The failure rate development consists of determining an industry generic pipe failure rates for RR and FW piping, respectively. Next a Bayesian update is performed to generate failure rates that best represent the design and operating conditions of Plant B. Rather than taking an apriori failure from some published source, the approach in this study is to derive apriori failure information from the PIPExp database. The information that is summarized in Section D.3 provides some insights into the time-dependency of failure rates. These insights are explored further below.

D.4.2.1 RR Pipe Failure Data — Programs to mitigate the effects of certain degradation mechanisms strongly influence the achieved piping reliability. As an example, all BWR plants commissioned prior to the early to mid-1980s have experienced IGSCC. Industry initiatives to mitigate or eliminate the influence by IGSCC were implemented by the mid-1980s, and thereafter the rate of IGSCC has dropped sharply. The trend in the IGSCC rate is established by normalizing the data displayed in Figure D.2 (IGSCC by Years of Operation). Calculating the rate of IGSCC per weld-year for a given system and pipe size performs the normalization. Before performing this normalization the database is subjected to additional processing to exclude from further consideration any IGSCC data not directly applicable to the RR System that is representative of the Base Case. Similarly the part of the database including plant population and weld population data must be processed in such a way that an appropriate exposure term is developed commensurate with the failure data. In developing the RR-specific exposure term the following exclusion criteria were applied:

Plant Population Exclusion Criteria Applicable to RR Piping

• BWR plants without external recirculation loops;

• BWR plants in which the RR piping is fabricated from IGSCC resistant material e. g., Nuclear Grade stainless steel.

Table D.3 includes selected weld counts used to derive an exposure term according to Equation (D.2). Organized by pipe size and years of operation, Table D.4 is a summary of weld failures in RR piping. Noteworthy is the observation that there have been no reported through-wall defects in any plant beyond T = 15 years of operation. Using the information in Tables D.3 and D.4, Figure D.16 shows the calculated rate of RR pipe failure per weld-year.

The failure rates in Figure D.16 assume that all RR welds of a certain size to be equally susceptible to IGSCC. As was shown in SKI Report 98:30 [D.15], a correlation exists between weld failure rate and weld configuration. This correlation is assumed to be attributed to the piping layout, complexity of welding operation, and the associated weld residual stresses. The chart in Figure D.17 shows the weld configuration versus fraction of weld failures.

Plant ID (NSSS Type)	Weld Count by Pipe Size (NPS)
4	6	8	12	14	16	22	24	28
1 (AA/3)[6] [7]	30	42	23	—	—	—	—	54	—
2 (BWR/4)	8	—	—	58	—	—	12	—	33
3 (BWR/4)	—	—	—	34	—	2	—	4	24
4 (BWR/5)	36	—	—	51	10	16	—	50	—
5 (BWR/5)	39	—	—	63	—	16	—	45	—
6 (BWR/5)	—	—	—	130	—	—	24	—	97
7 (BWR/5)	—	—	—	138	—	—	16	—	97
8 (BWR/4)	—	—	—	24	—	—	4	—	28
9 (BWR/4)	—	—	—	25	—	—	4	—	38
10 (BWR/4)	12	—	—	59	—	—	10	—	36
11 (BWR/4)	12	—	—	62	—	—	12	—	36
Plant B (BWR/4)	—	—	—	50	—	—	16	—	56
Mean:	23	—	—	63	—	—	12	—	49

Table D.4 Number of Through-Wall Flaws in RR Piping Attributed to IGSCC7

Pipe Diameter (0) [NPS]	Years of Operation
Total	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
3 < 0 < 6	15	0	1	1	2	4	1	0	2	2	0	2	0	0	0	0
6 < 0 < 12	9	0	0	0	0	1	0	1	0	1	1	2	1	1	0	1
12 < 0 < 22	4	0	0	0	0	0	0	1	0	0	0	2	1	0	0	0
0 > 22	16	0	0	0	0	0	0	0	0	5	0	6	4	0	1	0
44	0	1	1	2	5	1	2	2	8	1	12	6	1	1	1

Figure D.16 Rate of IGSCC-Induced RR Pipe Failure (‘Prior State-of-Knowledge’)

Pipe-to-pump

Pipe-to-cross

Elbow-to-pump

Pipe-to-valve О

Elbow-to-valve

=5

Pipe-to-end-cap c

Pipe-to-tee ~o

Pipe-to-reducer

§

Pipe-to-sweepolet Pipe-to-safe-end Pipe-to-pipe Nozzle-to-safe-end Elbow-to-pipe

Figure D.17 RR Weld Failures as a Function Weld Configuration

In addition to weld configuration, the likelihood of failure is also a function of pipe size. For a weld of type “i” and size “j” the failure rate is expressed as follows:

Aj = Fj/(Wj x T) (D.3)

and with

S = Fj / Fj (D.4)

Aij = Wj/Wjj (D.5)

the failure rate of weld of type “i” and size “j” is expressed as

Aj = (F x Sij)/(Wj x T) (D.6)

Aj = Sij x AijX A (D.7)

Failure rate of an IGSCC-susceptible weld of type “i”, size “j”

Failure rate of an IGSCC susceptible weld of size ‘j’

Number of size “j” weld failures

Number of type “i” and size “j” weld failures

Size “j” weld count

Type “i” and size “j” weld count

The service experience shows the failure susceptibility to be correlated with the location of a weld relative to pipe fittings and other in-line components (flanges, pump casings, valve bodies). For a given pipe size and system, the susceptibility is expressed as the fraction of welds of type “ij” that failed due to a certain degradation mechanism). This fraction is established from the PIPExp database; Table D.5.

In the above expressions the attribute (A) is defined as the ratio of the total number of welds of size “j” to the number of welds of type “i”. In expression (4.7), Ay is a correction factor and accounts for the fact that piping system design & layout constraints impose limits on the number of welds of a certain type. For example, in a given system there tends to be more elbow-to-pipe welds than, say, pipe-to-tee welds.

Combining the global (or averaged) failure rates in Figure D.16 with the information summarized in Figure D.17 and Table D.5 provides the apriori failure rates that are input to Equation (D.1). The results are summarized in Section D.4.3.

Table D.5 IGSCC Susceptibility by Weld configuration — Selected Parameter Values RR System [NPS] Weld Configuration Configuration Dependent Parameters Susceptibility (Sij) Attribute (AH) 12 Elbow-to-pipe 6.03E-01 2.8 Nozzle-to-safe-end 1.35E-01 5.0 Pipe-to-reducer 2.38E-02 25.0 22 Pipe-to-end-cap 2.71E-01 4.0 Pipe-to-sweepolet 8.33E-02 2.0 Pipe-to-cross 6.25E-02 4.0 28 Elbow-to-pipe 4.62E-02 5.6 Pipe-to-pipe 5.77E-02 3.1 Cross-to-reducer 9.60E-03 28.8

D.4.2.2 FW Pipe Failure Data — The estimation of failure rates for FW piping uses a non-informative prior distribution together with the weld population data in Table D.6. This approach is selected based on the available, limited service experience with ASME XI Class 1 FW piping; Table D.7. In developing the data summary in Table D.7 the following FW exclusion criteria were used to develop a point estimate of the failure rate:

Failure Data Exclusion Criteria Applicable to FW Piping

• Piping external to the drywell containment structure;

• Non-US data.

Plant ID (NSSS Type)	Weld Count by Pipe Size (NPS)
8	10	12	14	16	18	20	22	24
1 (AA/3)	1	61	—	14	—	—	—	—	—
2 (BWR/4)	—	—	28	—	—	25	—	—	—
3 (BWR/4)	—	—	28	—	—	26	—	—	—
4 (BWR/5)	—	—	40	12	—	6	—	—	38
5 (BWR/5)	—	—	41	6		6	—	—	42
6 (BWR/5)	13	—	66	—	8	—	7	—	22
7 (BWR/5)	9	—	68	—	8	—	7	6	19
9 (BWR/4)	—	—	50	—	—	—	6	—	45
10 (BWR/4)	—	—	32	1	—	30	—	—	—
11 (BWR/4)	—	—	30	1	—	29	—	—	—
Plant B (BWR/4)	—	—	63	5	—	—	53	—	—
Mean:	—	—	42	—	—	—	41[8]	—	—

Table D.7 Summary of FW Pipe Failure Data