ML20214U833
| ML20214U833 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 11/25/1986 |
| From: | Wilson R GENERAL PUBLIC UTILITIES CORP. |
| To: | Zwolinski J Office of Nuclear Reactor Regulation |
| References | |
| TASK-03-05.B, TASK-3-5.B, TASK-RR 5000-86-1086, NUDOCS 8612090415 | |
| Download: ML20214U833 (8) | |
Text
O GPU Nuclear Corporation NUOIM7 100 Interpace Parkway Parsippany, New Jersey 07054-1149 (201)263-6500 TELEX 136-482 Wnter's Direct Dial Nurnber:
November 25, 1986 5000-86-1086 Nr. John A. Zwolinski, Director BWR Licensing Directorate #1 Division of BWR Licensing U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Zwolinski:
Subject:
Oyster Creek Nuclear Generating Station Docket No. 50-219 License No. DPR-16 Postulated High Energy Line Break Within Isolation Condenser Penetrations This letter provides additional information concerning the subject issue, supplementing our letter dated September 17, 1986 on the same issue, and subsequent telept.one reviews in October and early November.
As a result of these discussions concerning the isolation condenser piping containment penetrations, we would intend to coordinate final isolation condenser piping containment penetration resolution with other issues related to the piping in the penetratiors: NUREG 0313, Revision 2 " Technical Report on Material Selection and Pmc.essing Guidelines for BWR Coolant Pressure Boundary Piping", and Systewcic Evaluation Program (SEP) Topic III-5.B. The concern in NllREG 0313, Revision 2, is with the welds inside the penetrations that are very difficult to inspect. SEP Topic III-5.8 is concerned with the two series outside containment isolation valves which are located in each of two isolation condenser steam supply lines at the penetrations. Any modifications resulting from the resolution of these issues should be implemented as a single integral change for each isolation condenser penetration. A physical modification, if needed, cannot be performed during the current Cycle llR refueling outage because of the time needed to finalize the necessary evaluations and do the required engineering.
Any modification will be scheduled for the next (Cycle 12R) refueling outage.
However, this schedule is dependent on the NRC staff review of previous GPU Nuclear submittals concerning SEP Topic III-5.B and the issuance of NUREG 0313,
)\\
8612090415 061125 DR ADOCK 0500 9
GPU Nucl]a Corporation is a subsidiary of GeneralPublic Utilities Corporation
~.
Revision 2, early in 1987.
This time constraint is needed so that project scope can be finalized, planning initiated and engineering documentation released for construction six months prior to the start of the Cycle 12R 1
outage.
It is our goal to have all Cycle 12R outage-related engineering released for construction six months prior'(based on the current schedule this is October 1,1987) to the outage in order that construction planning, material procurement, etc., can be finalized prior to the outage. This will allow outage work to proceed in an efficient and effective manner.
A justification for operation in Cycle 11 without modification,of the l
isolation condenser piping penetrations was provided in our Sep.tember.17, 1986 letter. Analysis results since the September 17, 1986 letter provide a more caccurate understanding of the conditions existent at the piping penetrations.
j i
A very recent unverified finite element analysis shows that gusset reinforcement (not considered in previous work) at three of the four penetrations significantly reduces the stresses on the drywell shell. We i
believe that further analysis,_ using more realistic isolation condenser steam line thermal-hydraulic modeling, will confirm that the drywell shell with l
gusset reinforcement at the steam line penetrations will not be overstressed 1
due to a HELB. We have also concluded that diagnostic techniques can be used to verify the source of an increase in unidentified leak rate if it is from a
~through-wall crack in a condensate return line weld in the penetration.
These i
measures will be initiated upon a 2 gpm increase in unidentified leak rate i
above an established baseline leak rate.
If our investigation cannot reasonably assure the absence of a leak in an isolation condenser condensate i
return line weld in the penetration, then a plant shutdown will commence to further investigate, analyze and repair as necessary.
The attachment to this letter contains the additional information reauested by the NRC staff, current analysis results and our intended action in response to l
f a 2 gpm increase in unidentified leak rate. Also included is further discussion concerning the isolation condenser penetrations.
We would like to meet with the NRC staff within several weeks following plant restart to initiate discussions concerning final resolution of the isolation condenser piping penetration issues.
If you should have any questions j
concerning the information provided in this letter or in the attachment, i
please contact the undersigned or Mr. M. W. Laggart of my staff at j
(201) 263-6205.
Y y tr ly yours, g%-
(,ilon
. F.
Vice President i
Technical Functions i
RFW/PC/ a cc's on next page 1388g t
i
_______j
a cc: Dr. Thomas E. Murley, Mministrator Region I U.S. Nuclear Regulatory Commission 631 Park Avenue King of Prussia, PA.
19406 NRC Resident Inspector Oyster Creek Nuclear Generating Station Forked River, N.J.
08731 Mr. Jack N. Donohew, Jr.
U.S. Nuclear Regulatory Consission 7920 Norfolk Avenue Phillips Building, Mail Stop 314 Bethesda, Maryland 20014 4
f 4
U J
a 1
i l
l 1
1388g i
-I i
.O ATTAC MENT Additional Information Requested by the NRC Staff Material and fabrication:
Isolation condenser 10 inch piping is 316 Schedule 80 stainless steel. The 18" x 10" flued head fittings are forgings also made of 316 stainless steel. Welding of the stainless steel pipe to the flued head penetration fittings was performed in the field for the steam supply lines and in the shop for the condensate return lines. No stainless steel butt welds were solution-annealed following either field or shop f abrication.
Stresses: The previously reported initial calculation of stresses in the drywell shell due to a postulated HELB was based upon loading derived from a thermal-hydraulic analysis of flow from a postulated guillotine-type rupture in the isolation condenser condensate piping weldment within the penetration.
The hydraulic model assumed the flow source was condensate at normal operating reactor coolant temperature and pressure exiting at the rupture location as two-phase flow from both ends of the pipe.
This corresponds to the condition existent at the condensate return line penetrations with the Isolation Condenser System in service. Biljaard analysis with this hydraulic load (and no containment shell reinforcement)' indicated primary local membrane (P ) +
L bending stress (P ) intensity exceeding the Service Level D allowable for b
Class MC components.
Per ASME Section III, the Service Level 0 allowable for PL+Pb shall be the greater of 1.8 Sac or 1.5 Sv, which amounts to 64 Ksi, utilizing actual CB4,I material test data foF the drywell shell.
A walkdown was performed on September 27 and 28,1986 and the data gathered during the walkdown confirmed that stiffening plates and gussets at three of the four penetrations (gussets and stiffening plates installed on the inside of the drywell shell at penetrations X3A/B (steam line penetrations) and X58 (condensatereturnlinepenetration))provideadditionalreinforcement.
Consequently, an evaluation of the new penetration geometry has been performed to more realistically calculate the stress level. A finite element analysis shows that the additional reinforcement significantly reduces the stress on the drywell shell. This analysis indicates a PL+Pb stress intensity of approximately 75 Ksi for the three reinforced penetrations, exceeding by 11 Ksi the allowable of 64 Ksi.
The hydraulic model discussed above which was used to obtain the load on the i
penetrations, takes into account the hydraulic conditions during a postulated l
HELB in a condensate return line penetration. We are currently in the process
}
of revising the model to accurately account for the conditions expected in a steam line penetration where steam exits from the pipe on one side of the j
rupture location while condensate exits from the other. We believe that the fluid momentum and pressure effects in a steam line penetration are reduced such that the resultant load on the penetration will not overstress the f
drywell shell considering the reinforcement provided at those two penetrations.
Due to the coglex nature of the hydraulic analysis we do not j
l
?
1388g i
t
_ _. _,. _ _ - _,. ~. - _ _.. _ _
- c,
expect complete results for two to three weeks. However taking into account the conservative nature of the bounding hydraulic analysis and the reduced i
stresses provided by the penetration reinforcement, we believe that the stress intensity on the drywell shell at the isolation condenser steam line j
_ penetrations will not exceed allowable during a postulated HELB.
i
}
Both condensate return line penetration nozzles have stiffening plates at their intersection with the drywell shell, however, only one of these penetrations benefits from gusset reinforcement. As stateri earlier, the load on a penetration with stiffening plates and gussets results in a drywell shell j.
stress intensity of about 75 Ksi.
For the condensate return line penetration j
without gussets the drywell shell stress intensity is as previously reported j
(about 120-130 Ksi). All analyses performed above are being reviewed and are
[
l subject to verification.
~
j In our September 17, 1986 letter, we stated that the allowable stress criteria j
for jet impingement loads was 0.9 Sy. This is true for original design.
i However, in Amendments 50 and 51 of the Oyster Creek Facility Description and e
l Safety Analysis Report (FDSAR) refined criteria were utilized to reanalyze i
penetration loading. The subject FDSAR amendments, which are incorporated by reference in the updated FSAR, state that the allowable stresses for jet impingement loads due to a postulated HELB on or in the penetration shall not 4
]
exceed the following:
Local membrane stress - 1.5 Sm 1
Local membrane + secondary membrane + secondary bending - 3.0 Sm j
Where Sm is the allowable tensile stress for the specified material per 1
Section VIII of the ASME Boiler and Pressure Vessel Code.
In performing the recent reanalysis, it was considered appropriate to use the updated version of the above criteria.
4 Crack Growth and Leak Rate Analysis: GPU Nuclear Technical Data Report (TDR)
No. 467 was previously submitted to the NRC by letter dated October 16, 1984 in response to SEP Topic III-5.B for Oyster Creek. This TDR presented a crack growth and leak rate analysis of isolation condenser piping outside i
containment. We have recently performed a preliminary crack growth and leak j
rate analysis of isolation condenser 10-inch condensate return piping inside containment using the same methodology. However, this analysis utilized pipe stresses determined during the recent Bulletin 79-14 reanalysis.
The stresses are lower for the piping inside containment.
The initial through-wall crack l
1ength was assumed to be eaual to 2t (t = wall thickness) with the crack i
l 13889 i
.. propagating to 90' of the pipe circumference.
Crack growth and leak rate results for 10-inch condensate return piping inside containment are shown below:
TIME CRACK LENGTH LEAK RATE (MONTf6)
(INCHES)
(GPM) 0 1.19 1.0 12 1.28 1.1 24 1.39 1.2 36 1.51 1.2 48 1.64 1.4 60 1.80 1.5 72 1.98 1.6 84 2.19 1.8 96 2.44 2.0 108 2.73 2.3 120 3.08 2.6 132 3.51 2.9 144 4.03 3.3 156 4.68 3.9 168 5.50 4.6 180 6.56 5.4 191 7.85 6.5 13889
... The results show that the crack growth rate is significantly slower for the inside containment piping compared to the outside containment piping due to the lower stresses. The instability of a crack was shown not to occur up to I
90* of pipe circumference in our previous analysis using pipe material properties. Recently, when more conservative properties (weld material) were l
considered we agree with the NRC staff that crack instability is possible at a conservative lower bound of approximately 4 inches. The rate of crack growth is not affected by material properties.
Oyster Creek Technical Specifications (TS) require plant shutdown if the unidentified leak rate exceeds 5 gpm total (TS 3.3.D.1.a) or if an increase of 2 gpm or greater is detected in a 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> period while operating at steady state power (TS 3.3.D.1.c).
A GPUN administrative procedure requires corrective actions and investigation if the unidentified leakage rate exceeds 4 gpm. An alarm annunciates in the control room when the unidentified leakage rate is 4 gpm or greater.
Nomally, the plant operates with approximately 0.5 to 4 gpm unidentified leakage over the fuel cycle based on past experience.
Thus, there could be up to 3.5 gpm between a normal startup leakage rate and the point at which the alarm annunciates. A leak rate flow integrator can measure a minimum of I gallon in a 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> period. This is considered sufficiently sensitive to assess an increase in unidentified leak rate. There is also a backup leak rate recorder in the control room which continuously reccrds unidentified leak rate to the nearest 0.2 gpm.
An additional restriction will be imposed during operation in Cycle 11 in order to alleviate the concern of a postulated HELB in the condensate return line penetrations. The action required by TS 3.3.D.3 will be taken if unidentified leakage increases 2 gpm or more in a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period as well as the 4 hour4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> period currently in TS 3.3.D.1.c to provide additional conservatism. This will ensure that corrective actions are initiated based on investigation of somewhat slower developing increases in unidentified leak rate.
Unidentified leak rate increases of 2 gpm or more over a longer period, i.e.,
weeks or months, will be investigated in an attempt to ascertain the source.
This change in leakage will be that which is above the baseline unidentified leak rate.
The baseline unidentified leak rate will be established after operating at steady state power with full recirculation flow when leak rate has stabilized.
If an increase of 2 gpm or more is detected and investigation does not provide reasonable assurance of the absence of a through-wall p'pe crack at either of the two condensate return line penetrations then a plant shutdown will commence to investigate, analyze and repair as necessary. An unidentified leak rate increase which is not attributed to the penetration location by investigation with the plant operating or after plant shutdown will be used in the establishment of a new baseline leak rate.
In order to avoid an unnecessary plant shutdown and subsecuent restart, with its attendant deleterious effect on plant components, a period of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> would be needed after the 2 gpm increase in order to allow a thorough investigation and evaluation. Since crack propagation would occur a*. a very slow rate, this time for evaluation is acceptable.
1388g
. Normally, unidentified leak rate is closely monitored and trended.
Experience has shown that various souices of unidentified leak rate increases can be attributed to certain systems or components. Radiochemical analysis of a sample of drywell sump fluid can provide some indication of unidentified leakage sources, i.e., feedwater, reactor coolant or closed cooling system water. Corrective actions such as isolating a recirculation loop or electrically backseating motor-operated valves in containment prnvide a means to verify and/or minimize various leakage increases. We are currently planning to install thermocouples on the isolation condenser condensate return line piping both inside and outside containment prior to restart.
The absence of a significant increase in the temperature differential above baseline (which would result from heatup of the in-containment piping) would provide an indication, coupled with confirmation of low Isolation Condenser System process flow, that a through-wall crack does not exist at the penetration.
The baseline temperature differential will be established within the first week of stable power operation. The heatup of the piping would be caused by hotter reactor coolant coming from the recirculation line (due to the through-wall crack) to which the condensate return line connects.
In addition, periodic monitoring of these thermocouples will be performed.
In conclusion, we believe, based upon our preliminary reanalysis that loads resulting from a postulated guillotine-type HELB in either isolation condenser steam line penetration will not result in exceeding the Code allowable stress for the drywell shell. Final confirmatory analyses are underway.
Therefore, our preliminary assessment is that an unreviewed safety auestion does not exist for these penetrations and our FSAR commitment is met.
The calculated stress intensity at the drywell shell for the condensate return line penetrations is significantly less for the one gusset-reinforced penetration, however, the stress intensity due to the postulated HELB exceeds the Code allowable for both. Therefore, our original assessment for these penetrations remains valid and an unreviewed safety auestion still exists as the stress exceeds the design basis. However, adeauate compensatory measures as discussed above will render highly improbable operating with a crack of critical size.
Therefore, the possibility of a HELB resulting in overstressing the isolation condenser condensate containment penetration is even nore highly improbable and operation of the plant through Cycle 11 is acceptable.
1388g
.