ML19344F045
| ML19344F045 | |
| Person / Time | |
|---|---|
| Issue date: | 08/31/1980 |
| From: | Frank L, Hazelton W, Hermann R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE), Office of Nuclear Reactor Regulation, NRC OFFICE OF STANDARDS DEVELOPMENT |
| To: | |
| References | |
| NUREG-0679, NUREG-679, NUDOCS 8009120337 | |
| Download: ML19344F045 (31) | |
Text
-
\\
NUREG-0679 l Pipe Cracking Experience in Light-Water Reactors l
l l
Manuscript Completed: June 1980 Data Published: August 1980 t
L. Frcnk, W. S. Hazelton, R. A. Hermann, V. S. Noonan, A. Taboada D;viason of Operating Reactors l
Office of Nuclear Reactor Regulation l
U.S. Nuclear Regulatory Commission Wsshington, D.C. 20555 7..,
3 '
a e
l
\\...../
l l
I 8009/90337
l i
l l
ABSTRACT Commercial light-water reactors have experienced pipe cracking since 1965.
This report summarizes pipe cracking experience in light-water reactors as reported in Licensee Event Reports from 1967 through 1979, other licensee and vendor reports, and Office of Inspection and Enforcement Bulletins.
I Pipe cracks which were environmentally induced, such as stress corrosion cracking of metal sensitized by welding and heat treatment, were most pre-valent.
Feedwater pipes experienced fatigue cracking from thermal stress i
and many small lines developed leaks as a result of fatigue caused by vibration.
l l'
Cracking incidents are separated into generic categories and listed by reactor type, pipe size, and systems affected.
I 1
1 l
l l
=
t iii
CONTENTS Page ABSTRACT............................................................
iii 1.
INTRODUCTION...................................................
1 2.
OPERATING EXPERIENCE WITH PIPE CRACKS..........................
1 2.1 Boiling Water Reactors..............................
1 2.1.1 Intergranular Stress Corrosion Cracking in Piping..
1 2.1.2 Intergranular Stress Corrosion Cracking in Safe Ends.............................................
4 2.1.2.1 Safe Ends Sensitized by Furnace Heat Treatments..............................
4 2.1.2.2 Crevice Environment.......................
6 2.1.3 Miscellaneous Small Lines..........................
7 2.2 Pressurized Water Reactors................................
7 2.2.1 Fatigue Cracking in Feedwater Piping...............
7 2.2.2 Intergranular Stress Corrosion Cracking in Piping..
10 2.2.3 Miscellaneous Small Lines..........................
13 i
3.
CORRECTIVE ACTI0N..............................................
14 3.1 Boiling Water Reactors....................................
14 l
3.1.1 Piping............................................
14 3.1.2 Safe Ends..........................................
15 l
3.1. 3 Miscellaneous Small Lines..........................
15 3.2 Pressurized Water Reactors................................
15 3.2.1 Fatigue Cracking in Feedwater Piping...............
15 3.2.2 Intergranular Stress-Corrosion Cracking in Piping..
16 3.2.3 Miscellaneous Small Lines..........................
16 l
4.
SUMMARY
17 REFERENCES..........................................................
18 i
l APPENDIX--
SUMMARY
TABLES OF PIPE FAILURE............................
21 GLOSSARY............................................................
31 v
'N FIGURE PaSe 1.
Summary of Intergranular Stress Corrosion Cracking Incidents in Boiling Water Reactors, by Pipe Size as of July 24, 1979.....
5 LIST OF TABLES 1.
Intergranular Stress Corrosion Cracking in Boiling Water Reactors by Line Type Through July 1979........................
4 2.
Intergranular Stress Corrosion Cracking in Furnace-Sensitized Safe Ends in Boiling Water Reactors in the United States.......
6 3.
Summary of Feedwater Pipe Cracks in Pressurized Water Reactors.............................................,.........
9 4.
Number of Welds, Through-Wall Leaks, and Ultrasonic Testing Indications in Three Mile Islaij Unit 1 Pipe System as of August 28, 1979................................................
12 5.
Systems and Pipe Sizes in Pressurized Water Reactor Plants Reporting Pipe Crack Leaks.....................................
13 l
l 6.
Plants and Systems Reporting Unconfirmed Indications of l
Cracking as of October 5, 1979.................................
13 l
Appendix Tables A.1 Summary of Intergranular Stress Corrosion Pipe Failures in Boiling Water Reactors Through July 1975.......................
22 A.2 Summary of Intergranular Stress Corrosion Pipe Failures in Boiling Water Reactors, August 1975 Through July 1979..........
23 A.3 Summary of Cracking in Boiling Water Reactors in Piping Smaller Than 4 Inches by Plant as Described in the Licensee Event Reports..................................................
26 A.4 Summary of Intergranular Stress Corrosion Cracking in Pressurized Water Reactors in Piping Smaller Than 4 Inches, by System, as Described in Licensee Event Reports Starting in 1969........................................................
28 l
Vi a
l l
PIPE CRACKING EXPERIENCE IN LIGHT-WATER REACTORS 1.
INTRODUCTION Pipe cracks in an operating nuclear power plant were documented as early as 1962 in a report (Ref. 1) describing the cause of failure of austenitic stainless steel recirculating piping in the Vallecitos boiling water reactor, an experimental power producer.
Pipe cracking in a commercial power plant was first observed in 1965 in the Dresden Unit I nuclear reactor.
In the intervening years, cracks in piping of light-water reactors (LWRs) have occurred as a result of fatigue, corrosion, stress corrosion, or a combina-tion uf these factors.
This report summarizes the incidence of pipe cracks in both boiling water reactors (BWRs) and pressurized water reactors (PWRs) as reported in licensee event reports (LERs), other licensee or vendor reports, and Nuclear Regulatory Commission (NRC) Office of Inspection and Enforcement (IE) Bulletins.
In all cases, attempts were made to identify pipe cracks that developed only as a result of service.
2.
OPERATING EXPERIENCE WITH PIPE CRACKS 2.1 Boiling Water Reactors Stress corrosion cracking in BWR piping and safe ends and cracking found in miscellaneous small lines will be discussed and summarized in this section.
2.1.1 Intergranular Stress Corrosion Cracking in Piping
(
A primary cause of failure in BWR piping made of austenitic stainless steel l
has been intergranular stress corrosion cracking (IGSCC).
IGSCC is a con-dition of brittle cracking along grain boundaries of metals caused by a l
combination of high stresses and a corrosive environment.
Although a number l
of corrodents such as chlorides, fluorides, hydroxides, and sulfates are known to cause IGSCC in stainless steel, in the BWR cases the corrodent has been reported to be dissolved oxygen in high purity primary coolant water.
The oxygen level will vary from approximately 8 parts per million (ppm) when exposed to air at room temperature to approximately 0.2 ppe during reactor operation. The latter is an equilibrium level resulting from oxygen being continuously generated by rrdiolysis and removed by deaeration.
Even the l
lower limit of 0.2 ppm has been demonstrated to cause IGSCC at BWR condi-tions in the laboratory.
IGSCC has occurred in BWRs and in the BWR environment in the laboratory only when the stainless steel pipe is in the " sensitized" condition.
In this condition, the material adjacent to grain boundaries has been depleted of chromium because of the precipitation of chromium carbides in the grain boundaries and has become subject to accelerated corrosion.
Sensitization may result from heat treatment, welding, or any other process that keeps 1
c.
the austenitic stainless steel in the temperature range of roughly 800*F-1600*F for a period of time.
The primary cause of sensitization in BWR piping has been welding.* In these cases, sensitization occurs only in an area of the pipe next to the weld commonly referred to as the heat-affected zone (HA7.) of the weld.
Welding also causes very high residual stresses in that part of the pipe that includes the HAZ.
These residual stresses also contribute to the IGSCC.
In BWR piping, IGSCC has occurred in the HAZ of austenitic stainless steel welds only where the material has been moderately sensitized and residual tensile stresses were postulated to be close to yield stress.
To get enough sensitization for cracking to occur under BWR conditions, it appears that the carbon level of the austenitic stainless steel pipe must be at the higher end of the acceptable range of the Type 004 stainless steel material.
In those cases of BWR pipe cracks in which carbon was reported, the carbon levels were higher than 0.05%.
An early series of IGSCC in BWR piping was experienced in Dresden Unit 1.
The cracks appeared in pipes of the reactor coolant pressure boundary begin-ning in December 1965 and occurred randomly through 1974. All the pipes were of Type 304 stainless steel with a diameter of 8 inches or less. With one exception, cracking occurred in the HAZ of welds that were reported severely sensitized from a high-heat-input welding procedure.
Cracking that occurred in one 20-foot section of pipe was attributed to cold working of the inside surface.
In September 1974, a rash of IGSCC occurred in other BWRs near welds in Type 304 stainless steel pipe containing nonflowing high purity coolant water.
Several small through cracks were first observed in September 1974 in 4-inch pipes that bypassed valves in the main recirculating loops in Dresden Unit 2.
The cracks were discovered during a search for the cause of an increase in the rate of primary water leakage being measured by the leakage rate surveillance system.
Examinations of welds in similar bypass lines in Quad Cities Unit 2 and Millstone Unit 1 also revealed cracks in j
each plant.
This situation prompted NRC/IE to issue a series of bulletins l
(Ref. 2) to all 21 BWR facilities that had been licensed at that time requesting that all welds in such bypass lines be examined for evidence of cracks.
As a result of these examinations, which included ultrasonic testing (UT), penetrant testing (PT), and visual examination, additional cracks were found in 6 plants at 15 locations.
Of the 15 locations, 5 had through-wall cracks and 10 had partial wall cracks initiating on the inner surface.
l l
Similar examinations were conducted in the subsequent two years; no new l
cracks were found.
Coincident with the cracking of the 4-inch bypass lines, cracking was observed in two 10-inch Type 304 sta'rless steel core spray injection lines in Dresden Unit 2 during a regular inservice inspection of the piping.
These cracks had all the characteristics of the cracks in the 4-inch bypass lines.
- Sensitization of safe ends due to furnace heat treatment is discussed in Section 2.1.2 on safe end cracking.
2
l
(-
In view of the generic nature of the pipe cracks, investigations were initiated by NRC (then the Atomic Energy Commission), the General Electric Company (GE), the Electric Power Research Institute (EPRI), and several a
utilities.
NRC formed a Pipe Crack Study Group (PCSG) to investigate the causes, extent, and safety implications of cracking occurrences.
These studies (Refs. 3 and 4) confirmed that all the cracks were of the l
intergranular type, initiating at the inside surface of the pipes and progressing to the outside through regions adjacent to (approximately 1/4 inch from) circumferential welds that had been slightly sensitized by the weld. The cracks propagated slowly and, when leaks developed, were still in most cases relatively small (1/2 inch to 4 inches long on the inside surface).
Laboratory work by GE indicated that time for such cracks to initiate and propagate depended on the level of dissolved oxygen in the water, the degree of sensitization of the steel, and the level of stress (stresses approaching the yield strength of the material were required to cause cracking). The primary and thermal stresses in the BWR piping were found to be well below yield stress.
However, the contribution to the residual welding stress was estimated to be high enough to raise the total stress le/e1 in localized weld regions to yield levels.
Since laboratory tests showed that slightly sensitized stainless steel could be cracked at the 0.2 ppm-oxygen level estimated to exist during normal reactor operation, t
a sufficiently aggressive environment exists in operating BWRs to cause i
Near-stagnant conditions should result in higher j
levels through concentration mechanisms.
Conditions conducive to cracking appear to be prevalent in the smaller diameter, thinner walled pipe. To
(
date, no cracking was found in any of the many thick-wolled, large-diameter pipes in reactor systems in the United States.
It is postulate 1 that lower residual stress levels did not coincide with the zone of maximum sensitization, as was apparently the case in thin pipes.
IGSCC at the HAZ of austenitic stainless steel welds is still occurring in operating BWRs in such systems as the recirculation bypass, core spray lines, reactor water cleanup (PWCU) lines, and the control rod drive (CRD) i l
returo lines.
Several cases of IGSCC in cold-worked pipe have also been reported (Nine Mile Point and Humboldt Bay).
The piping in these systems ranges in size from 1 to 10 inches.
IGSCC has been reported in larger pipe sizes in foreign countries (Ref. 5).
Such cracking has occurred in 14-inch-diameter Type 304 stainless steel pipe in recirculation risers in Japanese BWRs and in a 24-inch-diameter Type 304 stainless steel recirculation loop pipe in Germany.
Table 1 summarizes (Ref. 6) BWR crackin! e.1erience by systems and includes 1 1y 1979.
Figure 1 gives these both U.S. and foreign occurrences throug 3
data by pipe sizes and shows that IGSCC nas occurred mostly in the 4-to 10-inch pipe sizes. A detailed listing of the cracking incidents is given in Appendix Tables A-1 (Ref. 4) and A-2 (Ref. 6).
i l
3 1
l l
1 Table 1.
Intergranular Stress Corrosion Cracking in Boiling Water Reactors by Line Type Through July 1979 Through From July 1975 Line type July 1975 through July 1979 Total Recirculating bypass 30 12 42 Core spray 16 49 65 Control rod drive 1
2 3
Reactor water cleanup 10 29 39 Large recirculating 10-in.
0 13 13 Small, <3 in. (not control rod drive or reactor water cleanup) 0 8
8 Other 7
9 16 Totals 64 122 186 Note: As of October 1979, the total was 191.
2.1.2 Intergranular Stress Corrosion Cracking in Safe Ends intergranular stress corrosion cracking in BWRs was also detected in safe ends that were severely furnace-sensitized.
One instance was attributed to a crevice configuration in the thermal-sleeve / safe end attachment.
2.1.2.1 Safe Ends Sensitized by Furnace Heat Treatments Begianing in the late 1960's a series of intergranular stress corrosion cracks occurred in severely sensitized austenitic stainless steel safe ends of BWRs.
It was common practice at that time to construct ferritic steel vessels with austenitic stainless steel pipe extensions (safe ends) welded to the nozzles of the vessels to serve as transition pieces for field welding to stainless steel piping.
These pipe extensions were subjected to l
the stress relief treatment of the ferritic vessel, typically 1150*F (621*C)
(
for times up to 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.
This heat treatment severely sensitized the stainless steel safe ends; it was later determined that this condition contributed to a rash of incidents of IGSCC.
Two major incidents occurred at the Oyster Creek Nuclear Power Plant (Ref. 7) and the Tarapur Atomic Power Plant (Ref. 8) (India), where extensive cracking was found in sensitized components during the plant installation phase.
Cracks were found at locations of high residual stresses in most of the Type 304 stainless steel piping sleeves (stub tubes) through which CRDs entered the lower head region of the reactor vessels.
Extensive cracking was also found en many cf the Type 304 stainless steel safe ends attached to the vessel nozzles as well as in several other components.
Although the source of corrodent was not established, chlorides from saltwater mists were 4
60 -
7/24/79 TOTAL 186 l
50 -
49 l
I 40 -
36 d
30 -
8 6
2 22 l
20 -
l 14
_ I 10 -
6 I
3 4
6 8
10 12 14 24 PIPE DI AMETER (INCHES)
Figure 1 Summary of Intergranular Stress Corrosion Cracking Incidents in Boiling Water Reactors, by Pipe Size as of July 24,1979.
r l
l I
5
suspected (these plants are located near oceans).
Residual fluorides from welding fluxes were also suspected since it had been shown (Ref. 9) that
{
fluorides could produce intergranular stress corrosion in heavily sensitized Type 304 stainless steel at room temperature.
IGSCC in furnace-sensitized Type 304 stainless steel piping (safe ends) has occurred in operating plants at locations suspected of having high oxygen content in the coolant.
Such IGSCC has been reported in La Crnsse, Elk River, Humboldt Bay, Nine Mile Point, and Dresden Unit 2, as well as in several foreign reactors.
Details of the occurrences in the United States are given in Table 2.
Details on foreign plants are not readily available; Table 2.
Intergranular Stress Corrosion Cracking in Furnace-Sensitized Safe Ends in Boiling Water Reactors in the United States Plant Component Pipe size Oyster Creek Control rod housing steel tubes 6 in.
Reactor vessel nozzle safe ends Varied Nine Mile Point Core spray nozzle safe end*
6 in.
La Crosse Feedwater inlet nozzle safe ends 4 in.
Elk River Liquid level nozzle extension 1-1/2 in.
Emergency coolant inlet nozzle 1-1/2 in.
Steam outlet nozzle 10 in.
Humboldt Bay Instrument line 1-1/2 in.
Dresden Unit 2 Core spray 10 in.
Note:
Material was Type 304 stainless steel.
however, IGSCC of furnace-sensitized safe ends has been reported at Japan Power Demonstration Reactor (JPDR), Garigliano, Tarapur 1 and 2, Gemeenschappelijke Kernanenarqiecentrale Nederland NV (GKN), Melenor, and Kerkraftwerk RWE Bayernwerk un'H (KRB-1).
2.1.2.2 Crevice Environment On June 14, 1978, the Duane Arnold Reactor experienced the start of a 3 gpm
.(gallons per minute) leak of primary coolant.
The leak was located in one of the recirculation inlet nozzle safe ends, which are 10 inches in diameter and manufactured from Inconel 600.* The safe end is welded to a thermal sleeve of Inconel 600.
The leaking safe end and seven others were examined "Inconel 600 is the International Nickel Corporation's registered trademark for a nickel chromium alloy.
6
by UT in p' lace and after removal, and all were found to have cracks essentially completely around the circumference.
After completing an analysis of the Duane Arnold incident (Ref.10), it was concluded that crack initiation in the recirculation-inlet-nozzle safe ends resulted from a combination of high residual and operating stresses from the thermal-sleeve-attachment weld, the oxygen in the coolant, and a chemical environment resulting from a crevice formed by the safe-end/ thermal-sleeve-attachment configuration.
The configuration may have acted to increase the applied stresses since a significant discontinuity was present at the weld location.
2.1.3 Miscellaneous Small Lines
An update report (Ref. 6) of IGSCC incidents submitted to the Division of Operating Reactors by GE and dated 8/9/79 listed 14 occurrences in 3-inch lines; of these, six were in CRD and RWCU lines and are considered generic; therefore, they would not fit in the miscellaneous category.
Since 1978, LERs have been subcoded to separate pipe-related events into pipe-size categories ranging from less than 4 inches to 16 inches and more.
In reviewing the LERs for cracking in small pipes (i.e., <4 inches), causes are found to vary from fatigue due to vibration or thermal cycling, erosion-corrosion, and differential thermal expansion.
Vibration-induced cracks appear to dominate small-line cracking and occur in instrument, vent, and drain lines in socket welds at tees and nipples in the 3/4-inch to 1-1/2-inch size range.
Appendix Table A.3 contains a listing of small pipe crack incidents by plant as described in the LERs; in most cases' detection was made through observed leakage during routine plant surveillance.
2.2 Pressurized Water Reactors Unlike BWR pipe cracks, failures in PWR piping have not been observed in primary coolant piping.
Stress corrosion cracks, however, have occurred in austenitic piping containing relatively stagnant boric acid solutions.
These, cracks in steam generator feedwater pipes, and miscellaneous small-line cracks are discussed and summarized in this section.
2.2.1 Fatigue Cracking in Feedwater Piping On May 20, 1979, Indiana and Michigan Power Co. informed NRC of cracking in two feedwater lines in D. C. Cook Unit 2.
Leaking circumferential cracks were identified in the 16-inch-diameter lines in the immediate vicinity of the steam generator nozzles.
Subsequent volumetric examination (radiography) revealed crack indications at similar locations in all feedwater lines of both Units 1 and 2.
As a result of a letter sent to all PWR licensees and the issuance of Bulletin 79-13, inspections have revealed piping cracks, crack-like indications, or fabrication defects requiring repair in the vicinity of the feedwater nozzles at 16 of 25 Westinghouse PWR facilities.
To date, eight Combustion Engineering (CE) and two Babcock & Wilcox (B&W) facilities have been inspected.
The two B&W facilities were found to be
- Small lines, by definition, are less than 4 inches in diameter.
7
free of cracks in the vicinity of the feedwater piping-to-steam generator nozzle weld. Cracks were found in two of the eight CE facilities inspected.
The most severe crack identified was a through-wall crack 3-1/2 inches long at the outer surface in steam generator No.1 at D. C. Cook Unit 2.
This crack was located at a stress riser caused by a discontinuity introduced in the piping elbow from machining the weld end preparation counterbore. Metal-lographic examination of samples removed from the failed component revealed other less severe cracks in the tapered transition section of the elbow.
Also, liquid penetrant (LP) examination of the nozzle bore from the thermal sleeve to the nozzle end revealed light pitting and intermittent linear indications along the circumference of the nozzle bore.
Fractographic and metallographic examinations of the cracked piping removed from the D. C. Cook units revealed the following:
(1) The primary type of crack was a very straight transgranular single crack that had multiple fracture origin sites.
(2) The progression of the primary cracks was intermittent as indicated by the presence of beach marks (crack arrest lines) on the fracture face.
Fatigue striations were observed on the fracture face.
(3) There was little oxidation on the primary cracks indicating that stress played a primary role in propagating the cracks and corrosion played an adjunctive role.
From the metallographic and fractographic analyses, the mode of failure was identified as fatigue assisted by corrosion at the D. C. Cook units.
Similar characteristics have been observed and similar conclusions drawn regarding feedwater piping cracks at the other facilities with deep cracks (Beaver Valley and H. B. Robinson Unit 2) and at the majority of facilities with shallow cracks.
A summary of the cracking experienced in feedwater piping in operating PWRs is presented in Table 3.
Reanalyses of normal pipe system stresses and visual examination of the feedwater piping for evidence of distress, malfunctioning snubbers, or other j
support or restraint problems have not uncovered any significant anomalies that would be expected to cause cracking.
No significant deviations from proper feedwater chemistry have been discovered at the facilities where cracking has occurred.
From the results of data collected by means of additional instrumentation i
installed at facilities where cracking has occurred and thermal hydraulic flow modeling, thermal stratification, and stripping (high frequency oscil-lations) have been verified at low-flow conditions when unheated auxiliary feedwater is used for supplying condensate to the steam generators.
The temperature differences present during stratified flow conditions induce high local stresses in the areas where cracking has occurred.
The mixing of hot and cold water as well as the changes in the temperature profiles in the nozzle region are considered to be prime factors in inducing and propa-gating cracks by a thermal fatigue mechanism.
8
Table 3.
Summary of Feedwater Pipe Cracks in Pressurized Water Reactors Extent of cracking (nozzle vicinity)
Locations, No. of Maximum max. depth lines Piping Plant depth crack cracked component Probable cause Comment Westinghouse D. C. Cook Units 1, 2 Through wall Top 8 of 8 Elbow Corrosion-assisted fatigue 2 cracks through wall Beaver Valley 0.400 in.
9 o' clock 3 of 3 Elbow Corrosion-assisted fatigue 13 additional fabrication-related indications repaired Kewaunee 0.050 in.
7 o' clock 2 of 2 Pipe Corrosion assisted fatigue 3-in.-diameter auxiliary feed near SG inlet Point Beach Units 1, 2 0.047 in.
3 o' clock 2 of 2 Reducer Corrosion-assisted fatigue 3-in.-diameter auxiliary feed near SG inlet H. B. Robinson 0.750 in.
9 o' clock 3 of 3 Reducer Corrosion-assisted. fatigue Shallow cracking in nozzle Unit 2 under thermal sleeve Salem Unit 1 0.235 in.
4 of 4
- Elbow, Corrosion-assisted fatigue reducer y3 San Onofre Unit 1 0.100 in.
Lower half 3 of 3 Reducer Stress-assisted corrosion Multiple-branched cracks, of reducer evidence of some fatigue Surry Units 1 and 2 0.080 in.
2 and 6 of 6 Reducer Corrosion-assisted fatigue 5 o' clock Ginna 0.107 in.
8:30 o' clock 2 of 2 Elbow Stress-assisted corrosion /
Cracks also at deep machining corrosion fatigue marks Combustion Engineering Millstone Unit 2 0.250 in.
12 o' clock 2 of 2 Pipe Not analyzed Palisades 0.170 in.
3 and 2 of 2 Pipe Corrosion-assisted fatigue Cracks found also in weld at 9 o' clock vicinity of horizontal piping m.
Preliminary analytical results indicate that usage factors can be greater than unity in relatively short periods of time.
Other influences on the degree of cracking (such as shallow or deep cracking) are significant dif-ferences in the elevated temperature yield strengths of piping from plant to plant.
2.2.2 Intergranular Stress Corrosion Cracking in Piping No occurrences of IGSCC have been reported to date in piping for PWR primary cooiant systems.
Further, IGSCC is not expected to occur in'such piping because the conditions required for stress corrosion cracking are not present.
In particular, dissolved oxygen in the coolant, a major contributor to IGSCC in BWRs, is controlled to levels low enough to preclude IGSCC through the i
use of hydrazine additions and a hydrogen overpressure.
Other impurities i
that might cause stress corrosion cracking, such as halides and caustic, are also rigidly controlled.
Only for brief periods when the reactor is shut down and the coolant is exposed to air and during the subsequent startup are conditions considered even marginally capable of producing stress corrosion cracking in the primary system of PWRs.
Although not widespread, IGSCC has occurred in PWR piping in systems other than the primary coolant piping systems.
As in the BWR pipe cracking, these incidents have generally occurred in the HAZ of welds in austenitic stain-less steel pipe, notably Type 304, where the piping was sensitized by the welding process.
However, IGSCC has also been reported in base material that was otherwise sensitized.
In the PWR cracking, the corrodent has not been positively identified. Water with a high oxygen level as well as contaminants of chlorides, fluorides, caustics, and sulfur compounds are suspected because the cracking has occurred only in piping normally or intermittently containing stagnant water where contaminants may concentrate and become more aggressive. All the cracks occurred in pipes containing borated (H3B04) water with dissolved oxygen at air-saturated levels and low measured chloride levels.
The stresses that contributed to the IGSCC were probably residual stresses resulting from welding or fabrication.
IGSCC was first found in PWRs in 1974 at Arkansas Unit 1 when it resulted in leaks in 10-inch Schedule 10 Type 304 Stainless Steel pipe of the reactor building spray and decay heat removal systems.
In 1976, leaks due to IGSCC occurred in Surry Unit 2 in 10-inch Schedule 40 Type 304 stainless steel pipe of the containment spray system and at R. E. Ginna Unit 1 in 8-inch Schedule 10 Type 304 stainless steel pipe of the safety injection pump suc-tion line.
In view of the apparent generic nature of the problem, NRC/IE issued Bulletin No. 76-06, " Stress Corrosion Cracking in Stagnant, Low Pressure Stainless Piping Containing Boric Acid Solutions at PWRs" (Ref. 11),
requesting that all piping at PWR facilities be examined for this condition.
During this period, IGSCC was discovered at Arkansas Unit 1, R. E. Ginna, H. E. Robinson Unit 2, Crystal River Unit 3, San Onofre Unit 1, and Surry Units 1 and 2.
In early 1979, evidence of leaks (weepage) was observed during a refueling outage in spent fuel pool piping of Three Mile Island Unit 1.
A subsequent 10
visual examination revealed five leaks in 8-inch Schedule 40 Type 304 stain-less steel piping of the spent fuel pool cooling system and one leak in a 10-inch Schedule 40 Type 304 stainless steel pipe in the decay heat removal system.
In metallurgical analyses performed by General Public Utilities (GPU) Service Company on sections of the 8-inch piping, the leaks were attributed to IGSCC initiating on the inside surface of the pipe at the HAZ of welds where the Type 304 stainless steel material is sensitized.
This piping was exposed to boric acid solutions at temperatures averaging 75'F with a peak tempera-ture of 100*F for a period of 54 months.
l In view of the evidence of IGSCC, Metropolitan Edison initiated an extensive UT program of other Type 304 stainless steel piping containing stagnant or i
l near-stagnant boric acid solutions.
These pipes ranged in size from 2-1/2 inches to 24 inches.
Preliminary results using a highly sensitive nondiscriminating technique resulted in 415 UT indications of 1982 welds examined in the spent fuel, decay heat, building spray, and makeup systems.
l A screening procedure developed to distinguish between UT indications of IGSCC and reflections from geometric and other welding-related conditions was subsequently applied; this reduced to 42 the total number of UT indica-tions reportedly due to IGSCC.
A summary of the results of these examina-i tions as of August 28, 1979, is included in Table 4 (Ref. 12).
After removing and evaluating approximately half of the 40 weldments with UT indications from the screening examination, Metropolitan Edison found that many of the UT indications reported as IGSCC were, in fact, reflections l
from geometric discontinuities such as the weld root or machine marks on the inside diameter surface of represented lack of fusion in the weldsents.
Of 13 welds evaluated, 9 (including 7 found by leak detection) had IGSCC and 4 had no IGSCC but did have confirmed geometric discontinuities or weld defects. The remaining weldments continue to be investigated using radio-graphic and ultrasonic techniques to determine the cause of the reported indications.
In view of the TMI-1 findings, IE issued a second bulletin on the subject,Bulletin 79-17, " Pipe Cracks in Stagnant Borated Water Systems at PWR Plants" (Ref. 13), to alert PWR facilities of these findings, requiring that addi-tional examinations, including ultrasonic volumetric examinations, be made for IGSCC cricks on piping systems containing stagnant water. These exami-nations have not been completed.
l To date, pipe cracks of this nature have been reported in eight U.S. PWR l
plants. Table 5 lists these plants along with the systems involved and the l
pipe sizes. All the cracks occurred at locations of stagnant or intermit-tently stagnant, low pressure, low-temperature water containing approximately i
13,000 ppe boric acid and 8 ppm oxygen.
In addition, all the cracks occurred in the HAZ of welds in Type 304 stainless steel piping with relatively high carbon level.
Some indications of cracks found by volumetric examination have been reported in WR piping, but to date have not been confirmed to be IGSCC.
Plants and l
systems having these indications are listed in Table 6.
11 l
Table 4.
Number of Welds, Through-Wall Leaks, and Ultrasonic Testing Indications in Three Mile Island Unit 1 Pipe System as of August 29, 1979 Ultrasonic testing indication Weld Through-Using All UTs wall sensitive
^fter Geometric System Total completed leak tectinique screening reflector Spent fuel 566 566 6
149 22 127 Decay heat 408 408 1
104 11 93 Building spray 241 241 0
64 8
56 Makeup
- 1051 697 0
96 1
95 Core flood 31 31 0
1 0
0 Pressurizer surge 11 11 0
0 0
0 Pressurizer spray 28 28 0
1 0
1 Totals 2336 1982 7
415 42 372
- Does not include 354 2-1/2-inch welds in high radiation areas.
l Table 5.
Systems and Pipe Sizes in Pressurized Water Reactor Plants Reporting Pipe Crack Leaks Plant System Pipe size Arkansas Unit 1 Building (containment spray) 10 in.*
Decay heat removal 10 in.*
Crystal River Unit 3 Containment spray 8 in.**
Ginna Safety injection 8 in.*
H.B. Robinson Unit 2 Boron injection 4-6 in.***
San Onofre Unit 1 Cont'ainment spray 10 in.*
Surry Unit 1 Containment spray 10 in.**
Surry Unit 2 Containment spray 10 in.**
Three Mile Island Unit 1 Spent fuel pool cooling 8 in.**
Decay heat 10 in.**
Material was Schedule 10 Type 304 stainless steel.
- Material was Schedule 40 Type 304 stainless steel.
- Material unknown.
l l
Table 6.
Plants and Systems Reporting Unconfirmed Indications of Cracking as of October 5, 1979 Plant System Three Mile Island Unit 1 Spent fuel cooling Decay heat Building spray Makeup San Onofre Unit 1 Safety injection Mihama Units 1 and 2 Decay heat Takahama Refueling waste storage piping 2.2.3 Miscellaneous Small Lines In searching.the LERs from 1969 to the present for the occurrence of cracks i
in small pipes, one finds that almost every operating reactor is listed as l
having experienceo such cracking.
Cracks were found predominantly in socket welds in the 3/4-inch to 2-inch-diameter pipe-size range in vent, drain, and instrument lines.
Cracks were principally located near pumps where vibra-tions were present.
The mode of failure of the socket-weld cracks has been i
identified as fatigue due to vibration.
Table A.4 of the Appendix summarizes l
small-line pipe cracks by system, as described in LERs, starting in 1969.
13
i 3.
CORRECTIVE ACTION 3.1 Boiling Water Reactors 3.1.1 Piping When it became apparent (1974) that the IGSCC problem was generic in nature, NRC formed a Pipe Crack Study Group (PCSG) in January 1975 to (a) investigate the cause, extent, and safety implication of IGSCC; (b) make recommendations for operating plants; and (c) recommend corrective action to be taken at future plants.
In October 1975, the study group published NUREG-75/067,
" Technical Report Investigation and Evaluation of Cracking in Austenitic Stainless Steel Piping of Boiling Water Reactor Plants" (Ref. 3).
It was generally agreed that the cracking found (October 1975) did not-degrade the pipe to an extent that the piping lost the ability to perform its function.
Furthermore, since the material was Type 304 stainless steel with high inherent toughness, the cracks would not be expected to cause a rapid propagating failure resulting in a loss-of-coolant accident. Also, the leakage that occurred was not significant nor did it release any radia-tion to the environment.
It was also agreed that (a) cracks in these systems can be detected through inservice inspection and leak detection methods before they grow to sizes that might affect safety and (b) the probability is extremely small that the presence of cracks or leaks will present a significant safety hazard to the public.
However, it was also concluded that the presence of these cracks is undesirable and not consistent with NRC general design criteria and that steps should be taken to avoid or minimize such conditions.
It 4
was suggested that this might be accomplished by (a) changing to a material more resistant to stress corrosion cracking, such as clad ferritic steels or the low-carbon grades of austenitic stainless steel; (b) wherever possible, improving the control of water chemistry, particularly dissolved oxygen levels; (c) reducing residual stresses by controlled welding practice and joint designs; (d) reducing cyclic thermal and vibration stresses; aad (e) tightening leak detection restrictions and inservice inspection requirements for pipe weldment examination.
Subsequently (1977), the staff issued an implementing document, NUREG-0313 (Ref. 14), that contained staff positions consistent with the recommenda-tions of the first Pipe Crack Study Group.
I During 1978, the discovery of IGSCC for the first time on large-diameter piping in a BWR in Germany and some concerns about the interpretation of ultrasonic inspections led to the reactivation of a second Pipe Crack Study Group by the NRC.
The study group updated previously reported occurrences and reviewed new IGSCC incidents reported since the preparation of NUREG-75/067 (Ref. 3).
PCSG prepared specific conclusions and recommendations relevant to the current status of IGSCC, the significance of the problem, and the reliability of detection methods and measures available to correct or minimize IGSCC in existing and future plants.
These conclusions and recommendations, published in NUREG-0531, " Investigation and Evaluation of Stress-Corrosion Cracking in Piping of Light Water Reactor Plants" (Ref. 5), were generally consistent with those in NUREG-75/067 and NUREG-0313 and presented ideas and amplifi ation of areas covered in those reports.
The staff has reviewed NUREG-0531 and issued the implementing document, Rev.1 of NUREG-0313, for comment (Ref. 15).
The principal changes include:
14
Extension of the guidelines for reducing the susceptibility to IGSCC to ccver American Society of Mechanical Engineers (ASME) Code Class 2 piping.
Inclusion of augmented inseNice inspection requirements for noncon-forming safe ends.
Update of the inservice inspection sampling schemes to comply with the most recent NRC position.
Identification of NUREG-0531 (1978 PCSG, Ref. 5) recommendations that cannot be implemented immediately without further NRC evaluation.
3.1.2 Safe Ends The problem of IGSCC in furnace-sensitized safe ends was solved by either replacing the pipe with a low carbon stainless steel or Inconel 600 alloy, or by cladding the sensitized material with Type 308L stainless steel or a nickel alloy (Inconel) weld material. The only remaining BWRs with sensi-tized safe ends are at Dresden Unit 2, Big Rock Point, and Nine Mile Point Unit 1.
Augmented inservice inspections are being performed periodically on these safe ends.
l All the safe ends at Duane Arnold were removed and replaced with new safe-end/ttermal-sleeve-attachment configurations made without the long i
narrow crevice at the tip of the attachment weld.
Recommendations were made by the staff that, for operating plants or plants under review for an operating license where thermal-sleeve-attachment configurations with j
crevices are in place and are impractical to remove, an inservice inspection l
program be developed to examine the attachment weld and surrounding material to ensure that cracking would not go undetected during service.
i 3.1.3 Miscellaneous Small Lines All cracking incidents in small BWR lines have been s,ignaled first by a leak.
Subsequent visual inspection has verified i.he presence of cracks.
In the majority of cases where metallurgical evaluations have been performed, the mode of failure was identified as vibration-induced fatigue.
Corrective action has included weld repair and pipe replacement with rerouting of lines and the use of additional or new supports to preclude vibration.
Some piping has also been replaced with thicker sections to prevent cracking failures.
3.2 Pressurized Water Reactor 3.2.1 Fatigue Cracking in Feedwater Piping The utilities with plants affected by feedwater pipe cracks have formed an Owners Group with Westinghouse Corporation to provide support on generic and plant-specific bases to identify the mechanism of cracking, to collect and analyze operating data, and to perform thermal simulations for deter-mining the design and operational changes that may be required to eliminate 15
feedwater pipe cracks. The Owners Group and NRC staff have been meeting monthly to discuss the work performed to date and to exchange information.
The Owners Group has collected and analyzed mechanical and thermal data for transient conditions from startup to full power operation for seven facilities and has performed thermal hydraulic flow tests.
The results show that thermal stratification and stripping at low-flow conditions with auxiliary feedwater was the only unusual mechanical or thermal event that occurred.
Three-dimensional finite element stress analyses performed for changes of the thermal profiles show that large localized stress changes occur with temperature fluctuations.
The balance of the work of the Owners Group, including completing the stress analyses and making proposed design and operational changes, is scheduled for the spring of 1980.
The short-term corrective actions taken at the affected plants have been to replace or repair the cracked piping, using materials similar to the original construction.
In most cases, additional care has been taken to reduce the presence'of stress concentrators by increasing the radii at machining discon-tinuities, using more gradual transitions at section changes, and increasing wall thickness to lower the stresses.
The NRC staff has required that baseline ultrasonic and radiographic inspections be performed following the repair activities.
In addition, at the next refueling outage, the staff has required reinspection of the piping in the nozzle vicinity.
The staff has advised the licensees that other remedial measures and additional augmented inspections may be required at a later date.
3.2.2 Intergranular Stress-Corrosion Cracking in Piping The generic problem of IGSCC in PWRs is still under investigation.
Studies are being funded by the Electric Power Research Institute (EPRI), NRC, and others to investigate the cause of this condition in PWR environments and to improve detection methods.
Further, the examinations of piping required by IE Bulletin No. 79-17 (Ref. 13) are not expected to be completed for another year.
Therefore the total scope of the problem has not yet been i
I defined.
In_the past, the problem was corre'cted on a case-by-case basis.
Failed piping was removed and replaced, generally with a less-susceptible material such as a low-carbon stainless steel. Thin gauge piping was sometimes replaced with thicker schedule piping to improve the residual stress and sensitization effects. More recently, repairs have been made using a combi-nation of low-carbon stainless steel replacement material in conjunction with a corrosion-resistant cladding of the inplace piping to which the welds of the replacement spool pieces are attached.
This approach, which is recom-mended in NUREG-0313 (Ref. 14), would essentially eliminate any sensitization during welding, and thus the materials at the weldments would not be susceptible to IGSCC.
3.2.3 Miscellaneous Small Lines Corrective actions in instances of cracking in small lines have been under-taken on a case-by-case basis. Weld repair or replacement has been the 16 t
predominant fix.
Supports have been added when deemed necessary to reduce or eliminate vibration.
In some instances, lines have been rerouted, nippies removed, and heavier gauge material used as a substitute for the original piping.
4.
SUMMARY
(1) The NRC staff has considered the safety significance of the feedwater pipe cracks and has concluded, based on the available information, that the most severely degraded piping found to date (D. C. Cock Unit 2) would be unlikely to rupture in the event of an earthquake or limited water hammer loads. The staff considers it conceivable that a severely degraded feedwater line may rupture from a severe water hammer event.
At least one pipe break has occurred in the past in feedwater piping (not degraded) as the result of water hammer loads.
However, design changes such as J-tubes have been made and operational changes have been initiated to minimize the probability of water hammer.
It is considered unlikely that a severe water hammer event would occur in more than one feedwater line simultaneously. Thus the worst con-sequence that would be reasonable to expect in a facility with degraded feedwater piping in the event of a severe water hammer would be the rupture of a single feedwater line. This event has been considered as a design basis accident, and the facilities are designed to protect against the occurrence and to contain and control the accident should it occur.
i (2) The IGSCC of Inconel 600 in the Duane Arnold incident was due to the unfavorable stress and environmental conditions considered unique to l
the particular attachment configuration and the associated crevice formed by'this configuration.
Consequently, it is recommended that similar attachr. ant configurations and crevices be avoided.
(3) IGSCC in austenitic stainless steel piping in BWRs is a recurring problem caused by a combination of conditions, including (a) sensiti-l zation of the metal by welding, (b) high residual stresses due to welding, l
and (c) dissolved oxygen in high purity water coolant.
As of October l
1979, 191 cases of IGSCC have been reported.
These havo all occurred l
in the HAZ of welds, primarily in Type 304 stainless steel pipe. A l
Pipe Crack Study Group formed by the NRC to evaluate the problem has issued reports that include discussions of the cause, extent, and safety implications of IGSCL and recommendations for corrective action.
In addition, NRC staff has issued reports implementing the recommenda-tions. These include the Sse of more corrosion-resistant materials for plants in the construction stage and the instituting of augmented inservice inspection in IGSCC-sensitive lines in operating plants.
(4) IGSCC has occurred in BWR safe ends as a result of sensitization of the material during furnace heat treatment of the carbon steel component to which they are attached.
The problem was corrected either by replacing
'the piping with a material more resistant to corrosion or Dy cladding the sensitized material with weld material. Augmented inservice inspec-tion has been instituted for the remaining BWRs with furnace-sensitized safe ends.
17
(5)
In PWRs, IGSCC has occurred in piping other than the primary coolant piping, generally in low-temperature, low pressure, stagnant, or near-stagnant systems.
As in DWRs, cracking has occurred in the HAZ of welds of Type 304 stainless steel. However, the corrodent has not been firmly identified.
High-oxygen-level water'and chloride contaminants are sus-pected.
NRC has issued two bulletins on this problem calling for staff at PWR facilities to examine pipe in stagnant borated water systems for IGSCC (Refs. 11 and 13). To date, IGSCC has been reported in eight U.S.
PWR plants.
The problem is still under investigation.
Corrective action has been taken on a case-by-case basis.
Failed piping has been removed and replaced with more. corrosion-resistant material.
(6) Pipe cracks in miscellar.eous small. lines of both BWRs and PWRs are predominantly the result of vibration / fatigue and are not considered generic. Currently, these types of failures are anticipated to continue to occur randomly throughout plant life.
Repairs and replacement of small-line failures is handled routinely during maintenance shutdowns.
The staff will continue to review LERs on a monthly basis to collect and tabulate small-line cracking incidents to establish if trends occur that are considered safety significant.
REFERENCES 1.
W. B. Smith, Sr., and J. B. Violette, General Electric Co., Atomic Products Energy Division, "The Cause of Failure and the Repair of the VBWR Recirculation Piping," APED-4116, November 12, 1962.*
2.
U.S. Nuclear Regulatory Commission, Office of Inspection and Enforcement, Bulletin:
a.
74-10. " Failures in 4-Inch Bypass Piping at Dresden-2," September 8, 1974.*
b.74-10A, " Failures in 4-Inch Bypass Piping at Dresden-2," December 17, 1974.*
c.74-108, " Failures in 4-Inch Bypass Piping at Dresden-2," January 24, 1975.*
d.
75-01, "Through-Wall Cracks in Core Sprayed Piping at Dresden-2 "
January 30, 1975.*
e.75-01A, "Through-Wall Cracks in Core Spray Piping at Dresden-2,"
February 7, 1975.*
3.
U.S. Nuclear Regulatory Commission, " Technical Report Investigation and Evaluation of Cracking in Austenitic Stainless Steel Piping of Boiling Water Reactor Plants," USNRC Report NUREG-75/067, October 1975.**
4.
H. H. Klepfer et al., General Electric Co., Nuclear Energy Division,
" Investigation of Cause of Cracking in Austenitic Stainless Steel Fiping," NED0-21000-1, July 1975.*
' See footnotes at end of references 18
5.
U.S. Nuclear Regulatory Comission, " Investigation and Evaluation of Stress-Corrosion in Piping of Light Water Reactor Plants," USNRC Report NUREG-0531, February 1979.**
6.
Letter from H. T. Watanabe, General Electric Co., to V. Noonan, NRC,
Subject:
Piping improvement Program, dated August 9, 1979.*
7.
Jersey Central Power and Light Co., "0yster Creek Nuclear Power Plant No. 1, Report on Reactor Vessel Repair Program," Amendments No. 35, 37, 39, 40, and 43, U.S. Atomic Energy Commission Docket No. 50-219.*
8.
E. J. Corr and A. M. Hubbard, General Electric Co., Atomic Products Energy Division, " Preliminary Report, Cracking in Control Rod Drive Penetrations Reactor Vessels No. 1 and 2, Tarapur Atomic Power Plants,"
APED-5600, March 1968.*
9.
C. T. Ward, D. L. Mathis, and R. W. Staehle, "Intergranular Attack of Sensitized Austenitic Stainless Steel by Water Containing Fluoride Ions," Corrosion 25(9), 394-396 (September 1969).***
10.
J. S. Pevrin and V. Pasupathi, " Interim Report on Examination of Inconel Safe End From Duane Arnold," prepared for NRC by Battelle-Columbus, November 8, 1978.*
11.
U.S. Nuclear Regulatory Commission, Office of Inspection and Enforce-ment,Bulletin 76-06, " Stress Corrosion Cracking in Stagnant, Low Pressure Stainless Steel Piping Containing Boric Acid Solutions at PWRs," November 26, 1976.*
12.
D. M. Smith, Sr. et al., General Public Utilities, " Investigation of Intergranular Stress Corrosion Cracking at Three Mile Island Unit 1,"
GPU Report, January 4, 1980.*
13.
U.S. Nuclear Regulatory Commission, Office of Inspection and Enforcement,Bulletin 79-17 Rev. 1, " Pipe Cracks in Stagnant Borated Water Systems at PWR Plants," October 29, 1979.*
14.
U.S. Nuclear Regulatory Commission, " Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary l
Piping," USNRC Report NUREG-0313, July 1977.**
15.
U.S. Nuclear Regulatory Commission, " Technical Report on Material Selection and Processing Guidelines for BWR Coolant Pressure Boundary Piping," USNRC Report NUREG-0313, Rev. 1 (for comment), October 1979.**
Available in NRC PDR for inspection and copying for a fee.
The Public Document Room is located at 1717 H St.,
N.W., Washington, D.C. 20555.
- Available for purchase from the NRC/GPO Sales Program, U.S. Nuclear Regulatory Commission, Washington, D.C. 20555, and/or the National Technical Information Service, Springfield, Virginia 22161.
- Available from public technical libraries.
19
APPENDIX
SUMMARY
TABLES OF PIPE FAILURE i
l
{
l
e Table A.1 Summary of Intergranular Stress Corrosion Pipe Failures in Boiling Water Reactors, Through July 1975 i
Material other Weld zone Generator than Type cracking Size first 304 stain-Plant (MWe) synchronized less steel Incidents Lines
)
Experimental BWR 4.5 1957 Vallecitos BWR 5
10/15/57 Dresden Unit 1 200 4/15/60 24 9
Kahl 15 6/17/61 347 Big Rock Point Unit 1 72 12/8/62 316 Humboldt Bay Unit 3 70 4/18/63 Elk River 22 8/24/63 JPOR 11 10/26/63 1
1 Garigliano 150 1/23/64 KR8 (Gundreamingen) 237 11/12/66 La Crosse BWR 52.4 4/26/68 Lingen KWL 240
-5/20/68 GKN 48 10/26/68 316 Tarapur Unit 1 190 4/1/69 1
1 Tarapur Unit 2 190 5/7/69 Oyster Creek Unit 1 560 9/23/69 316 Nine Mile Point Unit 1 500 11/10/69 1
1 Tsuruga Unit 1 340 11/16/69 1
1 Dresden Unit 2 809 4/13/70 12 4
Fukushima Unit 1 440 11/17/70 8
3 Millstone Point Unit 1 652 11/29/70 3
2 NUCLENOR 440 3/2/71 Monticello Unit 1 545
-3/5/71 3
2 BKW (Huehleberg) 306 6/24/71 Dresden Unit 1 809 7/22/71 wurgassen 612 12/18/71 Oskarshamn Unit 1 440 1/72 Quad Cities Unit 1 809 4/12/72 4
2 Quad Cities Unit 2 809 5/23/72 4
2 Pilgrim 1 652 7/19/72 Vermont Yankee 540 9/20/72 Browns Ferry Unit 1 1075 10/15/73 Ringhals Unit 1 >
750 11/73 Shimane 442 12/2/73 Fukushima Unit 2 760 12/24/71 Peach Bottom Unit 2 1075 2/16//4 Cooper Unit 1 778 5/10/74 l
Duane Arnold 550 5/19/74 l
Oskarshamn Unit 2 580 8/74 l
Hamaoka Unit 1 513 8/1/74 1
1 Browns Ferry Unit 2 1075 8/28/74 Peach Bottom Unit 3 1075 9/1/74 1
1 Fukushima Unit 3 754 10/26/74 Hatch Unit 1 754 11/11/74 FitzPatrick Unit 1 786 2/1/75 Totals 64 30 l-l 22 l-
Table A.2 Summary of Intergranular Stress Corrosion Pipe Failures in Boiling Water Reactors, August 1975 Through July 1979 Detection Plant Date Component / system method
- Foreign 7/75 2-in. instrument line to Visual nozzle safe end Vermont Yankee 8/26/75 1-in. instrument line Leak Dresden Unit 1 9/1/75 6-in. pipe, cleanup Leak demineralizer and unloading heat exchanger return line Quad Cities Unit 2 10/3/75 4-in. pipe; loop 8 recir-UT culating bypass Foreign 11/75 3-in. reactor water cleanup Leak Foreign 11/14/75 8-in. core spray loop B UT Foreign 11/27/75 3-in. reactor water cleanup Leak Nine Mile Point 11/29/75 6-in. pipe, reactor water Leak cleanup Foreign 12/75 8-in. reactor water cleanup Hatch Unit 1 12/13/75 4-in. pipe; loop A recircu-UT lating bypass Cooper 1/13/76 2-in. instrument line Leak Quad Cities Unit 1 1/15/76 4-in. pipe; loop A recir-UT culating bypass 4-in. pipe; loop B recir-culating bypass Pilgrim 1/27/76 4-in. pipe; loop A recir-UT culating bypass 4-in. pipe; loop B recir-culating bypass l
Peach Bottom 3/14/76 1-in. instrument line Leak Unit 3 Humboldt Bay 3/17/76 2-in. vessel bottom drain Leak line; reactor water cleanup Nine Mile Point 3/19/76 6-in. pipe; reactor water Leak l
cleanup l
Nine Mile Point 3/22/76 6-in. pipe; reactor water Leak l
cleanup l
Nine Mile Point 3/24/76 5 ultrasonic indications in UT l
6-in. pipe (4 spools); reactor water cleanup **
2 ultrasonic indications in 2-in. pipe; headspray and head vent **
Brunswick Unit 2 3/24/76 4-in. pipe; loop A recir-UT culating bypass 4-in. pipe; loop B recir-UT culating bypass 4-in. pipe; loop B recir-UT culating bypass See footnotes at end of table.
23
1 Table A.2 (Continued)
Detection Plant Date Component / system method
- Foreign 7/76 10-in. loop A core spray UT Foreign 9/76 2-in. instrument line to Visual nozzle safe end Foreign 9/76 8-in. reactor water cleanup Pilgrim 9/5/76 4-in. elbow; reactor water Leak cleanup regenerative heat exchanger piping Peach Bottom 1/21/77 12-in. pipe; loop A core spray UT Unit 3 12-in. pipe; loop A core spray UT 12-in. pipe; loop A core spray UT 12-in. pipe; loop A core spray UT Foreign 3/77 12-in. recirculating jet pump UT riser 12-in. recirculating jet pump UT riser Foreign 3/77 10-in. recirculating jet pump Leak riser Foreign 3/77 14-in. shutdown cooling tie-in UT to recirculatOg system 12-in. recirculating jet pump UT riser Foreign 3/77 4-in. recirculating bypass PT weldolet loop A Hatch Unit 1 4/77 4-in. recirculating bypass PT weldolet loop B Foreign 4/77 10-in. recirculating jet pump UT pump riser Foreign 4/77 12-in, shutdown cooling inlet UT Quad Cities Unit J 4/15/77 4-in, recirculating bypass UT capped 16-in.-long stub Foreign 5/77 3-in. control rod drive return Nine Mile Point 5/15/77 10-in. isolation condenser PT safe-end-to pipe weld Nine Mile Point 7/1/77 10-in. isolation condenser Leak return Vermont Yankee 8/25/77 8-in. core spray loop A UT 8-in. core spray loop A UT 8-in. core spray loop A UT 8-in, core spray loop A UT 8-in. core spray loop B UT 8-in. core spray loop B UT 8-in. core spray loop B UT 8-in, core spray loop B UT 8-in. core spray loop B UT See footnotes at end of table.
24 i
l Table A.2 (Continued)
Detection Plant Date Component / system method
- Brunswick Unit 2 10/5/77 10-in. core spray Leak 10-in. core spray Leak Foreign 12/77 24-in. secondary steam UT generator loop; 6 HAZs Foreign 12/77 6-in. reactor water cleanup /
UT cooling tee Foreign 12/77 6-in. reactor water cleanup /
UT cooling tee Foreign 12/77 6-in. reactor water cleanup /
UT cooling tee Foreign 12/77 6-in. reactor water cleanup /
UT cooling tee Foreign 3/78 10-in. recirculating riser UT Foreign 3/78 10-in. recirculating riser UT Quad Cities Unit 2 3/5/78 4-in. head vent Leak Dresden Unit 3 3/9/78 3-in. control rod drive return UT Duane Arnold 3/25/78 4-in. reactor water cleanup Leak Hatch Unit 1 3/29/78 6-in. reactor water cleanup Leak Quad Cities 5/26/78 4-in. reactor water cleanup Leak Duane Arnold 12/15/78 4-in. reactor water cleanup UT Duane Arnold 12/15/78 4-in. reactor water cleanup UT elbow Duane Arnold 12/15/78 4-in. reactor water cleanup UT flow orifice; 2 HAZs Millstone 1/5/79 8-in. reactor water cleanup UT Vermont Yankee 1/10/79 2-in decontamination cleanup Leak I
flange Brunswick Unit 1 1/15/79 10-in. core spray loop A; UT 13 HAZs Brunswick Unit 1 1/15/79 10-in. core spray loop B; UT 10 HAZs Nine Mile Point 3/2/79 6-in. reactor water cleanup **
UT Duane Arnold 3/26/79 4-in. reactor water cleanup Leak Quad Cities Unit 2 4/5/79 4-in. reactor water cleanup Leak Quad Cities Unit 2 4/27/79 4-in. reactor water cleanup Leak Peach Bottom Unit 2 7/3/79 6-in. reactor water cleanup tee Leak Peach Bottom Unit 2 7/3/79 6-in. reactor water cleanup tee Leak 6-in. reactor water cleanup UT pipe
" A blank space in this column indicates that detection method was unreported.
- Cracking occurred in cold bends.
Notes: Material in all piping failures was Type 304 stainless steel.
Cracking occurred in the HAZ, except where otherwise noted.
UT = ultrasonic testing; PT = penetrant test; visual = visual examination.
25
Table A.3 Summary of Cracking in Bolling Water Reactors in Piping Smaller Than 4 Inches by Plant as. Described in the Licensee Event Reports Detection
/
Pfant System Pipe size method Description Big Rock Point Reactivity control 3/4-in. line Leak Pipe-to-valve butt (vibration)
Reactor coolant cleanup 3/4-in. vent pipe Leak 3/4-in. line to 3-in. pipe section weld, intergranular
~ stress corrosion cracking Reactor coolant cleanup 1-in.. drain line UT 1-in. Ilne to 3-in. line junctton Browns Ferry Unit 1 Coolant recirculation 1-in. recirculation Leak Line improperly supported; vibration-falled at toe system line of f,ilet weld i
Browns ferry Unit 2 Coolant recirculation system 1-in, instrument line Leak Instrument vibration / fatigue 3/4-in. bonnet vent Leak Vent line/ differential thermal expansion line Dresden Unit 1 Liquid radwaste management 3-in. line Leak At weld of recirculation line corrosion-erosion Dresden Unit 2 Control rod drive return (4-in line UT Thermal sleeve flange, thermal fatigue Control rod drive return
' 4-in. line Ui Safe end to pipe weld, thermal fatigue Dresden Unit 3 Control rod drive nozzle (4-in. line Visual. Control rod drive retainer with cracks by.
ro crack thermal fatigue C4 Control rod drive return
<4-in. line UT Safe end to pipe weld, thermal fatigue FitzPatrick Unit i Emergency core cooling Stub pipe Leak Crack in heat-af fected zone of weld joining stub pipe to first drain valve, vibration Main steam isolation 3/4-in. pipe Visual Crack in weld in pipe upstream of integrated leak test rate connection of residual heat removal I. Hatch Unit 1 Emergency core cooling 3/4-in. pipe section Leak Circumferential crack above weld between 3/4-in.
pipe and shop welded half coupling Humboldt Bay Reactor trip system 1-1/2-in. pipe UT Crack in U-bend of I-1/2-in. pipe in reactor water level instrument system Monticello Unit 1 Condensate and feedwater 3/4-in. vent connection Visual Butt weld tee cracked at 3/4-in branch connection, fatigue Condensate and feedwater 1-in. drain line Leak Poor weld cracked at 1-in branch connection, to moisture separator Monticello Unit 1 Residual heat removal 2-in. branch connection Leak Crack in residual heat removal line at boss connection, vibration Emergency core cooling 1-in, drain line leak Steam Erosion from flow of steam and water through leak 45" elbow leak valve Turbine bypass 1-in. drain line Steam Erosion from flow of steam and water through 90* elbow' leak restricted orifice
Table A-3 (Continued)
Detection Plant System Pipe size method Description Peach Botton Unit 2 Residual heat removal 1-in. line Leak Vibration cracked weld on 1-in. line to relief valve Residual heat removal line 1-in. vent Leak' Vibration cracked weld on 1-in. vent line Peach Botton Unit 3 Reactivity control (control Nozzle PT Cracks in nozzle blend radii, by thermal cycles' rod drive return nozzle)
Pilgrim Unit 1 Spent fuel pool
<4-in.
Leak Fatigue at pump discharge tieline near weldolet Quad Cities Unit 1 Coolant recirculation Bypass stub piece UT Intergranular stress assisted corrosion cracks of upper stub Residual heat removal 3/4-in. vent line Leak Crack in sockolet-to-vent line weld Quad Cities Unit 2 Reactor core isolation 3/4 in. drain equalizer Leak Fatigue failure cracking in nipples 3/4-in. pressure test Leak connection Reactor coolant cleanup 3/4-in. test connection Leak Vibration stress weld failure in heat-affected zone Reactor coolant cleanup
<4-in.
Leak Cleanup pump junction line had 3/4-in.-long visual crack Vermont Yankee Reactivity control Control rod drive Leak Transgranular stress corrosion crack hydraulic return crack line 3-in.
Coolant recirculation 1-in. instrument line Leak Crack in instrument line to jet pump riser Reactor coolant cleanup 2-in. line Leak Intergranular stress corrosion cracking in suction piping to T-weld crack
" Material unknown.
Notes: Material is stainless steel, except where otherwise noted; PT = penetrant test.
l
Table A.4 Summary of Intergranular Stress Corrosion Cracking in Pressurized Water Reactors in Piping Smaller Than 4 Inches, by System, as Described in Licensee Event Reports Starting in 1969 Licensee event report Usual crack Probable System Plant citations location cause Chemical Arkansas Unit 2 1
Near pumps Fatigue and control Calvert Cliffs Unit 1 4
caused by volume Calvert Cliffs Unit 2 5
vibration Haddam Neck 2
Fort Calhoun Unit 1 2
Indian Point Unit 2 3
Indian Point Unit 3 1
Kewaunee Unit 1 2
North Anna Unit 1 1
Palisades Unit 1 1
Point Beach Unit 2 1
R. E. Ginna Unit 1 3
Salem Unit 1 1
Surry Unit 1 1
Turkey Point Unit 3 1
Turkey Point Unit 4 3
Yankee Rowe 4
Zion Unit 1 1
Zion Unit 2 1
Coolant Arkansas Unit 2 1
Welds at Vibration recircu-Calvert Cliffs Unit 2 3
small tees, lation Fort Calhoun Unit 1 1
- nipples, Palisades Unit 1 2
etc.
Point Beach Unit 2 1
Salem Unit 1 1
Three Mile Island 1
Unit 1 Indian Point Unit 1 1
San Onofre Unit 1 1
Residual Arkansas Unit 1 3
Welds Vibration heat D. C. Cook Unit 1 2
near pumps removal Indian Point Unit 2 3
or valves Prairie Island Unit 1 1
Three Mile Island 1
Unit 2 Three Mile Island 1
Unit 1 Note:
Leakage occurred in all incidents.
28
Table A-4 (Continued)
Licensee event report Usual crack Probable System Plant citations location cause Reactor Calvert Cliffs Unit 1 3
Welds Fatigue coolant Calvert Cliffs Unit 2 3
in lines caused by cleanup Kewaunee Unit 1 1
near pumps vibration Trojan Unit 1 1
Yankee Rowe 1
Trojan Unit 1 Emergency-Arkansas Unit 2 1
Welds of Vibration core Beaver Valley Unit 1 1
vent or cooling Calvert Cliffs Unit 2 1
drain lines Farley Unit 1 1
Millstone Unit 2 1
Oconee Unit 2 1
Main North Anna Unit 1 1
Welds in Not deter-
-> steam supply instrument mined lines Condensate Three Mile Island 1
Socket weld V-bration feedwater Unit 1 Other engi-Turkey Point Unit 3 1
Drain line Vibration neered safety weld features Reactor core Oconee Unit 3 1
Weld in Vibration isolation sample line cooling Spent fuel Arkansas Unit 1 2
Welds Not deter-pool mined inter-Three Mile Island 1
HAZ granular Unit 1 stress corrosion cracking Containment Indian Point Unit 2 1
Vent-to-Vibration heat removal pump weld Note:
Leakage occurred in all incidents.
l 29
GLOSSARY ASME American Society of Mechanical Engineers BKW Bernische Kraftwerke AG B&W Babcock & Wilcox BWR boiling water reactor CE Combustion Engineering CR0 control rod drive EPRI Electric Power Research Institute GE General Electric Co.
GKN Gemeenschappelijke Kernanenergiecentrale Nederland NV GPU General Public Utilities Service Co.
HAZ heat-affected zone HX heat exchanger IE Office of Inspection and Enforcement IGSCC intergranular stress corrosion cracking ILRi integrated leak rate test (connection)
JPDR Japan Power Demonstration Reactor KRB Kernkraftwerk RWE-Bayernwerk (Gundremmingen)
KWL Kernkraftwerk Lingen GmbH LER licensee event report LWR light-water reactor NRC U.S. Nuclear Regulatory Commission PCSG Pipe Crack Study Group PT penetrant testing PWR pressurized water reactor RHR residual heat removal RWCU reactor water cleanup TMI-l Three Mile Island Unit 1 UT ultrasonic testing 31 n.
[ U""
U.S. NUCLEAR REGULATORY COMMISSION
~
BIBLIOGRAPHIC DATA SHEET
~
- 4. TITLE AN D SUBTITLE (Add Vobme Na, if oprepneer)
- 2. (tes,e aimkJ Pipe Cracking Experience in Light-Water Reactors
- 3. RECIPIENT S ACCESSION NO.
- 7. AUTHOR (S)
- 5. DATE REPORT COMPLETED L. Frank, W. S. HazeltUn, R. A. Hermann, V. S.
uOs m lvE4R V. S. Noonan, A. Taboada June 1980
- 9. PERFORMING ORGANIZATION NAME AND MAILING ADDRESS (Include Zep Codel DATE REPORT ISSUED 4ugust 19b UoS. Nuclear Regulatory Comission Office of Nuclear Reactor Regulation 8 **"* "" * #
Washington, D.C. 20555
- 8. (Leave omk)
- 13. SPONSORING ORGANIZATION NAME AND MAILING ADDRESS (/ne/ude Zip code /
Same as above
- 11. CONTRACT No.
Technical 1967 through 1979
- 15. SUPPLEMENTARY NOTES
- 14. (Leave umkJ
- 16. ABSTR ACT Q00 words or less)
Comercial light-water reactors have experienced pipe cracking since '1966. This report summarizes pipe cracking experience in light-water reactors as reported in Licensee Event Reports from 1967'through 1979, othi-r licensee and vendor reports, and Office of Inspection and Enforcement Bulletins.
Pipe cracks which were environmentally induced, such as stress corrosion cracking of metal sensitized by welding and heat treatment, were most prevalent.
Feedwater pipes experienced fatigue cracking from thermal stress and many small lines developed leaks as a result of fatigue caused by vibration.
4 Cracking incidents are separated into generic categories and listed by reactor type pipe size, and systems affected, l
i i
- 17. CIEY WORDS AND DOCUMENT AN ALYSIS 17a DESCRIPTORS Piping, safe ends, pipe cracks, leaks, Pipe cracking experience, stress corrosion cracking light-water reactors, LERs 17b. IDENTIFIERS /OPEN-ENDE D TERMS
- 18. AVAILASILITY STATEMENT
- 19. SECURITY CLASS (This reportl
- 21. NO. OF PAGES Unlimited availability jndass 22NRiCE gT C (rai,,.,,i l'ncla ssified s
NRC FORM 335 (7.77)