ML20149H127
| ML20149H127 | |
| Person / Time | |
|---|---|
| Site: | Maine Yankee |
| Issue date: | 11/14/1994 |
| From: | Whittier G Maine Yankee |
| To: | |
| Shared Package | |
| ML20149H125 | List: |
| References | |
| GDW-94-92, NUDOCS 9411170297 | |
| Download: ML20149H127 (100) | |
Text
_ __ _ _ _ _
Enclosure to letter GDW-94-92 Page 1 of 59 MAINE YANKEE 1994 S/G INSPECTION REPORT a /
/ / '
( ,,
Prepared By: i/ % W'hii/- !/ $in f V
/ //
Maine Yankee Principal Engineer
' /
I Reviewed By: b[D Maine Yankee Principal Engineer J
Reviewed By: Id Mov 9d Maine Yankee Corporate Engineering Section Head Approved By: N Manager, Corporate Engineering Department Approved By: / NV /4 V[7f Maine Yankee Vice President, Licensing & Engineering L:\94MN\94104 ;
9411170297 941114 PDR ADOCK 05000309 Q PDR
Enclosure to letter GDW-94-92 ,
Page 2 of 59 TABLE OF CONTENTS EAGE ;
1.0
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.0 GE0 METRY .............................. 6 3.0 CYCLE 13 AND 14 PRIMARY TO SECONDARY LEAK RATE ........... 7 4.0 JULY / AUGUST 1994 S/G MRPC INSPECTIONS . . . . . . . . . . . . . . . . 8 5.0 FUTURE S/G TUBE INSPECTION & ANALYSIS ENHANCEMENTS . . . . . . . . 11 ;
6.0 HISTORY OF CIRCUMFERENTIAL S/G TUBE CRACKS AT MAINE YANKEE . . . . 12 7.0 CIRCUMFERENTIAL CRACK INITIATION RATE . . . . . . . . . . . . . . . 13 8.0 CAUSE OF CIRCUMFERENTIAL CRACKS . . . . . . . . . . . . . . . . . . 14 9.0 CIRCUMFERENTIAL CRACK GROWTH RATE . . . . . . . . . . . . . . . . . 15 10.0 REG. GUIDE 1.121 ACCEPTABLE CIRCUMFERENTIAL CRACK SIZE . . . . . . 16 11.0 FUTURE RUN CYCLE DURATION JUSTIFICATION . . . . . . . . . . . . . . 17 12.0 MITIGATION OF PWSCC . . . . . . . . . . . . . . . . . . . . . . . . 18 13.0 IN-SITU PRESSURE TESTING . . . . . . . . . . . . . . . . . . . . . 19 13.1 1994 In-Situ Pressure Test Results .............19 13.2 Enveloping All S/G Loads by an In-Situ Pressure Test . . . . 20 13.3 Effect of Locked S/G Tubes on In-Situ Pressure Testing Results 21 13.3.1 Point of Application of Axial Load Effects ...22 13.3.2 Thermal Compressive Load During Operation . . . . 22 13.3.3 Residual Tensile Load During Cold Isothermal Conditions ...................22 13.3.4 Axial Flexibility of Horizontal Supports . . . . 23 13.4 Validity of In-Situ Pressure Testing . . . . . . . . . . . . 23 .
14.0 STRUCTURAL INTEGRITY OF 1994 DEGRADED S/G TUBES . . . . . . . . . . 24 15.0 EFFECT OF DEGRADED TUBES ON DESIGN BASES .............25 15.1 Anticipated Leakage During a MSLB . . . . . . . . . . . . . . 25 ,
15.2 MSLB Concurrent Leakage Effect on Offsite Dose . . . . . . . 25 15.3 S/G Tube Rupture Considerations . . . . . . . . . . . . . . . 27 L:\94MN\94104 l
l Enclosure to letter GDW-94-92 Page 3 of 59 TABLE OF CONTENTS (continued) l EAGE 16.0 S/G INSPECTIONS SCHEDULED FOR THE 1995 REFUELING OUTAGE . . . . . . 28 17.0 1995 TUBE PULL CONSIDERATIONS . . . . . . . . . . . . . . . . . . . 29
18.0 REFERENCES
............................30 19.0 TABLES ..............................32 20.0 FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 >
ATTACHMENTS Attachment A - Maine Yankee Run Duration Limit Evaluation for Circumferential Cracking.
Attachment B - Point of Loading Effect on Validity of In-Sito Pressure Testing.
t L:\94MN\94104
1 i
l l
1 Enclosure to letter GDW-94-92 !
Page 4 of 59 l
1.0
SUMMARY
l Due to increasing primary to secondary leakage approaching Maine Yankee's administrative limit of 50 gallons per day (gpd), Maine Yankee shut down on July 15,1994, well before the 216 gpd Technical Specification limit. A low pressure leak test was performed on steam generator (S/G) No. 2. Four leaking I tubes were identified. Motorized Rotating Pancake Coil (MRPC) inspections indicated that the four tubes had inside diameter (ID) cracks. The defect sizes ranged from 35 to 94% average through wall (Avg TW).
Reevaluation of the previous MRPC inspection data on these tubes indicated that flaws were present during earlier inspections but were not detected using ;
then current analysis methods. The defects previously went undetected because the cracks' eddy current signals were not distinguished from similar signals caused by tube geometry changes occurring at the same location. Review of the past data facilitated the development of improved data evaluation techniques to ensure that similar circumferential cracks would be detected in the 1994 S/G tube inspections. These improvements were incorporated into the ;
subsequent MRPC inspection of all hot leg tube expansion transitions in all three S/G's and will be incorporated into future MRPC inspections.
Three hundred and three (303) tubes with top of tubesheet circumferential t cracks were identified and plugged. The phase angles of the MRPC signals imply that these cracks initiated from the tube ID. There's a high likelihood the cracks are caused by primary water stress corrosion cracking (PWSCC) as ,
was the case for the circumferential crack in the tube pulled and analyzed in !
1990. l All but one of the 303 circumferential tube defects were present in preceding inspections. Ninety-two percent (92%) of the circumferential cracks detected '
in the July 1994 inspection existed in 1990. Reliable circumferential crack growth rates can be derived from the 1990, 1992, 1993, and 1994 inspection data. Review of the past data identified an average crack growth rate of 11.17% average through wall per effective full power year (Avg TW/EFPY).
The review of the previous inspection data revealed that the present crack initiation rate is low. Less than 11 of the 303 recently observed cracks initiated since April 1992 and only one initiated between October 1993 and i July 1994. ,
Maine Yankee's improved crack detection capabilities, low crack initiation rate, and moderate crack growth rate will allow Maine Yankee to run at least 500 effective full power days (EFPD's) between successive circumferential .
crack inspections with a low probability of cracks growing beyond the Reg.
Guide circumferential crack size limit of 79% Avg TW.
L:\94MN\94104 I
Enclosure to letter GDW-94-92 Page 5 of 59 Of the 303 circumferential defects sized by MRPC in the 1994 S/G inspection, 24 were larger than the 79% Avg TW Reg. Guide 1.121 circumferential crack size limit. In-situ hydrostatic pressure tests were conducted on 7 tubes that exceeded the Reg. Guide 1.121 size limit to further evaluate the remaining strength of these tubes. The tested tubes were representative of the largest defects in S/G's 1, 2, and 3. Five (5) of the above tubes demonstrated near compliance with the most limiting criteria of Reg. Guide 1.121 (3 times the normal operating pressure differential without bursting). The limited make up capacity of the testing apparatus limited the test pressure attainable in the other two tubes. These two tubes were pressurized to their maximum design load, which is the postulated main steam line break (MSLB) maximum differential pressure. .4either of the two tubes burst (i.e., their leakage was less than 0.5 gpm when retested at 1450 psi). It is probable that higher test pressures could have been achieved if the test pump had had a higher make up capacity.
Tubes with MRPC identified Avg TW defects approaching 100% may still remain intact during anticipated maximum operating loads because circumferential PWSCC defects generally are not a single uninterrupted crack but a number of microcracks. The microcracks are separated by ligaments of noncorroded material. The presence of the ligaments cannot be detected by the MRPC inspections but still contribute strength to the degraded tubes. MRPC identified circumferential crack sizes were shown, by the in-situ pressure testing conducted at Maine Yankee, to be conservative for PWSCC cracks relative to the remaining strength of the tubes.
The in-situ pressure testing of the large MRPC sized cracks to MSLB differential pressure demonstrated that significant structural margin remained in the most severely cracked tubes during normal operation differential pressure, and that the tubes would have been expected to remain intact through 1 a postulated MSLB with minimal additional leak &ge. A tube with PWSCC generally exhibits uneven defect growth through the tube wall. Based on the I in-situ pressure testing at Maine Yankee and the usually uneven PWSCC circumferential crack growth, the deepest crack penetration is expected to cause significant leakage before the tube would be expected to burst under design loads. This is reinforced by the fact that no plant has ever experienced a S/G tube rupture (SGTR) due to circumferential stress corrosion cracking.
Based on the results of the in-situ pressure testing, a maximum leakage of 6.0 gpm could potentially have initiated as a consequence of a postulated MSLB. A conservative analysis was performed demonstrating that Maine Yankee can accommodate a release of this amount following a MSLB without approaching the 10CFR100 offsite dose limits. The anticipated leak rate is less than the charging pump capacity such that reactor coolant inventory control is not challenged.
Maine Yankee does not presently plan to pull another tube with a l circumferential crack. The in-situ pressure testing provides a valid assessment of remaining structural integrity. The defect mechanism was identified from a previous tube pull in 1990. Little more is expected to be learned from an additional tube pull at this time.
L:\9 MN\94104
Enclosure to letter GDW-94-92 l Page 6 of 59 ]
2.0 GEOMETRY The 3 Maine Yankee S/G's were built by Combustion Engineering (C-E). They have a mix of characteristics from C-E's pre-Model 67 and Model 67 S/G !
designs. Each of the Maine Yankee S/G tube bundles (depicted in Figure 1) is made up of 5,703 tubes. The S/G tubes have a 0.75 inch outside diameter, are 0.048 inches thick, and made of SB-163 Inconel 600 material. The tubes are placed in a triangular pattern with a pitch of 1.0 inch. The tubes are !
explosively expanded (explanded) full length into a 20.31" thick tubesheet as shown in Figure 5. The tubes are supported in the axial flow region by 0.090" .
thick carbon steel lattice type (eggcrate) supports shown in Figure 2. The l first support is 48 inches above the secondary face of the tubesheet with five i other eggcrate supports equally spaced at 40 inches. There is one partial eggcrate support 32 inches above the uppermost full eggcrate support. There are two floating partial drilled carbon steel support plates at two elevations ,
as shown in Figures 1, 3, and 4. ;
It can be seen from Figures 1, 3, and 4 that the two (2) drilled support plates only partially span the cross section of the tube bundle. These partial drilled plates have been cut away from their original supporting tube bundle steel wrapper. They are now supported in the axial direction by tubes that have been intentionally expanded / locked into the plates and plugged.
I i
1 L:\94MN\94104
Enclosure to letter GDW-94-92 Page 7 of 59 3.0 CYCLE 13 AND 14 PRIMARY TO SECONDARY LEAK RATE j Trace amounts of primary to secondary leakage were detected at the end of Maine Yankee Cycle 13 just before its summer 1993 refueling outage. During the 1993 refueling, Maine Yankee inspected 75% of the S/G tubes with a classical eddy current test (ECT) bobbin coil. Fifty-four percent (54%) of I the hot leg top of tubesheet tube expansion transitions were examined with !
MRPC. A leak test was conducted in S/G's 2 and 3 by subjecting the secondary side to a 170 psi hydro test while looking for leaks from the primary side of the tubesheets. No S/G through wall (TW) defects were discovered in 1993.
The primary to secondary leakage continued after the 1993 outage. The leak rate increased for nine months as shown in Figure 6. Maine Yankee conservatively entered a controlled shutdown on July 15, 1994, as the S/G 2 ,
calculated leak rate approached 50 gpd and well before challenging the l Technical Specification limit of 216 gpd.
A low pressure hydrostatic leak test at 170 psi revealed four leaking tubes in S/G 2. Subsequent MRPC testing revealed that the defects were inside diameter (ID) circumferential cracks just above the top of the S/G tube sheet in the tube expansion transition region. See Figure 5. The leak test results and j subsequent MRPC identified defect characteristics are listed in Table 1. The ;
locations of the leaking S/G tubes in the tube bundle are shown in Figure 7. I 4
l L:\94MN\94104
l Enclosure to letter GDW-94-92 l Page 8 of 59 ;
i 4.0 JULY / AUGUST 1994 S/G MRPC INSPECTIONS During the Maine Yankee 1994 S/G inspection outage, MRPC inspections were performed on all in service hot leg tube expansion transitions in all three S/G's. A total of 303 top of tubesheet circumferential defects were found.
Two (2) pits were also discovered. All observed defects were plugged.
Figures 10, 11, and 12 depict the 1994 MRPC identified Avg TW circumferential crack sizes for S/G's 1, 2, and 3 respectively. The locations of the 1994 detected cracks in the S/G tube bundle may be seen from Figures 13, 14, and 15.
The MRPC eddy current test is used as the primary means of detecting top of tubesheet circumferential cracks in pressurized water reactor S/G's. Defects are identified and characterized as follows:
Maximum Depth:
Estimates of maximum crack depth are made by phase analysis of the primary frequency (400Khz). A three point phase angle curve is generated based on the known depths of defects in an ASME calibration standard. A defect signal will have a phase angle component. A software algorithm interpolates between the defect signal's phase angle and those from the known calibration signals to size the depths of indications in the steam generator tube.
Relative Volume Loss:
Voltage measurements are a relative measure of volume loss. The voltage ,
signal is calibrated at 5 volts for a flaw of known volume in the ASME l standard. A larger defect voltage indicates a larger defect volume loss.
i Length: l Estimates of crack length are made by isolating the vertical component of the flaw signal and comparing the time interval that the indication i is sensed relative to a circumferentially staged series of trigger pulses input from the rotating probe motor.
I L:\94MN\94104
Enclosure to letter GDW-94-92 !
Page 9 of 59 Avg TW:
Average cross-sectional wall loss is derived from the depth and length values described above. The % Avg TW = Depth (max %) x Length ,
.(degrees)/360. This method assumes that the entire arc length is uniformly deep (tubes removed from C-E steam generators have exhibited varying depths around the circumference) and that the area without MRPC defect indication is not flawed. A diagram which plots MRPC indcated Avg TW circumferential crack sizes against metallographic results for tubes removed from C-E units is.shown in Figure 16. Additional data from over 40 circumferential1y cracked lab samples also exhibits reasonable correlation between MRPC identified Avg TW crack sizes and the metallographic lab determined actual crack sizes (see Figure 17).
This method of describing the crack gives a reasonable approximation of ;
the crack size which may be compared to an associated Reg. Guiae 1.121 ,
circumferential crack size limit. In-situ pressure testing experience !
at Maine Yankee indicates that the Avg TW sizing technique is a ,
conservative means of quantifying the remaining strength and Reg. Guide '
1.121 compliance for tubes with PWSCC circumferential cracking.
Maine Yankee re-examined data from earlier (MRPC) inspections of the four i tubes which leaked in the 1994 low pressure leak test. The data review was 1 conducted to see if the defects could have been detected earlier and to identify signal interpretation techniques that would ensure similar defect ' i signals would be recognized in the future. Enhancements in the analysis ;
techniques were developed to better discern circumferential cracks. Using these enhanced techniques, three of these leaking cracked tubes could be detected in the 1990 data and the remaining leaking tube could first be detected in the 1992 data.
The past defect signals were missed in earlier inspections as they closely "
resembled the MRPC signal changes caused by tube geometry changes at the expansion transition region, uhich occurs at the same location. See Figure 5. :
The review revealed that the following practices enhance the detectability of top of tubesheet circumferential cracks and were incorporated into all of the 1994 MRPC inspections:
> Generation of terrain maps for all inspected tubes.
> Calculation of the ratio of the MRPC's circumferentially sensitive ,
coil voltage to the axially sensitive coil voltage at all defect-like indications. Ratios of 2.0 or larger require further investigation.
^
- Reviewing past inspection data for the subject tube when resolving ambiguous signal data and looking for changes.
L:\94MN\94104 l
l
l i
l i
Enclosure to letter GDW-94-92 !
Page 10 of 59
> Using recently available enhanced inspection equipment (0.115" l diameter three coil MRPC with low loss cable) to reduce signal to l noise ratios. This results in improved identification and i resolution of smaller defect sizes.
- Including past data signals from the 4 leaking tubes identified in l 1994 in the Maine Yankee S/G inspection data analyst guidelines, training, and performance demonstration program.
The 1994 inspection program was sufficient to identify very small cracks.
Fifteen percent (15%) of the 1994 circumferential cracks were less than or equal to 20% Avg TW.
The above mentioned data analysis enhancements were incorporated in a review of the past analysis data for the 303 tubes found with circumferential cracks in 1994. Two hundred seventy-eight (278) of the 303 July 1994 identified circumferential cracks existed in 1990. Figure 18 is a frequency plot of the defect sizes in 1990 and 1994. In 1990, 77% of the cracks were less than or equal to 20% Avg TW. This review of earlier inspection data further supports the position that small circumferential defects are discernable with the enhanced analysis interpretation techniques. 1 L:\94MN\94104 1
Enclosure to letter GDW-94-92 Page 11 of 59 5.0 FUTURE S/G TUBE INSPECTION & ANALYSIS ENHANCEMENTS Maine Yankee plans to upgrade its S/G inspection analyst guidelines, training, and performance demonstration program to include the inspection and analysis enhancements learned during the 1994 inspection. Specifically, the following upgrades are intended to be incorporated into the program before Maine Yankee's 1995 Refueling S/G Inspection:
- Past (pre 1994) inspection data from representative tubes with ,
cracks detected in the 1994 inspection will be added to the training and demonstration segments.
- Inspection data from other C-E plants with circumferential cracking will be added to the training and demonstration segments.
- Terrain mapping will be required for all inspected tubes.
- A three coil MRPC will be used in future top of tubesheet inspections. The ratio of the circumferential crack sensitive coil voltage to the axial crack sensitive coil voltage will be ,
calculated for all crack-like indications as a screen to distinguish geometry change induced signals from circumferential cracks. A ratio of 2.0 or higher will require further investigation for the possible presence of circumferential cracks.
- Primary and secondary analysts will be instructed to call any circumferential indications as cracks causing further review by the shift's lead analyst (s). The lead analyst (s) will review coil voltage ratios, past tube inspection results, and any other available data to assist in distinguishing a crack from a geometry signal.
- An inspection data noise criteria will be incorporated into the guidelines. Should the noise be too high, the tubes will need to be retested.
- MRPC probes with 0.115 inch diameter coils and low loss cable will be used for improved signal to noise ratio until a better probe becomes available.
- Maine Yankee will involve an independent S/G inspection vendor to assist in the oversight of future S/G tube inspections.
L:\94MN\94104
Enclosure to letter GDW-94-92 Page 12 of 59 6.0 HISTORY OF CIRCUMFERENTIAL S/G TUBE CRACKS AT MAINE YANKEE Maine Yankee first identified top of tubesheet circumferential cracks in 1990 when three cracked tubes were identified by a low pressure hydrostatic leak test. All top of tubesheet roll transitions regions on both the hot and cold leg sides of all three S/G's were examined by MRPC during the 1990 Refueling Outage. Subsequent MRPC inspections of a sample of the hot leg expansion transitions were conducted in 1992 and 1993. The tubes selected for inspection in 1992 and 1993 were such that each hot leg tube expansion transition would be inspected by MRPC at least once every two cycles. All circumferential cracks detected prior to 1994 were less than Reg. Guide 1.121 Avg. TW crack size limits. The extent of MRPC inspection and subsequent results are shown in Table 2. All circumferential cracked tubes ever 1 identified at Maine Yankee have been removed from service at the time they l were detected.
In 1990, Maine Yankee removed a tube with a top of tubesheet crack from S/G #1 for destructive examination. The pulled tube evaluation revealed that the l tube material was in compliance with the material specification, that the crack originated from the ID, and that the crack was most likely caused by PWSCC. The origination point (inside or outside diameter) of circumferential cracks may be inferred from MRPC defect signals. Nearly all of the 303 circumferential cracks detected in 1994 exhibit characteristics associated with ID initiated cracking. ,
I l
L:\94MN\94104
Enclosure to letter GDW-94-92 Page 13 of 59 l l
7.0 [1RCUMFERENTIAL CRACK INITIATION RATE Maine Yankee has reviewed all available hot leg MRPC inspection data (1990, ;
1992, 1993, and 1994) of the 358 circumferential1y cracked tubes ever detected at Maine Yankee utilizing the enhanced data analysis techniques developed and ,
implemented in the 1994 S/G tube inspection. The review revealed when a crack :
may have first initiated and the growth rate of the cracks during operation. -
The results of the crack initiation study are depicted in Figures 8 and 9. Of ,
the 358 circumferential cracks ever detected,10 could not be tied to a :
specific cycle in which they initiated because two cycles had elapsed between :
the inspections where the 10 were nondetectable and their next subsequait inspection where they first became detectable. Figure 8 assumes the cracks !
initiated one half way through the time between subsequent inspections. ;
Figure 9 assumes the 10 indeterminate tubes initiated in the latter half of l their two cycle inspection interval. j i
It can be seen from Figures 8 and 9 that only one of the 1994 detected circumferential cracks was not detectable (with enhanced data analysis :
techniques) in preceding inspections. Both Figures 8 and 9 reveal that the !
crack initiation rate since 1990 is very low and is decreasing.
The low incidence of crack initiation in recent examinations reduces the ,
number of cracks expected in future years and the likelihood that any crack I will grow beyond the Reg. Guide 1.121 acceptance criteria before being detected in refueling outage interval inspections. .
i i
f L:\94MN\94104
l Enclosure to letter GDW-94-92 Page 14 of 59 i 8.0 CAUSE OF CIRCUMFERENTIAL CRACKS The destructive examination of the S/G tube removed from service in 1990 at Maine Yankee identified PWSCC as the most likely cause of circumferential cracking. A review of Reactor Coolant System (RCS) operating conditions and the chemical analyses of the tube pulled in 1990 showed that the circumferential cracking could not be attributed to impurities in the RCS PWSCC incidence rate is customarily predicted by an increasing hazard Weibull time to failure function. The PWSCC initiation rate witnessed at Maine Yankee and shown in Figures 8 and 9 is not representative of the expected Weibull projection of a homogeneous S/G tube sample. Figures 8 and 9 suggest that only a subset of the total number of S/G tubes are most susceptible to PWSCC i and have nearly all initiated at this point (explaining ths very low present initiation rate) or a causative condition occurred prior to 1990 that has not existed since.
An effort has been made to try to identify subsets of the S/G tube population that have similar characteristics and have experienced top of tubesheet cracks at Maine Yankee. S/G's 1 and 2 seem to show some correlation between specific '
tube material heats and degraded tubes but no similar correlation has yet been identified for S/G 3. No correlation is apparent between tubes with circumferential defects and their sludge pile heights, presence of horizontal support denting, the number of times a tube has been eddy current tested over its life, and/or shapes of the expansion transition as discerned with an ECT probe.
The initiation rate of PWSCC generally increases with temperature and time.
Maine Yankee's average hot leg temperature was raised from 596.5 F to 599.5 F in 1986, and lowered back down to 598.5oF in 1989. These temperature changes are not believed to be significant enough to cause the cracks' initiation rate history shown in Figures 8 and 9.
L:\94MN\94104
Enclosure to letter GDW-94-92 Page 15 of 59 9.0 CIRCUMFERENTIAL CRACK GROWTH RATE As 278 of the 303 circumferential top of tube sheet cracks detected in 1994 ,
were present in 1990 and only one defect wasn't present in any preceding inspection, the growth rate of PWSCC induced circumferential cracks can be <
well established for Maine Yankee. The growth rate data was compiled and l analyzed by ABB C-E. The growth rate determination is detailed in Section 3.2 l of ABB C-E's Maine Yankee run duration evaluation for circumferential cracking. This evaluation has been included in its entirety as Attachment A.
The Maine Yankee's average circumferential growth rate is 11.17% Avg TW/EFPY.
The 95/95 upper confidence limit of the past Maine Yankee circumferential crack growth rate is 40.23% Avg TW/EFPY. The crack growth rate probability distribution function was developed from the entire set of available incremental crack growth data between July 1990, and July 1994, from steam generators 1, 2, and 3. The data set consisted of 683 apparent growth rates.
During the incremental periods over which growth rates could be determined, i only 80% of the cracks showed growth. There was no apparent relationship '
between growth rate and defect depth (i.e., the larger defects did not grow faster than smaller defects).
L:\94MN\94104
Enclosure to letter GDW-94-92 Page 16 of 59 10.0 REG. GUIDE 1.121 ACCEPTABLE CIRCUMFERENTIAL CRACK SIZE The maximum Maine Yankee Reg. Guide 1.121 acceptable circumferential crack Avg TW size was calculated by ABB C-E in Reference 7 to be 79%. This is consistent with the structural evaluation / testing of the laboratory induced circumferential crack program conducted by Maine Yankee in 1990. The most limiting Reg. Guide 1.121 criteria for Maine Yankee is to withstand 3 times the normal S/G tube normal operating pressure without bursting.
i 1
l 1
L:\94MN\94104 l
l l
Enclosure to letter GDW-94-92 Page 17 of 59 11.0 FUTURE RUN CYCLE DURATION JUSTIFICATION Maine Yankee's improved crack detection capabilities, low and declining crack initiation rate, and moderate crack growth rate permit Maine Yankee to run at least 500 EFPD's between successive circumferential crack inspections with a low probability of cracks growing beyond the Reg. Guide 1.121 Avg TW circumferential crack size limit. ABB C-E performed the run length evaluations for Maine Yankee. They are described in detail in Sections 3, 4 and 5 of their run duration report, which is included in its entirety in Attachment A. Both deterministic and prooabilistic run length evaluations were performed.
It is conservative to assume a circumferential crack can be detected by the time it is 20% Avg TW. Section 3.1 of Reference 8 documents the MRPC's demonstrated ability to detect and size circumferential cracks ranging from 4 to 20% Avg TW. If an MRPC inspection of 100% of the hot leg expansion transition is conducted at the beginning of a cycle and the 95/95 upper confidence level of the observed Maine Yankee crack growth rate is used in the deterministic run duration evaluation, Maine Yankee can expect to run 365 x (79-20)/40.23 - 535 EFPD's before a circumferential crack would exceed the 79% Avg TW Reg. Guide 1.121 size limit.
The deterministic run length is reinforced by a probabilistic evaluation which can incorporate the effects of the anticipated crack initiation rate as well as the actual variation in the anticipated crack growth rate. Maine Yankee can expect to operate a 500 EFPD cycle with a greater than 90% probability that one or less tubes will be found to have a circumferential crack exceeding 75% Avg TW. Additional probabilistic evaluations may support an even longer run duration with the same probability factor. These additional evaluations were not run since 500 EFPD's encompasses Maine Yankee's expected future cycle durations.
The run duration evaluations are consistent with the operating experience Maine Yankee has had with circumferential cracks present from July 1990 to July 1994. As can be seen from Figure 18, the cracks present in 1990 grew slowly. After 4 years, the majority of the tubes with cracks in 1990 still satisfied Reg. Guide 1.121 acceptance criteria and all tubes continued to have sufficient remaining structural integrit.v to resist bursting under their maximum design loads. In the future, enhanced crack detection techniques will ,
provide a high confidence that circumferentially cracked tubes will be removed l from service before they exceed the Reg. Guide 1.121 acceptance criteria. i L:\94MN\94104 l
l
Enclosure to letter GDW-94-92 Page 18 of 59 12.0 MITIGATION OF PWSCC Laboratory testing reported in paper C4 of Reference 13 indicates that Maine Yankee can expect the S/G circumferential crack initiation and growth rates to Iv one half (1/2) and one third (1/3) respectively of their present values when zinc is added to the RCS. Maine Yankee plans to start adding zinc after its 1995 refueling outage. This will make the run duration evaluations conservative and further reduce the likelihood of a circumferential crack exceeding Reg. Guide 1.121 acceptance criteria in the future.
L:\94MN\94104
Enclosure to letter GDW-94-92 Page 19 of 59 13.0 IN-SITU PRESSURE TESTING 13.1 1994 In-Situ Pressure Test Results Maine Yankee performed an in-situ pressure test on ten S/G tubes (4 tubes in S/G #1 and 6 tubes in S/G #2) during the 1994 S/G inspection program. Seven of the in-situ pressure tested tubes whose MRPC indicated Avg TW crack sizes exceeded the Reg. Guide 1.121 Avg TW size limits were tested as an alternate means of assessing compliance with Reg. Guide 1.121 criteria. These seven tubes had cracks representative of the largest MRPC indicated defects in all of the three S/G's. The 3 other tubes in-situ pressure tested were the remaining tubes in S/G #2 that were observed leaking during the previous low pressure led tcst. It was desired to pressure test all known leaking tubes to quantify the leakage that could have occurred during a postulated MSLB condition.
For 7 of the tubes tested, the in-situ test's purpose is to demonstrate compliance with Reg. Guide 1.121 S/G tube structural adequacy criteria.
The most limiting criteria for Maine Yankee is pressurization without burst to 3 times differential operating pressure (3 x 1450 - 4,350 psi).
The next most limiting criteria is pressurization without burst to 1.4 times the MSLB differential pressure (1.4 x 2520 - 3,530 psi).
Because of differences in material properties at room temperature and operating conditions, an Inconel 600 S/G tube at operating temperature (600 F) can be expected to fail at 87% of the burst pressure at room temperature per page 8-7 of Reference 3 and page 30 of Reference 14.
To show compliance with the Reg. Guide criteria with a room temperature test, the target test pressures must be increased by 15% to the ,
following values:
Operating differential pressure 1450 x 1.15 - 1,667 psi MSLB differential pressure 2520 x 1.15 - 2,898 psi 3 x Operating differential pressure 3 x 1450 x 1.15 - 5,003 psi 1.4 x MSLB differential pressure 1.4 x 2520 x 1.15 - 4,057 psi The results of the 1994 in-situ pressure testing are listed in Table 3 with their maximum test pressure, reported leak rate at the maximum achieved pressure, and MRPC identified defect characteristics.
Maine Yankee originally increased the target cold condition test pressure by only 10% rather than 15% above the 3 times the operating pressure value to demonstrate compliance with the 3 times differential pressure without bursting criteria. As none of the seven tubes that withstood 4800 psi leaked appreciably, it is expected they could have reached the higher differential operating pressure target of 5,003 psi and satisfy the most limiting Reg. Guide 1.121 criteria.
L:\94MN\94104 l
)
Enclosure to letter GDW-94-92
-Page 20 of 59 j Three of the ten tubes in Table 3 had a maximum test pressure limited by the maximum make up capacity of the testing apparatus. S/G Tube 89/46 !
had the largest leak rate of the original identified leaking tubes (1 .
drop every.2 seconds under a 170 psi pressure differential). Its defect i opened up early at 2700 psi to leak in excess of the pressure test !
pump's capacity. As tube 89/46's average TW defect size is less than '
the Reg. Guide's 1.121 identified limit, it is expected to have been able to withstand 5,003 psi had the test pump's make up capacity been -
large enough. S/G tube 49/122 came within 98.5% of satisfying the '
temperature corrected 1.4 x MSLB target criteria. S/G tube 70/55 met the temperature corrected maximum MSLB differential pressure target. Of the three pressure tested tubes which leaked, none burst. They all '
could be repressurized to 1450 psi leaking less than 0.5 gpm after being .
pressurized to their maximum value. Higher maximum test pressures could !
potentially have been reached if the test pump had a higher make up l capacity.
Tubes with MRPC identified Avg TW defects approaching 100% may still i remain intact during anticipated maximum operating loads because circumferential PWSCC defects usually are not a single uninterrupted crack but rather a number of microcracks. The microcracks are separated by ligaments of noncorroded material. The presence of the ligaments cannot be detected by the MRPC inspections, but do contribute strength i to the degraded tube. MRPC identified circumferential crack Avg TW ,
sizes at Maine Yankee were shown by in-situ pressure testing to be l conservative relative to the remaining strength of the tube. i 13.2 Envelopino All S/G Loads by an In-Situ Pressure Test In-situ pressure testing of a sample of the S/G tubes with top of tubesheet circumferential cracks was performed during the Maine Yankee 1994 S/G inspection to assess the remaining structural integrity of the e cracked tubes. As the in-situ pressure tests only impose pressure induced loads on the S/G tubes, it is prudent to review what effect '
other types of loadings [ Main Steam Line Break (MSLB), Loss of Coolant Accident (LOCA), safe shutdown earthquake (SSE), and fluid flow] may have on the usefulness of pressure only test results.
Per References 4, 6, and 7, S/G tube loadings other than pressure include LOCA rarefaction wave, MSLB rarefaction wave present on secondary side blowdown, pipe break (MSLB and/or LOCA) impulse response (the magnitude considered in this evaluation is the larger of the two),
Safe Shutdown Earthquake (SSE), and flow induced loads. The magnitudes ,
of these stresses are listed in Table 4 for various S/G tube circumferential crack Avg TW sizes.
L:\94MN\94104 i
I l
Enclosure to letter GDW-94-92 , l Page 21 of 59 ;
1 The MSLB rarefaction wave was investigated in detail in Reference 6. A fluid dynamic history has been generated for the secondary side of the S/G's during a MSLB event. The Maine Yankee eggcrate supports (thin steel lattice strips) and partial drilled suppnrt plates (which occupy I only 15% of available secondary side tube bundle inventory blowdown flow I area) provide minimal resistance to the exiting secondary side inventory during a MSLB. The forces acting on the horizontal supports are i resisted (locked tubes in the partial drilled support plates and l friction between tubes and eggerates) equally by all tubes passing i
through the supports resulting in a negligible increase in any !
l individual tube's loads / stresses. !
- \
During the blowdown phase of a MSLB event at Maine Yankee, the l l
i differential pressure between the primary and secondary side of the l faulted S/G increases to less than 1750 psi. Without rapid operator !
action, subsequent to the dryout of the faulted steam generator, continued Emergency Core Cooling System (ECCS) flow, combined with the l heatup of the RCS from decay heat, will cause a gradual repressurization !
of the RCS to a maximum value of 2520 psi (see figure 19). Emergency l procedures subsequently direct operators to throttle High Pressure Safety Injection (HPSI), establish a heat removal path, and reduce the ,
l RCS pressure to minimize loads on the RCS components. l l Table 4 shows that the increased stress imposed on the S/G tubes due to i
the MSLB differential pressure increasing from 1750 to 2520 psi is i larger than the additional dynamic stresses which act only during the initial blowdown when the tube differential pressure is only 1750 psi. :
The SSE stresses are not combined with the maximum MSLB stresses in l Table 4 as it is improbable an earthquake will occur during the !
anticipated short period that the differential pressure is at 2520 psi l l
after a MSLB. Table 4 shows that the maximum stress a S/G tube will see j is caused strictly by tube differential pressure during the MSLB. An l in-situ pressure test evaluating a tube's ability to resist the MSLB maximum pressure differential suitably envelopes / evaluates all the Maine l Yankee S/G tube design loadings.
13.3 Effect of Locked S/G Tubes on In-Situ Pressure Testina Results S/G tubes may become locked in their horizontal supports as magnetite l fills the space between the S/G tubes and their surrounding supports, !
squeezing the tube. An in-situ pressure tested tube being locked in one or more of its horizontal supports during an in-situ pressure test has a conservative effect on the validity of the in-situ pressure test results (evaluation of a tube's structural integrity). This is due to the ,
different points of application of the tube's axial load between l operation and testing, thermal compressive forces present during I operation, residual tensile forces existing during the cold in-situ I testing, and the vertical flexibility of the horizontal supports.
l l
L:\94MN\94104 l
i i
l Enclosure to letter GDW-94-92 Page 22 of 59 13.3.1 Point of Application of Axial Load Effects When a tube is locked into one or more of its horizontal supports, the in-situ pressure test imposes a larger axial load on a top of tubesheet circumferential crack than would be imposed by an equal operations pressure tube end load. The operating pressure imposed tube axial end load is applied at the top end of the tube. All of the tube's end load is available to load the locked horizontal supports before loading the defect above the tubesheet. The in-situ pressure test axial load is imposed on the S/G tube just above the tubesheet. The in-situ pressure test load is innediately resisted by the portion of the tube with the defect before a locked support can be loaded. Consequently, the reduction of the axial load passing through a top of tubesheet circumferential crack due to a locked horizontal support is larger during operation. The in-situ pressure test is a conservative means of evaluating a tube's remaining structural integrity. A rigorous mathematical analysis of this phenomenon is presented in Attachment B of this enclosure.
13.3.2 Thermal Compressive Load During Operation I
Because the Inconel tubes have a higher coefficient of thermal expanston than their carbon steel support structure (Reference 5),
a locked horizontal support will superimpose a thermally induced axial compressive force onto the pressure induced tensile force existing in S/G tubes during operation. This thermal compression reduces the pressure imposed axial tensile stresses. The in-situ pressure test is r.onservative in that the operating thermally induced compressive force is not present during the time of the in-situ pressure testing (tested at room temperature isothermal conditions). l 13.3.3 Residual Tensile Load During Cold Isothermal Conditions A tube becomes locked in a horizontal support during operation ,
when magnetite packs between the carbon steel support and the :
tube. When the S/G returns to room temperature, the locked tubes then experience a tensile force due to the differences in thermal expansion between the operating and cold conditions. This tension force (present only during cold conditions) adds to the axial force by the cold in-situ pressure testing. Reference 5 shows that this load varies with the location of the tube in the tube bundle.
L:\94MN\94104
l Enclosure to letter GDW-94-92 Page 23 of 59 13.3.4 Axial Flexibility of Horizontal Supports Maine Yankee's horizontal supports (steel lattice eggerates and S/G tube supported partial drilled support plates) are flexible in >
the S/G's axial direction. When S/G tubes are locked to the supports, there is little change in the axial load passing through the top of tubesheet S/G tube defects from when the tubes pass freely through the supports. This may be seen from Equation 7 of Attachment B, when K, becomes large and the predominate term causing f, to approach F.
13.4 Validity of In-Situ Pressure Testina In-situ S/G tube pressure testing is the desired means of determining the remaining S/G tube integrity where circumferential cracks exist at the top of the tubesheet. As detailed above, the in-situ pressure testing can impose loads on the S/G tube defects that are identical or conservative to the severest S/G tube loading encountered during operation. Testing in-situ ensures the same type of loads are present during testing as during operation. The defect during in-situ pressure testing has the same characteristics as during operation and is not aggravated by the tube removal process. The minimal radiation exposure and relatively inexpensive cost of in-situ pressure testing increases the number of tubes which can be tested. Remaining structural integrity evaluations may be based on a larger and n: ore accurate sample of test data than would be possible from succeeding burst testing of a limited number of pulled S/G tubes.
1 L:\94MN\94104
)
Enclosure to letter GDW-94-92 Page 24 of 59 14.0 STRUCTURAL INTEGRITY OF 1994 DEGRADED S/G TUBES Of the 303 circumferential cracks sized by MRPC in the 1994 S/G Inspection, 24 i exceeded the Reg. Guide 1.121 circumferential crack size limit (79% Avg. TW). '
Seven of the largest defect tubes were in-situ pressure tested to evaluate j their remaining structural integrity. Five of those tubes demonstrated near l compliance with Reg. Guide's most limiting criteria (3 times operating S/G '
tube differential pressure without bursting). One nearly satisfied 1.4 times the maximum MSLB differential pressure before leaking approximately 0.5 gpm (the make up capacity of the test pump). The remaining tube withstood maximum MSLB differential pressure before leaking in excess of the test pump makeup capacity. It is probable that higher test pressures could have been reached if the test pump had a higher make up capacity. As the pressure tested tubes were representative of the tubes with the largest reported cracks, it is expected that the majority of the tubes with MRPC sized defects greater than 79% Avg TW would have withstood the strictest Reg. Guide 1.121 criteria and ,
all would have withstood the maximum MSLB pressure differential (which is the largest Maine Yankee S/G tube loading) with minimal leakage.
L:\94M\94104
i I
I Enclosure to letter GDW-94-92 Page 25 of 59 15.0 EFFECT OF DEGRADED TUBES ON DESIGN BASES 15.1 Anticipated leakaae Durina a MSLB The in-situ pressure testing demonstrated that the circumferentially cracked tubes identified in the 1994 inspection would have oeen expected to remain intact during a MSLB. The testing did reveal t~nat some cracked tubes could have been expected to leak under the MSLB maximum pressure differential. As the three tubes that leaked appreciably during the in-situ pressure testing were pressurized at or near the MSLB maximum differential pressure, the anticipated leakage of any leaking tube in a MSLB is assumed to be the same as that reported in the pressure testing (approximately 0.5 gpm). The cold test conditions 0.5 gpm leakage rate corrected to a MSLB reference condition is 0.67 gpm.
The total number of tubes that would be expected to leak during a MSLB may be estimated by observing the MRPC identified defect characteristics of tubes that leaked during the in-situ pressure testing and observing 1 the number of tubes in each S/G with the same characteristics. A screening criteria of (reported defect volts) times (length in degrees) 2 2,430 includes all tubes that leaked during in-situ pressure testing at Maine Yankee and at another C-E plant. Using this criteria, Table 5 lists tubes inspected during 1994 that may have been expected to leak at maximum MSLB S/G tube pressure differential. As can be seen from Table 5, four (4) of the tubes predicted by the screening criteria to leak were actually pressure tested with negligible leakage. The screening criteria is conservative for Maine Yankee. From Table 5, a maximum S/G primary to secondary side leakage that would have been expected during a postulated MSLB may be calculated. This was determined to be 4.5 gpm at room temperature in-situ pressure test conditions, which is 6.0 gpm at the MSLB reference conditions (552 F and 2220 psi).
A leak rate of 6.0 gpm is well within the charging system's capacity and reactor coolant water inventory will be easily maintained.
15.2 MfiB Concurrent Leakale Effect on Offsite Dose A conservative analysis was performed to determine the offsite dose as a function of the primary to secondary leakage rate following a postulated MSLB, outside containment and upstream of both the Non-Return Valve (NRV) and Excess Flow Check Valve (EFCV) over a range of effective steam generator tube leakages. The analysis results are documented in Reference 10. The analysis addressed offsite doses following a MSLB as a function of the primary to secondary leakage rate from 0.1 to 100 gpm.
L:\94MN\94104 1
Enclosure to letter GDW-94-92 Page 26 of 59 The offsite dose analysis conservatively assumed rapid dryout of the faulted steam generator,100% of the leaking coolant flashes and a constant induced primary to secondary leakage from 0.1 to 100 gpm existed throughout the duration of the accident. Two cases were analyzed. A conservative case used Standard Review Plan (SRP) assumptions in the calculation of offsite doses. In the second (most probable) case, the offsite doses were calculatei using actual conditions (i.e., maximum plant coolant activity 'oncentrations observed since Cycle 4 were used).
Maine Yankee offsite doses (Rem) following a Postulated MSLB with 6.0 gpm primary to secondary leakage along with their associated 10CFR100 limits are shown in Table 6.
The limiting condition from the results presented in Table 6 is the Exclusion Area Boundary (EAB) 2-hour thyroid dose. From Table 6, the expected offsite dose is well within 10CFR100 acceptance limits for the anticipated leakage coincident with a MSLB. Figures 20 and 22 illustrate the sensitivity of offsite dose to leakage rate.
l l
1 L:\94MN\94104 l
l l
1 Enclosure to letter GDW-94-92 Page 27 of 59 15.3 S/G Tube Ruoture Considerations ;
The degraded tubes identified in the 1994 inspection would not have been expected to suddenly fail under c; erating conditions nor would they have been expected to contribute to a multiple tube rupture if any other tube were to rupture. A tube with PWSCC usually exhibits uneven growth ,
through the tube's wall and is expected to leak extensively before it is ;
likely to. rupture. The tube's remaining cross section free of cracks and the noncorroded ligaments present in PWSCC circumferential cracks i can be expected to provide significant remaining tube strength even after leakage starts. This was the case at Maine Yankee where the in-situ pressure testing demonstrated that.the leaking tubes and the tubes L with large MRPC indicated Avg TW circumferential cracks could still be !
expected to withstand 1.74 (2520/1450) times the normal operating differential pressure loads. lubes with this much remaining structural margin would not have expected to suddenly rupture at normal operations. "
This is consistent with the industry experiences noted in Reference 15 where no S/G tube rupture has been attributed to stress corrosion cracking circumferential defects. This type of corrosion progression coupled with Maine Yankee's low (50 gpd) administrative primary to secondary leakage limit should ensure Maine Yankee shuts down and makes repairs before any tube would not be able to withstand its design loads ;
with only minimal leakage.
During normal operations, the degraded tubes identifi.:d in the 1994 inspection would have been expected to withstand the impacting or -
jetting a neighboring ruptured tube may have imparted on them. Should ,
any tube rupture at the tubesheet, the impact force it may impart on a ;
neighboring tube was calculated by ABB C-E in Reference 11 to be I
pounds. The maximum jet impingement force a tube may expect to experience at operating conditions was calculated by ABB C-E in :
Reference 16 to be 136 pounds. Each of these loads is less than the ;
circumferential cracked tubes' normal operations remaining structural !
shear load margin of 215 pounds. This is 60% of the axial load normal !
operations structural margin demonstrated in the in-situ pressure ;
testing [0.6 X (S/G inside diameter cross sectional area) X (2520 - ;
1450) - 215 pounds). The degraded tubes would have been expected to !
maintain their integrity even if impacted by a neighboring whipping or jetting tube.
I L:\94MN\94104 l
I l
I Enclosure to letter GDW-94-92 l Page 28 of 59 l
16.0 S/G INSPECTIONS SCHEDULED FOR 1995 REFUELING OUTAGE j During the next refueling outage, Maine Yankee intends to conduct the following S/G tube inspections:
> 100% Bobbin probe full length exams of all three S/G's.
> 100% MRPC exams of the hot leg transition region for all three S/G's.
- Supplemental MRPC of Bobbin indications of interest to assist in signal characterizations.
> 100% MRPC exam of the cold leg expansion transition region of S/G 3.
L:\94Mh\94104
Enclosure to letter GDW-94-92 Page 29 of 59 17.0 1995 TUBE PULL CONSIDERATIONS Maine Yankee does not presently plan to pull a tube with a circumferential j crack during its 1995 Refueling Outage. A 1990 examination of a Maine Yankee i pulled tube with a circumferential crack is representative of the defective '
tubes identified in the 1994 S/G inspection. The 1990 pulled tube examination identified the degradation as PWSCC. The present defects are sized by ECT methodologies that have shown good correlation with destructive evaluation of circumferential cracks in pulled S/G tubes and 'aboratory. tube samples. The in-situ pressure tests conducted during the 1994 S/G inspection outage are the 3 most representative means of evaluating remaining structural integrity. ;
Little additional information on circumferential crack behavior beyond that I derived from the above mentioned examinations would be learned from additional l tube pulls. Maine Yankee will continue to evaluate the benefits of a tube i pull throughout the 1995 examination. '
1 i
l a
f L:\94MN\94104
- _ _ _ . . _ _ . __ ._ - _ . _ . . _ . . _ . = - - -
1 l
Enclosure to letter GDW-94-92 Page 30 of 59 REFERENCES
- 1. ABB C-E Letter M-PENG-94-008 dated 10/07/94 to C. Eames of Maine Yankee;
Subject:
In-Situ Test Results for Maine Yankee July 1994 Outage, M-PENG-TR-004, Rev. 00.
- 2. ABB C-E Report M-PENG-TR-004 dated October 1994;
Subject:
"In-Situ Pressure Test Results for Maine Yankee July 1994 Outage."
- 3. EPRI TR-101427 Volume I, " Examination of Trojan S/G Tubes."
- 4. ABB C-E Report CENC-1934, " Maine Yankee Steam Generator Analysis of Circumferentially Flawed Tubes at Tubesheet," January,1991.
- 5. ABB C-E Design Analysis MY-SS-900, " Steam Generator Tube Loads for Tube Lock-Ups," October 5, 1994.
- 6. ABB C-E Design Analysis MY-SS-901, " Steam Generator Tube Loads Due to Maine Steam Line Break," October 5, 1994.
- 7. ABB C-E Design Analysis CENC-1965, " Amended Analysis of Maine Yankee Steam Generator Circumferential1y Flawed Tubes at Tubesheet," Dated October 10, 1994.
- 8. ABB C-E Technical Report M-PENG-TR-002, " Maine Yankee Run Duration Limit Evaluation for Circumferential Cracking," October,1994.
- 9. Yankee Atomic Electric Co. Memo TAG-MY-94-051, " Maine Yankee S/G Tube Leakage Assessment - Summary of Results", dated October 27, 1994.
- 10. Yankee Atomic Electric Company Memo REG 203/94 "Offsite Doses as a function of Primary-to-Secondary Leakage due to MSLB with Induced Steam Generator Tube Leakage (Rev. 1)," dated October 26, 1994.
- 11. ABB C-E letter DJA-94-059TD to Maine Yankee dated November 8,1994.
Subject:
S/G Tube Whipping Impact Evaluation.
Subject:
Transmittal of Documents: MRPC Sizing of Circumferential Cracks. ,
- 13. Proceedings: 1992 EPRI Workshop on PWSCC of Alloy 600 in PWRs. EPRI TR-103345 dated December 1993: Paper " Potential Benefits of Zinc Addition to PWR Coolant" by R. E. Gold.
- 14. ABB C-E Report CE NPSD-957 (CEOG Task 729) "S/G Tube Degradation at the Support Plates" by G. C. Fink and S. M. Schloss; October 1994.
i L:\94MN\94104 l l
Enclosure to letter GDW-94-92 Page 31 of 59
- 15. ABB C-E Letter M-PENG-94-013 dated November 10, 1994 to C. Eames of Maine Yankee;
Subject:
S/G Tube Rupture in PWR Plants.
- 16. ABB C-E Letter DJA-94-060TD dated November 10, 1994 to C. Eames of Maine Yankee;
Subject:
S/G Tube Jet Velocity Loads.
l l
L:\94MN\94104
l Enclosure to letter GDW-94-92 Page 32 of 59 JJ!94 S/G #2170 osi SECONDARY SIDE HYDR 0 STATIC TEST RESULTS MRPC DEFECT CHARACTERISTICS OBSERVED
^
VOLTAGE LENGTH % Avg TW Row 89 Line 46 18.84 262 69 1 drop /2 seconds Row 49 Line 122 12.26 360 94 1 drop /55 seconds Row 81 Line 72 23.96 181 47 1 drop /20 minutes Row 49 Line 126 6.23 132 35 Damp Spot TABLE 1 L:\94MN\94104
\
f Enclosure to letter GDW-94-92 Page 33 of 59 CIRCUMFERENTIAL CRACKS DETECTED AT MAINE YANKEE S/G #1 S/G #2 S/G #3 HOT LEG COLD LEG HOT LEG COLD LEG HOT LEG COLD LEG
% 9 % 9 % 9 % 9 % 9 % 9 INSPEC. DEFECT $ IN1PEC. DEFECTS IN$PEC. DEFECT $ INSPEC. DEFECTS INSPEC. DffECT5 IN5PEC. OEFECTS 1990 200% 24 100% 0 300% 3 100% 0 200% 4 100% 1 REFVEL 1992 200% 12 0 --- 60% 3 0 --- 60% 6 0 --
REFUEL 1993 56% 1 0 --- 50% 0 0 --- 56% 1 0 ---
REFUEL 1994 5/G 100% 125 0 --- 100% 26 0 --- 100% 152 0 --
INSPEC.
TOTAL = --- 162 --- 0 --- 32 --- 0 --- 163 --- 1 358 TABLE 2
- L
- \94MN\94104 l
i -
Ya l
?
I Enclosure to letter GDW-94-92 Page 34 of 59 i 1994 IN-SITU PRESSURE TEST RESULTS
.2=w MRPC LEAK RATE !
MRPC % MAX. IN- 0 MAX ,
MAX % AVG. SITU PRESS l S/G # R0W LINE VOLTS TW LENGTH TW PRESS. GPM 1 62 27 9.57 97 360 97 4,800 2
- s.004 1 66 33 8.98 94 360 94 4,800 2 s;.004 1 106 63 14.82 98 320 88 4,800' :s.004 2 38 131 4.68 94 360 94 4,800 2 :s.004 2 49 124 6.36 92 360 92 4,800 2
- s.004 2' 49 126 6.23 95 132 35 4,800' :s.004 2' 81 72 23.96 92 181 47 4,800 2
- s.004 2 49 122 ?1 25 94 360 94 4,000 0.4877 2' 89 46 18.84 94 262 69 2,700 0.496 2 1 70 55 10.41 89 360 89 2,900 .5269' Maine iankee originally identified a 3 times operating differential pressure target ' lower than that identified subsequent analysis.
" Not measured, but assumed equal to the maximum pump capacity of .496 gpm at 2800 psi.
' Flow is reported for a maximum sustained pressure of 2700 psi as 2900 psi could not be maintained once leakage started at 2900 psi.
Selected for testing to identify an expected leakage rate at the maximum design pressure differential.
References 1 and 2.
TABLE 3 L:\94MN\94104 9
r Enclosure to letter GDW-94-92 Page 35 of 59 liAINE YANKEE S/G TVBE STRESSES VS. AVERAGE THROUGH WALL DEFECT SIZE LOADING TUBE STRESSES FDR 3 DIFFERENT CASES OF WALL LOSS 59% Average TW 79% Average TV 95% Average TW r
DYNAMIC PRESS. DYNAMIC PRESS DYNAMIC PRESS.
+ LOADS + LCADS + LDADS PRESS. ONLY PRESS. ONLY PRESS. CNLY LOADS LOADS LOADS 1750 psi 2520 psi 1750 psi 2520 psi 1750 psi 2520 pst S/G AP S/G AP S/G AP S/G AP S/G AP S/G AP Pipe Break Impulse 1.2 N/A 2.3 N/A 9.44 N/A
Response
Pressure AP 16.6 24.0 32.8 47.25 136.7 197.0 SSE 1.2 N/A 2.3 N/A 9.44 N/A Flow Vibration .5 N/A .9) N/A 3.4 N/A TOTAL 19.5 24.0 38.31 47.25 158.98 197.0 _
NOTES: All stresses are listed in ksi.
References 4, 6, and 7.
TABLE 4 i
i l
l
)
L:\94MN\94104 I
l
Enclosure to letter GDW-94-92 Page 36 of 59 1994 S/G TUBE DEFECTS WHOSE (VOLTAGE) X (EXTENT IN DEGREES) 2 2.430 R0W LINE IN-SITU IN-SITU ASSUMED LEAK TESTED PRESS TEST RATE LEAK RATE (IF NOT TESTED)
S/G #1 62 27 YES 0.0 ---
66 33 YES 0.0 ---
70 55 YES 0.53 ---
103 70 NO ---
0.5 106 63 YES 0.0 ---
106 73 NO ---
0.5 TOTAL !. 53 gpm S/G #2 i 49 122 YES 0.488 ---
81 72 YES 0.0 ---
89 46 YES 0.5 ---
TOTAL 1.0 gpm S/G #3 42 89 NO ---
0.5 49 84 NO ---
0.5 54 103 NO ---
0.5 60 99 NO ---
0.5 60 101 NO ---
0.5 65 48 NO ---
0.5 65 58 NO ---
0.5 82 111 NO ---
0.5 86 71 NO ---
0.5 TOTAL 4.5 gpm Adapted from Table A of Reference 1 TABLE 5 L:\94MN\94104
l l
Enclosure to letter GDW-94-92 Page 37 of 59 <
l 0FFSITE DOSES FOR C aom PRIMARY TO SECONDARY LEAKAGE CONCURRENT WITH A POSTULATED MSLB CONSERVATIVE CASES MOST PROBABLE 10CFR100 0FFSITE DOSES (REM)
Pre- Coincident CASE LIMITS existing Spike Spike EAB 2-hour Thyroid Dose 21 17 0.5 300.0 LPZ 8-hour Thyroid Dose 2.5 4.8 0.08 300.0 EAB 2-hour Whole-Body Dose 0.19 0.7 0.01 25.0 LPZ 8-hour Whole-Body Dose <0.02 0.12 0.001 25.0 Reference 10 Notes: EAB is Exclusion Area Boundary.
LPZ is low Population Zone TABLE 6 I
L:\94MN\94104 l
l l
l l
Enclosure to letter GDW-94-92 l Page 38 of 59 l
l FIGURES I
I L:\94MN\94104 l
Enclosure GDW-94-92 Page 39 of 59 MAINE YANKEE STEAM GENERATOR (5) >
TUBE & TUBE SUPPORT ARRANGDfENT ,
STEAM OUTLET N0ZZLE l
[ /[/
Q, __ _ _< >
lC i y.
1 m - - .- e 2
9 g-
// Q FEEDWATFR N0ZZLE l
1 ed
"' m TUBE SUPPORT NO. 9
-U ' ..[ l PARTIAL DRILL PLATE h!il . Q.-
gama.
I .!II ll //
g g , TUBE SUPPORT NO. 8 PARTIAL EGG CRATE
., wo ,
g ..3 I
p?
p 11; l- I j,! !!
w !p !n ,. ir m TUBE SUPPORTS NO. 1 THROUGH 6
! ; I -l
'l( f M ,, ,
I '
g' k N N f[ !
\ .
~s ,
t N*' q M i l, '
/
% )i 1 1
/ I
'E
=
-h i
i -)
/ ,
l i,
i' d j
]
- i' I l u
Q, j .
l l
-i l
i MAINE YANKEE EGGCRATE SUPPORT CONFIGURATION FIGURE 2 j
Enclosure GDW-94-92 Page 41 of 39 UPPER TUBE SUPPORT ARRANGEMENT U Y
~ i
- +
L ROW No.117 i
., g .. _
. __ i 0 ;- - ,
e -f r--
$ ~ "
{ ROW No. 96 .
- t- l ROW No.92 TI .._
y f_
i -TE {Z '
h i T=-
\ _
- ,I E
\ '
\
C iE-~
.-. ir -
, ORILL PLATE 7C-
\_ SUPPORT No.9 L [ h7 ROW No. 60 '
f e
-g ;
O+=-
- h== - '
i w T =- '
Y.
A, i
-TE !
8
}
C" E-- i s N
' [ TE.- t
@h s
\ PARTIAL If iY-' ROW No.32 .
N 'j'i 8 'EGGCRATE '
II SUPPORT No. 8 -
7~ ;
e isiljj 4 , a
)( 'ie " "T h ,r ' ll 's a DRILL PL ATE __/ 't,i [ ,
SUPPORT No.7 i e i .*
+
8 ei. .
,\ >
' I l
{ ,_ j 'l it,ii 8 e it k .li ' l ' ' ' s ' ii',
ll
~
l- l l l ll l l &
~l. l l : .
h ll LI. l 1 llIll ll ll l EGGCRATE
,s _ ROW No.192 (ROW No. I 1
ROW No.18 FIGURE 3
Enclosure GDW-04-92 Page 42 of 59 ra :
O O O O
- 0:0:0:00 to:o:0:0 0:Q:0
-i-
==i. en in mas EYE
/\
h w/
I
\ [
NOTE: LUGS WERE CUT AWAY IN 1978.
4 SEE NOTE DET. A g.
(TYPICAL-lO PLACES)
A N' ~ 7 ~ *nte gGAN
- r _ _ - --
e j f-ROW NO.GO
, . _ _r, _-
j/ r- ~ * .
...< a
- SA s--. _ _ ,,__ t\
cu uc. 52 270' + 90"
\ ~~
. \' a 4~ ~
. .' % . .~~-
.? '
~: ,,-, _ -. ^
\
\
paw kl
~
PLAN VIEW OF MAINE YANKEE S/G PARTIAL DRILLED SUPPORT Pi'ATES Figure 4
Enclosure GDW-94-92 Page 43 of 59 S/G TUBE TO TUBESHEET JOINT LOCATION OF CIRCUMFERENTIAL CRACKS I
l S/G TUBE
! - l 1
~j .
TUBE ROLL TRANSITION REGION u
l I TUBESIEET l /
l I
I CL !
NOT TO SCALE FIGUIE 5 i
? ! ..
.A 5 s . v'0, Y;c--
jn 9.t,. -
j
- s. - <
f
--l' :.
j; ;m,
- ..,. Na
, ,. +
' %.t::
- s ,
4 s
MAINE YANKEE ~
w c c:h '
.'y n . . > - og:
PillhlAltY - SECONDAltY LEAK IIATE '
. - u
- fn }
^ -
m p-l l -
> . w:, <
,q.p ..s
.na
~ :
a MC.< l s.,
,A Qy> *' W g:?~ ^
q2 ;. 5'0' -
om~
~
.mn
- - 9
- .y
- ~
, h, x .,
~p' f.
P- (
~40-
- f- ~ m-
=(.t. y s em ,
.m
' W s 4e : ,
, <5 ;.;y::
.M : .,..;y...p:j
? $$
W:!301: -
/ P' * : e' -
iQ[ ~ ;.~
?^ , s ~
m.... . .. sw -
- - M.B ,
j ' ~ e, 220;
- saw a u ,
g;r. ; :QQ::
- p+ s .>
- :
i !
4 I qf
' @:ce ,dSD
)h n ,
dif/) i 2 B 410 e -
u u
ft y.g
. p.:
em f.a
/q::fp p .<; <
t
'. .s ..- e 3- > -
- 44 ryf 7 7.G
-,/ # ,n .M/Ms w# /ar ~ ~
70 -j , i_. d ,....@ M.mL ~I.. .d. y M d,. l 1n. l j l l l } I [,. q "3 t'1
- y; .. .. s..n..._.,
_; .v s , ;............
.;; s:. m v . ..
., _ .,_...\._ 3. . s. '.
.....:...q.;,y- , . e : . .. q :
ua y p
- 11/06/92i J03/.1.0/931.
- K11/13/93 M, _ , ,.n; ~. . ; ,303/17/9E ^ c 407/19/94TD cp11/2.0/94&u ~ 303/2 m~
~
2.07/12./9.3?.. .. .
. :: : a. -
,n. . . . .
,= x.w 4/95. 6 o
- 01/07/93 '05/11/93? fo9/.12/93
- ;01./f.4/.94. P
- T0. 5/18/94 1 i: ..09,/19/9.4! _ :01./21/95 3;.- O ,E
- LDATEP ,. -
s
~
- ~ -
o 3 3 -
m C) s ::%. y< . ?l -.e. hs:29 - J NAU:. ... <
- 1. 5,Y r .,' U 7
.g g g
- .s v @ I
'^'# - STEAM GENEllATOlt LEAL { llATE ^Wt M 7% -
- k Y' n .a..6_ _ :, w m ;:n w . ,{%g . jg s yf .glc'l . ,, '
~ " '
w-
. 4:m; y u .>. u ,.wFigure
- 4 i m.
I ROM 410
..... ..... 100 90
..... ..... ..... GO
... .. . . . ..... ..... ..... ..... ..... ..... ..... ..... ..... 70
..... ...... ..... ..... ..... ..... ..... 60
..... 50
..... 40
. ... . . ..... 30
.. . . ..... ..... ..... 20
.......... .......... . .. .+w....
.. .. . . . ;.. .. . ..... la
.... . .. . . .l,.....
LINE la 20 30 40 50 60 70 80 90 100 lia 12n 130 ige 4: S/G #2 LEAKING TUDES t1AINE YANKEE UNIT 1 S/G #2 - 00/04/94 2$
W n LOCATION OF MAINE YANKEE JULY / AUGUST 1994 S/G #2 ABB CONDUSTION ENGINEERING *7m LEAKING TUBES $5.
O Figure / mo e
wc W I C
V 1
e N
- ~~~ -
f ~'~~~~ Total Number of CirECracksitiitidted & Ditectibliin all 3 S/G Pir Olitage g l Assuming the 10 Indeterminate Cracks Occuring 1/2 Through Their 2 Cycle inspection Period l t
i V
t e
400 - - - - - - - - - -
! 8 331 1
m l
cu 300 -
.s
.jg tx 200 - ,.
i O j; _ p ,
I e i
o lg100 -
26 1 0
! E i i a -
, i
! y 0 >i i
- : i i i 90 92 93 94 !
, ! Outage Year ;f 9 l i _ _ _ _ . _ _ -
% 0.
!_ _ __ _.___ __ _ ___ ki
- r i $
FIGURE 8
~ ~
~~ ~
~ ~ ~
~~ Total Number of Circ ~Craclis~lnitiated & Det5ctabidin all 3 SIG PeFoutage~ -- 's i
pssuming the 10 Indeterminate Cracks Occuring Late in Their 2 Cycle Inspection Period l l
t I
i 2 i G w
400 --
- o 331 >
u
- 300 -
! E i w i a ~
l i 200 g ts <
E o
j100-3
-]
17 g 1 i b i y 0 'h s > '
s 90 92 93 94 i Outage Year j[
_ _ _ _ _ _ . _ _ , ja g i
e:
SG #1 = SG #2 SG #3 J co ac I T I g FIGURE 9
1 F
i 1
i
[ Frequency o - ~ w . . . ~ . . 3 : : : :
5 i
! i 10 ! l ,
! l i !
, i !
15 m . . <
, ; l 20 mm
+ ! I i ; i I ;
25 i i i l i !
+ 1 1 1 l 1 1 I I I I I 35
+ i ! ! i l i i l I i l I I se 40 0 I i l I i l l l l l 1 l t
-' 45 ;
l l 1 I i l l c 50 A g .
n v I ; ; I i l I i 3 55
$ $ 1 l l I I I l l I l !
g
- 60 1 I l C ,m 65 m 1 l l t !
c -
m 70m '
0 + l I l ! ,
[
Q 75 M '
9 I i ! ,
e 80 l
8 + l i t i 1 85m '
(1 4 ! i I -
, 90 ,
{
r v 95 M
+ : ;
i i
i 1 '
l 100 M d i t i ,
i I I r l 65 Jo 87 a3ed 26-76-MGD a2nsotaug
I Frequency o - ~ w
- w a ~ . . 5 5 ,
E l 10 M l
l 15 20 M I !
25 M
+ 1 ,
30 M -
- i 35m
+ 1
- 0 en
~ 1' l O
p, 45M a i l l l a
8 50m m 1, I
i, 9 m
55 M ,
a n -
I
- > 60 M ,
< T ! !
- 65 ; I l e ~
i C i I G 70 C -
m !
O '
75lI E- -
, I a 8
~
h<
85 t c.o l
! N c0 '
i' E l l
w -
95 T '
i j 100 l
t ,
l ,
c i
l l
t 65 Jo 61 aSed E6-C6-MS a.tnsotou2
Enclosure CDW-94-92 Page 50 of 59 4
i-l l
i M oct E 96 06 1
, ; \~~
' . .! gg %>
j l 1 j
! E 1os N !
> i I . i
~
M91 c. .
_ v i
Q M 01a e l
a i G M99 5 l l
- a > N M09 x o
" , , _ m D
! i g SS h
I i i ! l 7 e 09 $
I I ! ! +
Mcv -
O I J ~
r v>
CV l l l l I l l Y 90 .
I l ! I y DC i l i l i Y 93 i
l ! l ! I t i I i ili 03
~
l l I l M 9L
. l l
E OL i
r 1
, EE f X % R 2
- 2 O
/,ouanbead S * * - ~
j
t p$ m O$:*7eN 0 0 0 m, 0 0 0 0 0 0 m@( g D
t
- n. - $
1 0 9 7 6 5 4 3 2 1 1 s
e
- _ G b j; i {f
) .
0 4
1 N
I R
E T
s u
Jg y :
E I
- _ w f~ . aN[" r 1
N I
s I
J ;r
- nnnII
, ib' h((7 (i ]n [0O ]
y (Oj[h,n"nn1
[ c
[
h ;n- '
- g" 7nt 3 h-J nhg.tVb[n
- m. "
nn
- .f Xn[b i E.
gn [
P 9 3U3N T4N I9E 1 E/I N/
0O G
r a
c n
i"r n'd ,n u n e
fM'n\-(g[y [ E0T r
c -
- 3. ( , - [ < - - o- Eg ( ., n K1S y
u <
g"
,r.
, O ;;" g nn g g M j s
'1 N U AB *
' "dSj '[ . J,h ,-,.d;MA" n M Y r
r p ' , gh [
. Sn lo
'n
,- 9Eac
) . q
- j. ,
N R.. IA/BGB ffn):fn't
- f or n,l ;
[1fjqJ'rhMQ g (y e gk'J , 1 dy ,yh j&y
, ffrj g.:VnMf[',.fj'ng#d"dX[
t r
hf i n j
_. M S A h
mfV, ofe~ y{Jhd 1
(nb i ; , i f n nf '
n -n nnn: n
) ;[ - : y l'i n iwS 0 1
'.'.e }nn
{ ,
1 c
} ,
o g
Nl f r g
gli.n7 n, J
[ n rn[
q j-s.gy 4
ni h, , n
[m a
j =
,. $g"nk(. rsE(([
m r
f
'nel n[y "1 gybU,enmn c.
i y4 f7%[y(s
- - t h
r 0
t,hye@[n g4 ( :y (u jp':[ : ;
l r
[/ J, Q , o 1
{t -
0 L: ,!1 : 1
. f1 3,M1.
, g
- f s,[1.nhh43ha ' n[nr;nf.[p$nnn[g'nh[nh[ oon
- s. 4, , .
S N
s[oO :h(
- t
.W
_ .3 L1 3 O
, . r n
- r (" 1n o I T
g -
r , r 9
I ,
i(:J .[ 9 A 4 f'jjq ;if j f' ,
- .' ,r3 ;" C L
u '
O qk(V ue yMhd'. ,'1(
c1 kk h ,,h\Y.1,{]r i,
- / kk
'3 i:li l L
K
- (y r . [e mh(y rc qM;{nNr ffr)p 1
\l[hi ,, ,,
)k jr :s 0 C
A r
f
'J g -
8 R
, 5 . .
C
', , gi i.
~,. '
y'L b L
u .L -A, 1..,,1 -
, , n n .
. !. ,y n'n' C
1 i: l
[1, / .f '
R
. /,
mn.Mhg I
pi p ':r 0 h % r{j i7: e r_ C 3 l
t
,ps$
/ff(7 g 3(();f y g
r 7
1 1
u,..rr$ . ,
i e
, r l) #
} i, k
, e - ,
r u
[
u -
G g
q- E i /
n S r " i 0
, L b . y,1s.k .W 6 F
' wL' ,.2- g , -
i :
. k :J 'n:
- i. r g 1, f q- I (,T .e 4
. i;fg . r 9 u
,. l :n "
1
)' .W "
C 9 e .
- 7. b' ffnjfn r' yG 1 I
q y fe Q. . ' ., : : -
i.
} h -
,5 3 Y
, n[n,Ji'c, u. M. f 7 Q r,r y1fr[f-[hy4n/ [ 0 L p
3dh ", 1 s ,
.h y=dy3 5 U J
}J. q.E , c' . w
-[4V
/
, , y
n ..f- .
(e s,iq" (n ? n - d. ;n n;" T
,i[5Vq. [
.(mMn:VNf) f
, . g -
g c , ' nV w N] -
S f- n , ' n< m Q [ [
U
(%1
- n -g i ,
(',,!n 1 myu O w [na:n' [ G
,,d* FI J[f'r'1(y,{+(}pgn.m . gkk.
e 0 U R/!1(ahp:
4
- I
- " n 1g n u I e 4 A
c,"1 f [n )[tr % nM y [n
- /
- f. . -
'n :,n pL' i -y a s3\gkf4h)' r
- 1 nn[g'$o t - r : ' : - 4-
- t fr . -h 1 gm]> ,. ..2 n ,h , .1o ., u o[;n t 3
,[
- t f Q)'< rn (( _
p'q( ym.jy n Nt' /h
.o n .n" u ,
n n
_. 0
, j ghn f, ,,q,-
i g ll
. r ,
s,hpLh j n r 3
_ i n" f - ~ M'- ?c 1v
}[of.pgfI
? i r
- . r :
][
t u[J . .uc{( v_I ,:
5 1
_ V .
, D[kl
&{
.pg": :' < ;k 1 ]v/0y[nh t n
_ C J 7 ,:.':
yil t
i/ t 0
2 Ug4li ii '
f, - i
.j" r
_ o, ,
e fy
.. . H. n Q ,
1 a i y [f
[= y '
'i f_
t cIi i, tL r
/ 1 pS
. ?,x i sj v e
l (i i<
J
[] r.( :3 pt ~
l .
,(h1"hl{}[
c 'n
[r-
_ _. _. d'i
_, _ r,$ ' j y
- [ J _
_ _ _ (( .
({:. _
h :L hg6b
([C N E _
N _
_ I _
L
,! l l ' 1 l
Enclosure GDW-94-92 Page 52 of 59
- 3 O O O O C O O O C G O O = = ' G D* CD N ED v3 9 t") N **
g .w ==
en
( (
j 4
l ! -m__q m O W H '
~ - - - - , H,
) "" ' N d IO O__O f~lO f !#NU i
' ' - F ' H I: 1 1. .7 j N 1 I l T ?Ell}ll?EEr?EEln e H Y ?
' V EE ' k nI I' r " T' I l's W l _n O5~(^????ElEEEEf?r^ '.^JMvaAJws+AJMU'J EE?EEEE1 M N
_t*.,
CWJuaJvs
^
^'3. _--~
" ~ * ^.
- _r,_
-~
_ _ _ ' ' { 'g [' ^ - [ 5 [ E T
^ Mb W%H y" y i O1 7 E m
- I .. L 1 yg 2
- L ' . 1..T.
' ' ' _ -_ IR i.I. ....AE..!!.II.!!'I.' r . -- __HT 1: pg I m IM. .m .. r : -L2. I : E I ..: a I. . i. 1
[^ [ .'
- l Z '-,
3 a:,_
' .^'gI ' _ .'} ^[ ,
Y' _I: ._J
- ': .' E . !_ m..I I: E ': .: E :J Ii E E I g j phh, h%t%.hENk%u N h U
- - Y . - . gr -. T _ . T. . ; - p r m._2--. eS M.h%&m s5 w + eu SN %- -
.: n : . M' . Fu'-........
ty 2 g :D DY _ " D'. _ N P IP[I: P P Y 4,.nJJJJAf J sh5A 7 = : =. v = > =l.= l =.=> = l l lv=> >7=ku, =l, =ll-J =E ~ h=,F Ep +ru =l 7 =>, =f, = i=>:ll, =, =lllll l, % =l>llh+vhll lOE r zr ' - 'r'"- _.. j - j r r L i- 5 I i- I i : E EE - 1 l' .; 1! U R Z I- 1: 1 1
-1 -
.' ' t ] E
.: L - -:1'- I i.II : -. Z. !
I T rI i 1: Tn I. EI
..II i I
' '--I, -
T I: I: E I: m I :
'1 O'
..__m:.-'"
.c _E,.. " .Z_: E- I e.g
- j. $n. . _ . " . . .. 1. C '
m:^_'^"": d ' '"T. E' 'If E' ^r[~r d E .I __'miI 1. II L. .. [-'KI.E E: J..L_ _ II' I u i. : IE I I'_- _
'n' n'~n' 'L LL 1: I: E I L _
..Z E 1 '
- ? : ' " ' "
T' VE EE F T' Y r: I- E T E T l
-" 7 E f E E " ' '
T II E E E E I E1 i fJy=J =h=, =l. % =.7 . -J, fJ. =. =ll. =ll. =::3J-ff, =l2,. =}ll-:lll-lllllll,
; ____ :_,_ l' '_1 :_ :
EJE ' '_ U_ I'_I
- ._C - _ _ - - _ . - . ---
p,.__'_'-__E_.'_l._* - _ .- J 7 _ . _ . _-= r
. p- ,-' e4 [
'-I- '. . I -- ': E Z Z F 'E' '.i L . I II ' 1 ?
PI: . .. _' l .~
.. E !.' E
.- - - - ' ~~': E_ .... l_
.m _
.. I
,.2.- L:
. r-
.__:E.' ' , _ _ I " - ' r' 'l L 1 : ~
"E I . E ..' ii .:1.l'y
'_'. --n' :- '1. _'.ti
-E i
.'.'E i - .: h II E. . . . id V V 1 ':- --
- -~ -'
I .1
- . r : :: ry M _ .I. F.. E
" 2 : E__"_ E . . ."Ei._. E :i.:E 'I-'- - --
-.-"1:7 -~1.T.Iy-jd I
' - _ Z l_ I _'
__I
- ..I. . . '_-'..' .yE : : I' : r4F j
k JJh.n. y - -f_' fl,%M., p s% ie% L r J avCYww- wM,vdlly i b g i _ _ . . _ _ .:_--.-__- - } - r .. /~p N
. . , .'" ".:v- Atea. :V R.I
*-n. - - v. - _ _ U. :s. -J.- . -"e,_.-v
.__.........~..._--__-.-7.: u . . - v: :A,*/- y
,p. , y v. .we_ ,
g _ r . . ' . :-- E- - -. -V
- .r 2 1_:
.I_"'_._:>. . ' . '1_-..
. o ,
I . _ H : I_I
~
. IL 1 H I E - I l' I t .
_ ___'_-:..I , -- - ,_ . (I _I- ": I: I- : , : I [; ' _ a _ ' : ' li : I :u..- 1 h __E I I { _ P . .. ' E , I. I l-. D 2 'T . U {v."' "",. - [. pk.' _:
- I[__ . U '7.'*yhb. y "[p ,} .- .-. .- #h r+5&&W YkYO"%V}?r& ^???ErY S EE M EE=Y}??El e
~
-vs.AJ.-JJJ.%~
3 - - "',; AwJAL. -
- 7,=:w--.
%,Ll~JU-NJJMAk.AA -
- 3 3 . : = " ; 79-- - - - - . -- ; ,
w J.40 r-p I :- -' r __m 4 - " : I . I _ .. . 'I r
. '^ ' '.
^
t - ~'-~E .- __'_ E ' : I !Y: t v ~-. W 1.
^
Z T I '
~~~-[~-'1' _.
-_ I s.'=4 1 u %% g __ I - I N I h_ ~ _ n
.L-- ' ** 3.Ah l t hL. ._ . b.. ."2- , _ , . . ' . L E ^ -. .i_[
1. I g g g _ = - - . O g 7,e w9eer=7,.* *^ =re#p* M **wV w' ,#,#4N e*w,'vi.r e.N'VV.* #N%#=*"%sN'"e#sq
%, bo t.k, ,JJJJJJ. h: F 5J. JN... JJ_,h_
_ .__ _ ._ _ KJA.k.h.A.JJJ.rNJ.h.M_ k hv.nH
-y,__. -
.._ .J.J.... ,JJ_ ,_ .J,. : U_ ff_ ,, =h._ : '
f%5,.he.PW"y*%.% %%%*%e%%*%hwh u r .h%he'^f, " , - ,-&^'!,- p,.'^-,'pR Q _2., :.) L 2.. Er r :y _.ng -_ _ ,. ,, ,;L :y i u y, og u - gj _ _ I . I u.
. ,.-. . -_ . - - _ _y _ _. ,.
_._y_', *_, m____ 1
, g @
- a-- , - -
.m g)
..I~.. .1__ _.I.. P :' . . ..: 1 ':'I.'E__
J1
- i. ..
' E
'*~I ..L
..I.. E IS
[10 n:..I..: :. : :..E..E II' -
- .. I: E.000 h
- m: n r I :'E E a r " ': in u
,-m~: 1 : 1
- 77I ': i : :..E_.-
T 1. ; .:' ' ' .. ' ' ' ..__.-E ..I E
Z Z '1 E '-
I L .. 1: E I: I L .! EZ1 E : 1.__4 _. 1. .. I :." JI. ' r T' u . - .. :_E. . .E i T..: . " I' -'.E hkN~%.k" UA -eA ~~N%~hh is . JA~,.FA.". t %%~h,~t% ~w'EEf?+^HJEEEEEEEEE}E?EEE?EEEEn 4 ~A , Q
-A.-Ass si.a as %v4JJJJA%A.AJ.N% %AS L L , H
- ' r. r n .. '
^
M
^^fE --^J
- n-r
.-._--r -_--_ r : : : - 1 _ 1 : - r: : 1 r --- r xi D z1 1 .. t u
- "L 1, g
_1_. 3g. ,. - 1 y L_. _. . ' ry _ E_ _ __E p
,- 7., - 1P ; _
. ,_ f -._I7 f_ y .(1 _h 4
, 1. .. n.n : : r g' , . .' . .. . . . . 'v. ,.-.,I.. ._..,7:vP. :1, .-
T
, y.
*m, or*rhee rA %*e Mh hs**es.%V ^ - .
I I E
.....E,. .-I, .I ' .I 2.- E,,N. r :
.E I
~ mM,.*e&M r . . .-"..",M..' , '-' ' ' . ^ - ~ -
7
.: .- ^.y
; = W "o# At y'"w%"4A[e*V "t= % '%*N 4"~e._U4=%'*.= N%#%r%"w.%M y#yr' _'va**T**Ve*4'.*4A ve 8%.%
l m%%YE!?????ETW !N Y??-Y??^nNE!?l$EEEErl VEE!!?EEl b l 9 A.a.>,.nl59 x=3, =y = =g=lJx= =f. =llg}, = =l?. u._= =ll I' &tr&1 7w .;& h~ W -+ W h~~ L 'E M E H +JEEEEEVr u-. um n: .. g r EJxE r.N I w, E.yy ,j .S ,
, wwg . e .--_c. . ~%% _
.#. 4pe.m. _... ^^ .wgwg.gwp_ : N
"- ' .' .. " ~
. ^^ ^^. . . . . . . . ' , , -
~- , .- ^ ^ ^. . _ .
- -}.^^^^
.- assh- v. g
,y _ _
. _ -- 3 - .. ._ . 1
-._'. 1 .
- 1 J *vhkuvPJi+*v%k, 7 . ' ,[. 'J, '. '^fy, Ll, =l,%%%+%.%% ,
.s.-w=,g%.-_'-_'
._,rA."._-_**^--- 7',,.,-,
m .f- -. edi . . _ __.._..71,.^. .. . c gh~% n.-.-_
~ v e%wg,av.w.w#sv.A#s.wA _ .-
g
^~~~~*t**'*?**%hY
'- - -*+^* ; .:v- YWYM l
~*
. .. _ xx EI I . . : I r r o
A, ,,; . . .w. . .. 'n n: . . . .m". y l n,,,h ,s - , ,X_. ": ,L} _ - - -__; 1 l j b)W e.d - . .. . . - ,-
^
I ' w w_ 3 r m, ._ .. m y l l 1JUVV 4 i l, 4 g
mywomcno ggi* eo t W 0 0 0 0 0 0 6 0 5 4 0 0 3 0 2 NS% uw om me C 1 0 9 8 7 1 R 1 1 s e G d N u
- - - '1 l 3 h) 0 4
1 I R T _
- 1iyJ J i: . 1 ;n .31 E E
s I
- 3q , 'e
,>]. yf.]N;u y nff'rWff(n xnL n J.n g r 1
%-- r N s 1 I I _
G
,ig': 'f d-) .n T4N
,Wa[J ..fn ' hv!1 '~ . OnW .
1 - I9E I k"j(1 n 0N/ a 1 c 3U3N g:J t
, p ' pN, . a R]T h,, ,,,K 1 0O A -
- qi [
t;h,/.y . pjy%M;g- 3 r t 1 nfr'
,'" W,E. ni w1J.
nny E/I E0T K1S d- .X o ,L (I n ynn: H jgr n;nv u . XN U y.
- u:fXt n,/.,nL O AB
, /
_J ]'%? ' ? ,S y.,c
, nn Y M 3 n.W j n.0Wv Hv~.:: t _ '- n 3O
,hpnff'r i
.H. 0E8C L)b 4' ' . 't . 'k - ,. o n, 2N 1 IGB I
- Onoy A/e
_%p<.[L. fJ .H. a g(V n [rn AMy f#". =+; n ~ 7;j .m e s W . tSn i
)fs , i
., . r V :s W ]4yJNhu: V 0,Y , oytp ,% ..njy [ 7r ,
ff n 0,n,f.rWn :?m ;K ~ yfg
- p. ya'] {h4}[ pneMgEjfn
- t n L
Y% N k anon 0 3 a{' -llr.r. u 1 wj' iJ< , ('- b. J
;E' . i 1 ,
1,E '
? . J o[1n,W#{0t c : n aJ'n c ?^ 1 I' 'u.
. .n ."
i , u" fyfL " n My T l
._Y: %nf. 'n.
kwfqj. ,a 4 -(< wf1 .u o[J i - ( r i/' I ; .
' qi , u bg aH{< , L" a nsnO;: ?Cm,.G 4 :
y' , 0
. .3VU JSy ' .:..uy< c.
H~ F , l. kq 0
. kf n o L ,f a
- . . 1 3u yo6 g/v
; i h 1 .
[% y? f
' ff4b I ]n1
(;J.d
- n g b1. n r,fff 4]%h.I :', qvehg,f4. fd r f ",Mhn. ; aMcVvpI%nfyy.3r3J - :.
. . c[. 'w}, , y . i nnr q
' a - ' ' . nn
.a:,N(iCc(p.,.C S s 1[ y .
n ,. . (X' n ' I ' if ,F N
, 7t y> g ,4 pvC' J
.M 7hc.
nni. L 0 O 9 I ks 'f : ( T 1ldl' r k.M #,M1pff4,m : A
< gu.Q1:;4 ' o
,, .r
[ m' s pQWub c j
.o .
# uq i
C O r L
,Q Vt ' W,:h. L n,.
nl . n J
. r i
,j' f , nh5 I
4@1 ,w 7
-v1. NW 1
1. [',. k,I ( n 1 st1
- ' 9?
h'nh+y.a . 7' C. I 0 8 K C A g4. ., AI , r a).IM . g C 4, i,
,$ R nVL i
$- C
'1', E' i .
g
,1 '" j'g- -
L
,n fg, - < .'-
.d'[. f 3ff4'pn,4nI .
- L b
%f' fyn A:,;,4Y N r, I:8QI C
( , P
- r. u h I 0
i
, - R ov 7 Jj l
' @wq~ ,s':g :i E. C , i #g< s1.'}
n i I
- n. .sn- .f J - C
., .' %L n#
;M w: ; %U , !N.
d i' :
-. n- q i .; : : (H" 3 i.
E ,:.. :.' 1y}/, 4wff', t, . , . .mL}g g: ;q
, . .7 a /- G i
p tr4( {L 4. / f A cl I 4-
..ti
. L '
).
0 6 / S 5 1 4' .u/J I ; 4 u' n
-d1nl[' i7r : .Q e n g n nfy y Q G ,4 n, a<r)7rgffV 4 9 r E
j g u 14%$71n#y4(n.,q,,n W,2 Mlp- [yt{UcA @' J%a p 9 [ r5 . i k nr 1nQKh: 1 g cn n ," A' J 1a(u : iL ' 3 .h - .J 0 i nr j[. f h.. I. 4,g-1.-. 5 Y F s t, .
._-t:glO L
?j'1
,1h:
'[ ,
1
- . r
,J J :
i o n ,: :., r U L d] u
, ,h41p4n%"jMfnr Du g J.,n1,rrW"fQo'. J m[WcL 4 4
, r
't n n r, /
.' ; . r, 1
T "f ,.; x i, v
.g N r nn ~
.r nn
.: 1 ;'n n rJ;n n ,u,.@4nt
,L' r.
W'f ' ..N n
. ', g: <
b,
.. 7 f Mj m'n .vv'n g
- J._ E ~ x n
. ? y-7r e 0 S
U _ m , ' r g < , yy1'!,* J 4 G Hh g .n'n ' i
- i 4 . i" Tpsk ;':
f- ;k ,n' 4 na. U
/:. p1l pL }\yjh:e, / 45(t{ Y #L 4QM C. - 4t{1
;h A W
n n. E i . 1 H
'q n '
ln iy , w;p' n- ;X w/. . ,
;n J :
[nW.,,o:1,u '" 7
$' c-
. . w, c;n , i 6y ,Oy: ,w<
e _ q. . y s . 0 _ , 1(fn: q} ,-:3 1I9;g Ifr. Xq, n,i(4yn,.
,/
,. 1 3 _
} '
(,n_
.,1 n,4 3nw 9Q3,n t
u o41 j' L
. i
,4 n -
- r. _ 1V ,K -
7- 4 , I u ~v : [ ym f;y '
-D
.{4 uy .
n
,f,Ng. r M(. ,w{' Lg,:t{ . , n f
,Cs%-{p(h~fq..Ot.
1 i.
.i s ji(
t < J
- - ~
fJ. - 0 _ jp~ ) 2 _ g.a . I, L t 4(U D QqJ[h .
<g.
i _ , ,ff[n JWJjc C
- cD r.
c y.,m .
- Q' .H o J p[n ,,- .-n, nn g" ,c>r .
r r
,, ,; ,Tkn,h
- JV .. ;[i ifmf;;
i a-1 n 1
.M, c ;
yy' peL - .m.e<_;4]n L. 5 ,vD,. n 0 l
- 1
- ' t
< cw; -
7, r t
- .77 .
t
?
_ f cO
- ,J.f g<Oxv.C[i ( J J,q . ; _
_ _ ((( ( E S
- I L
Enclosuste GDW-94-92 Page 54 of 59 Comoarison of MRPC Aversae Crac< Deotis witi Actua Deaths As Determined by Destructive Examination Ca cu atec Average vs. Actua Averag_e ' 100 90 - 80 - O 5 70 - i
< 0 ,
j $ 60 - U } 8 g 50 - m O
$ 40 -
2 0 30 - f 20 - 40 -
/ .
0,/ , , , , , , , , , 0 10 20 30 40 50 60 70 80 90 Met Lab 360 Degree Average Figure 16 Figure 7 of Reference 8
Enclosure GDW-94-92 Page 55 of 59 I t 10 ACTUAL S/G
'^ds*W % F/c5 Circumferendal Crack Data COMPARISONS 100.0 -
w.0 U l l l l alf
; ,7;
, l u a IM g:
. i i i i 7, i .-i izi i i h[l V
.I l . I I 20.0 3,',
J'i6) i l I I A - ' k Q,0 _ i l i I 0 10 2D El 40 50 60 70 80 90 100 FIPC Avg %TW i 1 l i E"T* 17 Reference 12 I
t 1 Crack Average Depth SIG 1,2, and 3 1990 vs 1994 >
#70 i b
- s. A 70l- ,
Q 1990 oo
-*-1994
- so x i E
8e er i E
- u. .
30 - - 20 -
" A o-I d ~: ~ t , : : : : : , ; . : ,
v-n ; s U E E E N E E E E E E O z Z E E E E E E E E E at . E$
- NOTES: hTD is no detectable discontinuity / cracks o NQ is non quantified (a crack is in its initiation stage and quantitative size *8
- characteristics cannot be identified) M '
E . 4, FIGURE 18 " .
. _ . , . _ . _ . - _ . _ - . . ~ . , . . ~ . _ _ . . . _ . - -
. _ _ . . _ _ _ _ _ _ , - ~ _ _ _ _ _ _ _ _ _ _ _ _ _ _
. Enclosure GDi?-94-92 i Page 57 of 59 l t
MSL3 DIFFERENTIAL PRESSURE ACROSS S/G TUBES ( ' 3750 , ,. , , g 2500 , 7 ,--- , , p 2250 rj , , , [2 4 1750 0 / i i t i Ly 7 , , , , 3 1500 , ; ; ; g 1250 , , , , e 1*o i i i i r.o 750
$ 5" ! ! ! !
c 250 , , , ; O 400 800 1200 1600 2000 TIME (SECONDS) k S/G SECONDARY SIDE P E SSURES 1200
^
5 1000 ' y8U l Inca w 400 N I l -~ l E O 200 { Faulted 0 400 800 1200 1600 2000 TIME (SECONDS) CURVES FROM REFERENCE 9 FIGURE 19
Figure 20 CONSERVATIVE EAB 2-Hour Thyroid Dose - Pre-Spike 1000:. i ; t = = t=t-i-; ; _ _ _ _ = = ===:
- rn :: c r c : -
10 CFR 100- 300 REM LIMIT - - - - - - - - - - -- 1 -
- c . .
/
100: .
= - - - -
= - - - -
X -
= _= =_=_.:
y _ _ u, y - f _ _ _ _ _ . ([ .
/
/ -
G) U) o -- -- - - - -
' /' - - -- - - - - - -
o 10:- _ = = == '
< M-_ _ _ _ _ = == :: :
o : :::: - x 1y ::: _ :: : :: g , /_ _
~ - - - - ~
r
$4 y __ '- ._
CONSERVATIVE ASSUMPTIONS _ _ _ _ . . I- /
/ - - - - _ _ _ _ - -
1-
/ .--
x x :: : s -.
~.
/ __. _ _ _ . _ - - - - _ _ _ _ .
/ x ::::: x ::::: x :::: :
tb :3 ya O a 01- - - -- - - - - - - - - - - - - - - M 0.1 1 10 100 eg Primary to Secondary Leakage Rate (gpm) $
=
gn ,y; a1E y ,e 0 e: - _ _ _;_ . = :. _ _- - 0
=: _ _- _ e: _ _ _:,__ = :_: -
g: _ _ _
- ___1 5: _ _ _ s=~ __ _: g _. _ _ = _: -
sr _ _ _ e = ~~ ._ _ _. ; _ E _ _ _ = z: _
- =___ -
_ _ s = ~ __ _ _;____ _ = r_: l*; =___ sr sx- _ _ e=Z__ _;; _~
~
'x _
= cx S N
O
=___ -
_ ' IT P M s U )
- S S m p
A y\ g e '., s E L B ( e k A t y i p - B A- O 0 a it v S -
=: _
=: _-
_ 5: _ _ _: _: _. _ =!_ 5r
- _ R :__-
1 R it c a
=:T _ _ P o =: _-
_ = _ Z _ _ _; _ Z _ _ =n: _ T :__ - e g t n C =x _ _ _ s= .__ _ _ _; _ Z _ _ = rx : _ S O r__ - a lo a _; _ _ - _ er,: _
== _ _ e=Z_ _ _
M =__ - k o4 c a rl e e s=
~ ;_-
____ __; __ __ _ sc__ : _
=__._ - ocy 1 s == _ _ -.
- e t c
2oD -
- L aC
' y eec r
e rd r a mi us n ui o : d imd gr , n x e i y o ar v F h #
- c mes T -
, e nb oo r
u S d o , T I _ e . I _ _ _:_: _ = :_: _
'y E _ 1 ot e
s a l
+ r- _ _ _; _ : _ _ - __ : _ :,E __ B 1
IM L _ e:Z _ _ _; _ Z _ _ - __:-
! __ y r
2 i _ ; _ Z __ _ a z__-
- M _
r ___- _ - _ ss=___ E B i
- R _ = _. __ __ _;__~.
__ - __- _ m A i 0 _ x=' - __ _;;_~~~ _ - _x_- _. _- i r E . 0 3 P l e . 0 0 , b 1 a R - b F . o r i C - 0 - P 1 t s ' 1 o _= _ - 0 M 0 5: _ _ _ 0 5: - _ _ 0 1 5: _: _ j
- E _ _
1 0 0 1 Q 0 0 1 0 . 1 . C pDO)C Do'( .
~
( C j
ATTACHMENTS Attachment A Maine Yankee Run Duration Limit Evaluation for Circumferential Cracking. Attachment B Point of Loading Effect on Validity of In-Situ Pressure Testing l l L:\94MN\94104
ATTACHMENT A MAINE YANKEE RUN DURATION LIMIT EVALUATION FOR CIRCUMFERENTIAL CRACKING i l L:\94MN\94104
MAINE YANKEE RUN DURATION LIMIT EVALUATION FOR ! CIRCUMFERENTIAL CRACKING , M-PENG-TR-002, Rev.02 VERIFICATION STATUS: COMI2L5I5 The sa'Wy4Wauxf dodgn Wernen emia!xe at t=et hat W eMod Y $ gn bs+:w alq C? o .tr / of CNA \Ct>
-n._.~.-.~.,-
yT):YE M ' tthi/ O -- mk e r:_p;m + s i J. F. HALL B. W. WOODMAN R. S. MAURER D. J. AYRES NOVEMBER 1994 ABB COMBUSTION ENGINEERING NUCLEAR OPERATIONS WINDSOR, CONNECTICUT
TABLE OF CONTENTS Section Title 1 INTRODUCTION.................................................... 1 2 BACKGROUND...................................................... 3 2.1 GENERAL................................................... 3 2.2 EXPERIENCE IN OTHER ABB-CE PLANTS......................... 4 3 DETERMINISTIC APPROACH TO CPACK GROWTH.......................... 5 3.1 MRPC SENSITIVITY TO CIRCUMFERENTIAL CRACKS................ 5 - 3.2 EVALUATION OF PLANT GROWTH RATES . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.3 LEFM ANALYSIS FOR PRIMARY SIDE CRACKING................... 8 3.4 SEMI-DETERMINISTIC OPERATING CYCLE LENGTH EVALUATION..... 10
~. PROBABILISTIC ANALYSIS......................................... 11
4.1 DESCRIPTION
OF PROBABILISTIC MODEL. . . . . . . . ! . . . . . . . . . . . . . . 11 i 4.2 MAINE YANKEE SPECIFIC STUDIES............................ 14 l 4.3 RESULTS.................................................. 14 4.4 OPERATING CYCLE LENGTH EVALUATIONS . . . . . . . . . . . . . . . . . . . . . . . 15 5 CON C LU S I ON S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 6 REFERENCES..................................................... 17 7 TAB LE S AN D FI GU RES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
- ~. . .. . _ ~ - . - - . - - _ . . . -- .- - . . - - - . . . .
Section 1 ' t INTRODUCTION i i Non-destructive testing of the Maine Yankee steam generators during a mid- l cycle outage in July-August 1994 resulted in 303 tubes being plugged because of circumferential indications in the tube expansion transitions at the top of i the tubesheet. Several of the tubes, based on eddy current testing (ECT), . potentially exceeded the criteria of USNRC Regulatory Guide 1.121(1) for ! preventing tube rupture. In-situ pressure tests of ten (10) of the tubes with i the largest defects indicated that seven met the Reference 1 criteria. Three -
.. of the tubes leaked to the extent that test pressures did not attain the required 3 times normal operating pressure. One leaking tube was pressurized to 1.4 times maximum expected pressure during an accident. The second. leaking ,
tube met the pressures postulated to be present during a main steam line break accident. The third leaker's average defect size satisfied the Reference 1 i criteria, but it's demonstrated test pressure was limited by the test system ' pump make-up capacity; this tube had the largest leak observed during the leak test.
- t i
Re-evaluation of historical motorized rotating pancake coil (MRPC) ECT data -i indicated that most of the defect indications had been present during prior { inspections in 1990, 1992 and 1993. However, their presence was not reported ! at that time because of geometric influences which affected ECT signal interpretation. The reanalysis of the ECT data also indicated that many of the defects exhibited growth from one inspection to another. As a result of the apparent defect growth, concern was expressed about the rate at which the defects could grow and the size of defects potentially present after the next cycle of operation. The defect size present after a period of operation must I be such that the defective tubes can be demonstrated to be in compliance with , the Reference 1 criteria. To address this concern, Maine Yankee Atomic Power Company (MYAPC) authorized ABB CE Nuclear Operations to perform a run duration (crack growth) analysis , using available field and laboratory crack initiation and growth data. The objective of the study was to deterndne appropriate inspection intervals for i tubes potentially susceptible to circumferential cracking at Maine Yankee by ; 1 F
, . . ._,_ ~-"- ~~"
=Y r f
considering the initiation and growth of defects. The analyses of Maine Yankee crack initiation and growth, which were performed by deterministic and probabilistic techniques, are described in this report. I P e S 5 t i l s I y e I ( B I
Section 2 BACKGROUND 2.1 GENERAL Each of the three Maine Yankee steam generators has 5,703 NiCrFe Alloy 600 heat transfer tubes with a nominal 0.750 inch outer diameter and a 0.048 inch nominal wall thickness. The tubes were explosively expanded into a 20.31 inch . (nominal) thick SA-508 Class 2 low alloy steel tubesheets. This process assured that there are not any deep tube-to-tubesheet crevices in the steam generators and that the transitiens between expanded and non-expanded regions of the tubes occur at the secondary face of the tubesheets. The design hot leg temperature at Maine Yankee is 602*F, but the plant has typically operated at 597-598'F(2). Low level primary to secondary leakage over the time period 1988 to 1990 prompted Maine Yankee to conduct leak tests during the 1990 refueling outage. The testing identified two leaking tubes in steam generator (SG) #1 and one in SG #3. Subsequent MRPC testing indicated the presence of circumferential crack-like indications in these and other tubes. A total of 31 hot leg and one cold leg tubes were identified as having circumferential cracks. None of these exceeded the Reference 1 criteria based on ECT analysis. One tube was removed for destructive analysis which confirmed the presence of primary side initiated circumferential IGSCC in the transition zone. Primary water stress corrosion cracking (PWSCC)was the postulated mechanism of cracking. Most of the other tubes also had distinctive indications of primary side cracking. At the next refueling outage (1992), there were 21 additional tubes with circumferential indications, one of which appeared to be OD ir.itiated and most of the remainder appeared to be ID initiated. A number of these indications had been, based on data review, present in 1990. At the 1993 refueling outage, there were only two circumferential indications. None of the 1992 or 1993 indications violated the Reference 1 criteria. 3
During the 1994 mid-cycle outage to locate and repair leaking tubes, re- l evaluation of ECT data determined that significantly raore indications had been present at earlier outages. Of the 303 tubes identified as defective in 1994, 279 had indications present in 1990. The data also indicated that 217 of these had exhibited some growth during subsequent periods of operation. Many but not all of these tubes were tested in 1992 and 1993 and many indications continued to exhibit growth. The evaluation of these data was the basis for ; the deterministic crack growth analyses described in Section 3. When the data , are re-evaluated, the number of new defects at each inspection is decreasing, which impacts the number of defects projected to initiate in the future. l 2.2 Experience in Other ABB CE Plants Circumferential cracks are not unique to Maine Yankee. Nine other plants supplied by CE have experienced transition zone cracking since the first incidence in 1987. Reference 3 provides a detailed chronology of this cracking. Significant cracking has occurred in three other plants. However, most of the cracking was OD initiated. In one plant, ECT data indicated that approximately half of 87 affected tubes had ID initiated defdcts with the ' balance being CD initiated. The presence of ID or OD defects has not been confirmed by destructive analysis. Two other plants also have several tubes , with ID initiated defects but these have not been confirmed by destructive examination. A re~ view of circumferential cracking data from ABB CE plants indicated that there are not any additional applicable data for crack growth analysis. The extensive PWSCC data in plants supplied by other vendors is not applicable to Maine Yankee because the metallurgical condition of whe tubing is different from that in ABB CE plants and the method of expanding the tubes into the tubesheet is also significantly different (hard roll partial depth versus full depth explosive expansion). Thus, the available data from Maine Yankee only , were used for the crack growth (run duration) analyses. 4 0
Section 3 ' DETERMINISTIC APPROACH TO CRACK GROWTH 3.1 MRPC SENSITIVITY TO CIRCUMFERENTIAL CRACKS Determination of appropriate times of operation between inspections is dependent on crack growth rates and the size of a crack that could be present without detection by MRPC. To estimate the sensitivity of the MRPC probe to circumferential cracks, ABB-CENO evaluated data from ten tubes removed from . ABB-CE steam generators. Pre-pull MRPC data and post-pull metallographic data (which provides data on actual defect dimensions) were available for analysis. Table 1 is a tabulation of the pulled tubes, the MRPC measured amplitude (voltage), MRPC measured depth and circumferential extent, the actual length ' and depth, the MRPC calculated average and the actual average depth. Table 2 provides additional information on the extent of cracking not detected by field MRPC. Figures 1 through 6 depict the actual crack length and depth profiles in ten degree increments of six tubes with the ECT dstimates superimposed. Figure 5 best illustrates the sizes of cracks that can and cannot be detected by MRPC. This figure shows that a crack with a continuous are length of 50 degrees and depth of 40 to 50 percent through-wall (TW) can be detected. This figure also shows that cracks of 20-40% TW can extend for 80 degrees without detection. Other figures show comparable findings, including the fact that very deep cracks may not be detected if the are length is short, 20' or less. , The average depth of corrosion (wall loss) for cracking extending 50% TW for an are of 50' is 7% TW. In addition to the 10 tubes removed from C-E steam generators, MRPC data has been collected on an additional 45 steam generator tube samples with laboratory induced circumferential cracks. This data is shown in Table 3. All of the listed tube samples were sectioned to determine actual average wall loss. Consideration of the combined data (55 tubes)shows that 26 of the 55 tubes have an actual average wall loss of 20% or less. In fact, the laboratory samples show that an average wall loss of 4% can be detected and quantified. 5
This supports the more conservative 20% average through-wall value used in the run time analysis. Additional support for this position in given in Table 1 which includes data from a tube removed from Plant A in October of 1989. This tube had a circumferential crack detected by MRPC with an average wall loss of 19.4%. There is no data, in either the pulled tubes or laboratory samples which would refute an MRPC detection capability at the 20% average through-wall level. The analysis in Section 3.4 assumes crack growth commencing at 20% average through-wall degradation and proceeding toward the maximum value consistent with the Regulatory Guide criteria. This is based on a conservative estimate of MRPC sensitivity to circumferential defects. . Below the MRPC detectability limit there is an unseen (hidden) population of small defects. They may have initiated late in cycle and be propagating at a high rate, or earlier and be propagating at a low rate. All defects must pass through this hidden area before they are sufficiently large to be detected. The hidden population will continue to contribute new defects as long as the PWSCC mechanism is active. In the fully probabilistic model (Section 4), the hidden population is automatically present in the simulation' corresponding to cracks which have initiated but are below the probability of detection i function. 3.2 EVALUATION OF PLANT GRCWTH RATES Calculation of crack growth rates in SG tubes is dependent on the ability of NDE techniques to detect and size defect indications. As discussed earlier, MRPC is the only ECT technique that reliably detects cirewnferential cracks. Crack size (circumferential extent and depth) information from successive inspections, supplemented by crack size data obtained from destructive evaluation of removed tubes are the only data frem which to deteomine growth rates. The ability of MRPC to size circumferential defects has been demonstrated using removed tubes and cracks produced in the laboratory. The MRPC data for each tube analyzed at Maine Yankee included the circumferential extent of the indication and an estimate of the maximum depth. From this data an " average" TW penetration was calculated as follows: 6
l AVG %TW = circumferential extent (degrees) X maximum depth 360 (1) In this calculation, that portion of the circumference without an indication is assumed to be non-degraded, which is non-conservative. That portion of the circumference with an indication is assumed to be uniformly degraded ct the maximum reported depth which is a conservative assumption. Figure 7 compares the MRPC average depth, calculated as described in equation 1 with the average determined by metallographic analysis during the destructive examinations for 9 tubes removed from C-E steam generators. Good agreement is evident. . f 3.2.1 Maine Yankee Crack Growth Rate Analysis These calculations have been performed and the results verified in Reference
- 8. The crack growth data from the 1994 outage was used to assess PWSCC crack growth rates for the run time evaluation. Average crack size data from steam generators 1,2 and 3 were used in this evaluation. Data was available for 1994, 1993, 1992, and 1990. Crack growth rates for each tube' were computed for the entire period (1990-1994) and for incremental periods.
This resulted in four data sets for steam generators 1&2 and six data sets for steam generator 3. The principal statistical features of each data set are shown in Table 4. The'first statistip.1 test performed on the data sets was a comparison between the average growth rate data sets (1990-1994) for steam generators 1 and 3 using the two-sample KS test (4) for distributional equivalence. The result indicated that the two data sets were distributionally equivalent and that growth rates could be pooled. As would be expected, The data sets from the incremental periods (1993-1994 etc.) are generally not poolable with the average data sets and exhibit a higher variability. Application of the KS test among the incremental data sets supports poolability for the incremental growth rate data, I i l The crack growth rate probability distribution function (PDF) was developed l from the entire set of incremental crack growth data from steam generators 1,2 ) and 3. This composite data set included crack growth rates computed on l operating intervals from: 7 l 4
l 1990-1992 l 1992-1993 1993-1994 l 1990-1993 1992-1994 The composite data set consisted of 683 apparent growth rates representing crack growth over time intervals from 0.5 EFPY to approximately 1.8 EFPY. This set was ;onsidered most appropriate for a one-cycle run time limit analysis. t o f *'..e 683 data points, 62 data points indicated zero growth. An additional 56 data points indicated small negative growth rates which are, of ccurse, a , result of NDE interpretation uncertainty. The distribution function fit to this composite data set is shown in Figure 8. A 95/95 probability / confidence limit was computed for this data for use in the run time limit calculation. A non parametric upper limit was established using the order statistic for the I data set. The value for the upper 95/95 upper confidence lindt for this data was 40.23%TW/EFPY. It should be noted that during the incremental periods, only 80% of the active cracks showed growth. In addition, there was no apparent relationship between growth rate and defect depth; 1.e., the larger defects did not grow faster than smaller defects. i 3.3 LEFM ANALYSIS FOR PRIMARY SIDE CRACKING ' The' apparent crack growth rate distributions extracted from field data reflect the high degree of uncertainty associated with limited observation capability particularly regarding reasonable initiation time estimates. As an overcheck, a physically based model for crack growth was developed based on a PWSCC model l used for crack growth assessment in CEDM nozzles (5). The basic PWSCC crack growth rate function is shown in Figure 9 as a function , of applied stress intensity for non-coldworked Alloy 600 material. The I function was developed from the data of Smialowska (6) and is applicable to operation at 626*F. Computations for circumferential cracking in the steam generator tubing were performed by integrating the growth rate function using an applied stress l intensity given by: 8
Kr = 1. 95 onYD (2) where: ex = residual stress
- D = crack depth
- assumed at yield state for tube naterial The computational algorithm used for this process includes effects for operating temperature and degree of coldwork. The initial crack depth assumed in the computations was approximately 20% through-wall. The degree of coldwork assumed was 2%.
The crack growth as a function of time based on calibrating the LEEM model to the mean of the MRPC data to adjust for temperature, residual stresses, and metallurgical conditions, is shown in Figure 10. A semi-probabilistic methodology was implemented to assess variability and provide an independent assessment of the crack growth rate PDF. This was accomplished using a Monte-Carlo simulation process in which the error distribution in the correlation of the Smialowska parent data was explicitly modeled. The resulting PDF adjusted to the mean value of the PWSCC field data is shown in Figure 11 in comparison with MRPC based distribution. The results suggest that a mechanistically based model can significantly reduce the variability of crack growth rate estimates when incorporated in an overall probabilistic model. The upper 95/95 probability / confidence value for the crack propagation rate is 39% TW/EFPY from this simulation. The lower variability of the laboratory crack growth data upon which the LEFM model is based, is a result of a more consistent crack growth measuring methodology. The crack growth information available from MRPC data is inherently more variable because of difficulty in interptttation. The LEEM , analysis results are supportive of observed MRPC variability. 9 N
i 4 3.4 SCHI-DETERMINISTIC OPERATING CYCLE EVALUATION I A high' confidence level estimate of the permissible operating cycle length can
)
be made for IDSCC using the upper one-sided 95/95 probability / confidence estimates of crack growth rates: tmax = AW/vn where: tmax = perudssible operating time AW = allowable crack progression Vn = upper 95/95 estimate on growth rate . (corrected for temperature and wall thickness) The allowable crack progression during an operating cycle is the difference between the Reg. Guide 1.121 limit and the MRPC detectability limit. The revised RG limit for Maine Yankee is 79% TW. The MRPC detectability limit for cirewmferential cracks is 20% TW for average crack depth (> 59%TW peak) . The allowable progression is therefore 59% TW based on average crack depth. ' The permissible operating length based on MRPC data is 535 Days. The permissible operating length based on the limited LErM analysis is 552 Days. 1 I I i i 10 b t
Section 4 PROBABILISTIC ANALYSIS
4.1 DESCRIPTION
OF PROBABILISTIC MODEL The probabilistic model developed for the Maine Yankee run time analysis was designed to reflect four basic processes: , Circumferential crack initiation a Crack growth MRPC inspection !
- Repair-removal of defect from service Each of these processes, with the possible exception of repair, has large components of associated uncertainty. This characteristic requires the implementation of a fully probabilistic overall model in order to realistically assess the permissible operating / inspection periods for the Maine Yankee steam generators.
4.1.1 Crack Initiation and Growth Models , The crack initiation process is simulated by the well known Weibull time-to-failure function given by: P (t) = 1. - EXP [- [ T Yl (3) l(B )J where: P (t) = proportion of tubes experiencing cracking by time T N = shape parameter B = scale parameter The parameters for the Weibull distribution have physical significance. The scale parameter (B) is the exposure time at which approximately 63% of the population at risk is expected to have failed. The shape parameter (N) has significance in terms of hazard rate as shown in Figure 12. For N<1, the 11
i hazard rate is decreasing such as is typical of electronic components during
?.he " burn-in" period. For N=1, the hazard rate is constant and the Weibull madel becomes equivalent to an Exponential model. For N>1, the hazard rate increases with exposure time, a situation typical of most corrosion phenomena.
Gorman (7) has provided an extensive catalog of Weibull parameter estimates based on worldwide operating plant data. This of course represents data from cracks which have pregressed well beyond initiation into the detectable regime. It can be experimentally shown, however, that this represents only a time delay function for reasonable crack growth rate distributions which can be easily accommodated by a minor adjustment to the scale parameter (B) . For Maine Yankee which has experienced approximately 350 circumferential IDSCC cracks, the parameters of the initiation models are given in Table
- 5. In both cases the scale and shape parameters were determined from the MRPC data obtained in the most recent Maine Yankee outage. The models are the result of two methods of obtaining Weibull parameters from the Maine Yankee data. The first method used a Least-squares approach to extract the model parameters from the data.' The second method was an exact fit to the 1990 and 1994 cumulative results. It should be noted that the estimated shape parameters are both less than 1, suggesting a decreasing hazard for initiation of new cracks. This is consistent with observation during the 1990-94 time period for Maine Yankee.
The crack growth model was developed as part of the deterministic analysis. A Gamma representation of the probability density function for the Maine Yankee crack growth data was utilized to describe the crack growth processes. The parameters describing the crack growth rate model is given in Table 6. 12
4.1.2 Inspection / Repair Model The inspection / repair process for circumferential cracking is completely dependent on the capabilities of the inspection methodology, in this ) case MRPC. Simulation of the inspection / repair process requires the probabilistic implementation of a probability of detection function (POD) such as that shown in Figure 13. Implementation of the inspection process is straightforward. At a given inspection time, each of a group of active cracks is subjected to a probability of detection function. . Corresponding to each crack is a POD based on the function. A uniformly distributed random number is chosen to decide if a given crack is detected. If the crack is detected it is removed from further service or subsequent inspection. 4.1.3 Simulation of Overall Process The computer simulation of the overall process is shown conceptually in Figure 14 and consists of steps reflecting the four basic processes discussed in Section 4.1. i The initial sequence involves the computation of crack population size, crack initiation times, and crack growth rates for given Monte-Carlo trial. The population size (NMAX) is determined from the Weibull initiation model and the final inspection time for a given case. The crack initiation times (TI t ) for a given sample are determined from a
~
randomized selection using the Weibull initiation model. Gamma random deviates are used to obtain crack growth rates (V i ) for each crack and are assumed to be constant throughout the propagation process. l l l The second sequence computes a matrix of crack sizes (Dt3) for each inspection time (T 3 ) in the simulation process: Di3 = Vt (T3 - TI ) (4) t Di3 = 0 if: tit > T3 (5) 1.e. crack has not initiated prior to inspection i = crack identity index j = inspection time index 13 1 i
The third sequence is the simulation of the inspection / repair process. The process proceeds in time sequence beginning with the first inspection time (T .2) . Each crack is subjected to a pseudo-inspeution 3
~~in which the POD function is used in conjunction with a unifoca random number to determine if a given crack is detected. An identification matrix (I 3) is used to track the status of crack detection and repair (removal from service.
It3 = 0 for undetected cracks It3 = 1 for newly detected cracks It3 = 2 for repaired cracks ! (In = 2 if I ).3 = 1) t The final sequence in the simulation is the examination of newly detected cracks (In = 1) to determine the number of exceedances of RG ' l.121 criteria. This is performed by examining crack depths (Dt3) for each inspection period. For Maine Yankee a 75% through-wall criterion was used. This number is conservative with respect to the 79% through-wall criterion for Regulatory Guide compliance. The above sequences, taken in total, represent one outcome or Monte-Carlo trial. The overall process is repeated 1000 times to develop distribution functions for the number of RG 1.121 exceedances for each inspection interval. f 4.2 MAINE YANKEE SPECIFIC STUDIES These calculations have been performed and the results verified in Reference 9. For the Maine Yankee studies two cases were run. Each case investigated the effect of operating run lengths varying from 425 day operation to 500 day operation . The cases each simulated one of the crack initiation models. Both cases were run with a POD function which permitted detection of no cracks with average depths less than 20% through-wall and permitted detection of all cracks with average depths greater than 21% through-wall. 14
I 4.3 RESULTS 1 l The output.for the cases is included in the appendix A tables. For each case, ) the probabilities of various numbers of tubes exceeding RG 1.121 limits are ! given for three run lengths and three future outages. 4.4 OPERATING CYCLE LENGTH EVALUATIONS The allowable operating cycle length can be obtained from the case set results using a probabilistic acceptance criterion. For each case set the following ; example acceptance criterion was applied: !
~
There must be a 90% probability that one or fewer tubes will be expected to exceed 75% average through-wall penetration during the current operating period. ! For both cases, the acceptance criterion is met for run time of 500 days. These estimates do not contrast sharply with those imposed by the semi-deteoministic analysis in Section 3 of this report. l
. l i
i l 15
Section 5 CONCLUSIONS This analysis of crack growth data from Maine Yankee resulted in these conclusions:
- 1. Primary water stress corrosion cracking nmy poes a relatively limited threat relative to exceeding Regulatory Guide 1.121 requirements at the end of an 18 month cycle. The apparent growth rates exhibited at Maine Yankee result in a acceptable situation using the semi-deterministic evaluation methodology.
- 2. The fracture mechanics (LEEM) approach to crack growth is supportive of the variability seen in the MRPC field data.
- 3. The fully Probabilistic approach results in an acceptable probability of one or less RGl.121 violations at the end of a 500 day run (18 month cycle).
16
Section 6 REFERENCES
- 1. US Nuclear Regulatory Commission Regulatory Guide 1.121, Bases for Plugging Degraded Steam Generator Tubes, August 1976.
- 2. P. Magee, J. F. Hall and M. L. Fortier, " Destructive Examination of a Steam Generator Tube From Maine Yankee", CE-NPSD-597, December 1990.
- 3. R. S. Maurer, " Chronology of Expansion Zone Circumferential Cracking," -
EPRI NDE Workshop, 1993.
- 4. M. Hollander and D. A. Wolfe, Non-Parametric Statistical Methods, New York, Wiley.
- 5. " Safety Evaluation of the Potential for and Consequence of Reactor Vessel Head Penetration Alloy 600 ID Initiated Nozzle 6dracking," CEN-607, May 1993.
- 6. R. 3. Rebak, A. R. McIlree, and Z. S. Klarska-Smialowska, Effects of pH and Stress Intensity on Crack Growth Rate in Alloy 600 in Lithiated +
Borated Water at High Temperature," Proceedines of the Fifth
. International Symposium on Environmental Degradation of Materials in Nuclear Power Systems - Water Reactors, ANS, 1992, pp 511-517.
- 7. J. A. Go rman, R. W. Staehle, K. D. Stavropoulos, " Statistical Analysis of Steam Generator Tube Degradation," EPRI NP-7493, Sept.1991. l l
- 8. Calculation Number: M-PENG-CALC-001, Revision 00, 11/11/94.
- 9. Calculation Number: M-PENG-CALC-002, Revision 00, 11/11/94.
i i
...-.wa ,.,9 -
P
-SECTION 7 TABLES AND FIGURES a
g E
. t
't t
4 h i e I l i e I
TABLE 1 MRPC CORRELATION'WITH METALLOGRAPHIC RESULTS ON REMOVED STEAM GENERATOR TUBES WITH CIRCUMFERENTIAL CRACKS PLANT DATE ROW LINE MRPC MRPC ACTUAL DIFF MRPC ACTUAL DIFF MRPC ACTUAL DIFF REMOVED VOLTS % MAX % LENGTH LENGTH CALCULATED AVG % AVG % A 01/88 19 25 68.8 88 100 -12 254 - 62 73 -11 10/89 22 52 13.2 83 91.3 -8.3 195 200 -85 45 47 -2 23 145 2.7 63 64.4 -1.4 104 210 -106 18.2 19.3 -1.1 14 118 ) 3.6 70 71.1 +0.9 249 350 -101 49.8 49 +0.8 B 04/92 13 147 80.3 89 100 -11 360 360 0 89 94 ~5 55 6' 39.8 88 100 -12 360 360 0 88 88 0 3G 130 7.1 80 100 -20 360 360 0 80 88 -8 09/93 79 83 1.0 87 71 +16 181 64 48 6.3 87 100 360 -179 43.7 44.5 -0.9~
-13 239 360 -121 57.8 65.1 -7.3 C 05/90 75 34 11.1 96 100 -4 126 360 -234 33.6 45 -
11.4 M t 9
. , . , , - - , - , , . ~
r, I t
, TABLE 2 MRPC LENGTH ESTIMATES VS. ACTUAL CRACK LENGTH DATE LONGEST
- ROW LINE MRPC ACTUAL PLANT REMOVED LENGTH LENGTH PEAK DEPTH ASSOC. ARC l CONTIGUOUS '
NO LENGTH ARC LENGTH DETECTED >40% NOT A 10-89 DETECTED 22 52 195 280 45% 10 30 A 10-89 23 145 104 210 A 1089 62% 10 10 14 118 249 350 57% t C 5-90 10 30 75 34 126 360 47% B 10 10 4-92 13 147 360 B 360 N/A N/A N/A 4-92 55 63 360 360 N/A B 4-92 36 N/A N/A 130 360 360 B 9-92 79 N/A N/A N/A 83 181- 360 48% 10 B 9-92 64 10 48 239 360 95 10 40 t M 5 b _ . . _ _ _ _ __ _ _ ___ _ _ - -- m _ . . _ . _ . _ . . , . . . _ _ . . . . , _ - . - . _ . . ..-e.. , . . , . . , . . . ,.e -__-,,...__.._e, ...,m,m. . . - , . . _ _ , _ , . . ~ _ . , . ~ . . . _ . . - , . . . . _ . . , _ _ _ _ . . , , .
.. - . = ~ - .. .-__ - - . . .
i TABLE 3 ' LABORATORY CRACK DATA NDD = NO DETECTABLE DEFECT SAMPLE I VOLTS MAX % i LENGTH t A VG% P ACTUAL 7 l 13.47 96 67 18 13.5 8 i 9.13 98 46 t 13 10.5 9 I 13.94 8 95 : 50 ' 14 i 10.8 10
- 14.93 i 100 50 i l 14 .
15.1 11 8 19.9 99 ! 62 i 18 9.9 ; 12 } NDO 1 1 i O 13 ' 1 9.54 85 . 62 14 10.63 I 82 l 86 15 20 1 l 12/n - 19.8 15 4.38 70 79 i 1 16 ! 22.1 16 8.38 86 67 17 1 1 13.2 17 11.81 8 81 104 ! 24 1 21.3 _ 18 i 39.22 8 89 118 ' 30 ' 28.4 19 a 16.07 e 87 58 a 15 1 10.4 20 l 5.2 89 39 10 ; 9.8 - 21 1 14.38 53 ' . _89 14 1 13.3 22 9.66 90 55 14 . 10.5 23 1 9.15 90 62 16 14.9 24 i NOD ' 25
! O
! 6.84 93 39 11 -
6.3 26 3.36 97 ' 30 i 1 9 i 7.1 27 9.13 . 99 34 i i 10 I 4.1 28 NDO i l ! i 1 0 29 6.13 I 82 1 90 ! 1 21 3.9 30 1.9 89 I 41 11 3.9 31 24.6 91 1 229 58 69.8
' 32 1 24.3 92 100
?
28 41.4 33 ) 2.94 t 75 44 ; 10 1 33 34 I 24.65 i 86 I 78 19 1 70.8 i' 35 1 29.8 94 i 93 25 i 15.7 36 1 2.03 81 ' 39 i 9 : 4.5 37 3 10.09 ! 88 I 46 i 12 17.1 38 3.97 89 27 i 7 i 5 39 I 6 21 ' 86 1 34 ! 9 l 10 4C l 20.14 I 95 ! 55 i 19.4 41 44.89 I 15_ _ t 92 1 90 t 23 22.8 j 42 63.12 95 229 i 61 48.6 M-4 28.99 80 145 32.2 36 M 11 27.34 96 176 46.9 53 M 13 20.81 90 219 ; 54.8 50 M 14 32.45 88 234 I 57.2 58 T8 7.12 74 139 28.6 58 i T 12 24.79 75 345 T7 71.9 77 19.61 64 160 28.4 31 T 10 10.07 81 331 1 74.5 75 T9 28.03 74 222 T 13 45.6 56 30.23 74 196 T 14 40.3 39 2.12 89 66 T 11 16.3 16 1 7.99 91 308 77.9 72
. , . . - . A. ,_ . .
.. .- .- _ _n , - - ~ . . . . . . - - . -- . . . . -. - - -- . . . - b TABLE 4 STATISTICAL
SUMMARY
OF CRACK GPOWTH RATE DATA STEAM TIME PERIOD MEAN STD SAMPLE SIZE i GENERATOR GROWTH
- DEVIATION
- 1 90-92 15.7 15.5 l 116 1 92-93 11.1 14.3 77 ,
1 93-94 6.3 15.4 77 i 2 90-92 6.9 5.9 11 i 2 92-93 9.7 21.0 ' 3 2 93.94 10.7 6.1 7 + 3 90-92 12.7 14.4 97 , 3 92-94 8.4 9.6 97 - 3 93-94 0.5 12.3 76 3 90-903 12.8 9.3 76 ! 3 92-93 9.1 13.1 22 1,2 & 3 92-93 17.6 ' i 13.7 23
% Through-wall / EFPY !
i TABLE 5 5 t WEIBULL PARA} TIERS FOR ' CRACK INITIATION MODEL CASE 1 CASE 2 ; GAMMA (SHAPE) 0.74 0.37
- SIGMA (SCALE) 2600. 500000 I l
I 6 i I l
~ ~ , , . . , , v., . . , . . , . - . . , -.
TABLE 6 PARAMETERS FOR GAMMA CRACK PROPAGATION MODEL ALPHA (SHAPE) 1.328 BETA (SCALE) 0.08826 b 4 l l l l 1 1 I
5 Figure 1 Depth anc Circumferentia; Extent of Cracx in P ant A - ube R22_52 (Wall Penetration vs. Angle) i 3gg Percent % Thruwall 90 - si .\ , 80 -
\
-- MRPC -
f.. f \ ' r.l\ !.
,0 t -
- p. l[ z+
4.k
\ .,
1, L., - 60 - l .. ... M- i { s, j ey - 1. . ,' a
.y (\, !,
50 - - . -
~
. . \.,
Nj . \11 'N__, l 40 - [ : l e , ;
. 1 r
~~ r-P ...
vi;. :" j.. . . 20 - "-
- mi ,
.;. . \
~ _ _ _
~
10 -
..._.m..
~
i \' v 0 ' 0 b N b b 160 150 lio 150 150 250 250 230 250 250 350 3 Degrees (MRPC = 83% Thruwall x 195 ) .=
Figure 2 Jepth anc Circumferential Extent of Craci in 7 ant A Tube R23L 45 (Wall Penetration vs. Angle) 100 Percent % Thruwall 90 - l 80 - . i 70 -
,, MRPC 60 - x v\ 1 MRPC
\ >}
\ g 50 -
\i l; r
- . - j, 40 - I '
- t s
' , ti .
1 30 - I, '
, g g;; \ l
. . . d.;y .
20 - ' i ,..
/
Mg ;>,.....--
\
. < v1.
, .; . -l. .
10 - '- j
\ -
1
.\
\ 1 .:
0- .
\'
0 20 t.0 60 50 . . . . . . . 100 120 140 150 150 200 220 23 250 250 3 Degrees i (MRPC = 63% Thruwall x 104 ) 23. l
.p.. -
Figure 3 Death and Circumferentia; Extent of Crack in ') ant A "u;ae R14L:18 (Wall Penetration vs. Angle) 100 Percent % Thruwall l l 90 - I! 80 - . MRPC MRPC i 70 - 3 (
~
)J \
60 - f; '
;;{ Yid
~
t 50 - \
~
\ ...'
j g
't ,, ,
n d
'{
\,
! . . 22:
i lg i ' ~ 40 - s {
!j
\
1 % ' 30 - 1 ! l" l
.+i.=..~ .
a., ga -- m:: -- 20 - '
.i . * * .
10 - ,
~
0 . . . , , , , , , , , . 0 20 4 60 50 100 120 10 160 150 200 220 2'O 250 250 300 320 340 3! Degrees (MRPC = 72% Thruwall x 249 ) at
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ = _ _ _ _ _ -
Figure 4 Jeath anc Circumferentia;. Extent of Crac i in Plant A l'u.ae 975; 34 ! (Wali Penetration vs. Angle)
,gg Percent % Thruwall
- G', A
~ 1 90 - ..+
MRPC 7 80 - !
~
I 70 - i _ .}
\
60 - l' l- ... 50 - i
* ==
- 1 A \ .
\
40 - ;\
\.,* ..\
i s..
~^
/
30 - ' _ 3
,-- 'Y.\ . =:
; .. g=. i n d.=I::S
- .g '. - - - -
\
~
s
- .g ..
20 - '
.; ~~'
i.e .
.."" .. i ..J3.; -
~...
.. .=n
- c. .
... :n
.. ,;:c..:.
c m. 10 - ..
.m 0 , ,
O 23 4 50 10 100 120 140 150 150 200 220 240 250 250 30 Degrees (MRPC = 96% Thruwall x 126 ) W44 i
Figure 5 Dept 1 anc' Circumferential Extent of Crac in 31 ant A ~~uae "179L83 (Wall Penetration vs. Angle) ' 100 Percent % Thruwall 90 - MRPC MRPC 80 - ~ i 70 - s b 60 - 'Ns '
.. i ..,
i *
.i 50 -- \ ^ "\ .
l
\
w/^\ / : , V . . . .
~
I t
- f. .. . _ . .
40 -
\
'/ \ ~ \.
i ( I
\l \i A I -
30 s S ,['\.y
~
~
\ .
;. , i.9>
20 -
; O . .:> -
X:_..
~~
1.,.. 10 - . 0- . , , , , 0 20 t.0 , , 60 50 100 120 100 150 150 200 220 240 250 25 Degrees (MRPC.= 87% 2 Cracks - 130 & 51 ) 23.; e _,_a-- " ' * " _ ._a
l Figure 6 l Death anc Circumferential 3 Extent of Crac! in ant A -ube R64L48 (Wall Penetration vs. Angle) ! 100 Percent % Thruwall . i d5 .
. "]: ;. !
..=.. .
( 90 - :.!" . ]
'+-
I ,MRPC A 80- ,' ' ' ' t .
... .\s i
~ ,
g ..
\
70 , , { '\.l- \
?
j .:.-,. ) ,a. r. ;
> . 3
(
\ \ =,.
60 - t. t
\ .=,.
ia... ,
- =- c .;
e T;:n j *
- /.:p:!.-
r . e 50 - {~ -_. -
\ .
;. .. . +
~. t<
. . .c ' =
i t 40 -
\ \
1
't j \ 's 30 - .,
'-- .. t.,- , :-:c#
k.:. ..
-J. '
~e...
- c. 2 - . v ...
. .pj: ; . . .
- E' 1.. ry:!dF.:
-e .
20 - s, -
- b.
. ,,.l m.
. . =i F ~
-- W:.
.a. -
10 -
..- ~
0 . . . . . . . . 0 20 4 50 50 100 120 140 150 150 200 220 240 250 250 300 320 34 Degrees - (MRPC = 87% Thruwall x 239 ) .
- -%,. ,- .-.a , ~ . - r. , -, ,. .e J. 4 J
Figure 7 Comparison of MRPC Average Crack Depths with Actual D As Determined by Destructive Examination Ca cu atec Average vs. Actua: Average 100 90 - 80 - . U
$ 70 -
a y 60 - cn O C2 o 50 - n b a 2 a , 30 - ) 20 - 1 10 - 0, , 0 10 20 30 1 40 50 60 70 80 90 i Met Lab 360 Degree Average s.fa
_, a ..a e. s&am .-. m _ w - FIGURE 8 DIST. FUNCTION FOR MY CRACK GROWTH RATES
" a 120 -
. ', l
?
100 - - -
~
> 80 - - - .J o - J j ~
l (
. Z j W '
D 60 - - - ._ O 4 m
% 1 1 40 .] .
7- J l
' 3 20 I-
- i J !
~ J l
- J
-- - - J l O'-a - > - - a._.a_2_ 1j 20 ;
0 40 60 80 100 1 I CRACK GROWTH RATE - %THRUWALUEFPY 1 1 I l 1 I l W' " h 7 4 y* /-----__--._-_-m
Figure 9 CEFM Crack Growth Function c 1.0E 9 - (From Smiaiowska Data) o : o : W s 2
- 1.0E 10 :
2m 5 1.0E 11 E
.c. -
3 : o u : g 1.0E , x : o : m _ d 1.0E-13 . . . . . . . . 0 10 20 30 40 50 60 70 80 Stress intensity MPa SQRT (m) i l
- (
CRACK GROWTH vs TIME OF OPERATION FROM 20% DEPTH to 75% DEPTH
-,....g..,..,.. , , .
..;....7._.. .g .g .. T-* 'I 1 e e je 't~ s '~ f * '" l 0.04 -
i ! to 0.035 - W
- I ,
O 0.03 j. ! x - r *-
- n. 0.025 -
m O
- I.-
w *. 0.02 0 o ..(.. 0.015 0.01 . t.- . . . . = ..l . i. ...u i i ....t. . ._4 . . ... . I _. .. . l .... . . . . t . i . l 0 10 20 30 40 50 a. TIME, MONTHS
5 9 k
' Figure i I
' COf1PARISON OF 11RPC AND LEFM CRACK RATES t 1 4 B.M - ....... ........... ........ . ... . . ............ ... . ....... . . _ . . .. .. . . .. . . . . . . . . . . . . . . . . _ r . . . . .
. f .\ .
~\~~~. --- "
- ~
0.0S -- - - ~
- !. ". I. i. ~NRPC B5. SED .
- i. -
/ .!. \ . .
g _: i I - -: 1. EFM : ll . . .
\ . . .: . :. .-
.. y g,g4 -.:..
.l~ .5. -
- . ~.;.
. ~ .
4 . I \ . 4 . ( \ . + _ . . . + .= Q .
.t . . . .
-j - -
\
- - t' '"':- - - -
t- '- -- . r 0.03 -. .t. - l
---"t.- . \. .
. l
.\
so - .. .
\
o f . o: 0.02 --. .- - . - . DL . . . . . .. - l
\ .
. . . i
. l .
l
. \ .
. . t g,gg - . . .
.j......... ...
.....7.
..g....................7..
t
/
. 1
. . s .
i / :. :. .
. ~ .
M =
"."".k"'"'~'I
- - ;~~ -
i. 8 .t~~ ~~- b "- ~ l-
. ". f
-- -" ~ ~ -
t . ...t ....t . . .' . t ....t . . . . I . . . . 1 . . . . 20 30 40 50 60 70 80 e le -; T CRACK GR064TH RATE - %TWEFPY .
- . . -. . - _ .......____.,_-..~.,.,._,...,m_. .- ,- . . . - - . _ _ _ _ _ _ _ . - _ - _ _ - _ _ _ _ . _ _ _ _ . _ . _ _
l l FIGURE 12 ! 1 HAZARD FUNCTIONS FOR WEIEULL :STRIBUTIONS l i
?
h{t) 5.0 . - ( n=4 a=1 n=2. , aai i 4.0 -
)
1 3.0 l - c . ; u ; c N 5
~ 2.0 --
e i i 1.0 n = 1. I a=1 i i n = 0.5.
' ami i 0 I I
't i 0 1.0 2.0 3.0 4.0 3 l TIME ;
l l I l l 8 - -- . ~ . , .. - . + . - ._ .
% m.. . ., , e .+...,.nww.e 4e< - * - * * .
T _ - - - - . s 0 g .... .:....:...:........:.... ....... ..; .........:......- ....:I 0 1
. w a
- .. . ... - . .. . . . .. ......:........- . :.1 0 g .. . ..- . . - . 8 t . r O I T . . L C . L 1 t
. A u .
M L F . R D H - O - - * . 0 T P g . .- ... .:..:. ..
.2..
. .. .. .! ...... ..:.. ..;....3l 6 %
F - , . - 3 O . l i 1 T N - . P E O - . . E I . R T D e U G AT - K I 1 t . C F E A M R E -
.- . ..1.. * . .. .- .. - . .. .3! 0 C g . . . .?. 4 L E P G M . A e I .
R - L E M A . V C . A m I P Y . T
- =
t 0 2
- g. .?
. s e
+
- +
.1 0 _
._ . ~ . - . . ..
- - - . - . - . - c 1 6 4 2 0 -
B. . 0 0 0 0
~=~O u d to o .>F" to m<)aDg.Q .
. ;li, t ,
FIGURE 14
- ' HEMAT:: : .R F : C H A E '. . . : T '. '
' RACK 'JCWTH 'S E'
. )
l
!COMPUTECRACKINITIATION
- TIMES FOR ALL EXPECTED CRACKS (WElBULL MODEL) i i
V COMPUTE CRACK DEPTHS FOR EACH INSPECTION
~
TIME (CRACK GROWTH RATE PDF) L I d
- U+I PERFORM !NSPECTION/
REPAIR PROCEES SIMULATION USING P.O.D. FUNCTION , I l 6 EXAMINE NEWLY DETECTED CRACKS FOR EACH INSPECTION TO DETERMINE NUMBER OF RG t.121 EXCEEDANCES (Nj) l INCORPORATE Nj INTO HISTOGRAM
' v i j J < 1000 L
END l l i l
i t i APPENDIX A l i MAINE YANKEE RUN TIME EVALUATION CASE SETS 9 i a l i
\
-....-,--......y. , , , .
~* g.
CASE SET 1 WEIBULL PARAMETERS: GAMMA = 0.74 SIGMA = 2600. 4 .- - ~ ~ - ---
- RUN Lzng7H IN DAYS = 25.0 PROBABILITY OF X 7 TUBES
/ TUBES OUTAGE 0 OUTAGE 1 CUTAGE 2 OUTAGE 3 0 .0000 .9130 .9030 .9280 1 .0000 .9970 .9970 .9970 2 .0000 1.0000 1.0000 1.0000
- 3. .0000 1.0000 1.0000 1.0000 4 .0000 1.0000 1.0000 1.0000 5 .0000 1.0000 1.0000 1.0000 6 .0000 1.0000 1.0000 1.0000 7 .0000 1.0000 1.0000 1.0000 8 .0000 1.0000 1.0000 )'
1.0000 9 .0000 1.0000 1.0000 1.0000 10 .0000 1.0000 1.0000 1.0000 11 .0000 1.0000 1.0000 1.0000 l 12 .0000 1.0000 1.0000 1.0000 ! 13 .0000 1.0000 1.0000 1.0000 [ 14 .0000 1.0000 1.0000 ; 1.0000 15 .0000 1.0000 1.0000 , j 1.0000 16 .0000 1.0000 1.0000
- 1.0000 17 .0000 1.0000 1.0000 1.0000 18 .0000 1.0000 1.0000 1.0000 19 .0000 1.0000 1.0000 1.0000 i
l l l 4 I i I l i l l
RUN LENGTH IN CAYS = 450.0 ; i PROBABILITY OF X iTUBES !
# TUBES OUTAGE O OUTAGE 1 OUTAGE 2 OUTAGE ' '
O .0000 .8900 .8900 .8800 1 .0000 .9930 .9970 .9940 2 .0000 .9990 1.0000 1.0000 3 .0000 1.0000 1.0000 1.0000 l 4 .0000 1.0000 1.0000 1.0000 5 .0000 1.0000 1.0000 1.0000 6 .0000 1.0000 1.0000 7 1.0000
.0000 1.0000 1.0000 1.0000 8 .0000 1.0000 1.0000 1.0000 9 .0000 1.0000 1.0000 1.0000 10 .0000 1.0000 1.0000 1.0000 11 .0000 1.0000 1.0000 1.0000 12 .0000 1.0000 1.0000 1.0000 13 .0000 1.0000 1.0000 ,
1.0000 14 .0000 1.0000 1.0000 1.0000 15 .0000 1.0000 1.0000 ' 1.0000 16 .0000 1.0000 1.0000 1.0000 17 .0000 1.0000 1.0000 1.0000 18 .0000 1.0000 1.0000 1.0000 19 .0000 1.0000 1.0000 1.0000 t I l l
RUN LENGTH IN DAYS = 500.; PROBABILITY OF X # TUBES 7 TUBES OUTAGE O OUTAGE 1 OUTAGE 2 OUTAGE 3 0 .0000 .8320 .8180 .8060 1 .0000 .9850 .9740 .9840 2 .0000 .9990 1.0000 1.0000 3 .0000 4 1.0000 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 5 .0000 1.0000 1.0000 1.0000 6 .0000 1.0000 1.0000 1.0000 7 .0000 1.0000 1.0000 1.0000 8 .0000 1.0000 1.0000 1.0000 9 .0000 1.0000 1.0000 1.0000 10 .0000 1.0000 1.0000 11 1.0000
.0000 1.0000 1.0000 1.0000 12 .0000 1.0000 1.0000 13 1.0000
.0000 1.0000 1.0000 1.0000 14 .0000 1.0000 1.0000 15 1.0000
.0000 1.0000 1.0000 1.0000 ~
16 .0000
- 1.0000 1.0000 1.0000 17 .0000 1.0000 1.0000 1.0000 18 .0000 1.0000 1.0000 19 1.0000
.0000 1.0000 1.0000 1.0000 4
9
,,,w,-. * ^ "
_, .,eeww,w.o~=- * * ' " ' " ' * ' "
CASE SET 2 WEIBULL PARAMETERS: GAMMA =0.37 - SIGMA = 500K e
.A- e-= * = - - - - - - ,*- --
' 4 se y,- s e -- -~-=--=.;-e-, . -
RUN LZ:tG :. ::: cAyg , ,., s .. PROBABILITY OF :: = TUBES
= TUBES OUTAGE O DUTAGE 1 OUTAGE 2 OUTAGE -
0 .0000 .8900 .9000 .3970 1 .0000 .9950 .9930 .9950 2 .0000 1.0000 3
.9990 .9990
.0000 1.0000 1.0000 1.0000 4 .0000 1.0000 5
1.0000 1.0000
.0000 1.0000 1.0000 6 .0000 1.0000 1.0000 1.0000 1.0000 7 .0000 1.0000 8
1.0000 1.0000
.0000 1.0000 1.0000 1.0000 9 .0000 1.0000 1.0000 1.0000 10 .0000 1.0000 11 1.0000 1.0000
.0000 1.0000 1.0000 12 .0000 1.0000 1.0000 1.0000 1.0000 13 .0000 1.0000 1.0000 1.0000 14 .0000 i 15 1.0000 1.0000 1.0000
.0000 1.0000 1.0000 l 16 1 0000 ,
.0000 1.0000 1.0000 1.0000 17 .0000 1.0000 18 .0000 1.0000 1.0000 19 1.0000 1.0000 _ .0000
.0000 1.0000 1.0000 1.0000 9 . ?
IG -
. i i
1 1
. . . , . . . . . . . . . . ... . . - +, - - - -
8 l
'UM LIMGT3 !" 2A3
=
-50.0 '\
pTUBES :?ROBABILITY C F *. 3 TUBES OUTAGE O OUTAGE 1 SUTAGE 2 ! O OUTAGE :
.0000 .8610 .5520 .8600 I
1 .0000 .9860 .9900 .9920 i 2 .0000 .9970 .9980 1.0000 3 '
.0000 1.0000 1.0000. 1.0000 4 .0000 1.0000 1.0000 1.0000 5 .0000 6
1.0000 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 7 .0000 1.0000 1.0000 1.0000 8 .0000 1.0000 1.0000 1.0000 9 .0000 10 1.0000 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 11 .0000 1.~0000 12 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 13 .0000 1.0000 !
14 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 15 .0000 1.0000 16 1.0000 1.0000 -
.0000 1.0000 1.0000 1.0000 17 .0000 1.0000 1.0000 1.0000 18 .0000 1.0000 19 1.0000 1.0000 '
.0000 1.0000 1.0000 1.0000 1
i
. f l
I l l l l 1
. _ _ _ _7 ,,
RUN LINGTH :N Cays = 500.0 PROBABILITY OF :'. # TUBES
- TUBES OUTAGE O OUTAGE 1 OUTAGE 2 DUTAGE 2 0 .0000 .7700 .8140 .7910 1 .0000 .9710 .9820 .9830 2 .0000 .9370 .9990 1.0000 3 .0000 1.0000 1.0000 1.0000 4 .0000 1.0000 1.0000 1.0000 5 .0000 1.0000 1.0000 1.0000 6 .0000 7
1.0000 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 8 .0000 1.0000 9
1.0000 1.0000
.0000 1.0000 1.0000 1.0000 10 .0000 1.0000 1.0000 1.0000 11 .0000 1.0000 1.0000 12 1.0000
.0000 1.0000 1.0000 1.0000 13 .0000 1.0000 1.0000 1.0000 14 .0000 1.0000 1.0000 1.0000 15 .0000 1.0000 1.0000 1.0000 16 .0000
'+
17 1.0000 1.0000 1.0000
.0000 1.0000 1.0000 1.0000 18 .0000 1.0000 1.0000 1.0000 19 .0000 1.0000 1.0000 1.0000 l
l l l l l l
ATTACHMENT B POINT OF LOADING EFFECT ON VALIDITY OF IN-SITU PRESSURE LOADING l 1 L:\94MN\94104 1
Page B1 POINT OF LOADING EFFECT ON VALIDITY OF IN-SITU PRESSURE LOADING : The in si'.u pressure test imposes the same or larger axial loadings on a top of tube neet defect as experienced during operation when the test pressure is the same as the operating pressure (the loadings are the same for the same pressures but it is customary to increase the test pressure at cold testing conditions to compensate for material property differences between testing and operating temperatures). A simple sketch of the in-situ pressure testing apparatus is shown in Figure BIA. It grips and seals to the tube above and below the circumferential crack. The spring between the top and bottom grippers is very flexible and provides negligible resistance to tube movement and/or loads. The test fluid fills and pressurizes the cavity between the two gripper seals. Provided there are no other forces acting on the tube when the pressure chamber is pressurized to the operating pressure of a tube, the axial load (hoop load as well) acting through the tube at the top of the tubesheet is the same for operations as it is for testing. See Figures BlB and B1C. This is the case when a tube can move freely in the axial direction through its horizontal supports. Should a tube be locked into one or more of its horizontal supports, the test imposed axial force acting through the top of the tubesheet circumferential crack is reduced by the resistance provided by the locked support. See Figure BlD. If a tube is locked into a horizontal support during in-situ pressure testing, it is also locked during operations (the differential expansion between the Inconel tubes and carbon steel supports is more likely to cause locking at operating temperatures) and a corresponding reduction in the operations imposed n ial load passing through the top of tubesheet circumferential crack is also experienced during operation. See Figure BIE. It is desired to identify the relationship between the axial load through the defect during in-situ pressure testing (F-f, of Figure BID) and during operations (F-f ' of Figure B1E) when a tube is locked in one or more of its 3 horizontal supports. Refer to Figure BlF. ! Definitions F pressure induced axial force on tube: end reaction load during operation: upper in-situ test apparatus gripper induced load during ; testing. L, length from the top of tubesheet to the point of application of F. L:\94MN\94104 i
i f Page B2 : L, length from the point of application of F to the first locked support. l A axial area of a S/G tube. ;
'E modulus of elasticity of the S/G tube material
. K3 flexibiiityofthetubebetweenthetopofthetubesheetandthepoint ! of application of F: it's the inverse of the stiffness and may be expressed as L /AE. 3 K, flexibility of the tube between the point of application of F and the : first locked horizontal support: it's the inverse of the stiffness and i may be expressed as L,/AE. I i K3 incremental flexibility of the locked horizontal supports in the axial ; direction of the S/G. l K, incremental flexibility of the finite portion of the tube right at the j defect. + 63 displacement at the point of application of F. 6, displacement at the first locked horizontal support. I f3 force applied to the tube through the horizontal support (s) locked to the S/G tube when F is applied anywhere along D (this is representative i of the in-situ pressure test imposed axial loading). l f3 2
. force applied to the tube through the horizontal support (s) locked to .!
the S/G tube when F is applied at or just above the locked horizontal l support (i.e., L - D & L, - 0: 3 this is representative of the ! operations imposed axial loading) ; f, force applied to the tube due to it being anchored in the tubesheet- ; IT'S THE FORCE ACTING THROUGH THE TOP 0F TUBESHEET DEFECT AND EQUAL TO F-f 3 l t NOTE 1: 3 3 8 f-f when L 3
- D & L, - 0 l NOTE 2: All displacements and forces are positive in the orientation i shown in Figure BlF.
L:\94MN\94104 s
PAGE B3 eq (1) 61 = (K 1 + K,) f 2 = (K4 + --) L* AE f eq (2) 52 "K 3 t f l l eq (3) i
$1 - 5 2 "Kf*(AE)f 2 t t l eq (4)
F=f 1
+f 2 1 l
i eq (5) i D=L 1
+L 2 Substitute eq's 1 & 2 into 3 ,
eq (6) L2 + AE K3
,1
[* = {AEKa+L1 Substitute eg's 4 & 5 into eq 6 L:\94MN\94104
Page B4 eq (7) b* b f, = { /UT( K4 + K) 4
+ D)y Equation 7 identifies the portion of F passing through the top of tubesheet defect. It may be seen from equation 7 that there is a greater portion of the axial load, F, passing through the top of tubesheet defect when F is applied between the tubesheet and first locked horizontal support (loading during in-situ pressure testing) than when F is applied above or at the first locked horizontal support (L, - 0 which is the loading during operations). The in-situ pressure test exerts more load through the defect than experienced during operation for the same pressure.
I I l l L:\94MN\94104
Page B5 v j steam /x generator i l tube gnppers p;[e seals (typ) f., _ 1
}
spring , e pressure
' / chamber
// /
/
k/ tubesheet quuwe,mna xx x x
/
a2 l 0~
/
/\ / 1
/ /
/ /
/ / \
\\\\ //// l tubesheet test pressure connection In Situ Pressure Test Device A !
FIGURE B1 (SHEET 1 OF 4)
- -_j
i i i Page 86 i F = (tube ID area) x j (ops press.) l NOTE: F = F WHENTHETEST PRESSURE EQUALS OPS PRESSURE F = (tube ID area) x (test press.) : steam x generator N tube
/
F' p 9F'
//
U C
\\\\
////[ /
h
/
OF F
/ / :
/ /
//// p p\\\\
\s - , ,
u/ tubesheet
/ O T /
/ /
/ /
Normal Ops j y , Axial Loading \ \ x x/ / ffff l (no locked horizontal supports) B In Situ PressureTest Loadings (no locked horizontal supports) , 1 C FIGURE B1 l (SHEET 2 OF 4) ! l l
Page 87 l I fk . [ x^ f ONE OR MORE HORIZONTAL 1 SUPPORTS LOCKED f, TO THE S/G TUBES
/
F O I [ o e '"s(
)
F-f, h F-f' \ 97 hF-f,
////j \\\
/ /
/ / ,
, ////p p\\\\
, , / /
/ /
\
/ ! / /
\\\\ / / / / ,(_i / /
/ /
/ /
\\\\ ////
l In Situ PressureTest Loadings I (locked horizontal supports) Normal Ops Axial Loading D l ( iockea horizontai supports)
- FIGURE B1 E (SHEET 3 OF 4) )
r Page B8 2
/ K 3 ]-
1"I b i
/
( 2 E l if -I-c t i 1 K 4 l
////l' tubesheet v
2 , NOTE: f = f' when L~ O & L = D 1 1 2 1 l Mathematical Model of Tube Loading F l l FIGURE B1 (SHEET 4 OF 4)}}