ML20214G521

Nonproprietary Addl Info in Support of Elimination of Postulated Pipe Ruptures in Accumulator Lines of South Texas Project Units 1 & 2

ML20214G521
Person / Time
Site:	South Texas
Issue date:	05/31/1987
From:	Chang K, Palusamy S, Swamy S WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML19292H294	List: ML20214G489 ML20214G521 ST-HL-AE-2165, Forwards Westinghouse Nonproprietary & Proprietary ...Info in Support of Elimination of Postulated Pipe Ruptures..., Per Util 861114 Request for Exemption from 10CFR50,App A,Gdc 4 Requirements.Proprietary Rept Withheld (Ref 10CFR2.790)
References
NUDOCS 8705270092
Download: ML20214G521 (29)

Text

(l. .?.

g ... L ... ... _ _. . . - _ . .

l WESTINGHOUSE CLASS 3 ADDITIONAL INFORMATION IN SUPPORT OF THE ELIMINATION OF POSTULATED PIPE RUPTURES IN THE ACCUMULATOR 1- .

LINES OF SOUTH TEXAS PROJECT UNITS 1 AND 2 May, 1987

_- S. A. Swamy Y. S. Lee F. J. Witt D. H. Roarty Verified:

K. C. Chan'g, Mvison? Kngineer Piping Technology V Approved: =,d[/ [

f5. 5. Pa)dsaiy, Kanager Structufal Material Engineering i

WESTINGHOUSE ELECTRIC CORPORATION Generation Technology Systems Division P.O. Box 2728 Pittsburgh, Pennsylvania 15230-2728 h(( DO 8

l INTRODUCTION The. justification for eliminating Class 1 accumulator line ruptures in the high pressure and high temperature portion for South Texas Project Units 1 and 2 has been provided in WCAP-11383. In that WCAP the scope of work sovered the Class 1 portion of the accumulator line from the RCS primary loop injection point to the first check valve (see figure 1). The pipe whip restraints being eliminated are in this portion of'the accumulator line. In addition, the loads in the remaining 12-inch diameter portion of the accumulator line (from the 1st check valve up to the accumulator nozzle) were compared with the loads in WCAP-11383. The maximum longitudinal stress at the critical location identified in WCAP-11383 was seen to envelope tho' stress in the 12 inch portion of the accumulator line between the first check valve and the accumulator tank nozzle. The operating temperature of the accumulator line after the first valve is 120*F and therefore the yield strength of the l

l material is significantly higher than the yield strength at the critical location identified in WCAP-11383. Therefore the results and conclusions of WCAP-11383 are applicable for the entire 12 inch span of the accumulator line (see figure 1) from the primary loop junction (anchor) to the accumulator tank nozzle junction (anchor). -

At the request of the NRC, additional information is provided in this report.

Answers to questions 8 and 9 were provided by Houston Lighting and Power.

l 1

L

I  !

I

! 2 '

! ,2 c .

j G

-4 E

E i

i  ?

8. ,

N hecuumlator R'+

i

IG Tank

M -

:r W ED us Classl++ Class 2 Sch140' Jch40

[ ,ans ANCHORPOINT)J

g ANCHOR POINT i

e V w N

I

=-

k i'

l g

~

! , /

OPERATING PRESSURE = 2304 psi OPERATING PRESSURE = 650 psi l: -

j 0 OPERATING TEMP = 120 F OPERATING TEMP = 5600F d

I

n. . .

7(

1. Question: Provide material specification and all fracture toughness 1 4

properties of the accumulator tank and its nozzle (and any safe end and associated weld). If the material is ferritic, determine whether the material is at the upper shelf of the fracture energy at the operating temperature, and compare the fracture properties to those of the weld.. Justify why welds at ,

valves are not limiting. Provide a sketch of the cross-section l of the valve-to pipe weld at limiting location 284 on figure 5-2 of WCAP-11351.

f Response: A. Material specification i 1. Nozzle-SA35OLF-2(forging,carbonandlow-alloy

= steel): o,=70-95 ksi (tensile strength),

ay=36 ksi (yield stress)

2. Tank - S4764 (stainless chromium-nickel steel clad plate, sheet and strip) contains SA537 C1 1 (pressure j vessel plates, heat-treatment carbon-manganese-silicon l

steel)o=70ksi(min.),o=50ksi(min.),and u y l SA 240-TP304 (heat resisting and Cr-Ni stainless steel

( plate and strip for fusion welded unfired pressure vessel), o,=75 ksi, oy=30 ksi.

l 3. Safe end - SA312-TP304 (seamless and welded austenitic stainless steel pipe): o,= 75 ksi, oy= 30 ksi B. Fracture Toughness The certified material test report (CMTR) for the nozzle provides the following test results:

. 3

. u. .

s. . .

Potential Test Result RT NDT Drop Weight No Break at O'F (Both heats) -10' Charpy Test Energy: 55,53,52,(1stheat) at 60'F 53, 50, 51 (2nd heat), ft-lbs O' Lateral Expansions: 50, 48, 46 -

(1stheat) 48, 45, 46 (2nd heat), mils Therefore the RTNDT of the nozzle is O'F. The drop weight i~: test for the tank material was not performed, however, Charpy V-Notch tests at 60*F were made and the test results are as follows:

Potential Test Heat Results RT NDT V Notch 1 Energy 53, 53, 52 ft-lb Test at Lateral 60'F Expansion 48, 49, 49 mils 2 Energy 51, 50, 51 ft-lb Lateral Expansion 49, 50, 50 mils 0*F 3 Energy 50, 50, 51 ft-lb Lateral Expansion 49, 48, 50 mils 4 Energy 55, 53, 53 ft-lb i Lateral ,

Expansion 49, 48, 49 mils 4

.L.. . _

The V-Notch test results of the tank material show very similar results with those of nozzle material. Thus RTNDT of the tank material is considered to be O'F.

The toughness of ferritic steel is taken from Appendix-A,Section XI of the ASME Code as the lesser of Klc = 33.2 + 2.806 Exp (0.02 (T-RTNDT + 100)) (1) or KIc = 200 ksi/in. (upper shelf toughness)

Substituting the operating temperature, 120' and RTNDT = 0*F in equation (1), the toughness of ferritic steel at 120*F is found to be 200 ksi/in. This is equivalent to Jg ,

= 1333 in-lb/in 2 where Jg , is assumed to be given by

=

JIc

• Ele /E where E = 30 x 106 p,g, TestsofactualSA350LF-2forgingmaterialhaveshownJ, g values of ( ']a c.e in-Ib/in at 50*F and [ .]a.c.e 2 ]a,c.e Since in-Ib/in at 75'F. The RTNDT was [ .

the tank and nozzle operate at 120*F with an RT NDT of O'F, ,

they operate at or near the upper shelf fracture toughness I tenperature. Thus the actual J Ie appropriate to the 120*F operating temperature is near the higher test result [

ja.c.e ,

Thus the accumulator hozzle and the accumulator tank material exhibit a greater toughness level as compared to the JIc "

2

[ ]a,c.e in-lb/in used as the criterion in the leak-before-break evaluation.

! 5 .

1,

~

. . , . .. s..-

As discussed in the introduction to this report, the leak-before-break evaluation is performed for the 12-inch .

diameter piping from the primary loop junction to the accumulatornozzlejunction. Location 284 in figure 5-2 of WCAP-11351 is on a branch line where the pipe diamettr'is 8 inches and need not be included in the leak-before break evaluation of the 12 inch diareter piping and therefore, a sketch of the cross section of the valve-to pipe weld is not included in this response.

I i

I .

6

. q b s' ' . = . . . . . ..-:..-.- .

2. Question: Provide justification for using the thermally-aged cast stainless steel in WCAP-10456 for the limiting weld.

Specifically, compare the base metal, weld metal, and welding process parameters between the limiting weld and the welds considered in WCAP-10456. ~.

Response: The limiting weld location is location 104E as identified in figure 5-1 of WCAP-11383. This is a field weld. The field welding was performed using the automatic gas tungsten are process (GTAW). The filler metal was 308L. The shop

_ vernacular for GTAW is TIG. All the welding was performed to applicable QA procedures.

, Traditionally nuclear pipe welds were made by first performing root passes using GTAW followed by either submerged are welding (SAW) for shop welds or shielded metal arc welding (SMAW) for field and shop welds to completion. Thus, typical welding data have been obtained on SAW and SMAW. For example in table 3 of reference 2-1, welds A, B C, and G are [ ]a,c.e while the ,

remaining (Dc E, and F) are [ ~)a c.e (reference 2-2).

Specifically in figure 2.4-8 (reference 2-3) of WCAP-10456 weld '

B is [ ]a,c.e while weld D is [ 3a,c.e. The general argument of WCAP-10456 that welds are not limiting compared to L [ Ja c.e when thermal aging is considered is based on the conclusions reached in reference 2-1 and will not be repeated here. Interestingly, the minimum KCU for the thermally aged welds B [ ]a,c.e and E [ .]C

2 noted above was [ ]a,c.e daJ/cm or greater after 10,000 hrsofagingat400'C(2-2)whichexceedsthe[ Ja.c.e 2

daJ/cm criterion of reference 2-4.

Toughness of GTAW have been evaluated by several investigators, one of the more recent being reference 2-5. This reference considers the fracture toughness of several welds including a 308LGTAW(i.e.,TIGweld). Significantly the toughness of the 7

. . _ ~ _ . . _ _ _ __ _ _ _ -

GTAW approached the toughness of the 304 base material

' nvestigated.

i 2 The Jg , exceeded 3000 in-lb/in at 550*F with a Tmat of 500. The Charpy V-Notch energy at 550*F exceeded the machine capacity (239 ft-lbs). At room toeperature the Charpy V-Notch energy was 140 ft-lbs,which -

corresponds to a KCU energy of around 90 ft-1bs (24 '

2 daJ/cm ). This value is well above the KCU value at room 2

temperature of around 7 daJ/cm noted for the unaged welds-

[ 3a.c.e discussed in WCAP-10456.

The ferrite content in GTAW is typically less than 10% thus it

~

~

is not expected that thermal aging degradation will signifi-cantly reduce the fracture toughness values. It is reasonable then to assume that the GTAW is at least as good as [- 1

! Ja.c.e having a gJ ,of [ .)***** in-1b/in2 and a T

ut of [ ] ' ** after 40 years of service.

4 References 2-1 Slama, G., Petroquin, P., Masson, S. H., and Mager, T. R.,

"Effect of Aging on Mechanical Properties of Austenitic Stainless Steel Castings and Welds," presented at SMIRT 7 Post Conference Seminar 6 - Assuring Structural Integrity of Steel Reactor Pressure Boundary Components, August 29/30,1983, Monterey, CA.

2-2 PWS 3-4, Effect of Aging on the Mechanical Characteristics of Austenitic-Ferritic Welds, Internal ID No. E EM DC 0

. 260, FRAMATOME (Industrial Property).

2-3 WCAP-10456, "The Effects of Thermal Aging on the Structural Integrity of Cast Stainless Steel Piping for W NSSS,"

i November 1983 (Westinghouse Proprietary Class 2). , ,

8

^

[.._..P.

c ;. \ s:. .

2-4 Witt, F. J. and Kim, C. C., Toughness Criteria for Thermally Aged Cast Stainless Steel, WCAP-10931 Revision 1, July 1986 (Westinghouse Proprietary Class 2). I 2-5 Toughness of Austenitic Stainless Steel Pipe Weids, EPRI NP-4768, Electric Power Research Institute, October 1986. l O

f a

l I

l l

l~

i l

l . 9

.- p ., .

l l

Question: The Code minimum yield and ultimate strengths of SA376 TP316 3.

are used for the base metal. However, from Table 4-6 in WCAP-11351, other materials in the accumulator line, i.e.,

SA312 TP304L, SA403 WP316L, and SA182 F316L have lower Code minimum yield and ultimate strengths. If the intent,1s to use the lower bound material properties, justify the selection of SA376 TP316 as the limiting material.

Responte: The Seuth Texas Project Units 1 and 2 material certification records are summarized for the 12-inch diameter piping in tables 4-1 and 4-2 of WCAP-11383 and tables 4-1 and 4-3 of WCAP-11351. In these tables all the heats of the piping material are made of SA376-TP316 except one heat of material i (HT11-226)whichisSA312-TP316L. This heat of material is located in the low pressure class 1 portion of the line operating at 120*F. The room temperature properties for this heat of SA312-TP316L material are oy = 41700 psi and o,

= 82100 psi. These properties are comparable with the room temperature properties of SA376-TP316. The location of maximum load occurs at the high temperature region where the yield point is significantly lower and the material is SA376-TP316.

Hence the selection of ASME Code minimum property of 19200 ps'i at 560*F for SA376-TP316 is considered appropriate.

1 l

9 1

10

.c, . ... .

4. Question: The portions of the accumulator line considered in WCAP-11383 and WCAP-11351 operate at 560*F and 120'F, respectively.

Provide the moduli and the stress-strain relationships of the limiting material (s) at the appropriate operating temperatures, i.e., 560*F and 120*F.

%,w ,

Response: The critical location for leak-before-break evaluation of the accumulator line is between the primary loop junction and the first valve. The temperature in this portion of the accumulator line is 560*F. The stress-strain relation for this location is provided (see figure 4-1). The modulus of

^ '

elasticity is 25.847 x 106p,g, e

e O

l c

I l

11 l

..i. ..... .

• ~

a.C p

.~

r

=

I l

i Figure 4-1. True Stress-True Strain Curve for SA376 TP316 Stainless Steel at 560*F 12

J

't

5. Question: In order to complete our safety avaivation, the staff needs to evaluate the accumulator line for potential degradation during service. Therefore, the applicant should provide a discussion on the potential and preventive measure for vibratory fatigue (socket welds, if any), flow stratification (ar.d associated low

, cycle thermal fatigue), wall thinning by erosion, and creep.

Also, discuss the experience of the accumulator lines of Westinghouse design with respect to in-service cracking.

Response: There has never been any service cracking or wall thinning identified in the accumulator lines of Westinghouse PWR design. Sources of such degradation are mitigated by the design, construction, inspection, and operation of the accumulator lines.

i Vibratory fatigue loads are monitored during the hot-functional testing of the project and are well below the high cycle I fatigue allowables. .

There is no mechanism for water hammer in the accumulator piping system.

Wall thinning by erosion and erosion-corrosion effects will not -

I occur in the accumulator line due to the low velocity, typically less than 10 ft/see and the material, austenitic stainless steel, which is highly resistant to these degradation mechanisms. Per NUREG-0691, a study of pipe cracking in PWR piping, only two incidents of wall thinning in stainless steel pipe were reported and these were not in the accumulator line.

Although it is not clear from the report, the cause of the wall thinning was related to the high water velocity and is therefore not a mechanism which would affect the accumulator line.

13

.  ; ^b..

1 Flow stratification, where low flow conditions permit cold and hot water to separate into distinct layers, can cause significant thermal fatigue loadings. This was an important issue in PWR feedwater piping where temperature differences of 300'F were not uncommon under certain operational conditions.

Stratification is believed to be important where low flow conditions and a tenperature differential exist. This is not an issue in the accumulator line where typically there is no flow during normal plant operation. During RHR operation the flow causes sufficient mixing to eliminate stratification.

I Finally, the maximum operating temperature of the accumulator piping, which is about 560*F, is well below the temperature which would cause any creep damage in stainless steel piping.

O I

l O

14

6 g'q . .

6. Question: A generic fatigue crack growth analysis was performed for the accumulator line. If the sequence of transients in the fatigue crack growth analysis was not selected randomly, provide the specific sequence used in the analysis. Fifteen transients were considered in the fatigue analysis with no Opergting Basis Earthquake (OBE). Since the staff considers it appropriate to include the OBE, justify this exclusion.

Response: The number of occurrences of a particular transient are equally spread over the lifetime so that the crack growth can be determined at intermediate times during the lifetime of the

~~

structure. The cycle input data is used to determine a schedular distribution by dividing each transient into event categories. Specifically, the total cycles of transients are scheduled into five events per year, two events per year, one event per year, one every fourth year, and one every eighth year. The crack size is updated by adding the incremental growth due to a transient to the initial crack size. The .

updated crack size is used as the initial crack size for the

. next transient loading. The calculations are repeated to l account for all the events during each year.

OBE was not included in the fatigue crack growth evaluation performed per WCAP-11383 because the seismic stress levels were very low. In fact, for SSE plus anchor motion, which has larger loads than OBE, the maximum nominal axial stress is

+_ 3.7 ksi. This level of stress will result in negligible l

crack growth, if any, for the small number of cycles due to OBE.

l

- 15 L o

_- . .,4 , ,-

7. Question: Because of a lower fluid temperature, the leak rate prediction in WCAP-11351 used a model different from that benchmarked in WCAP-11256 (Supplement 1). Provide a benchmark for the leak rate prediction model against leak test data at the appropriate fluid temperature. ,

Response: The critical location for leak-before-break analysis is found to be in the high pressure and high temperature segment of the accumulator line (temperature = 560'F and pressure = 2304 psi). At this location leakage through postulated through-wall flaw would be a two phase choked flow. The leak rate model used is the Fauske Henry Griffith model. Detailed benchmark calculations have been provided in WCAP-11256 Supplement 1 in support of this model.

S S

e 1

e 9

16

4-  ;-.3

8. Question: Relative to the system (s) that would be relied upon in this application, provide discussion of the leakage detection system (s) and procedures for detection and operator action at South Texas. Justify the workability of these systems and procedures in terms of operating experience at Westinghouse plants. Describe the background unidentified average leakage rate expected at South Texas and the basis for this prediction. Justify the use of 0.5 gpm as the minimum detect-able unidentified leakage. Discuss the Technical Specification and/or administrative procedure requirements for locating a 0.5

_ gpa unidentified leakage. Also discuss repair and system limiting conditions for operation associated with a leak in the i

accumulator line. Clarify if the entire accumulator line is

~

reactor coolant pressure boundary piping. Also, clarify if a

! leaking pipe in the leak-before-break scope will be declared inoperable.

Response: All LBB candidate lines at STP are located inside containment.

The STP leakage detection criterion includes a' detected

! unidentified leak rate of 1.0 gpm and, in accordance with NUREG-1061, Volume 3, a margin of 10 was applied to the leak rate to define the accumulator line leakage size flaw used in the stability analysis. The basis for the 1.0 gpm leak rate is the presence (inside containment) of diverse and redundant '

leakage detection systems to measure containment noble gas radioactivity, airborne particulate radioactivity, and containment sump level and flow rate. These systems are designed to alarm in the control room at a setpoint equivalent of less than or equal to 1 gpm. Indication of containment humidity is also provided in the control room. These methods are in compliance with Regulatory Guide 1.45 as discussed in FSAR Subsection 5.2.5.

17

,h Ib ._ .. ._ _ - .. __._.. __i l

. In addition to the above control room alarms and indications, technical specification 4.4.6.2.1 requires monitoring of containment gaseous or particulate radioactivity and normal l sump inventory and discharge at least once per 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. This section of the technical specification also requires -; .

performance of a reactor coolant system inventory balance at least once per 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

1 Detected leaks will be repaired within the system limiting conditions for operation established in either technical

_ :: specifications or administrative procedures. When laakage is detected in reactor coolant pressure boundary piping, technical specification 3.4.6.2 requires that the plant be in hot standby

= within six hours and in cold shutdown with the next thirty hours. Repair would be required before restart.

Experience at other Westinghouse plants indicates a normal background unidentified average leakage rate of between 0.1 gp's and 0.3 gpm, and it has also been demonstrated with pressurized pipe tests that leak rates above 0.1 gpm can be readily detected visually. The undefined leakage rate at STP is expected to be similar to other plants. Experience at similar l plants and the results of these tests indicate that a 1.0 gpm l . leak rate can be reliaHy detected and located during plant operation.

For the low pressure accumulator line piping located from the second check valve off the RCL and continuing to the accumulator tank, additional leak detection is available in the form of accumulator tank level and pressure indication and alarms in the control room. For each accumulator tank there are provided redundant level and pressure indication which alarm on both high and low pressure and high and low level in the control room. The tank range from high level to low level is 300 gallons. As required by technical specification 4.5.1.1 18

[ .,; . . n. . . . . . . . .

~. ._.____ . .. .-

l I~

! l l l i

a check to demonstrate operability must be completed every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. This ensures that over a 12-hour period a leak rate of 0.42 gpm can be identified and isolated to a specific accumulator piping system. A detailed description of these accumulator monitoring systems is discussed in FSAR Subsection 6.3.5.

~

9 e

l 9

6 19

5 _ ; 15.. . ._.....J

9. Question: In order to complete our safety evaluation, NRC needs to evaluate the potential of pipe degradation or failure from indirect causes such as fire, missiles, and component support failure as prevented by design, fabrication, and inspection.

Therefore, the applicant should for completeness prcyide a discussion on the compliance with Standard Review Plan 3.4.1, 3.5.1.2, 3.9.3, 3.9.6, and 9.5.1.

Response: Pipe degradation or failure from indirect causes such as fires, missiles, and component support failure is prevented by designing, fabricating, and inspecting reactor compartments,

~*

components, and supports, to NRC criteria that redu'ce to a low probability the likelihood of the events unacceptably impacting safety related components. STP complies with the criteria in Standard Review Plans 3.4.1, 3.5.1.2, 3.9.3, 3.9.6, and 9.5.1 as discussed in these same sections in the FSAR.

e t

20

.3 _ _ i ;

10. Question: It appears that a lower-bound stress-strain relationship was used in estimating the crack-opening area and thus, the leakage rate. The staff considers it more appropriate to use the average stress-strain relationships in estimating leakage rates. Provide justification that a lower-bound strass-strain .

relationship gives a conservative estimate of the leakage rate.

Response: For the South Texas Project accumulator line, the leak rate calculations and stability calculations have been performed at thegoverninglocation(seefigure5-1,WCAP-11383). At this location,thelowerboundmaterialproperty(ASMESectionIII

~

Code minimum) was used for the leak-before-break calculations (which includes crack stability analysis and leak rate predictions). Use of this property is judged to give conservative conclusions on leak-before-break. While it is true that use of higher yield strength would provide lower crack opening area and would result in lower leak rate predictions, it is equally true that the higher yield strength would result in lower J ,pp value and would result in larger critical flaw size. What is important is that the overall leak-before-break approach should be conservative. In addition, it is not advisable to use two different yield strength values at the same physical location simultaneously.

Such a use of different properties would yield unrealistic results and it will not be possible to assess'the actual margin available with respect to leak-before-break. Further, one should not loose sight of the margins of a factor of ten applied with respect to leak detection systems and the margin l

of a factor of two applied with respect to the leakage size i flaw. In sumary, it is judged that the properties used in WCAP-11383, for leak-before-break demonstration yielded conservative results, l

21 L -. _ - - _ _ - .

r, ..; .

c ;,+ sc'_ -

. z ;; ;__  ; __ _

In order to study the effect of yield strength on the leak rate

]

predictions, the yield strengths of the heats of the materials from WCAP-11383 were averaged (average yield strength = 27945 j psi). The crack opening area and the leak rates were '

calculated using this average yield strength value. ,The flaw size yielding a leak rate of 10 gpa was calculated to be about 3.6 inches. A crack stability analysis was performed for a 7.4 inch long through-wall flaw (> 2 times the leakage size flaw) using the EPRI plastic fracture handbook method. ASME Code minimum yield strengthy(o =19.2ksi)wasusedinthis calculat on. The J,ppj$ g was calculated to be 380 in-lb/in . Since the J,ppjgg is lower than the Jg , of

[ ]***** in-lb/in2 it is concluded that a 7.4 inch long

_ flaw would remain stable when subjected to normal plus SSE loadings.

i i

i l

l 4

22

g .

]

l l

11. Question: WCAP-11383 deals with the Class 1 portion of the accumulator line from the Reactor Coolant System (RCS) cold leg injection points to the first check valve, and WCAP-11351 deals with the Class 1 portion of the accumulator line operating at 120'F.

Provide clarification that the entire accumulator 11her from anchor to anchor, has been considered in the leak-before-break analysis as required by NUREG-1061 Volume 3. If the line contains other than Class 1 piping, the other piping Classes should be considered in the LBB analysis also. Clarify if the line between the accumulator tank and the check valve is Class

- - - 2. With reference to figures 5-1 through 5-3 in WCAP-11351, clarify if the branch piping is Residual Heat Removal (RHR) piping. Also clarify if there is any low energy line in piping, from anchor to anchor.

Response: The loads for the entire span of the 12 inch diameter pipe from the primary loop junction (anchor) to the accumulator tank nozzle junction (anchor) were reviewed to establish the critical location (see figure 11-1). The maximum stress

. location based on normal and SSE loads occurs in the section between the primary loop nozzle and the first valve. The temperature and pressure in this section are 560*F and 2304 psi respectively. The yield strength of the material in this I section at operating temperature is significantly lower than

• l theyieldstrengthintherestofthepiping(wherethe temperatureis120'F). The boundary between Class 1 and Class 2 piping and the location where the piping schedule changes from schedule 140 to schedule 40 are identified in figure 11-1. With reference to figure 5-1 through 5-3 in WCAP-11351, it should be noted that the branch piping is the Residual Heat Removal (RHR) piping. There is no low energy portion in the 12-inch piping between anchor to anchor as shown in figure 11-1, since the pressure and temperature exceed 275 psi and 200'F respectively.

23

! II ,i S.

i

! 2' E '

2 9

un i g I e hecumulator

[

. . Tank

! $ f Classl++ Class 2

~

+

, RCS COLD LES

' i 1 Sch140 Sch40 to o BRANCH SSIILE I

^

y" ANCHOR POINT ANCHOR POINT

\ I b ^

N en- I9 .

E /

i > d /

{ R OPERATING PRESSURE = 2304 psi OPERATING PRESSURE = 650 psi

E E.

OPERATING TEMP = 5600F OPERATING TEMP = 1200F a

8 -

C -

l il "

I l

l l

l l

l

t w rc - .

12. Question: Linear elastic fracture mechanics was used for the fracture stability analysis. However, from the calculated J ,pp, it appears that the associated Irwin plastic zone sizes are not small compared with the half-crack length "a". The staff considers it appropriate to use elastic plastic fracture mechanics. In fact, the applicant used elastic plastic fracture mechanics in the surge line analysis in WCAP-10489 where the J,pp and crack sizes are similar to those for the accumulator line. Justify the use of linear elastic fracture mechanics for the accumulator line or perform a reanalysis

_ using elastic plastic fracture mechanics.

Response: In WCAP-11383 crack stability was demonstrated for a 6.2 inch long through-wall circumferential flaw subjected to normal plus 2

SSE loads. The J,ppjgg was calculated to be 360 in-1b/in j using ASME Code minimum yield strength of 19200 psi. The Von Mises effective stress on the outer surface of the pipe was l 13.7 ksi which is significantly lower than the yield strength of the material. Hence, linear elastic fracture mechanics approach with plastic zone size correction was used to calculate K3 (w ich was converted to Japplied using the relation J = Kg /E). In order to obtain a conservative ,

estimate of Japplied, a conservatively high plastic zone size correction was applied recognizing that the actual plastic zone size would be much smaller. It should be noted that the Japplied without plastic zone size correction is about 150 2

in-lb/in . Therefore the actual J,ppjgg if calculated by ,

elastic plastic approach would be between 150 in-Ib/in (without plastic zone size correction) and 360 in-lb/in 2 (conservativeestimatewithplasticzonesizecorrection).

In section 4 of the report (WCAP-11383), the material certification records for the South Texas Project plants were summarized. As described in WCAP-11256 Supplement 1, the strength properties at operating temperature can be obtained by l

l 25

,, 6.* . ,

using the plant specific room temperature data and the typical primary loop data at high toeperature. The minimum yield strength based on all the heats of materials identified in

! WCAP-11383 is found to be 21540 psi. In connection with

another application, Westinghouse had performed a detailed elastic plastic finite element analysis to calculate the J,ppj $g because the stress level in that application was significantly higher. The finite element model used in that application is shown in figure 12-1. The variation of J with applied bending moment is shown in figure 12-2. The pipe size

- = and wall thickness were the same as in the South Texas Project Units 1 and 2 accumulator lines. Pertinent parameters are compared as follows:

i South Texas Project Available Finite Element Solution l

Crack length = 6.2 inches Crack length = 6.38 inches l Axial force = 194 kips Axial force = 205 kips Yield strength = 21540 psi Yield strength = 21500 psi l From figure 12-2, the Jappliedcorregpondingto1006in-kips

! bending moment is about 250 in-lb/in . Based on the comparison of parameters (shown above,) the Japplied I

the accumulator line in the South Texas Project will be about 250 in-lb/in 2, O

26

qL-}; . 3: .

\

l i

a,c.e l

! Figure 12-1. The Finite Element Model. R =

a = 3.19" (half crack lengtR), L6.375", = 45", 128 t = Elements, 1.005" 861 Nodes.

. 27

. . - - . . - - - . . - . . - . . . . ~ _ . _ - _ _ _ - . - - - . - - . - . - . . _ - - . - - . -

p -

.--... . . =- : ..  ;. . . . . - .

o.

u.

=

~

Figure 2-2. J yas a Function of Noment, N 4

28