ML20137C534

Evaluation of Indications Detected in Annulus Core Spray Piping at Pilgrim Nuclear Station

ML20137C534
Person / Time
Site:	Pilgrim
Issue date:	03/18/1997
From:	Bothne D, Mehta H GENERAL ELECTRIC CO.
To:
Shared Package
ML20137C531	List: ML20137C526 ML20137C534 ML20137C547
References
GE-NE-B13-01869, GE-NE-B13-01869-028, GE-NE-B13-1869, GE-NE-B13-1869-28, NUDOCS 9703250034
Download: ML20137C534 (6)

Text

,

O GENuclear Energy TECIINICAL MODIFICATIONS SERVICES GE-NE-B13-01869-028, Rev.0 March 1997 EVALUATION OF INDICATIONS DETECTED IN THE ANNULUS CORE SPRAY PIPING AT PILGRIM NUCLEAR STATION Prepared for Boston Edison Co.

800 Boylston St.

Boston, MA 02199 Prepared by j

GE Nuclear Energy 175 Curtner Avenue San Jose, CA 95125 9703250034 970328 PDR ADOCK 05000293 G

PDR

~

a GL-NE-Bl3-01869-028. Rev. O GENxclearE erty EVALUATION OF INDICATIONS DETECTED IN THE ANNULUS COIE SPRAY PIPING AT PILGRIM NUCLEAR STATION March 1997 Prepared by; D. Bofhrie, Engineer Engineering and Licensing Consulting Services Verified by

. HT%ehta, Principal Engineer Engineering and Licensing Consulting Services Approved by:

h _

11.S. Mehta, Principal Engineer Engineering and Licensing Consulting Services l

GE Nuclear Energy GE.NE.B13 01869-028. Rev. O TABLE OF CONTENTS Subiect fagt

1. INTRODUCTION AND BACKGROUND..-.

~

~ 3

2. CRACK GROWTH RATE CONSIDERATIONS.

4

3. STRUCTURAL EVALUATION METHODOLOGY 5

3.1. TilREE S1EP STRUCTURAL EVALUATION APPROACH _

5 3.2. TECHNICAL DESCRIPTION OF DLL METIIODOLOGY.

6 3.3. EVALUATION METHODOLOGY ASSUMING NO MOMENT LOAD CARRYING CAPABILI1Y.

7

4. STRUCTURAL EVALUATION-

-.-~

8

.~.--..-

4.1. EVALUATION OF COLLAR WELD SWI AN (IP8B).

9 4.2. EVALUATION OF SLEEVE WELD SW8AN (IPS) 9 4.3. EVALUATION OF COLLAR WELD SWl AS (2P88).

9 4.4. EVALUATION OF COLLAR WELD SWl BN (3 P8b).

10 4.5. EVALUATION OF SLEEVE WELD SW8BN (3P5)..

1I 4.6. EVALUATlON OF COLLAR WELD SWlB5 (4P8B)

.1I 4.7. DISCUSSION OF RESULTS..

1I

5. ADDITIONAL STRUCTURAL EVALUATIONS CONSIDERING COLLAR WELDS I1 S.I. EVALUATION ASSUMING HINGE BOUNDARY CONDmONS -

-12

-13 5.2. INSTALLATION TOLERANN5:.

5.3. FIV EVALUATION ASSUMINO COLLAR WELDS COMPLETELY DETACHED._

-14 5.4. DISCUSSION OF RESULTS.

-15

6. EVALUATION ASSUMING INDICATIONS IN SHROUD.

15

7. EVALUATION OF INACCESSIBLE WELD P9 15

8.

SUMMARY

& CONCLUSIONS..

.I6

9. R E FER ENC ES.

........-.~.....

.~...

..~.-.-... ~

I 8

10. APPENDIX FLAW EVALUATION HANDBOOK FOR PILGRIM NUCLEAR STATION.

28

(

2 I

i i

)

_., - -,. ~.

.._..u.-

, ~..

4 I~

b GE Nxeteer Exergy GE-NE-813-01869-028. Rev. O i

i l

1. INTRODUCTION AND BACKGROUND l

Cracking in the core spray line internal piping at several Boiling Water Reactor (BWR) plants I

has been recently observed. Cracking is believed to be intergranular stress corrosion cracking

{

(IGSCC) typically in the vicinity of the circumferential welds. A core spray line flaw j

evaluation handbook [1] for the Pilgrim Nuclear Station was developed to disposition any indications detected during the in-service inspection. A copy of the handbook is included as l

Appendix A to this report. The handbook followed the procedures outlined in the BWRVIP Core Spray Line I&E Guidelines document [2]. Figures 1 and 2 show the schematic of both loops A and B of the internal core spray lines.

The indications found in the core spray line during the 1997 in-service inspection are summarized in Table 1. The welds are identified in Figures I and 2. Of the six locations where indications have been identified in Table 1, four are at the welds between the collar and shroud. The collar to shroud weld is designated as weld P8b in Reference 2. Figure 3 from Reference 2 shows a schematic of this weld. The collar is not associated with the core spray l

injection flow. Its purpose is to prevent leakage from the inside to the outside of the core shroud at the location where the core spray piping enters the shroud. In addition, it provides the attachment of the core spray piping to the shroud. This adds stiffness to the piping and l

helps to assure that the piping will not be subject to flow induced vibration (FIV).

j

~ A key component of the structural evaluation is the assumed crack growth rate to determine the

{

projected crack length at the end of inspection interval. For this analysis, all evaluations are based on a 2 year operating cycle. The next section discusses the crack growth rate used in this evaluation.

I The objective of this report is to present the results of the structural evaluation of these indications. Two limiting assumptions can be made in evaluating the collar weld indications:

(1) the indications are located in the collar, and (2) the indications are located in the shroud.

The first assumption is more limiting since craciiug in the collar (unlike that in the shroud wall) is likely to be through thickness. Results are presented considering both the assumptions.

A discussion on the susceptibility to cracking of the hidden weld P9 is also included in the report.

i I

GE N: clear Energy GE-NE-BI3-01869-028. Rev. 0

2. CRACK GROWTII RATE CONSIDERATIONS l

The inspection results show that the indications are ofIGSCC origin. Therefore, a key input in the structural margin evaluation in the presence of these indications is the crack growth rate.

The BWRVIP Guidelines document (2] recommends a conservative crack growth rate of 5x10 in/hr.

The preceding crack growth is expected to be conservative for this application based on several reasons. The coolant conductivity at the Pilgrim station has averaged approximately 0.11 S/cm over the last three cycles as compared with a conductivity of ~0.52 S/cm during the first five cycles of operation. Figure 4 shows the crack growth rete predicted by the PLEDGE code as a function of water conductivity. An improvement of better than a factor of 2 is seen as the conductivity decreases from 0.3 S/cm to 0.1 S/cm. The current BWR normal water chemistry guidelines typically set action level 1 at conductivity greater than 0.3 S/cm. Thus, if one were to assume that the 5x10~5 in/hr growth rate is applicabic to a conductivity of 0.3 S/cm, the predicted crack growth rate expected at the Pilgrim station is approximately 2.5x10 5 in/hr (assuming an ECP of 200 nN SHE).

Another approach to calculate the average crack growth rate is to base it on the largest indication length divided by the operating hours since the assumed initiation. The largest indication length in Table 1 is 5.5 inches. The Pilgrim station has been in commercial operation since 1972. The field experience shows that cracking in the core spray piping or spargers has been seen as early as 8 years of commercial operation. If we assume that the initiation occurred after approx.12 years of commercial operation, then the crack growth period is approximately 96000 hours. This implies a crack growth rate of (5.5/[2x96000]) or 2.9::10 5 in/hr. The factor of 2 in the preceding calculation accounts for the fact that the crack grows at each end of the indication.

5 Based on the preceding discussion it is concluded that the crack growth rate of 5x10 in/hr used in this evaluation is conservative.

4

GEltclear Energy GE-NE-Bl3-01869-028. Rev. 0 l

3. STRUCTURAL EVALUATION METHODOLOGY The UT data show that the indications at the collars appear to be located in the HAZ of the core shroud itself, rather than in the collar. This in itselfis not a unique finding, in that three other BWR plants have also detected such indications and two other plants have detected indications in the shroud at locations where brackets have been attached to the shroud by fillet welds. The I

subject core spray collar is welded to the outside of the shroud with a somewhat similar weld.

Therefore, the structural evaluations presented in this report were conducted using two limiting assumptions: (1) the indications are located in the collar, and (2) the indications are located in the shroud. The first assumption is more limiting since cracking in the collar (unlike that in the shroud wall) is likely to be through thickness. The evaluation based on the second assumption is described in Section 6.

3.1. Three Step StructuralEvaluation Approach The stnictural evaluation approach assumed in this Section is that the indications at the collar welds are located in the collar. A three step technical approach, depending upon the inspection interval considered, was used in evaluating the indications at various welds.

The allowable flaw lengths at each weld are given in the flaw handbook [1]. In step one, a conservative evaluation approach is used. In this approach, the projected crack growth for the assumed inspection interval is added to each indication at a weld and a cumulative flaw length is obtained by summing the lengths of all the indications. A flaw proximity rule, as described next, is used in determining the cumulative flaw length. If two iudications are less than '2t' apart after accounting for crack growth, where 't' is the thickness of the collar or the pipe, then the remaining ligament between such indications is also included in calculating the cumulative flaw length. If this cumulative flaw length is less than the allowable flaw length given in the handbook for this weld, the indications at a weld are acceptable for continued service for the assumed inspection interval. This approach is conservative since all of the indication lengths are stacked together and thus no use is made of the added load carrying capability of the cracked section due to the fact that the remaining ligaments are distributed over the circumference.

5

GE Nxclear Ezergy GE-NE-BI3-01869-028. Rev. 0 If the first step evaluation produces unacceptable results, then the second step evaluation, as r

2 stated in the preceding paragraph, consisted of making use of the fact that the remaining ligaments between the multiple indications at a cracked section are distributed over the circumferential length of the weld. The 'DLL' computer program [3] was used in such evaluations. The technical background on the 'DLL' methodology is briefly described in Subsection 3.2.

Assuming the ligament geometry after accounting for crack growth, the smallest safety factor is determined in this approach and is compared with the appropriate j

required value for either normal / upset (2.77) or emergency / faulted (1.39) condition.

j If the 'DLL' evaluation does not result in an acceptable safety factor value for the assumed inspection interval, then the third step evaluation involves assuming ' hat the collar weld has no bending moment load carrying capability. This evaluation approach is described in Subsection 3.3.

3.2. Technical Description of DLL Methodology The DLL, or distributed ligament length methodology, was used to evaluate the structural margins for the various ligament lengths.

. In the limit load case, it is assumed that there are 1,2,...i,...n ligament lengths and that the i*

length is of thickness t and extends from an azimuth of G to On. The ligament length l of the i

ii i

i* ligament is related to the azimuth angles O and Oa by the following:

ii i

l = (D/2)*(0 - On)

(1) i ii where D is diameter. The calculation of the moment M that this ligament configuration can l

resist is complex since it is not clear which azimuth orientation of the neutral / central axis

~ would produce the least value of bending moment M. Therefore, the value of M is calculated for various orientations of the central axis from 0 to 360. First, a central axis orientation a is selected. The location of the neutral axis is determined by the following:

fa-..p> Rt(0)d0 -* '" Rt,d6 = (o, / o )* (2nRt,)

(2) r

.uo 6

1 GE N: clear E:ergy GE-NE-BI3-01869-028, gey. o

{

l Assumed azimuth angle of the central axis where a

=

Angle of neutral axis with respect to central axis, or sin" (S/R).

i p

=

Distance between central axis and neutral axis.

6

=

Mean radius.

i R

=

I t (thickness ofith ligament), if angle 0 is On < 0 < Oa, or 0.

t(0)

=

i Nominal thickness t

=

o Membrane stress o,

=

Material flow stress = 3S, or

=

This step defines the location of the neutral axis when the central axis is assumed to be at an 1

azimuth angle of a.

1 The second step, once the location of the neutral axis relative to the central axis is determined, is to determine the moment M.

This is found by integrating the bending moment j

o contributions frem individual ligament lengths. The orientation a that produces the least value of M is called a, min and defines the axis capable of resisting the limiting moment. Whether the specified set of uncracked ligament lengths provides required structural margin is verified f

by Ma, min /Z + P 2 SF(P + P + P,/SF)

(3) 3 Section modulus based on uncracked cross section.

where Z

=

Applied primary membrane stress.

P.

=

Applied primary bending stress.

I'd

=

Applied secondary bending stress.

P,

=

Safety factor.

SF

=

3.3. Evaluation Methodology Assuming no Moment Load Carrying Capability In this evaluation, it is assumed that the collar weld has no moment load canying capability.

The first step in this evaluation is to determine the maxircum value of the axial force at the subject collar weld location. Since the collar weld is now assumed to have no moment carrying capability, the finite element model of the core spray pipmg system, developed to determine the stresses from various operating condition loads for use in the development of flaw handbook, was modified to reflect hinged boundary conditions at both the shroud penetration points. All of the load combination cases are then rerun with this modified boundary condition 7

GE N: clear Energy GE-NE-Bl3-01869-028. Rev. O and a value of the highest axial force is calculated. The shear forces at this weld need not be considered since they are reacted by the core spray piping bearing against the surface of the hole in the shroud at the penetration. A safety factor is then calculated as the following:

SF

= lxa /(AxialForce)

(4) r

where, I

= Remaining ligament length The calculated value of SF is compared with the appropriate required values for normal / upset (2.77) and emergency / faulted (1.39) constions.

Fundamental frequency of the annulus core spray piping with the hinged boundary condition assumed at the collar welds, was also determined. The objective was to evaluate if the change in the fundamental frequency would pose any FIV concerns.

A very conservative evaluation would be to further assume that the collar weld has completely detached and that the core spray pipe in the shroud hole can move by a di.itance in the vessel radial direction permitted by the sparger installation tolerances. In addition, the core spray pipe can move laterally into the shroud hole. The concern then would be the FIV stresses produced in the core spray piping system. Based on the installation tolerances, the maximum values of these potential pipe movements were determined and the maximum attemating stress amplitude was calculated and compared to the threshold value of 10000 psi. Additionally, a primary plus secondary stress evaluation was conducted to show that even when the FIV stresses are added to the other load case stresses, the largest stress range is still less than the ASME Code allowable value of 3S.

m

4. STRUCTURAL EVALUATION This Section describes the results of structural evaluation of all of the indications reported in Table 1. At the collar welds, the structural evaluation conservatively assumed the indications to be on the collar side. The structural evaluation of collar indications assuming the indications to be on the shroud side is discussed in Section 6.

8

.,.AJ-.

A-.

4 5

S -, -.

J-*

a 4

--4A~-

+.,S-RJ a

-+L E

f.

GEN: clear Exergy GE-NE Bl3-01869-028. Rev. 0 4.1. Evaluation of Collar Weld SW1AN (1P8b)

A review of Table I shows that there are three indications at this weld. The flaw handbook [1]

l gives an allowable flaw length of 9.4 inches for this weld. The total cumulative length of the 4

l indications after accounting for crack growth (based on a crack growth rate of 5x10 in/ hour) for one fuel cycle of operation was calculated as 14.5 inches. This length exceeds the allowable value of 9.4 inches. However, this is a conservative approach, therefore, the second step approach of structural evaluation using the DLL computer code was then used.

The P,6, and P, stress magnitudes used in the DLL evaluation are given in Table 2. As stated earlier, this computer program makes use of the fact that the remaining ligainents between the multiple indications at a cracked section are distributed over the circumferential 4

length of the weld. A crack growth rate of 5x10 in/ hour was used in adjusting the lengths of the indications for input into the DLL code.

The results of this evaluation show that the calculated safety factor at the end of one fuel cycle of operation is 1.42 compared to the required value of 1.39 or higher. Thus, acceptable structural margin is available at this weld for one fuel cycle of operation. For operation beyond one fuel cycle, the calculated stmetural margin becomes less than the required value. This only l

means that the moment load carrying capability is not assured. However, an additional structural evaluation using the approach of Subsection 3.3, is described in Section 5 which demonstrates that operation beyond one fuel cycle in the as-is condition can be justified.

4.2. Evaluation of Sleeve Weld SW8AN (1P5)

Table I shows that there is a single indication of 0.8 inch length at this weld. The allowable i

flaw length per the handbook assuming two fuel cycles of operation is 7.0 inches. Clearly, the allowable value considerably exceeds the indication length and thus operation for more than two fuel cycles of operation is justified.

4,3. Evaluation of Collar Weld SW1AS (2P8b) 1 l

l A review of Table i shows that there are three indications at this weld. The flaw handbook (Reference 1) gives an allowable flaw length of 9.4 inches for this weld. The total cumulative t

9

. _. ~

GE N1 clear Ezergy GE-NE-B13-01869-028. Rev. O length of the indications after accounting for crack growth (based on a crack growth rate of 5x10-5 in/ hour) for one fuel cycle of operation was calculated as 10.8 inches. This length exceeds the allowable value of 9.4 inches. This is a conservative approach, therefore, the second step approach of structural evaluation using the DLL computer code was then used.

The P., P and P, stress magnitudes used in the DLL evaluation are given in Table 2. As 3

stated earlier, this computer program makes use of the fact that the remaining ligaments between the multiple indications at a cracked section are distributed over the circumferential length of the weld. A crack growth rate of 5x10 5 in/ hour was used in adjusting the lengths of the indications for use in the DLL code.

The results of this evaluation show that the calculated safety factor at the end of 3.5 years of operation (more than one but less than two fuel cycle of operation) is 1.58 compared to the f

required value of 1.39 or higher. Thus, acceptable structural margin is available at this weld for more than one fuel cycle of operation. For two or more fuel cycle of operation, the calculated structural margin becomes less than the required value. This only means that the moment load carrying capability is not assured. However, an additional structural evaluation i

using the approach of Subsection 3.3, is described in Section 5 which demonstrates that operation beyond one fuel cycle in the as-is condition can be justified.

4.4. Evaluation of Collar Weld SW1BN (3P8b)

A review of Table I shows that there are three indications at this weld. The flaw handbook [1]

gives an allowable flaw length of 9.9 inches for this weld. The total cumulative length of the indications after accounting for crack growth (based on a crack growth rate of 5x10-5 in/ hour) for one fuel cycle of operation was calculated as 8.5 inches. This length is less than the allowable value of 9.9 inches and is thus acceptable. The cumulative flaw length at the end of two fuel cycles of operation was calculated as 10.1 inches. This value slightly exceeds the allowable value of 9.9 inches. The DLL evaluation also showed similar results for the two fuel cycles case. Thus, it is concluded that more than one fuel cycle of operation is justified.

10

GE Nzclear E;ergy GE-NE BI3-01869-028. Rev. 0 4.5. Evaluation of Sleeve Weld SW8BN (3P5)

Table 1 shows that at this weld there are two indications of 1.8 inch and 0.4 inch lengths, respectively. Accounting for two cycles of crack growth, the cumulative crack length was calculated as 8.6 inches. The allowable flaw length per the handbook is 11.1 inches. Clearly, the allowable value considerably exceeds the cumulative crack length and thus operation for more than two fuel cycles of operation is justified.

4.6. Evaluation of CoIIar Weld SW1BS (4P?b)

Table I shows that there is a single indication of 1.8 inch length at this weld. The allowable flaw length per the handbook, including crack growth for two fuel cycles of operation, is 6.7 inches. Clearly, the allowable value considerably exceeds the indication length and thus operation for more than two fuel cycles of operation is justified.

4.7. Discussion of Results Table 3 summarizes the results of the preceding structural evaluations. It is seen that the least f

structural margin is at the collar weld SWI AN where one fuel cycle of as-is operation can be justified. At the other collar weld, SWl AS, more than one but less than two fuel cycle of as-is operation can be justified. At the remaining locations with indications, more than two cycles of as-is operation can be justified. Thus, additional analyses, as described in the next Section are necessary to justify as-is operation for two or more fuel cycles of operation.

5. ADDITIONAL STRUCTURAL EVALUATIONS CONSIDERING COLLAR WELDS The structural analysis results in Section 4 show that moment load carrying capability at two collar to shroud welds (SWl AN and SWl AS) is not assured in the as-is condition during the second and subsequent fuel cycles of operation. Subsection 3.3 describes the structural analyses to assure the adequacy of the remaining ligament and to address the FIV concerns.

Thus, these structural analyses can demonstrate that even though a collar weld may not have moment load carrying capability and even ifit is detached from the shroud, the FIV stress and Code stress range criteria are satisfied. Once these criteria are satisfied, further continued 11

GE N: clear E ergy GE-NE-BI3-01869-028. Rev. O l

t l

operation in the as-is condition can be justified. This Section describes the results of such j

structural analyses.

A review of the collar weld indications in loops A and B indicated that loop A presents the bounding case. Therefore, the finite element model of loop A was used to conduct these l

additional structural analyses.

5.1. Evaluation Assuming Hinge Boundary Conditions This evaluation was conducted in two steps. First the axial load carrying capability of the remaining ligament at the limiting collar weld (SWl AN) was determined using Equation (4).

In the second step, the finite element model of loop A was modified to reflect the hinge boundary condition at the collar welds and the magnitude of maximum axial force calculated.

The ratio of the axial load carrying capability determined in step I to the maximum axial force calculated in step 2, is the safety factor.

The remaining ligament at the end of two fuel cycles of operation is calculated as 2.5 inches.

The ligament area is (2.5x0.1525) or 0.39 in. The axial force carrying capability of this 2

ligament was calculated as 19770 lbs assuming a material flow stress of 50700 psi (S, = 16900 psi @ 550 F and flow stress = 3S ).

m The next step in the evaluation process was to determine the maximum axial force at the collar weld location. Since the two collar welds on loop A are now assumed to have no moment carrying capability, the loop A finite element model used in developing the flaw handbook was modified to reflect hinged boundary condition (i.e., all three rotation constraints released but the three displacement constraints retained) at both the shroud penetration points. Figure 5 l

shows the finite element model with the modified boundary conditions indicated. For clarity, Table 4a shows a summary of the original boundary conditions at various support points and Table 4b shows the modified boundary conditions for the assumed hinge at collar to shroud welds. All of the stress runs and the load combination cases were then rerun with these modified boundary conditions.

The maximum axial force was calculated as 909 lbs corresponding to the faulted load t

combination that includes, among others, the pressure, dead weight, LOCA and SSE loadings.

i I

l 12

GE Nxeled Energy GE-NE-BI3-01869-028. Rev. O Dividing the axial load carrying capability of 19770 lbs by 909 lbs gives a safety factor of 21.7 compared to the required safety factor of 1.4 for the emergency / faulted conditions.

The fundamental frequency of the original model was 24.9 Hertz. With the modified boundary condition, the fundamental frequency decreased slightly to a value of 23.9 Hertz. It was judged that this small change in fundamental frequency would not pose any FIV concems.

5.2. Installation Tolerances The preceding structural evaluation concluded that even though the collar weld may act like a hinge it would not tear off under axial loading during next two fuel cycles of operation. The next Subsection considers the unlikely scenario in which the collar weld has completely detached and evaluates the likelihood of the section of the piping near the collar rattling under flow induced forces during normal plant operation. A key parameter in this evaluation is the potential clearance between the sparger T-box and the inside surface of the shroud.

The ID of the Pilgrim core spray sparger is specified on GE drawing 719E427 as 184.25 minimum. If one adds eight inches for the two 3.5-inch NPS nominal pipe diameters, the expected OD, ignoring " egging" of the pipe cross section, is 192.25 inches minimum. The corresponding shroud ID from GE drawing 719E427 is 192.5/192.75 inches. The maximum gap, smallest sparger in the largest shroud, is 1/2 inch diametral (1/4 inch radial). Thus, the design drawings can not be used to prove that the sparger is in contact with the shroud ID.

However, there are other arguments which suggest that the sparger would be in contact with the shroud ID near the sparger Tee location.

The shroud drawing specifies graphically (depiction) that the sparger pipe OD should be line-on-line with the shroud ID. The sparger pipe support brackets are also shown in contact with the sparger pipe, thus the fabricator should have trimmed the brackets to fit as close to the pipe as possible. Weld shrinkage from welding the brackets to the shroud would tend to pull the pipe toward the shroud ID. The important control dimension for installation of the sparger into the shroud is the 184.25 inches minimum sparger inside diameter (including nozzles and fittings) as specified on GE drawing 719E427 Since the shroud ID equals or exceeds the expected formed sparger OD, it is expected that the installer would balance any sparger pipe to shroud ID radial gap at the ends of the semi-circular sparger sections and place the center of 1

the semi-circular sparger (near the sparger Tee) in contact with the shroud ID. The installer 13 4

j

l GEN clearEnergy GE-NE-BI3-01869-0.28. Rev. O may have also jacked the sparger pipe from the opposite side of the shroud to optimize his chance of meeting the 184.25 inches minimum.

Based on the preceding discussion, it can be concluded that the core spray pipe can move up to a maximum of 1/4 inch in the vessel radial direction if it is assumed that the collar weld is detached from the shroud.

)

The magnitude of translational moveinent of the core spray pipe in the shroud hole was determined as follows. GE's shroud drawing for Pilgrim is 719E427. In this drawing, Sections DD & EE, detail @ zone B-4, a radial clearance of 1/32 is specified between the 5 inch pipe and the hole in the shroud. The same information is also shown on the Sun Ship drawing.

Thus, a zero-to-peak displacement of 1/32 inch or a displacement range of 1/16 inch, was used in the structural evaluation desoibed in the next Subsection.

i 5.3. FIV Evaluation Assuming Collar Welds Completely Detached The discussion in the preceding Subsection indicates that if the collar welds are assumed to have completely detached from the shroud (i.e., zero remaining ligament), the pipe in the shroud hole could potentially move 1/4 inch in the vessel radial direction. A similar review of the diametral clearance at the shroud hole indicated that the core spray pipe can also move 1/16 inch in the lateral direction. Thus, if the collar were to get completely detached, the pipe at that location can move 1/4 inch in the vessel radial direction and 1/16 inch in the horizontal (tangential to shroud surface) and vertical directions. A conservative approach was used to calculate the FIV stresses as described next.

Figure 5 shows the finite element model of the loop A core spray system with the boundary conditions indicated therein for this analysis. For clarity, Table 4c also shov s the boundary conditions and displacements used in this analysis. The six displacement load cases (1/4 inch in the vessel radial direction,1/16 inch in the horizontal and vertical directions at each of the collar welds) were first run. The resulting moments from the six load cases were combined by the absolute sum method and the peak stress range at each of the weld was calculated using the appropriate stress concentration factor (C K ).

The stress concentration factor values used 2 2 were 4.2 at the fillet welds (e.g., at the vertical slip joint) and 1.8 at the groove welds. The maximum value of the calculated alternating stress (one-half of the peak stress range) was 2350 psi. Obviously, this value is considerably less than the threshold value of 10000 psi. Based on 14

GE-rVE-BI3-01869-028. Rev. O GENaclear Energy this, it is concluded that even if the collar to shroud welds are assumed detached from the shroud and the core spray pipe could move the maximum magnitude permitted by the clearances, it would not pose FIV concerns.

A primary plus secondary stress evaluation showed that even when the FIV stresses are added to other load case stresses, the largest stress range is still less than the allowable value of 3S.

5.4. Discussion of Results The stmetural analysis results presented in this Section demonstrate that even if a small ligament were to remain at the collar to shroud welds or these welds were to completely detach from the shroud, these conditions do not exceed any ASME Code or FIV stress criteria. Thus the operation in the as-is condition for the second fuel cycle is technically justified.

6. EVALUATION ASSUMING INDICATIONS IN SHROUD The nominal thickness of the shroud at this location is 1.75 inches. These indications are expected to be shallow (typically, less than 0.35 inch) based on field experience related to 4

shroud cracks at sparger brackets. A bounding crack growth rate of 5x10 in/hr has been used in the shroud integrity evaluations. This rate implies a crack growth of 0.8 inch for 24-month cycle of operation (~16000 hrs). This would mean a predicted crack depth of less than one inch. Thus, if these indications are assumed to be in the shroud, they are not expected to grow through the shroud wall during the next cycle of operation. Furthermore, even if it assumed that these indications become through-wall, the shroud flaw tolerance at this location is in excess of 100 inches.

Based on the preceding discussion it is concluded that shroud structural margins will be maintained if the indications are assumed on the shroud side.

7. EVALUATION OF INACCESSIBLE WELD P9 Although the BWRVIP Core Spray I&E Guidelines [2] refer to P9 as a crevice weld, it has very different characteristics than the P8b welds which were found to have indications. Field data supports the conclusion that this region does not behave like a true severe crevice and therefore significant cracking would not be expected in this weld. The P9 weld is expected to have a 15

4 GENxclear Exergy GE-NE-BI3-01869-028. Rev. O susceptibility similar to other girth butt welds in the core spray piping which have been inspected either ultrasonically or visually this outage and show no indications. The sleeve welds where

]

indications were noted are fillet welds which have a natural crevice geometry.

)

Since none of the ginh butt welds in Pilgrim internal core spray piping show incidence of

)

i cracking, there is a strong basis for concluding that the currently uninspectable weld P9 has not experienced significant IGSCC. Even if some amount ofIGSCC has occurred in weld P9, the Pilgrim Core Spray Line Flaw Evaluation Handbook [1] indicates that this location could tolerate l

1 i

an existing through wall crack of more than 9 inches in length. This is typical of other girth butt welds in the system where the allowable crack sizes are in excess of 9 inches. With the collar weld disengaged from the shroud, (i.e.1:ss cf-oment restraint) the allowable flaw size at weld.

P9 will still be bounded by that calculated with the collar weld in tact. This is typical in a piping system near a moment restraint (i.e. an anchor). If the restraint is released, bending moments in the piping near the restraint actually decrease as the load redistributes through other load paths.

In this vicinity, the reduced bending increases the flaw tolerance of weld P9 (i.e., increases the allowable flaw size).

In addition, the collar-to-shroud weld may have a greater susceptibility to IGSCC than other core spray piping welds because of the difficulty of the weld orientation which may have resulted in an inherent root defect. Therefore, the inspection findings of this weld are not expected to be a good indicator of the susceptibility of other welds in this location such as the P9 girth butt weld.

The susceptibility of this weld is expected to be similar to the other non-crevice groove welds in the core spray piping for which no indications have been detected in this outage. Therefore it is concluded that weld P9 is capable of performing its intended function without verification of its integrity by examination.

4 In summary, the results of this discussion show that this weld is not a true crevice weld, the i

favorable inspection results from similar welds, and the large flaw tolerance of weld P9, support the conclusion that no inspection of weld P9 is necessary.

8.

SUMMARY

& CONCLUSIONS An evaluation has been performed for two limiting assumptions: (1) the indications are located in the collar, and (2) the indications are located in the shroud. The first assumption is more limiting since cracking in the collar (unlike that in the shroud wall) is likely to be through thickness. The result of this evaluation shows that even for the most limiting assumption of 16

GE N::cicar Exergy GE-NE-B13-01869-028, Rev. O cracking in the thermal sleeve collar and using the previously NRC accepted conservative flaw 5

growth rate of 5 x 10 in/hr, the structural margin requirements are met in the as-is condition for at least one fuel cycle of operation. Additionally, the leakage from the collar is not a concern because (1) the core spray injection flow is not affected, (2) any leakage of core flow would be very small due to the tight tolerances between the piping and the core shroud at the I

penetration location. For operation beyond one fuel cycle of operation, the presence of the uncracked ligament assures that the natural frequency remains nearly unchanged, thus i

I eliminating any FIV concerns. Adequate structural margin is also maintained if the indications are located in the shroud.

The favorable inspection results from similar welds, and the large flaw tolerance of weld P9, support the conclusion that no inspection of weld P9 is necessary.

It is concluded that the stmetural integrity of the core spray piping will be maintained and the indications are acceptable for continued operation during at least one fuel cycle without repair.

All calculations for this analysis can be found in the corresponding Design Record File (Reference 4).

17

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

i f

GENzcle:s Exergy GE-NE-BI3-01869-028. Rev. O i

9. REFERENCES

[1]

" Internal Core Spray Flaw Evaluation for Pilgrim Nuclear Power Station," GE Report No. GE-NE-B13-01869-02, Rev.1, March 1997.

"BWR Core Spray Internals Inspection and Flaw Evaluation Guidelines," Prepared for

[2]

l the BWRVIP, Report No. GE-NE-B13-01805-21, June 1996.

7 i

l

[3]

"BWR Core Shroud Distributed Ligament Length Computer Program (Version 2.1, dated 09/19/96)," GE Report No. GE-NE-523-ll3-0894, Supplement 1, Revision 1, October 1996.

l l

[4]

Design Record File GE-NE-B13-01869-028, March 1997.

i I

l I

i I

l i

i t

l s

4 4

d 18 n

e GENxclear Energy GE-NE-BI3-01869-028. Rev. 0 l

l l

Table 1 Indication Geometry and Ligament Calculation l

I Weld Ind.#

Start (Deg.)

Stop (Deg.)

Total (Deg.)

Total (In.)

Scan Angle (Deg.)

SWl AN (IP8B) 1 325.18 64.02 98.84 5.5 70*

collar 2

118.12 186.66 68.54 3.8 70' 3

197.12 222.37 25.25 1.4 70*

SW8AN (IPS) 1 204.74 221.23 16.49 0.8 60' sleeve SWI AS (2P88) 1 308.95 10.27 61.32 3.4 70*

collar 2

27.94 49.95 22.01 1.2 70 1

3 132.55 168.63 36.08 2.0 70*

SWIBN (3P8B) 1 318.47 29.87 71.4 3.95 70' j

collar 2

265.97 269.59 3.62 0.2 70*

1 3

299.46 306.70 7.24 0.4 70*

SW8BN (3PS)

I 113.37 150.47 37.1 1.8 60*

]

sleeve 2

266.57 274.82 8.25 0.4 60*

i SWlBS (4P8B) 1 106.41 138.88 32.47 1.8 70*

collar l

i 19

GE Nuclert Exergy GE-NE-B13-01869-028. Rev. 0 1

i 1

Table 2 Collar Weld Stresses used in DLL Evaluations I

l Weld ID Governing Condition Stresses i

(psi) i Pm P

P, 8

SWlAN/SWlAS Emerg/ Faulted 738 697 10881 SWlBN/SWlBS Emerg/ Faulted 739 735 9347 l

~

l a

Table 3 Summary of Structural Evaluation Results j

i 1

l 1

i j

Analysis Evaluation Results -Requirea j

Component Weld Approach Struct. Margin Maintained for i

1. of Fuel Cycles Collar SWl AN (IP8b)

DLL 1

1 Sleeve SW8AN (1P5)

Handbook

>2 i

Collar SWl AS (2P8b)

DLL

>l Collar SWlBN (3P8b)

DLL

>2 l

Sleeve SW8BN (3PS)

Handbook

>2 j

Collar SWl AS (4P8b)

Handbook

>2 1

i i

1 1

1

.i i

l

~

]

1 20

,