NOC-AE-11002687, Response to Request for Additional Information for the South Texas Project License Renewal Application

Response to Request for Additional Information for the South Texas Project License Renewal Application

ML11193A016
Person / Time
Site:	South Texas
Issue date:	07/05/2011
From:	Gerry Powell South Texas
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
NOC-AE-11002687, TAC ME4938
Download: ML11193A016 (66)

Text

Nuclear Operating Company South Texas Prolect ElectricGeneratinS Station P. Box 289 Wadsworth, Texas 77483 - A AvA -

July 5, 2011 NOC-AE-1 1002687 10CFR54 STI: 32889793 File: G25 U. S. Nuclear Regulatory Commission Attention: Document Control Desk One White Flint North 11555 Rockville Pike Rockville, MD 20852-2746 South Texas Project Units 1 and 2 Docket Nos. STN 50-498, STN 50-499 Response to Request for Additional Information for the South Texas Proiect License Renewal Application (TAC No. ME4938)

Reference:

1. STPNOC Letter dated October 25, 2010, from G. T. Powell to NRC Document Control Desk, "License Renewal Application", (NOC-AE-10002607) (ML103010257)

2. NRC letter dated May 31, 2011, "Request for Additional Information for the Review of the South Texas Project, License Renewal Application (ML11140A015)

By Reference 1, STP Nuclear Operating Company (STPNOC) submitted the License Renewal Application (LRA) for South Texas Project (STP) Units 1 and 2. By Reference 2, the NRC staff requested additional information for the review of the STP LRA. STPNOC's response to the request for additional information is included in the Enclosure to this letter.

There are no regulatory commitments in this letter.

Should you have any questions regarding this letter, please contact either Arden Aldridge, STP License Renewal Project Lead, at (361) 972-8243 or Ken Taplett, STP License Renewal Project regulatory point-of-contact, at (361) 972-8416.

I declare under penalty of perjury that the foregoing is true and correct.

Executed on ,T1,4 6. ,otf bat6 G. T. Powell Vice President, Technical Support & Oversight KJT

Enclosure:

STPNOC Response to Request for Additional Information Ai Q7 om

NOC-AE-1 1002687 Page 2 cc:

(paper copy) (electronic copy)

Regional Administrator, Region IV A. H. Gutterman, Esquire U. S. Nuclear Regulatory Commission Kathryn M. Sutton, Esquire 612 East Lamar Blvd, Suite 400 Morgan, Lewis & Bockius, LLP Arlington, Texas 76011-4125 Balwant K. Singal John Ragan Senior Project Manager Catherine Callaway U.S. Nuclear Regulatory Commission Jim von Suskil One White Flint North (MS 8B1) NRG South Texas LP 11555 Rockville Pike Rockville, MD 20852 Ed Alarcon Senior Resident Inspector Kevin Polio U. S. Nuclear Regulatory Commission Richard Pena P. 0. Box 289, Mail Code: MN1 16 City Public Service Wadsworth, TX 77483 C. M. Canady Peter Nemeth City of Austin Crain Caton & James, P.C.

Electric Utility Department 721 Barton Springs Road C. Mele Austin, TX 78704 City of Austin John W. Daily Richard A. Ratliff License Renewal Project Manager (Safety) Alice Rogers U.S. Nuclear Regulatory Commission Texas Department of State Health Services One White Flint North (MS 011-Fl)

Washington, DC 20555-0001 Balwant K. Singal Tam Tran John W. Daily License Renewal Project Manager Tam Tran (Environmental) U. S. Nuclear Regulatory Commission U. S. Nuclear Regulatory Commission One White Flint North (MS O11F01)

Washington, DC 20555-0001

Enclosure 1 NOC-AE-1 1002687 Enclosure STPNOC Response to Request for Additional Information

Enclosure NOC-AE-1 1002687 Page 1 of 63 STPNOC Response to Request for Additional Information SOUTH TEXAS PROJECT LICENSE RENEWAL APPLICATION REQUESTS FOR ADDITIONAL INFORMATION REGARDING THE ANALYSIS OF SEVERE ACCIDENT MITIGATION ALTERNATIVES NRC Requested Information:

1. Provide the following information regarding the Probabilistic Risk Assessment (PRA) used for the Severe Accident Mitigation Alternative (SAMA) analysis:

a. Environmental Report (ER) Section F.2 states that the current PRA model (STPREV6) reflects the plant design configuration as of December 31, 2007.

Confirm that this applies to the analysis of all initiating events, both internal and external. Describe any significant changes made to plant design or operation since that date and their impact on the SAMA analysis.

STPNOC Response:

The review of STPREV6 Initiating Events Notebook confirmed that it reflects the plant design configuration and operating history as of December 31, 2007, including internal and external events. Table 5.4 of the Initiating Events Notebook details the plant operating history updated in STPREV6.

The review of plant design changes to be included in STPREV7 (through December 31, 2010) identified one plant modification that will require a revision to the PRA model currently being updated. The steam generator power-operated relief valves (PORV) are now failed closed upon loss of alternating current (AC) power to the hydraulic pumps that charge the accumulators for providing motive power to the PORVs to comply with accident analysis assumptions. An operator action is now required to allow the accumulators to operate the PORVs upon loss of AC power to the SG PORV hydraulic pumps. Although a final analysis is not complete yet, a Human Reliability Analysis will be performed to evaluate adding this operator action to the upcoming PRA model revision. It is expected that this design change will not significantly impact the PRA model results or the SAMA results currently under review.

NRC Requested Information:

b. The South Texas Project (STP) PRA appears to be a single unit model. Identify any significant design or operating differences between STP, Units I and 2, and, if there are shared systems between units, describe how these systems are

Enclosure NOC-AE-1 1002687 Page 2 of 63 modeled in the PRA. Provide an assessment of the impact of any significant differences between units or shared systems on the SAMA analysis.

STPNOC Response:

The STP PRA model is a single unit model because STP Units 1 and 2 are designed to be identical, and therefore, the STP PRA model results apply to both units. The common switchyard, the Main Cooling Reservoir, and the Essential Cooling Water Pond are shared between STP Units 1 and 2. The units were designed to be identical and have primarily remained so. There are currently two plant differences of significance.

The first difference was identified while reviewing a design change package prior to the STPREV6 update. A design change to implement automatic load tap changers for the Engineered Safety Features (ESF) transformers will be installed in all three safety trains in both units over several years. During the STPREV6 model development and update phase, a sensitivity study was performed to determine the impact to core damage frequency (CDF) and large early release frequency (LERF).

Table 1-1 below shows the overall model result differences. A change of 0.43% for CDF and 0.46% for LERF is small. The differences can be accounted for in loss of main transformer (LOMT) initiating event (-18%). The LOMT is a direct impact to the unit auxiliary transformer. The change for a general alignment scheme of Unit Auxiliary Transformer and Standby Transformer means that loss of the main transformer initiating event is slightly more important.

Table 1-1 OVERALL ESF TRANSFORMER ALIGNMENT SENSITIVITY RESULTS CDF (events/year)

BEFORE AFTER DELTA CDF  % Delta 1.1 5670E-05 1.16170E-05 5.OOOOOE-08 0.43%

LERF BEFORE AFTER DELTA  %

7.0824E-07 7.1153E-07 3.29000E-09 0.46%

Notes BEFORE: Prior to ESF transformer automatic load tap changer design change AFTER: After installing an automatic load tap changer in one of three ESF transformers per unit (i.e. current condition of the design change)

Given that the sensitivity results were so low, it has been determined that no change the Probabilistic Risk Assessment (PRA) model is necessary during the implementation of this design change. Based on the above sensitivity results, it is concluded that these alignment changes will not have significant impact on the SAMA analysis.

The second difference between the units is a much shorter time frame difference. When implementing a design change (see response to RAI question 1 .a) for the steam generator (SG) power-operated relief valves (PORV), it was identified that having control switches for

Enclosure NOC-AE-1 1002687 Page 3 of 63 operating the PORVs in the control room would be beneficial. This additional change was identified too late for implementation in Unit 1 the refueling outage 1RE16. The majority of the design change was implemented in the 1RE16 (April 2011). The entire design change will be implemented in the Unit 2 refueling outage 2RE15 (November 2011) and the additional control switches in the control room will be implemented in Unit 1 during 1RE17 (October 2012). The design changes for the SG PORV will require a human reliability analysis (HRA) for the proceduralized operator action to manually control the SG PORV.

Until the Unit 1 control switches are installed in the control room during 1RE17, the HRA will reflect only local actions in the field. The HRA will be updated accordingly when the additional control switches are installed. The above differences between Units 1 and 2 have minimal impact on the PRA and SAMA analysis.

NRC Requested Information:

c. The listing and description of initiators for the external events in ER Table F.2-1 appear to indicate that there are other initiators evaluated but not listed. For example, three control room fire scenarios (i.e., 18, 23 and 10) are listed that address only three fire zones (i.e., 047, 071, and 147). Discuss the initiators not included in the table and their contribution to core damage frequency (CDF), and assess their potential impact on the SAMA assessment.

STPNOC Response:

Environmental Report (ER) Table F.2-1 contains a complete listing of the external initiating events. Note that control room fires contain a prefix "FR" and fire scenarios for other areas are denoted with a prefix of "Z" followed by the fire zone number and a scenario letter.

Hence the control room fires listed are fire areas FA10, FA18, and FA23. The fire scenario with the highest frequency CDF is named Z047 and models a fire in zone 047. The response to RAI 3.d. contains additional information regarding a fire scenario and its impact.

Fire initiator Z047 is a fire in the cable spreading room train B.

NRC Requested Information:

d. ER Section F.7.1 states that the CDF of 6.39E-06 per year is a mean value from the RISKMAN Monte Carlo quantification. Confirm that all the CDF and release category frequency values given are also mean values. If so, describe why it appears that the sum of the initiating event contributor's mean values reported in Table F.2-1 equal the mean of the total distribution.

STPNOC Response:

The quantification of the Level 1 model results in over 62,000 sequences. Each sequence contains a complete set of failed, succeeded, and bypassed top events. The database cannot store all of the sequences. Thus a reduced set of sequences is used for Monte Carlo analysis to determine data uncertainty characteristics. However, due to the reduced set of saved sequences, the resulting Monte Carlo distribution is scaled so that the mean of

Enclosure NOC-AE-1 1002687 Page 4 of 63 the distribution matched the mean of the CDF point estimate. The CDF of 6.39E-06 per year is a point estimate.

NRC Requested Information:

e. Briefly describe the modeling of the planned and unplanned maintenance conditions assumed for the SAMA analysis. Specifically, indicate if the PRA results used in the SAMA analysis represent the results for the annual average unavailability of systems. If different than this, assess the impact of using the annual average maintenance and testing condition on the SAMA analysis.

STPNOC Response:

The PRA results used in the SAMA analysis represent the results for the annual average unavailability of systems.

The planned and unplanned maintenance conditions modeled in STPREV6 are described by the following information in the Planned Maintenance Event Tree (PMET) Notebook for STPREV6:

For the purposes of modeling in the PRA there are two kinds of maintenance: planned and unplanned. Unplanned maintenance is assumed to occur randomly, except as prevented by Technical Specification rules which would require a shut down. Planned maintenance occurs according to the rolling maintenance schedule and is organized to account for dependencies among systems to minimize system/train outage time.

The general rule for modeling unavailability in STP_REV6 (and previous models) is that planned maintenance is modeled in the event tree module in the PMET. Unplanned maintenance is modeled in systems module as system maintenance/test alignments or into the system fault trees as basic events, so that the effects are included in the master frequency file split fractions.

The majority of the planned maintenance is modeled in the PMET, following the functional equipment groups (FEG) which are organized by work control to account for dependencies among systems to minimize system/train outage time and thereby minimizing risk. At the time that PMET was developed, there were several modeling size constraints. Because of these constraints, only a limited number of FEGs could be modeled in PMET. Any system that does not have its planned maintenance modeled in PMET has unavailability (planned and unplanned) modeled as unplanned maintenance in the systems module.

Enclosure NOC-AE-1 1002687 Page 5 of 63 NRC Requested Information:

f. Provide a brief summary of the history of the STP Level 1 PRA that includes for each revision: the date released, the CDF contribution for internal events and each of the external event hazards [i.e., seismic, fire, tornado, and main cooling reservoir (MCR) breach], and the major changes in the revision that led to the change in the CDF, including identification of major changes or updates to the modeling for various initiator groups such as internal flooding, fire, and seismic.

Also, identify the STP PRA revision reviewed in the 2002 Westinghouse Owners Group (WOG) peer review.

STPNOC Response:

The summary will start with model STP 1999, which was the model peer reviewed by the Westinghouse Owners Group (WOG) in 2002. The model was released in October 2001.

Table 1-2 STP 1999 Core Damage Frequency (CDF) Groupings (eventslyear)

Total CDF Internal Events Contribution External Events Contribution 1.17E-05 8.84E-06 Fires 1.40E-06 Floods 1.41E-08 Flood MCR 2.88E-07 High Winds (i.e. 1.1E-06 tornados)

Seismic 7.29E-08 Total External 2.87E-06 Model STPREV4 was released in September 2003. The major changes that affected CDF and the containment response were:

• The plant specific data updated for train unavailability, initiating events update, and component failure data update

• Incorporation of the latest operator error modeling and improved loss of offsite power (LOOP) recovery modeling.

• Inclusion of Safety Injection accumulator modeling in Large and Medium loss of coolant accident (LOCA) event trees

• Inclusion of hot leg recirculation modeling in the Large LOCA event tree

• Removal of the 150-ton air conditioning chillers, and

• Improved modeling of support system initiating events.

Enclosure NOC-AE-1 1002687 Page 6 of 63 Table 1-3 STPREV4 CDF Groupings (eventslyear)

Total CDF Internal Events Contribution External Events Contribution 1.17E-05 9.08E-06 Fires 1.0E-06 Floods 1.48E-08 Flood MCR 2.88E-07 High Winds (i.e. 1.1E-06

_tornados)

Seismic 7.26E-08 Total External 2.48E-06 In November of 2004 STPRV41 was released. The major changes that affected the CDF and containment response were:

• Addition of "Operator Depressurization" to the Small LOCA event,

• Correction of modeling errors in Medium LOCA Long-Term response model,

• Re-quantification of the initiating event frequency for Inadvertant Opening of One and Two Pressurizer Safety Valves to reflect the failure to reclose in the initiating events,

• Correction of the Conditional Split fraction definitions used in the model to correct errors in the Basic Event importance calculations

• Re-binning of several maintenance duration data variables to correct input problems with RISKMAN Revision 7, and,

• Splitting of fault tree basic events containing several components into individual basic events to prepare for mitigating systems performance indicator implementation and to remove undue conservatism in basic event importance calculations used in the Graded Quality Assurance process.

STPRV41 reported a CDF 1.2% higher than STPREV4. A model STPRV42 was made to correct issues found during component risk ranking using STP_RV41. STPRV42 was released in February of 2005 and showed a negligible increase (9E-08) from STPRV41.

The CDF was 9.28E-06.

STPREV5 was released in September of 2005. This model incorporated plant modifications, procedure changes, and data update through 2004. In addition, modifications to the Class IE Vital AC system and the main steam isolation valves are modeled. The model incorporates a major change in the human reliability analysis (HRA) methods to use the Electric Power Research Institute (EPRI) HRA calculator. A Level 2 analysis update with a revision in the containment capability analysis was also included.

Enclosure NOC-AE-1 1002687 Page 7 of 63 Table 1-4 STP REV5 CDF Groupings (eventslyear)

Total CDF Internal Events Contribution External Events Contribution 1.04E-05 7.67E-06 Fires 9.72E-07 Floods 1.43E-08 Flood MCR 2.88E-07 High Winds (i.e. 1.1E-06 tornados)

Seismic 7.28E-08 Total External 2.73E-06 A STPRV51 was released in July, 2007 but did not result in any change to the CDF or the Large Early Release Frequency.

STP REV6 was released in July of 2008. The revision consisted primarily of data and groupings for planned maintenance and data variables for component failures and initiating events. The process for updating data variables describing component failures was augmented from that used in the past revisions. Consistent with previous revisions, the variables for which component failures occur were updated. Some selected, important variables for which no component failures had occurred were also updated.

Table 1-5 STP REV 6 CDF Groupings (eventslyear)

Total CDF Internal Events Contribution External Events Contribution 6.39E-06 3.89E-06 Fires 1.02E-06 Floods 1.26E-08 Flood MCR 2.90E-07 High Winds (i.e. 1.11E-06 tornados)

Seismic 7.31 E-08 Total External 2.50E-06 NRC Requested Information:

g. STP Nuclear Operating Company (STPNOC) risk managed technical specifications (RMTS) submittal of February 28, 2007, stated that there had been a follow on peer review of the human reliability analysis (HRA) of STPREV5 PRA which had identified one Level A and nine Level B Facts and Observations (F&Os). Describe these F&Os, their resolution status, and the impact of their resolution on the SAMA analysis.

Enclosure NOC-AE-1 1002687 Page 8 of 63 STPNOC Response:

Level A F&O HR-16 There is no evidence that an analyst reviewed the changes to the PRA model incorporated since the Individual Plant Examination (IPE) to decide if the current Human Failure Events (HFE) were adequate and that no new HFEs needed to be added.

Level B F&Os HR-08 The methodology for analyzing the dependence of time-sensitive actions does not identify groups of actions that involve both dynamic actions modeled as part of the system initiator and subsequent dynamic actions in response to the initiator on the failure of actions considered in the system initiator.

HR-09 Table 7-1 (Post Initiator Human Error Probability (HEP)Summary) of the report does not identify some dynamic actions though they are used in the sequence model. Table 7-1 also does not identify which actions are dependent on others or under what conditions.

HR-10 In one case in the HFE model, the importance of the action to start the positive displacement pump without centrifugal charging pumps (HERA6) will not be correctly reflected in the basic event importance report.

HR-13 The scenarios defined for evaluation of risk significant events to trip the reactor coolant pumps (RCP) after loss of component cooling water (HERCP1) and the action to start the positive -displacement pump and manually trip the RCPs (HERC6) appear inconsistent'with the sequence models for which they are used.

HR-14 Some judgments made in the annunciator response model used for evaluation of the failure to place the standby Electrical Auxiliary Building (EAB) HVAC train into service (HEEAB1) are questionable.

Enclosure NOC-AE-1 1002687 Page 9 of 63 HR-15 There is no obvious evidence (e.g. documentation) that a procedure requirement was performed to review model updates to ensure that any human errors that could be plant specific or industry specific are addressed.

HR-17 The step for verifying recovery actions for the failure to initiate residual heat removal cooling for a steam generator tube rupture (HEOC01) needs to be improved.

HR-18 The recovery execution for basic event modeling the failure to initiate bleed and feed cooling (HEOB02) appears to reference an incorrect procedure.

HR-19 The calculation used and the correct cognitive failures need to be identified for the failure to open doors for 2 of 3 EAB HVAC fan trains failed (HEOS01).

Resolution Status of the above F&Os A formal resolution for these F&Os is not complete at this time. The F&O draft resolutions are expected to be complete by the end of 2011. Preliminary review of the F&Os has determined that they are not likely to have a significant impact on the STP PRA model.

Based on this preliminary review of possible model impact, it is not expected that any resolution will impact the SAMA analysis.

NRC Requested Information:

h. Table F.2-1 does not include any internal flood initiators. Discuss the modeling and disposition of internal flood events and their contribution to the CDF.

STPNOC Response:

The internal flood initiators were screened during the Individual Plant Examination of External Events (IPEEE). STP is a relatively recent vintage plant with a high degree of separation between trains. This separation and the three train design is the reason that the flooding initiating events were able to be screened. A review of internal flood screening was performed in support of the Risk-Managed Technical Specification license amendment process and concluded that the previous IPEEE internal flood screening remains valid. This was accepted by the NRC in its original safety evaluation report (SER) (accession number 9201300172) for the STP PRA and in the SER (ML071780186) for Risk-Managed Technical Specifications (RMTS).

Enclosure NOC-AE-1 1002687 Page 10 of 63 NRC Requested Information:

i. Provide the contribution to CDF due to station blackout (SBO) and anticipated transients without scram (ATWS) events.

STPNOC Response:

Group CDF(events/year)  % Contribution SBO 2.23E-06 35.0%

ATWS 2.75E-07 4.3%

Total CDF 6.39E-06 100%

NRC Requested Information:

2. Provide the following information relative to the Level 2 analysis:

a. Provide a summary description of the current Level 2 PRA including: (a) the Level I to Level 2 linking, the containment event trees, the binning of Level 2 sequences to the 15 end-states cited in ER Section F.3.6 and (b) the process used to assign the 15 end-states to the four major release categories.

STPNOC Response:

A detailed description of the Level 2 PRA is provided in the following documents on file at STPEGS:

" STPEGS Probabilistic Risk Assessment Notebook, "Level 2 Analysis - Containment Event Tree" STPREV6, August 18, 2009.

" STPEGS Probabilistic Risk Assessment Notebook, "Level. 2 Accident Sequence Progression," STPREV6, August 8, 2009.

• STPEGS Probabilistic Risk Assessment Notebook, "Level 2 Results" STPREV6, September 14, 2010.

• STPNOC 2005 Level 2 PRA Update Reports (3 Final reports, Revision 0), ABSG Consulting Inc., "Level 2 Probabilistic Risk Assessment Update - 2005,"

"Containment Analysis Report - 2005," and "MAAP (Revision 4.05) Analysis Report,"

October 14, 2005.

• STPEGS Level 2 PSA and Individual Plant Examination, prepared by HL&P and PLG, Inc., August 1992.

The set of documents identified above were specifically developed to address peer review findings and to align the STPEGS Level 2 PRA with American Society of Mechanical Engineers (ASME) PRA Capability Category 2 (or higher) requirements.

The following excerpts from those documents address this RAI:

The entry to the containment event tree (CET) is characterized by the thermal-dynamic conditions in the reactor coolant system and containment at the time of severe core

Enclosure NOC-AE-1 1002687 Page 11 of 63 damage, and the availability of both passive and active plant features that can terminate the accident or mitigate"the release of radioactive materials.

The CET is different from the Level 1 event trees in that it deals primarily with severe accident phenomena rather than the success or failure of equipment and operator actions. This makes the progression of events and status of important physical parameters in the Level 2 analysis much more difficult to determine. The plant response to a severe accident consists of a sequence of complex interrelated physical phenomena that are not always adequately described by success or failure, but are sometimes a matter of degree or timing. The effects of these phenomena must be reduced to binary or multi-branch events in order to be evaluated by PRA methods.

Each accident sequence has a unique combination of top event successes and failures.

Ideally, each accident sequence that results in core damage should be evaluated explicitly in terms of accident progression and the release of radioactive materials to the environment. However, because there can be millions of such sequences, it is impractical to perform such analyses for each one. Therefore the individual sequences must be categorized into bins by some set of parameters that define the group. Each bin collects all of those sequences for which the progression of core damage, the release of fission products from the fuel, the status of the containment and its systems, and the potential for mitigating source terms are similar. Detailed analysis is then focused on specific sequences selected to represent each of these bins.

All of the plant model information on the operability status of active systems that is important to the timing and magnitude of the release of radioactive materials must be passed into the CET. This requires that, in addition to representing the systems and functions that are important to core cooling, the Level 1 event trees also address active systems and functions important to containment isolation, containment heat removal, and the removal of radioactivity from the containment atmosphere. The containment spray system is a good example of such systems.

The concept of "plant damage states" used in NUREG-1 150, "Severe Accident Risks:

An Assessment for Five U. S. Nuclear Power Plants," and discussed in the STP Individual Plant Examination (IPE) has been discarded. As far as RISKMAN is concerned, the CET is another tree just like the ones in the Level 1 part of the analysis.

Therefore the code passes the status of all the top events previously evaluated on to the CET as well as the values of all the macros. There is no need to define transitional states as the CET can directly determine the status of the plant through the application of split fraction rules.

The CET considers the influence of physical and chemical processes on the integrity of the containment and on the release of fission products once core damage has occurred.

The considerations that influence the progression of core damage, the time and mode of containment failure, and the release of radioactive materials to the environment fall into two categories:

" the physical conditions in the reactor coolant system (RCS) and containment at the time of core damage, fission product release and vessel breach, and

• the status and availability of containment systems for mitigating fission product release and removing decay heat.

Enclosure NOC-AE-1 1002687 Page 12 of 63 The considerations of physical conditions in the RCS and containment that are included in the CET are as follows:

• the pressure inside the reactor vessel at the onset of core damage and at vessel breach, .

• the availability of cooling on the secondary side of the steam generators, and

• whether or not the reactor cavity is flooded at the time of vessel melt-through.

The pressure inside the RCS at the onset of core damage is an important parameter because it influences the pressure at the time of vessel breach. The pressure at vessel breach is also a strong function of the core degradation process. If the reactor vessel fails by rupture of the bottom head at high RCS pressure, core debris will be dispersed beyond the confines of the reactor cavity. The pressure at vessel breach depends on the core damage process and whether or not any natural or operator induced mechanisms reduce RCS pressure as the accident sequence progresses. If the pressure at vessel breach exceeds approximately 200 psia, there exists a potential for ejection of dispersed core debris into the containment atmosphere, thereby increasing the containment loading at the time of vessel failure. High RCS pressures (particularly at the system set point) can be conducive to significant natural circulation in the RCS.

While natural circulation has beneficial effects for most transients; in a severe accident, natural circulation can transport hot gases from the core into the steam generators inducing tube rupture and a possible direct pathway to the environment. The availability of steam generator secondary side cooling will determine whether or not the steam generator tubes will be subject to high temperatures and potential failure if combined with high RCS pressure.

The presence of water in the reactor cavity at the time of reactor vessel melt-through is important to containment response because the interaction of this water with hot core debris can:

" fragment and disperse the core debris from the reactor cavity into other regions of the containment,

• cause the containment pressure to increase by vaporization of the water (i.e.,

steam spikes) and direct heating of the containment atmosphere, and

• enhance the release of fission products from the core debris due to oxidation of the particulates.

Functional containment status considerations included in the CET are:

" The state of the containment itself (intact, bypassed or failed) at the time when severe core damage starts (i.e., when the CET is entered). This distinction not only includes containment isolation failure and bypass considerations, but is also of particular importance for external events that can cause containment failure prior to core damage (e.g., earthquakes, severe storms, or external missiles).

" The availability of containment engineered safety features (such as containment sprays and fan coolers) for cooling the containment atmosphere and fission product removal before and after failure of the reactor vessel.

• The potential availability of filtration and/or other mechanisms for fission product removal in the containment leakage path (such as auxiliary building filters for

Enclosure NOC-AE-1 1002687 Page 13 of 63 interfacing systems loss-of-coolant accident (LOCA) or purge filters for sequences involving isolation failure) if the containment is failed at the time core damage is initiated.

Based on a review of the STP design and reference plant documents, the following specific items are considered for entry into the containment event tree:

RCS Pressure. The following four ranges of RCS pressure (P) at the time of core damage have been defined:

• P<200psia

• 200< P < 600 psia

• 600 < P < 2000 psia 0 P > 2000 psia (i.e., "pegged" at the system set point)

The primary system pressure is assumed to be a function of the following parameters that are determined in the Level 1 analysis.

" Availability of steam generator cooling.

• Type and size of LOCA.

• Position of pressurizer power-operated relief valves (PORV).

" Status of high head safety injection (HHSI) system.

• Success of RCS cooldown.

The assumed relationship between these parameters and RCS pressure at the time of core damage is shown in Table 2-1 on the following page.

Enclosure NOC-AE-1 1002687 Page 14 of 63 Table 2-1 Assignment of RCS Pressure at the Time of Core Damage for Sequences with Reactor Trip Successful Steam LOCA Size Two HHSI Secondary RCS Generator Pressurizer Available Depressuriz Pressure Cooling PORVs ation at Available Open IRCS Time of Cooldown Core Damage Yes RCP Seals only N/A Yes Yes Medium Yes RCP Seals only N/A Yes No High Yes RCP Seals only N/A No Yes High Yes RCP Seals only N/A No No High Yes Pressurizer PORV N/A Yes Yes Medium or Ruptured 1 Steam Generator( )

Yes Pressurizer PORV N/A Yes No High or Ruptured Steam Generator(1 _

Yes Pressurizer PORV N/A No Yes High or Ruptured Steam Generator(1 )

Yes Pressurizer PORV N/A No No High or Ruptured Steam Generator(1 )

No None No(2) N/A No(3) System Setpoint(4)

No RCP Seals only No(2) N/A No(3) System Setpoint No Pressurizer PORV No(z) N/A No(3 ) High or Ruptured Steam Generator-__

No N/A Yes(5) Yes(5 ) No3 High (1) The ruptured steam generator is not isolated so that leakage of reactor coolant through the secondary side to the environment occurs. HHSI is considered unavailable in the long term for such circumstances if it is initially successful, but makeup to the reactor water storage tank (RWST) is not provided.

(2) Bleed and feed cooling unsuccessful.

(3) No secondary depressurization possible if steam generator cooling unsuccessful.

(4) At or above the pressurizer PORV setpoint.

(5) Bleed and feed cooling successful.

Enclosure NOC-AE-1 1002687 Page 15 of 63 Steam Generator Heat Removal. Auxiliary feedwater is providing secondary side heat removal with or without secondary pressure control and the ability to cool down.

Water in Cavity Prior to Vessel Breach. This parameter addresses whether or not the contents of the RWST have been injected into the containment to "spill" into the reactor cavity.

Containment Isolation and Bypass Status. The following five situations are considered:

" Containment isolated and not bypassed.

" Containment not isolated or failed prior to core damage; leak area less than the equivalent of 3 inches in diameter.

• Containment not isolated with a leak area greater than or equal to the equivalent of 3 inches in diameter. The specific event causing this is a failure of the supplementary containment purge valve to close.

• Small containment bypass (i.e. a steam generator tube rupture (SGTR) or letdown isolation failure).

" Large containment bypass. This is an interfacing LOCA through the residual heat removal (RHR) or low head safety injection (LHSI) system.

Containment Spray Operation. Three combinations of containment spray (injection and recirculation) have been considered:

" Containment Spray Injection (CSI) and Containment Spray Recirculation (CSR) are available.

• Only CSI is available.

" Neither CSI nor CSR is available.

Containment Heat Removal. Heat is removed from the containment via the RHR heat exchangers in conjunction with LHSI or via the containment fan coolers.

Initiatinq Events In addition to the phenomena related "thresholds" discussed above, the classification of pressurized-water reactor (PWR) accident sequences into RCS pressure ranges at the onset of core damage has, in past PRA studies, been correlated to initiating event type, as shown in Table 2-2 below:

Enclosure NOC-AE-1 1002687 Page 16 of 63 Table 2-2 RCS Initiating Event Type Pressure Transients without vessel breach PORV Setpoint Transients with vessel breach (e.g., High Stuck open PORVs, RCP seal LOCA)

Small LOCAs High Medium LOCAs Low Large LOCAs Low The availability of steam generator secondary side cooling will determine whether or not the steam generator tubes will be subject to high temperatures and potential failure if combined with high RCS pressure.

To insure consistency between the Level 1 and Level 2 Probabilistic Safety Assessment (PSA), the same initiator definitions are used with the same lists of events trees. This is accomplished in RISKMAN by creating a new set of initiators for Level 2 that are identical to the Level 1 initiators, except that the CET is added to each Level 2 initiator event tree list.

The concept of a single containment event tree for all initiators is a difficult one to implement in RISKMAN because there are limitations in the RISKMANO program that make the transition from the many Level 1 trees to the one CET difficult. The reason is that each accident initiator evaluates its own particular series of event trees, which means that a particular top event may be evaluated for one initiator, but not for another initiator that takes a differen't path. However RISKMANe requires that in a tree all the referenced top events, even if not used for the current initiator, be previously evaluated.

Therefore the CET cannot reference top events that are not universally common to all sequences without producing RISKMAN execution errors.

The way this problem is handled is to construct "linking" macros, whose purpose is to convert top events statuses into macros, which can be universally defined. Each macro may have different definitions in different trees depending on which initiators use that tree. For initiators that do not need a particular macro, that macro can be set to a dummy value just to satisfy RISKMANO. As an alternative, the linking tree may contain dummy top events just so they can be referenced. The linking macros in the STP PRA model are in the plant damage state or "PDS" trees. These are the last trees in the Level 1 analysis and contain only the Level 1 success criteria and the links to the CET.

In the PDS trees, the SUCC "success" macro defines those sequences which do not lead to core damage. When the Level 1 analysis is used alone, this macro is used in the core damage binning rule, and when the Level 2 is added, the macro is used in the first top event of the CET to filter out non-core damage sequences.

Enclosure NOC-AE-1 1002687 Page 17 of 63 The CET considers the progression of severe accidents beginning with the onset of core damage and ending when a stable state is achieved. It addresses the events and physical processes that are important in determining the time, cause, and mode of containment failure, and the resultant release (via release categories) of radioactive fission products into the environment. The entry conditions for CET analyses are the Level 1 core damage sequences.

The STP CET addresses events occurring prior to vessel breach (including the potential for in-vessel recovery of the damaged core); the phenomena associated with both in-vessel and ex-vessel progression of the accident; containment integrity challenges; and the potential for containment failure. If containment failure does occur, the timing and mode (i.e., a small, controlled leak or a large break and the location of such failures) of failure are also addressed.

A detailed description of release category assignment to major release group is provided in the Containment Event Tree Notebook, and is summarized in Table 2-3 below.

Enclosure NOC-AE-1 1002687 Page 18 of 63 Table 2-3. Release Category Assignment Matrix RCS PRESSURE @ SPRAYS EX- CONTAINMENT FAILURE MODE Major RELEASE VB I=INJECTION VESSEL Release CATEGORY R=RECIRC DEBRIS Group COOLING PRE- EARLY LATE BYPASS None HI MED LO NO EXIST HI 1 EI1L1 VBI _ NONENYES NO EISMA LARsMA LARISMAi LARISMAJ VSEQ X X X X I ISGTR X X X I R01 X X X X I RO1U X X X X R02 X X_- -X X X I R02U X X-_-X X X I R03 X- -X X X X I R03U X- -_X X X X I R04 X- -X X- -X X X I R04U X- -X X- -X X __ X I BYCV X- -X- -X X II CICV X X R05 X- -X X- -X X X II R05S X- I -X X- -X X X II R05L X- -X X- -X X X X II R05SL X- -X X- -X X X X II R05SU X- -X X- - _X X X II R05SU X- -X X- -X _X X I

Enclosure NOC-AE-1 1002687 Page 19 of 63 Table 2-3. Release Category Assignment Matrix RCS PRESSURE @ SPRAYS EX- CONTAINMENT FAILURE MODE Major RELEASE VB I=INJECTION VESSEL Release CATEGORY R=RECIRC DEBRIS Group COOLING PRE- EARLY LATE BYPASS None HI MED LO NO EXIST VB I+R I NONE YES NO SMA LAR SMA LAR SMA LAR SMA RO5LU X- -X X- -X X X X II RO5SLU X- -X X- -X X X X II R06 X- -X X X X II R06S X- -X _X X X II R06L X- -X X X X X II R06SL X- -X X X X X II R06U X- -X X X X II R06SU X- -X X X X II R06LU X- -X X X X X II R06SLU X- -X X X X X II R07 X X- -X X X II R07S X X- -X X X II R07L X X- -X X X X II R07SL X X- -X X X X II R07U X X- -X X X- _II R07SU X X- -X X X II R07LU X X- -X X X X II R07SLU X X- -X X X X II

Enclosure NOC-AE-1 1002687 Page 20 of 63 Table 2-3. Release Category Assignment Matrix RCS PRESSURE @ SPRAYS EX- CONTAINMENT FAILURE MODE Major RELEASE VB I=INJECTION VESSEL Release CATEGORY R=RECIRC DEBRIS Group COOLING PRE- EARLY LATE BYPASS None HI MED LO NO EXIST VB I+R I NONE YES NO SMA LAR SMA LAR SMA LAR SMA_

R08 X X X X II R08S X X X X II R08L X X X X X II R08SL X X X X X II R08U X X X X II R08SU X X X X II R08LU X X X X X II R08SLU X X X X X II BYNCV X- -X- -X X III R09 X- -X X X X IliA R09U X- -X X X X AIlA R10 X- -X X- -X X X AliA R1LI X- _X X_ IX IX X 11Aa R1l X X X X 111A R 111 X X X X 111A R12 X X_ _X X X IlA R12U X X- -X X X liA R13 X- -X X- -X X X liA R13U X- -X X- -X X XliA X

Enclosure NOC-AE-1 1002687 Page 21 of 63 Table 2-3. Release Category Assignment Matrix RCS PRESSURE @ SPRAYS EX- CONTAINMENT FAILURE MODE Major RELEASE VB I=INJECTION VESSEL Release CATEGORY R=RECIRC DEBRIS Group COOLING PRE- EARLY LATE BYPASS None HI MED LO NO EXIST VB I I+R I NONE YES NO SMA LAR SMA LAR SMA LAR SMA R14 X- -X X X X IliA R14U X- -X X X X IliA R15 X X- -X X X IliA R15U X X- -_X X X IliA R16 X X X X IliA R16U X X X X IliA INTACT1 X- -X- -X_X X IV INTACT2 X X- -X- -X X IV NOTES to Table 2-3:

The symbols -X, -X-, and X- appear in the columns for RCS pressure and Sprays. The RCS pressure columns appear in two ways:

1. X- adjacent to -X with the other column blank. In the RCS pressure columns this means the pressure is between the two corresponding classes. For example R14 is at high-medium pressure at vessel breach. In the containment spray columns, this means either column is sufficient.

2. If all three symbols appear, then the variable (pressure or spray) is irrelevant.

Enclosure NOC-AE-1 1002687 Page 22 of 63 NOTES to Table 2-3 (Continued):

3. The last column shown in Table 2-3 is the major release group. The major release group, which has been used as a very informative high level plant risk characterization index, classifies sets or bins of large-early releases (Group I), small-early releases (Group II), late releases (Group Ill), and the benign case of no containment failure (Group IV). Of the major release categories, Group 1, the measure of which is large early release frequency (LERF), is the most important, because it has the greatest potential for adverse health effects. Large releases are those associated with a breach of containment equivalent to a 3-inch line or larger. This includes failure of the supplementary purge valve, a V-sequence break through the LHSI lines, and all energetic mechanisms such as failure coincident with vessel breach or hydrogen detonation. In addition, induced steam generator tube ruptures are assumed to be large releases because of the large potential driving head to pump out fission products provided by system pressure. As stated before, early releases are those that occur within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> of vessel breach. All top events connected with containment failure up through CE are classified as early.

Enclosure NOC-AE-1 1002687 Page 23 of 63 A large number of Modular Accident Analysis Program (MAAP) cases were run in support of the STP Level 2 PSA and IPE. Sensitivity studies were performed that show how the plant responds to different reactor coolant pump seal LOCAs, a spectrum of LOCA break sizes, and various differing core debris spread areas and numbers of operational reactor containment fan coolers. These cases are documented in Section 4.6 of the IPE report.

Within the bounds of the assumptions made, these analyses are still valid. What has changed in recent code revisions starting with the STP_1997 model is primarily associated with the frequency and importance of the various accident sequences as predicted by RISKMAN, rather than differences in the accident progressions as determined by MAAP.

The probabilistic portion of the STP Level 2 analysis for model STPREV6 is documented in the Containment Event Tree Notebook. In that analysis, the RISKMAN code is used to quantify the reference PRA Level 2 model and generate reports showing the end state frequencies and most important sequences.

The accident progressions leading to some of the highest frequency end states are verified with thermal-hydraulic analyses presented in the Accident Sequence Progression Notebook. At the same time, radioactive release fractions are generated for these cases. The end states selected, along with their frequencies, are shown in Table 2-4 below. The selected end states represent greater than greater than 80 percent of the frequencies of major release groups 1,11, and III respectively.

Table 2-4 Re presentative End States Analyzed in MAAP End State Group End State Description ISGTR I Induced Steam Generator Tube Rupture R05SU II RCS High Pressure, Pre-Existing Leak, No Debris Cooling CICV II Isolation Failure, Damage Arrested In-Vessel R07SU II RCS Low Pressure, Pre-Existing Leak, No Debris Cooling R15U III No Cooling, RCS at Low Pressure, Small Reactor Containment Building (RCB) Failure BYPASS III Small Containment Bypass R13U III No Cooling, RCS at High Pressure, Small RCB Failure R11U III No Cooling, RCS at Low Pressure, Large RCB Failure Accident sequences that lead to these end states were chosen for analysis. According to the probabilistic analysis, external events and support system failures dominate risk of release. Many of these initiators result in the loss of Class 1 E power so that they are effectively station blackouts. Therefore all the sequences selected for analysis, except one, are initiated in MAAP with loss of AC power. They are different in the timing of power recovery, mode of vessel failure, and mode and timing of containment failure.

The one non-blackout sequence is the small containment bypass, which is defined as either failure to isolate primary letdown or the spontaneous rupture of a steam generator tube.

Enclosure NOC-AE-1 1002687 Page 24 of 63 The release fractions for all the cases are summarized in Table 2-5 on the following page. The timing and size of release fractions from MAAP can be used to verify that each end state is placed in the correct major release group. In the IPE, the "small" containment bypass category was classified as a small early release, Category II.

However MAAP shows that if a steam generator tube rupture or letdown isolation failure is allowed to proceed to core damage, the releases for many of the species are just as big as those in the large category. However, the small bypass scenarios take many hours to develop, providing time to implement protective measures. Therefore the small bypass end state is reclassified as a late release, Category Ill.

Enclosure NOC-AE-1 1002687 Page 25 of 63 Table 2-5 Severe Accident Radioactive Release Fractions Species/ Noble Iodide TeU2 SrO Mo&2 CsOH BaO Lanthin CeO2 Sb Te2 U&

Case s Trans R05SU 0.50 0.0060 0.031 3.OE-4 6.7E-4 0.0056 4.OE-4 7.6E-5 2.5E-4 0.014 0.012 5.OE-7 BYPAS 1.0 0.50 4.4E-4 5.4E-4 .018 0.41 0.0049 2.7E-5 3.OE-5 0.14 7.OE-6 2.5E-9 S

CICV 0.29 0.0023 0 1.8E-6 6.OE-5 0.0021 1.7E-5 2.4E-7 2.4E-7 4.5E-4 0 0 R07SU 0.53 0.026 0.023 5.3E-4 0.0023 0.027 0.0011 1.OE-4 5.3E-4 0.030 0.0023 1.6E-6 ISGTR 0.25 0.13 0.0030 0.0023 0.040 0.090 0.020 0.0065 0.0066 0.11 0.0068 3.3E-8 R15U 0.55 2.OE-4 0.0025 8.2E-5 5.2E-5 1.3E-4 8.2E-5 7.2E-5 7.4E-5 0.0054 0.0041 2.2E-8 R13U 0.50 4.5E-4 0.0025 4.3E-6 1.5E-5 8.OE-4 1,4E-5 5.OE-6 7.1E-6 0.0030 0.0025 2.1E-8 R11U 1.0 0.0075 0.016 8.5E-4 2.8E-4 0.0043 5.8E-4 6.OE-4 6.1E-4 0.060 0.014 1.2E-7

Enclosure NOC-AE-1 1002687 Page 26 of 63 NRC Requested Information:

b. Describe any internal and external reviews of the complete update of the Level 2 model incorporated in STPREVS. Describe any unresolved F&Os from these reviews, their resolution status, and the impact of their resolution on the SAMA analysis.

STPNOC Response:

The Level 2 PRA update incorporated in STPREV5 is documented in three reports on file at STPEGS:

" STPNOC 2005 Level 2 PRA Update Reports (3 Final reports, Revision 0),

• ABSG Consulting Inc., "Level 2 Probabilistic Risk Assessment Update - 2005,"

"Containment Analysis Report," and

" "MAAP (Revision 4.05) Analysis Report," October 14, 2005.

These reports were reviewed by, ABSG Consulting Inc., prior to submittal to STPNOC; then they were reviewed and approved by STPNOC, and applied in the STPREV5 PRA Level 2 Analysis. The STPEGS STPREV5 PRA, including the updated Level 2 Analysis, and subsequent STPEGS STPREV6 PRA applied for SAMA analyses were reviewed and approved by STPNOC. Specifically, the review and approval of STPEGS PRA Level 2 Analyses is documented via the following three PRA notebooks:

" STPEGS Probabilistic Risk Assessment Notebook, "Level 2 Analysis - Containment Event Tree" STPREV6, August 18, 2009.

• STPEGS Probabilistic Risk Assessment Notebook, "Level 2 Accident Sequence Progression," STP_REV6, August 8, 2009.

• STPEGS Probabilistic Risk Assessment Notebook, "Level 2 Results" STPREV6, September 14, 2010.

It is important to note that the reports referenced above were developed to address then-recent industry/regulatory issues and advances in Level 2 PRA technology and to resolve Westinghouse Owners Group (WOG) Peer Certification containment performance technical elements that were graded less than 3 or contingent grade 3, and any associated Fact/Observations (F&O) with levels of significance of A or B. The specific F&O comments addressed are documented in the STP reports. The F&Os were resolved and incorporated into the STPREV6 PRA model used for SAMA analysis. A summary of issues addressed is provided in the following table.

Enclosure NOC-AE-1 1002687 Page 27 of 63 Table 2-6 Summary of F&Os Issue WOG Peer Certification Issue Description Number F&O Identifier 1 L2-01, Element 5, Sub-element 5 Early containment failure 2 L2-02, Element 2, Sub-element 5 Thermally induced steam generator tube rupture (SGTR) 3 L2-04, Element 2, Sub-element 5 Level 2 success criteria 4 L2-05, Element 2, Sub-element 11 Pressurizer PORV and fan cooler survivability 5 L2-06, Element 2, Sub-element 21 Pre-existing containment leakage 6 L2-06, Element 2, Sub-element 21 Assignment of spontaneous SGTR core damage frequency (CDF) sequences to the "late containment failure" category 7 L2-07, Element 2, Sub-element 23 Emergency action levels (EALs) not included in the evacuation model NRC Requested Information:

c. Provide a brief history of the Level 2 PRA and the major changes in modeling that impact the release category frequencies.

STPNOC Response : The following summary table represents the changes in the containment response over the last several approved PRA models:

Table 2-7 Summarv of Chanass in Containment Res~onse Model Large Early Small Early Late Containment Release Release Containment Intact Failure IPE (1992) 9.89E-07 6.67E-06 1.08E-05 2.56E-05 STP 1996 1.37E-07 2.93E-06 8.28E-07 4.99E-06 STP 1997 6.20E-07 2.05E-06 2.40E-06 6.42E-06 STP 1999 5.76E-07 2.14E-06 3.22E-06 5.39E-06 STP REV 5.37E-07 1.34E-06 2.46E-06 4.71E-06 STP RV42 5.12E-07 1.44E-06 2.18E-06 4.88E-06 STP REV5 6.06E-07 1.90E-06 2.61 E-06 4.98E-06 STP REV6 5.01E-07 1.16E-06 1.48E-06 3.1OE-06

Enclosure NOC-AE-1 1002687 Page 28 of 63 These results are different from the Internal Plant Examination (IPE) due to several changes in the model, as documented in the STPEGS PRA Level 2 Analysis Notebooks for each PRA revision. The IPE showed the largest contribution to Large Early Release Frequency (LERF) to be the failure of the containment supplementary purge valve to close on demand. It has been found that containment purge is not in service near as much as previously assumed, and that the database variable used for the air-operated dampers was "valve failure to close",

rather than the more appropriate "failure to return to fail-safe position". Correcting these parameters greatly reduced the large early isolation failure contribution.

A NUREG 1570, Risk Assessment of Severe Accident-Induced Steam Generator Tube Rupture, approach was used in previous models to evaluate ISGTR, which increased its importance. The NUREG 1570 approach was re-evaluated for STPREV5, and is no longer considered "overly conservative."

The current results show LERF to be dominated by ISGTR (Induced Steam Generator Tube Rupture) and VSEQ (Interfacing systems LOCA (ISLOCA)). The path of concern for ISLOCA is through the RHR suction motor-operated valves (MOV), rupturing the RHR/component cooling water (CCW) heat exchanger tubes, exposing CCW to high pressure. Credit for isolation of the CCW containment isolation valves was removed in STPREV6 because the operator action is not proceduralized and there is significant uncertainty regarding MOV isolation capability during the ensuing flow and pressure transient caused by the ruptured RHR tubes. An additional initiator with a large contribution to LERF, the loss of essential cooling water (ECW) from a tornado event (HWIND2), was not included in the IPE analysis, however was subsequently added to the PRA model.

NRC Requested Information:

d. Identify the version of Modular Accident Analysis Program (MAAP) used to determine the release fractions.

STPNOC Response:

MAAP Version 4.05 was used.

NRC Requested Information:

e. ER Section F.3.6 describes the selection of the representative accident sequencelsource terms for the major release categories. The one example discussed was that an accident sequence with a moderate frequency and severe release characteristics would be selected over an accident sequence with a relatively high frequency and a minor radionuclide release. From the information provided, none of the selected representative sequences (for those categories where multiple source term results are provided) follow this conservative example. For major Release Categories II and III, the selected sequences are not those with the most severe release characteristics. While the information provided in ER Table F.3-8 indicates that the representative sequences are appropriate for the base case, this is not necessarily true for a SAMA case where the Level 2 end-state distribution would be different from the base case. For

Enclosure NOC-AE-1 1002687 Page 29 of 63 example, if a SAMA primarily impacted sequences which have low reactor pressure vessel (RPV) failure pressure then the frequency of end-states R07SU and RI1 U would be reduced. Since these end-states have higher release fractions (and most likely, higher dose-risk and offsite economic cost risk per event) than the representative sequence chosen, the benefit could be larger than that assessed using the representative release fractions.

If the source term chosen for a release category is not the most severe of the significantly contributing end-states, the benefit could be underestimated for any SAMA which primarily impacts an end-state with a higher release fraction. For example, SAMA 4 impacts only the end-state VSEQ (interfacing system loss of coolant accident (LOCA)) portion of Release Category I. It is not clear that end-state VSEQ has a less severe release than the ISGTR end-state, which was chosen as representative for Release Category I. Release fractions for Inter-System LOCA (ISLOCA) are usually greater than that given for ISGTR. The STP IPE (Table 4.8.3-4) gives interfacing system Cs and I release fractions from 0.15 to 0.4 depending on the methodology. Similarly for SAMA 10, which impacts steam generator tube rupture (SGTR) sequences, the removal of these sequences from Release Category III will have a more significant impact since the release fractions for SGTR are three orders of magnitude greater than those for the representative sequence.

Provide further support for the selection of the representative sequences and their adequacy for the SAMA analysis.

STPNOC Response:

The intent of using a "representative source term" is to characterize the average nature of a major release category's numerous contributing sequences with a single term. The use of representative source terms can result in overestimating the impact of implementing some SAMAs while underestimating it for others; however, this simplified quantification process has previously been accepted in License Renewal applications.

The concern stated in this RAI is that STP's representative source terms are non-conservative and that they underestimated the impact of implementing some of the Phase II SAMAs. The implication is that some of the SAMAs that were determined not to be cost beneficial in the ER could have positive net values if the higher consequence sequences were mapped to their corresponding source terms rather than to a representative source term. In order to demonstrate that the representative source terms defined in the Environmental Report (ER) do not impact the conclusions of the analysis, a sensitivity case has been developed that shows none of the SAMAs would be classified as cost beneficial even when the most conservative source terms are assigned to each major Release Category.

Because the ER validated the use of the representative source terms for the baseline model, no changes are proposed to the SAMA identification process or Phase I screening analysis (no change to the MACR). The focus of the sensitivity analysis is on the Phase II

Enclosure NOC-AE-1 1002687 Page 30 of 63 quantifications in which the changes to the sequence frequencies may not be properly mapped to dose and off-site economic consequences.

The approach for this sensitivity analysis is to select the most conservative, relevant available source term for each major Release Category and to use the results to update the Phase II quantifications. The following Table 2-8 provides a summary of the conditional dose and off-site economic cost values for each of the source terms that were analyzed for STP. These values, which were taken from the STP Level 3 model, can be obtained from Table F.3-8 of the ER by dividing the dose-risk and off-site economic cost-risk values by the corresponding frequencies (there will be slight differences given that the results in Table F.3-8 were rounded for presentation).

Enclosure NOC-AE-1 1002687 Page 31 of 63 Table 2-8 Summary of Conditional Doses and Off-site Economic Costs for Available STP Source Terms Group I Group II Group II Group II Group III Group III Group III Group Ill Group IV Release Category (ISGTR) (RO5SU) (CICV) (R07SU) (RI5U) (RI3U) (RIIU) (Bypass) (Intact)

Dose (person-rem) 1.36E+06 5.12E+05 2.12E+05 7.50E+05 1.49E+05 2.85E+05 4.25E+05 2.22E+06 1.70E+04 Off-site Economic Cost ($) 2.40E+09 3.44E+08 8.60E+07 1.07E+09 7.14E+06 1.54E+07 4.02E+08 2.81E+09 4.68E+04 For Group I, there is only one analyzed source term, but the ISLOCA sequence, which is part of Group I, is a "bypass" scenario.

Group III includes a "bypass" source term, which has a larger conditional dose and off-site economic cost than the ISGTR source term from Group I. As a result, the Group III "bypass" source term is used for the Group I source term.

For Group II, source term R07SU has the largest conditional dose and offsite economic cost values of the small-early releases, so it has been assigned as the Group II source term.

For Group Ill, the "bypass" source term has the largest conditional dose and offsite economic cost values of the late releases, so it has been assigned as the Group III source term.

No alternate source terms are available for Group IV and no changes have been made to the source term for this major Release Category.

The same PRA results documented in the ER were used in conjunction with the source terms above to obtain the updated dose-risk and off-site economic cost-risk (OECR) values for each SAMA. Tables 2-9 and 2-10 summarize these results (note: the base case has also been updated to reflect the revised source terms).

Enclosure NOC-AE-1 1002687 Page 32 of 63 Table 2-9 Revised Dose-Risk Results Population dose- Group I Group II Group III Group IV Total Dose-risk, 0-50 mile (risk (Bypass) (R07SU) (Bypass) (Intact) Risk in person-remlyr)

Baseline 1.11 0.87 3.29 0.05 5.32 SAMA 3b 1.11 0.85 3.26 0.05 5.27 SAMA 4 0.83 0.87 3.29 0.05 5.04 SAMA 10 1.10 0.88 2.99 0.05 5.02 SAMA 12 1.11 0.87 3.29 0.05 5.32 SAMA 13 1.11 0.87 3.26 0.05 5.29 SAMA 15 1.10 0.86 3.22 0.05 5.23 Table 2-10 Revised OECR Results Total economic Group I Group II Group III Group IV Total OECR cost-risk, 0-50 (Bps)(RO7SU) (Bypass) (intact) TtlOC miles (risk in $1yr) (Bypass) (

Baseline $1,408 $1,241 $4,159 $0 $6,808 SAMA 3b $1,408 $1,209 $4,131 $0 $6,748 SAMA 4 $1,057 $1,241 $4,159 $0 $6,457 SAMA 10 $1,387 $1,249 $3,779 $0 $6,415 SAMA 12 $1,401 $1,241 $4,159 $0 $6,801 SAMA 13 $1,405 $1,241 $4,131 $0 $6,777 SAMA 15 $1,394 $1,220 $4,075 $0 $6,689

Enclosure NOC-AE-1 1002687 Page 33 of 63 The cost benefit analysis was updated using the above information and net values were recalculated for each SAMA. The tables below summarize the results for the baseline PRA results as well as for the 9 5 th percentile PRA results.

Table 2-11 Cost Benefit Results Using Revised Source Terms and Baseline PRA Results Cost of Total SAMA ID Implementation Averted Net Value Cost-Risk SAMA 3b $796,677 $6,518 -$790,159 SAMA 4 $100,000 $34,786 -$65,214 SAMA 10 $100,000 $29,868 -$70,132 SAMA 12 $100,000 $212 -$99,788 SAMA 13 $100,000 $3,874 -$96,126 SAMA 15 $100,000 $14,106 -$85,894 Table 2-12 Cost Benefit Results Using Revised Source Terms and 9 5 th Percentile PRA Results Total Cost of SAMA ID I Averted Net Value Implementation Cost-Risk SAMA 3b $796,677 $10,404 -$786,273 SAMA 4 $100,000 $55,527 -$44,473 SAMA 10 $100,000 $47,677 -$52,323 SAMA 12 $100,000 $338 -$99,662 SAMA 13 $100,000 $6,184 -$93,816 SAMA 15 $100,000 $22,517 -$77,483 In some cases, the percent change in the averted cost-risk was substantial, but because the absolute change in the risk was low, the averted cost-risk values remained low relative to the costs of implementation and none of the SAMAs are cost beneficial, which is consistent with the conclusions stated in the ER.

Enclosure NOC-AE-1 1002687 Page 34 of 63 NRC Requested Information:

3. Provide the following information with regard to the treatment and inclusion of external events in the SAMA analysis:

a. Provide a description of fire scenarios X, B, 18, BC and 23 as included ER Table F.2-1.

STPNOC Response The fire scenarios included in ER Table F.2-1 are described below in the following order:

18, 23, X, B, and BC. Details of the initiating event frequency quantifications can be found in Section 3.4.2 of the IPE report:

• South Texas Project Electric Generating Station, "Level 2 Probabilistic Safety Assessment and Individual Plant Examination," prepared by HL&P and PLG, Inc.,

August 1992.

The supporting Spatial Interactions Analyses are documented in Section 3.4 of the IPE report.

FR18 - Control Room Fire (Scenario 18)

In this scenario, EAB and control room envelope (CRE) ventilation trains fail due to fire.

The controls for all three trains are located on control room panel 22/4. The following assumptions are made:

• Failure of either supply fan or return fan of an HVAC loop fails that loop.

• Due to the close proximity of the CRE and the EAB circuits for each loop, it is assumed that the CRE and EAB ventilation trains fail together.

• Air supply, cleanup, and makeup equipment are not considered essential for the success of the ventilation system.

• The frequency of a hot short that spuriously closes a damper is considered to be much less than the frequency of any of the fans failing.

In Scenario 18, trains A, B, and C of both the EAB and CRE HVAC systems fail because of control room fire.

FR23 - Control Room Fire (Scenario 23)

The impact of this control room fire is failure of all auxiliary feedwater (AFW) trains.

Scenario 23 involves a fire on control room panel 6 that causes failure of four trains of steam generator controls.

For this scenario to occur, either the AFW isolation valve or the AFW pump control circuits must suffer an open circuit because of the fire in all four trains.

Enclosure NOC-AE-1 1002687 Page 35 of 63 Z047X -Scenario X This initiating event represents a fire in Cable Spreading Room Train B. Scenario impacts include AC Power Train B and C, direct current (DC) Power Trains B and C, Fan Coolers Train A, Recirculation Cooling Train A, RCP Seal Injection, PORV 656A, MSIVs, the centrifugal charging pumps (CCP) and positive displacement pump (PDP), and CCW supply to the RCPs.

Z071X -Scenario X This initiating event represents a fire in the Auxiliary Shutdown Area. Scenario impacts include AC Power Trains A, B and C, AFW Train D, Containment Isolation Trains A and C, and the PDP.

Z047B -Scenario B This initiating event represents a fire in Cable Spreading Room Train B. Scenario impacts include AC Power Train B, DC Power train B, Fan Coolers train A, Recirculation Cooling Train A, RCP Seal Injection and PORV 656A.

Z47BC -Scenario BC This initiating event represents a fire in Cable Spreading Room Train B. Scenario impacts include AC Power Train B and C, DC Power Trains B and C, Fan Coolers Train A, Recirculation Cooling Train A, RCP Seal Injection, PORV 656A and the CCPs and PDP.

NRC Requested Information:

b. In the May 9, 2007, STPNOC response to requests for additional information (RAIs) for RMTS, it was stated that a review of the fire frequency data presented in NUREGICR-6850 was planned for a future reanalysis of fire hazards at STP. If the results of this review have not been incorporated in STPREV6, assess the impact the fire frequency data on the SAMA assessment.

STPNOC Response:

A review of the fire frequency data presented in NUREG/CR-6850 has not been incorporated in STPREV6. Future updates to the Fire PRA and External Events PRA will be performed in accordance with the STP Risk Management Strategic Plan. STP intends to review the impact of NUREG/CR-6850 fire frequency data for incorporation into the STPEGS PRA in 2013.

Enclosure NOC-AE-1 1002687 Page 36 of 63 NRC Requested Information:

c. Identify the seismic hazard curves used to determine the seismic CDF in STP_REV6. If the seismic CDF is based on the Electric Power Research Institute (EPRI) hazard curve, provide the seismic CDF using the Lawrence Livermore National Laboratory (LLNL) hazard curve or the more recent USGS 2008 assessment and include a description of the dominant seismic CDF sequences.

Discuss the impact of these results on the SAMA assessment.

STPNOC Response The seismic hazard curves used to determine the seismic CDF in STPREV6 are the EPRI hazard curves. This same set of curves has been used since the original Internal Plant Examination of External Events (IPEEE). The NRC determined the use of these curves to be acceptable in approving the Risk Managed Technical Specifications for STP.

The NRC published Information Notice 2010-18 to inform Licensees about a August 2010 NRC document, "Safety/Risk Assessment Results for Generic Issue (GI) 199, Implications of Updated Probabilistic Seismic Hazard Estimates in Central and Eastern United States on Existing Plants," (ML100270582) The referenced NRC document stated that based on the assessment of the seismic performance of existing plants (using the seismic hazard information available at the time of the IPEEE), the NRC staff determined that the seismic designs of operating plants in the central and eastern United States still provide adequate safety margins. The regulatory assessment of GI 199 is ongoing including the proposal of a NRC Generic Letter. STP is monitoring the results of this assessment to determine the need to update their seismic CDF based on updated hazard curves. It should be noted that STP Units 1 and 2 reside in an area of very low seismic activity.

NRC Requested Information:

d. Provide a brief description of the latest fire and other external events models incorporated in the STP PRA.

STPNOC Response:

STPREV5 was reviewed by the NRC for issuance of the STP SER (ML071780186) in support of Risk Managed Technical Specifications. STPREV6 external events are unchanged from the NRC reviewed model.

An important class of common cause initiating events is due to external events that include events external to the plant and events such as fires, which can occur inside the plant but are external to the plant processes. External events analysis is a major task in the development of a risk model and separate sections have been reserved for its documentation (Individual Plant Examination (IPE) report, Section 3.4). As an integral part of this analysis, physical interactions that cause one or more initiating events and possible

Enclosure NOC-AE-1 1002687 Page 37 of 63 additional damage to one or more plant systems are identified. The analysis of these interactions is specific to the types of external events considered, which include:

• Seismic Events

• Fires and Explosions

• Internal and External Flooding

• Aircraft Crash

• Wind and Wind-Generated Missiles

• Turbine Missile

• Hazardous Chemical Releases

• Transportation Accidents A list of common cause initiating events for the above sources of physical interaction is identified in Section 3.4 of the IPE report - South Texas Project Electric Generating Station, "Level 2 Probabilistic Safety Assessment and Individual Plant Examination," prepared by HL&P and PLG, Inc., August 1992.

The 19 external event categories that were selected for quantification in the PRA are listed in Table 3-1. The term category is used in the sense that many specific events can be identified for each category by breaking down the category into cause, failure mode, degree of severity, etc. The distinctions made by specifying different categories are those necessary to account for the influence of the initiating events on the development or unfolding of the accident sequences in the event sequence model and to isolate key factors of importance in quantifying accident sequence frequencies and damage levels. For this reason, discrete seismic hazard intensity levels, distinct fire locations and magnitudes, and distinct flood scenarios are counted as separate initiating event categories..

Each class of seismic events was quantified for point estimate results at different, discrete levels of earthquake ground acceleration. The particular ground acceleration levels were selected on the basis of the seismicity and fragility curves. This is discussed more fully in Section 3.4.4 of the IPE report.

The different fire cases consist of different degrees of damage done by fires in the control room area (see Section 3.4.2 of the IPE) and the effects of fire in the Train B cable spreading room and the auxiliary shutdown panel areas (see the STPEGS Fire PRA Update Report). The Fire PRA Update Report is ST-RL-HL-0563, "PLG-1015 Fire Analysis Update for the South Texas Project Electric Generating Station PSA for Selected Zones," prepared by PLG, Inc., December 1994.

The external flooding events, described in IPE report Section 3.4.6, consist of the flood and flood-induced failure of equipment in different areas due to the flow of flood water across these areas.

The final external event category consists of severe winds (tornados) that disable the offsite grid (345kV and 138kV) and the Technical Support Center Diesel Generator (TSCDG), and dump sufficient debris in the vicinity of the essential cooling water (ECW) intake structure to cause plugging of the ECW pump traveling screens. This analysis is also presented in IP.E report Section 3.4.7.

Enclosure NOC-AE-1 1002687 Page 38 of 63 Table 3-1 below presents the external events, including fire scenarios, explicitly modeled in the current STPEGS PRA, STP_REV6.

Table 3-1 External Event Categories Selected for Quantification in the Current South Texas Project Risk Model (STPREV6)

Initiating Event Categories Selected for Code Group Separate Quantification Designator Seismic 0.1g Seismic Event SEIS1 Initiating Events 0.2g Seismic Event SEIS2 0.4g Seismic Event SEIS3 0.6g Seismic Event SEIS4 Plant Fires Control Room - Loss of All Three Motor-Driven FR10 AFW Pumps Control Room - Loss of CR HVAC and EAB FR18 HVAC Control Room - Loss of All AFW Trains FR23 Zone 31Z047 - Cable Spreading Room Train B, IZ047B Scenario B - Affects Train B AC and DC, RCFC A, Recirculation Cooling Train A, RCP Seal Injection, and PORV 656A Zone 31Z047 - Cable Spreading Room Train B, IZ47BC Scenario BC - Affects Train B and Train C AC and DC, RCFC A, Recirculation Cooling Train A, RCP Seal Injection, PORV 656A, and the CCPs and PDP Zone 31Z047 - Cable Spreading Room Train B IZ047X Scenario X - Affects Train B and Train C AC and DC, RCFC a, Recirculation Cooling Train A, RCP Seal Injection, PORV 656A, MSIVs, CCPs, PDP, and RCP CCW Supply Zone 07Z071 - Auxiliary Shutdown Area, IZ071X Scenario X - Affects Train A, Train B, and Train C AC Power, AFW Train D, Containment Isolation (CI) Trains A and C, and the PDP Zone 03Z147 - MAB 41' Corridor and Changing IZ1470 Scenario 0 - Affects CCW A, CCW B, CCW C, LHSI A, HHSI A, CS A, CI Train A, ECH C, CCP A, and SI Recirculation CoolinQ Train A

Enclosure NOC-AE-1 1002687 Page 39 of 63 Table 3-1 External Event Categories Selected for Quantification in the Current South Texas Project Risk Model (STPREV6)

Initiating Event Categories Selected for Code Group Separate Quantification Designator Plant Flooding LOOP, Technical Support Center DG (TSCDG) FL1

- Breach of leads to loss of PDP, Balance of Plant Diesel the Main Generator (BOPDG)

Cooling Reservoir LOOP, PDP and all Three Emergency Diesel FL26 Generators (4 item LOOP, TSCDG, PDP, BOPDG, All CCW and included in One Train (B) of Essential Chilled Water Initiator Chillers FL26)

LOOP, TSCDG, PDP, BOPDG, and Loss of ECW Intake Structure LOOSP, TSCDG, PDP, BOPDG, and One Train (B) of RCFCs LOOP, TSCDG, PDP, BOPDG, and Plugging of FLECW the ECW Pump Traveling Screen by Debris External Severe Wind (Tornado) - LOOP, TSCDG, HWIND Oventh BOPDG, and Plugging of the ECW Pump Traveling Screen by Debris NRC Requested Information:

4. Provide the following information relative to the Level 3 PRA analysis:

a. ER Section F.3.2 states that two previously identified sector population, land fraction, and economic estimation program (SECPOP) errors were corrected in the SAMA analysis. Clarify whether a third known error for incorrect column formatting of the output file was also corrected.

STPNOC Response:

The column formatting of the output file, referred to in this RAI, involved changing the regional economic data format to comply with MACCS2 input requirements, which has been completed.

Enclosure NOC-AE-1 1002687 Page 40 of 63 NRC Requested Information:

b. Provide a table of the sector population breakdown for the SECPOP rosette for the year 2000 and the projected rosette for the year 2050. ER Section 2.6.1 identifies a total population of 255,118 within the 50-mile radius. However, no population is provided for the year 2050. Provide the total population used for the year 2050 (and the year 2000 if different than ER Section 2.6.1, including explanation for the reason for the difference).

STPNOC Response:

Tables 4.b.li and 4.b.1ii show the year 2000 population breakdown by distance and direction (SECPOP rosette) to 10 and 50 miles from the site. Tables 4.b.2i and 4.b.2ii show the corresponding 2050 population breakdown.

Enclosure NOC-AE-1 1002687 Page 41 of 63 Table 4.b.1i YEAR 2000 POPULATION BY DISTANCE (TO 10 MILES) AND DIRECTION FROM STP Distance 0-1 1-2 2-3 3-4 4-5 5-10 (miles:

Direction N 0 0 15 0 0 32 NNE 0 0 0 0 498 168 NE 0 0 0 0 31 99 ENE 0 0 0 0 0 420 E 16 0 0 0 3 217 ESE 0 0 0 71 116 70 SE 0 0 0 3 100 1212 SSE 0 0 0 0 0 108 S 0 0 0 0 0 0 SSW 0 0 0 0 0 0 SW 0 0 1 0 0 76 WSW 0 0 0 4 6 116 W 0 0 0 5 0 114 WNW 0 0 0 0 4 2285 NW 0 0 0 19 30 219 NNW 0 0 0 0 0 34 Table 4.b.1 ii YEAR 2000 POPULATION BY DISTANCE (TO 50 MILES) AND DIRECTION FROM STP Distance 0-10 10-20 20-30 30-40 40-50 Total (miles):

Direction N 47 1237 536 14097 5445 21362 NNE 666 21441 1120 2540 10968 36735 NE 130 931 6687 11447 24758 43953 ENE 420 271 2480 16635 62994 82800 E 236 83 1243 87 46 1695 ESE 257 2 0 0 0 259 SE 1315 13 0 0 0 1328 SSE 108 117 0 0 0 225 S 0 0 0 0 0 0 SSW 0 1 0 0 0 1 SW 77 345 0 1111 628 2161 WSW 126 5671 1074 14758 3240 24869 W 119 261 829 1302 3614 6125 WNW 2289 1181 492 9669 1259 14890 NW 268 477 787 1455 222 3209 NNW 34 484 4469 11928 2211 19126

Enclosure NOC-AE-1 1002687 Page 42 of 63 Table 4.b.2i YEAR 2050 PROJECTED POPULATION BY DISTANCE (TO 10 MILES)

AND DIRECTION FROM STP Distance 0-1 1-2 2-3 3-4 4-5 5-10 (miles):

Direction N 0 0 20 0 0 44 NNE 0 0 0 0 677 228 NE 0 0 0 0 42 135 ENE 0 0 0 0 0 571 E 22 0 0 0 4 295 ESE 0 0 0 97 158 95 SE 0 0 0 4 136 1648 SSE 0 0 0 0 0 147 S 0 0 0 0 0 0 SSW 0 0 0 0 0 0 SW 0 0 1 0 0 103 WSW 0 0 0 5 8 158 W 0 0 0 7 0 155 WNW 0 0 0 0 5 3108 NW 0 0 0' 26 41 298 NNW 0 0 0 0 0 46 Table 4.b.211 YEAR 2050 PROJECTED POPULATION BY DISTANCE (TO 50 MILES)

AND DIRECTION FROM STP Distance 0-10 10-20 20-30 30-40 40-50 Total (miles):

Direction N 64 1681 706 19276 10482 32209 NNE 905 29160 1677 5277 29545 66564 NE 177 1266 12458 23466 51565 88932 ENE 571 369 4164 34102 129138 168344 E 321 113 1728 174 94 2430 ESE 350 3 0 0 0 353 SE 1788 18 0 0 0 1806 SSE 147 159 0 0 0 306 S 0 0 0 0 0 0 SSW 0 1 0 0 0 1 SW 104 469 0 1522 860 2955 WSW 171 7624 1446 20212 4503 33956 W 162 348 1078 1729 5512 8829 WNW 3113 1583 640 12570 1654 19560 NW 365 644 1030 1903 272 4214 NNW 46 653 5854 15626 2780 24959 The total 50-mile populations for years 2000 and 2050 are 258,738 and 455,418.

Enclosure NOC-AE-1 1002687 Page 43 of 63 The total year 2000 population from Tables 4.b.1.ii is 1.4% greater than that in ER Section 2.6.1. This small difference is because the population in ER Section 2.6.1 was taken from SECPOP's data base location of the existing units. As noted in ER Section F.3.1 for the SAMA, the population distribution projections were taken from an analogous study for the potential construction and operation of two new units at the STP site. The small difference in populations resulting from the small difference in STP Units 1 and 2 versus planned STP Units 3 and 4 plant coordinates was not deemed to warrant reworking the complex population projection calculations.

NRC Requested Information:

c. The Houston-Sugar Land-Baytown metropolitan area is just outside the STP 50-mile radius. However, the projected growth through year 2050 is expected to be high. Briefly explain how/whether the population studies addressed the potential for a step change in population within the 50-mile radius if/whether this metropolitan area expands to the southwest.

STPNOC Response:

The population projections were based on projections developed by the individual counties.

The multiplier developed for each sector was based on county-weighted population projections. Thus, those sectors near the Houston-Sugar Land-Baytown Metropolitan Statistical Area (MSA) would have already included any effect expected by local demographers from the high growth rate expected for this MSA.

NRC Requested Information:

d. The evacuation study was performed for the year 2007. Provide the year 2007 transient and total population used in the study.

STPNOC Response:

The 2007 evacuation time estimate was used only for calculation of the evacuation time parameter in MACCS2. As such, the 2007 populations were not relevant to the projected 2050 population. The total year 2007 population used to project the year 2007 evacuation speed to 2050 was based on the exponential growth rate from 2000 to 2050 of 0-10 mile population (the distribution of that population is shown in the response to RAI 4.b). The resulting 2007 population used in the SAMA study was 6,360. The transient component of this total 2007 0-10 mile population was not separately calculated. ER Table F.7.2-1 demonstrates that the baseline risk is insensitive to the evacuation speed (and thus the choice of 2007 population).

Enclosure NOC-AE-1 1002687 Page 44 of 63

6. Provide the following information with regard to the selection and screening of Phase I SAMA candidates:

a. In ER Section F.5.1.3.1, Wolf Creek SAMA 13, which provides for a gravity feed fuel oil supply, was screened from further consideration at STP based on an existing STP capability which requires a pump. The use of a pump has less capability than a gravity system. Provide further justification for the screening of this SAMA.

STPNOC Response:

The current STP fuel oil transfer system already uses a gravity feed line between the fuel oil storage tank (FOST) and the standby diesel generator (SBDG). Each SBDG has a dedicated FOST with a seven-day fuel oil supply. The FOST fuel inventory exceeds the PRA mission time and would support running the diesels for even the most protracted loss of offsite power (LOOP) events at STP.

In the event that the inventory is low in any of the six FOSTs, fuel oil makeup can be provided from either the auxiliary fuel oil storage tank (AFOST), the truck fill line, or the emergency fill connection. Each unit has an auxiliary fuel oil filtration skid that can pump makeup fuel to the FOST from either the AFOST or the truck fill line. Because the skid is powered by a pump, it can take suction from multiple fuel oil sources rather than being limited to one that is located at a higher elevation. In this sense, the existing FOST makeup configuration is considered to provide more capability than an additional gravity feed line. In addition, a new gravity feed line for FOST makeup would require a new tank at STP because the AFOST is at a lower elevation than the FOSTs.

The normal gravity feed configuration of the fuel oil system provides a reliable, passive method of providing fuel from the FOSTs to the SBDGs. The capability to refill the FOSTs from multiple sources exists, if necessary, but the capacity of the tanks limits the need to do so. Providing an additional gravity feed fuel line to either the FOSTs or directly to the SBDGs to supplement the existing gravity feed lines would have very limited benefit.

Finally, any potential benefit would be extremely difficult to measure for STP given that the fuel oil transfer system is within the SBDG component boundary and changes to the fuel oil transfer configuration would not be directly represented in the PRA model.

Based on the reasons presented above, Wolf Creek SAMA 13 is not required for STP and it has been screened from further analysis.

NRC Requested Information:

b. SAMA 16 involves using a portable engine driven instrument air compressor.

This SAMA was based on Prairie Island SAMA 22 which utilized nitrogen bottles rather than a portable compressor. STP SAMA 16 has an estimated implementation cost of $1.2M while Prairie Island SAMA 22 had an estimated implementation cost of $78K ($39K per unit). The cost of nitrogen bottles appears to be considerably less than that of an air compressor. Consider a SAMA that utilizes nitrogen bottles to provide an alternate air source.

Enclosure NOC-AE-1 1002687 Page 45 of 63 STPNOC Response:

For STP, the Instrument Air system is modeled in the PRA, but loss of Instrument Air was not identified as a significant contributor, which is supported by the diversity of the system and the ability to power compressor 14 from the BOP diesel. The implication is that any SAMAs designed to improve Instrument Air would have an averted cost-risk below that corresponding to the review threshold of $11,000 for the site.

In response to this question, the full importance list was reviewed for STP and there was only one Instrument Air split fraction with a Risk Reduction Worth value greater than 1.000, which was IAS14 at 1.016. That corresponds to an averted cost-risk of about $8,100 for the site, which is about an order of magnitude below the Prairie Island cost estimate for using nitrogen bottles as a backup air source. Even if the nitrogen bottles are capable of mitigating all loss of Instrument Air events and the SAMA can be implemented for a cost of

$78,000, the enhancement would not be cost beneficial ($8,100 - $78,000 = -$-69,000).

Because the cost of implementation is over 9 times the averted cost-risk, this SAMA would not be cost effective even if the 9 5 th percentile PRA results were applied.

NRC Requested Information:

c. At Indian Point the final SAMA evaluation included three cost beneficial SAMAs not evaluated in STP Section F.5.3.1.2. These are: SAMA 9 - create a reactor cavity flooding system to reduce the impact of core-concrete interaction from molten core debris following core damage and vessel failure, SAMA 53 - keep both pressurizer PORV block valves open, and a gagging device for SGTR events that would provide a means of closing a stuck open SG relief valve. At Prairie Island the final SAMA evaluation included three cost beneficial SAMAs not evaluated in STP Section F.5.3.1.5. These are: SAMA 3 - provide alternate flow path from refueling water storage tank (RWST) to charging pump, SAMA 19a -

provide a reliable backup water source for replenishing the RWST (for Unit 2), and a gagging device for SGTR events that would provide a means of closing a stuck open SG relief valve. Consider these SAMAs for STP.

STPNOC Response:

The following tables provide the disposition of the SAMAs identified in this RAI:

Enclosure NOC-AE-1 1002687 Page 46 of 63 Table 5-1 Review of Additional Indian Point Cost Beneficial SAMAs Industry SAMA Discussion for STP Disposition for Site Description STP SAMA List SAMA ID 9 Create a reactor The implementation cost of the reactor Cost of cavity flooding cavity flooding system is estimated to be implementation system $3,714,000 per unit in Appendix E, Table exceeds plant E.2-2 of the Indian Point Energy Center MACR. Screened Environmental Report [ENO 2007]. This from further cost, which is considered to be applicable to analysis.

STP, is over 14 times larger than the base STP MACR of $259,000 per unit. This SAMA could not be a cost beneficial for STP even if it eliminated all plant risk and the 95th percentile PRA results are applied.

53 Keep both STP normally operates with the PORV Already pressurizer power- block valves open. If a PORV is leaking, implemented.

operated relief the block valve may be closed to isolate the Screened from valve block valves leak, but the intent of this SAMA is further analysis.

open considered to be met.

NA Develop a Main A hydraulic ram device could be used to Cost of Steam Safety close a stuck open SV, but because an implementation Valve Gagging open SV at STP would create a local steam exceeds plant Device environment, no significant installation work MACR. Screened could be performed once a valve has from further opened. Hydraulic ram devices could be analysis.

permanently installed so they could be used in an accident scenario, but such a device would be required on each of the SVs (20 per unit). Because a single hydraulic ram device is estimated to cost about $75,000, implementation at the site would exceed $3 million for the hardware alone. This is over 3.5 times greater than STP's 9 5 th percentile MACR of $826,854.

Enclosure NOC-AE-1 1002687 Page 47 of 63 Table 5-2 Review of Additional Prairie Island Cost Beneficial SAMAs Industry SAMA Discussion for STP Disposition for Site Description STP SAMA List SAMA ID 3 Alternate flow path Parallel paths are already installed in the Already from RWST RWST suction path for the STP Charging implemented.

system. The two MOVs, which are Screened from powered from separate divisions of diesel further analysis.

backed power, provide an independent means of opening the RWST suction path apart from a common check valve.

Successful operation of either MOV can provide a suction source for all three charging pumps. The intent of this SAMA is considered to be met for STP.

19a Replenish RWST Providing the capability to provide makeup Cost of from large water to the RWST at a capacity that could implementation source support long term injection in an SGTR exceeds plant event is a significant hardware modification. MACR. Screened Prairie Island estimated the implementation from further cost to be $1.935 M per unit ($3.87M for the analysis site) [NMC 2008]. While the cost of implementation may be different than for Prairie Island, the scope of the change is considered to be consistent. Because the implementation cost is over 7 times greater than the STP site MACR of $518,000, providing makeup capability to the RWST could not be cost beneficial even if it eliminated all plant risk and the 9 5 th percentile PRA results are used.

NA Develop a Main Addressed above for Indian Point. Addressed above Steam Safety for Indian Point.

Valve Gagging Device

Enclosure NOC-AE-1 1002687 Page 48 of 63

REFERENCES:

ENO (Entergy Nuclear Operations, Inc.). 2007. Applicant's Environmental Report; Operating License Renewal Stage; Indian Point Energy Center.

Attachment E - Severe Accident Mitigation Alternatives Analysis. April.

NMC 2008 NMC (Nuclear Management Company, LLC). 2008. Application for Renewed Operating Licenses - Prairie Island Nuclear Generating Plant Units 1 and 2. Appendix F SAMA Analysis. March.

NRC Requested Information:

d. ER Section F.5.1 states that the industry based SAMA list from NEI 05-01 was used to identify the types of SAMAs that might address a particular issue. There is no further discussion of the use of this list in the ER. Clarify whether any SAMAs were developed from considering this list.

STPNOC Response:

One of the reasons that the NEI 05-01, Severe Accident Mitigation Alternatives (SAMA)

Analysis Guidance Document, guidance was developed was to move the industry toward a SAMA identification process that was based on plant specific risks. The development of the guidance was initiated after the NRC review of the H.B. Robinson SAMA analysis. During this review, the NRC explicitly stated that a review of a generic SAMA list was of limited benefit; the generic SAMAs had been analyzed by multiple plants and were consistently found not to be cost beneficial. The real benefit was considered to be in the development of SAMAs generated from plant specific risk insights. The STP SAMA identification process is consistent with this philosophy given that it is based on plant specific risk insights from the PRA models.

In addition, the generic SAMA list provided in NEI 05-01 may neither be complete nor applicable to STP. The list was derived from the body of SAMAs identified from previous SAMA submittals and other industry guidance (with duplicates deleted). There is no guarantee that the list of SAMAs is comprehensive or that it is relevant to any given plant beyond the fact that that it includes potential plant enhancements that may have been derived from similar plants.

As stated in Section F.5.1 of the ER, the generic SAMA list from NEI 05-01 was used as an idea source to generate SAMAs for the important contributors to STP risk. The process for developing the SAMAs is essentially the same as described above, but the SAMAs to be reviewed are derived from the PRA rather than using resources to disposition the entire contents of Table 14 of NEI 05-01.

Enclosure NOC-AE-1 1002687 Page 49 of 63 Examples of SAMAs that were derived from the concepts included in Table 14 of NEI 05-01 are summarized in Table 5-3 below:

Table 5-3 SAMA Summary STP SAMA JSAMA Concept Sources (from Table 14 of NEI 05-01)

SAMA 1: Use Portable 480V AC Generator for 003: Add additional battery charger or Long Term AFW Support and Protect the portable, diesel-driven battery charger to Technical Support Center Diesel Generator to existing DC system.

Support the Positive Displace Pump for 056: Install and independent reactor coolant Reactor Coolant Pump Seal Cooling pump seal injection system, without dedicated diesel.

SAMA 1 a: Use Portable 480V AC Generator 003: Add additional battery charger or for Long Term Auxiliary Feedwater Support portable, diesel-driven battery charger to and Install Westinghouse Shutdown Seals to existing DC system.

Preserve Primary Side Inventory 058: Install improved reactor coolant pump seals SAMA 2: Install a Core Spray System 028: Add a diverse low pressure injection system.

SAMA 6: Install an Additional Diverse, High 026: Provide an additional high pressure Head Safety Injection Pump injection pump with independent diesel.

SAMA 7: Provide Portable Fans and Ductwork 082: Stage backup fans in switchgear rooms for Alternate Electrical Auxiliary Building Room Cooling SAMA 8: Enhance Fire Barriers in Control 143: Upgrade fire compartment barriers Room Envelope Panel 22/4 1 NRC Requested Information:

e. ER Section F.6 states that site-specific cost estimates were developed for several of the SAMAs. ER Table F.5-3 cites Reference STPNOC 2009a as the source of the site-specific cost estimates. This reference is an e-mail from Engineering and Research Incorporate (ERIN) on High Head Safety Injection (HHSI). Briefly describe the process and level of detail used to develop the cost estimates (i.e.,

the general cost categories considered). Clarify the level of involvement and expertise of STP staff and ERIN staff in the development of the site-specific cost estimates. Provide the detailed cost estimates for SAMAs 3b and 11.

STPNOC Response:

The scope and definition of the SAMAs were initially developed by the PRA analysts (STP and ERIN personnel) and then reviewed/modified by the STP design staff to account for any plant specific issues that could interfere with or improve the SAMA designs. After finalizing the scope and definition for each of the SAMAs requiring cost estimates, the major cost

Enclosure NOC-AE-1 1002687 Page 50 of 63 contributors for each were identified and their magnitudes were estimated by STP design engineers. The STP design engineers perform cost estimates as part of their normal job duties.

The cost estimates that were developed for SAMAs 3b and 11 are provided below:

Table 5.4 SAMA 3b Install Fire Wrap on PDP Cables in Cable Spreading Room Description Oty. Unit Unit $ Total Engineering 1 LS $250,000 $250,000 Wrap B and C trays in Cable Spreading 242 LF $350 $84,700 Room With Fire Protection Wrap Wrap Conduits in Cable Spreading Rm. 23 LF $125 $2,875 With Fire Protection Wrap Sub-Total $337,575 Sub-Total per Unit X 2 Units 2 Units $337,575 $675,150 18% Capital & Corporate Overheads $121,527 Total $796,677 Table 5-4 Notes:

• LS-lumpsum

• LF-linear feet

Enclosure NOC-AE-1 1002687 Page 51 of 63 Table 5-5 SAMA 11 Modify Fire Protection System to Supply Containment Spray (CS) Headers Description Qty. Unit Unit $ Total Engineering 1 LS $50,000 $50,000 Procedure Revisions 1 Ea $40,000 $40,000 100 feet of 6 inch, 150 psi CS pipe with stainless steel (SS) Tee and double manual ASME SS valves at CS Pump discharge and CS tee and 1 single non-safety valve at fire header 1 LS $50,000 $50,000 Operator Training Sub-Total $360,000 Sub-Total per Unit X 2 Units 2 Units $360,000 $720,000 18% Capital & Corporate Overheads $129,600 Total $849,600 Table 5-5 Notes:

• LS-lump sum

• EA- each NRC Requested Information:

f. SAMA 17a, "install Westinghouse Reactor Coolant Pump (RCP) Shutdown Seals,"

has an estimated implementation cost of $7,611,000. This suggests that the single unit cost for this modification is estimated to be about $3.8 million. A recent submittal by Tennessee Valley Authority (TVA) for Watts Bar Unit 2 (TVA letter to NRC of July 23, 2010, ML102100588) estimated the cost to install improved Westinghouse RCP seals to be $1.1M per unit while this modification was estimated to cost $1.05M per unit in the Vogtle license renewal application.

Furthermore, the NRC staff is aware that the new seal package technology is

Enclosure NOC-AE-1 1002687 Page 52 of 63 being demonstrated at the Farley Nuclear Plant (TVA letter to NRC of January 31, 2011, ML110340040). Describe the difference between the "shutdown seals" assumed in SAMA 17a and the improved seals cited by TVA and Vogtle. Also, provide a more detailed description of the SAMA 17a modification and justification for the estimated cost to install the "shutdown seals" at STP.

STPNOC Response:

The "shutdown seals" identified in STP SAMA are not the same type of seal as the "improved seals" described by Watts Barr and Vogtle. STP utilizes Westinghouse Model 100 RCPs with a specific seal housing design. The specific pump seal design would result in an increased design and engineering cost that would be incurred by STP while other industry sites would share the design and engineering cost among a larger pool of contributors. It should be noted that at the time the SAMA analysis was developed, there was no finalized product or cost available from the vendor.

The following table provides the cost estimate developed by STP for installation of the "improved seals" for the RCPs:

Table 5-6 SAMA 17a: Install Westinghouse RCP Shutdown Seal Description Qty. Unit Unit $ Total 1 LS $100,000 $100,000 Engineering - RPE Procedure Revisions 1 Ea $25,000 $25,000 4 Ea $40,000 $160,000 Modified #1 Seal Housing New Emergency RCP Seal 4 $350,000 $1,400,000 CEM Emergency RCP Seal + Housing 1 $340,000 $340,000 Installation 4 $300,000 $1,200,000 Sub-Total $3,225,000 Sub-Total per Unit X 2 Units 2 Units $3,225,00 0 $6,450,000 18% Capital & Corporate Overheads $1,161,000 Total $7,611,000

Enclosure NOC-AE-1 1002687 Page 53 of 63 Notes for Table 5-6 above:

• LS-lump sum

• EA - each NRC Requested Information:

g. The cost of $4.5M given in ER Table F.5-3 for SAMA 14 seems very high given that an inter-unit cross-tie is already available. Provide a more detailed description of the modification and justification for the estimated cost. In the response, discuss the possibility of using existing breakers and buses to cross-tie buses in one unit under emergency conditions.

STPNOC Response:

The original intent of SAMA 14 was to provide the capability to perform a cross-tie between emergency 4KV AC buses within a single unit rapidly enough to prevent an RCP seal LOCA. It was assumed that the most effective means of providing this capability was through a direct bus to bus connection, which does not exist at STP.

In response to this RAI, STP's 4KV AC cross-tie capabilities were investigated further and it was determined that a viable 4KV AC cross-tie hardware path already exists on the affected unit via a back feed through an emergency transformer; however, this alignment is part of the mitigating strategies B.5.b program and the details are not available to the public.

While the hardware to support the cross-tie capability exists within each single unit, incorporating the capability into plant procedures would be significantly more resource intensive than other procedure changes that have been investigated as SAMAs (e.g., SAMA 4). The primary reason is because performing the cross-tie introduces the potential for a single failure to disable multiple divisions of equipment. The cost of the analysis that would be required to support the cross-tie alone would potentially range from $300,000 to

$400,000, the scope of which is beyond the more general analysis that supports the use of the cross-tie.

In addition to the cost of analysis, the cost of the relatively complex procedure changes associated with the cross-tie has been estimated to be an additional $300,000, for a total of about $600,000.

With regard to capability, it is estimated that the cross-tie could not be aligned in time to prevent a RCP seal LOCA, thus limiting the potential benefit of the SAMA to a small subset of LOOP scenarios in which equipment on the "powered" division has failed. These types of scenarios were not identified in the STP importance list review and a site specific AC cross-tie SAMA was not developed (it was identified through an industry review for a "similar" plant). The implication is that the benefit for the AC cross-tie SAMA would be below the

$11,000 review threshold established in Section F.5. 1.1 of the ER and should be screened from review. Even if the entire LOOP contribution were assumed to be mitigated by this SAMA, the maximum potential averted cost-risk would only be $205,060 when the 9 5 th

Enclosure NOC-AE-1 1002687 Page 54 of 63 percentile PRA results are applied (% contribution from LOOP

• 9 5 th percentile MACR =

averted cost-risk, or 0.24

• 856,854 = 205,060). This would result in a net value of -

$394,940 ($205,060 - $600,000 = -$394,940).

Finally, the manipulation time for the cross-tie is estimated to be driven by event classification and navigation of procedures rather than the physical process of aligning the buses such that providing a direct bus to bus cross-tie would provide no additional benefit for preventing an RCP seal LOCA. As a result, the original SAMA 14 design would not be required for STP and both are screened from further review.

NRC Requested Information:

6. Provide the following information with regard to the Phase II cost-benefit evaluations:

a. ER Section F.6.2 describes that SAMA 10 was modeled by reassigning the SGTR CDF contribution for Release Categories I (7.48E-09 per year) and III (1.35E-07 per year) to Release Categories II and IV, respectively. Neither of these contributions corresponds with any frequency values reported in ER Table F.3-5. Provide additional information on the source of each of these release frequency contributions and clarify that they represent a realistic assessment of the potential risk reduction for this SAMA.

STPNOC Response:

Because SAMA 10 is dependent on the availability of secondary side makeup to ensure inventory in the steam generators is maintained above the tubes, only a fraction of the steam generator tube rupture (SGTR) scenarios are relevant to the SAMA 10 quantification.

Specifically, the induced SGTR (ISGTR) contributors were excluded because the ISGTR classification implies lack of secondary side makeup capability in the STP model. This is one reason why the Release Category I frequency of 7.48E-08/yr and the Release Category III frequency of 1.35E-07 per year do not directly correspond to the frequencies listed in Table F.3-5 of the ER. Another reason is that the SGTR contributors are binned among different end states with contributors from other initiating events such that no single end state exclusively captures all of the SGTR scenarios that would be impacted by steam generator flooding.

In order to identify the relevant frequencies for this SAMA, the PRA model results were examined to determine how the SGTR events were distributed among the STP Level 2 release categories. This information was provided explicitly by the PRA software and the results are reproduced below:

Enclosure NOC-AE-1 1002687 Page 55 of 63 Table 5-7 PRA Model Results (eventslyear)

Group Name Frequency Description RELI 7.4767E-009 RELEASE CATEGORY 1, LARGE EARLY RELEASE RELII 7.6343E-008 RELEASE CATEGORY II, SMALL EARLY RELEASE RELIII 1.3498E-007 RELEASE CATEGORY III, LATE RELEASE RELIV 2.0146E-007 RELEASE CATEGORY IV, NO RELEASE It was assumed that all SGTR contributors had AFW available such that SAMA 10 could be credited with scrubbing the releases; however, because Release Category II is already a small release, no further reduction was assumed. Release Category IV represents the "no release" condition and those contributors were also assumed to not be impacted by SAMA 10.

In order to simulate the scrubbing effect of SAMA 10, the Large-Early releases were reduced to Small-Early releases and the late releases were assumed to result in "no release", as described in the ER. These assumptions are considered to adequately represent the impact of SAMA 10 implementation.

NRC Requested Information:

b. ER Section F.6.3, 5th paragraph, explains that the evaluation of SAMA 12 did not consider the condition in which non-condensable gases such as hydrogen are present since this condition is not modeled in the PRA, but that this condition is conservatively treated in the PRA. If this SAMA impacts this condition then the estimated risk reduction is potentially underestimated. Also, this same section of the ER states that SBO sequences were excluded in the modeling of this SAMA because AC power is needed to start a reactor coolant pump (RCP). This also potentially underestimates the risk reduction benefit for this SAMA since it does not appear to include SBO scenarios in which AC power is recovered. Discuss these issues and their impact on the SAMA analysis.

STPNOC Response:

Comments on the following paragraph in ER Section F.6.3 addresses Part 1 of RAI 6.b, regarding non-condensable gases.

"The presence of non-condensable gases such as hydrogen may occur, but is not typically modeled in nuclearplant PRA models. The use of discrete, success or failed events until explicitly recovered, which is the nature of PRA models is not sufficient to resolve periods of system failure followed by periods of later successful system response. The generation of hydrogen within the RCS that occurred following a temporary interruption of steam generatorcooling and high pressure injection at TMI-2 is not modeled in the STP PRA. It is instead conservatively represented by sequences in which high pressure injection and AFW are initially lost and not restored."

Enclosure NOC-AE-1 1002687 Page 56 of 63 Although the specific TMI scenario leading to hydrogen gas generation is not explicitly modeled in the STP PRA, it is, as noted, represented conservatively in induced SGTR scenarios, by sequences where high pressure injection (HPI) and AFW are initially lost and not restored resulting in core damage with the RCS at high pressure. These sequences are included in the assessment of the impact of this SAMA. The presence of non-condensable gases in the RCS at the high system pressures required for induced steam generator tube ruptures is not expected to significantly affect the heat transfer rate to the steam generator tubes relative to that of the hot legs. Therefore, the estimated risk reduction of SAMA 12 is not underestimated.

For Part 2 of RAI 6.b, regarding SBO scenarios:

The station blackout scenarios excluded from the assessed impacts of this SAMA (due to power being unavailable to the RCPs) is appropriate. Without offsite power, the RCPs could not be operated. For station blackout scenarios initially losing offsite power but with offsite power recovered in time to prevent core damage, such sequences are mapped to success and induced steam generator tube ruptures are not at issue.

For station blackout scenarios initially losing offsite power and with offsite power recovered only after core damage, the governing STP procedure would be OPOP05-E0-EC02, "Loss of All AC Power Recovery With SI Required". In this case, the critical function status trees are to be monitored for information only and the function restoration procedures are not to be implemented until after step 11 of the procedure.

While RCP seal cooling is to be established in preparation for pump operation, the initial steps of this procedure are to manually load ECCS pumps including HHSI pumps for injection, and if steam generator level is low, then to initiate AFW pumps to establish AFW flow. This procedure does not instruct the operators to start the RCPs to reestablish primary flow thereby potentially affecting loop seal clearing. To the contrary, if RCP seal cooling is lost long enough, the operators are specifically directed not to start the affected RCP. Therefore, elimination of station blackout sequences in which recovery of offsite power occurs after core damage, from the impacts of SAMA 12 is also appropriate.

STP Function Restoration Guideline OPOP05-EO-FRC1 may have been entered prior to the time of core damage in some scenarios, but only for non-station blackout scenarios.

These scenarios were considered in the benefit assessment of SAMA 12.

Therefore, the SAMA evaluation does not underestimate the impact of this SAMA due to conservatisms in the assessment.

NRC Requested Information:

c. ER Section F.6.5, for SAMA 15, states that "common cause failures were added, after the common cause data was edited." Explain what is meant by this statement.

Enclosure NOC-AE-1 1002687 Page 57 of 63 STPNOC Response:

It is believed that this question is commenting on a statement describing changes made to the top event model to develop the PRA sensitivity case for this SAMA. In RISKMAN, total failure rate data (in this case for HVAC fans to start and run), are entered in the common cause group modeling area. In RISKMAN, changes to the CCF model or data require "removing" CCF, making the desired changes, and then "adding" CCF. The statement, "common cause failures were added, after the common cause data was edited," simply refers to the procedure for CCF changes performed in RISKMAN, and is correct. It is simply defining the step-by-step process required to implement the change in the failure rates. No changes to CCF group inventory or logic in the associated fault tree(s) are associated with the referenced comment in this case.

NRC Requested Information:

7. Provide the following information with regard to the sensitivity and uncertainty analyses:

a. ER Section F.7.1.1.1 describes the PRA model changes made to evaluate SAMA 3b as deleting macros IZ47BC and IZ047X. Describe these macros.

STPNOC Response:

The macro IZ47BC describes the fire initiating event Z47BC. The macro IZ047X describes the fire initiating event Z047X.

b. NRC Requested Information:

The ratio of the 95th percentile CDF to the mean value CDF was reported to be 1.6 in Section F.7.1 of the ER. While this is a "typical" result for internal event CDF, it seems quite low for the fire and seismic CDFs which generally have wider uncertainty bands than internal events. Provide support to the adequacy of this distribution result given the expected wider distribution for external events and considering the impact of more current seismic hazard curves such as the USGS 2008 assessment.

STPNOC Response:

With the exception of the frequencies for the seismic initiating events, all of the initiating event frequencies utilize data variables that are described by probability distributions.

Because seismic CDF at STP is small, the use of point estimates for the peak acceleration frequencies has no visible effect on the data uncertainty curve. See response to RAI 3.c.

regarding consideration of more current seismic hazard curves.

Enclosure NOC-AE-1 1002687 Page 58 of 63 NRC Requested Information

8. For certain SAMAs considered in the ER there may be lower-cost alternatives that could achieve much of the risk reduction at a lower cost. In this regard, provide an evaluation of the following SAMAs:

a. SAMA 1, involving using a portable AC generator for long term auxiliary feedwater (AFW) support and protecting the Technical Support Center (TSC) emergency diesel generator (EDG) from tornado events, is identified as a means of mitigating a large number of important basic events. While the tornado protection is important for HWIND (i.e., Tornado Induced Failure of Switchyard) initiated sequences, many other sequences would be mitigated without the cost of the tornado protection. Consider such a SAMA.

STPNOC Response:

The cost of protecting SAMA 1 such that it could function in a high wind event, while not negligible, is not a critical factor in the Phase I screening for SAMA 1. The SAMA 1 cost estimate has been modified to eliminate the costs associates with providing high wind protection.

All costs associated with constructing the enclosures for the TSCDG and Load Center 1W (i.e. the load center supplied by the TSCDG) were eliminated. In addition, the engineering costs were reduced by 50 percent to reflect the reduced scope of work. The revised cost of implementation for the site is $2,419,000, which is over 4.5 times larger than the baseline MACR. Even if this modified version of SAMA 1 were assumed to eliminate all on-line risk and the 95th percentile PRA results were applied, it would not be cost effective.

The tables below provide the cost information for the baseline version of SAMA 1 and the version of SAMA 1 without high wind protection.

Enclosure NOC-AE-1 1002687 Page 59 of 63 Table 8-1 SAMA 1 Baseline Cost Estimate Description Qty. Unit Unit $ Total 1 LS $500,000 $500,000 Engineering 1 LS $100,000 $100,000 Procedure Revisions 1 LS $225,000 $225,000 480V DG set + fuel tank for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> 500 LF $300 $150,000 Conduit & Wire 1 Ea $25,000 $25,000 Switches/breakers to switch board Construct concrete block enclosure for $50,000 $50,000 TSC DG 1 LS Construct concrete block enclosure for $140,000 $140,000 Load Center 1W on EAB roof 1 LS New test procedures & Emergency $175,000 $175,000 Operating Procedures 1 LS 1 LS $25,000 $25,000 Testing 1 LS $75,000 $75,000 Operator Training Program Sub-Total $1,465,000 2 Units $1,465,000 $2,930,000 Sub-Total per Unit X 2 Units Capital & Corporate Overheads 18% $527,400 Total $3,457,400 Notes:

LS - lump sum LF - linear feet EA - each

Enclosure NOC-AE-1 1002687 Page 60 of 63 Table 8-2 SAMA I Cost Estimate Without High Wind Protection Description Qty. Unit Unit $ Total Engineering 1 LS $250,000 $250,000 Procedure Revisions 1 LS $100,000 $100,000 480V DG set + fuel tank for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> 1 LS $225,000 $225,000 Conduit & Wire 500 LF $300 $150,000 Switches/breakers to switch board 1 Ea $25,000 $25,000 Construct concrete block enclosure for $0 $0 TSC DG 1 LS Construct concrete block enclosure for $0 $0 Load Center 1W on EAB roof 1 LS New test procedures & EOPs 1 LS $175,000 $175,000 Testing 1 LS $25,000 $25,000 Operator Training Program 1 LS $75,000 $75,000 Sub-Total $1,025,000 Sub-Total per Unit X 2 Units 2 Units $1,025,000 $2,050,000 18% Capital & Corporate Overheads $369,000 Total $2,419,000 Notes:

LS - lump sum LF - linear feet EA - each NRC Requested Information

b. With respect to RAI 8.a, discuss the possibility for the TSC EDG to supply the positive displacement pump (PDP) and support AFW operation. If feasible, consider such a SAMA.

Enclosure NOC-AE-1 1002687 Page 61 of 63 STPNOC Response:

Assuming that the Technical Support Center diesel generator (TSCDG) could support its normal loads, the PDPs, and the loads associated with supporting long term AFW operation, the SAMA 1 implementation cost could be further reduced by eliminating the costs associated with the portable 480V AC diesel generators. The cabling and analysis costs are retained as a connection between the non-safety TSCDG and the safety related bus supporting AFW would be required.

The table below provides the cost information for SAMA 1 assuming that the TSCDG can be used to support AFW instead of a separate generator and assuming that no high wind protection is installed for the TSCDG.

Table 8-3 SAMA I Cost Estimate Using the TSC EDG for AFW Support and No TSC High Wind Protection Description Qty. Unit Unit $ Total Engineering 1 LS $250,000 $250,000 Procedure Revisions 1 LS $100,000 $100,000 480V DG set + fuel tank for 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> 1 LS $0 $0 Conduit & Wire 500 LF $300 $150,000 Switches/breakers to switch board 1 Ea $25,000 $25,000 Construct concrete block enclosure for 1 LS $0 TSC DG Construct concrete block enclosure for 1 LS $0 Load Center 1W on EAB roof New test procedures & EOPs 1 LS $175,000 $175,000 Testing 1 LS $25,000 $25,000 Operator Training Program 1 LS $75,000 $75,000 Sub-Total $800,000 Sub-Total per Unit X 2 Units 2 Units $800,000 $1,600,000 18% Capital & Corporate Overheads $288,000 Total $;_,888,000 Notes:

LS - lump sum LF - linear feet EA - each

Enclosure NOC-AE-1 1002687 Page 62 of 63 If the costs of the 480V AC generators are eliminated in conjunction with the TSC high wind protection measures, the cost of the SAMA is reduced to $1,888,000. This is over two times as large as the 9 5 th percentile maximum averted cost-risk of $826,854 and the SAMA would not be cost effective even if it could eliminate all plant risk. Without the high wind protection measures, however, it should be noted that this version of SAMA 1 would not address the single largest contributor to plant risk and that the potential averted cost-risk would be significantly less than the maximum averted cost-risk NRC Requested Information

c. The HWIND initiating event is the largest single contributor to CDF. For mitigating the HWIND sequence, consider a SAMA to provide an alternate intake structure for the essential cooling water (ECW) either in the essential cooling water pond (ECP) or the MCR that would minimize the likelihood of debris preventing ECW cooling and/or the possibility of using a temporary/portable pump with a movable suction that could provide water to the ECW system.

STPNOC Response:

The construction of an alternate intake structure or a modification to the existing structure that would not be susceptible to clogging in a high wind scenario constitutes a major plant modification. Based on a review of the existing intake structure design and the debris clogging failure mode, a modification to the intake structure is expected to be less resource intensive than the installation of an alternate intake structure. A potential modification would be to install a large surface area debris cage (136 feet x 25 feet x 16.5 feet) that would be less likely to be completely clogged in a high wind event. The modification would be seismically rated, but not safety related. STP has estimated the cost of the design, installation, and materials of the debris cage to be $827,800, which exceeds the STP MACR of $518,000 for the site. Even considering the 95th percentile Probabilistic Risk Assessment (PRA) results, this SAMA would not be cost beneficial.

The portable suction/pump SAMA would potentially be a less costly alternative; however, the same high wind event that caused failure of the intake structure may introduce Essential Cooling Pond access issues that would prevent alignment of a portable pump in time to prevent core damage. The time frame required for Essential Cooling Water (ECW) recovery is considered to be the time to core damage in a Reactor Coolant Pump (RCP) seal loss-of-coolant accident (LOCA) scenario, which may be around 40 to 60 minutes. Because recovery of the ECW system would restore emergency 4KV power, it is not necessary to prevent an RCP seal loss-of-coolant accident (LOCA), but 4KV power must be restored in time to provide injection to the reactor coolant system (RCS) to prevent core damage. For this evaluation, it is assumed that this is possible by using a "portable" ECW pump; however, the pump would have to be large enough to provide about 35,000 gallons per minute (total for both units for diesel cooling and decay heat removal), be self-powered, and be capable of being moved in conjunction with any required suction and discharge piping. These requirements imply a truck based pump. Systems with these capabilities are available, but the cost estimate for a 30,000 gallons per minute mobile pump, which was obtained from a vendor, is $300,000. The cost of installing hard piping at the structure to direct flow to the appropriate ECW bays is estimated to cost an additional $50,000, for a total of $350,000.

Even without including any costs associated with training and procedure updates to support this SAMA, the cost of implementation would exceed the potential averted cost-risk associated with the HWIND initiating event.

Enclosure NOC-AE-1 1002687 Page 63 of 63 The HWIND initiator contributes 17.3% of core damage frequency (CDF) and is associated with events that have similar risk reduction worth (RRW) values in both the Level 1 and Level 2 importance lists. This would correlate to about 17.3% of the maximum averted cost-risk (MACR), which is $89,614 ($518,000

• 0.173). Even if the 95% PRA results are applied, the potential averted cost-risk is only $143,046 ($826,854

• 0.173).

Based on these factors, neither of these potential enhancements would be cost-effective for STP.