ML14234A284

NRR E-mail Capture - STP risk-informed GSI-191 Meeting on 8/20/14

ML14234A284
Person / Time
Site:	South Texas
Issue date:	08/18/2014
From:	Harrison A South Texas
To:	Balwant Singal Division of Operating Reactor Licensing
References
MF2400, MF2401
Download: ML14234A284 (39)

Text

NRR-PMDAPEm Resource From: Harrison Albon [awharrison@STPEGS.COM]

Sent: Monday, August 18, 2014 6:55 PM To: Singal, Balwant Cc: Oesterle, Eric; Mitchell, Eliza

Subject:

RE: Out-of-Office (August 18 to August 29, 2014)

Attachments: 8-20-14 NRC Meeting - main.pptx Here is the main set of slides for the STP risk-informed GSI-191 meeting on 8/20/14.

Call me if you have questions.

Wayne Harrison STP Licensing (979) 292-6413 From: Singal, Balwant [1]

Sent: Thursday, August 14, 2014 2:45 PM To: Harrison Albon Cc: Oesterle, Eric; Mitchell, Eliza

Subject:

RE: Out-of-Office (August 18 to August 29, 2014)

Wayne, Wayne, Please copy the following NRC staff members on your e-mail forwarding the presentation slides:

Oesterle, Eric Eric.Oesterle@nrc.gov Mitchell, Eliza Eliza.Mitchell@nrc.gov Thanks.

Balwant K. Singal Senior Project Manager (Comanche Peak, STP, Diablo Canyon, and Palo Verde)

Nuclear Regulatory Commission Division of Operating Reactor Licensing Balwant.Singal@nrc.gov Tel: (301) 415-3016 Fax: (301) 415-1222 From: Harrison Albon [2]

Sent: Thursday, August 14, 2014 12:58 PM To: Singal, Balwant Cc: Lyon, Fred; Blossom, Steven; Kee, Ernie

Subject:

RE: Out-of-Office (August 18 to August 29, 2014)

Balwant, 1

You asked yesterday when we would have slides to you for the 8/20 meeting. Well have them to you (or Fred) by COB on Monday, probably before.

Regards, Wayne Harrison STP Licensing (979 292-6413 From: Singal, Balwant [3]

Sent: Thursday, August 14, 2014 10:30 AM To: 'Hope, Timothy' (Timothy.Hope@luminant.com); Sterling, Lance; Harrison Albon; Carl.Stephenson@aps.com; pns3@pge.com Cc: Lyon, Fred; Watford, Margaret; Oesterle, Eric; Markley, Michael

Subject:

Out-of-Office (August 18 to August 29, 2014)

I will be out-of-office from August 18 to August 29, 2014. Please contact the following NRC staff members for Project Manager assistance:

Fred Lyon at 301-415-2296 for Comanche Peak, South Texas Project, and Diablo Canyon.

Eric Oesterle at 301-415-1014 for Palo Verde.

Thanks.

Balwant K. Singal Senior Project Manager (Comanche Peak, STP, and Palo Verde)

Nuclear Regulatory Commission Division of Operating Reactor Licensing Balwant.Singal@nrc.gov Tel: (301) 415-3016 Fax: (301) 415-1222 2

Hearing Identifier: NRR_PMDA Email Number: 1522 Mail Envelope Properties (8C918BCF8596FB49BD20A610FA5920CF02083554)

Subject:

RE: Out-of-Office (August 18 to August 29, 2014)

Sent Date: 8/18/2014 6:55:12 PM Received Date: 8/18/2014 6:56:07 PM From: Harrison Albon Created By: awharrison@STPEGS.COM Recipients:

"Oesterle, Eric" <Eric.Oesterle@nrc.gov>

Tracking Status: None "Mitchell, Eliza" <Eliza.Mitchell@nrc.gov>

Tracking Status: None "Singal, Balwant" <Balwant.Singal@nrc.gov>

Tracking Status: None Post Office: CEXMBX03.CORP.STPEGS.NET Files Size Date & Time MESSAGE 2478 8/18/2014 6:56:07 PM 8-20-14 NRC Meeting - main.pptx 406744 Options Priority: Standard Return Notification: No Reply Requested: No Sensitivity: Normal Expiration Date:

Recipients Received:

South Texas Project Pilot Risk-Informed GSI-191 Li Licensing i A Application li ti Meeting with NRC Staff August 20, 2014

STP Meeting Participants

• Mike Murray, Manager Regulatory Affairs

• Steve Blossom, Blossom Risk-Informed Risk Informed GSI-191 GSI 191 Project Manager

• Ernie Kee, STP Risk-Informed GSI-191 Technical Team Lead

• Wayne Harrison, STP Licensing Lead

• Drew Richards, Richards STP Licensing

• David Johnson, ABS Consulting

• Bruce Letellier, Alion Science & Technology

• Janet Leavitt, Leavitt Alion Science & Technology

• Rodolfo Vaghetto, Texas A & M University

• Zahra Mohaghegh, University of Illinois, Champagne/Urbana

• Ed EdwarddDD. Blandford, Bl df d University U i it off N New M Mexico i

• David Morton, University of Texas, Austin

• John Hasenbein, University of Texas, Austin 2

Meeting Purpose and Desired d Outcomes

Purpose:

p

- Provide opportunity to gain understanding of Open Issues from the NRC Staff Review

- Initiate Resolution Process for Open Issues from the NRC Staff Review

- Gain alignment on topics the Staff wants the Licensee t presentt to to t the th ACRS on 9/3/2014

• Desired Outcomes:

1. Alignment on the nature of the open issues 1

2. Propose next action for each item

3. Alignment on ACRS topics 3

Agenda

• Role of CASA Grande

- PRA input application vs 50.46 application

- Describe Alion plan to revise the CASA Grande documentation ((V&V,, etc.))

• Discussion of NRC Staff identified issues (by Branch)

- Confirm agreement on what the issue is that needs to be resolved

- Determine the next step for each issue

• Discuss upcoming September 3, 2014 ACRS S b Subcommittee itt M Meeting ti

• Discuss NRC plans for audit of STP application material in September, p , 2014 4

Agenda - Branch Discussions

1. SSIB RAIs

- Head loss and chemical effects bump-up (RAI 15, 16, 17, 18)

• Includes purpose of L-Star correlation

• Includes purpose of VISTA correlation

- In In-vessel vessel and boric acid precipitation (RAI 37)

2. APLAB RAIs

- CASA Grande - Plant Configuration: RAI 1b, 2b, 3

- HRA: RAI 3, 5

- Uncertainties: RAI 1, 2, 4, 5, 6

- Stable end state: PRA Success - RAI 3c

- Use of different distributions

3. ESGB RAIs

- Coatings: RAI 1, 2, 6

- Chemical Effects: RAI 1 5

Role of CASA Grande

• Relative to PRA - discussion lead by Ernie Kee

• Alion plan for documentation - discussion lead by Bruce Letellier 6

Risk Assessment Overview

• Probabilistic Risk Assessment (PRA) is performed in a standard way with the existing plant PRA model of record in an assessment framework

- Uses standard plant procedures, processes, and programs for PRA analysis and assessment that support standard risk applications such as Maintenance Rule, RMTS, risk categorization, SDP

- Uses U existing i ti industry i d t standards t d d ffor PRA qualitylit

• To specifically analyze GSI-191 risk concerns, engineering judgment for recirculation risk input (failure probability) needed to be improved

- Risk (p (probability) y) of sump p failure needed improvements p - enhance the simplistic demand failure probability with engineering analyses

- A top event for in-vessel failures added for completeness as current PRA industry standards do not require consideration of long term cooling

- Inputs p for obtainingg conditional failure likelihoods are supported pp byy p prudent,,

peer-reviewed, data and engineering analyses

- Previously developed and accepted engineering models were adopted where possible to reduce development and regulatory review burden 7

In early 2011 when the flow chart was developed, the focus was on quantitative risk assessment to confirm existing qualitative evaluations indicating safe operation based on defense in depth measures and safety margins derived from plant changes already in place, such as strainer modifications, targeted insulation removal, and operations/procedure changes. Most plants have similar qualitative evaluations.

8

Figure 1 from RG 1.174 - Relationship of Regulatory Guide 1.174 to other risk-informed guidance 9

Progress since May, 2012 ACRS

• Regulatory interface

- License Amendment d Request (LAR)

( ) submitted b d 2013

- Set #1 Requests for Additional Information (RAIs) - 249 RAIs responded to in 3 packages

• Additional insights g and review support pp

- Quantifications: 12/2011, 05/2013, 11/2013, all show very small risk

- Sensitivity study methodology developed

- Additional confirmatory chemical effects tests completed that support and enhance previous chemical effects observations (ICET (ICET, T1 T1, T2 T2, bench top tests)

• T3 and T4 overloaded tests

• T5 repeat of T2 (LLOCA) with blender beds.

- New thermal-hydraulics capability and analyses

• 3D modeling

• coupling of RELAP5 and MELCOR models

• Advanced results visualization capabilities

• Quantification of safety margin in head loss, chemical effects further demonstrating safety margin with expanded experimental database. database 10

PRA input versus engineering analysis

• As opposed to a deterministic methodology, the objective of commercial nuclear power probabilistic risk assessment has been quantification of risk measures measures, typically CDF and LERF associated with plant operation and design

- RG1.174 defines (CDF, CDF) and (LERF, Delta LERF) measures, adding quantification of a change in risk due to some changed activity or plant modification

- One of the plant risk assessment inputs has (until recently) included a simplistic, all-inclusive failure mode demand failure likelihood for ECCS recirculation failure that is, a failure of LTCC based on engineering judgment

• The risk assessment should realistically assess exposure to CDF (as a precursor) and LERF as potential exposure to significant radioactive release

- The risk should be near zero - RG1.174 RG1 174 defines small small or very very small small

- A realistic risk estimate should be made to direct effective decision-making

• 10CFR50.46c should ask for the risk-informed evaluation to be done in a standard way using the plant PRA

- The ECCS failure likelihood is replaced with likelihoods for potential failure modes associated with the concerns raised in GSI-191 supported by prudent, peer-reviewed engineering analyses

- Standards, procedures, and processes that have been established for commercial nuclear power risk assessment (e.g., ASME PRA standard, RG1.200, existing plant risk assessment procedures for SDP, SDP RMTS RMTS, NEI Initiative 5b) 11

Role of CASA Grande, VISTA, and L* L

• In order to support the risk assessment, including RG1.174 requirements, engineering analyses and data are required

• RG1.174 asks for safety margin and defense in depth

- CASA Grande produces prudent, peer-reviewed engineering analysis of the response to hypothesized LOCA that includes safety margin in head loss

- VISTA and L* were developed in response to multiple RAIs that questioned the safety margin in retained in STPs head loss estimates

• In order to include more conventional head loss data, and address other comments with NUREG 6224,, VISTA was developed p and then compared p to STPs implementation of 6224 that includes safety margin.

• Similarly, to address comments about the formulation of the safety margin added in chemical head loss, L* was developed as a conservative representation of head loss including more data

• It is shown that, with more data included and addressing comments about the formulation of multiplicative bump up, STPs implementation of head loss retains safety margin 12

SSIB RAIs

• Head loss and chemical effects bump bump-upup (RAI 15, 16, 17, 18)

- Includes purpose of LL-Star Star correlation

- Includes purpose of VISTA correlation

• In-vessel I l and dbboric i acidid precipitation i it ti (RAI 37) 13

SSIB RAI 15 Conventional i l Debris b i Head d Loss Correlation l i (a) Validation for full range of debris loads and morphologies Correlations help identify combinations of concern, especially across the full spectrum of RI analysis, and strainer tests provide proof of performance for the most challenging conditions identified (b) Non-homogeneity Inhomogeneous composition through the thickness of the bed may exist that elevates the observed head loss.

loss Enclosure 1 relaxes this concern (c) Validation over full flow range Reassessed by looking at internal Reynolds number (Also 17 (c))

(d) Parameter uncertainty Direct measurements of constituents show that more accurate estimates help to confirm results 14

SSIB Head Loss, RAI 16 Verticall Loop Head d Loss Applicability l bl (a) Are STP vertical loop tests valid considering other facilities showed significantly different results?

Tests are important and are shown to be consistent with conclusions (b) Provide evidence that vertical loop tests conducted under site specific conditions will correlate to flume tests conducted under similar conditions Tests are important and are shown to be consistent with conclusions when internal flow conditions are considered (c) Provide a basis for using a correlation that has not been validated specifically for STP plant conditions and geometries includingg the sludge-limit g compaction p and a factor of 5 uncertaintyy bound is designed g to provide p

confidence that predictions bound realistic strainer performance for the majority of debris combinations and flow conditions that are experienced in the plant.

(d) How does NUREG/CR-6224 correlation predict the head losses that would be expected under conditions similar to those in the two flume tests conducted by STP in February and July 2008?

The July test is not representative of STP post-LOCA conditions and is not considered further.

February testing did represent conditions that could be expected at STP. It is unlikely, given the safety margin in the 5X bump up factor, that additional testing would challenge the conclusion that the 6224 correlation (as implemented) is acceptable 15

SSIB Head Loss RAI 17 Validation lid i iin Hypothesized h i d STP Conditions di i (a) Debris constituents in validation testing are not plant-specific The correlation (including safety margin) is used to reveal trends in strainer performance that may challenge risk informed success criteria. STP does not have micro-porous debris (b) Debris sizes in validation testing are not plant-specific.

The UNM vertical loop testing is considered to be part of the NUREG/CR-6224 validation effort which included a wide range of debris sizes including blender-processed fiberglass. HTVL testing conducted at Alion Hydraulics Laboratory used a modified NEI debris preparation. Therefore, prototypical debris sizes are considered to be well represented. d (c) Very little validation testing was conducted at STP velocities and none validated the correlation Addressed in RAI 15c (d) Validation testing did not include prototypical strainer geometries.

Validation testingg includes: (1)

( ) strainer testing, g, (2)

( ) HTVL testing, g, and (3)

( ) UNM vertical column testing.

g Strainer testing did include prototypical STP geometry, and one series of HTVL tests was patterned after the flume test conditions, debris types and debris loading (e) Validation testing performed in vertical loops does not simulate potentially important aspects of debris bed formation under plant conditions See (e) above. Also, uniform beds constructed in vertical column configurations provide a more consistent basis f validation for lid ti and d more conservative ti results lt (f) Conclusions from early testing must be limited Confirmatory analysis is primarily focused on recent data (since 2010) with sufficient documentation to demonstrate applicability to plant conditions.

16

SSIB Head Loss RAI 18 Head d Loss and d Chemical h i l Effects ff Bump Up (a) Provide justification that homogeneous beds represent of the plant Mi ing flo Mixing flowss and migration processes tend to homogenize homogeni e any an strata that might initially initiall be formed by sequential arrival.

(b) Describe why the LDFG density assumption is valid and why it does not significantly affect the results A sensitivity test using the sludge limit as a plausible compression condition resulted in a CDF increase by a factor of 1.8 (c) Explain how the NUREG/CR-6224 correlation compression function is applied The STP LAR intends to impose sludge-limit packing density for all cases. This was not implemented and a 1.8X increase would be expected when implemented (d) Explain why mass weighting is acceptable Calculations between mass and volume weighting show no significant increase in CDF with STP-specific inputs (e) Provide the potential ranges of packing factors for coating materials All coating i materialsi l h have a similar i il packing ki ffraction i to acrylic li coatings i (0 (0.39 39 as d described ib d iin STP LAR Enclosure 4-3, Reference 24) is reasonable because the constituents are comparable in size, approximately 10 microns. Non-coating particulate debris was assumed to have a packing fraction similar to iron oxide sludge (0.20). NUREG/CR-6224 cites the packed density of iron oxide sludge as 65 lb/ft3.

17

APLAB RAIs

• CASA Grande - Plant Configuration: RAI 1b 1b, 2b, 3

• HRA: RAI 3, 3 5

• Uncertainties: RAI 1, 2, 4, 5, 6

• Stable end state: PRA Success RAI 3c

• Use of different distributions 18

APLAB RAI 1b CASA Grande d - Plant l Configuration f

• Provide o de a tec technical ca justification just cat o for o us usingg oonlyy nominal values or calculate core damage frequency (CDF), large early release frequency

(

(LERF),

) d delta-CDF l (

(CDF), ) anddd delta-LERF l (

(LERF)

)

using time-temperature curves that maximize the probability of sump and core blockage for the entire assumed duration of the event

- The use of nominal ppool temperature p p profiles is consistent with a holistic risk-informed approach and provides a more realistic evaluation of risk.

19

APLAB RAI 2b CASA Grande d - Plant l Configuration f

• Provide a technical justification for assuming only nominal operating conditions or calculate CDF, LERF, CDF, and LERF using flow rates or other thermal-hydraulic conditions that maximize the probability of sump and core bl k blockage ffor the h entire i assumed ddduration i off the event

- Using U i nominal i l values l provides id results lt th thatt are reasonable, probable and that may be used in a holistic,, risk-informed evaluation 20

APLAB RAI 3 CASA Grande d - Plant l Configuration f

• (a)

(a),(b),(c)

(b) (c) Justify a combination of pumps failing in the same train is worse than the same set of pumps failing in different trains and clarify if an engineering analysis was performed in support of this assumption

- Details documenting the rationale for selecting pump state 22 to demonstrate the impact of the assumption, the analysis approach and results are found in Enclosure 1 21

APLAB RAI 3 Human Reliability l b l Analysis l (a) State if the CASA Grande models the plant conditions (e.g., sump flow rates, washdown rates,rates refueling water storage tank (RWST) drain-downdrain down times times, etc.)

etc ) that would occur if three containment spray trains were running (i.e., if the manual actions modeled by top event OSI are unsuccessful.)

Statistical sampling does not preclude selection of a very long task performance time that effectively represents failure of manual action but no strategies are employed to ensure that this condition is always represented in the statistical design (b) State if the CASA Grande models the plant conditions (e.g., sump flow rates, washdown rates, RWST drain-down times, etc.) that would occur if the operators fail to secure containment spray long term once containment pressure and iodine levels are suitably low (i (i.e.,

e the manual actions associated with OFFS are unsuccessful)

CASA Grande does not model the plant conditions that would occur if the operators fail to secure containment spray long term once containment pressure and iodine levels are suitably low (c) Provide a technical basis and explain how the PRA meets the ASME HLR-HR-G requirement The PRA model does include logic to represent failure to trip one running containment spray pump as well as failure to trip all containment spray pumps late in the sequence. However, there are no results from CASA Grande that are representative of these failure conditions 22

APLAB RAI 5 Human Reliability l b l Analysis l

• Please explain how the CASA Grande results were developed to address the various combinations of success and failure of these operator actions. Please also explain p how the consistencyy between the actual PRA scenario and the GSI-191 basic event failure probabilities developed in CASA Grande was assured

- The operator p actions on Page g 37 of Volume 3 are listed as:

1. Securing one Containment Spray System (CSS) pump if all three CSS pumps are successfully initiated

2. Securing all CSS pumps later in the event

3. Switchover to Emergency Core Cooling System (ECCS) sump recirculation after the Refueling Water Storage Tank (RWST) has been drained

4. Switchover to hot leg injection 23

APLAB RAI 1 Results l Interpretation-Uncertainty i i Analysisl i (a) Please identify all sources of key model uncertainty as defined by RG 1.200 1 200 Success criteria for fuel blockage and boron precipitation, Fiber penetration of the sump strainer, Head loss correlation at sump strainer, Debris ggeneration,, includingg size and shape p of zone of influence,, Debris transport to the sump, Ability of chemical precipitates to cause increased strainer head loss and fuel blockage (b) Please identify the key assumptions as defined by RG 1.200 T bl provided Table id d (c) Describe the potential effect of the key assumptions on the results Items addressed: discrete pump operability states; failure conditions at one sump; head loss correlation usedused, including bump bump up factors factors , is conservative; boron precipitation following Medium cold leg breaks is likely conservative; breaks occur on a leg equipped with SI yields a slightly conservative result; 24

APLAB RAI 2 Results l Interpretation-Uncertainty i i Analysis l i

• Provide CDF CDF, LERF LERF, CDF, CDF and LERF using the arithmetic mean aggregation of LOCA frequencies in NUREG-1829 NUREG 1829

- CASA has not been rerun, initially, Delta cdf is 1.54e-7 1.54e 7 and delta lerf is 2.56e 2.56e-10 10 25

APLAB RAI 4 Results l Interpretation-Uncertainty i i Analysisl i (a) Explain why the STP evaluation departs from the regulatory position in RG 1.174 regarding the use of mean values Mean values are used by the PRA for the initiating event frequencies, CDF and LERF actually use the mean values from NUREG-1829 (b) Provide a technical justification for the selection of the scale parameter Increasing the scale factor to a factor of 100 times the 95th percentile of the frequencies elicited from experts in NUREG-1829 produces less than a 2% increase in CDF (c) Provide the maximum expected difference between the CDF, LERF, CDF, and LERF developed from bounded Johnson distributions that consider alternative values of the scale parameter Estimates of CDF and LERF differ from CDF and LERF by values that do not depend on (d) Justify the apparent use of different bounded Johnson distributions in the PRA and CASA Grande The distributions derived from the fitted bounded Johnson distributions were scaled for use in the PRA This was done to match the resulting mean values with the means interpolated directly from PRA.

NUREG-1829 26

APLAB RAI 5 Resultsl Interpretation-Uncertainty i i Analysis l i

• Either calculate CDFCDF, LERF LERF, CDF CDF, and LERF accounting for the state-of-knowledge correlation or demonstrate that it is unimportant to this application

- Dependence of the PRA and CASA Grande on different parameters of the LOCA break frequencies q is sufficient so as to not warrant correlation between the PRA and CASA Grande 27

APLAB RAI 6 Resultsl Interpretation-Uncertainty i i Analysis l i

• Provide the results of an aggregate gg g analysis y that quantifies the integrated impact on CDF, LERF, CDF, and LERF from all sensitivity studies that were performed

- A sensitivity analysis was performed by 1) developing the scope of potentially important contributors to CDF and then 2) analyzing their individual contributions in a one-way sensitivity study. The study was then expanded to include aggregate contributions from the two highest contributors contributors, fiber penetration through the emergency core cooling system strainers and the success criteria for boron precipitation (boron fiber limit) 28

APLAB RAI 3c PRA Success

• State what plant conditions and configuration is assumed for the safe, stable end state in the PRA model.

state model Please describe

- To reach the one, safe stable end state in response to a medium or large LOCA requires

• Successful reactor trip,

• Successful safety injection actuation,

• MSIV closure or turbine trip or the reactor withstanding a potential PTS overcooling challenge,

• Sufficient accumulator injection,

• Low head pump injection to the RCS

• High head pump injection (not required if a large LOCA)

• Low head pumps in sump recirculation mode

• Sump available for recirculation considering GSI-191 issues

• No in-vessel flow blockage

• No boron precipitation leading to loss of core cooling, and

• Decay heat removal by either the RHR heat exchangers or the containment fan coolers

- The most likely safe, stable end state is that all containment fan coolers are operating and that cooling to the secondary side of each RHR heat exchanger aligned for low head pump sump recirculation is available. These are the only safe, stable end states credited in the STP PRA for medium and large LOCAs 29

ESGB RAIs

• Coatings: RAI 1 1, 2 2, 6

• Chemical Effects: RAI 1 30

ESGB RAI 1 Coatings

• Specify the epoxy coating in question and provide a basis (i.e., testing) for assuming it fails in pieces larger than fines

- The unqualified coatings size distribution of Table 2.2.18 (LAR Enclosure 4-3) was taken from LAR E l Enclosure 4 4-3, 3 Reference R f [12],

[12] Table T bl 7 7, Page P 32.

32 This calculation references the epoxy size distributions from ppaint chip p characterization of DBA coatings testing document TXU Paint Chip Characterization 31

ESGB RAI 2 Coatings

• The ZOI used to calculate these quantities is not provided. Please provide the ZOI used for both epoxy and inorganic zinc (IOZ) qualified coatings (e.g. epoxy

= 4D, IOZ = 10D)

- A ZOI was created for each of the four different break sizes i th in three b bounding di llocations ti which hi h d determine t i a realistic li ti maximum amount of surface area for the different epoxy, polyamide primer, and IOZ coatings. Due to the different inner diameters of pipes, the 4D ZOI radius is used for the epoxy and inorganic zinc (IOZ) qualified coatings as suggested by WCAP WCAP-16568-P 16568 P 32

ESGB RAI 6 Coatings (a) Describe what STP has done in terms of documentation review or testing of plant materials in order to ensure that the plant-specific plant specific unqualified coatings at STP are the same as the coatings used in the EPRI testing The generic type of each unqualified coating (i.e. epoxy, alkyd, etc.) is documented in the STP unqualified coatings inventory log (LAR Enclosure 4-3, Reference [12]). The specific product description, however, is unavailable for many unqualified coatings. Product descriptions are also unavailable il bl for f many off the th coatings ti ttested t d iin th the EPRI study t d (LAR EEnclosure l 5 5, R Reference f [3])

[3]), th the extent of comparison is made to applicable generic coating types that are available (b) The testers stated that they made no attempt to quantify debris on the filters, please provide additional justification for using this test data to assign a failure time to unqualified coatings Interpretation of the data implies that ranking filters by discoloration (visually dominated by heavily pigmented alkyds) may conservatively bias inferred failure timing to the maximum unqualified coatings failure rate of alkyds. A single estimated failure rate was applied for all upper-containment unqualified coatings types in the STP LAR Enclosure 4-3 analysis (c) Provide additional justification for the current analysis or provide a revised value for the failure timing Because alkyd coatings have the greatest influence on subjective interpretation of photographs (by virtue of distinctive coloration), and have the highest average substrate detachment, inferred failure rates can only be biased towards the maximum unqualified coatings failure rate of alkyds.

S di See discussion i iin response tto ESGB RAI 6b 33

ESGB RAI 1 Chemical h l Effects ff (a) Based on the NRC staffs experience observing testing, head loss for a given quantity of chemical precipitates should be related to both the type of precipitate and the filtering characteristics of the debris bed Although chemical effects have not been found to be significant contributors to STP post-LOCA sump conditions, CASA Grande evaluates a chemical contribution based on break size to provide safety margin in the analysis (b) A relative frequency plot of chemical effects as a histogram showing chemical head loss (feet) on the x-axis and number of occurrences on the y-axis would be very useful to the staff Histograms of chemical head loss vs frequency for CASA Grande Case 01 (all equipment operates) are presented by Figures 1 and 2 (c) Please supply the basis for choosing the exponential form of the PDF over others The single-parameter exponential PDF was chosen for shape and for convenience of fitting the desired statistics of the mean and a truncated tail probability (d) In general general, finer fiber beds tend tolead to greater head loss. loss Means of the exponential PDFs were determined from evaluation of STP ECCS strainer testing (LAR Enclosure 4-3, Reference [53]) as discussed in the ESGB RAI 1c response 34

ACRS Preparation

• Discuss ACRS Topics to be presented and discussed with ACRWS 35

Audit Plan

• Discuss scope of audit

• Agree on audit location and dates 36