ML20214P403
| ML20214P403 | |
| Person / Time | |
|---|---|
| Site: | Yankee Rowe |
| Issue date: | 05/27/1987 |
| From: | Papanic G YANKEE ATOMIC ELECTRIC CO. |
| To: | Mckenna E NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM), Office of Nuclear Reactor Regulation |
| References | |
| TASK-03-05.A, TASK-3-5.A, TASK-RR FYR-87-054, FYR-87-54, NUDOCS 8706030363 | |
| Download: ML20214P403 (14) | |
Text
.
s Telephone (617) 872-8100 TWX 710 3807619 YANKEE ATOMIC ELECTRIC COMPANY
. IEkb 1671 Worcester Road, Framingham, Massachusetts 01701
' YANKEE v.
May 27, 1987 FYR 87-054 United States Nuclear Regulatory Commission Document Control Desk Washington, DC 20555 Attention:
Ms. Eileen M. McKenna, Project Manager Project Directorate I-3 Division of Reactor Projects I-II
References:
(a) License No. DPR - 3 (Docket No. 50-29).
(b) Letter, USNRC to YAEC, dated February 20, 1987 (c) Letter, USNRC to YAEC, dated March 13, 1987
Subject:
SEP Topic III-5.A
Dear Ms. McKenna:
References (b) and (c) contain your remaining set of questions regarding SEP Topic III-5.A.
Attached are the responses to those questions.
As a result of our analysis and subsequent phone conversations with the staff, Operating Procedure No. OP-4232 will be revised to include a visual inspection of the 5-inch bypass line as part of the required periodic inspection of the Vapor Container which is normally performed bi-weekly. The bypass line will also be inspected whenever an unidentified main coolant leak is detected in j
the Vapor Container.
This committment is contingent upon satisfactory resolution of this issue.
It is our intention to revise this procedure after the staff has issued the Safety Evaluation Report for this topic.
We trust that you will find this information satisfactory; however, if you l
have any questions please contact us.
1 Very truly yours, YANKEE ATOMIC ELECTRIC COMPANY a4"4 G. Papa e, Jr.
Senior Project Engineer Licensing CP/gbc cc USNRC Region I USNRC Resident Inspector - YNPS
\\
B706030363 870527 l
PDR ADOCK 05000029 G
PF@ -
t
Attachment Letter 0570-043-002 May 22,1967 Page 1 of 13 REVIEW AND RARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INOi RECIRCULATION CROSSOVER LINE AT THE YANKEE NUaEAR POWER STATION Comment 1 Additional justification is required before the flaw length used in the Level D integrity analysis can be accepted for the leak before break evaluation. The staf f guidance indicates that the flaw length used in the Level D evaluation should be twice the length of a flaw that will be readily detected during normal operation.
Based on this criterton, confinn that the leakage corresponding to a flaw one-hal f the length used in the Level D evaluation can be detected reliably during normal operation using in-plant leakage detection equipment.
Describe the action to be taken when the applicable unidentified leak rate is detected.
Cl arification Detectable Crack Size and Integrity In the analysis described in the LBB report, a 3.0-inch-long crack was evaluated for detectability under normal operating conditions and for integrity under Level D conditions.
New detectability and integrity evaluations have been performed to follow the staff guidance that leak detection sensitivity should be based on plant experience and that the detectable crack size should be doubled for the integrity evaluation.
As noted in the discussion which follows, review of operating experience in the detection of leaks and review of the sensitivity of leak detection equipment show that a leak rate of 0.15 gpm can be readily and rel iably detected.
The detectable crack size corresponding rate has been determined to be 2.0 inches. Based upon the comparison of IM.EAK results to the EPRI code PICEP provided in response to Comment 2, an accuracy of twenty-five percent was applied to the IMLEAK results.
That is, the leak rate from a 2.0-inch-long crack was calculated by IM.EAK to be 0.23 gpm, which provides a margin for calculational accuracy of better than 25%.
(Note that for this detectability calculation, the axial load due to internal pressure and bending moments present under operating conditions were considered; the torsional noment was not included.)
i-
Attachment Letter 0570-043-002
- May 22,1967 Page 2 of 13 REVIEW AND Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INOi RECIRCULATION OtOSSOVER LINE AT THE YANKEE NUCLEAR POWER STATION As noted in Comment 4, the loads from the NRC confinnatory analysis performed by Cygna have been used to evaluate detectability under normal conditions and the integrity of the cracked pipe under seismic conditions.
The specific loads used are provided in the response to Comment 4.
The integrity evaluation was performed for a 4.0-inch-long crack which is twice the length of the detectable crack.
The f racture mechanics evaluation was performed using the method described in Section 4.3 of the LB8 report. The appl ied J-integral was calcul ated to be 168 psi-in. This is less than the J value of 960 psi-in and therefore stability yc is demonstrated for normal plus seismic conditions.
Review of Leak Detection Eauimment and Exnertence
- As noted in the following discussion of leak detection capability at YNPS, the leakage rate of 0.15 gpm can be detected by several leak detection systems.
In addition, actual operating experience with low leak rates confirms the ability to detect leaks of 0.15 gpm.
The YNPS Technical Specifications limit unidentified Main Coolant System (MCS) leakage to 1 gpm.
Should unidentified leakage exceed the 1 gpm value, the required action is to reduce the leakage rate to within the limit within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> or be in at least Hot Standby within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and in Cold Shutdown within the following 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br />.
As demonstrated by operating experience, operator action is taken upon detection of leakage rates much lower than the Technical Specification limit.
The existing systems at YNPS are capable of detecting small t
amounts of leakage within the Vapor Container. These systems meet the intent of Regulatory Guide 1.45 and have been accepted by the NRC staff (see NUREG-0825). There are five methods employed at Yankee that are used to determine the i
amount of leakage from various sources.
These are described l
in the following paragraphs.
i i
Attachment Letter 0570-043-002 May 22, 1987 Page 3 of 13 REVIEW AND CLARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INCH RECIRCULATION CROSS 0VER LINE AT THE YANKEE NUCLEAR POWER STATION 4
The Vapor Container Drain Tank (VCDT) collects all unidentified leakage from the vapor container sump.
Tank level is indicated and alarmed on the Main Control Board.
The tank level is logged once every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> and the indicator is available for continuous monitoring.
The alarm setpoint is manually maintained at approximately 2 inches above the normal level.
The tank holds approximately 9 gallons / inch. Therefore, it would take approximately 20 gallons to activate the alarm in one hour and a lower leakage of 0.04 gpm could be detected when the tank level is logged.
i The Primary Drain Collecting Tank (PDCT) collects leakage primarily from valve stem leak-off.
The tank level is indicated and alarmed in the Haste Disposal Building.
There is a common alarm connecting the Haste Disposal Panel with the Main Control Board.
The tank level is logged once every l
hour by the Auxillary Operator.
The Low Pressure Surge Tank (LPST) also provides a means for monitoring primary system leakage. Primary coolant bleed is coIIected in this tank. Charging pump suction is essentially taken from this tank after the water passes through the purification system. A loss in level in the LPST indicates an increase in the amount of makeup to the primary system.
Tank level is monitored hourly in the Main Control Room.
The level can be read in tenths of an inch, with approximately 8 gallons /0.1 inches. Tank level normally drops 0.1 inches / hour which corresponds to an identified leakage of 0.13 gpm.
Significant changes in t
leakage would alert the operator at the next hourly reading.
1 The plant has successfully found Main Coolant System leaks by monitoring airborne particulates using radiation monitors located inside the Vapor Container. While leakage sensitivity varies somewhat with the coolant activity level, Increasing airborne activity levels alert the operator to a problem in the Main Coolant System so that corrective action can be taken.
1
i Attachment Letter 0570-043-002 May 22,1987 Page 4 of 13 REVIEW AND Q.ARIFICATION OF BE LEAK BEFORE BREAK EVALUATION FOR THE 5-IN01 RECIRCULATION CROSS 0VER LINE d
AT THE YANKEE NUCLEAR POWER STATION Finally, a primary system water balance is performed once every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, with data collected every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />.
The water balance is perfor1ned in accordance with Operating Procedure No. OP-4220. This technique will detect unidentified leakage of much less than 1 gpm.
At the first indication of increased leakage within the Vapor Container, entry is made per Operating Procedure No. OF3629 to determine the location of the leak. An inspection of the Vapor Container is performed every two weeks per Operating Procedure No. OP-4232 which provides routine visual exams for 1eaks.
Operators will subsequently be notified of any i
findings and appropriate action will be taken based on Technical Specifications.
The following are exampies of how these leak detection capabilities have been used in the past.
l In July,1983 the air particulate monitors sensed increasing amounts of airborne radioactivity in the Vapor Container.
Entry into the Vapor Container confirmed an extremely small leak in a loop relief valve weld.
The leak was observed to be a vapor and thus, was not collected in the VCOT. The plant was subsequently shutdown to repair the weld.
In September,1984 the plant operators logged a VCOT level 3
j increase of approximately 9 gallons /hr. (0.15 gpm). A leak l
in the feedwater line check valve flange was discovered during a Vapor Container entry.
The VCOT level showed no 1eakage in the Vapor Container following the valve repair.
In June,1986 the VCOT detected an increase in leakage of approximately 200 gallons / day (0.14 gpm). Vapor Container entry confitned a leak in the No. 3 Steam Generator blowdown l ine.
The plant was subsequently shutdown to repair the 1ine.
The above operating experience and rev tow of detection capabilities demonstrates that sna11 anounts of leakage (0.15 gpm and lower) within the Vapor Container can be detected rel iably.
Operating Procedure No. OP-4232 will be revised to j
include a visual inspection of the 5-inch bypass line as part of the bi-weekly inspection of the Vapor Container. The bypass line will also be inspected whenever an unidentified leak is detected in the Vapor Container.
I
.,..____,__._._._____m._
Attachment Letter 0570-043-002 May 22,1987 Page 5 of 13 REVIEW AND Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INOi RECIRQJLATION m0SSOVER LINE AT THE YANKEE NUCLEAR POWER STATION Comment 2 It is our understanding that the computational methods in References 5 and 6 of the LB8 report are for BWR primary coolant conditions and may not be applicable for R(R fluid conditions.
Confinn that the IH.EAK sof tware contains computational routines that are appropriate for the fluid conditions in the 5-inch recirculation crossover line.
Al so, provide infonnotion in the report to denonstrate that the IM.EAK sof twaro has been benchmarked with experimental or service data, or with generally available, independently developed sof tware (e.g.
an updated version of the sof tware associated with Reference 5, available at EPRI).
Cl arif ication INLEAK is a version of the EPRI Code LEAK-01.
It has been benchmarked against the sample problems in EPRI NP-3395 and yields the same results.
It has also been benchmarked against experimental data reported in EPRI NP-3395 and the interim report on Battelle studies (Reference 6 of the LBB report). The benchmark problems oover a range of fluid0 conditions including fluid pressure of 1250 psi and 107 F subcool ing.
To confirm the applicability of the IM.EAK results, the Level A leak rates for 1.5-inch-long and 3.0-inch-long cracks were calculated using the EPRI PICEP code (4/3/87 version).
Leak rater of 0.82 and 0.077 gpm (compared to 1.07 and 0.097 gpm, resp:ctively, for IM.EAK) were calculated by PICEP.
This demonstrates that the PICEP results are within 25% of the IMLEAK results.
As noted in Comment 1, an accuracy of 25% was applied to reduce the leak rates calculated by IMLEAK to ensure that they are lower bound.
Attachment Letter 057 0-043-002 May 22,1967 Page 6 of 13 REVIEW MD Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INW RECIROJLATION OtOSSOVER LINE AT THE YANKEE NUCLEAR POWER STATION Comment 3 Spectfy the magnitude of the IM.EAK input variables used in the leak rate computations (e.g.
surface roughness and number of turns).
Clartftcation The input variables for IM.EAK:
Operating Pressure 2000 psi Operating Temperature 510/550 F Subcooled fluid conditions In = 3 Back Pressure 14.7 psi Depth 0.6250 in.
Roughness Height
.366 E-3 in.
Area Ratio 1.00 3
No. of 90 degree turns 0
No. of 45 degree turns 0
2 Area vary for different 0.00403 in Length detectability 3.000 in.
Width calculations 0.00134 in.
The mechantsm for development of a through-wall was assmed i
to be fatigue crack growth.
Accordingly, the roughness height was selected based on measurement of the roughness of f atigue fracture surfaces as reported in NURED/CR-1319 and the number of turns was set equal to zero. To confim the l
appropriateness of these asseptions, the guidance provided by EPRI Report NP-3395 was reviated.
The EPRI Report guidance confims that for fatigue cracks, the number of i
turns should be zero.
The effect of the roughness assmed was evaluated by comparing the roughness height and relative l
roughness recommended by the EPRI report (0.000197 inch and 0.1, respectively) to that used in IM.EAK (0.000366 inch and O.13, respectively). The higher roughness height and relative roughness used for the detectability evaluation results, with all other parameters held constant, in lower leak rates. Thus, for the detectability evaluation, the roughness height used is shown to be conservative.
i
4 Attachment Letter 0570-043-002 May 22, 1987 Page 7 of 13 REVIEW AND CLARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INCH RECIRCULATION CROSS 0VER LINE AT THE YANKEE NUCLEAR POWER STATION Comnent 4 Provide a table listing the individual bending moments (e.g., deadweight, thermal, seismic) for each of the principal coordinate directions and the axial loads at the limiting location. Also, confirm that the axial load entries in Table 4-2 of the report are correct.
Clarification The individual load components at the limiting location are provided in the following table.
The axial load entries in the following table or in Table 4-2 of the report do not include the axial forces due to internal pressure.
Following the method of Tada and Paris, the effect of internal pressure on the stress-intensity factor and crack opening area is calculated separately and then combined with the axial force and bending moment effects.
i The evaluation was performed initially with selsmic loads calculated using the YCS spectra and 2 percent damping.
Subsequent to the initial LB8 evaluation, the piping analysis performed for YAEC by Cygna was finalized.
The i
detectability and integrity evaluations have been revised to incorporate the results of the NRC spectra confirmatory i
analysis.
The loads from the confirmatory analysis are provided in the attached tables.
The results of the evaluations are provided in the response to Comment 1.
1 I
i l
l i
l
Attachment Letter 057 0.-043-002 May 22,1987 Page 8 of 13 REVIEW AND Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INCH RECIRCULATION CROSSOVER LINE AT THE YANKEE MJCLEAR POWER STATION LIMITING LOCATION LOADS Level A (Cygna DP 59)
(ft-lb)
Load Case M
M M
x y
z DW 640
-104
-7 8
M = l DW j = 640 Bending Moment, Mb x
M = l DW l =
7 z
Mb= SRSS (M ' bI x
Mb=
640 f t-lb 7680 in-lb
=
Attachment Letter 0570-043-002 May 22,1987 Page 9 of 13 REVIEW MD Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INOi RECIRCULATION CROSSOVER LINE AT THE YANKEE NUCLEAR POWER STATION LIMITING LOCATION LOADS Level D (Cygna DP 30)
(1b and f t-lb)
Load Case Fx Mx My Mz DW 23 0 71 1095 2125 SEISMIC 2097 2493 5357 5364 Axlal load, Fx:
Fx =
+
SEISMIC Fx = 2327 l b.
Resultant Moment, M :
Mx =
+ SEISMIC R
My, Mz similarly Mx = 2564 My = 6452 Mr = 7489
$ = SRSS (Mx, My, Mr)
MR = 10,212 f t-lb
= 122,546 in-lb
Attachment Letter 0570-043-002 May 22,1967 Page 10 of 13 REVIEW AND Q. ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INO4 fECIRQJLATION OtOS$0VER LINE AT THE YANKEE MJQ. EAR POWER STATION Comment 5 The one-degree rotation angle indicated in References 3 and 12 of the report is not necessarily applicable for all pipe system and plant configurations.
Information should be presented to demonstrate that a one-degree rotation angle is reasonable for the plant specific, extreme load waluation of the 5-inch line.
0 Cl arification The stability a large through-wall crack (90 circumferential length) under large-deformation bending conditions was re-waluated considering staff guidance that maximun displacements, rather than design conditions, should be waluated.
Based upon a rwfew of the main coolant system pf pt ng and components, the bounding extreme load condition was detemined to be steam generator snubber failure.
This f ailure would cause rotation of the steam generator and hot-leg and bending in the bypass 1ine.
Snubber failure modes of unexpected lock-up (i.e., failure to allow free thermal expanston) and lack of lock-up (i.e.,
failure to prwide restraint under dynamic loads) were constdered.
These result in an upper bound rotation at the end of the bypass line of 1.14'.
An elastic-plastic fracture analysts following the approach of EFRI Report NP-3607 was performed.
The strain hardening of the piptng material was constdered in the analyats.
The elastic-plastic formulation for through-wall circumferential flers under bendtng condtttons was used.
The applied J-integral was calculated for the maximum moment in the bypass 1ine under the applied ratation of 1.14'.
The maximum resultant moment in the piping due to the rotational displacement of the end of the bypass pf ptng was detemined using the SUPERPIPE finite element model of the piping.
The maximum mement was detemined to be 181 kip-in.
l Attachment Letter 0570-043-002 hky 22,1967 Page 11 of 13 i
REVIEW AND Q. ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INOi RECIRCULATION CROSSOVER LINE AT THE YANKEE NUCLEAR POWER STATION The Ramberg-Osgood parameters for the weld material were taken f rom EPRI Report NP-4668 where data for S>nW welds at 550 F are reported (specimen SG3). The Ramberg-Osgood parameters used in the evaluation of the weld are:
es = 2 n = 6.4 Ob= 45.7 ksi For an appiled moment of 181 kip-in, the J-integral is 128 p s t-i n.
This is bel ow the J fC value of 960 pst-in (see Comment 6) and demonstrates & substantial margin against f
l unstable crack extension.
I r
In addition, a similar evaluation was performed to ensure stablitty in the base metal.
The Ramberg-Osgood parameters used in the evaluation of the base metal ares es = 3.6 n = 4.8 Ob= 18.5 kst For an applied moment of 181 kip-in, the J-integral for the base metal (using the parameters 1isted above) is 573 ps t-I n.
This is below the J,c of 2569 pst-inch for the base-metal (which was providea in the LBB report) and demonstrates a substantial margin against crack extension.
l I
Attachment Letter 0570-043-002 May 22,1967 Page 12 of 13 i
REVIEW MD Q.ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-IN04 fECIROJLATION Ot0SSOVER LINE AT THE YANKEE WQ. EAR POWER STATION
[
Comment 6 Recent data for austenttic SW indicate that the J/T plot for SW in Figure 4-1 does not envelope the J/T behavior for austenttic S W.
Unless the J/T plot in Figure 4-1 of the L8B report can be justified for the actual SW welds in the 5-inch recirculation crossover line, the extreme load evaluation for SW should be based on the data indicated in (EPRI Report NP-4690-SR).
Clartf tcation The referenced data for austenttic SW was reviewed against that used in the LBB evaluation.
For the Level D integrity L
evaluation, a lower-bound valve of J tn of 960 psi-in was used.
The referenced EPRI data prw15es a value of 990 psi-in.
Therefore, the margins demonstrated for Level D integrity waluation are confirmed.
The revised data for i
austenttic SW welds has been used for the integrity ovaluation (provided in response to Comment 1) and for the extreme displacement evaluation (prwided in response to r
i Comment 5).
i l
i i
4 I
\\
1 i
1
Attachment Letter 0570-043-002 May 22,1967 Page 13 of 13 REVIEW MD Q. ARIFICATION OF THE LEAK BEFORE BREAK EVALUATION FOR THE 5-INm RECIRCULATION m0SSOVER LINE AT THE YANKEE MJQ. EAR POWER STATION Comment 7 Provide a comparison of the seismic loads used in the LBB evaluation with the five locations of highest stress in the confimatory seismic study perfomed using the NRC spectra to verify that bounding loads were used.
Cl arification The seismic loads f rom the location of highest stress (bypass line teminal end at the hot leg) from the confimatory analysis perfomed by Cygna using the NRC spectra have been incorporated in the integrity evaluation. The loads used for the integrity evaluation are provided in the responso to Comment 4.