ML20059F890
| ML20059F890 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 10/15/1993 |
| From: | MPR ASSOCIATES, INC. |
| To: | |
| Shared Package | |
| ML19311B173 | List: |
| References | |
| NUDOCS 9311050146 | |
| Download: ML20059F890 (22) | |
Text
-..
ATTACHMENT (11 Calvert Cliffs Nuclear Power Plant Units 1 and 2 l
i 1
Pressurizer Inconel 600 Sleeve Nozzle i
Safety Analysis Report for Alternative Roll / Plug Repairs (Non-Proprietary Version) i I
{
l l
Baltimore Gas & Electric Company Docket Nos. 50-317 and 50-318 November 1,1993
~
9311050146 931101 e;-
ty;d PDR ADOCK 05000317 g
P PDR l
MMPR ASSOCt ATE S INC E F4 G t N E E R S October 15,1993 NON PROPRIETARY VERSION CALVERT CLIFFS NUCLEAR POWER PLANT UNITS-1 AND 2 PRESSURIZER INCONEL 600 SLEEVE NOZZLE SAFETY ANALYSIS REPORT FOR ALTERNATIVE ROLI/ PLUG REPAIRS I
l d
l I
)
i 32D KING STREET ALEX ANDRI A, VA 22314 3238 703 519'C200 FAX; 703 519-0224
l a
i 4
CONTENTS i
Section BACKGROUND PURPOSE / SCOPE RESULTS 1.
Description of Repairs of the CCNPP Pressurizer Inconel 600 Sleeve Nozzles by Tube Rolling / Expansion Plugs 2.
Qualification of Repairs 3.
Examination of Leaking Sleeve Type Nozzles Before Repair 4.
Inspection Requirements to Ensure Repair is Satisfactory for Continued Service j
5.
Ejection Prevention 6.
Prior use of these Repairs 7.
Overall Summary REFERENCES APPENDIX A Izakage Inspection Criteria for Repaired Sleeve Nozzles i
2-
u l
BACKGROUND i
Pressure boundary penetrations fabricated from Inconel 600 have exhibited primary water stress corrosion cracking (PWSCC) in some pressurized water reactors. Cracking
~
has been reported in instrument nozzles and heater sleeves in pressurizers, hot leg instrument nozzles, steam generator drain nozzles, and reactor vessel head control rod drive mechanism nozzles (see Reference 16).
Pressurizer nozzles at the Calvert Cliffs Nuclear Power Plant (CCNPP) are susceptible to small leaks due to stress corrosion cracking (see Reference 3). These nozzles are mill annealed Inconel 600 and are of the welded sleeve type. PWSCC can occur in mill annealed Inconel 600 nozzles which become susceptible because of stresses induced during the nozzle to vessel weld. Therefore, it is desired to develop methods to safely repair and to respond to these leaks in a timely manner with a minimum of outage time, personnel radiation exposure and cost.
For these repairs, BG&E has developed and proposes to use design alternatives with positively constrained joints based on tube rolling and special expandable tube plugs / sleeves. Weld repairs can be used; but, result in greater personnel radiation exposure, forced outage time and preparation / implementation costs. Roll / mechanical plug repairs are also attractive in that they have no metallurgical effects on the vessel.
I In addition to development / tests performed by BG&E the roll and plug repair techniques discussed herein have been presiously developed and successfully used for other reactor coolant system applications. Such repairs have been reviewed extensively and have in the past been considered acceptable by the pertinent authorities. Roll repairs have been used in two BWR reactor vessel repairs of 26 stainless steel control rod drive housings where differential thermal expansion tends to loosen the joint much more than for the subject Inconel 600 sleeve nozzles at CCNPP. Also, roll repairs have been successfully used for repair of an instrument nozzle in another BWR reactor vessel. In addition, roll and plug repairs have been utilized in thousands of steam generator tubes in PWRs.
~
The repair designs discussed herein can be considered the same as rolled plugs or sleeves are considered for PWR steam generator tube repairs (i.e., as long as performance is satisfactory, the repair is permanent).
PURPOSFJSCOPE The purpose of this report is to provide technicaljustification to use the roll and plug type repairs described herein for any of the Inconel 600 sleeve type nozzles in the pressurizers at CCNPP-1 and CCNPP-2. Specifically, this report covers the following:
Descriptions of repairs Qualification of repairs Examinations before repair
a Repair inspection requirements Ejection prevention after repair Prior use of these repairs Overall summary References RESULTS 1.
Description of Repairs of the CCNPP Pressurizer Inconel 600 Sleeve Nozzles by Tube Rolling / Expansion Plugs The basic types of repair designs proposed herein for the CCNPP pressurizer are shown in Figures 1 and 2. These designs are to prevent leakage for the cracked nozzle configuration ofinterest per Figure 3.
Figure 1 illustrates a modification of a conventional type tube roll repair joint which seals the sleeve nozzle against the vessel by an interference fit and by surface feature deformations at the nozzle /vesselinterface. Also, this design includes an Inconel 690 retainer sleeve which has two key features:
Prevention of ejection by independent means for any possible future cracking of the original Inconel 600 nozzle.
Provision of a suitable barrier which prevents contact between the pressurizer water / steam and the original Inconel 600 sleeve nozzle inside diameter. This barrier prevents future corrosion cracking from the inside diameter of the sleeve nozzle. Future cracking of the sleeve nozzle is avoided because of the stainless steel safe end at one end and the compressive stresses and sealin the rolled / sealed area at the other end.
No stress corrosion cracking is expected within the stainless steel safe end or through the body of the rolled joint area as will be discussed later.
This type repair is planned to be used for instrument nozzles.
Figure 2 illustrates the combination of a tube roll repair plus a special plug design.
This type repair is planned to be used for a cracked pressurizer heater sleeve (if the heater can be omitted after the repair) and is considered to provide the greatest sealintegrity type repair. The same type design was used to plug instrument nozzles in the bottom of the TMI-2 reactor vessel, in order to remove metallographic sections for the Office of Regulatory Research of the NRC (see Reference 1). The design for CCNPP also includes independent means to prevent ejection as shown in Figure 2.
4
ASME Code materials will be used for the Inconel 690 plug and retainer sleeve as shown in Figures 1 and 2. The Inconel 690 plug is considered a replacement pressure retaining part and meets all pertinent Code requirements including stress analysis per Section III,1989 edition. These requirements fully meet the pertinent requirements of the original construction Code (1965 edition through Winter 1967 addenda). The retainer sleeve is not considered a pressure retaining part since the original nozzle performs this function. The nozzle ejection prevention capability of the retainer sleeve is demonstrated by tests (at greater than three times design basis ejection load).
The original Inconel 600 sleeve nozzle and its weld are left intact for each of the proposed repairs. As will be discussed later, no future cracking of structural significance is expected. These repairs provide two independent structural retention features even if the original Inconel 600 sleeve nozzle were to become severed at the original cracked area. First, the rolled and rolled / plugged joints above have been tested satisfactorily wellin excess of design pressure ejection loads. Secondly, these repair designs incorporate an additionalindependent f
retention feature as well, as shown in Figures 1 and 2. This feature is a roll i
expanded section of the retainer sleeve or plug inside the vessel. This roll expanded section will prevent ejection for any expected cracking of the original sleeve nozzle and/or its weld because of the resulting interference.
2.
Qualification of Repairs l
The repair designs in this report are qualified by tests and analyses; and, are essentially as strong as the original uncracked sleeve nozzles (due to the roll / plug i
joints). Analyses include consideration of normal operating and accident conditions including fatigue as well as any tendency for cyclic stresses to loosen the joint. Seismic loads are relatively insignificant. Vessel stresses are low and i
secondary in nature due to the interference fit at the roll / plug joints. There are no metallurgical effects on the vessel as would occur in a weld repair. The main load of concern is pressure which tends to cause ejection. Tests are used to ensure adequate ejection resistance.
Briefly, tests have been performed which cover the following which are the key parameters of interest:
Inconel 600 material (same as at CCNPP).
All existing Inconel 600 sleeve nozzle yield strengths at CCNPP.
j Two times vessel design pressure with no leakage allowed.
l Three times vessel design pressure load and no ejection.
a 5-i
a Thermal cycle to 700"F (for creep relaxation) and repeated above leakage i
and load tests (conservatively compares to 650*F maximum operating temperature).
l No permanen: deformation of the vessel.
1 Additionally, other tests on other rolled tube joints (for BWR control rod drive 1
housings) as well as actual operating experience (with steam generator rolled / mechanically expanded plugs) confirm that sleeve nonle through wall stresses within the body of these joints are compressive. No stress corrosion cracking is therefore expected within the body of the repair rolled areas described herein. Similarly, no stress corrosion cracking is expected within the expanded plug area since stresses will be compressive here also. Therefore, while some future stress corrosion cracking could be postulated to occur on wetted areas of the original Inconel 600 sleeve nonles at locations which are between the rolled / plugged area and the inboard ends of repaired sleeve nozzles (e.g., at roll transition areas) such cracking is not expected in the rolled / plugged areas used to maintain a sealed and load carrying joint (i.e., in the body of the joint). Future l
stress corrosion cracking of the original Inconel 600 sleeve nonle will not affeet the rolled region.
To help avoid any future cracking of the original sleeve nonles at roll transition areas, special attention has been given to the roll configuration and procedures, such that no cracking is actually expected within the roll / unrolled transition area.
For example, these transitions involve only gradual changes in diameter; and, the roll tooling is held in position by a fleure so as to prevent spiraling during removal. Also, existing cracks will not be significantly changed due to these roll repairs since the maximum expansion of the sleeve nozzle is limited by the bore in the vessel. Nevertheless, the repair design is based on the assumption that cracking could occur between the body of the roll joint and the sleeve nozzle weld; and, ejection is prevented even if such cracking were to occur.
One other effect of such cracking of the existing Inconel 600 sleeve nozzle is that the vessel material and the outside diameter of the sleeve nozzle will be wetted by l
reactor coolant or steam (inboard of the roll area). Accordingly, the effects of I
corrosion due to boric acid water / steam have been evaluated as follows:
I The original Inconel 600 sleeve nozzle may crack inboard of the roll area (which is not significant to the repair as discussed above).
Based on evaluations including actual operating experience with PWR reactor coolant systems, no significant corrosion of the vessel nozzle bore is expected because corrosion rates will be relatively low due to lack of oxygen. Also, once the annulus fills with corrosion products, corrosion should essentially cease. No significant hydrogen embrittlement will occur because of low corrosion rates and operations at relatively high temperatures.
l 3.
Examinations of Leaking Sleeve Type Nozzles Beibre Repair l
l Surface examinations will be performed to ensure that cracking of the sleeve nozzle is typical of previous experience (e.g., axial cracking up to about two inches long and located in/ adjacent the weld area). Based on experience thus far and i
expectations of future findings, the root cause ofleaking of these sleeve type j
nozzks has been determined with confidence to be due to stress corrosion l
cracking of the Inconel 600 and not the structural vessel itself. The repair designs
)
in this report will accommodate future cracking as discussed above.
Based on typical experience no significant erosion of the vessel is expected; l
however, erosion of the vessel will be detected / evaluated by pertinent visual inspections discussed herein.
As indicated above, no crack penetration past the Inconel material and into the vessel P3 material is expected or has been found. This is apparently due to the susceptibility of Inconel to stress corrosion cracking and the lack of susceptibility of the P3 vessel material. The deepest cracks found thus far (at St. Lucie 2) are reported in Reference 13. These cracks did penetrate through the 1-182 weld (an l
Inconel type weld material which is susceptible to stress corrosion cracking) but terminated at the I-182/P3 interface. Also, References 2 through 4, which discuss i
other instances of sleeve nozzle cracking in other pressurizers, do not indicate l
cracking into the P3 type vessel material as expected. The cracking found at CCNPP-2 was only within the sleeve and not even into the weld.
l l
No crack growth due to fatigue is expected in the vessel base material based on experience of others as well as structural evaluations per Reference 14. These evaluations were for crack growth for a reactor colant system high pressure injection nozzle at the Davis Besse plant and concluded, in essence, that fatigue crack growth from existing cracks would be insignificant. Notably, the cyclic duty on this high pressure injection nozzle is more severe than the sleeve nozzles in the pressurizer at CCNPP since the high pressure injection nozzle involves substantial thermal cycles and coolant flows which are not required for the pressurizer sleeve l
nozzles at CCNPP. Also, see Reference 15 which indicates no significant crack growth expected for PWR control rod drive nozzles (which are also Inconel 600 welded sleeve type nozz'es).
i As indicated above, we have evaluated the source of pressurizer leakage from heater sleeve and instrument nozzles and concluded that there is no indication of an impending gross failure of the pressure boundary. Since no stress corrosion cracking is expected in the vessel base material (and no fatigue crack growth as well) there is no need to inspect for crack penetration.
In summary, the nature of the pressurizer sleeve nazzle cracking is sufficiently well understood to ensure a slow increase in leak rate and detectable Icakage, by the normal reactor coolant leak detection systems, long before any possibility of a major leak (see Reference 3). Finally, the repair designs.in this report are o'
satisfactory irrespective of the co..dition of the existing Inconel 600 sleeve type nozzles. This conclusion is based on suitable inspections, qualification tests / analyses and hydrostatic leak tests at installation, as discussed herein.
Therefore, additional tests to determine the cause, nature, extent and location of the cracking are not needed.
l 4.
Inspection Requirements to Ensure Repair is Sntisfactory for Continued Service After a repair and before the plant is operated a post repair system hydrostatic pressure / leak test will be performed per existing procedures.
l After the plant has operated with a repaired pressurizer sleeve type nozzle, we consider the best type inspection to be a visual inspection for boric acid deposits.
l Also, if iron rust strains are present and appear to be recent, this can indicate corrosion of the vessel; and, any dripping of water (not associated with condensation or other sources) also indicates possible leakage.
To perform a valid inspection, the insulation must be such as to make the boric acid deposits or water drippage visible. Uor example, the insulation must either be well vented at least around the nozzie; or, it will be removed. Also, boric acid deposits due to leaks from the steam region of the pressurizer will contain only about 10% of the boric acid which water will contain. Accordingly, visual inspections for steam leaks will require more care since the amount of boric acid will tend to be less from a steam leak.
a.
Inspection Frequency Inspection frequency will be unchanged from that required fc non-cracked pressurizer sleeve nozzles since the repaired sleeve nozzles ar: considered to pr^ vide at least the same integrity as a non-cracked sleeve nozzle.
b.
I ch; o a t y pection Acceptance Criteria.
The repair designs discussed herein, with roll and plug joints with a mechanical metal-to-metal seal, provide a seal essentially the same as in a valve bonnet with a metal gasket seal. This mechanical sealjoint replaces the welded vessel to nozzle seal joint. Leakage from a repaired pressurizer sleeve nozzle is evaluated herein in a similar manner as for any metal gasket mechanical seal and is concluded to be not in siolation of the Technical Specification 3.4.6.2.a for CCNPP. Specifically, this Technical Specification can be summarized as follows:
No " pressure boundary leakage" is allowable.
" Pressure boundary leakage is defined as leakage (except steam generator tube leakage) through a non-isolable fault in a reactor coo &. t system component body, pipe wall, or vessel wd."
a' 1
1 Reason: " Pressure boundary leakage of any magnitude is unacceptable since it may be indicative of an impending cross failure of the pressure boundarv-and requires-promptly placing the unit in cold shutdown."
i Based on the above, we conclude that pressurizer sleeve type nozzle leaks for the renairs described herein should not be considered as a leak through a " component body, pipe wall or vessel wall" but through a mechanical seal joint. Leakage through a mechanical seal joint is not indicative of an impending gross failure. Therefore, leakage from these repairs, with mechanical seals, is not in violation of the Technical Specification since the Technical Specification is not applicable to such leakage, but applies to faults in a reactor coolant system component body, pipe wall or vessel j
wall. Finally, safety evaluation indicates no problem in any of the pertinent areas based on:
Reactor coolant boundary integrity i
Plant operation Safe shutdown In essence, leakage from pressurizer sleeve type nozzles as repaired herein, j
is not a safety concern nor is there any impenaing gross failure because:
Defects are not in the vesscl wall or body (no structural effect on integrity of vessel).
Inspection / leakage criteria for monitoring the area in question ensures no significant structural degradation of the vessel based on past experience with margm.
Leakage is small and can be monitored (not a run-away leak type problem).
Plant's existing Technical Specification leakage rate limits will allow safe shutdown if leaks should grow to any significant size.
There will be no ejection of repaired sleeve nozzles due to multiple special design features. (The roll repair and plug repair are sufficient to resist ejection even if existing nozzle cracking were postulated as 100% through wall for entire circumference.)
Even if :jection occurred, leakage is limited to previously analyzed accidents.
1 1
a To fully avoid the possibility of vessel erosion /cerrosion, virtually n i
leakage is permissible. To discriminate between true and possibly false indications of a leak and to ensure any such leakage is insignificant, the following acceptance criteria will be used. These criteria are applicable for hydrostatic leak test of a repaired sleeve nozzle immediately after repair as well as after subsequent operation of the plant.
i These leakage inspection criteria are developed / justified in Appendix A.
Based on this, the controlling parameter is the avoidance of damage to the carbon steel vessel due to erosion / corrosion mainly for commercial (and not safety) reasons. We plan the following minimum criteria be used to determine if a leak has been present during operation or is present during hydrostatic leakage testing and whether the nozzle is satisfactory for continued senice:
)
3 Greater than.1 in of boric acid deposit (found per inspection interval). We estimate this is less than.2% of the boric acid found after the water leak from a heater sleeve nozzle in the pressurizer at another plant in 1987 (this leak resulted in a small amount of vessel damage which was weld repaired). See Figure 4. Similarly it is only about 2% of the boric acid deposit we estimate was found at the bottom of the Calvert Cliffs Unit 2 pressurizer (which resulted in essentially no vessel damage). See Figure 5.
9 Greater than one drop per minute ofliauid (or condensed steam) leakage. This is about equivalent to the boric acid criteria discussed above (and has about the same large safety margin).
Any evidence of vessel damage due to erosion / corrosion. Will be evaluated on a case basis.
1 5.
Ejection Prevention Although future cracking (away from the original weld cracking area) is not expected because of special care in the roll transition areas, redundant support is provided to prevent ejection even if circumferential cracking were to occur in any region of the original sleeve type nozzle. As shown in Figures 1 and 2, redundant support is provided by the Inconel 690 retainer sleeve (for an instrument nozzle j
repair) and by the Inconel 690 plug (for a heater sleeve repair) fer any expected potential cracking of the sleeve nozzle or its weld. Thus, the repair designs are supported against ejection by an Inconel 690 member which is expected to be resistant to such cracking. See References 17 and 18 for test results which indicate such resistance even with crevice geometries, the use of cold worked Inconel 690 and stress / deformation conditions simulated by double "U" bend test specimens as well as rolled and kinetically welded joints. Also actual in-service use of Inconel 690 with short tube /tubesheet crevices at the top of the tubesheet in steam generators has not indicated a corrosion problem thus far... -
In addition to the above, the interference fit resulting from the tube rolling / plugging repair designs covered herein is fully capable of resisting ejection independent of all other means for all normal operating conditions. Specifically, qualification testing has been performed to confirm these interference fits are fully capable of sustaining vessel design pressure loads multiplied by a factor of 3 i
(i.e., 2500 psi x projected area x 3).
1 Finally, to help put the question of pressurizer sleeve nozzle ejection into perspective, the maximum blowdown rate from such an event is roughly equivalent to rupture of only about one steam generator tube. This is mainly because the j
break flow area from ejection, e.g., of a heater sleeve nozzle, is only due to a single ended break versus a double-ended pipe break for a steam generator tube-j rupture. Also, similar accidents have been evaluated by pertinent accident j
l analyses (see Reference 4 vgarding analysis of a stuck open pressurizer PORV and Reference 5 for other small break analyses).
6.
Prior use of these Repairs The roll repairs described herein have been reviewed / approved and used satisfactorily in the following domestic BWR's:
(a)
Oyster Creek in-core penetration housing in lower head of reactor
)
vessel (in 1974).
]
(b)
Big Rock Point control rod drive (CRD) penetration in lower head of reactor vessel (in 1979).
j (c)
Nine Mile Point Unit 1 CRD penetrations (25) in lower head of i
reactor vessel (starting in 1984. most recent in 1993).
The first domestic rolled-joint repair to a reactor vessel sleeve type penetration was performed in 1974; the most recent in 1993. These repairs have all
]
performed satisfactorily. NRC approvals for these applications were based on pertinent safety evaluations. NRC authorized Niagara Mohawk Power Corporation, initially per a Safety Evaluation and subsequently per 10CFR50.55a(a)(3), to utilize a roll repair alternative (see References 19 and 20).
i j
Leakage experience for the above repairs has been satisfactory. Notably, at NMP-1, the first few repairs with a relatively short roll joint did leak a small l
amount (with no serious consequence). All subsequent repairs at NMP-1 j
(including rerolling to achieve a relatively longer roll joint) have not indicated j
leakage (even after many years of service). The same good performance can be expected for CCNPP especially since thermal expansion effects at CCNPP will be j
substantially less. At CCNPP, the sleeve nozzles are Inconel which has a similar expansidty to the vessel material (versus a stainless steel sleeve nozzle at NMP-1).
)
Also, these sleeve nozzles at CCNPP are not subject to substantial thermal
- 1 i
s I
4 I
~
.._,..._._,..._.,..U
transients (versus NMP-1 which is subject to cooling water and periodic scram flow, as well as normal reactor operating temperatures).
With regard to senice experience with steam generator tube plugs, thousands have been installed; and all but a few have performed satisfactorily. For the few steam generator tube plugs that have not been satisfactory, the reasons for their inadequacies can be summarized as follows:
l Improper installation In-senice cracking above and below (but not within) the body of the expanded region.
7.
Overall Summary The following is an overall summary of the main issues covered in this report:
l The root cause of leakage from these pressurizer sleeve nozzles is stress I
corrosion cracking.
The cracking is limited to the Inconel 600 sleeve nozzle and is not in the pressure vessel.
l The nature of the cracking is well understood.
The repairs discussed herein virtually stop leakage and redundantly prevent ejection.
The basic roll repair approach for pressurizer sleeve type nozzles at CCNPP has been previously accepted and applied in the field in BWR reactor vessel nozzles (i.e., a historical precedent is available).
i The mli repair technique used successfully for BWR stainless steel nozzles is a more severe application than for PWR Inconel nozzles because of differential expansion effects.
Experience to date with the roll repair process has been very good.
The repair reduces personnel radiation exposure, outage time and cost as demonstrated by previous applications.
Overall, the repair approach discussed herein is considered to provide an acceptable level of quality and safety.. -
a REFERENCES 1.
"TMI-2 Reactor Vessel Lower Head Sampling Project Techniques and Results,"
Cole, N. M.; Lipford, B. L; Friderichs, T. J., MPR Associates, Inc., Washington, D.C., September 1990.
2.
" Cracking in Pressurizer Instrument Nozzles," presented during EPRI Workshop on Circumferential Cracking of Steam Generator Tubes, Charleston, February 1990.
3.
Calvert Cliffs Nuclear Power Plant " Basis for Determination, Unit 1 Safe Operation Relative to Unit 2 Pressurizer Heater Sleeve Leakage," September 18, i
1989.
4.
Licensee Event Report, Arkansas Nuclear One, Unit Two, " Pressurizer Heater Rupture Results in Reactor Coolant Pressure Boundary leakage and Pressurizer Vessel Corrosion," Rev.1, May 26,1987.
5.
Chapter 14 of " Updated Final Safety Analysis Report" for CCNPP-1 and 2.
6.
Nuclear Power Emerience Vol.1 PWR-2, VII. Safety Systems, A. ECCS, p.150, Paragraph 522, "HPI Nozzle, RCS Pipe Corroded - Unchecked Boric Acid Leakage from isolation Valve Concentrated Under Insulation," Arkansas One 1 -
Oct. '86 - Refueling PWRs in General - Dec. '86.
7.
Nuclear Power Emerience Vol.1 PWR-2, V. Reactor Coolant System, A. Pumps,
Calvert Cliffs 1 - Nov. 80 - Refueling.
8.
Nuclear Power Emerience Vol. PWR-2, III. Reactor Vessel, p.16, Paragraph 35, "RCS Pressure Boundary Degradation - Reactor Vessel Nozzle Flange Leak -
Degraded Flange Gasket Material-Boric Acid Corrosion," Arkansas One 1 -
Dec. 89, Cold Shutdown.
9.
Nuclear Power Emerience Vol.1 PWR-2, V. Reactor Coolant System, E.
Pressurizer, p. 34, Paragraph 119, "RCS Boundary Degraded - Corrosion from Boric Acid Leak - Pressurizer Heaters Ruptured - Inadequate Component Design," Arkansas One 2 - Apr. 89 - 100E 10.
Licensee Event Report, San Onofre Nuclear Generating Station, Unit 3,
" Pressurizer Instrument Nozzle Pressure Boundary Leak," Jan. 29,1987, Rev. 2.
1L United States Nuclear Regulatory Commission (Generic Letter 88-05) dated March 17,1988," Boric Acid Corrosion of Carbon Steel Reactor Pressure Boundary Components in PWR Plants." -
12.
NUREG/CR-2827, " Boric Acid Corrosion of Ferritic Reactor Components,"
July 1982.
13.
Letter from Albert F. Gibson, U.S. Nuclear Regulatory Commission to J. H. Goldberg, Florida Power & Light Co., April 13, 1993;
Subject:
" Notice of Violation (NRC Inspection Report No. 50-335/93-08)," Docket Nos. 50-335, 50-89.
i 14.
Ixtter from Donald C. Shelton, Toledo Edison to U. S. Nuclear Regulatory Commission, September 12,1990,
Subject:
"High Pressure Injection / Makeup Nozzle Updated Fracture Mechanics Analysis," Docket No. 50-346.
15.
Letter from NUMARC (Alex Marion) to U.S. Nuclear Regulatory Commission (William T. Russell), dated June 16,1993 (including safety evaluations of cracked control rod drive nozzles).
16.
Proceedings: 1991 EPRI Workshop of Alloy 600 in PWR's; Electric Power Research Institute EPRI TR-100852.
17.
Boshoku Gijustsu (Corrosion Engineering), Vol. 28, No. 2,1979, "Inconel 690: A New High Nickel Alloy for Corrosive Environments at Elevated Temperature,"
A. John Se.driks, J. W. Schultz and M A. Cordovi, The International Nickel Co.,
Inc.
18.
Proceedings of the Fourth International Symposium on Emironmental Degradation of Materials in Nuclear Power Systems-Water Reactors, August 6-10,1989, " Qualification of Kinetically Welded Alloy 690 Sleeves for PWR Steam Generators," J. M. Sarver; J. V. Monter; J. M. Helmey, (Babcock &
Wilcox Company).
19.
Letter from U.S. Nuclear Regulatory Commission (Domenic B. Vassalo) to Niagara Mohawk Power Corporation (B. G. Hooten) dated June 29,1984;
Subject:
" Control Rod Drive penetration Leakage from Stub Tube Cracking."
20.
Ixtter from U.S. Nuclear Regulatory Commissien (Rajender Auluck) to Niagara Mohawk Power Corporation (C. V. Mangan) dated March 25,1987;
Subject:
" Request to Utilize an Alternative to the Requirements of 10 CFR 50.55a(g) (TAC 61181). -
o' NON PROPRIETARY VERSION l
t t
INSTRUMENT N0ZZLE REPAIR DESIGN FIGURE 1 Proprietary information omitted.
0
~
NON PROPRIETARY VERSION
\\
1 I
i l
i 1
)
l 1
\\
1 I
l t
i i
1 l
l HEATER SLEEVE REPAIR DESIGN l
FIGURE 2 Proprietary information omitted.
l l
~.
i t
Clad
-i l
Existing Weld i
k Existing Weld Butter (e
\\
s
/
N Typical Location of.
/
In-Service h du t.
- g
/
Stress Corrosic,
/
Cracks '
/
/
i
/
/
/
s 7
/
/
i
/
Typical Leak Path Via Annular
/
Clearance With Cracked Sleeve -
l
/
/
/p Sleeve
/
Vessel Structural Wall
/
/
?"%d (1) Note: Only One of Several Different Weld Configurations Used is Shown Above
]
TYPICAL CRACK LOCATION IN EXISTING DESIGN WELDED SLEEVE T(PE NOZZLE (1) i F-90-84-31 FIGURE 3 9/21/93 j
~ _ _.. _... -, _ -,. _.. _. -......... _.. _
I l
l i
1 i
Boric Acid Heater V
t j
sj
(*c
^
s I
]
k
, ilk}-
y l,f' l
7 h;7 y 4; W-j j.,,
l l
l BORIC ACID ACCUMULATION OBSERVED BEFORE REMOVAL OF THE INSULATION AFTER HEATER SLEEVE LEAKAGE FROM
)
THE PRESSURIZER AT ANOTHER NUCLEAR POWER PLANT (SAME AS CALVERT CLIFFS)
[o'((f9[
FIGURE 4
l r
l Boric Acid (Outside insulation) er
,o
,v, x.
e d
. fl
,.b L
g
)
~
/kW A J";%
Vi
,N.
4
. y'yi/g!..y%gy
.c d
1 x
$^4pi y,.
,}$j
- J jj l kQ '
k
$[gh;M '
g 34,
!<.. N
., A 4
, af 4
Q'b;%;$_
a, Insulation Pressurizer Vessel Boric Acid
)
(Inside insulation)
CALVERT CLIFFS UNIT 2 PRESSURIZER INSPECTION FOR LEAKAGE
-(SHOWING BORIC ACID INSIDE AND OUTSIDE INSULATION)
F-90-92-5 10/4/93 FIGURE 5
1 a."
MMPR A S S O C I AT E S IN C.
ENGtNEERS Appendix A i
LEAKAGE INSPECTION CRITERIA i
FOR REPAIRED SLEEVE NOZZLES l
H i
?
Leakage inspection criteria have been developed to discriminate between true and posely false indications of a leak and to ensure any leakage is insignificant. Further, in case such " false" indications were an indication of an actual leak and to ensure any leakage is insignificant, we have evaluated the possible effects to ensure even these effects would be acceptable. Specifically, leakage inspection criteria have been developed by considering the various pertinent requirements as follows:
i 1
1.
About.4 gpm - based on a typical Technical Specification limit of 1 gpm unidentified leakage and typical plant leak rates up to about.6 gpm unidentified l
leakage.
2.
About 44 gpm - based on a typical normal makeup flow rate of about 44 gpm for each of three charging pumps.
l 3.
About 15 millionths of a com (1 drop per minute)--to provide a very large margin to ensure that there will be no damage to the vessel due to erosion / corrosion. As-indicated, this is the most restrictive criteria. Following is a discussion of the facts j
in support of this criteria.
Leakage of borated reactor coolant can, under certain conditions, cause substantial wastage of carbon steel vessels, pipes and bolting. Based on our review of data from actualin-plant incidents of such reactor coolant leaks (see References 4 and 6 through 12), we conclude the following:
Typically, in cases where wastage has occurred, time durations involve several months ofleakage (rather than a matter of days or even a few weeks, e.g.).
Wastage has always occurred in pipes and vessels, so as to be detectable by visual inspection from outside the vessel or pipe. No substantial effects of corrosion of the carbon steel are expected due to contact with borated reactor coolant unless damage is visible on external surfaces.
Substantial wastage can occur even at very low leak rates (as low as about
.001 gpm which is equivalent to about 60 drops per minute leakage (for an
.18 inch diameter drop size).
A-1 1
e
' Jr '
'6 In all cases we evaluated, any substantial wastage was associated with finding a substantial amount of boric acid deposits on nearby structures, insulation or equipment. In fact, we found no cases of vessel or pipe wastage where only small amounts of boric acid were found.
In cases where wastage has occurred, iron oxide stains have sometimes been found.
In no case have we found that the pressure vessel or pipe in question was even close to being damaged sufficiently for rupture to occur.
Accordingly and based on the above, we conclude that the acceptance criteria for possible leakage of the pressurizer at CCNPP should cover two parameters:
Only a low leakage rate is permissible (less than one t ap per minute).
3 Only small amounts of boric acid deposits are permissible (less than.1 in per inspection interval).
i We consider the amount of boric acid deposits found on nearby surfaces to be particularly important since this covers effects of leakage time as well as leakage rate.
These criteria may be conservative in cases where the leak is too small to maintain a liquid state necessary for high corrosion rates; however, it will not be nonconservative, provided the boric acid deposits are readily detectable during inspections. Finally, we consider it reasonable to expect that if a leak is indicated at as low as one drop per minute during hydrostatic test (per the criteria in this report) that leakage will be zero after power operation since such a small leak path will probably become stopped up with corrosion products. In any case, these conservative criteria ensure adequate safety margin for the vessel if the mechanical sealjoint were to leak.
i i
1 l
A-2
.. _.