ML19254B447
| ML19254B447 | |
| Person / Time | |
|---|---|
| Site: | Point Beach  |
| Issue date: | 08/28/1979 |
| From: | Fay C WISCONSIN ELECTRIC POWER CO. |
| To: | James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| NUDOCS 7909270826 | |
| Download: ML19254B447 (48) | |
Text
m g) 1 3
WISCONSIN Electnc eaara coursur 231 W. MICHIGAN, P.O. BOX 2046. MILWAUKEE, WI 53201 August 28, 1979 Mr. J. G. Keppler, Director Jffice of Inspection and Enforcement
~ Region III U. S. NUCLEAR REGULATORY COMMISSION 799 Roosevelt Road Glen Ellyn, Illinois 60137
Dear Mr. Keppler:
DOCKET NOS. 50-266 AND 50-301 FURTHER RESPONSE TO IE BULLETIN 79-13 POINT BEACH NUCLEAR PLANT, UNITS 1 AND 2 On June 25, 1979, the Office of Inspection and Enforcement issued IE Bulletin 79-13 which required PWR licensees to inspect at least a portion of steam generator feedwater piping systems and to report the results of these inspections to NRC. Our submittal of July 12, 1979 fulfilled the bulletin requirements for Point Beach Nuclear Plant Unit 2 and presented our initial plan for the Unit 1 inspectionL The final metallurgical evaluation report for the Unit 2 feedwater pipe reducer is provided in the enclosure to this letter.
In addition, during our review of this subject, we have revised the listing of transients and the piping stress isometrics from that initially provided with our July 12 sub-mittal. This revised information is also contained in the enclosure.
The metallurgical evaluation of the Unit 2 steam generktor reducer identifies the deepest crack as less than 0.047 inches into the pipe wall.
The cracks were circumferentially oriented and no microstructural dependence was chserved. These minor cracks have apparently been present for some cor.siderable length of time based upon the chemistry analysis of the deposited oxides. The report concludes that the most probable cause of this cracking is corrosion fatigue. As discussed in the enclosure, and in our previous submittal, we consider that the etistence of these cracks does not present an unsafe condition for continued plant operation.
During the Unit I refueling scheduled to begin September 28,.979, we intend to replace tne Unit 1 main feedwater piping from the steam gens rator nozzle back to the main check valve. Also, we intend to inspect, and repair as necessary, the auxiliary feedwater piping to a minimum of three feet upstream of the connection to the main feedwater piping.
9 g09 N 1044 235
Mr. J. G. Keppler - Page Two August 28, 1979 The design of the Point Beach steam generators utilizes a common feedwater inlet nozzle for the injection of both normal and auxiliary feed-As indicated on the stress isometrics, the auxiliary feedwater con-water.
nection to the main feedwater piping is close to the steam generator nozzle and is downstream of the main feedwater check valve.
Up to this point, water A
supply to the steam generators is provided by separate piping systems.
thermal evaluation of the auxiliary feedwater piping, which is contained in the enclosure, has conservatively shown that only three feet of auxiliary feedwater pipe closest to the main feedwater pipe experiences a.ignificant themal transient. Beyond this distance, there is no reasonable themal fatHue mechanism present. As you are aware, the preliminary conclu lon concerning the development of these cracks is that it is corrosion assisted fatigue and that the cyclic stress caused by thermal transients may be the Therefore, by assuring the integrity of the main feedwater prime cause.
piping downstream of the check valve and the auxiliary feedwater piping that is subject to thermal transients, the water supply to the steam generators Our repair program responds to the bulletin in assuring the inte-is assured.
Accordingly, the remaining grity of feedwater supply to the steam generators.
welds inside containment in the main feedwater system will not be inspected as the existence of small cracks in these welds is of no significance.
The Unit 1 main feedwater piping that will be replaced, will also We consider be removed from inside containment and inspected at a later time.
it highly probable that minor indications, such as those found on Unit 2, will be found during the Unit 1 inspections. We will provide a report of the results of the inspection after completion of the refueling outage.
Should you have questions, we would be pleased to discuss this program further.
Very truly yours, U
c [ L.'
C. W. Fay, Director Nuclear Power Department Enclosure Office of Inspection and Enforcement Copy to:
Division of Reactor Operations Inspection Mr. A. Schwencer, Chief Office of Nuclear Reactor Regulation e
1044 236
ENCLOSURE RESP 0f1SE TO IE BULLETIl4 79-13 AUGUST 28, 1979 This enclosure discusses and provides the following items for the Point Beach Nuclear Plant with respect to the subject bulletin:
a.
Revised main feedwater pipir.g stress isometrics for both nuclear units; these supersede the figures contained in Wisconsin Electric's July 12 submittal to NRC.
b.
..evised tables of Significant Operating Events for Feed-water Piping System Transients; these supersede the tables contained in Wisconsin Electric's July 12 submittal to NRC.
c.
The results of a thermal transient analysis of the Point Beach fluclear Plant auxiliary feedwater piping by Bechtel Power Corporation d.
A copy of Appendix B of the Southwest Research Institute Report, " Point Beach Nuclear Plant, Unit 2 Feedwater and Auxiliary Feedwater System Examination, Analysis, and Repair"; SwRI Project 17-4048, August 1979.
A.
PIPI!1G STRESS IS0 METRICS Figures 1 thru 4 attached present a summary of stresses, based upon the original analyses, of the four Point Beach Nuclear Plant main feedwater pipes.
The stresses are presented for piping beyond that inspected on Unit ' or to be inspecteu on Unit 1.
The Unit 1 program will encompass welds analogous to the welds inspected on Unit 2 (which are denoted on the Unit 2 figures).
i..a original piping analysis was performed te satisfy the B31.1 design requirements.
Because the piping gravity supports were designed indepen&ntly of the seismic supports, a conservatively assumed gravity stress of 3000 psi was used in the stress combinations.
To check the validity of this assumption, a gravity stress analysis was recently performed for the Unit 1 "B" piping loop. The highest calculated gravity stress was 647 psi at the data point representing the piping connection to the cor.tainment penetration (not shown on the attached figure). A comparison of the other calculated gravity stresses for the Unit 1 "B" loop to the assumed gravity stress utilized on the other three piping loops shows a substantial degree of conservatism in this assumption. Also, as discussed in our July 12 submittal, the original analysis assumed all 16-inch diameter piping; the larger 18 inch diameter
. reducer-to-nozzle joint section modulus was not considered. Thus, the stresse; for data point 5 on the attached figures are again conservative.
1044 237 1-
The stress results from the computer analysis of the entire main feed-water piping inside containment have been reviewed for all four pipelines.
A comparison of the total SSE stress results (which includes gravity and longitudinal pressure stresses) to an allowable stress of 27,000 psi (1.8 Sh, which is lower than the material yield strength) indicates the following:
1.
All total SSE stresses are less than 80% of allowable.
2 About 65% of the data points in all four loops have total SSE stresses less than 40% of allouable.
B.
TABLES OF SIGNIFICA!1T OPERATING EVENTS The attached Tables 1 and 2 replace Tables 2 and 3 of our July 12 submittal to NRC. These tables have been revised to further identify and clarify significant thermal transients experienced by the main feedwater piping.
The normal feedwater temperatures (No. 5 feedwater heater outlet) are as follows (100% power = 497 MWe net):
Power Level, %
Approximate Feedwater Temp., F 100 435 95 427 78 405 52 371 26 323 0
70 to 100 The zero percent power level represents a hot shutdown condition.
In this situation, feedwater may be supplied from different sources. The source of feedwater is selected by the plant operators dependent upon why the plant is in the hot shutdown condition.
The attached tables do not include events occurring at, or close to, the 0% power level since these events are not considered to cause significant thermal transients in the feedwater system.
C.
AUXILIARY FEEDWATER PIPING THERMAL EVALUATION The attached thermal evaluation of the auxiliary feedwater piping was performed by Bechtel Pcwer Corporation and presents the temperature profile during normal ;,teady-state plant conditions. The analysis conservatively shows that the temperature in the auxili Aty feedwater piping wall decreases to 150 F within a distance of 3 ft; the 3 feet is rounded off from the 1 foot mixing zone plus 1.5 feet from analysis.
This delta T (270 F) magnitude would decrease if the steady-state main feedwater temperature decreased.
" Assuming that the minimum auxiliary feedwater temperature is 32 F
{ service water is a potential source of supply), the maximum temperature differential in the piping upstream of this 3-foot point, in practical terms, would be less than ll5F.
In Timoshenko, " Strength of Materials",
Part II, Third Edition, Section 27 discusses thermal stresses in cylindri-cal shells.
Utilizing reasonable carbon steel properties, a temperature U44 238 2
difference of 115 F would be calculated to produce a maximum' bending stress (thermal) of about 15,000 psi.
Thus, it is concluded that tiie only portion of the auxiliary feedwater piping that is exposed to a significant thermal transient is the approximately 3 feet of pipe closest to the main feedwater pipe. During the Unit i refueling outage, at least this section of the auxiliary feedwater pipe will be inspected and/or replaced as deemed appropriate.
D.
FINAL METALLURGICAL EVALUATION REPORT
-The laboratory metallurgical evaluation of the Unit 2 "A" reducer was performed by the Southwest Research Institute (SwRI) of San Antonio, Texas. The final report by SwRI containing the inspection results, procedures used, procedure qualification records, certifications, etc.,
is available at the Point Beach Nuclear Plant.
Appendix B (33 pages) attached hereto entitled, " Metallurgical Evaluation of Feedwater Pipe Cracking" is the pertinent section of th.is SwRI report.
This appendix supersedes the preliminary information contained in our July 12 submittal to NRC.
Conclusions from the report are:
- the most probable cause of cracking is corrosion fatigue.
- the deepest penetration found was less than 0.047 inches
- the cracks appear to have been present for some time based upon chemistry analysis of the oxides.
Crack tips are invariably blunt and oxide filled.
- the cracks were invariably circumferentially oriented (absence in the axial direction) and no microstructural dependence was observed.
- the toughness of the reducer material at temperatures above 100 F and the ductility of carbon steel renders a catastrophic failure improbable.
These conclusions are basically the same as those presented in the preliminary report.
m A
S 1044 239
SHEET.
I cALCULAilON SHEET FILE No Bil A.d WOCLE A R 9 L A 14 T m A o f a v _IL'A..
.3.
ontE %4', :i sus >Eci
, " A
11Aih F E E B o A T E tt um.+
t cHxo. av _
_ o AiE
...-..7...---
\\
+.,,,
u.. ~,s s. e p <t g
A 7M 8
d li O,o..
R, A v c-t -
N, a' e. I d g
N
,o 5" D in. A ux.
(?
FecAweik< r h!
bd f u u t s ul,I f ;C,
g l
STRESS
SUMMARY
Data Thermal Gravity Longitud-Total Seismic REMARKS Point
3 t>21 3000 3608 12532 18456 Nozzle o
3/2d*
3000 3608 9546* 12484*
Aux. F-W Connec.
10 570o 3000 3603 13331 20054 E1 bow 12 455 3000 3603 8546 10484 End Bend
~
15
'6043 3000 3608 8289 9970 Begin Bend
- Stress intensification factor not included.
I F I G v R E.
1 1044 240
C ALCULATION $ HEE (
FILE N o SUBJECT
.. _b b E.-- U LE A P
Y M A D E S Y ___
._ D AT E t' t. t.
f_f D 'u /. t (I R -
U n.+
t,# B" CHKD.SY D A T E __ _ _ _ _. ___... _.
I 2f S e.s m.c.
s S u p p o pt
^
7 d
L gww o uuuuu w a s s
S,-
or io 3, o.s. A ux.
b 14, " Q. w. R sJ. o c.c r.
LO = L d.
5 STRESS SUiHARY Data Thernal Gravity Longitud-Total Seismic REMARKS Point
psi 5
4736 104 3608 5059 6406 Nozzle 10 4163*
65*
3608 8138*
12603*
Aux. F.W.
Connec.
Ib 6004 90 3608 5719 7740 Elbow 20 5934 183 3608 4159 4527 End Bend 25 6993 104 3608 4031 4350 Begin Bend
- Stress intensificat.1 factor not included.
I FIG U RE 2.
1044 241
CALCULATION SHEET FILE No D'
_ D ATE SU8 JECT f t.i__3 E A C H MUCLEAR P L A M T~
MADE BY vsa 1,
A" Cuxo.sy
_ o ATE
..._ _.._ Mhl M rasbeATr R l
or 3 0.o.. Aux.
grsas
.c G E N.
Fec<f w atee W
J sb Na%,
w S
i5 4
h
^W W
o c, o. a..
R e L o c. < c W =. M O
h N ' h 5 T ""
lo 12 M'IMIM T)
R:
y ijjt i p 'l nr rar;len-I: E Hj t,'-
.i) U d c' b it, _. __a STRESS
SUMMARY
Data Thermal Gravity Longitud-Total Seismic REMARKS Point
- Stress, Stress, inal Press-Stress, psi psi ps1 ure stress, OBl[.
SSE psi 1730 3000 3608 9692 12776 hozzle o
3 3114*
3000 3608 7559*
8510*
Aux. F.W.
Connec.
10 3114 3000 3608 11143 15678 Elbow 12 2129 3000 3608 7566 8524 End Bend 15 126 3000 3608 7379 8150 Begin Bend
- Stress intensification factor not included.
I 1
i l
FIG U R E 3
1044 242
CALCULATION SHEET FILE No SU BJ E CT _.[116 t P F A (. H MuC LE A R PLANT MADE BY Qw; D ATE bi,I/
Un t 1, " fA "
CHKD.BY D ATE.
M A 1 ti_f E EO W 7. '~ E R r_=
O
- s... -. c f
Support i
A us - In.s p c.hl h
w e.\\ d s
./^
u 6
\\*
e 3 D %. A ux.
M A w a.t=.-
STEAM [
W GEN.
g g g" Dia. R.Aveer WetA k
S S)
,[
- P
- ' ; ' ! [l i i
fD
}; I' T d L.L. C es. L Ju._ J _a STRESS
SUMMARY
Data T hennal Gravity Longitud-Total Seismic REMARKS Point
461) 3000 3605 8066 9524 Nozzle 8
5127*
3000 3608 7893*
9358*
Aux F.W.
Connec.
10 11331 300')
3608 8636 10664 Elbow 15 971 3000 3608 6976 7344 End Bend 20 7433 3000 3608 7059 7510 Begin Bend
- Stress intensification factor not included.
I' FIG U R E H 1 0 4 4 2 4 3
TABLE 1 LIST OF SIGNIFICANT OPERATING EVEl4TS FOR FLEDWATLR PIPING SYSTEM TRANSIENTS _
AUGUST 24. 1979 POINT BEACH NUCLEAR PLANT UNIT 1 DATE EVENT 7/10 Hot functional test completed, 11/2/70 Initial criticality.
11/22/70 Turbine trip from 40% power.
11/29/70 Power escalation to 70% and unit trip.
12/4/70 Turbine and reactor trip from 425 MWe.
12/6/70 Manual unit trip from 70% power.
12/18/70 Unit trip from 80% power.
12/19/70 Unit trip from 35% power (loss of feedwater).
1/4/71 Auxiliary feedwater injection during plant startup.
1/8/71 Unit trip from 82% power.
1/9/71 Turbine trip from 70% power.
1/27/71 Unit trip from 90% power.
1/28/71 Turbine trip from 92% power.
1/29/71 Unit trip from 72% pcwer.
2/3/71 Turbine trip from 90% power.
2/4/71 Unit trip from 90% power plus a turbine trip.
2/9/71 Unit trip from 80% power.
2/26/71 Load runback from 450 MWe to about 200 MWe.
7/2/71 Reactor and turbine trip.
8/29/71 Reactor and turbine trip.
9/7/71 Load runback from 425 MWe to 260 We.
9/18/71 Load runback from 480 Ne to 390 MWe.
12/3/71 Reactor and turbine trip.
1/3/72 Reactor and turbine trip.
1/19/72 100% load rejection test (reactor and turbine trip).
2/13/72 Load runback of 20%.
4/13/72 Load runback of 20%.
4/21/72 Reactor and turbine trip.
7/3/72 20% step increase in power.
9/11/72 Turbine and reactor trip from 99% power.
5/18/73 Unit trip from 72% power.
7/2/73 Reactor and turbine trip.
8/11/73 Reactor and turbine trip.
1/11/74 Reactor and turbine trip.
1/18/74 Reactor and turbine trip.
2M/74 Reactor and turbine trip, preceded by a turbine runback.
8/2/74 Reactor and turbine trip from 430 MWe.
9/25/74 Reactor trip from 99% power.
10/4/74 Reactor and turbine trip.
Paga 1 cf 2 jQ44
}ff
T ABLL 1 Poh ' 0h
- Cusi u AUGUST 24, 1979 vArt EVEl4T 2/27//9 Emergency shutdown with reactor and turbine trip, 11/16/75 Emergency shutdown with reactor and turbine trip.
1/10/76 Load runback of 20'u.
1/14/76 Reactor trip.
11/30/76 Load runback from 90k to 70'. power.
2/21/77 Reactor and turbine trip.
4/4/77 Reactor and turbine trip.
1/f/78 Reactor and turbine trip from 99" power.
2/9/78 Reactor and turbine trip from 99% power.
4/2/78 Reactor and turbine trip from 99% power.
Page 2 of 2 l
4
TABLE 2 i
LIST OF SIGNIFICANT OPERATING EVENTS FOR FEE 0 WATER PIPING SJ5 TEM TRAN5IENTS AUGUST 24. 1979 POINT BEACH NUCLEAR PLANT UNIT 2 DATE EVENT 12/22/71 First unit heatup, normal.
5/37/72 Initial criticality, power restricted to 1%.
7/28/72 20% power authorized.
8/3/72 Unit reaches 20% power.
8/4/72 Turbine trip from 20% power.
8/18/72 Unit trip from 10% power.
8/30/72 Load swin9 test 20% to 14% to 20%.
8/31/72 Unit trip from 20% power.
12/7/72 Turbine trip from 20% power plus a turbine and reactor trip from 10% power.
1/14/73 Reactor and turbine trip from 20% power.
2/18/73 Reactor and turbine trip from 20% power.
3/8/73 Received 100% power if cense.
3/9/73 Reactor and turbine trip, from 50% power.
3/10/73 Reactor and turbine trip.
3/14/73 Reactor and turbine trips - 2; one from 390 l4te.
3/24/73 Reactor and turbine trip from 92% power.
3/26/73 Reactor and turbine trip.
3/30/73 Reactor and turbine trip from 70% power.
4/8/73 Reactor and turbine trip.
5/30/73 Reactor and turbine trip.
6/19/73 Reactor and turbine trip.
10/13/73 20% load runback.
12/15/73 Reactor and turbine trip from 100% power.
12/27/74 Reactor and turbine trip from 430 lide.
2/11/75 Reactor and turbine trip.
2/24/75 Reactor trip from 10% power.
8/19/75 Reactor trip from about 10% power.
1/14/76 Manual unit trip.
2/21/76 Load runback of 20%.
4/8/76 Unit trip.
5/7/76 Load runback of 20%.
6/13/76 Load runback from 100% to 80% power.
9/3/76 Reactor trip.
1/b/77 Reactor and turbine trip.
6/28/77 Reactor and turbine trip.
7/7/77 Reactor and turbine trip.
1/10/78 Reactor and turbine trip from 60% power.
12/9/78 Reactor and turbine trip from 99% power.
4/11//9 Manual unit trip free ?2% power.
Page 1 af 1 l
Bechtel Power Corporation Engineers-Constructors Fifty Beate Strcet Project File No. 10h47-004 San Francisco, CaMornia 079-48 u.a4aar.s. e o Bo 3965. san rea.cisco. cA 94119 August 17, 1979 Mr. D. K. Porter Superintendent - Nuclear Projects Office Wtsconsin Electric Power Company 23 West Michigan Milwaukee, Wisconsin 53201 Attention: Mr. D. Dill
Subject:
Bechtel Job No. 10447 Point Beach Nuclear Plant Auxiliary Feedwater Line Temperature Distribution
Dear Mr. Dill:
On August 13, 1979, you requested that we provide a steady state temperature distribution in the 3 inch auxiliary feedwater line.
We prepared a calculation for this request with the following assumptions and results:
0 Main Feedwater Line (16 inch) temperature 420 F.
Auxiliary Feedwater (3 inch) line condition Stagnant (no flow)
The calculation was based on a cooling fin model assuming that the outside heat transfer coefficient is constant over the length of the pipe. We also calculated a single thermal conductivity that represents the pipe, water and insulation.
i.
-The insulation of the 3 inch auxiliary feedwater piping consists of 1 inch fiber glass. This is based on the records we have.
If thicker or Mifferent insulation is installed then these results are not conservative. The 3 inch pipe is also considered to be in the horizontal plane which is conservative.
For conservatism we assumed that the feedwater is mixed and at 420 for one foot into the 3 inch auxiliary feedwater line.
0 F.
The maximum length to where the 3 inch pipe wsil temperature reaches 150 under the above conditions in three (3) feet. This temperature is essentially ambient conditions in the containment near the piping. Attached are the temperature distribution curves for two auxiliary feedwater lines.
W trust this satisfies your request. If there are any questions, please call me.
Verr truly yours f '-
D. II. Clark k
Project Engineer CBil:ba Enclosure 1044 247
<s c
W
%o h '4 6 i
Og D%
34 t.r @u n@
T 5n F61 n
wy s
A o m ila,o; j o nr
%u ~ R
$m av E-uL% e g
~
W P,
~
m x=
N no g..
g It LD W
N a
_ O.
i 4
Z h.
m m
lt.
v til "X
I 11 O
n-B n
n
/
41 s-
%s lij cn t-u~
o U.
e e4
-,e cs
/
/
o o
U co crs, cocSe co coe co ovc cocoT coocT C.-D d H 3.L
_ 2 r-1044 248
n h
D9 n
p' ?m agf 449 jp a
w z
C;'
- r-r ll ~
lll?
1: 1.i n%o 1
CI t) c~
U*N i
Te K
gg o
0 g
o ouF gg v
W p'
O 11.
H 8m m
4-u) r O..
Vt i
2H y
oO
_0 E mE (1.
w O
s.
UZ
.J n<
H
_uF gg H
til O
Q 8
W s
d n
^
U.
{
L-U
~
1 co oct-co oce co*oce co o,c cocar coecr 8
Cd> BBniv63dW31 I
e 1044-249
e O
d 8
APPENDIX B METALLURGICAL EVALUATION OF FEEDWATER PIPE CRACKING Ol O
4 A
O 1044 250
METALLURCICAL EVALUATION OF TEEDWATER PIPE CRACKING I!URODUCTION on July 5, Southwest Research Institute (SwRI) was supplied one 18-in. X 16-in. reducer removed from the "A" steam generator of Point Beach Nuclear Plant, Unit 2.
The reducer was removed by cutting through
^the 18-in. reducer-to-nozzle and the 16-in. pipe-to-reducer welds.
Cutting was performed without the use of lubricants. The cut passed Ac through the weld crown at the 0.D. at both the 18-in. and 16-in. ends.
the I.D. of the 18-in. end, the cut passed through the root-pass weld bead At the I.D. of the or in the heat-affected zone (HAZ) of the reducer.
16-in. end, the cut pas ed through the weld bead or in the HAZ of the pipe.
On July 18, an additional sample was also submitted for analysis.
This sample was removed from a feedwater pipe of the "B" steam generator The sample contained a crack and was recoved as a con-of the same unit.
sequence of radiagraphic and ultrasonic inspection of repair welds to the feedwater piping system.
Radiographs performed at Point Beach by Superior Industrial X-ray Corporation and interpreted by Mr. S. Wenk of SwRI showed a crack-like indication from station marker #10 to #16 (63* - 100*) near the 18-in. reducu Two transverse indications 1/4 in long were also detected to nozzle weld.
at stations #42 and #43 running from the weld crown into the reducer base metal. Three linear indications were also detected at the 16-in. pipe-to-reducer veld on the pipe side.
Before removing any material from the reducer, ultrasonic inspection was performed to more accurately position the defect causing the RT indi-This examination confirmed the existence of a cations at the 18-in. cnd.
flaw which was clearly p esent from stations #10 to #21 (62* - 130*) and
- 23 to #25 (144' - 157*). The flaw was positioned approximately 3/8 in.
from the end of the reducer, that is, in the vicinity of the transition from the counter-bore to full reducer I.D.
Only the section from stations After location of the flaw giving
[ #8 to #26 was examined using ultrasonics.
rise to the RT indications, a ring 1-1/2 in. to 2 in, wide was cut from the O
B-1 1044 251
No lubricant was employed during this or any 18-in. end of the reducer.
subsequent cutting operation. The ring was then cut in half, through stations #0 and #28, 0* and 180*.
Each half was examined visually using A similar ring a stercoscopic nicroscope at r.agnifications up to 50X.
was also removed from the 16-in, end of the reducer.
No crack could be unambiguously identified at the countersbore tran-In the zonc from station #10 to #20 the oxide was unusually rough sition.
the Over most of the circumference the oxide at and porous in appearance.
counter-bore transition was not noticeably different from that in other but were not regions. Randomly distributed pit-like areas were present, Cracks could be positively iden-grouped linearly as at stations #10-#20.
tified at the root pass fusion line at stations v37 - #37.5, #42 - #45, and #4$.$ - #49 (*230* and 260* - 319*), Figure 1.
No evidence of cracks corresponding to the transverse RT indications could be seen at station
- 42. However, two stamped marks on the 0.D., adjacent to the weld crown, could be responsible for these indications.
The oxide was cicctrolytically removed from sc1cceed segments to examine the underlying surface. On the 18-in. end copious cracking was the inter-observed at and near the counter-bore transition, that is, at acetion of the flat counter-bore and bevc11ed counter-bore regions, Figure 2.
Most cracks were situated at the root of machining grooves, at No cracks in an the root-pass fusion line or at the base of deep pits.
axial orientation could be found in the base cetal, although several small axial cracks were present on the veld bead, Fit
'(b).
As will be seen below, no cracks were observed at the HA2 in r
- rgical sections.
the veld bead had Hocroscopic examination showed that grinding t
Some small virtually removed the machining marks in this arca, Figure 3.
tracks were present in the HAZ, Figure 3(b), at pits or small grooves.
Thus, the general absence of cracks in the HAZ is attributed to the elici-nation of stress-raisers, rather than to metallurgical structure or residual stress effects.
A C
B-2 1044 252
O Examination of the 16-in. end by magnetic particle inspection did not reveal any definite crack-like indications.
Two vcak indications were obtained at the fusion line of the root pass at stations #14 and
- 41 (100* and 295*).
Before oxide removal no cracks could be seen.
After partial removal of the oxide by cathodic polarization in Endex, a definite
$ copper-colored line could be seen at the counter-bore transition, Figure 4.
Similar copper-colored deposits could also be seen at pits in the weld bead and reducer section. Energy dispersive X-ray analysis positively identiffed the deposits as elemental copper, figure 4(b). After thorough cleaning, small cracks could be seen at tha veld fusion line and at the counter-bore cid. Cracks transition. Cracking was much less profuse than at the 18-in, were rarely present in machine grooves other than that which constituted the transition.
A second distinct groove was also present at the 16-in. end about 2-7/32 in. from the weld fusion line.
This groove appears to have been created by a final machining pass restricted to this end of the reducer.
A few minute cracks could be found in this area by careful visual inspection O
at 50X magnification after complete c1 caning. No other cracks could be found in other parts cf this section.
Several areas of the reducer I.D. temote to the counter-bore were Particular carefully inspected for either axial or circumferential cracks.
attention was given to areas at the 18-in end where the surface contained gouges in the axial direction. No evidence of axial cracks could be found.
One extremely small circumferentio.
rack, less than 0.001 in. long, was found at the base of a large pit at the 18-in, end. No other cracks were observed.
METALLOGRAPHIC ANALYSIS
The structure of the reducer and welds was normal for mild steel piping, Figure 5.
No metallurgical abnormalitics such as islands of martensite, heavy banding or gross inclusions were observed. Sections at stations #11, #14, and #41 (80*, 100* and 295*) of the 16-in. end revealed
^
a lack of fusion defect between the first ar.d second passes, Figure 6.
O 1044 253 B-3
This in-In some sections this defect also contained oxide inclusions.
insufficient shiciding of the first pass and insufficient dicates that input during the second pass were employed during deposit of the heat Despite the size of these defects, there original reducer-to-pipe weld.
v3s so~ evidence that flaw growth had occurred, dracts vete found in eli longitudinal sections taken from the s
The depth of the 66unter-bore region of the 18-in, and of the reducer.
idigest crack and the number of cracks in each section are given in in ali dases multiple cracking was observed, Figures 7 and 8.
table i.-
diacks were 6bsetved over the entire counter-bore region from the weld Cracks were also present for fusioh iine to the counter-bore transition.
66se distance bp the bevel of the counter-bore, Figure 8(a), but no cracks the unmachined reducer I.D., Figure 9.
Cracks were found were found at it the weld fusion line, figure 10, in sections from stations #0, #10, These were the only sections in which the root-pass fusion fi9, and #43.
in general, these cracks were oxide filled, similar to iine was present.
Only those cracks which could be seen visually those in the counter-bore.
Behr #43 were wide at the mouth.
to characterize the small axial cracks on the root-pass weld
~
bead, transverse serial sectioning was performed.
Typical defects and the deepest defect found are shown in Figures 11(a) and 11(b), respectively.
No cracks similar to those observed in the fusion line or counter-bore It is worth noting that this section was electrolytically could be seen.
The same technique was also cleaned to remove the oxide prior to tnounting.
As can be seen in employed on sections from the 16-in. end of the reducer.
Figure 12, this' procedure did not remove the oxide from within tight cracks.
'Fractographic examination of specimens, which had been cleaved open and elect'rolytically cleaned, showed that no dissolution of the base metal had therefore,.be assumed that the defects observed in the
'od'eur r ed. It can, It is most likely
' weld bead were different from those in the base metal.
t' hat the crack-like defects Vere formed during the welding and were sub-a sequently corroded.
R l
Shk.
\\
O TABLE I NUMBER AND DEPTil 0F CRACKS AT 18-IN. REDUCER END Maximum Depth Number of Cracks Section (in.)
Deeper than 0.002 in.
- 0 0.040 3
- 10 0.018 8
13
- 15 0.043,
- 19 0.026 7
- 22 0.014 4
- 25 0.004 1
- 28 0.008 3
- 34 0.007 2
- 43 0.042 7
e 4
A e
O B-5 1044 255
.= _
Two sections were taken from stations #14 and #41 of the 16-in. end the wcld where a careful magnetic particle test had indicated cracks at Cracks could be seen at fusion line on the reducer side of the veld.
both sides of the fusion line.
These cracks were tight, oxide filled and No cracks or oxide spikes were present in the branched, Figure 10(b).
Sections were therefore taken at areas where visual tounter-bore region.
Included for examination were the wcld bead, examination showed cracks.
transition zone and the step 2-7/32 in. from the weld. Micrographs showing typical cracks observed are shown in Figure 12.
Detailed exa=ination of crack tips showed that they are invariably blunt and oxide filled, Figures 7 through 13.
On many cracks, oxide spikes propagated perpendicular to the crack for a considerable distance into the bace metal, Figures 8 and 13.
This behavior was not micro-and must thus be attributed to changes in plant structurally dependent, for a prolonged period of operating conditions which caused crack arrest In general, the oxide within the cracks was dense for a considerabic time.
distance from the crack-tip, Figure 13.
In some cases, however, open fissures did extend in the oxide from the surface to the crack-tip, Figures The Pillings-Bodsworth* ratio for the oxidation of iron is always 8 and 10.
greater than 1, so that the oxide would tend to fill cracks even under the The cracks in which the ox4.de was influence of a moderate mean stress.
fissured were deeper than others in the section. The possibility that stresses applied during reducer removal may have opened these cracks cannot be ignored. It is thus difficult to draw any conclusicas based on this observation.
A transverse section through the saeple received on July 18 is shown in Figure 14.
The defect lies at a large angle to the pipe surface, as The fissure was predicted from the UT examination performed by SwRI.
Note that the pipe oxide filled and surrounded by a decarburized layer.
. O.D. was not decarburized. No distinct interface existed between the de-I carburized and normal zone, Figure 14(b). Upon opening, the oxide was found These characteristics are typical of a rolled-to be powdery and nonadherent.
in or forged-in lap.
It is, therefore, concluded that this defect was present prior to welding and resulted from the pipe fabrication process.
- The ratio of the oxide to metal volume.
jQffh B-6
~O FRACTOCRAPilIC EXAMINATION Specimens were taken adjacent to stations #15 and #43 of the 18-in. end for fractographic examination. Specimens were notched behind the counter-bore transition and breken open, Figure 15. Beach markings could
- be seen in several locations, Figure 15(c), indicative of periodic crack arrest. All cracks were heavily oxidi::cd. Even at the transition to laboratory cleavage fracture no detail could be seen, Figure 16(b).
In some areas, the fracture surface appeared to be intergranular, Figure 16(a), but electrolytic cleaning removed all such features. The cleaning procedure did not alter the appearance of the cleavage fracture. The intergranular facets are therefore ?.tributed to the oxide morphology. After cleaning, the surface of in-service cracks was pitted in appearance, Figure 16(c), but did not show signs of fatigue striations. ~ Energy dispersive X-ray analysis was performed on each of the fracture surfaces before electrolytic cleaning. For comparison, spectra were also collected on the region of laboratory overload. Iron was always the primary O. peak observed on the in-scrvice fracture, Figure 17. Small peaks indicating traces of phosphorus, sulphur, manganese and copper were also usually ob-served. An additional peak corresponding to titanium, but slightly shifted in energy, was also observed. Comparison with the spectrum from the cleavage fracture, Figure 17(c), indicated that the iron, manganese and titaniu= peaks originated from the base metal. The remaining elements, phosphorus, sulphur and copper must be attributed to service exposure. These peaks are not un-expected. Coordinated phosphate water treatment was used at Foint Beach for some time, accounting for the presence of phosphorus. Copper from corrosion of condenser and heat exchanger tubes is also co=monly found deposited on steci components, since iron displaces copper in solution according to the reaction: - Cu + + Fe Fe + + Cu 2 -Copper in solution therefore acts as a cathodic species. Sulphate is known to be a coc=ior. species in turbine deposits and, therefore, the presence of sulphur is not surprising. No evidence of halides was detected on any of the O 1044 257 r,-7
On one sample a small nickcl peak was also observed. specimens examined. Again, this is not surprising as nickel-base alloys are extensively used in steam generators. In some samples phosphorus was observed near the crack mouth, but not However, in one sample (at station #14 of the 18-in. end) at the crack tip. Lery little phosphorus was observed. One section from stations #11 to #11-3/4 of the 16-in. end was a This area was selected during visual examiantion as that con-broken open. Several small cracks could be seen, but taining the most obvious cracks. in length, exceeded 0.003 in. only four cracks, totaling less than 1/4 in. The deepest crack was 0.013 in. deep. A large segment of the 18-in. reducer was subjected to crack sizing using The segment from station the Satellite Pulse Technique developed at SwRI. Sub-and the crack depths plotted.
- 11.5 to #15 was examined in detail sequently the segment from #14 - #15 was broken open and the crack depths From this The comparison of these data is given in Figure '.8.
plotted. data, it is apparent that the crack depth is less than 0.050 in. along the therefore, seems unlikely that It, segment giving rise to the RT indications. cracks significantly deeper than 0.050 in. exist elsewhere. HECHANICA1 TESTS Charpy impact tests were performed to determine the fracture appearance Fourteen specimens were cut to simulate overload transition temperature. The long axis of these specimens was longitudinal; from the service cracking. In addition, four specimens the notch was cut parallel to the I.D. surface. were cut in the same orientation but with the notch perpendicular to the I.D. Two specimens were cut with the long axis tangential and the notch surface. The results of these tests are tabulated in parallel to the reducer I.D. Specimens notched parallel to the reducer I.D. often gave un-Table II. reproducible results because of delsmination in a plane parallel to the I.D. No major delamination was observed in specimens notched perpendicular . surface. (samples 4-1 through 4-4) or in the two tangential specimens 1 to the I.D. (2-1 and 2-2). ( 1044 258 "-8
TABLE II RESULTS OF CllARPY IMPACT TESTS ON REDUCER HATERIAL Fracture Temperature, 'F Specimen Energy, ft-Ibs. Appearance + ^ 440 V-3 191 D-L 440 V-4 167 D-L 155 D-L V-1 220 150 D-L 220 V-2 D-L 75 V-5 166.5 V-6 133.5 D 75 132 D-L 50 V-13 10% B V-14 107 50 125.5 D; V-7 32 47.5 70% B 32 V-8 95% B 0 V-9 30 V-10 155 D-L -20 V-11 29.5 95% B 0 V-12 32.5 95% B -20 4-2 96 20% B 100 76 30% B 4-1 75 84 30% B 4-4 75 4-3 56.5 80% B O 50 100 2-2 72.0 D 75 2-1 44.5 80% B cut with specimen axis longitudinal and notched + All V-speciment All 4-speci.sens cut as above but notched para 11c1 to I.D. All 2-specimens cut with specimen axis perpendicular to I.D. tangential and notched parallel to I.D. i D: ductile D-L: ductile with major delsmination B: brittle b 6 0 B-9 1044 259
SUMiARY OF FINDII;CS The crack-like RT indication at the 18-in. redecer end was found to No cracks deeper than 0.047 in. correspond to shallow, oxide filled cracks. were seen. Cracks were deeper and more numerous at the 18-in. end than at the 16-k.n.. cnd. The deepest cracks observed in any section at the 18-in. end were At the 16-in. end, cracks always at the counter-bore transition zone. initiated at the veld fusion line were deeper than elsewhere. No micro-Cracks were vide, oxide filled and frequently branched. structural dependence was observed. No cracks were Cracks were invariably circumferential1y oriented. detected on the full reducer I.D. -~ Beach markings were observed on the oxide-covered fracture surfaces, but no fatigue striations could be found. No anomalous species were observed in the crack deposit. A lack of fusion defect in the 16-in. reducer-to-pipe veld was the only microstructural abnormality seen. The reducer has a FATT below 100*F. Above that temperature the energy-to-fracture a standard charpy specimen exceeds 100 fr-lbs. DISCUSSION Little work has been performed on the stress-corrosion cracking and Veinstein has corrosion fatigue of mild steci piping alloys in pure water. water at 550*F stress-corrosion cracked SA 106B and SA 333 Gr. 6 in 8 ppm 02 However, slow strain rate tests at stresses above the yield stress. indicated that susceptibility to cracking was sensitive to oxygen content, level was reduced to 200 ppb. It is, decreasing significantly as the 02 therefore, probabic that at the lw oxygen levels present in PWR steam gelierator feedwater SCC would be extronely slow and require higher stresses than at 8 ppm 0
- 2 Several studies of the corrosion fatigue of pressure vessel steels and
- These, carbon steel piping alloys have been performed in PMR cnvironments.
in general, conclude that carbon steel behaves similarly to the quenched and ' "- 2 o' 1044 260
O A significant R-ratio and frequency effect is tempered grades (2-4). The threshold AK decreases with increasing R-ratio and a detri-observed. The ASME guide-mental influence of very low frequencies is observed. l lines for predicting crack growth rate have been found to be conservative significant crack branching was observed. (2-4). In one of these studies fn addition, it was reported that a thick oxide layer extended to the crack-tip. Tests in PWR primary coolant environments have generally observed similar corrosion fatigue behavior (5-8): low frequencies and high R-ratios Again, some workers report that a significant amount of are detrimental. Similar behavior has also been crack branching and oxidation occur (5,6). } observed in X-65 pipeline steel in water In view of the above, the cracking observed here is consistent with a The branching observed does not corrosion fatigue crack growth mechanism. The necessarily indicate that stress-corrosion cracking is involved. beach markings observed should not be interpreted as direct evidence of a corrosion-fatigue cracking mechanism, since such markings do not result from crack blunting on large stress excursions, but rather from crack-tip dis-Such marking could therefore be solution and changes in oxide corphology. produced during stress-corrosion cracking provided that operating conditions periodically altered and changed the crack tip chemistry or stress state. ' here is thus no direct evidence that stress-corrosion cracking has not 3 occurred, or ' hat fatigue loading is involved in the fracture process. 'ork of Weinstein( } the possibility of stress-corrosion crack-Based on the However, because of the low ing in mild teel piping cannot be ignored. icvels believed to exist in the feedwater pipin; system and the low strest Fatigue oxygen Icvels, stress corrosion cracking is thought unlikely. loadings have been theoretically predicted in the areas which have cracked, and thus corrosion fatigue is the prime choice for a cracking mechanism. Several observations are believed to indicate that the cracks seen for a considerable length of time. All crack tips were Eave been present In addition, no fractographic detail, such as blunt and heavily oxidized. O B-11
The environmental conditions seen in fatigue striations, could be seen. t the cracks were service are not severely oxidizing, so it is probable tr.aTo verify this suppos slow moving to allow such extensive oxidation. l s which have comparison should be made with cracks observed in the p anc Certainly it eqperienced rapid crack growth with observable striations.the time wh at would seem that some of the cracks were present ks could be phosphate treatment was used, since definite phosphorus pea identified in the X-ray spectra. l There is no indication that inappropriate water chemistry contro All the species detected are known to exist in was the cause of cracking. tice. The PWR feedwater and do not represent a departure from normal prac d crack heavy deposits of elemental copper selectively located at pits an t which iron is mouths may be regardad as nor=al, since these are sites a The degree of copper build-up is an exposed without a prot ective oxide. ficant con-the reduction of copper ions has made a signi Thus, the presence of copper ions indication that tribution to the iron oxidation process. There is no in feedwater must be viewed as contributing to the failure. h bsence of copper evidence, however, that cracking would not occur in t e a icns. since it The absence of axial cracks must be viewed as significant,gen-has been hypothesized that the cause of cracking is thermal stresses such th'ermal erated when colder feedwater enters the hot piping system. f a biasing stress, stresses would be biaxial and thus, in the absence o The presence of a biasing stress, result in randomly oriented cracks. f cracks at the axial in orientation, is also predicated by the absence o full reducer I.D. the counter-The stress level would be lower in this region than at the These results also imply that bore because of the thicker section. h ld value for cyclic stress amplitude is relatively small, below the thres o limi-All cracks were associated with stress raisers, so that the c i tress may I R = 0. nation of such geometries and reducticn in the axial bias ng s eliminate the problem in the future. / \\ 1044 262 n-u
O The results of the charpy tests indicate that reducer material has a fracture appearance transition te=perature no greater than 100*F. Speci-nei.s taken to simulate overload of the in-service crack gave a vide scatter because of delsmination in the plane parallel to the reducer I.D. Speci-mens 2-1 and 2-2 did not delaminate so that these tests should give a fair indication of the FATT for transverse specimens notched perpendicular fo the reducer I.D. This is slightly higher than that for the longitudinal specimens, as is normally observed. These results indicate that the re-ducer material has high toughness at temperatures above 100*F. Thus, there is little likelihood of catastrophic fracture in this temperature range. At lower temperatures, however, there exists the possibility of propagating a brittle fracture. CONCLUSIONS The most probable cause of cracking is corrosion fatigue. Cracks are believed to have been present during the period when coordinated phosphate water treatment was used. The association of all cracks with stress-raisers suggests that the fine polishing used in the repair procedure should significantly improve piping and re.?ucer longevity. The toughness of the reducer material at teuperatures above 100*F and the ductility of carbon steel renders a catastrophic failure improbable. O 4 0 B-13 1044 263
REFERENCES 1. O. Jonas, " Turbine Steam Purity," Southeastern Electric Exchange, Washington, D.C., April 1978, Document STD E-1528. 2. D. Weinstein, " BUR Environmental Cracking Margins for Carbon Steel Piping," EPRI Joint Projects Review Meeting on BWR Pipe Cracking, October 19 and 20, 1978, Palo Alt, CA. 3. D. A. Hale, J. L. Yuen and T. L. Gerber, " Fatigue Growth in Piping and RPV Steels in Simulated BWR Water Environment," U.S. Nuclear Regulatory Commission, March 1979, NUREG/CR-0390. 4. M. Suzuki, et al., "The Environment Enhanced Crack Growth Effects in Structural Steels for Water Cooled Nuclear Reactors," The Influcn.e of Environment on Faticue. 161, Institute of Mech. Engr., London, 1977. 5. W. H. Bamford, "The Effect of Pressurized Water Reactor Environment on Fatigue Crack Propagation of Pressure Vessel Steels," ibid, page 51. 6. W. H. Bamford, D. M. Moon and K. V. Scott, "Effect of High-Temperature Primary Reactor Water on the Subcritical Crack () Gro-'t> of Reactor Vessel Steels," HSST Quarterly Progress Report, Sept. 1977, ORNL/NUREG/TM-120. 7. T.'Kondo, et al., " Fatigue Crack Propagation Behavior of ASTM A533B and A302B in High Temperature Aqueous Environments," HSST 6th Annual Infor=ation Meeting, Paper No. 6, April 1972. 8. T. R. Mager, J. D. Landes, V. McLoughlin at.d D. M. Moon, "The Effects >f Low Frcquencies on the Fatigue Crack Growth Charnt.ristics of A533B C1. 1 Plate in an Environment of High Tcaperature Primary Grade Nuclear Reactor Water," HSST Report No. 35, Dec. 1973. 9. O. Vosikovsky, " Fatigue Crack Growth in an X-65 Line Pipe Steel at Low Cyclic Frequencies in Aqueous Environments," ASME J. Encr. Materials and Technoloev, 97, Series H, No. 4, 1975. e O B-14 1044 264
h*w } Qr Su gf ,e. .,x g kh 8 ,t .h,{: *, w
- 3. 4,,
n r.. bs 3 gg ' j. v. ? s vg r-y,'.- 'c ...:~. m r ** r gyj,.. 1,- i ;' 's n. x ~ 4.- 4X 2-37774 j q ,,,--y N ;!I'HIl?'- 'l jij j @ U lI(, a 1 ~ ?f r.,4 -
- 3.*
r.,
- go.
- - y.'
~ at ' W,, - f- ~ y 4 2, _.4,xg,,g y:. $ e l f.%c'
- . f 4,4 g, q,.R,
. b..,*1 ,e C-5 9.% \\*,., r 3 ,r ,~ .>.e h ~ '~ ?; ..e . 4. -
- ..,. a"'
- , *i. *m'.h
. _ f..wy%;f ~: "f;
- W
.' ;,. 't ~ ; ~. s 3..',,-, sc, Y,. g '..?*.,i, s g*,c 1 . 4 $ [. s,. _ ~/* 1 4' s s ~ .e
- y
. k..; ,,.~ .t, : : '. ~..:.. g.. 4 i% p.f - >, c ;
- . f.
b f,,,,; 'p., 3:' ; :$4 .'d. ' u; ,1 :. s.' ' ~ 4x w 2-3'; (b) FIGURE 1. INTERNAL SURFACE OF REDUCER AT 18" COUNTER-BORE REGION. Note the fusion-line crack in 1(b), 1(a), #19-20 segment, 1(b), #44-45 seg=ent. 1044 265
r _y y.. C J$ ** D
- f _
g. -q x ;3, M ; 1 4 %, t .- = A_ a - n. m A&%$$$1e..49 '; ~a - - r.,Q = 19X 2-37823 (a) %(IU):"I?il(p Fg [ f) tl "l: i : ;j,\\ [ o r;. p, ,3. 3 a i, n ,1 e ;, y g. ; ;..; p g,M,;t y-g;,gqq, c -u. ] *b'.h.
- a. {
f h... 05 ^ t 6- 'y W- .}g- ^ e.ly 4%:m_. s' v g e 9 .k M- ,i R.'
- -y
.si -97', A I,(,, .g 8'. .k> 5* ' p*
- i"a^f r
f*a . c. - .4
- fy',
M p s. l g 4 M 4.*. 3 # s r. Q '. ~ ~;
~l+ 30X 2-37930 (b) FIGURE 2. INTERNAL SURFACE OF 18" END OF REDUCER AFTER ELECTROLYTIC CLEANING. Note that cracking occurs at the root of machine marks near the counter-bore transition, (a), #15. Small axial cracks were occasionally ' resent on the veld bead, (b), #43, in additien to fusion-line cracking. 1044 266
'. 2 [ ms r EL =3') W49 3 f.l'{[EEEU Li ' ng mf 'I 9 k# 1 i 2-37827 10X ) f <.k ~ ~ A UNT E F 1E S TO TIIF. A LEFT. Note that little cracking 2-37825 SIX has occurred in the IIA 7,where (b) g ndin as removed the machining
.'... J
- g st ag %a-c u.*u%
.s: ~ .r:.. - ~ ,, K.. s"~.4 ory o... k am; 3-g = .'g.%,c.% 4** p,- ..aM
- e 3,* h.*.
o g4 , tig8"4 4*.
- y as t.
- .* =-
+- /
- 4 s
,-".V . %O% P--% em 49 L'4 m e% 1 g 9__'."***M .g. m---- m i '* k E- = a, e' :. O d t-2.- - -... _g F. O.-' #m n% r*.,.... r..,- n - m 0:i4li . *. _ - -w < ..Fl, M. i, " : / i~y n yl - p. =.. - 's 4 3 " ':,: ) j q\\,1 a '. 4.. i\\ + en. n h*
- L '
- O i.k wA 'eaniL ;.se. :...,> * *..,
! n. t ;i !. _ 1 :.,.,1,,p,,. ca. y. i 2. , s i i .t 4.,. ~
- N. ENa a M"((**
^ / h* i j i.j Y J q ...~. Q+%L'5yp~.*?.~'es s ~ .,.f. y :Qe?as'% ~e, ... ~ -~ 4 s ~~ t ,*~ 4 *
- ,, +
r y GMk - a,, ~ e ,e -A h.' '.e$,,** % %ema.. e . ~, ,~,$*3
- .v.
~ A,. ,a g '
- wMukJ ~%* * #" g\\
e-. g k W g g 2-37977 3.5X (a) W ,.g.-_ 4 + e 6 ',), y
- - 3 j
Y. / .s e, 'h.* a r e a ,s. dD"j v
- 1 4
9%,. o /1 y', 4 e y E d o y x .-s ' ^' ^'_ ' -.[ .w -i ~' w .w e a Q -e f g. ~,. F m m. t 2 m.,. 2-2136-S (b) , FIG dE 4. COUNTER-BORE TRANSITION (ARROWED) 0F 16-IN. REDUCER END AT WHICH A COPPER DEPOSIT (b) WAS OBSERVED AFTER SOME CLEANING. 1044 268
e a wywr em Q, r V. i D ~ N 4 M1 I 2-37658 500X 2-37660 (cr.O 500X (a) (b) 7] ~ ~ ~ . ' '.l ~ fl~ TE * -
- ye$
7.' ~ ~ ~ ' yf..].ki; k?* l , ;,
- 1 -
( \\,4
- y'
'_2 FIGURE 5. TYPICAL MICR0 STRUCTURES OF REDUCER AND f4, ,' r. WEi,D. Base metal, (a), and heat- .,w~ .'4 . h*g.*, / id' ' .~ affected zone, (b), of reducer; r 'a 4. Widmanstatten ferrite in weld head. (c).
- 5. ',, "h,,,'m/;..,.p,.~1'@ph
. 44 w-h he.% I % e n. ?:- .. Q 7: f.J3 2-37661 100X (c)
4 ?, . e. a.- ,oy*e,- .I a '.. 3 Q g.g. ."% W.h Q* .L: i. . $,4. = 3,, l s g,_ s...c
- l.,,
.n ya - (p', 'rK@ ?. l. t , \\. ' ]I,t oi- '3 [ ,h, i / L, l. ic - C w i_,- ~. T '. 1-( = } ',R s I ) -. s j
- y
,s. g. y. 5 \\.!q _ a. G,
- , u
+ u.. n 4 5 .r.,~..%-_'., ..._..m.._,...s.. ...... *- i ' -.....- - - - - _. t y 4X 2-37936 (a) 4 s' yfag 6:w.. 'm M wr2 y :n s W : :: a^; cixy..hY. m: $ $,. n y,if)$$>N.%..;n f g5%J2{i,dF[.3.'.'f., 'yA[h;j;gh',?y "l'lQ.&l;r.? b th.Ws. ' bHY'5'h..,3 ',h. E! ,m >g:.L ( ?;; < ~ - ! /,.:If.yd.i,.:tB
- &>.4f.,,T;...
/ h M b n >xu.t w py ..u.~..y-~; -s-e' ., W^ . f.Cta:,*.'. v.,sy. g g J y d % %. y'in p-:'.,. ....y Gzy, ..ja
- ,3 ;-
f <,;.:W. u..: t2. ; 4,- .s.i,v. ; 4.';g :?.r.e,y.y,: ~.2, . Q,.. M... y. - .s N,. a i ,y.:,,fy.. n.e,;..~. ;. W. 1 t + %. w :.';. t.v.t.O -w.x. g,f 3. '.,1 o :-- - -
- n. WW \\: V.,. vl.,,;.' %.!n p.e,,,]..n... L.., +,
. w+ m 4 ..up., :, % %. ,v.,.,. s v. x ,c.a g,+c.,7. g.s, a <..g :sp. !.,.,.. -.. ~, ' '. 7<w ,, s. ce-a ,;O:, A a# .~. :: ::. ' h5)g.Q. ',W....*n?$ N!?.&'o$ k.<:uul;f.iJ:. e .&.~5 ? ?.^.- Ew2a$$. ',
- ,,+
's..,,.., '.y D,. s,a; %:'N 50X 2-38114 '.15 (b) '!IGURE 6. WELD DEFECTS IN THF 17-IN. REDUCER TO PIPE WELD. The arrows in (a) indicate the extent of the lack of fusion defect. In some sections oxide or slag inclusions could be seen, (b). 1044 270
N N~ 1, r 2-37791 300X 2-37789 50X (b) (a) FIGURE 7. LONCITUDINAT,SECTION OF 18" REDUCER AT,(a),AND,(b), STATION #19, (c), STATION #28. x), u i ,. _ ~~ - 2 1 ^ __3 .] '~ } ~ (:. 4 [E _ _. pa 1 2 ('; '; FJ Sox 2-37617 (c)
e p6 s 2, ') L: - __3 L_ t; 1_ -s -;; ) r_ 8 I t f 1 ,f e l g-c.;. ~ s-__ k b 50X 2-37622-24 (a) FIGURE 8. LONGITUDINAL SECTIONS AT STATIONS #15, (a), AND #43, (b), IN Tile COUNTER-BORE TRANSITION 7,0N E. Note mtiltiple cracks. l I A L 4= N I J, w N s a A 50X 2-37627-28 (b)
F,"?22.N{t'Y!!A'ND7.'E$QI,%.id%DhM[ %# < ? I op o-L .J l.1 d ~ yyr.
- r ' y(8t$s;m gy;;y.,
. -sg-ff t?p,k IE ' w.' t a. 7 (.4 fl 5 $ h.Y Y 5. (h &l 2-37822 100X (a) __q 2-37819 300X (b) ^ FIGURE 9. TRANSVERSE SECTION OF REDUCER I.D. AT STATION #17. Typical structtere and surface finish, (a), and most severe attack observed, (b). 10~44 273
pU l @! ', \\! @! ' Q >4 I ) c.3 %} it
- d
.['c [ r1 >\\ mi i. t !U, I i lC; - u A 2-37626 50X (a) n w .a*,,4 e<.. d e, .>. yy y24.. R. a$. '. ?QG.,.W...%,L. d.V f);. > .~ r.. p., . f.l. Q. ' f.Th}..%y } 6...
- - 7*
. W '. '. b. mN., 's., -:.. M...-
- .,u h.
.+...s.. ,$. ;., n, a.'., % '.. y.< e,Q... y,,..?.. s 42.. M . p ~.. c.1. ~.. -,r ,o.
- w. p; y. y
~. y. .m... w e.... w ~, w... ^ * ~.i. . o,. ' 'Q, . 4 s, - .t 8. )N..... ' *.., y.y.?.. iW. t u.-;
- c.
y. <t;.1., e v.., ,e;, . g s i.Q. -4.gv.n, v s p .s s. .r' . a, y'., %,.. 9 h:% $-Y:k^K ~,Ih,T:-fQi.1$, @. 'ht -T..~
- & b $,s ?' '. $ h' h&$$ k5 V.'.' O,'-' '
$.h;gh.:.f. .f WWa/M @lti?R l. ?.t . s.%..th. !.&. :. Q,s .*.~....'.: ', s .e,. -Le.. + wa. A. 2-37934 100X (b) FIGURE 10. LONGITUDINAL SECTI0 tis AT THE REDUCER-N0ZZLE, (a), AND REDUCER TO PIPE WELDS. The veld bead is on the left in (a) and the right in (b). Sections were taken at stations #43 and #14, respectively. Section (b), etched in nital. 1044 274
Q' ' 7, U r1 I Il j. lij,,,, L
- s..: EL- -
--] 150X 2-37937 (a) \\ 250X 2-37941 (b) FIGURE 11. TRANSVERSE SECTIONS THROUGH THE WELD BEAD AT STATION #43, SHOWN IN FIGURE 2(b). No deep cracks, similar to those observed in the base metal, could be found. 1044 275
t I e t i puo m ,, r -.r (' +\\f\\ 0Y I ; ! I k i ( r .,s 4 t 2-38112 500X (a) lW ( Y 2-38113 300X } (b) FIGURE 12. CRACKS DETECTED AT THE COUNTERBORE REGION ) AT THE 16-IN. REDUCER END. 1044 276
n ( e l > 1 s o c 4 ? 2-37793 300X 2-37935 500X (a) (b) s __ _ _ J 1 y n n 6
- n...
s 2-37796 300X (c) FIGURE 13. DETAILED VIEW OF TYPICAL CRACK TIPS, ETCHANT: NITAL. Sections taken at stations #19 [see Figure 5(b)], (a), and #34 of 18" reducer end and station #14, (b), of 16" reducer to pipe weld (see Figure 7(b)]. 1044 277
l, .R ~ 75X 2-37864-65 (a)
- ~ -
'k,, ) 'q g* i o',e p#y, - .- y $'.4 FIGURE 14. TRANSVERSE SECTION TilROUCH CRACK AT
- g y
" k' h* (e-p.. 0.D. Note the decarburize d zone surrounding the oxide filled crack, ,,? g of. c'? (a), and the absence of a distinct i 'r 5 interface, (b), between this zone and j; g.4,p, *
- g "}g
'b the base metal. ,gd .r :.. -];., s g Qy;- d' 'I + s ' Q h, W '- N
- . i q-ag Y W -
m %.3 ~ y, u.3 %.; *';s." N':,J
- X '.-
?. ' t ' tj
- ?.
2-37863 300X (b)
l L
- i
(.. I s .,,, fe,*j o, r, k' 7 [ q,r e ' bl l. h L u. n :. J I i ! .,3 j,,, s, A p)* i. w.w. ,c w w +y:.;.n .. f } Q '. y _' "; ' ~- m ... s - - g :. g-y 3 A. ~ .- +. m y.,f: A.,. ~ ' ws ./ n __ _. 2-37620 4X 2-37619 4X (a) (b) r r.' ,e t' r p I ,1 r-r ? .u r 's l9' y f a i n 2-37619 12X (c) FIGURE 15. SAMPLES FROM STATION #15, (a) AND #43, (b) and (c), READY FOR FRACTOGRAPHIC EXAMINATION. 1044 279
e - e 3 He e o U.c m.c &a.1 v U tt) A V W =c C C
- u V
~ Um e 6 a g c E, e g w ac .n e ,A Kv W C j e. p c M
- 3 y
-e[ i i la2
- ED G.
> c M A 4-t D' mg u ,i. 4 a +- M U W G dW D C f l g,. la.1 W u W
- s I
A 3 O W tr) to ID C.O # o a w te t 9 L...
- e. g g
9 g ..a
- 3 L
U CM -r_] 4 ^ .j g g,c F . pee, ;, ,m ww M f~ -. i ) a, e W U C^4 4; .O La.1 u.C > t, "] I v 0 CV G ,4 - l a 4 ~w y c.4 uk M E> M C -) t ",a n', . 5 < wg wo 3,{ S... 3 >.c D l .i ( '1_ r -f 1 Hj 3 p o .-4 uM (c, s _ ~., g 3 g gf) y y ~ . y. &.- l A e ac @ C c c c..: w s. ~ c co u g -. y .e? ---.3 n .., -s a' ,y* b l[ .) d u___.- s, s~ e4 g I W m g 4 ~ ~ x g' ji ' $;
- r==
D Ch O p4 m ?. a ga ~, , r.. .___-... _ ~.~..&. ,m . -. _..... J N 1M ~% w -a i s % tW .c s. .g, 4 + 1 i g f "[ f ~[* [t
- g.
4 0 g. j .v + l 73 5. 'j ' I +
- . ^ i
- s
( y e .m 'm [ %. m 3 4' 3 ' 3, f( = { ; g. S 4 j., e, i s
- 3... {
l 4 .'gm. --. dK'5J ) n ~,, is g e .,x 4 U n ,4 --.7. 3 g a e g v 7 .+ y v .e k w iM ' 3. L g y. j% " x. a:.5 e s [ m: hf' a w j 4 g M I 1 g = m g ' r 4 m i.i ,f m f4 l i e> .~ r g 3
- y-
- .f... n Ch
,m, 0, ;..,
- =
t' q s. j ~ F,.... cp e -. j f MA N ~
- N, s
T g .a
- A*
e4 1 ( 4* g ~ e l ,j g g o N c .,.-.. T.. .. x. - -_ : ? 1044 280
,R* gl:l-Q.*' }_' WR y ;, % W,, s., c';
- y.
1g y y. ,s_ %K %,f,, $[q4 i, f C* _y;: \\ $hd3$i'J' / ' ? ;, ), 200X 500X 2-2067-S 2-1948-S (b) (n) s..e $g \\ .y ". FIcUltE 16. TYPICAI. FilACTURE SURFACES OF CRACKS IN /- 18-IN. REDUCER END, BEFORE OXIDE REMOVAL, , J,h/ 2.1 '** 4 N 6- \\[ ' 84 %s - ..h'.h ? t. ';!! ' ' Sg( d. (a) AND (b), MID AFTER OXIDE REMOVAL, (c). ,4 A s, ~.1 A 6d r, 'w. hk I..~ r, c_.0 er{: i. p j. 4 -,..;y '3. > (*f,,_,. w '- -1 3 q r a
- a h..,,, -
',Q A ' 2,:". a,- n v. , 5,qw>.: . $-f. , q....s. ,Y-( r - _z 'ul, )=. 1000X 2--2083-S
a e. O h u [.7, us >1 .J O; [T f Q' 'f I; I .' - i. i 1' ii: n l i g;i .I i 3 LJJLs uLJLTs.Q L to e 4 6 e 4 40 40 m 4 E 3 = 8 C 4 o. (-= C L a 6 n 4 OP') OV ~ C L Z M Q.= X
.<?"
04 a : zz C :. E- = J 4
- "F O
W $s. ~ * * $E S OA -= O W E C 'ssv O zwzo = zz z 'J J=
- G L
~< = z',.:.= ,I ? ~ N E l z zz-O O Q =9 e:paze ( O b a J.O w
- u w: =
um .c =xxoa s - w 2<cew c o uz oe O z I^W a L o: c O to. u e >, :.c w e >-. e :, ~ - - e.u as z .o-u -=
- =uxs z==
.: u c=s= c zsz s3 0 -ze-s O N 5 N O. ^- E. ? E-
- z ;: 2_.- 5 0
u-<su z O O =c O tg O O = 0 w A O O O a,, O i e e o O O O O N 10 T C01 x sayauT 'q2dag zoe23 1044 282}}