ML20041D057
| ML20041D057 | |
| Person / Time | |
|---|---|
| Issue date: | 02/28/1982 |
| From: | Cheng C Office of Nuclear Reactor Regulation |
| To: | |
| References | |
| NUREG-0886, NUREG-886, NUDOCS 8203040167 | |
| Download: ML20041D057 (67) | |
Text
NUREG-0886 Steam Generator Tube Experience r
U.S. Nuclear Regulatory Commission Offico of Nuclear Reactor Regulation C. Y. Cheng Task Leader p* "%q,
! s, >
1 3-l
+,
I 2
R88 808 W "
0886 R PDR
NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room,1717 H Street, N.W.
Washington, DC 20555
- 2. The NRC/GPO Sales Program, U.S. Nuclear Regulatory Commission, Washington, DC 20555
- 3. The National Technical Information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of Inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Comrnission papers; and applicant and licensee documents and correspondence.
The following documents in the NUREG series are available for purchase from the NRC/GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission issuances.
Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents available from public and special technical libraries include all open literature items, such as books, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries.
Documents such as theses, dissertations, foreign reports and translations, and non-NRC conference proceedings are available for purchase from the organization sponsoring the publication cited.
Single copies of NRC draft reports are available free upon written request to the Division of Tech-nical Information and Document Control, U.S. Nuclear Regulatory Commission, Washington, DC 20555.
Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfolk Avenue, Bethesda, Maryland, and are available l
there for reference use by the public. Codes and standards are usually copyrighted and may be l
purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY 10018.
l GPO Printed copy price:
$4.50
Steam Generator Tube Experience Manuscript Completed: December 1981 Dato Published: February 1982 C. Y. Cheng Task Leader Division of Licensing Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555 l
l
- ,s*****U,,
r cf' m
ABSTRACT This report provides information pertaining to the status of PWR steam generator tube experience and the resolution of unresolved safety issues A-3, A-4, and A-5 regarding steam generator tube integrity.
It provides an over-view of the types of problems which have occurred in PWR steam generators with particular emphasis on recent operating experience.
The report also discusses short and long-term corrective actions being pursued by the industry to resolve these problems, steam generator inspection and repair requirements which have been established to ensure the continued safe operation of PWR steam generators, and occupational radiation exposures associated with the above-listed activities.
It should be noted that information included in this report represents the current NRC staff understanding of each issue.
This report is intended to be a followup to the similar reports, NUREG-0523 and NUREG-0571, which discusses tube operating experience with the recirculation ("U" tube) type and once-through type steam generators designed by Westinghouse and Combustion Engineering, and Babcock and Wilcox, respectively.
i
i LIST OF CONTRIBUTORS AND ACKNOWLEDGMENTS C.Y. Cheng
- NRR D. Crutchfield
- NRR E. Murphy
- NRR C. McCracken
- NRR R. Serbu
- NRR K. Wichman
- NRR L. Frank
- RES J. Strosnider
- RES A. Herdt
- IE:II The authors wish to thank John Weeks of Brookhaven National Laboratory for critical review of the manuscript, Sherry Holden and Carolyn Wilson for their excellent typing job, and Walter Oliu, NRC, who edited and oversaw production of this report.
f r
.t.
~
+
x ii
TABLE OF CONTENTS P_ag
,s ABSTRACT................................
i 3
LIST OF CONTRIBUTORS AND ACKNOWLEDGMENTS................
ii 1.
INTRODUCTION............................
1
\\
2.
STEAM GENERATOR DESIGNS.............
1 3.
HISTORY AND DESCRIPTION OF STEAM GENERATOR OPERATING PROBLEMS..............................
2 4
3.1 Types of Steam Generator Tube Problems............
2 3.2 History of Operating Problems.................
11 3.2.1 Early Experience with Wastage Degradation and Stress Corrosion Cracking...
12 3.2.2 Denting........................
12 3.2.3 Row 1 U-Bend Cracking.................
13 3.2.4 Recent Corrosion Problems at the Tubesheet and Sludge Pile Locations..............
14 3.2.5 Recent Pitting and Localized Wall Thinning Problems....
15 3.2.6 Fretting Problems...................
15 3.2.7 Operating Problems with B&W Once-Through N
Steam Generators...................
16 L
3.2.8 Summary of PWR Steam Generator Operating Experience.
17 s
s 3.3 Recent Plant-Specific Problems................
17 3.3.1 Westinghouse Steam Generators.............
24 3.3.'2 Combustion Engineering Steam Generators........
28 13.3.3 Babcock and Wilcox Steam Generators..........
30 3.4 Corrective Actions for Operating Plants............
31 i
4.
STEAM GENERATOR TUBE SURVEILLANCE AND REPAIR............
32 4.1 Inservice Inspection..
32
- 4. 2 Tube Repairs.........................
34 4.3' ' Primary to Secondary Leakage Rate Limits...........
35 5.
LONG TERM CORRECTIVE ACTIONS....................
36 5.1 Improved Designs.......................
36 5.2 Improved Water Chemistry Control...............
37 i
I e
I iii w ::. -
~
TABLE OF CONTENTS (Continued)
Pay 5.2.1 B&W Recommendations..
37 5.2.2 Westinghouse and Combustion Engineering Recommendations...................
38 5.2.3 Current Secondary Water Chemistry Licensing Practice for Operating Plants............
39 5.3 Sleeving...........................
40 5.4 Replacement....................
41 6.
OCCUPATIONAL EXPOSURE ASSOCIATED WITH STEAM GENERATOR MAINTENANCE............................
41 6.1 Maintenance and Inspection..................
43 6.2 Repairs............................
43 6.3 Replacement.
46 6.4 Outage Duration........................
46 6.5 Exposure Reduction Techniques.................
47 7.
RELATED RESEARCH PROGRAMS...........
49 7.1 NRC Steam Generator Confirmatory Research Program.
49 7.1.1 Steam Generator Tube Integrity.............
49 7.1.2 Stress Corrosion Cracking of PWR Steam Generator Tubing...
49 7.1.3 Improved Eddy Current I' service Inspection for Steam Generator Tubing..............
50 7.2 Electric Power Research Institute - Steam Generator Research..........................
Si 7.2.1 Steam Generator Technology Subprogram..
51 7.2.2 Steam Generator NDE..................
51 8.1 TECHNICAL RESOLUTION OF UNRESOLVED SAFETY ISSUES A-3, A-4, and A-5 REGARDING STEAM GENERATOR TUBE INTEGRITY.......
52 9.
CONCLUSIONS............................
53 l
l l
iv
l LIST OF TABLES Table Title Page 1
Operating Experience With Westinghouse PWR Steam Generators Through November 1981...............
18-19 2
Operating Experience With Combustion Engineering PWR Steam Generators Through November 1981............
20 3
Operating Experience With Babcock and Wilcox Once Through Steam Generators Through November 1981........
21 4
Foreign Operating Experience With PWR Steam Generators..
22-23 5
Steam Generator Replacement Summary...............
42 6
Occupational Exposure Related to Steam Generator Maintenance, Replacement and Repair..............
44-45 7
Steam Generator Annual Dose As a Percentage of Total Annual Dose (Selected Pressurized Water Reactors 1974-1980)...................
48 4
V
i LIST OF FIGURES P_ age a
Figure Title 3
1 A Typical Westinghouse Steam Generator............
2 A Typical Westinghouse Preheat Steam Generator 4
(Model D-2)........................
3 A Typical Combustion Engineering Steam 5
Generator System (System 80)................
4 Babcock and Wilcox Once-Through Steam Generator........
6 7
5 Drilled Tube Support Design.................
6 CE Steam Generator Egg Crate Tube Support 8
Plate Design........................
7 B&W Tube Support Plate Design................
9 8
Problem Areas in PWR Steam Generators............
10 vi i -
i STEAM GENERATOR TUBE EXPERIENCE i
1.
INTRODUCTION In pressurized water reactors (PWRsi, water in the primary coolant system is kept under pressure sufficiently high to prevent it from boiling.
This high-pressure water passes through tubes around which water circulates in a secondary system where steam is produced to drive the turbine generators.
The assembly in which the heat transfer takes place is the steam generator.
The tubes within it are an integral part of the primary coolant boundary, keeping the radioactive primary coolant in a closed system, isolated from the environment.
These tubes form a principal part of the reactor coolant pressure boundary and constitute by far its largest surface area.
PWR steam generators have experienced a variety of tube degradation problems for a number of years that are caused by corrosion and/or mechanical conditions.
Corrosion and mechanically induced damage are caused by complex interactions of water chemistry, thermal-hydraulic design, materials selection, fabrication methods, and operations.
Various types of corrosion have affected steam genera-tors at most operating plants, which has resulted in scheduled and unscheduled outages for the repair or replacement of these steam generators.
In addition to the adverse effect on plant availability, these repairs and replacements have increased occupational radiation exposure.
The primary safety goal for steam generator tubes is that they retain adequate structural integrity to avoid excessive leakage over the full range of normal operational, transient, and postulated accident conditions.
To provide assurance that each plant can be operated safely, the plant Technical Specifications include limits on primary and secondary system activity and primary-to-secondary leakage levels.
Licensees are also required to perform periodic inservice inspections of steam generator tubes by the eddy current test (ECT)* method. Tubes degraded beyond the limit specified in plant Technical i
Specifications must be plugged.
For some plants, NRC has approved sleeve repairs as an acceptable alternative to plugging, thereby permitting these tubes to remain in service.
This report provides an overview of tube degradation problems in PWR steam generators with particular emphasis on recent operating experience, short-and l
long-term corrective actions being pursued by the industry to resolve these problems, and steam generator inspection and repair requirements which have been established by the Nuclear Regulatory Commission (NRC) to ensure the continued safe operation of PWR steam generators.
l 2.
STEAM GENERATOR DESIGNS t
Currently there are two major types of steam generators in use in PWRs in the United States:
The recirculating U-tube type, manufactured by Westinghouse
- ECT is defined in Section 4.1 I
(W) (Figures 1 and 2), and Combustion Engineering (CE) (Figure 3), and the once-through type, manufactured by Babcock and Wilcox (B&W) (Figure 4).
All commercially operating Westinghouse-designed steam generators are vertical shell recirculation type units with drilled tube support plates (i.e., the support plates contain drilled holes through which the tubes are inserted).
Figure 1 shows a Westinghouse steam generator typical of those operating today.
New generations of Westinghouse steam generators contain an additional preheater section, as illustrated in Figure 2.
All Westinghouse steam generators use a nickel-base alloy (Inconel-600) tubing, except for Yankee Rowe, which uses stainless steel tubing.
The use of drilled support plate (Figure 5) is significant because the annular space between the steam generator tube and the drilled support plate is the location of several forms of degradation.
All commercially operating Combustion Engineering steam generators are of the recirculating vertical shell type with Inconel-600 tubing and integral steam separation equipment (Figure 3).
The CE design has a combination of drilled carbon steel partial support plates, similar to that of W design, and carbon steel " egg crates" for tube support (Figure 6).
The driTled plates in these steam generators are two partial plates located near the top of the tube bundle, with the exception of those at the Palisades Plant, which has six full-support plates with drilled holes.
Unlike other PWR designs, the B&W once-through steam generator is a vertical, straight-tube-and-shell, once-through heat exchanger with shell-side boiling to produce super-heated steam (Figure 4).
Primary coolant from the reactor enters the steam generator through a nozzle at the top, flows downward through more than 15,000 Inconel-600 tubes, is collected in the bottom head, and exits through two outlet nozzles.
A unique feature of this design is the broached tube support plate concept (Figure 7).
These tube support plates are fabri-cated from carbon steel, drilled, and broached at.three points spaced 120 apart.
The broached design effectively reduces stagnant areas where solids can con-centrate by creating large openings for the flow of water and steam.
Accordingly, this design minimizes the propensity for some forms of tube degradation associated with other PWR designs.
3.
HISTORY AND DESCRIPTION OF STEAM GENERATOR OPERATING PROBLEMS This chapter presents an overview of adverse steam generator tube operating experience to date, including a definition of primary modes of tube degradation observed and a summary discussion of recent plant-specific problems.
A dis-cussion of surveillance, plugging, and leak-rate limit requirements is provided in Chapter 4.
3.1 Types of Steam Generator Tube Problems The primary modes of steam generator tube degradation observed to date are defined below.
Degradation refers to any chemical or mechanical mechanism affecting a tube's integrity.
Denting:
Plastic deformation of tubes resulting f rom the buildup of carbon steel support plate corrosion products (magnetite) in tube-to-tube support plate annuli (Figure 8).
2
l l
,/ _ gtv
~
hoi f
GENERAron i
' j'
[
W8 TWE SEP A A TM
,)
,f.
0 MANWAY (2) 180 Apany E_
- j
. !, i e,
/
sweat yang HolSTURE UPPER $Ntt
~,
V 90 /g Steam Quality (j
gig-vipar0g Ih 1;f,,lplj q i
g fuCE BUNDLt J
V
'O
/g Steam Quality Pagt 180N rube SUPPORTS j
ms.
((
ggggg MANWAy
- f. t.1 rube PLATE
\\. ll? h
$UPPORT f00f g
CHANNEL HEAD
'\\,
l
\\
Pft it4Ax y x
N$$"'
\\s
\\'
PRIMAyy (
gny Figure 1 A typical Westinghouse steam generator 3
(LA I
AUXILl&RY FEE 0wATEk g j-m-
N0ZZLE
, g
(,
2
- i I'
l r
a-g
//
/
l l
STEAM SEPARATORS
/
\\
\\
\\
\\
4 e
9 P
_,. 4 e,,
I' J,
d
- TU8E BUNDLE V
j y
90 /g Steam Quality M
/
I 11 l6 i
I i
J J.
n h n
,o L
a TUBE SUPPORTS
l I
l l
h ni T
i j
[60y/g Steam Quality l
i i
i LOWER SHELL g,
i flf n
.)
,..__r. [ FEEDWATER M0ZZLE l(
16' SCH. 80 (MAIM FEEDwATER) 9 a
n f
i?
I 41
( ( 4,{Q +ac:
h d
4 PREHEATER SECTION T
PREHEATER DUTLET L
l PRIMARY COOLANT IMLET
(
C00LAhi CHAkhEL 1
Figure 2 A typical Westinghouse preheat steam generator (Model D-2) l 4
g Steam Outlet to Turbine Restrictor I
Steam Dryers-a
-Integral Steam Separation Equipment Manway V
90 /g Steam Duality eq Feedwater Inlet (10% Flow)
V
- 60 /g Steam Quality b Egg Crate Support Structure Inspection Hand Feedwater Inlet (90% Flow)
Hole-Blowdown Line Manway-Primary Coolant Outlet Primary Coolant Stay Cylinder Inlet
}
Pedestal Figure 3 A typical Combustion Engineering steam generator (System-80) 5 m
PRIMARY INLET NOZZLE 1\\1kkD m
f AUXILLIARY
!. ' h ','?
m,
! ;, c'
.. r.
)
2 FEEDWATER o
HEADER a
INLET g
- ,i
=
wm
,,.,i.
l.. '.,C,.h H
6 d f, h ' n b' ;;pO _l l
W l' i
@ STEAM OUTLF.T
{
, ; p;,.
,q b
g OZRE FEEDWATER NOZZLE g
e
.e-y 100 "/" Steam--[E
. j,.i, :**,*,f. j2!,
h
! i. ; -
X: ',M,k e-g l
ANNULAR FEEDWATER 7,l3gSH LLp -[
HEATING CH AMBER V
60 /g Steam 4121Na'Ook i.
2 o
Quality SHELL h BROACHED PLATEg, E!,;;:0 TUBE SUPPORTS 3
6 l
l' SHROUD
=!
..,i.'6;'
o m
y
=b f
i i
6l*('
' I' fI
,[*
i, ORIFICE PLATE TUBE SHEET PRIMARY OUTLET NOZZLES i
Figure 4 Babcock and Wilcox once-through steam generator 6
!E z
Ow O A
F Q-O 3
m y) Q. (4
.b ev U
i l
6" o
E O
7
[
n g
ise d
e t
\\
lp a
tropp m
\\
us
\\
e bu e
t e
.y tarc gge ro tare
.mk neg mae ts
<y E
C k
6 e
ru g
ig F
f w.
l l
0.430" M AX. R ADIUS l
0.427" MIN RADIUS MINIMUM DRILL R ADIUS 0.320" i
STEAM i
l '
[
4 I
GENERATOR l
{
TUBES CROSS SECTION OF TUBE
.I SUPPORT PL AT E
+
l l j
t MINIMUM OUTSIDE TUBE PITCH TUBE RADIUS 0.3125"
+0.875"10.010"+
5/8" DI A.'
OO q'!%(
N(
i.1/2" r
l I
ll l
D i
f I
H l
f
%[
I
,l
'N
?
if[
t N
N l
Figure 7 B&W tube support plate design 9
U-bend 0 (Magnetite)
Fe3 4
Y-M 1
- rw
.(g
'd' Eg 5
5 l } &
\\g p:)
U-bend ovality resulting from s
Denting flow slot 3
jy 3
deformation
,l l
[
i
[I 9
i 4
R iT Denting-induced Support plate
] Il deformation of tube holes
.)
support plate N
J4 flow slots W
OE "4
N Tube-to-Tubesheet Thinning T
) Crevice f
Sludge Pile #
IGA Cracking Figure 8 Typical problem areas in PWR steam generators 10
Erosion-Corrosion:
The combined effect of corrosion and erosion caused by thermal-hydraulic conditions and the impingement of fluids which possibly contain suspended particles or highly reactive chemicals.
Fatigue:
Material failure resulting from the initiation of cracks and/or their propagation because of cyclic loads.
Fretting:
The loss of tube material caused by excessive rubbing of the tube against its support structure.
This can be caused either by primary side or secondary side flow-induced vibr ation of the tubes.
Intergranular Attack (IGA):
This is a general term denoting the corrosive attack of grain boundaries in Inconel-600 with no preferential (stress-related) orientation.
Pitting:
Localized attack on tubing resulting from nonuniform corrosion rates caused by the formation of local corrosion cells.
Stress Corrosion Cracking (SCC):
Intergranular cracking of stressed tubes, without reference to a causative chemical agent.
This term is used either to encompass a number of known SCC mechanisms or when the chemical causing the corrosion is not known.
In this report SCC is also used in conjunction with the causative agent or with reference to locations where the corrosion is occurring as follows:
Caustic Stress Corrosion Cracking (CSCC) is used when the specific SCC causative agent has been identified as caustic.
Primary Side Stress Corrosion Cracking (PSCC) is used to identify SCC on the reactor coolant side (inside of the tubes).
Although the causative agent has not been identified for this type of corrosion, for the purpose of this report PSCC is also used to indicate Pure Water (Coriou) Stress Corrosion Cracking, which is a mechanism whereby no known imourities are required as a causative agent.
Although boric acid and lithium hydroxide are present in reactor coolant, it has been postulated that PSCC may be Pure Water Stress Corrosion Cracking.
Secondary Side Stress Corrosion Cracking (SSCC) is used to identify SCC on the secondary side (outside) of the steam generator tubes when the specific causative agent is not known (i.e., bes1k water chemistry analysis does not indicate the presence of free caustic so it cannot be identified as CSCC).
Wastage:
A localized secondary side corrosion of Inconel-600 caused by chemical attack from acid phosphate residues concentrated in low flow areas.
3.2 History of Operating Problems Of 48 operating PWRs in the U.S., a total of 40 have experienced some form of steam generator tube degradation.
As of November 1981, there are 40 operating PWR units in the United States with recirculating-type steam generators (){-32 and CE-8).
Of these, 32 (W-25 and CE-7) have been found to have one or more forms of tube degradation.~
11 1
These figures do not include eight operating PWRs with once-through type steam generators designed by'B&W.
Of these B&W units, all have had some forms of adverse tube operating exper,ience.
The authority to operate TMI-2 has been suspended and this plant is not included in the total number of operating units.
3.2.1 Early Experience with Wastage Degradation and Stress Corrosion Cracking Degradation experience at Westinghouse and Combustion Engineering units before the mid-1970s included wastage (localized thinning of tube walls) and caustic stress corrosion cracking on the secondary side.
The predominant method of controlling the secondary water chemistry during this period was coordinated phosphate control.
These early problems have been attributed to difficulties in adequately controlling phosphate concentrations and to impurities carried into the steam generators by feedwater.
The establishment of all-volatile treatment (AVT) control in the mid-1970s succaeded in arresting any further significant wastage by phosphates, but caustic stress corrosion cracking has continued to be a concern, particularly in plants with significant periods of phosphate operation before conversion to AVT.
Caustic stress corrosion crack-ing caused a tube rupture above the tubesheet elevation in February 1975 at Point Beach Unit 1, resulting in a 125 gpm primary-to-secondary leak. With the exception of two Westinghouse-designed plants, Robinson Unit 2 and San Onofre Unit 1, all PWRs have been converted to or have operated exclusively with AVT control.
Robinson Unit 2 and San Onofre Unit I had not experienced phosphate wastage at the rate experienced at other plants using phosphate chemistry during the period when the other PWRs converted to AVT.
3.2.2 Denting Denting refers to the squeezing of tubes at support plate or tubesheet inter-sections caused by the corrosion of the carbon steel support plates and tubesheet.
The buildup of nonprotective corrosion product oxides, consisting mainly of iron oxide (magnetite), leads directly to tube distortion at the sup-port plate intersections and to distortion and cracking of the support plates.
The tube distortions at the support plate intersections have resulted in numerous instances of tube leaks caused by stress corrosion cracks initiated primarily from the inside (primary side) surface of the dented tube.
Denting was first identified in 1975, when a number of plants which had shifted from phosphate water chemistry control to AVT began to develop unidentified eddy current testing (ECT) signals at tube supports.
In some cases, inspection probes could not be passed through the tubes, which indicated significant tube distortion.
Subsequently, steam generators which had never operated with phosphate water chemistry developed denting.
Approximately 24 Westinghouse and Combustion Engineering plants have reported denting, including eight plants where denting is considered extensive.
With the exception of a few plants, all currently operating plants are potenti. illy susceptible to denting if sufficient condenser leakage occurs.
Because copper oxide has been demonstrated to be a catalyst in these reactions, those plants with copper in their secondary systems are even more susceptible.
All plants which are coming on line during the next two years have steam generators which utilize carbon steel support plates and are therefore potentially susceptible to denting.
The earliest startup date for a plant using all ferritic stainless 12
steel supports, which are not susceptible to denting, is 1983.
The Westing-house F models and some D models and the CE system-80 steam generators have been or are being fabricated with all ferritic stainless steel supports.
Although B&W continues to use carbon steel support plates, their steam genera-tors feature a quatrefoil support design (with minimum contact area), along with virtually copper-free secondary systems and full flow condensate polish-ing (Figure 7).
Consequently, no denting has been reported at any B&W units to date.
3.2.3 Row 1 U-Bend Cracking The small-radius U bends in the inner (or first) row of tubing in Westinghouse steam generators have been subjected to primary-side-initiated stress corro-sion cracking.
These cracks have occurred either at the apex of the U-bends or at the tangent point transition between the U-bend and the straight span portion of the tubing.
At domestic plants such as Surry Units 1 and 2 and lurkey Point Unit'4, apex cracks have occurred as a result of service-induced ovality of the tube as a result of the denting process.
Denting leads to support plate deformation and eventually to closure of the support plate flow slots.
Closure of the upper support plate flow slots induced bending and ovality at the apex of the inner row U-bends.
At Surry Unit 2 in 1976 this phenomenon caused a tube rupture, with a resulting primary-to-secondary leak of 80 gallons por minute (gpm).
The current industry practice is to plug all inner row (i.e., row 1) tubes as a preventive measure when upper flow slot closure is observed.
This has prevented similar failures at other plants where i
tubing was extensively dented.
Apex cracks have also been observed in at least two Westinghouse-designed foreign facilities.
Doel, in Belgium, experienced a large leak at the apex of an inner row U-bend.
Although there was no active denting at this unit that the staff is aware of, there was significant ovality of the tubing, which was believed to have been introduced during the fabrication process.
Apex cracks have also been reported for the 0brigheim facility in West Germany, which has the same Mannesmann supplied tubing used at Doel.
Another category of U-bend cracks includes stress corrosion cracks located in the transition area between the U-bend and the straight portion of the tubing.
These cracks have generally been observed at plants which have not experienced denting.
This tangent point cracking phenomenon has been responsible for numerous small leaks over the past three years affecting Westinghouse Model 51 l
steam generators, particularly those at Trojan Unit 1.
Other affected domestic plants using Westinghouse Model 51 include Surry Unit 2, North Anna Unit 1, Farley Unit 1, D.C. Cook Unit 2, and Zion Unit 1.
(Two foreign facilities with Westinghouse steam generators, Takahama Unit 1 and Ringhals Unit 2, have also reported U-bend tangent point cracking.) With the exception of those at l
Zion, which were fabricated somewhat earlier, the tubes affected by this particular problem were fabricated L.ound 1971 by Westinghouse Specialty Metals Division (SMD).
Study of the U-bend samples removed from the Trojan Unit I steam generators revealed them to be characterized by a smooth transition at one tangent point and well defined intrados and extrados transitions at the other tangent point l
(" opposite transition").
At the opposite transition, the extrados transition typically occurs 0.6 inches above the intrados transition.
Three of the 13 l
26 tubes examined at Trojan contained an array of branched cracks which were initiated from the primsry side.
These cracks were located on the extrados of the opposite transition, between the intrados transition and the extrados transi-tion.
The mode of failure most closely resembles " pure water" (Coriou) stress corrosion cracking observed in deoxygenated pure water in laboratory experiments.
It is believed that the " opposite side" transition geometries were introduced during the fabrication process and resulted in increased residual stress at this location.
The fabrication procedure includes the insertion of an internal ball mandrel through the U-bend during the bending process to prevent excessive tube ovality.
Westinghouse has reviewed the bending techniques used by SMD during the period in which U-bends exhibiting opposite side transitions were fabricated. Westinghouse has been unable, to date, to identify exactly why certain tubes were affected and others were not.
The microstructure of the alloy, in addition to the stress, is believed to be a variable affecting susceptibility to the tangent point cracking phenomenon.
However, the studies performed on the tubes at Trojan revealed no consistent or significant relationship between cracking and grain size, carbide distribution, minor element chemistries, and hardness.
Westinghouse is developing field techniques to thermally treat inner row U-bends.
The intent of the thermal treatment would be to reduce the suscep-tibility of the tubing to stress corrosion cracking by altering its grain structure and reducing the residual stresses from bending.
Additionally, Westinghouse is pursuing development of a shot peening process whereby the tube ID is blasted with small hard glass or carbon beads at high velocity to induce compressive rather than tensile stresses on the interior surface.
Westinghouse hopes to complete these development efforts by December 1982.
Eddy current techniques are also under development by the industry which may improve upon existing capabilities to address the U-bend cracking problem.
As an example, the Tennessee Valley Authority (TVA) has recently developed an eddy current technique which it believes capable of detecting tubes with
" opposite side" transition geometries that render the tubes susceptible to stress corrosion cracking.
This technique has been employed on an experimental basis at Sequoyah Unit 2, which has not yet begun commercial operation.
3.2.4 Recent Corrosion Problems at the Tubesheet and Sludge Pile Locations Corrosion of the steam generator tubes in the crevice between the tubes and the tubesheets was first identified in 1977 at Point Beach Unit 1.
In many early-generation Westinghouse (W) steam generators, the tubes were not expanded the full depth of the 24-inch-thick tubesheet.
Although the exact mechanism of the corrosion has not been identified, the tube-to-tubesheet crevice does pro-i vide a site for concentrating an aggressive environment which can lead to intergranular attack (IGA) and to eventual SCC of the Inconel tubing.
This phenomenon can affect units operating with either phosphate or AVT secondary water chemistry control.
Tubesheet crevice corrosion has occurred in at least
-seven of the 17 W plants where the tubes were not expanded the full depth of the tubesheet. Tube sheet crevice corrosion has affected an extensive number of tubes at both Point Beach Unit 1 and H.B. Robinson Unit 2.
i 14 I
i IGA at San Onofre Unit 1 and Point Beach Unit 2 occurred at or just slightly above the top of the tubesheet where sludge accumulated.
The IGA has generally been observed to occur in a 1/4-inch wide band around the tube circumference.
The degradation at San Onofre was quite extensive, necessitating sleeving and plugging repairs of approximately 7000 tubes.
San Onofre Unit 1 is one of two domestic plants still operating with phosphate secondary water chemistry control, although its operators have experienced difficulty in maintainirl the specified sodium-to phosphate ratios and in preventing condenser inleakage in recent years.
The IGA corrosion at Point Beach Unit 2 is in its early stages.
Point Beach Unit 2 r.onverted from phosphate to AVT secondary water chemistry control in 1974.
3.2.5 Recent Pitting and Localized Wall Thinning Problems Minor pitting (i.e., an occasional isolated pit) has occurred on some tubes which were removed from service earlier than 1981, but not to the extent that they were detectable by eddy current testing or that they constituted a concern for primary to secondary leakage.
A new pitting phenomenon has recently been observed at Indian Point Unit 3 where in excess of 1000 tubes were found to be affected during the September 1981 outage.
Indian Point Unit 3 is the first plant where significant pitting (readily detectable by ECT) has been identified.
In this case the detectable pitting is confined to the cold leg side of the tube bundle and concentrated within 6 to 20 in. above the tube-sheet, with decreasing degradation up to 36 in above the tubesheet.
The pitting was detected by ECT against background signals similar to those observed in the laboratory tubes containing surface copper deposits.
Because identifi-cation of this problem is relatively recent, the causes are still under investigation.
However, the unit has been subject to continuous condenser inleakage, and an examination of sludge has shown that it contains a high level of copper oxide, which is indicative of severe oxygen ingress through the condenser.
Localized wall thinning (or large pits) has been observed since 1979 at Prairie Island Unit 2, affecting in excess of 100 tuSIs at the periphery of the cold leg at the first and second tube support plates.
Localized thinning of the tube wall at the antivibration bar (AVB) supports has also been observed at this unit.
These difficulties are believed to be corrosion and possibly fretting related, although the exact cause has not yet been established.
Resin bleedthrough from the condensate polishers has occurred at this facility l
and may provide an explanation of the source of contaminants in the secondary water.
H.B. Robinson Unit 2, which continues to operate with phosphate secordary water chemistry control, has also experienced local wall thinning in the U-bends, which possibly is phosphate wastage related.
l l
3.2.6 Fretting Problems In the mid 1970s, tubes in early generation Westinghouse steam generators at j
San Onofre Unit 1 and Haddam Neck experienced fretting (wear) degradation l
at the anti-vibration bar (AVB) supports located in the U-bend region.
This problem was corrected by the installation of additional AVBs of a revised i
design.
The revised AVB design employs chromium plated Inconel bars with a j
square cross section that increases the area of contact and reduces the clear-ances between the bars and the tubes.
This type of AVB support has been incorporated into later generation Westinghouse steam generators, and no AVB-related fretting problems have been reported in recent years, with the possible 15
exception of Prairie Island Unit 2.
The extensive tube thinning at the AVB intersections observed in February 1981 at Prairie Island Unit 2 may be a fretting rather than corrosion-related phenomenon.
Fretting degradation as a result of a foreign object in the steam generator caused a gross tube rupture at Prairie Island Unit 1 on October 2, 1979.
The rupture caused a primary-to-secondary leak of 390 gallons per niinute.
The reactor was brought to a cold shutdown in a routine manner following the emergency procedures for such an event.
Subsequent inspection revealed that the tube rupture was caused by mechanical wear on the tube by a foreign object which eventually led to a pressure burst.
The foreign object was later iden-tified as a spring, jammed by the flow-blocking device.
It is believed that the spring came from sludge removal equipment that was inadvertently left in
- m steam generator during a previous outage.
Recent Tube Failures of Westinghouse Preheat Type Steam Generators Ringhals Unit 3, a three-loop Westinghouse plant in Sweden, was shut down on October 21, 1981 because of a 2.6 gpm primary-to-secondary leak.
Before the leak, the unit had been operating at power levels greater than 50% for approx-imately five months.
The steam generators, W preheat type (Figure 2), are similar in design (Model D) to those at McGuire Unit 1, the only domestic operating plant with this type of steam generator.
The leaking tube was located within the preheater section on the cold leg side of the steam generator.
The ECT results revealed numerous tubes with ECT indications localized within the preheater section at baffle plate loca-tions.
The tubes affected are in the peripheral rows (close to the steam generator shell) adjacent to the feedwater inlet.
There are approximately 100 tubes with ECT indications for each steam generator.
Approximately 45 of the tubes with ECT indications have wall reductions of greater than 50%.
The most recent eddy current testing of the steam generator tubes at Almaraz Unit 1 in Spain also revealed significant tube wall reduction at locations similar to those at Ringhals Unit 3.
Almaraz Unit 1, with steam generators similar to those at Ringhals Unit 3 and McGuire Unit 1, had been operating at various power levels, including full power, for about four months.
Westinghouse believes these ECT indications are attributable to excitation of the steam generator tubes from high fluid velocities and that the tube walls are being worn down from vibrational rubbing against baffle plates in the pre-heater sections of these steam generators. Westinghouse further believes that a reduction of flow velocity by controlling total feedwater flow should reduce the potential for vibration.
Duke Power Company completed eddy current testing of approximately 170 tubes in steam generator A at McGuire Unit 1 on November 19, 1981 to determine if similar problems are being experienced in the McGuire steam generators.
Pre-liminary findings show tube wall degradation no greater than 10%.
The NRC staff is closely following this problem and will tal:e appropriate action to ensure continued safe operation of this plant.
16
a p -
-u s_
2 2
l l
l 3.2.7 Operating Problems with B&W Once-Through Steam Generators Most of the leaks in B&W once-through steam generators have occurred in tubes adjacent to the inspection lane.
This 16ne consists of the area created when a row of tubes halfway across the tube bundle was omitted to facilitate inspec-tion and chemical cleaning of the tube bundle.
These leaks have occurred in the uppermost span at the intersection of the tube and the upper tubesheet or the intersection of the tube and the 15th support plate.
After fiber-optic inspections, through-wall circumferential cracks were reported as the source of the leakage.
Examination of tube specimens removed from the generator indi-cates fatigue, believed to result from flow-induced vibration, was the crack propagating mechanism.
In at least two instances (0conee Unit 3 in 1980 and Rancho Seco in 1981), the leaks have been observed and fiber optic inspection revealed a 360 crack around the tube circumference.
B&W believes that the full circumferential failures have occurred during plant cooldown when the tubes are subject to tensile differential thermal loadings.
The initiating mechanism for the circumferential fatigue cracks is believed to i
be a combination of surface damage from corrosion and normal tube loadings.
The open inspection lane provides a direct path for entrained corrosion pro-ducts and concentrated chemical agents carried by moisture during adverse secondary system conditions.
B&W is currently undertaking the design and qualification of a flow-blocking device in the inspection lane which can be attached to the tube support plate i
at several elevations.
The purpose of this device is to alleviate corrosive attacks on tubes adjacent to the inspection lane in the upper span by forcing the steam and water mixture out of the open lane and into the heated bundle, where it will be evaporated.
A tube degradation phenomenon which appears to be increasingly prevalent in B&W steam generators is localized wall thinning and is believed to be an impingement or erosion phenomenon.
This phenomenon has been observed at sup-port plater, particularly the 14th support plate, and has caused at least three leaks at Oconee Unit 1.
This phenomenon appears to be associated with debris found on the support plates and lower tubesheet.
The debris deposits l
also provide a medium for the concentration of adverse chemicals which can lead to corrosion of the tubing.
Samples removed from the field indicate that the i
debris is predominantly iron oxide with traces of other elements, although its origin has not been identified.
B&W and the affected utilities are evaluating chemical cleaning as a method for removing the debris, and thus reducing the potential for further tube degradation.
Chemical cleaning is discussed in Section 3.4(7).
i 3.2.8 Summary of PWR Steam Generator Operating Experience Tables 1, 2 and 3 indicate which PWR units designed by W, CE, anc B&W, respec-tively, are affected by the forms of tube degradation discussed in this chapter.
l These tables indicate those facilities which either have or are considering steam generator replacement.
Table 4 summarizes, to the best of our knowledge, those foreign plants that have experienced the kinds of tube degradation discussed.
A more detailed account of the early history of adverse steam generator tube i
17 i
i
Table 1.
Operating experience with Westinghouse PWR steam generators through November 1981 Previous Other corr.
induced wall i
model issuance water chem phosphate thinning or Plant name no.
date control control Wastage pitting Fretting l
Yankee-Rowe (a)(b) 12/63 AVT Yes San Onofre 1 27 3/67 Na/PO Yes X
X 4
Haddam Neck 27 12/74 AVT Yes X
X Ginna 1 44 9/69 AVT Yes X
y H.B. Robinson 2 44 9/70 Na/PO Yes X
4 Point Beach 1 44 10/70 AVT Yes X
Point Beach 2 44 11/71 AVT Yes X
Turkey Point 3 44 7/72 AVT Yes X
Indian Point 2 44 9/73 AVT Yes X
Surry 1 pre (c) 51 5/72 AVT Yes X
4 Surry 1 post (d) 51 7/81*
AVT No Surry 2 pre (c) 51 1/73 AVT Yes X
Surry 2 post (d) 51 9/80*
AVT No Turkey Point 4 44 4/73 AVT Yes X
4 Zion 1 51 4/73 AVT Yes Prairie Island 1 51 8/73 AVT Yes X
Kewaunee 51 12/73 AVT Yes Zion 2 51 11/73 AVT Yes
.'rairie Island 2 51 10/74 AVT No X
Cook 1 51 10/74 AVT No i
j Trojan 51 11/75 AVT No Indian Point 3 44 12/75 AVT No X
82 aver Valley 1 51 1/76 AVT No Salem 1 51 8/76 AVT No X
Farley 1 51 6/77 AVT No North Anna 1 51 11/77 AVT No Cook 2 51 12/77 AVT No North Anna 2 51 8/80 AVT No Ssquoyah 1 51 9/80 AVT No Sales 2 51 4/80 AVT No McGuire 1 D2 7/81 AVT No Farley 2 51 10/80 AVT No Stquoyah 2 51 9/81 AVT No Ditblo Canyon 1 51 9/81**
AVT No (a) No model number.
(b) Yankee Rowe employs 304 SS tubing. All other PWRs employ Inconel-600 tubing.
(c) Original steam generators.
(d) Replacement steam generators.
(e) Definition of extensive vs. moderate or minor denting is provided in Table 2.
- Startup date.
- License suspended 11/81.
18
l Table 1 (Cont.)
Primary side Secondary init. SCC in No. of No. (%)
Steam sida init.
small radius leaking of tubes Sleeve generator IGA / SCC U-bends Denting (5) tubes plugged repairs replacement X
116(1.8%)
X X-Extensive 31 948(8.4%)
X-6508 tubes X
X-Minor 4
69(0.5%)
X X-Minor 6
237(4%)
X-21 tubes i
X X-Minor 1068(11%)
Under consideration l
X X-Moderate 51 827(13%)
X-12 tubes X
X-Moderate 4
117(2%)
Planned X-Extensive 14 (21%)
In progress X-Extensive 5
477(3.7%)
i j
X-Extensive 40 1593(25.4%)
X-Completed i
X X-Extensive 27 2187(21.5%)
X-Completed X-Extensive 36 (24.8%)
Planned X
X-Minor 1
27 l
X 2
34(1%)
l X-Minor 0
0 i
X-Minor 1
15 X
1 61(<2%)
0 21(<1%)
X 12 368 X-Extensive 1
801(6.1%)
Planned 1
i 0
0 l
0 30(0.9%)
l X
8 8
X X-Minor 1
284(2.8%)
I X
3 24(<1%)
0 282(2.8%)
X-Minor 0
0 5
5 i
X I
19 l
l
a Table 2.
Operating experience with Combustion Engineering PWR steam generators through November 1981 Previous Other corrosion SG OL Secondary experience induced wall Secondary No. of No. (%)
model Issuance water chem.
w/PO thinning or side init.
leaking of tubes Sleeve Plant name no.
date control contfol Wastage pitting SCC Denting (b) tubes plugged repairs Palisades (a) 3/71 AVT Yes X
X X
X-Moderate 3758(22%)
X-33 tubes Maine Yankee 9/72 AVT No X-Minor 0
0 Ft. Calhoun 5/73 AVT No 0
3(<1%)(C)
.Calvert Cliffs 1 8/74 AVT No X-Minor 0
0 Millstone 2 8/75 AVT No X-Extensive 0 800(0.3%)
St. Lucie 1 3/76 AVT No X
X-Minor 1
76(1%)
l$
Calvert Cliffs 2 8/76 AVT No X-Minor 0
0 Arkansas 2 9/78 AVT No X-Minor 0
58(0.3%)
-(a) CE steam generators do not have specific model numbers. For the plants listed above, the steam generators are of the same basic design with the exception of Palisades. Palisades employs drilled hole support plates for the lower six tube supports instead of egg crate supports.
'(b) Denting is described as extensive, moderate, or minor as follows:
extensive denting - (a) presence of tube denting that is widespread throughout whole steam generator in which the average total reduction in tube diameter equals to or exceeds twice the tube wall thickness; (b) measurable support plate in plane deformations, such as hourglassing of flow slots in Westinghouse plants; (c) damage has caused leaking from dents.
moderate denting - (a) presence of tube denting that is widespread throughout whole steam generator in which the average total reduction in tube diameter exceeds 20% of the tube wall thickness; (b) no measurable support plate in plane deformation; (c) damage has not caused leaking from dents.
minor denting - (a) presence of tube denting is spotty to widespread, but the average total reduction in tube diameter is less than 20% of the tube wall thickness; (b) no visible support plate deformation; (c) damage has not caused leaking from dents.
(c) The nature of the tube degradation was not known.
i
Table 3.
Operating experience with Babcock and Wilcox once-through steam generators through November 1981 OL No. of No. (%)
Issuance Fatigue Erosion /
leaking of tubes Sleeves Plant name(a)'(b) date cracking Corrosion tubes plugged installed Oconee 1 2/73 X
X 11 311 (2%)
16 Oconee 2 10/73 X
X 3
30 (<1%)
Oconee 3 7/74 X
X 5
101 (<1%)
Arkansas 1 5/74 X
3 13 (<1%)
Rancho Seco 1 8/74 X
X 1
15 (<1%)
Three Mile Island 1 4/74 X
0(c) 19 (<1%)
Crystal River 3 12/76 X
0 32 (<1%)
Davis Besse 1 4/77 X
2 13 (<1%)
Three Mile Island 2 2/78 (NRC suspended authority to operate) 38 (<1%)*
(a) B&W steam generators do not have specific model numbers, but are of the same basic design.
(b) B&W plants have been operated exclusively with AVT secondary water chemistry control.
(c) Does not include recent TMI-1 steam generator problems currently under evaluation.
- Attributed to manufacturing defects.
i l
l l
21
t 4
Table 4.
Foreign operating experience with PWR steam generators (a)
Previous Wastage Secondary operation or other SCC / IGA SCC initiated SG model Tubing Start-up water ches.
with PO wall initiated from ID IN 4
NSSS Plant name (country) no.
material date control control thinning from OD U-bends Fretting Denting W
Seini NI NI 6/64 AVT No X(b)
A SENA (France) 14 SS 4/67 ATV No X(b)
X K
Obrigheim (W. Germany)
NI Inco 600 3/69 AVT NI X
X X
Yes X
X(b)
W Zorita (Spain) 24 Inco 600 8/(9 PO4 U
Beznau 1 (Switzerland) 33 Inco 600 12/69 AVT Yes X
X X
Y'S U/C Mihama 1 (Japan)
NI Inco 600 11/70 PO4 W
Beznau 2 (Switzerland) 33 Inco 600 3/72 AVT Yes X
X X
yK Stade (W. Germany)
NI Alloy 800 5/72 PO4 W
Nihama 2 (Japan)
N1 Irco 600 7/72 AVT Yes X
X K
Borssele (Netherlands)
NI Alloy 8,00 10/73 PO4 Yes X
d Takahama (Japan) 51 Inco 600 11/74 AVT Yes X
X X
X(b)
W Doel 1 (Belgium) 44 "I
K Biblis A (W. Germany)
NI Incoloy 3/75 PO4 W
Ringhals 2 (Sweden) 51 Inco 600 5/75 AVT Yes X
X U
Tihange 1 (Belgium)
NI Inco 600 9/75 AVT No X
X X
i W
Genkal (Japan)
NI Inco 600 10/75 AVT No X
U Doel 2 (Belgium) 44 W
Takahama 2 (Japan) 51 Inco 600 11/75 AVT No i
W Mihama 3 (Japan)
NI Inco 600 12/75 AVT No See footnotes on last page of table.
l
t Table 4.
(continued)
Previous Wastage Secondary operation or other SCC / IGA SCC initiated SG model Tubing Start up water chem.
with PO wall initiated from ID IN 4
NSSS Plant name (country) no.
material date control control thinning from OD U-bends Fretting Denting 5
Biblis (W. Germany)
NI Incoloy 12/76 NI NI X
j MHI Ikata (Japan)
NI Inco 600 9/77 NI NI i
F/C Fessenheim 1 (France) 51A Inco 600 12/77 AVT No
]
F/C Fessenheim 2 (France) 51A Inco 600 3/78 AVT No W
Ko-Ri I (Korea) 51 Inco 600 6/77 NI NI X
F/C Bugey 2 (France) 51A Inco 600 2/79 AVT No F/C Bugey 3 (France) 51A Inco 600 2/79 AVT No X(b)
O OHI OHI (Japan)
NI NI 3/79 NI NI H/M Bugey 4 (France) 51A Inco 600 5/79 AVT No K
Esenham (W. Germany)
NI NI 10/79 NI NI F/C Bugey 5 (France) 51A Inco 600 11/79 AVT No F/C Gravelines B1 (France) 51M Inco 600 7/80 AVT No W
Tricastin 1 (France)
SIM Inco 600 7/80 AVT No W
Almaraz 1 (Spain) 03 Inco 600 12/79 AVT No X(c)
F/C Danpferre 1 (France) 51M Inco 600 10/80 AVT No W
Almaraz 2 (Spain) 03 Inco 600 NI NI -
NI W
Ringhals 3 (Sweden).
D3 Inco 600 NI NI NI X(c) i (a) This summary is based upon information available to the NRC staff and may be incomplete.
Key: W - Westinghouse
~
(b) Fretting at anti vibration bar supports.
(c) Fretting in preheater section.
~
-H I
M - Mitsubishi Heavy Industries, Ltd.
I i
a E
g
+
operating experience is provided in NUREG-0523, " Summary of Operating Experience With Recirculating Steam Generators," and NUREG-0571, " Summary of Tube Integrity Operating Experience With Once-Through Steam Generators."
3.3 Recent Plant-Specific Problems This section provides a summary of plant-specific problems which have been experienced since November 1980.
Plants not listed below have not reported any significant operational difficulties with their steam generators during this period.
3.3.1 Westinghouse Steam Generators D.C. Cook Unit 2 This unit was shutdown on October 2,1981 with a primary to secondary leakage rate of 0.29 gpm in the steam generators.
This leakage rate was less than the 0.35 gpm limit in the plant Technical Specifications.
Hydrostatic tests revealed two tubes, leaking in the innermost row (row 1) of the tube bundle at the tangent point location of the small-radius U-bends.
Eddy current testing revealed one additional row 1 tube with an ECT indication,in the U-bend.
All three tubes were removed from service by plugging *.
This unit previously experienced a small (0.01 gpm) leak in October 1980.
Subsequent eddy current examination revealed a row-1 U-bend ECT indication; however, a hydrostatic test could not confirm this as the source of the leak.
Farley Units 1 and 2 The licensee is proposing to plug all tubes in row 1 in both Units 1 and.2 in order to reduce the frequency of unscheduled outages to repair steam gen-erator leaks, thereby improving the power generation reliability of these units.
Unit 1 previously experienced small, U-bend leaks (below the plant Technical Specification limit) in row 1 tubes in March 1979 and December 1980.
An additional leak was revealed in row 1 tubing during a leak test in December 1980.
No leaks from row 1 tubing have been reported for Unit 2 to date.
Unit 2 did experience a 1.1 gpm leak in June 1981 (away from row 1) which may have been caused by a hacksaw blade which was dropped into the general area'while upper inspection ports were being installed in the steam generators.
4 R.E. Ginna Unit 1 This unit has experienced a moderate amount of intergranular attack and stress corrosion cracking in the tubesheet crevices.
Fourteen tubes with tubesheet crevice ECT indications were found during an inspection in May 1981 and were sleeved **.
To date, 21 tubes have been sleeved.
Indian Point Unit 2 Approximately 23 tubes were plugged as a result of extensive denting, which was discovered between December 1980 and April 1981 during an extensive out-rPlugging is defined in Section 4.2.
- Sleeving is defined in Section 5.3.
24
3, -
age.
Following startup from the outage, this unit developed a leak that remained well below the 0.3 gpm Technical Specification limit, whereupon the unit 3
was shut down on August 21, 1981.
The leak originated in a cold leg tube that had previously restricted passage of a 0.610-inch probe, but allowed passage of a 0.540-inch probe.
In addition to plugging the leak, the licensee decided to plug four additional tubes in the cold leg that had also restricted passage of 1 4 a 0.610-inch probe. ? On the cold leg side, only tubes that restricted passage of a 0.540-inch probe had been plugged during the previous outage.
^
Indian Point Unit 3 This unit was shut down on September 24, 1981, as the result of a 0.7 gpm leak, which exceeded the Technical Specification leakage limit of 0.3 gpm.
Eddy current inspection revealed 1607 tubes with quantifiable ECT indications,
. 1 including 1091 tubes with indications exceeding the 40% plugging limit.
An additional 558 tubes were found with nonquantifiable indications.
These indi-cations were exclusively on the cold leg side, with the vast majority accurring several inches above the tubesheet.
The remaining indications (including those from the leaking tube) occurred above the first support plate at the periphery of the bundle.
All tubes with indications also registered distorted signals similar to those which have been observed in the laboratory for tubes contain-ing copper surface deposits.
Four tubes were removed from the steam generator for further laboratory study.
Preliminary laboratory results indicate that all tubes had local regions of pittingc.and no generalized wall thinning.
The cause of the pitting is still under examination.
However, this unit has been subject to continuous con-denser in-leakage, and an examination of sludge has shown it to contain a high j
level of copper oxide, which is indicative of severe oxygen ingress through the condenser.
Because,of the large number of tubes affected, the licensee may eventually have to sleeve as many of these tubes as possible.
In the interim, the licensee has asked for staff approval to increase the plugging limit from 40% to 65%.
As a basis for this request, the licensee has submitted burst test data for tubes containing arrays of simulated pits, as well as additional data to demon-s
- . strate that pits of structural significance are detectable by eddy current tes t.i ng.
The staff approved the licensee's request for a period of four months (until the next refueling outage).
lb Although denting apparently is not a factor in this latest leak occurrence, this unit has also experienced extensive denting.
The licensee had been administering boric acid treatment prior to the September 1981 outage in an attempt to reduce the rate of denting.
(Boric acid treatments are discussed in Section 3.4(6).) Although the addition of boric acid is believed not to be a factor in the pitting, boric acid treatments have been terminated pending further evaluation of the causes.
Point Beach Unit I t
l This unit has experienced extensive intergranular attack and stress corrosion l
cracking in the tubesheet cruices, which has resulted in the need to plug several hundred tubes.
However, by flushing the tubesheet crevices and reduc-ing operating temperatures, the licensee has apparently been successful in reducing the rate of the tubesheet crevice corrosion since November 1979.
i 25
'T Point Beach has operated without significant leakage since January 1980.
Steam generator inspections performed in December 1980 and July 1981 continue to show a decrease in the occurrence of newly degraded tubes.
The operation of this unit continues to be subject to a portion of the operating restrictions imposed by Commission Orders of November 30, 1979, January 3, 1980, and April 4, 1980.
These Orders include more restrictive limits on primary-to-secondary.
leakage, hydrostatic testing of tube bundles during steam generator inspection outages, and additional reporting requirements concerning steam generator l-inspection results.
In addition, these Orders require NRC approval for restart in the event that the unit is shut down because of leakage in excess of limits in the Technical Specifications.
The licensee has asked for staff approval to install 12 sleeves during the planned October 1981 outage as a demonstration program.
The staff approved this request.
The licensee ultimately plans to install large numbers of sleeves at both Units 1 and 2.
P Point Beach Unit 2 During an outage in April 1981, a tube containing a 41% eddy current indication at the top of the tubesheet elevation was removed for laboratory analysis.
Subsequent examination revealed a relatively narrow zone of IGA extending part way around the tube circumference, similar to what had been observed previously at San Onofre Unit 1.
The maximum depth of IGA penetration was determined to be 32%, with some evidence of localized tube wall thinning, which correlates well with the field indication.
Eddy current inspections over tra past few years have shown a large number of tubes with ECT indications at or above the top of'the tubesheet.
Most of these indications are believed to be attributable to wastage rather than to IGA as a result of early operation with phosphate control of secondary water.
1 The vast majority of irdications found have been less than the plugging limit and have not been increasing at a rapid rate between inspections.
The licensee ultimately plans to perform large scale sleeving of tubes with or without identified defect indications.
Prairie Island Unit 2 Results of the February 1981 steam generator inspection showed further progression of thinning indications first observed in January 1980 at the peri-pheral region of the cold leg, primarily at the first and second support plate elevations.
The February 1981 inspection revealed 211 cold leg ECT indications, of which 18 were pluggable.
4 This inspection also revealed for the first time extensive tube thinning at the AVBs affecting 25 tubes.
Of these, 23 tubes contained pluggable l
ECT indications, which ranged to a maximum penetration of 80%.
Most of these. tubes had not been inspected in the U-bend region since the initial base-line inspection.
However, of nine AVB intersections which were inspected in 1976 and 1977, and for which multi-frequency ECT now shows indications, six had previously exhibited distorted signals.
Fiber optic inspection of the AVB intersections suggested that the tube thinning may be a corrosion rather than a wear-related phenomenon.
The cause of the cold leg and U-bend indications has not yet been established.
j
'26 e
r w-y w -<
H.B. Robinson Unit 2 l-This unit was shut down as a result of a 0.3 gpm leak on July 30, 1981.
i Inspection of the leaking tube revealed that a through-wall stress corrosion crack above the top of the tubesheet elevation was the source of the leak.
In addition, evidence of general intergranular attack was observed below the top of the tubesheet in the crevice region.
The crack above the tubesheet had an axial orientation and was approximately 0.8 in. long.
The low leakage rate has been attributed to the restraining effect of the hard sludge on the tube, a phenomenon similar to one that has been observed previously at San Onofre Unit 1.
Both of these units operate with phosphate control of secondary water.
Eddy current inspection revealed a total of 212 tubes with pluggable indications (that is, of penetrations greater than 47%) above and below the top of the tubesheet elevation.
The distribution of the number of defects as a percentage of wall penetration indicated that the IGA and stress corrosion cracks were not being detected until after they had penetrated beyond 50%.
To provide additional assurance that the unit can be operated safely until its next steam generator inspection, license conditions have been imposed requir-ing periodic primary-to-secondary hydrotests during the current cycle and more restrictive limits on primary-to-secondary leakage.
The licensee has taken a number of steps to reduce the rate of corrosion, including sludge lancing, secondary side flushing prior to startup, and reduced operating temperatures.
In addition to its IGA and stress corrosion cracking problem, Unit 2 is experiencing active corrosion-induced wall thinning above the tubesheet on both the hot and cold leg sides and in the U-bends.
A total of 120 tubes were i
plugged in March and April 1980, 182 tubes in May 1981, and 213 tubes in August 1981.
To date, 1068 tubes (10.9% of the total number of tubes) have been plugged.
San Onofre Unit 1 Sleeving repairs of approximately 7000 steam generator tubes have been completed.
The unit has been approved for six months' operation following the start of Cycle 8, after which it must be shut down for its next steam generator inspection.
Subsequent restart will be subject to NRC approval of the inspection results and needed repairs.
Extensive repairs of the steam generators became necessary as the result of l
wide spread IGA at the top of the tubesheet.
Hot and cold leg flushing of the l
steam generator secondary sides, stricter controls of the secondary water chemistry, and reduced operating temperatures have been implemented to retard the. rate of further tube degradation.
l Surry Units 1 and 2 and Turkey Point Units 3 and 4 Extensive denting-related degradation of tubing at Surry Units 1 and 2 and Turkey Point Units 3 and 4 has necessitated the replacement of the steam gen-
~
erators at these facilities.
Steam generator replacement at Surry Units 1 and i
2 has been completed.
Hearings by the Atomic Safety and Licensing Board regard-I ing steam generator replacement at Turkey Point Units 3 and 4 have been completed, and replacement is currently underway at Unit 3.
Replacement of the Turkey Point 4 steam generators is scheduled for Fall 1982.
l
?
27
Surry Unit 2 has been operating at full power for approximately 95% of the time since the steam generator was replaced in August 1980.
There have been very few plant trips or unscheduled shutdowns.
Scheduled shutdowns have been pri-marily for snubber inspec,tions.
Steam generator condensate chemistry has been excellent.
Surry Unit 1 has operated well since the steam generator replacement outage in July 1981.
Steam generator operating experience has been similar to that at Surry Unit 2.
Unit I has recently experienced some fuel leaks (indicated by high reactor coolant system iodine activities), but this problem is believed to be unrelated to the steam generator replacement.
There has been no tube leakage in either unit since the replacement.
The Unit 2 tubes and tubesheet area will undergo inservice inspection during the outage in November / December 1981.
Trojan Unit 1 The unit was shut down on January 30, 1981 because of a small leak (approximately 270 gallons per day).
One leaking tube was in row 1 of steam generator A, and three leaking tubes were in row 1 of steam generator D.
All tubes leaked in the U-bend area.
In view of Trojan's history of frequent leaks of this type, the licensee elected to plug all remaining tubes in row 1 in these steam generators as a preventive measure.
The licensee also plugged the remaining row-1 tubes in steam generators B and C during the May 1981 refueling shutdown.
^
Zion Unit 1 Eddy current indications in the U-bends of tubes in both row I and row 2 were identified at Zion Unit 1 during steam generator inspections performed in Jan-uary 1981.
A total of 16 indications were found in row 2, all in one steam generator.
Because these indications occurred in the U-bend, they could not be quantified with any degree of certainty, and subsequently all 16 tubes were plugged. With one exception, all indications occurred at the tubes' tangent point elevation.
The indication for the remaining tube was located near the apex of the U-bend.
In addition, one of the 16 tubes plugged contained indica-tions at both tangent points.
Assuming that the indications found in row 2 identify cracks (rather than manufacturing defects, for example), this is the first known occurrence of U-bend cracking in tubes beyond row 1.
No indications were found in tubes in row 1 of this steam generator, but two indications in tubes in row 1 were identified at the U-bend tangent points of another steam generator.
The unit had a small (47 gallons per day) primary-to-secondary leak at the time it was shut down; i
however, the source of the leak could not be identified.
3.3.2 Combustion Engineering Steam Generators i
Millstone Unit 2 This unit has experienced a moderate amount of denting, oarticularly at the drilled support plates.
Tube inspections performed during the August 1980 outage indicate that the denting has apparently been stabilized as a result of corrective actions taken in 1977.
These included retubing the condensers with 90-10 cupro nickel (CuNi), installation of a full flow condensate polishing system, elimination of hardspot areas in the drilled support plates, and improved secondary water chemistry control.
During the August 1980 outage only one 28
I tube was plugged.
The total number of tubes plugged to date is 800, or approxi-mately 5% of the total number of tubes.
During the August 1980 outage, indications of dents were detected at a large number of egg crate support locations.
Egg crates were initially thought to be relatively immune to denting because of their excellent flow characteristics.
ECT indications of denting tend to be inaccurate at egg crate supports.
Denting at egg crate supports causes the tubes to become oval-shaped (rather than pro-ducing sharp dents such as occur in drilled-type support plates), which is difficult to detect with conventional ECT techniques.
To define the exact extent of denting, a device with multiple feeler gauges (a profilometer) was used.
The profilometer showed tube ovalities of 30-40 mils, whereas ECT indi-cated dents of 1-2 mils.
For even this, amount of denting, however, the corresponding tensile strain on the ID is low compared with strains necessary to produce stress corrosion cracking.
Therefore, the staff does not antici-pate primary side stress corrosion cracking (PSCC) as a consequence of egg crate denting.
Recently, Millstone Unit 2 has detected trace (3 500 milliliter / day) primary-to-secondary leakage.
A shutdown and ECT examination are scheduled for December 1981.
Palisades Eddy current inspections performed during the September 1981 outage resulted in the plugging of 49 tubes as a result of wastage degradation.
The number of tubes plugged is attributed primarily to eddy current data scatter and the fact that since the early 1970s, a large number of tubes at this unit have been de-graded in excess of 20% of the wall thickness.
Plugging limits implemented during the September 1981 outage were 55% in steam generator A and 50% in steam generator B in accordance with the plant Technical Specifications.
Comparative analyses of the eddy current data with the corresponding data from previous inspections indicates that phosphate wastage has been halted since phosphate chemistry was eliminated in 1974.
Additionally, the minor denting detected in 1975 has shown virtually no growth during the last three inspec-tions.
These improvements are attributed to the new condenser which was installed in 1975 and to a vigorous water chemistry control program with tight administrative controls for operation with condenser leakage.
St. Lucie Uqit 1 Prior to a refueling shutdown in November 1981, this unit had operated for I
approximately seven months with a trace primary to secondary leak of 200 to 500 milliliter / day.
At shutdown, the leaking tube, located in the U-bend region, was plugged.
During the recent eddy current testing of steam gener-ator tubes at St. Lucie Unit 1, Florida Power & Light Company found a significant number of defective tubes with wall thickness reduced by more than the 40% plant Technical Specifications plugging limit.
Initial ECT of tubas in steam generator B showed 50 tubes having indications on the outside surface, i
with indications greater than 40% in 39 tubes.
These tubes are in the U-bend radii of rows 8 to 12, with one in row 8.
ECT of tubes in steam generator A showed 36 having indications on the outside surface, with indications greater than 40% in 22 tubes.
These indication:. are for tubes located primarily in 29
rows 8 to 12.
The indications appear at the apex of the U-bends, at the intrados surface, and based on ECT, are suspected to be either cracking or continuous lines of pits.
The cause and significance of these indications are still being investigated by the licensee.
3.3.3 Babcock & Wilcox Steam Generators Arkansas Unit 1 Since Spring 1978, the full power steady-state operation level of steam generator A, as measured by the differential pressure of the steam gener-ator, has been steadily increasing while the full power operation level of steam generator B has remained constant.
The licensee believes that the steady increase in the operating level of the steam generator A is caused by increased flow resistance in the tube bundle as the result of a buildup of debris at the orifices of the tube support plates.
Further, the licensee believes that the buildup of debris in steam generator A is caused by contaminants from the turbine moisture separators drain system, the condensate polisher, and the drainage from the feedwater heater drain tank.
During the last refueling outage (January 1981) the licensee made several plant modifications which were expected to reduce the rate of operation-level increases.
The operation level has continued to increase to the point where now the licensee has reduced steady state power to 90% in order to reduce feedwater to keep the operating level of the steam generator within limits.
Long range plans for corrective action include chemical cleaning to remove debris.
Interim actions being taken include high temperature soaking and thermal shock during shutdown to break up the debris deposits.
The licensee will continue to reduce steady state power as necessary to maintain the steam generator water level.
The staff is keeping abreast of this condition, which appears to be unique to Arkansas Unit 1.
Tubes adjacent to the open inspection lane at the 15th support plate and upper tube sheet, respectively, leaked in July and September 1980.
Tubes removed and examined from other units indicate that such failures can be initiated by adverse chemicals concentrating under the deposits and attacking the tubes.
rack propagation by fatigue may also have been involved in the Arkansas Unit 1 leaks.
Oconee Unit 1 The unit was shut down in February 1981 as a result of a small (0.25 gpm maximum) primary to secondary leak in steam generator B.
Subsequent investigation revealed that the leaking tube was located in the second row of tubes beyond the open inspection lane.
Circumferential fatigue cracking is believed to be the most likely cause.
Steam generator inspections performed during the refueling outage in the fall of 1981 have revealed 41 tubes with pluggable indications, plus several hundred tubes with indications of less than the plugging limit.
The bulk of these indications occurred at the 14th support plate and are attributed to liquid impingement erosion.
c 30
l Oconee Unit 2 This unit was shut down on September 18, 1981 after a rapid increase in leakage from 0.35 gpm to 30 gpm.
This is the largest leak ever reported at a B&W unit.
The leak occurred in a tube at the 15th support plate elevation of the inspection lane.
The failure mechanism is believed to have been a circumferential crack propagated by fatigue.
A tube adjacent to the inspection lane exhibited a 30% ECT indication at the same elevation.
Fiber optic inspections to better characterize the nature of the tube damage were not performed because of the potential risk of radiation exposure to workers.
Oconee Unit 3 Steam generator inspections performed during a refueling outage in December 1980 revealed in excess of 300 ECT indications, the bulk of which occurred in the peripheral region of the tube bundle at various tube support locations, and are believed caused by erosion and corrosion.
Four tubes were plugged during the outage.
Rancho Seco Unit 1 This unit was shut down on May 17, 1Q81 after a rapid increase in leakage to 10 gpm.
The leak occurred in a tube located adjacent to the open inspection lane.
Three additional tubes adjacent to the lane were found with ECT indica-tions and were plugged.
Fiber optic inspection of the leaking tube revealed a 360 circumferential crack at the 15th support plate elevation.
The crack propagation mechanism is believed to be fatigue as a result of flow induced vibration.
3.4 Corrective Actions for Operating Plants The following corrective actions can be instituted in operating plants without physical modifications to the steam generator:
(1) Stop cooling water ingress through the main condenser:
These actions i
include improved leak detection and repair techniques, ECT of condenser tubes, replacement of condenser tubes with more corrosion-resistant materials, or, ultimately, total condenser replacement.
(2) Reduce air (oxygen) inleakage to the condenser:
By utilizing better detection and repair techniques.
(3) Eliminate copper heat transfer alloys in the secondary system:
In the presence of condenser leakage, copper is a major contributor in denting and pitting corrosion.
Its elimination from condensers, feedwater heaters-and moisture separators will significantly reduce corrosion potential.
(4) Flush steam generators with hydrazine treated water:
This flushing is performed at some plants prior to startup, after shutdown, and after severe condenser leaks to remove soluble impurities.
(5) Lance tubesheet sludge:
This method is used to remove tubesheet sludge l
by breaking it up and putting it into suspension with high pressure water l
jets.
Westinghouse is considering using chemical cleaning additives in 31
conjunction with lancing to clean tubesheet crevices in the 17 plants where they occur.
(6) Add boric acid or calcium hydroxide:
Boric acid has been added to the steam generators of four W units to reduce the rate of denting.
Denting is reduced by boric acid additions following a boric acid soak.
The mechanism, while not fully understood, is believed to involve formation of an iron borate complex which acts as a protective film and inhibits further corrosion.
Calcium hydroxide is being tested in model boilers and has the potential to neutralize the acidity which causes denting.
The use of calcium hydroxide may be tested within a year at one or two operating plants.
(7) Clean steam generators chemically:
This method of corrosion control is being evaluated for several potential applications:
(a) Pre-operational and periodic cleaning to keep crevices clean and reduce corrosion potential.
This method has already been employed pre-operationally by TVA.
(b) Sludge or deposit cleaning to remove these harmful products from plants which have little or no denting.
This method has yet to be used but development should be completed in less than one year.
(c) Dented crevice cleaning to remove the magnetite causing tubes to dent is being tested in models.
Model boiler development testing has been in progress for five years; however, this process involves the risk of severe corrosion from the chemical additives themselves and will not be feasible for several n ars.
(8) Reduce operating temperatures: Operating at reduced hot leg temperatures has been implemented at three Westinghouse plants to reduce the rates of IGA and CSCC, phenomena known to be strongly temperature dependent.
In November 1979, Point Beach Unit I reduced its operating hot leg temper-ature from approximately 600 F to 560 F following the discovery of extensive IGA and CSCC in the tubesheet crevices.
IGA and CSCC have not been observed to date on the cold leg side, which at Point Beach is normally operated at approximately 540 F at 100% power.
To reduce hot leg temper-atures, the licensee had to cut back to 80% of full power operation.
Inspections performed af ter November 1979 indicate that the rate of fur-ther IGA and CSCC appears to have been substantially reduced.
(Another corrective action performed in November 1979 involved hot and cold water flushing of the tubesheet crevices and sludge.) San Onofre Unit 1 and l
H.B. Robinson Unit 2 have also recently begun operating at reduced power and temperature rates.
H.B. Robinson is currently operating at 50% of full power and with a hot leg temperature of 576 F.
4.
STEAM GENERATOR TUBE SURVEILLANCE AND REPAIR l
4.1 Inservice Inspection The steam generator tubes form the boundary between the primary and secondary coolant systems in PWRs.
Periodic inservice inspections of these tubes are 32
essential to monitor their integrity for safe operation.
The primary safety consideration for degraded tubes is that they retain adequate structural integrity without excessive leakage for the full range of normal and postulated accident loadings.
At present, the Technical Specifications for nuclear power plants require that inservice inspections be performed every 12 to 40 months, depending on the con-dition of the steam generators.
In cases where the degradation processes are highly active, NRC has required that the inspections be performed at even more frequent intervals.
Eddy current testing (ECT) is the primary means for performing tube inspections.
This inspection method involves the insertion of a test coil inside the tube that traverses its length.
The test coil is then excited by alternating current, which creates a magnetic field that induces eddy currents in the tube wall.
Disturbances of the eddy currents caused by flaws in the tabe wall will produce corresponding changes in the electrical impedance as seen at the test coil terminals.
Instruments are used to translate these changes in test coil imped-ance into output voltages which can be monitored by the test operator.
The depth of the flaw can be determined by the observed phase angle response.
The test equipment is calibrated using tube specimens containing artificially induced flaws of known depth.
Geometric discontinuities along the tube length, such as the tubesheet, tube support plates, and dents, also produce eddy current signals, which makes dis-criminating defect signals at these locations difficult.
The recent development of multifrequency eddy current techniques (whereby the test coil is excited at multiple frequencies rather than at a single frequency) has substantially enhanced operator capabilities to detect relatively small-volume flaws in the presence of extraneous signals.
Very small volume flaws, such as those caused by intergranular attack, stress corrosion, fatigue cracks, and small pits, have traditionally been hard to detect with the single-frequency eddy current test method.
The use of multi-frequency techniques and specialized nonstandard probes has improved detection capabilities in this regard.
However, further improvements are necessary and are the subject of much ongoing effort by the nuclear industry and through NRC-sponsored research programs.
For the present, the staff concludes that small flaws of structural significance are generally detectable.
If such flaws go undetected and result in leaks, the initial leakage will generally be small and of little consequence, a conclusion confirmed by operating experience.
The restrictive leakage rate limits in the plant Technical Specifications provide assurance that the unit will be shutdown in a timely manner for the appropriate corrective action (see additional discussion in Section 4.3).
If necessary, preventive repairs (see Section 4.2), more restrictive limits on primary to secondary leakage, hydrotesting of the tube bundle, and corrective measures to retard the rate of further corrosion (see Section 3.4) are additional steps which can be taken to provide added assurance i
of safe operation.
Eddy current testing has also proven useful for detecting and monitoring the early stages of denting at drilled-hole support plates.
A dented tube will produce an eddy current signal which is generally indicative of the average 33
5 diametral reduction.
At the egg crate supports of CE steam generators, however, the dented tubes generally assume an oval-shaped geometry such that the average diametral reduction seen by eddy current testing is generally insignificant.
'For this reason, CE recommends to its customers that dent measurements at egg crate supports be performed using a profilometer.
A profilometer is a device capable of taking multiple diameter measurements around the tube circumference.
Severe denting will block the eday current probe.
The degree of denting can be quantified by inserting progressively smaller probes until one will pass through the dented location.
4.2 Tube Repairs The plant Technical Specifications provide limits (referred to as plugging limits) for the maximum allowable percentage of wall degradation beyond which the tubes must be removed from service by plugging.
The plegging repair tech-nique involves the installation of plugs at the tube inlet and outlet.
After plugging, the tube no longer functions as the boundary between the primary and secondary coolant systems.
The plugging limits are based upon the minimum tube wall thickness necessary to provide adequate structural margins (in accordance with Regulatory Guide 1.121) during normal operating and postulated accident conditions.
These limits make allowance for eddy current error and incremental wall degradation which may occur prior to the next inservice inspection of the tube.
These plugging limits are conservatively based upon an assumed mode of degradation in which the walls are uniformly thinned over a significant axial length of tubing.
These limits do not consiaer additional structural margins associated with defects that create small-volume thinning, such as pitting, nor do they con-sider the external structural constraints against a gross tube failure provided by the tubesheet and tube support plates.
Operating experience has demonstrated that avJitional plugging criteria are necessary to address tube denting.
Dented tubes are susceptible to stress corrosion cracking at the location of the dent, which is dependent on stress level, strain rate, time, and material variables.
Tests have shown that dented tubes with small through-wall cracks near the support plate have ade-quate margins to prevent bursting or collapsing during normal operating and postulated accident conditions.
Severe stress corrosion cracking (SCC) could, however, reduce the margins to an unacceptable level.
The objective of the plugging criteria for dented tubes is to remove from service any tubes which may develop through-wall cracks or become severely degraded before the next steam generator inspection.
These criteria are plant-specific and are generally based on operating experience that includes the maximum sized eddy current probe which can be passed through the dented location.
For plants with especially high rates of denting, additional criteria for plugging have been established based on the rate of denting and the interval of time before the next inspection.
Plugging criteria based upon the maximum sized probe which can be inserted through the dented location are justified by operating experience.
That is, operating experience has shown that dent-related leaks which have occurred in service have generally restricted the passage of a specific sized probe.
Such 34
.. ~.
criteria tend to be overly conservative since there is not a unique relation-ship between the maximum reduction in tube diameter and the susceptibility of a tube to PSCC.
Profilometry inspections can provide a more direct assessment of the strain level at the dent and may ultimately provide a better basis on which to set plugging limits.
Profilometry would exclude tubes with low strains from being plugged, although such tubes are subject to plugging under existing criteria.
Profilometry techniques and plugging criteria based upon observed strain levels are being used to a very limited extent at Indian Point Unit 2.
Profilometry techniques are also being used to quantify the relatively early stage of dent-ing which has been observed in the egg crate supports of CE steam generators.
Plugging tubes based on the magnitude of denting rather than on the observation of cracks is one example of a " preventive" repair approach which has been used at a number of plants to remove from service tubes that are believed to be susceptible to degradation, even though eddy. current inspection may not have identified such tubes as defective.
Preventive repairs are generally per-formed only at plants subject to small volume defects for which eddy current testing alone has not proven adequate for early identification of potential leaks.
For exam;.le, most Westinghouse plants that have incurred frequent, non-deating-related U-bend leaks in row 1 tubes have, as a preventive measure, plugged all tubes in row 1, regardless of whether or not they leaked or eddy current testing revealed them to be defective.
In addition, Westinghouse recommends to its customers that all tubes in row 1 be plugged when denting-induced closure (so-called hourglassing) of the support plate flow slots is observed.
In another instance, at San Onofre Unit 1, the IGA was very diffi-cult to detect with eddy current testing, and the licensee performed sleeving or plugging repairs on all tubes within the region where the IGA was determined to be most advanced, regardless of whether or not the tubes exhibited eddy cur-rent indications.
i For some plants, sleeving repairs have been approved as an acceptable alterna-tive to plugging.
The advantage of sleeving as opposed to plugging is that it permits the tube to remain in service.
Sleeving is discussed in additional detail in Section 5.3.
4.3 Primary-to-Secondary Leakage Rate Limits The primary-to-secondary leakage rate limits in the plant Technical Specifications l
provide additional assurance of adequate tube integrity during normal and postu-j lated accident conditions.
Should the leakage rate limit be exceeded, the licensee is required to shut down the plant, repair the leaking tube, and conduct a steam generator inspection.
For some plants with advanced tube degradation, NRC approval is required prior to restart.
In a practical sense the leakage rate limits provide a very important indication of the existence or rate of steam generator tube degradation.
Experience has shown that some forms of degradation can develop in a period of time shorter j
than the routine inspection intervals or may be difficult to detect with cur-l rent ECT techniques.
In the event that such degradation occurs, the leakage l
35
rate 1imits act to indicate when plant shutdown, ISI, and corrective actions should be taken.
From a practical standpoint, this is perhaps the most impor-tant function of the leakage rate limits.
5.
LONG-TERM CORRECTIVE ACTIONS As mentioned above, steam generator tubes are key components separating the primary and secondary coolant systems in PWRs.
These tubes, like many interface components, affect both systems, and their failure is an operational as well as a potential safety concern.
Therefore, the steam generator must be viewed as a part of the total system in which it operates.
Maintaining the integ-rity of the tubes thus requires a systems approach that should encompass mechanical, structural, material, and chemical considerations.
5.1 Improved Designs Westinghouse (W) and Combustion Engineering (CE) have developed and are fabricating advanced models of steam generators for future plants.
Babcock &
Wilsox, the other manufacturer of steam generators, is not currently fabricat-ing a new model.
The new Westinghouse (model F and advanced model F) and CE (System 80) steam generator models include multiple features to minimize operating problems.
The new features include:
1.
Ferritic stainless steel tube supports to minimize the potential for denting.
2.
Tube support designs which minimize crevices.
3.
Improved sludge' removal characteristics.
4.
Thermal hydraulic modifications to minimize areas of unequal heat transfer and steam blanketing.
5.
Elimination of open tubesheet crevices.
6.
Thermally treated Inconel-600 tubes (W only).
7.
Features to improve maintenance, repair, and ALARA conditions.
B&W has modified its existing designs in an attempt to eliminate the known mode of degradation.
For example, to reduce tube failures at the upper tube-sheet along the inspection lane, B&W has recommended that operating plants install five lane blockers between the 7th and 14th tube support plates to minimize the potential for moisture to enter the upper levels of the steam generator along the inspection lane during normal operations.
The installation of blockers, coupled with strict attention to secondary plant operations, should r
minimize the occurrence of this form of degradation.
Another area of improved design concentrates on the selection of more corrosion-resistant materials in the condenser because water leaks through the failed condenser tubing, when combined with air, can contaminate the condensate, feedwater, steam generator water, and steam.
This contamination in turn 36
i degrades the structural integrity of the steam generator tubes, turbine, and other components in the cooling system.
The utilities are eliminating the use of ammonia-sensitive alloys from the condensers and replacing them with more corrosion resistant alloy tubing. Where denting is a concern, steps are t
being taken to eliminate all copper alloys from the condenser, feed train, and moisture separator reheaters.
The copper alloys are being replaced by materials such as titanium, AL 6X, or stainless steel (for freshwater service).
5.2 Improved Water Chemistry Control The three PWR vendors currently recommend all volatile treatment (AVT) for steam generator water chemistry control.
AVT consists of the addition of hydrazine (N H ) to the condensate water (or at some plants the high pressure 7g turbine connect piping) for the purpose of scavenging oxygen.
Excess hydrazine (that amount stoichiometrically in excess of dissolved oxygen) thermally decom-poses to ammonia at steam generator operating temperatures, which will provide for pH control to reduce carbon steel corrosion.
If the thermal decomposition of hydrazine to ammonia does not sufficiently raise the pH, additional tanks and pumps are utilized so that other nonsolid additives such as ammonium hydroxide, morpholine, or cyclohexamine can be added.
These additives act to increase the pH throughout the entire condensate, feedwater, steam generator, and steam cycle to reduce corrosion of carbon steel-components throughout the secondary system.
Extensive experience in both the fossil and nuclear indus-tries has demonstrated the benefits of these additives for secondary cycle corrosion control in electric power generating plants.
The primary advantage of AVT is that no dissolved solid additives are used (such as phosphates) which can concentrate in the steam generators to induce corrosion, such as phosphate wastage of Inconel-600 tubing.
The disadvantage of AVT is that it provides no buffering capacity to mitigate the effects of impurities in the cooling water through the condenser or corrosion products.
Thus, when condenser leakage occurs, the resultant impurities can enter the steam gener-ators and cause severe changes in the pH, with resultant increases in corrosion rates.
Although the three PWR vendors currently recommend AVT, both Westing-house and CE had in the past recommended the use of phosphates to buffer impurities in their recirculating U-tube steam generators.
As a consequence of the discovery of phosphate wastage of Inconel-600 steam generators in 1973, AVT became the favored water chemistry control technique.
By 1975 all plants, with the exception of H.B. Robinson Unit 2, and San Onofre Unit 1, had shifted to AVT.
These two plants have not experienced phosphate wastage at the rate exhibited al other plants using phosphate chemistry control.
The specific j
reason (s) for a slower rate of wastage at these plants is not known.
All further discussion in this section will focus on those plants which are l
operating with AVT.
The AVT steam generator and feedwater chemistry control limits recommended by all three PWR vendors are virtually identical.
Their primary differences are in the recommended methods by which the limits are achieved.
l 5.2.1 B&W Recommendations l
l B&W recommends continuous full flow condensate polishing at all times, and blowdown only during startup, before a sufficient power level to produce super-37
i, heated steam is reached.
Both recommendations are prudent for the once-through superheating steam generator (OTSG) design.
Continuous full flow condensate polishing is necessary to preclude the possibility of condenser inleakage of hardness salts from entering the steam generator where, because of their low solubility as the steam becomes superheated, the salts will deposit on heat-transfer surfaces thus reducing efficiency.
The use of blowdown during startup only is also consistent with the OTSG design.
During low power operations 'he lower portion of the OTSG has internal recirculation, which tends to concen-trate feedwater impurities (similar to the normal concentration mechanism in 1
U-tube steam generators).
Therefore, blowdown is necessary during low power operation to mitigate the effects of concentrating these impurities.
- However, when the OTSG starts producing superheated steam, the internal recirculation and concurrent concentration of feedwater impurities stops.
Without this con-centration of impurities, blowdown then becomes simply a discharge of feedwater, which is an inefficient method for removing impurities.
5.2.2 Westinghouse and Combustion Engineering Recommendations Westinghouse and Combustion Engineering recommend that full flow condensate polishing be used sparingly and only after extensive design reviews are con-ducted to ensure that ionic impurities from the resin beds and the resin beads themselves are prevented from entering the steam generators.
Additionally, both vendors recommend detailed operating procedures to monitor performance of the condensate polishing systems.
The reluctance of these vendors to fully support continuous condensate polishing is based on laboratory test data and operating experience which demonstrate that an improperly operated or designed condensate polishing system can result in more carrosion damage to the materials of a recirculating U-tube steam generator than controllable quantities of condenser impurity.
The greatest difference between Westinghouse and Combustion Engineer-ing recommendations is in the area of steam generator blowdown.
Westinghouse recommends a maximum continuous blowdown less than or equal to 5.0 gpm.
This rate provides for the maximum concentration of impurities in the steam generator (which improves the ability to detect small condenser leaks).
Once a leak is detected, maximum available blowdown (0.75% of the maximum steaming rate of 12 to 150 gpm) is recommended until the leak is repaired and the steam generator chemistry restored to normal.
Combustion Engineering recnmoends a minimum continuous blowdown of 35 to 50 gpm, which is increased to 175 to 250 gpm when a condenser leak is detected.
This method reduces the maximum impurity concentration which is reached during condenser leakage, but it increases detection time for small leaks.
There are no data available which demonstrate the superiority of either blowdown method.
All three PWR vendors agree that:
(1) Condenser water inleakage is the most significant contributor to steam generator corrosion problems for plants with AVT.
Improved condenser designs, materials, leak detection procedures, and repair procedures are recommended.
Items to be considered include improved condenser tubes, better antivibration supports, double tube sheets, and welded tube /
tubesheet joints.
38
(2) Excessive condenser air ingress is the primary contributor to condensate and feedwater system corrosion.
Excessive corrosion of the condensate and feedwater system can result in corrosion product buildup in the steam generators and concurrent concentration of condenser cooling water impuri-ties to form sludge which enhances corrosion in the steam generators.
(3) Copper alloys should be eliminated from all areas of the condensate /
feedwater/ steam / condensation cycle.
Substantial evidence exists that copper oxides in the steam generators are an important catalyst in accelerating the rate of corrosion processes within steam generators.
5.2.3 Current Licensing Practices for Secondary Water Chemistry for Operating Plants In late 1975, the NRC staff incorporated provisions into the Standard Technical Specifications that required limiting conditions for the operation and sur-veillance of secondary water chemistry parameters.
The plant Technical Specifications for all PWRs issued an operating license since 1974 contain either these provisions or a requirement that these provisions be established af ter baseline water chemistry c:nditions have been determined.
The intent of the provisions is to provide added assurance that the operators of newly licensed plants will properly monitor and control secondary water chemistry to limit corrosion of steam generator components.
In the past, however, strict limits in the Technical Specifications have significantly restricted the operational flexibility of some plants with little or no benefit with regard to limiting degradation of steam generator l
tubes and the tube support plates.
Based on this experience and on the knowl-edge gained in recent years, the staff concluded that specific Technical Specification limits are not the most effective way of ensuring that steam generator degradation will be reduced.
Because of these considerations, in August 1979, the staff instituted license conditions that require the imple-mentation of a secondary water chemistry monitoring and control program that contains appropriate procedures and administrative controls.
Although specific operating limit requirements for secondary water chemistry control have been deleted from the existing Standard Technical Specifications, specific plant Technical Specifications will still retain requirements for prinary-to-secondary leakage rate limits, steam generator tube surveillance, and plugging criteria to ensure that tube integrity is not reduced below an ac rptably safe level.
The new approach requires that each licensee make liunsing amendments that incorporate an administrative control to implement a secondary water chemistry cor, trol program.
Any plant requiring minor changes to the program and procedures (i.e., limits of water chemistry parameters or frequency of sampling), would be handled under 10 CFR Part 50.59.
In addition, compliance with this water chemistry program and procedures is subject to audit by the NRC Office of Inspection and Enforcement.
To further refine the specifications for PWR steam generator and secondary system water chemistry, an industry committee of technical experts formed in mid-1980, under the auspices of the Electric Power Research Institute (EPRI).
This committee was directed to conduct a detailed review of the water chemistry specifications for all three PWR vendors.
The committee was responsible for 39
evaluating all available data supporting current chemistry specifications and developing generic secondary system chemistry specifications.
Committee members included technical representatives from each PWR vendor, from a mini-mum of four utilities, one consultant, and one EPRI representative as a nonvoting chairman.
In July 1981, the Committee completed the final draft of its generic report and submitted it to the PWR Owners Group Technical Review Committee for comment.
Comments from the Technical Review Committee are being resolved and the final draft generic guidelines will be issued to PWR owners for comment near the end of 1981.
The staff anticipates that the generic guidelines will be issued b; May 1982 for final comments.
The staff has requested that it be included in those asked for official comments at that time.
Pending a satis-factory resolution of staff commcnts, the staff intends to issue a Branch Technical Position which will incorporate the generic guidelines as the review basis against which all PWR secondary system water chemistry control programs are evaluated.
This position should be completed late in 1982.
5.3 Sleeving When tubes are severely degraded, often large numbers of them must be removed from service by plugging to ensure the generator's safe operation.
Plugging steam generator tubes results in a loss of heat transfer surface and can eventually necessitate a reduction in power levels.
Faced with this prospect, some utilities have elected to replace their steam generators.
Such replace-ments require a long outage, involve considerable cost, and entail significant occupational exposures.
To prolong the life of severely degraded steam gene-rator tubes, some utilities, with prior NRC approval, have elected to repair them by sleeving.
Sleeving not only decreases the olant downtime but also leaves the repaired tubes functional.
The tube sleeving procedure involves inserting a tube of smaller diameter (or sleeve) inside the tube to be repaired.
The sleeve is positioned to span the degraded portion of the original tube and is then either hydraulically or mechanically expanded above and below the degraded region.
The expanded joints are sometimes brazed to ensure additional leak tightress.
Sleeving has been used for two different purposes.
In Westinghouse and Combustion Engineering steam generators, sleeving has been used to repair degraded tubes as an alternative to plugging.
In one Babcock and Wilcox plant, tube sleeving has been used to stiffen the tubes so as to alter their natural frequency in an effort to eliminate or reduce flow-induced vibration.
Sleeving repairs to restore primary coolant boundary integrity have been per-formed, to date, on the straight portion of tubing degraded by wastage, intergranular attack, and stress corrosion cracking.
Although adequate for these purposes, at present such repairs do not appear to be a viable alterna-tive for tubes degraded by denting.
Tables 1, 2, and 3 summarize sleeve repairs that have been performed at Palisades, Ginna, Oconee Station, San Onofre Unit 1, and most recently, Point Beach Unit 1.
40
5.4 Replacement l
To avoid the need for derating the plant and excessive downtime to perform steam' generator inspections, some utilities have elected either to replace their severely degraded steam generators or are considering doing so.
Virginia Electric Power Company (VEPCO), for example, has successfully replaced the steam generators in its Surry Units 1 and 2 and returned to full power operation (see Table 5).
Florida Power and Light Company (FP&L) is currently replacing the steam generators in Turkey Point Unit 3 and is planning t
to do the same for Turkey Point Unit 4.
Replacement of three steam generators takes approximately 10 months.
i utilit.ies must consider the following factors before replacing steam generators:
1 (1) size of the equipment hatch opening, (2) vertical clearance within the con-tainment building, and (3) preference with respect to reactor coolant pipe cut i
or channel head cut.
The replacement steam generators in Surry Units 1 and 2 and Turkey Point Units 3 and 4 (Table 5) incorporate many of the design features of the new generation of Westinghouse steam generators (Model F) discussed in Section 5.1.
To min-imize the potential for several modes of tube degradation which have been identified to date, the replacement generators include the following improve-ments:
1.
Type 405 ferritic stainless steel quatrefoil tube support plate i
2.
Thermally treated Inconel 600 tubing and stress relief of the inner-most eight rows of the tube bundle to reduce the potential for SCC 3.
Expansion of the tubes to the full depth of the tubesheet to eliminate crevices i
4.
A flow baffle plate above the tubesheet to direct lateral flow across the tubesheet surface and thus minimize the number of tubes exposed to i
l sludge 5.
An improved blowdown system to increase blowdown capacity 6.
OCCUPATIONAL EXPOSURE ASSOCIATED WITH STEAM GENERATOR MAINTENANCE Occupational exposure has become a major consideration in the maintenance and repair of steam generators.
Corrosion, as a consequence of design and opera-tional problems, has created the need for routine inspection and maintenance, periodic repairs and modifications, and, in some circumstances, steam generator replacement.
The presence of radioactivity in steam generators -- mainly Cobalt-58 and Cobalt-60 -- and activated corrosion products deposited on internal primary system surfaces results in radiation exposures that amount l
to a major fraction of the occupational exposure received by contractor, l.
utility, and plant personnel performing steam generators maintenance.*
l l
I A
Information for this report was derived from licensee outage reports and data files, annual radiation exposure reports, final environmental statements, and NUREG/CR-1595, " Radiological Assessment of Steam Generator Removal and
. Replacement."
41 l
,.,,--,.n.-
-.,,--,..-,,,-,---.n-
Table 5.
Steam generator replacement summary No. of Moist sep.
Replace Site Site Plant S/G replaced Model method start comp Comments Surry 1 3
Yes (20")
51 F RC pipe cut 9/15/80 7/7/81 Operating at 100% power a
Surry 2 3
Yes (20")
51 F RC pipe cut 2/3/79 8/18/80 Operating at 100% power D
Turkey Point 3 3
Yes (7")
44 F Ch. hd. cut 7/1/81 4/15/82 Site effort in progress b
D Turkey Point 4 3
Yes (/")
44 F Ch. hd. cut 10/15/82 7/25/83 Replacement was considered complete on 12/31/79 and the unit was restarted on 8/18/80.
a b8est available information.
m
Information required by the Commission for annual exposure reports does not require licensees to record exposures by specific task, such as steam generator maintenance.
Steam generator work is typically combined with a wide variety of work categorized as "Special Maintenance" and is not separately described.
Data for steam generator tasks are available only where individual licensees have voluntarily compiled and maintained records of the activities, which they then made available to the Commission.
Table 6 gives occupational exposure data for steam generator inaintenance, repair, and replacement for 1974 to 1981 from seven utilities.
Table 6 includes the total yearly dose from steam generator work, and provides actual or esti-mated outage time (as available) for these activities.
In some cases, data are provided for specific units. Where information did not identify if the exposure resulted from maintenance or repair and replacement activities, a total occu-pational dose is provided.
Steam generator work is performed by contractor, utility, and plant personnel, but the available data are insufficient to deter-mine the extent of exposure for each group.
Radiological assessments of steam generator work are performed by NRC staff members in conjunction with license amendment reviews, and include a review of ALARA measures, dose estimates, cost / benefit analyses, and inspection of lic.ensee activities while in progress.
To date, such reviews and inspections have been conducted for steam generator replacements at Surry and Turkey Point, and for tube sleeving at San Onofre.
NRC also funded the following studies of the radiological aspects of steam generator work at Pacific Northwest Laboratory:
(1) NUREG/CR-0199, " Radiological Assessment of Steam Generator Removal and Replacement," (2) its revision, NUREG/CR-1595, and (3) NUREG/CR-1490, "Some Aspects of Cost / Benefit Analysis for Inservice Inspection of PWR Steam Generators."
6.1 Maintenance and Inspection Maintenance and inspection of steam generators typically involve the following activities which entail occupational exposure:
(1) Preparation of Work Area - surveying; removing interferences and lagging; shielding of " hot spots" and general area shielding; erecting scaffolding; installing tent and air breathing systems; removing access covers; decontaminating and shielding internal steam generator work area.
(2) Inspections and maintenance - eddy current testing; ultrasonic testing; dye penetrant testing; leak testing; cutting inspection ports; sludge lancing.
(3) Recovery and cleanup of work area - complete restoration to operating conditions; removing and processing of radioactive waste.
Maintenance and inspection may entail from two to four weeks of downtime, and entails doses typically less than 100 man-rems.
6.2 Repairs Repairs, frequently performed in conjunction with inspection and maintenance efforts, may involve the following activities:
tube plugging; tube pulling; 43 1
Tcble 6.
Occup2tional exposure related to steam generator caintenance, replacement and repair from selected PWRs (1974-1981) (dose in man-rems)
Plant 1981 1980 1979 1978 1977 1976 1975 Oconee 1, 2, 3 (2) (3)
(1) (3)
(1) (2) (3)
(1) (2) (3)
(1) (2) (3)
(1)
Maintenance 25 18 58 16 276 23 14 26 32 34 28 2
6 Repair / replacement 155 8
87 52 12 82 37 21 47 25 7
Total 206 161 377 232 115 44 Related outage time Robinson 2 Maintenance 212 97 120 61 Repair / replacement 91 130 95 Total 322 303 97 194 None 121 250 156 Related outage time 90d 96d 21d*
21d*
56d*
21d*
i San Onofre 1 81-80 e=
Maintenance 42 65 75 Repair / replacement 3451 250 Total 3493**
65 325 Related outage time 273d Indian Point 2, 3 (2) (3)
(2) (3)
(2) (3)
(2) (3)
(3)
Maintenance 39 --
99 65 120 --
346 25 Repair / replacement 4 157 10 90 15 22 41 Total 200 264 157 412 31 Related outage time Point Beach 1 Maintenance Repair / replacement Total 269 235 62 125 45 Related outage time 24d Surry 1, 2 (1)
(1)
(2)
Maintenance Repair / replacement Total 1430***
329***
2140***
788 1058*
1287 638 100 Related outage time 289d 331d See footnotes, last page of table, w
Table 6.
(continued)
Plant 1981 1980 1979 1978 1977 1976 1975 1974 Turkey Point 3, 4 (3) (3)
(3) (4)
Maintenance 75 92 114 Repair / replacement 239***68 46 173 Total 382 425 335 335 450 600 200 Related outage time 71d 103d 42d 17d 70d 40d
- Estimated.
tData for 1980-1981.
t Tube' sleeving effort 1
S/G replacement
' NOTE:
Figures are rounded off to nearest whole number.
Outage time is reported in days (d).
. Unit numbers (1), (2), (3), (4).
Of the 12 units covered in Table 6, nine are designed by Westinghouse, and three by Babcock and Wilcox.
tube weld repairs; tube honing and sleeving; and modifications to such compo-nents as feedwater rings, demisters, flow divider plates (and other internal components), and instrumentation, such as level gages.
A moderate repair ef fort typically lasts from two to five weeks, with doses ranging from 10 to 100 man-rems.
Extensive repairs and modifications increase outage time and doses (e.g., 100 to 3000 man-rems).
6.3 Replacement Replacing a steam generator involves the following four main steps, each of which entails occupational exposures in the ranges specified:
(1) Post shutdown preparation 270-310 man-rems (2) Removal of old steam generator 290-420 man-rems (3)
Installation of new steam generator 240-830 man-rems (4) Disposal of old steam generator 2.4-580 man-rems The estimated occupational doses for these phases, as provided in NUREG/CR-1595, range from 802.4 to 2140 man-rems per steam generator.
Estimates are based on the number of anticipated work hours in averaged radiation fields.
Total occupational radiation exposure expended during the steam generator replacement for Surry Unit 2 was approximately 2141 man-rems, which is almost 4% above the utility's exposure estimate of 2067 man-rems for each unit and 11% below the lowest estimate of 2407.2 man-rems provided in NUREG/CR-1595 for each unit.* Total occupational radiation exposure expended during the generator replacement for Surry Unit 1 was approximately 1759 man-rems, which is 15% below the original estimate and a 19% reduction from that expended for Unit 2.
6.4 Duration of Outage Steam generator maintenance and repair are typically scheduled to coincide with other major outages, such as refueling or a main turbine overhaul.
In such cases, the steam generator work may be performed simultaneously with " critical path" work and result in no exclusive outage time attributable to it.
In those instances where operating limits for the generator are approached or exceeded (e.g., as those set by Technical Specifications for tube leakage), an outage specifically for steam generator repair may be required.
A steam generator inspection and repair outage involving eddy current testing, sludge lancing, and modest tube plugging takes from two to four weeks.
Data are not suf-ficient to establish the frequency of unscheduled outages for steam generator repair.
In many instances where the outage duration for steam generator work is estimated, as in Table 6, available data included only total outage time for one or a combination of the following activities:
(1) all work conducted;
- Revision of NUREG/CR-0199, " Radiological Assessment of Steam Generator Removal and Replacement."
l l
46 L
l l
WW-w
(2) total manpower or manhours expended for supporting steam generator work; and (3) steam generator work categorized under "Special Maintenance" with many other types of work.
6.5 Exposure Reduction Techniques ALARA doses can be achieved through the practice of exposure reduction techniques.
The following subsections list recent exposure-reduction techniques for steam generator work.
4 6.5.1 Pre-work Preparation, Planning, and Training a.
Review prior similar work for ALARA approaches b.
Perform comprehensive radiation protection training for all workers, including contractors c.
Train workers on mockups and by rehearsing tasks d.
Conduct surveys to determine temporary shielding needs and their desirability e.
Establish a strong worker-management program so that worker time and effort in high dose rate zones is not wasted f.
Use adjustable platforms in lieu of scaffolding g.
Perform general area decontamination to reduce surface contamination h.
Use tents and glove bags to confine contamination i.
Optimize the numbers of workers assigned to a task i
j.
Position and control tools and equipment k.
Develop and test special tools and equipment to ensure that they will operate as expected 6.5.2 Minimizing On-the-Job Exposure a.
Decontaminate steam generators and/or nearby radioactive systems b.
Shield generators (when dose-effective) c.
Establish low background exposure waiting areas for workers d.
Use special tools and equipment, including automated equipment (e.g.,
eddy current tester) e.
Remove components to low background-exposure areas for maintenance f.
Maintain steam generator water levels for shielding where practical g.
-Conduct in progress ALARA reviews 47
h.
Use remote maintenance methods (e.g., closed circuit TV) i.
Use local ventilation of work areas (e.g., HEPA-filtered steam generator access tent) j.
Use an effective communications system for communicating with steam generator workers k.
Use anticontamination clothing and air breathing systems which afford maximum protection and worker comfort 1.
Use a positive exposure control administrative system to control dosage 6.5.3 Post-Work Recovery a.
Maintain a radioactive waste reduction program for solid and liquid wastes b.
Maintain general area cleanliness and decontaminate areas and materials to as low as practical levels c.
Conduct end-of-work ALARA reviews with evaluations and recommendations which can support further work In summary, the occupational exposure associated with steam generator maintenance, repair and replacement ranges from 10% to 60% of the total radia-tion dose per year in facilities where they must be tended.
The percentage of total annual dose attributable to steam generator work at these facilities is summarized in Table 7.
Where major repair or replacement efforts are required, dose expenditures may range from 2000 to 3500 man-rems.
The exposure reduction techniques itemized in Sections 6.5.1, 6.5.2, and 6.5.3 can be applied to achieve ALARA doses. The lack of an apparent solution to the tube degradation problems indicates that many more maintenance and repair efforts will be made on these and other PWRs.
Table 7.
Steam generator annual dose as a percentage of total annual dose (selected pressurized water reactors 1974-1980)
Plant 1980 1979 1978 1977 1976 1975 1974 Oconee 1/2/3 B&W 19.5 15.8 27.6 17.8 11.3 7.6 Robinson 2 W
16.4 8.2 20.3 16.9 15.2 23.2 San Onofre 1 W
Indian Point 2/3 W
20.6 17.2 20.7 2.9 Point Beach 1 W
10.6 36.5 20.1 29.1 12.2 Surry 1/2 W
45.9 59.8 45.4 4,5.8 40.7 38.7 11.3 Turkey Point 3/4 W
25.7 19.9 32.5 43.4 50.7 22.8 A
Steam Generator NSSS, W = Westinghouse, B&W = Babcock & Wilcox.
48 l
7.
RELATED RESEARCH PROGRAMS 7.1 NRC Steam Generator Confirmatory Research Program The NRC steam generator confirmatory research program addressing steam generator tube integrity began in 1976.
A study to develop a method for predicting stress-corrosion cracking in steam generator tubing began in 1977, and a research program to develop improved eddy current techniques for steam generator inservice inspection was initiated in 1978.*
The program objectives and scope of work for these continuing programs are described below.
7.1.1 Steam Generator Tube Integrity The original goals of this program were (1) to develop validated models for predicting margins to failure under burst and collapse pressures and leak rates for steam generator tubing found to be degraded in service, and (2) to establish the efficiency of eddy current testing for locating and characterizing defects in steam generator tubes.
Laboratory tests were conducted with tubing representative of that used in PWR steam generators and with flaws to simulate known or expected defects in operat-ing steam generators.
Burst and collapse tests were conducted in simulated PWR steam generator chemical and thermal environments using a high pressure auto-clave assembly.
These tests showed that burst and collapse pressures for the flawed specimens were higher than those that would occur during postulated accidents, such as a loss of coolant (LOCA) or a main steam line break (MSLB).
The eddy current tests showed that the single-frequency eddy current techniques presently in general use did not accurately characterize many of the machined defects.
However, it was also shown that current plugging practices are conservative from the viewpoint of margins-to-failure for the machined defects.
The present phase of the tube integrity program involves validation of laboratory test results on tubes that have been degraded in service in an operating reactor.
The program will perform research on a steam generator removed from the Surry Unit 2 nuclear station, which presents researchers with a unique opportunity to conduct tests on actual service-induced defects.
The program is scheduled to continue until the facility is decommissioned in FY 1987.
7.1.2 Stress Corrosion Cracking of PWR Steam Generator Tubing The overall objective of this laboratory experimental program is to develop deta and models which will be used to predict the stress corrosion cracking ser', ice life of Inconel-600 steam generator tubing under normal and abnormal service conditions.
The variables in the testing program include temperature,
^" Steam Generator Tube Integrity:
Phase I Report," (NURGE/CR-0718), Battelle Pacific Northwest Laboratory, September 1979.
49
stress, strain and strain rate, metallurgical structure and processing, and ingredients of the primary and secondary coolant.
Appropriate results and factors will be incorporated in the predictive models for SCC service life.
Through 1981, the program has investigated the effects of carbon content, temperature, stress level, cold work, and water conditions on the initiation and propagation of stress corrosion cracks at grain boundaries in Inconel-600.
Both constant extension and U-bend tests are being conducted.
The results of tests conducted to date appear in "Effect of Environmental Variables on the Stress Corrosion Cracking of Inconel-600 Steam Generator Tubing," by T.S. Bulischeck and D. Van Rooyen in Nucleac Technology (Vol. 55, page 383, November 1981).
Data from this program will be used to develop and refine predictive models for SCC behavior in primary and secondary side SCC performance for steam generator tubing.
The models will be applicable to tubes for which the proces-sing characteristics and service conditions are known together, as they are for example, for standard production tubes curtently in service.
The program will also indicate where additional information and testing are required in order to permit the predictive model to be applied.
7.1. 3 Improved Eddy Current Inservice Inspection for Steam Generator Tubing The objective of this program is to upgrade and validate eddy current inspection probes, techniques, and associated instrumentation for inservice inspection of steam generator tubing.
Furthermore, it is desirable to improve defect char-acterization as it is affected by variations in tube diameter and thickness, tube denting, probe wobble, tubesheet and tube support interference, and location and type of defect.
Preliminary results from this program indicate success in designing and developing improved eddy current equipment and techniques for the inservice inspection of steam generator tubing employing a multifrequency technique.
Using design calculations based on theoretical mathematical models, a proto-type three-frequency instrument with probes was constructed and laboratory evaluated.
It has the capability for either separating and measuring, or dis-criminating among variations in each of the following parameters:
(1) tube diameter, including denting at the support; (2) probe wobble; (3) the presence of supports around the tube; (4) tube wall thickness; (5) location (radial and axial) of defects in the tube wall; and (6) the size of defects.
By the end of 1981, improved instrumentation for field testing had been built, installed in a mobile inspection laboratory, and used successfully for inspec-tions at operating reactors.
As a result of field inspections, both hardware and software improvements have been made in the system.
Future efforts will include:
(1) completion of the development of probes, techniques and criteria for evaluating dented and cracked steam generator tubes; (2) development of correlations between eddy current readings from stress corrosion cracks and electro-discharge machine (EDM) notches; and (3) design, construction and evaluat. ion of pancake coil and multipurpose probes for the accurate detection and characterization of all possible flaw types experienced in service, including denting and circumferential flaws.
50
7.2 Electric Power Research Institute - Steam Generator Research The Nuclear Power Division of EPRI is also conducting research on steam generator issues.
The Analysis and Testing Program within EPRI includes the Steam Generator Technology Subprogram, which has as its objective the development of methods for alleviating steam generator corrosion and structural problems.
The Systems Integrity Program includes the Nondestructive Exam-ination and Evaluation Subprogram, which includes tasks for Steam Generator NDE.
Some of the tasks in both areas are supported by the ~ industry Steam Generator Owners Group.
7.2.1 Steam Generator Technology Subprogram Projects under this subprogram are divided into the following areas:
(1) chemistry and corrosion; (2) materials selection and testing; (3) thermal, hydraulic and structural testing and analysis; and (4) thermal and hydraulic code development.
Specific tasks in chemistry and corrosion involve determining the causes of denting, general intergranular corrosion, and cracking in cold-worked tubing (e.g., U-bends, square bends, rolled areas).
Other tasks involve quantifica-tion of water chemistry limits (i.e., allowable limits for concentrations of impurities, if any); evaluation of the effectiveness of water soaks in pre-venting corrosion; development of one or more neutralizers for crevice acids or for caustic deposits; and, the development of alternative water chemistry control systems (e.g., on-line chelant or inhibitor additions).
This effort includes the collection, evaluation, and correlation of operating plant data and pot and model boiler laboratory tests.
Tasks related to materials testing include the characterization of heat-treated Inconel-600, the investigation of the corrosion resistance of Inconel-690 and other candidate tubing materia;s, and corrosion testing in model boilers of various candidate structural material-tube material combinations.
In addition, it is anticipated that an examination of one of the steam generators from Surry Unit 2 will be conducted to assess the material problems experienced.
Environmental fretting and corrosion fatigue testing of Inconel-600 and other i
types of steam generator tubing will be performed as needed to extend the data l
base.
j 7.2.2 Steam Generator Nondestructive Examination (NDE) Program l
The principal focus of the steam generator NDE project is to develop inspection techniques which overcome the shortcomings of conventional eddy current systems used in the field. With that objective in mind major emphasis has been placed on the use of multifrequency/multiparameter eddy current technology for the inspection of steam generator tubing.
In addition, tasks have been undertaken to develop a variety of inspection devices and techniques for:
(1)
Inspection of tube conditions to supplement eddy current testing (2) Measurement of dent size 51
4 (3) Determination of the amount of corrosion product that has accumulated in the tube-to-support plate gap (4)
Inspection of the suppot t_ plates for damage Other tasks involve (1) developing automatic eddy current signal analysis to speed up signal interpretation and reduce operator dependence and (2) providing a better theoretical understanding of the interactions between eddy current and steam generator conditions.
These projects are reported annually.
The most recenttresults are reported in " Nondestructive Evaluation Program Progress in 1980" (EPRI NP-1690-SR), December 1980.
Another important element of the EPRI NDE program involves the establishment and operation of the F" $)E Center.
The purpose of the center is to provide the utility industry dedicated NDE capability that concentrates on acielerating the tra.
of research and development results into a form directly beneficial to the industry.
To accomplish this, three major goals have Deen defined.
The first is to conduct all those activities necessary to technology transfer; e.g., performance evaluation, engineering modification, calibration, performence data bases.
The'second is to train inspection personnel.1or the specific requirements of the industry.
The third is to
' develop working relationships with the academic community as a means of alleviating the long-term manpower shortage.
8.
TECHNICAL RESOLUTION OF UNRESOLVED SAFETY ISSUES A-3, A-4, AND'A-5 REGARDING STEAM GENERATOR TUBE INTEGRIT h i
In 1977, NRC established Ta:;k Action Plans (TAPS) A-3, A-4, and A-5 to' evaluate the safety significance of tube degradation in W, CE, and B&W steam generators, respectively.
These tasks were later designated as " Unresolved Safety Issues" in the NRC Annual Report for 1978, pursuant to Section 210 of the Energy Reor-i ganization Act of 1974.
This section summarizes the approach used in the TAPS to evaluate the safety significance of steam generator tube degradation and the resulting conclusions and requirements.
The TAPS integrated studies of systems analyses, inservice inspection, and tube integrity to establish improved criteria for ensuring adequate tube integrity and safe steam generator operation.
In the systems analyses studies, an evalua-tion of the consequences of steam generator tube failure during normal operation and postulated main steamline break (MSLB) and loss-of-coolant accident (LOCA) conditions has been performed.
The evaluation considers predicted fuel behavior, ECCS performance, radiological consequences, and containment building response.
Results of the systems analyses evaluation provide a basis for establishing a tolerable level of steam generator tube leakage during postulated accident conditions.
An evaluation of inservice inspection (ISI) techniques has been performed.
Acceptable-procedures for developing statistically based ISI programs have been developed which are intended to prov'Je adequate assurance that no more than the tolerable level of tube leakage, defined by the system analyses, would occur during normal or postulated accident conditions.
The evaluation of ISI techniques concentrated on establishing the ability of NDE techniques to detect and describe various defects and modes of degradation.
This effort included evaluation of eddy current testing accuracy in the region of the tubesheet and tube support plates.
52
The tube integrity assessment included evaluation of the behavior of degraded tubes during normal and postulated accident conditions and tube plugging cri-teria.
In addition, changes have also been identified in operating procedures and in steam generator and secondary system design to minimize tube degradation.
Results of these assessments have been used to develop and to evaluate the acceptability of new regulatory requirements.
Areas for which new requirements will be necessary include tube plugging aad repair criteria, inservice inspec-tion programs and techniques, secondary water chemistry monitoring, and condenser integrity.
Implementation of these requirements will make necessary analyses and evaluations by the licensees, as well as appropriate technical specification and license modificetions.
NRC will also need to update Regulatory Guides 1.121,
" Basis for Plugging Degraded Pressurized Water Reactor Steam Generator Tubes,"
and 1.83, " Inservice Inspection of Pressurized Water Reactor Steam Generator Tubes," and appropriate NRC Standard Review Plans and Branch Technical Positions.
A cost-benefit analysis for implementing TAP requirements has also been performed.
The objective of this task was to evaluate the potential impacts of implementing new ISI requirements and ensure that the costs of such require-ments, in terns of man-rem exposure, did not exceed the potential benefits.
In addition, the study focused on the practicality and financial costs of implementation.
A detailed description of all the evaluations and calculations performed, the resulting requirements and criteria, and the strategy for implementation are presented in a draft report currently under staff review.
The report,
" Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Generator Tube Integrity" (NUREG-0844), is expected to be issued for public comment in early 1982.
9.
CONCLUSIONS Steam generators manufactured by each of the three PWR vendors have experienced l
various forms of tube degradation resulting from a combination of inadequate l
design and fabrication, nonoptimal secondary system design and construction i
materials and poor operating practices, especially in secondary water chemistry control and condenser maintenance.
In addition, the inspection, repair, and replacement efforts needed to deal with these problems have also resulted in radiation exposures which account for a major portion of each facility's annual occupational radiation dose.
Industry-sponsored research has helped to identify the causes and mechanisms for several different types of tube degradation pheno-l mena which has subsequently led to some design and operating improvements.
It is anticipted that tube degradation will continue, but at a slower rate primarily i
because of better controls of variables leading to the problems rather than because of corrections to design deficiencies and construction materials.
Although some steam generator vendors have recently developed new steam generator models that are expected to provide significantly greater margins against tube degradation during operation, all plants scheduled to receive an operating license before 1984 have steam generators similar to those currently in service.
The NSSS vendors, the affected utilities, and the NRC staff are continuing to evaluate new areas where the potential for tube degradation exists and to l
53
Improve condenser integrity, secondary water chemistry control, steam generator and secondary plant designs and NDE inspection capabilities to minimize forced outages caused by steam generator tube failures.
In this regard, the staff is completing its Unresolved Safety Issues (USI) reviews, and summaries will be contained in the final edition of NUREG-0844,
" Resolution of Unresolved Safety Issues A-3, A-4, and A-5 Regarding Steam Gen-erator Tube Integrity." NUREG-0844 is expected to be issued in early 1982 for public comment.
Pending completion of Task Action Plans A-3, A-4, and A-5, the NRC staff has been evaluating adverse experience on a case-by-case basis and has concluded that continued operation and licensing do not constitute an undue risk to the health and safety of the public.
This finding has generally been based on the following considerations:
(1.4 Requirements for inservice inspections to monitor steam generator tube degradation have been established.
The frequency of inspection depends on previous adverse experience at each plant.
(2) Acceptance criteria (plugging limits) have been established to ensure that degraded tubing will retain adequate structural margins over the t
full range of normal operating, transient, and postulated accident conditions.
(3) Should complete (100%) through-wall degradation develop, the resulting leakage is generally small, as indicated by operating experience.
Allowable limits on primary-to-secondary leakage have been established beyond which the plant must be shut down for appropriate corrective action, and thus provide additional assurance of adequate tube integrity during normal and postulated accident conditions.
(4) Continued information from operating experience and USI Action Plan efforts will be utilized to update interim criteria and requirements.
(5) Wide dissemination of ALARA dose methods and techniques, based on up-to-date experience and further development efforts, can help minimize total doses when steam generator inspection, repair, and replacement are required.
i for plants with severe degradation, additional factors have been considered on a case-by-case basis:
(a) Additional inspections and/or preventive plugging (or sleeving) criteria have been implemented on a plant-specific basis, as necessary, to ensure that defective tubes have either been removed from service or repaired.
(b) For certain degradation phenomena, such as denting at tSe support plate intersections, and CSCC and IGA in the tubesheet crevices, the tubing is j
restrained against these structures to prevent a gross tube failure.
Leaks as a result of these degradation phenomena have been small and
- stable, i.e., no rapid failures.
Even if a LOCA or MSLB were to occur with some tubes containing through-wall or near through-wall cracks, the radiological consequences of such an event would not be severe.
r l
54
(c) Additional requirements, such as increased frequency of inspections, more restrictive limits on primary-to-secondary leakage, and hydrostatic test-ing of the tube bundle, have been established on a plant-specific basis, as necessary, to provide additional assurance of tube integrity.
(d) A small amount of leakage, less than the limits set in the Technical Specifications, can still be tolerated during normal operation without exceeding the offsite dose limits of 10 CFR Part 20.
(e) The probability of the design basis accident occurring during normal operation is small, and the probability that the accident would occur during the short period of time between the detection of a leak that exceeded the Technical Specifications leak rate limit and plant shutdown is even smaller.
(f) Corrective measures have been taken on a plant-specific basis to reduce the rate of further degradation.
These include improved controls of secondary water chemistry, sludge crevice flushing, boric acid treatments to retard denting, and reduced operating temperatures by reducing power levels.
The above rationale, which generally constitutes the basis for continued operation of licensed PWR facilities, also supports continued licensing of new facilities.
Further, to the extent that it is practicable for facilities not yet licensed for operation, state-of-the-art design improvements and operating procedures which are expected to decrease the potential for or rate of steam generator tube degradation are required by the staff.
The following design and operational factors are censidered by the staff in its reviews.
Designs should (1) Provide improved circulation to eliminate low flow areas, and to facilitate sludge removal (2) Minimize flow-induced vibration and cavitation l
(3) Provide increased flow around the tubes at the support plates (4) Use material for tube support plates with improved corrosion resistance (5) Use materials, processing and heat treatment to minimize the susceptibility of tubes to stress corrosion cracking (6)
Improve secondary system water chemistry control (7) Use improved secondary side materials (for condensers, feedwater heaters,
[
turbine discs and blades, elbows, etc.), and water cleanup systems to minimize erosion and its resulting sludge and corrosion product buildup.
55
f,",f""33S
' ^ " * ' '
U S NUCLE AR REGULATORY COMMISSION BIBLIOGRAPHIC DATA SHEET NUREG-0886 4 Tf TLE AND SUtsTITLE (Add Volume No. rf epreprosol
- 2. fleove blank l Steam Generator tube Experience a nEcie ENT S accession NO 1 AUT O H t::il
- 5. DATE REPORT COMPLE TE D f YE AR C.Y. Cheng and ethers M ON TH December 1981 9 PF HF OHMING OHGANI/ATION N AME AND MAILING ADDRESS (/nclude I<a Codel DATE HEPORT ISSUED MONTH l YEAR Division of Licensin9 February 1982 Office of Nuclear Reactor Regulation
,,t,,,,,,,,,,,
U.S. Nuclear Regulatory Commission Washington, D.C.
20555 a <Le, o+= s 12 SPONSOHING OHGANIZ AT:ON N AME AND MAILING ADDRESS IInclude I<a Codel h0 PROJECT /T ASK/WOHK UNIT NO Same as above
- n. CON T R ACT NO I 13 TYPE OF I4EPOHT PE RIOD Cove RE D II'tclus;ve dates /
Regulatory IS SUPPLEMEN TAHY NOTES
(
14 (L eme DIa" A l 16 A UST H AC T (200 words en less)
This report prov. ides information pertaining to the status of PWR steam generator tube experience and the resolution of unresolved safety issues A-3, A-4, and A-5 regarding steam generater tube integrity. It provides an overview of the types of problems which have occurred in PWR steam generators with particular emphasis on recent operating experience. The report also discusses short and long-term corrective actions being pursued by the industry to resolve these problems, steam generator inspection and repair requirements which have been established to ensure the continued safe operation of PWR steam generators, and occupational radiation exposures associated with the above-listed activities. It should be noted that information included in this report represents the current NRC staff understanding of each issue.
This report is intended to be a followup to the similar reports, NURE.G-0523 and NUREG-0571, which discusses tube operating experience with the recirculation ("U" tube) type and cnce-through type steam generators designed by Westinghouse and Combustion Engineering, and Babcock and Wilcox, respectively.
11 KE Y WOHDS AND DOCUMENT AN ALYS?S 1 74 DE SC HIP TOHS 14 'OL N il f t( HS OFE N E N DE D T E :".*S 18 AV AIL ABILITY ST ATE MENT 19 SE aH SS Tha reporr!
21 NO Or PAGE S Unlimited 20 Se cuHi Tv cuaSS <rss,,,,
22 enicE unclassified s
N RC F ORM 335 !? 771
UNITED STATES NUCLE AR REQULATORY COMMISSION POSTAGE AND FEES PAID W ASHIN GTON, D. C. 20S65 u.s. MUC LE A R R E GU L A tor Y Couwissson OFFICI AL CUSINESS PENALTY FOR FRIVATE USE,5300 L
j l
l l
l 123,5 5 SJ / ti8 7 /
2 9 <,9 9 9 il a + o. t A Cli aiv OF TIDC
['CLICY L PUBL IC AT 1 U?.5 MGT Ek e' O R '4UREG 0CPy L A 212 d A Stil t;G T O N CL 20555
.