NOC-AE-14003101, Enclosure 1 to Enclosure 6 Concerning Second Set of Responses to April 2014, Requests for Additional Information Regarding STP Risk-Informed GSI-191 Application

Enclosure 1 to Enclosure 6 Concerning Second Set of Responses to April 2014, Requests for Additional Information Regarding STP Risk-Informed GSI-191 Application

ML14178A485
Person / Time
Site:	South Texas
Issue date:	06/25/2014
From:	South Texas
To:	Office of Nuclear Reactor Regulation
References
GSI-191, NOC-AE-14003101, TAC MF2400, TAC MF2401
Download: ML14178A485 (133)

Text

NOC-AE-1 4003101 Attachment 1 Enclosure 1 Enclosure 1 to Attachment 1 Supporting resolution of APLAB, CASA Grande-LOCA Frequencies: RAI 2 STP-RIGSI191-RAI-APLA-III-2, Rev. 1, "RAI APLA-III-2: Modeling LOCA Frequency and Break Size under DEGB only Breaks", University of Texas

NOC-AE-14003101 Attachment 1 Enclosure 1 10 South Texas Project Risk-Informed GSI-191 Evaluation RAI APLA-III-2: Modeling LOCA Frequency and Break Size under DEGB-only Breaks Document: STP-RIGSI191-RAI-APLA-III-2 Revision: 1 Date: May 13, 2014 Prepared by:

John Hasenbein, The University of Texas at Austin David Morton, The University of Texas at Austin Jeremy Tejada, The University of Texas at Austin Reviewed by:

Zahra Mohaghegh, University of Illinois at Urbana Champaign Seyed A. Reihani, University of Illinois at Urbana Champaign Approved by:

Ernie J. Kee, South Texas Project

NOC-AE-14003101 Attachment 1 Enclosure 1 RAI APLA-III-2: Modeling LOCA Frequency and Break Size under DEGB-only Breaks John Hasenbein, David Morton, and Jeremy Tejada The University of Texas at Austin 1 RAI APLA-I1-2 The statement of RAI APLA-III-2 is as follows:

RG 1.174, Section 2.3.4, "Plant Representation," states that PRA results should be derived from a model that realistically represents the risk associated with the plant.

NUREG-1829 states that, in general, a complete rupture of a pipe is more likely than a partial rupture. It appears, however, that STP's methodology leads to the opposite result (i.e., a rupture of a given size is more likely to be caused by a partial rupture of a large pipe than a complete rupture of a smaller pipe). Please illustrate the results of your method by comparing the frequency of partial versus complete breaks for a set of representative pipe sizes. Please describe whether the methodology described in the STP pilot is consistent with the assumption of NUREG-1829 or provide justification for an alternate approach.

2 Introduction We modify the analysis in the hybrid LOCA document of Pan et al. (2013) to handle the case in which welds only experience double-ended guillotine breaks (DEGBs). We refer to the resulting model as the DEGB-only model. We compare these results with the results of Pan et al. (2013),

which we refer to as the continuum break-size model. We limit reintroduction of the notation developed in Pan et al. (2013), except as it is modified for the DEGCB-only situation. As in our earlier analysis, we maintain consistency with NUREG-1829's (Tregoning et al., 2008) initiating frequencies. First, we use the bottom-up conditional weights derived from Fleming et al. (2011),

which is the approach we recommend. (Using the bottom-up weights, in conjunction with the con-tinuum model and NUREG-1829's frequencies, constitutes the hybrid method that is implemented in CASA Grande.) Then, we study the results of employing a top-down approach, which allows us to highlight the effect on DEGB probabilities of using the bottomn-up methodology.

1

NOC-AE-14003101 Attachment 1 Enclosure 1 Let j - J index STP break sizes. As before we compute the probability of a LOCA being in category j (catj) using the formula P [catj Frequency[LOCA > catj] - Frequency[LOCA > catj+1 ]

Frequency[LOCA > cat] ,(1) where Frequency[LOCA > catj] is consistent with NUREG-1829. By consistentwith NUREG-1829 we mean the following: We model Frequency[LOCA > catjJ as a random variable using a bounded Johnson distribution, and we do so for the six size categories elicited in NUREG-1829 (see Table 1).

If we take the median of the Johnson distribution for a category j, which coincides with a category in NUREG-1829, then Frequency[LOCA > catj] is simply the median value from the elicitation (Tregoning et al., 2008, Table 7.19, page 7-55) for the current-day fleet. If j does not coincide with a NUTREG-1829 category, and we still focus on the median frequency, then we approximate Frequency[LOCA > catj] by interpolating linearly between adjacent NUREG-1829 categories.

When we consider a percentile of the Johnson distribution other than the median (or the 5th percentile or 95th percentile) then we interpolate linearly between the corresponding percentile in Johnson distributions fit to adjacent NUREG-1829 categories. The distinction between the analysis we conduct here and in Pan et al. (2013) concerns how we allocate these frequencies to specific weld cases.

Table 1: LOCA categories from NUREG-1829.

Effective Break Size (Inches) Notation 75 cat, 1-5 cat2 3 cat3 7 cat 4 14 cat5 31 cat6 We compute the probability that weld case i will experience a break of type j, using P[catj at weldi] wjP[catj], (2) where zu = P(weldiccatj) is the conditional probability of the break occurring at weld i given a category j break. We use set Ij to denote the set of weld cases that can experience a break of type 2

NOC-AE-14003101 Attachment 1 Enclosure 1 j, and we use' set I to index all 45 STP weld cases as described in Fleming et al. (2011) (see also Table 12 in the Appendix).

Computation of zv is as follows. The bottom-up approach of Fleming et al. (2011) generates the frequency of category j breaks for weld case i, which we denote Freqb,,[LOCA > catj at weldi].

Also, there are specific numbers of welds for each weld case i, and we denote the number of welds for weld case i by ni. Given these frequencies and the number of welds for each weld case, the wto values are computed as

_ (Freqb,.[LOCA > catj at 'weldj] - Freqb.[LOCA > catj+l at weldi]) x ni w Zicij (FreqbJ[OCA > catj at weldi] - Freqba[LOCA > catj-l at weldj]) x ?.7i (

Given P[catj] from equation (1) and wjufrom equation (3), we form P[catj at weldi] via equation (2).

As before, because the sum of all w} across i E Ij is equal to one, this approach matches the NUREG-1829 specified values for P[catj].

Mathematically, this approach is identical to that described in Pan et al. (2013), but the key distinction is in the definition of set Ij. In the continuum model, a weld can experience breaks of size up to its diameter and then can experience a larger (by a factor of v/2, as we discuss further below) DEGB break. However, by assuming that welds only experience DEGB breaks, the set Iy is redefined to be the set of welds that have size that corresponds exactly to the category j corresponding to the size of the DEGB.

3 Comparing Continuum Break-Size Model and DEGB-Only Break Model Table 2 provides the joint probability mass function (2) corresponding to the median for the con-tinuum break-size model with the six NUREG-1829 categories and with weld cases aggregated by pipe size. (All results we report here are for the median.) The first numerical entry in the table indicates, given that we have a break, there is a probability of 0.586 that it is a category 1 break on a 1-inch pipe. The table's right-most column indicates the marginal probability of having a break in each break-size interval: [1, 2), [2, 2.5), ... , [29, 31), [31, DEGB], where all values are in inches and DEGB is the largest possible DEGB break size, all using STP pipe sizes. The table's bottom row indicates the probability a break is in NUREG category 1, category 2, etc. Table 3 reports analogous results for the DEGB-only model. Note that if a pipe experiences a DEGB then because both ends of the pipe are exposed it is equivalent to a "one-sided break" on a pipe with a radius larger by a factor of v/2. This is why, for example, a DEGB break of a 10-inch pipe is classified as 3

NOC-AE-14003101 Attachment 1 Enclosure 1 a category 5 break in Tables 2 and 3. Tables 13 and 14 in the Appendix repeat the information in Tables 2 and 3 except at the resolution of STP's 45 weld cases.

As the tables indicate, the DEGB-only model increases the likelihood of breaks in small pipes.

In the continuum model, small breaks can occur in a 31-inch pipe. However, in the DEGB-only model small breaks can only occur in correspondingly small pipes. As a result the probability of a break in a 31-inch pipe decreases by a factor of about 16,500 (=1.75E-03/1.06E-07). As the bottom rows in Tables 2 and 3 indicate, we preserve NUREG-1829 frequencies, and hence the probabilities of a category 1, 2, ... , 6 break are identical in both models. Also, note that the category 6 columns are identical in Tables 2 and 3. This is because under both models a category 6 break must be a DEGB because STP's largest pipe size is 31 inches, which matches the low end of NUREG-1829's category 6 bin.

Table 4 repeats the right-most columns of Tables 2 and 3, and the table also includes, for each pipe size, the probability that weld cases of that pipe size experience a DEGB under both the continuum and DEGB-only models. The overall probability of a DEGB in the continuum model is 0.165, and the bulk of that comes from small pipes. Table 15 repeats the information of Table 4 at the resolution of STP's 45 weld cases.

Tables 5 and 6 depict the relative contributions by category for each pipe size under the con-tinuum and DEGB-only models. For example, under the continuum model, given that we have a category 2 break, the conditional probability it is in a 2-inch pipe is 0.1787. Tables 16 and 17 in the Appendix report the same information but at the resolution of the 45 weld cases. Figures 1-3 report these same conditional probabilities for categories 4-6 for the top few weld cases under both the continuum break-size model and the DEGB-only model.

4

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 2: LOCA probabilities for STP pipe sizes for current-day estimates when welds can experience a contimnum of break sizes. Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break) 1.00 5.86E-01 X X X X X 5.86E-01 2.00 4.06E-02 2.43E-02 5.14E-05 X X X 6.49E-02 2.50 7.56E-04 3.58E-04 5.03E-05 X X X 1.16E-03 3.00 1.53E-03 7.59E-04 3.37E-05 X X X 2.33E-03 4.00 4.28E-03 2.12E-03 9.13E-05 X X X 6.49E-03 6.00 4.07E-03 2.02E-03 7.66E-05 6.69E-06 X X 6.17E-03 8.00 6.56E-02 3.30E-02 1.14E-03 1.06E-04 X X 9.98E-02 10.00 6.46E-04 3.25E-04 1.13E-05 7.77E-07 3.23E-08 X 9.82E-04 12.00 1.17E-01 5.88E-02 2.04E-03 1.41E-04 5.85E-06 X 1.78E-01 16.00 1.7SE-02 8.44E-03 8.43E-04 1.31E-04 5.50E-06 X 2.72E-02 27.50 1.25E-03 3.84E-04 2.86E-05 3.71E-06 2.59E-07 1.O1E-07 1.67E-03 29.00 1.85E-02 5.03E-03 5.06E-04 7.96E-05 5.23E-06 1.70E-06 2.41E-02 31.00 1.31E-03 4.04E-04 3.01E-05 3.91E-06 2.73E-07 1.06E-07 1.75E-03 P(Break) 8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-05 1.90E-06 1.OOE+00 Table 3: LOCA probabilities for STP pipe sizes for current-day estimates when welds can only experience a DEGB. Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break) 1.00 8.59E-01 X X X X X 8.59E-01 2.00 X 1.36E-01 1.97E-03 X X X 1.38E-01 2.50 X X 1.51E-03 X X X 1.51E-03 3.00 X X 6.11E-04 X X X 6.11E-04 4.00 X X 8.18E-04 X X X 8.18E-04 6.00 X X X 4.65E-05 X X 4.65E-05 8.00 X X X 4.27E-04 X X 4.27E-04 10.00 X X X X 8.79E-08 X 8.79E-08 12.00 X X X X 1.09E-05 X 1.09E-05 16.00 X X X X 6.19E-06 X 6.19E-06 27.50 X X X X X 1.01E-07 1.O1E-07 29.00 X X X X X 1.70E-06 1.70E-06 31.00 X X X X X 1.06E-07 1.06E-07 P(Break) L8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-0.5 1.90E-06 1.OOE+00 5

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 4: The table shows the probability of a. break occurring at each pipe size, and tile conditional probability that a break at a given size is a DEGB, using both the continuum break-size model and the DEGB-only model. Below the table, the sums of the probabilities for the break sizes are shown to be one, and using those probabilities as weights we show the overall probability of a break being a DEGB under both models. Pipe sizes are in inches.

Pipe Size Continuous DEGB Only P(Break) P(DEGB) P(Break) P(DEGB) 1.00 5.86E-01 2.69E-01 8.59E-01 1.00E+00 2.00 6.49E-02 1.13E-01 1.38E-01 1.00E+00 2.50 1.16E-03 3.37E-02 1.51E-03 1.OOE+00 3.00 2.33E-03 6.85E-03 6.11E-04 1.00E+00 4.00 6.49E-03 3.29E-03 8.18E-04 1.OOE+00 6.00 6.17E-03 7.28E-04 4.65E-05 1.00E+00 8.00 9.98E-02 4.13E-04 4.27E-04 1.OOE+00 10.00 9.82E-04 3.20E-05 8.79E-08 1.OOE+00 12.00 1.78E-01 2.19E-05 1.09E-05 1.00E+00 16.00 2.72E-02 S.13E-05 6.19E-06 1.00E+00 27.50 1.67E-03 4.31E-05 1.01E-07 1.00E+00 29.00 2.41E-02 4.31E-05 1.70E-06 1.OOE+00 31.00 1.75E-03 3.53E-05 1.06E-07 1.OOE+00 1.OOE+00 1.65E-01 1.00E+00 :1.00E+00 6

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 5: Relative contributions by category when welds can experience a continuum of break sizes.

Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 1.00 68.19% X X X X X 2.00 4.72% 17.87% 1.05% X X X 2.50 0.09% 0.26% 1.03% X X X 3.00 0.18% 0.56% 0.69% X X X 4.00 0.50% 1.56% 1.86% X X X 6.00 0.47% 1.48% 1.56% 1.41% X X 8.00 7.64% 24.28% 23.32% 22.50% X X 10.00 0.08% 0.24% 0.23% 0.16% 0.19% X 12.00 13.60% 43.25% 41.55% 29.73% 34.13% X 16.00 2.08% 6.21% 17.19% 27.76% 32.09% X 27.50 0.15% 0.28% 0.58% 0.78% 1.51% 5.30%

29.00 2.16% 3.70% 10.33% 16.82% 30.48% 89.13%

31.00 0.15% 0.30% 0.61% 0.83% 1.59% 5.57%

Total 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

Table 6: Relative contributions by category when welds can experience only a DEGB. Pipe sizes are in inches.

Pipe Size Cat I Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 1.00 100.00% X X X X X 2.00 X 100.00% 40.15% X X X 2.50 X X 30.71% X X X 3.00 X X 12.45% X X X 4.00 X X 16.69% X X X 6.00 X X X 9.83% X X 8.00 X X X 90.17% X X 10.00 X X X X 0.51% X 12.00 X X X X 63.37% X 16.00 X X X X 36.11% X 27.50 X X X X X 5.30%

29.00 X X X X X 89.13%

31.00 X X X X X 5.57%

Total 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

7

NOC-AE-14003101 Attachment 1 Enclosure 1 Probability of Break By Weld Case - Category 4:

Continuum Break Size Model

• Weld 4- 2

• Weld 9 - 4A Weld 25 - 7A Weld 26 - 7B

• Weld 27 - 7C Total % = 89.99%

(a) Continuum model Probability of Break By Weld Case - Category 4:

DEGB Only Model

• Weld 13 - 5A

• Weld 26 - 7B Weld 27 - 7C Total % = 94.98%

(b) DEGB-only model Figure 1: This figure depicts the probability of a break occurring in each weld case conditional on the break being in category 4. Part (a) of the figure corresponds to the continuum break-size model and part (b) corresponds to the DEGB-only model.

8

NOC-AE-14003101 Attachment 1 Enclosure 1 Probability of Break By Weld Case - Category 5:

Continuum Break Size Model

" Weld 1 - IA

" Weld 4 - 2 Weld 9 - 4A

" Weld 25 - 7A Total % = 92.26%

(a) Continuum model Probability of Break By Weld Case - Category 5:

DEGB Only Model

" Weld 9 - 4A

" Weld 25 - 7A Total % = 93.16%

(b) DEGB-only model Figure 2: This figure depicts the probability of a break occurring in each weld case conditional on the break being in category 5. Part (a) of the figure corresponds to the continuum break-size model and part (b) corresponds to the DEGB-only model.

9

NOC-AE-14003101 Attachment 1 Enclosure 1 Probability of Break By Weld Case - Category 6:

Continuum Break Size Model 5.02% 5.02%

• Weld 1 - 1A

• Weld 4- 2 Weld 5 - 3A Weld 6 - 31 Total % = 98.86%

(a) Continuum model Probability of Break By Weld Case - Category 6:

Continuum Break Size Model 5.02% , 5.02%

• Weld 1 - 1A

• Weld 4 - 2

• Weld 5 - 3A Weld 6 - 31 Total % = 98.86%

(b) DEGB-only model Figure 3: This figure depicts the probability of a break occurring in each weld case conditional on the break being in category 6. Part (a) of the figure corresponds to the continuum break-size model and part (b) corresponds to the DEGB-only model. As we indicate in the text, these conditional probabilities are identical under the continuum- and DEGB-only model for category 6.

10

NOC-AE-14003101 Attachment 1 Enclosure 1 4 Continuum and DEGB-Only Models under a Top-Down Analysis A top-down analysis is discussed in Pan et al. (2013) and in that model u* = 1/J1Ij; i.e., we ignore relevant information regarding degradation mechanisms and simply say that each weld which can experience a break of a particular size is equally likely to have the break. As we discuss above, the definitions of the sets Ij differ under the continuum break-size model and the DEGB-only model, and hence the corresponding top-down models differ. In this section, we provide tables analogous to those from the previous section using the top-down models.

Table 7 provides the top-down joint probability mass function (2) analogous the one presented in Table 2 for the continuum break-size model, and Table 8 reports analogous results for the DEGB-only model. As with the hybrid approach in which we use bottom-up values, under the top-down approach the DEGB-only model increases the likelihood of breaks in small pipes. WVe again preserve NUREG-1829 frequencies, and hence the probabilities of a category 1, 2, ... , 6 break are identical in both models and identical to the values we obtain in Section 3. As in the previous section, the probabilities in the category 6 columns are identical under the continuum and DEGB-only models.

As Table 9 indicates, the probability of a DEGB in the continuum break-size model is 0.0746, with the bulk of that coming from small pipe sizes. This value is less than half of the value tinder the hybrid approach, indicating that using weights from the bottom-up approach increases the probability of a DEGB over simply using the NUREG-1829 values and assigning equal weights to pipes of equal sizes., Tables 10 and 11 specify the relative contributions by category for each pipe size under the top-down variants of the continuum and DEGB-only models.

11

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 7: LOCA probabilities for STP pipe sizes for current-day estimates when welds can experience a continuum of break sizes under the top-down model. Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break) 1.00 2.39E-01 X X X X X 2.39E-01 2.00 9.05E-02 2.01E-02 5.38E-04 X X X 1.11E-01 2.50 6.39E-03 1.40E-03 5.66E-05 X X X 7.84E-03 3.00 2.87E-02 6.29E-03 2.55E-04 X X X 3.53E-02 4.00 9.58E-02 2.IOE-02 8.49E-04 X X X 1.18E-01 6.00 9.05E-02 1.98E-02 7.29E-04 1.1OE-04 X X i.1E-01 8.00 5.75E-02 1.26E-02 4.63E-04 7.OOE-05 X X 7.06E-02 10.00 3.19E-02 6.99E-03 2.57E-04 3.74E-05 2.25E-06 X 3.92E-02 12.00 1.39E-01 3.05E-02 1.12E-03 1.63E-04 9.82E-06 X 1.71E-01 16.00 1.06E-02 2.33E-03 8.57E-05 1.25E-05 7.50E-07 X 1.31E-02 27.50 1.70E-02 3.73E-03 1.37E-04 1.99E-05 1.08E-06 4.76E-07 2.09E-02 29.00 2.13E-02 4.66E-03 1.71E-04 2.49E-05 1.35E-06 5.95E-07 2.62E-02 31.00 2.98E-02 6.52E-03 2.40E-04 3.49E-05 1.89E-06 8.33E-07 3.66E-02 P(Break) 8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-05 1.90E-06 1.OOE+00 Table 8: LOCA probabilities for STP pipe sizes for current-day estimates when welds can only experience a DEGB under the top-down model. Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break) 1.00 8.59E-01 X X X X X 8.59E-01 2.00 X 1.36E-01 2.38E-03 X X X 1.38E-01 2.50 X X 2.20E-04 X X X 2.20E-04 3.00 X X 8.10E-04 X X X 8.1OE-04 4.00 X X 1.49E-03 X X X 1.49E-03 6.00 X X X 3.57E-04 X X 3.57E-04 8.00 X X X 1.16E-04 X X 1.16E-04 10.00 X X X X 3.51E-06 X 3.51E-06 12.00 X X X N 1.30E-05 X 1.30E-05 16.00 X X X X 6.40E-07 X 6.40E-07 27.50 X X X X X 4.76E-07 4.76E-07 29.00 X X X X X 5.95E-07 5.95E-07 31.00 X X X X X 8.33E-07 8.33E-07 P(Break) 8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-05 1.90E-06 1.OOE+00 12

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 9: Under the top-down model, the table shows the probability of a break occurring at each pipe size, and the probability that a break at a given size is a DEGB, using both the continuum break-size model and the DEGB-only model. Below the table, the sums of the probabilities for the break sizes are shown to be one, and using those probabilities as weights we show the overall probability of a break being a DEGB under both models. Pipe sizes are in inches.

Pipe Size Continuous DEGB Only P(Break) P(DEGB) P(Break) P(DEGB) 1.00 2.39E-01 3.02E-01 8,59E-01 1.00E+00 2.00 1i.E-01 1.45E-02 1.38E-01 1.00E+00 2.50 7.84E-03 6.34E-03 2.20E-04 1.OOE+00 3.00 3.53E-02 5.18E-03 8.10E-04 1.OOE+00 4.00 1.18E-01 2.86E-03 1.49E-03 1.00E+00 6.00 1i.E-01 7.89E-04 3.57E-04 1.00E+00 8.00 7.06E-02 4.04E-04 1.16E-04 1.OOE+00 10,00 3.92E-02 5.69E-05 3.51E-06 1.00E+00 12.00 1.71E-01 4.33E-05 1.30E-05 1.00E+00 16.00 1.31E-02 3.12E-05 6.40E-07 1.00E+00 27.50 2.09E-02 2.28E-05 4.76E-07 1.00E+00 29.00 2.61E-02 2.28E-05 5.95E-07 1.OOE+00 31.00 3.66E-02 2.28E-05 8.33E-07 1.00E+00 1.00E+00 7.46E-02 1.00E+00 1.OOE-00 13

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 10: Relative contributions by category when welds can experience a continuum of break sizes under the top-down model. Pipe sizes are in inches.

Pipe Size Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 1.00 27.86% X X X X X 2.00 10.54% 14.77% 10.96% X X X 2.50 0.74% 1.03% 1.15% X X X 3.00 3.35% 4.63% 5.19% X X X 4.00 11.16% 15.43% 17.31% X X X 6.00 10.54% 14.58% 14.86% 23.29% X X 8.00 6.69% 9.26% 9.44% 14.80% X X 10.00 3.72% 5.14% 5.24% 7.90% 13.12% X 12.00 16.24% 22.47% 22.90% 34.51% 57.31% X 16.00 1.24% 1.71% 1.75% 2.63% 4.37% X 27.50 1.98% 2.74% 2.80% 4.21% 6.30% 25.00%

29.00 2.48% 3.43% 3.50% 5.27% 7.87% 31.25%

31.00 3.47% 4. 80% 4.89% 7.38% 11.02% 43.75%

Total 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

Table 11: Relative contributions by category when welds can experience only a DEGB under the top-down model. Pipe sizes are in inches.

Pipe Size Cat I Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 1.00 100.00% X X X X X 1.50 X X X X X X 2.00 X 100.00% 48.59% X X X 2.50 X X 4.49% X X X 3.00 X X 16.52% X X X 4.00 X X 30.41% X X X 6.00 X X X 75.45% X X 8.00 X X X 24.55% X X 10.00 X X X X 20.45% X 12.00 X X X X 75.82% X 16.00 X X X X 3.73% X 27.50 X X X X X 25.00%

29.00 X X X X X 31.25%

31.00 X X X X X 43.75%

Total 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

14

NOC-AE-14003101 Attachment 1 Enclosure 1 5 Summary The analysis we present here indicates that under the implemented "continuum" model, the proba-bility a pipe experiences a DEGB, given that it has a break, is 0.165. We also present a DEGB-only model in which all breaks are DEGBs, and we compare how that model allocates breaks to weld cases, relative to the continuum model. We preserve NUREG-1829 frequencies throughout, and hence the latter model dramatically decreases the probability that a large pipe experiences a break.

Finally, we put aside the bottom-up frequencies and instead assume each pipe within a size category is equally likely to experience a break, given we have a break in that category. Under this top-down approach for the continuum model, the probability a pipe experiences a DEGB, given that it has a break, is 0.0746, indicating that using the bottom-up frequencies increases the probability of a DEGB by more than a factor of two.

15

NOC-AE-14003101 Attachment 1 Enclosure 1 References Fleming, K. N., B. 0. Lydell, and D. Chrun (2011, July). Development of LOCA Initiating Event Frequencies for South Texas Project GSI-191. Technical Report, KnF Consulting Services, LLC, Spokane, WA.

Pan, Y.-A., E. Popova, and D. P. Morton (2013, January). South Texas Project Risk-Informed GSI-191 Evaluation, Volume 3, Modeling and Sampling LOCA Frequency and Break Size. Technical report, STP-RIGS1191-V03.02, Revision 4, The University of Texas at Austin.

Tregoning, R., P. Scott, and A. Csontos (2008, April). Estimating Loss-of-Coolant Accident (LOCA) Frequencies Through the Elicitation Process: Main Report (NUREG-1829). NUREG 1829, NRC, Washington, DC.

16

NOC-AE-14003101 Attachment 1 Enclosure 1 Appendix The appendix presents the results for each weld case as opposed to results which are aggregated across break sizes.

Table 12: Characteristics of the 45 Weld Cases at STP System 1 Weld # Weld Case 1 Damage Mechanisms # of T Pipe 1Welds Size (in.) DEGB (in.) Size Hot Leg 1 IA SC+D&C 4 29.00 41.01 Hot Leg 2 1B D&C 11 29.00 41.01 Hot Leg 3 1C TF+D&C' 1 29.00 41.01 SC Inlet 4 2 SC+D&C 4 29.00 41.01 Cold Leg 5 3A SC+D&C 4 27.50 :38.89 Cold Leg 6 3B SC+D&C 4 31.00 43.84 Cold Leg 7 3C D&,C 12 27.50 38.89 Cold Leg 8 3D D&C 24 31.00 43.84 Surge Line 9 4A SC+TF+D&C 1 16.00 22.63 Surge Line 10 4B TF+D&C 7 16.00 22.63 Surge Line 11 4C TF+D&C 2 16.00 22.63 Surge Line 12 4D TF+D&C 6 2.50 3.54 Pressurizer 13 5A TF+D&C 29 6.00 8.49 Pressurizer 14 5B TF+D&C 14 3.00 4.24 Pressurizer 15 5C D&C 53 4.00 5.66 Pressurizer 16 5D D&,C 4 3.00 4.24 Pressurizer 17 5E D&C 29 6.00 8.49 Pressurizer 18 5F SC+D&C 0 6.00 8.49 Pressurizer 19 5G D&C (Weld Overlay) 4 6.00 8.49 Pressurizer 20 5H D&C 2 4.00 5.66 Pressurizer 21 5I TF+D&C 2 2.00 2.83 Pressurizer 22 53 SC+TF+D&C 0 6.00 8.49 Small Bore 23 6A VF+SC+D&C 16 2.00 2.83 Small Bore 24 6B VF+SC+D&C 193 1.00 1.41 SIR 25 7A TF+D&C 21 12.00 16.97 SIR 26 7B TF+D&C 9 8.00 11.31 SIR 27 7C SC+TF+D&C 3 8.00 11.31 SIR 28 7D SC+D&C 3 12.00 16.97 SIR 29 7E D&C 57 12.00 16.97 SIR 30 7F D&C 30 10.00 14.14 SIR 31 7G D&C 42 8.00 11.31 SIR 32 7H D&C 2:3 6.00 8.49 SIR 33 71 D&C 5 4.00 5.66 SIR 34 7J D&C, 9 3.00 4.24 SIR 35 7K D&C 10 2.00 2.83 SIR 36 7L D&C 0 1.50 2.12 ACC 37 7M SC+D&C 0 12.00 16.97 ACC 38 7N TF+D&C 35 12.00 16.97 ACC 39 70 D&SC 15 12.00 16.97 CVCS 40 SA TF+VF+D&C 10 2.00 2.83 CVCS 41 8B TF+VF+D&C 19 4.00 5.66 CVCS 42 8C VF+D&C 47 2.00 2.83 CVCS 43 8D VF+D&SC 6 4.00 5.66 CVCS 44 8E TF+D&C 4 4.00 5.66 CVCS 45 8F D&C 1 4.00 5.66 775 17

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 13: LOCA probabilities for the 45 weld cases for current-day estimates when welds can experience a continuum of break sizes.

Weld Case Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break)

Weld 1 - IA 3.13E-03 8.44E-04 8.43E-05 1.27E-05 9.16E-07 2.80E-07 4.07E-03 Weld 2 - lB 4.17E-05 1.13E-05 1.13E-06 1.70E-07 1.22E-08 3.73E-09 5.43E-05 Weld 3 - 1C 2.43E-05 6.56E-06 6.55E-07 9.88E-08 7.12E-09 2.17E-09 3.16E-05 Weld 4 - 2 1.53E-02 4.16E-03 4.20E-04 6.66E-05 4.29E-06 1.41E-06 2.OOE-02 Weld 5 - 3A 1.18E-03 3.64E-04 2.71E-05 3.52E-06 2.46E-07 9.56E-08 1.58E-03 Weld 6 - 3B 1.1SE-03 3.64E-04 2.71E-05 3.52E-06 2.46E-07 9.56E-08 1.58E-03 Weld 7 - 3C 6.56E-05 2.01E-05 1.50E-06 1.95E-07 1.36E-08 5.29E-09 8.74E-05 Weld 8 - 3D 1.31E-04 4.03E-05 3.OOE-06 3.90E-07 2.72E-08 1.06E-08 1.75E-04 Weld 9 - 4A 1.65E-02 7.83E-03 7.82E-04 1.22E-04 5.10E-06 X 2.53E-02 Weld 10 - 4B 8.83E-04 4.18E-04 4.18E-05 6.50E-06 2.72E-07 X 1.35E-03 Weld 11 - 4C 4.11E-04 1.95E-04 1.95E-05 3.03E-06 1.27E-07 X 6.29E-04 Weld 12 - 4D 7.56E-04 3.58E-04 5.03E-05 X X X 1.16E-03 Weld 13 - 5A 2.51E-03 1.24E-03 4.76E-05 4.13E-06 X X 3.80E-03 Weld 14 - 5B 1.21E-03 5.98E-04 2.68E-05 X X X 1.84E-03 Weld 15 - 5C 1.72E-03 8.47E-04 3.80E-05 X X X 2.60E-03 Weld 16 - 5D 1.30E-04 6.39E-05 2.87E-06 X X X 1.96E-04 Weld 17 - 5E 9.39E-04 4.64E-04 1.78E-05 1.55E-06 X X 1.42E-03 Weld 18 - 5F O.OOE+00 0.OOE+00 0.OOE+00 0.OOE+00 X X 0.OOE+00 Weld 19- 5G 1.31E-04 6.48E-05 2.49E-06 2.16E-07 X X 1.99E-04 Weld 20 - 5H 6.48E-05 3.20E-05 1.43E-06 X X X 9.82E-05 Weld 21 - 5I 1.73E-04 1.14E-04 X X X X 2.87E-04 Weld 22 - 5J O.OOE+00 0.OOE+00 0.OOE+00 0.OOE+00 X X O.OOE+00 Weld 23 - 6A 3.77E-02 2.28E-02 X X X X 6.05E-02 Weld 24 - 6B 5.86E-01 X X X X X 5.86E-01 Weld 25 - 7A 1.10E-01 5.53E-02 1.92E-03 1.32E-04 5.51E-06 X 1.67E-01 Weld 26 - 7B 4.71E-02 2.37E-02 8.22E-04 7.65E-05 X X 7.1SE-02 Weld 27- 7C 1.75E-02 8.82E-03 3.06E-04 2.85E-05 X X 2.67E-02 Weld 28 - 7D 2.OOE-03 L.O0E-03 3.50E-05 2.41E-06 l.OOE-07 X 3.05E-03 Weld 29 - 7E 1.23E-03 6.17E-04 2.14E-05 1.48E-06 6.14E-08 X 1.87E-03 Weld 30 - 7F 6.46E-04 3.25E-04 1.13E-05 7.77E-07 3.23E-08 X 9.82E-04 Weld 31 - 7G 9.04E-04 4.55E-04 1.58E-05 1.47E-06 X X 1.38E-03 Weld 32 - 7H 4.95E-04 2.49E-04 8.63E-06 8.03E-07 X X 7.53E-04 Weld 33 - 71 L.OSE-04 5.41E-05 2.21E-06 X X. X 1.64E-04 Weld 34 - 7J 1.94E-04 9.74E-05 3.98E-06 X X X 2.95E-04 Weld 35 - 7K 2.15E-04 1.41E-04 X X X X 3.57E-04 Weld 36 - 7L 0.OOE+00 0.OOE+00 X X X. X 0.00E+00 Weld 37 - 7M O.OOE+00 O.OOE+00 0.OOE+00 0.OOE+00 0.OOE+00 X 0.OOE+00 Weld 38 - 7N 3.42E-03 1.72E-03 5.99E-05 4.15E-06 1.72E-07 X 5.21E-03 Weld 39 - 70 1.77E-04 8.90E-05 3.10E-06 2.15E-07 8.92E-09 X 2.69E-04 Weld 40 - 8A 8.09E-04 4.03E-04 1.68E-05 X X X 1.23E-03 Weld 41 - 8B 1.54E-03 7.65E-04 3.20E-05 X X X 2.33E-03 Weld 42 - 8C 1.66E-03 8.28E-04 3.46E-05 X X X 2.53E-03 Weld 43 - 8D 2.12E-04 1.06E-04 4.41E-06 X X X 3.22E-04 Weld 44 - 8E 6.03E-04 3.OOE-04 1.25E-05 X X X 9.16E-04 Weld 45 - 8F 3.54E-05 1.76E-05 7.36E-07 X X X 5.37E-05 P(Break) 8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-05 1.90E-06 I.00E+00 18

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 14: LOCA probabilities for the 45 weld cases for current-day estimates when welds can experience only a DEGB.

Weld Case Cat 1 Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 P(Break)

Weld 1 - 1A X X X X X 2.80E-07 2.80E-07 Weld 2 - 1B X X X X X 3.73E-09 3.73E-09 Weld 3 - IC X X X X X 2.17E-09 2.1TE-09 Weld 4 - 2 X X X X X 1.41E-06 1.41E-06 Weld 5 - 3A X X X X X 9.56E-08 9.56E-08 Weld 6 - 3B X X X X X 9.56E-OS 9.56E-08 Weld 7 - 3C X X X X X 5.29E-09 5.29E-09 Weld 8 - 3D X X X X X 1.06E-08 1.06E-08 Weld 9 - 4A X X X X 5.74E-06 X 5.74E-06 Weld 10 - 4B X X X X 3.07E-07 X 3.07E-07 Weld 11 - 4C X X X X 1.43E-07 X 1.43E-07 Weld 12 - 4D X X 1.51E-03 X X X 1.51E-03 Weld 13 - 5A X X X 2.86E-05 X X 2.86E-05 Weld 14 - 5B X X 4.86E-04 X X X 4.86E-04 Weld 15 - 5C X X 3.31E-04 X X X 3.31E-04 Weld 16 - 5D X X 5.20E-05 X X X 5.20E-05 Weld 17 - 5E X X X 1.07E-05 X X 1.07E-05 Weld 18 - 5F X X X O.OOE+00 X X O.OOE+00 Weld 19 - 5G X X X 1.50E-06 X X 1.50E-06 Weld 20 - 5H X X 1.25E-05 X X X 1.25E-05 Weld 21 - 5I X 6.08E-04 X X X X 6.08E-04 Weld 22 - 5J X X X O.OOE+00 X X O.OOE+00 Weld 23 - 6A X 1.35E-01 X X X X 1.35E-01 Weld 24 - 6B 8.59E-01 X X X X X 8.59E-01 Weld 25 - 7A X X X X 1.02E-05 X 1.02E-05 Weld 26 - 7B X X X 3.07E-04 X X 3.07E-04 Weld 27 - 7C X X X 1.14E-04 X X 1.14E-04 Weld 28 - 7D X X X X 1.86E-07 X 1.86E-07 Weld 29 - 7E X X X X 1.14E-07 X 1.14E-07 Weld 30 - 7F X X X X 8.79E-08 X 8.79E-08 Weld 31 - 7G X X X 5.88E-06 X X 5.88E-06 Weld 32 - 7H X X X 5.66E-06 X X 5.66E-06 Weld 33 - 71 X X 2.08E-05 X X X 2.08E-05 Weld 34 - 7J X X 7.30E-05 X X X 7.30E-05 Weld 35 - 7K X 6.66E-04 X X X X 6.66E-04 Weld 36 - 7L X O.OOE+00 X X X X O.OOE+00 Weld 37 - 7M X X X X O.OOE+00 X O.OOE+00 Weld 38 - 7N X X X X 3.18E-07 X 3.18E-07 Weld 39 - 70 X X X X 1.65E-08 X 1.65E-08 Weld 40 - 8A X X 6.45E-04 X X X 6.45E-04 Weld 41 - 8B X X 2.92E-04 X X X 2.92E-04 Weld 42 - 8C X X 1.32E-03 X X X 1.32E-03 Weld 43 - 8D X X 4.04E-05 X X X 4.04E-05 Weld 44 - 8E X X 1.15E-04 X X X 1.15E-04 Weld 45 - 8F X X 6.73E-06 X X X 6.73E-06 P(Break) 8.59E-01 1.36E-01 4.90E-03 4.73E-04 1.71E-05 1.90E-06 1.00E+00 19

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 15: This table shows probability of a break occurring at each weld case and the probability that a break at a given weld case is a DEGB using both the continuum break-size model and the DEGB-only model. Below the table, the sum of the probabilities for the weld cases are shown to be one, and using those probabilities as weights we show the overall probability of a break being a DEGB break across all weld cases.

WlCaeI Continuous I DEGB Only P(Break) P(DEGB) P(Break) P(DEGB)

Weld 1 - 1A 4.07E-03 4.21E-05 2.80E-07 1.00E+00 Weld 2 - 1B 5.43E-05 4.21E-05 3.73E-09 1.00E+00 Weld 3 - IC 3.16E-05 4.21E-05 2.17E-09 1.00E+00 Weld 4 - 2 2.OOE-02 4.33E-05 1.41E-06 1.00E+00 Weld 5 - 3A 1.58E-03 4.31E-05 9.56E-08 1.O0E+00 Weld 6 - 3B 1.58E-03 3.53E-05 9.56E-08 1.00E+00 Weld 7 - 3C 8.74E-05 4.31E-05 5.29E-09 1.OOE+00 Weld 8 - 3D 1.75E-04 3.53E-05 1.06E-08 1.00E+00 Weld 9 - 4A 2.53E-02 8.13E-05 5.74E-06 1.OOE+00 Weld 10 - 4B 1.35E-03 8.13E-05 3.07E-07 1.00E+00 Weld 11 - 4C 6.29E-04 8.13E-05 1.43E-07 1.OOE+00 Weld 12 - 4D 1.16E-03 3.37E-02 1.51E-03 1.00E+00 Weld 13 - 5A 3.80E-03 7.28E-04 2.86E-05 1.00E+00 Weld 14- 5B 1.84E-03 6.91E-03 4.86E-04 1.00E+/-00 Weld 15- 5C 2.60E-03 3.32E-03 3.31E-04 1.00E+00 Weld 16- 5D 1.96E-04 6.91E-03 5.20E-05 1.OOE+00 Weld 17- 5E 1.42E-03 7.28E-04 1.07E-05 1.00E+00 Weld 18 - 5F 0.OOE+00 0.OOE+00 0.OOE+00 0.OOE+00 Weld 19 - 5G 1.99E-04 7.28E-04 1.50E-06 1.00E+00 Weld 20 - 5H 9.82E-05 3.32E-03 1.25E-05 1.00E+00 Weld 21 - 5I 2.87E-04 1.13E-01 6.08E-04 1.00E+00 Weld 22 - 5J 0.OOE+00 0.00E+00 0.OOE+00 0.OOE+00 Weld 23 - 6A 6.05E-02 1.19E-01 1.35E-01 1.OOE+00 Weld 24 -16B 5.86E-01 2.69E-01 8.59E-01 1.00E+00 Weld 25 - 7A 1.67E-01 2.19E-05 1.02E-05 1.00E+00 Weld 26 - 7B 7.18E-02 4.13E-04 3.07E-04 1.00E+00 Weld 27- 7C 2.67E-02 4.13E-04 1.14E-04 1.00E+00 Weld 28 - 7D 3.05E-03 2.19E-05 1.86E-07 1.00E+00 Weld 29 - 7E 1.87E-03 2.19E-05 1.14E-07 1.00E+00 Weld 30 - 7F 9.82E-04 3.20E-05 8.79E-08 1.00E+00 Weld 31 - 7G 1.38E-03 4.13E-04 5.88E-06 1.00E+00 Weld 32 - 7H 7.53E-04 7.26E-04 5.66E-06 1.OOE+00 Weld 33 - 71 1.64E-04 3.31E-03 2.08E-05 1.00E+00 Weld 34 - 7J 2.95E-04 6.46E-03 7.30E-05 1.OOE+00 Weld 35 - 7K 3.57E-04 1.00E-01 6.66E-04 1.00E+00 Weld 36 - 7L 0.OOE+00 0.00E+00 0.OOE+00 0.OOE+00 Weld 37 - 7M O.OOE+00 0.OOE+00 0.OOE+00 0.OOE+00 Weld 38 - 7N 5.21E-03 2.19E-05 3.18E-07 1.OOE+00 Weld 39 - 70 2.69E-04 2.19E-05 1.65E-08 1.00E+00 Weld 40 - 8A 1.23E-03 1.37E-02 6.45E-04 1.00E+00 Weld 41 - 8B 2.33E-03 3.27E-03 2.92E-04 1.00E+00 Weld 42 - 8C 2.53E-03 1.37E-02 1.32E-03 1.OOE+00 Weld 43 - 8D 3.22E-04 3.27E-03 4.04E-05 1.00E+00 Weld 44 - 8E 9.16E-04 3.27E-03 1.15E-04 1.OOE+00 Weld 45 - 8F 5.37E-05 3.27E-03 6.73E-06 1.00E+00 1.00E+00 1.65E-01 1.OOE+00 1.OOE-00 20

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 16: Relative contributions by category when welds can experience a continuum of break sizes.

Weld Case Cat I Cat 2 Cat 3 Cat 4 Cat 5 Cat 6 Weld 1 - IA 0.36% 0.62% 1.72% 2.69% 5.34% 14.69%

Weld 2 - 1B 0.00% 0.01% 0.02% 0.04% 0.07% 0.20%

Weld 3 - 1C 0.00% 0.00% 0.01% 0.02% 0.04% 0.11%

Weld 4 - 2 1.79% 3.06% 8.57% 14.08% 25.03% 74.13%

Weld 5 - 3A 0.14% 0.27% 0.55% 0.74% 1.43% 5.02%

Weld 6 - 3B 0.14% 0.27% 0.55% 0.74% 1.43% 5.02%

Weld 7 - 3C 0.01% 0.01% 0.03% 0.04% 0.08% 0.28%

Weld 8 - 3D 0.02% 0.03% 0.06% 0.08% 0.16% 0.56%

Weld 9 - 4A 1.92% 5.76% 15.94% 25.74% 29.76% X Weld 10 - 4B 0.10% 0.31% 0.85% 1.37% 1.59% X Weld 11 - 4C 0.05% 0.14% 0.40% 0.64% 0.74% X Weld 12 - 4D 0.09% 0.26% 1.03% X X X Weld 13 - 5A 0.29% 0.91% 0.97% 0.87% X X Weld 14 - 5B 0.14% 0.44% 0.55% X X X Weld 15 - 5C 0.20% 0.62% 0.78% X X X Weld 16 - 5D 0.02% 0.05% 0.06% X X X Weld 17 - 5E 0.11% 0.34% 0.36% 0.33% X X Weld 18 - 5F 0.00% 0.00% 0.00% 0.00% X X Weld 19 - 5G 0.02% 0.05% 0.05% 0.05% X X Weld 20 - 5H 0.01% 0.02% 0.03% X X X Weld 21 - 5I 0.02% 0.08% X X X X Weld 22 - 5J 0.00% 0.00% 0.00% 0.00% X X Weld 23 - 6A 4.39% 16.78% X X X X Weld 24 - 6B 68.19% X X X X X Weld 25 - 7A 12.81% 40.72% 39.12% 27.98% 32.13% X Weld 26 - 7B 5.49% 17.45% 16.76% 16.18% X X Weld 27 - 7C 2.04% 6.49% 6.23% 6.02% X X Weld 28 - 7D 0.23% 0.74% 0.71% 0.51% 0.59% X Weld 29 - 7E 0.14% 0.45% 0.44% 0.31% 0.36% X Weld 30 - 7F 0.08% 0.24% 0.23% 0.16% 0.19% X Weld 31 - 7G 0.11% 0.33% 0.32% 0.31% X X Weld 32 - 7H 0.06% 0.18% 0.18% 0.17% X X Weld 33 - 71 0.01% 0.04% 0.05% X X X Weld 34 - 7J 0.02% 0.07% 0.08% X X X Weld 35 - 7K 0.03% 0.10% X X X X Weld 36 - 7L 0.00% 0.00% X X X X Weld 37 - 7M 0.00% 0.00% 0.00% 0.00% 0.00% X Weld 38 - 7N 0.40% 1.27% 1.22% 0.88% 1.01% X Weld 39 - 70 0.02% 0.07% 0.06% 0.05% 0.05% X Weld 40 - 8A 0.09% 0.30% 0.34% X X X Weld 41 - 8B 0.18% 0.56% 0.65% X X X Weld 42 - 8C 0.19% 0.61% 0.70% X X X Weld 43 - 8D 0.02% 0.08% 0.09% X X X Weld 44 - 8E 0.07% 0.22% 0.26% X X X Weld 45 - 8F 0.00% 0.01% 0.01% X X X Sum 100.00% 100.00% 100.00% 100.00% 100.00% 100.00%

21

NOC-AE-14003101 Attachment 1 Enclosure 1 Table 17: Relative contributions by category when welds can experience o0ly a DEGB.

22

NOC-AE-1 4003101 Attachment 1 Enclosure 2 Enclosure 2 to Attachment 1 Supporting resolution of APLAB, STP PRA Model - General: RAI 2 "Determination of Adequacy of Plant-Specific PRA to Support Risk-Informed Resolution of GSI-191, Rev. 1"

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 Determination of Adequacy of Plant-Specific PRA to Support Risk-Informed Resolution of GSI-191 Introduction The Probabilistic Risk Assessment (PRA) supporting a risk-informed submittal is expected to conform to specific technical adequacy requirements. Regulatory Guide 1.200' (RG 1.200) provides detailed guidance regarding these technical adequacy requirements by addressing the technical adequacy of the PRA.

RG 1.200 addresses initiating events from two distinct hazard categories: internal hazards and external hazards. Internal hazards involve internal events (e.g., turbine trip, loss of offsite power, etc.) and internal floods. External hazards include seismic events, high winds, external floods and 'other' external hazards. In RG 1.200, internal fires are addressed as an "external" hazard.

RG 1.200 addresses requirements for both Level 1 (i.e., that portion of a risk scenario that starts with an initiating event, addresses the response of the operators and equipment, and proceeds to the determination of either "successful termination" of the challenge or fuel damage) and Level 2 (i.e., that portion of a risk scenario that addresses the progression of events following fuel damage leading to a potential release of radioactive material from the plant).

Technical elements for the Level 1 internal events portion of the risk model are: initiating event analysis, success criteria analysis, accident sequence analysis, systems analysis, parameter estimation analysis, human reliability analysis, and quantification. Technical elements for the Level 2 portion of the risk model are: plant damage state analysis, accident progression analysis, source term analysis and quantification. Documentation and uncertainty characterization are important components of the technical element requirements.

RG 1.200 identifies the ASME/ANS PRA Standard' (the "Standard") as one acceptable approach to demonstrate technical adequacy. Appendix A3 of RG 1.200 provides a summary of the NRC position on each high level and detailed supporting requirement found in the Standard. The Standard covers the 1 U.S. Nuclear Regulatory Commission, Regulatory Guide 1.200, "An Approach for Determining the Technical Adequacy of Probabilistic Risk Analysis Results for Risk-Informed Activities," Revision 2, March 2009.

2 ASME/ANS RA-Sa-2009, "Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications," Addendum A to RA-S-2008, ASME, New York, NY, American Nuclear Society, La Grange Park, IL,February 2009.

3 Appendix A is entitled "NRC Regulatory Position on ASME/ANS Standard."

1

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 same scope as does RG 1.200 (i.e., at-power, internal events and external events). Note that per RG 1.200 and the Standard, a Peer Review is a required element of an acceptable PRA.

The level of compliance with the detailed supporting requirements articulated in the Standard is described in terms of three "capability categories." For a PRA to be considered technically adequate for risk informed applications specific supporting requirements must meet capability category II or higher.

The development of new editions to PRA standards continues as part of the processes for national consensus standards. Examples of draft standards currently being balloted or under acceptance review include Low Power, Shutdown PRA, Non-LWR PRA, Advanced LWR PRA (new plants), Level 2 PRA and Level 3 PRA. Additional standards are being considered (e.g., spent fuel pool PRA). If additional PRA standards are balloted and accepted, it is likely that NRC will review them and potentially incorporate their underlying requirements - with possible refinements - into future revisions of RG 1.200. These new standards are not part of the current revision of RG 1.200.

RG 1.200 endorses the framework articulated in Regulatory Guide 1.1744 (RG 1.174). RG 1.174 provides guidance on the use of PRA in the support of risk-informed plant-specific changes in regulation. RG 1.174 references the role of RG 1.200 in providing guidance on the adequacy of a plant-specific PRA to support regulatory decision making, including:

"...demonstration that the baseline PRA (in total or specific parts) used in regulatory applications is of sufficient technical adequacy..." s RG 1.174 also included guidance on addressing issues potentially important to safety that are not specifically addressed in the plant-specific PRA. These issues might include issues arising from limitations in the scope of the plant-specific PRA. For example, if the PRA does not include explicit consideration of scenarios originating during low power or shutdown conditions, then a submittal following the RG 1.174 guidance framework will have to provide either a justification of why such scenarios are not applicable to the specific submittal request or a bounding impact on risk from such scenarios. The relevant point is that RG 1.174 provides a path forward to address potential risk-informed plant-specific regulatory decisions even if the plant-specific PRA does not address all relevant risk scenarios explicitly, when such scenarios are not within the scope of the current revision of RG 1.200.

In summary: the technical adequacy of a PRA is to be measured against the current revision RG 1.200, for those parts of the PRA that are relevant to the specific regulatbry decision. Relevant and applicable 4 U.S. Nuclear Regulatory Commission, Regulatory Guide 1.174, "An Approach for Using Probabilistic Risk Assessment in Risk-informed Decisions on Plant-Specific Changes to the Licensing Basis," Revision 2, May 2011.

s RG 1.174, page 15.

2

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 technical supporting requirements identified in the Standard are expected to meet (at least) capability category II.

Approach to Determine Technical Adequacy to Address GSI-191 A five step process is outlined to determine and document the adequacy of the plant-specific PRA to provide risk-informed information to support resolution of GSI-191. These steps are:

1. Determination of the level of compliance of the PRA to the current version of RG 1.200.

2. Identification of the PRA elements relevant to GSI-191

3. Identification of relevant high level and supporting requirements from RG 1.200

4. Comparison of relevant PRA elements to relevant RG 1.200 requirements

5. Documentation of adequacy assessment Determination of the Status of Compliance of the Plant-Specific PRA (Step 1)

The first step in the process of assessing the technical adequacy of the plant-specific PRA is to document the status of the PRA. Key questions include

1. Has the PRA been peer reviewed?

2. To the current revision of the Standard?

3. Did the peer review process consider positions documented in Appendix A of RG 1.200?

4. What was the conclusion of the peer review?

5. Did the peer review result in any findings or observations? If so, what are they?

If the PRA has been successfully reviewed against the Standard and RG 1.200, with no significant relevant findings and observations from a peer review, then the technical adequacy evaluation process is complete.

However, many plants have not yet completed upgrades to their PRAs to meet all the technical requirements for fire and seismic modeling. These plant-specific PRAs would most likely have been peer reviewed to an earlier version of the Standard and RG 1.200. (The primary difference between revision 16 and revision 2 of RG 1.200 (as well as the primary difference between the 2002 version of the ASME 6 Revision 1 of RG 1.200 references the 2002 version of the ASME Standard, as augmented by Addenda A and B (2003 and 2005, respectively). These versions address at-power internal events Level 1 and limited Level 2 PRA.

They do not address fire or seismic scenarios.

3

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 Standard - as augmented in 2003 and 2005 - and the 2009 ASME/ANS Standard) is the inclusion of fire and seismic elements in revision 2.

A plant-specific PRA that has been successfully peer reviewed with respect to an earlier version of RG 1.200 can be used to support risk-informed regulatory decision making, as long as there are no significant differences in the technical requirements relevant to the submittal.

STP Status: The STP plant-specificPRA was successfully peer reviewed against the requirements of revision 1 of RG 1.200. This includes internal events and subsequent plant response for events such as LOCAs, which is the initiating event of concern relative to GSI-191. While the STP PRA does includefire and seismic scenarios, those portions of the PRA are not compliant with revision 2 of RG 1.200. As long as no fire or seismic induced scenarios contribute to GSI-191 fuel damage phenomena, the STP PRA is technically adequate to provide risk-informed information on this issue. Allfindings and observationsfrom the STP peer review againstrevision 1 of RG 1.200 were addressedduring the process of implementing (and receiving approvalfor)the STP Risk Managed Technical Specification program7.

Identification of the PRA Elements Relevant to GSI-191 (Step 2)

The PRA can be thought of as an organized set of scenarios. Each scenario begins with an initiating event that includes a representation of the response of the plant and operators to that initiating event.

The identification of relevant PRA elements therefore begins with a consideration of the initiators included in the PRA.

Selection criteria are established to characterize the PRA scenarios. The scenarios of interest in an evaluation of GSI-191 safety concerns must meet three criteria:

1. The scenario response model for the initiator (i.e., LOCA) includes taking credit for emergency core cooling system (ECCS) recirculation mode of operation to provide core cooling,

2. The scenario involves the potential to liberate insulation installed around associated reactor coolant system piping inside primary containment,

3. And, the scenario includes a mechanism that transports the liberated piping insulation and other postulated debris to the emergency containment sump(s).

A plant-specific evaluation is necessary in the evaluation of the PRA against these criteria. The identification of initiating events of interest is best illustrated by considering a specific example.

7 STP Nuclear Operating Company, "STP Project Units 1 and 2, Docket Nos. STN 50-498, STN 50-499, Response to NRC Requests for Additional Information on STP Proposed Risk Managed Technical Specifications (TAC Nos. MD 2341 & MD 2342)," NOC-AE-07002112, February 28, 2007.

4

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 STP Example:

Many scenarios in the STP PRA take creditfor ECCS sump recirculation. Specific initiatorswhose response model includes recirculationare:

1. RCP Seal LOCA

2. Very Small LOCA

3. Non-isolable Small LOCA

4. Isolable Small LOCA

5. Open safety relief valve (SRV) (one)

6. Open SRV (two or more)

7. Medium LOCA

8. Large LOCA

9. Steam Line Break Inside Containment

10. Steam Line Break Outside Containment

11. Other transientinitiatorsincluding support system failure initiators

12. InternalPlant Fires

13. Seismic Events

14. InternalFloods

15. Other LOCAs Inside Containment (e.g., breaks in the RHR system)

InitiatorGroups 1 through 5 involve modest openings in the primary system. These events do not meet the necessary Criteria2 and 3. Groups 1 through 4 result in only a modest amount of insulating materialbeing liberated(with the amount associated with Group 1 being small).

Small and very small LOCAs, by definition, do not result in containmentspray initiation,so they lack a mechanism to transportmaterial to the sump.

Groups 5 and 6 involve the opening of SRVs. One SRV opening is equivalent to a small LOCA, so the above argumentfor Groups 1 through 4 holdsfor this group also. In other words, the necessary Criteria2 and 3 are not satisfiedfor Group 5. In addition, the location of the SRVs is such that a relatively small amount of target insulation is found near the SRVs. This would mean that criterion 2 is not met for either Group 5 or 6. Thus, initiatorGroups 1 through 6 do not satisfy the criterianecessary to result in GSI-191 safety concern phenomena.

5

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 InitiatorGroups 7 and 8 involve plant responsesthat potentially meet all three necessary conditions. Thus, Groups 7 and 8 are therefore retainedfor further evaluation.

InitiatorGroups 9 through 13 include considerationof sump recirculationfor those sequences involving feed-and-bleed or a stuck open PORV whose flow is directed to the pressurizerrelief tank (PRT), which eventually overpressurizesto the pressure limit of the engineered rupture disk.

Engineeringassessments indicate that little insulation materialis found in the vicinity of the PRT rupture disk, so that a small amount of materialwould be made available to potentially be transportedto the containment sumps (necessary condition 2). In addition,for InitiatorGroups 10 through 13 the safety injection pumps will continue to operate in the injection phase until closed loop RHR cooling is established and the containmentsprays will not actuate so that no transportmechanism will be available to transportany liberatedmaterialto the sumps (necessary criterion 3). InitiatorGroups 10 and 11 are screenedfrom further evaluation at STP, as they also do not meet the necessary Criteria2 and 3. InitiatorGroups 10 and 11 do not result in conditions necessary to result in GSI-191 safety concern phenomena8 . For InitiatorGroup 9, containment spray is anticipatedto actuate in response to the break. The plant response could include sump recirculation,but only if heat removal via the steam generatorsis lost.

InitiatorGroups 12, 13, and 14 are associatedwith external events and generally have not been shown to be significant contributorsto initiatingevents that meet the requiredcriteria.

However, Internalfires and seismic events require additionalconsideration.As in the case of Groups 9 through 11, feed-and-bleed and stuck open SRV scenariosdo not meet the conditions necessary to result in GSI-191 phenomena. The necessary additionalconsiderationfor fires and seismic scenariosfocuses on the question as to whether breachesin the primarysystem can be caused directly by the initiator.

Forseismic events, it is a common and generally considered conservative assumption that even for modest accelerations,one or more primary side pressureboundary instrument tubes may fail resulting in the equivalent of a very small LOCA. This family of seismic induced LOCA scenariosis screened based on failure to meet the necessary Criteria2 and 3. The robust nature of the primarysystem makes other seismically induced LOCAs requiring sump recirculation (i. e.,

groups 7 and 8) very unlikely. In addition, while small, medium or large LOCAs are possible at sufficiently high seismic accelerations,the common assumption is that redundantcomponents arefully correlated. Such LOCAs are of very smallfrequency. Under this assumption,for example, a medium LOCA on one primary loop would be assumed to be accompanied by medium LOCA on all other loops. The result is that seismically induced medium and large LOCAs 8 In addition, at STP, the high head injection pump shutoff head is below the pressure necessary to open the PORV, reducing the likelihood of inducing a stuck open PORV.

6

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 are modeled as being excessive LOCAs-which have no success sequences by definition (i.e., they are mapped directly to core damage).

Fire induced LOCAs leading to opening of a pressurizerPORV or reactorvessel head vents are, in principle,possible. These scenariosare screenedfrom further considerationbecause they do not meet necessary Criteria2 and 3.

With these additionalconsiderations,InitiatorGroups 12 and 13 are screenedfrom further consideration. InitiatorGroups 12 and 13 do not result in conditions necessary to result in GSI-191 safety concern phenomena at STP.

Internalflooding (Group 14) represents a hazardgroup that, asfar as GSI-191 phenomena are concerned,is identical to other transientsor supportsystem failures that may degrade to feed-and-bleed (group 11). Internalflooding scenariosare screenedfrom further consideration because they do not meet the necessary Criteria2 and 3.

Group 15 considers other LOCAs inside containment. At STP, the RHR system is wholly within containment, so that under the very unlikely conditions of an interfacing system pressurization, the RHR piping could become overpressurized.Although the RHR pumps and heat exchangers are inside containment these components are within their own respective cubicles and therefore the amount of debris that could be liberatedis highly likely to be confined within the cubicle area itself and would not be readily transportedto the emergency sumps. Additionally, the RHR system is designed with relief valve capacity and associatedsetpoints (600 psig) are sufficient that limit pressure transientsfrom events such as interfacing system LOCAs. These considerationsresult in LOCA likelihoods that are bounded by InitiatorGroups already within the scope of the analysis.

==

Conclusion:==

From a plant response analysispoint of view, the initiatorsconsidered in the STP PRA were reviewed with respect to the conditions and criterianecessaryto potentially result in GSI-191 safety concern phenomena. Only Medium and Large LOCAs survived this screening process and are retainedfor further evaluation.

Identifying Relevant Supporting Requirements from Regulatory Guide 1.200 (Step 3)

As indicated above, RG 1.200 and the Standard provide guidance to measure the adequacy of a PRA to be used in regulatory decisions. The PRA scope addressed in RG 1.200 and the Standard covers Level 1 and limited Level 2, at-power conditions, internal events and external events.

7

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 Step 3 identifies those portions of RG 1.200 and the Standard that are applicable for the application under consideration. In step 3, the portions of RG 1.200 and the Standard pertaining to the relevant PRA elements, as identified in step 2, are identified.

The process is necessarily plant-specific and is illustrated by an example.

STP Example:

In step 2, the initiating event categories relevant to considerationof GSI-191 phenomena have been shown to be limited to two specific internal events: medium and large LOCA.

The step 3 process therefore begins with consideration of the high level and supporting requirementsfor the accidentsequence analysis of medium and large LOCAs. The assessment then turns to the assessment of high level and supporting requirementsfor the success criteria analysis, accidentsequence analysis, systems analysis, parameterestimation analysis, human reliabilityanalysis, and quantification,as these elements support the PRA modeling of medium and large LOCA. The evaluation continuesfor the limited Level 2 analysis, which contains technical elements addressing the plant damage state analysis, accident progression analysis, source term analysis and quantification. Documentation and uncertainty characterizationare importantcomponents of the technical element requirements. These are covered by the high level and supporting requirements of Parts2 and 3 of the Standard. All of these high level and associatedsupporting requirements have been met in the STP PRA.

Comparison of Relevant PRA Elements to Relevant RG 1.200 Requirements (Step 4)

In step 4, the relevant PRA elements are assessed for adequacy. This step essentially brings together the results of steps 2 and 3, using those relevant elements of RG 1.200 identified in step 3 as a basis to measure the PRA elements for their technical adequacy to support regulatory decisions.

To support regulatory decision making, each relevant technical supporting requirement must meet at least capability category II. Failure to meet at least capability category II for any relevant element should be documented as a finding. Any findings or observations should be addressed to determine their potential impact on supporting resolution of GSI-191.

STP Example:

The STP PRA has undergone a peer review under revision 1 of RG1.200. The relevant technical elements of revision 1 (in Parts2 and 3) as related and applicable to GSI-191 are substantially the same as in revision 2 of RG 1.200, so the peer review can be considered "current"with respect to the PRA elements relevant to GSI-191 phenomena. This peer review therefore completes the requirementsof step 4.

Findings and Observationsfrom that peer review have also been resolved.

8

NOC-AE-14003101 Final Attachment 1 Revision 1 June 3, 2014 Enclosure 2 Documentation of Adequacy Assessment (step 5)

The process followed to determine the technical adequacy of the PRA to support GSI-191 adequacy, including resolution or treatment of any findings or observations, is to be documented.

Conclusion A process to systematically determine the technical adequacy of a PRA to support resolution of GSI-191 has been developed. The process is based on the framework of RG 1.200. The framework first identifies those elements of the PRA that are relevant to the issue. These elements are then compared to the corresponding requirements of RG 1.200, revision 2. A PRA found to be technically adequate can then support a risk-informed submittal as described in RG 1.174.

The process has been demonstrated using the STP PRA as an example STP Example:

The STP PRA has been shown to be technically adequate to support the resolution of GSI-191.

RG 1.174 does impose additionalrisk information requirements. For example, we note that RG 1.174 requiresrisk informed analyses to address all relevant operationalmodes. All modes with the primary system pressurizedpotentiallycan result in the liberationof insulating materialand the need for sump recirculation. The STP PRA only addresses at power conditions, however, the PRA bounds the risk associatedwith pressurizedconditions when not at power (i.e., not in Mode 1 operations).

9

NOC-AE-1 4003101 Attachment 2 Attachment 2 Response to ESGB Request for Additional Information

a. Chemical Effects: RAI 3, 7,11, 17, 20, 22

NOC-AE-1 4003101 Attachment 2 Page 1 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 3

Please provide the technical basis for the 1 E-05 probability for the maximum chemical effect for each break size. The engineering judgment used to determine that value appears to be arbitrary and other expert assessors could easily reach different conclusions concerning a tail probability.

STP Response:

The exponential Probability Density Function (PDF) is a shifted, truncated, and single-parameter function that requires only the mean value to specify the entire continuous distribution for all x > 1 , where x is the chemical head-loss factor. The maximum chemical effect factors for small, medium and large breaks -15.3 for small break LOCA (SBLOCA), 18.2 for medium break LOCA (MBLOCA), and 24 for large break LOCA (LBLOCA) were calculated as percentiles of their respective distributions that preserve a tail probability of 1E-05. This means that only 1 in 100,000 random samples from the distribution would be greater than the reported maximum. The maxima were included in every Latin hypercube sample (LHS) replicate, and they were assigned a weight of 1E-05 to represent all chemical factors that might be higher.

Assignment of a maximum chemical factor at 1E-05 is only arbitrary in the sense that this choice controls the probability weight carried by any failures induced by chemical factors sampled from the higher range of the distribution. The weight of 1E-05 was chosen to correspond to percentiles that ensure a quantifiable number of chemically induced Emergency Core Cooling System (ECCS) failures. If no induced failures were observed, then the maxima would indeed be suspect, and the tail probability would need further reduction. With conventional debris head loss in the range of a few feet and a structural limit of only 9.35 ft., chemical factors exceeding 10 lead to failure. Therefore, the selected maxima would be considered conservative for the Loss of Coolant Accident (LOCA) spectra. Beyond the stated maxima, probability weights become vanishingly small, and sampling beyond the stated maxima would not induce any additional failures.

NOC-AE-1 4003101 Attachment 2 Page 2 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 7

CHLE Tank Tests 3 and 4 were performed with excessive quantities of aluminum relative to the plant and with a temperature profile intended to induce chemical precipitation.

These tests resulted in chemical precipitation and provided useful information related to head loss loop response to chemical precipitates. The existing tests do not appear to address the extent of deviation from the best estimate plant conditions that could result in chemical precipitation. One potential method to inform engineering judgment with respect to chemical effects probabilities could be a series of smaller scale tests designed to evaluate the threshold concentrations of species that could result in precipitation. For example, tests could be designed to evaluate how much aluminum or calcium in solution would cause precipitation that may result in significant increases in head loss. These types of tests were included in the original chemical effects test plans but were apparently cancelled. Please discuss any plans for smaller scale testing to investigate threshold values for precipitation and whether that information would provide greater confidence in determining the probability that a post-LOCA plant condition would result in chemical precipitate formation. If there are no plans for additional tests, please provide justification for this engineering judgment.

STP Response:

Chemical Head Loss Experiment (CHLE) Tank Tests 1 and 2 (the MBLOCA and LBLOCA tests) did not result in the formation of chemical precipitates. Based on the aluminum concentrations measured in solution, those results were consistent with the existing equilibrium-based model for the prediction of the threshold concentrations of species that could result in precipitation. Experiments conducted by Argonne National Laboratory (ANL) were identified that support the equilibrium-based model. Figure 1 is a reprint from "Aluminum Solubility in Boron Containing Solutions as a Function of pH and Temperature" (ADAMS Accession # ML091610696) (1). This figure shows the results of a large number of bench-top and vertical loop experiments and other literature data at various pH and temperature values, with separation of the data into a region where precipitation occurred and a region where precipitation did not occur. The authors presented equations for empirical lines separating the two regions with the exception of a few outliers, and a second set of equations for lines (shifted upward) that encompass all instances of precipitation.

The authors of Ref. 1 subsequently published the data in Nuclear Engineering and Design(2). The second publication included an additional line on the graph showing the prediction of the solubility of amorphous aluminum hydroxide based on an equilibrium-based model (Visual MINTEQ). Figure 2 is a reprint of the figure and demonstrates excellent agreement between ANL's empirical boundary lines and the predictions of Visual MINTEQ. Using the same approach as the original figure, an upward shift of the Visual MINTEQ line would encompass all instances of precipitation. The upward shift of the Visual MINTEQ line necessary to encompass the precipitation data is a 0.45-unit increase in 'pH + p[AI]T'; in addition, this upward shift encompasses all precipitation data throughout the entire temperature range with a single equation, whereas the empirical boundaries in Ref. 1 included a separate equation for the data above 72 0C (175 'F).

NOC-AE-14003101 Attachment 2 Page 3 of 16 12.5 12 -. 0 6061 Al Test 11.5 AA 11 K

10.5 10 9.5 No PPT, WCAP-16785 -

• PPT. non-ftlocculated ANL Loop, No PPT.

• PPT flocculated

• ANL Loop, PPT.

ICET-1&5 £ ANL, STB Benchtop, 9

60 80 100 120 140 160 180 200 220 Temperature (F)

Figure 1 (Reprinted from Ref. 1): Al stability map in the 'pH+p[AI]T' vs. temperature domain for solutions containing boron. Filled and open symbols mean the occurrence of Al hydroxide precipitation and no precipitation, respectively. 'pH' and 'p[AI]T' mean the solution pH at temperature and the negative log to the base 10 of the total aluminum content as dissolved or precipitate in units of mollkg.

NOC-AE-14003101 Attachment 2 Page 4 of 16 Temperature (OF) 60 s0 100 120 140 160 180 200 220 125 12 * , 6061 AITest 1100 AI Test 11,5

-- - " *,VMINTEQ

0. 105 10 95 No PPT. WCAP-16785

• PPT non-flocculated ANL Loop. No PPT.

• PPT flocculated ANL Loop. PPT U ICET.1&5 A ANL. STB Benclitop 20 30 40 50 60 70 80 90 100 Temperature (°C)

Figure 2 (Reprinted from Ref. 2): Al hydroxide precipitation map in the 'pH+p[AIIT' vs. temperature domain based on ANL's bench top and loop test data and literature data.

CHLE Tank Tests 3 and 4 were developed (as a replacement for the tests described in the original test plan) to confirm the validity of the literature data and the existing Visual MINTEQ model for amorphous aluminum hydroxide solubility with the plant-specific chemistry at STP. Figure 3 presents the results of all 5 CHLE tank tests in the same format as the ANL data. The data demonstrate that CHLE Tank Tests 1, 2, and 5 occurred in the non-precipitation region, and that the precipitation that resulted from excessive quantities of aluminum in CHLE Tank Tests 3 and 4 were consistent with the existing data and model. An upward shift of 0.45 units of "pH + p[AI]T" encompasses the precipitation data of Tests 3 and 4; this upward shift is similar to that necessary to encompass the precipitation region in the ANL data.

Since CHLE Tank Tests 3 and 4 adequately confirmed the threshold concentrations at which precipitation would occur based on existing models and data, no additional tests are planned. The consistency between the results of CHLE Tank Tests and existing data and model provides justification for the engineering judgment approach used for aluminum solubility in the license submittal. However, the engineering judgment was based on a direct application of the Visual MINTEQ amorphous aluminum hydroxide solubility prediction without the shift of 0.45 units, since the license submittal occurred before CHLE Tank Tests 3 and 4 had been conducted. The shift of 0.45 units results in a small reduction of the concentration at which aluminum precipitates in the range of pH 7.0 to 7.3 at 60 °C (140 OF).

NOC-AE-1 4003101 Attachment 2 Page 5 of 16 13.0 TestT5 O0 <- Test T2 12.5 *(c0

• 00 <- TestT1 12.0 0.<

11.5 0 S0 , o 0 aTest T3

. 11.0

÷ CT-est T4 0.

10.5 Prediction of Visual MINTEQ 10.0 Prediction of Visual MINTEQ shifted up by 0.45 units 9.5 9.0 20 30 40 50 60 70 80 90 100 Temperature (°C)

Figure 3: Al hydroxide precipitation map of the CHLE Tank Tests in the 'pH+p[AI]T' vs.

temperature domain. Open symbols (gray) indicate results where precipitation was not observed and closed symbols (black) indicate where precipitation was observed. Each test is represented by an approximately horizontal series of data points (near-constant values of 'pH+p[AI]T' as the temperature declined over the duration of the test).

References:

1. Bahn, C.B., Kasza, K.E., Shack, W.J., and Natesan, K. "Aluminum Solubility in Boron Containing Solutions and a Function of pH and Temperature. ADAMS Accession No. ML091610696, Argonne National Laboratory, September, 2008.

2. Bahn, C.B., Kasza, K.E., Shack, W.J., Natesan, K, and Klein, P. "Evaluation of precipitates used in strainer head loss testing: Part Ill. Long-term aluminum hydroxide precipitation tests in borated water." Nuclear Engineering and Design, vol. 241, no. 5, pp. 1914-1925, 2011.

NOC-AE-1 4003101 Attachment 2 Page 6 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 11a The conclusions contained in document CHLE-014, "T2 LBLOCA Test Report," (letter dated October 13, 2013, available in ADAMS Accession No. ML13323A673) state, in part, "Chemical products did form under the simulated STP LBLOCA conditions but primarily were adhered to the galvanized coupons." In addition, CHLE-020, "Test Results for a 10-day chemical effects test simulating LBLOCA conditions (T5)," states on page 10, "The high turbidity at the beginning of Tests T5 and T2 shown in Figure 3b might be caused by detachment of zinc particles from the zinc coupons and galvanized steel coupons due to the high temperature during the first 80 minutes of the test." Page 75 of Volume 6.2 states, "Although a zinc (Zn) product was observed to form under STP LOCA test conditions, it was not included in this analysis since the product was determined to be crystalline and mainly adhere to structures within containment as opposed to readily travel with solution." Based on international experience, Framatome ANP, Inc. report titled, "Influence of Corrosion Processes on the Protected Sump Intake after Coolant Loss Accidents," December 2006 (ADAMS Accession No. ML083510156), zinc corrosion product dislodged by falling water caused a significant increase in head loss, please discuss:

(a) If following a LOCA, water either falling from a pipe break or from other locations in the containment building could dislodge zinc corrosion product from galvanized steel surfaces that could transport to the strainer.

STP Response:

It is possible for falling water from a pipe break or other locations to impinge on the gratings or other galvanized surfaces within the containment building. Direct jet impingement from a break could cause damage to galvanized surfaces but would be for a relatively short duration and over a limited area. Continued impingement from the Containment Spray System (CSS) could occur for several additional hours until the system is secured. If galvanized surfaces are located below the break, falling water could impinge on the surfaces even after the system is secured, although the surface area impinged in this manner would be limited in extent.

Chemical Head Loss Experiment (CHLE) tank tests that contained galvanized steel or zinc surfaces experienced an initial peak in turbidity and zinc concentrations, as described in the CHLE-020 document (1). Testing conducted for another licensee after the STP license submittal demonstrated that the zinc release was associated with the initial period of low pH when Tri-sodium Phosphate (TSP) buffer was not present (2)

When TSP was present and the pH was circumneutral, the initial release of zinc did not occur. Thus, it is expected based on the current data that the period of time that zinc would be released from galvanized surfaces would be limited to the initial portion of the accident sequence before the TSP has fully dissolved into the containment solution.

NOC-AE-1 4003101 Attachment 2 Page 7 of 16

References:

1. UNM, CHLE-020: Test Results for a 10-day chemical effects test simulating LBLOCA conditions (T5), Rev. 3. University of New Mexico, Albuquerque, NM.

Feb. 22, 2014, ML14072A079.

2. UNM, CHLE-SNC-006: Bench Test Results for Series 2000 Tests for Vogtle Electric Generating Plant, Rev. 1. University of New Mexico, Albuquerque, NM.

Nov. 29 2013.

NOC-AE-1 4003101 Attachment 2 Page 8 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 11b (b) Whether chemical effects contributions from zinc should be considered as part of the STP chemical effects analysis.

STP Response:

Zinc from galvanized surfaces might be considered to contribute to head loss in two ways. First, zinc products dislodged from galvanized surfaces during the initial phases of the Loss of Coolant Accident (LOCA) (before the TSP dissolves), as observed in some of the Chemical Head Loss Experiment (CHLE) tank tests, may contribute an additional particulate source during the initial development of the debris bed. This contribution was considered to be small (less than 10 percent, based on concentrations measured in solution) compared to other sources of latent debris in the containment building and was not explicitly considered as a separate source of particulate. The second source of zinc from galvanized surfaces is the slow formation of zinc phosphate on the galvanized surface due to reactions between the zinc in the galvanized coating and the phosphate in the solution. Visual observations of the coupons in the CHLE tank tests indicated that this product formed slowly over a period of many days and remained largely adhered to the coupons. While quantitative rates of zinc phosphate formation were not obtained from the CHLE tank tests, qualitative observations indicated that the product would not be present until later in the accident sequence when temperatures were lower and strainer flow rates were lower, allowing additional margin for head loss through the strainer. Analysis indicated that the product was crystalline, which would be expected to contribute to less head loss than amorphous corrosion products. Based on the late formation of this material, its adherence to surfaces, and crystalline nature, this material was considered less significant in the STP chemical effects analysis and its formation was not explicitly considered in the analysis described in Volume 6.2. The bump-up factor used to apply chemical effects head loss was not calculated on the basis of individual chemical products, and the potential for zinc phosphate to be a contributor to chemical head loss was implicitly considered during the development of the bump-up factors.

NOC-AE-1 4003101 Attachment 2 Page 9 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 17

Page 187 of Volume 3 states "the chemical effects bump-up factor should never be less than one, and there is a practical maximum above which all events will lead to sump failure." Please discuss in more detail the approximate value of a bump up factor that will lead to sump failure. Please provide the values for conventional head loss that are assumed.

STP Response:

Every simulated break has its own time-dependent conventional head loss that is calculated based on debris accumulation and flow rate, and added to a baseline clean-strainer head loss of 0.22 feet of water. Chemical factors are applied to the conventional head loss when the temperature is less than 140+/- 5 OF and the fiber load exceeds 1/16 in. equivalent thickness. Total head loss is compared at every time step to the performance metrics of (1) NPSHAvaii, (2) void fraction, and (3) mechanical buckling.

For every break with a conventional head loss in the range of 1 ft. of water and a mechanical loading limit of only 9.35 ft., a chemical head-loss factor exceeding 10 will induce failures. A chemical head-loss factor of 43 would lead to buckling failure of the strainer for all simulated breaks in Case 01, full train operation. A chemical head-loss factor of 209 would lead to the violation of the NPSH margin criterion and failure for all simulated breaks in Case 01, full train operation.

These solutions were obtained by extracting the necessary data from the CASA Grande Case 01 simulation. An example for the large-break population of chemical factors needed to induce mechanical buckling failure is shown in Figure A. The cumulative distribution function (CDF) illustrates the percentage of LBLOCA cases that would fail for chemical factors < x .

1.00 0.90 0.80 0.70 CL 0.60 E 0.50

. 0.40 0.30 .

0.20 0.10 0.0.0. 1. 0 2 0 0 0 .

0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 Minium Bump-U~p Factor Required to Exceed Buckling limit Figure A. CDF for minimum chemical factor required to exceed strainer buckling limit for large breaks.

NOC-AE-14003101 Attachment 2 Page 10 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 20

Please discuss what benchmarking was performed with a) STP specific strainer tests and b) industry test data with similar conditions for the baseline head loss and chemical effects bump up factor.

STP Response:

Prototypical STP-specific strainer head loss tests were conducted at Alden Research Laboratory (ARL) in February (1 2) and July, 2008(3, 4),. The February head loss tests were superseded by the July head loss tests, because the February tests used walnut flour as a particulate surrogate. The reduced amount reflects most closely the amount of debris from the majority of the large breaks. The CASA Grande head loss population for Case 01 (all equipment starts and runs) was compared to all of the prototypical strainer head loss tests conducted at ARL. The maximum conventional CASA Grande head loss was 8.2 ft, which bounds the maximum tested head losses for all of the ARL tests, except Test 3 in February. Test 3 was terminated after large head losses, greater than 15 ft, were observed following the addition of fine fibrous debris(2); as stated above, this test used walnut flour as a particulate surrogate and was superseded. The maximum predicted chemical effects CASA Grande head loss was 154.9 ft, which bounds all the ARL tests. The maximum predicted total CASA Grande head loss was 161.9 ft, which bounds all the ARL tests.

Also, expected values of exponential distributions applied for chemical head-loss factors were chosen to be consistent with strainer test data showing chemical induced head-loss increases of approximately a factor of 2 Vogtle Electric Generating Plant (Vogtle) conducted prototypical strainer head loss tests at the Alion hydraulics laboratory 5( . Figure 1 displays the prototypical Vogtle and STP strainer modules. Both modules are PCI Sure-Flow@ designs.

rP (right) Prototypical

NOC-AE-1 4003101 Attachment 2 Page 11 of 16 All the STP strainer tests were conducted at an approach velocity of 0.0086 ft/s(2',4); the Vogtle strainer tests were conducted at an approach velocity of 0.0150 ft/s(5 ). Debris weights per prototypical strainer area for general debris types are displayed in Table 1.

However, the specific insulation products tested under each debris-type category differ between the two plants. For example, STP tested NUKON and Thermal Wrap under the low-density fiberglass (LDFG) category, whereas Vogtle only tested NUKON under the LDFG category.

2 Table 1: Debris Comparison between STP and Vogtle( '4, 5)

LDFG LDFG Fines Smalls Particulate per per per Sodium Strainer Strainer Strainer Calcium Aluminum Aluminum Area, Area, Area, Phosphate, Oxyhydroxide, Silicate, Utility Test Ibm/ft 2 Ibm/ft 2 Ibm/ft 2 Ibm/ft 2 Ibm/ft 2 Ibm/ft 2 STP Feb.2008 0.22 0.30 1.16 0.10 0.43 0 Test 4 STP Feb.2008 0.14 0.19 1.16 0.10 0.43 0 Test 5 STP July.2008 0.06 0.11 0.64 0.10 0.45 0 Test 2 Vogtle All Tests 0.31 0.14 6.60 0.09 0 0.14 The maximum conventional, chemical effects, and total head losses predicted by CASA Grande for Case 01 also bound Vogtle's prototypical strainer head loss testing, which measured 5.5 ft, 6.3 ft, and 11.8 ft, respectively(5 ).

For the head loss comparisons cited above, the STP tests at ARL and Vogtle tests at Alion were not corrected to a common flow rate and temperature, which is conservative.

The temperature range for these tests was 51'F to 11 7OF(1- 5 ) and the temperature range for CASA Grande is 117'F to 255°F (LAR Enclosure 4-3, Table 2.2.13) correcting the tests to a higher temperature would reduce the head loss. The ARL tests modeled the maximum STP flow condition. Case 01 of CASA Grande was simulated at the maximum STP flow condition, but containment spray pumps were secured during the LOCA.

Therefore, the CASA Grande head loss population may inherently include head losses at flow rates lower than the ARL test condition. The Vogtle tests were conducted at a higher flow rate than the STP tests. Therefore, correcting the Vogtle tests to a lower flow rate would reduce the head loss. All factors considered, the benchmark comparisons of maximal computed head loss meet or exceed all applicable test data for STP and Vogtle.

NOC-AE-1 4003101 Attachment 2 Page 12 of 16

References:

1. 0415-0100067WN / 0415-0200067WN. "South Texas Project Test Plan Feb 2008". Revision A. 11/24/2008.

2. 0415-0100069WN / 0415-0200069WN. "South Texas Project Test Report for ECCS Strainer Performance Testing Feb 2008". Revision A. 11/24/2008.

3. 0415-010007OWN / 0415-0200070WN. "South Texas Project Test Plan."

Revision A. 8/14/2008.

4. 0415-0100071WN / 0415-0200071WN. "South Texas Project Test Report for ECCS Strainer Testing July 2008". Revision A. 11/24/2008.

5. ALION-CAL-SNC-7410-005. "Head Loss Testing of a Prototypical Vogtle 1 and 2 Strainer Assembly". Revision 0. 12/31/2009.

NOC-AE-1 4003101 Attachment 2 Page 13 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 22a A total of five CHLE tank tests were performed to evaluate STP plant-specific chemical effects tests. CHLE Tests 1 and 2 were intended to evaluate an MBLOCA and an LBLOCA, respectively. Please address the following questions related to Tests 1 and 2:

(a) Please discuss why the test screen debris bed is an acceptable method for detection of chemical precipitates given the earlier test "CHLE-010, CHLE Tank Test Results for Blended and NEI [Nuclear Energy Institute] Fiber Beds with Aluminum Addition," that showed no head loss response even in the presence of large quantities of aluminum oxyhydroxide precipitate generated according to the WCAP-16530-NP-A protocol.

Note: Additional details are available a September 6, 2012, meeting summary dated October 4, 2012 (ADAMS Accession No. ML12270A055).

STP Response:

The tests described in "CHLE-010, CHLE Tank Test Results for Blended and NEI Fiber Beds with Aluminum Addition"1 did not contain aluminum oxyhydroxide precipitate generated according to the WCAP-16530-NP-A protocol. In the CHLE-010 test series, aluminum oxyhydroxide precipitate was generated by injecting an aluminum nitrate solution directly into the tank recirculation line. The aluminum nitrate solution was injected in periodic batches at a slow rate corresponding to an increase in aluminum concentration in the tank of 0.02 mg/L per minute. Precipitation occurred when the aluminum nitrate came in contact with the solution in the tank, which contained plant-specific concentrations of boric acid, lithium hydroxide, and tri-sodium phosphate (TSP) and was heated to about 45 0C (113 OF) at the time of aluminum nitrate injection. In contrast, the WCAP-16530-NP-A protocol involves a more rapid addition of solid aluminum nitrate and sodium hydroxide into normal potable water in a mixing tank at ambient temperature with a target aluminum oxyhydroxide concentration between 2,100 and 11,000 mg/L. The CHLE-010 tests were designed to simulate the slow release of aluminum during corrosion and the conditions for precipitate formation were substantially different from the WCAP-16530-NP-A protocol.

Earlier tests, described in "CHLE-008, Debris Bed Preparation and Formation Test Results"'2 and provided to the NRC Staff in STP Letter NOC-AE-14003075, dated February 27, 2014 (ML14072A076), did include aluminum oxyhydroxide precipitate generated according to the WCAP-16530-NP-A protocol. In those tests, addition of the WCAP precipitates to columns with the same type of debris bed as the MBLOCA and LBLOCA tests (the NEI-prepared debris bed) did result in substantial head loss. In CHLE-008 Test 5, the addition of WCAP precipitates to an NEI-prepared debris bed at an approach velocity of 0.093 ft/s caused such a rapid increase of head loss that the stainless steel support screen collapsed before the entire batch of precipitates was added. In CHLE-008 Tests 6 and 8, addition of WCAP precipitate to the column resulted in substantial head loss through the NEI-prepared debris bed at an approach velocity of 0.01 ft/s, which is comparable to that of the STP strainers. The quantity of aluminum oxyhydroxide precipitate that caused significant head loss corresponded to a screen loading of 246 g/m 2 . For comparison, the strainer testing done for STP at Alden Research Laboratory by AREVA 3 had a final aluminum oxyhyroxide precipitate screen loading of 2,200 g/m 2. Test 13 in CHLE-008 involved the addition of WCAP precipitates to the CHLE tank and excessive head loss was detected in all three columns. The

NOC-AE-1 4003101 Attachment 2 Page 14 of 16 CHLE-008 test results demonstrate that the debris beds used in the MBLOCA and LBLOCA tests were capable of detecting aluminum oxyhydroxide precipitates generated according to the WCAP-16530-NP-A protocol via a head loss measurement.

The discussion between the NRC and STP during the September 2012 conference meeting mentioned in the RAI focused on the relative degree of sensitivity between the NEI-prepared debris beds and the blender-prepared debris beds. The CHLE-008 and CHLE-010 test results demonstrated that the threshold loading rate for the detection of head loss in blender-prepared debris beds was lower than in the NEI-prepared debris beds. Unfortunately, the blender-prepared debris beds experienced significant head loss when chemical precipitates were not present and exhibited other forms of instability such as a non-linear response to fluid velocity or quantity of fiber, making them unsuitable for detection of chemical precipitates in the CHLE tests. In addition, a comparison of the loading rates at which head loss first occurred in the NEI-prepared debris beds in the vertical column and the mixed-debris bed in the AREVA strainer testing indicates that the threshold for detecting head loss is not as low in the NEI-prepared debris bed as it is in the mixed-debris bed. Thus, neither the NEI-prepared nor the blender-prepared debris beds provided a sufficient method of detecting chemical precipitates via a head loss measurement.

It is also important to note that the MBLOCA and LBLOCA tests employed multiple parameters to detect the presence of precipitates in addition to head loss through the debris beds. Samples were periodically analyzed for total and dissolved aluminum periodically during both tests. No significant difference between total and dissolved aluminum was detected, indicating that all aluminum was in a dissolved form. The measured total concentrations were below the saturation concentration predicted for amorphous aluminum hydroxide by Visual MINTEQ, corroborating the evidence from the total and dissolved measurements. Turbidity remained low throughout the MBLOCA and LBLOCA tests. Results reported in CHLE-010 demonstrated a linear response between the addition of aluminum and the turbidity of the solution, indicating that turbidity is capable of detecting precipitates that form in this experimental system. Thus, even in the absence of the column head loss data, the results from the MBLOCA and LBLOCA tests can be used to demonstrate that aluminum chemical precipitates do not form in the simulated MBLOCA and LBLOCA environments.

To account for the uncertainty associated with the head loss characteristics of various debris beds and the various ways of generating chemical precipitates, safety margin was added to the chemical effects contribution to head loss by applying a bump-up factor to the calculated value of conventional head loss as described in the LAR Enclosure 4-1, page 9 and in more detail in LAR Enclosure 4-3, Section 5.6.3. The chemical head loss bump-up factor did not directly use head loss data from the CHLE tests. As a result, the ability of the test screen debris bed to detect chemical precipitates has shown not to influence the results described in the STP license submittal.

NOC-AE-1 4003101 Attachment 2 Page 15 of 16

References:

1. University of New Mexico, CHLE-010: CHLE Tank Test Results for Blended and NEI Fiber Beds With Aluminum Addition, Rev. 3. Feb. 10, 2014. (ML14072A083)

2. University of New Mexico, CHLE-008: Debris Bed Preparationand Formation Test Results, Rev. 4. Feb. 3, 2014. (ML14072A082)

3. AREVA, South Texas-Project Test Report for ECCS Strainer Testing, Doc. 66-9088089-000.

NOC-AE-1 4003101 Attachment 2 Page 16 of 16 ESGB, Steam Generator Tube Integrity and Chemical Engineering - Chemical Effects:

RAI 22b (b) Please describe why the use of only aluminum and fiberglass in the MBLOCA test adequately represents the plant specific environment.

STP Response:

The objective of the CHLE testing program was to generate experimental data to support an overall risk-informed approach to the resolution of GSI-191, while also conducting a manageable number of tests. Each test within the program had multiple objectives, with the intent that the testing program as a whole provided data to support the resolution.

Inclusion of all materials in all tests would not necessarily have provided the most comprehensive data, since in some cases the presence of one material might reduce the contribution of chemical effects from another material. In the case of the MBLOCA and LBLOCA tests, the inclusion of zinc in the LBLOCA test and not in the MBLOCA test, coupled with comparisons of predicted release rates from the WCAP equations, demonstrated that the release of aluminum was greater when zinc was not present.

That important outcome would not have been recognized if zinc had been included in all tests.

While the tests included different materials and aspects of the LOCA to satisfy different objectives, the conditions within each test were representative of the plant-specific environment for the included materials. Many factors, including the quantities of boric acid, tri-sodium phosphate (TSP), and lithium hydroxide; the timing of TSP dissolution, acid generation, and spray duration; the temperature profile; and approach velocity through the screens were all representative of the plant-specific environment. For the MBLOCA test, the quantities of aluminum and fiberglass were also representative of the plant-specific environment during a MBLOCA.

NOC-AE-1 4003101 Attachment 3 Attachment 3 Response to SCVB Request for Additional Information:

RAI 1,2,3,4,5,6,7,8,9

NOC-AE-1 4003101 Attachment 3 Page 1 of 21 SCVB, Containment and Ventilation Branch: RAI la In support of Enclosure 2-3, "Request for Exemption from Certain Requirements of General Design Criterion [GDC] 38," please provide the following:

(a) Please list the specific STP plant systems that will not meet the requirements of GDC-38.

STP Response:

Section 1 of Enclosure 2-3 of the license amendment request (LAR) identifies the Containment Spray System (CSS) as the only system for which the proposed exemption to GDC-38 would apply. The CSS is the only system credited for meeting GDC-38 that takes suction from the containment sumps during the recirculation mode of accident mitigation and is consequently subject to debris effects, which are the focus of the risk-assessment provided in the STP licensing application.

NOC-AE-1 4003101 Attachment 3 Page 2 of 21 SCVB, Containment and Ventilation Branch: RAI lb (b) Please describe the specific requirements of GDC-38 that will not be met by each of the plant systems listed in response to item (a) above.

STP Response:

The second paragraph of GDC-38 prescribes the deterministic approach to the analysis that assures compliance with the criterion. The proposed exemption to GDC-38 utilizes the application of a risk-informed approach per Regulatory Guide 1.174 for the assessment of the effects of debris. Debris effects could affect all three emergency sumps and consequently all three trains of the Containment Spray System (CSS). In accordance with the definition of single failure in 10CFR50, Appendix A, this would be considered a single occurrence that could result in the failure of multiple components.

The risk-informed assessment demonstrates the change in core damage or large early release frequency from the debris effects is very low in accordance with the guidance of RG 1.174

NOC-AE-1 4003101 Attachment 3 Page 3 of 21 SCVB, Containment and Ventilation Branch: RAI 2a In support of Enclosure 2-4, "Request for Exemption from Certain Requirements of General Design Criterion 41," please provide the following:

(a) Please list the specific STP plant systems that will not meet the requirements of GDC-41.

STP Response:

Section 1 of Enclosure 2-4 of the license amendment request (LAR) identifies the Containment Spray System (CSS) as the only system for which the proposed exemption to GDC-41 would apply. The CSS is the only system credited for meeting GDC-41 that takes suction from the containment sumps during the recirculation mode of accident mitigation and is consequently subject to debris effects, which are the focus of the risk-assessment provided in the STP licensing application.

NOC-AE-1 4003101 Attachment 3 Page 4 of 21 SCVB, Containment and Ventilation Branch: RAI 2b (b) Please describe the specific requirements of GDC-41 that will not be met by each of the plant systems listed in response to item (a) above.

STP Response:

The second paragraph of GDC-41 prescribes the deterministic approach to the analysis that assures compliance with the criterion. The proposed exemption to GDC-41 utilizes the application of a risk-informed approach per Regulatory Guide 1.174 for the assessment of the effects of debris. Debris effects could affect all three emergency sumps and consequently all three trains of the Containment Spray System (CSS). In accordance with the definition of single failure in 10CFR50, Appendix A, this would be considered a single occurrence that could result in the failure of multiple components.

The risk-informed assessment demonstrates the change in core damage or large early release frequency from the debris effects is very low in accordance with the guidance of RG 1.174.

NOC-AE-1 4003101 Attachment 3 Page 5 of 21 SCVB, Containment and Ventilation Branch: RAI 3a , page 4, paragraph "Use of a Risk-Informed Approach to Resolving GSI-191",

states:

The design and licensing basis descriptions of accidents requiring ECCS operation, Including analysis methods, assumptions, and results provided in UFSAR [Updated Final Safety Analysis Report]

Chapters 6 and 15 remain unchanged. This is based on the functionality of the ECCS and CSS during design basis accidents being confirmed by demonstrating that the calculated risk associated with GSI-191 for STP Units 1 and 2 is "Very Small" and less than the Region III acceptance guidelines defined by RG 1.174.

The current licensing basis containment analysis methodology used to confirm the adequacy of the containment heat removal system (which complies with 10 CFR 50 Appendix A GDC-38) described in the UFSAR is different from the proposed methodology which resolves GSI-191 on a risk-informed basis and proposes an exemption request from GDC-38. For example: (a) difference in the single failure assumption in the proposed and current analysis; (b) computer codes RELAP for LOCA mass and energy (M&E) release, and MELCOR for the LOCA sump temperature response are used in the proposed analysis, and SATAN-VI, WREFLOOD, FROTH are used for M&E release and CONTEMPT4/MOD5 is used for sump temperature response in the current analysis; and (c) the proposed analysis inputs and assumptions are required to be conservative from GSI-191 perspective and also required to be conservative for sump temperature response whereas the current analysis inputs and assumptions are conservative for sump temperature response, (a) Please justify why the UFSAR licensing basis description of the methodology used for confirming the adequacy of containment heat removal system which complies with GDC-38 should not be replaced with the proposed licensing basis methodology which takes an exemption from GDC-38.

STP Response:

The licensing basis for the assessment of the effects of debris is being revised and the description of the risk assessment will be described in the STP UFSAR as discussed in the license amendment request, Enclosure 3, Attachment 2. The results of the risk-informed assessment demonstrate that the containment sumps are sufficiently reliable in support of the Containment Spray System (CSS) such that the function of the CSS with respect to containment analysis remains as currently described in the UFSAR.

NOC-AE-1 4003101 Attachment 3 Page 6 of 21 SCVB, Containment and Ventilation Branch: RAI 3b (b) Tabulate the differences between the inputs and assumptions between the current licensing basis containment analysis that calculates the most limiting sump fluid temperature profile for available NPSH calculation and the proposed containment analysis performed for risk-informed GSI-191. Please justify that the inputs and assumptions in the proposed methodology are conservative from both GSI-191 and sump temperature response perspectives.

STP Response Background and Reference to Submittal Documentation The proposed methodology is intended to be realistic, so it is not necessarily conservative with respect to inputs and assumptions. The proposed methodology is not proposed to replace the conservatively-based methodology described in the UFSAR.

More details are provided in the following paragraphs and tables.

As described in the LAR, the proposed exemptions from General Design Criteria (GDC)-

35, the "Emergency Core Cooling", GDC-38, "Containment Heat Removal", and GDC-41, "Containment Atmosphere Cleanup" are for approval of a risk-informed approach for addressing GSI-191 and responding to Generic Letter (GL) 2004-02 for STP Units 1 and 2 as the pilot plants for other licensees pursuing a similar approach. As further described, STPNOC seeks NRC approval based on a determination that the risk-informed approach and the risk associated with the postulated failure mechanisms due to GSI-191 concerns meets the guidance, key principles for risk-informed decision-making, and the acceptance guidelines in RG 1.174.

STP is not proposing to apply the risk-informed approach to revise the licensing basis for containment design described in the UFSAR. The proposed risk assessment evaluates a spectrum of Loss of Coolant Accident (LOCA) scenarios to quantify the amount of debris of various types that might be generated and transported to the emergency sumps, and how that debris might affect available NPSH for Emergency Core Cooling System (ECCS) and Containment Spray System (CSS) pumps taking suction from the sumps in the recirculation mode. It also evaluates potential transport of debris to the reactor core.

It calculates failure probabilities that are fed to the STP PRA.

Because the LAR uses a risk analysis, the engineering inputs in several areas are different from the methods used in the existing deterministically-based Licensing Basis (LB) analyses. The containment analysis is an example of such an area of difference.

The risk-informed approach to resolving GSI-191 applies the Probabilistic Risk Assessment (PRA) model to quantify the risk associated with GSI-191 concerns by calculating the difference in risk for two cases:

• The actual plant configuration for STP Units 1 and 2, with failures due to GSI-191 concerns, and

" The same plant configuration for STP Units 1 and 2, except for the assumption that there are no failures due to GSI-1 91 concerns.

NOC-AE-1 4003101 Attachment 3 Page 7 of 21 to the LAR provides the generic methodology for the proposed risk-informed approach to resolving GSI-191, consistent with RG 1.174 guidance. This enclosure describes the required inputs to the PRA model, the basic structure for appropriately modeling the inputs, and performance criteria used to calculate the risk. As described in , the risk-informed approach to resolving GSI-191 uses the plant-specific PRA with realistic modeling to quantify the residual risk associated with GSI-191 and to evaluate for acceptable sump design in support of successful ECCS and CSS operation in recirculation mode following postulated LOCAs with the debris effects discussed in GSI-1 91.

The Current LB Modeling Approach The licensing basis (LB) containment analysis is based on the CONTEMPT computer code and is documented in STP calculation NC07032. The CONTEMPT code is documented in two separate documents, NUREG/CR 3716 and NUREG/CR 4001.

The current LB containment analysis uses 260 OF for the sump fluid temperature and does not address strainer failure due to the concerns raised in GSI-191. The analysis is designed to maximize containment pressure (and temperature), which would actually improve net positive suction head available. The condition assumed in the LB analysis is very unlikely to be realized in operation. Because the LB CONTEMPT methodology is not intended to reflect realistic containment response behavior, and is based on an extremely unlikely scenario, the LB modeling approach differs from the approach used for compliance with RG 1.174 requirements.

In particular, CONTEMPT4 has been verified to perform two major analyses [4, Page LOCA U-1 1]:

• containment peak pressure and temperature analysis, and

• containment environmental thermodynamic conditions for equipment qualification and isolated pipe pressurization purposes.

The Risk Informed Modeling Approach The containment analysis is based on the STP MELCOR model that runs simultaneously with the STP RELAP5 model designed for use in the STP PRA for the evaluation required in the risk-informed approach for addressing the GSI-1 91 issue. As described in LAR Enclosure 5, Item 5.a.13: In-Vessel Fiber Limits, several parameters related to geometry, thermal hydraulics/heat transfer, and engineered safety features used in the MELCOR input were taken from a previously certified Modular Accident Analysis Program (MAAP) STP containment model. Hence, the certification document for the MAAP model is appropriately referenced throughout the text.

The following table lists the major qualitative differences by modeling subject area between the risk assessment and LB containment models. In the next section, numerical values are compared.

NOC-AE-1 4003101 Attachment 3 Page 8 of 21 Subject RELAP/MELCOR CONTEMPT Difference Subcompartment Yes No STP CONTEMPT Model has 1 large analysis volume (1 pool, 1 atmosphere) and is not validated for subcompartment analysis. MELCOR has several sub-compartments.

Modeling goals Containment pressure Peak Pressure analysis Fundamentally different modeling response Sump Pool (structural design goals drove the modeling decisions temperature testing, leak rate testing) for each code and explain some of response and containment the differences in assumptions for thermodynamic each model.

conditions Model variations Single containment 2 separate models: one The single MELCOR model works in model regardless of for each transient stage concert with RELAPS-3D for all primary side (injection, recirculation). transient stages. Primary side characteristics, break Also, different characteristics are the exclusive size, or stage of the assumptions for domain of RELAP5-3D during a transient (e.g. different steam coupled run. The CONTEMPT model before/after sump generator types, primary may assume several different forms recirculation) side characteristics, depending on transient stage, modeling goals, etc. primary side characteristics, etc.

Modeling Best-estimate Conservative ESF delays, free volume calculations, philosophy other model characteristics are best-estimate for the MELCOR model but are generally conservative for the CONTEMPT model Code execution Once-through, Can be an iterative There is no need for a collection or coupled run from process consisting of succession of runs for start of the transient initial runs, sensitivity MELCOR/RELAP. Under certain to conclusion runs, confirmatory runs, circumstances with CONTEMPT, the etc. user must perform several runs to ascertain set-points, switchover times, etc.

Code elements Control volumes, flow Control volumes, heat STP CONTEMPT model has no flow paths, heat sinks, engineered safety paths since it is a single-volume structures, features, control logic model. Both models have control engineered safety volumes, heat sinks/structures, and features, control logic engineered safety features. Details of the code elements differ.

Engineered Safety Fan coolers, sprays Fan coolers, sprays Both codes model fan coolers and Features sprays with some correlation or physics model (not identical ones).

Actuation set-points are based on differing sets of assumptions. Again, best-estimate scenarios are used for MELCOR and conservative scenarios are used for CONTEMPT.

NOC-AE-1 4003101 Attachment 3 Page 9 of 21 Subject RELAP/MELCOR CONTEMPT Difference Containment heat Neglect heat loss Include heat loss to Because CONTEMPT is not coupled in removal through containment environment, account real problem time to another code walls, no CSS or RHR for RHR heat exchangers that models the RHR heat exchangers heat exchanger and pumps in and details of LHSI/HHSI, these modeling (handled in CONTEMPT elements must be modeled. RELAP5-RELAP5-3D) 3D handles these aspects of the calculation in the MELCOR/RELAP5-3D coupled run.

Heat sinks/ MELCOR built-in Uchida and/or Tagami Condensation heat transfer is treated condensation correlations for both correlations used. Steel differently and heat sinks have atmosphere and pool liners included on different characteristics. Liquid pool heat transfer containment walls with heat transfer is calculated internally by coefficient air gap between liner MELCOR but assumed in the calculations. Concrete and concrete. Constant CONTEMPT model.

containment walls are heat transfer coefficient modeled without the with pool.

steel liner.

Sump pool No decay heat added. Only 1 large pool for The large, lumped pool of CONTEMPT treatment Mass and energy whole containment, vs. the smaller, annular sub-subtracted from the perhaps not intended to compartment'pool of MELCOR pool based on capture the true RELAP5-3D behavior of the sump instructions pool. Decay energy added directly to pool in 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

Pipe break Communicated from Mass and energy STP CONTEMPT model uses mass and mass/energy source RELAP5-3D via release methodology energy release directly into the coupling interface as described in WCAP Containment. RELAP/MELCOR uses problem time 10325-P-A up to 3600 coupled RCS and Containment progresses. The response model.

source is split by MELCOR into part recirculation liquid water, part methodology decay steam, and part "fog" heat model from 3600 seconds to 106 seconds.

NOC-AE-1 4003101 Attachment 3 Page 10 of 21 Summary Comparison of Main Parameter Values The main numerical input parameters controlling the initial conditions and boundary conditions for timing and actuation, etc., between RELAP5/MELCOR and CONTEMPT analyses are summarized in the following table.

CONTEMPT I Value [ ELAP5=

13D/MELCOR Value Single compartment Sum over all free volume 3.3E+6 ft 3 compartments 3329332.0 ft 3 Initial containment Initial atmosphere 0 temperature 114 F temperature 119.93 F Initial containment 14.5 psia (max T) or 15.1 Initial containment 0

pressure psia (max P) pressure 14.94 psia Initial relative humidity, partial Initial relative humidity 20 %

pressure of water vapor Initial RWST Initial RWST temperature 130 F temperature 85 F 12.0 psig (9.5 psig setpoint Spray pressure Spray setpoint (HI-3) + 2.5 psig uncertainty) setpoint 9.5 psig 4.E Spray actuation times Depends on time of HI-3. 15 s delay after setpoint, 85.0 s linear ramp to full flow 4-U-

Fan cooler setpoint 5.5 psig (3.0 psig setpoint + Fan cooler pressure 3.0 psig (HI-i) 2.5 psig uncertainty) setpoint (HI-i) 3.0_psig

.4-Fan cooler actuation Fan ti cooler actuation 03 45 s oes a15 s delay after setpoint times times

'a-Concrete - 0.8 BTU/hr-ft-F Concrete - 0.54 BTU/ft-hr-F Thermal conductivity Stainless Steel - 9.4 BTU/hr- Stainless Steel - f(T), varies ft-F Concrete - 0.208 BTU/Ibm-F Concrete - 0.20 BTU/Ibm-F Specific heat capacity Stainless Steel - 0.111 Specific heat capacity Stainless Steel - f(T), varies BTU/Ibm-F 3

Concrete - 144 Ibm/ft3 3 Concrete - 144 Ibm/ft 3 Density Stainless Steel - 488 Ibm/ft Density Stainless Steel - 495 Ibm/ft

NOC-AE-1 4003101 Attachment 3 Page 11 of 21

References:

1. STP Calculation NC07032, "Containment LOCA Pressure/Temperature Analysis",

STI 33686837.

2. NUREG/CR 3716, "CONTEMPT4/MOD4 - A Multicompartment Containment System Analysis Program"

3. NUREG/CR 4001, "An Improvement to CONTEMPT/MOD4 Multicompartment Containment System Analysis Program for Ice Containment Analysis

4. STP Guide "LOCA P/T ANALYSIS USER GUIDE", Revision 4, ST131870431

5. TAMU-GSI-002, "MELCOR Input Deck Certification: South Texas Project Large Dry Containment, STI 33647084

6. WCAP 10325-P-A "Westinghouse LOCA Mass and Energy Release Data for Containment Design March 1979 Version" dated May 1983.

NOC-AE-1 4003101 Attachment 3 Page 12 of 21 SCVB, Containment and Ventilation Branch: RAI 3c (c) In case the UFSAR licensing basis description of the containment heat removal system, including its related mass and energy release analysis methodology, is required to be replaced, please provide the revised UFSAR input for NRC staff review and approval.

STP Response:

STP proposes to supplement the existing UFSAR description with a description of the risk assessment of debris effects described in Enclosure 3, Attachment 2 of Reference 1 to the cover letter.

NOC-AE-1 4003101 Attachment 3 Page 13 of 21 SCVB, Containment and Ventilation Branch: RAI 4a The current licensing basis methodology for the iodine removal is documented in UFSAR Section 6.5.2, "Containment Spray System - Iodine Removal." The iodine removal is accomplished by the CSS which meets the requirements of 10 CFR 50 Appendix A GDC-

41. The proposed risk-informed GSI-191 methodology takes exemption from compliance with GDC-41 requirements.

(a) Please justify why the UFSAR licensing basis description of the iodine removal should not be revised with the proposed methodology which takes exemption from GDC-41.

STP Response:

The current licensing basis methodology for the iodine removal, as documented in UFSAR Section 6.5.2, is not being modified or replaced.

The licensing basis for the assessment of the effects of debris is being revised and the description of the risk assessment will be described in the STP UFSAR as discussed in the license amendment request, Enclosure 3, Attachment 2. The results of the risk-informed assessment demonstrate that the containment sumps are sufficiently reliable in support of the Containment Spray System (CSS) such that the function of the CSS remains as currently described in the UFSAR with respect to dose assessment.

NOC-AE-1 4003101 Attachment 3 Page 14 of 21 SCVB, Containment and Ventilation Branch: RAI 4b (b) Please tabulate the differences between the inputs and assumptions between the current licensing basis containment atmosphere cleanup method and the proposed containment atmosphere cleanup which takes exemption from the GDC-41 requirements.

STP Response The current licensing basis containment atmosphere cleanup method does not specifically address the effects of debris on the Containment Spray System (CSS). From the standpoint of GDC-41 and CSS, the parameter of interest is available NPSH in the recirculation mode. Other than the evaluation of the debris effects on CSS, the risk assessment does not evaluate containment atmosphere cleanup. The risk assessment shows that the probability of debris affecting available NPSH for CSS such that the CSS will not perform its function is very small in accordance with the RG 1.174 acceptance criteria.

NOC-AE-1 4003101 Attachment 3 Page 15 of 21 SCVB, Containment and Ventilation Branch: RAI 4c (c) In case the UFSAR licensing basis description of the iodine removal system is required to be replaced, please provide the revised UFSAR input for NRC staff review and approval.

STP Response:

STP proposes to supplement the existing UFSAR description with the risk assessment of debris effects described in the license amendment request, Enclosure 3, Attachment 2.

NOC-AE-1 4003101 Attachment 3 Page 16 of 21 SCVB, Containment and Ventilation Branch: RAI 5 In support of Volume 6.2, please list the differences between the heat sinks in the current licensing basis containment analysis documented in the UFSAR Tables 6.2.1.1-7 and 6.2.1.1-8 and in the proposed containment analysis for risk-informed GSI-191. Please provide justification in cases where the conservatism is reduced in the proposed analysis.

STP Response As discussed in the response to SCVB RAI 3.b, above, STP is not proposing to apply the RG 1.74 risk-informed approach to revise the licensing basis for containment design described in the UFSAR. The containment pressures and temperatures calculated in the risk-informed analysis depend on the specific cases evaluated and are time-dependent; however, the values that correspond to the current UFSAR design basis conditions are comparable to the current design and licensing basis results. The results of the analysis show that the probability that debris will prevent the Emergency Core Cooling System (ECCS) and Containment Spray System (CSS) from performing their required function is very small in accordance with the criteria of RG 1.174 and those systems are considered able to perform their functions as described in the UFSAR. There is no change in their design basis with respect to containment design.

NOC-AE-1 4003101 Attachment 3 Page 17 of 21 SCVB, Containment and Ventilation Branch: RAI 6 In support of Volume 6.2, please list the differences between the LOCA surface heat transfer model for heat sinks in the current licensing basis analysis documented in UFSAR Table 6.2.1.1-9 and the model in the proposed containment analysis for risk-informed GSI-191. Provide justification for the differences in case the conservatism is reduced in the proposed analysis.

STP Response See the response to SCVB-RAI 5, above.

NOC-AE-1 4003101 Attachment 3 Page 18 of 21 SCVB, Containment and Ventilation Branch: RAI 7 NUREG-0800, "Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants: LWR Edition" (SRP), Section 6.2.1.5, "Minimum Containment Pressure Analysis for Emergency Core Cooling System Performance Capability Studies,"

describes the minimum containment pressure analysis for ECCS performance capability.

RG 1.157, "Best Estimate Calculation for Emergency Core Cooling System Performance,"

May 1989 (ADAMS Accession No. ML003739584), Section 3.12.1, "Containment Pressure," provides guidance for calculating the containment pressure response used for evaluating cooling effectiveness during the post-blow-down phase of a LOCA.

UFSAR Section 6.2.1.5 documents the current minimum containment pressure analysis for performance capability studies of the ECCS. Please describe the proposed containment analysis, including assumptions and inputs, performed for the calculation of minimum containment pressure input for the ECCS analysis that calculates the peak cladding temperature for risk-informed GSI-191. Please justify that the inputs and assumptions are conservative for the purpose.

STP Response The risk assessment is limited to evaluating the effect of debris on Emergency Core Cooling System (ECCS) and Containment Spray System (CSS) in the recirculation mode. As discussed in the responses above, it is not proposed to replace the current design basis containment analyses. It is not proposed as a change to the ECCS evaluation model and STP does not propose to apply it to show that 10CFR50.46(b)(1) limits are met for peak cladding temperature. The results of the analysis show that the probability that debris will prevent the ECCS from performing its required function is very small in accordance with the criteria of RG 1.174 and the system is considered able to perform its function as described in the UFSAR. There is no change in the ECCS design basis with respect to containment pressure.

NOC-AE-1 4003101 Attachment 3 Page 19 of 21 SCVB, Containment and Ventilation Branch: RAI 8 Volume 6.2, page 117, Item 5.a.14, "In-Vessel Thermal Hydraulic Analysis," lists six scenarios simulated using the 3D Vessel-ID Core Model. Please describe and justify the basis for selection of these LOCA breaks scenarios.

STP Response Early in development of the risk-informed approach for GSI-191 investigation, extreme core blockage scenarios investigations for core and Reactor Coolant System (RCS) response were unavailable in the academic literature. STP therefore undertook basic re-search to understand such core and RCS responses in theoretically extreme scenarios.

The main idea behind these simulations was to investigate and understand, assuming that flow blockage could occur, which extreme theoretical scenarios would go to success and which would lead to failure. Results of the studies performed have since been published in peer-reviewed literature.

Scenarios were developed assuming instantaneous blockage at the time of recirculation switchover for hot and cold leg break locations and size (small, medium, and large) to account for different flow patterns and RCS response as described in the LAR. The break sizes were chosen at high values for small, medium, and large STP LOCA categories. All the cases assumed that one of the Emergency Core Cooling System (ECCS) trains is in the broken leg (STP Loop B) thereby minimizing effective injection flow to the core. This matrix results in six scenarios.

The primary objective was to study the core and RCS response for the main break locations and sizes to gain understanding of the severity of such responses under extreme conditions of blockage. Because flow from the sump would take a finite amount of time to carry any debris to the core, it is clear that such instantaneous blockage could only be realized in theory. Blockage, should it actually occur, would require some amount of time to build up. The scenarios selected therefore represent outcomes for theoretical extremes. The results are useful for success criteria in the Probabilistic Risk Assessment (PRA) and for safety margin asked for in RG 1.174. That is, by investigating these extreme theoretical scenarios, STP could see that the vast majority of hypothesized LOCA scenarios would go to success. STP investigated additional marginal cases as further described in the LAR Enclosure 5 (Volume 6.2, page 117, Item 5.a.14), where either there was flow through the barrel-to-baffle bypass region (without credit for LOCA holes in the baffles) or through a small opening (one fuel channel, either peripheral or center) in the core.

NOC-AE-1 4003101 Attachment 3 Page 20 of 21 SCVB, Containment and Ventilation Branch: RAI 9a In support of Enclosure 2-2, "Request for Exemption from Certain Requirements of General Design Criterion 35," please provide the following:

(a) Please list the specific STP plant systems that will not meet the requirements of GDC-35.

STP Response:

The license amendment request, Enclosure 2-2, Section 1 identifies the Emergency Core Cooling System (ECCS) as the only system for which the proposed exemption to GDC 35 would apply. The ECCS is the only system credited for meeting GDC-35 that takes suction from the containment sumps during the recirculation mode of accident mitigation and is consequently subject to debris effects, which are the focus of the risk assessment provided in the STP licensing application. The specific ECCS subsystems that are affected are Low Head Safety Injection and High Head Safety Injection.

NOC-AE-1 4003101 Attachment 3 Page 21 of 21 SCVB, Containment and Ventilation Branch: RAI 9b (b) Please describe the specific requirements of GDC-35 that will not be met by each of the plant systems listed in response to item (a) above.

STP Response:

The second paragraph of GDC-35 prescribes the deterministic approach to the analysis that assures compliance with the criterion. The proposed exemption to GDC-35 utilizes the application of a risk-informed approach per Regulatory Guide 1.174 for the assessment of the effects of debris. Debris effects could affect all three emergency sumps and consequently all three trains of the Containment Spray System (CSS). In accordance with the definition of single failure in 10CFR50, Appendix A, this would be considered a single occurrence that could result in the failure of multiple components.

The risk assessment demonstrates the change in core damage or large early release frequency from the debris effects is very low in accordance with the guidance of RG 1.174.

NOC-AE-1 4003101 Attachment 4 Attachment 4 Response to SNPB Request for Additional Information:

RAI 1,2,3,5

NOC-AE-1 4003101 Attachment 4 Page 1 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI la Please provide the following information for the STP Nuclear Steam Supply Systems (NSSSs):

(a) Volume of the lower plenum, core and upper plenum below the bottom elevation of the hot leg, each identified separately. Also, please provide the heights of these regions and the hot-leg diameter.

STP Response:

The volume of the lower plenum is 638.7 ft 3 .

The volume of the core is 715.1 ft 3.

3 The volume of the upper plenum below the bottom elevation of the hot leg is 520.51 ft Diagrams displaying these volumes and their respective elevations are shown below.

a. P.ottom elevation of sucllon leg iciossover leg*]- 31 -ID D0wg 14%6-01 4Cr2J-O2O42-( WN. typ- 8 pkl) b, Top eft.vation of cold !e.

<.Top e *ivotion MI or iheight 0 core)

d. Rooettore1va1oiknof down~crowr Hot W9gd~afflete?

Top Elevation of Cold Leg b_ 33'4.75 C(OP03-ZG '.,owci 121&54. Sh 1CFj

---

- - tRCS leg, 23.73' (Top of rcui (wri pipo! ID) -r C.UpINoplenvm rj suclioneg - 22'523l25'O-m2I&814, Si, 91, Suction Icdgirneler WIDI 1x~go hot Nog (42S BOorIo ElevTiono 5wCtiorn Leg a.

U;%vtOfreplate 2.11O7TI~weq 6117E0.0i Tcpfuel Nets 1V51VSTPPDIIUI Cie. ,d4'nvedi)

Top [IVwik~oýf Care, 2'W(%[SP NDR UI C18 -tlri-od)

Core REL13 ID Model:

t. AeI~vvVFu vI(141 V = 715:10ft 12.7%'(STP11 NDR U I C18 -dvrivd)

I lip kim fIdR b2. 75'457P NDRU I C19 -dmived) r -- -Lowi~enu - - 1 Core~ support top ....17.F 11)LOvg 61 11E69-Cl A , RLS J)I Bto elevation of dawncomrne d. - 50.81 (Dwg 6121 E87 -derived) 113model 5359&5,45 / .LerPesm(.l 3

V=638.7 t V.

- -. __________ ktoi elevration of lower plen umo - 4 Sal' D~vg ()121 ES7 IJFSAR Sec. ,. &rivd)

NOC-AE-1 4003101 Attachment 4 Page 2 of 25 Volume (Vup) of the Upper Plenum Below Hot Leg:

0.6039791 Vup = V8 45 + V865 x 3.625005 Vup = 464.1557298 + 0.16661469 x 338.2303665 Vup = 464.1557298 + 56.35415 3

Vup= 520.51 ft Volume 865 Hot leg inside diameter El Chng = 3.625005' tt RCSlegs --- 32.25'(Dwg6-C-18-9-N-S007) --- 29"1D

- ------- -31.04167' El Chng = 0.6039791' 3

T - Vup 520.51 ft Volume 845 El Chng = 3.6416876'

- Top Elevation of Core T 26.796'(STP NDR Ul C18 - derived)

Core RELAP5 ID Model:

605 &606 Active Fuel (168')

V=715.1 ft L J Bottom of Active Fuel I 12.796'(STP NDR U1 Cl8 - derived)

Reference:

1. T. Crook, A. Franklin, J. Scherr, R. Vaghetto, A. Vanni, and Y. Hassan, "South Texas Project Power Plant RETRAN-RELAP5-3D Conversion Tables", July 2013.

NOC-AE-1 4003101 Attachment 4 Page 3 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI lb (b) Loop friction and geometry pressure losses from the core exit through the steam generators to the inlet nozzle of the reactor vessel for steady state full power operation. Also, provide the locked rotor reactor coolant pump (RCP) k-factor. Please provide the mass flow rates, flow areas, k-factors, and coolant temperatures for the pressure losses provided (upper plenum, hot legs, Steam Generators (SGs), suction legs, RCPs, and discharge legs). Please include the reduced SG flow areas due to plugged tubes. Also, provide the loss from each of the intact cold legs through the annulus to a single broken cold leg and the equivalent loop resistance for the broken loop and separately for the intact loop. Please identify the flow area (hydraulic diameter) on which the k-factors are based.

STP Response:

Nuclear Steam Supply System (NSSS) parameters are provided here for steady state full power operation under nominal conditions. To examine the pressure losses at each section of the loop, individual pressures were taken from the RELAP5-3D steady-state simulation (1). Figure 1 shows the RELAP5-3D nodalization of the plant from Ref. 1. for reference in the following tables.

Figure 1. RELAP5-3D Nodalization Diagram of the Primary System (Accumulator not Shown)

NOC-AE-1 4003101 Attachment 4 Page 4 of 25 The steady-state plant operating conditions under which the pressure losses were calculated are summarized in Table 1.

Table 1. Steady-State Plant Operating Conditions PS.ar.eter Steady Stat.e C(n01ti669qunits)

Loop Mass Flow Rate 10108.453 (Ibm/s)

Upper Plenum Bypass Fraction 2.080 (%)

Core Flow Rate 36998.386 (lbm/S)

Corefessel Inlet Temperature 560.977 OF Average Core Outlet Temperature 626.354 OF Reduced Flow Area Due to SG tube Plugging 0% (No SG plugging)

Table 2 shows the total pressure drop through different sections of the primary system, calculated as difference of the total pressure in the nodes identified in column "RELAP5-3D Nodes ID."

Table 2. RELAP5-3D Loop Pressure Differentials Location RELAP5-3D Nodes ID RELAP5-36,Pressu're Change (psid) =*

Pump Inlet 11202-11301 50.3095 Pump Outlet 11301-11601 26.0005 Vessel Inlet 11601-50101 0.6746 Downcomer 53501-50101 4.2771 Upper Plenum Bypass 58501-50101 -41.9059 Core 84501-54501 -37.4984 Vessel Outlet 10001-84501 -17.648 Hot Leg 10402-10001 -0.4353 SG Plenum Inlet 10601-10402 5.4342 SG U-tubes Inlet 10801-10601 -5.3809 SG U-Tubes 10808-10801 -15.4235 SG U-Tubes Outlet 11001-10808 1.4743 SG Plenum Outlet 11202-11001 -9.3558 Vessel Inlet to Exit 50101-86501 38.8255 Table 3 shows the flow areas, hydraulic diameters, and k-loss coefficients at different locations of the primary system, identified as junctions between nodes in column "RELAP5-3D Junction ID."

NOC-AE-1 4003101 Attachment 4 Page 5 of 25 Table 3. RELAP5-3D Loop Flow Areas and Frictional k-loss Factors

=.:.*4=.....::

. * *Ju'nction ... ' ... * * .

,, oatRELAP5-3D Flow Jno Forward Reverse Location 2 Hydraulic Junction ID Area (ft)kos k-loss Diameter (ft) ______,___

Upper Plenum - Hot Leg jX21 (865 - XOO) 4.5869 2.41667 0.1194 0 Hot Leg - SG Plenum Inlet jXO5 (X04 - X06) 4.5869 2.58583 0.46464 0.279639 SG Plenum Inlet - U-tubes jX07 (X06 - X08) 15.2929 0.0506667 0.23 0.491709 U-tubes - SG Plenum Outlet jX09 (X08 - X10) 15.2929 0.0506667 0.491709 0.23 SG Plenum Outlet - Crossover Leg jXl 1 (Xl0 - X12) 5.241 2.58583 0.279639 0.46464 Crossover Leg - RCP Inlet X12 -X13 5.241 2.583 0.001 2.09 RCP Outlet - Cold Leg X13 - X14 4.1247 2.29167 2.09 0.001 Cold Leg - Vessel Inlet jX19 (X18 - 501) 4.1247 2.29167 0 0.1194 Upper Core Plate - Upper Plenum 845 - 865 51.0764 0.0365 0.5915 0.5915 The locked rotor Reactor Coolant Pump (RCP) k-factor, calculated with the RELAP5-3D steady-state input model is shown below:

Krotor = 5.86

Reference:

1. STP Power Plant RELAP5-3D Steady-State Model Verification. July 2013.

NOC-AE-1 4003101 Attachment 4 Page 6 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI Ic (c) Capacity and boron concentration of the RWST.

STP Response:

Refueling Water Storage Tank (RWST) (LAR Enclosure 4-3, Ref. 40, Page 6.3-32)

Full tank volume, gal 550,000*

Minimum volume (Technical Specification), gal 458,000*

Boron concentration (as boric acid), ppm 2,800-3,000

• Volumes include unusable volume.

NOC-AE-1 4003101 Attachment 4 Page 7 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI ld (d) Capacity of the condensate storage tank (CST).

STP Response:

Unit 1 - Secondary Make-Up Tank, gal 300,000 Unit 2 - Secondary Make-Up Tank, gal 300,000 Unit 1 - Auxiliary Feedwater Storage Tank, TS min, gal 485,000 Unit 2 - Auxiliary Feedwater Storage Tank, TS min, gal 485,000 Secondary Make-Up Tanks provide normal, non-safety related make up water to the secondary side.

The Auxiliary Feedwater Storage Tank is the TS required safety-related water source used by the Auxiliary Feedwater System to remove heat from the RCS via the steam generators.

NOC-AE-1 4003101 Attachment 4 Page 8 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI le (e) Flushing flow rate at the time of switch to simultaneous injection.

STP Response:

Components of the flushing flow rate include the cold leg injection flow rate, hot leg injection flow rate, and vapor generation rate that were extracted from the RELAPS-3D/MELCOR simulation of a cold leg double ended guillotine (DEG) break scenario, under nominal operating conditions. Details on the simulation conditions applied are available in LAR Enclosure 4-3, Reference 5, page 12. The RELAP5-3D nodalization diagram adopted for the simulation is depicted in Figure 2 of LAR Enclosure 4-3, Reference 5, page 7.

The total cold leg injection flow rate, the total hot leg injection flow rate, and the core boil-off rate are plotted in Figure A. These thermal-hydraulic parameters have been estimated as follows:

" The total cold leg injection flow rate was estimated as the sum of the cold leg injection flow rate from each of the safety injection (Sl) trains (sum of mass flow rates of the valve components 149, 249 and 349 of Figure 2 of LAR Enclosure 4-3, Reference 5, page 7).

" The total hot leg injection flow rate was estimated as the sum of the hot leg injection flow rate from each of the Sl trains (sum of mass flow rates of the valve components 148, 248 and 348 of Figure 2 of LAR Enclosure 4-3, Reference 5, page 7).

" The core boil-off rate was calculated as the sum of vapor generation rate in all the nodes of the core (Pipe components 605 and 606 (LAR Enclosure 4-3, Reference 5, page 5), total 42 nodes). The vapor generation rate in each node was estimated by multiplying the value of the parameter "vapgen" by the volume of the node.

NOC-AE-14003101 Attachment 4 Page 9 of 25 2000 1800 -Cold Leg Injection (Ibm/s)

-Hot Leg Injection (Ibm/s) 1600

-Vapor Generation (ibm/s)

- 1400 --- Switch to Simultaneous Injection E 1200 1000 U-FA InI 800 (A

S600 400 200 0 1.,

15000 17500 20000 22500 25000 27500 30000 Time (s)

Figure A. Rate of Injection Parameters across the Simulation Table 1 summarizes the flow rates at the time to simultaneous injection (hot leg switchover time). The values reported are averaged over a period of time specified in the table.

Table 1. Average Injection and Vapor Generation Before and After Switch to Simultaneous Injection I 15000s-Simultaneous Injection 1615.50 0.00 434.91 After Simultaneous Injection 556.54 1107.42 0.75

Reference:

1. RELAP5-3D User's Manual, INEEL-EXT-98-00834.

NOC-AE-1 4003101 Attachment 4 Page 10 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI If (f) High pressure safety injection (HPSI} runout flow rate.

STP Response:

High Head Safety Injection Pumps (HHSI) (LAR Enclosure 4-3, Ref 40, Page 6.3-31)

Max. (run-out) flow rate, gal/min 1,600

NOC-AE-1 4003101 Attachment 4 Page 11 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 1g (g) Capacities and boron concentrations for high concentrate boric storage acid tanks, if part of system.

STP Response:

The boric acid storage tanks are not a part of the STP Emergency Core Cooling System (ECCS), (LAR Enclosure 4-3, Reference 40) and are not considered within the CASA Grande Analysis.

NOC-AE-1 4003101 Attachment 4 Page 12 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI lh (h) Flow rate into the RCS from the boric acid storage tanks, if applicable.

STP Response:

The boric acid storage tanks are not a part of the STP Emergency Core Cooling System (ECCS) (LAR Enclosure 4-3, Reference 40) and are not considered within the CASA Grande Analysis.

NOC-AE-1 4003101 Attachment 4 Page 13 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 1i (i) Time to empty the RWST (all pumps operating).

STP Response:

The time to empty the Refueling Water Storage Tank (RWST) (time to initiate the sump switchover procedure) was calculated using the RELAP5 and MELCOR input models described in (LAR Enclosure 4-3, Reference 5, page 5). The value reported below was calculated under the following conditions:

" Cold leg double ended guillotine (DEG) break (27.5 inch break in loop 3)

" Nominal plant conditions (all Safety Injection (SI) and Containment Spray (CS) pumps operating) (See note)

" Usable volume of the RWST (volume of the water until the low-low level alarm is reached) equal to 413,735 US gal.

The time to empty the RWST was estimated to be:

TRWST = 29.5 min Note: One of the three containment spray pumps manually secured at the beginning of the transient.

Reference:

1. STP MAAP 4.04 Input File

NOC-AE-14003101 Attachment 4 Page 14 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI lj (j) Minimum containment pressure or containment pressure versus time graph.

STP Response:

Containment pressure was calculated using the MELCOR input model described in (LAR Enclosure 4-3, Reference 5, page 8). The pressure reported was calculated under the following conditions

" Cold leg double ended guillotine (DEG) break (27.5 inch break in loop 3)

" Nominal plant conditions (all Safety Injection (SI) and 2 Containment Spray (CS) pumps operating) (LAR Enclosure 4-3, Reference 5, page 14) (See Note)

The pressure of the containment was extracted as the total pressure of the upper compartment (node 4 of the MELCOR containment model nodalization diagram in Figure 4 of LAR Enclosure 4-3, Reference 5, page 9).

The containment pressure response during the time between the sump switchover and the hot leg switchover is plotted in Figure A.

20 I

19.5 I

-Containment Pressure 19 I

---

Sump Switchover Time 18.5

- - Hot Leg Switchover I (U

(0 18 I 0.

I 0 17.5 , I I..

0 I 0 17 0

I-0~ 16.5 I 16 I I

15.5 I ir 1700 6700 11700 16700 21700 Time [s]

Figure A. Containment Pressure (from MELCOR Upper Compartment)

Note: One of the three containment spray pumps manually secured at the beginning of the transient.

NOC-AE-1 4003101 Attachment 4 Page 15 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 1k (k) Sump boric acid concentration versus time.

STP Response:

The sump boric acid concentration over time was not computed. However, the sump boric acid concentration is tightly controlled. Approximately 88% of the sump boric acid concentration is provided by the Refueling Water Storage Tank (RWST) (between 2800 and 3000 ppm) and accumulators (between 2700 and 3000 ppm). The remaining contribution to boric acid concentration comes from the RCS which varies between 0 and 3500 ppm.

Reference:

1. South Texas Project Nuclear Operating Company. Westinghouse. CN-CRA 094 Rev. 1. Required Mass of TSP for LOCA Sump Solution pH Adjustment, November 2008

NOC-AE-1 4003101 Attachment 4 Page 16 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 11 (I) Minimum RWST temperature.

STP Response:

The minimum temperature for deterministic LOCA analysis is 50 0 F. This temperature is the design range minimum for the building ambient temperature.

Reference:

1. South Texas Project Electric Generating Station. MAB HVAC Design Basis Document 5V1 09VB001 10. Rev. 3. Table T-8. Page 266.

NOC-AE-14003101 Attachment 4 Page 17 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 1m (m) Injection temperature versus time from sump during recirculation.

STP Response:

The High Head Safety Injection (HHSI) discharge temperature and the Low Head Safety Injection (LHSI) discharge temperature were extracted from the simulation results of a 27.5 inch cold leg double-ended guillotine (DEG) break performed with RELAP5-3D/MELCOR. Information on the RELAP5 and MELCOR simulation conditions are reported in the Sump Sensitivity Analysis (LAR Enclosure. 4-3, Reference. 5, page. 14) for nominal conditions.

The RELAP5 nodalization diagram of the Emergency Core Cooling System (ECCS) is depicted in Figure A to facilitate the identification of the volumes (nodes) where the liquid temperature was read.

Hot Leg x=2,3, and 4 0, RWST I S

Containment U HHSI Pump sump Exchanger Figure A. RELAP5-3D Safety Injection Nodalization Diagram (X = loop number = 2, 3, or 4) (LAR Enc. 4-3, Ref. 5, Page 7)

Based on the diagram of Figure A, the discharge temperature of the HHSI (time-dependent junction x45) is the same as the sump pool temperature (temperature of the liquid in the time-dependent volume x91) because no heat structures were modeled along the HHSI flow path.

NOC-AE-14003101 Attachment 4 Page 18 of 25 The discharge temperature of the LHSI (time-dependent junction x46) was read as the liquid temperature at the exit of the pipe component x47, simulating the primary side of the Residual Heat Removal (RHR) heat exchanger.

The injection temperature, resulting from mixing of the liquid flows from the HHSI and LHSI, was read as the temperature of the liquid in the mixing branch (node x60).

Figure B shows the discharge temperature of the HHSI, the discharge temperature of the LHSI, and the temperature of the mixed liquid injected in the primary system during the time between the sump switchover and the hot leg switchover.

200 - LHSI 190 ",.*-,**HHSI/LHSI Mix 18SO -- HHSI

• " 170 - -- SumpSwitchoverTime

- Hort Leg Switchover Time 146160 150 CL 140 E

O 130 120 110 100 1250 3750 6250 8750 11250 13750 16250 18750 21250 Time (s)

Figure B. Safety Injection Temperature (From Sump Switchover to Hot Leg Switchover)

NOC-AE-14003101 Attachment 4 Page 19 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 2a Please provide the following elevation data:

(a) bottom elevation of the suction leg horizontal leg piping and cold leg diameter STP Response:

Inner Diameter 31",

Center Line El. 22' 5-5/16" Bottom EL. of ID 21.151' Top EL. of ID 23.7 3' (See response to SNPB-RAI-2b for cold leg information).

NOC-AE-14003101 Attachment 4 Page 20 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 2b (b) top elevation of the cold leg at the RCP discharge STP Response:

Center Line El. ,-Jr..

Bottom EL. of ID 31' 1.25" Top EL. of ID 33' 4.75"

NOC-AE-1 4003101 Attachment 4 Page 21 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 2c (c) top elevation of the core (also height of core)

STP Response:

Top Elevation of Core, ft 26.796' Height of Core, ft 14' (See the response to SNPB-RAI-la)

NOC-AE-1 4003101 Attachment 4 Page 22 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 2d (d) bottom elevation of the downcomer STP Response:

Bottom Elevation of Downcomer, ft 10.81' (See the response to SNPB RAI la)

NOC-AE-1 4003101 Attachment 4 Page 23 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 3 Please provide the limiting bottom and top skewed axial power shapes.

STP Response:

The maximum core axial offsets (AOs) for LOCA analyses must meet the requirements of the reload safety analysis checklist (RSAC). The RSAC requires a minimum AO of

-20% and a maximum AO of 13% for LOCA. The cores are designed to meet Technical Specification limits of Fxy and FnAH Power shapes are not a core design parameter, but are verified to be within the RSAC limits. Based on the design limits on AO, shapes (maximum allowed design LOCA AO) can be developed for bottom and top skewed axial power shapes using the following constraints:

1 ) Power=0 at bottom and top of core (no power generation outside of core)

2) The integral from 0 to 0.5 is 0.6 and the integral from 0.5 to 1.0 is 0.4 (bottom-skewed). The integral from 0.5 to 1.0 is 0.565 and the integral from 0 to 0.5 is 0.435 (top-skewed).

3) The slope is maximum (infinite) at x=0 (bottom-skewed), infinite at x = 1.0 (top-skewed)

4) The slope at x=1.0 is negative (bottom-skewed), slope at 0 is positive (top-skewed)

5) Uniform radial power distribution (fxy is everywhere 1.0), all channels have exactly the same axial profile.

The parameters above are chosen to maximize the power at the lowest (or highest for top-skewed) by setting the slope maximum at x=0 or x=1.0. Also, uniform radial peaking would minimize channel-to-channel mixing.

The resulting function for bottom-skewed profiles is:

f(x)= 4.05499*(x^O.574847)*(1 -x).

The resulting function for top-skewed profiles is:

f(x)= 4.66256*(x)*(1 -x)^0.717919 A conceptual illustration of the functions is shown below.

NOC-AE-14003101 Attachment 4 Page 24 of 25 fix) = 4.05499(x° 3 74" 7) (1-x) f(x) = 4.66256 (x)(1 -x) 0.717919 0 0.5 10 0:5 1 Bottom-peaked power shape Top-peaked power shape The actual minimum and maximum AOs measured for several core cycles has been well within the core design requirements. STPNOC Guideline REM-2 "Core Performance Trending Program" requires measured axial offsets to be plotted for over the cycle for Units 1 and 2. The minimum measured AO for was -7.5% in Unit 2 and the maximum measured AO was 6% in Unit 1.

NOC-AE-1 4003101 Attachment 4 Page 25 of 25 SNPB, Nuclear Performance and Code Review Branch: RAI 5 Justification and description of the methodology used to compute the sump boric acid concentration versus time.

STP Response:

The sump boric acid concentration over time was not computed. However, the sump boric acid concentration is relatively constant as the concentration is tightly controlled.

Approximately 88% of the sump boric acid concentration is provided by the Refueling Water Storage Tank (RWST) (2800 and 3000 ppm) and accumulators (2700 and 3000 ppm). The remaining contribution to boric acid concentration comes from the Reactor Coolant System (RCS) which varies between 0 and 3500 ppm. The RCS contribution to the sump boric acid concentration has little effect.

Reference:

1. South Texas Project Nuclear Operating Company. Westinghouse. CN-CRA 094 Rev. 1. Required Mass of TSP for LOCA Sump Solution pH Adjustment.

November 2008.

NOC-AE-1 4003101 Attachment 5 Attachment 5 Response to SSIB Request for Additional Information:

a. Transport: RAI 12

b. Head Loss and Chemical Effects Bump Up: RAI 25, 26 C. NPSH and Degasification: RAI 30, 31, 32, 34, 35

d. In-Vessel and Boric Acid Precipitation: RAI 37

e. Debris Bypass: RAI 39

f. Defense in Depth and Mitigative Measures: RAI 41

NOC-AE-1 4003101 Attachment 5 Page 1 of 30 SSIB, Safety Issue Resolution Branch-Transport: RAI 12 Based on description of Item 5.a.4 of Volume 6.2 (Page 47), it is assumed that debris will remain in the vicinity in which it was washed down until recirculation starts. Please provide additional justification for this assumption. Please state if the debris would be redistributed during pool fill including by potential sheeting flow and if this would affect the assumption that debris is mixed homogeneously in the pool at the start of recirculation. If so, please describe which types and sizes of debris are affected.

STP Response:

Item 5.a.4 of LAR Enclosure 5 (Vol. 6.2) was added as supplementary information describing details and assumptions of the supporting debris transport calculation. A description of the homogenous pool mixing assumption used in the STP CASA Grande evaluation is provided in the response to SSIB-RAI-1 1C in STP letter to NRC dated May 22, 2014, NOC-AE-14003103 (ML14149A434).

Because homogeneous mixing of all fine and small debris is assumed, the statement of local retention made in Vol. 6.2 has no impact on debris transport. Debris transport directly to the sumps under potential sheeting flow during pool fill is assessed as a fraction of the debris that is calculated to reside on the floor immediately after the break.

Similarly, the assumption of homogeneous mixing for all fine and small debris is not affected by the debris location following pool fill.

NOC-AE-1 4003101 Attachment 5 Page 2 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 25a Volume 3, Assumption 7.f, states that it is assumed that fiberglass will accumulate uniformly on the strainers but also states that the amount of debris that can collect on the bottom of the strainer is limited to two inches. This assumption seems to contradict itself.

(a) Please explain how the assumption is accounted for in head loss calculations or provide information that shows it is not significant to the results. Please explain how a non-uniform accumulation of fibrous debris, limited by the floor or pool height, would affect the head loss calculation.

STP Response:

Assumption 7.f is accurate. There is a gap of 2 inches from the bottom of the strainer to the floor. Initially, the strainer accumulates fiber uniformly, including the 2 inch gap.

Once the loading transitions to the circumscribed area and the 2-inch gap is filled, (it physically can no longer accumulate fiber), the remainder of the strainer continues to accumulate fiber in a uniform manner. The existing STP strainer layout and design ensures that the assumption of uniform accumulation of the transported debris over all active portions of the strainer is valid.

Equations 42 and 43 (LAR Enclosure 4-3, page 180) are used to calculate the incremental thickness increase and corresponding debris surface area. Once the 2-inch gap is filled, flow is considered to be restricted and Equations 42 and 43 calculate uniform debris accumulation on all sides of the strainer except the bottom where flow is set to zero. This decrease in flow surface area initially increases the head loss compared to previous time steps when considering Equation 33 (LAR Enclosure 4-3, page 175).

Assumption 7.f (LAR Enclosure 4-3, page 79) is used to calculate the incremental debris bed thickness increase and debris (flow) areas used in the head loss calculations according to Equations 40 through 43 (LAR Enclosure 4-3, page 180). Total strainer volumetric flow rate divided by available flow area determines approach velocity used in the head-loss correlation.

CASA Grande does not compare the debris height on the top of the strainer to the pool depth. In the rare condition that the bed can exceed this height, the current evaluation allows flow area to increase unrealistically at the top of the strainer. Although unconfined bed growth above the pool is non-conservative, it was stated in Assumption 7.f that this is an unlikely occurrence. Assumption 7.f can be validated by comparing the minimum pool height (input distribution lower bound) and corresponding fiber accumulation necessary to exceed the pool height to the probability distribution of debris volume (see response to SSIB-RAI- 26d).

NOC-AE-1 4003101 Attachment 5 Page 3 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 25b (b) Please provide an evaluation of how this affects Assumption 7.e. of Volume 3 regarding homogeneous bed formation.

STP Response:

This does not affect Assumption 7.e. Assumptions 7.e and 7.f are consistent in that homogeneously mixed debris will accumulate uniformly on the as-designed STP strainer.

Initially, the strainer accumulates the homogeneously mixed fiber uniformly, including the 2-inch gap. Once the 2-inch gap is filled, (it physically can no longer accumulate fiber),

the remainder of the strainer continues to accumulate fiber in a uniform manner, and the flow area is adjusted to account for loss of flow through the bottom. Because there is no flow simulated through the bottom of the strainer after the two-inch gap is filled, debris will be homogenously mixed on all remaining sides of the strainer that do support flow.

The existing STP strainer layout and design ensures that the assumption of uniform accumulation of the transported debris over all active portions of the strainer is valid.

NOC-AE-1 4003101 Attachment 5 Page 4 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26a The submittal calculates circumscribed bed surface areas based on debris loading

(

Reference:

Volume 3, Section 5.6.2). Please provide the following information:

(a) Please state if areas calculated for beds transitioned from thin bed to circumscribed.

STP Response:

Yes, the "strainer loading table" of Section 5.6 (LAR Enclosure 4-3, page 181, Table 5.6.3) includes areas calculated for all debris loadings including the transition from thin beds to circumscribed loads. This table was used to determine if the thin-bed loading criterion was exceeded. To evaluate surface areas of debris beds transitioning from thin bed to circumscribed, the debris was linearly interpolated using the arriving debris volume calculated using manufactured density. For thin beds, the interpolated debris volume range was between the 0 and 81.79 ft 3 with a corresponding area range of 1,818.5 and 419.0 ft 2, respectively from Table 5.6.3 (LAR Enclosure 4-3, page 181).

NOC-AE-1 4003101 Attachment 5 Page 5 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26b (b) When fibrous debris is deposited on the strainer its density will be significantly increased from the manufactured value. Please state how was this accounted for (Volume 3, Page 696, Section 5.6.2).

STP Response:

The increase in density from the manufactured value is not accounted for in the "Strainer loading table" (LAR Enclosure 4-3, page 181). This table is based on thickness alone, and reports associated areas and volumes. Debris compression is accounted for by the head-loss correlation that provides an effective thickness that can be used to interpolate the table to find effective surface area.

NOC-AE-1 4003101 Attachment 5 Page 6 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26c (c) Please clarify if there are any objects around the strainer that would prevent the debris bed from accumulating uniformly as assumed in the strainer loading (Volume 3, Table 5.6.3).

STP Response:

The general arrangement of the strainers is that one side of all 3 strainer trains faces the outer containment wall, and the bottom side faces the containment floor. One of the strainer trains additionally has a structural wall approximately 2 to 3 feet away. Visual inspection of the Emergency Core Cooling System (ECCS) strainer performance testing (LAR Enclosure 4-3, Reference 53, Figure 8-6, Figure 8-7, page 47) shows that even in confined spaces (test flume) the strainers load evenly.

The floor represents an obstruction that prevents circumscribed loading from exceeding the 2-in. gap that exists between the floor and the lowest edge of the strainers. When this gap is filled, the lower surface is no longer available for flow. The strainer loading table then assumes that debris accumulation continues uniformly in all unimpeded directions.

NOC-AE-1 4003101 Attachment 5 Page 7 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26d (d) The NRC staff is of the opinion that it is not realistic to assume the thickness of the debris bed on the strainer can be such that it will exceed the height of the water level in the pool. Please explain how this affects the debris loading calculation (Volume 2, Section 5.6.2).

STP Response:

CASA Grande does not compare the debris height on top of the strainer to the pool depth. In the rare condition that the bed can exceed this height, the current implementation allows flow area to increase unrealistically at the top of the strainer.

Although unconfined bed growth above the pool is non-conservative, it was stated in Assumption 7.f that this is an unlikely occurrence (LAR Enclosure, 4-3, page 79).

Assumption 7.f is validated by comparing the minimum pool height (input distribution lower bound) and corresponding fiber accumulation necessary to exceed the pool height to the probability distribution of debris volume.

A minimum containment pool volume of 39,533 ft3 (input distribution lower bound), and pool area of 12,301 ft 2 were used in the CASA Grande evaluation (LAR Enclosure 4-3, page 45). Dividing the pool minimum volume by the pool area yields the minimum possible sampled pool level of 3.2 feet or 38.6 inches. Subtracting the height of the top of the strainer 28.5 inches (LAR Enclosure 4-3, page 63) from the minimum pool level gives the minimum thickness needed for the debris to reach the surface of the pool (10 inches of fiber accumulation). This thickness equates to a volume of 328 ft3 necessary to accumulate on the strainer to meet or surpass the minimum pool level when interpolated from Table 5.6.3.

Complementary cumulative density distributions of the total fiber amount (before transport fractions are applied) are illustrated in Figure A for many Latin Hypercube Sampling (LHS) replicates, taken over the full break size range (SBLOCA, MBLOCA, and LBLOCA), using the Zone of Influence (ZOI) sizes described in Table 2.2.0 (LAR Enclosure 4-3, page 56). This quantity includes latent fiber and all ZOI destroyed fibrous insulation quantities. Figure B shows that on a closer scale the conditional probability of exceeding 328 ft3 of fibrous debris is less than 10-14.

NOC-AE-14003101 Attachment 5 Page 8 of 30 Distribution of LDFG Volume Before Transport 100 10 x

II A

a)

E

> 10 CD LL 0

10 1020,101 10., 10 102 10 ........ 1 . 104 3

Total LDFG Volume Generated - Before Transport (ift)

Figure A: CCDF of LDFG debris generated, before transport 101~

A S10,1 Total LDFG Volume Generated - Before Transport (ft3 )

Figure B: CCDF of LDFG present and generated, before transport (zoomed in)

NOC-AE-14003101 Attachment 5 Page 9 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26e (e) Please state how often the debris loading algorithm results in a circumscribed bed or one that is transitioning to circumscribed (fully or partially filled interstitial volume).

STP Response:

The interstitial gaps of a single strainer are filled (0.5 inch debris thickness) when the volume of debris meets or exceeds 81.79 ft3 (LAR Enclosure 4-3, Table 5.6.3).

Complementary cumulative density distributions of the total fiber amount (before transport fractions are applied) are illustrated in Figure A for many Latin Hypercube Sampling (LHS) replicates taken over the full break size range (SBLOCA, MBLOCA, and LBLOCA), using the Zone of Influence (ZOI) sizes described in Table 2.2.0 (LAR Enclosure 4-3, page 56). This quantity includes latent fiber and all ZOI destroyed fibrous insulation quantities. Figure B shows on a closer scale that the conditional probability of exceeding 81.79 ft3 of fibrous debris is less than 101 4 . A data tick has been added to the figure near the 81.79 ft 3 debris volume. Note that Low Density Fiberglass (LDFG) volumes in these figures represent generated debris volumes. Corresponding debris values that reach the strainer are smaller, so the associated probability of reaching the circumscribed load is also smaller.

Distribution of LDFG Volume Before Transport 100 A

1o'ik E0 0

-J 1010k-X: 81.06 Y: 3.331e-15 10" 101 102 103 10' 3

Total LDFG Volume Generated - Before Transport (0t )

Figure A: CCDF of LDFG debris generated, before transport

NOC-AE-14003101 Attachment 5 Page 10 of 30 10 , Distribution of LDFG Volume Before Transport 10`-1 x

II 1&O A

4)

E 10-12 0

LL

.In 8 9 14 81.06 10-124 1- 1 101 102 3

Total LDFG Volume Generated - Before Transport (f0)

Figure B: CCDF of LDFG debris generated, before transport (zoomed in)

NOC-AE-1 4003101 Attachment 5 Page 11 of 30 SSIB, Safety Issue Resolution Branch-Head Loss and Chemical Effects Bump-up:

RAI 26f (f) Please explain the significance of cases that result in the interstitial volume of the strainer becoming partially or completely filled with debris.

STP Response:

As the debris fills the interstitial volume of the strainers the flow area decreases from the clean strainer area to the limiting lowest flow area of the circumscribed faces surrounding the strainers. As the flow area decreases, the fluid approach velocity (U) increases (LAR Enclosure 4-3, Equation 34, page 176). This increase in fluid approach velocity (U) directly increases the calculated head loss (LAR Enclosure 4-3, Equation 33, page 175).

NOC-AE-1 4003101 Attachment 5 Page 12 of 30 SSIB, Safety Issue Resolution Branch -NPSH and Degasification: RAI 30 The STPNOC submittal states that the degasification caused by the pressure drop through the debris bed is calculated to determine if a pump failure criterion is met

(

References:

Volume 1, Section 1.1, "Structured Information Process Flow"; Volume 3, Assumptions 8 a. through i.; Volume 3, Section 5.7.2, "Degasification"; and Enclosure 6, Table 1). Please state if the degasification calculation credits containment accident pressure. If so, please explain how the pressure for each case or condition is calculated.

Please state what temperature is used for the degasification calculation and how this temperature was calculated for each case.

STP Response:

No, the degasification calculation does not credit containment accident pressure.

In all degasification calculations, bulk sump water temperatures below 2120 F assume atmospheric pressure (14.7 psia). At sump temperatures above 2120 F, the containment pressure is assumed to be equal to the vapor pressure of the sump water.

The overall temperature range used for the degasification calculation(s) is 102.50 F to 177.50 F for SBLOCA and MBLOCA. The overall temperature range for LBLOCA calculations is 86.00 F to 255.80 F for LBLOCA .The range was determined by break size basis and is based on time-dependent changes in the bulk sump water temperature from the time of recirculation through steady-state long-term cooling (LAR Enclosure 4-3, Reference 5).

NOC-AE-1 4003101 Attachment 5 Page 13 of 30 SSIB, Safety Issue Resolution Branch -NPSH and Degasification: RAI 31 The STPNOC submittal does not seem to evaluate the possible effects of the collection of gas bubbles in the strainer or ECCS pump suction piping (

Reference:

Volume 3, Assumption 8.h. and Section 5.7.3, "Gas Transport and Accumulation"). Please explain how it was determined that gas bubbles would not collect in the strainer, or piping between the strainer and ECCS and CSS pumps and eventually transport as large voids.

If gas pockets can become trapped in these locations, please explain its effect.

STP Response:

Reference 56, TDI-6005-07, Vortex, Air Ingestion & Void Fraction South Texas Project Units 1 & 2. Revision 3: November 24, 2008, evaluates the possibility of the collection of gas bubbles in the STP strainer. While Reference 56 concludes that there is no air ingestion or void formation, CASA Grande calculates a void fraction and applies the results in the calculation Emergency Core Cooling System (ECCS) and Containment Spray System (CSS) pump NPSHr.

As stated in Section 2.2.28, the acceptance criterion for a steady-state gas void fraction at the pump suction inlet is 2%. CASA Grande conservatively assumes that any void formed at the sump strainer is fully transported to the ECCS or CSS pump suction, Assumption 8.i.

The general transport of gas voids in the piping between the strainer and ECCS and CSS pumps is explained in Reference 58, VTD-G927-0001. Units 1 and 2 Acceptable Gas Void Volumes in ECCS and RHR Suction Piping.

NOC-AE-1 4003101 Attachment 5 Page 14 of 30 SSIB, Safety Issue Resolution Branch -NPSH and Degasification: RAI 32 The NRC staff could not determine whether the calculation of NPSH Available (NPSHA) includes containment pressure greater than the saturation pressure of the sump fluid.

Volume 3, Assumption 1.c indicates that containment pressure greater than the saturation (above 14.7 pounds per square inch absolute (psia)) is not credited in the NPSH calculations (

Reference:

Volume 1, Section 1.1, "Structured Information Process Flow"; Volume 3, Sections 3, "Assumptions," and 5.7.2, "Degasification"; and Enclosure 6, Table 1). Please clarify if the calculation for NPSHA includes containment pressure above the saturation pressure of the fluid. If containment pressure greater than the saturation pressure of the fluid is credited in the NPSHA calculation, please provide justification for its use and provide the methodology used to calculate the containment pressure and sump fluid temperature for each case.

STP Response:

The NPSHA module of CASA Grande does not include containment pressure above the saturation pressure, for coolant vapor pressure conditions greater than standard atmospheric pressure. For temperatures equivalent to or above boiling at standard atmospheric pressure, the containment pressure is set equal to the saturation pressure of the fluid for the NPSHA calculation. For containment coolant vapor pressures below boiling, the standard atmospheric pressure of 14.7 psi is used as the containment pressure.

NOC-AE-1 4003101 Attachment 5 Page 15 of 30 SSIB, Safety Issue Resolution Branch -NPSH and Degasification: RAI 34 The submittal lists minimum and maximum values for containment spray flow rates

(

Reference:

Volume 3, Section 2.2.8, "ECCS and CCS Flow Rates"). Please state how these values are used in the evaluation. If flow rates other than the maximum are used, please explain how the appropriate flow rate was determined for each case.

STP Response:

User entered minimum and maximum containment spray system (CSS) flow rates are applied as probability distribution bounds; the small, medium, and large break scenarios have the same bounds. For each simulated pipe break, the probability space between the user entered minimum and maximum system flow rates are randomly sampled to determine one individual CSS pump flow rate for the scenario. Total CSS flow rate is determined by multiplying the random pump flow rate by the number of operable CSS pumps.

CSS flow rates used in the CASA Grande evaluation are entered as probability distributions with equal probability between user entered CSS pump minimum and maximum flow rates. The maximum value is set to the FLOMAP calculated average design flows for train A and B operation during recirculation (LAR Enclosure 4-3 Reference 42, page A-39).

Values other than the maximum are appropriate because the bounding minimum flow rates, used as inputs, were selected from simulated probable events (LAR Enclosure 4-3, Reference 42, page A-40) for each operable-train state (i.e. 3, 2, or 1 train operable).

CASA Grande uses the higher two-trains-operation flow rates (LAR Enclosure 4-3, Table 2.2.15, page 54) for all events with two or three trains in operation (Cases 01, 09, 22 and 26). Events with one train operation (Case 43) used their respective minimum and maximum values from Table 2.2.15 (LAR Enclosure 4-3, page 54) to bound their probability distribution.

NOC-AE-1 4003101 Attachment 5 Page 16 of 30 SSIB, Safety Issue Resolution Branch -NPSH and Degasification: RAI 35 The STPNOC submittal calculates an equivalent break size of 38.9 inches for a 27.5 inch-DEGB in Volume 3, Section 2.2.8. Please describe how the equivalent break size of 38.9 inches was calculated and why it was necessary to calculate this value.

STP Response:

The double ended guillotine break (DEGB) values computed by Equation 22 of LAR Enclosure 4-3 (Section 5.3.1, Page 125) are used only to assign DEGB breaks to a LOCA category (S, M, L).

For a double ended guillotine break, the break exit diameter (De) is equal to the inner diameter (D,) of the pipe. A DEGB with full separation results in two jets (one from each ruptured side of the pipe). Thus, there are conceptually two break areas and two break volumes. The equivalent diameter (DDEGB) is determined by doubling the cross sectional area of the broken pipe and finding single equivalent diameter to represent the total area:

2- ( Di2 A DEGB 4,42'

• ,DDEGBz Di2) 4 2 4 )

Jon'Ten = V2 - D, Equation 22

NOC-AE-1 4003101 Attachment 5 Page 17 of 30 SSIB, Safety Issue Resolution Branch -In-Vessel and Boric Acid Precipitation: RAI 37 The STPNOC submittal uses 7.5 grams per fuel assembly as the fiber acceptance limit for cold-leg breaks (

References:

Volume 1, Section 1.1, "Structured Information Process Flow," Step 18; Volume 1, Sections 1.2.10, "Boric Acid Precipitation," and 1.2.11, "In-Vessel Fiber Limits"; Volume 3, Assumption 11.b; Volume 3, Section 4.2, "Structured Information Process Flow," Step 18; Volume 3, Section 5.11.2, "Acceptance Criteria:

Debris Loads"; and Volume 6.2, Items 5.a.13 and 5.a.15). The NRC staff stated in its SE on "Evaluation of Long-Term Cooling Considering Particulate, Fibrous, and WCAP-16793, Revision 2, "Chemical Debris in the Recirculation Fluid," October 2011 (ADAMS Accession No. ML13084A154), that the maximum amount of fiber that would be present in the limiting reactor design following a cold-leg break would be expected to be about 7.5 grams, if the hot-leg break fiber amount did not exceed 15 grams. The staff did not conclude that a fiber load of 7.5 grams was adequate to ensure that boric acid precipitation would not occur. The amount was projected as the potential maximum in the short term until industry completed a separate program on boric acid precipitation (BAP). In its evaluation, the staff considered that the plant calculation of the in-vessel debris load included the worst case debris load for the plant and that most plants would have much less than 7.5 grams of debris following a cold-leg break. Note that testing for the WCAP did show that the flow required to match decay heat boil off would reach the core following a cold-leg break with debris loads greater than 7.5 grams, but did not show that mixing credited to prevent BAP would not be affected. The limit of 7.5 grams per fuel assembly has not been technically justified as an acceptance criterion for BAP.

Please provide the technical basis for assuming that 7.5 grams is an acceptable limit for a cold-leg break at STP when considering the potential for boric acid precipitation.

STP Response:

7.5 g/fuel assembly (FA) of fiber is chosen as a threshold of concern for boric acid precipitation (BAP) based on previous results (LAR Enclosure 4-3, Reference [62]) that showed very little head loss when 15 g/FA with a full amount of chemical precipitates were applied during hot leg break flow conditions. With a debris load of 7.5 g/FA, the core is expected to remain full of water during a cold leg break (CLB) even though there is no opportunity for bypass flow credited in the analysis. However, during a CLB, at these low debris amounts (7.5 g/FA), STP would have significant flow through bypass pathways (LOCA holes and over the top of the core from barrel-to-baffle bypass region) as described in the LAR Enclosure 4-1, Section 2.1.2. LAR Enclosure 4-1, Section 2.1.2 describes thermal hydraulic analysis of extreme scenarios that show the core would continue to be supplied with adequate flow such that cooling is preserved, and the core would be reflooded early in the transient (ADAMS Accession No. ML14029A533).

The 7.5 g/FA threshold is applied under the assumption of full debris deposition on the fuel and takes no credit for debris that may actually deposit in the barrel-to-baffle bypass. The CASA Grande analysis records a scenario failure whenever an equivalent inventory of 7.5 g/FA enters the core. STP fuel assemblies are designed with a significant gap below the bottom tie plate that provides a large flow plenum between the bottom of the active fuel and the top of the bottom core plate. The flow-channel to barrel-to-baffle region has a large gap (approximately 2 inches) around the entire core periphery that is not credited for possible debris retention or for allowing low concentration flow to circulate though the bypass region through the Loss of Coolant Accident (LOCA) holes or over the top of the core.

NOC-AE-1 4003101 Attachment 5 Page 18 of 30 STP has shown the chemical contribution to head loss is insignificant prior to hot leg switchover (about 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />). Based on the STP reactor design, based on bounding experimental results from both the work done in support of LAR Enclosure 4-3, page 82 of 248, and from Reference (1), and based on bounding thermal hydraulic simulation results for blockage, it is reasonable to use 7.5 g/FA, defined by tests using a conservative chemical load, as a threshold of concern for BAP. Note that this limit is applied to all scenarios at all times regardless of realistic chemical loading and is therefore conservative even though higher thresholds could be considered, particularly early in the scenario. Although the core-fiber boron precipitation threshold could be reasonably described as a distribution having a mean much higher than 7.5 g/FA, the STP LAR applied a sharp, single-value threshold to maintain clarity on this reactor performance metric.

References:

1. CHLE-012 T1 MBLOCA Test Report Rev 4. Albuquerque, NM: University of New Mexico, February 18, 2014. (ML14072A084)

2. CHLE-014 T2 LBLOCA Test Report Rev 3. Albuquerque, NM: University of New Mexico, February 22, 2014. (ML14072A085)

NOC-AE-1 4003101 Attachment 5 Page 19 of 30 SSIB, Safety Issue Resolution Branch -Debris Bypass: RAI 39a The submittal states that debris bypass or penetration testing was completed to support modeling of the bypass of debris past the STP strainer (

Reference:

Volume 6.2, Item 5.a.16). Please provide additional details on how debris penetration testing for fiber was conducted. Specifically, please provide the following information:

(a) Provide details on the characteristics of the fiber that was added to the test facility.

i. How the fiber was prepared.

ii. State What the percentages were of each fiber classification as described in NUREG/CR-6808, "Knowledge Base for the Effect of Debris on Pressurized Water Reactor Emergency Core Cooling Sump Performance," February 2003 (ADAMS Accession No. ML030780733), Table 3-2 after the fiber was prepared.

iii. How was it ensured that agglomeration of the fiber did not occur prior to addition to the test loop?

STP Response:

i. The fiber was prepared with an ARL modified NEI protocol (LAR Enclosure 4-3, Reference 26, page 16). The general fiber preparation procedure was (LAR Enclosure 4-3, Reference 26, Attachment A):

1. Ensure drain valve in debris preparation tank is closed.

2. Ensure the debris filter is installed on the return line from the tank.

3. Weigh out the predetermined batch size of fiber and place into the scaled up debris preparation tank.

4. Record quantity and debris lot information in Table 4.

5. Fill debris preparation tank with water to at least the minimum dilution specified in the NEI debris preparation protocol (< 0.72 Ibm/gal). For 2.4 Ibm of fiber, the recommended dilution based on the shakedown testing results in 30 gal.

6. Start recirculating flow pump for the debris preparation tank.

7. Ensure the pressure washer nozzles are connected.

8. Change the height of the pressure washer nozzles to be within 2 in above or below the water surface.

9. Start the pressure washer and run approximately 15 min (determined from shakedown testing for 2.4 Ibm of fiber).

NOC-AE-1 4003101 Attachment 5 Page 20 of 30

10. Examine debris characteristics for the batch and note the "comments" section on the following page. If debris characteristics do not meet expectations, apply high pressure spray for an extended duration as needed, and document in comments section.

11. Stop the pressure washer and record the actual time over which the spray was applied in Table 4.

12. Place a clean barrel under the debris preparation tank and drain debris through the valve in the bottom of the tank.

13. Rinse tank walls and ensure filter is free of debris.

14. Label the barrel containing the debris with the batch number and debris weight.

Any deviations from this procedure were documented for each test in Attachment A of the South Texas Penetration Test Report (LAR Enclosure 4-3, Reference 26). For example, Tests 1 through 4 interchanged steps 10 and 11 which were noted with "ink and initial" modifications during testing.

ii. Analysis to quantitatively characterize the fiber after preparation and prior to testing was not conducted. However, step 10 of the fiber preparation procedure, as described above allows the fiber to be high pressure sprayed multiple times if the expectation of the mostly class 2 fibers was not met.

iii. Step 15 of fiber penetration test procedure requires "gently re-mixing the debris using a mixing paddle" before the debris is introduced into the hopper/test flume (LAR Enclosure 4-3, Reference 26, Attachment A).

The debris hopper also introduces additional mixing energy to the debris before it enters the flume.

NOC-AE-1 4003101 Attachment 5 Page 21 of 30 SSIB, Debris Bypass: RAI 39b (b) For tests that had more than one batch of fiber added to the test, please state what the timing was of each debris addition.

STP Response:

The subsequent batch of fiber was not added to the test until at least 5 pool turnovers were completed (LAR Enclosure 4-3, Reference 26, page 26). Since the tests were conducted at different flow rates, the duration of a pool turnover also varied. The durations of 5 pool turnovers for Tests 1 - 5, Test 6, and Test 7 were 11.8 minutes, 50.5 minutes, and 19.1 minutes, respectively (LAR Enclosure 4-3, Reference 26, Attachment A).

NOC-AE-1 4003101 Attachment 5 Page 22 of 30 SSIB, Debris Bypass: RAI 39c (c) Please describe the design of the test facility.

STP Response:

A schematic of the test configuration is below (LAR Enclosure 4-3, Reference 26, page 21).

A table of the dimensions critical to quality for Test 1 is shown on the subsequent page (LAR Enclosure 4-3, Reference 26, Attachment A).

NOC-AE-1 4003101 Attachment 5 Page 23 of 30 Dimension Description Prescribed Assigned Reference Figure Measurement Dimension Descriptio Dimension Tolerance Inflow Section Width 36" 1/2" Figure 2 35-7/8" Inflow Section Length 72" 1/2" Figure 2 71-3/4" Tapered Section Length 48" 1/2" Figure 2 48" Tapered Section Width 36" 1/2" Figure 2 35-7/8" (start)

Tapered Section Width 21-13/16" 1/2" Figure 2 21-7/8" (end)

Strainer Section Width 21-13/16" 1/2" Figure 2 21-7/8" Strainer Section 32" 1/2" Figure 2 31-3/4" Length Distance from Eastern Strainer Edge to 2" 1/2" Figure 2 2-1/8" Eastern Tank Wall Distance from Northern Strainer Disk to Northern 2-112" 5/8" Figure 2 2-3/4" Tank Wall Distance from Southern Strainer Disk to Southern 2-1/2" 5/8" Figure 2 2-1/4" Tank Wall Strainer Submergence 9-1/2" 1" Figure 3 ---

Total Strainer Height 30-1/2" 1/8" Figure 4 30-1/5" Strainer Disk Height 25" 1/8" Figure 4 25" Strainer Disk Width 28" 1/8" Figure 4 28" Total Strainer Width 31" 1/8" Figure 4 30-15/16" Disk Height off Ground 2-1/4" 1/8" Figure 4 2-3/16"

+ 1/16" Active Module Length 16-13/16" -3/16" Figure 4 16-5/8"

NOC-AE-1 4003101 Attachment 5 Page 24 of 30 SSIB, Debris Bypass: RAI 39d (d) Was the circulation of fluid within the tank turbulent? Did debris settle? If some debris did not reach the strainer, how was this accounted for?

STP Response:

Turbulence was introduced into the test tank via mixers. The mixers provided enough turbulence to suspend debris in the test tank and prevent debris from settling to the floor (LAR Enclosure 4-3, Reference 26, page 16). Since the fiber remained suspended, all debris reached the strainer.

NOC-AE-1 4003101 Attachment 5 Page 25 of 30 SSIB, Debris Bypass: RAI 39e (e) How was it ensured that fiber did not bypass the filters during the test?

STP Response:

The test setup as described in response SSIB RAI 39c ensures full flow through the filter bags during the test. Additionally, the NUKON was prepared as fines having a characteristic diameter of 7 microns (LAR Enclosure 4-3, Reference 44, Table 3-2). The nominal pores of the bags used for fiber collection was (were) 5 microns. The larger diameter fines, combined with the random orientation of the fiber as it contacted the filter bag, suggests that debris did not bypass the filters.

NOC-AE-1 4003101 Attachment 5 Page 26 of 30 SSIB, Debris Bypass: RAI 39f (f) Was the design of the strainer and the design of the test facility (flow rate, etc.)

prototypical with respect to the STP strainer?

STP Response:

Yes, the penetration test was prototypical with respect to the STP strainer. A STP prototypical PCI Sure-Flow strainer module was tested. The flow rate was scaled such that the maximum approach velocities of the tests and the STP strainer were equivalent at 0.0086 ft/s (LAR Enclosure 4-3, Reference 26, Table 3, page 19). Tests 1 through 4 were tested with a total of 2.4 lb of fine fibrous debris; Tests 5 through 7 were tested with a total of 9.6 lb of fine fibrous debris (LAR Enclosure 4-3, Reference 26, Table 3, page 19). The Design Basis Accident (DBA) test at ARL in July was conducted with a total of 5.5 lb of fine fibrous debris (1).

Reference:

1. 0415-0100071WN / 0415-0200071WN. "South Texas Project Test Report for ECCS Strainer Testing July 2008." Revision A. 11/24/2008.

NOC-AE-1 4003101 Attachment 5 Page 27 of 30 SSIB, Safety Issue Resolution Branch -Defense in Depth and Mitigative Measures:

RAI 41a Volume 1, Appendix C, Section C.5.4, lists mitigative measures that can be taken if the strainer becomes blocked. It is not clear how the mitigative measures identified to address strainer blockage are implemented at STP (note that these actions are also credited for prevention of inadequate core flow). Please explain the following to explain how the mitigative measures are capable of providing alternate flow to the required equipment.

(a) The mitigative actions identified to reduce flow through the strainers appear to actually be designed to conserve RWST volume. These measures may delay the initiation of recirculation, but except for securing CSS pumps will not reduce flow through the strainer. Please state at what point in the recovery these actions are performed. If not performed immediately, will the RWST inventory be conserved? If the reductions of flow through the strainer do not occur until after strainer blockage is evident, please state if these actions are effective.

STP Response:

A reduction in flow will occur (one (1) Containment Spray System (CSS) pump secured) before switchover to recirculation as directed by the conditional information page (CIP) in STP procedure OPOP05-EO-EOOO, "Reactor Trip or Safety Injection" (LAR Enclosure 4-3, Reference 32). Securing a single CSS pump will conserve Refueling Water Storage Tank (RWST) inventory. Per EOP OPOP05-EO-EO10, "Loss of Reactor or Secondary Coolant" all sprays can be secured after 6.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> based on Iodine levels low enough to support habitability but containment pressure would need to be less than 6.5 psig and TSC concurrence. Conservation of RWST volume and reduction of strainer flow are both beneficial strategies for LOCA response that can be achieved by securing spray pumps.

The third CSS pump is stopped shortly after the LOCA occurs, before the RWST is empty. As a consequence, when recirculation starts, that train will have approximately 40% less flow through the strainer resulting in much less debris accumulation and therefore head losses (on that train's strainer).

Stopping a CSS pump in the most likely plant state scenario (all CSS trains running) is intended to conserve RWST inventory. Additionally, the flow through the Emergency Core Cooling System (ECCS) strainer will be reduced by approximately 40% in the train that the CSS pump is stopped. This reduction in total flow through the strainer has the additional benefit of reducing debris buildup on that specific strainer. A potentially adverse effect of securing spray flow is that the lower debris bed inventory allows more fiber penetration to the core.

In the risk-informed methodology, the effect of the reduction in flow is taken into account as described in the LAR Enclosure 4-2, Section A.4.2, page 83 of 257 in the description of Top Event OFFS. The strainer loading is accounted for as found in LAR Enclosure 4-3, Equations 87 through 93 page 210 of 248 and Section 3 page 78 of 248 (6e: "It was assumed that the debris transportto each of the strainersis proportionalto the flow rate through each strainerdivided by the total flow rate through all of the strainers. This is a reasonableassumption since the debris transportswith the flow.")

NOC-AE-1 4003101 Attachment 5 Page 28 of 30 SSIB, Defense In Depth and Mitigative Measures: RAI 41b (b) Please state if STP has implemented operating procedures to secure the third train of ECCS/CSS if all three are initiated following a LOCA.

STP Response:

STPNOC did not revise the Emergency Operating Procedure (EOP) Emergency Core Cooling System-(ECCS) termination criteria to secure any trains of safety injection. The EOPs were modified to secure one train of Containment Spray System (CSS) if all three trains are injecting.

NOC-AE-1 4003101 Attachment 5 Page 29 of 30 SSIB, Defense In Depth and Mitigative Measures: RAI 41c (c) Please clarify if STP implemented operating procedures or other guidance to backwash the strainers, if necessary. If so, please provide details on the procedural controls for this action.

STP Response:

No. STPNOC does not have a procedure or guidance to allow backwashing of the ECCS strainers.

NOC-AE-1 4003101 Attachment 5 Page 30 of 30 SSIB, Defense In Depth and Mitigative Measures: RAI 41d (d) Please state when the RWST refill is started and how long it takes to refill RWST to the point where injection from the tank is viable. Please note that if the tank is not ready for injection when blockage occurs, this action may not be effective. The NRC staff notes that the STPNOC submittal states that most strainer blockage events occur within the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of the LOCA recovery.

STP Response:

As stated in the LAR Enclosure 4-1, page C14 the STP Emergency Operating Procedures (EOPs) contain steps to refill the Refueling Water Storage Tank (RWST).

STP procedure OPOP05-EO-EO10 "Loss of Reactor or Secondary Coolanf' directs the operator to enter procedure OPOP05-EO-ES13 "Transfer to Cold Leg Recirculation."

OPOP05-EO-ES13 directs the operators to refill the RWST in the step following completion of transfer to cold leg recirculation. Following a hypothesized large break LOCA, this would occur about 20 to 25 minutes following the start of the event.

The makeup flow rate to RWST is approximately 150 gpm. This equates to approximately 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br /> to make up 100,000 gallons. The minimum volume allowed in the RWST before securing the ECCS/CSS pumps is 32,500 gallons by the procedure.

In the unlikely event that debris prevents recirculation in all Emergency Core Cooling System (ECCS) trains in 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br />, the RWST would be sufficiently full of water. The refill rate (plus RWST drain down) would be able to meet core cooling requirements without interruption over a long period of time as shown in the LAR Enclosure 4-3, Table 5.10.1 page 224 of 248. Thus the actions to re-fill in the RWST will be effective for providing an alternative method of cooling water supply.

NOC-AE-1 4003101 Attachment 6 Attachment 6 Definitions and Acronyms

NOC-AE-1 4003105 Attachment 6 Page 1 of 2 Definitions and Acronyms ARL Alden Research Laboratory EOP Emergency Operating BA Boric Acid Procedure(s)

BAP Boric Acid Precipitation EPRI Electric Power Research BC Branch Connection Institute BEP Best Efficiency Point ESF Engineered Safety Feature B-F Bimetallic Welds FA Fuel Assembly(s)

B-J Single Metal Welds FHB Fuel Handling Building BWR Boiling Water Reactor GDC General Design Criterion(ia)

CAD Computer Aided Design GL Generic Letter CASA Containment Accident GSI Generic Safety Issue Stochastic Analysis HHSI High Head Safety Injection CCDF Complementary Cumulative (ECCS Subsystem)

Distribution Function or HLB Hot Leg Break Conditional Core Damage HLSO Hot Leg Switchover Frequency ID Inside Diameter CCW Component Cooling Water IGSCC Intergranular Stress CDF Core Damage Frequency Corrosion Cracking CET Core Exit Thermocouple(s) ISI In-Service Inspection CHLE Corrosion/Head Loss LAR License Amendment Experiments Request CHRS Containment Heat Removal LBB Leak Before Break System LBLOCA Large Break Loss of Coolant CLB Cold Leg Break or Current Accident Licensing Basis LDFG Low Density Fiberglass CRMP Configuration Risk LERF Large Early Release Management Program Frequency CS Containment Spray LHS Latin Hypercube Sampling CSHL Clean Strainer Head Loss LHSI Low Head Safety Injection CSS Containment Spray System (ECCS Subsystem)

(same as CS) LOCA Loss of Coolant Accident CVCS Chemical Volume Control MAAP Modular Accident Analysis System Program DBA Design Basis Accident MAB/MEAB Mechanical Auxiliary DBD Design Basis Document Building or Mechanical D&C Design and Construction Electrical Auxiliary Building Defects MBLOCA Medium Break Loss of DEGB Double Ended Guillotine Coolant Accident Break NIST National Institute of DID Defense in Depth Standards and Technology DM Degradation Mechanism NLHS Non-uniform Latin ECC Emergency Core Cooling Hypercube Sampling (same as ECCS) NPSH Net Positive Suction Head, ECCS Emergency Core Cooling (NPSHA - available, System NPSHR - required)

ECWS Essential Cooling Water NRC Nuclear Regulatory System (also ECW) Commission EOF Emergency Operations NSSS Nuclear Steam Supply Facility System

NOC-AE-1 4003101 Attachment 6 Page 2 of 2 Definitions and Acronyms OD Outer Diameter SBLOCA Small Break Loss of Coolant PCI Performance Contracting, Accident Inc. SC Stress Corrosion PDF Probability Density Function SI Safety Injection (same as PRA Probabilistic Risk ECCS)

Assessment SIR Safety Injection and PWR Pressurized Water Reactor Recirculation PWROG Pressurized Water Reactor SRM Staff Requirements Owner's Group .Memorandum PWSCC Primary Water Stress STP South Texas Project Corrosion Cracking STPNOC STP Nuclear Operating QDPS Qualified Display Processing Company System TAMU Texas A&M University RAI Request for Additional TF Thermal Fatigue Information TGSCC Transgranular Stress RCB Reactor Containment Corrosion Cracking Building TS Technical Specifcation(s)

RCFC Reactor Containment Fan TSB Technical Specification Cooler Bases RCS Reactor Coolant System TSC Technical Support Center RG Regulatory Guide TSP Trisodium Phosphate RHR Residual Heat Removal UFSAR Updated Final Safety RI-ISI Risk-Informed In-Service Analysis Report Inspection USI Unresolved Safety Issue RMI Reflective Metal Insulation UT University of Texas (Austin)

RMTS Risk Managed Technical V&V Verification and Validation Specifications VF Vibration Fatigue RVWL Reactor Vessel Water Level WCAP Westinghouse Commercial RWST Refueling Water Storage Atomic Power Tank ZOI Zone of Influence