ML17262A643

Nonproprietary Structural Evaluation of Re Ginna Pressurizer Surge Line,Considering Effects of Thermal Stratification.

ML17262A643
Person / Time
Site:	Ginna
Issue date:	05/31/1991
From:	Mel Gray, Palusamy S, Vora V WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17262A641	List: ML17262A640 ML17262A642 ML17262A643
References
WCAP-12929, NUDOCS 9111070191
Download: ML17262A643 (116)

Text

HCAP-12929 Structural Evaluation of Robert E. Ginna Pressurizer Surge Line, Considering the -Effects of Thermal Stratification Hay, 1991

~ 0 ~ ~

g 9111070i91 PDR ADOCV

Dli

HCAP-12929 Hestinghouse Proprietary Class 3 Structural Evaluation of Robert E. Ginna Pressurizer Surge Line, Considering the Effects of Thermal Stratification May, 1991 P. L. Strauch T. H. Liu C. B. Bond L. M. Valasek C. L. Boggess S. Tandon Verified by: Verified by:

M. . Gray V. V. Vora I

Approved by: Approved by:

S. S. alusamy, Manager R. . Patel, Manager Diagnostics and Monitoring System Structural Analysis Technology and Development Hork Performed under Shop Order RYNP-964 and RYNP-145 HESTINGHOUSE ELECTRIC CORPORATION Nuclear and Advanced Technology Division P.O. Box 2728 Pittsburgh, Pennsylvania 15230-2728 o 1991 Hestinghouse Electric Corp.

5429s/091

TABLE OF CONTENTS

~)~in Executive Summary 1.0 Background and Introduction 1.1 Background 1.2 , Description of Surge Line Stratification 1.3 Scope of Hork'.0 Surge Line Transient and Temperature Profile Development 2.1 General Approach 2.2 System Design Information 2.3 Development of Normal and Upset Transients 2.4 Monitoring Results and Operator Interviews 2.5 Historical Operation 2.6 Development of Heatup and Cooldown Transients 2'. 7 Axial Stratification Profile Development 2.8 Striping Transients 3.0 Stress Analysis 3.1 Surge Line Layout 3.2 Piping System Global Structural Analysis 3.3 Local Stresses - Methodology and Results 3.4 Total Stress from Global and Local Analysis 3.5 Thermal Striping 4.0 Displacements at Support Locations 5429s/091191 10

TABLE OF CONTENTS (Continued) ly

~f ign Ti )~l 5.0 Fatigue Usage 5.1 Methodology 5.2 Fatigue Usage Factors 5;3 Fatigue Due to Thermal Striping 5.4 Fatigue Usage Results 6.0 Summary and Conclusions 7.0 References Appendix A Computer Codes

~

~

Appendix B USNRC Bulletin 88-11 Appendix C Transient Development Details 5429s/091191:10

EXECUTIVE'UMMARY Thermal stratification has been identified as a concern which can affect the structural integrity of piping systems in nuclear plants since 1979, when a was discovered in a PWR feedwater line. In the pressurizer surge line, 'eak stratification can result from the difference in densities between the hot leg water and.generall'y hotter pressurizer water., Stratification with large temperature differences can produce very high stresses, and this can lead to integrity concerns. Study of the surge line behavior has concluded that the largest temperature differences occur during certain modes of plant heatup and cooldown.

This report has been prepared to demonstrate compliance with the requirements.

of NRC Bulletin 88-1.1 for Robert E. Ginna. Prior to the issuance of the bulletin, the Westinghouse Owners Group had a program in place to investigate the issue, and recommend actions by member utilities. That program provided.

the technical basis for the plant specific transient development reported here for the Ginna plant.,

~

~

This transient development utilized a number of'ources, including plant operating procedures, surge line monitoring data, and historical records for the plant. This transient information was used as input to a structural and stress analysis of the surge line for the plant.

The results of the structural analysis, and the fatigue analysis which .

followed, showed that the Ginna surge line meets the stress limits and usage factor requirements of the ASME Code for the remainder of the licensed operation of the plant. The support displacements resulting from stratification have also been provided and it was verifi,ed that sufficient travel allowance exists in spring hanger (RCH-l), to allow free pipe movement at all thermal conditions (including maximum system bT 210'F case). The structural analysis which resulted in this conclusion is discussed in Sections 3 and 4.

This work has led to the conclusion that Robert E. Ginna is in full compliance with the requi rements of NRC Bulletin 88-11.

5429s/091191:10

SUMMARY

OF RESULTS, AND STATUS OF 88-11,QUALIFICATION r 'n Hi r Date of commercial operation 7/1/70 Years of water-solid heatups 21 Years of steam-bubble heatups 0 System delta T limit from procedure 200'F Number of exceedances Seven xim m r n F r R 1 Equation 12 stress/allowable (ksi) 46.2/52.9*

(ksi)

Fatigue usage/allowable 0.9/1.0 Pr s riz r ur N zzl R s 1 Maximum stress intensity range/allowable (ksi) 32.4/57.9 Fatigue usage/allowable .47/1.0 mainin A i n ili None f -ll R uirmnt All analysis requirements met

• Future condi tion 5429s /091191: 10 iv

0 SECTION

1.0 BACKGROUND

AND INTRODUCTION Robert E. Ginna is a two-loop pressurized water reactor, which began commercial operation in July, 1970. This report has been developed to provide the technical basis and results of a plant-specific structural evaluation for the effects of thermal stratification of the pressurizer surge line for the plant.

The operation of a pressurized water reactor requires the primary coolant loop to be water solid, and this is accomplished through a pressurizer vessel, connected to the loop by the pressurizer surge line. A typical two-loop arrangement is shown in Figure 1-1, with the surge line highlighted.

The pressurizer vessel contains steam and water at saturated conditions with the steam-water interface level typically between 25 and 60'4 of the volume, depending on the plant operating conditions. From the time the steam bubble .

is initially drawn during the heatup operation to hot standby conditions, the

~

level is maintained at approximately 25K. During power ascension, the level

~ ~ ~

~

~

is increased to approximately 60'L. The steam bubble provides a pressure

~

cushion effect in the event of sudden changes in Reactor Coolant System (RCS) mass inventory. Spray operation reduces system pressure by condensing some of the steam. Electric heaters, at the bottom of the pressurizer, may be energized to generate additional steam and increase RCS pressure.

As illustrated in Figure l-l, the bottom of the pressurizer vessel is connected to the hot leg of one of the coolant loops by the surge line, a 10 inch schedule 140 stainless steel pipe, most of which is horizontal.

1.1 ~Ba k<~ri~n During the period from 1982 to 1988, a number of utilities reported unexpected movement of the pressurizer surge line, as evidenced by crushed insulation, gap closures in the pipe whip restraints, and in some cases unusual snubber movement. ~ Investigation of this problem revealed that the movement was caused by thermal stratification in the surge line.

~ ~ ~

5429s/091691:10

e Thermal stratification had not been considered in the original design of

~

any pressurizer surge line, and was known to have been the cause of

~

service-induced cracking in feedwater line piping, first discovered in 1979.

~

Further instances of service-induced cracking from thermal stratifi cation surfaced in 1988, with a crack in a safety injection line, and a separate occurrence with a crack in a residual heat removal line. Each of the above incidents resulted in at least one through-wall crack, which was detected through leakage, and led to a plant shutdown. Although no through-wall cracks were found in surge lines, inservice inspections of one plant in the U.S. and another in Switzerland mistakenly claimed to have found sizeable cracks in the pressurizer surge line. Although both these findings were subsequently disproved, the previous history of stratified flow in other lines led the USNRC to issue Bulletin 88-11 in December of 1988. A copy of this bulletin is included as Appendix B.

The bulletin requested util,i ties to establish and implement a program to confirm the integrity of the pressurizer surge line. The program required

~

4

~

both visual inspection of the surge line and demonstration that the design

~ ~

~

requirements of the surge line are satisfied, including the consideration of

~ ~

strati fi cati on effects. ~

Prior to the issuance of NRC Bulletin 88-11, the Westinghouse Owners Group had implemented a program to address the issue of surge line stratification. A bounding evaluation was performed and presented to the NRC in April of 1989.

This evaluation compared all the WOG plants to those for which a detailed plant specific analysis had been performed. Since this evaluation was unable to demonstrate the full design life for all plants, a generic. justification for continued operation was developed for use by each of the WOG plants, the basis of which was documented in references [1] and [23.*

The Westinghouse Owners Group implemented a program for generic detailed analysis in June of 1989, and this program involved individual detailed analyses of groups of plants. This approach permitted a more realistic

• Number in brackets refer to references listed in, Section 7.

5429s/09169li 1-2

0 approach than could be obtained from a single bounding analysis for all plants, and the results were published in dune of 1990 [3].

The followup.to the Hestinghouse Owners Group Program is a demonstration of the applicability of reference [3] to each individual plant, and the performance of evaluations which could not be performe'd on a generic basis.

The goal of this report is to accomplish these followup actions, and to therefore complete the requirements of the NRC Bulletin 88-11 for Ginna.

1.2 0 ri i n f r Lin Th rm 1 r ifi i n It will be useful.to describe the phenomenon of stratification, before dealing with its effects. Thermal stratification in the pressurizer surge line is the direct result of the difference in densities between the pressurizer water and the generally cooler RCS hot leg water. The lighter pressurizer water tends to float on the cooler heavier hot leg water. The potential for stratification is in'creased as the difference in temperature between the pressurizer and the hot leg increases and as the insurge or outsurge flow rates decrease.

At power,. when the difference in temperature between the pressurizer and hot leg is relatively small, the extent and effects of stratification have been observed to be small. However, during certain modes of plant heatup and cooldown, this difference in system temperature could be large, in which case the effects of stratification are significant, and must be. accounted for.

Thermal stratification in the surge line causes two effects:

o Bending of the pipe is different than that predicted in the original design.

o Potentially reduced fatigue life of the piping due to the higher stress resulting from stratification and striping.

5429s/091691:10 1-3

f Nrk The primary purpose of this work was to develop transients applicable to the Ginna plant which include the effects of stratification and to evaluate the structural integri'ty of the surge line. This work will therefore complete the demonstration of c'ompliance with the requirements of NRC Bulletin 88-11.

The transients were developed following the same general approach originally established for the Hesti nghouse Owners Group. Conservati sms inherent in the original approach were refined through the use of monitoring results, plant operating procedures and operator interviews, and historical data on plant operation. This process is detailed in Section 2.

The resulting transients were used to perform an analysis of the surge line, wherein the existing support configuration was carefully modeled, and surge line di splacements, stresses and support loadings were determined. This analysis and its results are discussed in Sections 3 and 4.

The stresses were used to perform a fatigue analysis for the surge line, and the methodology and results of this work are discussed in Section 5. The summary and conclusions of this .worg are summarized in Section 6.

5429s/091691:10 1-4

Surge Line Figure l-l. Typical 2-Loop Plant Loop Layout 5429s/091691:10 1-5

SECTION 2;0 SURGE LINE TRANSIENT AND TEMPERATURE PROFILE DEVELOPMENT The transients for the pressurizer surge line were developed from a number of sources, including the most recent systems standard design transients. The heatup and cooldown transients, which involve the majority of the severe stratification occurrences, were developed from review of the plant operating procedures, operator intervi ews, monitoring data and historical records for the plant. The total number of heatup and cooldown events specified remains unchanged at 200 each, but a number of transi ent events have been defined to reflect stratification effects, as descri bed in more detai later.

1 The normal and upset transi ents, except for heatup and cooldown, for the Ginna surge line are provided in Table 2-1. for each of the transients, the surge line fluid temperature was modified from the original design assumption of uniform temperature to a stratified distribution, according to the predicted temperature differentials between the pressurizer and hot leg, as listed in the table. The transients have been characterized as either insurge/outsurges (I/O in the table) or fluctuations (F). Insurge/outsurge transients are generally more severe, because they result in the greatest temperature change in the top or bottom of the pipe. Typical temperature profiles for insurges and outsurges are shown in Figure 2-1.

e Transients identified as fluctuations (F) typically involve low surge flow rates and smaller temperature differences between the pressurizer and hot leg, so the resulting stratification stresses are much lower. This type of cycle is important to include in the analysis, but is generally not the major contributor to fatigue usage.

The development of transients which are applicable to Ginna was based on the work already accomplished under programs completed for the Westinghouse Owners Group [1,2,33. In this work all the Westinghouse plants were grouped based on the similarity of their response to'tratifi cation. The three most important 5429s/091691 10 2-1

P factors influencing the effects of stratification were found to be the structural layout, support configuration, and plant operation.

The transients developed here, and used in the structural analysis, have taken advantage of -the monitoying data collected during the HOG program, as well as operator interviews and historical operation data for the Ginna plant. Each of these will be discussed in the sections whi ch follow.

2.2 mD i nInfrm i n The thermal design transients for a typical Reactor Coolant System, including the press'urizer surge line, are defined in Hestinghouse Systems Standard Design Criteria.

The design transients for the surge line consist of two major categories:

(a) Heatup and Cooldown transients (b) Normal and Upset operation transients (by definition, the emergen'cy and faulted transients are not considered in the ASHE Section III fatigue life assessment of components).

In the evaluation of sur'ge line stratification, the transient events considered encompass the typical normal and upset design events defined in the FSAR.

The total number of heatup-cooldown cycles (200) remains unchanged. However, transient events and the associated number of occurrences (" Label", "Type" and "Cycle" columns of Tables 2-1 and 2-2) have been defined to reflect stratification effects, as described later.

5429s/091691:10 2-2

2.3

~ r ific ti

~

n Eff ri ri nd0v 1 mn f Nrml n

~T

>a,c,e 5429s/091691 2-3

>a,c,e

>a,c,e 2.4 M ni rin R ul nd r r In rvi w 2.4.1 Monitoring Monitoring information collected as part of the Hestinghouse Owners Group generic detailed analysi sl:33 was utilized in this .analysis . - This included information from Ginna. The moni toring programs used existing and installed temporary sensors on the surge line piping, as shown in Figure 2-2.

The pressurizer surge line monitoring programs .utilized externally mounted temperature sens'ors (resistance temperature detectors or thermocouples). The temperature sensors were attached to the outside surface of the. pipe at various circumferential and axial locations. In all cases these temperature sensors were securely clamped to the piping outer wall, taking care to properly insulate the area against heat loss due to thermal convection or radiation.

The Ginna surge line temperature sensor configuration consists of two to five sensors mounted ci rcumferenti ally on the pipe at various axial l.ocations as shown in Figure 2-2. The multiple axial locations give a good picture of how the top to bottom temperature distribution may vary along the longitudinal axis of the pipe. In addition, displacement sensors were mounted at various axial locations to detect vertical and horizontal movements, as shown in 5429s/091691:10 2-4

Figure 2-2. Typically, data were collected at [

~

~ interval s or less, during periods of high system delta T.

~ ~ ~

Existing plant instrumentation was used to record various system parameters.

These system parameters were useful in correlating plant actions with stratification in the'surge line. A list of typical plant parameters monitored is given below.

>a,c,e Data from the temporary sensors was stored on magnetic floppy disks and converted to hard copy time history plots with the use of common spreadsheet software. Data from existing plant instrumentation was obtained from the utility plant computer.

2.4.2 Operational Practices An operations interview was conducted at Ginna on October 10, 1989. Since the maximum temperature difference between the pressurizer and the reactor coolant loop occurs during the plant heatup and cooldown, operations during these events were the main topic of the interview. Figure 2-3 describes the heatup process, and Figure 2-4 is the corresponding plot for the cooldown process.

In both heatup and cooldown, the plant has an administrative limit of 200'F on temperature difference between pressurizer and reactor coolant system (" system delta T").

5429s/091691:10 2-5

25~

~

i ri

~

1 ra i n n

A review of historical records from the plant (operator logs, survei.llance, test reports, etc.) was performed. From this review, two pieces of information were extracted: a characteristic maximum system delta.T for each heatup and cooldown recorded, and the number of maximum delta T exceedances of 210'F (as explained later, 210'F is used as the maximum system delta T limit in developing transients).

The known data for heatups and cooldowns experienced to date and their associated system delta temperature are described below for the plant.

P Percentage of Number of Historical System hT Heatups h Cooldowns Heatup & Cooldown Range ('F) Experienced to Oate Occurrences a,c,e Total 84 100 In addition to the 84 known events considered, there were an additional 45 events for which system delta T could not be determined. It was assumed that the 84 known events provide a sufficient characterization of the plant's

" operation, and that the 45 unknowns would fall within the distribution'etermined from the knowns.

This information was used to ensure that the transients analyzed for Ginna encompassed the known prior operating history of the plant with respect to system delta T. Comparison of the above table to the numbers used in the evaluation, as seen in Figure 2-5, confirmed applicability to the plant. [

>a,c,e 5429s/091691'10 2-6

h 2.6 D v 1 m n f H a n 1 wn Tr n i n The heatup and cooldown transients used in the analysis were developed from a number of sources, as discussed in the overall approach. The transients were built upon the extensive work done for the Hestinghouse Owners Group [1,2,33, coupled with plant specific considerations for Ginna.

The transients were developed based on monitoring data, historical operation and operator interviews conducted at a large number of plants. For each monitoring location, the top-to-bottom differential temperature (pipe delta T) vs. time was recorded, along with the temperatures of the pressurizer and .hot leg during the same time period. The difference between the pressurizer and hot leg temperature was termed the system delta T.

From the pipe and system delta T information collected in the WOG[1,2,33 effort, individual plants'onitoring data was reduced to categorize stratification cycles (changes in relatively steady-state stratifi ed conditions) using the rai nflow cycle counting method. This method considers delta T range as opposed to absolute, values.

>a,c,e The resulting distributions (for I/O transients) were cycles in each RSS range above 0.3-, for -each mode (5,4,3 and 2). A separate distribution was determined for each plant at the reactor coolant loop nozzle- and a chosen critical pipe location (location having the greatest cyclic activity and delta T pipe value). Next, one distribution bounding the number of occurrences for the RSS range for each plant considered was developed for each mode of 5429s/091691:10 2-7

operation, for the

~ critical

~

pipe location and also for the nozzle location.

This bounding distribution was formed using the method described in detail in

~ ~ ~ ~

reference [3]. The premise of th'e method is that a least severe but still

~

conservative distribution should bound all cycle occurrences.

Transients, which are represented by delta T pipe with a corresponding number of cycles, were developed by combining the delta T system and cycle distributions. For mode 5, delta T system is represented by a historical distribution developed from a number of HOG plants (HOG distribution). Using data from a number of plants is beneficial, as the resulting transients are more representative of a complete spectrum of operation than might be obtained from only a few heatups and cooldowns. As discussed in Section 2.5, this historical system delta T distribution was shown to encompass prior operating history data for Ginna. for modes 4, 3 and 2, the delta T system was defined by one maximum value for each mode. The values were based on the maximum system delta T obtained from the monitored plants for each mode of operation.

To determine total transient cycles, an analysis was conducted to determine the average number of stratification cycles per cooldown relative to the average number of cycles per heatup. [

The transient cycles for all modes were then enveloped in ranges of AT . , i.e., all cycles transi ents within each hT pipe'rom i range were added and assigned to the pipe pre-defined ranges. These cycles were then applied in the fatigue analysis with the maximum hT i for each range. The values used are as follows:

pipe For Cycles Nithin Pipe Delta T Range Pipe Delta T

>a,c,e 5429s/091691:10

This grouping was done to simplify the fatigue analysis. *The actual number of cycles used in the analysis for the heatup and cooldown events is shown i' Table 2-2.

The final result of .this complex process is a table of transients correspond-ing to the subevents of the heatup and cooldown process. A mathematical description of the process is given in Appendix C. [

] ' The critical location is the "location with the highest combination of pipe delta T and number of stratification cycles.

Because of main coolant pipe flow effects, the stratification transient loadings at the reactor coolant hot-,leg nozzle are different. These transients have been applied to the main body of the nozzle as well as the pipe to hozzle gi rth butt weld.

Plant monitoring included sensors located near the nozzle to surge. line pipe weld. Based on the monitoring, a set of transients was developed for the nozzle region to reflect conditions when stratification could occur in the nozzle. The primary factor affecting these transients was the flow in the main coolant pipe. Significant stratification was noted only when the reactor coolant pump was not operating in the loop with the surge line. Transients were then developed using a conservative number of "pump trips."

7

' Therefore, fatigue analysis of the nozzle was performed using the "nozzle transients" and the "pipe transients."

This accounted for both the stratification loadings from the nozzle transients, and the pressure and bending loads from the piping transients.

5429s/09169la 2-9

0 The total transients for heatup and cooldown are identified as HCl thru HC9 for the pipe, and HCl thru HC9 for the nozzle as shown in Tables 2-2(a) and 2-2(b) respectively. Transients HC7 thru HC9 for the pipe, and MC7 and HC9 for the nozzle represent transients which occur during later stages of the heatup.

As indicated in Section 2.5, based on a review of the Ginna operating records, there were seven events in which the system delta T exceeded the transient basis upper limit of -[

>a,c,e 2.7 xi 1 r fi nPr fil Dv 1 mn In addition to transients, a profile of the [

>a,c,e Two types of profile envelope the stratified temperature di stri butions

.observed and predi cted to occur in the line. These two profiles are [

ja,c,e 5429s/091691:10 2-10

Low flow profiles are characterized by a non-linear top to bottom temperature

~

distribution in association with low fluid velocities. A typical low flow

~ ~ ~ ~ ~ ~ ~

profile is shown in Figure 2-6. Low flow profiles are a function of the

~ ~

density difference. between the two fluids and the flow rates of each. During low flow conditions the two fluids do not mix, because of the density difference, but prefer to separate with the heavier (colder) fluid filling the lower portions of the pipe. 'he interface, the point at which the two fluids meet, has a constant elevation along its entire length for steady state conditions. This characteristic is present because stratification is a gravity induced phenomenon.

>a,c,e These three configurations are illustrated in Figure 2-7. I:

>a,c,e 5429s/091691:10 2-11

Review and study of the monitoring data for all the plants revealed a consistent pattern of development of delta T as a function of distance from the hot leg intersection. This pattern was consistent throughout the

. heat-up/cooldown process, for a given plant geometry. This pattern was used along wi th plant operating procedures to provide a realistic yet'om'ewhat conservative portrayal of the pipe delta T along the surge line.

• The combination of the hot/cold interface and pipe delta T as functions of distance along the surge line forms a signature profile for each individual plant analyzed. I:

>a,c,e 28 ri in Trni n The transients developed for the evaluation of thermal striping are shown in Table 2-3.,

L

>a,c,e Striping transients use the labels HST and CST denoting striping transients (ST). Table 2-3 contains a summary of the HSTl to HST8 and CSTl to CST7 thermal stripi ng transients which are similar in. their defi ni tion of events to the heatup and cooldown transient definition.

These striping transi ents were developed during plant specifi c surge line evaluations and are considered to be a conservative representation of striping in the surge line[3]. Section 5 contains more information on specifically how

~

the striping loading was considered in the fatigue evaluation.

~ ~ ~

5429s/091691:10 2-12

TABLE 2-1 SURGE LINE TRANSIENTS HITH STRATIFICATION NORMAL AND UPSET TRANSIENT LIST TEMPERATURES ('F)

MAX 'OMINAL LABEL TYPE CYCLES hTSgpat PRZ T RCS T

>a,c,e 5429s/091691:10 2-13

0 TABLE 2-1 (Cont'd.)

SURGE LINE TRANSIENTS HITH STRATIFICATION NORMAL AND UPSET TRANSIENT LIST TEMPERATURES ('F)

MAX NOMINAL LABEL TYPE CYCLES hTStrat PRZ T RCS T

~a,c,e

3) Nominal pressurizer and RCS temperature used for thermal anchor motion only..

4) I/O insurge/outsurge; F fluctuation 5429s/091691:10 2-14

TABLE 2-2a SURGE LINE PIPE TRANSIENTS NITH STRATIFICATION HEATUP/COOLDOHN (HC) 200 CYCLES TOTAL TEMPERATURES ('F)

MAX =

NOMINAL LABEL RCS T NOMINAL'TStrat TYPE CYCLES PRZ T

>a,c,e I

Nominal temperature used for thermal anchor motion only 54294/09169I 2-15

TABLE 2-,2b SURGE LINE NOZZLE TRANSIENTS NITH STRATIFICATION HEATUP/COOLDOHN (HC) 200 CYCLES TOTAL TEMPERATURES ('F),.

MAX NOMINAL NOMINAL LABEL TYPE CYCLES. BTSgpgg 'RZ T RCS T 5429s/091691:10 2-16

TABLE 2-3 SURGE L'INE - TRANSIENTS STRIPING FOR HEATUP (H) and COOLDONN (C)

S429s/091691 10 2-17

a,c,e Figure 2-1. Typical Insurge-Outsurge (I/O) Temperature Profiles 5429s/091691: 10 2-18

a,c,e Figure 2-2. Monitoring Locations for Ginna

a,c,e Figure 2-3. Heatup Curve for Ginna

a,c,e Figure 2-4. Cooldown Curve for Ginna

a,c,e Figure 2-5. Summary of Historical Data Distribution from Ginna Compared to Heatup and Cooldown Used for Analysis

a,c,e Figure 2-6. Example Axial Stratification Profile for Low Flow Conditions

a,c,e Fi gure 2-7. Geometry Cons i derati ons 5429s/091691:10'-24

aye,e Figure 2-8. Temperature Profiles for Low Flow Conditions in the Ginna Su'rge Line

'3 SECTION 3.0 STRESS ANALYSES The flow diagram (Figure 3-1) describes the procedure,to determine the effects of thermal stratification on the pressurizer surge line based on transients developed in section 2.0. [

>a,c,e 3.1 3

The Ginna surge line layout is documented in reference [53 and is shown schematically in figure 3-2. The Ginna surge line contains no vertical rigid supports or pipe whip restraints, which usually cause high thermal loads due

, to contact resulting from stratification. The only support is a spring hanger (RCH-l), located at analysis node 1110. The spring hanger is inconsequential to thermal loads, however, its stiffness and location are considered in the model. Nhen the spring hanger bottomed-out condition exists, the spring stiffness becomes rigid. The surge line pipe is 10 inch schedule 140 stainless steel. .Experience wi th the analysis of thermal stratifi cation has indicated that surge line. layout [

ja,c,e 3.2 i in m L 1 tr ur 1 An 1 i The piping system was modeled by pipe, elbow, and linear and non-linear spring elements using the ANSYS computer code in Appendix A. The geometric and material parameters are included. [

3

' The spring hanger is modelled even though it is somewhat inconsequential for the thermal condition. The potential for the spring hanger exceeding its displacement tolerance was checked. The Ginna surge line 5429s/091691:10 3-1

design with the existing support configuration was analyzed for the normal thermal and thermal strati fi cation 1 oadings.

The hot-cold temperature interface along the length of a surge line [

]a,c,e

. Each thermal profile loading defined in section 2 was broken into [

, ] ' Table 3-1 shows the loading cases considered in the analysis. Hi thin each operation .the [-

] ' Consequently, all the thermal transient

~

loadi ngs defined in section

~ ~

2 could be evaluated. ~

The pressurizer and RCL temperature listed in Table 3 reflect the approximate system hT. System temperatures are used only to define the boundary displacements at both RCL and pressurizer nozzles.

In order to meet the ASHE Section III Code stress limits, global structural codex'ach models-of the surge line for existing and future support configurations (spring hanger bottomed out and not- bottomed out) were developed using the information provided by reference [5] and the ANSYS general purpose finite element computer model'as constructed using [

]a,c,e For the stratified condition, [

5429s/09169'l~. 3-2

)a,c,e The global piping stress .analyses were based on one structural model for the Ginna surge line. The model represents the exi sting support configuration with one vertical spring hanger (RCH-1). The results of the ANSYS global structural analysis provides the thermal expansion moments. The ASME Section III equation (12) stress intensity range was evaluated. For the past condition, system delta T of 275'F was evaluated. For the future, a system delta T 210'F was evaluated as discussed in Section 2.0. The maximum ASME equation (12) stress intensity range in the surge line occurred at the hot leg branch nozzle (46.2 ksi), and was found to be under the code allowable of 3Sm (52.9 ksi) for the future condition with system hT limit of 210'F. Maximum equation (12) and equation (13) stress intensity ranges are shown in Table 3-2.

The pressurizer nozzle loads from thermal stratification in the surge line were also evaluated according to the requirements of the ASME code. The-

~

evaluation using transients detailed in Reference [123 plus the moment loading

~ ~

from this analysis, included the calculations of primary plus secondary stress

~

~

~

intensities and the fatigue usage factors. The maximum stress intensity range is 32.4 ksi, compared to the code allowable value of 57.9 ksi . The maximum fatigue usage factor will be reported in Section 5. It was found that the Gi nna pressurizer nozzle met the code stress requirements.

3.3 L 1 r s-M th d 1 n R 1 3.3.1 Explanation of Local Stress Figure 3-3 depicts the local axial stress components in a beam with a sharply nonlinear metal temperature gradient. Local axial stresses develop due to the restraint of axial expansion or contraction. This restraint is provided by the material in the adjacent beam cross section. For a linear top-to-bottom temperature gradient, the local axial stress would not exi st. [ >a,c,e 5429s/091691:10 3-3

>a,c,e 3.3.2 Finite Element Model of Pipe for Local Stress A short description of the pipe finite element model is provided below. The model with thermal boundary conditions is shown in Figure 3-4. Due to symmetry of the geometry and thermal loading, only half of the cross section was required for modeling and analysis.

>a,c,e 3.3.3 Pipe Local Stress Results Figure 3-5 shows the temperature distributions through the pipe wall [

>a,c,e 5429s/091691 10 3-4

>a,c,e 3.3.4 RCL Hot Leg Nozzle Analysis A detailed surge line nozzle fi nite element model was developed to evaluate the effects of thermal stratification. The model is shown in Figure 3-9. [

] .' A summary of stresses in the RCL nozzle location 1 due to thermal stratification is given in Table 3-3.

3.4 T 1 r f 1 1 n 1 An 1 4

3

' In order to superimpose local and global stresses, several stress analyses were performed using the finite element model. l:

>a,c,e

>a,c,e 5429s/091691:10 3-5

3.5 3.5.1 Background

't the time when the feedwater line cracking problems in PNR's were;first discovered, it was postulated that thermal osci llations (striping) may significantly contribute to the fatigue cracking problems. These osci llations were thought to be due to either mixing of hot and cold fluid, or turbulence in the hot-to-cold stratification layer from strong buoyancy forces during low flow rate conditions. (See Figure 3-10 which shows the thermal stripi ng fluctuation in a pipe). Thermal striping was verified to occur during subsequent flow model tests. Results of the flow model tests were used to establish boundary conditions for the stratification analysis and to provide oscillation data for evaluating high cycle fatigue. 'triping Thermal striping was also examined during water model flow tests performed for the Liquid Metal Fast Breeder Reactor primary pipe loop. The stratified flow

~

was observed to have a dynamic interface, region which oscillated in a wave

~

pattern. ~ These dynamic oscillations were shown to produce significant fatigue

~

damage (primary crack initiation). The same interface oscillations were

~ ~ ~ ~ ~ ~

~

observed in experimental studies of thermal striping which were performed in Japan by Hitsubishi Heavy Industries. The thermal striping. evaluation process was discussed in detail in reference [3], and is also discussed in references

[7], [8], and [9].

3.5. 2 Thermal Stri ping Stresses Thermal striping stresses are a result of differences between the pipe inside surface wall and the average through-wall temperatures which occur with time, due to the oscillation of the hot and cold stratified boundary. (See Figure 3-11 which shows a typical temperature distribution through the pipe wall). [

]a,c,e 5429s/091691:10

1' The peak stress range and stress intensity

~ ~

was calculated from a 3-0 finite element analysis. ~ [

] ' The methods used to determine alternating stress intensity are defined in the ASME code. .Several locations were evaluated in order to determine the location where stress intensity was a maximum.

Stresses were intensified by K3 to account for the worst stress concentration for all piping elements in the surge line. The worst piping element was the butt weld.

>a,c,e 3.5.3 Factors Nhi ch Affect Striping Stress The factors which affect striping are di scussed briefly below:

)a,c,e 5429s/091691 10 3-7

e

>a,c,e

>a,c,e

>a,c,e 5429s/091694, 3-8

TABLE 3-1 TEMPERATURE DATA USED IN THE ANALYSIS Max Type of System Analysis Pressurizer RCL T T Pipe Operation hT('F) Cases, Temp ('F) Temp ('F) ('F) ('F) BT ('F)

)a,c,e 5429s/091691 10 3-9

TABLE 3-2 Summary of Ginna Surge Line Thermal Stratification Maximum Stress Results ra in ondi i n HE Euain ~pi~+ ~Fi~r* All w 1 (ksi) (ksi) (ksi) 12 59.6 46.2 52.90 13 '8.0 48.0 50.1

• Future represents the operating condition with maximum system BT 210'F-

+ Past represents the operating'ondition with maximum system AT 275'F 5429s/091691:10 3-10

TABLE 3-3 ,

GINNA SURGE LINE MAXIMUM LOCAL AXIAL'TRESS AT [ )a,c,e (10" - 140)

Local Axial Stress (psi)

Location Surface Maximum Tensile Maximum Compressive

>a,c,e ja,'c,e 5429s/091691:10 3-11

TABLE 3-4 STRIPING FREQUENCY AT 2 MAXIMUM LOCATIONS FROM 15 TEST RUNS 1

Total Frequency (HZ) Duration

¹ Cycles 4 Lgth. in Min (Duration) Max (Duration) Avg (Duration) Seconds

>a,c,e 5429s/091691 10 3-12

I a,c,e Figure 3-1. Schematic of Stress Analysis Procedure 5429s/091691:10 3-13

03o30'I 1000 RC. IL 5020 11442

~~5.ZSr 1070

'CU-I 53'-IW QOlD 11'50 17 QS' 5150

• 517 R 1210 IC x Id'EDUCER 1.0LF 17.00'.1S7 R 5270 Figure 3-2. Pressurizer Surge Line Layout: Ginna

a,c,e Figure 3-3. Local Axial Stress in Piping Due to Thermal Stratification 5429s/091691 '10 3-15

a,c,e Figure 3-4. Piping Local Stress Model and Thermal Boundary Conditions 5429s/091691:10 3-16

a,c,e Figure 3-5. Surge Line Temperature Distribution at [ ' ' Axial Locations 5429s/091691:10 3-17

a,c,e'igure 3-6. Surge Line Local Axial Stress Distribution at [

Axial Locations 5429s/091691[." 3-18

a,c,e Figure 3-7. Su'rge Line Local Axial Stress on Inside Surface at

[ I ' Axial Locations 5429s/091691:10 3-19

a,c,e

. Figure 3-8. Surge Line Local Axial Stress on Outside Surface at 3

' Axial Locations 5429s/091691:10 3-20

a,c,e Figure 3-9. Surge Line.RCL Nozzle 3-D HECAN Model: 10 Inch Schedule 140 5429s/091691: 10 3-21

a,c,e Figure 3-10. Thermal Striping Fluctuation 5429s/091691:10 3-22

a,c,e Figure 3-11. Thermal Striping Temperature Distribution 5429s/091691:10 3-23

SECTION 4.0 DISPLACEMENTS AT SUPPORT LOCATIONS The Ginna plant specific support displacements were calculated under the thermal stratification and normal thermal loads for the existing support configuration. Tables 4-1 and 4-2 show the maximum values of the support displacements in the surge line'. These displacements were checked against-spring hanger RCH-1 travel allowance and it was determined that sufficient travel allowance exists under normal thermal and thermal stratification (with system hT 210'F) loadings.

54294/091691 4-1 h

TABLE 4-1 Haximum Support Displacement* (inch)

Onder Thermal Stratification Cases Di 1 mn r L i n Existing Future C ndi i 'nh n iti n+

~u)g)~or NoON DX DY D7 DX DY DZ ga,c,e X along plant East, Y vertically upward and Z by the right hand rule (see Figure 3-2).

+ Future represents the operating condition (including maximum system hT 210'F) with no spring can bottomed-out.

Existing represents the operating condition (including maximum system hT 275'F) with the spring can bottomed-out.

5429s/091691 10 4-2

TABLE 4-2 Haximum Support Displacement* (inch)

Under Normal Thermal Expansion Di 1 mn r L i n Existing Future niinh n i n+

$ gyj)~r ~N DY DZ DX DY DZ

]a,c,e Hith surge line uniform temperature of 653'F for both existing and future condition; and X along plant East, Y vertically upward and Z by the right hand rule (see Figure 3-2).

5,+ Existing and future condi tions are the same with no spring can bottomed out.

5429s/091691:10 4-3

SECTION 5.0 ASME SECTION III FATIGUE USAGE FACTOR EVALUATION 5,1 ~Mh Surge line fatigue evaluations have typically been performed using the methods of ASME Section III, NB-3600 for all piping components [

3

' Because of the nature of the stratification loading, as well as the magnitudes of the stresses produced, the more detailed and accurate methods of NB-3200 were employed using finite element analysis. for all loading conditions.

Application of these methods, as well as specific interpretation of Code stress values to evaluate fatigue results, is described in this section .

Inputs to the fatigue evaluation included the transients developed in section 2.0,,and the global loadings and resulting stresses obtained using the methods Qi described in section 3.0. In general, the stresses due to stratifi cation were categorized according to the ASME Code methods and used to evaluate Code stresses and fatigue cumulative usage factors. It should be noted that, [

>a,c,e 5.1.1 Basis The ASME Code, Section III, 1986 (Reference [4]) Edition was used to evaluate fatigue on surge lines with stratification loading. This was based on the requirement of NRC Bulletin 88-11 [6] (Appendix B of, this report) to use the "latest ASME Section III requirements incorporating high cycle fatigue".

5429s/091691 10 5-1

Specific requirements for class

~

1 fatigue evaluation of piping components are given in NB-3653. These requirements must be met for Level A and Level B type

~

loadings according to NB-3653 and NB-3654.

According to NB-3611 and NB-3630, the methods of NB-3200 may. be used in lieu of the NB-3600 methods. This approach was used to evaluate the surge line components under stratification loading. Since the NB-3650 requirements and equations correlate to those in NB-3200, the results of the fatigue evaluation are'reported in terms of the NB-3650 piping stress equations. These equations and requirements are summarized in Table 5-1.

The methods used to evaluate these requirements for the surge line components are described in the following sections.

\

5.1.2 Fatigue Stress Equations ifi i n The stresses in a component are classified in the ASME Code based on the nature of the stress, the loading that causes the stress, and the geometric characteristics that influence the stress. This classification determines the acceptable limits on the stress values and, in terms of NB-3653, the

. respective equation where the stress should be included. Table NB-3217-2 provides guidance for stress classification in piping components, which is reflected in terms of the NB-3653 equations.

The terms in Equations 10, ll, 12 and 13 include stress indices which adjust nominal stresses to account for secondary and peak effects for a given component. Equations 10, 12 and 13 calculate secondary stresses,'hich are obtained from nominal values using stress indices Cl, C2, C3 and C3'or pressure,. moment and thermal transient stresses. Equation ll includes, the Kl, K2 and K3 indices in the pressure, moment and thermal transient stress terms in order to represent peak stresses caused by local concentration, such as notches and weld effects. The NB-3653 equations use simplified formulas to 5429s/091691:10 5-2

~ ~

determine nominal stress based on straight pipe dimensions.'

>a,c,e For the RCL nozzles, three dimensional (3-D) finite element analysis was used as described in, Section 3.0.

>a,c,e

.Classification of local stress due to thermal stratification was addressed

~

~

with respect to the thermal transient stress terms in the NB-3653 equations.

Equation 10 includes a Ta-Tb term, cl assi fi ed as. "Q" stress in NB-3200, which represents stress due to differential thermal expansion at gross structural di scontinui ti es. [

] ' The impact of this on the selection of components for evaluation is discussed in Section 5.1.3.

5429s/091691 10 5-3

min in The stresses in a given component due to pressure, moment and local thermal stratification loadings were calculated using the finite element models described in Section 3.0. L 3

' This was done for specific components as follows:

1) L 5429s /091691: 1n 5-4

>a,c,e From the stress profiles created, the stresses for Equations 10 and ll could be determined for any point in the section. Experience with the geometries and loading showed that certain points in the finite element models consistently produced the worst case fatigue stresses and resulting usage factors, in each stratified axial location. [

>a,c,e 5429s/091691 10 5-5

E in12 r Code Equation 12 stress represents the maximum range of stress due to thermal expansion moments as described in Section 3.2. This used an enveloping approach, identifying the highest stressed location in the model. By evaluating the worst locations in this manner, the remaining locations were inherently addressed.

E i n 1 r Equation 13 stress, presented in Section 3.2, is due to pressure, design mechanical loads and differential thermal expansion at structural discontinuities . Based on the transi ent set defined for stratification, the design pressures were not significantly different from previous design transients. Design mechanical loads are defined by the design specification for surge lines built to the ASME Code.

The "Ta-Tb" term of Equation 13 is only applicable at structural discontinuities. I:

)a,c,e Th rm 1 re R h The requirements of NB-3222.5 are a function of the thermal transient stress and pressure stress in a component, and are independent of the global moment loading. As such, these requi rements were evaluated for controlling components using applicable stresses due to pressure and stratification transients.

5429s/091691:10 5-6

All w 1 r Allowable stress, Sm, was determined based on note 3 of Figure NB-3222-1. For secondary stress due to a temperature transient or thermal expansion loads

("restraint of free end deflection" ), the value of Sm was taken as the average of the Sm values at the highest and lowest temperatures of the metal during the transient. The metal temperatures were determined from the transient definition. When part of the secondary stress was due to mechanical load, the value of Sm was taken at the highest metal temperature during the transient.

5.1.3 Selection of Components for Evaluation Based on the results of the global analyses and the considerations for controlling stresses in Section 5.1.2, I 3

a,c,e

' The method to evaluate usage factors using stresses determined according to Section 3.0 is described below.

F r Cumulative usage factors were calculated for the controlling components using the methods described in NB-3222.4(e), based on NB-3653.5. Application of these methods is summarized below.

Tr n i n Load nd min in From the transients described in Section 2.0, specific loadcases were developed for the usage evaluation. [

>a,c,e Each loadcase was assigned the number of cycles of the associated transient as defined in Section 2.0. These were input to the usage factor evaluation, along with the stress data as described above.

5429s/091691 10 5-7

Usage factors were calculated at controlling. locations in the component as follows:

1) Equation 10, Ke, Equation 11 and resulting Equation 14 (alternating stress Salt) are calculated as described above for every possible combination of the loadsets.

2) For each value of Salt, the design fatigue curve was used to determine the maximum number of cycles which would be allowed if this type of cycle were the only one acting. These values, Nl, N2...Nn, were determined from Code Figures I-9.2.1 and I-9.2.2, curve C, for austenitic stainless steels.

1

')

Using the actual cycles of each transient loadset, nl, n2,...n ,

the usage factors Ul, U2...U from Ui i ni/N.. This n'alculate 2 n i done for all possible combinations.

i's Cycles are used up for each combination in the order of decreasing Salt. When Ni is greater than 10 ll cycles, the value of Ui is taken as zero.

>a,c,e

4) The cumulative usage factor, Ucum, was calculated as Ucum Ul +

U2 + ... + U. To this was added the usage factor due to thermal striping, as described below, to obtain total Ucum. The Code allowable value is 1.0.

5429s/091691:10 5-8

53,F i D Th rm 1 ri in The usage factors calculated using the methods of Section 5.2 do not include the effects of thermal striping. [

>a,c,e Thermal striping stresses are

~

a result of differences

~

between the pipe inside surface wall and the average through wall temperatures which occur with time, due to the oscillation of the hot and cold stratified boundary.

~

This type of stress is defined as a thermal discontinuity peak stress for ASME fatigue analysis. The peak stress is then used in the calculation of the ASME fatigue usage factor.

] a,c,e

' The methods used to determine alternating stress intensity are defined in the ASME code. Several locations were .evaluated in order to determine the location where stress intensity was a maximum.

5429s/091691:10 5-9

Thermal striping transients are shown at a bT level and number of cycles.

The striping hT for each cycle of every transient is assumed to attenuate and follow the slope of the curve shown on Figure 5-2. Figure 5-2 is conservatively represented by a series of 5 degree temperature steps. Each step lasts [ 1 ' seconds. Fluctuations are then calculated at each temperature step. Since a constant frequency of [ I ' is used in all of the usage factor calculations, the total fluctuations per step is constant and becomes:

>a,c,e Each striping transient is a group of steps with [ ] ' fluctuations .

per step. For each transient, the steps begin at the maximum hT and decreases by [ 1 ' steps down to the endurance limit of hT equal to

[ 3 a,c,e

' The cycles for all transients which have a temperature step at the same level were added together. This became the total cycles at a step.

The total cycles were multiplied by [. 3

' to obtain total fluctuations. This results in total fluctuations at each step. This calculation- is performed for each step plateau .from [

a,c,e to obtain total fluctuations. Allowable fluctuations and ultimately a usage factor at each plateau is calculated from the stress which exists at the hT for each step.'he total striping usage factor is the sum of all usage factors from each plateau.

The usage factor due to striping, alone, was calculated to be a maximum of

[ 3

' This is reflected in the results to be discussed below.

5.4 R 1 NRC Bulletin 88-11 requires fatigue analysis be performed in accordance with the latest ASME III requirements incorporating high cycle. fatigue and thermal stratification transients. ASME fatigue usage factors have been calculated considering the phenomenon of thermal stratification and thermal striping at various locations in the surge line. Total stresses included the combined 5429s/09169) 5-10

3

' The total stresses for all transients in the bounding set were used to form combinations to calculate alternating stresses and resulting fatigue damage in the manner defined by the Code. Of this total stress, the stresses in the 10 inch schedule 140 pipe due to [

)a,c,e The maximum usage factor for the Ginna surge line occurred at [

ja,c,e It is also concluded that the Ginna pressurizer surge nozzle meets the code stress allowable under the thermal stratification loading from the surge line and the transients detailed in reference [12], and meets the fatigue usage rements of ASME Section III, with a maximum cumulative usage factor equal 'equi to 0.47.

5429s/091691:10 5-11

TABLE 5-1

SUMMARY

OF ASME FATIGUE REQUIREMENTS Parameter Description Allowable (if applicable)

Equation 10 Primary plus secondary stress intensity; < 3Sm if exceeded, simplified elastic-plastic analysis may be performed K Elastic-plastic penalty factor; required for simpl i fied el asti c-pl asti c analysi s when Eq. 10 is exceeded; applied to alternating stress intensity Equation 12 Expansion stress; required for- simpl i fied < 3Sm elastic-plastic analysis when Eq. 10 is exceeded Equation 13 Primary plus secondary stress intensity < 3Sm excluding thermal bending stress; required for simplified elastic-plastic analysis when Eq. 10 is exceeded Thermal Limit on radial thermal gradient stress to Stress . prevent cyclic distortion; required for use Ratchet, of Eq. 13 Equation 11 Peak stress intensity Input to Eq. 14 Equation 14 =

Alternating stress intensity Input to Ucum Ucum Cumulative usage factor (fatigue damage) < 1.0 5429s/091691 10 5-12

a,c,e Figure 5-1. Striping Finite Element Model 5429s/091691:10 5-13

a,c,e Figure 5-2. Attenuation of Thermal Striping Potential by Molecular Conduction (Interface Have Height of One Inch) 5429s/091691:10 5-14

C SECTION 6.0

SUMMARY

AND CONCLUSIONS The subject of pressurizer surge line integrity has been under int'ense investigation since 1988. The NRC issued Bulletin 88-11 in December- of 1988, but the Hestinghouse Owners Group had put a program in place earlier that year, and this allowed all members to make a timely response to the bulletin.

The Owners Group programs were completed in June of 1990, and have been followed by a seri es of plant specific evaluations. This report has documented the results of the plant specific evaluation for Ginna.

Following the general approach used in developing the surge line stratification transients for the HOG, a set of transients and stratification profile were developed specifically for Ginna. A study was made of the hi stori'cal operating experience at Ginna, and this information, as well as plant operating procedures and monitoring data, was used in development of the transients and profiles.

As a result of the analyses, sufficient travel allowance exists for spring hanger RCH-1 under normal thermal and thermal stratification displacements for the future condition (system hT 210'F) . The results of this plant specific analysis along with support verification demonstrate acceptance to the requi rements of the ASHE Code Section III, including both stress limits and fatigue usage, for the full licensed life of the plant. This report demonstrates that Gi nna has now completely satisfied the requirements of NRC Bulletin 88-11.

5429s/091691:10 6-1

SECTION

7.0 REFERENCES

1. Coslow, B. J., et al., "Nestinghouse Owners Group Bounding Evaluation for Pressurizer Surge Line Thermal Stratification" Nestinghouse Electric Corp. HCAP-12277, (proprietary class 2) and HCAP-12278 (non-proprietary),

June 1989,.

2. Coslow, B. J., et al., Nestinghouse Owners Group Pressurizer Surge Line Thermal Stratification Program MUHP-1090 Summary Report," Nestinghouse Electric Corp. WCAP-12508 (propri etary class 2) and HCAP 12509 (non-proprietary), March 1990.

3. Coslow, B. J., et al., "Hestinghouse Owners Group Pressurizer Surge Line Thermal Stratification Generic Detailed Analysis Program NUHP-1091 Summary Report," Nestinghouse Electric Corp. HCAP-12639 (propri etary class 2) and HCAP-12640 (non-propri etary), June 1990.

4. ASHE BLPV Code Section III, Subsection NB, 1986 Edition.

5. RGE Letter, "Pressurizer Surgeline Evaluation Ginna Station," From D. Morgan to P. Strauch, August 2, 1990.

6. "Pressurizer Surge Line Thermal Stratification," USNRC Bulletin 88-11, December'20, 1988.

7. "Investigation of Feedwater Line Cracking in Pressurized Hater Reactor Plants," HCAP-9693, Volume 1 (Hestinghouse Proprietary Class 2).

5429 s/091691: ~ > 7-1

8. Hoodward, H. S., "Fatigue of LMFBR Piping due to Flow Stratification,"

ASHE Paper 83-PVP-59, 1983.

r

9. Fujimoto, T., et al., "Experimental Study of Striping at the Interface of Thermal H. Sun, Strati fication" in et al., (ed.)

Th rm 1 H r li in N ASHE, 1981, pp. 73.

1 r T hn 1, K.

IU. 9 I, 3. f., SUIt f"tt I,IIIII I I', 9 3993.

11. Yang, C. Y., "Transfer Function Method of Thermal Stress Analysis:

Technical Basis," HCAP 12315 (Hesti nghouse Proprietary).

12. Series 84 Pressurizer Stress Report, Section 3.1, Surge Nozzle Analysis, December 1974.

5429s/091691:10 7-2

APPENDIX A LIST OF COMPUTER PROGRAMS This appendix lists and summarizes the computer codes used in the analysis of stratification in the pressurizer surge line. The codes are:

1. PECAN

2. STRFAT2

3. ANSYS

4. FATRKlCMS A. 1 ~AN A.

NECAN is a Hestinghouse-developed, general purpose finite element program. It contains universally accepted two-dimensional and three-dimensional isoparametric elements that can be used in many different types of finite element analyses. guadri lateral and triangular structural elements are used for plane strain, plane stress, and axi symmetric analyses. Brick and wedge structural elements are used for three-dimensional analyses. Companion heat conduction elements are used for steady state heat conduction analyses and transient heat conduction analyses.

A.1.2 The temperatures obtained from a static heat conduction analysis, or at a specific time in a transient heat conduction analysis, can be automatically input to a static structural analysis where the heat conduction elements are replaced by corresponding structural elements. Pressure and external loads can also be included in the NECAN structural analysis. Such coupled thermal-stres's analyses are a standard application used extensively on an industry-wide basis.

5429s/091691:10 A-1

A.1.3 Pr

~ ~ r m V rifi

~

tion

~

Both the WECAN program and input for the WECAN verification problems, currently numbering over four hundred, are maintained under configuration I

control. Verification problems include coupled thermal-stress analyses for 2

the quadrilateral, triangular, brick, and wedge isoparametric elements. These problems are an integral part of the HECAN quality assurance procedures. When a change is made to WECAN, as part of the reverification process, the configured iaputs for the coupled thermal-stress verification problems are used to reverify WECAN for coupled thermal-stress analyses.

A.2 ~TRF T2 2.2.1 STRFAT2 is a program which computes the alternating peak stress on the inside surface of a flat plate and the usage factor due to striping on the surface.

The program is applicable to be used for striping on the inside surface of a pipe if the program assumpti,ons are, considered to apply for the particular pipe being evaluated.

For striping the fluid temperature is a sinusoidal variation with numerous cycles.

The frequency, convection film coefficient, and pipe material properti es are input.

The program computes maximum alternating stress based on the maximum difference between inside surface skin temperature and the average through wall temperature.

5429s/091691 10 A-2

A.2.2

~ ~

The program is used to calculate striping usage factor based on a ratio of actual cycles of stress for a specified length of time divided by allowable cycles of stress at maximum the alternating stress level. Design fatigue curves for several materials are contained into the program. However, the user has the option to input any other fatigue design curve, by designating that the fatigue curve is to be user defined.

A.2.3 r r m V rifi i n STRFAT2 is verified to Hestinghouse procedures- by independent review of the stress equations and calculations.

A. 3 AN'LYY.

A.3.1 ANSYS is a public domain, general purpose finite element code.

3 The ANSYS elements used for the analysis of stratification effects in the surge line are STIF 20 (straight pipe), STIF 60 (elbow and bends) and STIF14 (spring-damper for supports).

A33 r mVrifi i n As described in section 3.2, the application. of ANSYS for stratification has been independently verified by comparison to HESTDYN (Hestinghouse piping]

analysis code) and HECAN (finite element code). The results from ANSYS are also verified against closed form solutions for simple beam configurations.

5429s/091691:10 A-3

A.4 ~FATRK/ M

~ .

FATRK/CHS is a Westinghouse developed computer code for fatigue tracking (FATRK) as used in the Cycle Honi tori ng System (CHS) for structural components of nuclear power plants. The transfer function method is used for transi ent thermal stress calculations. The bending stresses (due to global stratification effects, ordinary thermal expansion and seismic) and the pressure stresses are also included. The fatigue usage factors are evaluated in accordance with the guidelines given in the ASHE Boiler and Pressure, Vessel Code, Section III, Subsections NB-'3200 and NB-3600.

The code can be used both as a regular analysis program or an on-line monitoring device.

A.2 FATRK/CHS is used as an analysis program for the'present application. The input data which include the weight functions for thermal stresses, the unit bending stress, the unit pressure'tress, the bending moment vs.

stratification temperatures, etc. are prepared for all locations and geometric conditions. These data, as stored in the independent files, can be appropriately retrieved for required analyses. The transient data files contain the time history of temperature, pressure, number of occurrence, and additional condition necessary for data flowing. The program prints out the total usage factors, and the transients pairing information which determine the stress range magnitudes and number of cycles. The detailed stress data may also be printed.

A.4.3 Pr r m V rifi i n FATRK/CHS is verified according to Westinghouse procedures with several levels of independent cal culations.

5429s/091691:10 , A-4

APPENDIX B USNRC BULLETIN 88-11 In December of 1988 the NRC issued this bulletin, and it has led to an extensive investigation of surge line integrity, culminating in this and other plant specific reports. The bulletin is reproduced in its entirety in the pages which follow.

5429s/091691 '" B-1

OMB No. 3150-0011 NRCB 88-11 UNITED STATES NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION WASHINGTON, O.C. 20555 Oecembcr- 20, 1988 NRC BULLETIN NO. 88-11: PRESSURIZER SURGE LINE THERMAL STRATIFICATION Addressees:

All holders of operating lfcenses or construction permits for pressurized water reacto'rs (PWRs).

~PUr ose:

h The purpose of this bulletin fs to (1). request that addressees cstabl1sh and implement a program to conf1rm pressurizer surge line integrity in view of the occurrence of thermal strat1ffcatfon and (2) require addressees to 1nform the staff of the actions taken to resolve this issue.

Oescrf tfon of Circumstances:

The licensee for the Trojan plant has observed unexpected movement of the pressurizer surge 11ne during inspections performed at each refueling outage since 1902, when monitoring of thc linc movements began. Ourfng the'ast refueling outage, the licensee found that fn addition to unexpected gap clo-surcs in the p1pe whip restra1nts, thc piping actually contacted two re-straints. Although the licensee had repeatedly adjusted shfms and gap s1zes based on analys1s of varfous postulated condft1ons, thc problem had not been resolved. The most recent fnvest1gatfon by the 11censec conf1rmed that the movement of piping was caused by thermal strat1ffcatfon in the line. This phcnomcnon was not considered fn the original pip1ng design. On October 7, 1988, the staff issued Information Notice 88-80, "Unexpected P1pfng Movement Attributed to Thermal Stratfffcatfon," regarding the Trojan experience and fndfcated that further generic coaeunfcatfon may be forthcoming. The licensee for Beaver Valley 2 has also noticed unusual snubber movemcnt and significantly larger-than-expected surge line displacement during power ascension.

The concerns ra1sed by the above observations are similar to those described in NRC Bulletins 79-13 (Revision 2, dated October 16, 1979), "Cracking fn Feedwater System Piping" and 88-08 (dated June 22, 1988), "Thermal Stresses fn P1pfng Connected to Reactor Coolant Systems."

8812150118

NRCB 88 ll Oecember 20, 1988 Page 2 of 6 Oiscussion:

Unexpected piping movements are highly undesirable because of potential high piping stress that may exceed design limits for fatigue and stresses. The problem can be more acute when the piping expansion is restr'icted, such as through contact with pipe whip restraints. Plastic deformation can result, which can lead to high local stresses, low cycle fatigue and functional im-pairment of the line. Analysis performed by the Trojan licensee indicated that therma'l stratification occurs in the pressurizer surge line during heatup, cooldown, and steady-state operations of the plant.

Ouring a typical plant heatup, water in,the pressurizer is heated to about

. 440'.F; a steam bubble is then formed in the pressurizer. Althouah the exact

=phenomenon is not thoroughly understood, as the hot water flows (at a very low flowr ate) from the pr'essurizer through the surge line to the hot-leg piping, the hot water rides on a layer of cooler water, causing the upper part of the pipe to be heated to a higher temperature than the lower part (see Figure I).

The differential temperature could be as high as 300'F, based on expected conditions during typical plant operations. Under this condition, differential thermal expansion of the pipe metal can cause the pipe to deflect si gnnifi-cantly.

For the specific configuration of the pressurizer surge line in the Trojan plant, the line deflected downward and when the surge line contact e d two p>pe whi p restraints, it underwent plastic deformation, resulting in permanent deformation of the pipe.

The Trojan event demonstrates that thermal stratification in the pressurizer surge line causes unexpected piping movement and potential plastic deformation.

The licensing basis according to 10 CFR 50.55a for all PWRs requires that the licensee meet the American Society of Mechanical Engineers Boiler and Pressure Yessel Code Sections III and XI and to reconcile the pipe stresses and fatigue evaluation when any significant differences are observed between measured data and the analytical results for the hypothesized conditions. Staff evaluation indicates that the thermal stratification phenomenon could occur in all PWR surge lines and may invalidate the analyses supporting the integrity of the surge line. The staff's concerns include unexpected bending and thermal striping (rapid oscillation of the thermal boundary interface along the piping inside surface) as they affect the overal.l integrity of the surge line for its design life (e.g., the increase of fatigue).

Actions R uested:

Addressees are requested to take the following actions:

1. For all licensees of operating PWRs:

a. Licensees are requested to conduct a visual inspection (ASME,Section XI, VT-3) of the pressurize surge line at the first available cold shutdown after receipt of this bulletin which exceeds seven days.

NRCB 88-11 December 20, 1988 Page 3 of 6 This inspect1on should determine any gross d1scernable distress or structural damage in the entire pressurizer. surge line, including piping, pipe supports, pipe whip restraints, and anchor bolts.

b. Within four months of receipt of this Bullet1n, licensees of plants in operat1on over 10 years (i.e.'; low power license prior. to January 1, 1979) are requested to demonstrate that the pressurizer surge line meets the applicable design codes* and other FSAR and regulatory commitments for the licensed life of the plant, conside~-

ing the phenomenon of thermal stratification and thermal striping in the fatigue and stress evaluations. Th1s may be accomplished by perform1ng a plant specif1c or generic bounding analysis. If the latter option is selected, licensees should demonstrate applicability of the referenced .generic bound1ng analysis. Licensees of plants in operation less than ten years (i.e., low power license after January 1, 1979), should complete the forego1ng analysis within one year. of receipt of this bullet1n. Since any p1ping distress observed by addressees in performing act1on l.a may affect the analysis, the licensee should verify that the bounding analysis remains valid, If the opportunity to perform the visual inspection in 1.a does not occur within the pe~iods specified in this requested item, 1ncorpora-tion of the results of the visual inspection into the analys1s should be performed in a supplemental analysis as appropriate.

Where the analysis shows that the surge 11ne does not meet the requirements and licensing commitments stated above for the duration of the license, the licensee should submit a justification for continued operation or bring the plant to cold shutdown, as appropri-ate, and implement Items I.c and 1.d below to develop a detailed analysis of the surge line.

C ~ If the analys1s in 1.b does not show compliance with the requirements and licensing comoitments stated therein for the duration of the operating license, the licensee is requested to obtain plant specific data on thermal stratification, thermal striping, and line deflec-tions. The licensee may choose, for example, either to install instruments on the surge line to detect temperature distribution and thermal movements or to obtain data through collective efforts, such as from other plants with a similar surge line design. If the latter option is selected, the licensee should demonstrate similarity in geometry and operat1on.

d. Based on the applicable plant specific or referenced data, licensees are reauested to update their stress and fatigue analyses to ensure compliance with applicable Code requirements, incorporating any observations from 1.a above. The analys1s should be completed no later than two years after receipt of this bulletin. If a licensee g Iyf III gf h Idh p I df d fh h 1<<111 I \ I Ighfgh yf f Ig

NRCB 88-11 December 20, 1988 Page 4 of 6 is unable to show compliance with the applicable design codes and other FSAR and regulatory coomitments, the licensee is requested to submit a justification for continued operation and a description of the proposed corrective actions for effecting long term resolution.

2. For all applicants For PWR Operating Licenses:

a. Before issuance of the low power license, applicants are requested to demonstrate that the pressurizer surge line meets the applicable design codes and other FSAR and regulatory ceanitments for the licensed life of the plant; This may be accomplished by performing a plant-specific or generic bounding analysis. The analysis should include consideration of thermal stratification and thermal striping to ensure that fatigue and stresses are in compliance with -applicable code limits. The analysis and hot functional testing should verify that piping thermal deflections result in no adverse consequences, such as contacting the pipe whip restraints. If analysis or test results show Code noncompliance, conduct of all actions specified below is requested.

b. Applicants are requested to evaluate operational alternatives or piping modifications needed to reduce fatigue and stresses to acceptable levels.

ce Applicants are requested to either monitor the surge line for the effects of thermal stratification, beginning with hot functional testing, or-obtain data through collective efforts to assess the extent of thermal stratification, thermal striping and piping deflections.

d~ Applicants are requested to update stress and fatigue analyses, as necessary, to ensure Code compliance.* The analyses should be completed'o later than one year after issuance of the low power license.

3. Addressees are requested to generate records to document the development and implementation of the program requested by Items 1 or 2, as well as any subsequent corrective actions, and maintain these records in accor-dance with 10 CFR Part 50, Appendix B and plant procedures.

Re ortin Re ufrements:

l. Addressees shall report to the NRC any discernable distress and damage observed in Action 1.a along with corrective actions taken or plans and schedules for repair before restart of the unit.

comp ance w t t e applicable codes is not demonstrated for the full duration of an operating license, the staff may impose a license condition such that normal operation is restricted to the duration that compliance is actually demonstrated.

NRCB 88>>11 Oecember 20, 1988 Page 5 of 6

2. Addressees who cannot meet the schedule described in Items 1 or 2 of Actions Re uested are required to submit to the NRC within 60 days of rece>pt o t ss ulleti n an alternative schedule with justification for.

the- requested schedule.

3. Addressees shall submit a letter within 30 days after the completion of these actions which notifies the NRC that the actions requested in Items lb, ld or 2 of Actions Re'ested have been performed and that the results are available for inspect on. he letter shall include the justification for continued operation, if appropriate, a description of 'the analytical approaches used, and a summary of the results.

, Although'ot requested by this'ulletin, addressees are encouraged to work collectively to address the technical concerns associated with this issue, as well as to share pressurizer sur'ge 'line data and operational experience, In addition, addressees are encouraged to review piping in other systems which may experience thermal stratification and thermal striping, especially in light of the'previously mentioned Bulletins 79-13 and 88-08. The NRC staff intends to review operational experience giving appropriate recognition to this phenome-non, so as to determine if further generic communications are in order.

The letters required above shall be addressed to the U.S. Nuclear Regulatory Comnissfon, ATTN: Oocument Control Oesk, Washington, O.C. 20555, under oath or affirmation under the provisions of Section 182a, Atomic Energy Act of 1954, as amended. In addition, a copy shall be submitted to the appropriate Regional Administrator.

This request is covered by Office of Management and Budget Clearance Number 3150-0011 which expires Oecember 31, 1989. The estimated average burden hours fs approximately 3000 person-hours per licensee response, including assessment of the new requirements, searching data sources, gathering and analyzing the data, and preparing the required reports. These estimated average burden hours pertain only to these identified response-related matters and do not include the time for actual implementation of physical changes, such as test equipment installation or component modification. The estimated average radiation ex'posure is approximately 3.5 person-rems per licensee response.

Comments on the accuracy of this estimate and suggestions to reduce the burden may be directed to the Office. of Management and Budget, Room 3208, New Execu-tive Office Building, Washington, O.C. 20503, and to the U.S. Nuclear Regula-tory Commission, Records and Reports Management Branch, Office of Adminsstration and Resource Management, Washington, O.C. 20555.

NRCB 88-11 December 20, 1988 Page 6 of 6 If you have any questions about this matter, please contact one of the techni-cal contacts listed below or the Regional Administrator of the appropriate regional office.

a les E. Rossi, Director Division of Operational Events Assessment Office of Nuclear Reactor Regulation Technical Contacts: S. N. Hou, NRR (301) 492-0904 S. S. Lee, NRR (301) 492-0943 N. P., Kadambi, NRR (301) 492-1153 Attachments:

I. Figure 1

2. List of Recently Issued NRC Bulletins

Attachment I NRCB 88-II December 20, l988 Page I of I Surge Line Stratification TpzR Hot Flow from Pressurizer hot 425'F Stagnant Cold Fluid Tc old 125'F

'Figure 1

APPENDIX C TRANSIENT DEVELOPMENT DETAILS 5429s/091691'10 C-1

5429s/091691:10 C-2

~a,c,e 5429s/091691:10 C-3

5429s/091691: 10 C-4

>a,c,e 54294/091691 '