separated into the four clock locations given header stub tube weld toe was plotted against tube number.
above. The depth of each crack was then plotted against The crack distribution and severity in shown in this graph
tube number. The crack distribution and severity in shown indicating that the most cracking occurs at the ends of the
in this graph. header.
12:00 to 3:00 2:00 to 6:00 6:00 to 9:00 9:00 to 12:00
20
SH 4 R ow A
Crack-like Indications
SH 4 R ow B
Total Num ber of
15
Figure 6: The total number of crack-like indications at the
SH4 header stub tube to weld toes is plotted against crack
10
position. Clearly, most cracking occurs between the 9:00 to
3:00 o’clock position. i.e. at the top of the tube.
5
0
12:00 to 3:00 2:00 to 6:00 6:00 to 9:00 9:00 to 12:00
Position of C rack-like Indication
4 ENGINEERING ASSESSMENT OF THE REHEATER OUTLET HEADER
The history for the RH2 header showed that an inspection in July 1993 of 198 stub/header welds
found that 36 were cracked and these were weld repaired. In August 1995 186 DMWs were
repaired using orbital weld overlays, and the remaining 12 DMWs were replaced with new
transition pieces. The RH2 header has also experienced cracking at the support brackets and these
have been successfully managed through remedial grinding.
During the 2001 inspection, reheater header stub tube 1C and 2C (south end) were found to have
pin hole leaks that had eroded the header body.
Magnetic particle inspection found extensive cracking at the stub tube to header weld toes.
Examples of this cracking are shown in Figures 7 and 8.
The distribution of the RH2 stub tube to header cracking was collected from NDT reports and is
shown in Fig. 9. For Row A tubes the cracks were predominantly located between the 9:00 to 3:00
o’clock positions and in elements 1 to 40. Row B tubes had more cracked tubes than Row A and C
with the cracks again predominantly between the 9:00 to 3:00 o’clock positions. For the Row B
stub tubes the cracking was found largely between elements 10 to 55. The least amount of cracking
was found in Row C stub tubes with the cracking predominating in elements 10 to 29 and 39 to 60.
SIF2004 Structural Integrity and Fracture. [Link]
For Row C tubes, cracking was evenly distributed around the tubes. As for the SH4 header, the
location of cracking was predominantly at the top of the tubes.
Figure 7: Cracking at the toe of the stub tube to header Figure 8: Close up of the cracking at the toe of the stub
weld of reheater outlet header, RH2. to header weld of RH2.
Stub Tube Row A
Indication of Cracked Stub Tubes
7
Stub Tube Row B
6 Stub Tube Row C
Figure 9: The distribution of cracks in the stub tube to
5 header welds for Rows A, B and C are shown for each
element across the RH2 outlet header. Clearly, the
4
majority of cracking has occurred between elements 10 to
3
61. The majority of cracking for Row A is between
elements 1 and 45, for Row B, between elements 10 and
2 61 and for Row C, between elements 10 and 29 and 39
and 60.
1
0
10 20 30 40 50 60
Elements across RH2
5 DISCUSSION
5.1 Stub Tube Weld Cracking
There was extensive and widespread cracking found at the toe of the stub tube welds on the SH4
and RH2 outlet headers. Once any cracking was identified it was removed by pencil grinding. Dye
penetrant inspection was used to determine when the cracks had been removed and once the cracks
were removed the crack depths were recorded. As the depth of the cracks varied considerably and
time and resources were scarce it was necessary to determine those tubes that must be repaired and
those that could be safely left. It was then necessary to determine the critical excavation depth,
below which weld repair was required.
During this inspection and the development of an engineering solution to manage this cracking,
the root cause of cracking was not identified. However, the location of cracking suggested that
there was an imbalance in the vertical movement of the stub tubes between the furnace wall
penetrations and the SH4 header or restraint to the longitudinal expansion of the tubing. Either
could induce considerable strain on the stub tubes when the header was heating during startup.
SIF2004 Structural Integrity and Fracture. [Link]
5.2 Minimum Thickness Calculations for the SH4 and RH2 Stub Tubes
To satisfy the statutory authority, the excavations to remove the cracking at stub tube weld toes
must leave a remaining wall thickness equal to or greater than the ASME Section I minimum design
thickness. This thickness is calculated using Equation 1, and any excavations greater than this
depth were to be weld repaired.
PD
t= + 0.005 D + e
2S + P ………………………. Equation 1
The minimum tube thickness based on ASME Section I was calculated from Eqn. 1 for the design
conditions listed below:
Tube material SH4 Header RH2 Header
SA213 T22 SA213 T22
P= Max. allowable working pressure 13.0 4.1 MPa
T= Design Temperature 541 541 °C
D= Tube Outer Diamater 50.8 54 mm
S= Max. allowable stress @ design Temp. 53.8 53.8 MPa
e= Thickness factor for expanded tubes 0 0
t= Min. design thickness 5.7 2.25 mm
Thus, provided the stub tube wall thickness remaining after the removal of the crack was greater
than the ASME minimum thickness, weld repair would not be necessary. Hence, an excavation
depth of 2.9 mm for the SH4 header or 2.3 mm for the RH2 header would be acceptable.
If, however, the excavation was deeper than these limits, the defect would require weld repaired.
This is shown in Fig. 10 below.
8.50 mm
5.7 mm
4.57 mm
Nominal tube
thickness = 8.6mm
5.00 mm
2.25 mm
Chase crack to
a max. depth
of 2.8mm Crack at weld toe
R 5.00 mm
65.1 mm
34.9 mm
Example of the crack at the weld toe. Recommended repair profile, SH4. Recommended repair profile, RH2.
Figure 10: Acceptable maximum excavation depth.
5.3 Creep Life of the SH4 and RH2 Stub Tubes including an Excavation
This repair procedure has provided a safe means to return the SH4 and RH2 headers to service,
however, those cracks ground out to the maximum allowable depth now have a stress concentration
built into the weld that could decrease the creep life of these welds.
SIF2004 Structural Integrity and Fracture. [Link]
The creep-rupture life of the stub tubes was calculated assuming removal of all evidence of
cracking through excavation to the minimum design thickness. The stress in the stub tube before
any effective wall loss through cracking and excavation was 32.35 MPa and 22.55 MPa for SH4
and RH2 stub tubes respectively, and the maximum stress level after excavation was 51.4 MPa and
47.2 MPa for SH4 and RH2 stub tubes respectively. However, the excavation has introduced a
stress concentration to the stub tube weld toe that is assumed to be 1.5.
The last partial inspection of the SH4 and RH2 stub fillet welds was in 1998 and no cracking was
detected. Hence it can be expected that the stub tube fillet welds had operated under operating
conditions up to this period without cracking. Hence the life fraction consumed can be assumed to
be the life fraction exhausted in the 200,000 hours to 1998 plus the life fraction exhausted once the
cracking had developed, and this cracking is assumed (conservatively) to have existed from 1998 to
2001. The total possible creep lives for each period are shown in Table 2 including the remaining
life.
Table 2: Creep-rupture life of the RH2 and SH4 headers.
SH4 Header Creep Life RH2 Header Creep Life
β
To 1998 1998-2001 Post 2001 Remaining To 1998 1998-2001 Post 2001β Remaining
Min >500,000 h 440,000 h 20,000 -2 years >500,000 h >500,000 h 60,000 7 years
Mean >500,000 h >500,000 h 117,000 13 years >500,000 h >500,000 h 315,000 36 years
β - life assuming 541°C, stress at excavation x 1.5SCF & operating pressure
Thus the SH4 header stub tubes that have been excavated to the maximum depth and not
repaired have a remaining life of between 2 to 13 years, and will need to be repaired or replaced to
achieve the 2021 planned life.
The RH2 header stub tubes have a remaining creep life of between 7 to 36 years and will
probably need to be replaced or repaired to operate satisfactorily until 2021.
5.4 DMWs between Outlet Header Stubs and Superheater and Reheater Tubing
The DMWs between the stub tubes and the secondary superheater tube bank and reheater tube
bank should have significant remaining life as the metal temperatures under which these joints
operate are relatively low. However, these stub tubes are clearly experiencing significant external
strains due to some restraint of movement during startup and shutdown and so the life of these
joints must also be in question. In fact the wholesale replacement of the RH2 DMWs in 1996
suggests that the system stresses on these joints are significant.
6 CONCLUSIONS
This investigation has shown that the superheater outlet header and reheater outlet header have
significant issues with cracking at the header stub to tube weld toes and with creep life limitations
for tubes that had cracks removed but that were not weld repaired. These issues would seriously
limit the life of the headers unless major repairs are implemented.
The reheater outlet header also had significant issues with cracking at the header stub tube
DMWs and until the root cause of the cracking is determined and corrected this will continue to
threaten the integrity of the headers.
In the cost benefit analysis undertaken by the station, the cost of repair of the stubs in situ was
compared to the cost of new headers and stubs. The superheater was to be replaced which provided
SIF2004 Structural Integrity and Fracture. [Link]
a window for major works. Furthermore, the access for repair was tight and fitup with the new
superheater was also an issue. Thus considering the difficulty of repair and the ensuing costs,
replacement of the each header with a new header and factory fitted stubs was found to be the most
economic.
ACKNOWLEDGEMENTS
The authors would like to acknowledge ATCO Power Canada Limited for allowing this paper to be
published and would like to acknowledge Canspec Inc for the replicas reproduced in this report.
Thanks are extended to Dick Coldham of HRL Technology P/L for his assistance in reviewing this
paper.
View publication stats