Boric Acid Corrosion Evaluation

REPORT SUMMARY
Boric Acid Corrosion Evaluation (BACE) Program

Phase l-Task 1 Report
Industry Experience Reference and Available Test Data Summary
Boric acid corrosion (BAC) represents a significant maintenance issue
affecting primary systems at most PWR plants. This report serves as
a continuation of previous EPRI efforts to provide nuclear plant per-
sonnel with topical information that improves their ability to evaluate
and predict BAC degradation of carbon and low-alloy steel fasteners
and pressure boundary components.
BACKGROUND In 1988, NRC issued Generic Letter 88-05 requesting that utili-
INTEREST CATEGORIES ties implement programs to address the BAC issue. Due to the large number of
BAC degradation incidents and NRC requirements for engineering evaluations of
BAC leaks, the nuclear steam supply system (NSSS) vendors and owners groups,
Nuclear plant operations NRC, and EPRI supported and/or performed a variety of BAC test programs.
and maintenance Primary program objectives focused on establishing corrosion rates for the differing
Nuclear plant corrosion geometry, temperature, pH, concentration, and leakage mode associated with vari-
control ous components or materials. Duquesne Light Company, Duke Power Company,
Nuclear component and Pacific Gas and Electric Company cofunded the work described in this report.
reliability
OBJECTIVES To provide PWR plant personnel with a comprehensive guide to
KEYWORDS industry experience with BAC; to compile a complete bibliography of available
literature in this topical area.
Corrosion
Boric acid
APPROACH The project team surveyed personnel at all U.S. PWR nuclear power
PWR plants to determine their experience with BAC incidents between 1988 and 1992.
Fasteners They augmented information from survey results by conducting follow-up interviews
Reactor components with plant personnel. The literature review included licensee event reports (LERs)
and other NRC documentation, NSSS owners group research results, previous
EPRI reports, and a query of numerous technical computer databases.
RESULTS The survey results indicated that most leaks were small and occurred
principally through valve packing and valve body-to-bonnet gaskets. Survey re-
sponses and follow-up discussions identified fastener upgrade programs, valve
packing design improvements, coating improvements, and walk-down inspections
as the principal means of reducing major plant BAC events.
Analysis of previous industry BAC testing revealed wide-ranging corrosion rates
under varying circumstances, from minimal to approximately 7 in per year. A diffi-
culty also arose in correlating data between programs. This report provides brief
descriptions of previous test programs and their data, which are divided into three
categories: (1 ) BAC-related degradation rates in carbon and low-alloy steel fasten-
ers, (2) BAC-related degradation rates in carbon and low-alloy pressure vessel
steels and, (3) boric acid crevice corrosion in pressure boundary components.
Appendix A summarizes the results of the BAC literature survey. This survey in-
cluded a database search of NRC Document Room, Idaho National Engineering
EPRI TR-101108s Electric Power Research Institute

Laboratory Technical Library, Massachusetts Institute of Technology (MIT)
Engineering Library, Nuclear Operations and Maintenance Information
Service (NOMIS), NRC publications and LERs, technical journals, and
contacts with owners group representatives.
EPRI PERSPECTIVE A future objective of the BAC Evaluation Program

is to perform real-time, component-based boric acid corrosion testing. The
resultant test data will be used to develop a method for accurate evalua-
tion of BAC incidents and to realistically predict in situ corrosion rates for
components and materials. Related EPRI research includes reports
NP-3784, NP-5769, NP-5985, and NP-6536.
PROJECT
RP2814-81
Project Manager: J. M. Jenco
Nuclear Power Division/Nuclear Maintenance Applications Center
Contractor: Altran Corporation
For further information on EPRI research programs, call

EPRI Technical Information Specialists (415) 855-2411.
Boric Acid Corrosion Evaluation (BACE) Program
Phase l-Task 1 Report
Industry Experience Reference and Available Test Data Summary
TR-101108
Research Project 2814-81
Final Report, December 1993
Effective December 6, 2006, this report has been made publicly available in accordance
with Section 734.3(b)(3) and published in accordance with Section 734.7 of the U.S. Export
Administration Regulations. As a result of this publication, this report is subject to only
copyright protection and does not require any license agreement from EPRI. This notice
supersedes the export control restrictions and any proprietary licensed material notices
embedded in the document prior to publication.
Prepared by
ALTRAN CORPORATION
200 High Street
Boston, Massachusetts 02110
Principal Investigators
T. Esselman
P. K. Shah
Prepared for
Nuclear Maintenance Applications Center
1300 Harris Boulevard
Charlotte, North Carolina 28262
Operated by
Electric Power Research Institute
3412 Hiliview Avenue
Palo Alto, California 94304
EPRI Project Manager

J. M. Jenco
Nuclear Power Division

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS REPORT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR
COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY
COSPONSOR, THE ORGANIZATION(S) NAMED BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF
ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS REPORT, INCLUDING
MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR
INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS
REPORT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL
DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES)
RESULTING FROM YOUR SELECTION OR USE OF THIS REPORT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR
SIMILAR ITEM DISCLOSED IN THIS REPORT.
ORGANIZATION(S) THAT PREPARED THIS REPORT:
ALTRAN CORPORATION
Price: $2500.00
ORDERING INFORMATION
Requests for copies of this report should be directed to the Nuclear Maintenance Applications Center
(NMAC), 1300 Harris Blvd., Charlotte, NC 28262, (800) 356-X48. There is no charge for reports requested
by EPRI member utilities.
Electric Power Research Institute and EPRI are registered service marks of Electric Power Research Institute, Inc.
Copyright © 1993 Electric Power Research Institute, Inc. All rights reserved.
EPRI Licensed Material
Boric Acid Corrosion Evaluation Program
Abstract
Boric Acid Corrosion (BAC) is a major maintenance issue for PWR power plant systems.
Leakage from piping and pressure vessels has resulted in varying degrees of external
corrosion of PWR components.
This report is a continuation of EPRI/NMAC efforts to evaluate and predict potential

degradation of carbon and low-allow steel fasteners and pressure boundary components
related to BAC. It presents an overview of industry experience in the area of boric acid
corrosion derived from utility survey responses, industry owners groups, government
reporting requirements, and previous EPRI research projects. The report provides a
characterization of various boric acid corrosion mechanisms found in nuclear power plants,
summarizes the results of previous industry boric acid testing programs, and includes a
current bibliography of related literature.
The guide provides plant personnel with a comprehensive reference to industry experience
in the area of boric acid corrosion. It assists plant personnel in characterizing instances of
boric acid corrosion by providing a uniform description of boric acid corrosion mechanisms,
and reduces personnel time required for assessment of boric acid information by presenting
a current bibliography of available information on the subject. In addition, it assists
personnel in evaluating the impact of boric acid corrosion by providing a summarization of
previous testing data in the area.
iii
Acknowledgments
This guide was prepared for EPRI's Nuclear Maintenance Applications Center under
project number 2814-81. As with the preparation of any published document, numerous
groups and individuals contributed significantly of their time and expertise. Their
important contributions included suggestions for improved document organization,
guidance in development direction, and technical as well as editorial peer review.
We would also like to acknowledge the time and effort provided by member utilities in
responding to surveys covering their boric acid corrosion experience in recent years. Many
utilities provided invaluable information, and through clarifications provided during
follow-up interviews, helped to establish an accurate perspective of current plant boric acid
corrosion problems and their approaches to prevention.
The following individuals and utilities deserve special recognition for their key role in
providing technical guidance and the core utility support for performance of this important
co-funded effort:
Mr. Dennis P. Weakland Duquesne Light Co.
Mr. David E. Whitaker Duke Power Co.
Mr. Henry J. Thailer Pacific Gas & Electric Co.
Additionally, we would especially like to thank the following individuals for their technical
assistance during the course of the project:
Mr. Steve Fyfitch B&W Nuclear Services
Mr. John F. Hall ABB-CE Nuclear Services
Mr. Dave Miskiewicz Florida Power Corporation
Mr. Raj Pathania EPRI
v
-
Table of Contents
1.0 Introduction 1-1
2.0 Industry Experience Summary 2-1

2.1 Utility Experience ................................................................................. 2-3
2.1.1 Licensing Event Report (LER) Summaries .................................. 2-3
2.1.2 Utility BAC Experience Survey Results ..................................... 2-11
2.2 NRC Guidance .................................................................................... 2-16
2.3 EPRI Information................................................................................. 2-19
3.0 BAC Test Data Summary 3-1

3.1 Westinghouse Test Data...................................................................... 3-1
3.2 NRC/Brookhaven National Laboratory Test Data .............................. 3-2
3.3 ABB-CE BAC Test Data on Fastener Material .................................... 3-4
3.4 ABB-CE BAC Test Data on Shell Material .......................................... 3-5
3.4.1 Borated Water Autocalve Tests................................................... 3-5
3.4.2 Beaker Tests................................................................................ 3-5
3.4.3 Borated Steam Tests ................................................................... 3-5
3.5 CE Owners Group Sponsored Test Data............................................ 3-9
3.6 B&W Owners Group (B&WOG) Test Data ........................................ 3-11
4.0 Boric Acid Corrosion Mechanisms 4-1

4.1 General Corrosion (Wastage) .............................................................. 4-1
4.2 Galvanic Corrosion............................................................................... 4-2
4.3 Crevice Corrosion - Metal Ion Concentration Cell ............................. 4-2
4.4 Crevice Corrosion - Oxygen Concentration Cell ............................... 4-3
4.5 Pitting .................................................................................................... 4-3
4.6 Intergranular Corrosion........................................................................ 4-3
4.7 Flow Assisted Wastage........................................................................ 4-4
4.8 Summary ............................................................................................... 4-4
vii
Table of Contents (cont.)
5.0 Summary 5-1
6.0 References 6-1
Appendix A-BAC Literature Summary A-1
Appendix B-NRC Licensing Event Reports (LERs) B-1
viii
List of Figures
Figure 2.1 Boric Acid Corrosion Evaluation Survey 2-13
Figure 4.1 Cross Section of an Uncorroded Metal Surface 4-5
Figure 4.2 General Corrosion Process 4-5
Figure 4.3 Galvanic Corrosion 4-6
Figure 4.4 Crevice Corrosion 4-6
Figure 4.5 Oxygen Concentration Cell Corrosion 4-7
Figure 4.6 Pitting Attack 4-7
Figure 4.7 Typical Intergranular Corrosion 4-8
Figure 4.8 Stress Corrosion Cracking 4-8
Figure 5.1 Boric Acid Corrosion Rate vs. Temperature 5-2
Figure 5.2 Boric Acid Corrosion Rate vs. Boric Acid Concentration 5-3
ix
List of Tables
Table 2.1 Bolt Rejections Through 9/30/84 2-1
Table 2.2 Summary of Components Damaged by Boric Acid Corrosion 2-2
Table 2.3 Material Degradaton Incidents-1987-1992 2-15
Table 2.4 Sources of Leaks-1987-1992 2-15
Table 2.5 Leakage Incidents of Borated Water-(1987-1992) 2-15
Table 3.1 Westinghouse BAC Test Data 3-1
Table 3.2 Acid Corrosion Rates in Aqueous Environment 3-3
Table 3.3 ABB-CE Drip Test Results 3-6
Table 3.4 ABB-CE Borated Steam Tests on Fastener Materials 3-7
Table 3.5 ABB-CE Autoclave Tests of Low Alloy Shell Materials at 300°F 3-7
Table 3.6 ABB-CE Short Time Breaker Tests at 220°F 3-8
Table 3.7 ABB-CE Borated Steam Tests Data 3-9
Table 3.8 CE Owners Group Crevice Corrosion Test Data 3-10
Table 3.9 B&WOG Borated Water Tests on Carbon Steels 3-12
xi
1.0 Introduction
Boric Acid Corrosion (BAC) is a major maintenance issue for pressurized-water reactor
(PWR) power plant systems. Since 1968, BAC has resulted in the degradation of ferritic
steels used for closure bolts in manways, valves, pumps, and in some cases, reactor
pressure vessels. BAC has also resulted in the malfunctioning of valves, and in the
degradation of components in CRDM assemblies.
This report serves as a single source document that compiles available BAC test data and
provides a bibliography of BAC topical literature. Follow-on EPRI programs will
supplement the existing industry data to produce a comprehensive, broad-based data
source. This data source, either in report form or as computer software, will allow users to
consider plant-specific conditions related to incidents of boric acid corrosion in accurately
evaluating and predicting the corrosion degradation of affected plant components and
systems.
Major instances of BAC-related degradation have been documented in earlier EPRI reports
and Nuclear Regulatory Commission (NRC) Regulatory Guides. Between 1968 and 1989, a
significant number of degradation events occurred in manway and reactor coolant pump
closure studs, and in body-to-bonnet closures bolts of valves. Small leaks of borated water
from instrument lines, valves, and CRDM assemblies have resulted in degradation of
reactor vessel head, reactor vessel bolts, and nozzles of primary coolant boundaries. These
components were remote from the leak locations and, in some cases, the leaks were
undetected for extended periods.
In order to alert utilities to the degradation and potential failures of critical components
due to BAC, the NRC issued several Information Notices, Bulletins, and Generic Letters
between 1980 and 1988. The NRC also sponsored one of the earlier BAC test programs,
conducted at Brookhaven National Labs. The data from this program was subsequently
published as NUREG-31098 in March 1982. Due to the number of BAC degradation
incidents and NRC requirements for subsequent engineering evaluation of BAC leaks, the
NSSS suppliers, NSSS Owners Groups, NRC, and EPRI supported or performed a variety
of BAC test programs. Primary program objectives were to establish rates of boric acid
corrosion for different conditions of geometry, temperature, boric acid pH, boric acid
concentration, and boric acid flow. Brief descriptions of these test programs and other test
data available in the industry are contained in Section 3.0 of this document. The test data
are in three different categories; (i) BAC related degradation rates in carbon and low alloy
steel fasteners, (ii) BAC related degradation rates in carbon and low alloy pressure vessel
steels and, (iii) boric acid crevice corrosion in pressure boundary components.
Test data was supplemented by surveying individual U.S. PWR nuclear power plants for
their experience with boric acid corrosion over the period 1987 through 1992. Most leaks
noted in the survey results were very small, principally in the area of valve packings and
valve body-to-bonnet gaskets. Utilities responding to the survey and follow-up interviews
attributed the majority of their success in reducing both the quantity and severity of BAC
incidents to fastener upgrade programs, implementation of leak reduction bolting
1-1
programs, improvements to valve packing designs and coatings, and to periodic walkdown
inspections.
This project is a continuation of previous EPRI efforts to provide useful information

relating to assessment of BAC degradation of carbon and low-alloy steel fasteners and
reactor coolant pressure boundary (RCPB) components. Included in this document are
sections summarizing industry experience with BAC, industry BAC component testing
programs and resultant test data, and a discussion of the various corrosion mechanisms
related to BAC. Also included as an appendix, is an updated topical bibliography resulting
from a survey of available literature on the subject.
Section 2.0 of this report presents a summarization of general industry experience with
BAC. This section is subdivided to present information on past NRC, EPRI, and utility
experience with BAC. Utility experience is represented by information taken from
individual licensing event reports (LERs), and supplemented by the survey results related
to this effort. Personnel interested in identifying general types of BAC incidents occurring
at other plants, or those interested in a broad overview of recent BAC incidents, will find
the information contained in this section of value.
Section 3.0 of this report presents a summarization of industry experience with BAC
testing programs, and their resultant test data. The section looks primarily at efforts
undertaken by NRC and NSSS Owners Groups to establish corrosion rates for specific
materials or components, within a narrow range of testing parameters. The information
contained in this section illustrates the range of corrosion rates found in individual test
programs, and the difficulty in correlating test data between these programs. This section
also highlights the complete lack of available test data addressing the effects of BAC on
polymers [cabling], composite materials [seals, gaskets], and structural steels. Personnel
interested in locating specific test data will find the information contained in this section of
value.
Section 4.0 of this report presents a brief discussion of the various corrosion mechanisms
associated with BAC, accompanied by graphics illustrating each mechanism type. These
include general corrosion (wastage), galvanic corrosion, crevice corrosion (metal ion
concentration cell), crevice corrosion (oxygen ion concentration cell), patting, intergranular
corrosion, and flow-assisted wastage. Personnel interested in better understanding
corrosion types, and the relative importance of each BAC mechanism will find the
information contained in this section of value.
Section 5.0 of this report briefly summarizes the overall relationship of industry test data
identified herein, and graphically illustrates those relationships as charted plots of
corrosion rate versus temperature, and corrosion rate versus boric acid concentration.
Personnel interested in broadly evaluating past corrosion test results and relationships by
either of these indices will find the information contained in this section of value.
Section 6.0 of this report presents a listing of all references used in the preparation of this
document. Personnel interested in pursuing specific areas of inquiry discussed in this
report will find the references listed in this section of value.
1-2
Appendix "A" of this report presents a complete bibliography of all references in literature
to BAC in nuclear power plant applications. The bibliography was developed through
queries of numerous databases and other technical sources, and refined to include only
those references dealing specifically with the subject matter of this report. An abstract
overviewing the precise content of each reference is included, where available. Personnel
interested in pursuing further inquiry in this topical area will find the references listed in
this section of value.
Apendix "B" of this report presents full text descriptions of all related Licensing Event
Reports (LERs) to NRC on the subject of BAC, for those reports where full text is available.
These reports are organized in reverse order by LER number. Personnel interested in
evaluating past BAC incidents and responses reported by other utilities will find the
information presented in this section of value.
1-3
2.0 Industry Experience Summary

Various components used in Pressurized Water Reactors (PWRs) such as pressure vessels,
piping, valves, pumps, threaded fasteners and other structural steels are constructed with
carbon and low alloy steels. Reference 15 provides detailed statistics on bolt degradations
through 9/20/84 that are summarized in Table 2.1. In the late 1970's and early 1980's,
numerous occurrences of severe degradation of these alloys by concentrated boric acid
solutions have been recorded [1-5]. These are briefly summarized in Table 2-2. A majority
of these degradation events occurred in bolted fasteners due to primary coolant leaks from
gaskets of flanged connections. However, there are also incidents of degradation of
pressure vessels, piping, and valves due to borated water leaks from instrumentation
fittings, CRDM seals, and degraded valve packing.
Table 2.1
Bolt Rejections Through 9/30/84
-
Component Total Rejects for BAC
Rejects
Bolts Rejected % of Total
Steam Generator Manway 310 116 37
Pressurizer Manway 20 5 25
Reactor Coolant Pump Flange 107 102 95
Valve (6" or larger) 334 264 79
CRDM Flange 392 198 51
2-1
Table 2.2
Summary of Components Damaged by Boric Acid Corrosion
Plant Date Corroded Component Source of Leakage
Connecticut Yankee 1968 Pressurizer Spray Valve Valve Packing Leakage
Palisades 1971 RV Closure Studs RV Head Gasket
Calvert Cliffs 1 1974 RV Closure Studs RV Head Gasket
St. Lucie 1977 SG Manway Closure Studs Gasket Leakage
St. Lucie 1978 Pressurizer Manway Gasket Leakage
Closure Studs
Surry 1978 Valve Valve Packing Leakage
ANO2 1979 SG Manway Closure Studs Gasket Leakage
Zion Unit 1 1979 CVCS Valve Unknown
Fort Calhoun 1980 RCP Studs Gasket Leakage
Calvert Cliffs 1980 RCP Studs Pump Seal
Calvert Cliffs 1980 SG Manway Studs Gasket Leak
Fort Calhoun 1981 RCP Studs Gasket Leakage
Calvert Cliffs 1981 RCP Studs Pump Seal
Calvert Cliffs 1981 Safety Injection Check Valve Gasket Leak
Studs
Calvert Cliffs 1981 Pressurizer Manway Studs Gasket Leak
Calvert Cliffs 1981 RCP Suction Piping Gasket Leak
D.C. Cook 1981 Check Valve Studs Gasket Leak
Indian Point 2 1981 RCP Studs Gasket Leak
1981 Check Valve Studs Gasket Leak
Kewannee 1981 Instrument Isolation Leaking Valve
Valve Bonnet Diaphragm
Oconee 1 2 and 3
, , 1981 RCP Studs Leaking Gasket
ANO 1 1981 SG Manway Closure Studs Gasket Leak
Robinson 2 1981 Pressurizer Spray Valve Studs Valve Packing
Maine Yankee 1982 SG Manway Closure Studs Gasket Leak
North Anna 1 1982 Check Valve Studs Valve Packing Leak
Arkansas 1 1986 High Pressure Injection Gasket Leak
Nozzle and Piping
Turkey Point 4 1987 RV Head, Closure Studs, and Leaking Tube Seal
CRDM
San Onofre 2 1987 Isolation Valve Failure Valve Packing Leak
Salem 2 1987 RV Head Seal Weld Leak m
Instrumentation Penetration
ANO 2 1987 Pressurizer Head Heater Penetration Leak
Millstone 2 1988 RV Nozzles RV "O" Ring Leak
1988 Leaking Containment
San Onofre 2 Stem Nut Threads Isolation Valve 2HV-0509
ANO 1 1989 CRD Housing Nut Rings CRDM Leaking Gasket
Flange
Fort Calhoun 1 1992 Flanges/Piping Supports Boric Acid Pump Flange
Waterford 3 1992 Packing Gland Studs RC Sample Valve-104
Waterford 3 1992 Packing Gland Studs RC Sample Valve-104 (2nd
occurence)
2-2
2.1 Utility Experience
2.1.1 Licensing Event Report (LER) Summaries
To gain a more comprehensive understanding of the nature of leaks and utility responses
to NRC guidance in this area, numerous licensing event reports (LERs) and NRC dockets
were reviewed. The following listing summarizes those documents relating to significant
utility experiences with boric acid corrosion.
1. EVENT DATE: 03/20/92 LER#: 92-018-00 REPORT DATE: 06/22/92
(FULL TEXT AVAILABLE - SEE APPENDIX B)
FACILITY NAME: Fort Calhoun Station Unit No. 1
TITLE: Corrosion of Boric Acid System Bolts
OPERATING MODE: 5 POWER LEVEL: 000
LICENSEE CONTACT FOR THIS LER:

NAME: Scott A Lindquist, Shift Technical Advisor
TELEPHONE: (402) 533-6829
On March 20, 1992, with the plant in a refueling outage, the removal of insulation and
heat tracing from the Boric Acid System during implementation of a modification
revealed severe corrosion of the carbon steel fasteners on the boric acid pump flanges
and piping supports. The corrosion was extensive enough to have led to a possible
failure of the fasteners during a seismic event. This event is being reported pursuant to
10 CFR 60.72(b)(2)(i). This report is being submitted pursuant to 10 CFR 50.73(a)(2)(ii)
and 10 CFR 50.73(a)(2)(vii).
The root cause of this event was the original design of the flange connections did not
anticipate corrosion problems due to boric acid leakage at the system flange
connections. The carbon steel fasteners were covered with glued heat tracing and
asbestos insulation, thus, sealing the fasteners in a potentially high corrosive
environment.
Based on the configuration of the charging header and the availability of operator
actions to mitigate the consequences of a complete loss of Boric Acid Storage Tank
(BAST) inventory, the safety significance of the degraded flanges was minimal. The
postulated loss of a BAST would not significantly affect plant safety.
As corrective action, the carbon steel fasteners were replaced. Long term corrective
actions include evaluating acceptability of replacing carbon steel fasteners with
stainless steel fasteners, incorporating the Boric Acid System into the Boric Acid
Corrosion Program and performing a review of the current program scope to determine
if other carbon steel fasteners are exposed to similar conditions and possible damage.
2-3
2. EVENT DATE: 07/11/92 LER #: 92-006-00 REPORT DATE: 08/07/92
FACILITY NAME: Waterford Steam Electric Station, Unit 3
TITLE: Reactor Shutdown due to Reactor Coolant System Leakage in excess of

Technical Specifications

NAME: J.G. Hoffpauir, Maintenance Superintendent
TELEPHONE: (504) 464-3138
At 0703 on July 11, 1992, Waterford Steam Electric Station Unit 3 declared an Unusual
Event as a result of unidentified reactor coolant system leakage greater than the
Technical Specification 3.4.6.2 limit of one gallon per minute. The reactor was shut
down and the source of the leakage determined to be the packing area of Reactor
Coolant Hot Leg Sample Valve RC-104.
This event resulted from the failure of a temporary leak repair made to RC-104 after
the valve's packing gland studs failed due to boric acid corrosion on March 25, 1992. As
a result, the root cause of this event and the earlier failure are the same: the use of a
material for the studs in an application for which it proved to be inadequate. That is,
the valve was installed in a system where possible packing leakage could expose the
studs to boric acid despite the fact that the studs were made of a material that is
susceptible to boric acid corrosion. RC-104 will be repaired or replaced during the next
refueling outage. Also, the stud material for similar valves will be evaluated. All
leakage resulting from this event was confined to the Reactor Containment Building
and therefore, the health and safety of the public and plant personnel were not
compromised.
3. EVENT DATE: 03/25/92 LER #: 92-002-00 REPORT DATE: 04/24192
FACILITY NAME: Waterford Steam Electric Station Unit 3
TITLE: Reactor Shutdown due to Reactor Coolant System Leakage in Excess of

Technical Specifications
2-4

NAME: W.R. Brian, Plant Engineering Superintendent
TELEPHONE: (504) 464-3127
At 0248 hours on March 25, 1992, Waterford Steam Electric Station Unit 3 declared an
Unusual Event due to Reactor Coolant System unidentified leakage in excess of the
Technical Specification 3.4.6.2 limit of 1 gallon per minute. The reactor was shut down
and the source of the leakage was subsequently determined to be the packing area of
Reactor Coolant Hot Leg Sample Valve RC-104.
The packing gland studs on RC-104 failed due to boric acid corrosion. The root cause of
this event was use of a material in an application for which it proved inadequate; that
is, the studs were made of a material that is susceptible to boric acid corrosion, and the
valve was used in a system where possible packing leakage could expose the studs to
boric acid. RC-104 will be repaired or replaced during the next refueling outage and
the stud material for similar valves will be evaluated. All leakage due to this event was
confined to the Reactor Containment Building, and therefore the health and safety of
the public and plant personnel was not compromised.
FACILITY NAME: San Onofre Nuclear Generating Station, Unit 2
TITLE: Containment Isolation Valve Technical Specification Not Met Due To Failure
of RCS Hot Leg SAMPLE VALVE 2HV-0508

NAME: H. E. Morgan, Station Manager
TELEPHONE: 714-368-6241
On 8/11/88, at 0705, containment isolation valve 2HV-0509 (Reactor Coolant System

(RCS) hot leg sample valve) was opened for RCS sampling and unanticipated RCS
leakage was observed. At 1245, it was determined that the redundant isolation valve
2HV-0508 may not have been fully closed and was thus causing the observed leakage.
In accordance with TS 3.6.3, 2HV-0509 was secured closed. Inspection of 2HV-0508 on
8/15 revealed excessively worn threads on the Limitorque actuator stem nut. This
condition had allowed the rotation of the stem nut without attendant valve stroke,
causing a false "close" position indication in the Control Room. It is postulated that the
valve failure occurred when a delayed close indication was received on 8/8188. Since the
condition of the valve had been unknown, 2HV-0509 was not maintained secured in the
closed position from 8/9 at 1315 to 8/11 at 1245, contrary to TS 3.6.3.
2-5
Valve 2HV-0508 was not equipped with a protective stem cover for its actuator.
Moisture, boric acid and debris had come in contact with the actuator stem which
resulted in abrasive thread surfaces, increasing the rate of wear of the softer stem nut
threads. This condition, along with frequent cycling of the valve, caused the actuator
stem nut threads to wear excessively. Valve 2HV-0508 was returned to service
following repair and the installation of a protective stem cover.
Various Limitorque actuators were previously identified as lacking protective stem

covers. The vendor manual neither provides the part identification nor direction for the
replacement of such covers. Although it was determined that these covers were not a
prerequisite for valve operability, an effort to determine suitable replacements was
undertaken. In May 1988, acceptable replacements for the missing covers were
identified and obtained. At that time, Maintenance procedures were revised to require
installation of such covers following maintenance activities. All safety related
Limitorque actuators identified as lacking a cover have now been corrected.
TITLE: Shutdown Cooling Isolation Valve Packing Gland Leakage

TELEPHONE # 714-368-6241
On August 13, 1987, at approximately 1900, with Unit 2 in Mode 5 and the Reactor
Coolant System (RCS) at approximately 350 psia and 127 degrees Fahrenheit, failure of
alloy steel packing gland follower studs during manual operation of motor operated
Shutdown Cooling System (SDCS) suction isolation valve 2HV-9378 resulted in leakage
estimated at l00 gpm through the packing gland. Operation of the SDCS continued via
a redundant flow path. RCS inventory was maintained by isolating letdown flow and
using charging pumps as Repressurization and venting of the RCS proceeded.
Containment closure was promptly restored and there was no effluent release from
containment above regulatory limits.
At 1100 on September 1, a temporary repair was completed which reduced the leak rate
to approximately 1/4 gpm, effectively terminating the event.
The cause of stud failure has been attributed to (1) packing leakage resulting in
wastage due to boric acid corrosion, (2) decrease in lubricating characteristics and
hardening of packing, and (3) the initial thrust required to open the valve. Corrective
2-6
actions include reduction of the maximum thrust necessary to open the valve by
installation of a modified packing gland assembly less susceptible to leakage and
hardening of packing, and replacement of packing gland studs with corrosion resistent
material.
The health and safety of plant personnel and the public were not affected by this event.
6. EVENTDATE:04/24/87 LER#: 87-003-01 REPORT DATE: 08/12/88
FACILITY NAME: Arkansas Nuclear One, Unit Two
TITLE: Pressurizer Heater Rupture Due to Primary Water Stress Corrosion Cracking
Results in Reactor Coolant Pressure Boundary Leakage and Pressurizer Vessel
Corrosion

NAME: Larry A. Taylor, Nuclear Safety and Licensing Specialist
TELEPHONE if- 501-964-3100
On 4/24187, an Unusual Event was declared and a reactor shutdown was commenced
due to a suspected reactor coolant system (RCS) pressure boundary leak of
approximately 60 drops per minute from the area of the pressurizer vessel lower head.
Subsequent investigation revealed that two pressurizer heaters manufactured by
Watlow Electric Company had ruptured resulting in damage to the heater sleeves
(which penetrate the vessel head and house the heaters). Damage to one heater sleeve
was sufficient to cause RCS pressure boundary leakage which resulted in a small area
of boric acid induced corrosion damage to the pressurizer carbon steel base metal.
Metallurgical analysis of the damaged components revealed that the heater sheaths
had undergone primary water stress corrosion cracking (PWSCC) which allowed water
to reach the magnesium oxide (MgO) insulation internal to the heater. Expansion of
the MgO due to hydration resulted in the subsequent rupture of the heater sheaths and
cracking of the heater sleeve. It has been determined that the manufacturing process
of the Watlow heaters induced susceptibility of the sheaths to PWSCC. As a result of
this event, all Watlow heaters have been removed from the pressurizer and the two
ruptured heater locations have been permanently closed.
7. EVENT DATE:04/24/87 LER#: 87-003-00 REPORTDATE:05/27/87
2-7
TITLE: Pressurizer Heater Rupture Results in Reactor Coolant Pressure Boundary

Leakage and Pressurizer Vessel Corrosion

NAME: Larry A. Taylor, Plant Licensing Engineer
TELEPHONE # 501-964-3100
On 4/24/87 an Unusual Event was declared and a unit shutdown and cooldown
commenced due to a suspected pressure boundary leak of approximately 60 drops per
minute from the area of the pressurizer vessel lower head. Examinations following the
cooldown revealed the leakage was due to a damaged heater sleeve which penetrates
the vessel lower head and houses a heater. A small area of corrosion damage to the
carbon steel material of the vessel was found adjacent to the leaking sleeve.
Subsequent investigation revealed the damaged sleeve resulted from rupture of the
heater sheath. The ruptured heater and other heaters manufactured by W atlow
Electric Company (EIIS Identifier = AB-EHTR-W065) have been removed from the
pressurizer (twenty-three of the total ninety-six heaters). An additional Watlow heater
was also found to have a ruptured sheath but had not resulted in sleeve leakage. The
two damaged heaters' sleeves have been removed and plugs inserted and welded into
the sleeve penetrations. The corrosion damaged area has been repaired. Due to limited
replacement heater availability, sleeves containing no heaters have been fitted with
dummy plugs welded in place. Investigations are continuing to determine the root
cause of the heater sheaths failure. This event is reported per the requirements of
50.73(a)(2)(i), 50.73(a)(2)(ii) and due to the generic implications of the Watlow heater
failure.
8. EVENT DATE: 10/23/86 LER #: 86 006 00 REPORT DATE: 12/09/86
FACILITY NAME: Arkansas Nuclear One, Unit One (ANO-1)
TITLE: Boric Acid Corrosion of Carbon Steel High Pressure Injection Nozzle and
Reactor Coolant System Cold Leg Piping
OPERATING MODE: N POWER LEVEL: 000

NAME: Patrick C. Rogers,Plant Licensing Engineer
TELEPHONE # 501-964-3100
Corrosion of a Reactor Coolant System (RCS) nozzle and adjacent cold leg piping was
discovered during refueling outage performance of RCS inservice inspection. RCS
leakage from a body-to-bonnet-seal on a normally open isolation valve in the HPI line
2-8
had run down the stainless steel HPI piping within the reflective line insulation to the
carbon steel nozzle and cold leg piping. The leakage was less than 0.10 gallons per
minute and had occurred over a period of about 6 months. This continuous leakage, the
pipintemperatures, and the piping to insulation fitup are believed to have resulted in
pooling of what became highly concentrated, corrosive boric acid along the bottom of the
nozzle and adjacent cold leg piping. The maximum depth of the corrosion was
approximately 1/2 inch in the area of minimum HPI nozzle thickness (3/4 inch
including the thickness of an interior stainless steel clad) at the HPI line to nozzle
weld. The galvanic effect due to dissimilar metals contributed to the severe corrosion in
this area. Repairs consisted of grinding the areas of corrosion and performing a weld
buildup in the area of the most severe damage. An evaluation and inspection of other
components where boric acid attack of carbon steel could occur or previously known
sources of borated water leakage was performed with no additional problems found.
Instructions have been issued such that when in-containment borated water leakage is
identified the potential for boric acid attack of carbon steel is thoroughly evaluated.
9. EVENT DATE: 10/20/83 LER 83-079/031-0 ISSUED: 11/17/83
On 831020, leakage observed from body-to- bonnet region of loop isolation valve 1-RC-
87. Caused by excessive boric acid corrosion in studs. Valve repaired & RCS pressure
reduced to 2,235 psig.W/831117 ltr.
Authors: HARRELL,E.W. Virginia Power (Virginia Electric & Power Co.)
Recipients: O'REILLY,J.P. Region 2 (RII, Post 820201)
Dockets: 05000338 50-338 North Anna Power Station, Unit 1, Virginia Electric
10. EVENT DATE: 11/27/80 LER 80-028/01T-0 ISSUED: 12/11/80
On 801127 primary system leak was discovered. Caused by failure of carbon steel
packing gland studs from boric acid corrosion due to erroneous installation of carbon
steel studs to replace stainless steel studs.
Authors: STARKEY,R.B. Carolina Power & Light Co.
Recipients: * Region II, RII (IE, 720101-810228)
Dockets: 05000261 50-261 H.B. Robinson Plant, Unit 2, Carolina Power & Light C
2-9
11. EVENT DATE: 08/04/82 LER 82-009/01T-0 ISSUED: 09/10/82
On 820804, during refueling inspection of reactor coolant pump, five closure studs
found degraded in size below Technical Specification minimum. Caused by boric acid
corrosion due to closure gasket leakage. Studs replaced.
Authors: PETTY,J.C. Duke Power Co.
Recipients: * Region 2 (RII, Post 820201)
Dockets: 05000287 50-287 Oconee Nuclear Station, Unit 3, Duke Power Co.
12. EVENT DATE: 04/22/82 LER 82-008/0lT-0 ISSUED: 05/06/82
On 820422,during refueling outage, leak repair procedures on valve SI-13B revealed

significant boric acid corrosion wastage of body-to-bonnet bolts. Caused by bolt
degradation. Affected bolts replaced.
Authors: * Wisconsin Public Service Corp.
Recipients: * Region 3 (RIII, Post 820201)
Dockets: 05000305 50-305 Kewaunee Nuclear Power Plant, Wisconsin Public Servic
13. EVENT DATE: 04/22/82 LER: 82-007/01T-0 ISSUED: 05/06/82
On 820422, during refueling outage, leak repair procedures on valve PS-1A revealed
significant boric acid corrosion wastage on body-to-bonnet bolts. Caused by bolt
degradation. Affected bolts replaced.
Authors: * Wisconsin Public Service Corp.
Recipients: * Region 3 (RIII, Post 820201)
Dockets: 05000305 50-305 Kewaunee Nuclear Power Plant, Wisconsin Public Servic
14. Preliminary Notification of Event, PNO II-87-19, Florida Power and Light,
Turkey Point Unit 4
Reports deposits of crystalline boric acid on the reactor vessel head.
2-10
15. Preliminary Notification of Event, PNO-IV-86-32, Arkansas Nuclear One,

Unit 1
Body-to-bonnet leak on a manual isolation valve in the HPI line caused boric
acid corrosion on the external surface of the HPI nozzle.
Numerous NRC Dockets and NRC Inspection Reports regarding Boric Acid Corrosion were
reviewed which primarily contained utility responses to Generic Letter 88-05 or NRC
audits of the utilities.
Based on these documents, it appears that all the utilities have boric acid leak detection
and monitoring programs that consist of walkdown inspections during hot shutdowns and
during leak testing prior to startup after refueling.
Many utilities have implemented material upgrade programs to reduce material

degradation incidents to subsequently reduce the number of leaks. Electricite de France
credits much of its success in removing instances of BAC to its material upgrade program.
Typical upgrade programs include:
• replacing carbon steel and low alloy steel fasteners with stainless steel
fasteners,
• replacing valve packings with constant load type valve packings,
• installing snubber boots to improve the reliability of snubbers,
• application of protective coatings on flange surfaces, and
• improved bolting procedures.
2.1.2 Utility BAC Experience Survey Results
As previously mentioned, numerous incidents of materials degradation due to borated

water leaks have been reported in the industry. The time period when these leaks
occurred ranges from the early 1970's to mid 1980's. Pursuant to NRC's requirements
contained in Generic Letters 82-02 and 88-05, all utilities initiated boric acid leak detection
and prevention programs, and as a result, the frequency of "reportable" leaks was reduced.
In addition, the utilities also started fastener and valve packing material upgrade
programs that also reduced the incidence of leakage and associated material degradation.
However, there were no available technical documents that characterize and report the
frequency of leakage, particularly small "non-reportable" leakage, during approximately
the last five years.
To gather current information on borated water leaks during the most recent five years, a
survey of utilities operating PWR power plants was conducted by NMAC. The survey form
is shown in Figure 2.1. The primary emphasis of the survey was on leak characterization
and the effect of leakage. Eleven responses were received. The data is summarized in
2-11
Tables 2.3, 2.4 and 2.5. Telephone interviews were also conducted with some of the
utilities who had responded to the survey. The following are the important items revealed
through the interviews.
• Analysis of the data shows that the occurrence of manway leaks has been
significantly reduced. Many plants now report no leakage at manway bolted
connections. This can be attributed to a high level of effort in leak reduction
programs by the utilities, supported by testing and evaluation efforts of EPRI [3,
5, 14, and 15] and the NRC [2, 20 and 21].
It is noteworthy that presently, most leaks are very small and occur
predominantly through valve packings and valve body-to-bonnet gaskets.
• Many utilities have fastener upgrade programs and are replacing carbon steel
and low alloy steel fasteners with more corrosion resistant fastener materials.
• Some utilities are replacing standard valve packings with improved designs
featuring "live load packing."
• Some utilities have initiated a program to replace all gaskets with upgraded and
more reliable gasket designs.
• Walkdown inspections are performed as soon as a unit is shut down and at the
completion of each outage to locate leaks. Critical components are inspected
during every outage. Inspection procedures generally include leak evaluation
criteria.
• One utility has modified its vent shrouds to add access doors for inspection of
the reactor vessel head area.
• Some utilities perform monthly or bimonthly walkdowns inside containment

and outside the bioshield to detect boric acid leaks and accumulation.
• One utility reports a program to coat unclad reactor vessel closure flange
surfaces with a corrosion resistant material to prevent BAC of the flange
surfaces during refueling outages.
• To prevent boric acid corrosion in the stud holes and on the flange surface of the
reactor vessel during refueling, one utility fills the stud holes with a 160 ppm
concentration of ammonium hydroxide prior to installing the stud pins.
2-12
Nuclear 1300 Harris Blvd.

Maintenance Charlotte, N.C.
Applications 28262
Center FAX: (704) 547-6168
NMAC
Return FAX Transmittal
Date: Time: Page 1 of: 2
To: John Jenco
From: Telephone:
Boric Acid Corrosion Evaluation Survey
Figure 2.1
Boric Acid Corrosion Evaluation Survey
2-13
Figure 2.1
Boric Acid Corrosion Evaluation Survey (cont.)
2-14
2-15
2.2 NRC Guidance

Based on numerous incidents of BAC reported in licensing event reports (LERs), the NRC
issued frequent Information Notices, Bulletins and Generic Letters to alert utilities to the
degradation and potential failures of critical components due to BAC. A brief chronological
history of NRC activities regarding boric acid corrosion follows [4].
• On June 11, 1980, the NRC issued IE Information Notice 80-27 [6]. This notice
described in detail the degradation seen on the closure studs of four Byron
Jackson reactor coolant pumps at Fort Calhoun Unit 1. The notice emphasized
the concern that the severe corrosion (seen at Ft. Calhoun), if left undetected,
could have led to a loss of integrity of the primary coolant system due to bolt
degradation. The notice also emphasized the need for supplemental visual
inspections (due to the unreliability of the ultrasonic techniques) and the use of
instrumented leak detection systems. No action was required by the utilities.
• A second IE Information Notice No. 82-06 [7] was issued on March 12, 1982,
after the failure of steam generator manway studs at the Maine Yankee Power
Plant. Although the failure was characterized by a stress corrosion cracking
mechanism (not wastage corrosion), the possible interaction of stud preload,
lubricant, an injection sealant, and primary coolant from leakage was
considered. No action was required by the utilities.
• IE Bulletin No. 82-02 [1] entitled, "Degradation of Threaded Fasteners in the

Reactor Coolant Pressure Boundary of PWR Plants," was issued by the NRC on
June 2, 1982. This Bulletin described the results of the failure investigation
performed on bolts from the Maine Yankee Plant (IE Notice 82-06). It also
required that certain actions be taken by PWR facility licensees. These actions
were to be limited to the reactor coolant pressure boundary (RCPB) and
encompassed threaded fasteners in (1) pressurizer and steam generator
manways; (2) control rod drive flange and pressurizer heater connections
(without seal welds); and (3) valve bonnets and pump flange connections on 6-
inch nominal diameter lines and larger. The following actions were required of
PWR operating licensees.
1. Develop and implement maintenance procedures for threaded fastener

practice for:
a. maintenance crew training of proper bolting/stud practices, tool

application, specifications and requirements,
b. detensioning and retensioning practices (torque Iteration), specified

tolerances, and other controls for disassembly and reassembly of
component closure/seal connections,
2-16
c. gasket installation and controls,
d. retensioning methods, and

e. quality assurance measures to be established for proper selection of
fastener lubricants and injection sealant compounds to minimize fastener
susceptibility to stress corrosion cracking (SCC) environments.
2. Threaded fasteners, when opened for component inspection or maintenance,

shall be removed*, cleaned, and inspected per IWA-2210 and IWA-2220 of
ASME Code Section XI (1974 edition or later) before being reused.
NOTE: *Fasteners "seized" or designed with interference fit, may be

inspected in place.
3. NRC Information Notice No. 80-27 and 82-06 were issued in regard to
corrosion problems associated with bolts/studs in RCPB closures. To assist
the Nuclear Regulatory Commission in its ongoing review and assessment of
the scope of the problem, utilities were asked to provide the following
information for closure and connections within the scope of this bulletin:
a. Identify those bolted closures of the reactor coolant pressure boundary

pump (RCPB) that have experienced leakage. Describe the inspections
and corrective measures taken to eliminate the problem.
b. Identify those closures and connections, if any, where fastener lubricants

and injection sealant materials have been or are being used and report on
plant experience with their application, particularly any instances of SCC
of fasteners.
A written report addressing these issues was required from the licensees.
• The NRC sponsored a BAC rate study program for carbon and low alloy steel
fasteners and results were published in NUREG/CR-2827 in July 1982 [8].
Details of the test data are discussed in the next section of this report.
• During January 1983, the NRC issued NUREG-0943, entitled, "Threaded-

Fastener Experience in Nuclear Power Plants" [2]. This report identified forty-
four (44) instances of threaded fastener degradation at nuclear plants between
October 1964 and March 1982. The report provided a comprehensive overview
of threaded fastener concerns and regulatory actions (both short-term and
ongoing) related to bolted connections at nuclear plants. The NUREG identified
berated water corrosion as the second most contributing mechanism for
threaded fastener degradation with thirteen 13 of the 44 instances attributed to
this wastage form of corrosion. The report stated that: "In almost every case,
the cause of threaded-fastener degradation was corrosive attack by borated
2-17
water leaking from closure gaskets or seals, and the degraded fasteners were
discovered during the process of correcting the leakage."
.
Responses to IE Bulletin 82-02 from utilities were evaluated and closed out in
May 1985 through issuance of NUREG-1095, "Evaluation of Responses to IE
Bulletin 82-02" [9]. This NUREG evaluated responses from 41 PWR licensees
and utilized statistical analyses to determine the important factors governing
the occurrence of bolting degradation in PW R s .
.On December 29, 1986, the NRC issued IE Information Notice No. 86-108,
entitled, "Degradation of Reactor Coolant System Pressure Boundary Resulting
from Boric Acid Corrosion" [10]. This notice describes the corrosion wastage
event at the ANO Unit 1 plant in October 1986. This document indicated that
boric acid corrosion in excess of 1-inch depth per year in ferritic steels is possible
where boric acid reactor coolant has impinged on a surface and remained wet.
No follow-up action was required of the licensees.
.
Supplement 1 to Information Notice No. 86-108 [11] was issued by the NRC on
April 20, 1987. This notice alerted PWR operating licensees of the conoseal
leakage found at Turkey Point Unit 4. The supplement described the incident
and stated that 3 of the 68 reactor vessel head bolts had to be replaced due to
the corrosion. Approximately 500 lbs. of boric acid crystals had accumulated on
top of the closure head.
.On November 19, 1987, the NRC issued Supplement 2 to Information Notice
No. 86-108 [12]. This supplement described two instances of severe boric acid
wastage corrosion -- the Salem Unit 2 conoseal pinhole leaks and an incident at
San Onofre Unit 2 where 18,000 gallons of reactor coolant spilled into the
containment from the discharge of a leaking shutdown cooling system isolation
valve.
.On March 17, 1988, the NRC issued Generic Letter 88-05, "Boric Acid Corrosion
of Carbon Steel Reactor Pressure Boundary Components in PWR Plants" [13].
The intent of the letter was to obtain information from PWR licensees relative to
assessing the safe operability of PWRs when reactor coolant leaks (below
technical specification limits) degrade carbon/low alloy steel components. It also
required an assurance that licensees will develop and implement a boric acid
corrosion prevention program to include the following:
1. The determination of the principal locations where leaks that are smaller
than the allowable technical specification limit can cause degradation of the
primary pressure boundary by boric acid corrosion.
2. Procedures for locating small coolant leaks (i.e., leakage rates at less than
technical specification limits). It is important to establish the potential path
of the leaking coolant and the reactor pressure boundary components it is
likely to contact.
2-18
3. Methods for conducting examinations and performing engineering

evaluations to establish the impact on the reactor coolant pressure boundary
when leakage is located.
4. Corrective actions to prevent recurrence of this type of leaks and corrosion.

• Responses were received from the licensees and evaluated by the Chemical
Engineering Branch of the NRC. A total of 50 responses were evaluated by the
NRC. All were found to be satisfactory and met the intent of the Generic Letter.
NRC audits were performed on ten licensees. The audit findings and response
evaluations were published in NUREG-6576 [4].
2.3 EPRI Information

In response to the number of bolting/fastener degradation incidents that took place at
nuclear power stations in 1982-1983, the Electric Power Research Institute (EPRI)
convened a three day seminar on "Bolting Degradation or Failure in Nuclear Plants," on
November 2-4, 1983 [14]. The purpose of this seminar was to describe to industry
representatives and regulators the EPRI projects, both ongoing and in the planning stage,
relative to bolting degradation. BAC related degradation was one of the major causes of
failures among other causes that included preload, stress corrosion, and improper heat
treatment of materials. Subsequent EPRI efforts resulted in the following publications
relating to boric acid corrosion:
• In December 1984, "A Survey of the Literature on Low-Alloy Steel Fastener

Corrosion in PWR Power Plants" [5] was issued. This EPRI document, NP-3784,
specifically addressed the issue of boric acid wastage corrosion.
• EPRI issued "Degradation and Failure of Bolting in Nuclear Power Plants," NP-
5769 in April, 1988 [15]. This two volume report discussed a myriad of topics
related to bolting degradation. Borated water corrosion was one of the
mechanisms among others such as stress corrosion, over torquing, and improper
heat treatment.
• In response to the Turkey Point Unit 4 reactor vessel head (1987) and Arkansas
Nuclear One Unit 1 reactor coolant system nozzle (1986) incidents, EPRI issued
the report, "Boric Acid Corrosion of Carbon and Low-Alloy Steel Pressure-
Boundary Components in PWRs," NP-5985, in August, 1988 [3]. This report
addressed the problems of boric acid corrosion through a literature survey of
boric acid corrosion events on carbon/low-alloy steel components other than
fasteners.
• Another document that addresses the boric acid corrosion problem is the EPRI
report entitled, "Elimination of Soluble Boron for a New PWR Design," NP-6536,
dated September, 1989 [16]. This report summarizes a limited scope design
study that investigated the feasibility of a commercial PWR that does not use
2-19
soluble boron for reactivity control. One of the advantages of boron elimination
listed in the report follows:
"Avoiding concerns with boric acid induced corrosion of pressure

vessels bolting and other critical components as a result of primary
system leaks..."
• In 1990, EPRI's Nuclear Maintenance Applications Center published "Good

Bolting Practices - Large Bolts" NP-5067V1 [17] and "Good Bolting Practices -
Small Fasteners," NP-5067V2 [18]. These bolting reference manuals were
designed to provide practical solutions for typical problems that equipment
operators and maintenance personnel might encounter. Each volume contains
sections pertaining to boric acid corrosion of bolting and small fasteners, and
provides specific information relating to material selection for fasteners.
• The EPRI document "Guide for Threaded Fasteners," NP-6316, July, 1989 [19]
provides detailed information on specific physical attributes of threaded
fasteners. It may prove valuable as a reference source during selection of
fasteners for boric acid service, and in assessment of fasteners degraded due to
boric acid exposure.
2-20
3.0 BAC Test Data Summary

In response to the experience gained in 1970 from a PWR in Europe that had experienced
reactor vessel head corrosion resulting from the leakage of reactor coolant containing boric
acid [22], and numerous incidents of fastener degradation that occurred in the 1970's and
early 1980's, the NSSS suppliers, NSSS Owners Groups, NRC, and EPRI supported and/or
performed a variety of BAC test programs. The main objective of these test programs was
to establish BAC rates for different conditions of geometry, temperature, boric acid
concentration, boric acid pH, and boric acid flow. This section presents descriptions and
summaries of the tests and results.
3.1 Westinghouse Test Data
These data were obtained from a Westinghouse letter to the NRC [22]. The data are
presented in Table 3.1.
The following provides the conclusions of the tests.
• Aqueous solutions of boric acid, when allowed to concentrate, are highly

corrosive to carbon steel surfaces at approximately 200°F.
• The presence of dry boric acid cannot be considered a benign condition.
• Data on the corrosion rate of carbon steel in oxygenated boric acid is scattered
and appears to be a function of test methodology, temperature, and other
contaminants present.
• Any aqueous boric acid that penetrates into a crevice would be expected to result
in some general corrosion of steel before reaching the dryout stage, at which
point the attack would cease or become negligible.
Table 3. 1
Westinghouse BAC Test Data
Boric Acid Test Temperature Carbon Steel

Concentration Condition (°F) Wastage Rate
(mils/month)
25% aqueous 200 400
aerated
25% aqueous 200 260
deaerated
15% Drip Test in Air 210 350-400
15 - 25% Dripping boric acid - Inconel 210 400
CRDM weld mock-up w/typical
crevice geometry
l
10% Boric Acid aqueous 194 400
+ 1% NaCl air
3-1
3.2 NRC/Brookhaven National Laboratory Test Data

Due to the increasing number of bolting failures reported in the early 1980's, Brookhaven
National Laboratory (BNL) performed a literature survey and conducted laboratory tests to
establish BAC rates. The results are reported in References 8 and 23 and are summarized
in Table 3.2.
References 8 and 23 do not provide the test methodology or other test details. Based on
telephone conversations with the author of the report, the test data were generated using
aqueous aerated boric acid.
The following are the conclusions:
• Ferritic material is susceptible to corrosion attack by H3 BO3 and LiOH

solutions up to at least 352°F.
• The rate of corrosion rapidly increases up to 212°F, then starts to decline as

water is boiled off, but the rate of metal loss is still high.
• A corrosion rate of at least 112 mils/yr. can be attained at 212°F. Faster rates
may be attainable if the ferritic material is kept wet as higher temperatures are
reached.
• Small additions of either KOH or LiOH (normally used as pH buffers) appear to

cause the corrosion to proceed at a faster rate than just the H3 BO 3 in solution
at higher temperatures.
3-2
3-3
3.3 ABB-CE BAC Test Data on Fastener Material
As part of a fastener material evaluation program, ABB-Combustion Engineering

conducted a series of corrosion tests to evaluate corrosion rates of fastener materials
exposed to berated water. Included in the test matrix were three low alloy steels (SA-193
Grade B7, SA-540 Gr B24 Class 3, and SA-540 Gr B23 Class 4), two high strength
corrosion resistant alloys (17-4 pH stainless steels and NiCrFe Alloy 718), and four
examples of low alloy steels with a surface coating or treatment applied to enhance
corrosion resistance.
To obtain a qualitative measurement of BAC, berated water drip tests were conducted.
Uniformly loaded, bent-beam specimens measuring 6" x 1" x 1/8" were employed. The
lower half of the specimen was dipped into a 190°F boric acid (1000-2000 ppm B) drip tank.
Boric acid solution (2000-5000 ppm B) at 70°F was dripped onto the upper portion of the
specimen. The results are presented in Table 3.3 [23].
In another series of tests, fastener materials were exposed to borated wet steam. The
specimens in this series of tests were small diameter laboratory studs bored from one end
to accommodate individual heaters and the other end to accommodate thermocouples. The
specimens were stressed to 1/2 to 2/3 of the yield strength. During the test, borated water
was directed through capillary tubes and impinged directly on the specimens. The steam
temperature was maintained at 600°F and the specimen temperature was maintained at
175°F - 400°F. The berated water (approximately 1000 ppm B as H 3 BO 3) flashed to wet
steam upon exiting the capillary tubes. Table 3.4 presents the summary of the test results
[24].
The ABB-CE test data established corrosion rates for various fastener materials and drew
the following conclusions.
• Corrosion rates in borated solutions do not differ greatly with changes in alloy
composition.
• The onset of corrosion requires that the leakage cool only a thin skin of fastener
material. This supports the conclusion of Reference 2 that RCP stud corrosion
can occur during all modes of operation.
• Relatively small leakage rates (0.05 gpm) are sufficient to cool localized areas of
fasteners to the temperature where corrosion will occur.
• Surface coatings and treatments, with the possible exception of electrolytic

nickel plating, are not effective in preventing BAC.
• Alloys such as Types 304 and 410 stainless steel, 17-4 pH stainless steel and
NiCrFe Alloy 718 are resistant to berated water corrosion.
3-4
3.4 ABB-CE BAC Test Data on Shell Material

In addition to boric acid corrosion of fasteners in bolted closures, boric acid corrosion was
also reported for reactor vessel heads, nozzles and piping, pressurizer shells, and valves.
BAC of carbon and low alloy pressure vessel steel occurred when these components were
exposed to berated water that had leaked through instrumentation fittings, seal welds,
valve gaskets, and bolted closures. The resulting corrosion ranged from an accumulation of
boric acid crystals with little or no corrosion to wastage of the material to the extent that
weld repair was required.
ABB-CE conducted three series of tests of low alloy pressure vessel steels exposed to
borated water that are briefly described in this section [25].
3.4.1 Borated Water Autoclave Tests

The specimen used in this test had dimensions of 4 1/2" x 1/2 x 1/8". Welded specimens
were prepared using a weld procedure typical of those used in fabrication of NSSS
components. The specimens were tested in a 10 liter capacity autoclave. The test solution
was deionized water with 723 ppm B as H 3 BO 3 , 1.8 ppm lithium, and 0.4 ppm ammonia,
and it was not deaerated. Each test lasted for 20 hours and the autoclave test temperature
was maintained at 350°F. Table 3.5 presents the results.
3.4.2 Beaker Tests
Earlier results reported in Reference 4 were reportedly from short-term (6 hours) tests that
appeared to be very conservative. Accordingly, ABB-CE conducted a series of beaker tests
in concentrated boric acid solutions to confirm the lower rates reported in their earlier
work on fastener materials.
The specimens used in this test were 4 1/2" x 1/2 x 1/8". The test specimens were placed in
flasks that contained 600 ml of test solution that contained deionized water with 22,000-
79,400 ppm B as boric acid at 220°F. Each flask contained water-cooled condensers that
prevented the test solutions from evaporating. Table 3.6 summarizes the test results. The
data shows a wide variation in corrosion rates with corrosion rates decreasing markedly
with time that also corresponded to increased pH resulting in decreased corrosion rates.
These tests, which were not representative of field conditions, also showed that the
corrosion rates were dependent on the test procedure utilized.
3.4.3 Borated Steam Tests
Since the above test programs may not have been representative of a situation in which
wet steam originating from borated water impinges on a hot carbon steel surface, ABB-CE
conducted a third laboratory test program to obtain BAC rate data for a low alloy steel
under more representative conditions.
The test specimens consisted of two sets of square blocks of metal, approximately one inch
thick. One block contained capillary tubes with an ID of 0.003-inches that carried borated
water from the heat exchanger. The other block, the deposit sample, was located at a
3-5
distance of 1/2"- 2" from the capillary block. Both blocks had penetrations for heaters and
thermocouples to maintain the desired temperature. The source of the steam was a high
temperature test facility that circulated borated PWR water at 600°F. The corrosion rates
were measured for the deposit as well as the capillary block samples. The changes in
diameters of the holes determined the corrosion rate for the capillary tube and the loss of
volume in areas adjacent to the steam impingement established the BAC rate for the
deposit block. Details of the test methodology are presented in Reference 25. Table 3.7
presents a summary of the test data.
The report [25] presents the following findings:
• Corrosion rates greater than 500 mils per year are possible under alternate
wetting and drying conditions of boric acid.
• Borated water corrosion will occur at normal full power operating temperatures
since relatively small leakage rates are sufficient to cool localized areas of
material to the temperature where corrosion will occur.
Table 3.3
ABB-CE Drip Test Results
Material Number of Exposure Extent of

S p e c i m ens (hours) Corrosion
A-516, Grade 70 4 775 Severe
AISI 4340 1 2155 Severe
Electroless Nickel 3 765 Severe corrosion to AISI 4340
Plated
Chromium Coated 3 1000 Severe corrosion to AISI 4340
Electrolytic Nickel 3 1000 Light attack to AISI 4340
Type 304 SS 2 775 None
Type 410 SS 3 1540 Minor petting, no general
2 1785 corrosion
• Beam Specimen 6" X 1" X 1/8"
• Uniform Tensile Stress Beyond Yield - 168°F Temperature
• Bottom Half Immersed in Boric Acid (1000 - 2000 PPM B) 190°F
• Drip Tank (2000 - 5000 PPM B)
3-6
Table 3.4
ABB-CE Borated Steam Tests on Fastener Materials
Test Corrosion Rate (in/yr.) at Test

Material Temperature (°F)
175 300 350 400
SA-193 B7 Cr-Mo alloy 0.639 0.190
SA-540 B23 Class 4 Cr-Ni-Mo alloy 0.731 0.660
SA-540 B24 Class 3 Cr-Ni-Mo alloy 0.265 0.722 0.495 1.690
SA-540 B24 Class 3 Cr-Ni-Mo alloy 0.393 0.432
17-4 pH SS 0.000 0.000
17-4 pH SS 0.000
Alloy 718 Ni-Cr-Fe Alloy 0.000 0.000
Alloy 718 Ni-Cr-Fe Alloy 0.000
Cr-ion Implanted Low Alloy Steel 1.542
TiN plasma Coat Low Alloy Steel 0.753
Enamel Low Alloy Steel 2.032
Solid Film Lubricant Low Alloy Steel 0.925
• Small diameter studs with a bore to accommodate heaters and thermocouple,

175°-400°F
• Stressed to 112 to 1/4 yield strength
• Capillary tubes with ID of 0.003" to flash water into steam
• Steam impinged directly on specimen, 500 hours
• Deoxygenated and demineralized water at 600°F, 2200 psi, 1000 ppm B. LiOH
and hydrazine added
Table 3.5
ABB-CE Autoclave Tests of Low Alloy Shell Materials at 300°F
Material Number of Specimens Average Corrosion Rate
(mils/yr.)
SA-633 Grade B 3 17.1
SA-508 Class 2 3 16.8
SA-508/SA-533 Weld 2 16.1
• SA 533B, SA 508 C12 steel and weld specimen

• 4.5" x 0.5" x 0.125" specimens placed in autoclave
• Deionized water solution with 723 ppm boric acid added to autoclave
• Autoclave temperature at 350°F - 70 hours test duration - not deaerated
3-7
Table 3.6
ABB-CE Short Time Beaker Tests at 220°F
Material Exposure Concentration Corrosion Rate pH@ pH @

Time (hrs) (ppm Boron) * (in/yr.) Start End
SA106 Gr B 4 79,400 7.250 3-4 5
SA106 Gr B 6 26,200 1.630 4 6-6
SA106 Gr B 27 26,200 0.396 4
SA106 Gr B 24 22,000 0.752 3 6
SA106 Gr B 96 22,000 0.241 3 6
SA533 GrB 6 26,200 1.416 4 5-6

SA533 GrB 27 26,200 0.305 4 5
SA533 GrB 24 22,000 0.651 3 5
SA533 GrB 96 22,000 0.358 3 6
SA508Cl2 6 26,200 0.911 4 5-6

SA508Cl2 27 26,200 0.214 4 5
SA508Cl2 24 22,000 0.519 3 5
SA598Cl2 96 22,000 0.238 3 6
* A s H3 B O 3
• Specimens placed in flasks with boiling boric acid
• Time varied from 4 to 96 hours
• Fresh solution of 22,000 ppm B replaced the original solution
3-8
Table 3.7
ABB-CE Borated Steam Tests Data
Sample Sample Target Exposure Corrosion Corrosion

No. Type Flow Time Rate @ Exit* Rate @ ID**
(gpm) (hours) (in/yr.) (in/yr.)
C1 Capillary 0.05 392.6 0.4464 0.9660
C2 Capillary 0.05 392.5 0.3348 1.1626
C3 Capillary 0.05 392.5 0.3571 0.8106
C4 Capillary 0.05 392.5 0.8035 1.0414
C5 Capillary 0.10 369.0 1.0683 0.8510
C6 Capillary 0.10 369.0 0.6172 0.9359
D1 Deposit 0.05 392.5 0.580 N/A
D2 Deposit 0.05 392.5 0.3571*** N/A
D3 Deposit 0.05 392.5 0.5133 N/A
D4 Deposit 0.05 392.5 0.3794*** N/A
D5 Deposit 0.10 369.0 0.5698 N/A
D6 Deposit 0.10 369.0 0.6172 N/A
* Deposit Block
** Capillary Block
*** Unheated Samples
• SA-533 Grade B Steel Material
• 1" blocks drilled to accommodate a 0.003" ID capillary tube blocks were heated
at 650°F, and 1000 ppm B as boric acid solution at 600°F was circulated
• The borated steam exited the capillary samples and exposed another sample
placed at a 2" distance from the capillary sample
3.5 CE Owners Group Sponsored Test Data
Some of the ABB-CE pressurizer heater penetrations developed leaks due to the
application of magnesium oxide as an insulator between heater elements and Alloy 600 Ni-
Cr-Fe sheaths [26]. In one of the units, a borated water leak degraded the pressurizer
bottom head material near the penetration that required a weld repair. In other cases,
Alloy 600 instrument nozzles developed primary water stress corrosion cracking (PWSCC)
[27] and these incidents resulted in borated water leaks which could have degraded the
reactor coolant boundary materials. As a result, the CE Owners Group sponsored boric
acid corrosion test programs to quantify the pressurizer shell material corrosion/erosion
rates for representative conditions. The test methodology and results are described in
Reference 28.
A total of four blocks of low alloy steel SA-633 Grade B Class 1 were tested. Each block
was 4" x 4" by 5" thick and had a 3/4" diameter hole to accommodate Alloy 600 tubing.
3-9
Each block had a diametric clearance around the Alloy 600 tubing which varied from
0.0026" to 0.0099". The 3/4" diameter Alloy 600 tubes had artificially implanted cracks
which provided varying levels of leak rates. The specimen temperature was maintained
around 600°F and the berated water was provided from a recirculating test loop at 600°F
and 2250 psig. The boric acid concentration was 600-1000 ppm B as boric acid with 1.2-1.4
ppm Li as lithium hydroxide. The water was deaerated with hydrazine and the O2 and Cl2
were maintained at < 0.1 ppm. The test results are presented in Table 3.8 [28]. [Data in
Table 3.8 are proprietary data and are reproduced with the permission of CEOG.]
The following conclusions are available:
• The vast majority of metal loss occurred at the exit of the block. The surface
degradation in the area where the borated water/steam impinged was minimal
compared to degradation at the exit.
• Small leak rates caused greater damage.
• Boric acid and corrosion product buildup affects the flow rate.
• The degradation remote from the annulus can be significantly high (up to 1340
mils/year or 0.67 in 3 /year).
Table 3.8
CE Owners Group Crevice Corrosion Test Data
(Proprietary Data)
• SA-533B CL1 Steel - 5" Thick
• 2 to 9 Mils Diameter Crevice
• Representative Condition of Pressurizer Nozzle Except No Insulation Present
• Specimen Temperature of 600°F
3-10
• Leak Rates .007 GPM to .05 GPM
• Test Duration 500-600 Hours
3.6 B&W Owners Group (B&WOG) Boric Acid Program

Several incidents of borated water leaks were reported, as shown in Table 2.2. Carbon
steel and low alloy steel degradation occurred in components remote from the leak
locations. There, boric acid would crystallize and subsequent wetting and drying would
result in metal degradation. B&WOG test programs investigated the effect of flowing
berated water on the inside and outside of carbon steel pipes.
The material used for these tests was SA-106 Grade B 4-inch Schedule 80 carbon steel
pipe. In two experiments, a 304L stainless steel pipe was attached to the carbon steel pipe
to simulate degradation from galvanic effects.
Each trough/pipe was mounted at an angle to allow flow of the solution. In one of the tests,
the externally heated trough was enclosed so that the humidity could reach levels similar
to that found in containment. The internally heated pipe apparatus was insulated to
simulate on-site conditions of a leak traveling along a carbon steel pipe under the cover of
insulation. This also allowed the local humidity to reach higher levels.
The test temperatures ranged from 300°F to 550°F. Thermocouples were used to monitor
the temperature of the trough/pipe. Initial boric acid concentration was controlled within
the limits of 13,000 to 15,000 ppm (as H3 BO3) with 1.0-2.0 ppm lithium (as LiOH) added
to duplicate the primary water chemistry. A boric acid flow rate was maintained at 0.03
gpm (0.01 per outlet in the heated pipe arrangement) [29].
The test results are reported in Table 3.9 [29].
The following conclusions were drawn:
Summary of Results
• In the first type of tests (where the open trough was exposed to air as the boric
acid solution flowed along the length of the trough), the highest degree of
damage (4.9 in/yr) for carbon steel pipe heated to 300°F occurred at the top of
the pipe, at the location where the borated solution impinged upon the pipe.
• The trough made from carbon steel welded to stainless steel had the highest
degree of corrosion (5.1 in/yr) at the weld on the carbon steel side. (Galvanic
corrosion may have occurred in this case, but this could not be proven with the
tests completed in this study.)
• The trough enclosed in the "humid air" environment had the second highest
degree of corrosion (4.7 in/yr) approximately two-thirds of the way from the top.
3-11
• In the closed (insulated) pipe tests, temperatures approaching 500°F led to

lower levels of corrosion (1.3-4.1 in/yr), while a glassy residue remained on the
surface and higher temperature (530°F) led to a much lower corrosion rate (0.3
in/yr).
• The highest levels of corrosion occurred at the interface between the dry boric
acid crystals and the flowing solution. It is believed that the solution kept the
crystals wetted at the interface, causing a very high concentration of boric acid
that led to the corrosion levels observed.
• In the "humid air" environment, the highest level of corrosion occurred further
along the length of the pipe since the crystals were not able to dry out as quickly
where the solution impinged on the pipe.
• Based on the performed tests, the effect of flow rate was not determined.
• The experiments indicate that increased temperature and, to some extent,

decreased hydration increased the rate of boric acid corrosion of carbon and low
alloy steel.
• The relationship between temperature and boric acid concentration could not be
quantified.
Table 3.9
B&WOG Borated Water Tests on Carbon Steels
• Dripping boric acid on pipes; inside flow and outside flow
• Pipes heated with electric heaters
• Data are not sufficient to establish relationship of flow, temperature, and boric
acid concentration
• SA 106 Grade B Carbon Steel Pipe
3-12
• 13,000 to 15,000 ppm (as H 3 BO3 ) 0.03 gpm flow
• Boric acid heated to 200°F
• Time 200-300 hours
3-13
4.0 Boric Acid Corrosion Mechanisms

Corrosion of a metal component can occur through a number of mechanisms that may lead
to localized or general loss of material. The corrosion mechanism that is active in any
given situation is a function of the metal composition, its heat treatment and purlty,
corrosive environment, presence of dissimilar metals, and other factors. Corrosion may
proceed via two or more processes simultaneously. Some metals are immune to certain
corrosion mechanisms while they may be aggressively attacked by others. In the boric acid
corrosion studies previously cited, there was no clear identification of the active corrosion
mechanism(s) in all cases.
As with all types of corrosion, many factors influence the rate of attack. The acidity,
temperature, concentration, corrosive agent flow rate, oxidation potential of the corrosive
agent, aeration, and the presence of inhibitors or accelerators are some of the more
important factors. In the case of boric acid corrosion, these factors can include steam
impingement force, wicking effect of non-metallic insulation, flow rate through leaking
gaskets, presence of dissimilar metals, metal impurities, applied mechanical stresses, and
residual mechanical stresses. Many of these factors interact and this interaction may be
very complex.
The dominant form of corrosion reported from both past and recent inspections has been
identified as "wastage" (or general dissolution corrosion) of carbon and low alloy steel.
Most BAC test programs described in Sections 3 and Appendix A address "wastage"
degradation. Other factors that influence the rate of corrosion include force of steam
impingement, wicking effect of non-metallic insulation, flow rate through leaking gaskets,
galvanic corrosion, crevice corrosion, and pitting. However, these boric acid corrosion
mechanisms were not fully explored in the test programs cited. This section briefly
describes potential BAC mechanisms [30].
4.1 General Corrosion (Wastage)

The dominant form of corrosion reported from both past and recent nuclear plant
inspections has been identified as "wastage" (or general dissolution corrosion) of carbon
and low alloy steels. Most of the BAC test programs reported have focused on "wastage"
degradation. Wastage, or general dissolution corrosion, is associated with a metal surface
being uniformly attacked through a dissolution process. A typical metal consists of
crystalline regions or grains that may be interspersed with small precipitates of higher
melting point constituents, materials with limited solubility, or contaminants. Iron
carbides form fine precipitates in steels, for example. The term "inclusion" is used to
identify entrained bits of contaminating materials.
In the general corrosion process, an aggressive environment attacks an exposed metal

surface, leading to overall material loss. The anode areas on the metal surface shift to
different positions until the entire metal surface has been anodic at some time. It is
possible that at a given point, corrosion occurs on a group of readily dissolved atoms until
this group is depleted. Then the point of attack moves to some other point, resulting in a
uniform thinning of the metal. Surfaces attacked in this manner may appear to have some
4-1
increase in texture due to small differences in the rate at which the components of the alloy
are attacked. Bulk corrosion may be difficult to detect visually since there may not be a
convenient point of reference with which to assess the loss of surface material.
The rate of general corrosion is typically expressed in units of inches penetration per year
(ipy) or mils penetration per year (mpy). Since this process is relatively homogeneous, its
rate can often be predicted with a high degree of confidence. Based on corrosion rate
charts for a specific metal and/or experimental data, design allowances can be used which
make provision for significant material loss during the life of a component. Rate
calculations are based on experimental data that determine corrosion rates at various
temperatures and corrosive agent concentrations. As with all types of corrosion, many
factors influence the rate of attack. The acidity, temperature, concentration, motion
relative to metal surface, degree of oxidizing power and aeration, and presence or absence
of inhibitors or accelerators are some of the important factors. Most of these factors
interact and often this interaction is very complex.
4.2 Galvanic Corrosion

Galvanic corrosion may occur when two dissimilar metals are in direct contact in a
corrosive environment. The potential for galvanic corrosion is principally driven by the
magnitude of the electrochemical potential between the metals. Some combinations, such
as aluminum and copper, have very high electrochemical potential and corrode at a very
high rate. Two slightly different steels would have a low electrochemical potential and a
low corrosion rate.
In the galvanic corrosion process, an electrochemical cell is established between the metals
and the electrolyte (boric acid). Due to differences in the electrochemical potentials, one
metal will be anodic (negative) with respect to the other and will be preferentially corroded.
Local galvanic corrosion may occur if a metal is welded with a dissimilar metal filler, for
example. Dissimilar metal weld corrosion has occurred due to boric acid corrosion.
Galvanic corrosion has been tested by the Babcock and Wilcox Owner's Group. The key
BAC parameters addressed in the studies previously cited include the pH of boric acid,
temperature, and level of oxygen (aerated or deaerated. The BAC utility survey results,
shown in Table 3.1, reveals numerous occurrences of dissimilar metal weld corrosion.
Except for B&WOG limited data, no other data exists in the reviewed literature on BAC
rates for galvanic corrosion.
4.3 Crevice Corrosion - Metal ion Concentration Cell
The local rate of corrosion may be enhanced as a consequence of geometry. For example, a
flat plate may corrode at an acceptable rate at all locations except where a crevice is
present. The crevice may be the consequence of a shielded area under a washer, space
between overlapping plates, or other situations in which a shielded area, with limited
direct access to the corrosive environment, is created. Crevice corrosion may be
particularly aggressive when an aerated, halide-containing solution is present. Chlorides
4-2
(C1 -) are particularly aggressive in promoting crevice corrosion. There are various theories
on the reason why crevices are subject to enhanced corrosion rates. The common
understanding suggests that the corrosive solution within the crevice area will become
saturated with metal ions more quickly than the bulk solution. This leads to an
electrochemical potential difference within the solution that subsequently makes the metal
within the crevice anodic. Detection of crevice corrosion is often complicated by the lack of
access for inspection. Radiographic inspection and ultrasound techniques are quite
effective in detecting this type of degradation.
4.4 Crevice Corrosion - Oxygen Concentration Cell
In the oxygen concentration cell corrosion process, crevices again play an important role.
Within a crevice, the oxygen concentration may be lower than in the bulk solution since the
oxygen within the crevice can react with the metal and may not be replaced readily due to
limited access to the oxygen source. Since the cathodic reaction consumes oxygen, the area
within the crevice becomes anodic with respect to the metal outside. Thus, the corrosion
rate within the crevice is enhanced. Corrosion will proceed until the oxygen concentration
becomes uniform at all locations or it may continue until the metal is destroyed if oxygen
continues to be supplied to the area outside the crevice (see Figure 4.2).
In practice, both metal ion and oxygen concentration cells can exist at the same time and
may work in opposition to each other. The location of the more severe corrosive attack
would then be determined by the cell that has the greater electrochemical potential [30].
Most of the observed boric acid crevice corrosion damage has occurred in carbon and low
alloy steel fasteners. The configuration of fasteners makes them susceptible to both types
of crevice corrosion attack. Considering the recent incidents of pure water stress corrosion
cracking (PWSCC) of CRDM housings and instrumentation and heater penetrations,
development of a full understanding of BAC crevice corrosion should become an important
aspect of future BAC test programs.
4.5 Pitting
This type of corrosion is not a common BAC mechanism. Under certain conditions, anodic
areas on a metal surface remain stationary rather than shifting about. When this occurs,
corrosion takes the form of pits rather than general thinning. Pitting occurs through
localized galvanic corrosion between a base metal and inclusions or segregated components
of the alloy.
4.6 Intergranular Corrosion

Intergranular corrosion is a localized attack along the grain boundaries of a metal or alloy.
Corrosion can proceed to the point where whole grains of metal fall away and the metal
loses its strength and metallic characteristics. An example of this type of corrosion is
shown in Figure 4-3.
4-3
Intergranular corrosion is usually caused by an improper heat treatment or heat from

welding that causes the precipitation of certain alloy components (mainly chromium) at the
grain boundary. This precipitation causes a depletion of corrosion-resisting elements in
the area surrounding the grain boundary, and this area becomes anodic to the remainder of
the grain.
The austenitic stainless steels comprise the class of materials in which this form of attack
is most common. The precipitate is a chromium carbide that appears at the grain
boundaries during heating between 800°F and 140°F. The depleted component is
chromium and attack occurs in the chromium depleted areas.
The review of LERs revealed that in a number of cases, stainless steel piping in boric acid
systems have experienced intergranular stress corrosion cracking primarily due to chloride
contamination. The details will not be discussed in this report since the subject project
addresses only BAC related degradation.
4.7 Flow Assisted Wastage
In the case of boric acid leak incidents, it would appear that the leaking stream directly
impinging a vessel or pipe wall would accelerate general wastage. An active flow of
borated water over the pipe and vessel walls would also accelerate general wastage type
Corrosion. Limited data exist [25] [28], some of which were taken in a representative jet
impingement environment, and others were obtained using different boric acid flow rates.
These data do not clearly establish the relationship with general corrosion rates (wastage)
of boric acid.
4.8 Summary
There are a number of specific BAC corrosion mechanisms affecting primary plant systems.
The dominant mechanism reported from nuclear plant inspections has been identified as
"wastage" (or general dissolution corrosion) of carbon and low alloy steels. Many general
factors influence the corrosion rates, including acidity, temperature, concentration,
corrosive agent flow rate, oxidation potential of the corrosive agent, aeration, and the
presence of inhibitors or accelerators. Specific factors include steam impingement force,
wicking effect of non-metallic insulation, flow rate through leaking gaskets, presence of
dissimilar metals, metal impurities, applied mechanical stresses, and residual mechanical
stresses. Many of these factors often interact and this interaction may be very complex. It
is important to note that the reviewed literature does not identify the fundamental boric
acid corrosion mechanisms that initiate wastage type degradation.
4-4
Figure 4.1
Cross Section of an Uncorroded Metal Surface
Figure 4.2
General Corrosion Process
4-5
Figure 4.3
Galvanic Corrosion
Figure 4.4
Crevice Corrosion
4-6
Figure 4.5
Oxygen Concentration Cell Corrosion
Figure 4.6
Pitting Attack
4-7
Figure 4.7
Typical Intergranular Corrosion
Figure 4.8
Stress Corrosion Cracking
4-8
5.0 Summary
Boric acid corrosion has been an issue of significant concern for operators of PWR
powerplants since 1968. The period during the 1970's-80's is characterized by numerous
events associated with significant leakages, principally from reactor vessel flanges and
penetrations, steam generator manways and reactor coolant pumps. These events led to
an increasing industry awareness of problems associated with boric acid corrosion.
During the 1980's, NRC provided guidance in this area, in the form of numerous
Information Notices and Supplements, NUREG documents, and culminating in Generic
Letter 88-05. The intent of GL 88-05 was to obtain information from PWR licensees
relative to assessing the safe operability of PWRs when reactor coolant leaks below
technical specification limits degrade carbon or low alloy steel components. It also
required an assurance that licensees would develop and implement a boric acid corrosion
prevention and degradation damage evaluation program. As a result, industry NSSS
owners groups, individual utilities, and EPRI responded to develop additional information
on the effects of BAC, through industry evaluations, surveys, and materials and component
corrosion test programs.
Boric acid corrosion test data presented in Section 3 were reviewed. Figures 5.1 and 5.2
illustrate plots of BAC rates as a function of temperature and boric acid concentration.
There is no adequate data for the temperature ranges between 212°F to 300°F and 350°F
to 500°F. Similarly there is not adequate data for boric concentration in the range of 9000
ppm to 22000 ppm. There is wide variation in the corrosion rates at 212°F and 350°F and
for the entire range of boron concentrations. It is not possible to show any kind of
correlation between the corrosion rates and temperature or boric acid concentration.
Moreover, the methodology employed by each testing organization is different. The
corrosion rates also appear to be sensitive to the methodology.
A number of combinations of factors present beyond the temperature and boric acid
concentration, such as component geometry, boric acid flow rate, and addition of LiOH or
KOH, make it difficult to accurately predict which environments and combination of
factors will result in the maximum BAC rate. The available test data generally measures
degradation rates due to wastage and/or galvanic attack on carbon and low alloy steels.
However, no data has been found which addresses the underlying mechanism(s) of boric
acid corrosion.
In the absence of such data which characterizes the fundamental BAC mechanisms, it is
likely that a very large quantity of component test data will be required to establish a data
base from which predictions of degradation rates in service conditions can be made. At the
same time, it is clear that component geometry, especially as concerns local flow and/or
liquid (or steam) impingement, can strongly influence wastage rates.
5-1
Figure 5.1
Boric Acid Corrosion Rate vs. Temperature
ABB-CE-ABB - Combustion Engineering generated data
BNL - Data gathered or generated by Brookhaven National Laboratory
B&WOG - B&W Owners Group generated data
WSTNG - Westinghouse generated data
5-2
Figure 5.2
Boric Acid Corrosion Rate vs. Boric Acid Concentration
ABB-CE-ABB - Combustion Engineering generated data
BNL - Data gathered or generated by Brookhaven National Laboratory
B&WOG - B&W Owners Group generated data
WSTNG - Westinghouse generated data
5-3
6.0 References
1. NRC IE Information Notice No. 82-02, "Degradation of Threaded Fasteners in the
Reactor Coolant Pressure Boundary of PWR Plants',' June, 1982.
2. W. H. Koo, NUREG-0943, "Threaded-Fastener Experience in Nuclear Power Plants,"

Table 2-3, page 2-12.
3. A. S. O'Neil and J. F. Hall, "Boric Acid Corrosion of Carbon and Low Alloy Steel
Pressure Boundary Components in PWRs," EPRI Report NP-5985, August, 1988, Table
3-1, page 3-3.
4. C. J. Czajkowski, "Survey of Boric Acid Corrosion of Carbon Steel Components in

Nuclear Plants" NUREG/CR-5576, June, 1990, page 2.
5. J. F. Hall, "A Survey of the Literature on Low Alloy Steel Fastener Corrosion in PWR
Power Plants," EPRI Report NP-3784.
6. NRC IE Information Notice 80-27, June, 1980.
7. NRC IE Information Notice 82-06, March, 1982.
8. C. J. Czajkowski, "Boric Acid Corrosion of Ferritic Reactor Components," NUREG/CR-

2827, July, 1982.
9. W. Anderson and P. Sterner, "Evaluation of Responses to IE Bulletin 82-02, NUREG-

1095," May, 1985.
10. NRC IE Information Notice No. 86-108, "Degradation of Reactor Coolant System
Pressure Boundary Resulting from Boric Acid Corrosion," December, 1986.
11. NRC IE Information Notice No. 86-108, Supplement No. 1, April, 1987.
12. NRC IE Information Notice No. 86-108, Supplement No. 2, November, 1987.
13. NRC Generic Letter 88-05, "Boric Acid Corrosion of Carbon Steel Reactor Pressure
Boundary Components in PWR Plants," March, 1988.
14. Bolting Degradation or Failure in Nuclear Plants Seminar, Sponsored by EPRI,

November 2-4, 1983, Knoxville, Tennessee Unpublished Collection of Proceedings.
15 R. E. Nickell, "Degradation and Failure of Bolting in Nuclear Power Plants," EPRI

Report NP-5769, April, 1988.
16 Combustion Engineering, "Elimination of Soluble Boron for a New PWR Design," EPRI
report NP-6536.
6-1
17.Daniel Van Duyne, "Good Bolting Practices-Small Bolts and Threaded Fasteners,"
EPRI report NP-5067V2, December, 1990.
18. John Bickford, Michael Looram, "Good Bolting Practices-Large Bolts," EPRI report NP-
6067V1, 1987.
19. J.E. Looram, M.J. Donovan, Y. Chung, R.A. White, "Guide for Threaded Fasteners,"
EPRI report NP-6316, July 1989.
20. J. Czajkowski, "Bolting Applications," NUREG/CR-3604, May, 1984.

-
21. Alfred Goldberg and Mary C. Jahos, "Lower Bound KISCC Values for Bolting
Materials- A Literature Study, NUREG/CR-2467, February, 1982.
22. "Degradation of Reactor Coolant System Pressure Boundary Resulting from Boric Acid
Corrosion," W. J. Johnson of Westinghouse letter to NRC NS-NRC-87-3260, dated
October 16, 1987.
23. J. F. Hall, "Corrosion of Low Alloy Steel Fastener Materials Exposed to Borated Water,"
Third International Symposium on Environmental Degradation of Materials in Nuclear
Power Systems - Water Reactors, August, 1988.
24. C. J. Czajkowski, "Corrosion and Stress Corrosion Cracking of Bolting Materials in

Light Water Reactors," Proceeding of the International Symposium on Environmental
Degradation of Materials in Nuclear Power Systems - Water Reactors, August, 1983.
25. J. F. Hall, R. S. Pathania, etc., "Boric Acid Corrosion of Carbon and Low Alloy Steels,"
Proceedings of the Fifth International Symposium on Environmental Degradation of
Materials in Nuclear Power Systems - Water Reactors, August, 1989.
26. V. N. Shah and P. E. MacDonald, "Residual Life Assessment of Major Light Water
Reactor Components - Overview," NUREG/CR-4731, November, 1989.
27. "CEOG Work on Alloy 600 Penetration," Presented to Coordinating Group on Alloy 600
Penetrations in PWRs, Palo Alto, California, January 8, 1991.
28. Thomas P. McGee and J. F. Hall, "Boric Acid Erosion-Corrosion of Pressurizer Shell
Material," A Presentation Given by ABB-CE at Altran Corporation in Boston on
October 22, 1992. Proprietary Material Reprinted By Permission.
29. "Summary of B&W Owners Group Boric Acid Corrosion program," A Letter from
Christine A. Campbell of B&W to P. K. Shah of Altran Corporation, September 9, 1992.
30. "Resistance to Corrosion" Inco Alloy International Publication.
31. Reports concerning Boric Acid Corrosion obtained from Halliburton NUS Nuclear
Operations and Maintenance Information Service (NOMIS), March 1992.
6-2
32. "Protect Nuclear Plant Fasteners from Boric Acid Corrosion," N. Moisidis, ABB Impell
Corporation, Power Engineering, March 1992.
6-3
Appendix A—BAC Literature Summary

The BAC literature survey included an extensive search of available databases that
included NRC Document Room, Idaho National Engineering Laboratory Technical Library,
Massachusetts Institute of Technology (MIT) Engineering Library, Nuclear Operations and
Maintenance Information Service (NOMIS), NRC publications and license event reports,
technical journals, and contacts with Owner Group representatives.
The following provides a brief summary of the important BAC related literature. Detailed
discussions on many of the references are presented elsewhere in this report.
1. "Boric Acid Corrosion of Ferritic Reactor Components," by C. Czajkowski, NUREG/CR-

2827, July, 1982 [8].
This document compiles prior BAC data on AISI 4136 steel at 70°F and 140°F.
Additional tests performed by Brookhaven National Laboratory on A193-B7 and AISI
4130 material at higher temperature are also reported. It also describes BAC
degradation incidents that occurred in late 1970's and early 1980's. This study was
sponsored by the NRC.
2. "Threaded-Fastener Experience in Nuclear Power Plants," by W. H. Koo, NUREG-0943,

January 1983 [2].
This report identified forty-four (44) instances of threaded fastener degradation at

nuclear plants between October 1964 and March 1982. The report provided a
comprehensive overview of threaded fastener concerns and regulatory actions (both
short-term and ongoing) related to bolted connections at nuclear plants. The NUREG
identified berated water corrosion as the second most contributing mechanism for
threaded fastener degradation with thirteen of the forty-four instances attributed to
this wastage form of corrosion.
3. Bolting Degradation or Failure in Nuclear Power Plants Seminar, An Unpublished

Proceeding of the Seminar Held in Knoxville, Tennessee by EPRI, November 2-4, 1983
[14].
Numerous presentations were made by the participants covering different types of

bolting failures experienced by the industry. The types of failures reported involved
stress corrosion cracking due to contaminants, fatigue, improper torquing,
erosion/corrosion, and borated water corrosion. Reactor coolant pressure boundary
components affected included steam generator manway closures, reactor coolant
pumps, pressurizer manway closures, reactor vessel closures, chemical and volume
control systems, isolation valves, safety injection check valves and other check valves.
Ongoing and planned efforts of EPRI, NRC, owner groups, vendors, and utilities were
presented in the seminar. (Boric acid corrosion related degradation was one among the
types of degradations that were discussed in the seminar.)
A-1
4. "Corrosion and Stress Corrosion Cracking of Bolting Materials in Light Water

Reactors," by C. J. Czajkowski, Proceeding of the International Symposiums on
Environmental Degradation of Materials in Nuclear Reactors, August 1983 [16].
Brookhaven National Laboratory test data on boric acid corrosion rates for low alloy
fastener material is presented. Failure analysis of cracked bolts from service and the
results of laboratory tests that demonstrate the corrosion mechanisms of bolt cracking
are presented. Stress corrosion cracking due to contaminants such as chloride and
molybdenum disulfide is discussed.
5. "A Survey of the Literature on Low-Alloy Steel Fastener Corrosion in PWR Power
Plants" by J. F. Hall, EPRI NP-3784, December 1984 [5].
This EPRI sponsored report summarizes fastener corrosion events that have occurred
in domestic PWRs since 1968. Laboratory test data related to fastener corrosion are
also discussed.
6. "Evaluation of Responses to IE Bulletin 82-02," by W. Anderson and P. Sterner,

NUREG-1096 [9], May 1985.
In June 1982, NRC issued IE Bulletin 82-02 that required all nuclear utilities to
develop and implement maintenance procedures for threaded fasteners in the reactor
coolant pressure boundary. The bulletin emphasized the importance of training, tool
application, proper torque practice, lubricant applications, gasket installation, and
controls. The scope of action items was limited to threaded fasteners in (1) steam
generator and pressurizer manway closures, (2) valve bonnets, (3) pump flange
connections installed on lines having a nominal diameter of 6-inches or greater,
(4) CRD flange connections, and (6) pressurizer heater connections.
Responses to IE Bulletin 82-02 from utilities were evaluated and closed out in May
1985 through issuance of NUREG-1095. This NUREG evaluated responses from 41
PWR licensees and utilized statistical analyses to determine the important factors
governing the occurrence of bolting degradation in PWRs.
7. NS-NRC-87-3260, A letter from W. J. Johnson of Westinghouse to U.S. NRC dated

October 15, 1987 [22].
The attachment to this letter provides the results of the BAC tests performed by
Westinghouse. It includes limited test data on crevice corrosion due to boric acid.
Corrosion of the reactor vessel head of a European PWR resulting from the leakage of
reactor coolant containing boric acid on to the reactor head is reported.
8. "Degradation and Failure of Bolting in Nuclear Power Plants, Volume I and II," by R.
E. Nickell, EPRI NP-5769, April 1988 [15].
These two volume reports provide an extensive documentation of EPRI organized
Generic Bolted Joint Integrity Program. Numerous topics are covered in this report
that include:
A-2
• Service history of pressure boundary and structural bolts;
• Guidelines for ASME and ASTM Codes and purchase specifications;
• Good bolting practice and preload evaluation;
• Heat treatment and lubricant applications;
• Reference curves for by versus K I S C C and K IC.
Extensive statistical data characterized causes of failures for steam generator manway,
reactor vessel, reactor coolant, valves, and pressurizer man way One of the sections
provided BAC data on A302B plate materials.
9. "Corrosion of Low Alloy Steel Fastener Materials Exposed to Borated Water," by J. F.

Hall of ABB-CE, Third Informational Symposium on Environmental Degradation of
Materials in Nuclear Power Systems - Water Reactors, August 1988 [24].
ABB-CE conducted a series of corrosion tests to evaluate corrosion rates of fastener

materials exposed to berated water. Tests were conducted in a prototypical
environment and steam from a high temperature test facility that calculated berated
water at 600°F impinged directly onto the test specimens that were stressed to 1/2 to
2/3 of yield strength.
The test data also includes corrosion rates of high strength corrosion resistant alloys
and fasteners with surface coatings and treatments.
10. "Boric Acid Corrosion of Carbon and Low-Alloy Steel Pressure Boundary Components in
PWRs" by A. S. O'Neill and J. F. Hall, EPRI NP-5985 Project 2006-18, August 1988 [3].
This report presents a literature survey of the effects of borated water leakage on
carbon and low-alloy steel components (other than fasteners) in PWR applications. The
affected components included a reactor vessel head, pressurizer safety valve, a high-
pressure injection nozzle, and piping. Laboratory BAC test data of ABB-CE for carbon
and low-alloy steel plate and piping products are presented.
11. "Boric Acid Corrosion of Carbon and Low Alloy Steels," by J. F. Hall, R. S. Pathania,
et.al., 5th International Symposium on Environmental Degradation of Materials in
Nuclear Power Plants - Water Reactors, August 1989 [25].
This paper reviews several reported incidents of boric acid corrosion of reactor vessel
heads, nozzles, piping, pressurizer shells, and valves. It also reports results of three
Combustion Engineering conducted series of tests of low alloy steels exposed to borated
water.
12. "Survey of Boric Acid Corrosion of Carbon Steel Components in Nuclear Plants -
Responses to Generic Letter No. 88-05," by C. J. Czajkowski, NUREG-5576, June 1990
[4].
A-3
On March 17, 1988, the NRC issued Generic letter 88-05. The intent of the letter was
to obtain information from PWR licensees relative to assessing the safe operability of
PWRs when reactor coolant leaks (below technical specification limits) degrade
carbon/low alloy steel components. It also required an assurance that licensees will
develop and implement a boric acid corrosion prevention and degradation damage
evaluation program.
Responses were received from the licensees and evaluated by the Chemical Engineering
Branch of the NRC. A total of 50 responses were evaluated by the NRC. All were
found to be satisfactory and met the intent of the Generic Letter. NRC audits were
performed on ten licensees. The audit findings and responses were summarized in this
report.
13. Reports concerning Boric Acid Corrosion obtained form the Halliburton NUS Nuclear
Operations and Maintenance Information Service (NOMIS), March 1992 [31].
These reports are an exchange of information among the participating utilities on BAC
that provide information on maintenance and inspection procedures and practices
related to BAC.
14. "Protect Nuclear Plant Fasteners from Boric Acid Corrosion," by N. Moisdis, M.
Popesca, M.Ratiu ABB Impell Corporation, Power Engineering, March 1993 [32].
This article provides lists of valves and pumps potentially exposed to boric acid
corrosion damage. It also presents guidelines for bolt replacement material and criteria
for valve retrofit program.
15. "Aging Assessment of BWR Standby Liquid Control Systems," by G.D. Buckley, R.D.
Orton, A.B. Johnson, L.L. Larson, Battelle Pacific Northwest Labs., Richland, WA.,
Report No. PNL-8020, August, 1992.
Pacific Northwest Laboratory conducted a Phase I aging assessment of the standby

liquid control (SLC) system used in boiling-water reactors. The study was based on
detailed reviews of SLC system component and operating experience information.
Sources on sodium pentaborate, borates, and boric acid, as well as the effects of
environment and corrosion in the SLC system were also reviewed to characterize
chemical properties and corrosion characteristics of borated solutions. Relatively few
SLC component failures were attributed to sodium pentaborate buildup or corrosion.
The leading aging degradation concern to date appears to be setpoint drift in relief
valves, which has been discovered during routine surveillance and is thought to be
caused by mechanical wear. Degradation was also observed in pump seals and internal
valves, which could prevent the pumps from operating as required by the technical
specifications. In general, however, the results of the Phase I study indicate that age-
related degradation of SLC systems has not been serious.
16. Department of Energy team analyses of Soviet designed VVERs. Revision 1, Analysis
appendices, Report No.: DOE/NE-0086-REV. I Analysis Appendices, September 1989.
A-4
This document provides Appendices A through K of this report. The topics discussed
respectively are: radiation induced embrittlement and annealing of reactor pressure
vessel steels; loss of coolant accident blowdown analyses; LOCA blowdown response
analyses; non-seismic structural response analyses; seismic analyses; "S" seal Integrity;
reactor transient analyses; fire protection; aircraft impacts; and boric acid induced
corrosion
17. "Dissolution and Deposition at the Surface of Authentic PWR Primary Circuit
Corrosion Product Oxide: Results for Cobalt and Manganese at 300°C," by E.W.
Thornton, Central Electricity Generating Board, Berkeley (England), Berkeley Nuclear
Labs, Report No.: CEGB-RD-B-6162/R89, February, 1989.
Metal oxides dissolve and deposit around the primary circuit of a PWR. These processes
are widely believed to be responsible for the net transport of corrosion products from
one part of the circuit to another and for the transport of radioactivity from the core to
out-of-core radiation fields at shutdown. In this report, results are given for the rates at
which cobalt and manganese dissolve from and deposit on a sample of authentic
corrosion product collected from the primary circuit of an operating PWR. The report
discusses the effects of solution pH and boric acid concentration on equilibrium solution
concentration and rates of dissolution and deposition at 300°C.
18. "Survey of Boric Acid Corrosion of Carbon Steel Components in Nuclear Plants," by C.J.
Czajkowski, NUREG/CR-5576, June 1990.
A review of licensee responses to Generic Letter No. 88-05 was performed by the U.S.
Nuclear Regulatory Commission (NRC). The review encompassed 50 satisfactory
responses from the affected licensees. A series of ten (10) joint Brookhaven National
Laboratory (BNL) and NRC audits were performed on a selection of utilities. All of the
licensees audited had program implementations that met the intent of the Generic
Letter. A review of the available literature and the plant audits has led to the
conclusion that the requirements of the Generic Letter have essentially been met for
the 10 plants audited and it is recommended that resident inspectors verify that a
documented and implemented program is in effect at their own plants.
19. "High Temperature Studies of Simulant Fission Products: Part 5. Interaction of Cesium
Iodide with Boric Acid over the Temperature Range 400 to 1000°C," by B.R. Bowsher,
A.L. Nichols, Atomic Energy Establishment, Report No.: AEEW-R-1973, July 1985.
A thermal gradient system has been used to study the interaction of cesium iodide with
boric acid in various atmospheres over the temperature range from 400 to 1000°C.
Specific analytical techniques have been used to determine the reaction products, and
differential thermal analysis-thermogravimetric studies have been undertaken to
assess the reaction kinetics. Solid cesium iodide and molten boric acid react in a
-1
diffusion-controlled manner with an activation energy of 190 +- 30 kJ mol to give
complex cesium berates (e.g., Cs 2 B10 O16 ) and hydrogen iodide. This volatile iodine
A-5
species interacts with 304 stainless steel to produce nickel and iron-based iodides
within the surface oxide layer of the metal.
20. "Review of pH Calculation and Corrosion Product Solubilities under PWR Primary
Coolant Chemistry Conditions," by E. W. Thornton, M. V. Polley, Central Electricity
Generating Board, Berkeley (England), Berkeley Nuclear Labs, Report No.: CEGB-
TPRD/B-0878/R86, December, 1986.
The calculation of high temperature pH in boric acid solutions is discussed and various
relationships for the ionization constant K(W) or ion product Q (W) for water are reviewed.
It is shown that the boric acid equilibria of Mesmer, Baes and Sweeton remain virtually
unaltered when Marshall and Franck's relationship for K (W)is substituted in a re-
analysis of Mesmer, Baes and Sweeton's original experimental data. Magnetite
solubility data and Westinghouse's studies of iron, nickel and cobalt solubility from
mixed ferrites are collated and consideration is given to experimental difficulties that
could have contributed to the variability in the data. Thermodynamic model fits have
been computerized and used to compare different studies and to determine pH values
at which the temperature dependence of solubility is predicted to be zero.
Consideration is given to the differing dependencies of solubility on dissolved hydrogen
concentration in the three model fits. Two models for predicting iron and nickel
solubility with respect to non-stoichiometric nickel ferrites are briefly discussed
showing that only one of these is likely to be credible.
21. "Formation of Magnetite (Fe 3 0 4 ) in Aqueous Media and Properties of the Interface
Magnetite/Solution," by A.E. Regazzoni, Comision Nacional de Energia Atomica,
Buenos Aires (Argentina), Report No.: CNEA-NT-12/84, December, 1984.
The formation of Fe 3 0 4 particles in aqueous media and the properties of the Fe 3 0 4

/aqueous solution interface are studied. This system is of particular interest in nuclear
reactor chemistry, since Fe 3 0 4 was identified as the main component of the corrosion
products of nuclear power plants cooled with pressurized water, of the Atucha I and II,
and Embalse type. Four methods for the synthesis of Fe 3 0 4 are described: a) controlled
oxidation of Fe(OH) 2 in the presence of NaN0 3 at 25°C; b) controlled oxidation of
Fe(OH) 2 in the presence of NaN0 3 and N 2 H 4 and at 100°C; c) alkalinization of F2+ and
Fe 3 + solutions at 80°C; d) simultaneous oxidation and alkalinization of a Fe 2+ The
interfacial properties of Fe 30 4 particles suspended in aqueous solutions of indifferent
electrolytes are described. These properties are essential for the activity transport
associated with the corrosion products. Finally, the adsorption of H3 B0 3, H(n)P04 (n-3)
and n Co(II) in the Fe 3 0 4 /solution interface at 30°C. It is concluded that the adsorbed
species are chemically bonded to surface metal ions.
22. "Investigations Related to the Chemical Behaviour of Methyl Iodide at Severe PWR-
Accidents," by R. Borkowski, Kernforschungszentrum Karlsruhe G.m.b.H. (Germany,
F.R.), Inst. fuer Radiochemie, Report No.: KFK-968, September, 1985.
The decomposition velocity of methyl iodide in aqueous solutions of boric acid has been
measured at temperatures up to 423°K and at chemical conditions which are expected
A-6
to exist in the sumpwater pool during a severe reactor accident. The decomposition was
due only to hydrolysis that increased by the expected amount at high temperature. No
influence of the cooling water additives was observed. Treatment of the available
kinetic data indicated that the influence of polluting material expectedly present in the
sump is likely to be negligible too. A possible exception may be the enhancement of the
decomposition rate by particulate and dissolved silver. The resistance of methyl iodide
to gas phase decomposition by steam and oxygen at 423°K was investigated and only
slow decomposition in the order of 10 -7/s observed which is explained by reaction with
steel surfaces. Neither gas phase oxidation nor hydrolysis occurs at this temperature.
The resistance to oxidation is of kinetic nature. Gas phase hydrolysis is not possible due
to thermodynamics. This was confirmed by the observed gas phase formation of methyl
iodide from hydrogen iodide and methanol at 423°K. The kinetics of this reaction are
best explained by two parallel reactions, one of second order with a kinetic constant of
1.25 x 10 -5/kPa s, and one of third order under action of steam with a constant of 2 x 10-
6 /kPa 2 s.
23."Fluid Composition in the Tube Sheet Crevices of a Nuclear Steam Generator," by O.

Weres, L. Tsao, University of California at Berkeley, Lawrence Berkeley Lab., Report
No.: LBL-20252; CONF-851129-1, October, 1986.
A useful understanding has been gained of the conditions needed for a crevice liquid to
exist and what determines how alkaline this liquid will be. We believe that corrosive
agents other than hydroxides, silica and organic salts must play a role in crevice
corrosion. The presence of organic anions and silica in the condensate argues against a
strongly alkaline crevice environment. In many cases, there may be insufficient caustic
in the crevice fluid to account for the corrosion observed, and another explanation must
be sought. Among the organic compounds, only acetate, formate, and lactate are
quantitatively important, and acetate is innocuous. Formate decomposes to produce
carbonate, which is corrosive. If formate is an important ion in the condensate or
carbonates are found in the crevice, remedial measures should focus on eliminating
formate and bicarbonate from the condensate. Hydrazine and other AVT compounds
should be screened for ability to reduce carbonate to formate and selected accordingly.
The possible corrosiveness of decomposition products of lactate also deserves some
attention. In principle, acetic acid or silica might be added to the condensate in order to
decrease the alkalinity of the crevice liquid, but this benefit must be balanced against
possible harmful effects elsewhere in the system. Adding boric acid to the condensate
will cause a sodium borate liquid to form in the crevices, likewise reducing alkalinity
but with less likelihood of harmful side effects. The activity of NaOH in a sodium
acetate liquid will be controlled by the sodium silicate buffer system in many cases.
Therefore, adding silica will always enable the alkalinity of a sodium acetate liquid to
be reduced to that of the sodium disilicate - quartz buffer. In the absence of acetate,
adding silica will cause the crevice fluid to dry up completely.
24."Relative Importance of Temperature, pH and Boric Acid Concentration on Rates of H2

Production from Galvanized Steel Corrosion," V.M. Loyola, J.E. Womelsduff, Sandia
Labs., Report No.: SAND-82-1179, January, 1984.
A-7
The corrosion of galvanized steel, to produce hydrogen gas, will occur if sprays operate
during a Loss-of-Coolant Accident in a Light Water Reactor. The rates of hydrogen
generation, however, are variable and dependent on accident and post-accident
conditions. This report describes a study designed to identify the important parameters
(temperature, pH, and boric acid concentration) in determining the rates of hydrogen
generation from Light Water Reactor containment building spray solutions. The data
are gathered over a wide range of temperature, pH, and boric acid concentration,
and are used in a two-level, three-factor factorial experiment to determine the relative
importance of the three parameters to the hydrogen generation process. A statistical
treatment of the data gives an indication of the relative importance of the parameters
(temperature, pH, boric acid concentration) and of their interactions. It attempts to fit
the data to a relatively simple equation to model the interactions of the various
parameters.
25. "Degradation of B7 Threaded Fasteners in Borated-Water Solutions at Elevated

Temperatures," Singleton Materials Engineering Labs., Report No.: TVA/OP/EDT-
83/27, July 1983.
The enclosed test program was initiated principally because of NRC IE bulletin No. 82-
02, which reported severe degradation of Grade B7 fasteners in the reactor coolant
pressure boundary due to boric acid attack. Two phases of tests were run at 550°F in
boric acid solution. A third phase test is planned, and the results will be issued as a
supplement to this publication.
26. "Material and Water Chemistry for a Ferritic Reactor Coolant System in Pressure
Water Reactors," by L. Stieding, Bundesministerium fuer Forschung und Technologie,
Bonn-Bad Godesborg (Germany, F.R.), April, 1979.
The use of unplated, low-alloy steels in a boric acid controlled PWR is not considered
possible without changing the water conditions during the start-up and shut-down
periods of the reactor. The significant pH reduction of the water due to boric acid
during these periods most probably leads to damage of the magnetite protective layers
followed by selective corrosion. As this highly important process has not been
sufficiently evaluated with respect to our specific application problem, more detailed
information will be necessary. KWU test facilities provide a means of performing such
tests. In order to avoid corrosion attack during the above operating conditions, an
inhibition of the water with exp 7 Li-borate is recommended which will amount to
approx. DM 60.000, per period of use. In German.
27. "A Comparative Evaluation of Containment Spray Additives. Detrimental Impacts of

an Inadvertent Spray Actuation," by A.K Postma, P.S. Tam, Benton City Technology,
May 1979.
This report presents the results of an engineering analysis of detrimental health and
economic impacts of inadvertent operation of a containment spray system. Injection
spray solutions of plain water, boric acid, boric acid with hydrazine, boric acid with
sodium hydroxide, and sodium borate with sodium thiosulfate were considered in the
present study. It was concluded that none of the solutions would represent an acute
A-8
toxicity problem to people in the containment vessel at the time of an unwanted spray
system trip. Also, corrosion to safety-related equipment was found to be negligible. The
major effect of spray additive type in an inadvertent spray operation appears to be the
level of effort required to recover from the incident. None of the detrimental effects
brought to light in this study appear to play a determining role in the selection of spray
additive type.
28. "Evaluation of Replacement Bolting in Critical Applications," by H. McShane,

Conference on Improved Technology for Critical Bolting Applications, July, 1986.
SA-193 Grade B7 carbon steel bolting has found numerous applications in the nuclear
power industry. The high tensile strength of this material makes it ideally suited to the
task of holding together flanged connections that may experience pressures as high as
2250 psia. A problem often experienced with this material is that it is highly
susceptible to corrosion in a boric acid environment. This is especially true at the
elevated temperatures typically encountered in a nuclear power plant. This paper
provides a calculational method for determining the adequacy of lower strength
replacement bolting or degraded carbon steel bolting in typical nuclear power plant
applications.
29. "Effect of Boric Acid Treatment on the Secondary Cycle at ANO 2," November, 1985.
This report reviews EPRI report NP-4270 that contains data from one plant on the
effect of boric acid use. Preliminary indications show that boric acid addition may have
minimal impact on secondary-cycle chemistry, except that corrosion of copper alloys
may be increased.
30. Computer Program for Constructing Stability Diagrams in Aqueous Solutions at

Elevated Temperatures, by C.M. Chen, K. Aral, Babcock & Wilcox Co., April, 1982.
When a metal is placed into an aqueous solution containing various chemicals (oxygen,
boric acid, sulfur compounds, chlorides, fluorides, phosphates, etc. ), corrosion problems
are often encountered especially under the high temperature and high pressure
conditions present in nuclear power systems. Stability diagrams of metals in aqueous
solutions (better known as potential-pH or Pourbaix diagrams) which depict the
equilibrium phases of metal, its oxides, its ions, or metal compounds associated with
the solution's additives are a valuable means of predicting corrosion. A computer
program has been developed for calculating electrode potentials and pHs of reactions
and delineating the predominant area of each chemical or ionic species in aqueous
solutions. Input is kept to a minimum through an expandable library of species
(established as a data base), allowing the retrieval, by name, of thermodynamic and
free-energy data. The library contains various compounds of elements Fe, S. Ni, Cr. C,
Ni, Ti, B. Zr, and Cu. The free-energy value can be retrieved at any temperature from
25 to 300°C. The stability diagram can be obtained continuously from room
temperature up to 300°C and plotted on a digital plotter.
A-9
31. "Evaluation of Corrosion Resistance of Materials of Sealing Surfaces for Steam and
Nuclear Power Plant Fittings," by B. Ya Ivnitskii, L.G. Umanskaya, R.I. Babkina,
Energomashinostroenie n 4, April 1981.
Results of investigation and evaluation of corrosion resistance in the medium of steam,

distilled water, and boric acid of materials of sealing surfaces of fittings (the alloys,
TsN-6, TsN-12, V3K, KhN80SR2) are presented. It is shown that corrosion resistance of
these surfacing alloys in distilled water and boric acid is high and practically equal. In
Russian.
32. "Investigation of Corrosion Resistance of Steels in the Aqueous Heat Carrier of a VVER
Nuclear Reactor with Boric Control," by V.I. Nikitin, T. Ya Karpova, S.P. Adamova,
A.G. Timofeev, Energomashinostroenie n 3, March, 1980.
Results of an investigation of the regularities of the corrosive effect of an aqueous heat
carrier with boric acid on steels are presented. Steels of different types -- austenitic
stainless, martensitic-ferritic stainless, pearlitic, etc. -- were studied. The tests were
carried out in autoclaves at the temperature of 20-300° for 100 to 3000 h. It is found
that austenitic and martensitic-ferritic stainless steels are corrosion resistant. Pearlitic
steels are prone to corrosion. An increase in chromium content up to 5% improves
corrosion resistance of steels. In Russian.
33. "The Solubility of Magnetite and Nickel Ferrite in Aqueous Solutions," by Jin Yop
Chung; Kun Jai Lee, Proceeding of the JAIF International Conference on Water
Chemistry in Nuclear Power Plants - Operational Experience and New Technologies for
Management, April, 1988.
The effects of temperature, pH, and hydrogen concentration on the solubility of major
corrosion products, magnetite and nickel ferrite, were studied. Magnetite is the main
corrosion product generated in steam generators, while nickel ferrite is the one
depositing on the fuel cladding of pressurized water reactors. Measurements about the
solubility of magnetite by a static method were made in an autoclave. The experimental
results are compared with the values obtained by the theoretical calculation with a
thermodynamic analysis. The calculations were performed for magnetite and
stoichiometric nickel ferrite as a typical corrosion products generated in PWR coolants
by utilizing the known and published thermodynamic properties. Input data for the
thermodynamic analysis is examined in detail because it affects the results critically.
The pH values that give the minimum solubilities of the nickel ferrite were determined
by the temperature coefficient of solubility being zero. These pH values can be adjusted
by controlling the concentrations of lithium hydroxide since the boric acid are used as
chemical shim. In addition, the effects of pH and the boron concentrations on the
deposited iron oxide to the primary system were calculated using an empirical
relationship. The result showed that the deposition of the corrosion products can be
controlled by a combination of lithium and boron concentrations. Other analyses about
the solubilities are discussed with respect to the deposition of corrosion products to the
primary systems in nuclear power plants.
34. "Study of the Hydrogen Evolution Accompanying Zinc Corrosion", by H. Grunert, J.

Hagedorn, H. Stewig, H. Wilhelm, Akademie der Wissenschaften der DDR, 1991.
A-10
As a result of testing in autoclaves and in streaming water vapor, it was shown that the
corrosion of Zn at temperatures > 100°C is accompanied by H evolution not only in pure
boric acid solution and in nearly neutral potassium borate solution but also in water
vapor. In immersion tests at temperatures < 100°C a marked corrosion of Zn was
observed only in pure boric acid solutions and not in potassium borate solution. The
results obtained enable the estimation of the amount of evolved H by the corrosion of
hot-dip galvanized structural steelwork during an accident in a nuclear power plant
operating with pressurized water reactors. In German.
35. "Stress Corrosion Cracking of Austenitic Stainless Steels in Borated Water," by H.

Tsuge, N. Maruyama, S. Nagata, H. Nagano, Conference: Environmental Degradation
of Materials in Nuclear Power Systems--Water Reactors, August, 1983
Intergranular stress corrosion cracking (IGSCC) in the heat affected zone of welded 304
stainless steel in stagnant borated water in PWR plants has been reported. However,
the effect of the boric acid itself on the IGSCC is still ambiguous. The IGSCC of
sensitized 304 and 316 stainless steels in borated water at low temperature has been
studied using the double U-bend SCC test, the slow extension rate test (SERT) and the
newly developed potential controlled double U-bend test. The IGSCC of stainless steels .
is accelerated in borated water (1500 ppm B as H 3 BO 3 ) in both the double U-bend SCC
tests and potential controlled double U-bend SCC tests compared with in pure water.
36."French Experience in Water Chemistry Tests and the Monitoring of Corrosion

Products in PWRs," by P. Berge, P. Jeanson, G. Lecointre, P. Besly, G. Frejaville,
Conference: Water Chemistry and Corrosion Problems in Nuclear Power Plants.,
November, 1982.
Special requirements are imposed on the chemical conditioning of the primary coolant
of PWRs in order to limit the activity of the circuits due to corrosion products. The
"decreasing lithia " criterion, which makes the lithia concentration dependent on the
boric acid content, is applied at all Electricite de France PWR power plants. Measuring
dose rates at different reactors will confirm that this treatment has a beneficial effect.
The first results are expressed through a primary circuit activity index that is obtained
from measurements of the dose rate at various points in the circuit. In French.
37. "Experimental Modeling of Hydrogen Evolution Rates From Surfaces With Exposed
Zinc Metal or Zinc Primer Coating in Contact With Containment Sprays in PWR
Nuclear Power Plants," by L.E.S. Smith, R.O. Lane, W.A. Van Hook, June 1981.
The interiors of pressurized water reactor containments have some construction

materials and equipment surfaces coated with Zn primer and galvanized Zn. In the
event of a loss-of-coolant accident, the surfaces may be exposed to hot spray solutions.
The spray solutions recommended by the nuclear industry contain boric acid buffered
with either sodium hydroxide or trisodium phosphate. These solutions react with Zn to
produce gaseous H and Zn compounds. Rates of H evolution from Zn surfaces immersed
in spray solution containing boric acid and sodium hydroxide (pH = 9.8), and solution
containing boric acid and trisodium phosphate (pH = 7.5) are correlated
mathematically.
A-11
38. "The Effect of Boric Acid, Chlorides and Potash Lye on the Corrosion of Austenitic
Stainless Steels in Hot Water Solutions," by J. Fliz, M. Ziomek-Moroz, July 1980.
Corrosion problems of construction materials exposed to aggressive agents mentioned

in the title occur in nuclear power plant installations using boron control. A series of
corrosion tests has been done. From the results, the most resistant steels have been
chosen. The effect of chloride ions on the corrosion of steels tested at various pH values
and various aggressive agent contents have been discussed. In Polish.
A-12
Appendix B—NRC Licensing Event Reports (LERs)
EVENT DATE: 03/20/92 LER #: 92-018-00 REPORT DATE: 06/22/92
FACILITY NAME: Fort Calhoun Station Unit No. 1
DOCKET NUMBER: 05000285
TITLE: Corrosion of Boric Acid System Bolts
OTHER FACILITIES INVOLVED: N DOCKET NO: 05000
THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR

SECTION: 50.73(a)(2)(ii) & 50.73(a)(2)(vii)

NAME: Scott A Lindquist, Shift Technical Advisor
TELEPHONE: (402) 533-6829
COMPONENT FAILURE DESCRIPTION:

CAUSE: SYSTEM: COMPONENT: MANUFACTURER:
REPORTABLE NPRDS:
SUPPLEMENTAL REPORT EXPECTED: NO
ABSTRACT:
On March 20, 1992, with the plant in a refueling outage, the removal of insulation and
heat tracing from the Boric Acid System during implementation of a modification revealed
severe corrosion of the carbon steel fasteners on the boric acid pump flanges and piping
supports. The corrosion was extensive enough to have led to a possible failure of the
fasteners during a seismic event. This event is being reported pursuant to 10 CFR
50.72(b)(2)(i). This report is being submitted pursuant to 10 CFR 50.73(a)(2)(ii) and 10
CFR 50.73(a)(2)(vii).
anticipate corrosion problems due to boric acid leakage at the system flange connections.
The carbon steel fasteners were covered with glued heat tracing and asbestos insulation,
thus, sealing the fasteners in a potentially high corrosive environment.
Based on the configuration of the charging header and the availability of operator actions
to mitigate the consequences of a complete loss of Boric Acid Storage Tank (BAST)
inventory, the safety significance of the degraded flanges was minimal. The postulated
loss of a BAST would not significantly affect plant safety.
B-1
As corrective action, the carbon steel fasteners were replaced. Long term corrective actions
include evaluating acceptability of replacing carbon steel fasteners with stainless steel
fasteners, incorporating the Boric Acid System into the Boric Acid Corrosion Program and
performing a review of the current program scope to determine if other carbon steel
fasteners are exposed to similar conditions and possible damage.
(END OF ABSTRACT)
TEXT
The major components of the Boric Acid System are two concentrated boric acid storage
tanks (BAST) CH-llA and CH-llB, and two boric acid pumps, CH-4A and CH-4B. The
safety function of the system is to provide a flow path for concentrated boric acid from the
BASTs to the charging pumps for injection into the Reactor Coolant System (RCS). Each
concentrated BAST containing 2.6 to 4.5 weight percent boric acid has sufficient boron to
bring the plant to a shutdown condition with the most limiting control rod stuck in the
fully withdrawn position. Each boric acid pump has sufficient capacity to feed all three
charging pumps.
On March 20, 1992, as part of modification MR-FC-89-076, which removed system relief
valves and heat tracing from the Boric Acid System piping, insulation was removed from
the piping around the boric acid pumps (CH-4A and CH-4B). After removal of the
insulation, it was discovered that some carbon steel restraint components (U-bolts) and
pump flange bolts were corroded from contact with leaking boric acid. The affected piping
is considered to be seismically qualified as Critical Quality Equipment (CQE) pressure
retaining components. The piping has been designated per the FCS CQE List as an
American Society of Mechanical Engineers (ASME) Section III, Class 3 subsystem.
Engineering performed an inspection of the corroded fasteners and restraints to evaluate

the structural integrity of the piping system. This evaluation concluded that given the
condition of the fasteners on the boric acid pumps, during a seismic event the flange
connections may have failed. The corrosion damage on the restraints was judged to have
no adverse effect on supported piping since undamaged adjacent restraints provided
adequate support. The inspection disclosed that all of the fasteners associated with the
inlet and outlet flanges of both pumps were severely corroded. The joints, although
degraded, have been demonstrated by past surveillance tests to be free of significant
leakage during normal operation. Although not verifiable, a seismic event could cause a
common mode failure of the joints of both pumps with the potential for a complete loss of
boric acid inventory from the BASTs (CH-11A and CH-11B) due to failure of the system
pressure boundary. In addition, continued degradation would likely have resulted in
failure of the joints if the condition was not discovered and corrected.
At the time this condition was discovered, the unit was in a refueling outage and the Boric
Acid System was out of service for modification. With the plant shutdown, the FCS
Technical Specifications require a single operable boric acid flow path. This requirement
was met by another flow path which was unaffected by the corrosion damage. After
discovery of this condition, the degraded fasteners were replaced with new carbon steel
B-2
fasteners. Therefore, no safety concern existed at the time of discovery and the condition
was corrected prior to start-up.
Following the refueling outage, a thorough review of the as-found condition of the flange
bolts determined that they could not perform their design function. This would cause both
trains of boric acid to be considered inoperable per the requirements of Technical
Specification 2.2. Thus, this condition was determined to be reportable and a four-hour
phone call was made to the NRC on May 22, 1992, pursuant to 10 CFR 50.72(b)(2)(i). This
report is being submitted pursuant to 10 CFR 50.73(a)(2)(ii) and 10 CFR 50.73(a)(2)(vii).
The safety function of the piping/Flanges and related pumps is to provide a flow path from
the BASTs through the boric acid pumps to the suction of the charging pumps. The safety
concerns raised by the postulated failure of the flange bolts are:
1) Part or all of the inventory of the BASTs is lost through leakage. This reduces the total
amount of boric acid injected into the RCS to maintain 4% shutdown margin.
2) Penetration M-3 (charging header containment penetration) relies on system pressure

from the discharge of the charging pumps to assure containment integrity during
approximately the first thirty minutes of an accident. During accident conditions, the
initial source of suction for the charging pumps is the BASTs which automatically align
on a Safety Injection Actuation Signal (SIAS). The BASTs could potentially be lost by a
failure of the affected flanges.
To address the affect on shutdown margin, the Updated Safety Analysis Report (USAR)
Chapter 14 accidents and events were evaluated to assure that the BAST inventories were
not credited in any of the analyses. The results of the evaluation determined that no credit
is taken for the mitigation of any USAR Chapter 14 accidents or events with the BAST
inventories.
The BAST inventories are also used for cooldown scenarios. The cooldown of the RCS
requires the use of the BAST inventory for coolant shrinkage makeup and for RCS boration
to maintain a 4% delta K/K shutdown margin. However, an alternate source of berated
water is available from the Safety Injection and Refueling Water Tank (SIRWT) via the
safety injection pumps which has been shown to be sufficient to provide a 4% shutdown
margin for actual operation in previous cycles. The path from the SIRWT is not affected by
the corrosion problem.
The loss of BAST inventory due to a failure of the flanges would not affect the ability of the
plant to ensure that adequate shutdown margin is maintained during accident conditions.
For shutdown scenarios, the ability of the plant to cooldown would not be significantly
affected. Therefore, the loss of the BAST(s) does not significantly affect plant safety.
The second safety concern evaluated is related to the adequacy of penetration M-3 to
maintain containment integrity. The penetration was originally designed with one check
valve (CH-198). The design of the penetration relied on pressurizing the penetration to
greater than containment pressure with the charging pumps or a water seal of sufficient
height to prevent exfiltration from containment once the BAST and/or SIRWT were
B-3
depleted and the charging pumps were stopped (approximately 30 minutes, post-LOCA). A
change in the containment transient analysis in 1991, increased the pressure in
containment at the time the charging pumps were stopped. The pressure in the new
analysis exceeded the head of the water seal and invalidated the original design basis.
Therefore, Penetration M-3 was determined to be outside the Fort Calhoun Station (FCS)
design basis and LER 91-003 was issued. LER 91-003 concluded that if the check valve
(CH-198) failed to isolate containment atmosphere, there were multiple other check valves
in the system to prevent backflow out of containment. There are check valves on the
discharge side of the charging pumps and check valves internal to the charging pumps that
are leak tight. These check valves are routinely verified to be operable in the reverse
direction because one charging pump is normally idle and leakage would be indicated by
system performance.
As part of the corrective actions identified in LER 91-003, changes were incorporated into
the Emergency Operating Procedures (EOPs) to ensure that the penetration remains
pressurized using the High Pressure Safety Injection Pumps once the charging pumps are
stopped. The EOPs also require that the charging pumps be aligned to the SIRWT on low
BAST level alarms. This ensures that the penetration will remain pressurized until
containment pressure drops below the water seal head.
Based on the configuration of the charging header and the availability of operator actions
to mitigate the consequences of a complete loss of BAST Inventory, the safety significance
of the degraded flange fasteners is minimal.
anticipate corrosion problems from the boric acid leakage at the system flange connections.
Carbon steel fasteners were covered with glued heat tracing and asbestos insulation, thus,
sealing the fasteners in a potentially high corrosive environment.
At the time the condition was discovered, the unit was in a refueling outage (Mode5) and
the Boric Acid System was out of service for modification. After discovering the problem,
the degraded fasteners were replaced with new carbon steel fasteners prior to returning
the Boric Acid System to service. The highly corrosive environment has been partially
eliminated by the implementation of modification MR-FC-89-076 which removed the heat
tracing and asbestos insulation and replaced them with removable non-asbestos insulation.
To ensure that no other fasteners were similarly affected by boric acid leakage, the
remainder of the Boric Acid System was inspected and the results indicated that one other
flanged valve had corroded flange bolts which were replaced.
The following corrective actions will be completed:
1) Applicable components, as determined by Engineering evaluation, in the Boric Acid

System that are not currently part of the Boric Acid Corrosion Prevention (BACP)
program will be added to the scope of the BACP program. This will ensure that
periodic inspections are performed to identify areas of leakage before the fasteners
become damaged. This will be completed by November 27, 1992.
B-4
2) A review will be performed of the existing scope of the BACP program to ensure that
equipment in other borated water systems are not similarly affected by unidentified
leakage (i.e., under insulation). This will be completed by November 27, 1992.
3) Design Engineering will evaluate the acceptability of replacing the carbon steel
fasteners in the Boric Acid System with stainless steel fasteners to eliminate corrosion
problems due to boric acid leakage, by August 15, 1992. If acceptable, these fasteners
will be replaced during the next Refueling Outage.
There have been no previous similar events at FCS concerning corrosion of Boric Acid
System bolts.
FACILITY NAME: Waterford Steam Electric Station, Unit 3
TITLE: Reactor Shutdown due to Reactor Coolant System Leakage in excess of Technical
Specifications
OTHER FACILITIES INVOLVED: N/A DOCKET NO: 05000

SECTION: 50.73(a)(2)(i)

NAME: J.G. Hoffpauir, Maintenance Superintendent
TELEPHONE: (504) 464-3138

CAUSE: SYSTEM: COMPONENT: MANUFACTURER:
REPORTABLE NPRDS:
SUPPLEMENTAL REPORT EXPECTED: No
ABSTRACT
At 0703 on July 11, 1992, Waterford Steam Electric Station Unit 3 declared an Unusual
Event as a result of unidentified reactor coolant system leakage greater than the Technical
Specification 3.4.5.2 limit of one gallon per minute. The reactor was shut down and the
source of the leakage determined to be the packing area of Reactor Coolant Hot Leg
Sample Valve RC-104.
B-5
This event resulted from the failure of a temporary leak repair made to RC-104 after the
valve's packing gland studs failed due to boric acid corrosion on March 26, 1992. As a
result, the root cause of this event and the earlier failure are the same: the use of a
material for the studs in an application for which it proved to be inadequate. That is, the
valve was installed in a system where possible packing leakage could expose the studs to
boric acid despite the fact that the studs were made of a material that is susceptible to
boric acid corrosion. RC-104 will be repaired or replaced during the next refueling outage.
Also, the stud material for similar valves will be evaluated. All leakage resulting from this
event was confined to the Reactor Containment Building and therefore, the health and
safety of the public and plant personnel were not compromised.
(END OF ABSTRACT)
REPORTABLE OCCURRENCE
At 0703 on July 11, 1992, Waterford Steam Electric Station Unit 3 entered Technical
Specification (TS) 3.4.5.2. Action Statement 'b' as a result of Reactor Coolant System (RCS;
EIIS Identifier AB) unidentified leakage greater than the limit of one gallon per minute.
In accordance with the action statement, a reactor shutdown commenced at 0658. The
reactor was manually tripped in accordance with Operating Procedure OP-010-001,
"General Plant Operations," at 1027. 10CFR50.73(a)(2)(i)(A) indicates that the completion
of any nuclear plant shutdown required by the plant's Technical Specifications is a
reportable event.
INITIAL CONDITIONS
Plant Power: 100% Mode: 1
Procedures Being Performed Specific to this Event: None
Technical Specification LCO's in Effect Specific to this Event: None
Major Equipment Out of Service Specific to this Event: None
EVENT SEQUENCE
At 0634 on July 11, 1992, Off-Normal Procedure OP-903-023, "Reactor Coolant System
(RCS) Leak," was entered based on a number of indications which pointed to the existence
of a RCS leak: decreasing level in the Volume Control Tank (VCT), a mismatch between
equilibrium charging flow and combined letdown and reactor coolant pump controlled
bleedoff flow, and a high leak rate indication from the Containment Liquid Leak Recorder
(SP-IFR-6710). At 0658, a water inventory balance performed over a period of 13 minutes
indicated the presence of a leak of about 7.65 gallons per minute (gpm).
At 0703, Waterford 3 declared an Unusual Event and entered Technical Specification (TS)
3.4.5.2, "Reactor Coolant System; Operational Leakage," as a result of RCS unidentified
leakage greater than the limit of one gallon per minute. Boration for the required reactor
shutdown began at 0658; the shutdown was complete at 1027.
B-6
At 0727, personnel entered the Reactor Containment Building (RCB; EIIS Identifier NH) to
perform a RCS inspection. A video camera installed to monitor RC-104 after the March
event was checked but steam in the area and/or condensation on the camera lens prevented
it from providing useful information. At 0734, RC-104 was identified as the source of the
leak based on the RCS inspection.
RC-104 is a 3/4 inch diameter globe valve with a pneumatic operator. The valve is located
in the Reactor Containment Building in piping connected to the RCS loop 1 hot leg. Only a
flow-restricting orifice separates RC-104 from the RCS. The valve is operated remotely
from a panel in the Reactor Auxiliary Building (RAB; EIIS Identifier NF) and functions to
isolate the RCS from the Primary Sample System (EIIS Identifier AB). In addition, the
RCS loop 1 hot leg, via RC- 104, is also the source of reactor coolant samples to be analyzed
by the Post- Accident Sample System (PASS; EIIS Identifier IP). Two containment
isolation valves (EIIS Identifier JM-ISV), Reactor Coolant Sample Header Inside
Containment Isolation Valve PSL-105 and Reactor Coolant Sample Header Outside
Containment Isolation Valve PSL-107, are located in the piping downstream of RC-104.
As described in Waterford 3 LER 92-002, RC-104 failed previously on March 26, 1992. In
that event, packing gland stud corrosion from exposure to boric acid led to stud failure, loss
of packing material, and eventually to a reactor shutdown because of unidentified RCS
leakage greater than that allowed by Technical Specifications. Evaluation later revealed
that the cause of the stud failure was that the studs were made from carbon steel (a
material that is susceptible to boric acid corrosion) and used in an application where
exposure to boric acid was possible. RC-104 was eventually leak repaired by sealant
injection and left in the open position in order to obtain reactor coolant samples from the
RCS loop 1 hot leg. As reflected in LER 92-002, permanent corrective action- either repair
or replacement- is scheduled for the upcoming refueling outage.
In this most recent event, with the source of the leakage identified and the magnitude of
the leak rate understood, efforts focused on reducing the leakage from the damaged valve
until repairs could be made. The primary objective of these efforts was to close RC-104 in
order to isolate the leaking packing area from RCS pressure. This process was complicated
to a certain extent by the temporary repairs made after the March 25, 1992 event but the
valve was eventually shut at 2215. Shutting RC-104 reduced but did not eliminate the
leakage from the packing area, in part because the RC-104 valve stem is bent (noted
during the March event; see LER 92-002).
With the valve now closed and the leak rate reduced, leak repair efforts began. RC-104
had been left open after the March event to provide a coolant sample path. A strongback
and a clamp assembly were installed to hold the packing and packing retainer plate in
place and a leak repair sealant was injected into the packing area by means of an existing
threaded connection (Condition Identification/Work Authorization 279662/01091834). The
strongback and clamp assembly were still installed when the temporary leak repair
performed after the March event failed.
Given the failure of the first leak repair, the decision was made to perform the second leak
repair with RC-104 closed. This approach takes advantage of the RC-104 mainseat which,
with the valve shut, provides a barrier between the RCS and the damaged parts of the
B-7
valve. Because of valve damage identified after the first event, the mainseat does not
provide absolute isolation. Nevertheless, it will serve to limit any future leakage and
increase the integrity of the temporary leak repair sealant. A temporary sample path was
identified by which PASS samples may be obtained from the pressurizer surge line (EIIS
Identifier AB-PZR). In any event, with RC-104 closed leak repair sealant was injected into
the packing area through an existing threaded connection. The leak repair to RC-104 was
complete by 0000 on July 12, 1992. At 0507 on July 12, Operating Procedure OP-903-024,
"Reactor Coolant System Water Inventory Balance," was completed indicating a total RCS
leak rate of 0.527 gpm of which 0.067 gpm was identified leakage and 0.460 gpm was
unidentified leakage.
At 0508, with repairs to RC-104 complete and the unidentified leakage rate less than the
Technical Specification 3.4.5.2 limit, Waterford 3 exited the action statement and
terminated the Unusual Event.
EVENT CHRONOLOGY
July 11, 1992
0634 Entered Operating Procedure OP-903-023, "Reactor Coolant System (RCS) Leak,"
based on leak indications provided by lowering VCT level and mismatch between
charging and letdown flow.
0658 Leak rate calculation performed over 13 minutes indicates presence of 7.65 gpm leak.
Commenced boration for reactor shutdown.
0703 Declared Unusual Event as a result of the RCS leakage. Entered TS 3.4.5.2.
0727 Personnel entered Reactor Containment Building for RCS inspection.
0734 RCS inspection revealed RC-104 as the source of the leak.
0830 Attempt made to close RC-104 remotely is unsuccessful.
1027 Reactor shutdown complete.
1135 Commenced reducing RCS pressure.
1226 Personnel enter Reactor Containment Building to evaluate the scope of the necessary
repairs
1907 Commenced RCS cooldown.
1934 Completed OP-903-024. Total RCS leakage: 10.0031 gpm, Unidentified leakage:
9.9478 gpm, Identified leakage: 0.0553 gpm.
2130 Secured RCS cooldown.
B-8
2215 RC-104 closed by remote indication.
2250 RC-104 verified closed.
July 12, 1992
0000 (approx) Leak repair of RC-104 complete.
0507 Completed OP-903-024. Total RCS leakage: 0.527 gpm, Unidentified leakage: 0.460
gpm, Identified leakage: 0.067 gpm.
0508 Exit TS 3.4.5.2. Secure from Unusual Event.
CAUSAL FACTORS
The RCS leak and subsequent reactor shutdown described in this PRE is an outgrowth of
the RC-104 failure that occurred on March 27, 1992. As such, the root causes of the two
events are the same. As described in LER 92-002, the packing gland studs on RC-104
failed due to boric acid corrosion because they were made from material that was not
appropriate for the application. The July 11, 1992 unidentified leakage and subsequent
reactor shutdown were caused by the failure of a temporary repair made to RC-104 after
the earlier event. Although an element of risk was known to exist, it was anticipated that
the temporary leak repair made to RC-104 would besufficient to allow operation of the
plant until the valve could be reworked during the upcoming refueling outage. Given the
difficulty involved in repairing or replacing RC-104 (for example, the valve is non-isolable
from the RCS without a freeze seal), the decision to delay permanently repairing RC-104 to
allow for thoughtful work and contingency planning was in the best interests of safety.
CORRECTIVE MEASURES
RC-104 was closed and leak repaired by sealant injection (Condition Identification/Work
Authorization 279796/01096110). The strongback and the clamp assembly installed after
the March 25, 1992 failure are still installed. The leak repair sealant was injected into the
packing area through an existing threaded connection.
Temporary Alteration Request (TAR) 92-020 was installed to provide a temporary flowpath
around RC-104 to obtain routine reactor coolant samples as well as samples for the Post
Accident Sampling System (PASS). Both PASS and routine coolant samples will be drawn
from the pressurizer surge line until RC-104 is repaired or replaced (This approach was
discussed with NRC representatives after the March, 1992 failure of RC-104 but was not
implemented at that time.). RC-104 will remain closed to isolate the damaged area of the
valve until permanent corrective action can be taken. An engineering evaluation
performed as part of the development of TAR 92-020 demonstrated that the pressurizer
surge line was an acceptable alternative to the RCS loop 1 hot leg for obtaining PASS
samples.
No additional corrective action beyond that described in LER 92-002 is necessary to

address this event. As discussed there, RC-104 will be permanently repaired or replaced
B-9
during the next refueling outage. Also discussed there are plans to prioritize and then
evaluate other valves whose studs may be susceptible to boric acid corrosion and can
possibly be exposed to boric acid leakage.
SAFETY SIGNIFICANCE
Plant operators took appropriate action to shutdown the reactor, stabilize RCS parameters,
minimize RCS leakage, and ensure that individuals and organizations were notified of the
event in a timely manner. As a result, leakage was confined to the Containment Building.
The health and safety of the public and plant personnel were not compromised by this
event.
SIMILAR EVENTS
LER 92-002 reported a failure of RC-104. The event described in this LER is an outgrowth
of the earlier event. Details of that earlier event are included throughout this LER.
LER 91-008 reported unidentified RCS leakage of approximately 20 gpm while in Hot
Shutdown. TS 3.4.5.2 was entered and an Unusual Event was declared. Leakage was
determined to be from a bonnet bolt hole on Pressurizer Spray Check Valve (EIIS Identifier
AB-V) RC-303 and was caused by improper assembly of the gasket retaining ring due to
underdeveloped maintenance training. There is no similarity in causal factors.
LER 88-015 reported unidentified RCS leakage of 4 gpm which was attributed to leakage
past two in-series Reactor Coolant Loop Drain Valves (EIIS Identifier AB-DRN) RC-206A
and RC-207A. The valves had been hand closed and were not fully shut. After shutting
both valves with a valve wrench, the leakage dropped to within TS limits. This event is
similar only by the fact that it also required entry into TS 3.4.5.2 and declaration of an
Unusual Event.
LER 85-018 reported unidentified RCS leakage of 6.1 gpm which was later determined to
be from the packing glands of Pressurizer Spray Valves (EIIS Identifier AB-FCV) RC-301A
and RC-301B. TS 3.4.5.2 was entered and the reactor was shutdown. Station Modification
926 was issued to improve the reliability of the packing design on these valves by utilizing
a different type of packing and adding additional rings of packing. There is a similarity in
the leakage source but the root cause of the leakage is different. There were no generic
concerns identified in LER 85-018.
LER 85-013 reported unidentified RCS leakage of 1.7 gpm which was later attributed to
failure of Letdown Volume Control Tank Relief Valve (EIIS Identifier AB-V) CVC-115 to
properly seat after it inadvertently lifted a day before. CVC-115 was subsequently
repaired. This event is similar only by the fact that it also required entry into TS 3.4.5.2
and declaration of an Unusual Event.
B-10
FACILITY NAME: Waterford Steam Electric Station Unit 3
TITLE: Reactor Shutdown due to Reactor Coolant System Leakage in Excess of Technical
Specifications
OTHER FACILITIES INVOLVED: N/A DOCKET NO: 05000

SECTION: 50.73(a)(2)(i)

NAME: W.R. Brian, Plant Engineering Superintendent
TELEPHONE: (504) 464-3127

C A U S E : SYSTEM: COMPONENT: MANUFACTURER:
REPORTABLE NPRDS:
ABSTRACT:
At 0248 hours on March 25, 1992, Waterford Steam Electric Station Unit 3 declared an
Unusual Event due to Reactor Coolant System unidentified leakage in excess of the
Technical Specification 3.4.5.2 limit of 1 gallon per minute. The reactor was shut down
and the source of the leakage was subsequently determined to be the packing area of
Reactor Coolant Hot Leg Sample Valve RC-104.
The packing gland studs on RC-104 failed due to boric acid corrosion. The root cause of this
event was use of a material in an application for which it proved inadequate; that is, the
studs were made of a material that is susceptible to boric acid corrosion, and the valve was
used in a system where possible packing leakage could expose the studs to boric acid. RC-
104 will be repaired or replaced during the next refueling outage and the stud material for
similar valves will be evaluated. All leakage due to this event was confined to the Reactor
Containment Building, and therefore the health and safety of the public and plant
personnel was not compromised.
(END OF ABSTRACT)
B-11
REPORTABLE OCCURRENCE
At 0028 hours on March 25, 1992, Waterford Steam Electric Station Unit 3 entered
Technical Specification (TS) 3.4.5.2 Limiting Condition for Operation (LCO) as a result of
Reactor Coolant System (RCS) (EIIS Identifier AB) unidentified leakage greater than the
TS limit of 1 gallon per minute (gpm). In accordance with the TS Action requirements,
shutdown of the reactor (EIIS Identifier AB-RCT) commenced at 0247 hours and the
reactor was manually tripped per Operating Procedure OP-010-001, General Plant
Operations, at 0545 hours. 10CFR50.73(a)(2)(i)(A) states that completion of any nuclear
plant shutdown required by TS is a reportable event.
INITIAL CONDITIONS
Plant Power: 100% Mode: 1
Procedures Being Performed Specific to this Event: NONE
Technical Specification LCOs in Effect Specific to this Event: NONE
Major Equipment Out of Service Specific to this Event: NONE
EVENT SEQUENCE
At 2345 hours on March 24, 1992, a RCS chemistry sample was obtained. In accordance
with Chemistry Procedure CE-003-327, Operation of the Primary Sample Panel (EIIS
Identifier AB-PL), the shift chemist opened Reactor Coolant Hot Leg Sample Valve (EIIS
Identifier AB-SMV) RC-104 to purge the sample line and collect a sample.
RC-104 is a 3/4 inch globe valve with a pneumatic operator. The valve is located in the
Reactor Containment Building (RCB) (EIIS Identifier NH), in piping connected to the loop
1 hot leg of the RCS. Only a flow-restricting orifice separates RC-104 from the hot leg.
RC-104 is operated remotely from a panel in the Reactor Auxiliary Building (RAB) (EIIS
Identifier NF), and functions to isolate the RCS from the Primary Sample System (EIIS
Identifier AB). Two containment isolation valves (EIIS Identifier JM-ISV), Reactor
Coolant Sample Header Inside Containment Isolation Valve PSL-105 and Reactor Coolant
Sample Header Outside Containment Isolation Valve PSL-107, are located in the piping
downstream of RC-104.
At 2355 hours, the shift chemist attempted to close RC-104 using the remote switch, but
the valve's closed indicator on the remote panel did not light. The switch was moved to the
open position and then back to the close position and a closed indication was received.
At 0014 hours on March 25, 1992, RCB atmosphere particulate activity was noted to be
rising and Operating Procedure OP-901-020, High Activity in Containment, was entered.
The chart recorder indicating the leakage flow rate (EIIS Identifier NH-FR) to the RCB
sump (EIIS Identifier BP-COL) was noted to indicate a 7 gpm leak rate. TS LCO 3.4.5.2
was entered due to unidentified leakage in excess of the 1 gpm limit. Personnel
subsequently entered the RCB to search for the source of the leak.
B-12
Operating Procedure OP-903-024, Reactor Coolant System Water Inventory Balance, was
completed at 0140 hours, and determined that RCS leak rate was 13.16 gpm, 13.135 gpm
of which was due to unidentified leakage. The leakage flow rate to the RCB sump
fluctuated between 12 gpm and 15 gpm for the next 4 hours.
At 0220, personnel in the RCB noted a large amount of steam in the area of the Reactor
Coolant Pump (RCP) (EIIS Identifier AB-P) 1B controlled bleedoff valves (EIIS Identifier
AB-FCV). Although RC-104 is located in that area, it was not possible to pinpoint the
source of the leak at that time due to radiation levels and steam. Operators commenced a
plant shutdown in accordance with TS 3.4.5.2 at 0247. An Unusual Event was declared at
0248 hours. The reactor was manually tripped, in accordance with OP-010-001, at 0545
hours.
At approximately 0500 hours, the shift chemist opened RC-104 to obtain a RCS chemistry
sample. Since frequent samples would be required following plant shutdown, RC-104 was
left open to maintain flow through the sample panel. The leakage flow rate to the RCB
sump peaked at 16 gpm at 0545 and then dropped sharply, reaching 5 gpm at 0630 hours,
before leveling off at 1.5 gpm. At the time of the event, there was no apparent correlation
between the drop in leakage flow rate to the RCB sump and the operations performed by
the shift chemist.
Personnel entered the RCB at 0930 hours, but still could not get close enough to the leak to
identify the source due to steam. At 1025 hours, in an attempt to minimize the RCS leak
rate, operators began depressurizing the RCS to 1300 psia. Cooldown of the RCS to 520
degrees commenced at 1040 hours to maintain adequate subcooling margin.
At 1130 hours, personnel again entered the RCB. The rate of leakage had slowed
considerably since the previous entry and at 1216 hours, the leakage was determined to
originate from the packing area of RC-104; the valve was shut from the remote panel and
the leakage flow rate to the RCB sump remained at 1.5 gpm. RC-104 was subsequently
shut via the local handwheel operator and leakage from the valve slowed to a few drops per
minute.
Inspection of RC-104 revealed that the packing gland studs were sheared off, the packing
was blown out, and the stem was bent against the outside wall of the gland area. The
packing gland follower was jammed against the stem above the bonnet and could not be
inserted into the packing gland area. Replacement of the packing gland studs for RC-104
was attempted, but there was insufficient room to allow drilling out the remaining portion
of the sheared studs.
To avert the thermal transient associated with a plant mode change to Cold Shutdown, and
allow for an expeditious return to power operations, a temporary leak repair was
completed until the valve can be repaired or replaced during the next refueling outage.
The affects of the leak repair were analyzed and determined to be within acceptable
parameters.
To facilitate the leak repair, the jammed packing gland follower was cut from the stem. A
strongback and clamp assembly was then installed to hold the packing and packing
B-13
retainer plate in place while an approved leak repair sealant was injected into the packing
area through an existing threaded connection. RC-104 was stroked open and sealant was
again injected until the leakage stopped. RC-104 will remain open continuously to allow
RCS chemistry sampling.
EVENT CHRONOLOGY
March 24, 1992
2355 Shift chemist attempts to close RC-104 after sampling but does not receive the closed
indication; switch for RC-104 is taken to open position and then back to close; valve
indicates closed.
March 25, 1992
0014 Noted RCB atmosphere particulate activity monitor (EIIS Identifier IL-45) reading
3.04E-14 micro-curies-per-milliliter; entered OP-901-020, High Activity in
Containment.
0018 Started Charging Pump (EIIS Identifier CB-P) A due to increasing RCS leakage.
0019 RCB sump flow chart recorder in the Control Room (EIIS Identifier NA) indicates 7
gpm and rising; entered OP-901-023, RCS Leakage.
0020 Secured Charging Pump B due to increased seal leakage. Pressurizer level is steady.
(This leakage was not a significant contributor in the event and the pump was
repaired and returned to service.)
0028 Entered TS LCO 3.4.5.2 due to unidentified RCS leakage greater than 1 gpm.
0121 Personnel entered RCB to look for leakage source.
0140 Performed OP-903-024, RCS Water Inventory Balance; unidentified leakage - 13.135
gpm, identified leakage - 0.025 gpm. RCB sump flow chart recorder indicates 12
gpm.
0220 Leakage noted to be in area of RCP 1B controlled bleedoff valves; unable to pinpoint
exact location due to radiation levels and steam.
0241 Performed OP-903-024; unidentified leakage - 12.874 gpm, identified leakage - 0.026
gpm. RCB sump flow chart recorder indicates 12 gpm.
0247 Commenced plant shutdown.
0248 Declared Unusual Event. Source of leakage is believed to be in the area of the RCP
1B seal package.
B-14
0500 Shift chemist opens RC-104 to obtain sample; valve is left open due to increased
frequency of sampling following plant shutdown.
0546 Manually tripped reactor following manual turbine trip per OP-010- 001. RCB sump
flow chart recorder indicates a peak at 16 gpm.
0600 RCB sump flow chart recorder indicates 12 gpm and dropping. RCS pressure
approximately 2260 psia.
0630 RCB sump flow chart recorder indicates 5 gpm and dropping. RCS pressure
0830 RCB sump flow chart recorder indicates 1.5 gpm and steady. RCS pressure
0930 Personnel enter RCB but still are unable to get close enough to the leak to identify
the source due to steam.
1026 Commenced depressurizing RCS to 1300 psia in an attempt to further reduce RCS
leakage.
1040 Commenced RCS cooldown to 620 degrees to maintain adequate subcooling margin.
1130 Personnel enter RCB with list of valves that Plant Engineering identified as possible
sources of RCS leakage.
1216 RCS leakage determined to be from RC-104 packing; RC-104 closed remotely by shift
chemist to stop leakage. RCB sump flow chart recorder indicates a small spike and
steadies at 1.5 gpm.
1224 Secured from Unusual Event due to identification of leakage source.
1345 RC-104 closed locally using handwheel. Leakage from the valve slowed to a few
drops per minute.
March 26, 1992
0154 Performed OP-903-024; unidentified leakage = 0.24 gpm, identified leakage = .03
gpm. Exited TS 3.4.5.2 for RCS leakage. Repair of RC-104 in progress.
CAUSAL FACTORS
Design Engineering conducted a metallurgical evaluation of a stud fragment from RC-104

which indicated that the stud failed by boric acid corrosion. It is believed the corrosion
caused the stud to waste away until the cross sectional area was insufficient to take the
applied loads. The surface of the fractured end indicates failure of the reduced section of
the stud by tensile overload. The studs were exposed to boric acid from packing leakage
and apparently failed during normal operation of RC-104. A magnet test determined that
B-16
the studs were either carbon or low alloy steel, both of which are susceptible to boric acid
corrosion
The root cause of this event was use of a material in an application for which it proved
inadequate; that is, the packing gland studs on RC-104 were made of a material that is
susceptible to boric acid corrosion. Although RC-104 met all approved design requirements,
the effect of possible packing leakage on the packing gland stud material was not
considered when the valve was designated for use in the RCS.
RC-104 was last worked in October 1988, when a body to bonnet leak was identified on the
valve. A field inspection of the valve noted that although minor corrosion was apparent on
the carbon steel bonnet studs, their integrity was maintained. As a result, Spare Parts
Equivalency Evaluation Report (SPEER) 88-452 was initiated and concluded that stainless
steel studs and nuts were an acceptable substitute for the existing bonnet studs and nuts,
and would resist the affects of boric acid. The bonnet studs on RC-104 were replaced with
stainless steel parts as part of the body to bonnet leak repair completed in October 1988;
the packing gland studs were also evaluated under SPEER 88-452 but were not replaced.
While completing the body to bonnet leak repair, it was also noted that the stem on RC-104
was bent. Work instructions were added and the valve stem was replaced. RC-104 was
subsequently stroked to verify proper operation.
The current holder of the design for RC-104, Anchor/Darling Valve Company, indicated
that the hand wheel accessory on RC-104 provided a means for tightening the valve which
could result in bending of the stem. A force of 100 pounds on the rim of the handwheel
would exert 22,600 pounds force on the valve stem. it is believed that shutting the valve
via the handwheel led to bending of the stem noted in October 1988, and during this event.
Together with representatives from Chesterton Packing Company and Anchor/Darling

Valve Company, Plant Engineering examined RC-104, its packing, and frequency of
cycling. It was determined that increased packing wear due to frequent cycling and the
horizontal orientation of the valve may have contributed to the amount of boric acid to
which the packing gland studs were exposed.
CORRECTIVE MEASURES
A temporary leak repair was completed on RC-104 in accordance with the station's Work
Authorization Preparation and Implementation Program. The work included installation
of a strongback and clamp assembly to hold the packing and packing retainer plate in place
while an approved leak repair sealant was injected into the packing area through an
existing threaded connection.
To facilitate RCS chemistry sampling, RC-104 was successfully stroked open after the
initial injection of leak repair sealant. Sealant was again injected until the leakage
stopped. Chemistry Procedure, CE-003-327, Operation of the Primary Sample Panel, was
revised to allow RC-104 to remain open continuously. The two containment isolation
valves downstream of RC-104 provide isolation of the sampling line and an additional
B-16
valve exists at the Primary Chemistry Sample Sink (EIIS Identifier AB-SNK) for sample
control.
Following the leak repair, two Temporary Alteration Requests (TARs) were implemented
in accordance with the station's Temporary Alteration Control program. TAR 92-006
installed a video camera to allow remote monitoring of RC-104 to detect any degradation of
the leak repair. TAR 92-007 installed a cylindrical leakage containment device beneath
RC-104; included is a stainless steel funnel and tygon tubing that will carry any leakage to
the floor drain so that it may be quantified.
RC-104 will be permanently repaired or replaced during the next refueling outage. Only
stainless steel studs will be used on the bonnet and packing gland to resist the effects of
boric acid: Estimated completion date for repair of RC-104: November 30, 1992
While RC-104 was being repaired, five other valves were inspected. Two of the valves,
Pressurizer Surge Line Sample Valve RC-316 and Pressurizer Steam Space Sample Valve
RC-319, have a function similar to RC-104. The other three valves, Reactor Coolant
Sample Header Inside Containment Isolation Valve PSL-105, Pressurizer Sample Header
Inside Containment Isolation Valve PSL-203 and Pressurizer Steam Space Inside
Containment Isolation Valve PSL-303, are containment isolation valves for RC-104, RC-
316 and RC-319, respectively. All five of the valves have the same manufacturer as RC-104
and are similar to RC-104. All of the valves inspected appeared to have carbon steel studs
and showed evidence of minor rust buildup, but were intact.
A plan will be developed to evaluate other valves whose studs may be susceptible to boric
acid corrosion and can possibly be exposed to boric acid leakage.
Evaluation of these valves will be prioritized according to their safety significance, their
ability to be isolated and their similarity to RC-104 in construction and application. The
plan will allow for replacement of the stud material on a case by case basis dependent on
the evaluation results and valve availability: Estimated completion date for the evaluation:
January 31, 1993.
SAFETY SIGNIFICANCE
Plant operators performed appropriate actions to shutdown the reactor, stabilize RCS
parameters, minimize RCS leakage and ensure individuals and organizations were notified
of the Unusual Event in a timely manner. As a result, leakage was confined to the RCB,
and the health and safety of the public and plant personnel was not compromised by this
event.
SIMILAR EVENTS
LER 85-013 reported unidentified RCS leakage of 1.7 gpm which was later attributed to
failure of Letdown Volume Control Tank Relief Valve (EIIS Identifier AB-V) CVC-115 to
properly seat after it inadvertently lifted a day before. CVC-115 was subsequently
repaired. This event is similar only by the fact that it also required entry into TS 3.4.5.2
and declaration of an Unusual Event.
B-17
LER 85-018 reported unidentified RCS leakage of 6.1 gpm which was later determined to
be from the packing glands of Pressurizer Spray Valves (EIIS Identifier AB-FCV) RC-301A
and RC-301B. TS 3.4.5.2 was entered and the reactor was shutdown. Station Modification
926 was issued to improve the reliability of the packing design on these valves by utilizing
a different type of packing and adding additional rings of packing. There is a similarity in
the leakage source but the root cause of the leakage is different. There were no generic
concerns identified in LER 85-018.
LER 88-015 reported unidentified RCS leakage of 4 gpm which was attributed to leakage
past two in-series Reactor Coolant Loop Drain Valves (EIIS Identifier AB-DRN) RC-206A
and RC-207A. The valves had been hand closed and were not fully shut. After shutting
both valves with a valve wrench, the leakage dropped to within TS limits. This event is
similar only by the fact that it also required entry into TS 3.4.5.2 and declaration of an
Unusual Event.
LER 91-008 reported unidentified RCS leakage of approximately 20 gpm while in Hot
Shutdown. TS 3.4.5.2 was entered and an Unusual Event was declared. Leakage was
determined to be from a bonnet bolt hole on Pressurizer Spray Check Valve (EIIS Identifier
AB-V) RC-303 and was caused by improper assembly of the gasket retaining ring due to
underdeveloped maintenance training. There is no similarity in causal factors.
EVENT DATE: 08/09/88 LER it: 88-021-00 REPORT DATE: 09/14/88
FACILITY NAME: SAN ONOFRE NUCLEAR GENERATING STATION, UNIT 2
TITLE: CONTAINMENT ISOLATION VALVE TECHNICAL SPECIFICATION NOT MET

DUE TO FAILURE OF RCS HOT LEG SAMPLE VALVE 2HV-0508
OTHER FACILITIES INVOLVED: DOCKET NO:

SECTION: 50.36(c)(2), 50.73(a)(2)(i)

TELEPHONE: 714-368-6241

CAUSE: X SYSTEM: JM COMPONENT: ISV MANUFACTURER: L200
REPORTABLE NPRDS:
SUPPLEMENTAL REPORT EXPECTED:
B-18
ABSTRACT:
On 8/11/88, at 0706, containment isolation valve 2HV-0509 (Reactor Coolant System (RCS)
hot leg sample valve) was opened for RCS sampling and unanticipated RCS leakage was
observed. At 1245, it was determined that the redundant isolation valve 2HV-0508 may
not have been fully closed and was thus causing the observed leakage. In accordance with
TS 3.6.3, 2HV-0509 was secured closed. Inspection of 2HV-0508 on 8/15 revealed
excessively worn threads on the Limitorque actuator stem nut. This condition had allowed
the rotation of the stem nut without attendant valve stroke, causing a false "close" position
indication in the Control Room. It is postulated that the valve failure occurred when a
delayed close indication was received on 8/8/88. Since the condition of the valve had been
unknown, 2HV-0509 was not maintained secured in the closed position from 8/9 at 1315 to
8/11 at 1245, contrary to TS 3.6.3.
Valve 2HV-0508 was not equipped with a protective stem cover for its actuator. Moisture,
boric acid and debris had come in contact with the actuator stem which resulted in
abrasive thread surfaces, increasing the rate of wear of the softer stem nut threads. This
condition, along with frequent cycling of the valve, caused the actuator stem nut threads to
wear excessively. Valve 2HV-0508 was returned to service following repair and the
installation of a protective stem cover.
Various Limitorque actuators were previously identified as lacking protective stem covers.
The vendor manual neither provides the part identification nor direction for the
undertaken. In May 1988, acceptable replacements for the missing covers were identified
and obtained. At that time, Maintenance procedures were revised to require installation of
such covers following maintenance activities. All safety related Limitorque actuators
identified as lacking a cover have now been corrected.
(END OF ABSTRACT)
Plant: San Onofre Nuclear Generating Station Unit: Two

Reactor Vendor: Combustion Engineering
Event Date: 08-09-88
A. CONDITIONS AT TIME OF THE EVENT: Mode: 1, Full Power
B. BACKGROUND INFORMATION:
Technical Specification (TS) required Reactor Coolant System (RCS) (EIIS System Code
AB) chemistry samples are obtained by opening a 3/4 inch sample line from either one of
the two RCS hot legs. There are three normally closed containment isolation valves (EIIS
Component Code ISV) associated with the RCS sampling system which receive a close
signal upon a Containment Isolation Actuation Signal (CIAS) (EIIS System Code JM).
2HV-0508 and 2HV-0517 are both Limitorque Motor Operated Valves (MOVs) which are
located inside containment on separate lines which lead from the No. 1 and No. 2 hot legs,
B-19
respectively. Downstream of these valves, the two sample lines join together into a single
line prior to penetrating containment. 2HV-0509 is an air operated valve located outside
containment on the RCS sample line.
TS 3.6.3, Containment Isolation Valves, requires that with one containment isolation valve
inoperable, the penetration be isolated within 4 hours by securing the redundant valve(s)
in the closed position, or be in hot standby within the next 6 hours and in cold shutdown
within the following 30 hours.
C. DESCRIPTION OF THE EVENT:

1. Event:
At 0930 on 8/8/88, following a RCS chemistry sample, closure of 2HV-0508 was observed to
have taken approximately five minutes (as determined by illumination of the closed
position indicating light in the Control Room). The observed stroke time was considered
abnormally long. The valve was declared inoperable and in accordance with TS 3.6.3, 2HV-
0509 was secured closed by the removal of its air supply to the valve operator. At 1315 on
8/9/88, 2HV-0508 was de-energized and declared operable for containment isolation since
the valve was thought to be secured in its safety related position. Following this action, the
air supply to 2HV-0509 was returned to service.
At 0705 on 8/11/88, with 2HV-0508 and 2HV-0517 indicating closed, RCS leakage was
observed when 2HV-0509 was opened to facilitate RCS sampling via 2HV-0517. At 1245, it
was determined that 2HV-0508 may not have been fully closed and therefore, the valve
was considered inoperable. In accordance with TS 3.6.3, 2HV-0509 was secured closed by
removing the air supply to the valve operator.
On 8/15/88, a containment entry was made and inspection of 2HV-0508 revealed that the
valve was not fully closed. Since the valve had operated properly until it was last cycled on
8/8/88, it is postulated that the valve failure occurred at that time. Therefore, from 1315
on 8/9/88 until 1245 on 8/11/88, TS 3.6.3 was not met since during this period the damage
to the valve was unknown, and consequently, 2HV-0509 was not maintained secured in the
closed position.
2. Inoperable Structures, Systems or Components that Contributed to the Event:
None.
3. Sequence of Events:
DATE TIME ACTION

8/8 0930 2HV-0508 declared inoperable. 2HV -05 09 secured in
closed position by removal of air supply to the valve
operator.
8/9 1315 Air supply to 2HV-0509 returned to service.
B-20
8/11 0705 2HV-0509 opened to sample RCS. RCS leakage

- observed.
8/11 1245 Investigation determined that 2HV-0508 may be

inoperable, 2HV-0509 secured in closed position.
4. Method of Discovery:
On 8/15/88, following a containment entry which revealed that 2HV-0508 was not fully
closed and identified that the stem nut in the Limitorque valve actuator had stripped
threads, it was determined that 2HV-0508 had been inoperable since 8/8/88, at which time
it was last operated.
5. Personnel Actions and Analysis of Actions:
Upon determining that 2HV-0508 may not be fully closed, operators promptly secured
2HV-0509 in the closed position in accordance with TS 3.6.3.
6. Safety System Responses:
Not applicable.
D. CAUSE OF THE EVENT:

1. Immediate Cause:
The actuator stem nut threads in the Limitorque actuator of 2HV-0508 had worn enough to
lose engagement with the actuator stem threads. The actuator stem is driven by rotation
of the actuator stem nut to provide movement of the valve components (i.e., valve stem and
disc), thus opening and closing the valve. With excessively worn threads in the stem nut,
rotation of the stem nut continued without the accompanying axial movement of the
actuator stem necessary for valve stroking. Likewise, the actuator rotors, which contain
the limit switch contacts for providing remote position indication in the Control Room,
continued to rotate and thus resulted in the false "close" position indication.
2. Root Cause:
Valve 2HV-0508 was not equipped with a protective stem cover which provides protection
for the actuator stem and other internals from the surrounding environment. As a result,
moisture, debris and boric acid (from a leak which no longer exists) were able to come in
contact with the actuator stem. The combination of these conditions resulted in abrasive
thread surfaces, increasing the rate of wear of the softer stem nut threads. This condition,
along with frequent cycling of the valve, caused the actuator stem nut threads to wear
excessively.
During the last Unit 2 refueling outage, various Limitorque actuators, including the
actuator for 2HV-0508, were identified as lacking a protective stem cover, and an
engineering review was initiated in December 1987 to resolve this concern. The vendor
B-21
manual neither provides the part identification nor directions necessary for the
undertaken. In May 1988, acceptable replacements for the missing protective stem covers
were identified and obtained. At that time, Maintenance procedures were revised to
require installation of such covers following maintenance activities. Subsequently, except
for 2HV-0508, all safety related Limitorque actuators identified as lacking a protective
stem cover had been corrected. The next maintenance activity for 2HV-0508, including the
installation of the cover, had been scheduled to be performed during the next refueling
outage.
E. CORRECTIVE ACTIONS:
1. Corrective Actions Taken:
a. Valve 2HV-0508 was returned to service following repair and the installation of a
protective stem cover.
b. As described in SCE's response to Generic Letter 88-05, "Boric Acid Corrosion of

Carbon Steel Reactor Pressure Boundary Components in PWR Plants", programs are now
in place for detecting and evaluating the impact of boric acid leaks inside containment. It
appears that the boric acid intrusion observed in this case resulted from leakage which has
since been eliminated and probably occurred prior to implementation of the existing
controls.
2. Planned Corrective Actions:
No further corrective action is required since the presence of protective stem covers is
verified during maintenance of the Limitorque actuators in accordance with maintenance
procedures.
F. SAFETY SIGNIFICANCE OF THE EVENT:
There was no safety significance to this event since during the period in which 2HV-0508
was not fully closed, 2HV-0509 remained in its closed position, with the exception of two
brief occasions when 2HV-0509 was opened for RCS sampling. Furthermore, 2HV-0509
remained fully operable and capable of closing upon a CIAS signal.
G. ADDITIONAL INFORMATION:
1. Component Failure Information:
The MOV for 2HV-0508 is a Limitorque (Manufacturer Code L200) actuator - Model No.
SMB-000.
2. Previous LERs on Similar Events:
None.
B-22
3. Results of NPRDS Search:
A review of the NPRDS industry data for Limitorque actuator (Model No. SMB-000)
revealed several instances in which stem nut threads were excessively worn. The majority
of these failures were caused by the lack of lubrication and/or the collection of dirt around
the stem.
TITLE: Shutdown Cooling Isolation Valve Packing Gland Leakage

SECTION 50.73(a)(2)(ii)

TELEPHONE # 714-368-6241

CAUSE: B SYSTEM: BP COMPONENT: ISV MANUFACTURER: W255
REPORTABLE TO NPRDS: Yes
CAUSE: B SYSTEM: BP COMPONENT: MO MANUFACTURER: L200
REPORTABLE TO NPRDS: Yes
ABSTRACT:
On August 13, 1987, at approximately 1900, with Unit 2 in Mode 5 and the Reactor
Coolant System (RCS) at approximately 350 psia and 127 degrees Fahrenheit, failure of
alloy steel packing gland follower studs during manual operation of motor operated
Shutdown Cooling System (SDCS) suction isolation valve 2HV-9378 resulted in leakage
estimated atl00 gpm through the packing gland. Operation of the SDCS continued via a
redundant flow path. RCS inventory was maintained by isolating letdown flow and using
charging pumps as depressurization and venting of the RCS proceeded. Containment
closure was promptly restored and there was no effluent release from containment above
regulatory limits.
At 1100 on September 1, a temporary repair was completed which reduced the leak rate to
approximately 1/4 gpm, effectively. terminating the event.
B-23
The cause of stud failure has been attributed to (1) packing leakage resulting in wastage
due to boric acid corrosion, (2) decrease in lubricating characteristics and hardening of
packing, and (3) the initial thrust required to open the valve. Corrective actions include
reduction of the maximum thrust necessary to open the valve by installation of a modified
packing gland assembly less susceptible to leakage and hardening of packing, and
replacement of packing gland studs with corrosion resistent material.
The health and safety of plant personnel and the public were not affected by this event.
(END OF ABSTRACT)
TITLE: RCS Pressure Boundary Leak via Shutdown Cooling System (SDCS) Isolation
Valve Packing Gland
Plant: San Onofre Nuclear Generating Station (SONGS) Unit: 2

Reactor Vendor: Combustion Engineering
Event Date: 08-31-87
Time: 1845
A. PLANT CONDITIONS AT THE TIME OF THE EVENT:
Mode: (5) Cold Shutdown, SDCS in operation

RCS Temperature: 127 degrees Fahrenheit
RCS Pressure: 350 psia
B. BACKGROUND INFORMATION
SYSTEM OPERATION
The Reactor Coolant System (RCS) (EIIS System Code AB) is connected to the suction of
the Shutdown Cooling System (SDCS) (EIIS System Code BP) pumps via parallel flow
paths which contain redundant, motor-operated isolation valves (EIIS Component Code
ISV). After the RCS is cooled down to approximately 285 degrees Fahrenheit using the
steam generators (EIIS Component Code SG), the SDCS is placed into operation by
opening the motor-operated isolation valves. Both of the parallel flow paths are opened in
order to provide redundant access for the RCS to the Low Temperature Overpressure
Protection (LTOP) relief valve (EIIS Component Code RV) and to provide redundant
shutdown cooling flow paths.
TECHNICAL SPECIFICATION REQUIREMENTS
Limiting Condition for Operation (LCO) 3.4.8.3.1 requires that both flow paths from the
RCS to the SDCS be available when the RCS temperature is at 235 degrees Fahrenheit or
less with the reactor vessel head on, in order for the Low Temperature Overpressure
Protection (LTOP) relief valve, located between the isolation valves and the SDCS pumps,
to be considered OPERABLE.
B-24
The inability to align both flow paths places the Unit in Action Statement (b) for this LCO.
This Action Statement requires the valve to be opened within 7 days or (1) the RCS must
be reduced to less than 200 degrees Fahrenheit, (2) depressurized and (3) vented through a
greater than, or equal to, 5.6 square inch vent within the next 8 hours.
WORK AUTHORIZATIONS
Administrative control of activity performed on safety related equipment is defined in

Station Procedure. In general, activity involving maintenance is pre-planned and
documented in the San Onofre Maintenance Management System (SOMMS) and
eventually authorized by a Work Authorization Record (WAR) issued by the Equipment
Control section of the Operations Division. The WAR (1) identifies the work to be
accomplished, (2) documents the impact the specific activity will have on the operability of
the equipment involved, (3) identifies the plant conditions that must be established to
perform the activity safely and (4) specifies the functional testing requirements that must
be completed before the affected equipment is returned to service. The process includes
independent review and concurrence by all disciplines involved.
In those instances where the Operations Shift Superintendent determines that an activity
must be accomplished in an expedited manner in order to avoid an unsafe condition, or to
comply with time limits imposed by Technical Specification Action Statements, the Shift
Superintendent's Accelerated Maintenance (SSAM) process is implemented. The process
requires that the same interdisciplinary review and concurrence be obtained with
documentation of the activity taking place concurrent with the activity.
NONCONFORMANCE REPORT (NCR) SYSTEM
Station Procedures require that non-conforming conditions be identified in a NCR and that
once initiated, further activity to operate the affected equipment be prohibited until the
non-conforming condition is evaluated and resolved by appropriate personnel.
C. DESCRIPTION OF THE EVENT:
1. Event:
At 0025 on 8/29, 1987, Unit 2 was taken off line to commence a scheduled refueling outage.
At 1235 on 8/30, 1987, the redundant isolation valves in one of the two parallel flow paths
were opened from the Control Room. However, one of the isolation valves in the other
parallel flow path (2HV-9378) could not be opened remotely. The Shift Superintendent
initiated and authorized a Shift Superintendent Accelerated Maintenance (SSAM) Order to
investigate and, if possible, correct the conditions which prevented the opening of 2HV-
9378. The Maintenance Order (MO) generated for the SSAM required initiation of a
Nonconformance Report (NCR) for any identified nonconforming condition.
At 1655, the SDCS was placed in service using the single flow path and cooldown was
continued. At 1730 on 8/30, RCS temperature decreased below 235 degrees Fahrenheit and
Technical Specification Limiting Condition for Operation (LCO) 3.4.8.3.1 was entered.
B-25
Action Statement (b) of this LCO requires that the second flow path be in service within 7
days or the RCS must be cooled below 200 degrees Fahrenheit, depressurized and vented.
Mode 5 was entered at 1940.
At 0245 on 8/31, it was determined that the motor operator had overheated and failed.
Manual operation of the handwheel moved the valve to an indicated 5% open position
where it could not be further opened. Subsequent inspection by the cognizant engineer
overseeing the investigation to determine the cause of the failure of the valve to open,
identified that extensive boric acid crystals and corrosion were present in the area of the
packing gland and valve stem.
Further efforts to operate the handwheel confirmed that the valve could not be opened
beyond 5%. The inspection findings were provided to Control Room personnel and, as
provided by the SSAM, documented in an NCR.
The boric acid crystals/rust was removed from the valve and, with approval from the
Control Room, the valve was further opened using the manual handwheel to about the 10%
open position where the handwheel again became difficult to turn. During the course of
this evolution and after the handwheel had been rotated approximately 10 to 12 turns, the
maintenance crew heard what was believed to be the sound of a stuck valve disc being
freed from a jammed position in its seat.
The maintenance worker who was operating the manual handwheel to open the valve
indicated that, because of limited valve accessibility and the awkward position necessary to
operate the valve, he decided to extend his reach by use of a pipe wrench used as an
extension on the hand wheel The wrench was engaged and re-engaged to the rim of the
handwheel in quarter turn increments.
At approximately 1845 on 8/31, after rotating the handwheel approximately 3 additional

turns using the extension, the packing gland follower studs failed. Water immediately
began leaking from the valve packing at a rate estimated to be 100 gpm.
The Control Room was immediately informed of the leak. Following initial unsuccessful
attempts to close the valve, the area was evacuated. Control Room operators initiated
action to evacuate all personnel from containment and to establish containment closure by
closing all containment penetrations providing direct and unmonitored access from the
containment atmosphere to the outside atmosphere. Pressurizer level was maintained by
operating the charging pumps and isolating letdown flow. The operating reactor coolant
pump was stopped and action was initiated to cooldown the pressurizer and depressurize
the RCS to reduce the leakage and permit a temporary repair.
By 2020, containment closure had been restored and the purge and associated radiation
monitors remained in operation. RCS leakage from 2HV-9378 was being collected in the
containment sump and directed to the Liquid Radwaste System for processing. The
pressurizer was taken solid at 2140 on 8/31.
B-26
During the period from 2000 on 8/31/87 until 0800 on 9/1/87, RCS depressurization
continued and the valve leak rate slowly decreased. The SDCS remained in operation and
effectively maintained RCS temperature below 127 degrees Fahrenheit.
At 0800 on 9/1/87, with the RCS depressurized to approximately 24 psia, leakage from the
valve had decreased to approximately 20 gpm. An additional attempt to stop the leakage
was initiated and by 1100 was completed. This temporary repair was successful in
reducing the leakage to approximately 1/4 gpm, effectively terminating the event.
Although individual exposure to radiation remained well within regulatory limits, the
event resulted in the contamination of several workers with reactor coolant. The Health
Physics aspects of the event are being evaluated and the results of this evaluation will be
reported in a separate informational LER.
2. Inoperable Structures, Systems or Components that Contributed to the Event:
None.
3. Sequence of Events:
DATE TIME ACTION
8/29 0025 Unit 2 taken off line to commence scheduled refueling

outage. Main Breakers open.
8/30 1655 RCS placed on Shutdown Cooling.
8/30 1730 RCS cooled below 235 Fahrenheit TS LCO 3.4.8.3.1

Action Statement (b) entered.
8/31 1845 RCS leak of 100 gpm developed while attempting to

open 2HV-9378. RCS at 127 Fahrenheit and 350 psia.
8/31 1915 Packing gland follower studs found broken.
8/31 2020 Containment closure established. Purge and associated

radiation monitors remained in operation.
8/31 2140 Pressurizer taken solid.
9/1 0800 Began temporary repair to valve packing gland. Leak

rate is approx. 20 gpm with RCS at 24 psia.
9/1 1100 Temporary repair completed. RCS leakage reduced to

approx. 1/4 gpm.
9/4 2300 Pressurizer manway removed. TS LCO 3.4.8.3.1, Action

Statement (b) exited.
B-27
4. Method of Discovery:
Visual observation of the packing leakage by the maintenance crew during manual
operation of the valve.
5. Personnel Actions and Analysis of Actions:
Operator Actions:
Proper actions were taken to evacuate all personnel from containment, restore
containment closure by closing the equipment hatch and the personnel hatch, maintain
pressurizer level using charging pumps, and initiate RCS depressurization to reduce the
leakage and permit a temporary repair.
Maintenance Actions:
Proper maintenance actions were taken in attempting to contain the valve packing leakage
and in effecting the temporary repair which terminated the event.
Health Physics Actions:
No personnel were exposed to radiation above regulatory limits. There was no effluent
release from containment above regulatory limits. The radiological controls applied and
Health Physics actions taken during this event will be the focus of a separate informational
licensee event report.
6. Safety System Responses:
No safety system actuations occurred during this event.
D. CAUSES OF THE EVENT:
1. Immediate Cause:
Failure of the packing gland follower studs while attempting to open 2HV-9378 manually.
Evaluation of the corroded studs (diameter was less than 10% of its designed value) led to
the conclusion that the strain produced by manual operation of the valve, along with forces
resulting from the packing gland follower plate drag and boric acid wastage of the studs,
resulted in the failure of the studs. The "popping" sounds heard by the maintenance crew
are, in fact, believed to be related to the stud failures.
Based on analysis considering the as found conditions of the studs, operation of the
handwheel without the use of an extension would still have resulted in stud failure.
However, use of an extension in this case was inappropriate and could have led to stud
failure with even less severe stud corrosion.
B-28
2. Intermediate Causes:
Inspection of the packing gland follower studs demonstrated that they had been subject to
severe wastage due to boric acid corrosion. Contrary to Section 5.4.12.2 of the Updated
Final Safety Analysis Report (UFSAR), the studs were alloy steel. Packing leakage from
the RCS had reduced the diameter of the studs. The possibility that such wastage existed
and could result in packing gland failure due to further efforts to open the valve, was not
recognized during the initial inspections.
3. Root Causes:
a. Packing Gland Stud Failure:
Use of carbon and alloy steel bolting in this application was consistent with the vendor
drawing, but it was contrary to the UFSAR description referenced above. Also, these
valves do not have diaphragms or leak-off connections, as described in the UFSAR. The
reasons for these deviations are under continuing investigation. The consequence of this
improper material was corrosion of the packing gland studs which resulted in the
immediate cause of the failure as mentioned above.
b. Motor Operator Failure:
Subsequent to core offload and RCS drain down to the bottom of the hot legs, 2HV-9378
was disassembled and inspected. The following factors, in combination, led to the failure of
the valve to open using the motor-operator:
(i) A high required initial valve opening thrust.
Based on an analysis of the MOVAT test results obtained in May, 1986, the initial thrust
required to move the gate for these WKM double wedge gate valves is high in comparison
with other gate valves. It was determined that the closing torque setting for these valves
was conservatively set to close the valve against a measured running thrust and a stem
blow out force calculated at 2485 psia. Thus, when the valve is closed, the wedges are
forced into the seat with greater force than necessary to ensure proper sealing. Since the
maximum output of the motor for the 10 inch valves (2HV-9377 and 2HV-9378) is limited
by the size of the inverter which supplies motive power, no margin was available in the
opening force to overcome substantial increases in packing drag. The 16 inch valves (2HV-
9337 and 2HV-9339), which are powered from a 480 volt motor control center (MCC), are
capable of developing the force necessary to overcome substantial increases in packing
drag.
The metallic seal between the body and the bonnet provided evidence of somewhat uneven
torquing of the bonnet to valve body. This would have resulted in uneven forces being
exerted on the packing gland and stem, contributing to the increase in force required to
open the valve.
B-29
(ii) Packing leakage.
The existence of solidified boric acid between the packing and the stem increased the static
friction coefficient between these parts sufficiently to prevent opening by the motor
operator. Of the 4 Unit 2 SDCS isolation valves, only 2HV-9378 and 2HV-9337
experienced previous packing leakage.
(iii) Grafoil packing material.
When the grafoil packing (which is a porous material) is subjected to boric acid, it hardens
and loses its original lubricating characteristics. As a result, the static friction coefficient
increases. The grafoil packing was last replaced in December 1984.
4. Contributing Weaknesses
a. Contrary to the Station procedure governing nonconforming materials, work continued

despite the fact that an NCR was required and written for the deficiencies identified.
Management controls were therefore, not effective in limiting efforts to open 2HV-9378 to
properly planned and defined activities. Specifically, management controls were lacking in
enforcing the procedure requirements to ensure that the NCR was dispositioned, and the
disposition activity was complete, and verified complete, prior to operating and opening the
valve. This failure bypassed the process of obtaining independent review of the
nonconforming conditions by Operations, several levels of Technical supervision, and
QA/QC, which may have resulted in postponement of further attempts to operate the valve
until the RCS was depressurized.
b. In reviewing the SSAM process and the SSAM used during this event, several
weaknesses were noted:
(i) The work plan for the SSAM contained a general statement regarding implementation
of corrective action on the valve. Although this part of the plan was not implemented and,
consequently, had no effect in creating the subsequent leak, such a work plan step is overly
broad in scope and vague concerning the limits placed on permissible efforts to correct
equipment deficiencies.
(ii) Although tailboard of the SSAM work scope was conducted between Operations and
Maintenance, and Operations approved the work activities performed as the maintenance
efforts progressed (except use of the extension), contrary to the maintenance procedure
governing the SSAM process which was revised a few days prior to the event, the
Operations Shift Superintendent was not formally asked to review and therefore did not
actually review the SSAM work plan and certify his approval of the work scope and the
tailboard by his signature in the SSAM MO. Failure of maintenance personnel involved
with this SSAM to check and verify that the latest procedure revision was used resulted in
this deficiency. This deficiency, in turn, bypassed a formal review and approval of the
work scope by the Operation Shift Superintendent and precluded the considerations of
plant conditions to eliminate work plan steps which may be incompatible with these
conditions.
B-30
(iii) Given the 7-day action statement under TS LCO 3.4.8.3.1, (b), although allowed by the
Operations procedure governing the SSAM process, the initiation of a SSAM in this event
was not warranted. Similarly, this procedure does not provide sufficient direction
regarding the length of time a SSAM may exist. The SSAM was allowed to progress well
beyond 24 hours. Essentially, this procedure deficiency created the atmosphere of an
emergency when, in fact, there was no emergency.
E. CORRECTIVE ACTIONS:
1. Corrective Actions Taken:
The leakage was reduced by implementing a temporary repair.
Subsequent to core offload and RCS drain down to the bottom of the hot legs, a design
change was implemented to all four Unit 2 SDCS suction isolation valves (2HV-9337, 2HV-
9339, 2HV-9377, 2HV-9378). This design modification replaced the alloy steel packing
gland follower studs and nuts, bonnet studs, lifting ring, bonnet nuts, and carbon steel
packing gland follow plate with high strength, boric acid corrosion resistant, stainless steel
components, and boric acid corrosion resistant nuclear grade epoxy paint was applied to
the carbon steel yokes of these valves.
In addition, based on the "Valve Stem Packing Improvement Study" performed by the
Electric Power Research Institute (EPRI) described in EPRI Report NP-4255, which was
issued in February, 1986, this design modification also replaced the grafoil packing
material with "live load" packing of a different configuration, having a lower static friction
coefficient and added springs to the packing gland follower studs between the packing
gland follow plate and nuts. The springs are designed to maintain proper preread on the
packing gland as the packing ages.
Based on this EPRI study, this packing modification has been shown to significantly reduce
packing leakage, and is less susceptible to hardening and loss of lubrication as the packing
ages.
2. Planned Corrective Actions to Prevent Recurrence:
The closing torque switch for 2HV-9377 and 2HV-9378 will be reset using a value for stem
blow out force calculated at 410 psia vice 2485 psia. This revised setting will provide
additional thrust in the opening direction for overcoming increases in packing drag. After
adjustment, the margin available will be greater than any anticipated increases in packing
drag with the new packing material and live loading configuration.
During the current refueling outage for Unit 2 and the next refueling outage for Unit 3, all
valves which are unisolable from the RCS will have their carbon/alloy steel components
inspected and if necessary, these components will either be replaced with stainless steel or
have a boric acid corrosion resistant epoxy applied to ensure they will not suffer any
significant boric acid wastage over a fuel cycle.
B-31
During the current refueling outage for Unit 2 and the next refueling outage for Unit 3, all
WKM gate valves which isolate the RCS will have their packing modified in a manner
similar to that implemented on the four Unit 2 SDCS suction isolation valves. These
modifications, and other requirements concerning valve packing, will be reflected in the
next Final Safety Analysis Report update. The remaining valves, which are unisolable
from the RCS, were inspected; the gland follower studs and bonnet-to-body studs were
found to be made of stainless steel not susceptible to boric acid corrosion.
The ability of the remaining SDCS isolation valves motor operators to unction properly in
the absence of packing leakage indicates that by minimizing packing leakage and using a
packing material not susceptible to hardening, the probability of motor failure will be
greatly reduced regardless of the high required initial thrust of WKM gate valves.
A containment entry has been made for Unit 3 to determine if there is any evidence of
packing leakage from the corresponding four valves. Although the packing area of the
valves can not be inspected directly due to lagging and access limitations, there was no
evidence of packing leakage seen beneath or around the valves, except for 3HV-9337 where
there was some evidence of prior leakage which is not continuing. Valve 3HV-9337 is
normally isolated from the RCS by 3HV-9339. At the next entry to Mode 3, the missile
shielding and lagging will be removed, as required, to permit direct inspection of the
packing area. It is planned to implement the modifications made to the Unit 2 valves on
Unit 3 at the next refueling outage. Also, prior to the next opening of the valves, they will
be fully inspected and the opening evolution will be carefully controlled.
All cognizant engineers and technical supervisors; maintenance crafts personnel, planners,
and supervisors; and operations personnel including supervisors, will receive additional
training focusing on the procedure requirements to ensure that when an NCR is being or
has been initiated, it must be dispositioned, and that disposition satisfactorily
implemented, prior to using or operating the affected component or equipment.
In addition, the maintenance order program will include a step in the work plan to ensure
that when an NCR is being or has been initiated, the component or equipment involved
shall not be used or operated until the NCR has been satisfactorily resolved.
To improve the SSAM control process, both the Operations and Maintenance procedures
governing SSAM are under revision. The revision will (1) define limits on the use of and
provide termination criteria for a SSAM; (2) place emphasis on ensuring that the work
plan of a SSAM receives sufficient, formal review; (3) provide appropriate precautions in
the SSAM; and (4) provide an appropriate feedback mechanism to operations as the work
progresses for ongoing evaluation. Guidance and criteria regarding the use of extensions
on manual valve operators will be addressed to ensure all consequences of valve failure are
properly considered prior to their use.
In addition, all maintenance personnel will be reinstructed on the need to check and verify
that the latest revision of a procedure is used for all occasions. This concern has received
and will continue to receive ongoing emphasis.
-
B-32
F. SAFETY SIGNIFICANCE OF THE EVENT:
No personnel were injured or exposed to radiation above regulatory limits.
Effluent release calculations indicated that this event did not result in any releases to the
environment above regulatory limits. Evaluation of Containment Purge monitor readings
and air samples inside containment indicated that the highest value attained at the Site
Boundary was less than 0.21 MPC during the entire event.
Valves in the parallel SDCS suction flow path had been opened and were providing the
flow path for SDCS operations, as well as, LTOP protection. In addition, the charging
pumps were able to maintain RCS inventory during the entire event. It is therefore
concluded that neither the health and safety of plant personnel nor the health and safety of
the public was affected by this event.
G. ADDITIONAL INFORMATION:
1. Component Failure Information:
Valve 2HV-9378 is a 10"-by-8"-by-10", WKM, Pow-R-Seal, motor operated gate valve. The
valve operator is a Limitorque, Model SB-1.
2. Previous LERs on Similar Events:
None.
3. IE Bulletin 82-02:
"Degradation Of Threaded Fasteners In The Reactor Coolant Pressure Boundary Of PWR

Plants", dated June 2, 1982, required all PWR licensees to review procedures associated
with the identification, inspection, removal and installation of Reactor Coolant Pressure
Boundary (RCPB) threaded fasteners (studs or bolts) in (1) steam generator and
pressurizer manway closures, (2) valve bonnets, and pump flange connections installed on
lines having a nominal diameter of 6 inches or greater, (3) control rod drive flange and
pressurizer heater connections that do not have seal welds to provide leak-tight Integrity,
and (4) reactor vessel head closure studs not complying with the provisions of Regulatory
Guide 1.65, "Materials and Inspections for Reactor Vessel Closure Studs."
As a result of that Bulletin, Inservice Inspection Program procedures, as well as

maintenance procedures, were modified to ensure all such RCPB fasteners were, as a
minimum, visually inspected for boric acid induced corrosion whenever the fastener is
removed for maintenance or other reasons. Valve packing gland studs, however, are not
considered as part of the RCPB and were therefore, not included in that program. The
program established to address the Bulletin will be expanded to include inspections of
additional RCPB valve components subject to boric acid corrosion.
B-33
4. IE Notice 86-108 and Supplement 1
"Degradation of Reactor Coolant Pressure Boundary Resulting From Boric Acid Corrosion",
dated December 29, 1986 and April 20, 1987, respectively, informed Licensees of instances
where leakage from the Reactor Coolant System had resulted in boric acid corrosion of
portions of the RCPB fabricated from ferritic steel alloys; e.g., external surfaces of safety
injection nozzles, Reactor Vessel Head mounted components, etc. Although this
phenomenon may result from packing gland leakage depositing boric acid on RCPB piping
surfaces, no deposition of boric acid on ferritic steel surfaces of the RCPB was experienced
in this case. Increased inspections of such ferritic steel surfaces is under consideration in
response to this Information Notice.
TITLE: Pressurizer Heater Rupture Due to Primary Water Stress Corrosion Cracking
Results in Reactor Coolant Pressure Boundary Leakage and Pressurizer Vessel Corrosion

SECTIONS 50.73(a)(2)(i), 50.73(a)(2)(ii), Other- Generic

NAME: Larry A. Taylor, Nuclear Safety and Licensing Specialist
TELEPHONE #- 501-964-3100

CAUSE: B SYSTEM: AB COMPONENT: EHTR MANUFACTURER: W065
REPORTABLE TO NPRDS: Y
CAUSE: C SYSTEM: AB COMPONENT: AE MANUFACTURER: C490
ABSTRACT:
On 4/24/87, an Unusual Event was declared and a reactor shutdown was commenced due
to a suspected reactor coolant system (RCS) pressure boundary leak of approximately 60
drops per minute from the area of the pressurizer vessel lower head. Subsequent
investigation revealed that two pressurizer heaters manufactured by Watlow Electric
Company had ruptured resulting in damage to the heater sleeves (which penetrate the
vessel head and house the heaters). Damage to one heater sleeve was sufficient to cause
RCS pressure boundary leakage which resulted in a small area of boric acid induced
B-34
corrosion damage to the pressurizer carbon steel base metal. Metallurgical analysis of the
damaged components revealed that the heater sheaths had undergone primary water
stress corrosion cracking (PWSCC) which allowed water to reach the magnesium oxide
(MgO) insulation internal to the heater. Expansion of the MgO due to hydration resulted
in the subsequent rupture of the heater sheaths and cracking of the heater sleeve. It has
been determined that the manufacturing process of the Watlow heaters induced
susceptibility of the sheaths to PWSCC. As a result of this event, all Watlow heaters have
been removed from the pressurizer and the two ruptured heater locations have been
permanently closed.
(END OF ABSTRACT)
I. DESCRIPTION OF EVENT
A. Plant Status
At the time of discovery of the event on April 24, 1987, Arkansas Nuclear One, Unit 2
(ANO-2) was operating at 100% power with a reactor coolant system (RCS) average
temperature of 580 degrees Fahrenheit and pressurizer pressure of 2250 psia. The unit
had been in continuous operation for 167 days.
B. Component Identification
The ANO-2 pressurizer (AE) (Figure 1) is a component of the RCS (AB) which functions to
maintain system operating pressure and compensate for changes in coolant volume during
load changes. The pressurizer is a vertically mounted, cylindrical vessel made of SA-533,
Grade B. CL-1 (P-3) alloy steel and was designed and fabricated in accordance with the
ASME Boiler and Pressure Vessel Code. Minimum vessel wall thickness is 3-7/8 inches in
the area of the lower and upper head with a minimum thickness of 4-7/8 inches in the
cylindrical portion of the shell. The interior surface of the cylindrical shell and upper head
is clad with a weld deposited 304 stainless steel. The lower head is clad with a weld
deposited NiCrFe alloy (Inconel). All cladding material has a minimum thickness of 1/8
inch. The pressurizer was manufactured by the NSSS vendor, Combustion Engineering
(CE) .
In order to compensate for heat losses through the vessel wall and to maintain the
programmed pressure, the pressurizer is equipped with 96 single unit, direct immersion,
electrical resistance heaters (EHTR) (Figure 2) which protrude vertically into the
pressurizer through sleeves welded in the lower head. Of these 96 heaters, 73 were
manufactured by General Electric (GE) and the remaining 23 were manufactured by
Watlow Electric Manufacturing Company (manufacturer code: W065). The inconel heater
sleeves are welded to the internal pressurizer inconel cladding thus forming a system
pressure boundary. Heater elements are inserted into these sleeves with the heater
sheaths welded to the lower end of the sleeve also forming a system pressure boundary.
The failed heaters described in this event were manufactured by Watlow Electric
Manufacturing Company. The heaters are 88 inches in length, 0.875 inch outside diameter
and are rated at 12.5 KW minimum to 13.75 KW maximum at 480 VAC.
B-35
C. Sequence of Events
On April 22, 1987 at 1216 hours, a containment building entry was made on ANO-2 by
operations and health physics personnel to assess a suspected small leak on one of the four
safety injection tanks (BP) as indicated by a higher than normal makeup rate to the tank.
This initial containment building entry was performed because of Arkansas Power and
Light's (AP&L) previous experience with small RCS leaks and the potential for boric acid
corrosion of components as a result of such leaks. While in the containment, a general
walkdown of the building was conducted. During this inspection, leakage in the area of the
pressurizer was noted by the observation of boric acid crystal accumulation on the floor
beneath the vessel. Additional entries were made on April 23 and 24 to identify the source
of the leakage. As a result of these additional inspections it was determined that two small
RCS leaks existed: one non-pressure boundary leak of approximately 100 drops per minute
(dpm) from the open end of a pressurizer atmospheric high point vent line located above
the pressurizer and a potential pressure boundary leak of approximately 60 dpm from the
pressurizer lower head, originally thought to be from a pressurizer level instrumentation
nozzle (Figure 1). An Unusual Event was declared at 1720 hours on April 24, 1987, and a
unit shutdown and cooldown was commenced per the requirements of Technical
Specification 3.4.6.2. The unit was placed in Mode 3 (Hot Standby) at 2110 hours on April
24, 1987 and Mode 5 (Cold Shutdown) was achieved at 1028 hours on April 25, 1987.
Upon reaching cold shutdown, work was initiated to determine the actual location of the
leakage on the pressurizer lower head. On April 25, 1987, following removal of insulation
and visual inspection, the leakage source was determined to be the heater sleeve for the X1
pressurizer heater (Figure 2 and 3). As a result of the leakage around the X1 heater
sleeve, a small area of corrosion damage to the pressurizer carbon steel base metal was
found. This area was located about one-half inch away from and adjacent to the X1 heater
sleeve and was approximately one and one-half inches in diameter and three quarters of an
inch deep. The wastage mechanism was boric acid induced corrosion of the carbon steel
material. The pressurizer lower head was completely drained of water by removing the C1
heater located in the low point of the head. Attempts were then made to remove the heater
element from the X1 sleeve as well as two other heaters, AA1 and T4 which indicated failed
electrically. The AA1 heater was easily removed; however, the X1 and T4 heaters could not
be extracted. A video camera was inserted into the AA1 heater sleeve, adjacent to X1
(Figure 3), in order to determine why the X1 heater could not be removed. This inspection
revealed that the X1 heater sheath had ruptured. Several portions of the heater sheath
and internal heater components (conductor wires and resistance wires) were observed
laying in the area adjacent to the X1 sleeve. Inspection of the heater sleeve internal to the
pressurizer indicated cracking in several locations around its circumference. The portion of
the heater sheath immediately above the sleeve was ruptured and expanded outward
making it impossible to pull the heater from the sleeve.
On April 30, 1987 the pressurizer upper head manway was opened for additional visual
inspection of the X1 heater. At this time damage to the T4 heater in the area of the upper
heater support plate was also discovered. A heater adjacent to T4 was removed and a
video inspection of T4 performed. This inspection revealed damage similar to that of the
X1 heater (Figure 4). However, the T4 sheath rupture had not caused sufficient sleeve
damage to produce leakage.
B-36
II. EVENT CAUSE
A. Event Analysis
The leakage from the pressurizer high point vent line was due to seat leakage past the
solenoid valves (PSV) in the line. A small amount of leakage past valves of this type is not
uncommon and this leak presented no operational problems. The concern for corrosion
damage to components being wetted by the borated water leaking from the vent line
prompted discovery of the leak on the bottom head of the pressurizer.
The heaters originally installed in the ANO-2 pressurizer during plant construction were
manufactured by GE. In June 1981, as a result of pressurizer level instrumentation
problems, the pressurizer heaters were operated with less than the required water volume
to ensure they were covered. This resulted in the failure of 23 heaters. These defective
heaters were bypassed, and the unit was operated during the next cycle without these
heaters in service. In September 1982, during the 2R2 refueling outage, the 23 failed GE
heaters were removed and replaced with new heaters supplied to AP&L by CE. CE
purchased the new replacement heaters from Watlow Electric Corporation.
The RCS pressure boundary leak discovered on April 22, 1987, was determined to be
caused by the failure of the X1 heater sheath wzich resulted in damage to the X1 sleeve
and the internal sleeve to pressurizer cladding weld. Damage was sufficient to produce a
leak path allowing water from the pressurizer to pass between the sleeve and vessel wall
and eventually to atmosphere.
B. Safety Significance
The basis for ANO-2 Technical Specification 3.4.6.2 states that pressure boundary leakage
of any magnitude is unacceptable since it may be indicative of impending gross failure of
the pressure boundary. Evaluation of the damaged heater sleeves and the corroded area
on the pressurizer external surface indicated that the probability for gross failure of either
of these components at the time of discovery was small. The pressurizer wall is
approximately 4 inches thick in the corroded area. The corrosion damage had penetrated
to a maximum depth of approximately 3/4 of an inch or 18 percent of the wall thickness.
Also, only a very small fraction of the internal sleeve to vessel cladding weld was degraded.
Additionally it is important to note that in this application, the sheaths of the pressurizer
heaters are components of the RCS pressure boundary. The design of the Watlow heater
utilizes an internal high pressure seal constructed of a MICA-LAVA combination (Figure
2). The design specification for the heater states that this seal is a secondary pressure
boundary and during this event the seal demonstrated its capability to withstand full
system operating pressure and temperature without failure.
The length of time that the leak existed prior to detection is not known. In October 1986,
ANO-2 was placed in cold shutdown to perform repairs to a leaking main feedwater line
check valve. Prior to heatup after each cold shutdown outage, an inspection is conducted of
the containment building. It is unlikely that the leakage could have existed at this time
without being detected. RCS pressure boundary leakage detection systems at ANO-2 are
provided as required by 10CFR50, General Design Criteria 30. These systems include
B-37
redundant containment air monitoring systems capable of detecting radioactive gas and
particulates in the containment air; a containment sump level monitoring system; and
containment humidity, pressure and temperature monitors. Also reactor coolant system
inventory calculations are performed as required by Technical Specifications to detect
possible RCS leakage. The leak rate resulting from this event was estimated to be
approximately .002 gallons per minute, which is well within the normal makeup capacity of
plant systems. However, it is significantly below the threshold value of leakage which can
be detected by the installed leakage detection systems. Should the leakage have continued
undetected, the boric acid corrosion would have resulted in further degradation of the
pressurizer carbon steel material.
C. Root Cause
In order to determine the exact mechanism that resulted in the pressure boundary leak,
one ruptured heater sheath (X1), one non-failed heater manufactured from the same heat
of material as the failed heater (G2), and two heater sleeves (X1 and T4) were
metallurgically analyzed by CE.
The metallurgical examination revealed that the microstructure for the particular heat of
material used to fabricate the X1 heater had characteristics that indicated susceptibility to
primary water stress corrosion cracking (PWSCC). It is suspected that the manufacturing
process of the Watlow heaters induced the susceptibility for PWSCC. The Watlow heaters
were constructed utilizing an inconel heater sheath surrounding internal heater conductor
wires and resistance heating coils. Compacted MgO was used as an insulator between the
inconel sheaths and heating elements. The inconel heater sheath undergoes Waging to
reduce its diameter, simultaneously compressing the MgO insulating material to an
acceptable density. However, following the swaging operation no annealling of the heater
sheath was performed to eliminate residual stresses induced by the swaging operation,
remove the defects of cold work, or improve the microstructure Hence, residual stresses
from the cold work of the swaging operation remained and a cold worked microstructure
was present.
As the result, it was determined that heater X1 had failed by means of PWSCC. Once the
X1 heater sheath cracked through wall, the magnesium oxide (MgO) insulation within the
heater was wetted and transformed to magnesium hydroxide. The insulation material
subsequently was able to swell to approximately 150 percent of the original volume.
Stresses induced by this swelling caused further cracking of the X1 heater sheath. This
swelling also applied sufficient stresses to the X1 heater sleeve to cause it to crack,
resulting in the primary system leakage.
Additionally, as part of the root cause determination of this event, a review was conducted
of the design and manufacturing process used for the GE heaters installed in the
pressurizer. The GE heaters were fabricated by a multipass swaging step to reduce the
heater size followed by an anneal at 1950-2050 degree Fahrenheit, another cold swaging
step, a second anneal, and a final reduction step. The GE manufacturing process appears
to be designed to minimize the detrimental influences of cold work and residual stresses.
The dissolution and redistribution of the carbide precipitates caused by the annealing
process would also result in a microstructure which is less susceptible to PWSCC.
B-38
D. Basis for Reportability
This event is being reported per 10CFR50.73(a)(2)(i) and 50.73(a)(2)(ii).
Upon discovery of the potential RCS unisolatable pressure boundary leakage, a plant
shutdown was initiated as required by ANO-2 Technical Specification 3.4.6.2. This was
reported per 10CFR50.72(b)(i)(A) at 1742 hours on April 24, 1987 as the initiation of a
nuclear plant shutdown required by the plant's Technical Specifications. The completion of
any nuclear plant shutdown required by the plant's Technical Specifications is reportable
per 10CFR50.73(a)(2)(i).
Guidance provided in NUREG 1022, Licensee Event Report System, indicates that the
reporting requirements of 10CFR50.73(a)(2)(ii) includes material problems (e.g.,
metallurgical, chemical) that cause abnormal degradation of principal safety barriers (i.e.,
reactor coolant system pressure boundary); therefore, the provisions of this section are also
applicable to this event. This event is also being reported because of the generic
implications.
III. CORRECTIVE ACTION
A. Immediate
Upon verification that the RCS leakage was pressure boundary leakage associated with a
pressurizer heater, AP&L assembled a team of personnel to assess the extent of damage,
determine the cause, and develop a repair plan. This team included personnel from
various AP&L organizations, NSSS vendors (including Combustion Engineering and
Babcock and Wilcox), and several consulting firms. As work progressed and more
information became available as to the extent and cause of damage to the pressurizer, it
became apparent that a permanent repair to restore the vessel to an original condition was
not feasible at this time. This conclusion was reached primarily based on the facts that the
tooling and repair techniques needed to repair the heater sleeves internal to the
pressurizer remotely from outside the vessel were not available and would take several
months to develop and qualify. Also, a permanent repair would require personnel access to
the lower head area inside the vessel which would necessitate removal of the internal
heater support plates and all installed heaters, and exposure of personnel to very high
radiation levels inside the pressurizer. A decision was made to pursue a temporary repair
which would allow plant operation until the following refueling outage.
B . Subsequent
A summary of the temporary repairs performed follows:
• All heaters manufactured by Watlow, except for X1 and T4, that were installed
in the pressurizer (21 total) were removed. The empty sleeves were examined
and their integrity verified. Six new heaters manufactured by GE that were
available in spare parts inventory were inspected and installed in the
pressurizer. The remaining empty heater sleeves (15) were fitted with dummy
inconel heater plugs welded to the sleeves.
B-39
• The heater sleeves for the X1 and T4 heaters were cut off approximately three
eighths inch below the internal welded area and removed from the vessel. The
remainder of the sleeves were then drilled out and the area of sleeve removal
inside the vessel was inspected by nondestructive examination (NDE) to verify
that no defects existed in the cladding or clad to base metal interface. After
thorough examination of this area and the bores of the holes through the vessel
wall, plugs were inserted into the X1 and T4 holes and welded to the outside of
the vessel (Figure 5) utilizing the temper bead welding repair process of Section
III, ASME Code, 1983 Edition Winter 85 Addenda. The plugs were designed and
manufactured in accordance with the ASME Code and were made from the same
type of material as the vessel wall, SA-533, GR B, CL-1 alloy steel. All welding
and inspections were performed in accordance with paragraph NB-4622.9 of the
referenced ASME Code.
Because the plug and walls of the sleeve penetration are made of a non-clad
carbon steel alloy and would be exposed to the borated water environment of the
pressurizer during operating and shutdown conditions, independent
assessments of the affects of corrosion of the various materials for the remainder
of the current fuel cycle were conducted. These evaluations considered general
corrosion of the exposed P3 material, galvanic corrosion of the P3 material
coupled to inconel clad, crevice corrosion of the P3 plug and pressurizer base
metal, and stress corrosion cracking and hydrogen embrittlement of the P3
material and weldment. These evaluations were appropriately conservative in
nature, and concluded that the predicted small levels of corrosion were
acceptable for the remainder of the current fuel cycle.
• The corroded area located on the external surface of the pressurizer adjacent to
the X1 heater sleeve was repaired as an integral part of the process used to
install the X1 plug. In order to perform the temper bead weld of the plug to
vessel wall, the metal around the X1 hole was ground out in an oval shaped
pattern. Maximum final dimensions of the excavated area were 3-15/16 inches
in length and 1 inch in depth. This process effectively removed all of the
corrosion damaged area. The area was verified to be free of defects by NDE and
filled in when the temper bead weld was made on the X1 plug.
Final NDE, including ultrasonic examination, was performed on both plugs and welded
areas. The results of these inspections showed no unacceptable indications.
The final hole diameters for the X1 and T4 plugs (Figure 5) were 1.188 and 1.181 inches,
respectively. In the highly unlikely event that either of the repaired areas were to fail
catastrophically during operation, it would result in a hole in the bottom of the pressurizer
of approximately 0.0076 square feet. A review was conducted of small break Loss of
Coolant Accident (LOCA) cases previously analyzed by Combustion Engineering to insure
this postulated failure was bounded by analysis. This review identified an analysis which
specifically addressed the results of a stuck open pressurizer power operated relief valve
(PORV) with an effective area of 0.00754 square feet. The results of the analysis show that
reactor vessel level never decreases below the bottom of the hot leg and the core remained
covered throughout the transient. The only difference between the analyzed case and the
B-40
postulated plug ejection would be the location of the hole. This minor difference would
have little or no effect on plant response or the overall results of the analysis. Therefore,
even assuming failure and total ejection of either plug from the vessel, the core would
remain covered throughout the event. Also, the current ANO-2 Emergency Operating
Procedures address LOCAs of all sizes.
An extensive effort was initiated to remove the damaged X1 and T4 heater parts from
inside the pressurizer. Remote video cameras were inserted into sleeves in the lower head
where heaters had been removed and down through the upper vessel manway opening to
identify the location and type of debris present. As expected, these inspections revealed
debris consisting of the constituent parts of the heaters including inconel sheath material,
MgO insulation, 80-20 NiCr resistance wire, and nickel clad copper conductors.
Dimensions of the parts varied randomly ranging from less than 1 inch to several inches in
length with various geometries. Special tools inserted through the upper manway opening
were utilized to remove as much debris as possible; however, physical obstructions such as
heater support plates and intact heaters prevented complete removal of all debris. A final
inspection was performed to quantify, as close as possible, all parts left in vessel. This
information will be useful should inspections or assessments become necessary in the
future.
An evaluation of the effect of leaving loose heater parts in the pressurizer was conducted
by CE. This evaluation considered effects on reactor coolant chemistry, heat transfer
characteristics of steam generator tubes and fuel rods, potential damage to internal
pressurizer components such as operating heaters, and the possibility that small parts
might exit the pressurizer through the surge line and be transported through the RCS
loops.
The evaluation showed the following:
1) The materials are inert and will not combine or react with any known material in the
RCS.
2) Magnesium oxide will exist as a suspended solid in the RCS but the small quantities
remaining will make its presence insignificant.
3) Loose parts lodged against the surface of an operating heater will not result in damage
due to vibration/flow indicated wear or the creation of hot spots along the surface of the
heater.
4) The probability for very small pieces to exit the pressurizer and be transported into the
loops is remote. The surge line extends approximately 18 inches into the pressurizer
lower head and is covered with a capped screen containing 792, 1/2 inch diameter holes
(Figure 4). Flow velocity in the pressurizer is low and debris located in the bottom head
will not be affected by flow through the screen. If a small piece were to exit the
pressurizer through the screen, it most likely would be reduced to harmless particles
due to abrasive actions in the loops prior to reaching the reactor core.
B-41
It was concluded that no adverse conditions would exist as a result of heater debris
remaining within the vessel. This conclusion is further substantiated by the fact the
pressurizer has been operating for some time with debris present without any visible
damage to the vessel, reactor coolant system, or the core.
The process used to repair the ANO-2 pressurizer and the results of the many evaluations
performed to support this extensive effort demonstrated that ANO-2 could be returned to
operation without undue risk. However, to further assure the safe operation of the unit
pending final repairs to be completed during the next refueling outage, several
compensatory actions were implemented. These include the following:
•. The Containment Air Monitoring (CAM) System (IL) has the capability to draw
air samples from various locations within the containment building. A suction
line is located in the south steam generator cavity which contains the
pressurizer. This line was modified by rerouting the line from its current
location in the cavity to an area above the pressurizer. This was done to
increase the capability of the system to detect very small leakage rates that
might exist should either of the temporarily repaired pressurizer areas develop a
leak. Grab samples were taken at least weekly from this sample location and
analyzed for radioactivity. Although the minimum level of leakage that can be
detected utilizing this system will vary with overall RCS leakage, calculations
indicate that the system should be able to detect leakage from the pressurizer on
the order of .01 gpm. This is a significant increase in leak detection capability of
this system.
•. Procedures were developed to perform routine containment building entries

during power operation to inspect accessible portions of the RCS for the purpose
of leak detection and assessment. These tours were conducted on a monthly
basis unless prohibited by ALARA considerations due to abnormal containment
conditions. The inspections included specific examinations of the area beneath
the pressurizer while the temporary plugs were installed in the vessel.
In a meeting with the NRC staff on October 26, 1987, AP&L presented its proposed plans
for a permanent repair to the pressurizer at heater locations X1 and T4. During the last
refueling outage (2R6) completed in May 1988, permanent repairs to the pressurizer
locations X1 and T4 were completed. The details of the final permanent repair plant were
submitted to the NRC in our letter of May 27, 1988 (2CAN058810). Additionally, six
dummy inconel heaters plugs were removed and new heaters installed.
The pressurizer solenoid operated high point vent valve determined to be the source of
leakage above the pressurizer was replaced with a new valve. The disc and seat of the new
valve were lapped to minimize the potential for leakage past the seat. In addition, a plant
modification was made to install a collection funnel underneath the open end of the high
point vent line to collect any leakage past the valves in this line. A drain line from the
funnel was routed to a floor drain in the containment basement. This should prevent the
possibility of any future RCS leakage from this location contacting equipment in this area.
B-42
A detailed inspection was performed to determine if any components had been affected as a
result of the leakage from the high point vent line to atmosphere located above the
pressurizer. Insulation was removed from the top and side of the pressurizer and
inspections made of the upper head and vertical shell. Small amounts of boric acid crystals
were found in various locations under the insulation; however, no corrosion damage was
found to the shell or other pressurizer components such as safety valves flanges, supports
or instrumentation nozzles.
C. Future
None.
IV. ADDITIONAL INFORMATION
A. Similar Events
A similar event involving boric acid induced corrosion resulting in a degradation of the
RCS pressure boundary was reported in LER 50-313/86-006 (lCAN128607).
B. Supplemental Information
While reconnecting the pressurizer heaters electrically on May 22, 1987 two heaters, AA2
and R1 (both GE heaters) were found to indicate shorted electrically. Both heaters were
removed from the pressurizer and dummy inconel heater plugs inserted and welded in the
empty heater sleeves. The sheaths on these heaters were inspected visually by NDE and
no defects were found. Further investigation did not reveal any leakage of the heater
sheaths. The unit was heated up and reached Hot Standby by May 26, 1987.
On July 6, 1987 a through wall defect in the seal weld on the dummy inconel heater plug at
location Y4 resulting in an RCS pressure boundary leak was discovered. The details of this
event and the corrective actions taken were submitted in LER 50-368/87-006-00
(2CAN078704) and supplemental LER 50-368/87-006-01 (2CAN068810).
Energy Industry Identification System (EIIS) codes are identified in the text at (XX).
ANO 2 PRESSURIZER
FIGURE 1
FIGURE OMITTED - NOT KEYABLE (DRAWING)
WATLOW HEATER CONSTRUCTION

FIGURE 2
HEATER LOCATION MAP

FIGURE 3
B-43
ANO 2 PRESSURIZER
FIGURE 4
HEAD PENETRATION PLUG AS-BUILT GEOMETRY

FIGURE 6
TITLE: Pressurizer Heater Rupture Results in Reactor Coolant Pressure Boundary

Leakage and Pressurizer Vessel Corrosion

SECTIONS 50.73(a)(2)(i), 50.73(a)(2)(ii), Other Generic

NAME: Larry A. Taylor, Plant Licensing Engineer
TELEPHONE #- 501-964-3100

CAUSE: B SYSTEM: AB COMPONENT: EHTR MANUFACTURER: W065
CAUSE: C SYSTEM: AB COMPONENT: AE MANUFACTURER: C490
SUPPLEMENTAL REPORT EXPECTED: Yes
EXPECTED SUBMISSION DATE: 11/01/87
ABSTRACT:
On 4/24/87 an Unusual Event was declared and a unit shutdown and cooldown commenced
due to a suspected pressure boundary leak of approximately 60 drops per minute from the
area of the pressurizer vessel lower head. Examinations following the cooldown revealed
the leakage was due to a damaged heater sleeve which penetrates the vessel lower head
and houses a heater. A small area of corrosion damage to the carbon steel material of the
vessel was found adjacent to the leaking sleeve. Subsequent investigation revealed the
damaged sleeve resulted from rupture of the heater sheath. The ruptured heater and other
heaters manufactured by Watlow Electric Company (EIIS Identifier = AB-EHTR-W065)
have been removed from the pressurizer (twenty-three of the total ninety-six heaters). An
B-44
additional Watlow heater was also found to have a ruptured sheath but had not resulted in
sleeve leakage. The two damaged heaters' sleeves have been removed and plugs inserted
and welded into the sleeve penetrations. The corrosion damaged area has been repaired.
Due to limited replacement heater availability, sleeves containing no heaters have been
fitted with dummy plugs welded in place. Investigations are continuing to determine the
root cause of the heater sheaths failure. This event is reported per the requirements of
50.73(a)(2)(i), 50.73(a)(2)(ii) and due to the generic implications of the Watlow heater
failure.
(END OF ABSTRACT)
A. Unit Status
At the time of discovery of the event on April 24, 1987, the unit was operating at 100%
power with a reactor coolant system (RCS) average temperature of 580 degrees Fahrenheit
and pressurizer pressure of 2250 PSIA. The unit had been in continuous operation for 167
days.
Pressurizer: EIIS Identifier = AB-AE-C490.
The AN0-2 pressurizer (Figure 1) is a component of the RCS which functions to maintain
system operating pressure and compensate for changes in coolant volume during load
changes. The pressurizer is a vertically mounted, cylindrical vessel made of SA-533, Grade
B, CL-1 (P-3) alloy steel and was designed and fabricated in accordance with the ASME
Boiler and Pressure Vessel Code. Minimum vessel wall thickness is 3-7/8 inches in the
area of the lower and upper head with a minimum thickness of 4-7/8 inches in the
cylindrical portion of the shell. The interior surface of the cylindrical shell and upper head
is clad with a weld deposited 304 stainless steel. The lower head is clad with a weld
deposited NiCrFe alloy (Inconel). All cladding material has a minimum thickness of 1/8
inch. The pressurizer was manufactured by the NSSS vendor, Combustion Engineering.
Pressurizer Heaters: EIIS Identifier = AB-EHTR-W065.
In order to compensate for heat losses through the vessel wall and to maintain the
programmed pressure the pressurizer is equipped with 96 single unit direct immersion
electrical resistance heaters (Figure 2) which protrude vertically into the pressurizer
through sleeves welded in the lower head. The inconel heater sleeves are welded to the
internal pressurizer inconel cladding thus forming a system pressure boundary. Heater
elements are inserted into these sleeves with the heater sheaths welded to the lower end of
the sleeve also forming a system pressure boundary. The failed heaters described in this
event are 88 inches in length, .875 inch outside diameter and are rated at 12.5 KW
minimum to 13.75 KW maximum at 480 vac. The heaters were manufactured by Watlow
Electric Manufacturing Company.
B-45
C. Sequence of Events
On April 22, 1987 at 1216 hours a containment building entry was made on ANO-2 by
operations and health physics personnel to assess a suspected small leak on one of the four
safety injection tanks (EIIS Identifier = BQ-TK) as indicated by a higher than normal
makeup rate to the tank. This initial containment building entry was performed because
of AP&L's previous experience with small RCS leaks and the potential for boric acid
corrosion of components as a result of such leaks. While in the containment, a general
walkdown of the building was conducted. During this inspection, leakage in the area of the
pressurizer was noted by the observation of boric acid crystal accumulation on the floor
beneath the vessel. Additional entries were made on April 23 and 24 to identify the source
of the leakage. As a result of these additional inspections it was determined that two small
RCS leaks existed: one leak of approximately 100 drops per minute (dpm) from the open
end of a pressurizer atmospheric high point vent line located above the pressurizer and
another of approximately 60 dpm from the pressurizer lower head. Evaluations of the
leaks were conducted, and it was determined that the lower head leak was potentially RCS
pressure boundary leakage from a pressurizer level instrumentation nozzle (Figure 1). An
Unusual Event was declared at 1720 hours on April 24, 1987, and a unit shutdown and
cooldown was commenced per the requirements of Technical Specification 3.4.6.2. The unit
was placed in Mode 3 (hot standby) at 2110 hours on April 24, 1987, and Mode 5 (cold
shutdown) was achieved at 1028 hours on April 25, 1987.
Upon reaching cold shutdown work was initiated to determine the actual location of the
leakage on the pressurizer lower head. On April 25, 1987, following removal of insulation
and visual inspection the leakage source was determined to be the heater sleeve for the X1
pressurizer heater (Figure 2 and 3). As a result of the leakage around the X1 heater, a
small area of corrosion damage to the pressurizer carbon steel base metal was found. This
area was located about one-half inch away from and adjacent to the X1 heater sleeve and
was approximately one and one-half inches in diameter and three quarters of an inch deep.
The wastage mechanism was boric acid induced corrosion of the carbon steel material. The
pressurizer lower head was drained completely of water by removing the C1 heater located
in the low point of the head. Attempts were then made to remove the heater element from
the X1 sleeve as well as two other heaters, AA1 and T4 which indicated failed electrically.
The AA1 heater was easily removed, however, the X1 and T4 heaters could not be
extracted. A video camera was inserted into the AA1 heater sleeve, adjacent to X1 (Figure
3), in order to determine why the X1 heater could not be removed. This inspection revealed
that the X1 heater sheath had ruptured. Several portions of the heater sheath and
internal heater components (conductor wires and resistance wires) were observed laying in
the area adjacent to the X1 sleeve. Inspection of the heater sleeve internal to the
pressurizer indicated cracking in several locations around its circumference. The portion of
the heater sheath immediately above the sleeve was ruptured and expanded outward
making it impossible to pull the heater from the sleeve.
On April 30, 1987 the pressurizer upper head manway was opened for additional visual
inspection of the X1 heater. At this time damage was also discovered to the T4 heater in
the area of the upper heater support plate. A heater adjacent to T4 was removed and a
video inspection of T4 performed. This inspection revealed damage similar to that of the
B-46
X1 heater (Figure 4). However, the T4 sheath rupture had not caused sufficient sleeve
damage to produce leakage.
II. EVENT CAUSE
A. Event Analysis
The leakage from the pressurizer high point vent line was due to seat leakage past the
solenoid valves (EIIS Identifier = AB-PSV) in the line. A small amount of leakage past
valves of this type is not uncommon and this leak presented no operational problems. The
concern for corrosion damage to components being wetted by the berated water leaking
from the line prompted discovery of the leak on the bottom head of the pressurizer.
The pressure boundary leak was determined to be caused by the failure of the X1 heater
sheath which resulted in damage to the X1 sleeve and sleeve to pressurizer cladding weld.
Damage was sufficient to produce a leak path allowing water to pass between the sleeve
and vessel wall and eventually to atmosphere.
The basis for ANO-2 Technical Specification 3.4.6.2 states that pressure boundary leakage
of any magnitude is unacceptable since it may be indicative of impending gross failure of
the pressure boundary. Evaluation of the damaged heater sleeves and the corroded area
on the pressurizer external surface indicated that the probability for gross failure of either
of these components at the time of discovery was small. The pressurizer wall is
approximately four inches thick in the corroded area. The corrosion damage had
penetrated to a maximum depth of approximately three quarters of an inch or 18% of the
wall thickness. Also, only a very small fraction of the internal sleeve to vessel cladding
weld was degraded. However, should the leakage have continued undetected the boric acid
corrosion would have resulted in further degradation of the pressurizer carbon steel
material. Additionally, it is important to note that in this application the sheats of the
pressurizer heaters are components of the RCS pressure boundary. The design of the
Watlow heater utilizes an internal high pressure seal constructed of a MICA-LAVA
combination (Figure 2). The design specification for the heater states that this seal is a
secondary pressure boundary and during this event the seal demonstrated capability to
withstand full system operating pressure and temperature without failure.
The length of time that the leak existed, prior to detection, is not known. In October 1986,
ANO-2 was placed in cold shutdown to perform repairs to a leaking main feedwater line
check valve. Prior to heatup, after each cold shutdown outage, an inspection is conducted
of the containment building. It is unlikely that the leakage could have existed at this time
without being detected. RCS pressure boundary leakage detection systems at ANO-2 are
provided as required by 10CFR50, General Design Criteria 30. These systems include
redundant containment air monitoring systems capable of detecting radioactive gas and
particulates in the containment air, a containment sump level monitoring system, and
containment humidity, pressure and temperature monitors. Also reactor coolant system
inventory calculations are performed as required by Technical Specifications to detect
possible system leakage. The leak rate, resulting from this event, was estimated to be
approximately .002 gallons per minute, which is well within the normal makeup capacity
of plant systems. However, it is significantly below the threshold value of leakage which
B-47
can be detected by the installed leakage detection systems. Similar instances of boric acid
corrosion attack of RCS carbon steel components as a result of system leakage are
described in I.E. Information Notice 86-108 and ANO Licensee Event Report 50-313/86-
006.
B. Root Cause
The heaters originally installed in the ANO-2 pressurizer during plant construction were
manufactured by General Electric (GE). In June 1981, as a result of pressurizer level
instrumentation problems, the pressurizer heaters were operated with less than the
required water volume to insure they were covered. This resulted in the failure of twenty-
three heaters. These defective heaters were bypassed, and the unit was operated during
the next cycle without these heaters in service. In September 1982, during the 2R2
refueling outage, the twenty-three defective GE heaters were removed and replaced with
new heaters supplied to AP&L by Combustion Engineering (CE). Due to unavailability of
identical GE heaters, CE purchased the new replacement heaters from Watlow Electric
Corporation.
The Watlow heaters were constructed utilizing an inconel heater sheath surrounding
internal heater conductor wires and resistance heating coils. A compacted magnesium
oxide (MgO) was used as an insulator between the inconel sheaths and heating elements.
Based on investigations performed to date, it has been postulated that the inconel sheaths
for heaters X1 and T4 experienced stress corrosion cracking, allowing water to enter the
internal portion of the heater and contact the MgO insulating material. MgO exhibits a
high affinity for water absorption, and when exposed will hydrate and expand to
approximately two to three times its original volume. This phenomena produced swelling
and eventual rupture of the heater sheaths. The rupture produced significant damage to
the heater sleeves of the X1 and T4 heaters. Damage to the X1 sleeve and internal sleeve
to cladding weld was sufficient to produce a leak path allowing water to pass between the
sleeve and vessel wall and eventually to atmosphere.
The exact mechanism that caused the heater sheaths to crack initially has not been
determined at this time. However, based on conversations with representatives from
Watlow Manufacturing Company and internal evaluations, it is suspected that a step
utilized in the manufacturing process of these heaters may have caused the heater sheaths
to become susceptible to stress corrosion cracking. This step involves swaging of the
inconel heater sheath to reduce its diameter, simultaneously compressing the MgO
insulating material to an acceptable density. The swaging process induces high stresses
into the inconel material making it susceptible to stress cracking. Annealing of the
material to reduce induced stress was not performed after the swaging process. Portions of
the damaged heater sheaths have been removed from the pressurizer and sent to
Combustion Engineering and Babcock and Wilcox for metallurgical analysis to confirm this
hypothesis.
As part of the root cause determination of this event a review was conducted of the design
and manufacturing process used for the GE heaters installed in the pressurizer.
Conversations with personnel from GE concerning the construction of GE heaters indicated
that a similar swaging process is performed on the heater sheaths, however, the sheaths
B-48
are stress-relieved by annealing at approximately 2050 degrees Fahrenheit during

swaging.
C. Basis for Reportability
This event is being reported per 10CFR50.73(a)(2)(i) and 50.73(a)(2)(ii).
Upon discovery of the potential RCS pressure boundary leakage, a plant shutdown and
cooldown was performed as required by ANO-2 Technical Specification 3.4.6.2. The
completion of any nuclear plant shutdown required by the plant's Technical Specifications
is reportable per 10CFR50.73(a)(2)(i). The event was reported per 10CFR50.72(b)(i)(A) at
1742 hours on 4/24/87 as the initiation of a nuclear plant shutdown required by the plant's
Technical Specifications. This event is also being reported because of the generic
implications.
Guidance provided in NUREG 1022, Licensee Event Report System, indicates that the
reporting requirements of 10CFR50.73(a)(2)(ii) includes material problems (e.g.,
metallurgical, chemical) that cause abnormal degradation of principal safety barriers (i.e.,
reactor coolant system pressure boundary), therefore, the provisions of this section are also
applicable to this event.
A. Immediate
Upon verification that the RCS leakage was pressure boundary leakage associated with a
pressurizer heater, AP&L began assembling a team of personnel to assess the extent of
damage, determine the cause, and develop a repair plan. This team included personnel
from various AP&L organizations, NSSS vendors, including Combustion Engineering and
Babcock and Wilcox, and several consulting firms. As work progressed and more
information became available as to the extent and cause of damage to the pressurizer, it
became apparent that a permanent repair to restore the vessel to an original condition was
not feasible at this time. This conclusion was reached primarily based on the facts that the
tooling and repair techniques needed to repair the heater sleeves internal to the
pressurizer remotely from outside the vessel were not available and would take several
months to develop and qualify. Also, a permanent repair would require personnel access to
the lower head area inside the vessel which would necessitate removal of the internal
heater support plates, all installed heaters and exposure of personnel to very high
radiation levels inside the pressurizer. A decision was made to pursue a temporary repair
which would allow plant operation until the next refueling outage presently scheduled for
January 1988.
B. Subsequent
The pressurizer solenoid operated high point vent valve determined to be the source of
leakage above the pressurizer was replaced with a new valve. The disc and seat of the new
valve were lapped to minimize the potential for leakage past the seat. In addition, a plant
modification was made to install a collection funnel underneath the open end of the high
B-49
point vent line to collect any leakage past the valves in this line. A drain line from the
funnel was routed to a floor drain in the containment basement. This should prevent the
possibility of any future RCS leakage from this location contacting equipment in this area.
A detailed inspection was performed to determine if any components had been affected as a
result of the leakage from the high point vent line to atmosphere located above the
pressurizer. Insulation was removed from the top and side of the pressurizer and
inspections made of the upper head and vertical shell. Small amounts of boric acid crystals
were found in various locations under the insulation, however, no corrosion damage was
found to the shell or other components such as safety valves flanges, supports or
instrumentation nozzles.
A summary of the temporary repairs performed follows:
• All heaters manufactured by Watlow, except for X1 and T4 that were installed in
the pressurizer (21 total), were removed. The empty sleeves were examined and
their integrity verified. Six new heaters, manufactured by GE, that were
available in spare parts inventory since initial construction of the unit, were
inspected and installed in the pressurizer. The remaining empty heater sleeves
(15) were fitted with dummy inconel heater plugs welded to the sleeves.
• The heater sleeves for the X1 and T4 heaters were cut off approximately three
eighths inches below the internal welded area and removed from the vessel. The
remainder of the sleeves were then drilled out and the area of sleeve removal
inside the vessel was inspected by nondestructive examination (NDE) to verify
that no defects existed in the cladding or clad to base metal interface. After
thorough examination of this area and the bores of the holes through the vessel
wall, plugs were inserted into the X1 and T4 holes and welded to the outside of
the vessel (Figure 5) utilizing the temper bead welding repair process of Section
III, ASME Code, 1983 Edition Winter 85 Addenda. The plugs were designed and
manufactured in accordance with the ASME Code and were made from the same
type of material as the vessel wall, SA-533, GR B, CL-1 alloy steel. All welding
and inspections were performed in accordance with paragraph NB-4622.9 of the
referenced ASME Code.
Because the plug and walls of the sleeve penetration are made of a non-clad
carbon steel alloy and would be exposed to the borated water environment of the
pressurizer during operating and shutdown conditions, independent
assessments of the affects of corrosion on the various materials for the
remainder of the current fuel cycle were conducted. These evaluations
considered (1) General corrosion to the exposed P3 material, (2) Galvanic
corrosion of the P3 material coupled to inconel clad, (3) Crevice corrosion of the
P3 plug and pressurizer base metal, and (4) stress corrosion cracking and
hydrogen embrittlement of the P3 material and weldment. These evaluations
were appropriately conservative in nature, and concluded that the predicted
small levels of corrosion were acceptable for the remainder of the current fuel
cycle.
B-50
• The corroded area located on the external surface of the pressurizer adjacent to
the X1 heater sleeve was repaired as an integral part of the process used to
install the X1 plug. In order to perform the temper bead weld of the plug to
vessel wall, the metal around the X1 hole was ground out in an oval shaped
pattern. Maximum final dimensions of the excavated area were three and
fifteen sixteenths inches in length and one inch in depth. This process
effectively removed all of the corrosion damaged area. The area was verified to
be free of defects by NDE and f fled in when the temper bead weld was made on
the X1 plug.
Final NDE, including ultrasonic examination, was performed on both plugs and welded
areas. The results of these inspections showed no unacceptable indications.
An extensive effort was initiated to remove the damaged X1 and T4 heater parts from
inside the pressurizer. Remote video cameras were inserted into sleeves in the lower head
where heaters had been removed and down through the upper vessel manway opening to
identify the location and type of debris present. As expected, these inspections revealed
debris consisting of the constituent parts of the heaters including inconel sheath material,
magnesium oxide insulation, 80-20 NiCr resistance wire and nickel clad copper conductors.
Dimensions of the parts varied randomly ranging from less than one inch to several inches
in length with various geometries. Special tools inserted through the upper manway
opening were utilized to remove as much debris as possible, however, physical obstructions
such as heater support plates and intact heaters prevented complete removal of all debris.
A final inspection was performed to quantify, as close as possible, all parts left in vessel.
This information will be useful should inspections or assessments become necessary in the
future.
An evaluation of the effect of leaving loose heater parts in the pressurizer was conducted
by Combustion Engineering. This evaluation considered effects on reactor coolant
chemistry, heat transfer characteristics of steam generator tubes and fuel rods, potential
damage to internal pressurizer components such as operating heaters and the possibility
that small parts might exit the pressurizer through the surge line and be transported
through the RCS loops.
The evaluation showed the following:
1) The materials are inert and will not combine or react with any known material in the
RCS.
2) Magnesium oxide will exist as a suspended solid in the RCS but the small quantities
remaining will make its presence insignificant.
3) Loose parts lodged against the surface of an operating heater will not result in damage
due to vibration/flow induced wear or the creation of hot spots along the surface of the
heater.
4) The probability for very small pieces to exit the pressurizer and be transported into the
loops is remote. The surge line extends approximately eighteen inches into the pressurizer
B-51
lower head and is covered with a capped screen containing 792 one half inch diameter
holes (Figure 4). Flow velocity in the pressurizer is low, and debris located in the bottom
head will not be affected by flow through the screen. If a small piece were to exit the
pressurizer through the screen, it most likely would be reduced to harmless particles due
to abrasive actions in the loops prior to reaching the reactor core.
It was concluded that no adverse conditions would exist as a result of heater debris
remaining within the vessel. This conclusion is further substantiated by the fact the
pressurizer has been operating for some time with debris present without any visible
damage to the vessel or reactor coolant system.
The process used to repair the ANO-2 pressurizer and the results of the many evaluations
performed to support this extensive effort have demonstrated that ANO-2 can be returned
to operation without undue risk. However, to further assure the safe operation of the unit,
several compensatory actions are being implemented. These include:
• The Containment Air Monitoring (CAM) System has the capability to draw air
samples from various locations within the containment building. A suction line
is located in the south steam generator cavity which contains the pressurizer.
This line has been modified by rerouting the line from its current location in the
cavity to an area above the pressurizer. This was done to increase the capability
of the system to detect very small leakage rates that might exist should either of
the temporarily repaired pressurizer areas develop a leak. Grab samples are
planned to be taken at least weekly from this sample location and analyzed for
radioactivity. Although the minimum level of leakage that can be detected
utilizing this system will vary with overall RCS leakage, calculations indicate
that the system should be able to detect leakage from the pressurizer on the
order of .01 gpm. This is a significant increase in leak detection capability of
this system.
• Procedures are being developed to perform routine containment building entries

during power operation to inspect accessible portions of the RCS for the purpose
of leak detection and assessment. These tours are planned to be conducted on a
monthly basis unless prohibited by ALARA considerations due to abnormal
containment conditions. Specific examinations of the area beneath the
pressurizer are planned while the temporary plugs are installed in the vessel.
The final hole diameters for the X1 and T4 plugs (Figure 6) are 1.188 and 1.181 inches,
respectively. In the highly unlikely event that either of the repaired areas were to fail
catastrophically during operation, it would result in a hole in the bottom of the pressurizer
of approximately .0076 square feet. A review was conducted of small break LOCA cases
previously analyzed by Combustion Engineering to insure this postulated failure was
bounded by analysis. This review identified an analysis which specifically addressed the
results of a stuck open pressurizer PORV with an effective area of .00754 square feet. The
results of the analysis show that reactor vessel level never decreases below the bottom of
the lot leg and the core remained covered throughout the transient. The only difference
between the analyzed case and the postulated plug ejection would be the location of the
hole. This minor difference would have little or no effect on plant response or the overall
B-52
results of the analysis. Therefore, even assuming failure and total ejection of either plug
from the vessel, the core would remain covered throughout the event. Also, the current
ANO-2 Emergency Operating Procedures address LOCAs of all sizes.
C . Future
Future actions planned as a result of this event include:
• A final confirmation of the root cause of this event is in progress. Preliminary

results of the metallurgical testing performed to date on the heater material
removed from the pressurizer are being reviewed. A detailed review of all
information related to the event has been performed. This included evaluation
of documents such as purchase orders, material specifications, quality assurance
and quality control records, heater installation procedures, and others spanning
a time period from the date of manufacture of the Watlow heaters to the date of
leak discovery. Discussions with the NSSS vendors and consulting firms
assisting AP&L in this effort are continuing. A summary document addressing
the findings of these efforts is being developed. Additional information will be
provided in an update to this LER which is planned to be submitted by
November 1, 1987.
• An action plan will be developed with the goal of developing the elements
necessary to provide a permanent repair of the ANO-2 pressurizer. The details
of this plan are scheduled to be submitted to the NRC by November 1, 1987.
A. Similar Events
A similar event involving boric acid induced corrosion resulting in a degradation of the
RCS pressure boundary was reported in 50-313/86-006.
B. Supplemental Information
While reconnecting the pressurizer heaters electrically on 5/22/87, two heaters, AA2 and
R1 (both GE heaters) were found to indicate shorted electrically. Both heaters were
removed from the pressurizer and dummy inconel heater plugs inserted and welded in the
empty heater sleeves. The sheaths on these heaters were inspected visually and by NDE
and no defects were found. Further investigation into the cause of these heater failures is
in progress.
The unit was heated up and reached hot standby on 5/26/87.
ANO 2 PRESSURIZER
FIGURE 1
B-53
WATLOW HEATER CONSTRUCTION

FIGURE 2
HEATER LOCATION MAP

FIGURE 3
ANO 2 PRESSURIZER
FIGURE 4
HEAD PENETRATION PLUG

AS-BUILT GEOMETRY
FIGURE 5
EVENT DATE:10/23/86 LER#; 86-006-00 REPORT DATE:12/09/86
FACILITY NAME: Arkansas Nuclear One, Unit One (ANO-1)
TITLE: Boric Acid Corrosion of Carbon Steel High Pressure Injection Nozzle and Reactor
Coolant System Cold Leg Piping
OPERATING MODE: N POWER LEVEL: 000
THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10CFR:

50,73(A)(2)(ii)

NAME: Patrick C. Rogers,Plant Licensing Engineer
TELEPHONE # 501-964-3100

CAUSE:X SYSTEM:BQ COMPONENT:NZL MANUFACTURER:B015
REPORTABLE TO NPRDS:Y
C A U S E : X S Y S T E M : B Q C O M P O N E N T : V MANUFACTURER:V085
REPORTABLE TO NPRDS:Y
SUPPLEMENTAL REPORT EXPECTED: NO
B-54
ABSTRACT:
Corrosion of a Reactor Coolant System (RCS) nozzle and adjacent cold leg piping was
discovered during refueling outage performance of RCS inservice inspection. RCS leakage
from a body-to-bonnet-seal on a normally open isolation valve in the HPI line had run down
the stainless steel HPI piping within the reflective line insulation to the carbon steel nozzle
and cold leg piping. The leakage was less than 0.10 gallons per minute and had occurred
over a period of about 6 months. This continuous leakage, the pipintemperatures, and the
piping to insulation fitup are believed to have resulted in pooling of what became highly
concentrated, corrosive boric acid along the bottom of the nozzle and adjacent cold leg
piping. The maximum depth of the corrosion was approximately 1/2 inch in the area of
minimum HPI nozzle thickness (3/4 inch including the thickness of an interior stainless
steel clad) at the HPI line to nozzle weld. The galvanic effect due to dissimilar metals
contributed to the severe corrosion in this area. Repairs consisted of grinding the areas of
corrosion and performing a weld buildup in the area of the most severe damage. An
evaluation and inspection of other components where boric acid attack of carbon steel could
occur or previously known sources of borated water leakage was performed with no
additional problems found. Instructions have been issued such that when in-containment
berated water leakage is identified the potential for boric acid attack of carbon steel is
thoroughly evaluated.
(END OF ABSTRACT)
A. Unit Status
At the time of discovery of the event, the unit was in refueling shutdown (1R7).
High Pressure Injection (HPI) nozzle "A"- EIIS Identifier = BQ-NZL-B015.
An HPI nozzle is attached directly on the side of each of the four reactor coolant system
(RCS) cold legs downstream of the reactor coolant pumps (RCP). The HPI nozzle for the
"D" RCP is also used for normal RCS makeup. The HPI nozzle and RCS cold leg piping are
constructed of carbon steel with a stainless steel interior cladding. The typical HPI nozzle
configuration for ANO-1 is shown in Figure 1. The RCS, including HPI nozzles, were
manufactured by Babcock and Wilcox.
HPI manual isolation valve MU-45A; EIIS Identifier = BQ-V-V085.
MU-45A is a normally open, manual isolation valve for the "A" HPI line. This valve is
provided to allow isolation or the RCS from the HPI system for performance of hydrostatic
testing of the HPI system. A similar valve is located in each of the other three HPI lines to
the RCS. MU-45A is located approximately 10 feet above the "A" HPI nozzle. This valve is
a 2-112 inch gate valve, model SAPS manufactured by Velan Valve Company.
B-55
C. Problem Description
During the ANO-1 1R7 refueling outage, a special inspection or the four HPI nozzle safe-
ends and thermal sleeves was performed as part of the augmented inservice inspection
program. These inspections were required a- a result of followup actions from an unrelated
event previously reported in LER 50=313/82-09. On 10/23/86, during the inspection of the
"A" HPI nozzle safe-end and thermal sleeve, severe corrosion damage was found along the
bottom exterior surface of the nozzle. This damage extended the length of the nozzle and
then onto the RCS cold leg piping.
Figure 2 shows the location or the damage. As shown in Figure 2, the corrosion damage
began adjacent to the nozzle to safe-end weld in the carbon steel material of the nozzle.
The most severe corrosion attack was at this point and produced two deep grooves or
cavities. The maximum depth of the corrosion in this area was approximately 1/2 inch. At
the area of most severe corrosion the nozzle has a wall thickness of 3/4 inches, including
the stainless steel clad thickness. Between the two initial corrosion grooves was a small
island of carbon steel material that sustained minimal damage. The damage at the weld
interface extended approximately 25 percent of the nozzle circumference (nozzle outside
diameter is 3-1/2 inches at this weld).
The two deep grooves extended longitudinally from the weld, Converged and narrowed to
approximately 1-1/2 inches in width at the beginning of the nozzle thickness transition
area. At this point the two grooves branched into four parallel paths, with depths of 1/16
inch to 6/16 inch, which continued along the nozzle bottom in the nozzle thickness
transition area. The two inner paths abruptly stopped at the end of the nozzle thickness
transition, while the outer two paths continued parallel along the nozzle bottom surface,
across the nozzle to RCS pipe weld and then vertically down the RCS cold leg piping for
approximately 6 inches. The maximum depth of the corrosion on the cold leg was
approximately 1/4 inch.
II. EVENT CAUSE
A . Event Analysis
The corrosion was initiated by RCS leakage from the HPI manual isolation valve
MU145A. Leakage from the body-to-bonnet seal ring for valve MU-45A was first
discovered by licensed plant operators on 8126186. The leak rate at that time was
measured to be 0.078 gpm and was described as a leaking fluid between 100 and 130
degrees Fahrenheit traveling down the HPI piping inside the piping insulation. Given that
the leakage rate was well within the technical specification limits and that the leakage was
believed to only be contacting the stainless steel piping of the HPI system, operation of the
unit was continued. On 9/18/85 while the unit was shutdown for an unrelated cause, the
seal ring for valve MU-45A was replaced. Following startup from this shutdown, leakage
was again observed from MU-45A on 9/26/85. The leak rate at this time was measured to
ba 0.094 gpm. Continued operation of the unit was still considered appropriate based upon
the initial justification. The leakage continued until successful repairs to MU-45A on
2/2/86 during another unrelated unit shutdown.
B-56
As the leakage of the RCS berated water flowed down the HPI piping, it was heated due to
the increasing pipe temperatures causing evaporation of the water from the solution. The
concentration or the boric acid increased due to the water loss which resulted in a decrease
of the pH of the remaining solution enhancing the corrosive potential of the fluid. It is
believed that pooling of the highly concentrated boric acid solution occurred due to the close
fit of the HPI piping insulation (especially in the area of the HPI nozzle thickness
transition) and possible plugging of the insulation seams by boric acid crystals. This
phenomenon caused the carbon steel of the outside of the nozzle to be continually wetted
with a hot, low pH highly concentrated boric acid solution, and may have involved various
chemical forms of molten boric acid, resulting in the corrosive attack. Acceleration of the
corrosion at the nozzle to safe-end weld is believed to have occurred due to the galvanic
effects or the dissimilar metals.
During normal operation and hot standby conditions, with the RCS at greater then 500
degrees Fahrenheit, it is believed the boric acid solution could not have been in a liquid
state as far as the RCS cold leg. Thus the corrosion on the RCS cold leg probably would not
have occurred at that time. However, during the period of leakage from MU-45A several
unit heatup and cooldowns occurred. During these times it is believed the leakage would
be enriched from the previous deposits of boric acid in the nozzle area and then deposited
on the areas not affected during periods of operation or hot standby.
B. Root Cause
The root cause of this event was the failure to fully recognize the potential communication
path of the leakage to carbon steel components and the corrosive attack potential of even
small quantities of allowable borated water leakage. Very few cases of berated water
leakage are expected to produce the damage experienced. The observed corrosion incident
is believed to have been due to the complex interactions of:
the continuous berated water leakage overtime the geometries of the leakage source and
system components the temperatures of the components involved the material properties of
the components exposed to berated water leakage
C . Basis for Reportability
This event is reportable under the provisions of 10 CFR 50,73(a)(2)(ii) since this involves a
material problem that caused abnormal degradation of the RCS pressure boundary. The
event is also being reported due to the potential generic implications of the event. This
event is considered significant for the degree of damage experienced over a relatively short
time frame. The event was reported per 10CFR 50.72 at 1220 hours on 10/23/86.
A . Im m e d i a t e
Upon discovery of the corrosion, visual and radiographic (RT) examinations were
performed to identify the nature and extent of the damage. Preliminary results from these
inspections indicated no through wall penetration and no stainless steel cladding exposure.
B-57
Grinding of the affected areas to sound metal was then performed. Dye penetrant (PT) and
ultrasonic (UT) examinations were performed $0 verify sound metal had been reached. A
maximum depth of 1/2 inch of material has required to be removed, confirming the
preliminary results. Areas which did not meet minimum wall thickness were repaired by
weld buildup. The weld repair was confined to the 3-1/2 inch O.D. portion of the nozzle.
The weld repair was performed in accordance with the requirements of ASME Section XI.
Repair of the remainder of the nozzle and the affected portion of the cold leg was by
grinding only. The grinding was such that the metal surfaces were blended uniformly into
adjacent surfaces with at least a 3 to 1 taper. Repairs were inspected in accordance with
ASME Section XI requirements. PT, RT and UT examinations were performed on the
ground areas which were not weld repaired. Examination results verified repair
acceptability and thermal sleeve integrity and tightness. A conservative stress analysis
was performed which qualified the repaired nozzle for at least one cycle of operation.
B. Subsequent
An inspection plan was developed to identify other potential problem areas within the
plant which could be or had been subject to boric acid leakage. Initial screening criteria
included:
engineering design review of systems to identify carbon steel components, identification of

previous sources of known borated water leakage, and system walkdowns to look for boric
acid/rust deposits
Sixty-nine (69) locations were identified as potential candidates for inspection. Of these
69, 24 were identified as requiring visual inspections based on the following screening
criteria:
physical location of carbon steel components in relation to leakage sources material

composition at locations, current inspections performed during the refueling outage, and
inspections which would determine whether related areas would need to be inspected.
Insulation was removed from these locations (if necessary) and visual inspections were
made. Although rust was present at some of the locations, no excessive general corrosion
attack, pitting or grooving of carbon steel surfaces was noted.
Additionally, until procedure modifications can be made, interim guidance has been
provided to the operations staff for assessment of the potential impact of boric acid solution
leakage. Emphasized in this guidance was that bright red iron oxide staining of the
exterior of the insulation, indicative of iron oxide removal by corrosion should be identified
during inspections of in-containment systems.
C . Future
Additional actions to be taken include:
A detailed stress analysis will be performed for the repaired nozzle to further qualify
extended operation.
B-58
Operating procedures are being modified to address the analysis and required corrective
actions to prevent recurrence of this type of event. The primary prevention methods will
be to minimize leaks and promptly repair leaks when they are detected.
A design modification will be made to the HPI line insulation. The insulation below MU-
45A will be modified in such a manner as to deflect any potential future leakage from the
valve to the outside of the insulation. The other HPI lines will be modified in a similar
manner.
No similar occurrences have been reported for ANO-1. A follow-up report is not planned.
FIGURE 1 ANO-1 Typical HPI/MU Nozzle Configuration
FIGURE OMITTED - NOT KEYABLE (GRAPHICS)
FIGURE 2 - AREA OF CORROSION DAMAGE
FIGURE OMITTED - NOT KEYABLE (GRAPHICS)
B-59

Boric Acid Corrosion Evaluation

Diunggah oleh

Informasi Dokumen

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Boric Acid Corrosion Evaluation

Diunggah oleh

Hak Cipta:

Format Tersedia

REPORT SUMMARY

Boric Acid Corrosion Evaluation (BACE) Program

EPRI TR-101108s Electric Power Research Institute

EPRI PERSPECTIVE A future objective of the BAC Evaluation Program

For further information on EPRI research programs, call

Final Report, December 1993

EPRI Project Manager

Nuclear Power Division

Boric Acid Corrosion Evaluation Program

This report is a continuation of EPRI/NMAC efforts to evaluate and predict potential

Boric Acid Corrosion Evaluation Program

Mr. Dennis P. Weakland Duquesne Light Co.

Mr. David E. Whitaker Duke Power Co.

Mr. Henry J. Thailer Pacific Gas & Electric Co.

Mr. Steve Fyfitch B&W Nuclear Services

Mr. John F. Hall ABB-CE Nuclear Services

Mr. Dave Miskiewicz Florida Power Corporation

Mr. Raj Pathania EPRI

1.0 Introduction 1-1

2.0 Industry Experience Summary 2-1

3.0 BAC Test Data Summary 3-1

4.0 Boric Acid Corrosion Mechanisms 4-1

Table of Contents (cont.)

5.0 Summary 5-1

6.0 References 6-1

Appendix A-BAC Literature Summary A-1

Appendix B-NRC Licensing Event Reports (LERs) B-1

Figure 2.1 Boric Acid Corrosion Evaluation Survey 2-13

Figure 4.1 Cross Section of an Uncorroded Metal Surface 4-5

Figure 4.2 General Corrosion Process 4-5

Figure 4.3 Galvanic Corrosion 4-6

Figure 4.4 Crevice Corrosion 4-6

Figure 4.5 Oxygen Concentration Cell Corrosion 4-7

Figure 4.6 Pitting Attack 4-7

Figure 4.7 Typical Intergranular Corrosion 4-8

Figure 4.8 Stress Corrosion Cracking 4-8

Figure 5.1 Boric Acid Corrosion Rate vs. Temperature 5-2

Table 2.1 Bolt Rejections Through 9/30/84 2-1

Table 2.2 Summary of Components Damaged by Boric Acid Corrosion 2-2

Table 2.3 Material Degradaton Incidents-1987-1992 2-15

Table 2.4 Sources of Leaks-1987-1992 2-15

Table 2.5 Leakage Incidents of Borated Water-(1987-1992) 2-15

Table 3.1 Westinghouse BAC Test Data 3-1

Table 3.2 Acid Corrosion Rates in Aqueous Environment 3-3

Table 3.3 ABB-CE Drip Test Results 3-6

Table 3.4 ABB-CE Borated Steam Tests on Fastener Materials 3-7

Table 3.6 ABB-CE Short Time Breaker Tests at 220°F 3-8

Table 3.7 ABB-CE Borated Steam Tests Data 3-9

Table 3.8 CE Owners Group Crevice Corrosion Test Data 3-10

Table 3.9 B&WOG Borated Water Tests on Carbon Steels 3-12

Boric Acid Corrosion Evaluation Program

Nuclear Maintenance Applications Center

This project is a continuation of previous EPRI efforts to provide useful information

Boric Acid Corrosion Evaluation Program

Boric Acid Corrosion Evaluation Program

2.0 Industry Experience Summary

Nuclear Maintenance Applications Center

Boric Acid Corrosion Evaluation Program

2.1 Utility Experience

2.1.1 Licensing Event Report (LER) Summaries

1. EVENT DATE: 03/20/92 LER#: 92-018-00 REPORT DATE: 06/22/92

(FULL TEXT AVAILABLE - SEE APPENDIX B)

FACILITY NAME: Fort Calhoun Station Unit No. 1

TITLE: Corrosion of Boric Acid System Bolts