ML14202A615

ASME Section XI Inservice Inspection Program Request for Alternative REP-SI: Proposed Alternative to Requirements for Repair/Replacement Activities for Certain Safety Injection Pump Welded Attachments

ML14202A615
Person / Time
Site:	Diablo Canyon
Issue date:	07/21/2014
From:	Pacific Gas & Electric Co
To:	Office of Nuclear Reactor Regulation
References
50-275-OL, 50-323-OL
Download: ML14202A615 (30)

Text

predicted to grow when residual stresses are considered. For conservatism, K in the absence of residual stresses are included in the figures and are employed in the present analyses of crack stability and fatigue crack growth.

Since this evaluation does not involve real flaws detected in service but supports operability with postulated flaws, ASME Service Level D conditions are assumed in the evaluation, consistent with the philosophy in USNRC Inspection Manual, Part 9900 [17] which states that the criteria of Appendix F of ASME Section III (which deals with Service Level D conditions) can be used for operability determination.

Structural factors are provided in ASME Section XI, Article C-2620 [2]. For Service Level D conditions, structural factors of 1.3 and 1.4 are provided for membrane and bending stresses respectively for circumferential flaws. For conservatism, a structural factor of 1.4 will be used for evaluation of allowable flaw size. For residual stresses, ASME Section XI, Article C-7300, stipulates a structural factor of 1.0 [2]. Stress intensity factors as a function of OD flaw depth for maximum loads, excluding compressive residual stresses but including applicable structural factors, are shown in Figure 3-3.

3.2.5 Fracture Toughness Properties A material's linear elastic fracture toughness K1c is required to evaluate allowable flaw size for the LEFM failure mode. The K1c for Type 410 stainless steel in the as-quenched or un-tempered condition is not readily available. As recommended by ASME Section XI Article C-8330 [2],

data from the literature are used to select conservative toughness values, as described in the following paragraphs. These data are obtained from impact testing and elastic-plastic fracture toughness testing of Type 410 stainless steel. Referenced data is applicable to Type 410 in the un-tempered condition or tempered within the embrittling temperature range.

Industry data for mechanical properties of martensitic stainless steels indicate that toughness and ductility are affected by heat treatment, particularly tempering. As discussed in [1], the probable condition of the Type 410 stainless steel weld heat-affected zone is primarily un-tempered ReportNo. 1301620.402.R2 3-6 SJ Structural Integrity Associates, Inc.

martensite. Figure 3-4 and Figure 3-5 show data for impact energy as a function of tempering temperature [ 18, 19]. The data show a minimum in impact energy when tempering is done in the temperature range 850-885°F. Impact energy of as-quenched 410 is estimated to be about equal to this minimum impact energy, approximately 18 ft-lbs for Type 410.

Various empirical correlations exist to estimate fracture toughness from impact energy measurements. Barsom and Rolfe [20] report a conservative lower-bound correlation between room-temperature Charpy V -notch ( CVN) impact energy and room-temperature fracture toughness (Kic), according to the below equation:

(4) where K1c is in ksi"in and CVN is in ft-lbs. This correlation is applicable to low-alloy steels, however fracture and fatigue resistance of martensitic stainless steels is similar to martensitic low-alloy steels having comparable strength and hardness [21]. In the case of as-quenched Type 410 stainless steel, according to Equation 4, a lower-bound CVN of 18 ft-lbs corresponds to a K1c of 58 ksi"in.

Fracture toughness measurements were obtained for specimens from failed Type 410 valve studs presumed to have been temper-embrittled, i.e. per this reference, tempered in the range 1 000-1 050°F [22]. Charpy impact data indicated low toughness when tested at temperatures between 40 and 120°F, with a lower-shelf CVN of approximately 10 ft-lb. Elastic-plastic fracture toughness tests were conducted, but no valid K1c measurements were obtained due to pop-in behavior and measured toughnesses too close to the precracking level. Fracture toughness tests did allow the calculation of KQ and K1c, which are fracture toughness parameters calculated under elastic-plastic conditions. KQ is calculated at the point of maximum load, while K1c is calculated at the initiation of crack growth. These KQ and K1c values suggest that the lower-shelf (lower-bound) fracture toughness of presumed temper-embrittled Type 410 is about 25-30 ksi"in (Figure 3-6).

Report No. 1301620.402.R2 3-7 e

Structural Integrity Associates, Inc.

These values are similar to K1c measurements [21] available for Type 403, a martensitic stainless steel similar to Type 410, tested at a range of temperatures (Figure 3-7). The data suggest a lower-shelf toughness in the range 30-35 ksiv'in for Type 403, which supports the values estimated for Type 410.

Based on the reviewed data, a K1c of 25 ksiv'in is assumed for un-tempered Type 410 in subsequent allowable flaw size calculations.

3.3 ID Flaw 3.3.1 Evaluation Methodology The evaluation methodology for determining acceptability of postulated ID flaws is the same as for OD flaws, described in Section 3.2.1, with the exception that the relevant geometry for the flaw is a semi-elliptical circumferential flaw originating on the ID of the pipe and growing toward the pipe OD.

3.3.2 Flaw Geometry Residual stresses (Figure 2-6a) are found to be small or strongly compressive near the OD but strongly tensile at the ID, suggesting that a flaw at the ID should also be considered. Residual stresses would not contribute to fatigue crack growth. However, for the evaluation of allowable flaw size, a flaw at the ID (Figure 3-1 b) is also evaluated to address tensile residual stresses at the ID.

3.3.3 Operating Loads For the evaluation of allowable ID flaw size, loads and the procedures used to combine loads and obtain equivalent axial force are the same as described in Section 3.2.3.

Report No. 1301620.402.R2 3-8

~Structural Integrity Associates, Inc.*

3.3.4 Stress Intensity Factor versus Crack Size As discussed in Section 3.2.2, residual stresses are more strongly tensile at the ID than at the OD, suggesting that a flaw located at the ID (Figure 3-1 b) should also be considered for the evaluation of allowable flaw size. The influence function from API-579 [3] has been incorporated into pc-CRACK' [23], and this fracture mechanics code is used to compute stress intensity factor K as a function of ID flaw depth for applied loads due to internal pressure, maximum operating loads, and residual stress. The applied stress intensity factors for ID flaws are linearly superimposed to determine the total stress intensity factor, including applicable structural factors, in a similar manner to OD flaws (see Section 3.2.4). Results are plotted in Figure 3-8. Residual stresses are included in the evaluation of K for flaws at the ID, because residual stresses increase K for ID-connected flaws.

3.3.5 Fracture Toughness Properties A fracture toughness of25 ksi.Yin, as described in Section 3.2.5, is used in the evaluation of allowable ID flaw size.

3.4 Results Crack instability of OD flaws and ID flaws in the Type 410 stainless steel pipe nipple is determined by calculating the crack depth at which applied stress intensity factor K exceeds material fracture toughness K1c. Based on fracture mechanics properties of the Type 41 0 stainless steel described in Section 3.2.5, a postulated flaw must produce a K greater than the assumed K1c of25 ksiv'in to produce unstable crack growth.

3.4.1 OD Flaw The critical OD flaw size is evaluated using Figure 3-2, which shows stress intensity factor K as a function of OD flaw depth a/t for semi-elliptical, circumferential, OD-connected flaws under Report No. 1301620.402.R2 3-9 SJ Structural Integrity Associates, Inc.

maximum operating loads and pressures, i.e. including seismic/abnormal events. By definition, structural factors are not included in calculation of critical flaw size. Under these maximum loads, including residual stresses, the stress intensity factors do not exceed 25 ksi.Vin for flaws as deep as 80% of the wall thickness, the maximum flaw depth for which the stress intensity factor solution is valid. The residual stresses are primarily compressive and are beneficial in reducing the stress intensity factors. With residual stresses, the critical flaw depth is shown to exceed 80%

of the wall thickness, corresponding to a depth a of 0.123 inch and a surface length 2c of at least 0.246 inch for crack aspect ratio cia= 1. For added conservatism, residual stresses are not considered in the determination of allowable crack size.

The allowable OD flaw size is calculated by applying structural factors specified in ASME Section XI, Article C-2620 [2], with results shown in Figure 3-3. For an assumed K1c of25 ksi.Vin, the allowable flaw depth alt is 0.716 when cia= 4, corresponding to a flaw depth a of 0.110 inch and surface length 2c of at least 0.880 inch. When cia= 1, K does not exceed 25 ksi.Vin.

3.4.2 ID Flaw As discussed in Section 3.2.2, although the most likely postulated flaw is located at the pipe OD due to the stress concentration introduced by the weld, tensile residual stresses at the pipe ID suggest that a flaw on the ID should also be considered. Stress intensity factors for a semi-elliptical, circumferential, ID-connected flaw under maximum operating loads and pressures and applicable structural factors are shown in Figure 3-8. The applied stress intensity factors for ID flaws do not exceed the assumed K1c of 25 ksi.Vin for flaws as deep as 80% of the wall thickness, the maximum flaw depth for which the stress intensity factor solution is valid [3].

Report No. 130 1620.402.R2 3-10 SJ Structural Integrity Associates, Inc.

Table 3-1. Summary of Forces and Moments on Welds [15,16]

Transient Fx (lbs)

Fy(lbs)

Fz (lbs)

Mx (in-lbs)

My(in-lbs)

Mz(in-lbs)

(1) Pump I-I, Design Calculation 9-323, Node 5 (Pump Suction Drain)

DL I3 6

-4 24

-5 42 THRMNI

-7

-244

-I7 I75

-36 388 THRMN2 5

203 29

-I63 59

-329 THRMAI

-8

-344

-75 52 I

-I69 556 DE 2

8 6

I7 8

I5 DDE 4

I6 II 34 I7 30 HOSGRI 9

2I I2 47 I8 57 (2) Pump I-I, Design Calculation 9-323, Node 50 (Pump Body Drain)

DL

-28 0

7 I

-8I I

THRMNI 0

-6

-4

-IOO

-25 I9 THRMN2 0

-I6

-3

-I78

-9 85 THRMAI 5

I 65

-47 239

-I7 DE 5

I5 I6 66 68 72 DDE IO 30 32 I32 I36 I45 HOSGRI 2I 38 35 276 I 58 I86 (3) Pump 2-I, Design Calculation 9-537, Node 85 (Pump Underside)

DW 28

-7 0

0

-24

-I80 NI

-7

-II 0 7

60

-I2

-I20 N2

-6 90

-6

-48 I2 96 AI

-4I

-636 42 360

-I08

-720 DE 3

7 IO 60 24 24 DDE 7

I4 2I I32 36 48 HOSGRI I7 I3 I8 I08 48 I20 (4) Pump 2-I, Design Calculation 9-537, Node 5 (Suction)

DW 25 0

0 0

I32 I80 NI

-I 0

-3 0

0 0

N2 I

0 2

0 0

0 AI

-4 0

-I5 I2 I2 36 DE 5

7 I6 60 72 48 DDE IO I3 32 I08 I44 I08 HOSGRI 29 II 23 96 204 240 (5) Pump 2-I, Design Calculation 9-537, Node 95(a) (Vent)

DW I4 0

0 0

0 0

NI 0

0 0

0 0

0 N2 0

0 0

0 0

0 AI 0

0 0

0 0

0 DE 0

0 7

0 60 0

DDE 0

0 I5 0

I32 0

HOSGRI 8

0 II 0

84 0

Report No. I30I620.402.R2 3-II e

Structural Integrity Associates, Inc.

Table 3-1, cont'd.

Transient Fx (lbs)

Fv (lbs)

Fz (lbs)

A{, (in-lbs)

My(in-lbs)

Mz(in-lbs)

(6) Pump 2-2, Design Calculation 9-536, Node 5 (3/4" drain from pp discharge)

DW 27 6

4 36 120

-156 N1

-3

-1 1

12 0

24 N2 1

1 1

12

-12 0

ACC1

-19

-7 8

72

-48 108 ACC2

-8

-3 2

12 0

36 DE 16 16 12 108 72 180 DDE 32 32 24 216 144 348 HOSGRI 46 34 23 360 216 528 Notes:

(a) Load calculation sheet identifies location as node 85; per system sketch node is identified as 95.

Table 3-2. Equivalent Axial Loads (lbs)

Location (1)

(2)

(3)

(4)

(5)

(6)

Pump 1-1 1-1 2-1 2-1 2-1 2-2 Node 5 Node Node 85 Node 5 Node NodeS (Pump 50 (Pump (Suction) 95 (3/4" Suction (Pump Underside)

(Vent) drain from Drain)

Body pp Drain) discharge)

DL 262 441 952 1159 14 1045

~ DL +Normal thermal 2656 (I) 1132 1684 1158 14 964

~

.~

DL+DE+

2750 (2) til 1018 1644 1617 321 1329

~ Normal thermal ro

~ DL+ HOSGRI +

5275 (3)

Abnormal thermal 4880 2280 3018 451 3893 Notes:

(1) Normal operating load used for fatigue crack growth calculation. Represents 7,000 thermal cycles.

(2) Normal operating load used for fatigue crack growth calculation. Represents 400 DE cycles.

(3) Maximum operating load used to calculate allowable crack size. Also represents 20 Hosgri cycles in fatigue crack growth calculation.

Report No. 1301620.402.R2 3-12 e

Structural Integrity Associates, Inc.

Table 3-3. Stress Intensity Factors at the Deepest Point of Semi-Elliptical Circumferential Flaw on Pipe OD for Crack Aspect Ratio cia= 4 for Various Load Cases Stress Intensity Factor K (ksi...Jin)

Crack depth K due to K due to K due to aft Residual Unit Axial Pressure Stress Load 0.01 0.194 0.515 0.783 0.05

-4.149 0.947 1.421 0.1

-9.335 1.145 1.690 0.15

-12.365 1.271 1.843 0.2

-14.013 1.365 1.944 0.25

-15.269 1.489 2.081 0.3

-15.908 1.606 2.201 0.35

-16.028 1.717 2.305 0.4

-15.710 1.822 2.395 0.45

-15.436

• 1.967 2.535 0.5

-14.959 2.109 2.663 0.55

-14.386 2.249 2.780 0.6

-13.736 2.386 2.884 0.65

-13.217 2.560 3.024 0.7

-12.443 2.730 3.147 0.75

-11.312 2.896 3.251 0.8

-9.733 3.055 3.334 Report No. 1301620.402.R2 3-13 SJ Structural Integrity Associates, Inc.

Table 3-4. Stress Intensity Factors at the Deepest Point of Semi-Elliptical Circumferential Flaw on Pipe OD for Crack Aspect Ratio cia= 1 for Various Load Cases Stress Intensity Factor K (ksi.Yin)

Crack depth K due to K due to Kdue to aft Residual Unit Axial Pressure Stress Load 0.01 0.082 0.331 0.504 0.05

-3.169 0.582 0.871 0.1

-6.574 0.683 1.001 0.15

-8.239 0.739 1.060 0.2

-8.887 0.776 1.087 0.25

-8.961 0.805 1.099 0.3

-8.587 0.827 1.098 0.35

-7.875 0.845 1.089 0.4

-6.912 0.858 1.070 0.45

-5.728 0.862 1.035 0.5

-4.453 0.863 0.991

• 0.55

-3.192 0.860 0.941 0.6

-1.961 0.854 0.882 0.65

-0.761 0.847 0.821 0.7 0.504 0.837 0.752 0.75 1.909 0.823 0.676 0.8 3.513 0.805 0.592 Report No. 1301620.402.R2 3-14

~Structural Integrity Associates, Inc.

2c (b)

Figure 3-1. Crack geometry for semi-elliptical, circumferential, surface flaw.

(a) OD-connected flaw (b) ID-connected flaw.

30 25 20 c

$ *;;; 15

.lll:: -

~

.~ 10 ell c cv

+J

• =

5 ell ell cv...

+J V) 0

-5

-10 0

I

___ L_

---r-----+---~

i

+---~---

I c/a = 4 0.1 0.2 0.3 0.4 aft j

l l

0.5 0.6 0.7 OD Flaw 0.8 Figure 3-2. Maximum stress intensity factor (Kmax) as a function of OD flaw depth for calculation of critical crack size. Maximum operating load and pressure (Feq = 5,275lbs),

no structural factor included. Results are shown both including and excluding residual stress.

Report No. 130 1620.402.R2 3-15 S} Structural Integrity Associates, Inc.

30 -r--*****-**********-** **--***--*-*******- ----*-----*-*- * ---------* -------*-***----

• • ----*-*-*----* -----*-**-**- -----------------l I

J I

~~'

1*

25 I I

II c/a = 4,, ""

I

,...... l i

I :: -!1---+-----r-; -::-- ~

~ '" ~ /

I

~

1 I

I

..., ' -.... t. excl. residual stress I

I

• ~ 10 ---~ -~-*

f-- ~*

l*-------1

~

...... i...................!....................

1.. '.~~*~*~**** ** ***

,I I

• ~

5 I

i

~ _:

---

~-~~~ -~-=-~--=------,===!

SF= 1.4 for axial and pressure stress 00 II

-10 0

I 1

1 1

Flaw,

0.1 0.2 0.3 0.4 aft 0.5 0.6 0.7 0.8 Figure 3-3. Maximum stress intensity factor (Kmax) as a function of OD flaw depth for calculation of allowable flaw depth (Structural Factor= 1.4). Maximum operating load and pressure (Feq = 5,275 lbs). For conservatism, excludes residual stress.

0 z

l 12()

f100 0 0 "'eo u :

2 Ci()

0

~

r--

• 1-- --

Kouhillnt Trnta t:nt

• 1100 f

• OU O.tltc*

I

/r TtNttlnt Tr****tM-4 thuttl'"- ~~~ C4ol I

I I

I VI

• OI.di. lint " TJ.. 10 s ut nu-roc

-.......-~

_JJ _/

L.

~

I I

.,..,.......... -~

~"':,

~

'" ~'

__../.'

V I

I

.1'

/

.... ~

GOO 700 80(1 900 1000 11()0 1100 1100 1400 TOIPfRIHC TEVPEitATUR

• *r Figure 3-4. Charpy V-notch impact energies of Type 410 stainless steels quenched from 1850°F and tempered 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> at indicated temperatures [18].

Report No. 1301620.402.R2 3-16 SJ Structural Integrity Associates, Inc.

J fr-!b 100 120~---------------------------,

120 80

]

.0 0

1/1 J:j 00

<It

~

CJ c w

40 20 0

('

c 300 500 700 Tempering Temperature Figure 3-5. Izod impact energies of Type 410 stainless steel quenched from 1800°F and tempered 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> at indicated temperatures [19].

TEMPERATURE (0 f")

sa 189 150 200 158 T~~e 410 Stainless Stee l 125 l25 UID IB9

~

~

75 to 0..

75

.:E:

.:::c.

0

(,)

50

~

~

sa c Piece 2 25 0 Pjece 3 25 0

c e

2B

~ 8 69.

80 100 TEMPERATURE (°C)

Figure 3-6. Fracture toughness of coupons from Type 410 stainless steel valve studs. The failed studs were presumed to be embrittled by tempering in the range 1000-1050°F [23].

Elastic-plastic fracture toughness parameter KJc reported in lieu of K1c.

Report No. 130 1620.402.R2 3-17 e

Structural Integrity Associates, Inc.

Testing temperature, oF

-100 0

100 200 300 180

~

160 160 140 140

~

• en 120 120 ""'

en

.c (I) 100 100 '5 C1> c:

c:

.c.

80 80 C1>

C) o2TCT l=

)

o3TCT (I) 0 60 60

'0

• 4TCT Cii

~

o6TCT 40

.2 40 0

1 Ktc calculated from J1c

~ 20

• Equivalent K1c 20 u..

0 0

-100 -50 0

50 100 150 Testing temperature, oc Figure 3-7. Temperature dependence of fracture toughness (Kic) for Type 403 martensitic stainless steel, which has similar properties to Type 410 [22].

25 c/a = 4 0 +---~----~----~---r----+----+----;----~ **--*---*-*--- ---------*--t----------**-+*--*-------*-+----------t******-******------11 __ ~~- ------

10 SF= 1.4 for axial and pressure stress 1

Flaw

-10 _.___ _

_._l __

l.____ _

_._l _

0 0.1 0.2 0.3 0.4 a/t 0.5 0.6 0.7 0.8 Figure 3-8. Maximum stress intensity factor (Kmax) as a function of ID flaw depth for calculation of allowable flaw depth (includes residual stress, Structural Factor= 1.4 for stresses other than residual). Maximum operating load and pressure (Feq = 5,275 lbs).

Report No. 1301620.402.R2 3-18

~Structural Integrity Associates, Inc.

4.0 EVALUATION OF FATIGUE CRACK GROWTH 4.1 Objective The objective of the fatigue crack growth analysis is to evaluate subcritical crack growth of postulated flaws due to fatigue in the Type 410 stainless steel joints under anticipated cyclic loads during the evaluated time period. The calculated maximum flaw dimensions at the end of the evaluation period are determined to assess the acceptability of the welds for continued service.

The purpose of this analysis is to evaluate fatigue crack growth of two types of postulated flaws:

a flaw located on the pipe OD and a flaw located on the pipe ID.

4.2 OD Flaw 4.2.1 Evaluation Methodology The methodology for determining acceptability for continued service is based on linear elastic fracture mechanics. A semi-elliptical, circumferential flaw at the pipe OD is postulated. The stress concentration at the OD increases both the likelihood of a flaw and the magnitude of cyclic stresses at this location. Postulated flaws are evaluated by comparing the maximum flaw dimensions at the end of the evaluation period with the allowable flaw size. Flaw growth analysis is based on cyclic fatigue crack growth.

A fatigue flaw growth calculation is performed using applicable fatigue crack growth rates and operating conditions and transients that apply during the evaluation period. Cyclic stress intensity factor M, the maximum range of K fluctuation, is determined for each transient. The stress intensity factors for each type of load are computed as a function of postulated flaw depth and superimposed for the various operating states. Growth of postulated flaws is determined from the fatigue crack growth rate equation for the applicable 11K for each transient, and the final Report No. 1301620.402.R2 4-1 e

Structural Integrity Associates, Inc.

flaw size at the end of the evaluation period is obtained and compared with the allowable flaw SIZe.

4.2.2 Fatigue Crack Growth Properties 4.2.2.1 Fatigue Crack Growth Rate Law Fatigue flaw growth rate due to cyclic loading in piping can be characterized by a generalized equation ofthe form:

(5) which relates the rate of flaw growth da/dN in inches per load cycle to the range of applied stress intensity factor Min ksiv'in, where Co and n are parameters dependent on material and environmental conditions.

Information on the fatigue crack growth characteristics of martensitic stainless steels such as Type 410 was sought from the literature. Data was available for Type 403, a martensitic stainless steel of similar composition to Type 410, and this data from Reference [22] is summarized in Figure 4-1a. Fatigue crack growth behavior is affected by temperature, stress ratio, and environment. The fatigue crack growth properties will not be highly dependent on heat treatment, strength level or composition [21 ], so the results shown in Figure 4-1 a for Type 403 are believed to be representative of the Type 410 welded material under consideration.

A water environment is conservatively assumed for fatigue crack growth of OD flaws to account for the potential for humid air or condensed moisture on pipe external surfaces. Figure 4-1 a shows that the growth rate in water is about twice that in air. The fatigue crack growth data in air in Figure 4-1a corresponds to the following equation:

Report No. 1301620.402.R2 4-2

~Structural Integrity Associates, Inc.

da = 7.55x1 o-9 Mt.9~ (air, R = 0.5) dN The growth rate is in inches per cycle when f.,.K is in ksi.Yin. This relation is for stress ratio R = Kmin!Kmax= 0.5. This is the R-ratio used in generating the test data on which Figure 4-1a is based. The R-dependence of f.,.K will be taken to be the same as in the ASME Code XI, Appendix C, Section C-841 0, for austenitic stainless steel [2], which is expressed as the factor S(R), where Co = C

• S(R). The factor S(R) is defined as follows:

1 S(R) = 1 + 1.8R

-43.35+57.97R R<O 0<R5,0.79 0.79 < R < 1 (6)

(7)

Incorporating the R-ratio effect from Equation 7 in the fatigue rate equation in Equation 6, for an air environment and R-ratio of0.5, the factor S(R) = 1.9, and the factor C= 3.97 x 10-9* Doubling the growth rate Co in Equation 6 to compensate for water environment, the factor C = 7.95 x 10-9.

Therefore, the fatigue crack growth rate in a water environment as a function of R is expressed by the following equation:

4.2.2.2 Fatigue Threshold da = 7.95x10-9 S(R)Mt.95 dN (8)

In addition to fatigue crack growth rate, information is required on fatigue crack growth threshold stress intensity factor Mth* At cyclic stress intensities below M 1h, fatigue crack growth will not occur. ASME Section XI [2] does not define a threshold for Type 410 stainless steel, and information on /5,.K1h for steels was sought from the literature. Figure 4-1 b summarizes information from Reference [21] on fatigue crack growth threshold M 1h in steels. Figure 4-1 b Report No. 1301620.402.R2 4-3 SJ Structural Integrity Associates, Inc.

shows that the fatigue crack propagation threshold is largely independent of mechanical and metallurgical properties of these steels. Hence, the data of Figure 4-1 b, which indicates a threshold of approximately 5 ksiv'in at R-ratio of 0.5, will be assumed to be applicable to the Type 410 martensitic stainless steel under consideration.

4.2.3 Cyclic Loads Loads are as defined in Table 3-1. The procedures used to combine loads and obtain equivalent axial force are the same as described in Section 3.2.3, and load combinations for fatigue crack growth are summarized in Table 3-2.

The fatigue crack growth calculation considers three types of transients, in which load is assumed to cycle between maximum and minimum values of equivalent axial force Feq for the assumed number of cycles. For all transients, deadweight is taken as the minimum load, corresponding to the smallest possible minimum load condition under all transients considered.

The equivalent axial loads and cycles for the three transients are defined in Table 4-1.

For the thermal transient, 7,000 cycles are assumed, with the maximum equivalent axial load defined as the sum of normal thermal loads (the maximum ofTHERMN1 and THERMN2) and deadweight. Pressure is also included in the evaluation of cyclic stress intensity factors.

Additionally, 400 cycles are assumed for design earthquake (DE), with the maximum equivalent axial load defined as the sum of normal thermal, deadweight, and DE loads.

Finally, 20 cycles are assumed for the Hosgri earthquake, with the maximum equivalent axial load defined as the sum of abnormal thermal, deadweight, and Hosgri loads.

Rather than treat the six locations included in Table 3-2, attention is focused on the location with the highest cyclic load, which is Location 1, the pump suction drain on SI pump 1-1. However, the largest combination of [DL + Hosgri +Abnormal Thermal Load] is used for the Hosgri event (Location 3, the pump underside on SI pump 2-1).

Report No. 1301620.402.R2 4-4 e

Structural Integrity Associates, Inc.

4.2.4 Cyclic Stress Intensity Factor versus Crack Size The equivalent pipe loads and cycles summarized in Table 4-1 are used in the fatigue crack growth analyses. Stress intensity factors due to pressure, residual stress and a unit axial tension load of 1,000 lbs are included in Table 3-3 and Table 3-4, for crack aspect ratios cia= 4 and 1 respectively.

The total cyclic stress intensity factors are obtained by adding these individual K-contributors, accounting for the magnitude of the applied tensile loads associated with each transient, as provided in Table 4-1. Cyclic stress intensity factor 11K is defined as Kmax minus Kmin, which correspond to the maximum and minimum total stress intensity factor versus flaw depth.

Contributors to Kmax include design pressure and equivalent axial force Feq due to deadweight, thermal loads, and where applicable, DE or Hosgri earthquake loads. For all transients, Kmin corresponds to Feq due to deadweight only.

Recall from Figure 3-2 that stress intensity factors associated with an OD flaw are either compressive or very small when residual stresses are included. In the presence of residual stresses, the applied cyclic stress intensity factor is less than the fatigue threshold 11Kth of 5 ksi-v'in for almost all crack sizes considered. Consequently, postulated cracks would not be predicted to grow when residual stresses are included. For added conservatism, 11K in the absence of residual stresses is included in the figures and employed in the fatigue crack growth analysis.

Figure 4-2 presents cyclic stress intensity factors Mas a function of OD crack depth alt for crack aspect ratio cia of 4 and 1. Figure 4-2 is based on the most frequently occurring transient, the 7,000 cycles of normal operating load and pressure with a maximum equivalent axial force Feq of 2,656 lb. Cyclic stress intensity factors associated with the 400 DE cycles, with a maximum Feq of2,750 lb, are close in magnitude to those in Figure 4-2. Kmax associated with Hosgri event cyclic loading, with a maximum equivalent axial force of 5,275 lbs, correspond to Report No. 1301620.402.R2 4-5 e

Structural Integrity Associates, Inc.

those in Figure 3-2 (excluding residual stress). Note that the stress intensity factor solutions are valid for crack depths a/t up to 0.8 [3].

4.3 ID Flaw 4.3.1 Evaluation Methodology The evaluation methodology for evaluating fatigue crack growth of postulated ID flaws is the same as for OD flaws, described in Section 4.2.1, with the exception that the relevant flaw geometry is a semi-elliptical circumferential flaw originating on the ID of the pipe and growing toward the pipe OD.

4.3.2 Fatigue Crack Growth Properties 4.3.2.1 Fatigue Crack Growth Rate Law The same fatigue crack growth rate equation (Equation 8) is used for the evaluation of fatigue of postulated flaws located on the OD and on the ID, with the exception that values forM and S(R) will correspond to the cyclic stress intensity factors obtained for the ID flaw.

4.3.2.2 Fatigue Threshold The fatigue threshold f1Kth of 5 ksiv'in (described in Section 4.2.2.2) will be used in the evaluation of fatigue crack growth of ID flaws.

4.3.3 Cyclic Loads Cyclic loads and transients for the evaluation of fatigue crack growth of ID flaws are the same as those described in Section 4.2.3.

Report No. 130 1620.402.R2 4-6 e

Structural Integrity Associates, Inc.

4.3.4 Cyclic Stress Intensity Factor versus Crack Size Figure 4-2 compares cyclic stress intensity factors for ID-and OD-connected flaws of the same aspect ratio for the most frequently-occurring transient. The results demonstrate that 11K for an ID flaw is smaller than for an OD flaw up to the allowable flaw depth a/t of 0.716. Additionally, 11K for ID flaws is below the fatigue threshold 11K1h of 5 ksiv'in, indicating that postulated ID flaws would not be predicted to grow under cyclic loads. Consequently, the fatigue crack growth evaluations are limited to OD-located flaws.

4.4 Results Fatigue crack growth of postulated OD flaws in the Type 410 stainless steel pipe nipple is evaluated by predicting flaw growth under anticipated cyclic fatigue loading and comparing the final flaw depth at the end of the evaluation period to the allowable flaw depth. Based on fatigue crack growth properties of the Type 410 stainless steel described above, the cyclic stress intensity factor must exceed about 5 ksi.Vin in order for fatigue crack growth to occur (at R = 0).

In the present calculation, a threshold cyclic stress intensity factor or 11K1h equal to 5 ksiv'in is used.

The fracture mechanics code pc-CRACKTM [18] is used to compute crack growth produced by the three transients in Table 4-1. The maximum and minimum K for cia= 4 for each transient as a function of crack depth a is input in tabular form. The fatigue crack growth rate relationship in Equation 8 is used.

The extent of fatigue crack growth is computed for four initial OD flaw depths, with results summarized in Table 4-2:

(1) 10% through-wall flaw (2) Flaw whose surface length is the threshold detection limit for PT (1116")

(3) Flaw just exceeding the Mrh of 5 ksi.Vin

( 4) Initial flaw required to grow to the allowable flaw depth by the end of the evaluation period Report No. 130 1620.402.R2 4-7 SJ Structural Integrity Associates, Inc.

First, a flaw depth alt = 0.1 is assumed. For the nominal pipe thickness of 0.154 inch, the 10%

through-wall flaw depth a= 0.0154 inch. A flaw aspect ratio cia= 4 is used. For the 10%

through-wall flaw, M < Mth for all load cycles and is not predicted to grow, except for the 20 cycles assumed for Hosgri earthquake. The associated crack extension at the end of the evaluation period 11a = 8.3 x 1 o-6 inch, corresponding to the crack growth produced by the 20 cycles of Hosgri earthquake.

Second, a flaw is assumed whose surface length is that of the threshold detection limit for surface examination by PT, equal to 1/16" [24]. This size of flaw is intended to represent the largest flaw that could have been missed during inspection. For conservatism in this case, a flaw aspect ratio cia= 1 is used, for a flaw depth a of 0.03125 inch. Note that per Figure 3-2 and Figure 4-2, the stress intensity factors for cia= 1 are smaller than for cia= 4. For a flaw just at the threshold detection limit for PT, M < 11Kth for all load cycles, even those associated with the Hosgri earthquake, and the flaw is not predicted to grow.

Third, a flaw depth just exceeding the fatigue crack growth threshold Mth of 5 ksi.Yin for all transients is assumed. The 11Kth is exceeded for cracks having an aspect ratio cia= 4 at a crack depth a= 0.0260 inch. Flaws smaller than this would not be predicted to grow under anticipated cyclic loading. The final flaw depth at the end of the evaluation period a1= 0.0275 inch, and the amount of crack extension 11a = 0.0015 inch.

Fourth, the depth of initial flaw required to reach the allowable flaw size alt = 0.716 at the end of the evaluation period is determined, corresponding to a final crack depth a= 0.110 inch. Recall that unstable crack growth should not occur for crack depths less than this size, even if a seismic/abnormal event occurs, as shown in Figure 3-3. A flaw aspect ratio cia= 4 is used. The initial flaw depth a that will reach the allowable flaw depth during the evaluation period is 0.104 inch, and the crack extension during the evaluation period 11a = 0.006 inch. This corresponds to a surface half-length c = 4a = 0.416 inch. Hence, for a flaw to grow to the allowable flaw depth during the evaluation period, a crack having a depth of 0.104 inch and surface length 2c = 8a = 0.832 inch is required, corresponding to 25% of the outer circumference of the pipe.

Report No. 1301620.402.R2 4-8 SJ Structural Integrity Associates, Inc.

The results summarized in Table 4-2 show that for the four different postulated flaws, crack growth under anticipated cyclic loading during the evaluation period is minimal.

Report No. 1301620.402.R2 4-9 SJ Structural Integrity Associates, Inc.

1.

2.

3.

1 2

3 4

Table 4-1. Cyclic Loads for Fatigue Crack Growth Transients Minimum<a)

Maximum Contributors to Transient No.

Equivalent Equivalent Stress Intensity Factor Cycles Axial Force, Axial Force, Kmin Km£Lr:

Min Feq (lbs)

Max Feq (lbs)

Normal thermal cycles 7000 262 2656 Design earthquake (DE) 400 262 2750 Min Feq MaxFeq + P Hosgri earthquake 20 262 5275 Note:

(a) For all transients considered, min F eq corresponds to deadweight only.

Table 4-2. Results of OD Fatigue Crack Growth at Weld Toe (Stress Path 1)

Initial flaw size Final flaw size Flaw case Depth Half-Depth length ai (in) a/t Ci {in) ar(in) 10% through-wall flaw (a) 0.0154 0.100 0.0616 0.0154 Threshold size flaw 0.03125 0.203 0.03125 0.03125 for PT detection (b)

Flaw just exceeding 0.0260 0.169 0.104 0.0275 11Kth (a)

GJ = Gallow at end of evaluation period (a) 0.104 0.675 0.416 0.110 Notes:

(a) cia= 4, where a is flaw depth and cis surface half-length of flaw.

(b) cia= 1 (c) M < M 117 except during Hosgri load case.

(d) 11K < 11K117 including during Hosgri load case.

Half-length alt cr(in) 0.100 0.0616 0.203 0.03125 0.179 0.110 0.716 0.441 Aa (in) 8.3 X 10-6 (c) 0 (d) 0.0015 0.006 Report No. 1301620.402.R2 4-10 e

Structural Integrity Associates, Inc.

310 1~

IlK, ksi..fm.

20 30 40 6080 Air Type403 In H20 pH 7, 25 °C

- -. pH 1 0, 25 oc pH 7, 100 oc pH 10, 100 °C In 1 M NaCI solution

• • • • • • • pH 2 to 1 o. 1 00 oc 1o-s~--~--~~--~~~--~

10 20 30 40 60 80 100 llK.MPa~

(a) 10-5

0) g,

~

. 5

~

~

"t5 10~

~

.c

<l w

Cl 01: <

a:

a:

0....

u

~

I >...

Z' kl

~.

Ul Ul 0 -1 0 :r Cll w

a:

l:

I Cl ;::

it V

MILO STEEL 0

lOW-ALLOY STEEL A 18/9 AUSTENITIC STE EL 0

A517-F STEEL

.i. 9310 STEEL B A508 CLASS 2 STEEL A533 GRADE B CLASS I STEEL

. 1 V

24 Cr-1 Mo STEEL AK,n* fi.4p-O.Il!IR)

FOR R:.O. I 61< 1 ~ = 5.!1 hiJ'ii'Cii FOR R!: O.I lh~

• I.0 9 9t.CN/ml/l RATIO OF MINIMUM STRESS ANO MAXIMUM STR ESS, ;~~~

, A (b)

Figure 4-1. Fatigue crack growth rate and threshold data for steels. (a) Fatigue crack growth rates in Type 403 stainless steel in air, water, and a 1 M NaCl solution at 10 Hz and R-ratio of 0.5. Compact specimens (0.5 in thick) obtained from L-T orientation of plate austenitized at 1750°F, cooled in air, and tempered 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> at 1200°F (crys = 100 ksi) [22].

Type 403 has similar properties to Type 410.

(b) Fatigue crack thresholds for steels [21].

Report No. 1301620.402.R2 4-11 e

Structural Integrity Associates, Inc.

30 25 c

-? 20

'Vi

~ -

~ 15

<I

.~

Cll c 10 (I)

+'

.5 Cll Cll 5

(I)...

+'

Cll

.~

u 0

(j

-5

-10 I

I I

Excludes residual stress I

I I

I I

I ----*--------1----

I :

c/a = 4 ~

L --- '" -- -

~ - - -

k --- --,

c/a = 4 Q--

L-1----

ca-l

-~**** ~ *** ** **** j *.t...~.T. ***j~~:i~

~-.:........... -.-;.

f.---r---

I

-- -r t m c/a = 1 I

00,10 I

Flaws 0

0.1 0.2 0.3 0.4 aft 0.5 0.6 0.7 0.8 Figure 4-2. Cyclic stress intensity factor (AK) as a function of flaw depth for normal cyclic operating load and pressure, i.e. not including DE loads (Max F eq = 2,656 lbs). (Residual stresses do not influence cyclic stress intensity factors.) Results for both OD-connected and ID-connected flaws are presented. AK for ID flaws is lower than for OD flaws and below the threshold AK1, of 5 ksi.Vin.

Report No. 1301620.402.R2 4-12 SJ Structural Integrity Associates, Inc.

5.0 CONCLUSION

S AND RECOMMENDATIONS A stress and fracture mechanics evaluation has been performed for Type 410 stainless steel nipples welded to Type 316 valves and Type 304 fittings on the DCPP Safety Injection Pump vent and drain lines. This evaluation consisted of:

Stress analysis Evaluation of allowable flaw size under maximum loading Evaluation of crack propagation of postulated flaws under cyclic fatigue loading The depths of OD and ID flaws located along stress path 1 required to cause crack instability under maximum operating loads and pressure, including seismic/abnormal loads and applicable structural factors, are evaluated. The allowable flaw depth for an OD flaw is determined to be 0.110 inch, approximately 71.6% ofthe wall thickness of0.154 inch. The allowable flaw depth for an ID flaw exceeds 80% of the wall thickness.

Postulated ID flaws are not predicted to grow under cyclic loading, as cyclic stress intensity factors 11K are below the fatigue threshold Mth* Consequently, fatigue crack growth evaluations are limited to OD flaws.

For an OD crack to grow by fatigue under cyclic operating loads and pressure to the allowable flaw size in the evaluated number of cycles, an initial crack of at least 0.104 inch depth (corresponding to a surface length of0.832 inch for crack aspect ratio cia= 4) is required.

Fatigue crack growth will not occur for a postulated OD flaw whose surface length is equal to the PT threshold detection limit (2c = 1116 inch), since 11K < M 1h for all load cycles, even those associated with the Hosgri earthquake.

For a 10% through-wall OD flaw (a= 0.0154 inch), 11K < 11K1h for all load cycles and is not predicted to grow, except for the 20 cycles assumed for the Hosgri event. The associated crack extension under anticipated cyclic loading at the end of the evaluation period 11a = 8.3 x 1 o-6 inch.

Report No. 1301620.402.R2 5-1

~Structural Integrity Associates, Inc.

For an OD crack 0.026 inch deep, just exceeding the fatigue crack growth threshold M 1h of 5 ksi.Yin, the amount of crack extension under anticipated cyclic loading during the evaluated period ~a= 0.0015 inch.

The evaluations of the postulated OD and ID flaws show that crack growth under anticipated cyclic loading during the evaluation period is minimal.

These findings are applicable to the subject Type 410 stainless steel safety injection pump nipples and are based on material properties developed from industry data and information on dimensions, operating loads, and NDE inspection procedures provided by DCPP. SIA recommends that DCPP review available plant-specific data for loads, weld dimensions, and material properties to verify their applicability.

Report No. 130 1620.402.R2 5-2 e

Structural Integrity Associates, Inc.

6.0 REFERENCES

1. SIA Report No. 1301620.401, Revision 0, "Safety Injection Pump-Probable Condition of Type 410 Stainless Steel Weldments," December 19, 2013.

2. ASME Boiler and Pressure Vessel Code,Section XI, 2001 Edition with Addenda through 2003.

3. API Standard 579-1/ASME FFS-1, Fitness-For-Service, Second Edition, June 2007.

4. Diablo Canyon Power Plant email transmittals, dated 1212412013, with attachments that confirm key pipe dimensions, and consist of DCPP SI Pump photographs and as-built walkdown information.,

Subject:

"RE: DIT for pipe forces at SI Pumps-Part 11213 of 3",

SIA File No. 1301620.204.

5. ANSYS Mechanical APDL and PrepPost, Release 12.1 x64, ANSYS, Inc., November 2009.

6. SIA Calculation 0800777.304, Rev. 0, Residual Stress Methodology Development and Benchmarking of a Small Diameter Pipe Weld Overlay, Using MISO (Multi-Linear Isotropic Hardening) Properties.

7. Crane Technical PaperNo. 410, Flow ofFluids through Valves, fittings, and pipe, 1976.

8. ASME B16.11-2005, Forged Fittings, Socket-Welding and Threaded.

9. Ladish General Catalog No. 55, Forged and Seamless Welding Pipe Fittings, 1971.

10. M.W. Kellogg Company for P. G. & E. Diablo Canyon Project, Specification No. P81P1-K1-F5-SMA W-6G, "Weld Procedure Code No. 149, Revision date 12103173," SIA File No.

1301620.202.

11. ASME Boiler and Pressure Vessel Code,Section II, Part D - Properties, 2001 Edition with Addenda through 2003.

12. ASME Boiler and Pressure Vessel Code,Section II, Part A-Ferrous Material Specifications, 2001 Edition with Addenda through 2003.

13. "AISI 410," High-Temperature Property Data: Ferrous Alloys, M.F. Rothman, ed., ASM International, 1985, p. 9.75. SIA File No. 1301620.205.

14. Technical Data Sheet, "ATI 410TM I ATI 420TM I ATI 425 ModTM I ATI 440ATM I ATI Report No. 1301620.402.R2 6-1 e

Structural Integrity Associates, Inc.

440CTM", Allegheny Technologies Inc., March 2013. SIA File No. 1301620.205.

15. DCPP SI Pump Vent and Drain Piping Stress Analysis, Design Input Transmittal No.

5 0600119-01-00, email correspondence "FW: D IT for pipe forces at SI Pumps" from Chris Beard (DCPP) to George Licina (SIA), December 19,2013, 19:23 PST., SIA File No.

1301620.201.

16. Email correspondence "FW: DIT for pipe forces at SI Pumps" from Chris Beard (DCPP) to Heather Jackson (SI), December 27, 2013, 08:16 PST. SIA File No. 1301620.204.

17. NRC Inspection Manual, Part 9900: Technical Guidance, April 16, 2008.

18. H. Tanczyn, "Properties of 12 per cent chromium alloys modified with small columbium additions," Advances in the Technology of Stainless Steels and Related Alloys, ASTM STP 369, ASTM International, 1965.

19. "Design Guidelines for the Selection and Use of Stainless Steel," Publication No. 9014, Nickel Development Institute.

20. J.M. Barsom and S.T. Rolfe, Fatigue and Fracture Control in Structures, 2nd edition, Prentice-Hall, Inc., 1987.

21. "Fatigue and Fracture Properties of Stainless Steels, ASM Handbook Vol. 19: Fatigue and Fracture, ASM International, 1996.

22. Fracture Toughness Characterization of Type 410 Stainless Steel, EPRI, Palo Alto, CA:

1987. NP-5511.

23. pc-CRACK for Windows, Version 4.0.1.0, Structural Integrity Associates, 2011.

24. Email correspondence "RE: DIT for pipe forces at SI Pumps" from David Gonzalez (DCPP) to Heather Jackson (SIA), December 30, 2013, 14:53 PST. SIA File No. 1301620.206.

Report No. 1301620.402.R2 6-2 S} Structural Integrity Associates, Inc.

OUTGOING CORRESPONDENCE SCREEN (Remove prior to NRC submittal)

Document:

PG&E Letter DCL-14-060

Subject:

ASME Section XI lnservice Inspection Program Request for Alternative REP-S I: Proposed Alternative to Requirements for Repair/Replacement Activities for certain Safety Injection Pump welded attachments File Location: S:\\RS\\CLERICAL\\DCLs-Finai\\DCL-14-060, Relief Request (SI Pump weld issue)\\DCL-14-060.docx FSAR Update Review Utilizing the guidance in XI3.1D2, does the FSAR Update need to be revised?

Yes D No 1Z1 If "Yes", submit an FSAR Update Change Request in accordance with Xl3.1 02 (or if this is an LAR, process in accordance with WG-9)

There are no commitments in this letter Statement of Commitment: