ML20244A842
| ML20244A842 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 05/31/1989 |
| From: | Cordle H, Straube P TOLEDO EDISON CO. |
| To: | |
| Shared Package | |
| ML20244A845 | List: |
| References | |
| PROC-890531, NUDOCS 8906120167 | |
| Download: ML20244A842 (61) | |
Text
{{#Wiki_filter:._ _ -_ _ _ _ _ _ _ .
- t :: ..3 - q
[. . l , 1 i DAVIS-BESSE PRESSURIZER SURGE LINE THERMAL STRATIFICATION PHASE I PROGRAM May, 1989 Surge Line Thermal Stratification Phase I-Study Group R. J. Gradomski, Mechanical Design H. J. Cordle, Principal Engineer E. C. Matranga, Systems Engineering P. R. Wohld, Performance Engineering P. H. Straube, Civil Design S. J. Osting , Civil Design T. B. Ridlon, Civil Design Prepared by: '
~
I 2 Bf H. Cordie Date Reviewed by: - _[ M
- f. H. Straube Dat'e Approved ' .
k [77 Date 690bi20167 090$o2 gDR ADOCK05AOgg6
~
TABLE OF CONTENTS I. -INTRODUCTION
- II.
SUMMARY
, III PHASE I. PROGRAM
.' A . Pressurizer Surge Line Configuration and Condition.
B. Review of Restraint Gap History
'b.
C. Whip Restraint Design Basis D. Review of Inspection Results E. . Review of Plant Operational Data F. Deflection / Stress Analysis G. Fatigue Life Estimate H. Surge Line Monitoring
- 1. Instrumentation
- 2. Results
- 3. Comparison with Analytical Results
- 4. Estimated Number of Stratification Transients I. Evaluation of Potential Interference IV. REFERENCES V. TABLES VI WIGUAES i
4 t I. INTRODUCTION This report summarizes the first phase of investigations on the effects of thermal stratification on the Davis-Besse Unit 1 Pressurizer Surge Line and L related components. Background-l During start-up tests of the Huelheim-Kaerlich (M-K) plant, a B&V 205 Fuel Assembly design in West Germany, large thermal gradients across the pressurizer surge line elbov below the pressurizer nozzle were observed. Since the stress and fatigue analyris of the pressurizer surge line and the nozzle for all B&W plants assume a homogeneous fluid temperature, the. observed thermal stratification represents a concern. Review of industry literature has revealed documented occurrences of thermal stratification and its effects: (1) USNRC Information Notice 84-87, " Piping Thermal Deflection Induced by Stratified Flow", and (2) USNRC Information Notice 88-80, " Unexpected Piping Movement Attributed to Thermal Stratification". In the reported cases, unexpected movement of the pressurizer surge line occurred. These piping movements are undesirable because resultant piping stress may exceed design limits. The problem can be more acute when the piping deflection is restricted, such as through contact with pipe whip restraints. In December, 1988, the NRC issued Bulletin 88-11. " Pressurizer Surge Line Thermal Stratification". This bulletin is intended to focus industry attention to the stratification issue and provide assurance of surge line integrity over the design life of the plants. Stratification has been observed during all phases of plant operation i.e. heatup, cooldown and steady state operation. Plant heatup had been noted to contribute the largest thermal differentials (gradients) in stratified flow. In normal steady state operation, a small flow of reactor coolant water (~3 gpm) is recirculated through the pressurizer and surge line by means of the continuous minimum pressurizer spray. At power, normal reactor coolant volume variations
.vith the associated control of pressurizer level result in low flow rates through the surge line ranging from zero to approximately 80 gpm. These flows are small compared to the 8.75 inch inside diameter of the surge line. Hotter water from the pressurizer tends to ride on top of a quiescent cold water layer in the bottom of the pipe during these lov flow conditions. Colder vate.r settles to the bottom of the pipe and cools further due to heat loss through the insulation. The thermal differential between the hot upper section and cold lower section of the pressurizer surge line could be as much as 350'F.
Davis-Besse Program During the last quarter of 1988, a three phase program was developed to understand the phenomena and effects of thermal stratification, and to maximize 1 the pressurizer surge line service life. 1
L 1 o Phase 1 vas the short-term program to determine the present condition of the pressurizer surge line, determine the additional effects of deflections and radial gradient stressos caused by thermal stratification events, and to. estimate the remaining fatigue life. Figure 1 is a diagram of these activities. The investigation of potential thermal striping effects is to be covered separately in conjunction with the B&W Owners Group. o Phase 2 will continue from Phase 1 and provide the formal revisions of the fatigue analysis needed to document the design for operation in the stratified service conditions. This phase of the program will include the further short term work needed for response to NRC Bulletin 88-11,
" Pressurizer Surge Line Thermal Stratification". It will result in recommendations for actions to ensure that the predicted service life extends beyond the design lifetime of Davis-Besse Unit 1.
L o Phase 3 will be the implementation of the service life extension recommendations. l L i-
II.
SUMMARY
Phase I of the Davis-Besse Pressurizer Surge Line Thermal Stratification Program was a careful assessment of the existing piping condition and the deflections, stresses and related fatigue effects considering thermal stratification conditions. The Phase I program included: o Review of the surge line configuration and design bases of the piping, supports and restraints. o Review of the history of restraint clearances o Review of In-Service Inspection results o Engineering inspection (Valkdown) of the surge line and supports for evidence of gross discernable distress or structural damage. o A stress and deflection sensitivity study using finite element analysis and estimates of heatup/cooldown stratification transients considering experience at Muelheim-Kaerlich. o Evaluation of fatigue usage based upon the stress analysis and considering Davis-Besse past operating history. o Monitoring of surge line deflection and plant parameters during plant heatup to confirm conclusions based upon analysis. History Review of the first recorded measurements (1980) of surge line pipe whip restraint gaps showed an out-of-tolerance condition at several locations. In 1984, the vest snubber on support PSU-R1 was found to have a broken piston rod, apparently resulting from the deflection of the surge line and interference with the vest vall. The support was redesigned at that time by permanently removing the damaged snubber and increasing the size of the east snubber. There have been no further problems with support, PSU-RI. Since 1982, clearance measurements did not show any changes that would be considered indicative of continued displacement. Inspections During the 5th Refueling Outage in 1938, the Inservice Inspection schedule included voluroetric and surface examinations of several key areas of the surge line, This included the pressurizer nezzle-to-vessel veld (UT), the pressurizer nozzle safe-end weld (PT), the outboard veld on the elbov beneath the 3
pressurizer (PT), a veld at the lower elbov connecting to the riser section (UT, PT), inspections of welds on the 45 degree elbov near the hot leg (PT). No reportable indications were found. Additional engineering visual inspections were made of the surge line toward the end of the Fifth Refueling Outage. No evidence of gross discernable distress or structural damage to the surge line, supports or restraints was found. The positions of the spring hanger and snubbers were found within expected ranges. Restraint gap measurements showed no significant changes from cold measurements made since 1982. Deflection / Stress Analysis Data obtained from instrumentation of the surge line in the Huelheim-Kaerlich (M-K) plant formed the basis for prediction of the thermal stratification conditions. Additionally, experience at M-K and at other plants suggested that the conditions during heatup and cooldown would present the largest stratification temperature differentials. Thus, the evaluation of Davis-Besse conditions centered around the heatup and cooldown transients in which large temperature differentials exist due to the difference in temperatures between hot leg and pressurizer water. Additional cycles of smaller stratification , transients during hot operations would not be expected to add significantly to l the fatigue estimate. A sent.itivity study of the potential surge line deflections, stresses and moments at end connections was performed by Impell Corporation using ANSYS, a finite element structural and thermal analysis code. The thermal stratification I conditions for the study were centered around estimates of conditions based upon Muelheim-Kaerlich data and included a variety of assumptions on extent of stratification in various parts of the surge line. 1 Fatigue Evaluation A fatigue evaluation was performed by B&V based upon the moments supplied by Impell from cases constructed from the M-K data. A review of Davis-Besse startup procedures showed that the number of stratification cycles experienced at M-K was generally consistent with Davis-Besse operations. The most limiting i calculated fatigue usage factor, predicted through the end of the current Sixth l Fuel Cycle (40 plant heatups and cooldowns) is 0.7. The remaining margin is considered adequate to continue operations beyond the end of the cycle and provide for uncertainties. Deflection Measurements During Heatup Transverse piping deflections were measured during the heatup after the 5th refueling outage as a cross check on the deflection analyses. Twelve lanyars potentiometers were installed at seven pipe whip restraint locations. The measured deflections fell within the estimates used in the fatigue evaluatien based upon the Muelheim-Kaerlich data. At the end of the heatup, a potential interference was detected at the location of whip restraint SL2 from inspection l 4 i
l of deflection data. Later visual inspection of the restraint did not show evidence of recent hard contact. The interference, if it occurred, would have been minor based on the measurements, and analysis showed-the additional stress from such an interference-to be small and not significant to fatigue l accumulation. { Conclusions l' It was concluded that the accumulated fatigue of the surge line and its end connections will not exceed design limits during the Sixth Fuel Cycle. Sufficient margin remains to permit operation with safety.until improved data are obtained in conjunction with the B&V Owners Group program, and a more detailed fatigue' analysis demonstrates compliance with Code requirements over the full design life of the unit. 5
e III. PHASE 1 PROGRAM A. Pressurizer Surge Line Configuration and Condition A systematic review of-the pressurizer surge line design and installation was made in order to determine the actun1 condition of the surge line and to provide an accurate basis for the construction of an analytical model. Drawings 12501-HL-PS/PSU Rev. 0 (Hanger Location Drawing, Pressurizer Spray and Surge.Line), and 7749-C-189, Revs. 14 & 15 (Pressurizer Surge line Vhip Restraints) vere reviewed to. determine the geometry of the pressurizer surge line including the location of pipe hangers, snubbers and whip restraints. Additionally Babcock & Wilcox Calculation 620-0014-50, Final Stress Report - Pressurizer Surge Line, was reviewed. The pressurizer surge line is of all velded construction, schedule 140, 10-inch nominal' diameter, SA 376 Type 316 pipe material. The original design code is USAS B31.7, 1969 Edition. A fatigue evaluation was - performed originally, as required by code and did not. include the effects of thermal stratification or thermal striping. . The pressurizer surge line configuration consists of two horizontal pipe sections (an upper and lower section) with a connecting vertical riser. Eight pipe whip restraints, one spring hanger, and three snubbers provide support for the pressurizer surge line under.various loading conditions considered in the design basis. Figure 2 illustrates the surge line configuration. Typically, the pipa whip restraints are of I-beam, box-type construction that are bolted tc poured concrete valls. At each whip restraint location, an impact collar which acts as a spacer is affixed to the pressurizer surge line. Free movement of the pressurizer surge line is determined by preset gaps between shims, applied to the inside of the whip restraint, and the pressurizer surge line collar. On October 17, 1988, an inspection was conducted by Engineering personnel in accordance with Inspection Plan IP-M-017, Rev. O, (Ref. 4). The inspection plan was designed to obtain clearance (gap) measurements at the points most likely to cause interferences of pressurizer surge line movement during plant heatup. The measurements were taken at 90' intervals around the circumference of the pipe collar at each of the eight pipe whip restraints. These measurements were taken during cold, and hot standby conditions. More detailed discussion of the history of restraint gap measurements is given in Section III.B. The pressurizer surge line was visually examined. When sighted axially along the top of the upper horizontal section, a deflection of the pressurizer surge line was noted. This deflection could be characterized as concave dovnvard sad was noted on the lover horizontal section also. This deflection was consistent with the recorded whip restraint gap 6
measurements i.e. increasing gap measurements at the top of each restraint as viewed from the attachment nozzles and progressing toward the vertical rise. The whip restraints appeared in good condition with no evidence of structural damage, e.g. mislocated, loose, signs of impact with the pressurizer surge line. The three snubbers and spring hanger vere also inspected and appeared to be in good condition with no signs of damage. The positions of the spring hanger and snubbers were found within expected ranges. No insulation was removed from the pressurizer or the surge line at the time of this examination. However, several sections of insulation had been removed and replaced during the course of the Fifth Refueling Outage to accommodate the Non-Destructive Examinations (NDE) discussed below. No evidence of gross distortion or structural damage to the surge line, supports or restraints was found. B. Review of Restraint Gap History A review of clearance measurements between the pressurizer surge line and the related pipe whip restraints was made (Ref. 23). Table 3 is a historical summary of the restraint gap measurements. Table 4 is a summary of the changes in gap dimensions to October, 1988. There are no measurements of record from the initial startup of the plant. It is reasonable to assume that the gaps were initially set within the tolerances shown on drawing 7749-C-189, Rev 14. In May, 1980, measurements indicated that some restraint clearances were outside tolerance. The out-of-tolerance condition was resolved as stress analysis indicated the expected movement would be less than the clearances (Ref. 26). In June, 1982 small additional deviations were found, and some shims were adjusted (Restraints SL3, SL4, SL6 and SL7). Also, the design clearances on Drawing 7749-C-189 vere evaluated and several vere revised via Drawing Change Notice. The drawing was reissued as Rev. 15. Changes between the measurements in May, 1980 and June, 1982 are bounded by the 0.38 inch change indicated for restraint, SLS. Analysis of the gap history showed that very little change has occurred I since June, 1982, indicating little or no effect of shim adjustments or the interaction with support PSU-R1 discovered in 1984. l However, measurements of gaps made during the inspection in October, 1988, shoved that the cold gap dimensions were inconsistent with the specified collar-to-restraint gaps on Drawing 7749-C-189, Rev. 15. PCAO Report 88-0859 was initiated to document and analyze the out-of-tolerance condition. In order to resolve the apparent inconsistency with the gap history discussed above, a more detailed review of the tolerances in Draving No. 7749-C-189 was made. This revealed that the original 7
requirements of the DCN in 1982 were not entirely reflected in Rev. 15., resulting in the erroneous observation that an out-of-tolerance condition existed. As a precautionary measure in November, 1988, the decision was made to remove selected shims in restraints SL4 and SL6 to provide increased clearance. The locations of the removed shims are shown in Table 3. The shim removals were based upon initial results from the deflection / stress analysis and after review of the whip restraint design bases described in Section III.C. With one exception, described in Section III.H, measurements of pressurizer surge line deflection, made with lanyard potentiometers during the heatup of the unit for Cycle 6, showed no interferences. Measurements showed surge line deflections fell within the gap clearances measured after the shim adjustments in June, 1982. A possible slight interference with restraint SL2 was detected from examination of potentiometer deflection curves after reaching hot standby. The interference could not be otherwise confirmed. Should it have occurred, it appeared to have been the result of the normal thermal expansion of the reactor coolant hot leg piping. Analyses have shown additional stresses resulting from such an interference to be slight and to have a negligible effect upon the fatigue analysis. Clearance changes shown on Table 4 are not entirely consistent, either between restraint locations or from one period to the next. Recent measurement experience indicates that a considerable variability is possible and may have been responsible for some of the discrepancies. Errors of the same order of magnitude as the observed gap measurement changes in recent years are both possible and likely. Although accuracy of the gap measurements prior to 1988 cannot be determined, the changes in the same direction over the years suggests some systematic change took place, particularly early in plant life. The history of gap measurements points to the probability that an original permanent set of the surge line, and possibly a subsequent one, took place. The relatively small clearance changes measured in recent years and the deflection measurements from the lanyard potentiometers indicate that stresses since 1982 have remained in the elastic range, and no process of continuing deformation exists. 1 C. Whip Restraint Design Basis l l The design basis of the surge line pipe whip restraints was reviewed along with the review of the surge line and related whip restraint gaps. The review indicated that the surge line did not entirely meet the commitment ! in the USAR to have the ability to withstand a pipe rupture, either I longitudinal or circumferential at any point on the line. The concern and l its resolution was documented in PCAOR 88-0915, (Ref. 6). 8
l-The " break anywhere" criterion in the USAR has been changed by the NRC since completion of the Davis-Besse design. Review of the current Davis-Besse design according to the current USNRC Standard Review Plan (SRP) 3.6.2, Rev. 1 including Branch Technical Position MEB 3-1, shoved that the eight restraints with closely controlled gaps are not required for whip protection. Retention of five of.the restraints vill be sufficient to ensure piping stability after a break, should it occur. D. Review of Inservice Inspection Results Inservice inspection records were reviewed for evidence of abnormalities. Figure 3 summarizes the surface and volumetric inspections performed to date as well as forecast inspections to be conducted during the' Sixth Refueling Outage. Inspections made in 1988 during the 5th Refueling Outage covered many of the highly stressed regions of the surge line as well as the pressurizer nozzle. Close attention was paid to the pressurizer nozzle velds as the nozzle location was expected to have one of the highest cumulative usage factors. Additionally, the pressurizer nozzle velds were the primary areas of interest based on the M-K thermal stratification data. The pressurizer nozzle velds starting at veld No. 7 and ending at the nozzle to pressurizer veld, excluding veld VJ-33 (elbow to safe-end), had been ultrasonically. examined on a four year frequency beginning in 1980. The most recent inspections, conducted during the Fifth Refueling Outage, show no reportable indications for the ultrasonic examination of the nozzle-to-pressurizer veld.and the penetrant tests conducted on velds VP-23 and 7. Volumetric and surface examinations for the line and nozzle attachment velds revealed no reportable indications. Snubber PSU-R2 was reported, in 1983, to have a defective viper ring (Ref. 8). PSU-R2 is located near whip restraint SL-5 in the middle of the pressurizer surge line vertical drop (upper to lover horizontal run transition). Visual inspection in 1984 disclosed a broken piston rod on the vest snubber on support, PSU-R1 (Ref. 7). The support was redesigned at that time by permanently removing the damaged snubber and increasing the size of the east snubber. Consistent with current observation,s and analysis, with the original design, the vest snubber vould have been pinched between its support and the vall. Since the repair, there have been no problems reported with support, PSU-Rl. Review of the restraint gap history indicates that the interference which damaged the snubber caused no significant change in gap clearances (or permanent deformation of the surge line). 9
1
. E. Review of Plant Operational Data Plant operational data, coupled with the M-K experience, were used to estimate magnitude and frequency.of occurrance of thermal stratification transients. This information, in turn, was used to develop inputs for deflection / stress sensitivity analyses and for evaluation of fatigue effects due to stratification.
1 The focus of the effort was upon the heatup and cooldown transients since the maximum differential between pressurizer water temperature and hot leg water temperature occurs during these operations. Bounding values of stresses and moments occurring during'the heatup and cooldown operations formed the basis-for an evaluation of any increased fatigue usage due to stratification. Although stratification transients are not limited only to the heatup and cooldown operations, additional occurrences during other operating modes vould involve smaller magnitudes and incur relatively small fatigue life penalties. In order to define temperature transients upon which flexibility and fatigue analyses could be based, the D-B Plant Operating Procedures and the temperature stratification data from Muelheim-Kaerlich vere reviewed. Plant-specific thermal stratification and deflection data were not available for Davis-Besse. Muelheim-Kaerlich data reported in BAW-2053P (Reference 10) were considered close enough to Davis-Besse conditions to provide initial. estimates of maximum values of surge line stratification temperature differentials at various stages of heatup and cooldown. These values were then compared to the Davis-Besse operating procedure, DB-OP-06901 which governs the heatup of the plant from cold shutdown. Figure 4 is a plot of the various temperature hold points / restrictions as a function of time. As demonstrate _d by this plot, there are a number of sequential plant evolutions and equipment restrictions which govern the reactor coolant and pressurizer temperatures. This plot permits comparison of the Muelheim-Kaerlich data against Davis-Besse operations, and adjustment of the stratification temperature differentials between top and bottom of the surge line considering Davis-Besse limitations (Refs 11 and 12). Figures 5.and 6 are plots of the Huelheim-Kaerlich temperature data from three thermocouple located around the outside diameter of the surge line in the horizontal run near the pressurizer nozzle. The top, side and l bottom locations are represented by-T-6, T-7 and T-8 respectively. Additionally, the pressurizer and hot leg temperatures are shown. Figure 5 shows the heatup operations and Figure 6 covers cooldown. From the Huelheim-Kaerlich data, and later confirmed by measurements at Davis-Besse, the temperature distribution in the surge line is affected by the layout of the surge line. The horizontal run of the surge line, located below the pressurizer, is at a lover elevation than the hot leg connection. The result is a " siphon" or " cold trap" region which tends to cool belov hot leg temperature during periods when the surge line flow is I 10
7-_ quiescent. Thus, the stratification temperature differentials for Davis-Besse are somewhat larger, when the plant is hot, than.vould be the case were the coolest location to remain at hot leg temperatures. Definition of Temperature Matrix Based upon the stydy of Muelheim-Kaerlich data and the review of limiting Davis-Besse operating conditions during plant heatup,.a number of conditions were defined to bound the probable stratification conditions. Plant heatup and cooldown conditions were selected as the limiting conditions since experience at Trojan and at Muelheim-Kaerlich suggested that maximum differential temperatures of stratified coolant in the surge line occurred during these operations. Three groups of conditions covering the 19 original load cases considered in the scoping sensitivity study were identified as shown in Table 1.
- g. o Group 1 consists of conditions defined consistent with the maximum stratification differentials assuming the temperature is bounded by the hot leg temperature on one hand and the pressurizer water temperature on the other.
Due to the layout of the Davis-Besse surge line, which is separated into an upper and lower horizontal run joined by a vertical riser, Group I was separated into a base case and an alternate case. The base case assumed that the upper horizontal run is essentially uniform in temperature, and all the stratification is local to the lower horizontal run. The alternate case assumes both upper and lower horizontal runs are equally stratified. o Group 2 is based upon observations of Huelheim-Kaerlich and more conservatively considers the bottom of the lover horizontal run containing water cooled below the hot leg temperature. The upper horizontal run is assumed to be uniform (unstratified) at about hot leg temperature. This set of assumptions was used as the basis for later fatigue evaluation, and correlated reasonably well with observations during plant heatup . o Group 3 conditions were defined in order to extend the bounds of the analysis to test for sensitivity to larger temperature differentials. F. Deflection / Stress Analysis A sensitivity study of surge line deflections, stresses and moments was made using a compilation of conditions derived from estimates of conditions considered possible at Davis-Besse. The objectives were:
- 1. To obtain the maximum deflection to determine potential for interference with the restraints.
11
l L I. ,
- 2. To obtain the maximum stresses to determine the potential for
! repeated deformation (thermal ratcheting). l- 3. To obtain maximum loads and moments for use in evaluation.of nozzles at the end connections. l 4. To obtain necessary input from the above for an estimate of fatigue usage. Information generated in reviews of surge line and restraint configurations, and in reviews of plant operational data, described above, was used to perform the analyses. The sequence of work performed was the following.
- 1. Determine the matrix of temperature conditions. required to bound the probable stratification transients.
2.- Using finite element piping analysis, determine the deflected configuration and stresses in the line in an unrestrained condition.
- 3. Determine whether the deflection would result in interferences with restraints or other structure.
- 4. Redetermine the deflection and the resultant maximum stresses in the event that a restrained (interference) condition is identified.
Impell Corporation was contracted to perform the deflection / stress analyses, having had prior experience in evaluating the Trojan Nuclear Power Station pressurizer surge line thermal stratification effects. This work served to identify potential interference conditions which could be corrected prior to restart of Davis-Besse from its fifth refueling outage. Analysis A finite element model of the surge line was constructed using i ANSYS, a finite-element computer code for structural, thermal analysis. The mathematical model of the surge line showing node points is illustrated in Figure 7. The analyses employed both elastic and non-linear techniques and are documented in the Impell j report, Ref. 13. i A major assumption in the model is the use of a linear i approximation of the thermal gradient from top to bottom of the I stratified run cf piping. This approximation and the assumption of 12
. j that the affected run of pipe is uniformly stratified from end to end indicate some of the limitations of the analysis.
Nevertheless, the analyses have produced some good comparisons.to the~ monitoring data. 1 The sensitivity analysis considered loe.d cases covering thermal stratification magnitudes for the range of conditions described above. The initial study consisted of 19 cases but was subsequently expanded to 28. . Cases 20 through 28 vere run in order to compare analytical results with measured displacements. The analysis included two cases run for baseline purposes covering normal unstratified thermal expansion of the surge line to reactor full power conditions (Load Cases 2A and 2B). The remainder of the Load Cases cover a' conservative spectrum of thermal conditions and assumed restraint clearances. Results Results of the original load cases are summarized in the Impell report (Ref. 13). Table 2 shows the summary of stress values directly evaluated by Impell, the 3Sm stress limit (USAS B31.7, Equation 12), and.the margin to the 3Sm limit adjusted to consider the CMTR properties. o The analyses support the observation from the history of restraint gap measurements that no plastic strain has occurred since the original heatup of the plant. (See discussion of gap measurement history in Section III.B)
- o Surge line stress levels fell belov 3Sm Stress limit per USAS l B31.7, Equation 12, based upon code allowable stresses in the cases selected for Davis-Besse fatigue evaluation. These cases were based upon the temperature distributions observed at M-K and restraint gaps as existing since 1982, (Load Cases 11, 13B, and 14).
o Only three alternate cases in the sensitivity study resulted , in surge line stresses exceeding the 3Sm stress limit after correction for actual (CMTR) material properties. Even these cases were less than 10 percent above the limit. Load Case 8 involved an assumed vorst-case set of restraint clearances. l Load Cases 18 and 19 assumed unrealistically large stratification temperature differentials (from 506 to 95'F, and 575 to 205 'F, respectively). o The analyses predicted that some contact vould occur between the surge line and restraints SL4 ar.d SL6, and under some conditions SL5. Where such contact would have occurred with the restraint gap assumed in the sensitivity study, the i 13
i resultant stress shown in Table 2 includes the effects of any interference. o For those cases relating to the Huelheim-Kaerlich data, (Load cases 11, 13A & B, 14 and 14A) only the transients expected to occur midway into Heatup, Cases 13B and 14A, vould have resulted in a calculated interference considering the clearance between the surge line collar and the voll at restraint SLS as reported in October, 1908. The lanyard potentiometer measurements during the recent heatup shoved no evidence of contact. A subsequent review of the north and vest clearances at restraint SL5 showed them to be substantially larger than reported in October 1988.. (Ref. 4 and Ref. 18) Resultant Changes Based upon the envelope of calculated deflections, shims were removed from three locations as a precaution prior to the heatup for Cycle Six. This eliminated any likelihood of future interferences with restraints SL4 and SL6. The removal of shims was made possible by the reevaluation of the whip restraint design bases, and the conclusion that the whip restraints in this location are not required. The locations of the removed shims are noted in Table 3. G. Fatigue Life Estimate The additional thermal stratification stresses in the surge line and nozzles at its end connections had not been considered in the original fatigue analysis performed during the design phase. A revised fatigue calculation (Ref. 15) was performed by B&V utilizing the design fatigue calculation from the Code stress report and superimposing the stratification effects. Considering the anticipated cumulative effects of operation through the Sixth Fuel Cycle, the most limiting fatigue usage factor was found to be 0.7 at the hot leg to surge line nozzle material discontinuity. While the fatigue estimate was not based upon a full estimate of all stratification events tnroughout all modes of plant operation, it was based upon evaluatien of the most severe transients seen to date in similar plant operations (M-K). The remaining expected transients are expected to be of lesser magnitude, not substantially affecting fatigue results. It was concluded that ample margin remained for Cycle 6 sperations , and probably beyond. Additional conservatism is contained in the methodology employed in the calculation, deriving mainly from the addition of fatigue I 14 i
effects of the thermal stratification transients to the results in the original stress report. This approach did not permit those economies one vould expect if the actual transients were more realistically treated in a combined design transient evaluation. Evaluations of actual deflections measured during the heatup for Cycle 6 found the transients used as basis for the fatigue evaluation had larger deflections than measured, providing additional conservatism. The measurements made during the heatup for Cycle 6 are covered in Section III.H below. Table 5 is the summary of the B&V fatigue evaluation results for the surge line and for the nozzle connections to pressurizer and hot leg. Figure 8 shows the location of the high-stress points evaluated in the analysis. The fatigue evaluation was based upon the assumption that some plastic deformation of the surge line occurred on two separate occasions early in plant life, a condition equivalent to the piping
" shakedown" period described in the ASME Code. Subsequent heatups and cooldowns, consistent with both observation of restraint gap stability and analysis of the stratification transients discussed above, vould have had stress levels within the elastic range.
Impell load cases 11, 13B and 14, based upon the Muelheim-Kaerlich data, were selected as being conservative and realistic bases for the evaluation of the fatigue effect of the stratification events (Ref. 12). Table 6 is the summary description of the load cases forming the input for the evaluation. Table 7 provides the estimate of the thermal cycles applied. (See Ref. 14.) Table 8 lists the moments applied for each of the load cases used in the fatigue evaluation. H. Surge Line Monitoring The heatup of the Unit for Cycle 6 offered an opportunity to obtain data to improve understanding of operational effects upon stratification, and to confirm conclusions based upon the analytical results, inspections and restraint gap clearance records. A program was und2rtaken in the short time available to obtain information on displacements cf the surge line during the heatup period. The pregram was undertaken October 5, 1968 and data vere taken during the heatup te hot standby from November 10 to November 25, 1988. Deflection data collection was terminated prior to criticality and the temporary instrumentation was dismantled. 15
l ' l
'1. Instrumentation l Twelve Josyard potentiometers were installed at seven
! restraint locations to obtain data on deflections of-the surge line in two dimensions transverse to the piping run. Figure 9 is a diagram of.the locations of the lanyard potentiometers and the location of the surge line thermocouple. Near-continuous deflection records were-
- obtained throughout the period by a small digital data l acquisition system supplied by Minerva Research Corporation.
System output was recorded by a lap-top computer located-in' the containment. Plant computer records were obtained for a record of i important plant parameters, including: 1 Computer Parameter Definition Point Instrument Reactor coolant hot leg temperature T782 TE-RC3A5 l Pressurizer water temperature- T777 TE-RC15-2 Pressurizer water level L768 LY-RC14 Surge line temperature T775 TE-RC16 (Top of lower horizontal run at the midpoint) Occasional scans of surge line temperature using a contact pyrometer were made to observe the extent of the stratification at various points along the surge line. Plant operating logs were collected to provide understanding of the operational origin of various transients, and to provide the understanding needed both for future estimates of cyclic stratification frequency and for consideration of operational remedies. Installation of the Lanyard Potentiometers The lanyard potentiometers were clamped firmly to the vide flange sections of the pipe whip restraints. The end of each lanyard vas connected to the piping. In the case of the instruments e.t SL2 and SL3, the lanyards were nttached to an angle iron section affixed to the surge line. These angle iron sections were strapped firmly around the insulation parallel with the pipe axis and bearing firmly against the spacer collars which serve as impact points for the piping should it whip against the restraint. Each collar is bolted together in halves to clamp around the surge line piping. Thus extension or retraction of the lanyard connected to the 16
l l angle iron section vould reflect the pipe deflection at the . location of the restraint. The installation is described in Ref. 16. J l Potentiometer Calibration 1 The lanyard potentiometers were calibrated against slotted aluminum blocks with dimensions traceable to the National Bureau of Standards. The calibration process is described in
.Ref. 17. . Deviations in measurements presented by the lap-top computer were generally within 1/32-inch over the'six-inch '{ ,
range of extension of the lanyards. The calibration was checked both on Novetaber 2, and November 25, 1988, before and after the heatup. At the. final calibration, the lanyard potentiometer .l installations were checked. All potentiometers were found i secure with no evidence of slippage of the device within the clamps on the restraints. Review of the data obtained, however, indicated.that the collars to which the lanyard were attached on restraints SL2 and SL3 may have slowly rotated during the data collection. This rotation could have introduced an error, later determined to be approximately 0.3 inch. Additionally, one potentiometer (Channel 12) at restraint SL8 was shown unreliable. Surge Line Temperature Indication, TE-RC16 The surge line temperature indication is from a strap-on thermocouple located approximately at the midpoint of the lower horizontal run of the surge line. The instrument is located under the insulation at the top of the pipe. Readout was from the MODCOMP computer point, T775. This thermocouple is used in the plant for trending only, and has not recently beer. calibrated. Cross checks of readings were made occasionally during the heatup using a contact pyrometer inserted through a joint in the insulation adjacent i to the thermocouple location near restraint SL7. The results l showed agreement within a few degrees early in .the period when temperatures were lov, but read approximately 80 degrees lover than contact pyrometer readings in cross checks made toward the end of heatup. This thermocouple is very valuable for identifying trends and stratification events, however, it was not considered reliable for accurate measurement of surface temperature. [ i l 17
o' u s g-u: k' q: a h q. 2. Results
. Two stratification transients, one occurring early in the heatup, at 13:08 on Nov. 10th, and one occurring late in the heatup at 05:58 on Nov. 24th, were representative of-the L
largest observed and provided a. good basis for comparison with the analytical predictions based upon the Impell l-
' analyses (Input to the fatigue analysis performed by B&W).
The transients during plant heatup offer the largest differential temperatures between bottom and top of the surge line. 'Early in the heatup, the pressurizer must be raised to a temperature.in excess of 400 'F in order to raise pressure to permit reactor coolant pump operation. Later in the heatup, pressurizer temperature is raised as necessary to maintain RC pump.NPSH and maintain coolant subcooling margin. During this time, quiescent conditions in the lover regions of the surge line and the effects of heat loss through the insulation result in large differential temperatures between pressurizer temperature and the bottom of the lover horizontal run of the surge line. Small adjustments in pressurizer level cause hot vater.to flow into the surge line generating large stratified thermal differentials and significant deflections. The stratification differentials during heatup and cooldown are the largest expected. Reactor Coolant System and surge line parameters from the MODCOMP plots for Nov. 10 representing the initial phase of heatup are shown in Figure 10. The stratification transients are evident from the curve of surge line temperature, (T775). The deflection curves derived from the lanyard potentiometers on restraint, SL4, are shown in Figure 11. The surge line at this location experiences the largest deflection in response to a stratification transient. Comparison of the surge line temperature trace in Figure 10 with the deflection curve in Figure 11 reveals a recognizable similarity. Figure 12 shows Reactor Coolant System and surge line conditions for the period of operation toward the end of heatup on Nov. 24th. The value of minimum pressurizer spray flow vas not mearured during the plant heatup, but based upon preset valve position, was probably in excess of 0.75 gpm when all reactor coolant pumps vero in cperation. The floppy disks containing the data and copies of the sof twa>:e are filed with Nuclear Records Management. Figures 13 and 14 shev the results of contact pyrometer plots j taken on two different occasions. Both of these plots were i i 18 i j
n' . ? .. for occasions in which a very slow insurge from the reactor-coolant system was underway. 'These data tend to support the assumptions made for the analyses of Impell load cases 11, 13B and 14. The lower horizontal run tends to be the most-highly stratified since its lov elevation makes it a natural collection point for the higher density cool water resulting from heat losses through insulation. Table 9 presents the thermal and deflection data from the two major transients compared to the deflections predicted from the Impell analysis. Deflection diagrams illustrating the various cases, and providing means for visual comparison are given in Figures 15 through 19. At the end of the heatup on November 25th, reviev of the deflection curves showed a number of truncated peaks from readings of the potentiometer record $r.g vertical deflection on the second restraint from the hot leg connection (SL2). The data are presented in Figure 20 as direct output from the acquisition system software, and show upward deflection in the negative direction. Such truncations are possible indications of an interference between the piping and the restraint at that location. The analysis of the condition is discussed in Section III.I. 3 Comparison of Measurements with Analytical Results The measurements of deflection by the lanyard potentiometers serve as a benchmark for comparison with the analysis performed by Impell, and serve to add confidence to the results of the fatigue evaluation. The measurements demonstrated that the deflections and resulting stresses predicted through analysis were bounding values. The fatigue evaluation discussed in Section III.G vas based upon the stress and moment data derived from the analysis contained in load cases 11, 13B and 14 summarized in Tables 6, 7 and 8. The measurements taken on Nov. 10th and 24th, 1988 and shown la Figs 10 through 12, are of value in comparison vfth these load cases. Table 9 permits comparison bervsen the estiested thermal conditions and the deflection measurements for the transients. Figures 15 and 16 are the plots of the deflections for the transient early in the heatup (Load case 11) arid the corresponding measured deflection. Figure 17 shows the data obtained from a repeat analysis (Load Case 22) using the best estimate temperatures in the nurge line at the time of the measured deflection. The similarity of this calculated . 4 l 4 19
{
- L i
I deflection curve to that of the actual deflection l measurements in Figure 17 shows a reasonable approximation of I the calculation to the actual transient. Stress values from i this shoved levels were approximately 70 percent of the values in Load Case 11 used in the fatigue analysis. I Table 9 also offers comparison of calculated and measured l l deflection data for the transient late in the heatup. Both
-measurements for load cases 13B and 14 are included. These load cases were used as input for the fatigue evaluation. The deflection data on Nov. 24th representing the largest transient deflection recorded in this phase of the heatup is l used for comparison. The largest recorded transient l'
deflection fell between the thermal conditions used for these load cases in the analysis, hence, the single transient serves for comparison to both conditions. Figure 18 shows the measured deflection for the transient occurring late in the heatup. Figure 19 shows the corresponding deflection calculated for load case 13B. The shapes of the deflection curves indicate strong similarities. There are, however, differences particularly at SL2, SL7 and SL8, locations near the end connections, shoving larger than expected deflections. The potentiometer (Channel 12) at SLB was found at final calibration to be unreliable. Later observations of difference in measured deflection and restraint clearances led to the conclusion, discussed in more detail below, that the spacer collars to which the lanyards were attached at SL2 and SL3 may have slowly rotated (slipped on the pipe) during the data collection. This rotation is a probable cause of those deviations which cannot be readily explained by reasonable distributions of temperatures or moments. The overall magnitudes of the largest displacements (at SL4 and SL6) suggest, however, that the deflections in Load Cases 13B and 14, used in the fatigue analysis effectively bound the measured deflections. Impell observations and conclusions are contained in Ref. 13. Apart from measurements from Channel 17 vhich was known te have malfunctioned, review of the data from the Irnyard potentiometers did not reveal any obvious pcints at which the potentiometers slipped or changed unexpectedly. In fact, in the evaluation of the results surrounding the apparent interference with restraint SL2, (Nov. 25th), relation', hips between the potentiometer records on three adjacent locaticins shoved very good consistency. Additionally, the dat.a shaved I 20
pr-
, ..(
c L. e j
+
- reasonably good' consistency with the calculated deflections for the thermal conditions measured on Nov. 10th (Load Case p - 22).
- f Rotation of the ~ collars at restraints SL2 and SL3 probably
. occurred over time, responding.to vibration, dead weight and the steady pull of.the lanyards. Such rotation probably affected the measurements later in the heatup more than ' hose
. made earlier. The data from the transient early in the heatup sppear consistent with analysis.
Even later in-the heatup, there are no points at which sudden changes in measured' deflection vere observed. The potentiometers appeared to be following deflections repeatably. However, absolute values of. deflection measured over the long term may be in error.
- 4. Estimated Number of Stratification Transients The detailed transient' record of the heathp of Davis-Besse Unit 1.shows surge line temperature continuously varying with time. The temperatures are reflected in the deflection of the surge line in response to the thermal stratification. In the estimation of. fatigue, the largest transients were evaluated and the smaller magnitude deflections vere assumed to have negligible effect.due to the lover alternating stress range. . The number of cycles which'can be tolerated increases rapidly with reduction. in magnitude of the alternating stress.
Review of the Muelheim-Kaerlich data indicated a sequence of operations similar to the procedures used for heatup of Davis-Etsse Unit 1.- Thus Davis-Besse vould experience similar motion of vster through the surge line and similar stratfiication transients over equivalent ranges of the heatup process. Although there would be variations between the plants and from one plant startup to the next, on average, the major transients over the heatup at Davis Besse would be expected to fall within the numbers of cycles given in Table 7. Early in the heatup, during steam bubble' formation, four significant transients were observed corresponding to the three defined in Table 7. Of these, two showed surge line temperature variation (MODCOMP point T775) approximately equivalent to the transient of Nov. 10 at 13:08 hrs., and the other two were smaller. Impell calculations of stresses and moments of these larger transients (estimated in Load Case
- 22) were approximately 70 percent of the value used in Load Case 11 for the fatigue evaluation. The conclusion is that f
21
( the fatigue evaluation, based upon three larger transients during the early phase of heatup, conservatively represents the average'of past heatups. Later in'the heatup, a single cycle was assumed for a transient with top / bottom thermal stratification-differentials of 506'F/120'F (Load Case 13B) and a. single. cycle of 506'F/205'F (Load Case 14). Only one large transient-deflection was experienced during this portion of the heatup with best-estimate thermal stratification temperatures estimated at 490'F/210'F. The deflections observed in this cycle were bounded by those in Load Case 13B. The observed cycles in the heatup are sufficiently close to the estimate in Table 7 to consider the estimated' count to be l representative of the plant performance'over time. Additionally, considering the magnitudes of the measured transients, the actual fatigue usage vould be less.than calculated by B&W. I.- Evaluation of Potential Interference A slight interference with restraint SL2 was indicated from reviews of lanyard potentiometer data but was not positively confirmed. The deflection curves (Figure 20) revealed truncated peaks, indicating possible contact between the surge line and restraint i' SL2 during several transients the morning of the 25th November beginning about 8 hours after the reactor coolant system achxeved hot standby, Tavg = 532'F, Tprz - 650'F. There was no indication that other restraint interferences had occurred. References 19 and 20 discuss the potential interference and the related post-test inspections of restraints, respectively. During plant'heatup, the Surge Line/ Hot Leg connection moves upward
., as a result of Hot Leg thermal expansion, reducing the clearance to the upper members of restraints SL1, 2 and 3. The lanyard potentiometer data show the trend, and later calculations by B&W of the hot-leg connection thermal expansion motion indicate very small clearances remaining.
The data show that contact with restraint SL2, would have been very slight at the hot standby temperature conditions. Contact, if it occurred, was at the end of an upward motion, relaxing from the effect of a stratification transient. 22
r l4
-There is remaining uncertainty as to whether the potentiometer' data '
actually reflect an interference.
- The potentiometer data reflect a deflection vell in excess of
'the available clearance (1.27 in. compared to a cold gap of' O.94 in.).
- Post-test examination indicated a looseness between the surge
- line and the collar to which the lanyard was connected.
- Although marks existed on the underside of restraints SL1, 2 and 3, they did not confirm hard contact between the restraint and-the pipe collar.
Even though an interference could not be positively confirmed, the potential effects were evaluated. It was calculated (Ref. 21) that the additional stress due to the additional heatup to full' power conditions could result in an additional stress of approximately
'5.1 ksi in the most highly _ stressed' portion of the line. A later calculation by B&W,(Ref. 24) concluded that the increase would be between 2.5 and 5.1 ksi.' This additional stress and the added moments on-the hot-leg nozzle would have a negligible effect upon fatigue life. The effect of thermal stratification is generally to bovlthe surge line dovnvard, a direction tending to relieve such an interference. This type of interference would not have any influence on the surge line stresses during a hot standby thermal stratification condition.
i 23
l 1 IV. REFERENCES 1 Bechtel drawinE number 12501-HL-PS/PSU, Rev. O, " Hanger Location Drawing, Pressurizer Spray and Surge Line". 2 Bcchtel drawing number 7749-C-189, Rev. 15, " Pressurizer Surge Line L Vhip Restraints" 3 Babcock & Wilcox Calculation No. 620-0014-50, " Final Strest Report
- Pressurizer Surge Line" 4 Davis-Besse Unit 1, Inspection Plan IP-M-017, Rev. O, Measurement of Pressurizer Surge Line Restraint Clearances" 5 PCAOR 88-0859, G-t. 14, 1988, Inadequate clearances with some surge line whip restraints.
6 PCA0R 88-0915, Oct. 26, 1988, Whip restraint designs do not meet the USAR criteria. 7 Nonconformance Report 84-180, Broken strut on surge line snubber., PSU-R1. ! 8 Babcock &Vilcox, Inservice Inspection Report for 1983, Surface Examination Evaluatico Report, ER-83-029, dated Aug. 24, 1983. Deals with defective viper ring on snubber PSU-R2 9 Davis-Besse Operations Procedure, DB-0P-06901, Plant Startup 10 BAV-2053P, July 1988, " Report for the Evaluation, Assessment, and Resolution of Pressurizer Surge Line Temperature Cradients", The B&W Ovners Group Materials Committee, Toledo Edison Calculation No. C-ME-64.02-204, "!ressurizer Surge ! 11 Line Temperature Stratification" 12 Toledo Edison Calculation No. C-ME-64.02-205, " Surge Line Stratification - Input to B&V for Fatigue Life Study" 13 SURGE-01, Rev. 2, " Davis-Besse Surge Line Thermal Stratification Sensitivity Study", Impell Corporation, Job No. 1040-102. 14 NEG-88-01037, Nov. 2, 1988, " Pressurizer Surge Line Stratification Study - Input for Fatigue Evaluation", Letter to E.J. Domaleski, ' B&V from H.J. Cordle, Toledo Edison. 15 B&W Calculation No. 86-1173619-00, "DB-1 Surge Line Fatigue Evaluation (Considering Thermal Stratification)"; , B&W Calculation No. 32-1173620-00, Non-Proprietary version. 24 I l _j
16 MVO 7-88-0859-01, Lanyard potentiometer installation on the surge line. Includes Temporary Modif:ication, TM 99-0965 and Safety Evaluation 88-0609. 17 NE0-89-00059, Jan. 19, 1989, " Pressurizer Surge Line Thermal Motion Instrumentation Project, November, 1988", Memo E. C. Caba to R. J. Gradomski. 18 NED-89-10121, Apr.il 3, 1989, " Pressurizer Surge Line Thermal Stratification Clearances to North and Vest Valls at Restraint SL5", V.M. Watson 19 NEG-88-1041, Dec. 9, 1988, " Actions Resulting From Lanyard Potentiometer Measurements", H. J. Cordle 20 NEG-89-01000, Jan. 4, 1989, " Inspections of Surge Line Restraints for Signs of Contact", H. J. Cordle 21 NED-0294, Jan. 11, 1989, " Pressurizer Surge Line", T. S. Swim 22 Letter, Serial No. 1-494, dated January 18, 1985, to J.G. Keller, U.S. NRC transmitting information regarding failure of Gnubber, PSU-R1. 23 NED 89-10071, Feb. 20, 1989, "RCS Surge Line Whip Restraint Gap Measurements", T. Swim. 24 TED-89-085, dated Feb. 16, 1909, " Davis-Besse Nuclear Power Station, Unit 1, Master Services Contract, Effective February 15, 1098, Reference Nos.: B&W No. 582-7151, TED No. 80-008, Task 973 -- Surge Line Stratification". Letter describes results of B&V evaluation of potential interference with whip restraint, SL2. 25 Drawing No. 7749-C-189, Rev. 15, " Pressurizer Surge Line Vhip Restraints" 26 NCR 371-80, Sept. 2, 1980, " Reactor Coolant System - Package 127". I 1 I i l j l 1 1 1 l l l l l 25 _ _ _ _ _ _ _ _ _ . ._________________A
---,-n,-,,--,-----,--,-,----nr- , - , - -- - ,,--_--- , - - W 9 V. TABLES i l I 1 26
> 10077.7 4 / ' 35/22 04204 e
9 0 6 0 56/4& 5 7 0
~
- r e
n
> 4 - - 5 5 5 -
z i
- > - l
> nn :
i X > 7 muu rrr 0 nnm uu 9 . iiu-mmn r u e g I > 3 1 1
== rrr 1 rrr - -
s r
- - =
s == s u R P U s CE eet n.e r.r r .r s r.r e s O rp Ct orwrp aete Ct zeL e r . T R d H pc r aep dhpoep rur p poep - p rp - eu R . G : lLTr LVU l l : I LTTLVU dhTLVU L t : : ! :! :: !
- n e
wt / oa Mi te - 0 5 e _ l e 56257 w a. h h 70172 t t o n ) 35/2/ - e a tf C, - 6 5 D o E e T 0 7
. b y
G hra LA. 0, 5 3 - e c h c td in N F 2- ( 0 nn n ive uum - i
. 2 - e l
= = rrr 1
A s - r e r e td ir e P - ~ s r.r a sa R e U O ~ Ct zet e - F F r g R
- orwrp - - i am eonw
- dhpoep - - d st E
EDe G-0
- LTTLVU -
0 e i e un N R n -
- 09000 56255 09000 5 r u l h oo hi
~
U i 001 1 4/1 001 70177 35/33 04200 56/55 t a e st T won 9 0 6 0 A 0 0 r b l oca sh L 4 e ti 5 A hS nnn B. A nnn 4 1 5 p m m o 2i f R uuu uuu nnn e r t E - 2 1 rrr 3 rrr 4 uuu t F Pa E 1
== -
- 1
==
1
==
- rrr Ur G s r.r s e d Ot P
5 P U Ctzet e Ct zet e r.r s r.r h e RG s. R E O arwrp arwrp Ct zet e t t orwrp a M R dhpoep dhpoep dhpoep y m me r U R G LTTLVU LTTLVU LTTLVU E ; l! :: ! l I!!l!!! : b i t oh rg n I 5' U - 0 0 - 0 0-l n s T - 00454 - -004070- - 09090 0 e e fi h E L T
- 1 41 1 71 4/2/ 1,01 46/5/- 404 - 04626-66/6/ - i v
s d el B R R )
- 8, 0 0 9, 4 4 - 9 r A N T 1 1 1 1 -
4 9 - 4 d n da o ni T E R L 7, 4 4 6, 6 6 6 6 at
~
O A n i i pnxe t Z I E ( 4, nnn 5, nnn nnn o i P uuu uuu uuu- i d et I T 1 - 1 rrr 3 rrr- rrr- t n o
== - == - == a o sp g F s
R M P r .r- s r.r r . r c c n R U . CL rwrp zet e Ct orwrp zet e t zeL e od O - i U E orwrp n C f in 5 I < T R - G -;!t dhiLVU poep dhpoep L t : :: LTTLVU hpoep TTLVU! i t a o ta i i t de 5 . 0 l ! :!0 - - 0 r a nr
'F 09455 04077 - - 09000 t c ou E I - 401 77 01 400- 04622 s i ct R
1 4 /22 9 46/55 - 66/66 f a T . 0 4 1 9 4 l a i t oenr o P .
. 4 6 6 . a i p R <
m rg r t m R < imin m nnn nnn ee t ae
- 5 rrr 6 uuu uuu hl s ct T <
1 - 1
- ==
1
== rrr rrr t i
< P . s r .r s -r .r
== t l fl S < U - Ct zet e Ct zet e r . r- so a ia
< t zeL e= rh m ti O- rwrp g oep rwrp poep to.rwrp poep- e r at t
R - G - L dhTLVU
! I : ! !: : l dhTLUU L t : l ! :::::::l - iTLVU- Ca vd e on h T
ri t n Si i c 9 0 n - M - r i o - - : : : R p 1 2 3 A t a p -- - E - m p u - P P P 7 t - U U U H 2 r o t u a e
-- - O O O E R R R T T F a e H - - G G G R e H F - -
T l o - - Sb o - -
- b t d - r -
G - u n n - e N.B - i
- E u -
I T = f y -
- d
-o
-- P R o a = r R a E - d P ~ n d
i w w o
- 1
- u l -
O~E H - T - F 4 o t" i
.I
TABLE 2 ELASTIC ANALYSIS MAXIMUM CODE STRESSES (Se)- Stress (ksi) Max Stress Umits Load Temp Un. intensified USASCode USAS CMTR Ratio Case (*F) itde fil I21- I31 I41 ISI 1 410. 240 20.69 54.82 57.84 68.19 0.80 2A 627.5 240 5.69 15.07 50.64 59.70 0.25- i 2B 627.5 240 6.45 17.09 50.64 59.70 0.29 3 614. 240 16.20 42.94. 50.96 60.08 0.71 4 410. 240 22.62 59.95 57.84 68.19 0.88 5 614. 240 16.29 43.17 50.96 60.08 0.72 ' 6 614. 240 16.92 44.83 50.96 60.08 0.75 7 410. 240 25.12 66.58 57.84 68.19 0.98 8 410. 240 28.10 74.47 57.84 .68.19 1.09 9 614. 240 19.34 51.25 50.96 60.08 0.85 10 614. 240 22.13 58.65 50.96 60.08 0.98 11 409. 50 15.90 42.13 57.88 68.24 0.62 12A 506. 55 17.79 47.14 54.40 64.14 0.73 12B 506. 60 19.43 51.50 54.40 64.14 0.80 12C 506. 55 19.22 50.94 54.40 64.14 0.79 ! 12D 506. 65 24.25 64.27 54.40 64.14 1.00 13A- 506. 40 19.65 52.06 54.40 67.89 0.77
*13A 506. 50 19.63 52.02 54.40 64.14 0.81 13B ~506. 40 19.68 52.16 54.40 67.89 0.77
- 13 B 506. 50 19.66 52.10 54.40 64.14 0.81 l 13C 506. 35 25.62 67.89 54.40 67.89 1.00
- 13 C 506. 60 24.17 64.06 54.40 64.14 1.00 14 649. 40 16.02 42.44 50.12 62.55 0.68
*14 649. 50 15.91 42.15 50.12 59.09 0.71 14A 649. 35 19.71 52.22 50.12 62.G5 0.83 i
*14A 649. 50 18.37 48.68 50.12 59.09 0.82 .
f<mS.
- 1. Un. intensified stress which results in the maximum intensified stress.
- 2. Calculated code stress per USAS B31.7,1969 Edition, Equation No.12 (Se - C2(Do/21)MI). )
C2 - 2.65 for all elbows (nodes 25-40, 50-65, 105-120, 155 170, & 225 240).
- 3. Code stress limit of 3Sm is per USAS B31.7 Code,1969 Edition for A376 TP316 material.
- 4. 3Sm stress limit per CMTR mechanical properties [Ref. 27), calculated as 1
3Sm x (Sy per CMTR/Sy per B31.7).
- 5. Ratio is Code stress /CMTR stress limit.
Maximum stress ratio if different then maximum code stress location. RMERENCE : cMTR :S EX T*R$c7'Ep plapM c.Al Ct4t. o01~%&W hAGrE'~Olp RE V O2 . 2.0
p< , I ] TABLE 2 (Cont'd)- I
. El.ASTIC ANALYSlS MAXIMUM CODE STRESSES (Se)
Stress (ksi) Max Stress t.imits Izad Temp Un-intensified USASCode USAS CMTR Ratio Cae (*F) rede fil f21 131 f41 f51 15 410. 40 13.48 35.71 57.84 72.18 0.49
*15 410. 50 13.46 35.68 57.84 68.19 0.52 16 614. 40 12.19 32.30 50.96 63.60 0.51
*16 614. 50 12.01 31.83 50.96 60.08 0.53 17 409. 60 20.55 54.45 57.88 68.24 0.80 18 506. 35 27.30 72.34 54.40 67.89 1.07
*18 506. 60 26.31 69.71 54.40 64.14 1.09 19 649. 35 24.67 65.37 50.12 62.55 1.05
*19 649. 65 24.18 64.08 50.12 59.09 1.08
- 20. 373. 50 10.54 27.92 58.69 69.20 0.40 21 373. 50 8.09 21.44 58.69 69.20 0.31 22 373. 50 11.02 29.19 58.69 69.20 0.42 23 373. 60 15.50 41.07 58.69 69.20 0.59 24 600. 40 14.17 37.55 51.30 64.02 0.59
*24 600. 60 14.15 37.50 51.30 60.48 0.62 25 600. 40 12.42 32.91 51.30 64.02 0.51
*25 600. 50 12.26 32.48 51.30 60.48 0.54 26 600. 35 18.31 48.53 51.30 64.02 0.76
*26 600. 60 17.46 46.28 51.30 60.48 0.77 27 650. 35 11.12 29.46 50.10 62.52 0.47
*27 650. 50 10.83 28.69 50.10 59.07 0.49 28 650. 35 18.06 47.87 50.10 62.52 0.77
*28 650. 50 16.62 44.03 50.10 59.07 0.75 PCTES.
- 1. Un intensified stress which results in the maximum intensified stress.
- 2. Calculated code stress per USAS B31.7,1969 Edition, Equation No.12 (Se = C2(Do/21)Mi).
C2 = 2.65 for all elbows (nodes 25-40, 50-65, 105-120, 155-170, & 225-240).
- 3. Code stress limit of 3Sm is per USAS B31.7 Code,1969 Edition for A376 TP316 material.
- 4. 35m stress limit per CMTR mechanical properties [Ref. 27), calculated as 3Sm x (Sy per CMTR/Sy per B31.7).
5. Ratio is Code stress /CMTR stress limit. Maximum stress ratio if different then maximum code stress locations. REFERENcs : Dnrn :s EXTRncrec VROM CFLCuLnr:cu SMR GrE-ol, Rev. C2 , 2s4
I t h hh I. 1 - l' g W i un g - i ..
. = .
g-i g- o. i 1 o g CEG8 FSER 8E8R 0030 8888 CAR 8 E838'$588
- g w :::a sasa sasa sasa saa saaa ::a : :a m .'
?
H gg . I E l, - L =W'R8: ! ER ! 8Eli CE!! G8ll 28ll E8!! $0!i c gyl :: sa sa sa sa sa : : 0 -Ev b' J 54 0
- 0 0 0 sg
~
- s 0
=
M.
=
~E% i H
g k*0 ! ' I i
- RRi! 20li SR!! USI!!GRI! C C ! ! , M C 1 l' $0l l l
- gW,f g,9
~E2 sa sa isx ls s
- E gg. mv W
I r 28ii EE!l $2: 1 C;!! TEI EE?i 08R! R$R! H g? : sa sa sa sa saa a. :: J -e i W kB l . t Q g . . . . . .
.. .j . . . . . . .j .. .. ...j.... .. .j . . . . .. .. .. .. . . . . . . . .
3 E R R R R 22 re2 Irm2 red M 7 RC lRC 40 RE Ass 'AAS Ass lYI RF>
~
7.55
~~~
-~~
i hh
-~~
~ 55
~
~
-~~
i hh EQ 3 2m -- 2 s ish s ii) iiis iii)
-- 2 -- 2 W
u gkW vg-m $g-u $g-u [g-u ggum rqqu ryou --- vqgu
-, es* ss ss ss ss" --- ---
5 --- --- ~~~ H. L . . . . . . . . . . . . . . . . .. . . . . .. .. .. .. . . . . . . . . l".......
$ ~W "W ~W ~W gW g g' g W
g p < < < < a e e c L N b3 Q n, e s e * < d 4 g i g o - a o - -7 - g r e < v o 3"g y y r g I 7 e = r w
- r n & % %
a R 2 8 8 8 s s u
. j 1
TABLE 4 ! l PRESSURIZER SURSE LINE RESTRAINT BAP CHANGE ANALYSIS 15-Feb-89 CLEARANCE BAP-IN. ' START END SHIM RESTRAINT PER10D BAP DIRECTICH FERIOD FERIDD CHANGE ADD N DIRECTION DF CHANGE CCMMENT SL1 S/U TO 5/B0 A HORIICNTAL ? 1.31 ? UNKNOWN WITHIN TOLERANCE B VERTICAL 1.50 9 L'NKNOWN WITHIN TOLERANCE sib 0 TD 6/02 A HOR!!CNTAL 1.31 1.3B 0.07 WESTWARD WITHIN MEAS. ERROR S VERTICAL 1.50 1.50 0.00 NO CHANGE 6/B2 TO 10/BB A HORIZONTAL t.38 1.25 -0.13 EASTWARD WITHIN MEAS. ERROR B VERTICAL 1.50 1.75 0.25 UFWARD WITHIN MEAS. ERROR NET ID 10/20 A HOR!!CNTAL -0.D6 EASTWARD WITHIN 91EAS. ERROR B VERTICAL 0.25 UFWARD WITHIN MEAS. ERROR SL2 E/U TO 5/B0 A HORIZONTAL ? 0.94 ? UNKNOWN WITHIM TOLERANCE B VERTICAL ? 1.94 ? UNKNOWN WITHIN TCLERANCE 5/B0 TD 6/B2 A HORIZDNTAL 0.94 0.88 -0.06 EASTWARD WITHIN MEAS. ERROR B VERTICAL 1.94 1.75 -0.19 DOWNWARD WITHIN MEAS. ERROR 6/02 TO 10/B8 A HORIICNTAL 0.BB 0.BB 0.00 NO CHANGE B VERTICAL 1.75 2.00 0.25 UPWARD WITHIN MEAS., ERROR NET TO 101BB A HORIICNTAL -0.06 EASTWARD WITHIN MEAS. ERROR 8 VERTICAL 0.06 UPWARD WITHIN MEAS. ERROR SL3 S/U TD 5/B0 A HOR 12CNTAL ? 0.56 ? EASTWARD OUT OF TOLERANCE B VERTICAL ? 2.31 ? UFWARD DUT OF TOLERANCE 5/B0 TO 6/02 A H0RI!CNTAL 0.56 0.50 -0.06 EASTWARD WITHIN MEAS. ERROR B VERT! CAL 2.31 2.50 0.19 0.63 UPWARD WITHIN MEAS. ERROR 6/22 TD 10!BB A HORl!DNTAL 0.50 0.50 0.00 NO CHANGE B VERTICAL 1.EB 1.00 0.00 NO CHANGE NET TD 10/2B A h0R!lCNTAL -0.06 EASTWARD WITHIN MEAS. ERRDR B VERTICAL 0.19 UPWARD WITHIN MEAS. ERROR
.. .. .._ ..... .. .. . . = ........ -- .... .
SL4 S/U TO 5/B0 A HORilCNTAL ? 0.19 ? EASTERD OUTOFTDLERANCE B VERTICAL ? 2.94 UPWARD DUT OF TDLERANCE 5/B0 TO 6/E2 A HOR!IONTAL 0.19 0.13 -0.06 EASTWARD WITHIN MEAS. ERROR B VERTICAL 2.94 3.00 0.06 1.13 UPWARD WITHIN MEAS. ERROR 6/E2 TC 10/8B A HORI!DNTAL 0.13 0.25 0.13 WESTWARD WITHIN MEAS. ERR 0li B VERTICAL 1.BB 2.13 0.25 UFWARD WITHIN MEAS. ERRDR NET TD 10/EB A HORIZONTAL 0.06 WESTWARD WITHIN MEAS. ERROR D VERTICAL 0.31 L'PWARD
e TABLE 4 PRESSURIZER SURGE LINE RESTRAINT GAP CHANGE ANALYSIS 15-Feb-89 CLEARANCE 9AP-1N. START END SHIM RESTRAINT PERIOD BAP DIRECTION PERIDD FERIOD CHANGE ADD 01 DIRECTION OF CHAMBE COMMENT SL5 ~ S/U TD f/B0 A EAST ? 0.94 ? EASTWARD DUTSIDETOLERANCE D SOUTH ? 1.BB ? NORTHWARD DUTSIDE TOLERANCE 5/B0 TD 6/82 A EAST 0.94 0.63 -0.31 EASTWARD D . SOUTH 1.BB .1.50 -0.38 SOUTHWARD 6/82 TO 10/BB A EAST 0.63 0.50 -0.13 EASTWARD WITHIN MEAS. ERROR D SOUTH 1.50 1.50 0.00 NO CHANSE l NET TD 10/EB A EAST 0.44 EASTWARD D SOUTH -0.38 SOUTHWARD SL6 S/U TO 5/B0 B MORIZONTAL' 1.81 S NORTHWARD DUTSIDE TOLERANCE A VERTICAL ? 9.31 ? UPWARD OUTSIDE TDLERANCE 5/B0TD6/22 B HORIZDNTAL 1.21 1.75 -0.06 0.25 SOUTHWARD WITHIN MEAS. ERROR l A VERTICAL 0.31 0.25 -0.06 UFWARD WITHlWMEAS. ERROR 6/82 TO 10/BB B HORI!ONTAL 1.50 1.3B -0.13 SOUTHWARD WITHIN MEAS. ERROR l A VERTICAL 0.25 0.13 -0.13 UPWARD WITHIN MEAS. ERROR NET TO 10/BB. B HOR!!CNTAL -0.19 SOUTHWARD WITHIN MEAS. ERROR-A VERTICAL -0.19 UPWARD WITHIN MEAS. ERROR SL7 5/U TO 5/B0 B HOR!!CNTAL ? 1.56 ?' UNKNOW WITHINTOLERANCE A VERTICAL ? 1.25 ? UNKNOWN WITHINTOLERANCE. 5/B0 10 6/82 B HORI!ONTAL 1.56 1.75 0.19 0.25 NORTHWARD WITHIN MEAS. ERROR A VERTICAL 1.25 1.31 0.06 T'WWARD WITHIN MEAS. ERROR 6/B2 TO 10/SB B HORIZONTAL 1.50 1.56 0.06 NORTHWARD WITHIN !!EAS. ERROR A VERTICAL 1.31 1.19 -0.13 UPWARD WITHIN MEAS. ERROR NET TO 10/GB B HM!IONTAL 0.25 NORTHWARD WITHIN MEAS. ERROR A VERTICAL -0.06 UPWARD WITHIN MEAS. ERROR SLB S/U TO 5/B0 B HORIZONTAL ? 1.56 ? UNKN0H WITHINTOLERANCE A VERTICAL ? 1.75 ? UNKNOW WITHIN TOLERANCE 5/B0 70 6/B2 2 HORI!ONTAL !.56 1.63 0.06 NORTHWARD WITHIN MEAS. ERROR A VERTICAL 1.75 1.56 -0.19 UPWARD WITHIN !!EAS. ERRDR 6/52 70 10/SB B HORIZDNTAL 1.63 1.81 0.19 NDRTHWARD WITHIN MEAS. ERROR ! A VERTICAL 1.56 1.56 0.00 NO CHANGE NET TO 10/BB B HORIICNTAL 0.25 NORTHWARD WITHIN !!EAS. ERROR A VERTICAL -0.19 UPWARD WITHIN REAS. ERROR I yi l L_______.__-_
P .-
]
d b TABLE 5 PRESSURIZER SURGE LINE THERMAL STRATIFICATION STUDY. > RESULTS OF FATIGUE EVALUATION TO END ADD FOUR LOCATIONS ' MAT. CYCLE 5 HTUP/CLDNs TOTAL HL/SL N0ZZLE AS A BRANCH CONN. CS 0.602 0.017 0.619 HL/SL N0ZZLE MAT. DISCONTINUITY CS 0.683 0.021 0.704 HL/SL N0ZZLE MAT. DISCONTINUITY SS 0.335 0.008 0.343 .
' HOST CRIT. SURGE LINE LOCATION SS 0.057' O.006 0.063 (B&W loop joint 5)
PRZR N0ZZLE MAT. DISCONTINUITY SS 0.265 0.032 0.297 PRZR N0ZZLE MAT. DISCONTINUITY CS 0.566 '.068 0 0.634' The maximum Total Fatigue Usage Factor is equal to 0.704. (.< 1.0). l 32
TABLE 6 LOAD CASE DESCRIPTION
'HEATUP OPERATIONS:
Steam Bubble Formation Stratified flow ~in lower run,-vertical and upper horizontal pipe at uniform temperature'. ' Existing restraint gaps, no restraint contact found. Load Lower Hoz. Vertical Upper Hoz. Press Hot Leg Case Run Temp. Run Temp. Run Temp. Temp. Temp.
'F 'F 'F 'F 'F 11- 409/100 100 100 409 100 Midway in Heatup Stratified flow over lower run, upper and vertical run, at uniform temperature. Consideration of current existing restraint gaps is included. Contact at restraint SLS (-Z) based on 1.75" gap. .
Load Lower Hoz. Vertical Upper.Hoz. Press Hot Leg Case Run Temp. Run Temp. Run Temp. Temp. Temp.
.op op op op op 13B 506/120 275 375 506 375 Toward End of Heatup Stratified Flow over lower run, vertical run and upper run at uniform temperature. Consideration of existing restraint gaps. No restraint contact.
Load Lower Hoz. Vertical Upper Hoz. Press Hot Leg Case Run Temp. Run Temp. Run Temp. Temp. Temp.
'F 'F 'F 'F 'F 14 506/205 500 500 649 500 1
33
..z ,
J
. TABLE 6,(Continued)-
COOLDOWN OPERATIONS:-
, -Transient at RC Pump Trip and Pressurizer Venting Stratified flow in lower run, vertical and upper-horizontal pipe at-uniform temperature. ' Existing restraint gaps,.no restraint contact.
- j.
l Load Lower Hoz.- Vertical Upper Hoz.- Press. : Hot Leg l Case- Run Temp. Run Temp. Run Temp. Temp. Temp. op. *F 'F 'F 'F
- 11- '409/100' 100 100 409 100 l '.
l' t-l l i l l h 34
1, . yi I l
-TABLE 7 ESTIMATE OF THERMAL STRATIFICATION CYCLES APPLIED THROUGH CYCLE 5 HEATUP OPERATIONS:
Steam Bubble Formation LOAD CYCLES HEATUP TOTAL CASE PER HEATUP CYCLES CYCLES 11- 3 36 108 Midway in Heatup LOAD CYCLES HEATUP TOTAL CASE PER HEATUP CYCLES CYCLES 13B. 1- 34 34 Plastic Deformation Assumed 2 2 Toward End of Heatup LOAD CYCLES HEATUP TOTAL CASE PER HEATUP ' CYCLES CYCLES 14 1 36 36 COOLDOWN OPERATIONS: Transient at RC Pump Trip and Pressurizer Venting LOAD CYCLES C00LD0VN TOTAL CASE PER C00LD0VN CYCLES CYCLES 11 1 36 36-
. 35
C* - A' TABLE 8 STRESS DATA FOR LDA0 CASES !!. 138. .it LDAD CASE 11. LDAD CASE 130 LDAD CASE 14 PRESSURIZER N0ZILE COM. STRESS = 7971 PS! STRESS = 9906 PSI STRESS = 8326 PSI B & N NO K i 2 Ms 59752 I M BS Ms 1.01E5 IM86 mis-3.92M 5 I M BS Ms 1.025E6 I M BS Ms 1.2K6 IMIS H s 1.03E6 I M IS M s 1.616E5 ! M IS N!= 1.59E5 I M BS M!s 9.32E4 IN-LBS ELB0FB 6 N NODES 2 & 7 STRESS = 49450 PS! (M K 21 STRESS s 61302 PS! (N0DE 2) STESS = 50734 PSI (M0E 21 Ms 1.07E6 IN-LIS (NOK 7) mis 1.34E6 I M is (NODE 71 M s-l.!!E6 'I M BS (NODE 71 Ms 1.38E5 IMll Ms-1.!!E5 I M IS Ms 4.65E4 IN-LIS !, Ms 1.03E5.,!N-LIS Mis-3.66E5 IN-L46 M1= 2.24E5 IN-L36 ELBON-B & N NODES 5 & 15 STRESS = 50700 PSI (NOK 15) STRESS = 62739 PSI (NODE 15) STRESS : 50713 PSI (NDDE 15) M -1.07E6 IMBS (NODE 5) Ms-l.342E6 I M BS IN0DE 5) Ms-1.00E6 IN-LIS INODE 51 Ms 1.9E5 I M BS
- Ms 9.24E4 I M IS
- M s-2.18E5 I M BS .
- Ms 8.08E4 IN-LBS MI: 3.77E4 IN-LBS M1s 3.15E4 IN-LBS 1 ELBON-B k N NODES 13114 STRESS = 34840 PSI (N0DE 131 STRESS 44817 PS! (N00E 13) STRESS = 37389 PSI (N00E 13) "
-* Ms-6.08E4 IN-LBS '
L Ms 51906 IMBS (N0DE 13) Ms 1076 - I M 85 1-Ms 7.11E5 I M 95 M s B.88E5 I M BS Ms 7.22E5 ! M BS MI 1.64E5 IN-LBS Mis-3.04E5IFLBS Ms-2.99E5 IN-LBS l
'. BON-B & N NODES 16 n 3 STRESS 23267 PSI (NOK 16) STRESS = 20949 PSI (NODE 16) STRESS = 24693 PSI (NOK 16)
Ms 1.58E5 IMBS Ms-3.16E5 !MBS M s-3.36E5 ! M BS L * '
- Ms 2.40E4 I M RS Ms 5.79E4 I M BS MYs 1.16E5 IN-LIS M 4.73E5 ' IMIS
- M!* 5.45E5 ! M BS MI: 4.14E5 IN-LBS
=
ELBOWB k N N0 DES 23136 STRESS : 43298 PSI (N0DE 36) STRESS = 55117 PSI (NODE 36) STRESS 4S145 PSI (N00E 26) I Ms 6.!!E4 I M BS Ms 1.54E5 IRBS M s 2.12E5 ! M BS Ms 6.03E5 I M BS Me 2.2BE5 IMBS M 6.91E5 IMBS MI: 6.B0E5 IN-LBS M s B.52E5 !N-LBS M!s 7.29E5 ! M IS HOT LES CDMECTION. STNESS = 10171 PSI STRESS 13353 PSI STRESS = 12120 PSI 9 L U NODE 54 Ms 6.11E4 I M BS Ms 1.53E5 IMBS Me 2.12E5 !MBS M 6.90E5 I M BS M 9.5E5 I M BS H s 9.!!E5 I M BS M!s 7.75E5 IN-LBS M!s 9.72E5 I M BS M s B.24E5 I M RS NOTES:
- 1. STRESS VALUES(ARE g-, -C2 e--[
* (-t h - 6'-[ f. (0 q
- 2. PRESSURIZER N022LE STRESSES ARE BASED ON AN ASSUMED 14.75' DD AND 3' WALL THICKNESS PIPES.
- 3. HDT LES N0!!LE STRESSES ARE BASED ON AN ASSUMED 13.75' 00 AND 2.5' NALL THICKNESS PIPES.
- 4. ALL MOMENTS ARE PER TNE B & N C00RDINATE SYSTEM.
'6
n _ - . _ _ _ - _ - _ - _ - _ - - - - _ _ . , _ . - - - - - - - . - - . _ _ _ - _ - - _ - _ . - _ -- _ -- . _ _ - _. - _ _ - - -_ 1- - .---I 8---t 1 1 1 1.I y y i I I I i i i i i l i 1. C W IT l @01000l i 14 @ @ l m T C) i C T C i N m f6 i e @(O i a N a i m =(0IibO Y l
- l- I 'TOO00I I no m io I (010 @ l N a T I O N Y I (e T 101 - T @ l t0 0 m I ib @ N I T Z I ai @ l0 N m l0 l i * @ Q l T
- T I lb @ a 1 O IO (D 1 to - T I O ld Y I a m N 1
- O a i e i m.e l i N I I .. 1 ...I . .1 . .4 ...I ..8 ...i ...I e H IyeI i @ l Iooo1OC0iOOai * * * ,8 O N a i O*OiOOOIOa0I L.
I al O I 91 1 I I ! 1 I I I I I i 1 I I I i l I t' J' 1 8 40E4 0 1 8 4 I I I I i i i OL a l Jai 1 I I I I I I I I I-3 M i--------- I--I---1---1---I---1---1---I---8---1 44 2 1,1 Im i 1 1 1 1 1 1 1 1 I i 34
- W I (0 1 2 0 010 m i 1 O rb T l T T m i CD (D a i N O ID i N m c 1 O N (D I 10 (D IO I m (O N I C L m m- Iw I Olb Nfb N I I (D w N I lb @ (0 8 @ e lo 1 W tb d Iw@nIlbNinI(DmaiSNibI O9 J
- .U 1- ai 10 (0 a (0 (O I IOYOINa@lTmYI@m(0Ialofb I (0 @ T I Q @ (O I O tD a 3 0 C. 2 I I/t w I l \ l l . ..l ...I . ..l . ..l ..4 . ..j . . 1 . ..I EE C 1OeI i @ l i 000i000100* IO-NIQNa 1O*OIC00l00OI Le F I al O I I I I I I I I II I I i I i l i 1 i i i i I i 1,1 I to.A 1OEI M i a I i l I I i i iI W m iJai i l I i l I i i i I4a K yt !---------I~~l---1---!---1---8---I---1---8---I
- 1F i i 1\\\lN N4\ l\ l \ l \l NI \\l48 1 Z > I (D 010 0 01 1%NNINNNIN001%TTINN@ImmNIf-AINNNI D3 y iW01OO*aOmi i N N N I N T N I \ c @ l \ @ l0 i T \ (O I T 10 % i @ a \ l T N N I C E I a Z l = @ T N a(0 1 i \ \ \ l \ fb \ l N N Y I N m (0 I10 \ a 1 Nib N l T N N I T N \ l 4 0 1,1 i lA I 10 NN I I\\\l\ .\lN .1% ..I .Nel ..NI . eN I .N \ l Ca W m8ZYIC ON I INNNINONIN0*INONiDN*ICONICONIONNl0*
at WiCN4 50 0 1 I\\\lNI\lNI i iNI li Nl I iNI i NI \ \ l 0 0. IV ylt 1 Ym I l\\\l\ \l\ I\ l N I \1 \1 \\l , Ea>= 1 10 i i\\\l\ \l\ l\ l N l \l \l \\l 98 M 3, Dg---------l--l---l---l---l---l---l---l---l---lyd. e i i i i i i i i i i i0 m si i l i I i i l I e I I 8 .D E 5 WI I I I I I i i i i ! Ie* 4 1,1 I I l l i I I I I I I X m Wl i I l 4 I I I I I I lVO VI - 1 I I I a l 4 i I I I I eL kl I i i i i i i i i i lea th I i i i i i i i i e i l ea 8,8 8 1 i l i l 1 i i i i iE4 mW m.8-------~~l--I-~~l---1---I~~~l---1---1---I---1e y 11 I I I I I I I I I I I40 E . IN I (O N O D O I I w O m I m O lb I P O 10 i m N w I @ m (D l m tC a l N tD Y l m (D N I M4 a f" i N a I tb (0 0 N N l I a N Y ! a m l01 T O N I ib @ (4 I a Y Y I m * @ l O Y O I N iO N l 4 W L seg.II m w1 (0-***I 1 C O O I O N (0 I O lb ib I O - N I N Y m i 10 m N 4 - T N I Q 10
- i L a I i A 4 I . ..l . .1 . ..l ..ei ...i ..el ..el ...I ee J miUat i O I IO00i0001000iOad IQw0i000i00010001*7
- Ul Ea .e I I I I I i I i i i i i lI I i 1 I I I I I i i i i l i I L m l. 8O=I (0 i i i i 1 i I I i I 90 II i
- i I l I i l i i i IL g .,1-------~~l--1---I---I---1---I---I---I---1---1 3C
&i I I i 1 -1 I I i i I Ida F d I SO0001 s@mNimmNIONAlematonnlovNiponimmele Osw f-ai C0000, i m O O i a m e. i C o @ i a N T I O e D i N m C i n - m l N a m i c o I ai T w w a -* I i 0 0 - 1 O Y 10 10 0 w 1 O lb e i T O T i @ N T I N IO N S O @ w I eC N 1meiI N I e . .1 . ..I e ..I . ..I ...i ....I e .el ...Iat b t i U al I m i I000i0001OaaiCawIONwIowcIO00I0001 E*
ri EI O l i l l I I i iii lI l i I .8 I lI l i 1 1 I i i I I I I 9R H es t Q a l T I I i 1 e i I i 1 14C e UtJ l i i i i i i i i i 1 Z u 1 -- -- -- -- -- - - -- 1 -- - - - -- - - - I - - - - I -- - - - t 0 L
= WIF i 1 i\\\lN Nl\ lN 1 \ l \l Ni \ \ l a4 J l I Z > 1 (D (O N O O O I l N N N i N N N I N m T I N (O lb I O N W l N IO N I a m \ l l0 N N I
- 1 I W 0 i O N t0 0 N N I I \ \ \ l N a N I N m m i N O C l T Nib I O N N I N O N 1
- N N I Q.C I, I a Z l (0 * = a a i i N N \ l \ a \ l \ (O m i A m @ l (0 \ @ l (0 @ \ l Q N N I N \ \ l I*
u E I m i I (O \ l I\\\l\ .\ l \ . .I\ **1 *\ *I .*\l **\l . \ \ l 44-W plZOla m i l\\\lNO\lNCOINOalONQlOCAIDONIONNIOC m .M I C e i Y I I\\\l\l\I\l I INI I I IN1 II IN3 i \l \\I I 3 Ul miE I N i INN \lN \lN IN I \ l Ni NI \\l 0. 0 gg WIF I I INNNIN NI\ lN I N I NI Ni NNI 0 .C vt I--~~~----1--I---1~~~l---1---8---3---I---1---1FM y 1----~~---l--I---I~~~l---1---1---I---1---I---1* W i i l i i i 8 a l i i 1 m . I 1 CC I I I I I I i 1 1 I mW I I 33C1 a l i i i i i a i i a i i I LL3I I I I i l I I I A I I I Li I I a i i l 4 i I e m i I LL i 1 I I I I I I i I l I L Ul i 1 4448 mi a i 1 I I I I i 8 3aLi Wl 1 I i 1 l i i i#
@ M IW IWN40101 FI I l 1 1 1 I i 19
- g im IELO-3>l El i I I I I I i 14
/ IC i = 0.1 I I I I Zl i I I I I I I I O
( IU l 1 l H H k- P P- F I =i X Ut N I X DI N 1 X3NI X3 NIX 3N1 X3Nl XJnNI X 3NIZ i 3 ...I i Of El I i I i I i I I I I i I i i l 1 0 i i 444 Oi- iN im if in le lJ iN ll im 83 I i i 11 n m l OfJ lJ lJ lJ lJ 8 I I WWWI Ulm im im im im im im im i N
- , - - - - - - , - _ - - - - . - - - _ - - - - - _ - - . ---_.,r- -_-_-- - ,- - -- -- - _- - - -- - - - - - _ - _ - - - -r-- ----r, 4 .
VI. FIGURES l 38
- ,;, g
, n p in
- u ie af n
,, t a
e mil ee o H ' Ru g eg n a it t i t i r a u mF a D it s i c E f i t a r t S me ui p
/ s nt oi n *-
a l ia t m lmP ct r e a u u ao Mht t s les f eR e Au m r ein inWiar t D p ei nh n e c n i p i g n mst i io se ha t e t r s s e er e mW r e n " s r e pwire et Sen R ep ef ai t pm h t esp D / t teP i r r t e St GiryC ae T - oW n ih D h t i t eI n ay i mit r s i t ni n Gl wm p n ,a tra f e 1 it Ec e 2 Ess W l uB ad ve C o ps Nd ei enn i n Sl Ae CD f Se Ar CS t Es u a
.e eR c i f
e a a D L C e . =4 s ~ g l r u - ~ - l S s d lo t a dnl r t n m u Cs aaa Drerm e L i e s m t t nn ie l aba i z Sh P e s t n a r i x a a t r s er m nmh o uT s T M i e r l n t p's Rseu taNfo r e lc u e i@ m r en n o TA a e l pa ie et pad Omt u e C y s s t e o l P v eD e hM Wp i i wts ng s a DC n i d na iG e iEa v e rh f Da en t b O a edM Rna PP
~ ~
g
iiI l\
,- h -,ox s
*g - E 4
m
- - l Npl ~
. N I
L E _E- G R 7 U S s _ 4 _ _ _<g 1 _ T i _ a +p I N U n iI g'A
& y, 1
s
., E S
S
/ 1 E B
r /j / - - 4' S
- ~ I V
z - A n - ~ D u l-NW.e I s s 1 e r g
, - P ~
3 0
-3 s
f~ i I w\ \-
~ ~
=,- - ,L J 2
e 1
~ -
/p q i F
r u g
- X
,7 4 ~
,,, 4 h
~
/
t r
. o y, 1 R
~ ~ N s
- ~
2 % ,'
> i
- = _
- H f(r -
f - nl _ g-
. n 1Qy,.
, a y, s o s i . s L .
1 E ~
' .,3 v.
2 4 2 5 g
* <9 p
,0
,if
\
i:
)= o- g! %
aa 5 u ..-
% i 1 h
Et m@ o,n S ? 3, g, . "
.y u
iR.i,? g ~ y e
- 3
/
j(4 m e 3 .. g ops ,
~ -. i 1 4 -
st t N $. tg T / *= es il,9 4e t, *s ea wu
) gh f*,o git 0 fS6
= we =e .
s i f*r %% N %t j 5 , e *k0 - C E f30- ag 3
t e -
x ( $$d x"s h 3I p q 5 g5 o, 9 n q,
- SE (t 5 4 * @
S J E h i c5
$ t i~
-m =E m k aea e!
c r8
* & 'it o 4*
m? dck' gu 4$ e w
$e Dw I gd. 4 o se se W
.k pb hp
'I '{ sMn
/ i na Lag' ev_ g.,
s og
<x d
$$k tret
)
D m g - - 2 E
?
d
$20Q g
k > e =gc18s Rs YI
[., j m ,
- 5
>,nos we -
_ i N aumr a.e -
*n s i r o n ~ > aus i 1 _
1 - a f I s I a a% I elig' - I k
== iu e
- gVe akces % earaunuen so,,s i !* 3 u>s eth I~ i
'g$U y.- l
,A*lNu""s""A,1l1 - E o
ws* 2 I-- all ,,,, se a .,e ww,, 3 y - g I o n s sws -n - ssws ouaanuun *- ' na nasu m - Ir \ i 2
!! \
\
2 i g
\ t _
i g ~
=
et \ { w 8 i \ " -
- 5-w I
e
% g \
t ar es
~
warwo.m sa .s,
, \ 3 a,=,
a s ~> > ma nu . .- , os es ou ne a,ue - - % y n,s - surnau unos n. nu q - y
. e e, , Toi*TrlTz %#- %
Y
> > .e s usa , , ., so una, - I s \* -
t= j - a w. -
- t \
,8,1
, , , , .un uan e e,nomer g i t
E sas 'ensas n event it r w w **d s*w msw *rtwens E . -
! \
nu m ,., un. I- s N-- 3
,, n ,,, .4-u
" ' \o; -
*'** en nM&d*:.;'.*h% --
** + : t"l :n??.t':%t'R*Ts 'tas,
,v. , . u n o, , . .. \ _
\
, , , , saun., ,,, _
,o . u,n , ~ ., az,,,
8 a ** l l l l l l
** ,a ,e.
s"eaarms ay somnu-n I - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _62
13$ ,g '
$g$
g .uh ,;t r 3 g 4 yg=*g ) gg. 2 )."gw.;C'. 34gS O[ :,w* Ee 2s g?.g , ,_ .g
- on'
'O o g gn" C o" g - ,9* 4 6 7 F T T I 4
1 S r r 3
, * *5 A u 6 8
)
P C 0 o 0 3 '
, 5 t 6 e 4
D 5 0 r A ,
, 3i 2 & O ,
g L / F s o p / b
- s, N 8 G g <
, ' 0 4 a 6 c
. TH U g y
s r 1 c D 8 o S g
- A R o 6 A 7C i
D g o o M L W B T M% - s d l
, t 4L e 1
3 P . m 6M F !
- ,' J M
F o r TTE%s - 0 G r - 3: EN U d e ! e d . 5 1 t c a A
*LT LLUrp i n
]
I r T : )sI ES H , t b s A C TC 1 m m . T 0s s 2eO a0 oSEG t I A u L g a F A; a 1 ne r _ t SS U t *F L *o ( )H a . S t e (T S SA D p l o wI
. s .t E 4
h c AEi E a p p e s ' i T Z i l r 1
'l MD ,
, a LT 7
T e a n N K. m r< g - i i h l e rN a[ ' e r u
- f.
M e [,f 4 ( g e =
- - - O
- - - - ~ - -
0 4 02o 0 o 8 0 4 0 0 08 6 0 04 02 00 806 0 04 02 2 0 1 1 1 1 3 3 a 2 2 2 2 1 7"n3p f . p28* *
.w i 1I l !
iiiIi n:kTt {
,. :3 ; g n sa3 N g,,S ) ,T @ N$mt%1 k% hr
- M x,,.:I5n{
n.n .$u~ ",c" .'.r' e. n
,2 sE 1g e3 2g{u m nk .
G 4 2 0
. 2 l
)
P - g 3 5 y . 0 g 2 s p o 8 G ( g 6 N THU p - l i
. 1 6
8 H c 7S S p P A D o TEM (p u 7 W 6- . u . i B TP Mp M A m E T p . o [ % r EG y 8 0
- i 4
wLH F .I 2 1 1 B U 7 l d LT g 7 e T I p r
)=
1 t c a
/EE l
l L i s J4 \
]
- l l
t r cTC o l t
/
r T F 4O
- O h
I b s L S E. l f i r 7 ' o (
)
A x
.S5 GGi o
9
, C t (I I
ta a A S SUA f
- r t 4 I
8 h :l i D eel hf- t n s 1 i 7 h i l c MMD i
~
~
~
p e f
,o G
E L r e 7 a 0 K. ~ 9 1 m . 4 _ i ' e ' _ h ' l e ' u M
- 4 i t,
( f A e - _ - - - - - - ' O 0 0 0 0 o 0 0 0 8 0 6 04 02 00 O D 0 60 40 0 2 6 2 0 2 3 0 3 8 2 2 u2
. 2 1 1 1 1 1 g" n. 2 f
_ iI 2j . l
2 ( w i% y r
*L l
.(R g
3
- s. -
)*
1 I i ag a gg I 4 _ [l A M q 4, ( i.pg .
//, ".
0 % 2
/ .
O pp y a s E 9
' 1 L
%(,[
L S
- a d S
/
O . Y 2 . S L '
'p
/
7L . > - S% 3'
/y I
. /[ -
G l [. ,' t
,7 p
g
- g ,
7/ g (
' s ~ ,T
. _;h , f' (4 '
- M g O.
\ G L I
g g
'g s .
N R s i S
/ t.
A so7o t N%
. i
' 'r R v. .
% .ms / t M, s '19 w 8
R. 'H ( 1 F L ro
- 9
,'S5
*- 9
,t E onoC G( -.L c%m n r e
t P l# 4 Cc
. 5
@sc io r
1 e A f cet wwTut.
%, 5 au r
f ( a re CSd( . JI P L P * #N
. Gs o aL *w t
G>E t S $ t l _ s Hes 1o f e L E -
- - M f c (
l gw$"pMgepBe t3 4,w o.
- l
_ 5 I 1 1
/
01 3 1 l z - n" gth I\ 1
3
.. E I
gO
.- i
\ V :
n n N _b 1 t i I' g ol g/ '1 V: 4 sn tU n r
/,/ gd nw ie o e z
s
/~
~ .
Ps _ i r u s e - s sB M e t r s Sav i s e r w r - g
, - P Z hD g
in
. ~
I I q-
- \ \C' YL; /
Ho _
/:-
8
- s _ r_ O ei nd iE L
5 1 X
%ER U
G I F e o ' g d - c' h re t r o ST ulo 1
~
~_'
~
y/ , - N' 1
" 0 a 6' -
' 0 7 -
6' , 3 5 L E 9 bg [l
~ . - _ _ -
\ /
.T 1 i
5 d - w W
~
5 l 7
!! "i g (' , .
e \' 5 0 E s o a i s
- 3 $
J !l! h I I* y l f5 (\IN b h! ; E v
&v
\ i #!
se ' li \ d_ ieE! I J C8 ' 4- i
* ~e
\ \ .
'\
e x%= 4 i
\ k C
\ 7 " X A FIGURE 9
,t. ' vXX ~
s
^iX r
47
_ _ _ _ _ _ _ . _ ..._._____._._..m.. _l___.
' m__.__i
. . . .s.
3 ,0 ,0 . . L [
- r . . .
00222.. . . . . .
~ ~ '
I
.....f
>( . . .
. . . . . . . . . . . .g 00 21 81- . . . . . .
'< . . . <g 00 . mt- .
.Cd 5
S M
~
N p o ./ . 00 .. .,_ . . . . . . g........
.g E --
1
. . . . . . ~. ;~ Y Z
r e ic o
.g
- 9. . R g00e0 =s .. . . . . .
8 [. .) %( I k n
- e:
b dhw O s. 00 m aS- . . . . f f........ . . p...
.)g
~l p m
.g 00 21 :70- .
b 00 ..._ . . . . . . . . . . . . . . . . . 3
)( . . .
0
. . . . . w . . . . .
.g 00 43t0- . . .
Nd (
. . . . . . . . .. g . . . . .
0
;<e 00 t00t04.= ^
^
00 00 0'3 0 04'3 01'3 00'1 001 031 00'3 07'2
)NI( LVL PMOC RZRP CR 857L}E oga 0$a 071 03s ot 08 004 0s4 07s ees 00a
)F( i HC,PMET RW GLH 2 POOL CR 287T0 8 S 012 071 031 09 05 073 033 002 002 004 01 4 g$(% )F( PMET ENIL EGRUS RZRP CR 577T A 09 OS 4
031 014 ,. ( 033 03 002 012 071 004 073
)F( 1-5iCR,PMET RZRP CR 677TO pu s5 o5
=- :
bG l 1
.' : 5
!_ a)s r
l' .
. '0 G0 M2w
- o. g 0 Zo i e rt c m $ 4Gl
- r. g)3, ' 2RS ai a im g m '
S D p is gr a . p i ' eee 4 3mo ' cmd 4-L o L aao d o S o ' o pNM Mc c S y 7 4 o rra ' ea e 1 t T u. Nou. n ' kecl Acc aC a sl ap RtL a nE e ' T PH rL iire S W. t s8 s ' DFTR E t G r ' SO( - iD( f hhhh
' cccc
_ nnnn iiii
- \ '
s -
\ '-
~\ ' 1111 N < '
~
EEEE 4390 9188 "s '
' 9813 s .
\ 1717,
-\s
,\ 's e 6249
- - t + - ++
1
*:\ y .
i
-n t
5678 0000 g HHHH CCCC i EFGH
\s x- \s,
.' ..\
' i
'l,; eu / la h ? l M ge (; -
i
' s_ 1
- o. i.
BBC,D
" i ,'
e. i ., ! l i
- i. - 0000
' . ~
~
0000
" ._ m 2300 0900
+ - ++
i 2266 0011 xxxx
-g hhhh cccc nnnn iiii
~
(B g 0 B 12
++++ B 00ff EEEE . HHff CCOO 2222 B ++ -
ABCD 5
b s
.a0 00 00 . . . .
l
)(
00 .i, _ . . . . . . . . . . . . . . . . . ..g l . a
. )( . . . . . . . . . .
g 00 as as.. . o. . . . . . . . . . . . . . .
. )( . . . . . . . . . . . .
o . . . . .
, I..
00 ted et- . .
"n.t w . . . . . . . . . . . . . . . . . m o 8 n; 9 c
..g g ot ne a.p M . . . . . . . . . . . . . . .
Z ne M 3g . o - g 00 :00 21- . . . . . . . . . . . . . . . . .
,{ m c O
C Z ,H,, d
>< 0 00 2 20= . > . . . . . . . . . . . . . . ,,
m
~ ni
< U p"
g . . - I. . . . . . . . . . . . . . . . c
, .g 00 21:70- . . . - . . .
w
?, . . . . . .. .
00 0 . . . . . . . . . . . . . . . .
.g
.Q.
00.a mO- . . . . . . . . . . . . . . . . .p
)( . .. .
g g ,, - - 03 0 09 081 051 1 021 OS 003 072 042 012
)NI( LVLPMOCR2hPCR867L3U 007 058 000 055 005 054 l 004 053 003 052 002 8 8 )F( B HC,PMET RW GLH 2 POOL CR 287TO H
l' , . . . g ; 007 0sa 00e ous 00s 0s4 m 0s3 003 0s2 0c2 ' gD% )F( PMET ENIL EGRUS AZRP CR 577T A
= = = = = = = = = =
1;;3,{,g007
!N N )F( 2-51CR PMET RZRP CR 777T 0
[ eD. M 50
a r
. C. n
- O H
>.2-Jg 2 - .t c-r Gr-o
)$ $ 'h 2;,;s-
.i - l _
$ ;; \ /
8 4
- h. . .h. -
3 5 0 4 a f= 1.
.u.ag** 2-
~
n b 5
-- $- 15 'i ---
- s. ,
,,\r _
d of o . g w F3 g\ N \ 3 6
~
G 6 c. 4 E-
- .s 3 0 (k 0 a / I
- g
- 71/'/
* \E s-i-
~/ T\/ <
\. O E 6 en h e D E S $
{ f 0m
- 3 E := u D CC -
Z h, $_ Y k .KQ "$.- -h { a
, k 7 \\M \
7 \a Y E i E" b 1 ; 3 t2.i.. 3 g w g \s ~
/ 1
.e ! :-
E. - \ b^4 - O $ \b C D w L \, - o s b $ 4
.nx j ' 1 s
S o y b o r, z ., s .-
\ s-,A \. h '-
\/*
SI
g, t n e' O h f s he f'-!
$af2 r
'e se g "w w g i-0
, ., 2
" \ /
e d.i,a, . g A M n% 2 s
' S 1
. 2- a sE J6 %
- l S
g i l' I $g - s ,
\ ;n~ \* i iE 1i- ? N N e
" 1
, s
')
n e 5 = j ,$
% i s' .
;\ x . ( ,2 s-44:( *S V\" i !2 E
s +' J\/ 7.ef NA \ g j t, -
" Nb g 5 N
- o
* $ t
- f v
*c> /i -
5 3"
$ Y/W \)- '
11 g a 3 / N-c w w 3 e *
, m N 3 1 9 d
{ en c e a
, a , e 0$
3 v E T T
- C 2 g. I
' k S g ,-
3
; [-
D' s Y
\,f
/
s IV"
- t i
- . y L
l
/ \/ / p t 1 o'
5Z !
t "l o' t0 0 1 P'
\ /
w g O u Lh N - IT 4
/ .\'
5
!i V \
.ho
, / \1 - '
if i) / .
/
zs .', - (' . ( e
; e 2 I n
\
w
- og =
. O n .A .
Y< - - t - 4 f F *:. 7 4
- E H ' /
". 4-y
=/q -- -
".. .../_.
- g. g.
- RB -
i
/
r .c e oo 2
\\
~1 a y:!
8 .
- =
a a v. < a
~
yw e f-5 l 1 g d. , o er 1 1
}w 30 C
vXAA\i l
* /
1 9 j 4 L f g
\/
9 h 2e [ si Jf L
,k' s '.
0 2 4 . F
\
/
.e, 4
V
.c e Q .
_ . h. .
, w i i b i
~ * % 0.
! I: \, -- ; - _ _ s.
. e. o3 : __.._
A : H .N \ e
.z < 'x E 6 z 3-
% Y
\ yb \ s I
I e ~
/
)
l l <l / ! 3 2 g 8
/.
/\:'.7/
/:
x m t
. \
i : 8
\.
< u.
. I i \" /
j//\V
/ - \
/
\
\
/
\ %/\/ '
[ N/wy s, z( / / v - ys-
.p 5+
g_._- _ _ _ _ _ _ _ .. - _____________- __- _____-_ __________ _ _ ____ _ __ _ _ ______
N *. 1 o" E 6 e H j. l
/
\
(~ ~
; -, ~
7 1
- , N- ,.
\3
! . \ * 'e t- 0 K* 2%
! \ \
') I
/ 5 N
<.c e 4 5 :
N} V /; $ \ E
,e y v gV[g: c e
5
- ?
o
\;
h . .
- g. _ ..
i d/ N ', l \ s
? \
5 \ g * /
\;/'/ /
-i
! x N/\
72 ~ h
,/A/ ;.
s u 3 s jN P 3 \ . t i \/ o SS
7x ) 1 ; l-I' i
+ l ,
l
.. j
'g l b
I I 1 I O O h C
\- p
--u - -- - - .
Og s ' I . kt L / 1" F
/
/ \ L r
s
-fi cf !
l ? j iY ! $$ t ,
- /
( 1 .k
/ ::s 9
,0/'y;- '
/
~
If w d / 4 -5 8 5 $ s's y 2. c
- k /7
~
/ /
3 T
- % l o *
- e w g /
(; k 1 i to e a 3 m d o2
.j- t ba
=
r e 7 \ if
!I I -/ -
g ,E -s . 4 y
?' ~n h- $$
Z n
\I - . -
E e \ N
% \ 4 E A x 3 f ~L 4 / . v5 X a r t \mty __
g %A" 1 . / s 54
-.)
6 1 (, 4
*i e
k g > 1 lP l \ / I ) I : I o-I f i 9 . i 9 I'
! Y-g l
J
% ! $ a k 3 "
7
, lt y
\
t N / \s 3 !
>i .
/ ; --
- -e '; ;
.\ L -
/ e \
- I
/ J l
j 1, s -
\\ -
1 /v / \% ; 5 l
., / /\\( \
/
\ * !
C 8
.l
$ a -l 5 j ao .
tf
a t \ "
/
g
- a ij /\V / / ,
e - - F r/ u <
. - l
.} u s !
- x i 8 ,E o
J b ( 5 I 3 a
- - \ - /\*r .
/ ..
4 1 0
; k/ \ /
Al
- e .. , u
$ e I \
- v.
', N
- ~
3 s 0 .a ir _
// /, s fI <
K. e . o u s < tg 7
\/- g
)
t. 57 ( __ _ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ - . 1
.f l .
l2w lZo M uGl FRS
. S eee cmd
--. . ~ N aao pNM
< S y
.. 0 .
ea 1
- T .
kecl
... C' . sl ap N .,. i ire O M ,
DFTR I \ . T p C _*.- t . hhhh cccc l E !. A . R t
.?. W . nnnn iiii O .. P ,
..- O 9 -
. 2011 S
l l .
~
E . - + - - L T . EEEE
- A 7526 E .
C , 3216 N ..- L i 5470 t .A N s' . - O . 6109 S 4 N . s , , , . c t e e 7165 t - +++ T N e.
... N o
. u
.. n l234 I t i080 A r Rs t
T c h HHHH CCCC T s S o re R E i I
, EFGH Rt .. D u ...
T a -
- E A r .- N .
o T v - A N -
. G ,
o t E . T N C ~ E . P L ~; : . ABCD E E D
. :.- T 0090 O .
. H 5000 2000 0000
++ ++
8411 s~ ,- 0000 I xx xx
~ . hhhh
.~ 9 cccc nnnn iiii l
ifff Hfff COOO ABCD
%4 l il,
'.p
. ATTACHMENT 2 B & V REPORT SUBMITTAL IN RESPONSE TO NUCLEAR REGULATORY COMMISSION BULLETIN 88-11
" PRESSURIZER SURGE LINE THERMAL STRATIFICATION"
BAW-2085 MAY 1989 atemWFEgimaggEEEErFsMNE5FBBBISHIMbMHama 1
"AOWNERS GROUP MATERIALS. COMMITTEE
. d,as s 'Y '
SUBMITTAL IN RESPONSE TO l NUCLEAR REGULATORY COMMISSION BULLETIN 88-11
" PRESSURIZER SURGE LINE THERMAL STRATIFICATION" s'_ _ ___ _ ._ . - -- - ~ i l
_'c Babcock & Wilcox
, a McDermott company
BAW-2085 l- May 1989 SUBMITTAL IN RESPONSE TO NUCLEAR REGULATORY COMMISSION BULLETIN 88-11
" PRESSURIZER SURGE LINE THERMAL STRATIFICATION" I
Prepared for Arkansas Power & Light Company Duke Power Company Florida Power Corporation General Public Utilities Nuclear Sacramento Municipal Utility District Toledo Edison Company ; l l (See section 10 for document signatures) Prepared by R. J. Gurdal i J. R. Gloudemans The Babcock & Wilcox Company Nuclear Power Division P. O. Box 10935 Lynchburg, Virginia 24506-0935 l f
.1
L
+
i CONTENTS Page
- 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
- 2. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-l'
- 3. REGULATORY REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . 3-1
- 4. B&W OWNERS GROUP THERMAL STRATIFICATION PROGRAM ... . . . . . . . 4-1
- 5. FATIGUE ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 5.1. Oconee Unit 1 Bounding Fatigue Analysis . . . . . . . . . . 5-2 5.2. Davis-Besse Bounding Fatigue Analysis ...........5-4 5.3. Comparison of Analysis Assumptions to Oconee Test Data . . . 5-6
- 6. COMPARISON OF . Pl ANT SUP.GELINES . . . . . . . . . . . . . . . . . . 6-1 16.1. Dimensions, Configuration, and Thermal-Hydraulics . . . . . 6-1 6.2. Operating Procedures . . . . . . . . . . . . . . . . . . . . 6-5
- 7. THERMAL STRIPING . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 7.1. Definition . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 7.2. B&W Owners Group Program . . . . . . . . . . . . . . . . . . 7-1 7.2.1. Evaluation of Surgeline Hydraulic Mechanisms . . . . 7-2 7.2.2. Surgeline Pipe Wall Heat Transfer Analysis . . . . . 7-3 7.2.3. Evaluation of Oconee Field Data . . . . . . . . . . 7-3 7.2.4. Assessment of Available Industry Data . . . . . . . 7-4 7.2.5. Structural Evaluation of Surgeline for Thermal Striping . . . . . . . . . . . . . . . . . . 7-4 7.3. Preliminary Results . . . . . . . . . . . . . . . . . . . . 7-4 7.3.1. Assessment of Available Industry Data . . . . . . . 7-6 7.3.2 Evaluation of Oconee Test Data . . . . . . . . . . . 7-16 7.3.3 Structural Evaluation of Thermal Striping Based on Oconee-1 Measured Data . . . . . . . . . . . . . . . 7-17
- 8.
SUMMARY
AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . 8-1
- 9. REFERENCES . ... . . . . . . . . . . . . . . . . . . . . . . . . . 9-1
- 10. DOCUMENT SIGNATURES . . . . . . . . . . . . . . . . . . . . . . . 10-1
- ii -
Contents (Cont'd) l Page APPENDIXES A. Oconee Unit 1 Surge Line Measurement Program . . . . . . . . . . . . A-1 B. Verification of the Bounding Fatigue Analyses by Using Oconee Unit 1 Temperature Measurements . . . . . . . . . . . . . . . . . . B-1 C. Justification for Use of Cyclic Strain-Hardened Yield Strength . .. . C-1 List of Tables Table 5-1. B&WOG Plant Heatups and Cooldowns . . . . . . . . . . . . . . . 5-7 5-2. Top to Bottom Temperature Differences . . . . . . . . . . . . . 5-7 6-1. Surgeline Dimensions . . . . . . . . . . . . . . . . . . . . . 6-7 6-2. Insul ation Compari son . . . . . . . . . . . . . . . . . . . . . 6-8 6-3. Surgeline Hydraulic Conditions . . . . . . . . . . . ... . . . 6-8 6-4. Limiting Loop to Pressurizer Temperatures for A Representative B&W-Designed Plant . . . . . . . . . . . . . . . . . . . . . . 6-9 6-5. Heatup Procedures Affectina the Surgeline - Davis-Besse Unit 1 . 6-10 6-6. Heatup Procedures Affecting the Surgeline - Rancho Seco Unit 1 . 6-11 6-7. Heatup Procedures Affecting the Surgeline - Crystal River Unit 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-12 6-8. Heatup Procedures Affecting the Surgeline - Oconee Nuclear Uni ts 1, 2, and 3 . . . . . . . . . . . . . . . . . . . . . . . 6- 13 6-9. Heatup Procedures Affecting the Surgeline - Arkansas Nuclear One Unit 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 6-14 7-1. HDR Test Series TEMR-PWR: Ranges of Conditions and Conditions of Test 33.19 . . . . . . . . . . . . . . . . . . . . . . . . . 7-19 7-2. Striping Cases and Results . . . . . . . . . . . . . . . . . . 7-20 B-1. Top to Bottom Temperature Differences (Temperatures in F) . . . . B-2 List of Fiaures Figure 6-1. Oconee 1 PZR Surgeline TC Locations . . . . . . . . . . . . . . 6-15 6-2. Surgeline - Toledo Edison Davis-Besse Unit 1 . . . . . . . . . 6-16 6-3. Oconee Unit 1 Surgeline Stratification Data . . . . . . . . . . 6-17 6-4. Oconee Unit 1 Surgeline Stratification Data . . . . . . . . . . 6-18 7-1. Frequency of Occurrence Versus Amplitude . . . . . . . . . . . 7-21 7-2. Frequency of Occurrence Versus Amplitude of Near-Wall Fluid Temperature Fluctuations . . . . . . . . . . . . . . . . . . . 7-22 7-3. Buoyant Effects in the Pressurizer Surgeline . . . . . . . . . 7-23
- iii -
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ = _ _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ - . >
w 1 Fiaures (Cont'd) Figure- Page 7-4. Oconee Surgeline Stratification at Location 11 (Day 1 of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7 7-5. Oconee Surgeline Stratification at Location 11 (Day 2 of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7-25 7-6. Oconee Surgeline Stratification at Location 11' . H (Day 3 of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7-26 ' 7-7. Oconee Surgeline Stratification at Location 11 (Day 4 of ' Heatup) . . . . . . . . . . . . . . . . . . . . . . . . 7-27 7-8. Oconee Surgeline Stratification at Location 11 (Day 5. of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7-28 7-9. Oconee Surgeline Stratification at Location 11 (Day 6 of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7-29 7-10. Oconee Surgeline Stratification at Location 11 (Day 7 of Heatup) . . . . . . . . . . . . . . . . . . . . . . . 7-30 7-11. Duration Above Different Magnitudes of Stratification During Heatup . .. . . . . . . . . . . . . . . . . . . . . . . ._. 7-31 C-1. Strain History Beyond Yield . . . . . . . . . . . . . . . . . . . C-8 1
- iv -
i _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ d
i i
- 1. INTRODUCTION The , purpose of this report is twofold: first, to describe the - B&W Owners Group program and plans for addressing the surgeline thermal stratification and thermal striping issue, and second, to present the results of the preliminary work done to justify continued operation until the final program results are available. A portion of the Owners Group plan was presented to the Nuclear Regulatory Commission staff during a Regulatory Response Group meeting on September 29, 1988. A more detailed description of the Owners i Group plan, including preliminary observations from the Oconee test program, was presented to the staff on April 7, 1989.
The. report first provides the background of the thermal stratification and striping issue and describes the current regulatory requirements. The B&W Owners Group program is then described and the results obtained to-date are presented. These include the results of the bounding fatigue analyses, preliminary results from the measurement program at Oconee Unit 1, 'and a comparison of the configuration and dimensional characteristics of the various B&W plants. The latter is used to justify the use of 0 conte Unit I as the plant on which measurements are taken. Together these elements of the program are used to justify continued near term operation. The thermal striping program currently- underway is then described along with preliminary results. Upon completion of the striping program, a final report will be prepared which will provide justification for operation for the remaining life of the plants or an action plan that will lead to this ultimate goal. I 1-1 1 .__ ___ _____-_-_ ____- --_ _ -____-__- _____- _ __ _ ____ - _
- 2. BACKGROUND During heatup of a pressurized water reactor, the pressurizer is heated until a steam bubble is formed. The resulting pressurizer fluid temperature is thus significantly higher than the average fluid temperature in the reactor coolant system. Temperature differences along the pipe axis and vertical cross section of horizontal runs can also be established due to natural or forced convection between the two fluid volumes. At low velocities, the hotter fluid can flow along the upper portion of the pipe leading to strati-fication along all or a portion of the surgeline length. In addition, because the pipe connecting .the two parts of the system is also at an elevated temperature it can develop radial temperature gradients simply due to heat losses.
An associated phenomenon .which may occur during thermal stratification is called thermal striping. Since the warmer water is flowing across the cooler water, possibly creating interfacial waves and turbulent effects, a moving temperature interface may exist at the boundary between the two layers. These phenomena can alternately warm and cool the metal where they contact the inner surface of the pipe. The amount of alternate warming and cooling of the metal is dependent on the amplitude and frequency of the fluctuations as well as on the temperature difference between the hot and cold layers and the effective heat transfer coefficient to the pipe wall. In the extreme case, metal fatigue may result from this alternate warming and cooling of the pipe. The most obvious effect of thermal stratification can be substantial bowing, either up or down, of the surgeline due to the vertical thermal gradient in the pipe. This resultant bowing and possible contact with adjacent struc-tures was not considered in the original stress analysis since stratification was not an identified design basis condition at the time of the original stress analysis. 2-1
The effects of this bowing have been observed in the surgeline of the Portland General Electric Company Trojan plant during each refueling outage since 1982. The piping was observed to have lessened gaps on pipe whip restraints and, in some cases, actually contacted the restraints. Similar effects were also noted at Beaver Valley Unit 2. Both plants are Westing- , house PWRs. During hot functional testing, the surgeline of the B&W plant at Muelheim-Kaerlich in West Germany was instrumented and temperature readings were taken during startup testing. The Muelheim-Kaerlich plant is different from the 4 domestic B&W plants in terms of power level, surgeline layout, diameter and thickness. The measurements taken indicated that stratification, which was not part of the plant design basis or analysis, did occur in the surgeline. It was determined that this stratification was greatest during startup from cold conditions. Subsequent fatigue evaluations of this condition have shown that the Muelheim-Kaerlich su.seline meets its design goal of a forty year life. Based upon the Muelheim-Kaerlich observations, the B&WOG defined a program to instrument one of the domestic B&W plants to determine if stratification was present and, if so, to determine its magnitude. During the latter portion of 1988, based upon information related to surgeline motions, the NRC began preparing a bulletin requiring further investigation. Each of the PWR owners groups were requested to meet with the NRC to discuss their knowledge of the surgeline concerns and to provide feedback on the content and schedule of the proposed bulletin. As a result of this meeting the B&WOG expanded the scope of their program to include measurements of surgeline movement as well as thermal striping. 2-2 4
- 3. REGULATORY REQUIREMENTS On December 20, 1988 the Nuclear Regulatory Commission issued NRC Bulletin Number 88-11, Pressurizer Surceline Thermal Stratification. This bulletin requires certain actions of licensees of all operating PWRs. The applicable actions are paraphrased below.
- 1. At the first available cold shutdown after receipt of the bulletin, and which exceeds seven days, conduct a visual inspection of the pressurizer surgeline.
- 2. Within four months of receipt of the bulletin, licensees of plants in operation over ten years are requested to demonstrate that the pressurizer surgeline meets the applicable design codes- and other FSAR and regulatory commitments for the licensed life of the plant, considering thermal stratification and thermal striping in the fatigue and stress evaluations. (For licensees of plants which have been in operation less than ten years, this action must be completed within one year of receipt of the bulletin.)
- 3. Update the fatigue and stress analyses to ensure compliance with the applicable Code requirements.
If the above schedule could not be met, licensees were required to submit an alternate schedule within 60 days with justification of the new dates. This was done by letter from the B&WOG Materials Committee Chairman on February 24, 1989. This letter stated that the thermal striping portion of the program would extend beyond the dates requested in the bulletin for item 1.b, and would be forwarded by October 31, 1989. The results of the other parts of the program and preliminary results of the thermal striping evaluation are contained herein. IFor fatigue analysis the latest ASME Section III requirements incor-porating high cycle fatigue. 3-1
- ~-
- 4. B&W OWNERS GROUP THERMAL STRATIFICATION PROGRAM The B&WOG program currently in place to resolve the pressurizer surgeline thermal stratification issue consists of the following four parts.
- 1. Bounding fatigue analyses of the two types of B&W plants operating domestically.
- 2. Measurement of surgeline temperatures and movements during plant heatup, power operation, and cooldown at Oconee Unit 1.
- 3. Comparison of pl ant configurations, operating procedures, and specific plant and operator practices.
- 4. Thermal striping evaluation.
The bounding fatigue analyses were performed on two operating plants: Oconee Unit I and Davis-Besse. The surgelines on all domestic B&W operating plants are similar in configuration, geometry, and materials of construction with , the exception of Davis-Besse. All plants have the same surgeline diameter, thickness, and materials. The pipe routings are all similar with the excep-tion of Davis-Besse. The Davis-Besse plant has a nozzle supported reactor vessel and raised reactor coolant loops while the other domestic plants have skirt supported reactor vessels and lowered reactor coolant loops. This results in a different surgeline routing at Davis-Besse, therefore two separate fatigue analyses were required. A measurement program was initiated at Oconee Unit 1 to determine the surgeline temperatures and motions during plant heatup, and full power operation, plant upsets, and during cooldown. The results of this program for plant heatup and power operation are used to confirm that the fatigue analyses are indeed bounding and to permit definition of realistic surgeline transients for use in updated fatigue analyses. To date, no Oconee Unit I upsets or complete cooldowns have occurred. Therefore, no data for these 4-1
' types of events is available to integrate into this evaluation. The ap-placability of Oconee Unit I data to the other operating domestic plants from an' operational standpoint is discussed ' below. The applicability from a configuration standpoint is discussed in Section 6.0. The applicability frem a structural standpoint is addressed by the use of two fatigue analyses discussed above.
The plant operating procedures at Oconee Unit I have been compared to those of the other B&W domestic plants. Since the magnitude of thermal stratifica-tion in the surgeline is dependent . on the plant heatup procedures, the
-similarity of procedures ensures that the measurements taken at Oconee Unit I are representative of all domestic B&W plants. The evaluation of operating
. procedures is discussed in Section 6.2.
The effects of thermal striping on the fatigue life of the surgeline are being evaluated in a longer term effort. Preliminary results of this work are reported in this submittal. The measurement program at Oconee Unit I was designed to determine the magnitude of thermal stratification in the surge - line and the plant parameters that affect the stratification. Since the temperature instrumentation was mounted on the outside surface of the
;surgeline (a 1" thick pipe), it has inherent limitations for the determina-tion of inside wall temperature oscillations. Detailed heat transfer
- f. analyses of the surgeline wall have shown that at the relatively high frequencies associated with thermal striping (i.e. approximately 0.1 to 10 Hz) the outside mounted thermocouple will generally not detect the inside surface temperature changes and, in fact, striping phenomena have not been observed with the Oconee instrumentation.
A detailed description of each of the four parts of the program is contained in the following sections of, and appendices to, this report. Also included are the results of these efforts "to date. 4-2
- 5. -FATIGUE ANALYSES ,
i The bounding fatigue analyses were performed on two operating plants, Oconee Unit I and Davis-Besse. B&W performed all aspects of the Oconee analysis on behalf of the B&W Owners Group for the purposes of providing a generic result representative of all the lowered loop plants; Toledo Edison supplied key assumptions and loading inputs to B&W for the fatigue analysis of Davis-Besse. Toledo Edison's surgeline program, described in additional detail in Section 5.2, preceded the B&W Owners Group program and was initiated indepen-dent of the owners group efforts. As a result, there are some differences between the Oconee and Davis-Besse analyses in regard to assumptions and the application of the Muelheim-Kaerlich data to the structural evaluations done on the surgelines. These differences are minor and do not affect the con-clusions regarding unit operating lifetime. The Oconee analysis results are described in Subsection 5.1. Toledo Edison's program and results are discussed in Subsection 5 2. A comparison of the bounding assumptions for both the Oconee and Davis-Besse analyses to the Oconee test data is included in Subsection 5.3. The two thermal stratification analyses have been performed using the following codes:
- 1. For Davis-Besse Unit 1: USA Standard B31.7, 1969 Edition, " Nuclear Power Piping."
- 2. For Oconee Unit 1: ASME Code Section III 1977 Edition, with Addenda Through Summer 1979.
These codes were chosen since they had been used for the previous surgeline analyses. NRC Bulletin 88-11 requests the use of the latest ASME Section III Code incorporating high cycle fatigue (10 6 to 1011 cycles). The latest ASME Section III requirements are less restrictive than the ones used herein except for the fatigue requirements. The latest ASME Section III, Figure I-9.2 fatigue requirements are extended up to 1011 cycles for low stress 5-1
t values. However, 'the new requirements do not effect surgeline cumulative usage factor with thermal stratification, since the number of stratification cycles is very low compared with the cut-off value of 106 cycles (low cycle fatigue). Migh ' cycle fatigue is being analyzed in the thermal striping evaluation. 5.1. Oconee Unit 1 Boundino Fatiaue Analysis The Oconee Unit I surgeline geometry is shown in Figure 6.1. This geometry is typical of the B&W lowered loop plants with the exception of Davis-Besse. The surgeline was modeled on the ANSYS finite element computer code using piping elements which allow a linear temperature- gradient to be applied across the pipe diameter. The loadings consisted of pressure, seismic, deadweight, and thermal expan-sion from the original stress report combined with new thermal stratification loadings. The . thermal loading cases used are shown below. These tempera-tures were derived from those measured at Muelheim-Kaerlich. Surgeline data measured at Oconee indicates much smaller top to bottom temperature diffe-rences than were assumed in the bounding analyses. A comparison of the loading case assumptions to the Oconee data is provided in Section 5.3. LOAD LOAD LOAD Case 1 Case 2 Case 3 (Pre-HeatuD) (Heatuo) (Cooldown) Pressurizer Temp. (OF) 451 579 432 Hot Leg Temp. (OF) 109 369 93 Horizontal Run Top Temp. (DF) 439 531 399 Horizontal Run Bottom Temp. (OF) 109 109 93 Delta T Between Top and Bottom (OF) 330 422 306 Thermal stratification was assumed to occur over the entire lower horizontal pipe run. The pipe at the pressurizer end of the line was assumed to be at
- the pressurizer temperature while the pipe at the hot leg end of the line was assumed to be at the hot leg temperature. Load Case I was assumed to occur three. times during each heatup-cooldown cycle while Load Cases 2 and 3 were assumed to occur once per heatup-cooldown cycle. The end motions of the 5-2
surgeline at the hotL leg and at the pressurizer were calculated by applying RCS loop temperatures to the model. The resultant cyclic loads at each joint were calculated and applied to a T3 PIPE model to calculate stresses and determine fatigue usage factors. The T3 PIPE computer code calculates pipe stresses and fatigue usage factors using ASME Code methods' for Class I piping. The appropriate stress indices are automatically included for the selected ASME Code dates. For the case in point, the 1977 Edition with Addenda through the Summer of 1979 were used. All Code criteria were met with the exception of the requirement that the expansion stresses not exceed three times the design stress intensity (3Sm) which, for austenitics, is equal to two times the material yield strength. This is ' intended to prevent the material from being cycled in the plastic ' range by thermal expansion. The 3Sm limit assumes elastic-perfectly plastic material behavior when, in fact, most steels used in nuclear power plants exhibit considerable strain hardening. Therefore the strain hardened yield strength was substituted for the virgin yield strength for purposes of this preliminary analysis. This meets the Code intent to prevent cycling in the~ plastic range. Refer to Appendix C for the technical justification for the use of twice the cyclically strain-hardened yield strength in place of the 3Sm limit specified in Section III of ASME Boiler and Pressure Vessel Code.
'It is expected that the final anaiy:is will meet the more conservative 3Sm
~
requirement of the ASME Code. The fatigue usage factors were calculated for the surgeline, surgeline drain - nozzle, hot leg nozzle, and pressurizer nozzle. The usage factors for the thermal stratification load cases were combined with those from the stress analysis of record to obtain the total usage factors. These include thermal stratification effects during all heatup-cooldown cycles, ir.cluding those which occurred in the past. This was done using the specified number of 'heatup-cooldown cycles of 360 for the 40 year life of the plant or nine heatup-cooldown cycles per year. From the operating experience of these i plants, the nine heatup-cooldown cycles per year is a conservative number. The fatigue life of each part affected by surgeline stratification was then calculated in terms of allowable number of heatup-cooldown cycles and is presented below: 5-3
Hot leg nozzle (carbon steel portion) . . . . . . . . . . . . 270 cycles Hot leg nozzle (stainless steel portion) ......... 162 cycles Surgeline (straight or elbow) . . . . . . . . . . . . . . . . . . . . 153 cycles Surgeline drai n . nozzl e. . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 cycl es Pressurizer nozzle (stainless steel portion) ..... 341 cycles Pressurizer nozzle (carbon steel portion) ........ 396 cycles The B&W domestic unit which has the most heatup-cooldown cycles to-date is Oconee Unit 2 with approximately 96 (see Table 5-1). Thus, it can withstand another 39 cycles of heatup-cooldown without fatiguing to its limit the surgeline drain nozzle, the most limiting case, using the conservative analysis described above. This translates into five more years of operation using the specified heatup-cooldown cycle accumulation rate of 360 per 40 ,
-years which is, in itself, conservative.
5.2. Davis-Besse Boundino Faticue Analysis Toledo Edison initially became aware of the NRC's surgeline thermal strati-fication and striping concerns in September 1988 while Davis-Besse was in cold shutdown for a refueling outage. A program was immediately developed and implemented to assess the condition of the surgeline and to verify that the unit could be safely returned to power. The orogram included a broad spectrum of inspections, maintenance reviews, and analyses. The analyses were aimed at determining the remaining useful life of the surgeline. The results of this evaluation are reported in this section. In order to define temperature transients upon which fatigue analyses could be based, Toledo Edison reviewed the Davis-Besse operating procedures and the temperature stratification data from Muelheim-Kaerlich. Surge 11ne conditions were estimated from the Muelheim-Kaerlich (M-K) data. Plant-specific adjustments to the M-K data were made to account for differences between Muelheim-Kaerlich and Davis-Besse. The following text addresses the specific analysis assumptions and results of this work. The Davis-Besse surgeline geometry is shown in Figure 6.2. It can be seen that this geometry is different from other domestic B&W plants. The analysis was conducted in the same manner as the Oconee Unit 1 analysis. The thermal stratification stresses were combined with the stresses due to all other i 5-4
_- - ___ _ = _ _ _ _ - _ _ _ - _ _ _ - _ _ _ _ . - - _ _ _ - _ - _ _ _ _ _ _ _ _ . . specified transients and determined the total usage factors for the surgeline , and the nozzles at each end. The stress analysis loading consisted of pressure, seismic, deadweight, and thermal expansion loadings from the original stress report combined with new thermal stratification loadings. Davis-Besse's unique configuration lends I itself to different stratification conditions than Oconee 1. The 7.25 feet rise near the center of the surgeline will reduce transient thermal gradients that exist in either horizontal line depending upon the direction of flow. The thermal loading cases used are shown below. These temperatures used are derived from the temperatures measured on the surgeline at Muelheim-Kaerlich, modified to account for Davis-Besse operating limits. Steam Midway At End Bubble in of Forms Heatun Heatuo Cooldown Hot leg Temperature OF 100 375 500 100 Upr. Horiz. Run Temp. OF 100 375 500 100 l' Vertical Run Temperature OF 100 275 500 100 Lwr. Horiz. Run Temp. - Top 0F 409 506 506 409
- Btm 0F 100 120 205 100 Delta T Between Top & Bottom 0F 309 386 301 309 Pressurizer Temperature OF 409 506 649 409 Thermal stratification was assumed to occur over the full length of the lower l horizontal pipe run. Stratification transients were assumed to occur three times during the bubble formaticn of each heatup-cooldown cycle while the remaining cases occur once per heatup-cooldown cycle. The end motions of the surgeline at the hot leg and at the pressurizer were taken from the existing stress report and applied to the model. Impe11 Corporation performed the L deflection / stress analysis of the thermal stratification events using an ANSYS model similar to that used by B&W for Oconee Unit 1. The Davis-Besse surge line met the stress criteria of 3Sm limits of USA. Standard B31.7, l Subsection 1-705 Equation 12.
B&W performed the fatigue evaluation utilizing the output of the Impell analysis. The total fatigue usage factors were calculated by B&W for the i 5-5 l
i' surgeline, hot leg nozzle, and pressurizer nozzle. The usage factors due to thermal stratification during all heatup-cooldown cycles, including those whkn occurred in the past, were combined with those from the stress analysis of record to obtain the total usage factors including thermal stratification effects. The resulting fatigue usage factors for the total of 40 heatup and cooldown cycles projected through the end of current Fuel Cycle Six are: Hot leg nozzle (as a branch connection) . . . . . . . . . . . . . 0.619 , Hot leg nozzle (carbon steel portion) . . . . . . . . . . . . . . . 0.704 Hot lec nozzle (stainless steel portion) . . . . . . . . . . . . 0.343 Surgeline (straight or el bow) . . . . . . . . . . . . . . . . . . . . . . . 0.063 Pressurizer nozzle (stainless steel portion) ........ 0.297 , Pressurizer nozzle (carbon steel portion) ........... 0.634 The above results show that the limiting component for fatigue is the carbon steel portion of the hot leg surgeline nozzle. The 0.704 usage factor for this nozzle is based on 40 heatup-cooldown cycles. Hence, the nozzle (and other parts of the surgeline) can withstand 57 (i.e. 40/0.704) heatup-cooldown cycles without exceeding ASME Code criteria. With only 37 cycles accumulated by Davis-Bcm; to date, 20 additional cycles remain. These 20 < cycles will provide approximately seven additional years of operation at the rate of three cycles per year which Davis-Besse has been experiencing during the past 12 years. Even at the conservatively specified rate of six cycles per year, 31/2 years of additional operation are assured. 5.3. Comparison of Analysis Assumptions to Oconee Test Data Comparative fatigue evaluations have been performed for both Oconee Unit 1 ' and Davis-Besse, taking into account the temperature measurements from the February 1989 heatup of Oconee Unit 1. Table 5-2 gives an overview of the temperature differences assumed in the bounding fatigue analyses described in Subsections 5.1 and 5.2, compared to the ones measured during the February 1989 Oconee Unit I henc9. The fatigue results from the bounding analyses have bee F und to envelope the fatigue using the Oconee Unit 1 temperature measurements for the most critical locations. A description of the fatigue comparison is included in Appendix B of this Document. 5-6
n" , V Table 5-1. B&WOG Plant Heatuos and Cooldowns Number of Limiting ., , Pl ant Heatuos and Cooldowns Number of Heatuosi Arkansas Nuclear One 86 135 - Crystal River Three 29 135 Davis-Besse 37 57 o Oconee One 84 135 - Oconee Two 96 135 Oconee Three 66 135 Rancho Seco 35 135 IAs determined by this evaluation. ,, Table 5-2. Top to Bottom Temperature Differences '- (Temperatures in F) Poundino Fatioue Analyses Measurements at Oconee Unit ] Oconee Unit 1 p_ avis-Besse (February 1989) llEATUP: , 422 386 280 , 330 309 250 330 309 250 . 330 309 240 301 220 4 23 additional cycles with temperature differences rang-ing from 206F to 65F. 000LDOWN: 306 309 No full cooldowns have occur-red to date. 5-7 , o
- 6. COMPARISON OF PLANT SURGELINES The factors affecting surgeline performance have been evaluated to assess the potential for thermal stratification in the surgelines of B&W domestic plants as observed at Oconee Unit 1. The evaluation addressed two different types of factors: those that are inherent in the base design and the operating procedures that may influence the surgeline conditions. The following two subsections summarize these evaluations.
6.1. Dimensions. Configuration. and Thermal-Hydraulics As shown in Figures 6.1 and 6.2 and tabulated in Table 6.1, the domestic B&W plants employ two different surgeline configurations. On the Davis-Besse plant the surgeline has a vertical drop of 7'3" instead of 13' froin the hot leg connection elevation to the bottom of the surgeline. Davis-Besse's surgeline has a long horizontal run from the hot leg before turning downward to the low point of the surgeline. Hence, the overall run of pipe that constitutes the surgeline at Davis-Besse is essentially divided into two horizontal runs by the 7' vertical section. In the icwered loop plant configuration the single significant vertical run of pipe is very near the hot leg. The resulting short horizontal section entering the hot leg is only 21" iong. The surgeline for each configuration is 10" schedule 140 stainless steel pipe (inside diameter 8.75") with a wall thickness of 1*'. The surgeline is insulated, but not identically, at each plant. Table 6.2 summarizes some key insulation data for the plants. The number and type of surgeline supports and restraints varies from plant to plant. Davis-Besse has several pipe whip restraints with surgeline mount-ings. These differences influence the heat losses from the surgeline (because of interruptions or discontinuities in the insulation), and the 6-1
structural evaluations that must consider the effects of surgeline displace-J ments. , The surgeline hydraulic conditions are similar from plant to plant, but vary
~'"~
significantly depending on the pl ant's operating mode and evolutions or upsets in nrogress. Each B&W plant is controlled to approximately 2155 psig 1',. which requires a saturation temperature in the pressurizer of about 647 F. Hot leg temperatures at full power vary a few degrees from plant to plant, but are all between 600 and 605 F. Therefore, the typical pressurizer to hot leg temperature differential is about 50 F. During normal power operation, the surgeline is exposed to very small flow rates from the pressurizer to the ." hot leg. This fl ow, provided by the pressurizer spray bypass line, is approximately 1.5 gpm and serves to minimize thermal cycling on the spray j, " line and to promote chemical equilibrium in the pressurizer. Continuation of . this flow from the pressurizer into the surgeline provides a steady heat input for warming the lino. However, long transport times in the large line and heat' losses through the insulation result in establishment of an equili- 1 brium stratified condition in the absence of flow transients in the surge- ri line. The 1.5 gpm bypass spray flow has been utilized since plant startup . and is a generic value. Small deviations can exist from plant to plant because of the accuracy involved in setting the needle valve that controls this flow. The bypass flow rate is also a function of the running reactor coolant pump combination. If both pumps are running in the loop connected to the pressurizer, the bypass flow is at or near the nominal value. With either of thene pumps secured, the bypass flow is diminished. If neither pump in the pressurizer loop is running, the spray bypass flow may be near zero. The vast majority of plant operations involve running four pumps. Operation at power is not permitted with two pumps out of service in the same loop. The frequency and magnitude of upsets is similar for the lowered and raised loop plants. Depending on the sequence of events, insurges or outsurges may take place that impose moderate to high flow rates through the surgelines. Table 6.3 summarizes the range of anticipated flow conditions that may occur for both raised and lowered loop plants. The values shown for purposes of 6-2 ,
illustration are arbitrarily based upon surgeline conditions at hot, full , power. In the 1.5 gpm bypass flow condition the velocity through the line is quite low, but the fluid displaced from the pressurizer by the bypass flow provides - the heat source necessary to support long term stratification. Preliminary a results frors the Oconee test program confirm that stratification does occur $, in this mode of operation as well as in others where the surge flow rate is higher. The Oconee test data shows that during power operation the water leaving the pressurizer surge nozzle is approximately 590 to 600 F. This suggests that the water in the lower part of the pressurizer is below the saturation temperature even if some allowance is made to account for errors in the measurement. This is because the lower most pressurizer heaters are I about 52 inches (Ref.10 and 11) above the bottom cf the pressurizer.2 The ' upper part of the surgeline remains near this temperature in all horizontal sections of the line while the lower part of the surgeline may be signifi-cantly cooler depending on the plant conditions. Figure 6.3 displays a typical set of data at power for Oconee Unit 1. Top-to-bottom delta T is between 40 and 70 F in the lower horizontal piping sections. Figure 6.4 shows a typical top-to-bottom temperature profile at two horizontal sections of the se geline. These data suggest the temperature gradient in the surgeline is relatively linear during normal power operation. A sharp temperature gradient is not discernible. When an upset occurs that causes a large insurge or outsurge, the stratified conditions are swept out and the line becomes isothermal. This process imposes a thermal transient on the surgeline. The surgeline volume for the '
-lowered loop plants is about 20 ft3 (23 ft3 at Davis-Besse). A pressurizer level change of about 6 to 8 inches is sufficient to displace the surgeline fluid. Once steady state conditions are reestablished in the reactor coolant system, even if at a new operating condition, the surgeline will restratify 2F urther evidence that the pressurizer liquid is stratified during equilibrium conditions with only the 1.5 gpm bypass spray flow is the Oconee data taken at hot zero power with full spray (278 gpm) for an extended period. In this higher flow condition, the top of the pressurizer surge line reached about 6400F, very near saturat.an temperature.
6-3
and come to a new equilibrium condition assuming the bypass spray flow is operational. . The above information and mechanisms are applicable to each of the plants and the thermal conditions are expected to be quite similar. Davis-Besse's unique configuration lends itself to somewhat different overall stratifica- , tion conditions. The vertical rise near the center of the surgeline will reduce somewhat downstream transient thermal gradients resulting from a transient flow condition involving a temperature change in the flow field.
~
This is particularly true if the surgeline is near isothermal conditions when the surge transient occurs. This statement is based on tests done in a laboratnry environment with an inverted loop and water as the test medium (reference 1). The Oconee data also support the effectiveness of the vertical run in reducing transmission of stratification gradients. During ', quiescent periods with significant stratification in the horizontal runs, the ,' vertical rise at Oconee shows very small temperature differences between the three thermocouple located at the two measurement planes (refer to location
- 4 data on Figure 6.3). As a result, the short (21") horizontal run between the vertical section and the hot leg will experience a smaller degree of stratification than the lower horizontal section during transient outsurges.
During steady state conditions only the lower horizontal run at Oconee will tend to stratify. Davis-Besse is expected to demonstrate similar behavior. A primary difference between the Oconee and Davis-Besse configurations is that at Oconee the upper horizontal piping is quite short and appears entirely mixed by the effects of the hot leg flow. This eliminates stratifi-cation at Oconee as evidenced by the data. However, Davis-Besse's relatively long upper horizontal run is not expected to be as strongly influenced by hot leg flow. Some stratification should occur in this part of the surgeline, although to a lesser degree than in the lower horizontal run. The main conclusions at this point are that:
- 1. Davis-Besse's thermal stratification is expected to be of similar magnitude to that observed at Oconee.
- 2. Each of the lowered loop plant surgelines are nearly identical configurations and should have similar, if not identical, thermal-hydraulic conditions during normal power operation.
6-4
Related to the second conclusion, the similarity of conditions in the surgeline during shutdown operation is a function of the operational evolu-tions performed with the plant shutdown. The next subsection addresses this point. 6.2. Operatino Procedures Operational evolutions have a significant impact on the steady state and transient thermal conditions experienced in the surgeline. The Oconee data shows that evolutions which affect the inventory control in the RCS have the most influence. Pressurizer level changes are good indicators of transients in the surgeline. Since the operational evolutions are controlled by procedure, plant to plant procedural differences could have a strong bearing on the surgeline transients and conditions experienced during plant heatup and cooldown. Peak thermal stratification is expected during the initial pressurization of the reactor coolant system (plant heatup). All of the plants first esta-blish a steam bubble in the pressurizer and then increase system pressure by energizing the pressurizer heaters. The heaters have virtually no influence on the temperature of the rest of the system. Thus, the temperature dif-ference between the pressurizer and reactor coolant system i'hcreases as the system is pressurized. During these early parts of the heatup procedure, the surgeline conditions are determined by the pressurizer temperature control (which is manual), reactor coolant system inventory control (primarily by the makeup and letdown systems), and auxiliary spray from the decay heat removal system if it has been in service. With a moderate outsurge from the pres-surizer the surgeline stratification will be greater than it is during any other operating procedure. Surgeline stratification is determined by the following three factors:
- 1. Surge flow rate
- 2. Cooling of the surgeline (ambient losses)
- 3. Surgeline boundary conditions (hot leg and pressurizer temperatures)
The degree of stratification is dependent on the direction of the surge and its magnitude. Outsurges of the hotter pressurizer fluid result in greater 6-5
stratification than does an insurge. This was observed at Oconee. As discussed earlier, a large surge will flush the surgeline decreasing thermal stratification. Initial review of the Oconee data shows that the surgeline bottom temperatures are lower than the reactor coolant hot leg temperature by up to about 50 F when the plant is at power. This temperature difference depends on the quality of the installed insulation. As shown in Table 6.2, Oconee Unit l's surgeline insulation type and thickness is representative of the insulation installed at other B&W plants. There are operating restrictions that limit the maximum pressurizer to hot leg temperature differential . During heatup and cooldown, when the tempera-ture differences are largest, the reactor vessel pressure / temperature curves limit the pressure for low temperature reactor coolant system operation. Since the pressure is controlled by the pressurizer temperature, the maximum pressurizer temperature is indirectly limited by this limit on RC pressure. Table 6.4 provides representative values for these limits from one B&W-designed plant. As the plants age, the pressure limits may be lowered. Industry activities are attempting to relax these limits in order to allow higher pressure limits which would simplify operation of the plants. There are other operational limits that bear on the typical differences between pressurizer temperature and RC loop temperature. However, the P/T limits provide a representative bound. A preliminary comparison of operating plant procedures has been completed with the focal point being those evolutions encountered during the initial pressurization and heatup of the reactor coolant system. Each pl ant's controlling procedure for plant heatup from cold shutdown to hot shutdown was reviewed. . For each evolution that has a potential impact on the surgeline, the approximate coolant system pressure and temperature were estimated and the loop to pressurizer temperature differential was calculated as a gauge of the temperature extremes that the surgeline could experience at its end points. Before reactor coolant pumps are started, there is no pressurizer spray line pressure differential to cause normal spray flow. Tables 6.5 through 6.9 list the specific steps involved in the plant heatup for several stations. The tabulated values for coolant temperature and pressure are approximate. However, the Oconee data shows that the estimated 6-6
temperature differentials correlate well with the observed peaks in stratifi-cation. The evaluation shows that there are similarities in the plant evolutions for startup although some differences exist. Based on this evaluation, the plants should experience similar surgeline transients in regard to frequency and the magnitude of the temperature differences that might exist in the surgeline. This comparison shows that the units are operated similarly enough that gross differences should not exist from one plant to another. A detailed evalua-tion of the surgeline transients as they were noted during the Oconee measurement program will be included in the final evaluation of the struc-tural effects of surgeline stratification as part of the effort for Bulletin Item 1.d. This will include a more detailed comparison of plant !.pecific , procedures. Table 6-1. Surceline Dimensions Section Identifier Plant Reference A B C D E Oconee 1 2 21 13/32" 12'3" 25'7" 8'-3 11/32" 14 61/64" Oconee 2 3 21 13/32" 12'3" 25'7" 8'-3 11/32" 14 61/64" CR 3 4 21 13/32" 12'3" 26'1" 8' - 21/32" 14 61/64" ANO-1 5 21 13/32" 12'3" 25'7" 8' - 21/32" 14 15/16" Oconee 3 6 21 13/32" 12'3" 25'7" 8'-3 11/32" 14 61/64" Rancho Seco 7 21 13/32" 12'3" 26'1" 8'-3 11/32" 14 61/64" Plant Refer. A B C D E F Davis-Besse 1 8 3' 9/16" 21' 7'-2 13/16" 16'9' 4'-11 15/16" l'3" Note: 1. Refer to Figures 6.1 and 6.2 for the pipe section identifiers used in this table. 6-7
Table 6-2. Insulation Comparison Plant Installer Tvoe Thickness Oconee Units 1, 2, 3 Diamond Power Reflective 3 inches Arkansas Unit 1 Transco Reflective 3 inches e Crystal River Unit 3 Transco Reflective 3-1/2 inches Davis-Besse Unit 1 Diamond Power Reflective 3 inches , Rancho Seco Diamond Power Reflective 4 inches Table 6-3. Suraeline Hydraulic Condition 1 177FA Plants 8.75" ID Plant Condition Bypass Full Hild Spray Spra;,l Upset Flow,lbm/s 0.12 15.66 500 Velocity, ft/s <0.1 1.01 32.4 Reynolds No. 4E3 5.3E5 1.7E7 3C orresponds to 190 gpm. On the Oconee units this full spray flow rate is approximately 280 gpm. p NJ e 6-8 ,
Table 6-4. Limiting Loop to Pressurizer Temperatures for a Representative B&W-Desianed Plant Crystal River Unit 3 RC Temo RC Press Pzr Temp Pzr RC Temo 0F psig 0F OF 70 300 422 352 85 157 300 422 265 175 235 236 528 476 240 377 2250 654 277 378 Notes:
- 1. Data taken from reference 9 as tabulated on figures providing pressure /
temperature limits.
- 2. The above values of the temperature difference between the pressurizer and loop are upper bounds. Other plant limits and operating procedures provide even lower limits to this temperature difference. The above tabulation provides an easy way to demonstrate that there are practical limits on the magnitude of this temperature difference.
6-9
jli' 0001 6638 29903 2222222 4 1 1 1 7 11 1 9 4 363 071 677 333 7 7 8 r i
)
F 2 1 1 222222221 1 1 1 1 21 1 1 31 11 v E,
)
2 '( 4 e T v s t r eamL o n s u. . E e r r 4 rt <D
)
F 1 09 67 7 7 7 8 (' 0001 1 1 3822222227 P 000222 1 1 1 58 99 9 9 994 99 3333333333333 4 41 6666664 8 1 33 3 34 M E T E p R N t u r Z I P L ta S
- E t n )
la ig G P s R ( p 0005555000000005600 5G0005 26 70 20 0 005 U S 333777201 S 1 1 1 1 1 1 2 22222222236700001 1 S ty e E 1 22 222 E1 f a R t s P Hi t n a esC T n lpn cR U ra G 0
)
r ot s Ne0 I S ve h
/
F f dmqo T Se1 R 0 5 Y s c '( ee ) r ir F 000055 00000000 000 02 000000 5 6E - C B0 9 6 (Psl nFEnrelaP i l 00002255 1 1 1 1 1 5888 1 1 1 880588 1 1 1 1 51 000003 1 222 2 3 4 555 55 EF S - 0
/5 u o M m f L
B FiPvO FrFF
' o E T AAa- 3 f
T S DDB 8.8514 00 C 0324 R E O R T Ug U Ai s D d n ep c 6 E g in a la 6 d C S s is p 0 m e P2 t ep g O T 0" e n g 1
<d &w i es f
l mis e%pmVl o I 5 R MI L 8
~
i l S 7 is p otcB eb 5 T pV B s n P TE le M
~ r u 5 osT 8C 00 T io PS E v ne5 s ol y st eS e si n g f oR0 gO it P YU e d alp 2 d 2 int d U BY R l n nm1 it eemt on r it l osU n o
T WA U R D Z a eia Sat oe omr n Ptlt a al os e C, m ia imo h t elcn oA C R dlb m yr l H ivtoruy i s r A OP L S E P C /w nb o uot ru s it au ue h e st h yT Oe RP I n w E L AR O t yn cBt sis s uc caenncUn - t a F C R Vs epe ssb r o H CZ RP R nt r i me ePe r med aV r v d eieA' R ZP BTPt auaeP npsDn0 r t dt u
& R P es mat e ah ir sEGooP e i 8 d - h Vmu iF nCeH CS i
r S RF O dnh ec uSheCCupG t R ReqmSOiB SCsC T nVRt Vot 2 p hV2 h* eR H RRt o N Z 2 uy eat e5 Oe seZ aCvsahR uPth4H O PTL i ZP Zsaf eR u a r P 8 wtn T t ph 3 I :NE T UITED AME S k CcbbP l aa eet l le1 ai i r a eta erear n r ci ut >S ly an f t t i et y w etn c r r f CUWE E TB L Reh sss l l t t a ek CEEUTiVC n& ac5 . . W c. r n c4 DMhSS.VetaCocola c ey gd HSRSiCPSV l n a ire t
- E REE PATB.W A i 61 46677056777822071 F
1 4 4 56661 1 1 1 224 5 7257 4 8 455677789 R DRFO 33444 4 4 4 4 44 44 44 4 4444 4 4 4 4 4 U NPDLL I D STP.A AU 4 4 4 4 4 4 4 4 4 55555555 55555555 T: E N C A OiEEA T AMX L R P P L MtHTM
,4o l if( iJlll 1lfll1
a F nin
- - L
', 2 i i.2 ro E xs 3 w nei2. 3 D
err o ) n csr as p= F ( 0 7 76 2 2 2 2 2 22 23 1 9 5 0 4 2 3 8 8 8 8 P 0 7 7 6 9 9 9 9 99 9 9 9 2 5 6 7 1 3 3 4 4 M1 E 22 333333 3333 4 4 4 4 66 666 T E p R N t u r Z I a P L S t
)
E t n ty e g la G P fa is p R s ( S 33 003 50 1 01 10000 00 00 5 100 0 55 7 0000 25 5 505 55 5 5 t n U la S 1 1 1 1 1 1 2 22 2 2 2 2 2 2 3 4 4 5 7 9 9 11 1 S p r E 1 1 22 2 1 o R E 2 it B. f t P H nP p e C T U2 50 c e x R G o0 F N C D. 0 5
)
F I T eP 2
- S3 9 (
00 0 5 555 00 000 0 0 000 000 0 0 22
- 6. C o2 1
P0 0 1021 2 1 1 2 12 15 15 15690 1 1 2 2 0 3 4 6 0 3 1 1 1 2 2 2 2 2 4 4 4 4 5 5 5 3 6E h P M1 EF C W. h r E LF B n7 4 / FFF T AA a v. '0 '5 '0 C T S RR e 061 134 R E F
- X R g0 in5 R U lc<
1 ip D y cp g t er a T. E pm is r n C S e e p pa T c mT u um s O I M i v pS 0 0 a d n V t r R I L : TE ES g e r oR C 4 e e ABO P PS R ni n f H n H v h s TT U YU o d r / P Uit n S ia S s b e ewAO T U BY WAR Dme Et e v P t n P N V a B opehisue i r t a enU p8 T OP leNia T r eA mu Cy sd n t gd p A LS OS ayb ebd m n e / emd ur R a lv t n E L AR R ggsB ett r ud o it n dep u m w r p s uPfoZ P pV e saiCTP r V nemP CZ n e m esC s/ r H R P Piini lpl l mmneme uCryS e r R V w mye eat R U
& R ea et s B r a h i
O mf hCte h y sC h S ePRtsHT P dT sT pe . t Cd im C Ci h er eC SH u r S RF aC t S o S pR oR n t D r N Z SR o O PT R Rt H m et o e h eR Rt h PRRR e ee eCeZZ f zPP heT L I : T NE CteZ ZnD ut es t ch t t UITED RepyleP P at epaf P le s ir eCez s y r sh i at lt Ri t AMEE ceec l l s et tobb c mime jc i i t mr r i c ub et t CLWL f f r I ul r r r r E TB la oe o ncd y oc aar e et r r a c arh a t aaa t a l t
- REE A l
t t nSSVSSiSTSSPPSPV n t l E PABW T PCVFiACCI R DRFO U NPD I L 1 9 D AU T: E STP.L A 6566 1 2 3 4 4 5 4 5 7 8 6 6 7 8 3 N C TAM X 1 2 2 2333 36 78 1 1 2 2 2 2 3 3 3 3 4 A OIEEA L R IHTMM 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 33 3 3 3 3 P P L 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 T[
m c. m e _ e m L c o o. r-w E D _ a p o ot o
)
F ( 66 3 5 5 7 7 5 8 1 1 - 9 7 0 668 P 66 9 9 92 2 3 4 6 6 7 9 1 2 2 4 4 4 4 4 4 4 4 4 6 6 6 6 M33333 E . T E p R N I t u r Z P a L t S ) E t n i g la G P s p R ( U ty S 005000000 0000 0005 S 551 22 2 2 5 066 5 5 5 5 5 5 E 1 1 222333 4 4 4 5 66 8 8 1 e S fa . 1 1 1 2 s R E 3 t n P H it a lp C T n r R U fo G r d e N e ir u
)
F I T i v 9 . Y q P er ( C R .7v F E ss P0 0 005500 5 01 2 2 5 5 7 000 0 000 4 4 4 7 0 05 02 0 3
- 7. E 6 l aR e r u 8 e M1 1 1 1 1 1 1 1 2 2 2 2 34 4 5 5 EF t ,
h o ro ln u E L B F S y2 02 / F F F f T AA r -P '0 2 0 C T S CO 0 5 5 1 3 2 R E R _ r o n U t
) c w 2 a o D
g0 e d E i3 s - R t u _ S pP _ C T I M 5O g ip T r h s O _ 1 t I L 2 o is d o R T E E t ae t t g p n ac h f P P S f Si s 0 YU R ne 6 lyta i i r P B Y U iaR H p F4 F l a e n pe2D 0 um (
* ~
- v WA o t
U OP R D iaor 0 4 oS7 0 nt d Mlou 3 Ee a u n T L S Cb l c~ 2 t C2 mA pF a A L A R oreR o E C Z Ob d ee nnsi s d t uf t s n ' Ru ps op V isr m2 P e R B aod e m sel em r _ H & R P snnr m y eia u s iu q er pe B T e u3 P5 O TP t S RF Z a r r s psp t e isPP e m T / a we Co t N P LT R p O I : T TE D N E S eR t mCRCmreCRSSPuP Rfi RRyR CCSC aS V uRt pH h u t UI Rh o CrdS r e RtaZ uoH r a e A C ME I L WL E ZCe P k at s wDH yRiCeeth t s Raas s im aPh sOeP f E T B ee et pc et mcet er pir r nr r er s oa ar on t t t R TEEB A r f oh c aoreta e cch n c et E P A W t t n iCiTSPSC n l t R DRF.O L FCInSSVSVI U N P D L I D A U P.A 072 5 T E S 7 4 224 4 6 N C T 'A! MX 272 4 4 4 4 4 1 3 4 94 1 1 2 2 A O A 23333 3 3 3 4 4 4 4 5 5 5 5 5 L R MI HT E EM P P I L 555 5 5555 5 55 55 5 5 5 5 _ ,4e a
utt w umm.m raaa E D ,
)
F ( 5 5 5 5952 2 2 2 9 5 5 2 2 2 8 0 4 8 8 8 8 P 9 999 0 3 4 4 4 4 4 3 3 4 4 4 4 2 3 4 4 4 4 M 2222 34 4 4 4 4 4 4 4 4 4 4 4 5 56 6 66 r E h T E3 / F R Nd 0' 9 Z L I n - P a 0 ig E 8 S G2, 2 T ( p R1 S 5 5 5 5 0 055553 00 5 5 500 0 5 5 5 5 U S ty e S 4 4 4 4 6 5 7 7 7 7 0 5 5 33333 4 3 3 7 7 7 00 0 5 5 5 5 33 3 4 8 91 1 Si t fa E 2 22 n s R - E U t n a P H r lp C Ta f r o R Gle d e NC u I rPq Yiu r ) F T N1 8 C e2 0 hFe
/Er F s
( 0 0 00 000 0 0 0 0 00 5 5 5 0 5 0 5 5
/
' 5s P 0 5 5 5 5 5 8 8 8 8 9 5 5 2 2 2 5 7 0 2 3 3 3 3 34 5 2
5 6 E e1 0 5 4 r 1 le n M 1 1 1 1 1 1 1 1 1 1 1 2 2 EF L n/ 1
- fo u E B F o/ A 0FF '
8 T A A C 1
/
P 2'
<80 C T
SOO T32 65 R E R U C D F 0 t n A T E le1 9 n t s si o C p M r v e S en O O T I 0 0 Lia m-F h- tes 6 T 8 U R MI 5 1 1 gt gnin q 5 4r g A init a e L 2 P TE s p
=
p
=
n a mgr 3 o is p in PS E iop m o o f P YU R 0 m e m e n e upt r osp i dt o t R5 r te U BY U 3 3 T T r h u 0 s Z 8 d n e osV 0 Se 5 P "p P 8 a e T WAR D F a C C n la 5 i l s 2C r u y w
/
h A OP L S E *0 5 R y s R mo u mT B 3 s e mR B a pO r d E L C 1 na n er d s od n ia u v onn io " AR O o r M c p r ir sT a H et ia p anea f 0 d CZ R t=
= D t a e 1 pt h o . U y RP bS rho Vl dop O'F b mA a l 3 e
l b y l a md yo r
& R P m
u b o R t r r e n p r CS d Pt oGnR &0 s l e O h S u oZ t F ent F h i s S RF d BtS P S aZ 0 T as i eP c s mH xts nNS T sP S 52 im5 R N Z lo C 2 C d mr C yV O PTL m e a r a oH OP e H t e e s a R n e o RbB oh C Z TITED:NE F e eDu eSt D hC sDah F r e c h . T P UI A TC es le isR e set k a h C s eCmt dqe e Sl e t 2 AM E E 1 4 eV Rat b er c lofy c l 4 a sk r 4 e c c CIL WL e t t t e et c ot r la
= rmtoaRt n e oal cnC i e e ev E TB u r e r r ef r a h etaf PP r
REE A
- oh r loOe c n h s r t yI r
eh d o c n t I E: PTDA .W F mFTIsPV1 TEC3 2.V C E n iSCPS3 9 8.C R DRFO 4 4 5 56 6 67 8 8 U N PDL I L Mt2 1 222221 1 22223 2222 5 5 5 667 8 9 1 9 2 2 22 222 2 2 2 T: DEASTP U .A 2 2 0 1 91 1 t N C TAMX A OIEE A r fA fA I l M t f L R P P L I HTM i P O P O e&w lll
r mn L _ a s s s n A) 4 o oo o E t m o o nn D uhw.t c o e A( ar r t n rt
==
t
)
F ( 066 7 7 7 4 1 9 5 5 22 8 7 7 8 8 7 7 7 7 P 066 9 9 9 1 2 2 3 3 4 4 4 5 5 8 4 4 4 4 4 4 4 4 5 66666 1 4 4 4 M1 3 3 3 3 3 4 4 E T E p R N tu r Z I a P L 1 S t
)
E i t t n a g G nl
; P is p
R I ( S 00 5 0 05 50 0 00 000 0 0 U e S 3 5 50 2 52 5 2 750 5 0 2 5 5 7 7 0 4 4 0 5 5 5 5 5 1 1 2 22 2 3 3 33 33 4 4 4 4 7 1 1 S n E 1 1 2 22 2 E O R P r H a C T e R G lc ) N u h
/
t
)
I N F F T S 2 ~ 0Y 0P ( 00 0 00 00 0 0 0 0 00 0 0 0 5 5 5 0 2 2 9 C a 4v (EF 1 P 00 101 151 5 5 5 6 7 8 9 07 8 3 3 3 3 6 0 3 3 1 1 1 2 2 2 33 4 4 4 5 5 5 6 E S e n M1 1 1 1 . EF L nR a2 m1i5 i E B F k 2 0 Fr foFF T A T A r 0 '76'0 *0 6.231 C S A1 1 1344 R E R U s F D t s e 5 E 1 s r p 3 4 C S T r s lo F . O I M 6" n s f e s s a p y r t r 0 inO a T n 0 U R I L 8 ia r n aV o t is o 1 t n P T E P S E
~
d t d r B lo mF co < FiA a n R t, T e e' e '5 mis P YDU h 0
" t le e a 2 o m t
BEY U lo s t a 7 y C 2 ar a3 4 le r
. U WWA D t r er 2
~V s s St a p r
pt h i e t T OOP R E no o mD e C s u aw ae A f t LLS C h n E LL AAR O lb e c V g I n S e d uuA t n Ri&a h s h a iai e n gH sdn mp pe CCZP s c m n p t t R is i b H RR P u ie /l Mo avrop p in bl atea me r p pi a05 yaP ul
&& R S B Hb wcr f m m )s a m u S RRF O
Cm Sl e P v mr e elo u u P P (p s oh t e P k.em7 ac o1 r pR Ra u a sentnCCm Et r N ZZ PPT R Ct ot O T :NNE I L t nt e Ne l usdn oRRu nV d s Z R sV st oS h t UITEED AMEE E eS VhV eZ toR aB cV oes c dt P aB Pdr V B zeR B TiZ e 4 P ufi2Dl nH ps eTt at ei di sR sP V T cr r h T r u Ph i CIL WWL nb E ayb y nst l l t t ns s et r t E TTB pmj e r h r eaa o ad pa pdjear ac ia e us t t
- REEE A es r r u r EP TBB A . .W l aat s ef i s e t r i r afi c es t r e t FETiVEVWpSSSSAOSOAPPSMV l t t t t l PDRFFO n UN DA UPOD t I L L
. .A 5 6 1 2 3 5 T: ESTPP 1 2 N
A CTAMMX OI HTTM EEEA 4 2 6 7 05 2 2 3 5 55 73 8 1 1 5 6 73 3 1 L RM 1 1 00 0 0000 1 2 2 2 4 4 4 51 151 5 P PL I 8 9 999 1 1 1 1 1 1 1 1 1 1 1 1 1 i$ ll
4 m -
+-
z .i i o i n
~"-
=
.f .
/ ^ ,n .
j g, m , O/, c, . . d.~ m m f ,. g d n s in ,:'I' 3
= Na /
ve v i u s m u
-=xgm , 51 /
0 ' . 5 i
'G 5
g. 3/ ,
,W-c ,.-
\
S [ " e .
,isn N /
7 7 m O I T A s m f x,s, m C =
= R e
4 O E B L =w D U C T E a c i r a A NT nw am r om N S g N 1U YE
- 6. E R v ,a /L s
3wa r o r w c o em mu m E P N F O t s. xj\ i EGNO m a o m S . RRR T r g a UUAT v,- , m r . v n h-m o e GS I N \.No Ts s e o s g, , FRIE ZR Tm r o E N t s i '% n N , n PO C c o I L Y A e m s o p
- g.
i 1 / T m R P c e S / E ,'m- m u 7 a h E k m- - g h.g u N v,n s g w vf T O f 0 1 N s m % in o C ii, O cs
I l l I ga 6
- m _
s. 4 iI k %
- k
%[.
m A-I yi
,/ /
/ *'
1 R - E _ Z - I
- ~' -
R L S " S E _ R P
~
~
N E 2
. - ~
\
2 q'A I ,I Y / 6 9
-~
1 E R U G I F
'/
/., , X N
9l/
, T i
i a ~ R O NH%
~ -
m/i I%
~ I H % ,
lI 3 s
- o r o
$*6N/'gNg~, N 0 t .
l 9 an ni 5 g
~ ' .i o f ~ st 1 e
E ,I 4 k* dd . e6 rr ere t el tfb eea L rT E T
- O 6 N
- ,; m m w I;
c: 1 OCONEE UNIT .P FUM 1 sco 'f,' t ._. " LOCATION 7 v-
- 620 t-
. [1, W Top W
' 600 580 _ . -
,_, $ ^ - ^ . -. ' ' ' ;.MA:,- ^,- :: w'rk W 560 E =^- - A- . _:-F _ ;c. m 0 SanamT e 5*
' N, bW 520 V.
-l l' Li 500 . i . . .
10:48 1498 17:28 20:48 e -- 11ME [HH.MM] PRESSURIZE 9-t .- 6s0
~
s40 LOCATION 11 I W
- 620 - i b -
Top Themeco@ r 5 580 - _ _ ,
- ;; a , _ ;/-V- _ u-z , _p f
g560-cr b 540 -
- a .
- - nn
$20 -
^
500 . . . . . 10-48 14:08 17:28 20A8 TIME [HH.MM] .. I I 1 f L! [L-
FIGURE 6.3 - 1 SURGE L NE STRATIFICATION DATA POWER OPERATION TEST DATE: 2/21/89 SAMPLE TIME: 20 SECONDS 660 s40 LOCATION 1 5620
. M Thwmoca44 s
Ge00 s '~ ~~ / Bonom Thwmocasm# k a g 560 - W C 540 - Hot Leg 520 - t f) , 500 . . . . . r y !
. STEAM 10:48 14:08 17:28 20:48 f l TIME [HH.MM)
/ conesuration not to scue. l Refer to Figure 6.1 for cmensions of confguratiort d%
N t> 660 s40- LOCATION 4 620
!5 g 600 -
s %_, m 5 580 -
-:_==.
m... . . sJ. j. E, E ,: - V APERTURE CARD l $" 5'0 ;
,ue ,inaitNt O"
...... ~ , . - .- - Acertin? O2M l %0 . . . . .
10.48 14:08 17:28 20:48 TIME [HH.MM] wwee i
~
6-17 $ 9 0 's 3 1 0 l b l 'h< i i
. - = g
=-
%h 'meb$r*M '. m.7 i-
*I n
, /
4 L: Si . l,l- '
- f. -
OCONEE UNIT! i3 t s , FulA,
-us ,
!f ' .
p e-'
.?.
I
- r. . No T1
- 1 00=10.75 in i t H;. .- ' T2
'II
- -{.
'W T3 -
e.. 1,
~
, (~h c! - ID=8.75 in - '
= ,
.-[]T4:
. ! -J -
1 . v l 1 Ts d . T6 7 T7 -j
,, SECTION A-A i
'6!
.a <
s 10P ll . ~ ( [ e
,. Location 11- '
13 m.. F-Te E !
- S- i
- f. ~ ,
ga - J~ t- 8* 1 T6 i
. i S* .- --. t*
le j- *
,? 80T1048 0 w ,
530
.,,,,,,4 bee tet SFS tes 8'I f-AVG. TEMPERATURE DEGREES FAHRENHEIT ( *F l r.
l--Qf w r .
B SURGE JhE STRATIFICA~~lOM DATd OWEROPERATION TESTDATE:2/21/89 3 1. Data points are average thermocouple readings HotLeo
' taken between 10:43 and 19:48.
g I
- 2. Surge line configuration not to scale. See I Figure 6.1 for dimensions.
f) ,
\
sumu GENERATOR ! N l
/
C i PRESSURIZER- --+- Q ', u N^ f
^
j i, a >>
..- = l t
,. Location 5 l
,- T3 i S1 i "
APERTURE
.- CARD
#go ,stanai en 0" rs
;sperture Card f4 BOTTOM O* a . . ' ' ' ' '
W 660 Sa0 UU 500 500 AVO. TEMPERATURE.pEGREES F AHRENHEIT l *F 1 - l , L c-18 8 90 C.3 i O l 6 l 'h)
9 1
- 7. THERMAL STRIPING
!J - 7.1-. Definition Thermal striping is the localized metal stress caused by repetitive 'fluctua-
- tions of. the temperature at a fluid-metal interface. The fluid temperature fluctuations are due. to the interactions between forced flow and buoyancy.
L The buoyant forces tend to stratify the fluid, obtaining vertical segregation (in a horizontal flow component) by temperature and density. The fluid shear forces associated with forced convection and fluid viscosity, on the other hand, tend to mix the fluid. The combination of these effects can generate undulations of the fluid-fluid interface, resulting in thermal striping. The existence and characteristics of these fluctuations depend primarily on the simultaneous occurrence of buoyant forces and fluid shear forces which are of comparable magnitude. 'The fluctuations are also responsive to the flow
' geometry; for example, they may be greatly amplified by the helical, secon-dary fluid motion induced by axial flow in a pipe bend. Finally, the fluctuations of--the pipe surface temperature are directly responsive to the interactions between convective and conductive heat transfer at the fluid-pipe interface.
The basic concern associated with thermal striping is that it is a mechanism for crack initiation. 7.2. B&W Owners Group proaram The objective of the B&WOG program is to thoroughly evaluate and quantify the effects of thermal striping on the integrity of the surgeline. In order to resolve the surgeline thermal striping issue, a multi-element plan has been initiated. An important element in the plan is the measurement program at Oconee Unit 1. Thermocouple were installed around the surgeline outside circumference at several locations during the measurement program at Oconee Unit 1 (see Appendix A). These thermocouple were intended to measure the 7-1
temperature distribution in the surgeline during stratification. The data from the measurement program are currently being evaluated. In order to detect thermal striping by temperature measurements on the outside of a pipe, the temperature oscillations must have both a large amplitude and a long period. Assessments performed to date on the Oconee test data have not detected the thermal striping phenomenon. This ccnclusion is supported by results from the thermal striping literature survey which show that the
. predominant striping frequencies are too high to be detected on the outside surface. In any case, thermal striping of some magnitude could have occurred without being detected by outside thermocouple.
The B&W Owners Group program has five basic elements that will contribute to resolution of the thermal striping issue. These are:
- 1. Evaluation of Surgeline thermal-hydraulic conditions
- 2. Surgeline Pipe Wall Heat Transfer Analysis (damping effects of !
surgeline wall on. measured temperatures) l
- 3. Evaluation of Oconee field data
- 4. Assessment of available industry stratification and striping data
- 5. Structural analysis of striping effects on the surgeline The first four of the above elements serve to develop and justify the thermal striping input for the fifth element. The program is laid out to maximize what can be learned from all possible sources short of performing a labora-l tory test specifically aimed at striping phenomena. The following paragraphs briefly describe the work to be completed to support a submittal (technical report) to the Staff in October 1989.
7.2.1. Evaluation of Surceline Hydraulic Mechanisms The objective of this task is to identify and understand the relevant thermal-hydraulic mechanisms involved in the stratification and striping phenomena of the surgeline. This knowledge will help to ensure that impor-tant aspects of these phenomena are accounted for in the final fatigue evaluation of the surgelines. 7-2 L- - - - - - - - - - - - - - - - - _ - - - _ _ _ - ---- _ _ ---_--- -_-____----_-- _ _ _____i
L L l This task will make use of the open literature z.nd previous experimental l work. Code ~ analyses results, as they apply to the surgeline, will be reviewed. Hand or code calculations may be used to assess the impact of important variables that influence the degree of stratification. Variables
~
of interest are likely to include the surgeline insulation characteristics, effects of pressurizer spray bypass flow, localized effects of pipe discon-tinuities (such as elbows), and surgeline endpoint conditions. This latter item includes the temperature differential between the ends of the surgeline and the various flow conditions that can arise over the range of temperature differences. 7.2.2. Surceline Pioe Wall Heat Transfer Analysis This analytical task will provide two basic types of information. The first is the steady state temperature distribution of the inside pipe wall given the outside pipe wall temperatures. The second, and perhaps more critical information it will yield, is an assessment of the transient heat transfer characteristics of the pipe wall. The relationship of inside fluid tempera-ture amplitude and the associated period of oscillation will be investigated. This result will help to determine reasonable upper limits on the thermal oscillations that cause striping. This latter result may not be necessary depending on the results achieved from the survey of industry data (discussed in Subsection 7.2.4). 7.2.3. Evaluation of Oconee Field Data The data evaluation will follow the data reduction and plotting process currently in progress. Some meaningful relationships and factors probably have yet to be considered, but the kinds of information of interest include maximum and minimum temperatures and temperature differentials for the various plant conditions involved in heating up, operating, and cooling down the RCS. Correlations between different locations in the surgeline will be considered. The correlation of plant operations to observations in the surgeline are important, including the assessment of differences in operating procedures and operator practices between B&W-designed plants that may have an impact on thermal stratification and striping. A by-product of this effort will be the identification of potential changes to operating proce . dures that can eliminate or reduce stratification or striping. 7-3
1 l-. l The Oconee, data evaluation in conjunction with the thermal hydraulics evaluation of the surgeline should yield a physical model of the surgeline s hydraulics and the processes which caused the Oconee surgeline transients.
.This -is expected to provide . insights that will lead to a better overall' understanding of the striping and stratification phenomena and measures that can. effectively. reduce them.
7.2.4. Assessment of Available Industry Data Literature searches on thermal striping have identified several papers that directly address the thermal-hydraulic conditions that lead to the oscil-latory behavior of the fluid interface. These will be reviewed in detail for applicable data and correlations. If the review concludes that it is techni-cally justified to .do so, analytical / scaling work will be done to develop striping frequencies and amplitudes appropriate for surgeline conditions. The analysis may conclude that the potential for or existence of thermal striping in the surgeline is nil for certain plant conditions and not others. Such results will be extremely useful in determining the ultimate disposition of the striping issue for.the B&W configuration and will be considered in the ultimate input provided to the structural evaluation. 7.2.5. Structural Evaluation of Surceline for Thermal Stripino The successful disposition of the thermal striping issue is contingent on a demonstration that striping, if it does exist in the surgeline, has a negligible impact on the structural integrity of the pressure boundary. Until the earlier tasks are completed, the extent of this analysis can not be defined. Preliminary results, as discussed in the next section, show that j the likelihood of initiating a crack in the surgeline from the thermal striping phenomenon is minimal. A complete accounting of surgeline condi-tions and transients is planned for the final submittal. 7.3. Preliminary Results 1 A conservative interim assessment of the cyclic thermal stress due to striping has been made based on information from the available literature and from measurements at Oconee. Results from this striping evaluation show that the fatigue impact on the surgeline is approximately 10 percent of the 7-4
- ._. _ = - . -. . _ . -- __ _ __ _ _ _ _ _
7_ ) ct I E -1 allowable usage factor. The evaluation shows that temperature differentials
. in~ the surgeline with the plant at power are not large enough to affect the !
l usage factor; those existing during the early parts of plant heatup are most significant. The conclusion of this ' interim work is that potential fatigue _, effects of thermal striping are not sufficient to cause concern for continued plant operation. The-information in the literature suggests, however, that buoyant effects are important considering the temperature differences and flow ranges of concern to the pressurizer surge line, and therefore, the possib,ility of striping must be considered. The following four areas of research provide valuable insight into the process of thermal striping: o BWR Feedwater Nozzle Tests o LMFBR Tests (Westinghouse) o Argonne National Laboratory (ANL) Tests o Project HDR (FRG) Tests Based upon the reported ranges of frequencies observed in these experiments, , and the proportion of the top-to-bottom temperature differences actually imposed on the wall, a conservative estimate of the pipe wall thermal exposure was made by assuming a single frequency and amplitude of the inside wall thermal fluctuations. The frequency of occurrence of 0.25 Hz and a value of 45% of the imposed fluid temperature differences were selected for
. the interim analysis. This interim set of characteristics of the striping phenomenon, was chosen to be representative of the available striping information, and to lie on the conservative side of the published ranges of striping amplitudes.
Oconee test data provided an important input to the interim assessment of striping fatigue effects. The data were processed to determine the duration during the plant heatup over which temperature differences of various magnitudes were experienced. Using the interim values extracted from the literature, the frequency and magnitude of the thermal exposure of the pipe 7-5 i
wall were determined. From -this information, thermal stress ranges were evaluated, and' fatigue impact was assessed. The contents . of _ the published literature leading to' this. approach are described and the methodology ' leading to these interim conclu.sions is discussed in more detail in the following paragraphs. 7.3.1. ~ Assessment of Available Industry Data The estimated characteristics of surgeline striping are based on the avail-able striping research. This research is described in detail .in section-7.3.1.1. The four major areas of research are presented individually in this l subsection and are then summarized. The surgeline striping characteristics are defined and' discussed in section 7.3.1.2. The conservatism associated with,these assumed characteristics are also addressed. 7.3.1.1. Available Research The thermal-hydraulic characteristics of thermal striping have been examined in the following four major areas of research:
- 1. BWR feedwater nozzle tests
;- 2. LMFBR tests (Westinghouse) =
- 3. Argonne National Laboratory (ANL) tests
- 4. Project HDR (FRG) tests These programs are summarized below, and are described in more detail in the
. subsequent paragraphs.
The BWR feedwater n'ozzle studies were extensive, but the geometry of interest was quite' unlike -that of the pressurizer surgeline. The BWR studies did demonstrate the ability to combine low-temperature data with high-temperature data and with plant striping data, by suitably adjusting the low-temperature results. The application of the BWR feedwater nozzle results illustrated the use of probability density functions. Thermal fluctuations were analyzed to determine the frequency of occurrence of cycles having discrete ranges of amplitudes. These incremental-amplitude analyses were carried through the nozzle stress analysis by introducing plant time-at-conditions data. These BWR feedwater nozzle studies are described in detail later in this section. l 7-6 1 l l 1-
i 1 Woodward examined the fluid temperature fluctuations in transparent horizon-tal pipes. The amplitudes of the near-wall fluid temperature fluctuations reached 60% of the imposed fluid temperature difference, with most of the cycles exhibiting amplitudes of 10 to 35%. The frequency of fluctuations ranged from 0.1 to 0.5 Hz. A film heat transfer coefficient was needed to determine the wall thermal fluctuations from those of the near-wall fluid. Woodward referred to the studies of Fujimoto et al. Fujimoto et al also studied striping in a transparent horizontal pipe. A fluid density dif-ference was imposed by adding calcium chloride to the warmer fluid stream. Thin squares of copper were used to measure wall striping. The striping amplitude was less than 10% of the imposed temperature difference (the temperature difference between the interacting hot and cold fluid streams). The film heat transfer coefficient was 1.25 to 7 times the Dittus-Boelter coefficient. The studies of Woodward, and of Fujimoto et al, are described in detail after this section. Kasza et al at ANL have tested extensively therm'al stratification ,and striping in transparent horizontal piping with bends in both the vertical and horizontal planes. Based on a limited amount of published power-spectral-density information, the higher-amplitude fluctuations occurred at lower frequencies, 0.1 to 0.6 Hz, with amplitudes of 30 to 40% cf the imposed temperature difference. These ANL studies are outlined later in this section. Wolf et al, in the TEMR test series of Project HDR, measured striping in metal, horizontal pipes at plant-typical temperatures. The complete results of these tests have not been published, however. Typical striping frequen-cies were 0.1 to 10 Hz. The amplitudes of the wall temperature fluctuations were generally from 10 to 40% of the imposed fluid temperature difference, with peak amplitudes from 25 to 50%. Examining the single test presented in the published results 4f the nine "PWR" tests), the maximum striping amplitude was approximately 30% and the frequency of occurrence of the larger fluctuations was approximately 0.2 Hz. Wolf et al noted the interactions between convective and conductive effects, and hence the difficulty of extrapolating to a plant the results of tests performed using a transparent model. They also noted the insensitivity of temperatures measured at the 4 7-7
k yi outside of a metal pipe to inside interactions. The TEMR tests are described later in this section. BWR Feedwater Nozzles Tests l The BWR feedwater nozzle configuration was examined in relation to observed
- l. feedwater line cracks. The thermal striping of this configuration has been
- l. .obtained .from two test facilities, Two-Temperature and Moss Landing, as well as from plant measurements.12 The Two-Temperature Test Facility was limited to atmospheric pressure; hot and cold fluid temperatures of 160 and 70F were used. The Moss Landing Test Facility, on the other hand, achieved plant-typical temperatures. The results of the two test facilities were combined with plant data by adjusting the Two-Temperature results to account for the changes of fluid density, viscosity, and thermal conductivity between the test and reactor conditions.
The . composite data was processed to obtain the number of cycles having discrete ranges of stress amplitudes. These amplitude ranges, or windows, were prescribed to be relatively small at the higher amplitudes, and up to 20% wide at the smallest amplitudes. The results of this analysis were presented in tabular form.12 This data has been restated in terms of windows of equal amplitudes, 10%, and plotted in Figure 7.1. The frequency of occurrence decreases rapidly and regularly up to a stress amplitude of 50% of the maximum stress, and then more slowly at the higher amplitudes. Although most of the fluctuations had low amplitudes, approximately 1% of the metal temperature fluctuations obtained stress amplitudes approaching the maximum stress. It is estimated that the observed amplitudes.of the near-wall fluid tempera-ture fluctuations were reduced by one-half to obtain the amplitudes of the wall temperature fluctuations and hence the wall stress amplitudes. The maximum stress of Figure 7.1 thus corresponds to a wall temperature fluctua-tion of approximately 50% of the imposed temperature difference, the tempera-ture difference between the two fluid streams of unequal temperatures and densities. Most of the fluctuations had stress ampStudes less than 50% of the maximum stress, which corresponds to wall temperature fluctuations less than 25% of the imposed fluid temperature difference. 7-8 i
LMFBR Tests 13 Woodward investigated stratification and striping in a 1/5-scale model of an LMFBR at the Waltz Mills Test Facility. The model was plexiglass, therefore the hot and cold water temperatures were limited to 130 and 70F. Two lengths of horizontal piping, of 4" and 6.5" inside diameter, were examined. The tests were conducted in the turbulent transition range, with Reynolds Numbers (based on half-pipe flow areas) of 2 x 103 to 8 x 103 . Dye and thermocouple were used, the thermocouple were typically inserted 1/32" into the fluid. The thickness of the interface region, over which the fluid temperature changed from hot to cold, ranged from 0.6" to 2". The striping frequency was 0.1 to 0.5 Hz and the fluctuations were approximately sinusoidal. The amplitudes of the temperature fluctuations (of the near-wall fluid) approach-ed 60% of the imposed temperature difference, and were most pronounced at low Richardson Numbers. Probability-of-occurrence information was presented for only three ranges of amplitudes (or windows). This information has been converted to the fractional occurrence for constant window widths of 10% amplitude, and plotted in Figure 7.2. Most of the fluctuations had mid-range amplitudes, 10 to 35%. The probability of occurrence dropped rapidly at the higher amplitudes, approaching zero at 60% amplitude. A heat transfer coefficient was needed to obtain well temperature information from the near-wall fluid temperature measurements. The heat transfer coefficients determined by Fujimoto et al were referenced by Woodward. Fujimoto et al l4 tested striping in a 14.2" horizontal pipe made of acrylite. Calcium chloride was added to the warmer fluid stream to obtain plant-typical density differences. (The fluid temperatures were used simply to track the streams of differing densities.) Wall striping was measured on thin squares of copper. The amplitude of the fluid temperature fluctuations was observed to decrease near the wall. The fluctuations at the interface between the fluids of differing densities evidenced frequencies of 0.3 to 3.0 Hz. The amplitude of the wall f?uctuations was less than 10% of the imposed tempera-ture difference. The convective heat transfer coefficient was calculated to be from 1.25 to 7 times that of the Dittus-Boelter correlation (for forced convection in tubes). The information obtained by Woodward and by Fujimoto 7-9
et al was referenced in the evaluation of thermal stratification of the pres-surizer surgelines of the South Texas Project power plants. ANL Tests Kasza et al have conducted extensive experimental studies of stratification and striping at ANL 15-25 . These studies have generally used water flowing turbulently in transparent pipes of 6-in inside diameter. Combinations of horizontal ,and vertical piping. lengths have been tested, including bends in the horizontal plane. Vertical lengths of piping were observed to eliminate stratification. Stratification in horizontal lengths began at a Richardson - Number of approximately 0.05; flow stagnation and reversal occurred near a Richardson Number of 0.7. Kasza et al applied the buoyancy index of Jackson and Fewster, namely [= Ri Re-0.625 f p Kasza et al observed a correlation between buoyant effects and this buoyancy index. The threshold of buoyant effects was found to correspond to a Y on .theorderof10-4;aY-ontheorderof10-2 or larger obtained strong buoyant effects. The more-recent investigations of Kasza et al 18-25 obtained some details of the thermal fluctuations. Whereas the bulk fluid temperature fluctuations were about 75% of the imposed temperature difference, the amplitude of the wall fluctuations was 30 to 40% of the imposed temperature difference. On the basis of limited power-spectral-density information, most of the signal energy was concentrated below 1 Hz, peaking between 0.1 and 0.6 Hz, and decaying approximately exponentially with increasing frequency. These maximum fluctuations were observed approximately one diameter downstream of a horizontal elbw. .HDR Wolf et al have conducted extensive examinations of thermal mixing in the HDR project at Karlsruhe, FRG.26-35 The TEMB test series examined pressurized thermal shock using a large-scale pressure vessel and various high-pressure injection configurations; the experimental results were compared to the - predictions of many codes and correlations.26-33 Fluid temperature fluctua-tions were observed and recorded, but received little emphasis. 7-10
1 The TEMR test series concentrated on thermal stratification in horizontal feedwater lines.34-35 The test section was a 20-foot length of 15.6 in l' inside diameter metal pipe, extensively instrumented with 11-ms thermo-L couples. Cold water entered one end of the horizont61 run through a bend from vertical upflow, the opposite end of the horizontal run was attached to a reservoir of hot fluid. The TEMR tests consisted of 3 subseries, 2 of which were labelled "BWR" and "PWR." In the BWR tests, a plate with slit orifices was installed at the junction of the horizontal run with the reservoir, to simulate the holes of a typical BWR feed sparger. The horizon-tal -to-reservoi r junction was unobstructed in the PWR tests. The third subseries of TEMR tests considered the buildup and decay of hot water pockets. The horizontal-to-reservoir junction was blocked except for a horizontal slit at the bottom'of the pipe cross-section. The PWR -tests of the TEMR series are most relevant to the surgeline con-figuration. The ranges of conditions of the 9 PWR tests are listed in Table 7.1. The average fluid temperature ranged from approximately 200 to 300F, and the imposed fluid temperature differences ranged from approximately 200 to 400F; the volumetric flow rates spanned 10 to 200 gpm. The Reynolds Numbers based on the flow area (rather than on a reduced flow area to account for stratification) were in the turbulent range. Kasza and Kuzay have employed a buoyancy index which is dependent on the Reynolds, Richardson, and Prandt1 Numbers. In an order-of-magnitude sense, the threshold of buoyant effects occurs at an index of 10-4, and strong buoyant effects occur for an index of 10-2 and larger. Applying this index to the PWR test conditions, all the PWR tests of Wolf et al were in the strong buoyant range. The interface between the fluid of unequal densities was characterized as being wavy, with typical frequencies between 0.1 and 10 Hz.22 Within the mixing layer, the fluid temperature fluctuations were not damped near the wall .21 The fluid mixing did reduce the local maximum temperature difference from the imposed temperature difference, however. The amplitude of the wall temperature variations, expressed as a fraction of the amplitude of the fluid temperature fluctuations, was stated in two contexts.22 For all the BWR and PWR tests, the fractional amplitude was 10 to 40%, but the peak fractional amplitude was 25 to 50%. 7-11
Measurements from one of the pWR tests, Test 33.19, were presented.22 The conditions of Test 33.19 are listed in Table 7.1. Test 33.19 was charac-
.terized by a relatively high flow rate and ratio of inertial to viscous
. forces (Re)., and by a mid-range temperature difference. The resulting ratio of buoyant to inertial forces (Ri) was low compared to that of the other PWR tests,aswastheindexofbuoyanteffects(f). The temperatures measured in the fluid, on the inside pipe metal surface, and on the outside pipe surface were presented.21,22 Examining these figures, the fluid temperature fluc-tuated with an amplitude which was almost equal to 'the imposed temperature difference, the difference between the temperatures of the hot and cold fluid streams. The temperature of the inside pipe wall fluctuated with an inter-mediate amplitude, but the temperature of the outside surface of the pipe evidenced no fluctuations.
The inside pipe surface temperature exhibited irregular fluctuations. The maximum amplitude of these fluctuations was approximately 30% of the imposed - temperature difference; the larger-amplitude fluctuations occurred at intervals of approximately 5 seconds. This interval corresponds to a frequency of occurrence (of relatively large fluctuations) of 0.2 Hz. This, frequency of occurrence must be distinguished from the characteristics of the individual fluctuations. Because the larger-amplitude variations generally occurred within groups of fluctuations of much smaller amplitude, the frequency of all fluctuations was approximately 1 Hz. That is, the larger-amplitude fluctuations, which occurred at intervals of aprraximately 5 seconds, each persisted for only approximately I second. These characterize-tions were obtained by examining the figures presented for PWR Test 33.19. The HDR experimentalist drew the fc11owing conclusions from the TEMR re-sults:21,22 e The extrapolation of model data to a plant, using a transparent model, is made difficult by the complex interactions between convective and conductive phenomena. e There is no simple, unique correspondence between the thermal response of the exterior of the pipe and that of the interior. i 7-12
l Comparison of Thermal Strioina Conditions Figure 7.3 provides an overview of thermal striping. Both dimensional and dimensionless axes are presented. The dimensional axes, flow rate versus temperature difference, apply specifically to the surgeline geometry and conditions. The dimensionless axes, Reynolds Number (Re) versus Grashof Number (Gr), both correspond to the surgeline quantities and provide a more general basis with which to assess the thermal-hydraulic interactions. The Reynolds Number indicates the ratio of inertial to viscous forces whereas the Grashof Number provides a measure of the ratio of buoyant to viscous forces. The information presented in Figure 7.3 is to be regarded in an order-of-magnitude sense. For example, flow rates were converted to velocities using the whole-pipe flow area, rather than reducing the area to accommodate stratification; and the surgeline fluid properties were evaluated at 300F and slightly subcooled - they are relatively insensitive to pressure, but quite sensitive to temperature. Regions of relatively weak and of relatively strong buoyant effects, compared to inertial effects, were estimated by evaluating the Richardson Number (Ri) and the buoyancy index (f) which has previously been described. The condi-tions of interest to surgeline strati'Tication and striping lie in the " strong buoyant effects" range shown in Figure 7.3. The range of interest is further refined by considering the surgeline temperature difference (DT): there is no fatigue concern for DTs less than 90F, and the maximum DT is approximately 'j 300F. The conditions of interest are thus approximately 1011 < Gr < 1012 and Re < 105. (There is probably a lower-Re bound, below which the buoyant effects predominate to the extent that interface instabilities and thus striping are suppressed; this limit has not been quantified, except that Kasza et al have observed such a limit for the case of fluctuations down-stream of a bend in the horizontal plane.) The dimensionless axes of Figure 7.3 provide a convenient basis on which to i compare the several investigations of striping. These are the singly cross-hatched regions in the figure. The conditions of Woodward, and of Kasza et al, lie far below the range of Grashof Numbers of interest. Both these striping investigations were conducted at atmospheric pressure, thus the reduced thermal expansion coefficient and DT, as well as the increased 7-13 l l l ___________-____-______________-___O
\
viscosity, resulted in relatively small Grashof Numbers. The conditions of 1 Wolf et al, on the other hand, are just on the high-Gr side of the conditions of interest, due only to their larger pipe diameter compared to that of the surgeline. Several data sets are not shown on the figure. The data of Fujimoto et al was obtained at atmospheric conditions, but with the inter-fluid density difference enhanced toward that encountered at surgeline conditions. The viscosity remained in the low-temperature range, however, and the data evaluation of Fujimoto et al depended on a correspondence between mixing and ' diffusion within a thermal gradient and a concentration gradient. The EDF data is also not shown. This data, although unpublished, was apparently obtained at atmospheric pressure and correspondingly low Grashof Numbers. Finally, the conditions of the BWR feedwater nozzle research are not shown because of the pronounced geometric dissimilar des between the nozzle and the surgeline. It is uncertain whether the low-Gr data of Woodward and of Kasza et al apply at the conditions of interest for the surgeline. Certainly the visualiza-tions available with the low-Gr tests provide valuable insight regarding striping mechanisms, characteristics, and limiting regions, but these insights may apply only at the tested conditions, even if the Richardson Number is preserved in the translation of conditions to those of interest. The work of Wolf et al thus seems singularly pertinent to surgeline applica-tions. It should be recognized, however, that the most-applicable subseries of tests by Wolf et al included only nine test conditions, utilized only a horizontal pipe, and each. involved a transient obtained by injecting cold fluid into an initially hot and isothermal pipe. Also, the detailed results of Wolf's research are as yet unpublished. Notwithstanding their dissimilar Grashof Number ranges, the published striping charact' eristics of the three investigations plotted in Figure 7.3 were quite similar. Woodward obtained frequencies from 0.1 to 0.5 Hz, and amplitudes of the near-wall fluid oscillations which generally ranged from 10 to 35% of the imposed fluid temperature difference, and which peaked near 60% of the imposed DT. The characteristics of wall temperature oscillations were not available in Woodward's research. Kasza et al obtained the following 7-14 ___.____.___.________m.______ _ _ _ _
it i t F _.; _1 0 ' i characteristics of vall ' temperat' u re fluctuations in the horizontal pipirig o- dowr. stream of a horizdntal bend: amplitbdes fmn 30 to';40% of tho' imposed , , fluid t'emperature difference, with the frequencies of thejlarger-amplitude ' l I fluctuations generally between 0.1 and 0.6 Hz. Wol f et al also _optained wall ; fluc'tuation characteristics. These fluctuations were reported to occur over z' pe E frequency range from 0.1 to 10 Hz: there amplitudes were generally between '10 and 40% of the imposed ' temperature ' differt.nces with peak ampli .
, tudes of 25 to 500 Examining the single published wall temperature trace ,
(hf the PWR series), the peak amplitude was approximately 30% of the imposed
, temperature difference; the fluctuation frequency was roughly 1 Hz, the frequency of occurre.nce of larger-amplitude fluctuations was roughly 0.2 Hz.
' ' Finally, the characteristics of thermal striping were; addressed in a- 1980 report 'by the NRC which summarized pipe cracking lin PWRs.36 The range of-frequencies was 0.1 to 10- Hz. The reduction of amplitude due to film heat transfer was described, resulting "... in a peak metal temperature variation at the surface of roughly one-fourth to one-half the wa'er ' temperature
- vari atior.. "~
7.3.1.2. Surceline Sttipino Analysis ~ Ing,Legumptign_1 Surgeline striping encompasses a range of frequencies and amplitudes, and would be well characterized by a probability density function which defined the frequency of occurrence versus incremental amplitude. Moreover, this probability density function' would be responsive to the ongoing interactions between the fluid inertial, viscous, and buoyant forces as defined principal-ly by the temperatures and flow rates of the interacting fluid streams. Because the available striping data is wholly insufficient to develop such a probability density function and its dependencies, an alternative characteri-zation of surgeline striping has been adopted. Surgeline striping has been characterized by a single frequency of occurrence and amplitude, namely 0.25 Hz and 45% (i.e., the amplitude of the wall temperature fluctuations is 45% of the temperature difference between the hot , and cold streams). These characteristics were selected to be realistic and conservative. The selected amplitude of 45% lies on the high side of the ' observed ranges of amplitudes. The selected frequency-of-occurrence of 0.25 Hz corresponds roughly to the observed frequencies of occurrence of higher-7-15
amplitude fluctuations. It should be noted that the larger-amplitude fluctuations . generally occur within groups of fluctuations of much lesser amplitude, such' that the prevailing fluctuation frequency (of all fluctua-tions) is much higher than the frequency of occurrence of the larger-ampli-tude fluctuations. It should also be noted that amplitude-versus-frequency information, such as the power spectral densities obtained by Kasza et al, clearly indicate an inverse relation which appears entirely logical; namely, the amplitude of fluctuations drops off sharply with increasing frequency. ~ In summary, a single frequency and amplitude of fluctuation have been estimated based on the research available. Striping is better represented by a probability density function, giving the frequency of occurrence for
' discrete ranges of amplitude. This function is expected to vary somewhat with surgeline conditions, specifically flow rate and perhaps _the imposed fluid temperature difference. The probability density functions would then be sampled using plant times at conditions. The resulting striping charac-teristics are expected to be more realistic, and less severe, than the single amplitude and frequency . estimated herein. Another inherent conservatism involves the pipe location affected by~ striping. The interface between hot and cold fluids, and hence the striping-affected zone, slowly vary throughout a transient. This effect is ignored in this striping characterization and the subsequent structural evaluation.
7.3.2. Evaluation of Oconee Test Data The Oconee test data provides important input to the fatigue evaluation, namely the magnitude of the top to bottom temperature differential, the duration of various top to bottom temperature differentials, and the number of transient temperature cycles that occurred during the heatup. These data, along with the assumptions resulting from Section 7.3.1, provide the basic input for the surgeline fatigue evaluation. The following paragraphs describe the Oconee test data reduction and the results of this process as they were input to the fatigue evaluation. Surge line data was taken at Oconee Unit I during the 2/89 heatup, power escalation, and subsequent full power operation. Approximately 150 paramet-ers were sampled every twenty seconds and saved in eleven data file sets. These parameters included reactor coolant system parameters, displacement 7-16
h measurements, and thermocouple readings. This process was totally automated for approximately eight. hours before the data file had to be stopped and a l- new data file started.
= Spreadsheet software macros aided in the conversion of the raw data into more easily evaluated information. Twenty-six plots were created for each data . set for evaluation. These plots were evaluated for insight into the phenome-non of thermal stratification. - After a preliminary evaluation of the data, additional plots were created to aid in the structural evaluation of the surgeline. Plant operations between cold shutdown and hot shutdown resulted in. the greatest stratification as expected. The greatest stratification was well represented by surgeline location 11 which is located near the middle of the horizontal run nearest the pressurizer (see Figure 6.1). Files were combined to- plot the top and bottom thermocouple readings at location 11 for each day of the heatup (see Figures 7.4-7.11). These figures were used to estimate the number of thermal cycles throughout the heatup for the stress analysis.
An additional evaluation determined the length of time different magnitudes of stratification. (delta Ts) existed during the heatup. A spreadsheet was used to determine the amount of stratification at location 11 each time step (every 20 seconds) throughout the heatup. The length of time stratification existed above certain delta Ts shown in Figure 7.11. The location 11 plots and the length of time stratification existed were considered when making assumptions for the thermal stratification and thermal striping evaluations. The thermal striping fatigue evaluation is discussed in the following section. 7.3.3. Structural Evaluation of Thermal Striping Bas _ed on Oconee-1 Measured Data To account for thermal fluctuations in the wall of the surgeline an ANSYS model was built to assess the temperature distribution through the thickness of the pipe. The temperature of the wall was assumed to be 45% of the peak amplitude of the stratified fluid temperature profile. Four stratified fluid temperature differentials were analyzed: 280F, 250F, 225F, and 200F. The stratified fluid temperature profile for this analysis is a sine wave for a 7-17 ______-______-a
period of 4 seconds. This sine wave was closely approximated as a " cut-sawtooth" wave'. For each case the actual inside wall temperature range is 45% of the stratified fluid temperature profile. The average temperature of l the inside wall was based on a minimum fluid temperature of 123F plus one half the stratified fluid temperature range. The 45% inside wall temperature then fluctuates about this average temperature. From the ANSYS computer runs, the data was reduced to produce a temperature profile for the nodes versus the distance through the thic.kness for a given time where the inside wall temperature is at a maximum and a minimum. From this temperature distribution, the linear and nonlinear thermal gradient stresses are calculated. These calculations are based on the piping equa-tions in paragraph NB-3653.2 of the ASME Code. These stresses result in a peak stress range and thus an alternating stress for each of the temperature cases run. An allowable number of cycles is calculated from the fatigue curves in the appendices of the ASME Code for stainless steel. A typical number of striping cycles is based on the data taken at Oconee as described in Section 7.3.2. The time in which the fluid is stratified for a given temperature range is presented in this data. The time duration for which surgeline fluid was stratified between 2500 and 2800 F was utilized for calculation of number of striping cycles for 2800F profile case. The total time duration between 2250F to 2500F was utilized for 2500F profile case. This process was repeated for other temperature profiles. These temperatures and times are representative of a typical heatup and cooldown cycle at a B&W operating plant. From this information, a typical number of cycles can be calculated based on the multiplication of: 1) the time (minutes) in which the fluid is stratified for a given temperature range; 2) twice this value to account for both heatup and cooldown; 3) 240 heatup and cooldown cycles in the design life; 4) 60 seconds per minute; and 5) one over the period (seconds). The actual number of cycles is then divided by the allowable number of cycles resulting in a fatigue usage factor for each temperature profile. The usage factors are then added to give a total usage factor due to fatigue. This information is presented in Table 7.3. The cumulative usage factor due to thermal striping for this analysis is 0.10. l 7-18
p
~1 ~,
From' Table 7.3, it can be seen that striping during fluid stratification temperature ' ranges of less than 200F result in no fatigue damage. Since fluid stratification at power was less than this 200F temperature range, it
- is concluded that fatigue will only be impacted during heatup and cooldown
. transients.
Table 7-1. HDR Test Series TEMR-PWR: RggesofConditions and Conditions of Test 33.19 e The extreme conditions are listed for any of the 9 tests, rather than for the. tests having extreme combinations of conditions. e The ' flow rates and dimensionless numbers use the flow area of the whole pipe; properties are evaluated at the average fluid temperature. e Xis the buoyancy index used by Kasza and Kuzay, where I>10-2 obtains strong buoyancy effects. Condition Minimum Maximum Test 33.19 Fluid temperatures, F Hot 314 486 417 Cold 79 130 130 (Hot-Cold) 201 391 287 Average 198 290 274 Flow Rates Volumetric, gpm 10 200 200 Velocity, ft/s 0.016 0.34 0.34 Re = vdN 104 2 x 105 1.8 x 105 Ri-g4Tfd/v2 (= 1/Fr2) 5 x 101 3 x 104 5.5 x 101 Y= Ri Re-0.625 / [ 0.025 84 0.026 7-19 m_____m__.._.- ____________.__m _ . _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ . _ _ . _ . _ . . _ _ _ . _ . _ . _ _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ _ _ __________.________.._____m . _ _ _ . _ _ _ . _ _ _ _ _ . - _ _ _
- , , .1 1.
l n ' L- Table 7-2. Strioino Cases and Results-
. Temperature .
. Typical #
' Range (F) Period Sa Allowable Cycles (Based Usage Case Fluid- Metal (sec) (ksi) Cycles on Oconee Data)- Factor
.A.- 280 126 4.0 22.62 2.11E+06 93600 0.0444 B- 250 112 4.0 20.11 4.52E+06 120000 0.0265.
C- 225 101 4.0 18.13 1.37E+07 444000 0.0324 0- 200 90- 4.0 16.16 INFINITE ------ 0.0000 E- METAL TEMP RANGES BELOW 90F INFINITE ------ 0.0000
. Cumulative Usage Factor = 0.10 7-20
t f Figure 7.1 Frequency of Occurrence Versus Stress Anplitude 1: Data based on NED0-21821-A adjusted for uniform amplitude windows of 10%.
--100
' 50 \
O
- 20 \
== Q
- 10 oJ -
t l- U l
- e 5 o 1 . B 8
- i r2 .
~
I
-1 l
-0.5 -
\
n* Stress Amplitude, % of Maximum 7,5 100
,0 2,5 5,0 7-21 i
sn r uo o si f ~ rt - ea d Vu e t t ~
"ecu c s u
nl j . eF d% r a0-re3,1 ur I cu f ct do Oa r
- r as f e op do ww ye m od on
_ n cT Wi
.. ed , ui qu n
oed w \
. el du rF et F . si l al l bp
- 2. W 7 -
a ta r - m \ r am ea h r re co uN f g ri if an F o Bu . 1
\. g -
x3 ~- [ [
/
. d = 8_ [
5 0 5 0 1 1 2 2 5
- !IiI1 i- i(
tl
?M
\
s .\ a $_ o I
\ \ \ \ \ ' d O0E- 17
, iv autLaSans mnetxem xoadde e8 ,
;B C '
EG sc o a5 e 5 5 S-e uaaouco anStae3 ou 'd06>17
'9. \
.k 7"
\
~
e
- \
.g
\
$0 W
[
'o 5 $
\, ..
i \ o w e o. e-3 E I
$ \\s o l t ,
o o 5 i 0* ~
\
h - o- %m I.
# x E e 4 o 6
.5 mU r 4s 3g $ c)
% uo 3 2" y%
8 u _^ 5 s E F $8 S-- E 8M , ;- 3 ;;- g R> N i
#3 x\i T
9 %U \ // e,
~ 8E o-g
= 11 m- N
*+to f
2 uo
- u. e ce N
x\s\ \- / e un \\ e C 38 \\ uotatsueal
, quatngani i ,
ad6 'aaeg mot 3 autta6ans E o i f 7-23
;l 1 0 0 _
1 ~ 4: 2 _ N _ O I T A , C _ O
- L 0 G
O T ,0 8 L W A 1 B L W N r-B O I 1 T i 0 2/ 0 A 1 1 J A _ C
/
_ t I
] P _
O - I M F9 8P
/
M O T _ 4 I 7U 0 1 7 T/T: E 2 A ,0 : H: H 4 E . ANH R 2 [ K U _ e G 00 1 E 0 S r g u REAK O F 8 S44 P 0 4 00035 5 444 4 7 7 M 0O 8 L G C _ TT1 I c T ON . i n 00 0 000055 L E AY F STA AD Y89 8 888888 3S R R OO D T T E i A A R R E E N P P _ _ I OO-L E L
)
G T OO I S N _ P E _ E D D57 PPP P EM TO E V 0 R 0 G MMM M R ET C ,0 E _ UUU U O L(F - 6: L P NPPI IP Pt F PNFN R E L OU OtPPP PE MOOO LL O C SSS SLLL P C RBAA 9B RnR U E1 1 1 1 AD0DDUSSTTT T EE BTTEEE EEEOEEAAA U O OM C R S H TPTPM RPRPARRRR RAOAOE TTEEE
/VHHH OE MH R T OZZZ EM E ZTTTTTMPPPP PSSSSS ,
HO T T E 59 20 5 4 M70775 1 350 M P T OO T O N 4~ 8 1 1 0 2 1 22233 222222 O L L LLL . WL L EWLWWW _ BO O OBOBBB C
~
0 _ 0: O
~ ~ - - - - 0 _
0 0 0 0 0 _ 7 E 0 4 0 3 0 2 0 1
"% FQy u 5u.n(EK0oO t
y2 _ l~ l l ill
. 1 0 0:
1 . 4 2 N i O I
- k. -
T A i C O - - L 0 T 0 i 8: A 1 N O I T i A C I
]
M F -8 9 . P M I 8U
/
0 T /T 2 A / i 0 2: H H 5 A :E E NH . 1 [ E R AK O 7 F G 5 00o02 M S7 80 s88 P I e . T T2 C-T r AY u S TA F 22222 T85 99999 g AD E D i - i F N _ I " L . 0 _ 0 1 T O E S TT T - E SSD 0 _ T PP EEEE ,0 _ G MMMSTT UUEAEEA S PPTEVVE 6 .. R I P I SRLLR PYCE AA LLS VVEC
~
U B ATDTTD 1EA L AA L . S 1D D EEH VEEEEHH V TPYEYYE RPALAAL AOCRCCR E TTEZEEZ SSDPDDP ' . E N 4 5 3 0 06560 1 3 024::::.6033 6 001 11 O L LLLLL WWWWW p C O BBBBB 0 0 0: O 0 0 0 0 0 0 0 0 7 0 6 g 0 4 0 3 0 2 0 1 go"hm y eII ) u$mOwQ wk*
, i '
0 1 0: 1 4
- 2 N
O I T A gi C O L 0 T i 8 0: A 1 K O O B N G O L O I R G O T L O T i A RW A E & P B OC CI
]
M L W O B F9 8P 0 M : I 9U
/
H 6 T/T: E 2 A
,0 :
- 2 H E L
7 ANH e REK 1 [ E E U L O P u r q TT3AAY F O c G 0000000 I S8766666 P 00 6 e M I T P C U O OM C R n O E i F STA AD 0000000 599 999 99 00 99 M H R T E M E D i HO T P T OO T N I t T B L ' 0 3 1
^ T oJR oW E O T
D P PP sF i1sB 0 0 G E S AP MMM U UU mLA wG 6 R E F MTUSI P PP I P PP EPELLLOOOn1 l F FNF C1 F S0 U P iF ILL D1 T A A8PPPoO 1 1 MMMT NA EA EH EDDDUUUP S L V1 ET S PT EEPPPuT T I RDPRRPPPeH l I U RAAOAA L(LcD A A E ZTOTTT8cA PSLSSS1 t 1 u1 E 030$478 9s N 01: :31 10 3 4 90000 0001 1 1 2 0s 2a 0 O L L W LWLLLL Lt C B OBOOOO OO 0 0 O
- 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 6 4 3 2 1
" % w HgIzuEId muuEOoO
)
t tt t ub'
r- 8 r- < (' a . Z ERRRRRRESSR E 9 8
*sssss8EEEE !*
O E lW O i - J mE gp j( 8 s~ E5 l<- esDSSsgikk gl* Z saisi9 si s 9 l$!!!!E!! assamar sa - p< enessesan: y i O sesseeee== { - g 5 hs 8888888888 g h 0 V ng u 8 lg
'g [$$
H s, E8 e .,, EREEE 88S gB F .- y co s t
- t gg "g 5 _ gg Z b. 55 85 a 192 W * *5 O
[ III "8 - s -!
- EEE -
-) 555 alkk l as -
~
r CO ll9 w Ba!$asll@@ilma
- s 1)
W -nasseau b8 Z **susnass 888fsin888 8 g O ' ' ' ' ' 8 8 8 8 8, 8 a 8
~
~ w w -
[ d. } IISHN3BHyd S338030 7-27
l l u 1 0 1 0 1 4: 2 N G 0000000 I S5555 889 P 3333333 O I C n 00000 4 4 4 4 6 91 22222 23 05 T A T sS G E ST C " 0 0
)OVS P
CBS T O 1 O T n1 M D&R NAO L "0 5 Ot F CF P 0 P P1 PE U ,0: T M MLM(O U UE U S NTA 8 K A P P EV PPWE P l P LLRH ILI PRD COHS 1 WC 1 O O O G N C CZA8OH 1 D EEE 1 P DDODOD t 1 EENP E L O L S O I TPCTT RPURR AODAAGO TTETTET AP R G O T O L T A SSMSSBS R W E & A 0051 25201 51 336 C" B-L W C L 33571 1 3 } O B 1 1 1 1 222 I 9 M F/P U 8 LLWLLLL L M x I 1 OOBOOOO 0 H 8 T/T ,0: 1 E 2 A 2 H L P ANH 7 [
- E E L OU 1
e G 5 0000000 E r u REKA O F c I S2 5555555 P 3 3 33 3 3 3 3 M I T P C U O OM g TT5 AY n 0 0000000 COE R STA 4 4 4 4 4 4 4 i F T422222222 MH R T AD E M E D , HO T P T OO T N I n T B L E e O E R E v L A ,0: 0 G Z P N P PP PPPv OCCCCCCA v 6 R R D 2t21 2 R RR iAnn P R1 A88AAsS U 1 D D D D rD:0DZ 1 f 1 iR EEe S EEEE RTPT ARPR EAOA rRPt tAOC TPs E LTTT CSSSSSSC i TTv , E 956353G0 04 4001 : i5 N 0t : 1 668es 0O0000C0 O LLLLLLtw OOOOOOOe t 0 C - 0: 0 O 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 6 5 4 3 2 1
.. $@Ex6 n
.. u* . ( s Ym
{l jil1 l{ il
q l' 0 1 0: 1 4 2 o055550599 N S0666B5 P0T77T8222 566 1 1 1 OI
= F6 6300 0000 1 1244 6333 4 44444555 T
A D , C O
/ _
P R
)
C ORS H UOs T E TN Pt H C) A EE AIT (e - L ' R U NDO r 0 G O ONSNT ,0: L T PPP P PST FOROU ( ECA 8 W A ~ CCCCCR RRRRREEW 2 2t22 ANN A H.O E O 1 8 L W -. 8 8AAA 1 T EY DC B N 1 1 1 DDDDDHMUL tEEEERf 1 HP TA O I rTTPTZ eRRPRPSSS vAAOALR iTTTTLZOB P TV 1 0/ SSSSSAPHT T 2 0 1 A 59453641 5 44 : 03520.1 3 1
/
A
/
C 1 345567333 / 1 1 1 1 1 1 222 ] P I M O 9 L O F/P U I 8 2 1 LLLLLWELE OOOOOBOOO 0 M H T 1 4
,0:
T/T 2 A 2 H [ E R 9 ANH
- E
- o000005001 1 1
E K U OS OO 7 e RKA OF E S 99999955 66 P 3333334 4 77 M I T O L G C O N r TT6 = 557F215566 52222344 4 1 L E u AY 3333333S41 4 S S i g STA AD T R FOO T T D F E T
, A A R R E E P P NI )TT7 PE72 S
E 7 T 1 O. L E OO 4 L CI 2394 RA3 23 0 RTM2TP E nxAOOA aPnTPL F 0 HEU F LUO ,0: G V T(TPET' ( P V AUVAKU P C SETEENTP SHALHAACC P 6 E L OU E P L R ( R 2 MS C HZCD 8DRSP ROSA E RSTERR H2 2A P C U O U . 1 DEE E CED TD RS LRDD ELPNAEF OC 1 EE 1 OM C OE R S T P PA R A M OG ORGAO TOTECTOETT E P R P EA RP NTP MH R T E M E g' SCSBI SCBSS N , HO T P T T E 895035259m 5 5 1, 3424 4: 5 OO T B N 1 12334 6801 000000001 1 O LLLLWLELLL OOOOBOOOOO L C - 0 0: O 0 Q 0 0 0 0 0 0 Q M 4 0 3 0 2 0 1 7
- 5 y e I <u. g y e 0 o
yt* _
\!fjl1l lf' vIl li l[i(1I ( 1i I
1 0 0: 1 4 2 5 G55555 N - 66666 S1 1 1 1 1 P22222 6 1 2 O I C n 44 333 22333 V$5555 2 3 5 T _ E A L . P U C -_ O C O O - T P NR OE M L 2 P R DH E T NR 0 T - O F P L TPUP RO UFN 0 8: A . C EYYCT C R VAARA R ERRATA 2G2 1 2LPP N d A 1 2R S S,D DP DS 1E S1 VD O I E DF T R EUTPKR ARCAOAA TAEETET EREEPNE T T SVSH$LS , A - 1 574 0 22554 9 4: C I 9 C 1 41 4 56 1 11: 2 1 2 ] LLL F 8
/P LWWWL L I 3 U OBB9O O 0 1
H T 2A W
/T ,0:
0 H 1
. A :NH E 555 55555 G666 66666 2
1 [ R O E - 7 F S1 1 1 1 1 1 1 1 K P 22222222 u e r T AT7 AY C R 222 22257 3 33 33 322 Y555 5$555 i g S TAAD T F E D T S E , N I EF U S O L . L C E E V L A 0 G .,=_ Y A RP POP V ST C CC C C C L PAPP P
,0 6
R - R SR stRTR 2M18 I M2A R i C1 2A 2 AA _ U - P A DETD 1 DEOEEE RDDD 1 1 S _ E _ TSE PT TPT P RNP RCPT A AAONOP T ARTETTT _ E SMTSRsSS , E 0331: 98 1 0292 3324 N _ 3 4 568t33 00000t 1 1 O _ LL L WWLLWLLL BBOOBOOO C 0 0: O 0 0 0 0 0 0 0 0 0 _ 0 0 0 _ 7 5 4 2 1 l w hWyEz4
- gggmO ymo l
0 0 3 y % s e q ., U Ft u
'5 n 7i 2 m '-
( O
>4 MT A s A e E R A t T E Fu 0 SN E G
C )% ' , d, e 0' 5m 2 3 in} 1 - 52 Rs Z' 1 t a >1 P 1 u la S-1 S n '" 7i o " v e s e ' T- Tta "" e r . Fu t 1 1c " eS '5 n i F 1o 8 1 L w7 - 2m 2 p T- 9 ) u1 >2 7 t 1 T a 0 1 ed ' 0 1 h n 2 - r a 1 uS o - T 1 h1 1 ( 6 4 1 T- e 1 1 c eh t n h i e r t f w f e on f i D s 7
- rse e Sh a k e r
d 0 u t u p u E T 7. 0 t at I 5 1 t a i t O 3 a r e n a g e . N1 D p a H m e M g T t ni r n . 1 1 e r u . 7 e f D . ef n r i . uD o g i t 0 iF e v a c I 0 1 oi f b i t A ar nt i oS t f a r o u D
. - - - - - 0 t
5 0 0 0 0 0 0 0 7 5 4 3 2 1
, o h35 2N oge e$$ o i y 8i ii 8
~L-
- 8. SUMMfo.Y AND CONCLUSIONS This report describes the B&W Owners Group program for addressing the surgeline thermal stratification and thermal striping issue and presents the results of the preliminary work done to justify continued operation until the final program results are available. The Owners group plan is the same as was presented to the Nuclear Regulatory Commission staff on September 29, 1988 and April 7,1989. It consists of three parts: bounding calculations to justify near term continued operation, a measurement program to quantify the phenomenon, and the final analysis using the plant data.
The results to-date show that near term safe plant operation is assured. The measurement program results are currently being assessed. When this is completed, and the work on thermal striping is finalized, the entire program will be in place. The striping analysis to comply with the requirements of NRC Bulletin 88-11 Item 1.b is expected to be completed by October 31, 1989. Preliminary results from the striping evaluation show that the fatigue impact on the surgeline is estimated to be approximately 0.10 of the usage factor. The evaluation shows that temperature differentials in the surge line with the plant at power are not large enough to affect the usage factor; those existing during the early parts of plant heatup are most significant. Based on the interim results contained herein, it is concluded that the domestic B&W plants can continue operating safely in the near term until the final analyses are in place. Davis-Besse can be expected to operate for 3-1/2 to 7 more years without exceeding the ASME code limits while the oldest lowered loop plants can operate for an additional 5 years without exceeding the ASME code limits with the exception that the cyclic strain hardened yield strength was substituted for the virgin yield strength in the fatigue evaluation (see Section 5.1). 8-1
- 9. REFERENCES
-1. Kasza, K.E., Kuzay, T.M. , and Oras, J.J. , " Overview of Thermal-Buoyancy-Induced Phenomena in Reactor Plant Components", Proceedings of the Third International Conference on Liquid Metal Engineering and Technology, Oxford, England, Vol. 1, pp. 187-194, April 1984.
- 2. (a) B&W Drawing No. 13191016 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 131911E9 - Reactor Coolant Piping Assembly Elevation (c) B&W Drawing No. 131912E7 - Assembly and Details for 10" Surgeline
- 3. (a) B&W Drawing No. 146615E12 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 146616E4 - Reactor Coolant Piping Assembly Elevation (c) B&W Drawing No. 146617E5 - Assembly and Details for 10" Surgeline
~
- 4. (a) B&W Drawing No. 141583E11 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 141584E8 - Reactor Coolant Piping Assembly Elevation (c) B&W Drawing No. 141585E4 - Assembly and Details for 10" Surgeline
- 5. (a) B&W Drawing No. 131982E10 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 131983E6 - Reactor Coolant Piping Assembly Elevation (c) B&W Drawing No. 131984E2 - Assembly and Details for 10" Surgeline
- 6. (a) B&W Drawing No. 150142E13 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 150143E6 - Reactor Coolant Piping Assembly Elevation (c) B&W Drawing No. 150144E4 - Assembly and Details for 10" Surgeline
- 7. (a) B&W Drawing No. 143493E13 - Reactor Coolant Piping Assembly Plan (b) B&W Drawing No. 143492E10 - Reactor Coolant Piping Assembly Eleva-tion (c) B&W Drawing No. 143503E6 - Assembly and Details for 10" Surgeline 9-1
- 8. (a) Reactor Coolant Piping Assembly Plan View - 152055E7 (b) Reactor Coolant Piping Assembly Elevation View - 152056E4 (c) B&W Drawing No. 152030E8 - Assembly and Details for 10" Surgeline
- 9. Plant Technical Specifications for Crystal River Unit 3, Amendment 95, Docket No. 50-302.
- 10. B&W Dwg.149767E7, Heater Bundle Details.
- 11. B&W Dwg. 25482F3, Pressurizer General Arrangement.
- 12. H. Watanabe, " Boiling Water Reactor Feedwater Nozzle /Sparger Final Report," NED0-21821-A (February 80).
- 13. W.S. Woodward, " Fatigue of LMFBR Piping Due to Flow Stratification,"
ASME Pressure Vessel and Piping Conference, Portland, Paper 83-PVP-59, CONF-830607-27 (June 83).
- 14. T. Fujimoto, K. Swada, K. Uragami, A. Tsuge, and K. Hanzawa, "Experimen-tal Study of Striping at the Interface of Thermal Stratification,"
Thermal Hydraulics in Nuclear Technoloov, K.H. Sun et al (ed), ASME (81).
- 15. K.E. Kasza, J.P. Bobis, " Thermal-Transient Induced Pipe Stratification,."
ANS, 11, 675-676 (80).
- 16. K.E. Kasza, J.P. Bobis, W.P. Lawrence, and T.M. Kuzay, " Heat Exchanger Thermal-Buoyancy Effects: Design and Performance Comments (Phase I),"
ANL-CT-81-31 (April 82).
- 17. J.J. Oras and K.E. Kasza, " Thermal Transient Induced Buoyant Flow Channeling in a Vertical Steam Generator Tube Bundle," ANL-83-109 (Oct 83).
- 18. K.E. Kasza, T.M. Kuzay, and J.J. Oras, " Overview of Thermal-Buoyancy-Induced Phenomena in Reactor Plant Components," Proc 3rd Intl Conf-Oxford, pp. 187-194, (April 84).
- 19. K.E. Kasza, J.P. Bobis, W.P. Lawrence to J.C. Liljegren, " Thermal Transient Induced Pipe Flow Stratification Phenomena and Correlations (Phase II)," ANL-CT-81-19 (Feb 81).
9-2
l l- l
- 20. K.E. Kasza and T.M..Kuzay " Thermal Transient Induced Pipe and Elbow Flow L Stratification Phenomena and Correlations (Phase III)," ANL-82-85 (Oct 82).
- 21. T.M. Kuzay and K.E. Kasza, " Thermal Oscillations Downstream of an Elbow in Stratified Pipe Flow," FBR Thermal Hydraulics, Trans ANS, 45,794-796 (June 84).
- 22. T.M. Kuzay and K.E. Kasza, " Experiments and Analysis of ' a Horizontal Pipe Elbow in Stratified Pipe Flow," Liouid Metal Thermal Hydraulics, pp 459-460.
- 23. T.M. Kuzay and K.E. Kasza, " Thermal Striping Downstream of a Horizontal Elbow Under Thermally Stratified Flow Conditions," Joint Mto. ANS aD.d AIF. Washinoton, CONF-841105-10 (Nov. 84).
- 24. K.E. Kasza, J.J. Oras and R. Kolman, " Measurement of Velocity Profiles in a Stratified Pipe Flow Recirculating Shear Zone using Laser Flow Visualization" Liouid Metal Reactor Thermal Hydraulics, pp. 458-460.
- 25. T.M. Kuzay and K.E. Kasza, " Resolution of Thermal Striping Issue Down-stream of a Horizontal Pipe Elbow in Stratified Pipe Flow," ANS Annual .
Meetino. Boston, CONF-850610-4 (June 85).
- 26. L. Wolf, K. Fischer, W. Hafner, and W. Baumann, U. Schygulla, and K-H Scholl, " Overview of HDR Large Scale PTS Thermal Mixing Experiments and Analyses with 3-D Codes and Engineering Models," Trans 8th SMIRT, },' pp.
359-365 (85).
- 27. L. Wolf, U. Schygulla, W. Hafner, K. Fischer, and W. Baumann, "Results of Thermal Mixing Tests at HDR-Facility and Comparisons with Best-Estimate and Simple Codes," Trans 8th SMIRT, E, BE.1-9 (85).
- 28. L. Wolf, U. Schygulla, F. Gorner, and G.E. Neubrech, " Thermal Mixing Processes and RPV Wall Loads for HPI-Emergency Core Cooling Experiments in the HDR-Pressure Vessel," EfD, 25, pp. 337-362 (Oct. 86). !
- 29. L. Wolf, U. Schygulla, W. Hafner, K. Fischer, W. Baumann, and T.G.
Theofanous, " Application of Engineering and Multi-Dimensional, Finite 9-3 _________________-___-_-____--________a
. Difference Codes to HDR Thermal Mixing Experiments TEMB," Proc 14th WRSRIM, NUREG/CP-0082 1, pp. 396-416 (Feb. 87).
'30. L. Wol f, . W. Hafner, U. Schygull a, W. Baumann, and W. Schne11 hammer,
" Experimental and Analytical Results for HDR-TEMB Thermal Mixing Tests for- Different HPI-Nozzle Geometrics," Trans 9th SMIRT, g, pp. 319 324 (87).
- 31. U. Schygulla, E. Hansjosten, H.J. Bader, and K. Jansen, " Assessment of Heat' Transfer and Fluid Dynamics in Cold Leg and Downcomer of HDR-TEMB Experiments," Trans 9th SMIRT, g, pp. 313-317 (87).
- 32. W. Hafner, K. Fischer, and L. Wolf, " Computations of the HDR Thermal Mixing Experiments and Analysis of Mixing Phenomena," Trans 9th SMIRT, g, pp. 301-312 (87).
- 33. L. Wolf, W. Hafner, K. Fischer, U. Schygulla, and W. Baumann, "Applica-tion of Engineering and Multi-Dimensional Finite Difference Codes to HDR Thermal Mixing Experiments TEMB," FlQ,193, pp.137-165 (June 88).
- 34. L. Wolf and U. Schygulla, " Experimental Results of HDR-TEMR Thermal Stratification Test in Horizontal Feedwater Lines," Trans 9th SMIRT, D, pp. 361-366 (87).
- 35. L. Wolf, U. Schygulla, M. Geiss and E. Hansjosten, " Thermal Stratifica-tion Tests in . Horizontal Feedwater Pipelines," Proc 15th WRSRIM, NUREG/CP-0091, 1, pp. 437-464 (Feb 88).
-36. Investigation and Evaluation of Cracking Incidents in Piping in Pres-surized Water Reactors," NUREG-0691 (Sept 80). 9-4
E: . 1
.i
- 10. DOCUMENT SIGNATURES This document has been prepared by:
J.(/t. Glouder6ans R. J. Gurdal O / This report has been reviewed for technical content and accuracy. W. D. Maxham Materials & Structural Analysis Unit
.h C. W. Tall (y Performanc H nalysis Unit Verification of independent review.
<!l : m m R. J) Schomaker, Manager Performance Analysis Unit l
A. D. McKim, Manager Materials & Structural Analysis Unit This report has been approved for release. F. R. Burke d A Program Manager 10-1
APPENDIX A Oconee Unit 1 Surge Line Measurement Program l 1 A-1
0
. )
1
- 1. PURPOSE
-{
The-purpose of the test program was to collect data on the pressurizer surge .q line during plant heatup, power operation, and plant cooldown in order to determine:
- a. The magnitude and extent of thermal stratification in the pressurizer surge line, and b.- The magnitude and direction of surge line displacement.
The thermocouple data will be used to evaluate the need for redefinition of the design bases for the surge line and additional stress and fatigue analysis and/or modifications to operating procedures to minimize stratifica-tion effects. The displacement data will be ured to confirm analytical predictions of the surge line displacements. F<illowing this validation, computer simulations will be used to resolve concerns about closure.of pipe whip restraint gaps and snubber travel.
- 2. TEST PLAN The program was oriented toward maximizing the recording of data throughout the entire plant evolution taking the plant from a cold, depressurized condition to hot full power and back down to cold conditions. The recording of surge line sensor data began prior to initial energizing of the pres-surizer heaters and continued through power escalation. Data will be recorded during upsets and cooldown to cold shutdown. No alterations to the plant's normal startup procedures were made because of this data recording program. During periods of steady operation, such as zero power physics testing,. the data collection system was .re-configured to record selected thermocouple at higher scan rates (from 0.6 to 1.2 seconds) than the configuration allowed with all temperature and displacement sensors being recorded. These selected data acquisition periods, called " striping runs",
optimized the system's ability to detect temperature oscillations at the exterior of the surge line wall. The data acquisition included major plant parameters to enable correlation of plant conditions to the surge line conditions, especially with regard to transients that occur in the surge line. Data was continuously recorded at a sample rate of 20 seconds for the entire data acquisition period. Short A-2
b pa'uses- in the data. collection process to enable' downloading of data were ! acceptable,.but were done to the extent possible during steady state periods.
- 3. DATA'RE0UIREMENTS 1'
3.1. Measurement Sensors The locations of thermocouple along the surge line pipe are shown on Figure 6.1. A total of 54 thermocouple were-installed as specified in the orienta-tion key at each of the locations shown. The number and distribution of thermocouple are set to observe the temperature distribution in a plane perpendicular to the pipe length with emphasis on the bend closest to the pressurizer and the horizontal sections of the line. The location of displacement sensors along the surge line are shown in Figure A1. In order to provide displacement measurements in all directions at each of the locations shown, a total of 25 sensors are installed. Two additional sensors at locations 6Y and 10Y are provided for redundancy and signal comparison between sensor types. The number and distribution of displacement meters along the surge line were set to observe the amount and direction of pipe displacement during plant heatup and cooldown. 3.2. System Data Numerous reactor coolant parameters were recorded simultaneously during the data acquisition period. These included reactor coolant loop temperatures (hot and cold legs), pressurizer temperature, level, and pressure, reactor coolant pump status, and selected balance of plant parameters. This data will be correlated with changes in surge line conditions to identify what plant evolutions have significant impacts on stress conditions in the surge line. 3.3. Preliminary Results A preliminary evaluation of the Oconee data has shown that the typical range of surge line temperatures with the plant at 100% power is 490 to 600 F with only one location registering temperatures as low as 490 F. All other locations show minimum values of 530 to 540 F. The evaluation of the data will explore the validity and significance of this deviation. In the meantime, a traditional fatigue assessment has shown that for the surge line i A-3
material (stainless 316) that a zero-to-peak temperature fluctuation of 450F will result in an alternating stress equal to the endurance limit from the ASME design high-cycle fatigue curve. This result is derived using the relationship - 1.43E T and an endurance limit of 16,500 psi at 10"- cycles. This temperature fluctuation is equivalent to a peak-to-peak temperature difference of 900F. Therefore, if thermal striping is an operative phenome-
. non at power, the maximum temperature differential in the surge line cannot cause the endurance limit of the material to be exceeded.
A more detailed evaluation of the data, as described in Subsection 7.2, is expected to show that the maximum temperature oscillation at any one point on the surge line inside surface is significantly less than the top-to-bottom temperature differential. During the vast majority of the plant's life, the reactor coolant system will either be at power conditions or at cold shut-down. At cold shutdown there are no thermal gradients in the surge line to contribute to a striping effect. At power, the temperature differential is stable and relativi ; small. These factors suggest that for the quiescent conditions which characterize the largest fraction of the surge line operat-ing life the thermal striping effect may be rather small. A-4
Y 3 in ' - 1
- fbi 0 1
3 - b b n wM h w i g e
/ i n
m. P S O P 8 5 5 5 5 1 L 0 t o m , H[ 3/ y h / I 4 n
- 1
+ D e
O R \ R M T ) E W MA L C g AR A U ? T D E E T N OS - T NA N SE G - 3 R (2 T S N T
+
O N E OME I T A I T C T A A N C PL E OS I L D M P Y U 1 S/ T I R # T C O R T D L E S T I V E T B N N L VRPO B U 1 I U A N A E E N R L T I E E ONI EG UR T N
- T S
E P e D S PI-U E AN V O LSP G G I C I F R O U S S, i n 7 R O' N 1 E C 1 Z P IR D R E ' U Z I S R ' S E R P U S S E R
%o i 6
1 n P
APPENDIX B Verification of the Bounding Fatigue Analyses by Using Oconee Unit 1 Temperature Measurements I 4 I B-1
J
.1 l
In. light . of the surge line temperature measurements - performed 'during the February 1989 Heatup'of Oconee Unit 1, it was decided to review the bounding fatigue analyses previously performed (see description in Subsections 5.1 and 5.2).
. First of all, the effect of the non-linear temperature- profile was compared '
to the assumed linear temperature profile. The non-linear temperature profile which corresponds to the maximum top to bottom . temperature' difference was evaluated (403F at the top and 123 F at the bottom, for a difference of. 280F). A finite element surge line model was built to determine an equivalent linear temperature profile. This surge line model was loaded respectively with a linear temperature profile varying from 403F at.the top to 123F at the bottom, and with the.actually measured non-linear temperature profile. It was found that the resulting rotation of the pipe cross-section due to the actual non-linear temperature profile is 25% higher than the one due to the linear temperature profile. In a subsequent step, the temperature variations measured during the Oconee Unit I heatup have been scanned and 28 thermal stratification cycles have been counted and retained. The 28 peaks range from a 280F to a 65F top to bottom t.emperature difference. These 28 thermal stratification cycles represent a good picture of the temperature variations to be expected during a plant heatup. Table B-1 gives an overview of the temperature differences assumed in the bounding fatigue analyses, compared to the ones measured during the February 1989 Oconee Unit I heatup. Table B-1. Too to Bottom Temperature Differences (Temperatures in F) Boundina Fatiaue Analyses Measurements at Oconee Unit 1 Oconee Unit 1 Davis-Besse" (February 1989) HEATUP: : 422 386 280 ) 330 309 250 330 309 250 330 309 240 301 220
+ 23 additional cycles with temperature differences rang-ing from 206F to 65F.
C00LDOWN: 306 309 No complete cooldown data available to date. 1 B-2
5 The comparative study described below has been performed. . I First, the peak stress ranges calculated in the bounding fatigue analyses for the maximum tdp to bottom temperature differences (422F and 386F respective-ly) are:
- 1. ' scaled down in accordance with the different top to bottom tempera-ture differences measured at Oconee Unit 1,
- 2. multiplied by 1.25 to reflect the increased rotation due to the non-linearity of the temperature profile,
- 3. added to the corresponding thermal striping peak stresses (described in Subsection 7.3.3).
An alternating stress range is then calculated for each measured top to bottom temperature difference, leading to an allowable number of cycles from the fatigue curves given in Appendix I of the Section III ASME Code. The heatup thermal stratificati,on usage factor results from the summation of the products " number of heatups times number of cycles per heatup" divided by the allowable number of cycles (for each measured top to bottom temperature difference).
.The. revised cumulative usage factor is the sum of the usage factors from:
- 1. heatup thermal stratification (see above),
- 2. cooldown thermal stratification (from the bounding analyses),
- 3. stress reports for all functional specification transients (from the bounding analyses),
- 4. thermal striping (as described in Subsection 7.3.3).
The above described fatigue evaluation has been performed for both Oconee Unit I and Davis-Besse. The fatigue results from the bounding analyses (Subsections 5.1 and 5.2) were found to envelope the fatigue results using the Oconee Unit I temperature measurements for the most critical locations. B-3 _ _ _ . - ____ __ -_________ - ______- _ _ _ ___ _ __ _ D
APPENDIX C Justification for Use of Cyclic Strain-Hardened Yield Strength l C-1 l _ . _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ -
Ef.
- 11. PURPOSE
.The purpose of this discussion .is to provide a technical justification for the use of twice the cyclically strain-hardened yield strength in place of the limit 3' S m specified in Section III of the ASME Boiler and Pressure Vessel Code.
- 2. SIGNIFICANCE OF 3Sm LIMIT In the design of pressure vessels, the applied loads frequently result in stresses that exceed the yield strength of the material. This is particular-ly true of stresses that arise due to the constraint of the material when subjected to a temperature gradient. Since a reasonably exact analysis of such non-linear cyclic stresses would be a formidable and expensive task even
~
with the analytical . tools available today, simplified methods have been developed to ensure adequate design margins. The authors of ASME Section III chose the method of elastic strain invariance as the basis for the design
. procedures in Section III. This method permits the use of calculations based on elastic material behavior if certain requirements are - satisfied. In Section III, this requirement is specified by limiting the range of primary
- plus secondary stress intensities to 3Sm. The basic idea is to ensure that, after a few cycles of limited plastic deformation, the structure " shakes down" to elastic action, i.e., after a few cycles, a residual stress pattern will develop' about which subsequent stress cycles behave linearly. The following description is taken from Ref. 1.
In the study of allowable secondary stresses, a calculated elastic stress range equal to twice the yield stress has a special significance. It determines the borderline between loads which, when repetitively applied, allow the structure to " shake down" to elastic action and loads which produce plastic action each time they are applied. The theory of limit design provides rigorous proof of this statement, but the validity of the concept can easily be visualized. Consider, for example, the outer fiber of a beam which is strained in tension to a strain valueE , somewhat 1 beyond the yield strain as shown in Fig. I by the path OAB. The calculated elastic stress would be S - Si = EEi. Since we are considering the case of a secondary stress, we shall assume that the nature of the loading is such as to cycle C-2
the strain from zero to El and back to zero, rather than cycling the stress from zero to S 1 , and back to zero. When the beam is returned to its unde-flected position, 0, the outer fiber has a residual compressive stress of magnitude S1-S. y On any subsequent loading, this residual compression must be removed before the stress goes into tension and thus the elastic range has been increased by the quantity S1 -S.y If S1 = 2Sy, the elastic range becomes 2S y , but if S 1 > 2Sy , the fiber yields in compression, as shown by EF in Fig.1(b) and all subsequent cye.es produce plastic strain. Therefore, 2S y is the maximum value of calculated secondary elastic stress range which will " shake down" to purely elastic action.
- 3. FATIGUE ANALYSIS WHEN 3Sm LIMIT IS EXCEEDED As explained in Paragraph II, a prerequisite for a valid fatigue analysis is satisfaction of the 3S m limit for the range of primary plus secondary stress intensities. As further discussed in Paragraph II, the limit on these stresses is (1) to ensure that the use of linear elastic analyses will yield reasonably accurate results even though the yield strength of the material may be exceeded locally and (2) to ensure that shakedown occurs, i.e., that after a few cycles of limited plastic deformation, the structure behaves linearly with no progressive distortion during each load cycle. If the range of primary plus secondary stress intensities is in fact exceeded, neither of these goals can be assumed to have been met and the procedure for calculating the usage factor must be modified.
From the structural viewpoint, the principal concern with cyclic stresses in the plastic region is that high values of strain concentration can occur which, if not properly accounted for, can lead to fatigue failure earlier than would be predicted from a purely elastic analysis. Since an accurate means of calculating these strain concentrations was not available to the authors of the Code rules twenty years ago, simplified methods (Simplified Elastic-Plastic Fatigue Analyses) were developed to account for any strain concentrations that might occur. Each of these methods involves the calcula-tion of a factor to be applied to the alternating stress before entering the fatigue curve. This factor represents a strain concentration factor and is applied in the fatigue analysis in a manner similar to a stress concentration factor. See, for example, Ref. 2, Para. NB-3653.6. C-3
I 1 While the simplified procedure defined in Section III is conservative and l easy to use, it includes simplifying assumptions with regard to the behavior of material in the inelastic range. In deriving the 3Sm limit, use was made of a linear-elastic / perfectly-plastic (i .e., horizontal) stress-strain diagram (see Fig. 1) and it was shown that elastic behavior is assured provided that the stress range does not exceed 2 Sy , where Sy is the static yield strength. The use of a horizontal stress-strain diagram above S y is conservative since the actual stress-strain curve for austenitic stainless steels exhibits strain harden-ing. More important, however, is the fact that austenitic stainless steels have a pronounced tendency to strain-harden under cyclic loading. The basic idea is that the strength of the material increases under cyclic plastic
. deformation so that the stress-strain curve shifts upward relative to the static curve. To take advantage of this effect, it would be reasonable to base the limit of elastic behavior on twice the cyclic strain-hardened yield strength (2Sb ) rather than on twice the static yield strength (2Sy). This is not a new concept, as evidenced by the fact that it was incorporated in an earlier pressure vessel design code, " Tentative Structural Design Basis For Reactor Pressure Vessels and Directly Associated Components", December 1958 (Ref. 3).
The following definition of the limit of elastic behavior is taken from Ref. 3, Appendix B.
"B.2.5 Limit of Elastic Behavior. The stress intensity Sb iS defined as the limit of elastic behavior and is used in conbetion with the fatigue diagram to determine the reduction in mean stress produced by plastic flow. For some materials the yield stress, Sy, is higher than the endurance limit, Se, In these cases Sb"S-For materials which strain harden by appreciable amounts, tNe endurance limit may be appreciably higher than the yield stress of the annealed material. Thus Sy is not the true limit of elastic behavior and the endurance limit is a more realistic value to take for S b. For this purpose, the endurance limit of polished speci-mens without safety factor is taken as the best estimate for S e b because the use of too low a value results in an unconservative estimate of the alleviating effects of the plastic flow."
l From this definition of the limit of elastic behavior, a reasonable approxi-mation to the cyclically strain-hardened yield strength (S )bmay be obtained C-4 ___-_- _ -__ _ _ _ _ _ __ _ L
by using the alternating stress to -failure at 106 cycles of polished spect-mens without safety factor. Since the value of S a at 106 cycles on the-design fatigue curve is equal to one-half the value of the actual failure S a (the design fatigue curve contains a factor of two on the failure stress at 106 cycles), the cyclic strain-hardened yield strength may be approximated by twice 'the value of Sa on the design curve at 106 cycles. From the design curve (Ref. 4, Fig. I-9.2.1), Sb = 2Sa - 2(26) = 52 ksi.. The limit of elastic behavior, 2 S b, would then be 2(52) - 104 ksi. Since the design fatigue curve is based on tests conducted at 70F, the value at 70F must be reduced by the ratio E550/Eyo to obtain the equivalent value at 550F. This results in a limit of elastic behavior at 550F of (25.55 X 10 6/28.3 X 106 ) X 104 = 93.9 ksi. Therefore, it can be concluded that the Code limit of 3Sm (" 58 ksi) could be replaced with the 2Sb limit (= 93.9 ksi) and still meet the
. stated intent of the limit, which is to ensure linear behavior after a few cycles.
Reference 5 summarizes the results of a study conducted for the ASME Coda Subgroup on Fatigue Strength. The purpose of the study was to review the fatigue design methods and curves and to make improvements where possible based on new technology and data. During the course of the study, it was necessary to make some assumption' concerning the cyclic yield strength of austenitic stainless steels. Based on a review of test data, a value of 44 ksi was chosen to represent the cyclic yield strength over the range 70F to 800F. This implies a limit of 'l elastic behavior equal to 2 X 44 = 88 ksi at 800F. To adjust this to 550F , the value 88 ksi is multiplied by the ratio E550/E800 - 25.55/24.1 - 1.06. This results in 88 x 1.06 - 93.2 ksi., which.is only 0.8% less than the value 93.9 ksi. derived above. A quantitative verification that the use of 2Sb as the elastic limit is reasonable can be provided by estimating the plastic strain per cycle. The Coffin-Manson equation (Ref. 6, page 412) relating plastic strain range per cycle to the number of cycles to failure (N) is hE p N1 /2 = C (Constant) C-5 Y
a < Assuming that the static tensile test consists of- one-quarter cycle.,and that L the total plastic. strain is the true strain at fracture, one obtains, Ln ( 100 ) , (3f4)1/2 -C 100-RA T where RA is the percent reduction in area at fracture. From Ref. 1, Fig 11, RA - 72.6%. Then , 100 C - (1/4)I/2 Ln ( ) 100-72.6 Using Eq. I with'N = 10 6 cycles, (106 )1/2 hep - 0.6473 AC p - 0.0006 The' quantity AE p is the plastic strain range per cycle based on an elasti-cally' calculated stress range of 93.9 ksi. This is less than one-third the plastic. strain used to define the 0.2% offset yield strength for steels (G-0.002). This result shows that the plastic strain per cycle based on the use of 2Sb as the elastic limit is inconsequential 'and that the use of linear elastic mechanics is reasonable as long as the stress range does not exceed 2Sb -
- 4. CONCLUSION Use of twice the strain-hardened yield strength in place of the limit 3Sm specified in ASME Section III, while less conservative than the Code value, satisfies the intent of the Code limit.
=5. REFERENCES l.
- 1. Criteria of The ASME Boiler and Pressure Vessel Code for Design by finalysis in Sections III and VIII, Division 2., The American Society of l
Mechanical Engineers,1969.
- 2. ASME Boiler and Pressure Vessel Code, Section III, Subsection NB,1983 i Edition.
- 3. Tentative Structural Design Basis for Reactor Pressure Vessels and Directly Associated Components (Pressurized, Water Cooled Systems), U.S.
Department of Commerce, Office of Technical Services,1958. C-6 l-1
m . l
- 4. ASME Boiler and Pressure Vessel Code, Section III, Appendices, 1983 Edition. !
- 5. C. E. Jaske, W. J. O'Donnell, " Fatigue Design Criteria for Pressure Vessel Alloys", Journal of Pressure Vessel Technology, November 1977. ;
I
- 6. G. E. Dieter, " Mechanical Metallurgy", 2nd Edition, McGraw-Hill Book Co., i 1976.
i C-7
ei Si l
/l 1 -
I S ---~~ , i I ll
/i / i 1
l I l l l 1 l . y i Sy -- d ' I s Sy ..A# : !p n i 1 e i
. ] l . I E I l 2 1 y i I w I
- I E I
- I a 1 i i l' 1 I O 0 ty to tg STRAIN, t $T AAIN.C C
(a) Sy < 5:< 25y -Sy (b)5:>25y E F
. STRAIN HISTORY BEYOND YlELD Figure C.1 C-8 1
_. _ _ - _ _ _ _ - - - - . _ _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ l}}