RC-12-0142, License Amendment Request - LAR-06-00055, License Amendment Request to Adopt NFPA 805, Response to Request for Additional Information

License Amendment Request - LAR-06-00055, License Amendment Request to Adopt NFPA 805, Response to Request for Additional Information

ML12297A218
Person / Time
Site:	Summer
Issue date:	10/10/2012
From:	Gatlin T South Carolina Electric & Gas Co
To:	Martin R Document Control Desk, Office of Nuclear Reactor Regulation
References
RC-12-0142, TAC ME7586
Download: ML12297A218 (166)

Text

Thomas D. Gatlin Vice President,Nuclear Operations 803.345.4342 A SCANA COMPANY October 10, 2012 RC-12-0142 U.S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555-0001 Attn: R.E. Martin

Subject:

VIRGIL C. SUMMER NUCLEAR STAT!ON_(VCSNS) UNIT 1 DOCKET-NO. 50-395 OPERATING LICENSE-NO. NPF-12 LICENSE AMENDMENT REQUEST - LAR-06-00055 LICENSE AMENDMENT REQUEST TO ADOPT NFPA 805 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION

References:

1. Letter from Thomas D. Gatlin to NRC Document Control Desk, dated November 15, 2011, License Amendment Request - LAR-06-00055, "License Amendment Request to Adopt NFPA 805 Performance-Based Standard for Fire Protection for Light Water Reactor Electric Generating Plants (2001 Edition)"

2. NRC Letter from Robert E. Martin to Thomas D: Gatlin dated July 26, 2012, "Virgil C. Summer Nuclear Station Unit 1 (VCSNS) - Request for Additional Information (TAC NO. ME7586)" ADAMS Accession No. ML12202A027

Dear Sir or Madam:

South Carolina Electric & Gas Company (SCE&G), acting for itself and as agent for South Carolina Public Service Authority, submitted a License Amendment Request per Reference 1 to adopt NFPA 805 as a basis for the VCSNS Fire Protection Program. Following review and audit of this request the NRC determined that additional information was required and issued a Request for Additional Information (RAI) per Reference 2. The attachments to this letter provide SCE&G's response to the RAIs.

There are no regulatory commitments in this letter. If you have any questions about this submittal, please contact Mr. Bruce L. Thompson at (803) 931-5042.

Virgil C.Summer Station

• Post Office Box 88 . Jen.kinsville, SC. 29065 . F (803) 345-5209

Document Control Desk LAR 06-00055 RC-1 2-0142 Page 2 of 2 I certify under penalty of perjury that the foregoing is correct and true.

/JO -) a- 2e',

Executed on a __ _CuAW __

Thomas D. Gatlin JMW/TDG/wm Attachments: 6

1. Fire Modeling (FM) Request for Additional Information (RAI) Responses

2. Safe Shutdown (SSD) Request for Additional Information (RAI) Responses

3. Fire Protection Engineering (FPE) Request for Additional Information (RAI) Responses

4. Monitoring Program (MP) Request for Additional Information (RAI) Responses

5. Probabilistic Risk-Assessment (PRA) Request for Additional Information (RAI)

Responses

6. Radioactive Release (RR) Request for Additional Information (RAI) Responses

Enclosure:

1

1. CD Rom: Input Files c: K. B. Marsh S. A. Byrne J. B. Archie N. S. Carns J. H. Hamilton R. J. White W. M. Cherry V. M. McCree R. E. Martin NRC Resident Inspector S. E. Jenkins Paulette Ledbetter K. M. Sutton NSRC RTS (CR-06-00055)

File (813.20)

PRSF (RC-12-0142)

Document Control Desk RC-12-0142 Page 1 of 37 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit 1 DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 1 Fire Modeling (FM) Request for Additional Information (RAI) Responses

Document Control Desk RC-12-0142 Page 2 of 37 FM RAI 01 NFPA 805, Section 2.4.3.3, states: "The PSA [probabilistic safety assessment] approach, methods, and data shall be acceptable to the AHJ [authority having authority] ... "The NRC staff noted that fire modeling comprised the following:

The Consolidated Fire and Smoke Transport Model (CFAST) was used to calculate control room (CR) abandonment times.

CFAST was used to calculate Hot Gas Layer (HGL)-temperature for damage determinations in selected fire zones throughout the plant.

Fire Dynamics Tools (FDTs) was used to calculate Zone of Influence (ZOI) dimensions in selected fire zones throughout the plant.

Section 4.5.1.2, "FPRA Quality" of the Transition Report states that fire modeling was performed as part of the FPRA development (NFPA 805, Section 4.2.4.2). Reference is made to Attachment J, "Fire Modeling V&V," for a discussion of the acceptability of the fire models that were used.

Specifically regarding the acceptability of CFAST-for the CR abandonment time study:

a. Provide the input files in electronic format for 6 selected CFAST runs that were conducted, i.e., the input files for the cases with the highest heat release rate (HRR) in Tables 6, 8, 10, 13, 15 and 17 in the Control Room Risk Calculation Report (Section 8 of Calculation No. DC0780B-100, Fire Modeling: CB17.01).

SCE&G Response Input files are provided in an enclosed CD ROM.

FM RAI 01b Specifically regarding the acceptability of CFAST for the CR abandonment time study:

b. Due to presence of a large number of cabinets and control boards, the effective volume of the MCR will be less than if calculated using the length, width and height.

Explain why the presence of this equipment was not considered in the volume estimation.

SCE&G Response The control room CFAST models are based on the actual dimensions of the fire zone without deducting volumes associated with panels as there is no generic guidance in industry fire modeling documents recommending the use of actual as compared to corrected volumes. The volume of the cabinets in the control room was estimated to be slightly less than 14% of the total room volume, based on drawing 201-189 and a cabinet height of 6 ft, as shown in the following table:

Document Control Desk RC-12-0142 Page 3 of 37 Table 1. Cabinet Volume Calculation (Dimensions from Drawing 201-189)

Rhombus shape = area of rectangle + area of triancie

Document Control Desk RC-12-0142 Page 4 of 37 1 XCP6114 I 2488.8 1530 I 4018.8 109089.3 Total Area (inA2) 757.6 Total Area (ftA2) 6 Height of cabinets (ft) 4545.4 Volume cabinets (ftA3)

Volume control room 33500 (50 ft x 67 ft x 10 ft)

_ _- 13.6 = Cabinet'Volume (%)

As a sensitivity case, the dimensions in the control room were reduced such that the volume in the room was reduced by 15%, while maintaining the height of the control room and the-total surface area of the ceiling and walls. The floor is-reated as an adiabatic surface-in CFAST as recommended by model developers at the National Institute of Standards and Technology (NIST) [Peacock et al. 2008]. The conserved quantities may- be transformed into effective enclosure dimensions for use in CFAST using the following equations:

W.D.H=V W, -D =o.85. V 2.H.(W + D)+W,.D, = AB 2.H.(W + D)+W.D=AB where H is the actual height of the enclosure (m), W, is the effective width (m), D, is the effective depth (m), V is the volume of-the control room (M 3), and AB is the boundary surface area (M 2). The equations above are reduced to two equations and two unknowns using the original values of W, D, and H from the control room report (DC0780B-1 00, Table A.2); and the effective width and depth for CB-17.1 are calculated, with the results presented in Table 2. The only values that changed for the other compartments (elevator, office, and kitchen) were the values of the X-position. Table 3 and Table 4 show the revised locations of the targets (upper left corner and lower right corner of compartment) and the fire (center of the control room) for the sensitivity calculation.

Document Control Desk RC-12-0142 Page 5 of 37 Table 2: Revised Compartment Geometry Considering Cabinet Volume Compartment Width, Depth Height 'X-P*sition Y-Positioni Z-Position ftt(m) ft (m) ft(ýft (m) f m CB-17.1 24.13 117.99 10(3.05) 0 0 0 (7.36) (35.96)

Elevator Corridor (63-13) 20 (6.096) 8.5 (2.59) 19 (5.79) 24.13 (7.36) 0 0 Office Space (63-06) 20 (6.096) 17 (5.18) 10 (3.05) 24.13 (7.36) 50 (15.24) 0 Kitchen (63-07) 20 (6.096) 21 (6.40) 10 (3.05) 24.13 (7.36) 28.5 (8.69) 0 Table 3: Revised Heat-Flux Target Locations Number ,Compartmen .Locationx Lo*cation, y Locatipn,'z X Normal Y Normal Z Normal 1 CB-17.1 6.033 88.49 0 0 1 CB-17.1___

_ (1.84) (26.97) 6(1.83) 2 CB-17.1 18.10 29.50 0 0 1 2 CB-17.1 (5.52) (8.99) 6 (1.83) 1 In Table 5, the CFAST results shows a time to abandonment of 8.25 minutes versus 8.50 minutes, which is a reduction of 3% for the case of loss of HVAC, 10 minutes for fire brigade arrival. In addition the probability of abandonment increased by 11%. These results show that the impact is small compared to other variables. The control room report is being updated with the sensitivity results (DC0780B-1 00, Appendix A.8).

Table 5: Summary of Results for Sensitivity Case-Volume of Cabinets included nt...Probability of MCR Average Abandonment.

C onditions Abandonment . Time Base Case 8.99E-02 8.50 minutes Volume of Cabinets 1.00E-01 8.25 minutes Included 1.00E-01 _8.25_minutes

Reference:

Peacock, Richard D., Walter W. Jones, Paul A. Reneke, Glenn P. Forney, CFAST-ConsolidatedModel of Fire Growth and Smoke Transport (Version 6) - User's Guide, NIST Special Publication 1041, Gaithersburg, MD, December 2008.

Document Control Desk RC-12-0142 Page 6 of 37 FM RAI Olc Specifically regarding the acceptability of CFAST for the CR abandonment time study:

c. In Section 4.3.3 of Calculation No. DC0780B-100, it is discussed how the horizontal natural ventilation flow areas are determined. It appears that a characteristic opening fraction from the study of Klote and Milke was used to calculate the average opening fraction for every door. Explain why an average characteristic value for all opening gaps is used instead of the actual door gaps.

Explain why this approach is conservative and describe the results of the parametric analysis to show the dependence of abandonment time on the-ventilation opening area, if such an analysis was performed.

SCE&G Response The table below provides a summary of the Probability of Main Control Room Abandonment and the Average Abandonment Time for the sensitivity case in which the door gap was varied below and above the base case value of 0.02. The case with the fire brigade arriving after 10 minutes and the loss of HVAC was chosen because it had one-of the lowest abandonment times. As the results show, the size of the gap below the door does not have a significant impact on the probability of abandonment or the average time to abandonment. The MCR Abandonment report (DC078OB-1 00) is being updated with the sensitivity results (Appendix A.7).

Summary of Results for Sensitivity Case Doorway Opening Fraction Probability of MCR Abandonment Average Abandonment Time 0.001 9.08E-02 8.46 Minutes 0.02 (Base Case) 8.99E-02 8.50 Minutes 0.03 8199E-02 8.50 Minutes FM RAI Old Specifically regarding the acceptability of CFAST for the CR abandonment time study:

d. Explain why the default drop-off and zero flow pressure values in CFAST were used.

Explain how these values are consistent with the plant-specific HVAC data.

SCE&G Response The drop-off and zero flow pressures are only used in the functioning HVAC cases. The purpose of the drop-off and zero flow pressures are to determine at what point the ventilation is turned off. The bounding result is when the HVAC is not functioning at all. The results from the

Document Control Desk RC-12-0142 Page 7 of 37 non-functioning HVAC cases show the lowest abandonment times and are therefore the most conservative. Therefore, the pressures are not an important input to the overall results.

FM RAI Ole Specifically regarding the acceptability of CFAST for the CR abandonment time study:

e. Calculation No. DC0780B1OO mentions (page 16) that the fire location was chosen such that it produces worst case fire scenarios. Explain the justification for this assertion.

In addition, it is noted (page 21) that the fire location for all simulations is at floor level in the center of the room. It is stated that the central location of the fire limits the effects of the walls on the fire growth and the plume-behavior. It is not clear to the NRC staff what is meant by 'limits the effects of the walls on fire growth and plume behavior.' Explain this statement in greater detail.

SCE&G Response The statement is intended to clarify that the analysis is based on plume correlations for entrainment and is not considering the wall or corner effects. Plumes are-entraining air throughout the circumference since the version of CFAST used- does not account for such effects.

The fire base height is assumed to be at the floor of the control room, rather than 0.3 m (1 ft) below the top of the electrical panel as is recommended in NUREG/CR 6850, Supplement 1

[2010]. Unlike fire exposure analysis to overhead components, a habitability calculation is typically dominated by the visibility threshold parameter and the elevation of the hot gas layer, which in turn are sensitive to the smoke production rates. The smoke production rate is not generally maximized when the fire height is maximized but rather when the height between the fire base and the ceiling is maximized due to a longer entrainment distance [Heskestad, 2008].

Minimizing the fire elevation will be the most conservative when the visibility threshold is exceeded before the temperature threshold regardless of the rate at which the hot gas layer descends. As mentioned in the report, visibility is the dominant abandonment criterion and a minimized fire elevation, therefore, is conservative.

Reference:

Heskestad, G. (2008), "Fire Plumes, Flame Height, and Air Entrainment,"

Section 2-1, The Society of Fire Protection Engineers (SFPE)Handbook of Fire ProtectionEngineering,4 th Edition, Society of Fire Protection Engineers, Bethesda, MD, 2008.

FM RAI Olf Specifically regarding the acceptability of CFAST for the CR abandonment time study:

f. The abandonment time analysis has not considered fire spread from one cabinet to adjacent cabinets. However, Section 9.2 and Table 20 of Calculation No. DC0780B-

Document Control Desk RC-12-0142 Page 8 of 37 100 discuss the phenomenon of fire propagating to adjacent cabinets. Provide justification for not considering a scenario involving multiple cabinet fires in the calculation for time to abandonment.

SCE&G Response The Fire PRA does include the risk contribution (i.e. CDF and LERF) scenarios where fire propagates from panel to panel in the quantification (DC0780B-1 00, Section 9). At the same time propagating fires were not included in the CFAST "time to abandonment" analysis. The reason for not including the propagation is that heat release rates up to 2.2MW reached in 12 min (right-most tail of the distribution for HRR, Case 5) were included in the analysis, which-bounds heat release rates of more than one panel having more "likely" fire sizes based on the probability distribution. In addition, for these higher heat release rates of 2.2 MW, abandonment times are calculated to be relatively short (5 to 10 min) and before the fire reaches the peak, which suggest that similar times would be obtained if the heat release rates are higher due to propagation.

FM RAI Olg Specifically regarding the acceptability of CFAST for the CR abandonmenttime study:

g. It is assumed that the HRR of the transient fires is bounded by the large fixed ignition source fire (electrical cabinets with unqualified cable, fire in multiple cable bundles).

i. The growth time for the cabinet fire is 12 minutes, whereas according to FAQ-52, the average growth time for transient fires could be 0, 2 or 8 minutes, depending on the actual transient combustible. Provide additional justification for the assumption that the large fixed ignition source fire is bounding for transient fires.

ii. Though the HGL temperature and optical density is not dependent on the location of the fire, targets are used to assess the radiant heat flux criterion. It would seem that the location of a transient fire could affect these calculated target heat fluxes. Provide additional justification for not postulating any transient fires for the purpose of calculating CR abandonment.

SCE&G Response

i. The time to abandonment results listed in Table 10 of the Main Control Room abandonment study suggests that a heat release rate of over 500kW is necessary to force abandonment (DC0780B-100). The peak heat release rate for transients is 317 kW. Therefore, it was assumed that the CFAST simulations available would bound the results for the transient regardless of the faster growth rate as a peak of 317 kW is not expected to produce abandonment conditions based on CFAST results for the different ventilation conditions evaluated.

Document Control Desk RC-12-0142 Page 9 of 37 ii. For determining abandonment conditions based on the criteria listed in NUREG/CR-6850, the heat flux from the room surfaces and hot gas layer to the targets was calculated as noted in the question. The localize fire effects are not considered for abandonment. Notice that these localized effects are not only limited to a transient fire but also to any fixed ignition source panel. That is, it is assumed that operators will not perform actions other than fire suppression and control in the immediate vicinity of a relatively large fire, fixed or transient. In contrast, the abandonment analysis is based on global room conditions. In addition, the transient fires are failing the panels nearby them, which is further indication that no actions could be taken in those panels.

FM RAI 01 h Specifically regarding the acceptability of CFAST for the CR abandonment time study:

h. The report provides results of a sensitivity study for the -times to abandonment for different arrival times of the fire brigade with and without forced ventilation.

Provide the results from any additional sensitivity studies performed to address fire location, ambient temperature, etc.

SCE&G Response For the control room abandonment calculation, in addition to the original sensitivity cases included in the report ffire brigade arrival times and forced ventilation operability), two additional sensitivity cases were added: (1) the impact of including cabinet volume and (2) the impact of varying the gap below the door. The details of these sensitivity studies are given in the responses to FM RAI 01 (b) and (c).

For the cabinet volume sensitivity, the volume of the cabinets was estimated to be about 15% of the total room volume based on the measurement of the cabinets in Drawing 201-189. The CFAST results show a time to abandonment of 8.27 minutes versus 8.50 minutes, which is a reduction of 3%. These results show that the impact is small compared to other variables. The control room abandonment report (DC0780B-1 00 Rev 1) was updated with the sensitivity results (Appendix A. 8).

For the door gap sensitivity, the doorway opening fraction was varied from the base case of 0.02 to a low value of 0.001 and a high value of 0.03. The results showed no difference between the base case and the high value. For the low value, the results differed only slightly

(-1%). The control room abandonment report (DC0780B-100 Rev 1) was updated with the sensitivity results (Appendix A.7)

For the control room, a sensitivity calculation for ambient temperature was not performed because the room temperature is controlled to within a narrow temperature range. In addition, the temperature was not the limiting factor that led to abandonment; it was the visibility.

Document Control Desk RC-12-0142 Page 10 of 37 FM RAI 01i Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

i. Calculation DC078OB-001 suggests that an opening of 0.5-in. is assumed below every door if the leakage area is not known, as per NUREG-6850 Section F.2. It appears that NUREG-6850 prescribes this assumption only for HGL calculations. Clarify whether this assumption was used for every fire modeling calculation or just for HGL calculations. Also, the methodology used in the MCR risk calculation seems to differ from this approach. Provide further justification for this approach to door leakage areas.

SCE&G Response The 0.5-in opening, which corresponds to a leakage area fraction of 0.003, was used only in the hot gas layer calculations for the performance-based calculations. The leakage fraction area fraction of 0.03 was used in the individual zone reports for hot gas layer calculations. For the control room report (DC0780B-001), a characteristic opening fraction of 0.02 was used.

However, to address the uncertainty in the values, a sensitivity study for the door gap size in the main control room abandonment calculation was performed. As mentioned in the response to FM RAI 01c, the sensitivity results show no significant impact on the time to abandonment or probability of abandonment when ranging the door gap fraction from 0.001 to 0.03. The results of the sensitivity case are being incorporated into the control room report (DC0780B-1 00, Appendix A.7).

For other fire areas, several sensitivity cases were run for the hot gas layer CFAST calculations in which the equivalence ratio for the natural ventilation case was outside of the validation range (See DC0780B-366). In cases when the leakage area fraction was increased to bring the equivalence ratio into the validation range, the conclusions were unchanged.

FM RAI O1j Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

j. It is not clear how the dimensions of a "transient zone" are determined. The report states that the transient zones are larger compared to the traditional ZOI dimensions.

Clarify how the dimensions for the transient zones are determined.

SCE&G Response The location of the transient zones are selected based on judgment by inspecting the fixed ignition source and cable tray layout of the zone. The dimensions are calculated based on distanced produced by the PCCKS queries once column line coordinates are assigned to each

Document Control Desk RC-12-0142 Page 11 of 37 transient zone. For a specific example, see section 7.3.1.5 in DC0780B-035. These reports, also include figures depicting the location of the transient zones. The dimensions of the transient zones are fairly large compared to the zone of influence calculated using fire modeling for the 98th percentile heat release rates. This selection of larger zones ensures that targets near sources are not missed in the analysis as the zone expands a relatively long horizontal distance from the ignition sources and vertically extends from floor to ceiling (CB06: DC0780B-088, Figure 2 and Section 7.3.1.5).

FM RAI 01k Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

k. On page 12 of Calculation DC078OB-001 (bullet No. 4 of the second paragraph described under the engineering judgment scenario), it is assumed that some fire suppression activity will be initiated before fire spreads from one transient zone to others. Explain how this is ensured when determining the transient zone size.

SCA&G Response The assumption of some fire suppression activity will be initiated before fire spreads from one transient zone to another is based in part on the recognition made in the latest manual suppression curves described-in Chapter 14 of the Supplement 1 to NUREG/CR-6850 where the fire brigade response time of an "average plant" is included within the response time to control and suppress fires. The FAQ "recognizes that manual suppression is a continuous activity that can begin once the fire is detected, rather than rely primarily on fire brigade suppression efforts". In addition, since the transient zones are relatively large, it is expected that a large fire will be necessary to propagate between adjacent transient zone, which suggests that detection and activities to control the fire will likely be under way.

FM RAI 011 Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

1. Based on the last paragraph of page 13 (Calculation DC0780B-001), it appears that the decision to perform detailed fire modeling for a transient zone was not based on some numerical screening threshold, but was based on the judgment of the PRA analyst. Clarify what criteria were used to screen areas where fire modeling would be performed. Include a justification for why the criteria are acceptable.

SCE&G Response During the process of FPRA model quantification using the transient zone level target mapping suggests locations of concern (i.e. transient zones with relatively high risk) that require further detailed analysis. As a result, some transient zones received further detailed analysis and

Document Control Desk RC-12-0142 Page 12 of 37 others don't based on the overall plant quantification results and risk ranking. An example of a transient zone requiring additional analysis is the ABO1.21.02 in document DC0780B-027, Section 7.3.1.1.

FM RAI Olm Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

m. Page-22 (Calculation DC0780B-001) describes the method, by which targets have been identified. It is stated that if the specific location of the target is not known, it has been assigned to all the scenarios. Clarify how the location of the target was determined for every scenario that it has been assigned to and how the~guidance on fire location discussed on page 27 has been applied for this case.

SCE&G Response In the case of cable trays (i.e. raceway sections), the cable and raceway database maintains a-set of plant coordinates for each of the raceway sections. This capability allows the FPRA analysts to define scenarios by plant coordinates that serve as inputs to a query in the cable and raceway database that retrieves the cable trays within the set of coordinates. Coordinates are defined by six points in space characterizing a volume (xl, x2, yl ,y2, zl ,z2). In the case of conduits and cable end points, the cable and raceway database do not have plant coordinates associated with them. Therefore, mapping of conduits and end points was done "by hand" by identifying their location on electrical drawings and assigning a transient zone to them.

FM RAI Oln Regarding the acceptability of the PSA approach, methods, and data in general as discussed in the paragraph above:

n. Clarify what is meant by "bounding vertical and horizontal distances were used for calculating detection time" (page 27 of Calculation DC0780B-001). It is not clear how these values were calculated for a room that has numerous obstructions that couldsignificantly delay detection and activation. Provide examples from specific analyses in which these activation times are calculated.

Also, elaborate on how the suppression time was conservatively estimated as stated in Section 6.2 2(b) of Calculation DC0780B-001.

SCE&G Response The statement of bounding "vertical and horizontal" distance was intended to indicate that for smoke detection calculations, the height of the fire zone was used as the vertical distance, and 10 to 20 ft was used as horizontal distance. The smoke detection model in NUREG 1805 was used so there is no consideration for obstructions in smoke detection calculations. When

Document Control Desk RC-12-0142 Page 13 of 37 suppression was credited, there are two approaches: 1) detailed DETACT calculations are available to justify activation (examples in CB06: DC0780B-088, Appendix E; 1B25.01.02:

DC0780B-1 92, Appendix E) or 2) the sprinkler heads are within the cable trays (e.g. CB10) and it is assumed that sprinklers will activate before fire damages the nearby tray as the sprinkler has a lower activation temperature than the cable damage threshold. It should be noted that conservatism is achieved by assuming the ZOI targets are damaged before suppression starts.

That is, there is no credit for suppression for the ignition source and targets within the ZOI. For example, in the approach 2 described earlier, the sprinkler will not be credited in protecting the nearby tray. However, it will be credited in preventing propagation after damage to the nearby tray is assumed.

FM RAI 01o

o. What are the criteria used in the generic methodology described in Calculation DC0780B-001 for determining when a more complicated fire model is necessary for evaluation of a given problem (e.g., prediction of sprinkler activation in a space with complicated geometry).

SCE&G Response During the process of FPRA model quantification using the transient zone level target mapping suggests locations of concern (i.e. transient zones with relatively high risk) that require further detailed analysis. As a result, some transient zones received further detailed analysis and others don't based on the overall plant quantification results and risk ranking. Example of a transient zone requiring additional analysis is the ABO1.21.02 in document DC0780B-027, Section 7.3.1.1.

FM RAI 01p

p. On page 27 of Calculation DC0780B-001 there is discussion about uncertainty of fire location. Location factor is not explicitly discussed here. Provide additional information about how the generic methodology treats fires located close to walls or corners.

SCE&G Response The fire location factor is assigned to each ignition source in the fire modeling database, based on walkdowns. The fire location factor (a value of 2 for fire near a wall, 4 for fire near a corner) is multiplied by the critical heat release rate to calculate the severity factor. The critical heat release rate is calculated based on a fire condition and a shortest distance.

Document Control Desk RC-12-0142 Page 14 of 37 FM RAI 01q

q. On page 54 of Calculation DC078OB-001 there is a discussion about the targets. It is reported that since there is approximately 10% unknown cable and the rest is known to be thermoset, the whole plant is considered to have only thermoset cable. Provide additional justification why it is not necessary to address this 10% of unknown cable more conservatively, if any non-thermoset cables are located in the fire zones for which detailed fire modeling was conducted.

SCE&G Response Recent cable/raceway identification work has reduced the number of unknown cable from 10%

to less than 2% of the total linear feet of cable installed. This study is captured in TR0780E-004 Attachment-Admin-03 entitled, "Fire Protection Engineering Equivalency Evaluation- Existing Electrical Wiring-Cabling." See additional information in FM RAI 02.

FM.RAI 01r

r. On page 60 of Calculation DC0780B-001, there is discussion about the assumption of an 'incubation time' of ten minutes from hotwork fire initiation to propagation to the next tray. This same approach is used on page 63 in Section E.4 on Junction Boxes.

Describe whether this is the methodology utilized and if so, provide additional justification for this assumed time period.

SCE&G Response In the specific case of junction boxes, there is no incubation time assumed- in the FPRA.

Section E.4 of calculation DC0780B-001 has been updated to reflect this. All the damage times in the fire modeling database for JB fires are set to zero minutes. However, the modeling approach in the FPRA assumes an incubation time for hotwork fires of 10 minutes. This is primarily based on the expectation that these fires will start small due to welding slags or sparks. The manual suppression time constant- for hot work fires in Table 14-2 of the supplement 1 to NUREG/CR-6850 of 0.188 suggests that the average time for extinguishing these fires is 1/0.188 = 5.3 min. As listed in Fire Modeling: Generic Methodology, Table E-1, all the fires in the Fire Events Database were controlled or suppressed before propagation to a full section of a cable tray or a combustible package, which are the established fires postulated in a Fire PRA. Therefore, a value larger than 5.3 was selected. The use of 10 min results in a manual non suppression probability of 0.15. In the VC Summer Fire PRA, this indicates that there is an 85% chance that the fires will not "engulf' or damage a full cable tray section. Since no hotwork fire has generated such damage by inspection of the fire events database, a 15%

damage probability was assumed to be bounding.

FM RAI Ols

s. On page 26 of Calculation DC078OF-096 for Fire Modeling Fire Area CB10 VFDR, it is discussed that this staging area also requires that no combustibles be placed within

Document Control Desk RC-12-0142 Page 15 of 37 one foot of the adjacent walls, so that wall effects of the fire source do not have to be accounted for. Provide additional justification for this critical dimension of one foot of separation, beyond which wall and corner effects can be ignored. Expand the response to this question for the other fire areas where this critical separation distance is utilized (CB12, CB18 and IB3).

SCE&G Response Fires that are located near a wall boundary or near a corner will experience a reduced air entrainment and a force imbalance on the plume that tends to push the flames against the boundary [Williamson et al., 1991]. The offset distance between a fire and a wall or a corner boundary at which the boundary influences the entrainment likely varies with the fire size. It has been observed to range from as little as a few inches to several feet (Williamson et al., 1991, Nuclear Regulatory Commission (NRC), 2005). When a fire is considered to-be influenced by a wall or a corner boundary, the image method is used to quantify the effect on the entrainment.

The image method treats a fire that is located in a corner as one-quarter of a virtual fire with a-symmetry plane coincident with each wall forming the corner (NIST-GCR-90-580, 1990).

Likewise, -a fire that is located near a wall is treated as one-half of a virtual fire with a symmetry plane along the wall.

A sensitivity case was investigated for each of the performance-based fire modeling reports in which the edge of the staging area was located 1 ft from a wall or corner (CB1 0, CB1 2, and CB1 8). A fire was simulated by increasing the following inputs by a factor of two for a wall effect or a factor of four for the corner effect: the fire size, the fire enclosure volume, the fire enclosure wall and ceiling areas, and the natural ventilation area. The peak temperatures for the maximum expected case and the sensitivity case are listed in the table below. The results show that by using the image method to determine the impacts of a wall or corner fire, the peak temperature in the room is still below the damage temperature of 205 °C. These results were added to the performance-based fire modeling reports for each fire area (CB1 0: DC078OF-096, CB12: DC0780F-097, CB18: DC0780F-103).

compaftment (wall - Peak Temperature Maximum Peak Terriperature with wall or corner impacts or corner) Expected Fire .inclUded CB10 (corner) 141 TC 187 °C CB12 (wall) 125 C 155 °C CB18 (wall) 151 °C 183 *C IB131 (neither) N/A; Staging Area Not Near Wall or Corner

References:

NIST-GCR-90-580 (1990), "Development of an Instructional Program for Practicing Engineers Hazard I Users," Barnett, J. R. and Beyler, C. L., National Institute of Standards and Technology, Gaithersburg, MD, July, 1990.

Nuclear Regulatory Commission (2005), "Appendix F - Fire Protection Significance Determination Process (SDP)" Inspection Manual Chapter (IMC) 0609, Nuclear Regulatory Commission, Washington, DC, February 28, 2005

Document Control Desk RC-12-0142 Page 16 of 37 Williamson, R.B., Revenauqh, A. and Mowrer, F.W., 1991. Ignition Sources In Room Fire Tests And Some Implications For Flame Spread Evaluation. Fire Safety Science 3: 657-666.

FM RAI Olt

t. During the walkdown of fire zone 1B25, the staff reviewed the analysis of the fire scenarios in 1B25.01.02 subzone, which credited fire suppression/detection. The staff has the following questions about the analysis performed in this fire zone:

During the audit, the staff noted that it is discussed in the detailed fire modeling report that standard response sprinklers are used in the fire zone and therefore a Response Time Index (RTI) of 130 (m-s)0.5 was used for the analysis. The licensee justified this value for the RTI by way of reference to NUREG 1805, "Fire Dynamics Tools", "which provides a-generic RTI value of 130 (m-s)0.5 for standard response heads with a fusible link." However, in Chapter 10 of NUREG 1805, there is a note about selecting the RTI of a sprinkler element which states, "the actual RTI should be used when the value is available." Provide justification for the RTI value chosen for this analysis and describe how that value compares with the RTI of the actual sprinklers in the fire zone. Apply this response to any additional fire zones where detailed fire modeling was conducted to address credited suppression systems.

For at least one of the fire scenarios in this zone, a sprinkler activation calculation was performed to determine if/when a sprinkler would actuate. In this fire zone, there are sprinklers and smoke detectors at ceiling level and additional sprinklers at an intermediate level several feet below the ceiling. The staff has the following questions about how that configuration may impact the existing analysis:

The FDT calculation for sprinkler activation assumes that the fusible link is located a short distance (a few inches) below a flat ceiling extending in all directions. The current analysis credits intermediate level sprinklers activating, however, these heads are located several feet below the ceiling overhead. In addition, these intermediate level sprinklers were not observed to have "heat collectors" over the heads. Provide justification for the assumption that this configuration does not affect the conclusions of the analysis.

During the audit, the staff was informed that the sprinkler system is a preaction msystem and therefore requires a signal from the smoke detection system in order for water to flow to the heads. Therefore, it is important for the analysis to include the timing of smoke detection activation with respect to intermediate level sprinkler activation.

Provide further clarification about the timing between smoke detector activation and intermediate-level sprinkler activation.

Document Control Desk RC-12-0142 Page 17 of 37 SCE&G Response In 1B25.01, the sprinklers are only credited in subzone 1B25.01.02, and only for the large pump oil fires, with a heat release rate of 1917 kW. Notice that sprinklers are not credited for smaller size fires. The sprinklers near the ceiling, not the intermediate sprinklers, are credited for preventing propagation from the transient zone, in which the pump is located, to the rest of the compartment and credited to occur within 20 minutes of the start of the fire. All the targets within the transient zone are considered damaged, but the targets outside of the transient zone are credited with sprinkler protection. As a result, the time to smoke detection and sprinkler activation is expected to be sooner than 20 minutes as suggested by the activation calculations.

For the time to activation calculation, the sprinklers near the ceiling are used, rather than the intermediate sprinklers because the sprinklers are credited for preventing the spread of the fire beyond the transient zone. They are not credited for preventing damage within the transient zone. A separate time to detection calculation for the smoke detectors was not performed.

Since sprinklers are not credited for activating until after 20 minutes, it is assumed that all targets within the transient zone would be damaged and that sufficient smoke would be generated to allow for smoke detection and sprinkler activation.

The sprinklers in 1B25.01 are Standard Fusible Link type sprinklers. The sprinklers were manufactured in the late 1970s or early 1980s by ASCOA, and the specific RTI values for these sprinklers are not available; therefore, the generic RTI was selected, 130 (m-s) 112. A sensitivity case was investigated in which the RTI was increased to 235 (m.s) 11 2, the highest value in the FDT spreadsheet for sprinkler activation (NUREG-1805, Chapter 10). The time to activation increased from 2.4 minutes to 4.4 minutes for the electrical cabinet fire with a heat release rate of 702 kW (EC-2) and increased from 1 minute to 2 minutes for the large oil fire (ZO1). This calculation shows that the time to activation of the sprinklers is soon enough to prevent spread-of the fire to other transient zones, as it is less than 20 minutes.

CB06 is another fire area in which an estimate of RTI was needed to calculate the response time of the heat detectors (DC0780B-088, Appendix E). VCS Document 1MS-55-04, Sheet 1, indicates that the activation temperature of the heat detectors in CB06 is 140°F (600C).

NUREG-1-805 indicates that the RTI of the heat detector is a function of the activation temperature and the listed spacing. The maximum radial distance between detectors is approximately .10 feet, calculated from the sprinkler drawing (VCS Document 1MS-55-40, Sheet 9, Revision 2) which shows detector spacing of 18' 10" (east-west orientation) and 6' 5" (north-south orientation). The RTI of the heat detectors in CB06 is estimated to be approximately 321 (m-sec) 1 2, based on data presented in Table 2 in Chapter 12 of NUREG-1805 by assuming an activation temperature of 145 0 F and a UL listed spacing of 10 ft. Assuming RTI of 321 (m.sec)1 2, the time to detection for an electrical cabinet fire with a peak heat release rate of 702 kW, was calculated to be 2.37 minutes (DC0780B-088, Appendix E). As a sensitivity calculation, the RTI was increased to 490(m.sec) 2 , which is the highest value in the FDT spreadsheet (NUREG-1805, Chapter 12) for heat detector response time, and the detection time was calculated to be 3.6 minutes. The time to detection, even if the higher value of 3.6 minutes is used, is small compared to the time at which suppression is credited, which is 11 minutes. In the Fire PRA calculations, all targets within the zone of influence of the cabinet fire (i.e., within the same transient zone) are assumed to fail in the first 11 minutes; and the suppression system is credited in preventing the spread of the damage outside of the transient

Document Control Desk RC-12-0142 Page 18 of 37 zone. Since the calculated time to detection of the heat detection system is small (<4 min) compared to the time at which suppression is credited (11 min), the actual value of the RTI is not significant.

FM RAI 01lu

u. During the audit, the staff walked down fire zone CB10, which is one of the performance- based fire modeling zones. In this zone a "transient staging area" has been proposed in the fire modeling analysis. In this southwest corner of the compartment, the staff observed three vertical pipes directly adjacent to the staging area. One of the pipes is steel, but the other two are copper pipes that are jacketed with some insulation that is non-metallic. It was not obvious what the material flammability characteristics of this insulation are and if it could be ignited from an adjacent transient fire. Provide additional information about this insulation and how it could affect the conclusions of the analysis.

SCE&G Response The vertical pipes identified during the walk down are hot water lines and the insulation is used as a freeze protection on indoor lines. The insulation is fiberglass, per ASTM C547, having a thermal conductivity less than or equal to 0.3 Btu-in/hr-ft2-°F (0.043 W/m-K) at 200 OF (93 °C).

The insulation is jacketed with a vapor barrier laminate of aluminum foil and glass cloth, with lap adhesive. Due to the non-combustible nature of the material, the insulation is not considered an intervening combustible for modeling purposes. This information was added to the performance-based fire modeling report for CB10 (DC0780F-096 Rev. 2).

FM RAI 01v

v. During the audit, the staff walked down fire zone IBI 1, which is one of the performance-based fire modeling zones. In this zone a "transient staging area" has been proposed in the fire modeling analysis. This staging area is identified as being located 4.5-ft from a nearby VFDR cable tray. However, during the walkdown, the staff observed several exposed cables extending from this variance from deterministic requirement (VFDR) tray in the direction of the staging area. The staff confirmed that the licensee measured the distance to the staging area from the edge of the VFDR cable tray, rather than this exposed cable. Provide additional information about this configuration and how it could affect the conclusions of the analysis.

SCE&G Response Figure 2 of performance based fire modeling report for IB131 (DC0780F-173 Rev 2) has been revised to reflect the exposed cables that were observed during the walkdown. The staging area has been moved 2 ft towards the south, resulting in a distance of 6.5 ft to the nearby VFDR cable tray (changed from 4.5 ft), and 8.5 ft from the staging area to the fixed ignition source

Document Control Desk RC-12-0142 Page 19 of 37 (changed from 9 ft). The revision had no impact on the results and the conclusions were unchanged.

FM RAI 01w

w. During the audit, the staff discussed the fire modeling analysis performed in fire zone CB4. Part of this analysis utilized CFAST to calculate the HGL temperature, which was used to determine whether damage would propagate outside the originating transient zone. Since CFAST is a zone fire model, the HGL temperature that is calculated is an average temperature for the upper gas layer in the compartment. In reality, a higher HGL temperature will be observed closer to the fire source than far away from the fire source. The staff request that the licensee provide additional information to confirm that the fire postulated will not lead to damage beyond the originating transient zone.

It is of interest that the limits of applicability, with respect to room dimension aspect ratios for CFAST, were slightly exceeded for the L/H ratio. Explain why this is acceptable.

SCE&G Response To justify the use of CFAST for the fire areas that were subjected to detailed fire modeling in support of the Fire PRA, the dimensionless parameters that describe the fire scenarios were calculated and compared against model validation parameters, as specified in NUREG-1824.

For CB04, the compartment aspect ratio and the equivalence ratio were outside of the validation range. A sensitivity case was run with CFAST, in which the room dimensions and the air flow openings into the room were adjusted to be within the validation range. The results for the sensitivity case and the base case showed no hot gas layer temperatures above the damage criteria. Since the results for both the sensitivity case and the base case, were the same, the use of CFAST is considered justified (See DC078OB-366).

Since the hot-gas layer is not expected to reach damage temperatures, the damage is assumed to be limited to the targets within the transient zone; and, damage does not impact the whole compartment. The fact that a higher HGL temperature will be observed closer to the fire source is included in the calculations; because when a transient fire occurs, it is assumed that all targets in that transient zone are damaged. The CFAST results are used to determine if damage outside of the transient zone are impacted.

FM RAI Olx

x. Provide the CFAST input files in electronic format for 9 selected fire zones (that were walked down during the site audit on June 5, 2012) that utilized detailed fire modeling calculations, i.e., the input files for CB10, CB12, CB18, IBIl, CB15, CB04, CB06, ABO1 (21.02), and IB25 (01.02).

Document Control Desk RC-12-0142 Page 20 of 37 SCE&G Response The CFAST model input files for the cases indicated are provided as requested, except for ABO1.21.02 and 1B25.011.02, for which no CFAST modeling was performed. In those cases, the Excel spreadsheets for the hand calculations are provided.

FM RAI 02 NFPA 805, Section 2.5, requires damage thresholds be established to support the performance-based approach. Thermal impact(s) must be considered in determining the potential for thermal damage of SSCs. Appropriate temperature and critical heat flux criteria must be used in the analysis.

Section C.10 of Calculation DC078OB-001 includes a discussion about the targets. It is reported that there is approximately 10% unknown cable and the rest is known to be thermoset, and the whole plant is considered to have only thermoset cable. Provide the following information:

a. For the 10% of unknown cabling, characterize the installed thermoset and thermoplastic cabling in the power block. Specifically with regard to the critical damage threshold temperatures and critical heat flux threshold as described in NUREG/CR-6850. Provide a statement regarding the extent of installed thermoplastic cable insulation.

b. If necessary, explain how raceways with a mixture of thermoset and thermoplastic cables were treated in terms of damage thresholds and HRR and fire propagation.

c. If thermoplastic cabling is present, discuss the impact on ZOI size due to increased HRR and fire propagation.

d. If thermoplastic cabling is present, discuss self-ignited cables and their impact to additional targets created.

e. Explain if and how covered, partially covered, or holes in closed raceways affected the damage thresholds of cables used in the analysis.

f. Describe how the damage thresholds for non-cable components (i.e., pumps, valves, electrical cabinets, etc.) were determined. Describe whether there were any non-cable components that were assigned damage thresholds different from those for cables and, if so, explain how those thresholds have been evaluated for damage and ignition.

g. If more targets are identified, describe the impact to core damage frequency (CDF) and large early release frequency (LERF), as well as changes in CDF (delta (A) CDF) and changes in LERF (delta (A) LERF) for those fire zones affected.

Document Control Desk RC-12-0142 Page 21 of 37 SCE&G Response

a. Recent cable/raceway identification work has reduced the number of unknown cable from 10% to less than 2% of the total linear feet of cable installed. This study is captured in TR0780E-004 Attachment-Admin-03 entitled, "Fire Protection Engineering Equivalency Evaluation- Existing Electrical Wiring-Cabling." Per NUREG 7102, All Kerite FR3 or equivalent cable is being treated as thermoplastic for damage criteria analysis purposes.

b. Although, recent cable/raceway identification work has reduced the number of unknown cable from 10% to less than 2% of the total linear feet of cable installed, if a raceway did have a mixture of both thermoplastic and thermoset cabling, the damage criteria for thermoplastic cabling was used throughout the plant. Because at least 98% of the cables are known to be IPEEE383-174 or better with >90% being Kerite, which is a thermoset cable, the flame spread rate and the HRR for thermoset cables were used in the analyses, as mentioned in DC0780B-001, Sections A.8.1 and A.8.3. However, due to guidance in Section A.5.2 of NUREG-1 805 and NUREG 7102, the failure criteria for the cables are taken to be equivalent to the ignition temperature of thermoplastic cables (205'C). This is due to the high concentration of Kerite cable throughout the entire plant complex.

c. N/A. Cables in the plant are assumed damaged at the thermoplastic damage criteria.

Refer to answers A & B.

d. N/A. Cables in the plant are assumed damaged at the thermoplastic damage criteria.

Refer to answers A & B.

e. Cables are assumed damaged at the thermoplastic damage criteria even if they are routed in closed raceways. Metal covers for raceways are assumed not to have any effect on damage thresholds.

f. In locations with sensitive electronics, CB06 and 1B14, the panels are assumed damaged in the first-damage state in the progression of the fire scenario, where suppression -is not credited. Therefore, the damage threshold is not applicable because the panels within the particular transient zone impacted by the fire are failed immediately.

g. No additional targets were identified.

FM RAI 03 NFPA 805, Section 2.7.3.2, "Verification and Validation," states: "Each calculational model or numerical method used shall be verified and validated through comparison to test results or comparison to other acceptable models."

Section 4.5.1.2, "FPRA Quality" of the Transition Report states that fire modeling was performed as part of the FPRA development (NFPA 805 Section 4.2.4.2). Reference is

Document Control Desk RC-12-0142 Page 22 of 37 made to Attachment J, "Fire Modeling V&V," for a discussion of the verification and validation (V&V) of the fire models that were used.

Furthermore Section 4.7.3 "Compliance with Quality Requirements in Section 2.7.3 of NFPA 805" of the Transition Report states that, "Models and numerical methods used in support of compliance with 10 CFR 50.48(c) were verified and validated as required by Section 2.7.3.2 of NFPA 805."

Regarding the V&V of fire models:

a. It has been mentioned in both Generic Methodology calculations (DC0780B-001 and DC078OC-001) that a tool called VC Summer Fire Modeling Database has been enveloped using Microsoft@ Access. Explain how it was developed along with the underlying mathematical bases for this package and the supporting spreadsheets that were used to perform the FDT calculations. In addition, explain how this package and the supporting spreadsheets were verified.

b. Provide technical details to demonstrate that the fire models have been applied within the validated range of input parameters, or to justify the application of the model outside the validated range reported-in NUREG-1824 or other V&V basis documents.

c. In Attachment J of the LAR, the normalized parameter Lf/Hc was calculated and reported to be within range for the four fire areas that had performance-based calculations, per Section 4.2.4.1 of NFPA 805. The most recent draft of NUREG 1934, "Nuclear Power Plant Fire Modeling Application Guide," which explains how to calculate these parameters, corrects a clerical mistake for this normalized parameter.

This parameter should be (Lf+Hf)/Hc (to account for fire elevation). Please confirm that this normalized parameter was calculated correctly and if not, describe whether the new value is within the parameter's range of applicability. Provide justification for the use of the fire model in cases when the new parameter value is outside the validation range.

d. It appears that these normalized parameters were only calculated for fire zones where the methodology described in Section 4.2.4.1 of NFPA 805 was used. Provide.a discussion of the normalized parameters for all fire zones where fire modeling was conducted.

e. During the audit, the staff observed that the spreadsheets included in Attachment 1 (Supporting Documents) of the reports for the zones where detailed fire modeling was performed show hand calculations. The equation that was used to calculate the plume temperature at a specified height shown on some printouts (e.g. one of the printouts in the report for zone AB101.21.02) is different from that shown on the 09_PlumeTemperature_ Calculations FDT spreadsheets. Provide the origin of the version of Heskestad's equation that was used and explain how this equation was verified and validated.

Document Control Desk RC-12-0142 Page 23 of 37 SCE&G Response

a. A validation study for the main quantification module of the database has been completed and is provided below. After the verification and validation was completed, additional modules, including a module for performing the screening of multi compartment fires, and the module for generating FRANX input files were used and evaluated routinely during the cut set review and Fire PRA quantification process.

Verification and Validation: A Test Case The fire modeling functions were validated as follows: 1) input values in FMDB for two test cases were compiled and listed in Table land Table 3 respectively, 2) the functions were solved by hand and results compared with the ones calculated by FMDB and reported in the Zone of Influence Form, as illustrated in the screen captures in Figure 1 and Figure 2. Finally, Table 2 and Table 4 summarize the V&V results for the fire modeling functions. The spreadsheet FDB-VandV worksheet FixedSum shows the development and disposition of two fixed ignition cases. The cases demonstrate source frequency development and scenario development. The spreadsheet FDB-VandV worksheet TranSum demonstrates the same process but for two transient cases.

Table 1: Input values for V&V case 1

&VSudy-fr on "o Inlunc An tblEquipmentCount Fire location factor 1 tblEquipmentCount Automatic detection type Spot type smoke detection tblDetType Probability of detection failure 0.05 tblEquipmentCount Vertical distance to detection (ft) 8 tblEquipmentCount Horizontal distance to detection (ft) 0 tblEquipmentCount Automatic suppression type Automatic sprinklers tblSuppType Probability of suppression failure 0.02 tblEquipmentCount System initiation Automatic tblEquipmentCount Sprinkler type Standard response bulb/Ordinary tblSprinklerTypes RTI 235 tblSprinklerTypes Activation temperature (F) 165 tblEquipmentCount Vertical distance to suppression (ft) 8 tblEquipmentCount Horizontal distance to suppression (ft) 0 tblEquipmentCount Sprinkler credited as a heat detection TRUE tbllgnitionSources Heat release rate (kW) 317 tblTableE-1 Gamma dist Alpha 1.8 tblTableE-1 Gamma dist Beta 57.4 tbllgnitionSources Fire diameter (ft) 2 tbllgnitionSources Radiation fraction 0.3 tbllgnitionSources Shortest distance (ft) 6 tbllgnitionSources Fire condition Flame or Plume tbllgnitionSources Fire duration (min) 60

Document Control Desk RC-12-0142 Page 24 of 37 tbllanitionSources Taroet material Thermoset tblTargetMaterial Target damage temperature 330 tblTargetMaterial Target damage heat flux 11 tbllgnitionSources Supp curve Cable fires tbllgnitionSources Time to prompt suppression (min) 5 tbllgnitionSources Time to delayed detection (min) 15 tbllinitionSources Time to manual activation (min) 15 tbllinitionSources Human error probability 0.1 tblignitionSources Credit prompt detection TRUE tbllgnitionSources Credit prompt suppression TRUE tbllgnitionSources Credit auto detection TRUE tblignitionSources Credit auto suppression TRUE tblCompartmentAreaCalc Room length (ft) 49 tblCompartmentAreaCalc Room width (if) 46 tblCompartmentAreaCalc Room height (ft) 10 tblCompartmentlnfo Wall material Concrete tblWallMaterial Thermal conductivity (kW/m-K) 0.001 tblWallMaterial Specific heat (kJ/kg-K) 0.88 tblWallMaterial Density (kg/m3) 2000 tblCompartmentlnfo Wall thickness (ft) 1 tblCompartmentlnfo Opening height 8.9 tblCompartmentlnfo Opening width 1.4 tblCompartmentlnfo Height of opening above floor 8.9 tblCompartmentlnfo Ambient temperature (C) 20 tblRooms Brigade response time (min) 20

Document Control Desk RC-12-0142 Page 25 of 37 Critical Distances Negative *alies suggest er6rrs related to inap~ropriatC inputvaluesi Room Tewerawe [C]

R ~j 12jf (F] .Zonte ofrInfluemce for 'EOI 012 Flame Height (ft]) ______

Thermoset: Targets' :  : Thermoplastic Targets: Solid.State Targets.

Plume TemFfp (ftI] [f .lme Tm(t] Plu~e Tm It 0 P'uim 'lnHGL : .. [ Plu:le.in HGGC[ft) 18.2ý: -Plume.in HGL:( 00 Faeadaon []f ~ Flamfe Raato (I] adiation diatio itr

Ptlume Diameter iff)] 2.  !.PlumaeDiameter (i..It  ::j 2. .- :i;*: ?.: :;;: i ii;::::! :i:--[

SeverityFactor ..... Non Suppression Probability'. Activ'ation Timesi.

Critical HRR.[kW] 238 T1me T D [m .-

Severity Facor ... .0062 1: " A . . Time:autodetomi] 001 c P: u 00..... Time auto sujpP[mli]- 08.

Pn anual

- 10

.....i_.... __

cO'se Figure 1 Table 2: Summary of V&V results Case 1

__ Irtout's iHnd,, M hc

____~ ~~~AQ, -a FD nitCopvejr~ioiI1 Commef~nt _ _ 2Cek Room Temperature C 129 264 Ok Flame Height (ft) 1.73 5.7 Ok Plume Temp (ft) 2.13 7.0 Ok Plume in HGL (ft)

Flame Radiation (if) 0.83 2.7 Ok Plume Diameter (ft) 0.65 2.1 Ok Critical HRR (kW) 238 This should be the value given the fire condition. Ok Severity factor 0.063 This should be the value given the fire condition. Ok Screen No This should be the value given the fire condition. Ok This should be the value given the fire Time to target damage (min) 1 condition based on a plume convective Ok flux of approximately 28 kW/m2 Pns Auto 0.02 Ok Pns Manual 1 Ok Time to auto detection (Min) 0.35 0.01 Ok Time to auto suppression (Min) 51 0.85 Ok

Document Control Desk RC-12-0142 Page 26 of 37 Table 3: Input values for V&V case 2 tblEguipmentCount Fire location factor 1 tblEquipmentCount Automatic detection type Spot type smoke detection tblDetType Probability of detection failure 0.05 tblEguipmentCount Vertical distance to detection (ft) 8 tblEguipmentCount Horizontal distance to detection (ft) 0 tblEguipmentCount Automatic suppression type Automatic sprinklers tblSuppType Probability of suppression failure 0.02 tblEguipmentCount System initiation Automatic tblEguipmentCount Sprinkler type Standard response bulb/Ordinary tblSprinklerTypes RTI 235 tblSprinklerTypes Activation temperature (F) 165 tblEguipmentCount Vertical distance to suppression (ft) 8 tblEquipmentCount Horizontal distance to suppression (ft) 0 tblEquipmentCount Sprinkler credited as a heat detection TRUE tbllgnitionSources Heat release rate (kW) 317 tblTableE-1 Gamma dist Alpha 1.8 tblTableE-1 Gamma dist Beta 57.4 tbllgnitionSources Fire diameter (ft) 2 tbllgnitionSources Radiation fraction 0.3 tbllgnitionSources Shortest distance (ft) 6

-tbllgnitionSources Fire condition Flame or Plume tbllgnitionSources Fire duration (min) 60 tbllgnitionSources Target material Thermoset tblTargetMaterial Target damage temperature 330 tblTargetMaterial Target damage heat flux 11 tbllgnitionSources Supp curve Cable fires tbllgnitionSources Time to prompt suppression (min) 5 tbllgnitionSources Time to delayed detection (min) 15 tbllgnitionSources Time to manual activation (min) 15 tbllgnitionSources Human error probability 0.1 tbllgnitionSources Credit prompt detection TRUE tbllgnitionSources Credit prompt suppression TRUE tbllgnitionSources Credit auto detection TRUE tbllgnitionSources Credit auto suppression TRUE tblCompartmentAreaCalc Room length (ft) 49 tblCompartmentAreaCalc Room width (ft) 46 tblCompartmentAreaCalc Room height (ft) 10 tblCompartmentlnfo Wall material Concrete

Document Control Desk RC-12-0142 Page 27 of 37 V&V Studjrfor Zdhb"d'f Influence&Analysi EgujipmnittY frj

-Atable:Locationh 7~j4A ui&e. 44I-wiii EI0 tblWallMaterial Thermal conductivity (kW/m-K) 0.001 tblWallMaterial Specific heat (kJ/kg-K) 0.88 tblWallMaterial Density (kg/m3) 2000 tblCompartmentlnfo Wall thickness (ft) 1 tblCompartmentlnfo Opening height 8.9 tblCompartmentlnfo Opening width 1.4 tblCompartmentlnfo Height of opening above floor 8.9 tblCompartmentlnfo Ambient temperature (C) 20 tblRooms Brigade response time (Min) 20

• CriticalDistac .... .... . .. : . ae d:to:inappr6priateý input vale Negative values siuggest errofrts6 i Romrn 9M.f, r.*tu[e-[C) 326 = :69[I) Zon.e. of Influence. for: E00641:

r Therh*moet Taigets Thermoplastic Targets Solid State Targets.

Plmeerpf]12 Plume 16Tmp~t 5 lm~mit Plume in HGL:{ Ift Plee in HGL {ft]. 00 Plume in HL tt] ..

Flame Radiation {ftt) [II 6.6 Flame Radiationr*{ft 6.9 . F*ame Radiation (ft) [ 15.:

Plume Diametei;[ft [ 3.41' P:lme Diameter[tt-: Ptume dia'iet. s ae calcuated at the I _____ ~critical vertical dlistian'celisted for the :ZOI Severity Fator. ! ........ Non Suppression f1r6babity .  : Activaion Timdesi '.:

Critical tARP fkW)1`

521 .l ITmeTe T'o Dm[mn Dam 241 u Tm auto.. d tm S0097 PnAut 1'000 F n. Manual100 2.22 ~~ ~ ~ ~ ~ ~ ieauosp

.2 ~

< . rnr) i z:2220 Figure 2: FMDB solutions for input values in table above Table 4: Summary of V&V results case 2 Room Temperature C 326 619 Ok Flame Height (ft) 2.79 9.2 Ok Plume Temp (ft) 4.78 15.7 Ok Plume in HGL (ft)

Flame Radiation (ft) 2.70 8.9 Ok Plume Diameter (ft) 1.25 4.1 Ok Critical HRR (kW) 521 This should be the value given the fire condition. Ok

Document Control Desk RC-12-0142 Page 28 of 37 Severity factor 0.09 This should be the value given the fire condition. Ok Screen No This should be the value given the fire condition. Ok Time to target damage (min) 1 This should be the value given the fire condition on a plume convective flux of approximately 28 kW/m2 Ok Pns Auto 1 Ok Pns Manual 1 Ok Time to auto detection (Min) 1.33 0.02 Ok Time to auto suppression (Min) 20 Should be 20 since auto supp is not Ok credited. 20 is the time to manual act.

b. For all compartments in which fire modeling was used, a verification-and validation check was made to ensure that the CFAST model was used within the model validation range. The dimensionless parameters that describe the fire scenarios were calculated and compared against model validation parameters, as specified in NUREG 1824.

When parameters fell outside the validation range, a sensitivity case was run with CFAST to demonstrate that the conclusions were unchanged and to show that the use of CFAST in the current application is justified. The results will be documented in (See DC0780B-366).

c. The calculations did not account for the fire elevation. Recalculating the normalized parameters showed that the values were within the validation range for CB10, CB12, and CB18. For IB131, in the case of the limiting fire scenario (transient fire), the normalized parameter is inside the range when the transient fire is assumed to be at floor elevation. But when the transient fire is assumed to be 1 m above the floor, the normalized parameter is 1.3, which is above the upper limit of 1.0. If the transient fire were assumed to be 0.3 m above the floor, the normalized parameter would be within the range. A CFAST run using the value of fire elevation of 0.3 m showed a peak temperature of 216 0 C rather than 2260C when the fire elevation was 1 m. Both results are above the damage criteria of 2050C; therefore there is no change in conclusions.

The reports are being revised such that the correct equation for the normalized parameter is used. For IB131, the sensitivity calculation is being added to the document (DC0780F-1 73). In addition, Attachment J of the LAR will be corrected.

d. For all compartments in which fire modeling was used the normalized parameters that describe the fire scenarios were calculated and compared against model validation parameters, as specified in NUREG 1824. When parameters fell outside the validation range, a sensitivity case was run with CFAST to demonstrate that the conclusions were unchanged and to show that the use of CFAST in the current application is justified. The results are being presented in a new calculation (See DC0780B-366).

Document Control Desk RC-12-0142 Page 29 of 37

e. The source of the plume temperature correlation used in the spreadsheets supporting ABO1.21.02 is listed in the spreadsheet as Heskestad, G. Section 2-1, Fire Plumes, Flame Height, and Air Entrainment, SFPE Handbook of Fire Protection Engineering, 3rd Edition (P.J. DiNenno, Editor-in- Chief), National Fire Protection Association and The Society of Fire Protection Engineers, Quincy, MA, 2002. The Heskestad plume temperature correlation is validated in NUREG-1824, as a part of the FIVE Rev 1 software package.

FM RAI 04 NFPA 805, Section 2.7.3.3, "Limitations of Use," states: "Acceptable engineering methods and numerical models shall only be used for applications to the extent these methods have been subject to verifications and validation. These engineering methods shall only be applied within the scope, limitations, and assumptions prescribed for that method" Section 4.7.3, "Compliance with Quality Requirements in Section 2.7.3 of NFPA 805," of the Transition Report states that "Engineering methods and numerical models used in support of compliance with 10 CFR 50.48(c) were and are used with the same limitations and assumptions supported by the V&V for the methods as required by Section 2.7.3.3 of NFPA 805."

Regarding the limitations of use, identify uses, if any, of the fire modeling tools outside the limits of applicability of the method and for those cases explain how the use of the fire modeling approach was justified. An example of this limit of applicability issue can be referenced in RAI 01, item v.

SCE&G Response The use of fire modeling tools was justified by the method described in draft NUREG 1934, in which dimensionless parameters were compared against the V&V applicability ranges for the various models. When the dimensionless parameters were outside of the range, a sensitivity calculation was performed to demonstrate the appropriateness of using the model for the application. The V&V results are documented in the performance based fire modeling reports (DC0780F-096, DC0780F-097, DC0780F-1 03, and DC0780F-1 73), in the new calculation that is being prepared (See DC0780B-366).

FM RAI 05 NFPA 805, Section 2.7.3.4, "Qualification of Users," states: "Cognizant personnel who use and apply engineering analysis and numerical models (e.g., fire modeling techniques) shall be competent in that field and experienced in the application of these methods as they relate to nuclear power plants, nuclear power plant 'fire protection, and power plant operations."

Document Control Desk RC-12-0142 Page 30 of 37 Section 4.5.1.2, "FPRA Quality" of the Transition Report states that fire modeling was performed as part of the Fire PRA development (NFPA 805 Section 4.2.4.2). This requires that qualified fire modeling and PRA personnel work together. Furthermore, Section 4.7.3, "Compliance with Quality Requirements in Section 2.7.3 of NFPA 805," of the Transition Report states:

"Cognizant personnel who use and apply engineering analysis and numerical methods in support of compliance with 10 CFR 50.48(c) was competent and experienced as required by Section 2.7.3.4 of NFPA 805.

This requirement will continue to be met by adherence to SCE&G procedures and project management of contractor support staff. For personnel performing fire modeling or FPRA development and evaluation, VCSNS and contract personnel developed -and maintained project instructions to be used by individuals assigned various tasks, to ensure consistency of the engineering and PRA products. These instructions were developed by personnel with intimate knowledge and experience in the task subject matter. Task specific instructions were developed to identify and document required training and mentoring to ensure individuals are appropriately qualified per the requirements of NFPA 805 Section 2.7.3.4 to perform assigned work."

Regarding qualifications of users of engineering analyses and numerical models:

a. Describe what constitutes the appropriate qualifications for the VCS, Unit I staff and consulting engineers to use and apply the methods and fire modeling tools included in the engineering analyses and numerical models.

b. Describe the process/procedures for ensuring the adequacy of the appropriate qualifications of the engineers/personnel performing the fire analyses and modeling activities post-transition.

c. Explain the communication process between the fire modeling analysts and PRA personnel to exchange the necessary information and any measures taken to assure the fire modeling was and will be performed adequately.

SCE&G Response

a. Appropriate qualification for VCS staff and consultants to use and apply the methods and fire modeling tools are as follows:

" Required Reading including Fire Modeling Project Instructions, NUREG/CR-6850, User's Guides

• Classroom training in Fire Analysis

• Classroom Training in Fire Modeling (for Detailed Fire Modeling)

" Demonstration of comprehension for Fire Modeling

• Demonstration of proficiency in Fire Modeling

Document Control Desk RC-12-0142 Page 31 of 37

b. The contractor QA process assures that the people that performed the fire modeling analysis for the VC Summer project are qualified and trained, as follows:

" The fire modeling task leader was selected by the contractor project manager (PM) and Technical Advisor and the SCE&G PM based on full consideration of his expertise and experience.

• In accordance with the QA program, the task manager (TM) prepares a project instruction (PI) for the task. The PI is reviewed by the contractor PM, Technical Advisor, and the SCE&G-PM and modified as necessary to achieve consensus. It is then approved by the PM, the QA Manager, and the Division Manager.

• Staff are selected to work on the task by the TM, and approved by the contractor PM.

• The names of the selected staff are provided to the contractor QA administrator, who maintains all the QA records, including the training and qualifications records. All staff are required to read and certify their knowledge and understanding of the QA Manual, the QA procedures, and the project-specific PIs relevant to their assigned project tasks. All the signed forms are retained, and the matrix shows the date of certification.

• Whenever a PI is modified, all staff-that will be working on the task after the modification are to be re-certified to the PI.

VCS unit I staff performing Fire Modeling are enrolled in the Engineering Personnel Training Program in accordance with site procedure TQP-707. Under this procedure, a Training Module Worksheet (TMW) has been developed for "NFPA 805 Fire Modeling" to provide the qualification program for Fire Modeling. The TMW requirements have been summarized in

a. above. Personnel performing Fire Modeling are required to have completed this TMW prior to performing the task as stand alone originator or-verifier.

c. The fire modeling analysts populated the targets associated with each fire scenario. The database produces the input files for the quantification process. Therefore, the inputs to quantification are direct inputs from the fire modeling analysts identifying targets for each scenario. In addition, fire modeling analysts were part of the cut set review meeting.

FM RAI 06 NFPA 805, Section 2.7.3.5, "Uncertainty Analysis," states: "An uncertainty analysis shall be performed to provide reasonable assurance that the performance criteria have been met." NFPA 805, Section 1.6.9, defines "Completeness Uncertainty" as "uncertainty in the predictions of the model due to model scope limitations. This uncertainty reflects an unanalyzed contribution or reduction of risk due to limitations of the available analytical methods."

NFPA 805, Section 1.6.40, defines "Model Uncertainty" as uncertainty in the predictions of a model related to the equations in the model being correct, whether or not they are appropriate to the problem being solved, and whether or not they are sufficiently complete."

Document Control Desk RC-12-0142 Page 32 of 37 NFPA 805, Section 1.6.43, defines "Parameter Uncertainty" as uncertainty in the predictions of a model due to uncertainties in the numerical values of the model parameters." Section 4.7.3, "Compliance with Quality Requirements in Section 2.7.3 of NFPA 805," of the Transition Report states that "The impact of important uncertainties on the Fire PRA results was established using extensive, well formulated sensitivity studies to provide reasonable assurance that the performance criteria have been met as outlined in Section A of 2.7.3.5 of NFPA 805."

Regarding the uncertainty analysis for fire modeling:

a. Describe how-the uncertainty associated with the fire model input parameters (compartment geometry, radiative fraction, etc.) was accounted for in the analysis?

b. Describe how the "model" and "completeness" uncertainties were accounted for in the analysis. An example of completeness of an uncertainty issue is described in FM RAI 01, item b.

SCE&G Response

a. The input parameter that has the most impact on results is the heat release rate. The uncertainty is accounted for by using bounding heat release rates, as demonstrated in the individual zone report for CB06 (DC0780B-088, Section 7.3.2.6). Heat release rates are selected to be the screening values (98th percentiles) of the distributions recommended in NUREG/CR-6850. In addition, cable tray fires postulated as secondary combustibles do not assume fuel burnout. The scenario with the largest number of panels and cable trays was selected as the fuel package for the CFAST calculations.

b. The individual zone reports (see for example the report for CB06, DC0780B-088, Section 7.3.2.7) include a discussion of margin based on the model uncertainties (biases) listed in NUREG 1824. Completeness uncertainties were handled with sensitivity cases, such as whether HVAC is functioning or not in the control room abandonment -calculations. In addition, sensitivity calculations were performed for those fire scenarios with normalized parameters outside of the validation range for CFAST (See DC0780B-366).

A sensitivity analysis was conducted to look at ambient temperature impacts on the hot gas layer temperatures predicted by CFAST. The objective was to determine if higher ambient temperatures would lead to damaging hot gas layer temperatures. Three zones were identified that had the highest HGL temperatures calculated by CFAST, but also had temperatures that were below the damage criteria of 205 0 C. In those cases the CFAST runs were rerun using a higher ambient temperature ( 9 0th percentile), based on room temperatures reported in document S-021-018, EnvironmentalZone Information (Westinghouse). In all three cases the peak temperatures rose 10 degrees or less, but were still below the damage criteria, indicating that the higher ambient temperature, did not influence the results. The results are listed below:

Document Control Desk RC-12-0142 Page 33 of 37 ABO1.17 170 40 170 CB20 165 36 171 IB20 165 31 174 FM RAI 07 Section 4.5.2.1, "Fire Modeling Approach" of the Transition Report discusses the performance-based- approach via fire modeling. Reference is made to Attachment J, "Fire Modeling V&V," for a discussion of the acceptability of the fire models that were used.

Directly after this reference to Attachment J, there is a note, which states, "At VCSNS, the use of the fire modeling option (see NFPA 805, Section 4.2.4.1) to disposition potential variances from deterministic requirements follows a predefined process documented in the NFPA 805 project instructions or SCE&G design guides. The objective of these documents is to provide the framework for the use of fire modeling both during the NFPA 805 transition and in the future while the plant operates under NFPA 805 licensing basis. Consistent with the fire modeling requirements in NFPA 805, these documents allow for the use of fire models that are verified and validated within a range of applications. Consequently, the fire models available for use at VCSNS when operating under NFPA 805 are not limited to the ones selected for supporting the transition. Fire models that are verified and validated (e.g., FDS, CFAST, FDTs, etc.) and are exercised within the corresponding application range may be used in the future following the process outlined in the project instructions and in accordance with the requirements of NFPA 805."

The staff does not agree that a fire modeling methodology, which has not been previously reviewed by the NRC, can be used for a future plant change evaluation. The note in the Transition Report should be revised to only allow the future use of fire model methodologies that have been reviewed by the NRC for use in performance-based fire modeling applications in accordance with NFPA 805 Section 4.2.4.1.

SCE&G Response The note in Section 4.5.2.1 Fire Modeling Approach will be revised as follows:

"At VCSNS, the use of the fire modeling option (see NFPA 805, Section 4.2.4.1) to disposition potential variances from deterministic requirements follows the requirements in NFPA 805 and the recommended processes in NUREG-1824 and NUREG-1934. These documents provide the framework for the application of fire modeling both during the NFPA 805 transition and in the future when the plant operates under NFPA 805 licensing basis. Consistent with the fire modeling requirements in NFPA 805, only fire models that are verified and validated within a range of applications are used. Fire models that are verified and validated (e.g., FDS, CFAST, FDTs, etc.) and are exercised within the corresponding application range may be used in the

Document Control Desk RC-12-0142 Page 34 of 37 future. This approach meets the requirements of NFPA 805 and is consistent with the process followed in the performance based fire modeling calculations reviewed by the NRC staff during the transition process."

FM RAI 08 NFPA 805, Section 2.4.4.3, "Safety Margins," states: "The plant change evaluation shall ensure that sufficient safety margins are maintained. The deterministic approach for meeting the performance criteria shall be deemed to satisfy this safety margins requirement."

NFPA 805, Section 4.2.4.1.5, "Protection of Required Nuclear Safety Success Path(s)"

discusses the required safety margins in the context of the difference between the maximum expected fire scenario (MEFS) and the limiting fire scenario (LFS). This margin is to account for uncertainties and unknowns in the analytical process and to ensure adequate defense-in-depth is provided and nuclear safety performance criteria are met and-maintained.

Provide a discussion of the uncertainties and margin that exist for fire areas utilizing the performance-based approach and a technical justification for how the analysis meets the regulatory requirements.

SCE&G Response For the performance-based fire modeling, the uncertainty in the fire properties are location of the fire, quantity of combustible material, and heat release rate of the transient material stored.

The uncertainty for each fire property is handled as follows:

1. Location of the fire. The combustible storage location in each fire area is dealt administratively by marking the floor to indicate where transient storage is allowed.

2. Quantity of combustible material. The quantity of material is administratively controlled by limiting the quantity (mass) of material and by limiting the size of the storage area and height of the storage area.

3. Heat release rate of combustible material. For three of the zones (CB10, CB12, and CB1 8), no controls on the type of material is specified, since the calculations use a bounding heat release of polystyrene. For IB131, an additional administrative control is in place in which the type of material is limited to cardboard and wood, which have a lower heat release rate than plastic, such as polystyrene.

The safety margins associated important parameters used in the performance based fire modeling, such as the fire properties, model bias of CFAST, the presence (or not) of a fixed ignition source, smoke detection in the compartment, and the time allowed for the fire brigade to arrive, are described below and are also listed in the table below for additional clarity.

Document Control Desk RC-12-0142 Page 35 of 37

1. Peak heat release rate. The second column gives the margin between the values of peak heat release rate for the maximum and the limiting scenarios. The factors that would cause the fire to reach the limiting scenario heat release rate are described by the next three columns: types of transients, mass of material, and the area of base of the fire.

2. Types of transients. For CB1 0, CB1 2, and CB1 8, as mentioned above, the CFAST calculation assume the transient fire is composed of polystyrene, a plastic with one of the highest heat release rates. Therefore, instead of showing a margin, a bounding value of HRR is used. For IB131, the margin is that instead of only wood or cardboard, a small amount plastic (<3 kg) would be tolerated without reaching damaging conditions.

3. Mass of combustible material. For the room to reach a damaging temperature, the mass of combustible material would need to be greater than the quantity specified for the maximum expected-fire. The margin is listed in the table below. As-an example, for CB10, the mass would need to double from the allowed maximum of 9 kg (100% margin) to approximately 20 kg of material (-100% margin) since the temperature in the compartment does not reach 1920C until after approximately 20 kg of material has burned, approximately 15 minutes after the start of the fire.

4. Area-of the Base of the fire. In the CFAST calculations, the base of the fire is assumed to cover the entire staging area. The larger the area of the fire, the larger the fire intensity is expected to be. For example, for CB10, the area of the fire would need to increase by 80%

so that the HRR would be large enough to reach 900 kW. The results for the other zones are given in the table below.

5. Model bias of CFAST. Another type of margin is the difference between-the temperature simulated by CFAST and the actual temperature expected in the zone. NUREG-1934 (Table 4-1) shows that CFAST calculations tend to over predict the HGL temperatures by 6% (bias factor=1.06, standard deviation= 0.12). The implications are that, for example, in CB10, for the limiting fire scenario in which the temperature reaches 1920C, the probability of exceeding 2050C only 12%. The results for the other zones are given in the table below.

6. Smoke detection. A margin of safety exists since all fire areas -have smoke detection, which would increase the likelihood of early fire brigade arrival.

7. Time to damage. The longer the time to damage, the greater margin of safety exists.

8. Proximity to wall. In CB10, CB12, and CB18, the staging location is located within one foot of a wall or corner. A level of safety is demonstrated by sensitivity cases in which the fire is modeled by using the mirror method, in which a fire at the wall is assumed to have twice the heat release rate, twice the area, and twice the volume. For the corner impact, the procedure is the same, except using a factor of four rather than two. The sensitivity cases showed no damage in the maximum expected scenarios.

9. Fixed ignition source. For CB10, CB12, and CB18, an additional margin of safety is that no fixed ignition sources are present in the zone. In IB131, the fixed ignition source scenario is a motor fire with a maximum expected fire of 69 kW. A large margin exists for the motor fire because the only additional combustibles nearby are two cable trays. The radiant heat

Document Control Desk RC-12-0142 Page 36 of 37 flux shows that the nearest cable is not likely to catch fire from the electrical motor fire. In addition, the calculations show that even if the nearby cable trays are ignited, the resulting fire would be approximately 400 kW-- too small to cause temperatures high enough to cause damage to the VFDR cables.

900 kW (Bounding) (100%) 0.67 m'-

(80%) (80%)

CB12 758 kW to Polystyrene 9 kg to 36 kg 0.56 M2 to 5% Yes 20 No No 1500 kW (Bounding) (400%)_ I., M2 (100%) (100%)

CB18 500 kW to Polystyrene 9 kg to 16 kg 0.37 rW to 30% Yes 13 No No 2

900 kW (Bounding) (78%) 0.67 M (80%) (80%)

IB11 93 kW to Cardboard Cardboard Ca rd boa rd 30% Yes 6.3 N/A Yes 600 kW and wood and wood: 9 and wood:

only (margin: kg up to no 0.37 M2to (500%) plastics mixed limit 2.4 m' in)- (>500%)

polystyrene: Polystyrene:

3 kg is limit 0.37 M2to 2

0.44 M (19%)

FM RAI 09 During the audit and based on review of the detailed fire modeling reports supporting the performance-based fire modeling approach in fire areas C1310, C1312, CB18 and 113111, the staff learned that the use of a detailed administrative controls methodology in the performance-based fire modeling analyses is being credited.

However, it is not clear whether this approach has adequately considered all of the necessary aspects to ensure that a post-transition administrative controls program will be implemented such that the technical assumptions included in the fire modeling

Document Control Desk RC-12-0142 Page 37 of 37 calculations are not violated. For instance, it is not clear whether a height limitation will be included in an administrative procedure to ensure that combustible materials stored in transient staging areas do not exceed the values used in the analysis.

Provide a discussion on the bases used to develop the performance-based fire modeling scenarios and a description of and a commitment for how the post-transition procedures will maintain a program that is consistent with the fire modeling calculations.

In addition, provide a discussion of the DID concept that was applied in areas where performance-based fire modeling was -used such that administrative control procedures are not the only form of protection in a given fire area.

SCE&G Response The process of developing applicable procedures will be based in part in the results and insights generated by the fire modeling studies. The performance based fire modeling reports give details regarding what is required to be included in the administrative procedure. For example, the mass-limit- the type of material allowed (plastics or not), and the height limit are specified. A figure includeý in-the reports gives the location of the combustible storage area. In addition, the defense-in-depth is discussed in each report. Sensitivity calculations for doors open or closed, for normalized parameters outside of the validation range, and for wall and corner impacts were added to the reports.

Document Control Desk RC-12-0142 Page 1 of 16 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit 1 DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 2 Safe Shutdown (SSD) Request for Additional Information (RAI) Responses

Document Control Desk RC-12-0142 Page 2 of 16 SSD RAI 01 License Amendment Request (LAR) Section 4.2.1.1, "Compliance with NFPA 805 NSCA (Nuclear Safety Capability Assessment) (Section 2.4.2)" states: "The NSCA methodology review evaluated the existing NSCA methodology against the guidance provided in NEI 00-01, "Guidance for Post-Fire Safe Shutdown Analysis", Revision 1, Chapter 3, "Deterministic Methodology," as discussed in Appendix B-2 of NEI 04-02, "Guidance for Implementing a Risk Informed Performance Based Fire Protection Program under 10 CFR 50.48(c)." NEI 00-01, Revision 2 is the current version cited in Regulatory Guide 1.205, "Risk Informed Performance-Based Fire Protection for Existing Light-Water Nuclear Power Plants", Revision 1. The LAR references section lists Revision 1 and 2. Provide a gap analysis on the differences between the alignments using NEI 00-01, Rev. 1, as the basis for transitioning the NFPA 805 nuclear safety capability as indicated in NEI 04-02, versus using NEI 00-01, Revision 2, which is the current version cited in Regulatory Guide 1.205, Revision 1.

SCE&G Response:

A gap analysis for NSCA methodology has been performed. Technical Report TR08620-014, is being revised to incorporate the gap analysis. No adverse or non-conservative conditions were identified with respect to revised methods in Revision 2 of NEI 00-01, as applicable to NFPA 805. The circuit analysis effort for the VC Summer NFPA 805 Project began in April 2008, with a majority of the work completing late spring 2009. Revision 2 to NEI 00-01 was issued May 2009 and thus most circuit analysis work was completed prior to the release of Revision 2.

However, eminent changes slated for inclusion in NEI 00-01, Revision 2; were considered in the development of the circuit analysis project instructions. The gap analysis for circuit failure modes addressed in Appendix B of NEI 00-01 identified no shortcomings in the circuit analysis methodology. Additionally,. the special exclusion techniques contained in NEI 00-01, Revision 2 for "Important to safety" equipment were not used in the VC Summer circuit analysis.

SSD RAI 02 LAR Section 4.2.1.2, "Safe and Stable Conditions for the Plant" states that "For the plant to be in a safe and stable condition, it may not be necessary to perform a transition to cold shutdown as currently required under 10 CFR 50, Appendix R. Therefore, the unit may remain at or below the temperature defined by a hot standby/hot shutdown plant operating state for the event." If there are any fire areas where it is necessary to perform a transition to cold shutdown, provide a detailed discussion of why the transition to cold shutdown is needed, and what time limitations mare required for achieving cold shutdown. If yes, describe the fire areas/zones along with a list of cold shutdown systems and equipment for NFPA 805 compliance. Define the word "event" as it applies to the above statement.

Document Control Desk RC-12-0142 Page 3 of 16 SCE&G Response:

The NSCA currently does not require any fire areas to transition to cold shutdown in order to remain safe and stable. "Event" is referring to the fire event. However, in the event the Station decides to transition to cold shutdown, the NSCA has conservatively identified the necessary actions to initiate cool down to cold shutdown mode for each Fire Area.

SSD RAI 03 LAR Section 4.2.1.2, "Safe and Stable Conditions for the Plant," states that "If evacuation of the Main-Control Room (MCR) was required due to a significant fire in the Control Complex,_the Control Room Evacuation Panel (CREP) is designed to provide the instrumentation and controls to maintain Hot Standby, as a Primary Control Station (PCS)."

LAR Attachment G provides a list of recovery actions (RAs) required to be taken in the plant to maintain hot standby conditions. Describe any instrumentation-and controls such as local equipment operation or instrument indications, not part of the CREP, used to maintain Hot Standby that are not listed as RAs in Attachment G.

SCE&G Response:

TR08620-312, Attachment 6 identifies the credited NSCA PCS and Recovery Actions required for a control room evacuation.

Upon evacuation of the main control room, the Primary Control Station is shifted from the Main Control Panel to the Control Room Evacuation Panel. To transition the PCS to the CREP, actions are required to be performed at the Main Control Panel prior to leaving the Main Control Room, and at the CREP.

In the event-that a fire inside the Main Control Room prevents operators from opening all of the quick disconnect switches at the Main Control Panel-prior to leaving, the alternate switches in the cable spreading room would need to be opened. Because the quick disconnects are likely to be operated in the Main Control Room and if not, can be operated on the way to the CREP in the cable spreading room, these actions are not considered Recovery Actions. This is the only case where a potential action outside of the PCS is not considered a Recovery Action.

LAR, Attachment G will be revised to clarify the use of the quick disconnects and their classification as PCS actions.

Document Control Desk RC-12-0142 Page 4 of 16 SSD RAI 04 Deleted SCE&G Response N/A SSD RAI 05 LAR Section 4.2.1.2, "Safe and Stable Conditions for the Plant (Results)" states that "Some systems, such as the Chemical and Volume Control System (CVCS), serve multiple goals of coolant inventory addition and boric acid addition for long term reactivity control. Following the initial coping/assessment period at the start of a fire, the operators will maintain safe and stable conditions as follows:"

Define the term "initial coping/assessment period" as it is used in the above section.

During this period, describe if and how any operators will be performing any safe shutdown-operator RAs as a result of the fire. Describe what guidance documents and procedures operators use to determine when to exit the initial coping/assessment period.

SCE&G Response:

The "initial coping/assessment period" is the period following a receipt of a fire alarm or report of fire to the control room when the operations make initial assessments of the alarm/annunciation/report. During this period, Operations is determining which fire zone/area is affected based on information coming from the Fire Detection and Control System and local observations in the plant. Operations is also evaluating the consequences of any other annunciator alarms that may occur based on potential impacts of fire on cables in that area. It is not expected that-Operations will perform any safe shutdown operator RAs during this period.

After the exact fire zone is determined, the initial coping/assessment period-directs operations personnel to obtain the correct procedure to maintain nuclear safety. Once entry into a procedure is determined, then the actions of the procedure are followed.

The exact sequence through station operating procedures (e.g. AOP) has not yet been finalized, but will be completed during the NFPA 805 transition period, which would include associated Training of the Station Operations Organization on these changes.

Document Control Desk RC-12-0142 Page 5 of 16 SSD RAI 06 LAR Section 4.2.1.3, "Establishing Recovery Actions" states "The discussion below provides the methodology used to define and assess the Recovery Actions necessary to support the goals of the NFPA 805 Nuclear Safety Capability Assessment (NSCA) for VCSNS. This process was initially based on FAQ 07-0030 (MLl10070485) and consists of the following steps:"

Describe what this process is currently based on.

SCE&G Response:

The review of Recovery Actions followed the guidance provided in FAQ 07-0030. The term "initially" will be deleted from the LAR.

SSD RAI 07 LAR, Attachment G, "Recovery Actions Transition" states that "Based on the definition provided in RG 1.205, and the additional guidance provided in FAQ 07-0030, the following location is considered taking place at the primary control station:"

A list of the actions is in the Results of Step I Section of the LAR.

RG 1.205, Section 2.4-states: "The staff has identified two cases where operator actions taken outside the MCR may be considered as taking place at a primary control station.

These two cases involve dedicated shutdown or alternative shutdown controls, which have been reviewed and approved by the NRC. In either case, the location or locations become primary when command and control is shifted from the MCR to these other locations."

a. Describe whether all of the actions in both cases have been reviewed and approved by the NRC and include the references indicating approval, including excerpts of statements of explicit NRC approval.

b. Describe whether the location or locations of all of the actions become primary when command and control is shifted from the MCR to these other locations.

c. Describe whether the actions in both cases meet the criteria in RG 1.205, Section 2.4 a. and b.

SCE&G Response:

a) The NRC reviewed and approved the Alternate Shutdown Strategy for V.C. Summer per NUREG- 0717, Supplement SER-3, dated January 1982, pages 9-10 through 9-12.

Below are exerts from the SER-3 where the NRC stated:

Document Control Desk RC-12-0142 Page 6 of 16 "The applicant'sanalysis indicated that alternative shutdown measures were requiredfor the control complex (main control room, cable spreadingroom, relay room, intermediate cable chase and basement cable chase) in orderto assure the availabilityof the safe shutdown systems because the control complex contains more than one division of safe shutdown system cabling. In the event that fire disables the control complex, the control room evacuationpanel (CREP), which consists of two separatepanels located in a separate fire protected area in the intermediate buildingis provided as an alternativeto providing fire protection (refer-to Section 7.4.2 of the Safety Evaluation Report). The control functions and indicationsprovided at the CREP which are necessary for safe shutdown are electrically-isolatedfrom the control room."

............... The design objective of the CREP is to provide a central point to monitorplant shutdown and provide certain control functions ............ Emergency procedures delineatingthe shutdown sequence utilizing the CREP are provided........ plant operatorsverify that all supportequipment necessary to achieving safe shutdown has come on automatically, or turn on this equipment at separatepanels."

"The above CREP indications and control functions are either electricallyisolated from the control-room through transferswitches or are provided with power from cables routed separatelyfrom the control complex to assure their availabilityin a control complex fire."

"Basedin the above, we conclude that the CREP and alternate safe shutdown capability complies with the requirements of Section //I.L of Appendix R to 10 CFR Part50, and is therefore, acceptable."

b) The CREP becomes the Primary Control Station (PCS) when command and control is shifted from the Main Control Room.

c) Per the requirements of RG 1.205 and supplemental information provided in FAQ 07-0030 Revision 5 (approved by NRC letter dated February 4, 2011, ML110070485),

these actions meet the requirements of RG 1.205 Section 2.4.b.

LAR Attachment G, page G-3 has an incorrect reference cited. In the first paragraph on page G-3 it is stated that approval is referenced in NRC letter "Approval in 1981, 12-22, NRC Approval of III.G.3 and II.L with Add'l Requirements." Attachment G will be revised to remove the incorrect reference.

SSD RAI 08 LAR Section 4.2.1.2, "Safe and Stable Conditions for the Plant (Results)" states that "Reactor coolant system (RCS) pressure control is maintained by the ability to increase pressure by an emergency bus supplied pressurizer heater bank or by control of the charging rate and by the ability to reduce pressure by pressurizer power operated relief valve (PORV) operation."

Document Control Desk RC-12-0142 Page 7 of 16 LAR Attachment B, Section 3.1.2.2, "Pressure Control Systems" states that "Although utilization of the pressurizer heaters and/or auxiliary spray reduces operator burden, neither component is required to provide adequate pressure control. Pressure reductions are made by allowing the RCS to cool/shrink, thus reducing pressurizer level/pressure. Pressure increases are made by initiating charging/makeup to maintain pressurizer level/pressure. Manual control of the related pumps is acceptable."

a. -Based on the performance-based analyses, describe whether there are any fire scenarios where both pressurizer heaters and auxiliary spray would not be available to support reactor pressure control nuclear safety performance criteria (NSPC). If there are-any scenarios, provide a discussion of how pressure control NSPC will be achieved.

b. Describe whether there are any fire areas where long term pressure control would require the use of solid plant conditions. If so, describe how the operators have been trained in performing these actions under post-fire conditions. Describe the procedures that control post-fire solid plant conditions.

c. Provide a justification for the generic statement in Appendix B, Section 3.1.2.2 that states "Manual control of the related pumps is acceptable."

d. When pressurizer PORV operation is used for reducing RCS pressure, describe how it is analyzed as "required" for safe shutdown.

SCE&G Response:

Section 4.2.1.2 only describes the vitally backed pressurizer heaters as an option for pressure control. This is in agreement with Attachment B which identifies that heaters are preferred, but not the only way to maintain RCS Pressure Control. When possible, the NSCA credited pressurizer heaters to ease operator burden, but in some cases, they were not available and RCS Pressure Control was achieved by one of the other means described in Attachment B.

a. Auxiliary spray is not credited to support the reactor pressure control NSPC for achieving safe and stable. Where protection of a redundant train was not possible, performance-based Fire Risk Evaluations were performed to address loss of all pressurizer heaters located in Fire Areas AB01, 1B25, and RB01. If the heaters are unavailable for post-fire shutdown, RCS pressure would be maintained using charging makeup and emergency feedwater to increase/decrease RCS pressure respectively.

b. If the pressurizer heaters are lost due to fire, and are incapable of being returned to service, the pressurizer steam bubble could eventually collapse and the plant would operate at a solid plant condition. Currently, Fire Emergency Procedures (FEP) 2.1 and 3.1 describe maintaining long term RCS pressure control under this condition. Operator requalification for the FEPs is performed every two years. The contents of the FEPs may be changed and/or relocated to another appropriate set of procedures as a result of the NFPA 805 transition.

Document Control Desk RC-12-0142 Page 8 of 16

c. These are the words of NEI 00-01 Rev 1 related to makeup and charging as related to RCS Pressure Control. This statement is referring to manually controlling charging/makeup pumps from the main control board. A Control Room action such as this is -acceptableunder NFPA 805.

d. Utilizing the pressurizer PORVs for pressure reduction is only credited for transitioning to Cold Shutdown and is not required to maintain Safe and Stable. The PORVs are simply required to stay closed to maintain Safe and Stable.

SSD RAI 09 LAR Table S-1, "Plant Modifications Committed", indicates that a modification will provide instrument air auto start capability for the Diesel Driven Air Compressor (XAC0014).

Explain the process used to analyze possible fire damage to instrument air equipment such as air filters that could create an air leak and affect the instrument air system pressure and/or capacity.

SCE&G Response:

The proposed modification to provide instrument air auto start capability for the Diesel Driven Air Compressor is required for Fire PRA. The modification includes a review for the need of subcomponents, such as instrument air tubing, air filters, etc. Basically, the Diesel Air Compressor is a self contained skid mounted air compressor which sits in the Yard. All subcomponents (air filters, battery, etc.) with the exception of the auto start Pressure Switch (TBO1.01 412') are mounted on the skid. The Diesel Air Compressor, its pressure switch, and connecting cables have been modeled in the Fire PRA. The fire scenarios that require the Diesel Air Compressor to auto start will not damage the Diesel Air Compressor, its pressure switch, and connecting cables.

In support of the Fire PRA, a calculation was performed to determine what size Instrument Air pipe break each type of compressor (Normal, Standby, Supplemental & Diesel Driven) could supply and continue to provide sufficient air pressure. Drawings were marked up and walkdowns were performed to ensure that the large piping (> 1 YA") is not in an Instrument Air required zone. All subcomponents such as air filters, moisture separators which might contain components that fail in a fire are either in the same fire zone with the air compressors or are downstream of the >1 1/2" piping. The documentation is in the process of being issued as calc DC06510-007.

SSD RAI 10 Deleted SCE&G Response:

Document Control Desk RC-12-0142 Page 9 of 16 N/A SSD RAI 11 LAR Table 4-2 "NEI 04-02 Improvements, Post Transition Alignment (3.3.2 B Common Power Source Cables)" states "The NFPA 805 transition project has analyzed common power supplies required to be energized for the NSCA function to ensure compliance with NEI 00-01. Cases where breaker coordination has been determined to be insufficient, entries will be made into the Corrective- Action Program as a part of NFPA 805 implementation."

a. Provide the status for the cases where breaker coordination was determined to be

-insufficient for the current electrical system. Include in the discussion when compliance to the common power supply requirement will be achieved.

b. Describe whether entries were made into the Corrective Action Program (CAP) for these cases, including compensatory measures for the current license, if necessary.

SCE&G Response:

a. During the development of the various evaluations for the transition to NFPA 805 various apparent breaker/fuse coordination concerns were identified. The evaluations documented in Attachment L to Attachment A to TR07800-009 determined that the identified concerns had no impact on the NSCA evaluation.

Subsequent work has identified the following additional evaluations and resolutions:

Main Control Board-Sub-Panel 08 Technical Report TR08620-030 was developed to document a "Comparison of V. C.

Summer Station to Base Case of NEI 00-01 for Multiple High Impedance Faults." During the development of this report a common power supply concern was identified as the result of a pair of 20 A fuses in termination cabinet XPN07130 and several pairs of 15 A fuses on Main Control Board sub-panel 08 (supplying 125 DC to various devices). An internal preliminary evaluation was prepared that demonstrated that coordination exists between the two fuse sizes and types for the credible range of fault current. (Fault current is limited due to the resistance in the DC cables between the power source and locations subject to fire damage where power from sub-panel 08 is credited.)

Therefore the concern about fuse coordination for power from Main Control Board sub-panel 08 has been resolved.

APN1FA TR08620-030 and TR07800-009 both identify that coordination does not exist between the 480 VAC 70A breaker feeding 480/120 VAC single phase transformer XTF1 FA and

Document Control Desk RC-12-0142 Page 10 of 16 the 150A branch breakers in APN1FA which is 120 VAC single phase power panel supplying various devices.

The NSCA does not credit panel APN1FA and therefore this coordination issue has no impact on the NSCA evaluation.

The Fire PRA presently credits power from APN1 FA as a control power source for motor driven instrument air compressor 3A. However during the review, it was recognized that modification ECR 50538 changed the source of power for both the control scheme and the compressor motor.

Therefore the PRA will be revised to reflect the changes in power supplies for instrument air compressor 3A. Once corrected the Fire PRA will not credit APN1 FA and the coordination issue will have no impact-on the Fire PRA results.

b. Corrective Action Program CR-1 2-02306, Action 3, has been established to direct the addition of the evaluation of the fuse coordination for Main Control Board Sub-Panel 08 into the applicable calculation.

SSD RAI 12 LAR 4.2.1.2 "Safe and Stable Conditions for the Plant, Safe and Stable Summary Description" states "An important part of maintaining RCS inventory is maintaining Reactor Coolant Pump (RCP) seal integrity. RCP seal cooling is maintained by either the charging pump seal injection path or the Component Cooling (CC) flow to the RCP thermal barrier heat exchanger. Modifications are planned (see Table S-1 in Attachment S) to provide a redundant seal injection system that is independent of the existing system and not affected in the problem fire areas. Second, a new seal material is planned (see Table S-1 in Attachment S) so that the loss of seal cooling does not lead to significant loss of RCS inventory. Until new seal materials are installed, procedures for seal cooling interruptions are in place to address the issue as a part of the existing appendix R analysis."

During the NFPA 805 audit, discussions indicated that there is the possibility that the new RCP seal installation may be staggered such that one or more seals may be installed during the implementation window, and the remaining seals installed post-transition.

If all of the new seals are not scheduled to be installed during the transition period, describe whether the procedures for seal cooling interruptions are planned to be part of the approved NFPA 805 license.

SCE&G Response:

As per the LAR, attachment S, Table S-1, the new RCP seal materials will be installed under modification ECR 50799 NFPA 805 RCP Seal Replacement. The installation of these seals is

Document Control Desk RC-12-0142 Page 11 of 16 expected to extend beyond implementation of the NFPA 805 Program (180 Days following Safety Evaluation receipt).

Necessary procedures will be in place for seal cooling interruptions until the new seals are fully installed. These procedures would continue to be maintained under the approved NPFA 805 program until the planned seal modifications are complete.

SSD RAI 13 LAR Section 4.2.1.2, "Safe and Stable Conditions for the Plant, (Safe and Stable Summary Description)" under the bullet discussing Support Systems states "Systems typically not credited (but potentially-available) include instrument air, secondary side support, industrial cooling, and other plant systems not associated with a safety function."

a. The above discussion addresses systems not typically credited. Provide the areas where any of these systems are credited and the analysis that documents that they would be available.

b. Describe how the probabilistic risk assessment (PRA) modeled these systems.

SCE&G Response

a. The NSCA does not credit Industrial Cooling, secondary side support, industrial cooling or other systems not associated with a safety function included in the NSCA. However, Instrument Air was-included in the NSCA model and is credited to ensure certain Nuclear Safety Performance Goals are achieved in all areas.

The NSCA identifies fire areas where instrument air is lost due to a fire, and a DROID was generated for the loss of that function.

b. The FPRA also credits instrument air as a dependency for any components that require instrument air in order to perform a necessary function to support the success criteria of the front line systems modeled in the FPRA. The system is modeled similar to other systems in the model. Equipment locations are mapped and used as targets for the various fire scenarios as appropriate. Circuit analysis is performed for active components in the instrument air system, and the associated cables that can cause failure are mapped as targets.

SSD RAI 14 Describe the basis for Section 4.5.2, "Proper Polarity Three Phase Hot Shorts on alternating current (AC) Power Conductors" of the "NFPA 805 and Fire PRA Circuit Analysis, Task 4.4", TR07800-009.

Document Control Desk RC-12-0142 Page 12 of 16 Describe whether the 13 open items in Appendix C, Circuit Analysis Opens Table, of "NFPA 805 and Fire PRA Circuit Analysis, Task 4.4", TR07800-009, have been entered into the CAP and whether they have been completed.

SCE&G Response The basis for section 4.5.2 is NUREG/CR-6850 (Sept 2005) Vol 1 section 2.3.2.1. This is further discussed in NEI 00-01, Appendix C, section C4 which discusses the response to Generic Letter 86-10 question 5.3.1. As a conservative measure Case 2 of NUREG/CR-6850 section 2.3.2.1 was applied for Kerite cables. Other cable manufacturers are analyzed per their individual cable qualification.

The 13 open items in Appendix C, Circuit Analysis Opens Table have been addressed acceptably and are considered closed. The response to the open items was inadvertently left out when the other open items were addressed/closed in Attachment L to Attachment A.

TR07800-009 will be revised to correctly address/close these open items.. As recommended in the response to several of the open items, the Medium Voltage Breaker Coordination Calc DC08040-012 is being revised. These actions have been entered into the Corrective Action Program as CR-1 2-02306, Actions 1 and 2 and are expected to be completed first quarter of 2013.

SSD RAI 15 LAR Section 1.2, "Historical Perspective and Discussion", states that "Molded case circuit breakers should be periodically exercised and inspected and on a rotating refueling outage basis, should be sample tested to-confirm drift remains in acceptable limits."

Describe how the above testing and exercising of the molded case breakers is procedurally controlled for breakers that are credited for meeting NFPA 805.

SCE&G Response The BKR maintenance database (maintained by VCS Electrical Maintenance Supv) provides the records of BKR performance and maintenance. Safety Related MCC (Motor Control Center)

Circuit Breakers are cycled a minimum of five times each per EMP-280.006 at a minimum of every 5 years per EMP-280.004. All of the MCC Circuit Breakers at VC Summer are molded case circuit breakers. Additionally, at each scheduled PM (Preventive Maintenance) all MCC Breakers reaching 25 years of age prior to the following scheduled PM are being replaced.

SSD RAI 16 Section 2.0 "Scope" of "Safe Shutdown Separation Fire Protection Engineering Evaluations," TR 0780E-001, states "The scope of this TR is for the fire protection engineering evaluations (FPEEs) that are related to the NSCA and Safe Shutdown Separation requirements..."

Document Control Desk RC-12-0142 Page 13 of 16 Describe whether all open items, such as reworking existing fire protection systems, have been completed. If not, describe how they are being tracked and provide the latest schedule for completion. Describe whether items that are not in compliance with the current license have been entered into the CAP and confirm that compensatory measures have been implemented.

SCE&G Response Open items not yet completed are tracked in the NFPA 805 Change Management Plan, which will serve as a clearing house for implementation issues relative to the Transition. Many of the items are tracked in the station corrective action program, which is also referenced in the NFPA 805 Change Management Plan.

The schedule for completion of Modification related activities is found in Table S-1 of the NFPA 805 LAR. The projects are also included in the NFPA 805 Change Management Plan.

Corrective actions have been written for items found to be in non compliance with the current licensing basis, and compensatory-measures initiated, as applicable.

SSD RAI 17 The alternate shutdown RAs of Attachment 6 "(CREP Shutdown Transition (PCS) and Recovery Actions)", of "Nuclear Safety Capability Assessment Report Fire Shutdown Analysis" TR08620-312, differ from the alternate RAs in LAR Table G-1, "Recovery Actions and Activities Occurring at the Primary Control Stations."

Describe which document is correct, explain the apparent discrepancy, and describe any corrections made.

SCE&G Response Both documents are correct. The actions in the NSCA are credited for recovery of a safety function (keep one train free from fire damage). These may or may not also be included in the FPRA, since the FPRA addresses undesirable conditions using different criteria. The actions credited in the FPRA are those used to reduce the risk of core damage/LERF (as opposed to assuring that a given function, such as instrumentation, is available). This may result in the need for more actions, fewer actions, and/or different actions than cited in the NSCA. In the case of VC Summer, the alternate shutdown areas ultimately used NFPA 805 Section 4.2.4.2 for 805 compliance. Therefore, the recovery actions that needed to be credited in the FPRA to assure compliance under 4.2.4.2 were the ones included in Table G-1. This turned out to be more actions than were identified in the NSCA. Table G-1 conforms to the FPRA credited actions required for compliance.

LAR, Attachment G and TR08620-312, Attachment 6 will be revised to clarify FPRA and NSCA transition and recovery actions. Reference response to SSD RAI 03.

Document Control Desk RC-12-0142 Page 14 of 16 SSD RAI 18 NEI 04-02 Table B-3, "Fire Area Transition" Attachment 14, Fire Area Assessment Table (Table B-3); Attachment 14, Fire Area 1B08, and LAR, Attachment C, Fire Area 1B08, discuss water from fire suppression activities draining to adjacent fire zones in the 412-foot main floor area.

Describe whether the 412-foot main floor area has been analyzed for this additional water draining into the area. Describe if there are similar areas where water from fire suppression activities could affect other areas and whether they have been analyzed.

SCE&G Response The 412' elevation main floor area has been analyzed for additional water draining into the area from outside areas. This is found in Calculation DC07810-033 Evaluation of Fire System Flooding Effects Outside the Reactor Building" in the-intermediate Building section beginning on page 66. Other areas similar to this case have been analyzed as well.

SSD RAI 19 Section 4.4.4.3.2, "Chilled Water Mechanical Chillers" of TR 08620-015 "Nuclear Safety Equipment Report", Part 1, states that "A RA frecovery action] is available to start the B chiller or the C chiller on B train at the respective chiller control panel after taking the transfer switch to local."

This RA does not appear on LAR Table G-1. Discussions of RAs not in Table G-1 are found throughout TR 08620-015. Explain this discrepancy.

SCE&G Response The potential Recovery Actions discussed in TR08620-015 are not indicative of Recovery Actions credited for Nuclear Safety. This discussion is regarding the control capability of the equipment from the control room and locally. Recovery Actions are discussed as potential options but not methods of compliance. The method of compliance for each area, including any credited Recovery Actions, is included in the NSCA Report TR8620-312.

SSD RAI 20 Section 4.3.1.3.5.c, "Heating Ventilation and Air Conditioning (HVAC) Systems" of TR08620-015 "Nuclear Safety Equipment Report", Part 1, states that "Plant-specific evaluations are necessary to determine which HVAC systems are essential to safe shutdown equipment operations."

Document Control Desk RC-12-0142 Page 15 of 16

a. Describe whether all HVAC systems evaluations been completed.

b. Provide a list of all HVAC systems that have been determined to be "essential to safe shutdown equipment operations."

c. Describe the actions, modifications and procedures that have been identified and include a schedule for completion.

SCE&G Response

a. All safe shutdown HVAC systems evaluations have been completed. Calculations for this task are noted in the table below:

Calc Number Title DC00020-226 CONTROL ROOM HEAT UP - LOSS OF CHILLED WATER DC00020-227 CONTROL/TSC/CABLE SPREADING SEPARATE ROOM HEAT-UP CALCULATION COMPLETE LOSS OF HVAC DC00020-228 RELAY/CABLE SPREADING ROOM CB-425 HEAT-UP CALCULATION DC00020-229 CHARGING/Sl PUMP-ROOM AB-12 HEAT-UP CALCULATION DC00020-230 AUXILIARY BUILDING 412 SWITCHGEAR ROOM HEAT UP DC00020-231 AUXILIARY BUILDING 463 SWITCHGEAR ROOM HEAT UP DC00020-232 INTERMEDIATE BUILDING 436 AND 463 SWITCHGEAR ROOMS HEAT UP DC00020-233 BATTERY AND CHARGING ROOM HEAT-UP CALCULATION DC00020-234 SPEED SWITCH AND EVAC PANEL ROOM HEAT-UP CALCULATION DC00020-235 CREP A/B ROOM HEAT-UP CALCULATION DC00020-236 RHR/SPRAY PUMP ROOM HEAT-UP CALCULATION DC00020-237 SERVICE WATER BOOSTER PUMP ROOM HEAT-UP CALCULATION DC00020-238 EFW / TDEFW PUMP ROOM HEAT-UP CALCULATION DC00020-239 TEMPERATURE RISE IN THE SERVICE WATER PUMPHOUSE DC00020-240 TEMPERATURE RISE IN THE DIESEL GENERATOR (DG) ROOMS

b. Currently, per TR08620-015 Attachment 1, the following are credited as essential HVAC Systems for the NSCA:

Reactor Building Cooling Vital 7.2 KV Switchgear Room cooling Service Water Pump and Switchgear Area cooling Control Room Ventilation Vital Battery charger rooms cooling Relay Room cooling Vital Battery Room cooling Diesel Generator Room cooling Charging Pump Room cooling RHR Pump Room cooling Service Water Booster Pump area cooling Emergency Feedwater area cooling

Document Control Desk RC-12-0142 Page 16 of 16 Control Room Evacuation Panel and Speed Switch Room cooling Vital Auxiliary Building Switchgear Room cooling Because the equipment selection was performed prior to the heat-up calculations above, all of the HVAC systems were originally screened in as essential. Based on the results of the evaluations, some of these systems may be removed from the required equipment list, but the current list of credited systems is conservative.

c. There are no additional actions, modifications nor procedures required.

Document Control Desk RC-12-0142 Page 1 of 16 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit I DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 3 Fire Protection Engineering (FPE) Request for Additional Information (RAI) Responses

Document Control Desk RC-12-0142 Page 2 of 16 FPE RAI 01 Incipient Detection is described as a necessary modification in LAR Attachment S for electrical cabinets. FAQ 08-0046 (NUREG 6850, "EPRI/NRC-RES Fire PRA Methodology for Nuclear Power Facilities," Supplement 1) provides detailed discussion of the elements and limitations that are necessary to credit incipient detection systems for Fire PRA (FPRA). Provide more details regarding system design features, NFPA code(s) of record, installation, acceptance testing, setpoint control, alarm response procedures and training, and routine inspection, testing, and maintenance that will be implemented to credit the new incipient detection system. Provide details regarding availability and reliability. Describe whether this installation and the credit that will be taken will be in compliance with each of the method elements, limitations and criteria of NUREG/CR-6850 Supplement 1, Chapter 13, or will there be deviations. Provide justification for any deviations.

SCE&G Response To ensure proper credit is provided for the Fire PRA for installation of this modification, FAQ 08-00046 EPRI 101-6735, and NUREG-6850, Supplement 1 will be used as design inputs in system selection, design, and installation. This includes -design features, acceptance testing, and routine inspection testing and maintenance. Conceptual layout and design alternatives were available to the NRC Inspection Team review for protection of Relay Room (CB436) and Upper Cable Spreading Room Panels (CB448) as described in the Fire PRA. In accordance with FAQ 08-0046, NFPA 76- 2009, "Standard for the Fire Protection of Telecommunications Facilities" will be used as the code of record for the design of the system at V.C. Summer.

Deviations from these documents for the system installation are not expected. A code compliance document will be developed and maintained by V.C. Summer personnel.

Sensitivity settings will meet those defined in the Nov 23, 2009 FAQ 08-0046 Closure Memorandum.

Regular Functional Testing and Maintenance will be performed in accordance with NFPA 76, NFPA 72 and applicable vendor recommendations.

This system will be included within the overall NFPA 805 Condition Monitoring Program.

FPE RAI 02 LAR Attachment S "Plant Modifications Committed" Table S-1 for ECR1553, identifies a modification to "improve availability and reliability of station communication system(s) during fire scenarios."

Provide more detail with regard to what specific modifications are being done. Describe whether these modifications are assumed to be in place when determining feasibility for RAs.

Describe whether there are open items in the feasibility evaluation identifying the need to

Document Control Desk RC-12-0142 Page 3 of 16 complete these modifications.

Table S-1 indicates that communications is "implicitly considered in credit for operator actions." Describe how communications was evaluated during RA feasibility evaluation and in the FPRA (establishment of Human Error Probabilities (HEP) and Human Reliability Analysis (HRA)).

SCE&G Response LAR Attachment S, Table S-2 item 8 for 3.4.6 Communications states: "Table B-I"- Complete communications study and define strategies to ensure viable communications exists to support the fire brigade and other plant personnel during the course of a fire emergency. (NFPA 805 Section 3.4.6)" That study will identify any needed modifications and provide recommendations for improvement. Any modifications identified as a result of this study will be designed and installed in accordance with the commitment date for ECR 71553.

LAR Attachment G, states that there are no Recovery Actions for shutdown from the Control Room. For Alternate Shutdown, modifications to the Communications Systems were assumed to be completed when determining feasibility for Recovery Actions.

The HRA was performed based on the post-transition plant. As stated in Table S-1, for ECR 71553, SCE&G has committed to upgrading the communications system. The HRA assumed that communication would be available and standard performance shaping factors associated with ex-MCR actions were applied on that basis.

FPE RAI 03 LAR Attachment S "Plant Modifications Committed" ECR50810 is supposed to "provide mitigation strategies to address fire initiators or limit fire propagation."

a. Provide more detail for this modification(s).

b. Describe the modification, including location.

c. Describe the applicable design standards.

d. Describe the elements of defense-in-depth (DID) that are being satisfied.

e. This modification is identified as being in the Fire PRA. Describe specifically what is "in the Fire PRA".

SCE&G Response The scope of the ECR includes Fire Protection feature enhancements identified during the course of transitioning the stations Fire Protection Program to one based on NFPA 805. The scope of the modification package presently includes, as applicable:

Transient Materials: Signage/ Floor Demarcations (Admin Controls)

Document Control Desk Attachment 3 RC-12-0142 Page 4 of 16 AB 400 Sprinkler System Rework NFPA 13 AB 463 Sprinkler System Rework (Minor) NFPA 13 Junction Box Pillows (Embedded Conduit)

Flammable Liquid Storage Cabinets (Seismic Considerations)

Portable Fire Extinguisher Additions NFPA 10 Smoke Detectors (Based on final walkdowns of engineering documentation) NFPA 72E The elements of defense in depth being satisfied range from fire prevention to fire suppression and detection.

Concerning the Statement: "In the Fire PRA" relates to crediting fire protection features that meet specified standards, which would include NFPA Codes. In selected cases, enhancements to existing systems and features were identified during this review, and are being addressed accordingly.

FPE RAI 04 The LAR states- in Table B-1 Section 3.2.3(2), Compensatory Actions will be revised and updated incorporating NFPA 805 insights. Describe what compensatory actions will be revised and whether actions, impairment duration limits, or reporting requirements will be created or modified. If yes, provide a detailed list of those changes being made, and the justification for those changes in compensatory measures being considered.

Describe whether the Technical Requirements Manual (TRM) (or comparable plant document) is expected to transition or will it be superseded.

SCE&G Response The overall Fire Protection Program is defined in SAP-131, Fire Protection Program. This Station Administrative Procedure describes the responsibilities, program elements, and procedures required to ensure effective implementation of the program. SAP-131 references Fire Protection Procedures (FPPs) that implement the program. The FPPs document the conditions for operability, actions, impairment durations, and reporting requirements for systems currently credited by the fire protection program. The action section includes compensatory measures required when the systems cannot perform their intended function.

SAP-131 and the associated FPPs will be revised to address the new license condition for the plant including new or existing systems required as part transition to NFPA 805. Existing compensatory actions will be transitioned to these NFPA 805 credited fire protection systems. It is anticipated that a new Station Administrative Procedure or subordinate procedure will be generated to document actions, impairment duration limits, or reporting requirements. Details regarding the specific changes have not yet been developed. However, the basis or justification for compensatory measures will be consistent with existing actions. Necessary changes to plant procedures are included in the NFPA 805 Change Management Plan [tracked via CR-i1-03925 Action 001].

Document Control Desk RC-12-0142 Page 5 of 16 FPE RAI 05 The LAR states in Table B-1 Section 3.3.1.2 states that "controls of limited duration" for untreated wood will be put in place ... Attachment Ll identifies the need to use non-treated wood in limited quantities in an attempt to address unique situations.

Provide a description of the process that would control these deviations. Describe whether the future process will use the "Plant Change Evaluation" defined in NFPA 805 to evaluate the impact to the Fire.Protection Program (FPP). Provide a specific description of engineering and administrative procedure changes that the fire protection engineer (FPE) will use to control these elements of the requirement 3.3.1.2. Describe whether temporary non-compliances of limited duration will be considered program non-compliances requiring compensatory measures. Explain how instances of untreated wood will be controlled such that the assumptions of the FPRA will remain valid.

SCE&G Response The Attachment Ll Approval Request was submitted to address two (2) expected situations in response to a prescriptive NFPA 805 Chapter 3 requirement regarding transient combustibles.

Section 3.3.1.2 provides a "shall" requirement for treated fire retardant lumber with only one exception allowed for large cribbing-timbers. Plant maintenance and operations will- at times require untreated wood in the form of small hand tools and maintenance equipment and untreated wood for temporary material / equipment transport (e.g. pallets or equipment crates).

This approval request is to address planned activities when acceptable alternatives to treated wood do not exist.

The transient / combustible control program will address deviations from this Chapter 3 "shall" requirement by identifying the limits and controls for these conditions when necessary.

Bounding limits in regards to amount, duration, and compensatory measures for the use of untreated wood will be provided on a specific fire scenario basis in order to maintain the assumptions in the Fire PRA. Any changes to these bounding limits will be processed via the "Plant Change Evaluation" to maintain the integrity of the Fire PRA. The administrative controls will be developed based on this engineering input.

Details regarding the specific changes to the engineering and administrative procedures have not yet been developed. Controls will include limits on amounts, location, duration, available fire protection features, and compensatory measures to ensure that the temporary use of untreated wood does not adversely impact the Fire PRA.

Non-compliances with these administrative controls will be addressed as program non-compliances under the plant's corrective action program with condition evaluations and corrective actions taken to address and restore compliance. The "Plant Change Evaluation" is not expected to be used to evaluate these types of program non-com pl iances.

The established administrative controls will be bounded by the combustible quantities and heat release rates assumed in the Fire PRA.

Document Control Desk RC-12-0142 Page 6 of 16 FPE RAI 06 The LAR states in Table B-1 Section 3.3.9 that "A Fire Hazard Evaluation of the Transformer area considered drainage alternatives to that cited in this section. (Table S-2,,Item 2)."

It is not apparent how Table S-2, Item 2 addresses this requirement. Provide a detailed description of the Fire Hazard Evaluation of the Transformer area drainage alternatives.

Provide adequate justification for any alternative compliance approaches.

SCE&G Response VCSNS relies upon a function test in lieu of a periodic inspection to ensure that the oil collection basins are capable of performing their design function. During Transformer Deluge System Testing, flow is maintained for a 10 minute water spray discharge plus an added duration based on the oil volume of the subject transformer. The test confirms that deluge water is contained in the transformer catch basin collection system.

FPE RAI 07 LAR Table B-1 Section 3.3.7.2 identifies that the outdoor high pressure flammable gas storage is evaluated in TR0780E-006. There is no bulk gas storage in this evaluation.

Clarify this discrepancy.

SCE&G Response The correct document is TR0780E-004.

FPE RAI 08 LAR Attachment L2 requests approval for "existing wiring in suspended ceilings."

Describe what fire area(s) constitute "...suspended ceilings is limited in risk significant areas important to the NSCA, FPRA and NPO analysis." The request states that "...wiring is specified to be within metal conduits, cable trays, armored cable, or rated for plenum use." Wiring that is in cable trays without metal tops and bottoms does not meet the requirements.

The current justification in Approval Request L2 is not adequate. A request to comply with chapter 3 using performance-based methods [in accordance with 10 CFR 50.48(c)(2)(vii)] must satisfy criteria for risk, DID and safety margins (SM). The current request only addresses DID and SM. Provide either a qualitative or quantitative risk evaluation that demonstrates the ability to meet the acceptance criteria.

Explain whether this describes the "as-built" condition or the "as designed" condition.

Describe the types of cables, the types of service, and the fire protection features (e.g.

detection) that are currently installed above the suspended ceilings.

Document Control Desk Attachment 3 RC-12-0142 Page 7 of 16 SCE&G Response There are some areas of the VCSNS power block that have suspended ceilings (CR, TSC/Tagging desk area, RCA access point and other office style areas). The predominant cable use routed above these ceilings is for lighting. This configuration consists of conduit routed to a junction box with armored flexible cable from the junction box to the light. Most other circuits (fire/smoke detectors, ventilation, etc.) are routed in conduit with no allowance of open tray installation in these overhead areas. The only installations allowed are enclosed tray or square conduit. The exception to the above statements is communication cables (phone, computer and security). Communication cables are not routed with power/control cables. Communication cables routed above suspended ceilings are rarely routed in conduit, but due to their low energy are not considered a potential source of fire. These cables' predominant PVC jacketing combined with the lack of conduit results in an additional smoke concern for a fire, but is not a contributing factor to potential fire spread as these cables are routed through fire seals similar to other cables. Per plant design, the majority of areas with suspended ceilings are designed with fire/smoke detectors located above and below the suspended ceiling with corresponding sprinkler systems (i.e. RCA access area, TSC, etc.). Most cables located above suspended ceilings are directly compliant and there is a limited installation of noncompliant communication cables, whose additional hazards are mitigated by the presence of fire seals, fire/smoke detectors located either above or below the suspended ceilings and associated sprinkler systems.

FPE RAI 09 LAR Attachment L3 "Electrical Cable Construction" does not identify the as-built condition of currently installed "existing non-compliant" cables.

Describe the cable flame spread or other cable construction standards that were used for the currently installed cables. Describe whether these standards are addressed in FAQ 06-0022 and whether the current installation is different from the NFPA 805 requirement.

Describe the hazards that are being presented and whether those elements are alternatively meeting the intent of the requirement.

SCE&G Response Based on the known and approximated values of qualified cable installed in the VCSNS power block, it is known that the non-IEEE 383-1974 cable installed in the plant comprises less than 2.0% of the total linear feet of cable installed. The majority of the unqualified cable is communication cabling (phone, computer or security) and therefore is routed predominantly through/to areas that are non-impacting to NFPA 805 support circuits/equipment due to the lack of NFPA 805 support circuits/equipment located in these areas. Where communication cabling is located in a potentially impacting area, these areas are continuously manned or consist of office areas (i.e. CR TSC, Engineering or Security). Additionally, the majority of the non-IEEE 383-1974 cabling is routed in conduit. Non-qualified cable which is routed in conduits is considered to meet the intent of IEEE 383-1974 due to the reduced availability of oxygen in conduit which will limit the potential fire growth and flame spread. In addition, the lack of

Document Control Desk Attachment 3 RC-12-0142 Page 8 of 16 openings in the conduit restricts the exit of any smoke generated. The major impact of non-IEEE 383-1974 cables installed-in the plant is the additional smoke hazards due to the polyvinyl chloride (PVC) of the cables in a fire scenario. The non-qualified cables in conduit do not contribute to additional smoke or fire probability. The additional smoke hazard from the PVC is mitigated by the fact that the VCSNS Fire Brigades don self-contained breathing apparatuses (SCBA's) for all firefighting efforts. With a large portion of the non-qualified cables located in plant office areas (OPS, Engineering, or Security), the potential impact of these cables is further reduced as a fire event in these areas is likely to be discovered in the incipient phase. Also, some of these areas are located outside of the power block. Additionally, there currently are no situations that involve the use of non-qualified cables routed within train specific NSR cables in trays or conduits.

FPE RAI 10 LAR Attachment L4 "Bulk Gas Storage" presents an alternative method to meet a NFPA 805 requirement.

A request for approval of the use of a performance-based method to meet a requirement in NFPA 805 Chapter 3 must contain an evaluation of risk, DID and SMs. The current request only addresses DID and SM.

The justification should include details regarding the extent of the hazard, the quantity and capacity of the tanks, a description of the refilling process and a-description of the features that reduce the hazards associated with this alternative method. Since the tanks may form a missile hazard, include a discussion of the substantial buildings and structures between the tanks and any equipment required to meet the nuclear safety performance criteria.

SCE&G Response An FPEEE was performed to determine the adequacy of the existing hydrogen storage tank area (Ref: TR0780E-004, Attachment Admin-01). The hydrogen storage tank area-is north of the Turbine Building. The bulk hydrogen storage facility consists of six hydrogen storage cylinders located outside the protected area south of the turbine building. The storage bottles at 2400 psig have a total capacity of 44,800 scf. The tanks are filled at the storage location by a vendor using a Tube Trailer per procedure SOP-217. The tanks are surrounded by fencing and cement vehicle barrier posts. There are no missile shields surround the tanks.

Explosion of one of the bulk hydrogen storage tanks may result in a release of energy equivalent to 348 lb-TNT. A 66 ounce, 1965 ft/sec fragment missile could be generated by such an explosion. Since hydrogen tanks are not located within 100 feet of any Seismic Category I structure, the striking velocity is reduced to 1780 ft/sec. Such a missile could penetrate up to 8.0 inches into a concrete structure. All Seismic Category I structures at this site are 2 feet thick, double re-enforced; therefore, this missile may result in spalling and cracking of the concrete, but will not penetrate and damage components housed within.

Document Control Desk RC-12-0142 Page 9 of 16 The adjacent Fire Areas to the hydrogen bulk storage area are YD03, SWYD-02, TB01, TB02, TB03, and TB05. There is no safe shutdown equipment located in the adjacent fire areas that is required to remain operable in order to keep the plant safe and stable. Equipment associated with the availability of off-site power are located within these Fire Areas. In the event off-site power was lost, the plant would use the emergency diesel generators to provide power.

FPE RAI 11 LAR Attachment request L5 "Fire Brigade Notification" identifies the need to allow for delayed fire brigade response based on verification of a "direct visual contact with the fire".

Describe how-this delay will be factored into the assumptions for time-to-damage and non-suppression probability for the FPRA assumptions.

SCE&G Response The time to "direct visual contact with the fire" is not accounted for as an explicit parameter.

The analysis follows NUREG/CR-6850 Supplement 1, Fire Probabilistic Risk Assessment Methods Enhancement as stated in calculation DC0780B-001, Fire Modeling: Generic Methodology, Section 6.1.3.3 (4.). NUREG/CR-6850 Supplement 1, Section 14.2.1 states "When used in computing the probabilityof fire-induced damage, t refers to the -time available before damage to fire PRA targetsoccurs. Thus, in this application, t is replacedby the estimated time to damage minus the estimated time to detect the fire, i.e., <Tdamage> - <Tdet>.

This difference represents the estimated time available to suppress the fire. Methods to compute <Tdamage> and <Tdet> are described in NUREG/CR-6850. Note also that this definition of time available to suppress the fire differs from NUREG/CR-6850 in that it does not require an adjustment for Tfb (the fire brigade response time)."

Under this approach, the "direct visual contact with the fire" is an action after "fire detection" and actions to control and/or suppress the fire start at detection.

FPE RAI 12 LAR Attachment request L9 "Hose Station" identifies that pressure and flow rates are not reduced by pressure reducers and requests that the NRC approve this configuration.

Provide justification why not reducing water pressure at hose stations is acceptable.

Include a description of the system including the minimum and maximum calculated pressures and flows, and the impact to personnel and equipment of not having the pressure reducing valves installed. In addition also include a description of any training provided to plant personnel using the hose stations at this higher pressure.

SCE&G Response

Document Control Desk Attachment 3 RC-12-0142 Page 10 of 16 The hose standpipe system serves as a primary means for fire suppression in many areas of the plant, and serves as a backup suppression system in areas with automatic suppression systems. The 65 psig (Reference NFPA 14) hydraulically remote system outlet pressure and compliance with the minimum pipe sizing of the pipe schedule for standpipes and supply pipe sizing requirement insures that adequate hose nozzle pressure is available to a person to effectively attack a fire with a flow of 100 gpm through the fire hose.

The fire hose standpipes are fed from headers located inside the buildings, which are ultimately connected to the yard fire main loop. The water supply connections are arranged so that no single failure can impair both a primary fixed fire protection systems and backup hose station standpipe system, by the use of isolation valves in the water supply system. Hose reels are provided throughout the station as indicated on the fire protection layout drawings.

Individual standpipes fed by the water supply system are 4-inch diameter for multiple hose connections and 2-1/2-inch diameter for single hose connections. Hose stations are generally located outside of or near the entrances to normally unoccupied areas; are accessible from all plant areas; and are capable of reaching all plant areas.

DC07810-036, "Nozzle Pressure at Hose Reels", calculated the nozzle pressure at various hose reels throughout the plant. The locations in the plant that would require the most demanding pressure based on losses, which would represent the minimum pressure, were indentified. The results indicate that adequate pressure is available to produce the required 100 gpm. Each individual fire pump has a rated capacity to provide at least 2500 gpm at 125 psig. The maximum pressure that can be produced is 150 psig, which is the Fire Pump Relief Valve setpoint.

High pressure lines have higher reaction forces Which may impact unsuspecting fire brigade members. The hydrants used for training (initial and refresher) at the South Carolina Fire Academy are supplied by two fire water pumps rated to provide 1500 gpm each, and are computer programmed to supply a constant water pressure of 150 psi. Fire Brigade members are trained and qualified on the use of high pressure (> 125 psig) hose lines.

FPE RAI 13 LAR Attachment request LI I appears to ask for approval of all existing fire detector layouts throughout the plant in accordance with NFPA 72E. The fire alarm and detection system was upgraded to NFPA 72, however, fire detection device layout was not upgraded and remained in accordance with NFPA 72E.

Describe whether the intention of this request is for the NRC staff to approve the individual locations of the detectors throughout the plant as being compliant with NFPA 72E as opposed to NFPA 72. If this is the intention, provide a gap analysis of the differences between the detector location requirements under NFPA 72E and NFPA 72 from the respective code years. For any location requirements that the detectors will not meet under the updated code, provide a justification for not updating the detector locations in accordance with the newer code.

Document Control Desk Attachment 3 RC-12-0142 Page 11 of 16 SCE&G Response The intention of this request is not to request approval for the individual locations of detectors throughout the plant per NFPA 72E. The intent of this request was to clarify that the code of record at the time of the installation of the automatic fire detectors was NFPA 72E-1 978, Standard on Automatic Fire Detectors. NFPA 805 Section 3.8.2 indicates that required automatic fire detection be installed in accordance with NFPA 72, National Fire Alarm Code.

The first edition of NFPA 72 was not published until 1993.

The fire annunciation and proprietary alarm system portion of the system was upgraded in 1992 and installed in accordance with NFPA 72 - 1990, Standardfor the Installation, Maintenance, and Use of Protective Signaling Systems; the code of record at the time of the design I installation of this upgrade. At the time of this upgrade, NFPA 72E - 1978 was still the applicable code of record for the fire detector locations.

The first edition of NFPA 72, National-FireAlarm Code was a consolidation of several existing fire alarm codes including NFPA 72 - 1990, Standard for the Installation, Maintenance, and Use of Protective Signaling Systems and NFPA 72E - 1990, Standard on Automatic Fire Detectors.

NFPA 805 - 2001 Section 1.8, Code of Record indicates that the codes and standards referenced in NFPA 805 refer to the edition of the code or standard in effect at the time the fire protection systems or feature was designed or specifically committed to the authority having jurisdiction.

NFPA 72E-1 978, Standard on Automatic Fire Detectors and NFPA 72 - 1990, Standard for the Installation, Maintenance, and Use of Protective Signaling Systems are the applicable codes of record for the required fire alarm and detection system. Design Calculation DC0780D-005, NFPA Code Compliance - Fire Alarm and Detection Systems documents the code compliance review for NFPA 72- 1990 and NFPA 72E -1978.

FPE RAI 14 Table B-1 compliance statement CA - is defined as either 1. Clarification of the requirement or as a 2. Submitted request for approval. Approvals are listed in Attachment L and presented as (C)(2)(vii)'s.

NEI 04-02 Section 4.3.1 provides specific compliance statements for each Chapter 3 attribute. Two of the means to comply are:

Complies with Clarification - Items that are not in 'literal compliance' with the requirement as listed in NFPA 805 but should be transitioned as complies, and 10 CFR 50.48(c)(2)(vii) allows licensees to use performance-based methods to demonstrate compliance with NFPA 805 Chapter 3 requirements.

Document Control Desk RC-12-0142 Page 12 of 16 These two compliance strategies are combined into "Complies by Alternative (CA)." This is interpreted as some CA's in table B-1 are merely editorial clarifications and others require separate NRC approval as a 10 CFR 50.48 (c)(2)(vii) listed in Attachment L.

In an effort to ensure clarity, provide a listing of each Chapter 3 requirement that uses "CA" as the compliance strategy and explicitly delineate'which strategy is being chosen.

If the strategy is intended to be an editorial change then explicitly state the editorial clarification or change. If the strategy is a request for approval then state which approval in Attachment L is being credited for compliance in the specific NFPA 805 Chapter 3 requirement.

SCE&G Response The items from Table B-1 that have been identified as CA are listed in the below table and the 11 items (either CA or CE) for which NRC approval is sought are identified with reference to the Attachment L approval request..

B-1 Section CA or CE 10CFR50.48(c)(2) (vii)- Approval Request 3.2.3(1) CA No No- As committed in Table S-2 item 2, Fire Protection Program Preventative Maintenance and Surveillance procedures will be revised to improve alignment to scope and frequencies-associated with NFPA Code Requirements.

3.3.1.2(1) CA Yes Li 3.3.6 CA No No- Metal roof construction is class1 per FM Global property Loss Prevention Data Sheets 1-31 3.4.1 (d) CA Yes L5 3.4.2.4 CA Yes L6 3.4.3.(a) (4) CA Yes L7 3.6.2 CA Yes L9 3.3.5.1 CE Yes L2 3.3.5.3 CE Yes L3 FPE RAI 15

Document Control Desk Attachment 3 RC-12-0142 Page 13 of 16 Table B-1 identifies Chapter 3 elements where credit is being taken for "Previous NRC Approval (CNRC)". Compliance statements are identified in Table B-I, but the referenced Safety Evaluation Reports (SERs) are not identified. The results summary only indicates

"...were previously found to be acceptable and has been summarized in the defined evaluation." NEI 04-02 Section 4.3.1 and associated Figure 4-2 delineate the method to identify those requirements where previous approval has been used. The Section 4.3.1 and the simplified flowchart identify the need to provide verbatim excerpts from the submittal documents and the approval documents in the compliance basis field.

For those requirements in Table B-1 of the LAR where CNRC credit was used, provide appropriate SER references and excerpted submittal and approval statements of explicit NRC approval. Provide statements regarding the continued validity of the original basis of approval.

SCE&G Response Per Table B-1 of the LAR, CNRC credit is used for NFPA 80, Standard for Fire Doors and Fire Windows, and NFPA 90A, Standard for the Installation of Air-Conditioning and Ventilating Systems. Excerpts addressing prior NRC review and approval are provided below and detailed analysis is provided in TR0780E-006 Rev. 0, Fire Protection Features, Fire Protection Engineering Equivalency Evaluations. The previously approved NRC deviations remain valid and the noted deviations would not adversely affect the capability to achieve the NFPA 805 performance criteria.

NFPA 80, Standard for Fire Doors and Fire Windows:

NRC Branch Technical Position (BTP) APCSB 9.5-1 AppendixA (Reference 9.7) states:

"Floors,walls and ceilings enclosing separate fire areas should have minimum fire rating of 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. ... Door openings should be protected with equivalent rated doors, frames, and hardware that have been tested and approved by a nationally recognized laboratory."

Per FPER Rev. 2, dated 11/30/78, during a site review by the NRC, the NRC questioned the labeling of some fire doors indicating "...fire doors or door frame labels were not to be in place where credit for rated fire doorhas been utilized in analysis..."

VCSNS's response to this NRC review question (Reference FPER Rev. 2, 11/30/78) indicated that any door label discrepancies would be noted and a mutually agreeable solution would be provided. The submitted response included letters from the manufacturers of the bullet resistant and pressure resistant doors indicating that the door units, although not rated, were manufactured of similar materials and construction to rated fire doors. The bullet resistant and pressure doors are constructed of steel, do not have any openings or ports, and are self closing.

NUREG-0717, Supplement No. 3, NRC Safety Evaluation Report dated January 1982, indicated the NRC's review of VCSNS's response stating:

"We have reviewed the placement of fire doors and dampers to assure properfire ratinghas been provided. Fire doors carry an Underwriters'Laboratorieslabel except for certain pressure and bullet resistantdoors. The pressure and bullet resistantdoors are acceptable to us in the areas in which they are used."

Document Control Desk Attachment 3 RC-12-0142 Page 14 of 16 And concluded:

"Basedon our review, we conclude that the fire doors and dampers provided are, or will be, in accordance with the guidelines of Appendix A to Branch Technical Position ASB 9.5- 1 and are, therefore, acceptable."

NFPA 90A, Standard for the Installation of Air-Conditioninq and Ventilating Systems:

NRC Branch Technical Position (BTP) APCSB 9.5-1 Appendix A (Reference 9.8) states:

"Floors,walls and ceilings enclosing separatefire areas should have minimum fire ratingof 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. ... Penetrationsfor ventilation system should be protected by a standard "fire door damper"where recognized. Refer to NFPA 80, "FireDoors and Windows. ") The fire hazard in each area should be evaluated to determine barrierrequirements. If barrierfire resistance cannot be made adequate, fire detection and suppression should be provided, such as: (i) water curtain in case of fire, (ii) flame retardantcoatings, (iii) additionalfire barrier" Per FPER Rev. 2, dated 11/30/78, during a site review by the NRC, the NRC questioned the rating / labeling of some fire dampers indicating

" ... During our site visit we observed two 1-1/2 hour rated fire dampers installedin vertical penetrationsof the floor/ ceiling construction. Ducts or openings in fire rated barriersshould be provided with appropriatefire dampers. Modify your design accordingly" VCSNS's response to this NRC review question (Reference FPER Revision 2, 11/30/1978) indicated that "Approved accordion-type fire dampers (2- 1-1/2 UL fire resistantrating)are provided for floor

/ ceiling assemblies that are provided with blade lock and stainless steel negatorclosure springs.

NUREG-0717, Supplement No. 3, NRC Safety Evaluation Report dated January 1982, indicated the NRC's review of VCSNS's response stating:

"We have reviewed the placement of fire doors and dampers to assure properfire ratinghas been provided. ... The applicanthas provided three-hourventilation fire dampers for most three-hour wall ceiling / floor assemblies. Certain locations have two 1-1/2-hour fire dampers. These cases were analyzed and found acceptable where the fire loading was small and the estimated fire duration was well below the damper rating."

which concluded that:

"Basedon our review, we conclude that the fire doors and dampers provided are, or will be, in accordance with the guidelines of Appendix A to Branch Technical Position ASB 9.5- 1 and are, therefore, acceptable."

Note: the reference document listed in Table B-i, page A-42, for the above evaluations should be TR0780E-006 and not TR0787E-006. Table B-1 will be revised to correct.

FPE RAI 16

Document Control Desk Attachment 3 RC-12-0142 Page 15 of 16 In the definition of the Power Block, Attachment I, Table I-1, the Yard is identified as a power block structure included in the FPP. Confirm that this designation includes the following structures/features:

" Condensate Storage Tank

• Refueling Water Storage Tank

• Manholes

" Diesel Generator Fuel oil Tanks

" Auxiliary Boiler Oil Storage Tanks

• Transformers

-if any of these are not included provide justification for their exclusion.

SCE&G Response These components are in the Yard as defined in Table I-1.

FPE RAI 17 In plant partitioning for FPRA, confirm that no active fire barriers (e.g. water curtains),

thermal wraps, ERFBS, or fire resistant coatings are used to credit partitioning boundaries. If any of these fire protection features are credited in this manner, identify what type, fire area / zone location, and duration of rating credited.

SCE&G Response No active fire barriers (e.g. water curtains), thermal wraps, ERFBS, or fire resistant coatings are used to credit partitioning boundaries. The FIRE PRA PLANT BOUNDARY DEFINITION AND PARTITIONING indicates the following. "It should also be noted that no active fire barrier elements, thermal wraps, or fire resistant coatings are credited as partitioning elements in the Fire PRA and they were not credited in the plant regulatory fire protection program." Ref Doc TR07870-018, Section 4.2 Page 11.

FPE RAI 18 The LAR states in Table B-I, Section 3.11.3, that NFPA 101, "Life Safety Code", is exempted from the scope of the NRC review per 10 CFR 50.48 C.2 (i) regarding Life Safety.

The NFPA 805 requirement relates to the features of penetrations in fire barriers, specifically doors and dampers. Identify the strategy of compliance with Section 3.11.3 Fire Barrier Penetrations with regard to NFPA 101.

SCE&G Response

Document Control Desk RC-12-0142 Page 16 of 16 The code compliance document and Table B1 will be revised to incorporate applicable sections of NFPA101, related to maintaining the integrity of fire rated barriers via protection of personnel and equipment openings including fire protection devices such as doors and dampers.

Document Control Desk RC-12-0142 Page 1 of 6 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit 1 DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 4 Monitoring Program (MP) Request for Additional Information (RAI) Responses

Document Control Desk RC-12-0142 Page 2 of 6 MP RAI 01 National Fire Protection Association Standard 805, "Performance-Based Standard for Fire Protection for Light Water Reactor Electric Generating Plants", 2001 Edition, (NFPA 805), Section 2.6 "Monitoring" states that "a monitoring program shall be established to ensure that the availability and reliability of the fire protection systems and features are maintained and to assess the performance of the fire protection program in meeting the performance criteria" and that "Monitoring shall ensure that the assumptions in the engineering analysis remain valid."

Specifically, NFPA 805, Section 2.6 states that (2.6.1) "Acceptable levels of availability, reliability, and performance shall be established;" (2.6.2) "Methods to monitor availability, reliability, and performance shall be established; the methods shall consider the plant operating experience and industry operating experience;" and (2.6.3) "If the established levels of availability, reliability, or performance are not-met, appropriate corrective actions to return to the established levels shall be implemented. Monitoring shall be continued to ensure that the corrective actions are effective.

Section 4.6.2, "Overview of Post-Transition NFPA 805 Monitoring Program" of the Transition Report states that the NFPA 805 monitoring program will be implemented "after the safety evaluation issuance as part of the fire protection .program transition to NFPA 805" (Table S-2, Implementation Items, Item 4 of the Transition Report). The Transition Report also states that a monitoring program consistent with the NRC approved version of FAQ 10-0059 will be implemented.

The staff noted that the information provided in Section 4.6.2 of the Transition Report is insufficient for the staff to complete its review of the monitoring program, and as such, is requesting that the following additional information be provided:

a. A description of the process by which structures, systems, and components (SSCs) and programmatic elements will be identified for inclusion in the NFPA 805 monitoring program, including the approach to be applied to any fire protection SSCs that are already included within the scope of the Maintenance Rule program.

SCE&G Response SSCs for inclusion in the NFPA 805 Monitoring Program will be determined by following the guidance in FAQ 10-0059 for Phase 1 - Scoping.

Fire Protection systems, features, and programs will be monitored by the NFPA 805 Monitoring Program and include:

i. Required by the Nuclear Safety Capability Assessment ii. Modeled in the Fire PRA iii. Required by Chapter 3 of NFPA 805 iv. Fire Protection Programmatic Elements An Expert Panel will review NFPA 805 Analysis Documents, Reports, Calculations, and Fire PRA for SCOPING equipment meeting the requirements identified in FAQ 10-0059.

Document Control Desk RC-12-0142 Page 3 of 6 a) Results from Scoping, Screening, and Risk Determinations by the Expert Panel will be documented in procedure PSEG-91 1. Expert Panel meeting minutes will be retained through the CR process.

b) The expert panel will be composed of members from the Fire Protection Team and the Maintenance Rule Coordinator.

" Fire Protection Principle Engineer (Chair)

• Fire Protection Coordinator

• Fire Protection Plant Support Systems Engineer

• Fire Protection Design Engineer

" Fire PRA Engineer

" Maintenance Rule Coordinator c) Expert panel charter includes the following tasks (initially and any future changes):

• Develop / Approve Scoping

" Develop / Approve Screening

• Develop / Approve Risk Determination

" Develop / Approve Performance Levels

• Develop / Approve Corrective Action Plans

" Review Monitoring and Assessment Results

" Review PRA Changes for Monitoring Impact NSCA equipment for systems other then Fire Service, if not already monitored, will be added to the respective systems Maintenance Rule program and include:

i. Nuclear safety equipment ii. Fire PRA equipment iii. NPO equipment iv. SSCs relied upon to meet radioactive release criteria

Document Control Desk RC-12-0142 Page 4 of 6

b. A description of the process that will be used to assign availability, reliability, and performance goals to SSCs and programmatic elements within the scope of the monitoring program including the approach to be applied to any SSCs and programmatic elements for which availability, reliability, and performance goals are not readily quantified.

SCE&G Response Availability, reliability and performance in the NFPA 805 Monitoring Program will be determined by following the guidance in FAQ 10-0059 for Phase 3 - Risk Target Value Determination.

The Expert panel will determine levels of availability, reliability and performance for the monitoring program taking into consideration:

• risk significance of In-Scope items,

• values specified in the Fire PRA model,

• recommended values in EPRI Technical Report 106756,

• Industry Operating Experience 0 plant specific historical data

• monitoring cycle The initial values will be monitored for appropriateness and may be adjusted by the expert panel.

c. A demonstration of how the monitoring program will address response to programmatic elements that fail to meet performance goals (example discrepancies identified in programmatic areas such as combustible controls programs).

SCE&G Response The corrective action process, in accordance with SAP-999, will be used to improve and address:

i. Fire Protection Programmatic Elements that do not meet Performance Criteria.

Document Control Desk RC-12-0142 Page 5 of 6

d. A description of how the monitoring program will address fundamental fire protection program elements.

SCE&G Response Condition reports (CRs) and Permits will be monitored and compared to provide a metric for Hot Work, Transient Combustibles, and Storage Area program performance.

Fire Brigade performance will be monitored by Drill performance scores and response times.

e. A description of how the guidance in EPRI Technical Report (TR)1006756, "Fire Protection Equipment Surveillance Optimization and Maintenance Guide" if-used, will be integrated into the monitoring program.

Note that such changes permitted by NFPA Section 3.2.3 require NRC approval with an appropriatejustification in accordancewith 10 CFR 50.48 (c)(2)(vii).

SCE&G Response Expert Panel will utilize the EPRI Technical Report 106756 to assist in determining levels for availability, reliability, and performance for Fire Service related components in the NFPA 805 monitoring program.

The document will also be used to optimize surveillance and preventative maintenance of Fire Service components along with a review of historical performance of the components.

f. A description of how periodic assessments of the monitoring program will be performed taking into account, where practical, industry wide operating experience including whether this process will include both internal and external assessments and the frequency at which these assessments will be performed.

SCE&G Response A periodic assessment shall be performed as described in section 4.6.2 of the NFPA 805 LAR, taking into account, where practical, industry wide operating experience. This may be conducted as part of other established assessment activities. Issues that will be addressed include:

Review Systems with Performance Criteria and determine if the performance criteria effectively monitors the functions of the system.

ii. Review the performance during the assessment period and identify trends in system performance that should be addressed.

Document Control Desk RC-12-0142 Page 6 of 6

g. A confirmation that periodic NFPA 805 assessments (audits) of the fire protection program will be conducted under the existing Fire Protection Quality Assurance Program. If not, describe the process that will be used to conduct these assessments.

SCE&G Response SCE&G confirms that it performs audits in accordance-with the requirements in the Quality Assurance Program Description (QAPD) and Regulatory Guide 1.189 as addressed in the QAPD and the UFSAR, Appendix 3A.

Document Control Desk RC-12-0142 Page 1 of 87 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit 1 DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 5 Probabilistic Risk Assessment (PRA) Request for Additional Information (RAI)-

Responses

Document Control Desk Attachment 5 RC-12-0142 Page 2 of 87 PRA RAI 01 Describe how the evaluation includes the possible increase in HRR caused by the spread of a fire from the ignition source to other combustibles. For example, on page 60 of 73 in the Generic Fire Methodology report, a qualitative argument is made that the second cable tray will not get ignited for a fire in the initial cable tray from a hot work induced cable tray fire. Perform a quantitative calculation to verify this conclusion. The same type of conclusion is drawn for a junction box fire. Perform thesame type of quantitative verification. If you are unable to support these qualitative arguments, provide a sensitivity study as to the importance of these conclusions to your PRA results.

SCE&G Response Consistent with the guidance in NUREG/CR-6850, the VC Summer Fire PRA includes the following ignition- sources:

" General transient fires

" Transient fires due to welding and cutting (i.e. hotwork activities)

" Cable fires due to welding and cutting (i.e. hotwork activities)

" Self ignited cable fires (not postulated in VC Summer due to the presence of qualified cable)

" Junction box fires The frequency of these ignition sources is not apportioned by counting ignition sources like in the case of fixed sources such as electrical cabinets, pumps, etc. Instead, theifrequency of individual fire scenarios postulated in a given fire zone is apportioned using the following factors when applicable: 1) the floor area ratio (also referred as the geometrical weighting factor), 2) the maintenance, storage, and occupancy influence factors, and 3) cable load weighting factors. As a practical and convenient approach, the contribution from these ignition sources is rigorously assigned to the same transient fire locations (i.e., transient zone) so that no location in the plant is left without a contribution from these ignition sources. Although in general terms, the frequency of these scenarios is apportioned in the same way (i.e., using influence factors or cable load factors), and they are postulated in similar locations (as apportioned by floor area ratios), the behavior of these fires is different and they are modeled as such in the Fire PRA.

That is, the timing of the scenario development process and the extent of target damage is not modeled the same way for each of these ignition sources.

General transient fires and transient fires due to hotwork are assumed to be fires occurring near floor level affecting nearby plant equipment, cable tray and conduits. These fires have informally been characterized as "trash can fires" to represent combustibles that are not permanently located in the fire zones and can be moved within the open floor areas in the zone.

Generally, these fires are postulated so that they propagate through the cable trays in the transient zone and damage conduits that are also located in the transient zone. This approach

Document Control Desk Attachment 5 RC-12-0142 Page 3 of 87 ensures that no pinch points in the fire zones are missed. In summary, these are relatively large fires that are assumed to propagate.

In-contrast, junction boxes and cable fires due to hotwork are postulated in the Fire PRA as relatively smaller fires when compared with the general transient and transient due to hotwork ignition sources.

• Junction box fires, as evidenced in incident numbers 1369, 745, and 665 in the EPRI's Fire Events Database, did not propagate out of the junction box. Based on the descriptions, fire appeared to remain within. the junction box or conduits associated with them.

• Cable fires due to hotwork, as evidenced in incident numbers 2126, 2113, 1392, 640, 391, 385, 117, and 110 in the EPRI's Fire Events Database, did not propagate out of a single cable tray. In addition, most, if not all, of these fires (some of them have vague descriptions) were limited to damage to the insulation of very few number of cables and not the entire cable tray.

Consequently, these ignition sources are modeled differently in the Fire-PRA when compared to general transient and transient fires due to hotwork. Junction box and cable fires due to hotwork are modeled as smaller fires affecting only one cable tray or conduit based on the evidence available in EPRI's Fire Events Database. This approach has the advantage of distinguishing, when necessary, the risk contribution of ignition sources or fire scenarios that behave differently. The approach has the added advantage of consistently including the contribution from all the ignition sources in similar locations throughout the fire zone.

Because the treatment of junction boxes and cable fires due to hotwork assumes only one cable tray or conduit is damaged, the selection of this target is based on a bounding approach. The approach consists of assigning to each scenario the cable tray or conduit with the highest Conditional Core Damage Probability (CCDP) in the transient location (i.e., transient zone) where the fires are postulated. Therefore, the junction box and cable fires due to hotwork are not only failing one cable tray or conduit, but they are failing the highest CCDP tray or conduit in the location. This approach solves the question of "which cable tray or conduit" to assign as targets to these ignition sources.

It should be noted that the junction box fires and the cable fires due to hotwork are postulated in the same location as the transient fires and the transient fires due to hotwork. Therefore, the transient location is currently analyzed in the Fire PRA for propagating fires through more than one tray. Although the junction box fires and the cable fires due to hotwork assume only one tray damaged, the location in the plant where they are postulated is analyzed for fires that propagate. Consequently, the assumption that junction boxes and cable fires due to welding are limited to one cable tray does not limit in any way the ability of the Fire PRA in identifying pinch points or locations in the plant where a propagating fires in cable trays may provide risk insights. It simply differentiates the risk contribution of these sources given the general

Document Control Desk Attachment 5 RC-12-0142 Page 4 of 87 understanding of the behavior of these fires as described by the event descriptions in EPRI's fire events database.

Given how the fire scenarios for these ignition sources are structured in the Fire PRA, the impact of the assumption that junction box and cable fires due to hotwork are limited in damage to one cable tray could be assessed by comparing the CCDP of these ignition sources with the CCDP that is calculated for general transients and transient fires due to hotwork in the corresponding locations. Although higher CCDPs are expected with this comparison, the "cable fires due to hotwork" will not present much of a change in the contribution to the plant CDF. The reason for not observing a contributing change in risk for cable fires due to welding is due to the credit for manual suppression using the hotwork suppression curve. For cable fires due to hotwork, there is no damage postulated for a period of time reflecting the incipient state of a hotwork fire associated with a spark or welding slag igniting and propagating through a cable-tray. Failure to suppress the fire on this stage, results in damage to a full cable tray. The duration of the incipient stage is uncertain and is only credited for this type of ignition source (i.e. cable fires due to welding) in the Fire PRA because the cables are not exposed to a relatively large vigorous fire like the ones assumed in the Fire PRA for general transients, electrical cabinets, etc. That is, cable fires due to hotwork start with a small spark or welding slag falling on cables.

As a comparison, consider the model for fire propagation in cable trays in Appendix R of NUREG/CR-6850. The propagation from the first cable to the second is assigned to be 4 minutes, which assumes that the first tray, which is exposed to the ignition source, is in flames and engulfing the second one. In the case of cable fires due to welding, the cable fires-are initiated by small sparks or welding slags that take some time to propagate to cables. It is then reasonable to assume that a spark or welding slag will require more than four minutes to damage a cable tray. Furthermore, in all cases reviewed in the ERPI Fire Events Database the time is long enough to allow successful manual suppression activities before a full tray is damaged. As a practical approach, the incipient state for this ignition source is assumed to be 20 minutes. At this point in time, a full cable tray, the one with the highest CCDP in the transient zone is assumed damaged. This assumption is roughly based on the argument that it would take a relatively long time to ignite one full cable tray when the fire starts with a spark or an ignited welding slag compared with the accepted model of four minutes for ignition of a tray exposed to a vigorous fire.

Currently in the Fire PRA, there is no suppression credit for junction box fires. The cable tray or conduit associated with the highest CCDP in the transient zone is assigned as the target and is assumed failed at time zero-hence resulting in non suppression probabilities of 1.0. This is considered to be a conservative assumption because 1) junction box fires start small, 2) they occur inside a metal box or conduit, and 3) propagation outside the box or conduit needs to happen before it can grow to an intensity that can propagate to full cable trays. In most cases,

Document Control Desk RC-12-0142 Page 5 of 87 this initial propagation process will occur at low oxygen levels inside the junction box or conduits further slowing the process.

In summary, the assumption that the junction box fires and cable fires due to hotwork will generate damage to one full cable tray is not based on quantitative fire modeling. A relatively large fire postulated in a cable tray is likely to propagate to nearby trays and analytical fire models (or even the empirical model for fire propagation in cable trays in Appendix R of NUREG/CR-6850) suggest such behavior (again, the behavior of propagating fires between cable trays is considered in the Fire PRA for transient fires and transient fires due to hotwork).

Instead, the assumption is based on the behavior of the fires labeled as challenging in the fire modeling database and is implemented in the Fire PRA with the following bounding considerations: 1) a full cable tray or conduit is failed (as opposed to damage to a small number of cables in a tray as suggested by the description of cable fire due to hotwork fire events), 2) the cable tray or conduit with the highest CCDP in the location is selected as the target, and 3) these fires are located in the same place as the general transient and transient due to hotwork fires so that the propagation through cable trays suggesting the CCDP of the location is included in the Fire PRA. Therefore, every transient location in the plant includes-the contribution of general transients, transient fires due to hotwork, junction box fires, and cable fires due to hotwork. Based on the evidence available in EPRI's Fire Modeling Database, it is considered that modeling ALL of these ignition sources with similar fire propagation behavior would overestimate the CDF in such locations. Alternatively, the modeling approach implemented in the Fire PRA is intended to characterize the different types of ignition sources based on their specific characteristics.

PRA RAI 02 Transient fires should at a minimum be placed in locations within the plant physical access units (PAUs) where conditional core damage probabilities (CCDPs) are highest for that PAU, i.e., at "pinch points." Pinch points include locations of redundant trains or the vicinity of other potentially risk-relevant equipment, including the cabling associated with each. Transient fires should be placed at all appropriate locations in a PAU where they can threaten pinch points. Hot work should be assumed to occur in locations where hot work is a possibility, even if improbable (but not impossible), keeping in mind the same philosophy. Describe how transient and hot work fires are distributed within the PAUs. In particular, identify the criteria that determines where an ignition source is placed within the PAUs. Also, if there are areas within a PAU where no transient or hot work fires are located since those areas are considered inaccessible, define the criteria used to define "inaccessible." Note that an inaccessible area is not the same as a location where fire is simply unlikely, even if highly improbable.

Document Control Desk Attachment 5 RC-12-0142 Page 6 of 87 SCE&G Response For fire zones with no detailed analysis, the full transient and hotwork frequency is included in the ignition frequency assigned to the zone.

For fire zones receiving detailed fire modeling analysis, the transient and hotwork frequencies have been assigned to the transient zones and apportioned using the floor area ratio concept.

Since the transient zones are assigned to all the open floor areas in the fire zone, transient and hotwork frequencies are assigned "everywhere". The location of the transient zones is available in the individual zone reports. (See for example for CB06: DC0780B-088, Figure 2.) That is, there is no selection of where the transient and hotwork fires are possible or not possible, as they are assumed in all the open floor areas of the fire zones.

PRA RAI 03 Discuss the calculation of the frequencies of transient and hot work fires. Characterize the use of the influence factors for maintenance, occupancy, and storage, noting if the rating "3" is the most common, as it is intended to be representative of the "typical" weight for each influence factor. It is expected that the influence factor for each location bin associated with transient or hot work fires will utilize a range of influence factors about the rating "3," including the maximum 10 (or 50 for maintenance) and, if appropriate, even the rating "0." Note that no PAU may have a combined weight of zero unless it is physically inaccessible, administrative controls notwithstanding. In assigning influence factor ratings, those factors for the Control/Auxiliary/ Reactor Building are distinct from the Turbine Building; thus, the influence factor ratings for each location bin are to be viewed according to the bin itself.

SCE&G Response All PAU's have a transient frequency as required by the Fire PRA standard and is recommended in NUREG/CR-6850. Each fire zone has been mapped to one of the transient classifications (e.g. turbine building, plant wide components, etc). The ratings for the influence factors are assigned following guidance in NUREG/CR-6850 and are listed in the ignition frequency report (TR07800-007). The transient frequencies are calculated considering the generic values for each of the mappings. For fire zones receiving no detailed analysis, the full transient frequency assigned to the fire zone based on the influence factors is included in the total frequency assigned to the zone. For fire zones receiving detailed analysis, the full frequency assigned to the transient zone has been apportioned to the different transient postulated transient fires using the floor area ratio. The frequency includes general transients, transient due to hot work and cable fires due to hotwork.

PRA RAI 04

Document Control Desk RC-12-0142 Page 7 of 87 If any influence factors that were used were outside of the values identified in Table 6-3 of NUREG/CR-6850, identify the values used, identify the PAUs that use these factors, and justify the assigned factor(s).

SCE&G Response Only the factors listed in Chapter 6 of NUREG/CR-6850 have been used. See Section 4.8.2 in the ignition frequency report (TR07800-007).

PRA RAI 05 Describe the methodology that was used to evaluate DID and that was used to evaluate SMs. The description should include what was evaluated, how the evaluations were performed, and what, if any, actions or changes to the plant or procedures were taken to maintain the philosophy of defense-in-depth or sufficient safety margins.

SCE&G Response The methodology for evaluating defense-in-depth and safety margins is described in the Fire Risk Evaluation Report (Evaluation 11-4, Revision 1), and summarized below.

Based on NFPA 805 (Section 1.2), defense-in-depth is considered to be achieved when an adequate balance of each of the following elements (or echelons) is provided:

. Echelon 1: Preventing fires from starting. As indicated in the VCS Fire Protection Procedure FPP-022, administrative controls are in place to ensure control of combustibles, including, among others, combustible amount and class. Controls apply to in situ and transient combustible materials, with specific provisions for handling and storage of combustible liquids, compressed gasses, and oxidizers. Additionally, FPP-022 discusses administrative controls in place to govern the use of ignition sources during maintenance activities.

Echelon 2: Rapidly detecting fires and controlling and extinguishing promptly those fires that do occur, thereby limiting fire damage. The assessment of this echelon involves the evaluation of the adequacy of fire detection systems, automatic fire suppression systems, portable fire extinguishers, hose stations and hydrants located in the fire areas and the fire pre-plans.

Echelon 3: Providing an adequate level of fire protection for structures, systems, and components important to safety, so that a fire that is not promptly extinguished will not prevent essential plant safety functions from being performed. The adequacy of this echelon involves evaluating whether walls, floors, ceilings, and structural elements are rated or have been evaluated as adequate for the hazard, and whether openings in the fire area barrier are rated or have been evaluated as adequate for the hazard. This echelon is also assessed for supplemental barriers (electrical raceway fire barrier system, cable tray covers, etc.), fire rated cables, and the adequacy of the guidance

Document Control Desk Attachment 5 RC-12-0142 Page 8 of 87 provided to operations personnel detailing the required success path(s), including recovery actions to achieve nuclear safety performance criteria.

Insights of the Fire PRA are also used to evaluate the balance of the three defense-in-depth echelons. In particular, Echelons 1 and 2 are captured in the total fire scenario frequency for the fire area of interest, which accounts both for fire frequencies from ignition sources (which reflect combustible and hotwork controls implemented in Echelon

1) and fire non-suppression probabilities (which reflect potential credit taken for automatic and manual fire suppression capabilities). Echelon 3 is captured via the conditional core damage probability (CCDP) in the fire area.

It should be noted that even if automatic fire suppression features exist in a given fire area, they may not have been credited, due to modeling reasons, in the Fire PRA. The same goes with manual fire suppression (fire brigade).

A rule of thumb is employed in which a fire scenario is considered to be of high frequency if it can be expected to occur in average once every 10 years, that is, if its frequency is higher than 0.1/yr. This number is used as a criterion for the adequacy of Echelons 1 and 2.

In addition, a fire scenario that has a frequency greater than or equal to 1E-06/yr and a CCDP of at least 0.1 is considered to be of high consequence, and therefore this criterion can be used as a measure of the adequacy of Echelon 3. Because a picture of defense-in-depth at the fire area level is preferred, this criterion is adjusted by considering the average CCDP in the fire area, weighted by the frequencies of the associated fire scenarios (that is, the ratio of the CDF in the fire area over its total fire scenario frequency).

The results of the defense-in-depth evaluation for the fire areas addressed with a fire risk evaluation are documented in the Fire Risk Evaluation Report (Evaluation 11-4, Revision 1).

The evaluation finds that for these fire areas, adequate defense-in-depth is maintained. No actions or changes to the plant or procedures are found to be needed to maintain the philosophy of defense-in-depth.

Methodology for the evaluation of safety margin

-The maintenance of safety margin for the analyses supporting the fire risk evaluations of the fire areas considered in the Fire Risk Evaluation Report (Evaluation 11-4, Revision 1) is evaluated based on consideration of the following elements:

Fire Modeling: Fire modeling performed in support of the fire risk evaluations was done utilizing codes and standards developed by industry and NRC staff to provide realistic yet conservative results. For example, the heat release rates used in the transition analysis were based on NUREG/CR-6850, Task 8, Scoping Fire Modeling. These heat release rates are conservative and were used to screen out fixed ignition sources that do not pose a threat to the targets located in the fire areas.

Plant System Performance: Plant system performance parameters were not modified as a result of the fire risk evaluations.

Document Control Desk Attachment 5 RC-12-0142 Page 9 of 87 PRA Logic Model: The bases for the application of the codes and standards supporting the Fire PRA were not altered in support of the fire risk evaluations.

The safety margin evaluation finds that for the fire areas addressed with afire risk evaluation, an adequate safety margin is maintained. No actions or changes to the plant or procedures are found to be needed to maintain sufficient safety margins.

PRA RAI 06 The Transition Report summarizes equipment available to achieve and maintain safe and stable conditions but provides limited information about how long the facility can be maintained in those conditions using equipment that is free from fire damage and actions that can be taken from the CR. The submittal indicates that safe and stable conditions cannot be maintained indefinitely using this equipment. Provide a discussion about how long safe and stable conditions can be maintained using the undamaged equipment, and actions necessary to maintain these conditions once realignment and/or resupply activity such as refilling fluid tanks or re-aligning systems is required. If safe and stable conditions cannot be maintained indefinitely, evaluate at least qualitatively, and preferably quantitatively, the risk associated with the failure of actions and equipment necessary to extend these conditions indefinitely.

SCE&G Response The Fire PRA used a mission time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, per the ASME/ANS standard. For that 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period, only equipment that is "free from fire damage" is credited. At the conclusion of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period, each sequence in the FPRA is in one of three "safe-and-stable" states. These are:

1. On closed-loop RHR cooling.

2. On recirculation from the containment sump, with cooling to the heat exchangers.

3. Being cooled by the steam generators For the first two cases, there is no need for actions beyond the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> time frame. The equipment in operation at that time (at t+24 hours) simply needs to continue operating beyond that timeframe.

For the third case of (heat removal from the steam generators) it will be necessary to refill the condensate storage tank at some point after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. All other equipment in operation at t+24 hours simply needs to remain in operation. Since the mission time for EFW is 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, success means that it has operated for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and the SGs are full at that time. It is theoretically possible that there is still some water in the CST and that refill would not have to be immediate, but we can assume for this qualitative assessment that water is exhausted at that time. Given that the decay heat level would be low at this time, there would be an extremely long period of time after that to provide CST inventory, and the actions required are no different than for internal events and would also not be affected by a fire (the fire would long ago have been extinguished). The operating EFW pumps, at least one of which would be free from fire

Document Control Desk Attachment 5 RC-12-0142 Page 10 of 87 damage in the scenario, will trip automatically. For the most limiting case, the operators would need to add water to the CST and restart a pump (in some scenarios, they may only need to start RHR cooling, since just because the sequence is one where EFW is the success path doesn't mean that RHR is not available, but rather only that the question was not asked).

Again, because the SGs are full and the decay heat is low, the time available to perform the necessary actions is long. From a qualitative perspective, there is effectively no risk associated with failures beyond 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

PRA RAI 07 Section 10 of NUREG/CR-6850 Supplement I states that a sensitivity analysis should be performed when using the fire ignition frequencies in the Supplement instead of the fire ignition frequencies provided in Table 6-1 of NUREG/CR-6850. Provide the sensitivity analysis of the impact on using the Supplement 1 frequencies instead of the Table 6-1 frequencies on CDF, LERF, ACDF, and ALERF for all of those bins that are characterized by an alpha that is less than or equal to one. If the sensitivity analysis indicates that the change in risk acceptance guidelines would be exceeded using the values in Table 6-1, justify not meeting the guidelines.

SCE&G Response This sensitivity analysis on CDF and LERF was performed and is documented in the Task 15 report.

wftaingwinca tanbl for 2,quondensti alyses Amh ruFn for vreMV L Ph5 ant ion"d using the Tfve cD4F ffequseconda. vui .ItRtithe 8O 4suplemIntate tmat ad M94MMerte FIRANMY Mbe&

sarinft~vy showed tht1~e~ d~igkdorte mort

.!EuLt~elWw~n frmazndu cm be d#*,k~pe&

Table 2, "Sensitivity Analyses for VCS Fire PRA Quantification" The first value is CIDF. The second value is LERF. Note that while the supplement states that a sensitivity analysis should be performed, it does not state that the results of the sensitivity analysis should be used as the baseline value. Therefore, the wording of the last sentence, that if the results of the sensitivity analysis had exceeded the risk acceptance guidelines to "please justify not meeting the guidelines" is incorrect. The guidelines are met if the baseline value meets the guidelines. The sensitivity analysis is simply a piece of information to be weighed in the overall assessment of compliance under 4.2.4.2 (similar to consideration of defense-in-depth and safety margins). It is concluded that the FPRA is not overly sensitive to any of the sources of uncertainty associated with the existing plant. The FPRA is sensitive to the effectiveness of the proposed plant modifications, so there is a confirmatory item to update the FPRA after the modifications are actually implemented to assure that they achieve the predicted benefits.

Document Control Desk Attachment 5 RC-12-0142 Page 11 of 87 With regard to performing sensitivity analysis on the ACDF, and ALERF, there is no mention in either RG 1.200 or RG 1.174 (which are referenced by RG 1.205) for performing sensitivities on Arisk calculations either for demonstrating PRA quality (RG 1.200) or risk-informed compliance (RG 1.174). While such may be mentioned in the supplement to NUREG/CR-6850 with regard to ignition frequency, this does not make it a requirement (either for ignition frequencies or any other sensitivity cases). Therefore, with regard to requests that sensitivity analyses performed for the FPRA also include the sensitivity on the VFDR delta-risk calculations, we believe that there is no regulatory basis for requiring that information to be provided. The requirement to perform the VFDR delta-risk calculations, and compare them to the criteria in RG 1.174, is per RG 1.205. RG 1.205 refers only to the post-transition plant (the plant that contains alternatives to deterministic compliance - i.e., containing VFDRs, or the "variant" plant) and the "deterministically compliant plant." By reference to RG 1.174 as the basis, this clearly requires only that the baseline mean risk of the base plant (in this case the variant plant) be compared to the baseline mean risk of the modified plant (in this case, the plant with the theoretical changes that would bring it into full compliance under NFPA 805 Section 4.2.3, i.e., the deterministically compliant plant). There is no requirement to compare sensitivity cases for the base and modified plant. The only requirement for sensitivity analysis is in RG 1.200, which endorses ASME/ANS RA-Sa-2009, and requires that sources of uncertainty be evaluated using sensitivity analysis on the baseline PRA in order to develop an understanding of how these sources of uncertainty can affect the base PRA results. This is part of the process of assuring that the base PRA has sufficient capability to support the application. NRC staff has stated on numerous occasions that the role of RG 1.200 is to assure the technical adequacy of the base PRA for application to risk-informed regulation.

PRA RAI 08 Describe how CDF and LERF are estimated in MCR abandonment scenarios. Describe if any fires outside of the MCR cause MCR abandonment because of loss of control and/or loss of CR habitability. Describe whether "screening" values for post MCR abandonment are used (e.g., conditional core damage probability of failure to successfully switch control to the PCS and achieve safe shutdown of 0.1) or have detailed human error analyses been completed for this activity. Justify any screening value used.

SCE&G Response The post-transition abandonment procedure that will be used at VC Summer will allow the shift supervisor to order abandonment of the control room fire fires in the control complex (defined as CB-04, CB-06, CB-1 5, and CB-1 7) when they determine that the damage is such that shutdown from the MCR is not possible). With regard to the HRA for MCR abandonment, no screening values were used. Detailed HRA was used for two specific aspects of abandonment -

• Failure to abandon the control room when there is a loss of control (inability to successfully shut down from the MCR), which is assumed to lead directly to CD and to LERF. CB-17 is the control room. CB-04, CB-06, and CB-15 are outside the

Document Control Desk Attachment 5 RC-12-0142 Page 12 of 87 MCR, but have high potential to result in a loss of control if a large fire should occur in those areas.

Failure to successfully shut down from outside the control room given that MCR abandonment has been ordered.

Failure of either of these actions leads to CD and LER.

The final aspect was the probability of forced abandonment due to habitability problems in the MCR resulting from the fire (applies only to MCR fires, CB-17). The probability that habitability will be lost as a result of a fire in the MCR has been determined based on CFAST runs (see MCR Abandonment CFAST report). The HFE for failure to abandon the MCR is not applied in this situation (force abandonment is assumed), only the failure to successfully shut down after abandonment is applied.

It should be noted that (conservatively) no credit is given for actions in the MCR in the case of control room abandonment - all actions credited are associated with the remote shutdown panel (CREP), the actions outside the MCR required to enable the panel, and the other ex-MCR

-recovery actions required to achieve plant shutdown. While it is probably the case that some actions would be successful from the MCR before abandonment would occur, the possibilities are too complex and the different scenarios too numerous to model. Therefore, successful shutdown is modeled for the limiting case of no MCR actions. Note also that the remote shutdown procedure cannot achieve successful shutdown for the case of a LOCA- or SLB, so whether the abandonment is forced or voluntary if the scenario results in either of those conditions CD and LER is assumed.

The details of the model structure for control room abandonment are shown in Attachment 1 to this submittal.

PRA RAI 09 It was recently stated at the industry fire forum that the Phenomena Identification and Ranking Table Panel (PIRT) being conducted for the circuit failure tests from the DESIREE-FIRE and CAROL-FIRE tests may be eliminating the credit for Control Power Transformers (CPTs) (about a factor 2 reduction) currently allowed by Tables 10-1 and 10-3 of NUREG/CR-6850, Vol. 2, as being invalid when estimating circuit failure probabilities. Provide a sensitivity analysis that removes this CPT credit from the PRA and provide new results that show the impact of this potential change on CDF, LERF, ACDF, and ALERF. If the sensitivity analysis indicates that the change in risk acceptance guidelines would be exceeded after eliminating CPT credit, please justify not meeting the guidelines.

SCE&G Response The circuit failure probabilities for those circuits containing CPTs that were originally modeled with credit given for the CPT with a probability of 0.31 have been changed to remove that

Document Control Desk Attachment 5 RC-12-0142 Page 13 of 87 credit. The results of this change are shown in the table below. The results of this sensitivity run show that the impact on CDF and LERF are not significant. The conclusion of this analysis is that the post-transition plant still complies with NFPA 805 Section 4.2.4.2.

CDF LERF Baseline 7.75E-05 5.7E-07 CPT Sensitivity 7.93E-05 6.7E-07 Change from Baseline 1.80E-06 1.OOE-07 PRA RAI 10 Attachment W of the LAR provides the ACDF and ALERF for the VFDRs for each of the fire areas, but the LAR does not describe either generically or specifically how ACDF and ALERF were calculated. Describe the method(s) used to determine the changes in risk reported in the Tables in Appendix W. The description should include:

a. A summary of PRA model additions or modifications needed to determine the reported changes in risk. If any of these model additions used data or methods not included in the FPRA Peer Review, describe the additions.

b. Identification of new operator actions (not including post MCR abandonment which are addressed elsewhere) that have been credited in the change in risk estimates. If such actions are credited, how is instrument failure addressed in the HRA.

SCE&G Response The methodology for evaluating the ACDF and ALERF for the variance from the deterministic requirements (VFDRs) is described in the Fire Risk Evaluation Report (Evaluation 11-4, Revision 1), and summarized below.

For a given fire area, the ACDF and ALERF are calculated as the difference in risk between the fire area CDF (LERF) in the variant plant, which is the transitioning plant with VFDRs, deviating from the NFPA 805 deterministic requirements in the fire area, and the fire area CDF (LERF) in the compliant plant, which is the plant in which the VFDRs for the fire area have been assumed deterministically resolved.

The CDF (LERF) of the variant plant is evaluated in the Fire PRA. The fire PRA was developed in accordance with the ASME/ANS RA-Sa-2009 standard and was subjected to a peer review.

The CDF (LERF) of a given fire area is the sum of the CDF (LERF) of the fire scenarios belonging to that fire area.

Document Control Desk Attachment 5 RC-12-0142 Page 14 of 87 A VFDR is characterized by one or more functional states representing the discrepancy between the initial and desired state of the fire-affected equipment associated with the VFDR. In turn, these functional states are, as relevant, mapped to basic events included in the Fire PRA.

In the variant plant, the VFDRs contribute to the CDF and LERF of the fire area to which they belong.

In contrast, the VFDRs do not contribute to the CDF and LERF of the compliant plant, because in the compliant plant the VFDRs are assumed to be deterministically resolved. Thus, to model the Fire PRA of the compliant plant, the basic events associated with the functional states of VFDRs are 'toggled' to prevent them from contributing to the CDF and LERF of the compliant plant Fire PRA model.

There are cases where some VFDR-related equipment is not modeled in the Fire PRA because

-it has an insignificant risk contribution. In such cases, an explanation is provided to justify that the VFDR has an insignificant contribution to the ACDF and ALERF.

Attachment 2 to this response summarizes, for the VFDRs addressed with a fire risk evaluation, the Fire PRA modeling needed to determine the reported change in risk. This table does not include the VFDRs associated with Alternate Shutdown (ASD) Fire Areas, i.e., CB04, CB06, CB15, and CB17. The evaluation of the ACDF and ALERF for these specific fire areas is described in the response to the second part of the RAI, below.

Regarding the second part of the RAI, there are no recovery actions credited in the FRE other than those associated with alternate shutdown (i.e., MCR abandonment). In Fire Areas CB04, CB06, CB15, and CB17, which are the ASD fire areas, the ACDF and ALERF are attributed entirely to recovery actions. Recovery actions are modeled in the Fire PRA with a set of human failure events (HFEs) associated with ASD activities. In the Fire PRA, such HFEs are documented in Attachment C to the Fire PRA Human Reliability Analysis Report, Task 5.12 (Attachment 10 to DC00340-001). Table 1 of Attachment C shows the location for each step of the remote shutdown procedure, AOP-600.1. This table was used in conjunction with Table 3 of the same attachment, which shows the assigned human error probability (HEP) for each relevant step of the remote shutdown procedure. This table is reproduced below with additional information for the compliant plant analysis.

In the compliant plant, any action that takes place in the main control room prior to abandonment or at the evacuation panel (CREP) after abandonment maintains its originally assigned failure probability (that is, the HEP for these actions are the same in the variant and compliant plant). The recovery actions, that is, the local actions that take place outside the primary control station, are assumed to always be successful in the compliant plant and thus assigned a failure probability of zero. Assuming that these actions are always successful yields conservative delta risks.

Document Control Desk RC-12-0142 Page 15 of 87 Table. Human failure events associated with ASD activities Variant AOP AOP-600.1 Action Functional HEP (Baselin Complian 600.1 Summary Action Calculati e)Plan t Plant Step on HEP HEP A1-2 Trip the Reactor Reactor Trip FPVCS1 0.00023 0.00023 from the Main Control Board A1-3 Trip the Establish FPVCS1 0.00023 0.00023 Turbine/Generator Natural from the Main circulation Control Board A2-3 Trip all Main Establish FPVCS1 0.00023 0.00023 Feedwater Pumps Natural circulation Main-13 TRP RCPs Locally Establish FPVCS3x 0.0044 0 Natural circulation Main-16 Adjust the EFW Maintain SG FPVCS4x 0.0076 0.0076 flow to maintain level x each SG level between 60&65%.

Main-8 Transfer the Coordinate hot FPVCS2 0.0086 0.0086 following to LOCAL shutdown (CREP XPN- locally 7200B)

Main-4 Align controls Establish FPVCS2 0.0086 0.0086 Natural circulation Main-5 Transfer the Establish FPVCS2 0.0086 0.0086 following to LOCAL Natural (CREP XPN- circulation 7200A)

Main-7 Align the following Establish FPVCS9 0.0054 0.0054 controls (CREP Natural XPN-7200B) circulation Main-17 Establish natural Establish FPVCS11 0.0053 0 circulation cooling Natural xx SG pressure above circulation 1000 psig A2-12 Condensate control Establish FPVCS10 0.00095 0 Locally at (TB- Natural xx 412) circulation Total for all scenarios - Variant Plant 0.05 Adjusted HEP for ASD FRE 0.0395

Document Control Desk Attachment 5 RC-12-0142 Page 16 of 87 The Fire PRA that was submitted for the peer review was the post-transition (i.e., variant) plant, which included the proposed modifications. Therefore, none of the Fire PRA modifications required to evaluate the reported change in risk and summarized in Attachment 1 and the Table above differ from the data and general methods included in the Fire PRA peer review. The peer reviewers were made aware of the modifications that were included and the modeling used for those modifications.

All the operator actions that are credited in the change of risk estimates are associated with ASD activities (i.e., post MCR abandonment) and are described in Table 2 above. These operator actions were included in the variant plant model submitted to the peer review. The peer reviewers were made aware of the credited operator actions and the HRA used.

PRA RAI 11 Describe whether the peer reviews for both the internal events PRA (IEPRA) and FPRAs consider the clarifications and qualifications from Regulatory Guide (RG) 1.200, Revision 2, "An Approach for Determining the Technical Adequacy of Probabilistic Risk Assessment Results for Risk-Informed Activities," March 2009 to the American Society of Mechanical Engineers/American Nuclear Society (ASME/ANS) PRA Standard. If not, provide a self-assessment of the PRA model for the RG 1.200 clarifications and qualifications and indicate how any identified gaps were dispositioned.

SCE&G Response With regard to the fire PRA, the review was conducted consistent with RG 1.200, Rev 2. The peer review team for the Fire PRA used an industry (PWROG) peer review database that incorporates the SRs from AMSE/ANS RA-Sa-2009 (the version of the standard cited in RG 1.200 Rev 2.) For each SR where NRC provided a clarification or qualification for an SR in the RG, the database also contained the RG 1.200 clarifications and qualifications, and those were considered by the peer reviewers.

The industry peer review of the internal events PRA in 2002 pre-dated RG 1.200. However, the PRA has subsequently been reviewed against revision 1 of RG 1.200. The important Fact and Observation (F&O) suggestions from this review resulted in changes to the fault tree model. The VCSNS PRA model with these changes is the basis for the model that was used for the VCSNS Fire PRA.

Subsequent to initiating the VCSNS Fire PRA project, RG 1.200 was revised (Rev. 2). The VCSNS internal events model has been reviewed against this revision. Gaps between the VCSNS PRA and RG 1.200 Rev. 2 will be addressed in subsequent revisions to the internal events PRA. Changes made to the internal events PRA will also be incorporated into the fire PRA. The gap review did not significantly affect the fire PRA results.

Document Control Desk Attachment 5 RC-12-0142 Page 17 of 87 PRA RAI 12 Identify if any VFDRs in the LAR involved performance-based evaluations of wrapped or embedded cables. If applicable, describe how wrapped or embedded cables were modeled in the FPRA including assumptions and insights on how the PRA modeling of these cables contributes to the VFDR delta-risk evaluations.

SCE&G Response The use or credit of existing Fire Barrier wraps (ERFBS) within the NFPA 805 analysis atV. C.

Summer has been analyzed on a case by case basis, when performing the NSCA or Fire PRA analysis. Original passes through the analysis did not take credit for the existing, installed fire barrier wraps at the station. When required to resolve a DROIDs deterministically, existing wraps were "credited" as a means to disposition the Droid. When credited in-the NSCA, the associated circuits were not "affected" in the Fire PRA fire scenarios.

Embedded conduit containing required circuits were also considered to exist within the fire area/

zone, as appropriate, and dispositioned accordingly to address separation requirements via a Fire Protection Engineering Equivalency Evaluation (FPEEE).

When ERFBS were credited, they are credited to their full hourly rating of the wrap for the installed configuration. Testing basis and references were used to determine the adequacy of existing and new ERFBS (e.g. Supp 1 GL 86-10 testing criteria). When identified, enhancements to existing features (e.g. ERFBS) are addressed in Table S-1 of the NFPA 805 LAR. None have been identified for existing Fire Barrier Wraps, although additional wraps have been identified within the scope of the project to achieve resolution to DROIDs/ VFDRs.

PRA RAI 13 Identify any plant modification (implementation item) in Attachment S of the LAR that has not been completed but has been credited directly or indirectly in the change-in-risk estimates provided in Attachment W. When the effect of a plant modification item has been included in the PRA before the modification has been completed, the models and values used in the PRA are necessarily estimates based on current plans. The as-built facility after the modification is completed may be different than the plans. Add an implementation item that, upon completion of all PRA credited implementation items, verifies the validity of the reported change-in-risk. This item should include the plan of action should the as-built change-in-risk exceed the estimates reported in the LAR.

Document Control Desk RC-12-0142 Page 18 of 87 SCE&G Response As of this date, no plant modification in Attachment S has been fully implemented, although several are in various stages of development. Conservative modeling assumptions were made in the Fire PRA to develop the model. An implementation item to validate/update the Fire PRA model to reflect the as built modifications will be added to Table S-2 [Action]. This will also include a verification that the resulting change in risk is either less than that currently estimated in Attachment S or is within the guidance of RG 1.174.

Relative to the Fire PRA, this activity is covered under the SCE&G PRA maintenance and update program. Any modifications that are not yet installed at the time of transition, as agreed to upon approval of the LAR, will be evaluated as to the risk significance under the SCE&G PRA MU program, and implemented either immediately (if determine to be significant) or at the next regularly scheduled update.

PRA RAI 14 Identify any changes made to the IEPRA or FPRA since the last full-scope peer review of each of these PRA models that are consistent with the definition of a "PRA upgrade" in ASME/ANS-RA-Sa-2009, as endorsed by Regulatory Guide 1.200. Also, address the following:

a. If any changes are characterized as a PRA upgrade, identify if a focused-scope peer review was performed for these-changes consistent with the guidance in ASME/ANS-RA-Sa-2009, as endorsed by Regulatory Guide 1.200, and describe any findings from that focused-scope peer review and the resolution of these findings for this application.

b. If a focused-scope peer review has not been performed for changes characterized as a PRA upgrade, describe what actions will be implemented to address this review deficiency.

SCE&G Response The definition of PRA upgrade from ASME/ANS-RA-Sa-2009 is:

"the incorporation into a PRA model of a new methodology or significant changes in scope or capability that impact the significant accident sequences or the significant accident progression sequences. This could include items such as new human error analysis methodology, new data update methods, new approaches to quantification or truncation, or new treatment of common cause failure."

No changes made since the initial peer review of the VCSNS PRA meet this definition.

Document Control Desk Attachment 5 RC-12-0142 Page 19 of 87 PRA RAI 15 FSS-A4-01: According to DC00340-001, the multi-compartment analysis (MCA) postulated that a fire must burn 20 minutes to damage equipment across spatial separation or into an adjacent compartment. Provide justification for this assumption. Ensure that you discuss why damage may not occur in less than the time assumed, in this case 20 minutes.

SCE&G Response For the multi-compartment analysis, all pathways, even those with low barrier failure probabilities, are initially included in the multi-compartment analysis. During the screening process, as explained in Design Calculation DC00340-001 (Section 6.2), all multi-compartment combinations are considered for screening-out based on (1) qualitative screening (i.e. no PRA targets), (2) risk contribution (i.e., low core damage frequency), or (3) fire modeling (no full compartment burnout).

A sensitivity calculation was performed in which all the multi-compartment combinations that were screened out due to ignition frequency below 1.0E-7 were included. The additional contribution of CDF is approximately 3.0E-1 3, which is much lower than the total plant CDF of 5.62E-05. Therefore, the sensitivity analysis associated with brigade response time is not relevant to these scenarios.

As stated in DC00340-001, Section 5.2: "For the screening based on risk contribution, it is assumed that a fire that burns for more than 20 minutes may cause fire damage in an adjacent fire zone across a spatial separation. The rationale for this assumption is as follows. It was judged that a fire burning for 20 to 40 minutes could reasonably be expected to cause damage across a fire zone boundary that is based on spatial separation.

A sensitivity case was run for all multi-compartment combinations that were not screened out due to qualitative screening, risk contribution, or fire modeling. The time to damage for all of the multi-compartment scenarios was reduced from 20 minutes to 10 minutes. The resulting ignition frequency was computed in the fire modeling database and multiplied by the conditional core damage probability (CCDP) to get the core damage frequency (CDF) for both cases. The table below shows that the increase in CDF was found to be 7% above the CDF of the whole plant of 5.62E-05 (Attachment 12 to DC00340-01) [(7.82E 4.04E-06) - 5.62E-05 x 100=7%].

.Ignton Ilgntion D.Damage a o rreqUenc

-eqUency- rI-Freq20uenc- 10 mi 20rnn 0min ABO1.04-TO-ABO1.08.02 1.70E-03 4.77E-04 9.32E-04 8.1 OE-07 1.58E-06 AB01.04-TO-ABO1.09 1.28E-03 4.76E-04 9.31 E-04 6.09E-07 1.19E-06 ABO1.04-TO-ABO.10 1.02E-04 4.76E-04 9.31 E-04 4.87E-08 9.52E-08 ABO1.04-TO-ABO1.17 1.70E-03 9.23E-07 9.23E-07 1.57E-09 1.57E-09

Document Control Desk Attachment 5 RC-12-0142 Page 20 of 87 ABO1.08.01-TO-ABO1.08.02 1.28E-03 1.24E-05 2.42E-05 1.58E-08 3.1OE-08 ABO1.29-TO-ABO1.18.02 6.02E-03 1.27E-07 1.27E-07 7.65E-10 7.65E-10 ABO1.29-TO-ABO1.21.02 6.02E-03 9.11 E-07 9.11 E-07 5.48E-09 5.48E-09 ABO1.29-TO-CB20 6.02E-03 1.27E-07 1.27E-07 7.65E-10 7.65E-10 ABO1.30-TO-ABO1.21.02 6.OOE-02 2.26E-07 2.26E-07 1.36E-08 1.36E-08 ABO1.30-TO-ABO1.29 9.35E-05 2.26E-07 2.26E-07 2.12E-1 1 2.12E-1 1 DG01 .02-TO-TBO1.02 5.89E-05 3.24E-07 3.24E-07 1.91 E-11 1.91 E-1 1 FHO1.04-TO-ABO1.21.02 6.OOE-02 2.02E-07 2.02E-07 1.21 E-08 1.21 E-08 IB01-TO-IB22.02 6.29E-04 8.28E-07 8.28E-07 5.21 E-i10 5.21E-10 IB21-.02-TO-IB21.01 3.36E-05 1.86E-05 3.64E-05 6.27E-10 1.22E-09 1B25.01.01 -TO-IB25.01.03 2.01 E-04 5.54E-06 1.08E-05 1.11 E-09 2.17E-09 IB25.01.01-TO-1B25.01.05 3.92E-07 5.54E-06 1.08E-05 2.17E-12 4.24E-12 1B25.01.02-TO-1B25.01.05 1.01 E-01 2.37E-05 4.63E-05 2.38E-06 4.66E:06 1B25.01,04-TO-1IB25.01.03 1.79E-04 3.40E-06 6.64E-06 6.08E-10 1.19E-09 IB25.04-TO-!B25.01.05 1.56E-03 7.93E-06 1.55E-05 1.24E-08 2.42E-08 SWPH02-TO-SWPHO1 2.29E-05 1.05E-07 1.05E-07 2.40E-12 2.40E-12 TBO1.03-TO-IB20 1.27E-03 9.84E-07 9.84E-07 1.25E-09 1.25E-09 TBO1.03-TO-IB21.01 3.OOE-03 9.84E-07 9.84E-07 2.95E-09 2.95E-09 TB01.03-TO-IB21.02 3.OOE-03 6.07E-06 6.07E-06 1.82E-08 1.82E-08 TBO1.03-TO-TBO1.02 2.92E-05 3.13E-03 6.12E-03 9.14E-08 1.79E-07 TB03-TO-TBO1.01 2.47E-05 1.52E-07 1.52E-07 3.75E-12 3.75E-12 TB03-TO-TBO1.02 2.34E-06 2.80E-06 2.80E-06 6.54E-12 6.54E-12 YD03-TO-TBO1.02 6.41 E-03 8.30E-07 8.30E-07 5.32E-09 5.32E-09 Total 4.04E-06 7.82E-06 In conclusion, the assumption 6f 20 minutes for a fire to propagate to an adjacent compartment

-is a conservative choice; however, even if the time is lowered to 10 minutes, the CDF increases only a small amount.

PRA RAI 16 FSS-A4-02: According to the plant disposition, the individual fire zones have been updated to address the peer review concern that suppression credit is not adequately justified. The plant disposition indicates that this discussion has been added, but provides no technical response about the various types of suppression included, the criteria for its credit, and how it is credited. As a result, provide a technical response which addresses these issues for your crediting of fire suppression in your FPRA.

Document Control Desk Attachment 5 RC-12-0142 Page 21 of 87 SCE&G Response All individual zone reports contain a discussion of the suppression credit. For example, CB135, Section 7.1.3, states the following:

The scenario quantification in the Fire PRA credits the following detection and suppression features as specified in the fire modeling database for the appropriate scenarios:

• Incipient detection (see Supplement to NUREG/CR-6850, Chapter 13)

• Smoke detection

• Pre-action sprinkler system

• Fire brigade

• Prompt detection and suppression by hotwork fire watch for hotwork scenarios The incipient detection, the fire brigade, and the fire watch are all independent detection and suppression systems. The dependencies of the fire detection and suppression systems are discussed in Ref. 3.2.1, Section 6.1.3.3.

Currently-the Fire PRA models failures of fire protection features (i.e. smoke detection and fixed suppression systems) with generic failure probability values that are available in Appendix P of NUREG/CR-6850. These generic failures are considered bounding based on a comparison with plant specific data as calculated in, "PLANT SUPPORT ENGINEERING GUIDELINE-NFPA 805 FIRE PROTECTION MONITORING PROGRAM GUIDE, Revision 0". Under this calculation, plant specific values for reliability and availability are determined on a routine basis.

In all cases, the reliability and availability values for the smoke detection and fixed suppression systems are higher-than the generic values in Appendix P of NUREG/CR-6850. Technical details on the plant specific reliability and availability calculations for fire protection features are available in the Condition Monitoring Program (under Development). This information is currently housed in the VC Summer NFPA805 Share Point Site under Fire Prot NFPA 805-Task Documents -T39 Condition Monitoring.

There are both transient and fixed ignition sources in this fire zone. The sprinklers are credited to prevent hot gas layer formation after the initial target set is damaged by fire. Since some sprinklers are located within cable trays, it is assumed that the sprinklers will activate within the time-to damage of the initial target set (which is conservatively assumed damaged by fire). This assumption is based on the comparison of cable damage temperature and the sprinkler activation temperature. Since the sprinklers activation has a lower temperature, and some of the sprinklers are located within the trays, it is expected that activation occurs before fire can spread out of the initial target set and generates room wide damage.

PRA RAI 17 FSS-C7-01: The peer review finding indicates that dependencies between automatic and manual suppression systems are not addressed in the FPRA. The plant disposition indicates this is addressed, but provides no technical response. Provide a technical

Document Control Desk RC-12-0142 Page 22 of 87 justification that the dependency between automatic and manual suppression is addressed in the FPRA.

SCE&G Response All individual zone reports contain a discussion of the dependencies between suppression systems. For example, CB15, Section 7.1.3, states the following:

The dependencies of the fire detection and suppression systems are qualitatively discussed in Design Calculation DC0780B-001, Rev. 0,"Fire Modeling-Generic Methodology," Section 6.1.3.3. In the specific case of Fire Zone CB15, there are dependencies between the smoke detection system and the sprinkler system as a smoke detection signal is necessary to open the pre-action valve. In addition, there are dependencies between the operation of the sprinkler system due to availability of water supply and fire brigade actions requiring fire water. In the former, it is assumed that the probability of failure of the pre-action system modeled in the Fire PRA captures this dependency as this is a generic value calculated specifically for this type of system. In the later dependency case, failures in the fire water supply system in VC Summer can be isolated in and water re-directed in a number of points within the system (see for example drawing 302-231 for a flow diagram of the fire pumping system) to ensure water availability for the fire brigade (see also section 5 of the Fire Protection Evaluation Report).

PRA RAI 18 FSS-D3-01: The peer review indicated that in many cases detailed fire modeling was not performed, leading to conservative results. The plant disposition indicates that a set of scenarios in the auxiliary building (AB) had only received preliminary screening analysis at the time of the peer review, yet now has been done. No technical response has been provided in the plant disposition. As a result, describe those scenarios which were added to evaluate the AB, characterizing their ignition, damage, suppression, CCDP, and CDF/LERF.

SCE&G Response Detailed fire modeling for ABO1.21.02 in T3 and T5 was conducted. The fire modeling was documented in Fire Modeling: Fire Subzone ABO1.21.02, DC0780B-027. The following excerpt is from Section 7.3.1.1:During the Fire PRA analysis, the preliminary fire risk values were high enough in transient zones T3 and T5 that site-specific inputs to the fire modeling database were warranted. Fire modeling was conducted for five small electrical cabinets and one transformer in transient zones T3 and T5. The locations of the fixed ignition sources in transient zones T3 and T5 are shown in Figure 3.

Document Control Desk RC-12-0142 Page 23 of 87 Figure 3. Location of Fixed Ignition Sources in T3 and T5 The fire modeling hand calculations were performed to determine (1) if a fire in one of the fixed ignition sources would spread to overhead cable trays (temperature > 205'C); and if so, (2) if the resulting fire (fixed ignition source plus ignited overhead- cable trays) would generate enough heat to exceed the radiant heat flux damage criteria (6 kW/M2 ) and damage cable trays located 5 to 8 feet away: Figures 4 and 5 depict the relationships between the fixed ignition sources and the nearby cable trays.

Document Control Desk RC-12-0142 Page 24 of 87 2032 3124 '

3063 2024 Cable Trays 4108 4107 45039 R 5038 D

9' Ignition Source NOTE: The distances "D"-and "R" are given in Table 4. See Figure 3 for location of fixed ignition sources Figure 4. Fixed Ignition Sources and Nearby Cable Trays in T3 and T5, Side View, Looking North 4109 2032 3124 4691 [l 3063 Cable Trays C[a 4108 5039 81 6'

9' NOTE: See Figure 3 for location of Fixed Ignition Source XXS0061-PS.

Figure 5. Fixed Ignition Source XXS0061-PS and Nearby Cable Trays, Side View, Looking North Based on the walkdown notes in Appendix E, the distance to the closest target (overhead cable tray) was determined for each fixed ignition source and listed in Table 4 (distance "D" as shown in Figure 4). The location of the fire in each electrical cabinet is modeled to be 1 ft below the top of the cabinet, as recommended in Supplement 1 to NUREG/CR-6850 (Ref. 3.1.2, Chapter 12).

For the transformer, the fire is modeled to be at the top of the transformer. The time to damage of the closest overhead cable tray and the heat release rate

Document Control Desk RC-12-0142 Page 25 of 87 (HRR) at which damage occurred were calculated in file VC Summer Hand Caics ABO1.21.02.xlsx for each ignition source and are listed in Table 4. The values of (1) the heat generated by the fixed ignition source plus the three overhead cable trays and (2) the distance to the next closest set of cable trays were calculated and are presented in Table 4. Based on the radiant heat flux calculation, a fire involving any one of the ignition sources plus three overhead cable trays would not result in damage to any other cable trays in the transient zones; therefore, the "No suppression required" box was checked (TRUE) under Task 11 - Fire Modeling in-the fire modeling database, indicating no further targets are damaged.

Table 4. Inputs-to Fire Modeling Database for T3 and T5 Fixed Ignition.Sources 13 XI[-9u14 1.1 1 3 b. lb 474 TRUE T3 APN9014 2.2 3.5 20 5.75 468 TRUE T5 XPN7268A 3 5 40 5.75 550 TRUE T3 XPN98-FS 3 4 25 5.75 460 TRUE XXS0061- 8 51D T3 PS 6 5.5 45 TRUE T3 XPN7258A 4 7 70 5.3 531 TRUE Source: Walkdown notes in Appendix E; calculations in VC Summer Hand Cacs ABO1.21.02.xlsm and HRR CFAST ProfilesABO1.21.02.xlsm (Attachment 1)

NOTE: 1The distances "D"and "R" are shoym in- Figures 4 and 5.

PRA RAI 19 FSS-D8-01: The peer review found that credit for detection and suppression was not documented and thus rendered a Not Met to the corresponding SR. Your response only references dispositions to other facts and observations (F&Os) which also need additional technical discussion added. Provide a general technical description of the extent of crediting these fire protection features.

Document Control Desk Attachment 5 RC-12-0142 Page 26 of 87 SCE&G Response A discussion of crediting detection and suppression has been added to each individual zone report. For example, Design Calculation DC0780B-099, Rev. 0, "Fire Modeling-Fire Area CB15", Section 7.1.3, states the following:

The scenario quantification in the Fire PRA credits-the following detection and suppression features as specified in-the fire modeling database for the appropriate scenarios:

• Incipient detection (see Supplement to NUREG/CR-6850, Chapter 13)

• Smoke detection

• Pre-action sprinkler system

• Fire brigade

• Prompt detection and suppression by hotwork fire watch for hotwork scenarios The incipient detection, the fire brigade, and the fire watch are all independent detection and suppression systems. The dependencies of the fire detection and suppression systems are discussed in Design Calculation DC0780B-001, Rev.0, "Fire Modeling-Generic Methodology,"

Section 6.1.3.3.

Currently the Fire PRA models failures of fire protection features (i.e. smoke detection and fixed suppression systems) with generic failure probability values that are available in Appendix P of NUREG/CR-6850. These generic failures are considered bounding based on a comparison with plant specific data as calculated in "PLANT SUPPORT ENGINEERING GUIDELINE-NFPA 805 FIRE PROTECTION MONITORING-PROGRAM GUIDE, Revision 0". Under this calculation, plant specific values for reliability and availability are determined on a routine basis.

In all cases, the reliability and availability values for the smoke detection and fixed suppression systems are higher than the generic values in Appendix P of NUREG/CR-6850. Technical details on the plant specific reliability and availability calculations for fire protection features are available in the VC Summer NFPA Share Point Site-under Fire Prot NFPA 805-Task Documents-T39 Condition Monitoring.

There are both transient and fixed ignition sources in this fire zone. The sprinklers are credited to prevent hot gas layer formation after the initial target set is damaged by fire. Since sprinklers are located within cable trays, it is assumed that the sprinklers will activate within the time to damage of the initial target set (which is conservatively assumed damaged by fire). This assumption is based on the comparison of cable damage temperature and the sprinkler activation temperature. Since the sprinklers activation has a lower temperature, and the sprinklers are located within the trays, it is expected that activation occurs before fire can spread out of the initial target set and generates room wide damage.

Document Control Desk RC-12-0142 Page 27 of 87 PRA RAI 20 FSS-D9-01: According to the plant disposition, the discussion on smoke effects treatment in the FPRA was expanded as a result of the peer review finding. Provide a more complete technical response identifying the short term damage which you included in your FPRA.

SCE&G Response Design Calculation DC0780B-001, Rev. 0, "Fire Modeling: Generic Methodology," Section C.7, addresses smoke damage as follows:.

The approach for incorporating smoke damage in the Fire PRA follows the guidance available in NUREG/CR-6850, Appendix T. Long-term equipment damage from exposure to smoke, such as induced corrosion over a time scale ranging from days to months, is not a primary concern in a fire PRA because-risk-significant fire scenarios are resolved on a time scale ranging from minutes to hours. -Although some components are vulnerable to short-term smoke damage (electrical transmission equipment of 15 kV and higher, instrumentation with fine mechanical motion, and unprotected electronic components and circuit boards, etc), short-term damage requires severe smoke exposure. In accordance with the discussion in NUREG/CR-6850, Appendix T, ambient smoke exposures that are expected to be encountered in a compartment during a fire would not be severe enough to cause short-term damage. According to the guidance in NUREG/CR-6850, the only situations inside the plant in which smoke conditions are expected to be severe enough to cause short-term smoke damage involve:

Vulnerable components inside a burning electrical panel or cabinet or in adjacent -panels in an interconnected bank Very high voltage (15 kV) electrical transmission equipment exposed to a large oil fire.

In both of these situations, there may be incidental smoke damage, but the associated thermal conditions already cause the vulnerable equipment to fail even in the absence of smoke damage. That is, the Fire PRA assumes that the equipment will completely fail if the fire is postulated within the cabinet enclosure. At VC Summer, the following targets have been identified as vulnerable to short-term smoke damage:

0 Interconnected cabinet banks in the Relay room, Fire Zone CB06. Currently, the Fire PRA fails the entire bank of cabinets at the ignition of any of the panels within the bank. A bank of cabinets is considered to be counted (i.e., counted as directed by Chapter 6 of NUREG/CR-6850) interconnected vertical sections that are not separated by a double wall and an air gap.

Consequently, the short-term smoke damage due to high smoke concentrations within the cabinet enclosures is accounted for as a failure mode in the analysis.

Electrical buses (i.e. Power boards) in the AC and DC distribution systems. Examples of these panels include MCC's and switchgear cabinets as listed in the table below. Currently, the Fire

Document Control Desk Attachment 5 RC-12-0142 Page 28 of 87 PRA assumes that the entire bus is de-energized (i.e., assumed failed) for fires in any of the breaker cubicles. Consequently, the short-term smoke damage due to high smoke concentrations within the cabinet enclosures is accounted for as a failure mode in the analysis.

PRA RAI 21 FSS-G2-01: The plant disposition indicates that damage to sensitive electronics has been incorporated in-the FPRA. Indicate those components that meet the criteria for sensitive electronics and justify why this damage is not assumed beyond the CR and relay room as stated in Appendix V. For example, in many cases, switchgear has electronic undervoltage, overvoltage, under/over frequency and other types of sophisticated relays that are justas temperature sensitive as control modules. In your response, indicate what other electrical cabinets are installed in plant areas other than the MCR and relay room. As appropriate, provide an assessment on the additional effects of sensitive electronics on your FPRA.

SCE&G Response The damage criterion for sensitive electronics was used for the relay room, CB06. For several of the switchgear rooms (ABO1.29, 1B17, and 1B22.02), a hot gas layer scenario was postulated, in which the whole compartment was damaged. In addition, for several switchgear rooms (ABO1.29, 1B17, 1B21.01, 1B22.02, TB02) the panels are assumed damaged in the first damage state in the progression of the fire scenario, where suppression is not credited.

PRA RAI 22 FSS-H5-01: The peer review identified that parameter uncertainty evaluations were not performed. Counter to the plant disposition as stated in Appendix V, discussions during the audit revealed that these evaluations were not performed. Since this analysis is needed to achieve CCII, perform these studies and provide the results to achieve CCII. If CCI is all that is desired, justify the adequacy of CCI. In this justification, explain why CCI does not lead to underestimation of ACDF, ALERF, CDF, and LERF.

SCE&G Response There are two issues that are being confused. One relates to the uncertainty evaluations within the FSS task (the subject of this finding) and one pertains to the propagation of uncertainties in the final quantification (the subject of UNC findings). It is this latter case where it was discussed in the audit that the evaluations were not performed. This is addressed, and justified, in PRA RAI 85. For the finding that is the subject of this RAI, the disposition states:

"A discussion on the parameter uncertainty associated with the fire modeling results (when applicable) has been included in the individual fire modeling zone reports. It should be noted that not all fire zones receive fire modeling analyses. Consequently, this discussion is added only in the reports in which analytical fire modeling has been conducted for determining if hot

Document Control Desk Attachment 5 RC-12-0142 Page 29 of 87 gas layer scenarios are postulated in the fire zone. The parameter uncertainty discussion includes a qualitative listing of the uncertain parameters and when applicable the quantification of the uncertainty generated by key parameters as applicable to the scenario."

For the control room, uncertainty analyses have been performed for the heat release rate of the fire, the mechanical ventilation on and off, the fire brigade arrival time, and the door gap opening. In addition, individual fire zone reports contain sensitivity analyses for uncertain parameters, such as ambient temperature (e.g., the individual zone report for SWPHO1, DC0780B-290), and contain sensitivity analyses for dimensionless parameters that are outside of the validation and verification range for the fire model CFAST (e.g., the report for CB04, DC0780B-086).

Therefore, this finding has been addressed and the subject SR is now met at CC II.

PRA RAI 23 CS-A10-01: Provide a discussion on when routing is assumed. In particular, what is the

-routing assumed between cable trays in a single PAU or between PAUs. Describe how the assumed routing is treated when establishing fire PRA targets.

SCE&G Response The only use of assumed routing in the FPRA was, in actuality, assumed "non-routing" (i.e.,

exclusion). That is, credit was taken for offsite power from the 230kV supply when it could be assured that no offsite power cables could reasonably be expected to be affected by the scenario. Other than this single case (described in detail below) no routing was assumed. If routing was not available in the PC-CKS cable routing database for any cable included in the FPRA (for example, field routed conduit), then the routing was field verified by walkdown and entered into PC-CKS.

As discussed in Step 7b of the FPRA Task 5 report (Attachment 4 to DC00340-001):

230 kV Power Supply The 230 kV power supply provides redundant off-site power and, for the interest of the FPRA analysis, powers the CCW booster pumps which affects the mitigating function of RCP thermal barrier cooling. Because of the magnitude of the effort required to trace all 230 kV power system cables, it was decided early in the Fire PRA process to conservatively assume that 230 kV power is not available for any fire events. After progressing with the Fire PRA analysis, it became apparent that this assumption was overly conservative, and it would be beneficial to take credit for the availability of 230 kV power in areas where it could be easily verified that 230 kV power cables are not present and that faults in cables in the area would not propagate upstream and result in total loss of 230 kV power. A total of three fire zones were identified in which credit for 230 kV power could be taken. The following three fire zones were verified by plant walkdown to not contain 230 kV power system cables:

Document Control Desk Attachment 5 RC-12-0142 Page 30 of 87

" IB16 IB ESF SWGR Cooling Unit Room A

" IB17 IB ESF SWGR Cooling Unit Room B

• TB04 TB Switchgear Room 463 In addition, it was demonstrated by circuit analysis (see Task 5.9 Circuit Analysis Report) that fire damage to cables in the reactor building would not result in circuit faults that would propagate back and result in loss of 230 kV power. Thus, availability of 230 kV power was also credited for fire zone RB01.

Thus, the 230 kV power supply was uniquely modeled to allow credit in only a few fire zones. A "230 kV power exclusion" was modeled in which the 230kV power supply is assumed to be failed in all zones except for 1B16, IB17, and TB04. Credit can be given to the 230kV power supply remaining available in these zones when the 11-5kV power supply is lost. Because a full circuit analysis was not done for the 230kV power supply, a pseudo-target "FLAG-230KV-EXCLUDE" was created to represent the cables that map to the 230kV power supply. This pseudo target is mapped to all areas where credit cannot be given for the 230kV power supply and it therefore is-failed. This pseudo target is not mapped to 1B16, 1B17, and TB04. The pseudo target is also mapped to the basic events that fail the 230kV power supply or directly correlated to the failure of-the 230kV power.

PRA RAI 24 PP-B2-01: This fact and observation (F&O) is listed as a suggestion, but it is the only F&O describing PP-B3 which is attributed a "Not Met" by the peer review. The peer review suggestion identifies that spatial separation as a partitioning feature is not justified for certain cases. The justification is not provided in the plant partitioning document. Provide a description of the configurations for those configurations identified in the F&O, and justify spatial separation as a partitioning feature. The plant disposition does not address the partitioning, yet indicates that the multi-compartment analysis takes into consideration scenarios which accounts for the interactions across this spatial separation.

SCE&G Response Barrier failure probabilities of 1 were used for multi-compartment combinations with spatial separation as the barrier. The multi compartment combinations are quantified. For example, the multi-compartment combination of IB25.01.02-TO-1B25.01.03 has a barrier failure probability of 1.0 and is included in the quantification (DC0780C-001, p. A-77 and p. C-1). Spatial separation was used when no barrier exists, such as between subzones (like 1B25.01.02 and 1B25.01.03) or if there are non-rated barriers due to openings in the walls or floor/ceiling assemblies, such as between ABO1.09 and ABO1.04 (DC0780C-001, p. A-8). Using spatial separation as a barrier, with a failure probability of 1.0, within the Fire Modeling Database is useful to keep track of cables that are mapped to subzones.

Document Control Desk Attachment 5 RC-12-0142 Page 31 of 87 PRA RAI 25 ES-A6-01: Item 4, pg. 14 of the FPRA plant final report indicates that at a maximum 2 spurious operations are included. With respect to ES-A6, this implies that CCII was not met. The peer review rating in table V-16 confirms that CCI was assigned, but the F&O ES-A6-01 is not provided in-the LAR. Provide this F&O and the plant disposition. Justify the achieved capability category.

SCE&G Response This statement in the report is incomplete. The limit of two MSOs was applied only to the identification of new initiating events that do not lead directly to plant trip and also containment bypass (per ES-A5). For new initiating events that lead to containment-bypass (ES-A6) up to three failure were considered. This-is-documented in the FPRA Task 5 report (Attachment 4 to DC00340-001) under Step 8.. For example the below item shows where three spurious operations were included.

a .. Lpno*

• ,,p.ingcq rfi- ;t ) tjtif e s, E Le= tdown tLi& i fI s L i a ve Muti~plle spuwi~us openng of XV'T-153, -8154 and HCV-137 is reqquiredct fai islation ofr te excess letdown pa-ih. NormalIy dosed ,,al',e: HCV-137 ras; not been-electrica liy a*nalyzedý and ýL' not induded in the logir-L Ts is a conrei-vative au:mptlon in the MSO rmoddljia because ii esenti-all assumes a normal!ly close vake always spuriously opens- Sputious opening of XVT-8153 a nd XVT-8154 has been included u-nder gate

MS08-LET OWN. w hi is rncluded in gate LET-002 for random=ailure, of letdown-The limitation on spurious operations only applied to the identification of new initiating events (and the ES associated with them). -For all equipment selected, applicable spurious actuations were included in the FPRA model and the model quantified with no limitation in the number for spurious actions that could appear in a cutset.

PRA RAI 26 For F&O PRM-A4-01, summarize how, for a given fire scenario, the appropriate initiator is selected (due to impacted equipment) and the related mitigation system fault tree logic is valid.

Document Control Desk RC-12-0142 Page 32 of 87 SCE&G Response The approach to initiator selection was changed after the peer review to address the F&O. The revised approach is discussed in the FPRA Task 5 report (Attachment 4 to DC00340-001) under Step 2, as follows:

lo ea!ch ti-re caa the nue ntiirasge is %FIRL, This new initratr wa-s added toe caIinto pei atora -dutndued iniitorTis a turbinre trip (3d!). o atgt hr %zTT a~ppears in'tie modal, F! was added alng with t (Hunder OR7te g-ej_ Whenr Te equipment faiha-es are set de.ing quanilhiation gat th 'Itim dhcaie aprurin tite 5617 oitL Th 'ER nitiat-or was -als weea g aLd6 toý zhat caiEr,- other typs i varq These in~clude ctonscqequwitil s7af v C me& mLLOCA (gavte MOFR eoia side ,gt -FlREJ, end S1Ibreak (gate SISr-F1RE). The %`ýFE ini a Tr wa dded in those ares e te so hthe eect of cerrau equi'ent falue would propagate Through the model withthe approrat'e initiator, In this way, each fire scenario can- result in either one specific initiating event (if only direct failures are involved) or various initiating events (based on spurious operation probability, if applicable). As an example of how this works, the logic is shown in the screen capture of the SLB-FIRE gate below. It was determined using the MSO expert panel report that MSO 23 could contribute to a SLB. This logic includes spurious failures of IPV-2020 (basic event IPV-2020-FIRE). This event is mapped in the FRANX model to fail the event with a hot short probability.

The other BEs in the logic are mapped either to fail (set to TRUE) or fail with a hot short probability. If the gate-SLB-FIRE-INIT is true, the scenario cutsets can include initiating events of this type. Any cutsets that are included in the model that are associated with this initiator will include the event SLB-FIRE for identification purposes. Other initiating logic is developed and structured similarly and is described in the Task 5 report.

Document Control Desk RC-12-0142 Page 33 of 87 PRA RAI 27 F&O PRM-A4-02 identifies improper modeling of multiple-spurious operation (MSO) scenarios. The disposition only states the model was reviewed for accuracy, but does not identify if the errors flagged by the peer review team are -corrected, if other errors were corrected, etc. Clarify the disposition of this item.

SCE&G Response The specific MSO-related errors identified during the peer review were fixed on-the-spot throughout the week of the peer review. To address the specific F&O-associated MSO errors, MSO-27 logic was updated in the fault tree model and can be found in gate EFW-SBO in sub-gates MSO-27-A, MSO-27-B, and MSO-27-C. Additional review of MSO-28 determined that the valves in question are locked open and therefore do not require modeling.

After the peer review was complete, the MSO panel report was reviewed against the MSOs included in the model to ensure that they were modeled appropriately. It was concluded that the errors identified during the peer review were the only ones that required correction in the model. Step 8 of the FPRA Task 5 report (Attachment 4 to DC00340-001) includes a description of modeling for each MSO and documents the logic captured in the fault tree.

Document Control Desk Attachment 5 RC-12-0142 Page 34 of 87 PRA RAI 28 F&O PRM-A4-04 identifies MSO modeling errors based specifically on a sample and not a 100% review. The disposition of this item identifies correction of errors and a review of MSOs for additional issues. Describe the scope of the review performed to determine the extent of condition.

SCE&G Response After the peer review was complete, the MSO panel report was reviewed against the MSOs included in the model to ensure that they were modeled appropriately. -]twas concluded that the errors identified during the peer review were the only-ones that required correction in the model.

Step 8 of the FPRA Task 5 report (Attachment 4 to DC00340-001) includes a description of modeling for each MSO and documents the logic captured in the fault tree.

PRA RAI 29 For F&O PRM-A4-05, the disposition of this item-only states that a particular set of logic was modified and included in the fault tree appropriately. Discuss how each technical point raised by the peer review team was addressed, as well as how the extent of condition was determined given that the peer review only identified an example.

SCE&G Response This is discussed in the FPRA Task 5 report (Attachment 4 to DC00340-001) under Step 2:

lie S12-F-IRE was addet he moisel 5c! hat spu:ou aduatiua asaaed pi-h VIc ASS-FR n~de~e4 sub-tree~was crated inCAFTusing t~~r'c d t, Th loi

,Dt ur!~a Eth data P IMD-lnhfron ;Erawing 1zS A. -11 08, cuin RBpicsredy, puo E p ogic. aTsorit, naisthetprg gatra is F-3erepreANmodel Mmvdet. ?r' 1ýTEres intos rom iaents tha=t faldeto suiusho so of C0Forn ac-tJonS car, =-,nd SI5 odr ocatr tlhese, e;eTEr- kic va de to the r~~~~0~~~~I.~~  ;~beo oh I~ t7 r~iI i cha0rl zr~~~wr C aiIn I-ar Trha jr ti d sP,ua s oTs c,rTs ffrn ronbntkxs Pi r 1 3i ' 0 'j A SIS-FIRE sub-tree was created in CAFTA using data from drawing lMS-041 -011 Sheet 0008, including RB spray pump logic. This logic was run as the top gate in the FRANX model. No cutsets were generated when Gate SIS-FIRE was quantified, i.e. there were no scenarios where 2 out of 3 RB high pressure transmitters were present. Therefore, no additional fault tree modeling is required for the spurious operation of RB Spray Pumps.

Document Control Desk Attachment 5 RC-12-0142 Page 35 of 87 The final top level logic is shown below. As can be seen it is in an OR relationship with the random inadvertent SIS. The entire logic is available in the CAFTA file. This is not an extent of condition issue; it is a specific case that is not indicative of a broad modeling issue.

PRA RAI 31 IGN-B5-01. According to the peer review finding, a qualitative discussion of the sources of uncertainty is needed to address this standard support requirement (SR). The peer review cites Appendices U and V of NUREG/CR-6850 as guidance for this issue. The plant disposition solely identifies the distributions used for the calculation, but does not address the peer review comment. Identify the assumptions and sources of uncertainty

Document Control Desk RC-12-0142 Page 36 of 87 associated with the fire frequency analysis as required by the supporting requirement, IGN-B5, of the ASME/ANS PRA Standard.

SCE&G Response Please see the response to PRA RAI 85.

PRA RAI 33 For F&O HRA-C1-01, describe how dependencies were removed from-the charging pump swap in the recovery action.

SCE&G Response To clarify, the issue is that the charging pump swap actionprovides additional time for the completion of the action to start a SW-Pump. That is, the HEP for failure to start the SW pump assumes that charging pump swap is being done, thus providing the additional available time.

The proposed solution in the F&O is to remove the credit from the SW start action (reduce the available time by half). However, this solution is overly conservative for the following reason.

The two actions are performed by different individuals using different cues and different

-procedures. Therefore, the actions are essentially independent (at most a low dependence). If the actions were to be modeled in detail, then the "cut sets" for failure to start the SW pump would be:

Fail to Swap CP

• Fail to Start SW (short time frame)

Fail to Start SW (long time frame)

That is, the failure to start the SW in the short time frame would apply only if the CP swap failed. The resolution suggested for the F&O does not acknowledge this. Rather, it proposes-never crediting the longer time frame.

Because the two actions are essentially independent, the first situation is not modeled because it was concluded that the dominant contributor to failure to start the SW pump is the second cutset (the single element cutset for the long time frame case). While the F&O makes a valid point when looking at the approach in the vacuum of the HRA SRs only, when considering the overall picture of the entire PRA this simplification of the model does not affect the results, so it is not necessary to model the non-dominant combination.

PRA RAI 34 For F&O HRA-C1 -02, provide the basis for the time required to perform a manual action inside and outside the CR for a fire, including validation or reference to an accepted methodology. Describe when and where the basis will be documented.

Document Control Desk Attachment 5 RC-12-0142 Page 37 of 87 SCE&G Response The basis is documented in the VC Summer Operator Manual Action Feasibility and Reliability Study, which is Attachment A to Fire PRA Human Reliability Analysis Report, Task 5.12 (Attachment 10 to DC00340-001).' As noted by the peer reviewers, the main body of the HRA report only states that 5 minutes and 10 minutes were used. The detailed discussion of this approach is in the feasibility and reliability study, and the peer review team apparently did not review this document. The following discussion of the basis is-taken verbatim from the feasibility and reliability study. Since this basis is fully documented in this location, no further action is planned.

NUREG/CR-6850 Task 12 describes the process for performing both screening and detailed analysis of post-fire human actions identified in the Internal Event PRA or IPEEE. It also describes how to identify and quantify new actions to be performed as part of the plant fire mitigation plans and procedures. The guidance identifies performance shaping factors (PSFs) and fire effects to consider in performing detailed Human Error Probability (HEP) calculations.

These include staffing, training, procedure controls, timing, environment, and accessibility.

EPRI/NRC Fire HRA Guidelines (NUREG-1921), was developed-to provide a methodology and supporting guidelines for estimating human error probabilities (HEPs) for human failure events following fire-induced initiating events for probabilistic risk assessments (PRA). It provides a basis for the process of identifying Human Failure Events (HFEs), the methodology for assigning quantitative screening values, and initial considerations of performance shaping factors which need to be addressed during fire events. The FHRA guideline was intended to be a stand along reference which supplements the guidance provided in NUREG/CR-6850 Task 12.

As part of this guidance, demonstrating feasibility and the use of time margins are important to show that the human action times are estimated based on demonstrations that-are realistic as possible. Additional feasibility and reliability guidance is referenced and provided in NUREG-1852. This guideline identifies criteria for demonstrating the feasibility and reliability of operator manual actions (OMAs) outside the main control room (MCR). This guidance provides technical-information justifying operator manual actions are feasible and can reliably be performed under a wide range of plant conditions that an operator may encounter during a fire event. It additional provides performance shaping factors to consider in performing detailed HEP calculations.

At VC Summer, certain Internal Event human-failure events were screened during the review of relevant fire plant initiators. Actions associated with sequences such as ATWS, SGTR, ISLOCA, etc. were screened out consistent with discussions in NUREG/CR-6850. After the initial screening, a detailed human error probability (HEP) was performed utilizing the EPRI HRA calculator. For each fire human action, an additional 5 minutes was added (over and above the Internal Events timing) to the T1/2 (Median Response Time) for fire initiating events where the HFE was performed solely within the Main Control Room. This additional time is to account for delays associated with operators trying to gather and diagnosis the issues at hand. For HEPs being performed outside the Main Control Room, an additional 5 minutes (total of 10 additional I For clarification, the entire Fire PRA report is designated as DC00340-001, and the main body is a summary of the entire FPRA. The individual FPRA task notebooks are designated as attachments to that report. The HRA notebook is Attachment 10. In addition, these notebooks often have attachments. The feasibility and reliability report is thus Attachment A to Attachment 10 to DC00340-001.

Document Control Desk RC-12-0142 Page 38 of 87 minutes) was added to the T1/2 (Median Response Time) to account for environmental, travel path, emergency lighting, etc. It should be noted NUREG -1852 states that increasing the timing by a factor of 2 would demonstrate a "high confidence of a low probability of failure" for local operator manual actions in response to a fire. However, a factor of 2 was only added for actions outside the MCR where numerous rooms need to be accessed to perform the OMA.

HRA analyst consensus felt that adding a factor of 2 for all ex-control room actions is overly conservative. Note - For certain OMAs requiring multiple room entry points, an additional 10 minutes (total of 20 minutes) was added to these specific actions.

Additionally, for each detailed HEP, the Emergency Operating Procedures (and other relevant procedures) were reviewed to identify key indications which were deemed necessary for the success of each fire human action. These key indication cables were-routed and placed in the model (i.e. fire initiators that fail the OMA) or dispositioned for exclusion from the model.

Lastly, for each detailed HEP, travel paths were identified. This review identified pertinent rooms which must be accessed to successfully perform the human action. If an operator has multiple ingress/egress paths, then these rooms were screened. Using, for example, a fire event requiring an operator to perform a task in the Auxiliary Building (AB), the travel path from AB Turnover Station was identified. If multiple paths exist, these rooms (paths) were not incorporated into the list and hence will not be included in the FPRA as impacts. When rooms were identified as having an impact, they were placed in the model as failing the HFE. Given a fire initiating event in that area, the human action is failed because there is no "safe" pathway for the operator to reach the location to perform the action.

PRA RAI 35 Disposition in Attachment V, Table V-18 for supporting requirement SF-A4-01 indicates plant procedures for fire brigade drills and natural emergencies have not been updated to address seismically induced fire. Confirm that procedures are now updated. The supporting requirement in the ASME/ANS Standard also requires a qualitative assessment of the potential that a seismically induced fire, or spurious operation of fire suppression systems, might compromise a post-earthquake plant response. Provide this assessment based on the updated procedures.

SCE&G Response The Disposition in Attachment V, Table V-18 for supporting requirement SF-A4-01 actually states that sentences in DC00340-001, specifically referred to in the F&O, were revised to provide recommendations to address seismic issues in plant procedures and training. This closed the open ended sentences as directed in the proposed disposition outlined in the F&O.

An implementation item to incorporate these recommendations into plant procedures and training will be added to Table S-2 [Action].

Document Control Desk RC-12-0142 Page 39 of 87 PRA RAI 36 Finding in Attachment V, Table V-18 for supporting requirement ES-A4-01 establishes a deficiency in the linkage between the Component-BE table in the FRANX database and computer aided fault tree analysis (CAFTA) for spurious operation of pressurizer spray valves. The response is not clear as to whether the spray valve failure mode was in the model (and the finding is invalid) or whether the model was corrected. Clarify this response. In addition, the disposition addresses one particular scenario the peer reviewer discovered and provides small confidence this problem will notappear again for other components. Describe the root cause of this problem and provide verification that this issue will not be repeated for other scenarios and that all pertinent components included in the FRANX database are linked to the CAFTA fault tree.

SCE&G Response The commenter has misinterpreted the finding. This was not a fundamental issue of the linking of components to the tree, but rather one of the treatments of a specific MSO in the analysis.

The MSO 35a modeling was added in response to this finding as a cause of inadvertent SI. The is documented in ATTACHMENT 4 TO DC00340-001 FIRE PRA PLANT-RESPONSE-MODEL REPORT, TASK 5.5, as shown in the following description and fault tree.

RCS PRESSURE CONTROL 35 RýIOS)JelSl Pi0u eP-S~I Orr~' spa a~esAW

~I (lY ttreis, rioeratio ýif, RCPAND.

Based on the MSO evaluation rep<ort (Attachment I of the TASK 5.2 technical report),

this scenario requires spurious operation of multiple combinations of valves, pumps and heaters, In addition, operational experience has shown that all heaters fundioning-would not overcome one spray valve failed open. Scenario screened based on more than two required spurious 7W Reo ancl spus oucs i ofIImal ISpufuuspraypnin 11 in '.

_6 This combination of component failures will cause an SIS and is modeled in the #SIS-FIRE initiating gate. This MSO is specifi~cally modeled in gate MSO=35A.

Document Control Desk Attachment 5 RC-12-0142 Page 40 of 87 PRA RAI 37 Finding in Attachment V, Table V-1 8 for supporting requirements ES-BI -01 and ES-B1 -03 identify apparent significant errors in the underlyingdata used in quantification of fire risk. Specifically, the peer review identified records referring to nonexistent basic events in the PRA model, and PRA model basic events representing potentially legitimate fire failure modes which were never identified as being impacted. Concerns were also identified as to the manual methods of generating these files, as opposed to a database query method. While the disposition of these F&Os identifies a review of the mapping results and addition of documentation to any unmapped basic event, there is no discussion as to the impact of any corrections to the risk results. Further, the staff assumes that the original databases were reviewed, but the errors still were not found. It is not clear if the review conducted to resolve the F&Os was somehow more comprehensive than prior reviews to ensure identification of errors. Discuss the impact of these F&Os on the FPRA due to model changes resulting from the re-review of the mapping.

SCE&G Response The errors identified in the F&O were predominantly associated with unmapped basic events (BEs) that did not require mapping. All BEs were reviewed and any that remained unmapped in the model were dispositioned to reflect the reason why mapping was not required (i.e., common cause failures, flag events, HRA events, etc.). The majority of BEs identified did not require mapping in the model. After the review was complete, the model was updated with any necessary changes. Additionally, a column was added to the BE table of the CAFTA database

Document Control Desk RC-12-0142 Page 41 of 87 to cite the disposition of each unmapped BE for reference. These changes were included in the version of the model used to develop the LAR.

PRA RAI 38 Finding in Attachment V, Table V-18 for supporting requirement ES-B3-01 identifies that containment penetrations screened in the IEPRA based on small size were not fully reviewed for inclusion in the FPRA. It also identifies differences in the CDF model and the LERF model used for the FPRA, and changes needed to screening criteria for containment penetrations. The disposition only states that additional penetrations have been identified, but does not discuss either that the FPRA model has been revised to correct the issues or provide any basis for why the existing FPRA-LERF model is adequate. It does not discuss if the CDF and LERF models remain different, nor does it discuss any changes to screening criteria. Discuss how this finding was addressed, including identifying if changes to the model were made, and justify the technical adequacy of the LERF model used for the FPRA.

SCE&G Response The revised consideration of penetrations is discussed in the FPRA Task 5 report (Attachment 4 to Design Calculation DC00340-001) under Step 13, as follows:

Step 13: Model Any New Accident Progressions identified to determine the Fire -

induced LERF In Step 12, the only new scenario identified is the multiple spurious operation of RHR suction valves (two in series). The scenario is within-bounds of the existing accident progression modeling for ISLOCA; therefore no new accident progressions are needed to be modeled.

Several previously screened containment isolation valves were evaluated for applicability to the Fire PRA LERF analysis. Based on the IDCOR Technical Report 86.3.A2 paper, pipes with a 2 inch or less diameter can be screened from the LERF analysis. The equation for compressible flow through a pipe is:

1 q'= 6.87 (Yd1 C/Sg)*((Pi-P 2) Ppl) /2 2

(Crane Technical Paper 410 equation 3-22). For an equivalent flow through multiple orifices, ratio the d1 2 of the pipe inside diameters. Table 12 shows equivalent flow.

Table 12. Equivalent Flow Through Multiple Orifices for LERF Analysis Nominal Pipe/Tube ID d, 2 # Equivalent to 2" 2 sch 40 2.067 4.273 1 sch 40 1.049 1.100 3 1/4 sch 40 .884 .782 5 3/8 tube .083 wall .209 .0437 97

Document Control Desk RC-12-0142 Page 42 of 87 Table 13 summarizes the containment isolation valves and the application of each in the LERF analysis.

Table 12 shows how the penetrations screened from the internal events PRA were re-evaluated for inclusion in the FPRA LERF model. The four pipe sizes identified are shown in the table.

The table also shows the number of penetrations of a given size that would have to fail open in order to be a LERF. As is shown, for all the pipe sizes less than two inches, at least three penetrations would have to fail open to cause a LERF. Since each penetration contains two valves, and they "fail closed," it would take six spurious failures to cause a LERF. Therefore, penetrations smaller than 2" can be screened. The two inch penetrations would need to be included unless they can be screened on another basis. The final disposition of the penetrations for the LERF model are given in Table 13 of the report. The basis for the use of 2" equivalent penetration size as the definition of LERF is based on Westinghouse Owners Group, WCAP 16378 "Westinghouse Owners Group Definition of Large Early Release" November 2004.

PRA RAI 39 Finding in Attachment V, Table V-18 for supporting requirement ES-B4-01 establishes a deficiency in modeling of support equipment whose fire-induced failure could adversely affect primary equipment. The disposition addresses one particular finding associated with power dependency and provides small confidence this problem will not appear again for other support equipment. Describe the root cause-of this problem and provide verification that this issue will not be repeated for other support equipment listed in SR ES-B4.

SCE&G Response A complete check of the dependencies identified in the circuit analysis database was performed to assure that they were incorporated into the FPRA. The circuit analysis database includes dependency considerations for interlocks, power supplies, support system, instruments, etc.

The missing dependency was found to be a singular occurrence (an "oversight").

PRA RAI 40 Disposition in Attachment V, Table V-18 for supporting requirement CS-A8 identifies tasks to address the item, but there is nothing to indicate these items have been completed or what the resulting model impacts were, or if not completed why the deficiency is not significant to this application. Provide additional details for this item.

Document Control Desk RC-12-0142 Page 43 of 87 SCE&G Response Per specific guidance on treatment of Kerite as a qualified cable, it is a qualified cable with a lower damage threshold. The Fire PRA follows the generic guidance for Kerite cable: its treated as a qualified cable (no jacket or insulation melting) for determination of circuit failure mode probabilities, no need for postulating self ignited cable fires, and flame spread criteria, but with a thermoplastic damage threshold (see p. A-12 in NUREG-1805). The basis for screening these valves from modeling for interfacing systems LOCA is document in the Task 5 report (ATTACHMENT 4 to DC00340-001, FIRE PRA PLANT RESPONSE MODEL REPORT, Task 5.5 under Step 8, discussion of MSO-16) as follows:

16 Interfacing Spurious opening of multiple series RHR suction valve from System LOCA RCS An interfacing system LOCA (ISLOCA) resulting from spurious opening of multiple series RHR suction valves from the RCS is identified in Task 5.2, Step 7 as the only high consequence event that is not screened.

Power cables associated with these valves are "thermoset" cables and part of a grounded AC system (Ref: PC-CKS Electrical Analysis Database), therefore, they are not analyzed for three-phase proper polarity hot shorts. From NUREG/CR-6850, Task 9, Section 9.5.2:

Document Control Desk RC-12-0142 Page 44 of 87 Three-phase properpolarity hot shorts on AC power systems:

Case 1: Grounded AC system with thermoset-insulatedcablelO. Three-phase properpolarity hot shorts are evaluated as extremely low-probabillty events for grounded three phase AC power systems. Based on observed characteristicsand behaviorof fire-induced cablefailures, an estimated upper bound on the probabilityof occurrencefor a three phase circuit utilizing *hermoset-insulated triplex cable (one 3-conductor cable) located in a Pypicol cable tray or conduit is 5E-8/yr. This bound considers:

- The likelihood of rnulticonductor-to-multiconductor hot shorts for thermoset insulatedcable, The likelihood of concurrentand independent phasefaults,

- The likelihood of phase faults of the properpolarity, i.e., phase rotation, and

- Typicol fire ignition and severty frequencies and suppressionfailure probabilities.

On this basis, the three-phaseproperpolarity hot shortfailure mode is not consideredrisk-significantin accordancewith the defined screening criteriaof lE-7/yrfor "potentiallyhigh consequence equipment,' as defined by Section 2.5.6. It is recommended that this failure mode not be included in the Fire PRA cable selection processfor groundedthree-phaseAC circuits involving thermoset-insulatedcable.

in addition, the 480V power supply breakers to the valves are locked open during plant operation (Ref: VCSNS Technical Specification 4.5.2. "Emergency Core Cooling Systems"). A surveillance is performed each shift (shift logs) in Modes I - 3 to verify that the valves are closed and tlfe breakers are open (power removed). Therefore, spurious opening of these valves is not postulated.

Circuit analysis for these cables is documented in the Task 4.4 NFPA 805 and Fire PRA Circuit Analysis Report (SCE&G TR07800-009).

PRA RAI 41 Finding in Attachment V, Table V-1 8 for supporting requirement CS-B1 -01 identifies open items in Attachment D of a licensee report need to be addressed. The disposition does not state that these open items are addressed. Clarify the disposition of this item.

SCE&G Response

Document Control Desk RC-12-0142 Page 45 of 87 Attachment L (Resolution of Open Items Identified in TR07800-009 Attachment A) of Attachment A (Common Power Supply & Common Enclosure Associated Circuits Review) of TR07800-009 (NFPA 805 AND FIRE PRA CIRCUIT ANALYSIS REPORT, Task 4.4) documents the resolution of the open items cited in this F&O, plus additional open items subsequently identified during the closure process. The result was that all of the issues were resolved such that no breaker coordination issues will exist in the post-transition plant. No changes to the FPRA were required. Attachment V, Table V-18 for CS-B1-01 will be revised to identify resolution of the open items is addressed in Attachment L of Attachment A in TR07800-009.

PRA RAI 42 For F&O DA-02, the acceptable basis for using plant data only is that the quantity of data is sufficient to characterize the parameter value and its uncertainty. Provide confirmation that this is the basis for accepting plant data as a sole source-and ignoring generic evidence, as required to conform to capability category II of the standard. Illustrate that this "does not impact the development of a FPRA" by performing a sensitivity analysis to-demonstrate that the 3 cases identified by the peer review as having significant differences would not impact the FPRA results.

SCE&G Response VCSNS PRA guideline PSA-05, "Data Update Guideline with Emphasis on Bayesian Updating,"

uses the quantity of data as the basis in determining that plant specific data is sufficient to characterize the parameter value and its uncertainty. A sensitivity analysis was run in which the at power internal events model was re-quantified with the parameter values for the three cases replaced with Bayesian updated values. As an example, the case with the largest CDF increase was the Service Water Pump Failure to Run. The plant specific data for this case had a very large amount of data, consisting of over 390,000 run hours. The change in CDF for the 3 cases was an increase of 2.7E-07, not significant compared to the base CDF of 1.6E-05.

PRA RAI 43 F&O DA-03 identifies that "non-fatal" common cause failures were added to the model, but does not affect the FPRA. If there are more than two components, then a fire-induced failure of one component combined with a non-fatal common cause failure (CCF) of the remaining components would now be a possible fire failure scenario. Describe the basis as to why this resolution does not impact the FPRA.

SCE&G Response F&O DA The original F&O concerned missing non-fatal combinations. Common Cause is modeled at the component level. Both fatal and non-fatal combinations are captured. These are no longer missed in the modeling so the original problem does not impact the fire PRA because it no longer exists.

Document Control Desk Attachment 5 RC-12-0142 Page 46 of 87 For the case of a fire event failing one component, the common cause for one or more redundant components would be the same fire. This is captured in the PRA fire analysis.

PRA RAI 44 For F&O DA-08, describe whether the first issue was resolved via a RA or data screening.

Provide a description of the basis for the resolution. For the second issue, provide justification that a common suction path that may lead to steam binding, air binding or debris clogging of both pumps does not exist. Provide justification for the CCF data used to address the second issue. For the third issue, data for this failure mode is in the Idaho National Engineering and Environmental Laboratory (INEEL) CCF database. Provide justification for why it was not used. For the fourth issue, describe whether CCF for emergency diesel generator (EDG) fuel transfer pumps is included in the station blackout SBO model.

SCE&G Response F&O DA VCSNS addressed issue 1 by accounting for the fact (as noted in the F&O) that the beta factor from industry data is conservatively high (given actual total losses of CCW), while maintaining the mission time at its present value. This was accomplished by introducing a coupling factor that accounts for the portion of failures that are postulated to fail both the running and standby pumps. Use of this "severity factor' (as recommended in the F&O) provides a better model for the Loss of CCW initiating event common mode failures.

Of 225 events for pump failures from industry data in CCFWIN, none were attributed to CCW pumps. The coupling factor looked at lethal failures of pumps, excluding high pressure systems. The industry data showed 15 lethal shocks in 113 events. This alpha (coupling factor) was incorporated into the model for normally running and standby CCW pumps.

To address issue 2, VCSNS utilized data screening to account for the possibility of common failure modes between the motor driven and steam driven pumps. By introducing a coupling factor based on the results of this screening, the VCSNS model now accounts for the possibility of both random and common cause failures of the three Emergency Feedwater pumps. The Westinghouse report that describes this listed NUREG/CR-5485 as a reference.

The recommendation to consider and account for CCF events between the motor and turbine driven pumps is given in NUREG/CR-4780. NUREG/CR-5485 describes the modeling of asymmetrical COF events, including as an example, two motor driven pumps and one turbine driven pump. The fault tree incorporates the symmetrical and nonsymmetrical causes for the motor and turbine pumps, per NUREG/CR-4780, and supplemented by NUREG/CR-5485.

To address issue 3, VCSNS has added basic events for both the random failure probability of the screens failing as well as the common cause modeling for screen failures. This modeling change was made in both the normal Service Water model and the Loss of Service Water Initiator, utilizing INEEL data.

For issue 4, the fuel oil transfer system including common cause failure was added to the model

Document Control Desk RC-12-0142 Page 47 of 87 PRA RAI 45 For F&O HR-02, describe whether miscalibration common cause events were added to the model.

SCE&G Response F&O HR Mis-calibration common cause events were added to the model, but they were added as common cause events rather than as HRA events.

PRA RAI 46 For F&O HR-03, provide the justification and reference for the time window used in the HRA calculation for feed and bleed actions.

SCE&G Response F&O HR Different feed and bleed operator actions are now used for different initiating events.

For LOCA and Steam Generator Tube Rupture, 15 minutes is used. This was based on a thermal hydraulic computer program.

Loss of main feedwater, secondary breaks and loss of instrument air use 30 minutes.

Thermal hydraulic runs (Westinghouse document LTR-ESI-02-105, task 36, and Westinghouse Calc. Note CN-RRA-99-032, Rev. 1 Appendix I) determined that 45 minutes was available for general transients. However, based on Westinghouse Document DAR-OA-06-4 this was shortened to 30 minutes to account for lower initial steam generator level on loss of feedwater and related initiating events.

PRA RAI 47 For F&O HR-05, provide a description of the revision made to the HRA calculation to address these issues, including the basis for assigning the level of dependency.

SCE&G Response A review of dependence for post-initiator HRA events was performed. Levels of dependence were re-assessed. Event combinations were assessed on a case by case basis.

The notes explaining levels of dependence are shown below:

Document Control Desk RC-12-0142 Page 48 of 87 1 Loss of one CCW train triggers an alarm response. If the opposite train of SW is lost, all CCW is not lost for at least 30 minutes due to Service Water heat up. Loss of the other train of SW also starts out as an alarm response. AOP 118.01 does not need to be entered right away. The AOP provides back-up actions for the alarm response actions. Because of the separation in time between loss of CCW trains, this is judged to have low dependency. For loss of offsite power case, the EDG could probably run for a while on fire service so here too, starting Service Water and CCW pumps is assigned low dependence.

2 AADRXSW1 DA--HE or ABDRXSW1 DB--HE are not performed in the control room and are in response to an alarm that should be separated in time due to switchgear room heat-up. Also, 80 minutes (after the assumed alarm time) is available for this action. These have CD with each other and ZD with other events.

3 OAAC is performed outside of the control room upon a loss of CCW. It must be accomplished within a short time frame. Because of the ability to run charging pumps for a while without cooling, this is now assigned "low dependence" for SW and CCW HE cases.

For the case where it occurs ONLY with the swap of the non essential header valves (OACCW...) or the trip of RCPs (RCPTHE) OAAC has zero dependence since the manual action to start CCW did not fail.

4 Combination contains a recovery event. Actual dependence is assumed to be negligible because these are performed outside of the CR in response to alarms and conditions by personnel performing testing or maintenance. Also, where-dependence was determined for recovery events, it is modeled by "marking as recovery" in the recovery rule file. See note 21 about recovery NVERT1 and note 19 about recovering TDEFW start failure.

5 When an initiating event other than %LCC or-%LSW is in progress the dependence between the actions to restore CCW and Service Water are assumed to have medium dependence.

This is higher than the low dependence assumed for the initiating case, see Note I for example.

6 Isolating letdown has some dependence on the other actions for loss of CCW, but the problems occur out in time and the presence of additional alarms mean that this only has low dependence. It has no dependence on depressurization and other actions that do not involve loss of CCW.Also, no dependence was assigned for isolating letdown and aligning non essential header valves or alternate cooling to the charging pumps or tripping the reactor coolant pumps since successful start of a CCW pump must have occured in these cutsets.

7 Failure to depressurize the secondary side occurs late in a consequential seal LOCA scenario. Secondary depressurization events assumed to have zero dependence with other HEPs.

8 For the scenario of needing to start a charging pump on loss of CCW, cooling one train of charging pump cooling has succeeded. The intervening success makes starting the charging pump a ZD.

9 Intervening success occurs before this so in has ZD. In some cases this means EFW actions were a success, but EFW failed for some other reason.

10 Conditional dependence is already accounted for in the quantification of this action so it has ZD for additional dependence.

11 Tripping the reactor coolant pump must occur within 10 minutes of a loss of CCW to the bearing coolers.Beacuse of the extra time to run charging pumps after loss of CCW, this has only LD for starting CCW and SW pumps and ZD for other actions.

Document Control Desk RC-12-0142 Page 49 of 87 12 HEP events to open LCV 115 valves, 8801 A & B valves and 8701 and 8708 are performed in the same step of the same procedure and so are completely dependent. These are assumed to be independent of HEPs not called for in EOP1.0 Attachment 3.

14 Events for failure to initially react to an ATWS in time are assumed to be independent of other events that have a much longer time frame. ATWS events are assumed to be completely dependent upon each other.

15 Complete dependence is assumed for the actions to start the A, B and C charging pumps or the A,B and C CCW or SW pumps.

17 Actions to restore instrument air compressors (for %LIA) are assumed not to have a dependence with other HEPs that occur much later. Starting the standby air compressor and the diesel air compressor are both called for in AOP 220.1. However, since the standby is started in step 1 from the CR and the diesel compressor is started locally per step 2 alternate action, these are assessed as medium dependence. Starting the supplemental compressor has complete dependence on ensuring the standby compressor started.Supplemental and diesel have complete dependence.

18 The action to swap the CCW non-essential header to the other train, OACCWSWITCH, is assumed to have a low dependence with loss of service water or loss of CCW HEPs, zero dependence with of failure to start charging pumps (see note 8) and no dependence with other HEPs 19 Recovery of EFW TD Pump start failure is assumed to be independent of events that lead to an RCP seal LOCA because this occurs somewhat further out in time.

20 Recovery of an EDG failing to start is only credited on %LSP cutsets that do not go to SBO.

There should be sufficient time to takethis action from the initiating event because the action recovers a train of mitigating equipment. Zero dependence on other actions is assumed.

Also, this is quantified at 0.5 so dependence doesn't make much difference in this case.

21 BCPM--XPP39CHE needs to be "Marked as a recovery for NVERT1 (corrected in 4E model) 22 Entering EOP 15.0 on loss of EFW should come quickly, but there should also be some time available due to time for SG dryout. MD assumed between D-TRANOPSTRTHE and feed and bleed. ZD assumed with D-TRANOPSTRTHE or feed and bleed and other actions except EFW for ATWS, see note 32.

24 Entering recirculation (OAR4, OAR5) is from RWST level and is on the Reference page. Due to this explicit guidance it is assumed to be independent of depressurization. Also independence is assumed between OAR4, OAR5 and other actions.

25 Event OAEFC has to occur at a specific time in the SBO (4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />). Other events occur once offsite power is restored. Also OAEFC is a very specific Severe Accident action. It is assumed to be separated in time and have no dependence to other actions.

26 SGTR is slow moving. SGTR actions and non-SGTR specific HEPs are assumed to be separate in time and have no dependence. SGTR recoveries are assumed to have low dependence to one another. Refilling the RWST and initiating normal RHR are assumed to have no dependence on other events since they should be separated by hours in time.

27 Entering high head hot leg recirculation (OAL1) is required after 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> following a large LOCA. It is assumed that this has zero dependence with other actions because of the long time frame and because some other actions must have succeed for this to be important.

28 Failure to terminate SI during SGTR (OATC) already has dependence accounted for in the calculation of the HEP. However, LD is applied to OAT1 or OAC and other SGTR actions.

Document Control Desk RC-12-0142 Page 50 of 87 29 Restoring instrument air to the RB is performed via monitoring of critical safety functions (EOP 15). This is assumed to be independent because it is monitored by someone other than the CRS. This is assumed to be independent of closing a PORV block valve (OABV).

30 LP recirculation HEPs are to be eliminated because LP recirculation swap is now automatic.

32 Failure to initiate feed and bleed for an ATWS is assumed to be medium dependent on the action to initiate EFW on an ATWS. Feed and bleed is a CSF action, otherwise these would be highly dependent. These are assumed to be independent of other actions.

33 Complete dependence is assumed for CS valve manual actuation events.

34 The need to isolate containment is assumed to be far enough removed in time so that event O--CNTMISOL-HE is independent of events not associated with manual actuation of components that should have actuated automatically.

35 The OAESF recovery events are performed inside the control room in response to EOP 1.0 Attachment 3. Dependency is assumed to be zero for these -because EOP 1.0 is separated in time from subsequent EOPs and because Attachment 3 is extremely well practiced and understood. An exception is manual containment isolation which is also performed per attachment 3. The manual Cl dependency is handled in the recovery rule file.

36 OA_AAC_SBO is separated in time from other events. The only way equipment recovery can happen after this failure is for offsite power to be recovered subsequent to one hour. The earliest time this is asked in the model is 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />.

40 Steam generator isolation for secondary side break is isolated in time and is in different procedures than actions to start charging pumps or respond to loss of CCW. Therefore, UC-FTIFLTSGCHE is assigned zero dependence on other actions.

41 There are 45 minutes available to restore electric power to a PORV block valve (WOABENRGZMOVHE). This is separated in time from other events so it has zero dependence.

NM These combinations- are not currently modeled. Some of these will be included in future updates.

N/A There are no dependent events in these combinations.

PRA RAI 48 For F&O HR-06, characterize the human actions reviewed with the operators and summarize the results of the interviews.

SCE&G Response The initiating events and human actions reviewed with operators were those that were judged to be possibly influenced by a "full plant level perspective" (as mentioned in the F&O). Human actions reviewed included failure of alternate cooling to the charging pumps and response to switchgear room high temperature. Initiators reviewed included loss of a DC bus, loss of component cooling water, loss of service water and pipe rupture in component cooling water.

The result was a modeled dependency between response to switchgear room high temperature and the loss of a DC bus. The fire PRA includes this modeling.

Document Control Desk Attachment 5 RC-12-0142 Page 51 of 87 PRA RAI 49 For F&O HR-08, provide a description of the time-reliability models that were performed for human failure events (HFEs) with short time windows. Describe whether this is the same model used in the FPRA HRA. Provide a basis for why only one HRA probability had to be updated.

SCE&G Response F&O HR-08 was mainly concerned with events following an ATWS. Since they have to be completed in a very short time frame, but were assigned low probabilities based on usual HRA methods, the reviewer thought that they could benefit from using a time reliability model. These were checked with the "HCR/ORE" method that uses "skill", "rule" and "knowledge" response curves. Using the ratio of time available to time required as a parameter, the HEP is read from the intersection on the appropriate curve. When checked with the HCR/ORE method, only one ATWS related HEP produced a slightly greater probability with the curve, so only one value was changed.

Since the only basic event identified for this F&O is manual trip following an ATWS and-ATWS is not included in the fire PRA model, this issue does not affect the fire PRA model.

The HCR/ORE model was used for time reliability. This model uses "Skill", "Rule" and "Knowledge" based curves or formulas. The ratio of time available to time required is read from a graph of the curves or inserted into the appropriate formula to get an HEP. This F&O only affected ATWS related actions. ATWS was not a part of the fire model so there is no sensitivity to this.

PRA RAI 50 For F&O HR-01-2007, SR HR-01 requires that once the overall HRA has been completed, the plant should perform a review of their HEPs for internal consistency with respect to scenario, context, procedures and timing. Describe when this requirement will be incorporated into plant guidance. If it will not be incorporated into plant guidance, provide a basis for the statement, "This review is performed at each HRA update, although it is not currently a specific requirement in the guideline."

SCE&G Response CR-12-02238 has been written to add performing an internal consistency review to the PRA guideline for human reliability (PSA-04).This action will be complete by the end of 2012.

Document Control Desk RC-12-0142 Page 52 of 87 PRA RAI 51 For F&O QU-04, provide a description of the changes made to the PRA guidance to ensure multiple operator action strings are evaluated for dependence after each change in the PRA HRA.

SCE&G Response F&O QU The following is an excerpt from PSA-04:

5.6 Evaluate Multiple Operator Action Strings - Task 4 Once the cut sets are available, they should be reviewed to identify sequences that, but for low human error rates in recovery, would have been dominant contributors to core damage frequency.

Setting human action basic event probabilities to one, solving the model and reviewing the resulting cut sets identifies multiple operator action strings. The review of cut sets for dependencies is a necessary step for performing major updates of the PRA model.

The PRA modeling or the recovery rule file is changed to ensure that the resulting strings of human error events do not have unaccounted for dependencies.

PRA RAI 52 For F&O QU-06, provide a basis for the conclusion that performing these updates after each major revision does not have a negative impact on the FPRA.

SCE&G Response QU-06 discusses updating the sensitivity analysis and parametric uncertainty analysis. The PRA structure, quantification and results are not affected by updating the sensitivity and uncertainty studies. Performing these updates cannot have a negative impact on the fire PRA.

PRA RAI 53 For F&O QU-07, the disposition does not address the documentation portion. Describe what actions have been taken to ensure that insights about the contributors to risk, key plant features that impact the results, any unique or specific modeling approaches that influence the results, and results of parametric uncertainty are included in this and future results summaries.

Document Control Desk RC-12-0142 Page 53 of 87 SCE&G Response QU Insights associated with PRA model revisions are listed in the calculation documenting the revision as part of the normal model revision process.

PRA RAI 54 For F&O L2-02, it is not clear from the disposition whether the PRA model includes early containment over pressure failures as discussed in NUREG/CR-6595, "An Approach for Estimating the Frequencies of Various Containment Failure Modes and Bypass Events".

Describe what is so unique about the containment that this failure mode can be discounted. Provide a summary discussion of the plant-specific basis as to why this failure mode cannot occur (or is -very unlikely to occur), and include a discussion of fire-induced core damage sequences and their relation to this containment failure mode.

SCE&G Response F&O L2 See calc notes CN-RRA-02-42, "Documentation of LERF for V. C. Summer Nuclear Station Probabilistic Safety Assessment," ,CN-RRA-02-51, "V. C. Summer Nuclear Station Severe Accident Phenomena," and ,CN-RRA-02-80, "V. C. Summer Nuclear Station Level 2 Upgrade." Fire induced core damage sequences fail equipment in a manner similar to other core damage sequences. These calc notes also apply for fire.

PRA RAI 55 For F&O HR-04-GA, summarize the discussions referenced in the response that 1)

Discuss the specific factors considered when evaluating the dependency between actions; and, 2) Indicate the basis for assigning the dependency levels for the second and subsequent actions in a set, especially for the LD and ZD dependencies.

SCE&G Response The factors considered when evaluating dependence between actions (from VCSNS Design Guide PSA-04, Human Reliability Analysis) are:

1. Two operator action failures separated in time by an essential successful action are regarded as independent (ZD).

2. The time available for most operator actions varies from minutes to hours. The degree of dependence between OAs is varied accordingly. The EPRI HRA calculator tool contains suggested levels of dependence based on timing.

Document Control Desk RC-12-0142 Page 54 of 87

3. Response to memorized IMMEDIATE ACTION steps are independent (ZD) of actions taken later in the procedure. Similarly, the IMMEDIATE ACTION steps are independent (ZD) if they are performed by different crew members.

4. For cases where an operator action failure significantly reduces the time window for a subsequent operator action, high dependence would be assessed for the second operator action.

For cases where an operator action failure guarantees failure of a subsequent operator action, complete dependence would be assessed.

PRA RAI 56 For F&O DA-01 -GA, provide a markup of the revised data update guideline so that the reviewer may understand the redefined process and rules used.

SCE&G Response In response to F&O DA-01-GA-, PSA-08, "PSA Model Updates," was revised to-incorporate Fire PRA.

The Data Update Guideline (PSA-05) was not revised as a result of DA-02. The items identified are more appropriately updated using plant-specific rather than Bayesian updating. The F&O describes the difficulty of using a formula to use plant specific data vs. Bayesian. The difficulty is justified for obtaining a more accurate value.

PRA RAI 58 For F&O QU-01 -GA, describe whether there are any cut sets containing multiple maintenance actions that were inappropriate. If so, describe what action was taken.

SCE&G Response The review of cut sets is described in the calculation for PRA updates. The top 50 cut sets (or total if less than 50 are available) for each initiating event were reviewed for revision 6.

Inappropriate combinations of multiple maintenance actions are excluded by mutual exclusive logic in the VCSNS PRA fault tree.

PRA RAI 61 For F&O QU-05-GA, describe whether the definition of "significant" in your quantification guidelines is consistent with Section 2 of the Standard. Describe whether the definition of "significant" impacted guidance for documenting assumptions and sources of

Document Control Desk Attachment 5 RC-12-0142 Page 55 of 87 uncertainty as well as the review of significant cutsets and accident sequences. If so, describe how. If not, describe why not.

SCE&G Response "Significant Accident Progression", "Sequence, Significant Accident Sequence", "Significant Basic Event" and "Significant Cutset" are all defined in section 4 of guideline PSA-03. These are consistent with section 2 of the ASME standard. These definitions did not change the documentation of assumptions or uncertainty or the review of significant cutsets or accident sequences. To the extent possible, words other than significant are used in the documentation in order to avoid running afoul of strict definitions. The definitions in PSA-03 are based on the guidance in the ASME Standard.

PRA RAI 62 Finding in Attachment U, for supporting requirement AS-01 identifies several deficiencies for the ISLOCA analysis centering around giving credit for successful prevention of core damage and large releases. Specifically, the F&O identified 1) crediting success of high pressure injection and recirculation given depressurization and makeup long term from an external source, 2) probabilistic treatment of low pressure piping failures with inadequate documentation, 3) non-piping failure modes not significant, 4) flooding impacts not considered, 5) open-ended makeup without sump recirculation,_6) credit of recirculation and containment cooling. The disposition of the finding states only that large pipe breaks were-added to the analysis as going directly to core damage. This does not appear to address the above issues. Provide additional information that fully addresses the scope of the F&O.

SCE&G Response After the Westinghouse peer review that generated the F&O for this supporting requirement, SCE&G contracted with Westinghouse to provide a modeling approach that addresses these issues. The modeling in the resulting Calc Note, CN-RRA-02-81, "V.C. Summer Nuclear Station ISLOCA Modeling and Success Criteria," was incorporated into the VCSNS PRA model. A model with this modified ISLOCA approach was used as the basis for the Fire PRA. An excerpt from CN-RRA-02-81 is provided below.

Evaluations, Analysis, Detailed Calculations and Results ISLOCA Frequency - Quantification of Low Pressure Piping Exposure to High Pressure The ISLOCA portion of the calculation note is intended to supersede the ISLOCA calculations documented in CN-RRA-02-53 for VCSNS ISLOCA, reference 19 sections 2.0 and 6.3.4 and the ISLOCA calculations in the VCSNS IPE Initiating Event Notebook, reference 2.

A limited survey of other PRAs has been made to evaluate the approach taken in the VCS IPE for ISLOCA. This has been done as part of this effort as well as a review of more recent

Document Control Desk RC-12-0142 Page 56 of 87 NUREGS covering this topic namely references 12 and 14. Other plants that have the RHR pumps and system components located inside containment consider other component failure paths (heat exchanger and other non-pipe components). Much of the RHR system at VCSNS is outside containment. The VCSNS ISLOCA analysis considers the ISLOCA paths specific to the VCSNS plant layout. The conclusion is that the items captured in the peer review cover the items that need to be addressed further in the VCSNS ISLOCA evaluation. Section 2 and the following sections document resolution of these peer review comments. Much of the RHR system at VCSNS is outside containment.

ISLOCA through the Seal Water Return and Excess Letdown Line Gross amounts of primary system coolant could potentially be lost outside of containment through the Reactor Coolant Pump (RCP) no. 1 seal water return and excess letdown line. This process line is part of the Chemical and Volume Control System, described in Reference 3b.

Reference 2, Figure 2.1 (and shown in Appendix F) shows the potential ISLOCA flowpaths through the seal water return and excess letdown line.

Two different events could expose the seal water return and excess letdown line to RCS pressure. First, one of the three RCP no. 1 seals could develop excessive leakage. Second, the valves controlling RCS excess letdown flow could develop excessive leakage. However, an ISLOCA through a RCP no. 1 seal is much more probable than an ISLOCA through the excess letdown line for the following two reasons. First, the excess letdown line isolation and throttle valves are normally closed (XVG-8154, XVG-8153, and XCV-1137) while the no. 1 seal-water return lines directly downstream of the RCPs are normally open (XVG-8141A, B, & C). Second, the isolation valves on the excess letdown line all fail closed while the isolation valves on the no.

1 seal water return lines fail open. Therefore, only the ISLOCA frequency through the RCP no. 1 seals will be quantified here.

The seal water return line is equipped with a flow meter (FE-1 56B, FE-1 55B, or FE-1 54B) and a high flow alarm. The probability that the high flow alarm fails to sound in the control room is assumed to be the frequency that a pressure switch fails to open in Reference 4, 2.OE-07 per hour. From Reference 5 the high flow alarm is tested at least every other refueling outage (3 years = 26280 hours).

Upon receipt of a seal water return line high flow alarm, precaution 5 of SOP-1 02 (Reference 6) instructs the operators to isolate the seal water return line from the affected RCP. Reference 7 identifies the probability that the operators will not attempt to isolate the appropriate seal water return as Seal Water Return Human Error (SWRHE) as 1.55E-02 per demand.

If the operator does not attempt to isolate the seal return line, this LOCA would lead to low pressurizer level and pressure, which would generate an "S" signal and a Containment Isolation Phase A ("T" signal). This "T" signal would then automatically close the two motor operated isolation valves installed on containment penetration 410. Given low pressurizer pressure, the probability of "T" signal failure is taken as the Representative Safeguards Actuation Signal probability reported in Reference 8, 4.60E-04 per demand. However it is generated, this isolation signal will close motor operated valves XVG-8112 and XVG-8100.

Document Control Desk Attachment 5 RC-12-0142 Page 57 of 87 This calculation can be seen in the worksheets of Appendix B and the resulting frequency is summarized below. The calculation presented in the IPE included only catastrophic ruptures of the valves in question. The calculation has been revised to include significant leakage through the valves as a credible failure mode.

ISLOCA through Containment Penetrations between the RCS and Centrifugal Charging Pump Discharge Header Gross amounts of primary system coolant could potentially be lost outside containment through piping on the suction side of the centrifugal charging/safety injection pumps. The centrifugal charging/safety injection pumps are part of both the Chemical and Volume Control System and the High Head Safety Injection System. Reference 3b describes the Chemical and Volume Control System, Reference 3c describes the High Head Safety Injection System, and Reference 2, Figure 2.2 shows the convergence of these flowpaths on the Charging Pump Discharge Header.

The charging pump discharge piping and high head safety injection lines are rated for a pressure of 2500 psig, the RCS charging line is rated for a pressure of 2350 psig, and the RCP seal injection lines are rated for-a pressure of 2340 psig. These process lines are considered potential ISLOCA pathways because the charging pump intake piping is only rated for a maximum pressure of 170 psig. Component misalignment and failure could expose this low pressure piping to RCS pressure, potentially leading to a loss of primary system coolant outside of containment.

During normal power operation the charging pump discharge header is maintained at a higher pressure than the RCS to accommodate charging and seal injection flow; Under this condition any valve misalignment or failure in the lines between the charging pump discharge header and the RCS would result in coolant flow into containment rather than out of containment during normal operation. Therefore, any ISLOCA through these eight containment penetrations is assumed to begin with the exposure of the charging pump intake piping to the pressure of the

-pump discharge header.

The operating charging pump is rotated weekly during power operation following System Operating Procedure 102 Section III.E. In step III.E.2.3 of SOP-102, Revision 12, the operator stops the operating charging pump. If the check valve on the discharge side of this stopped charging pump fails to close (XVC8481A, B, or C), the intake piping of the stopped charging pump will be exposed to the pressure of charging pump discharge header. It is assumed that the charging pump which was just stopped can not be restarted due to backflow of water through it.

The testing interval of the check valves downstream of the charging pumps is assumed to be 3 weeks (the check valve was closed from an open state when the charging pump was last stopped).

A charging pump suction line can also be exposed to the discharge header pressure through the rupture of one of the check valves on the discharge line of the two idle charging pumps (XVC-8481A, B, or C).

Document Control Desk Attachment 5 RC-12-0142 Page 58 of 87 Given the exposure of low pressure CVCS piping to the Charging Pump Discharge Header pressure, there are three types of flowpaths through which primary coolant can reach the charging pump discharge header, the normal charging line, the RCP seal water injection line, and the high head safety injection lines. However, only the ISLOCA frequency through the normal charging line is quantified here because the RCP seal water injection line was determined to be an insignificant ISLOCA flowpath in Reference 2 Section 2.2 (included in Appendix F) and an ISLOCA through the normal charging line is much more probable than an ISLOCA through either the high head safety injection (HHSI) lines for the reasons discussed below.

An ISLOCA through the charging line is much more probable than through the HHSI for two reasons. First, the check valves in the high head safety injection lines are closed when the RCS is pressurized. Technical Specification 4.4.6.2.2 insures the leakage of these closed check valves is within acceptable limits. However, the check valves in the normal charging line are normally open during power operation. Second, the motor operated valves between the charging pump discharge -header and the high head safety injection lines are closed during power operation. Technical Specification 4.5.2 directs the operators to check the position of these closed valves at least once each 31 days. However, the motor and air operated valves between the charging pump discharge header and the charging line are normally open during power operation. For these two reasons the frequency of an unisolatable ISLOCA through the high head safety injection lines is assumed to be insignificant in comparison to the frequency through the normal charging line.

The normal charging line is designed with three check valves in series to prevent backflow of primary coolant if the charging pump discharge header loses pressure (XVC-8378, XVC-8347, and XVC-8381 shown in Reference 2, Figure 2.4, included in Appendix F). Check Valve XVC-8381 is tested for leakage in the reverse direction every refueling outage per STP-1 15.006, Reference 9. However, no regularly scheduled surveillance verifies the ability of valves XVC-8378 and XVC-8347 to close given a reversal in pipe flow. Therefore, a conservative probability that all three check valves fail to close is taken as the probability that the tested check valve fails to close. Conservatively this accounts for CCF of the check valves.

If gross amounts of primary coolant were lost from the RCS through the charging line, low pressurizer level or pressure should generate a reactor trip and an "S" signal. If for some reason the "S" signal is not automatically generated, the operators would manually generate an "S" signal per step 5 of EOP-1.0, "REACTOR TRIP/SAFETY INJECTION ACTUATION", Reference

6. Two MOVs in series on the charging line are designed to close on an "S" signal, and this would isolate an ISLOCA through the charging line. These two MOVs are XVG-8107 and XVG-8108, and they are shown in Reference 2, Figure 2.2, included in Appendix F. If for some reason these MOVs do not close on the "S" signal, the operators would attempt to close them manually per step 18 of EOP-1.0, "REACTOR TRIP/SAFETY INJECTION ACTUATION",

Reference 6. Since there is considerable redundancy in the generation of the charging line isolation signal, the probability that the charging line is not isolated in this scenario is assumed to be no greater than the probability that two MOVs fail to actuate.

Document Control Desk Attachment 5 RC-12-0142 Page 59 of 87 This calculation can be seen in the worksheets of Appendix B and the frequency is summarized below. The calculation presented in the IPE included only ruptures of the valves in question.

The calculation has been revised to include leakage through the valves as a credible failure mode.

ISLOCA through RHR Discharge Lines Gross amounts of primary system coolant could potentially be lost outside containment through the discharge lines of the RHR (Low Head Safety Injection) System. Reference 3 describes the RHR system, and this system is shown in Reference 2, Figure 2.3, included in Appendix F.

The piping and valves on the RHR discharge lines from the RCS up to and including the outboard containment isolation MOVs are rated for a maximum pressure of 2485 psig, the piping and valves from these MOVs up to and including the RHR pumps are rated for a maximum pressure of 600 psig, and the piping and valves on the suction side of the RHR pumps is rated for a maximum pressure of 450 psig. This calculation of RHR discharge ISLOCA frequency assumes any exposure of RHR piping outside of the containment isolation MOVs to RCS pressures constitutes initiation of an ISLOCA event.

The failure mode of the normally closed valves isolating the RHR discharge from the RCS is catastrophic rupture or significant leakage. Technical Specification 4.5.2 insures that motor operated valve XVG-8889 on the line to the hot legs (shown in Reference 2, Figure 2.3, included in Appendix F) is closed and its power is locked out while the plant is in modes 1 through 3. Technical Specification 4.4.6.2.2 insures closure of the RHR check valves inside containment while the plant is in modes 1 through 4. These check valves are shown in Reference 2, Figure 2.3, included in Appendix F (XVC-8974A & B and XVC-8973A, B, & C) on the lines to the cold legs and in Reference 2, Figure 2.5, included in Appendix F (XVC-8988A &

B and XVC-8993A & B) on the lines to the hot legs.

The calculation of RHR discharge line ISLOCA frequency is broken into two steps, calculation of an ISLOCA through the RHR discharge line to the RCS hot legs and calculation of an ISLOCA through the RHR discharge lines to the RCS cold legs.

The frequency of an ISLOCA through the RHR discharge to the hot legs is calculated as the failure frequency of the normally closed MOV times twice the failure probability of two check valves in series leaking. The factor of two represents the fact that the RHR discharges to two hot legs. The MOV has several failure modes which are considered. These are summed to determine the failure frequency of the MOV. These failure modes are as follows:

" Random Failure of the MOV

" Failure of the MOV due to being left open

" MOV Disk Rupture The dynamic effects of accidental pressurization of the RHR system are not modeled as an important consideration. This is consistent with the IDCOR analysis, Reference 34, and with the

Document Control Desk RC-12-0142 Page 60 of 87 RHR piping and heat exchanger strength analysis provided as part of the Seabrook Station Risk Management and Emergency Planning Study, Reference 30.

A MOV failure mode for the valve to spuriously open is not considered since the valve is closed and the power is locked out per Technical Specification 4.4.6.2.2 discussed above.

In calculating the rupture/leakage of the second check valve given the rupture / leakage of the first valve, three things are assumed:

1) the rupture/leakage of the first valve is equally likely to occur at any time in the testing interval,

2) the rupture/-leakage of the second valve can occur at any time in the testing interval,

3) on the average failure will occur at the midpoint of the testing interval.

Common cause failure of the check valves is not considered since these check valves are exposed to-quite different environments. The check valve closest to the hot legs is exposed to RCS pressure and temperatures constantly. The next check valve in series is protected by the first check valve and sees lower pressure and temperatures. The second check valve is exposed to RCS conditions upon failure of the first check valve but duration of the exposure of the component to these conditions differs significantly from that of the first valve.

The frequency of an ISLOCA through the RHR discharge to the cold legs is calculated as six times the frequency of three normally closed check valves in series significantly leaking/

rupturing open. The factor of six represents the fact that the RHR discharges to three RCS cold legs, and these three RCS discharge lines are-linked by a common header to the two RHR trains. Common cause failure of these check valves is not considered using the same rationale as the hot leg check valves.

This calculation can be seen in the worksheets of Appendix B and the frequency is summarized below. The calculation presented in the IPE included only ruptures of the valves in question.

The calculation has been revised to include significant leakage through the valves as a credible failure mode.

ISLOCA through RHR Suction Lines Gross amounts of primary system coolant could potentially be lost outside containment through the discharge lines of the RHR (Low Head Safety Injection) System. Reference 3d describes the RHR system, and this system is shown in Reference 2, Figure 2.3, included in Appendix F The piping and valves on both RHR suction lines from the RCS up to and including the two isolation MOVs inside containment are rated for 2485 psig. There is no isolation capability between the inboard isolation MOVs and the RHR pumps, and this piping is rated for 450 psig.

Five relief valves connected to the RHR piping provide overpressure protection for the RHR piping. Two 3-inch relief valves (9708A and 8708B) installed on the RHR suction lines are sized to relieve 900 gpm at a set pressure of 450 psig, reference 26, included in Appendix F. These valves relieve to the Pressurizer Relief Tank (PRT). Three %-inch relief valves (8865, 8864A,

Document Control Desk RC-12-0142 Page 61 of 87 and 8864B) are installed on the RHR discharge lines. They discharge to the boron recycle system.

Given overpressurization of one RHR train, one 3-inch and three 3/4-inch relief valves are expected to open. The combined capacity of these four relief valves is approximately equivalent to a 3.3 inch pipe break.

At V. C. Summer Nuclear Station the most credible mechanism for initiation of an RHR suction ISLOCA during power operation is the catastrophic rupture or significant leakage of the closed MOVs isolating the RHR pump suction from the RCS (XVG-8702A & XVG8701A or XVG-8702B

&-XVG-8701 B). Per Technical Specification 3.5.2, V.C. Summer Nuclear Station utilizes a RHR/RCS valve interlock to prevent inadvertent opening the RHR hot leg suction valves while the primary system pressure is greater than or equal to 425 psig. This valve interlock is verified operable at least once every 18 months per Surveillance Test Procedure STP-105.010, which fulfills Technical Specification 4.5.2.d.1. Furthermore, Technical Specification 4.5.2 directs the operators to verify that the RHR suction MOVs are closed every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> when the plant is in modes 1, 2, and 3.

The ISLOCA through the RHR suction lines is calculated for leakage through two trains of two closed Motor Operated Valves in series. The calculation can be seen-in the worksheets of Appendix C and the frequency is summarized below.

RHR Piping Rupture vs. RHR Pump Seal LOCA The ISLOCA initiating event frequency is split between RHR Seal LOCA and Piping Rupture Outside of Containment. The highest contributor to the ISLOCA IE frequency is ISLOCA through RHR Suction Lines. The piping downstream of the inboard isolation MOV is rated for 450 psig.

The piping has pressure relief valves installed inside containment downstream of outboard containment isolation MOV prior to the containment penetration. Two 3-inch relief valves (9708A and 8708B) installed on the RHR suction lines are each sized to relieve 900 gpm at a set pressure of 450 psig, reference 26, included in Appendix F. In addition, three 3/4-inch relief valves (8865, 8864A, and 8864B) are installed on the RHR discharge lines. Given overpressurization of one RHR train, one 3-inch and three 3/4-inch relief valves are expected to open. The combined capacity of these four relief valves is approximately equivalent to a 3.3 inch pipe break. The piping is not challenged until the capacity of all 5 relief valves is exceeded (approx. 1000 gpm) for a leak through the suction piping.

The RHR discharge piping is challenged once the capacity of the three 3/4-inch relief valves (8865, 8864A, and 8864B) are installed on the RHR discharge lines. No credit can be taken for the RHR suction piping relief valves for overpressure events on the discharge piping due to the RHR discharge check valves. The piping is not challenged until the capacity of all 3- relief valves is exceeded (>100 gpm) for a leak through the suction piping.

The RHR piping Fragilities (PWR and BWR) was investigated in Reference 32. It is included in Appendix I with Stainless Steel Material Properties from Reference 32. The piping information and material properties have been summarized in the two tables that follow.

Document Control Desk RC-12-0142 Page 62 of 87 Nominal Schedule D/t (As D/t Pipe Size Built) (Corroded)

(in)

• 10 40 29.5 37.7 12 40 31.4 39.1 14 40 32.0 39.2 Notes:

Pipe size, schedule, D/t information from Reference 32

• All wall thickness reduced-by 0.08 inches for corrosion allowance regardless of pipe size 90% Yield Ultimate Ultimate Strength Material Temp. Strength Strength (ksi)

Type (Deg. F) (Median) (ksi) (Median) 304 SS 400- 74.0 66.6 23.0 304SS 600 70.0 63.0 19.5 304 SS 800 65.5 59.0 16.5 Notes:

Stainless Steel Material Properties from Reference 32 The RHR discharge and suction piping was evaluated using the rationale provided in Reference

31. It states that the mean failure stress for the piping is approximately at 90% of the ultimate stress. The parameters presented in the tables above where used to evaluate the RHR suction and discharge piping. The analysis looks at the parameters at 600 degrees fluid temperature for 10-inch, 12-inch, and 14 inch 304 SS schedule 40 RHR piping. VCSNS has verified that the piping is schedule 40 304 SS grade SA312. The suction piping is primarily of 12-in diameter and smaller with a small section of 14 inch piping. The RHR discharge piping is primarily 10 inch piping and smaller. A brief review of information provided in reference 31 shows that for the same schedule of piping, smaller diameter piping does not have less strength at the same pressure and temperature, so a this review was limited to the larger diameter pipe.

The internal piping pressures at 600 deg. F are calculated for a hoop stress pressure equal to 90% of the ultimate stress. These values are shown in the table that follows.

Document Control Desk RC-12-0142 Page 63 of 87 Nominal Schedule D/t (As Hoop Internal D/t Hoop Internal Pipe Built) Stress Pressure (Corroded Stress Pressure Size (in) (600F, at Mean )* (600F, at Mean 1000psi) Failure 1000psi) Failure (ksi) Strength Strength (psi)** (psi) 10 40 29.5 14.7 4278 37.72 18.9 3340 12 40 31.4 15.7 4012 39.11 19.6 3222 14 40 32.0 16.0 3938 39.2 20 3218 Notes:

Pipe size, schedule, D/t information from Reference 32

• 'All wall thickness reduced by 0.08 inches for corrosion allowance regardless of pipe size

• • Internal Pressure calculated condition where Hoop Stress = 90% x Ultimate Strength (psi), Reference 31 Ultimate Stress value used for 304 SS at 600F Piping Hoop Stress is calculated using the following formula:

• • Pressure that Hoop Stress = 90% x Ultimate Strength (psi), Reference 31. Ultimate Stress value used for 304 SS at 600F S= PD/2t Where S = Hoop Stress (psi)

P = internal pressure in piping (psi)

D = Outside diameter of Piping (inches) t = Wall thickness (inches)

The Internal Pressure at Mean Failure Strength is calculated by rearranging the piping hoop stress formula as follows:

P= S * (2t) / D The value for stress input for S is 90% of the Ultimate Strength for 304 SS. The internal pressures are calculated for both as built piping and for piping with a corrosion allowance of 0.08 inches as used in Reference 31. This corrosion allowance is also conservative.

The calculations show for all the RHR discharge and suction piping considered in the ISLOCA case, the calculated system pressure at mean value for failure stress is greater than the RCS system pressure. A piping failure fraction of .01 is selected for ISLOCA resulting from pipe breaks outside containment. It is felt that this failure probability conservatively covers the possibility of piping component failures. This failure probability is consistent with the value used in reference 30 (6E-03), for piping components (valves, heat exchangers, flanges, etc). The remaining fraction of .99 is assigned to ISLOCA though the RHR pump seals. The discussion from Reference 30 supporting the selection of a conditional failure probability of 0.01 for pipe breaks outside containment is included in Appendix K.

Document Control Desk Attachment 5 RC-12-0142 Page 64 of 87 ISLOCA Initiating Event Frequency Sensitivity to Use of Generic Mean Failure Data The sensitivity of the ISLOCA Initiating Event Frequency was reviewed to evaluate the use of generic mean failure data for Check Valves and Motor Operated Valves in applications with these components in series. This effort is in response to the first issue in F&0 1 E-06, the quantification of the ISLOCA frequency taking into account the variance on the mean failure rates.

The spreadsheets developed for the ISLOCA Frequency calculations described in the previous sections were modified-for use of higher failure rates defined by taking into account the variance for the number-of independent events in the calculation cutset using the same failure rate. The failure frequencies were replaced-with values calculated using statistical error propagation (or the delta method) described in Reference 10 section 11.5.4.2. To facilitate the analysis, error factors were assigned to the mean failure probabilities assigned using rationale provided in references 13 and 27 and included in Appendix C.

As outlined in the calculations performed on the Excel worksheet "VCSNS ISLOCA Sensitivity" included in Appendix B, the following formulas from Reference 10 were used for the failure rates for which-a number of independent events were associated in the same calculation cutset:

" For 2 components in series: (P*P) = (P**2 + V)

• For 3 components in series (P**3) (P**3 + 3*P*V)

Where P is the component failure rate (or probability)

V is the variance The sensitivity of the ISLOCA Initiating Event Frequency was reviewed to evaluate the use of generic mean failure data for Check Valves and Motor Operated Valves in applications with these components in series. The results of the analysis are shown in- the table below. The value calculated-can be used-in uncertainty analysis for the VCSNS PRA.

Description ISLOCA Frequency ISLOCA Frequency Calculated with Mean Calculated with Values Values (per Year) from Sensitivity Section (per Year)

F(Seal Water Return ISLOCA) = 1.96E-07 6.71 E-07 F(ISLOCA Through Charging Line) = 1.36E-09 4.94E-09 F(RHR Discharge LOCA) = 7.68E-09 1.57E-05 F(ISLOCA Through RHR Suction Lines) = 1.15E-06 8.16E-06 Total ISLOCA Initiating Event Frequency = 1.36E-06 2.45E-05 The ISLOCA Frequency calculated with mean values is the best estimate and should be used in the VCSNS IPE. The calculation makes use of best data available and the calculated ISLOCA frequency is the best estimate.

Document Control Desk RC-12-0142 Page 65 of 87 Suggested ISLOCA Event Tree Structure and Applicable Success Criteria The current ISLOCA event tree that represents ISLOCA modeling in the current VCSNS PRA Model @3HUP is shown in Figure 1. A suggested ISLOCA event tree structure to support results of section 6.3.1 of this calculation note is shown in Figure 2. The primary difference in the two event trees is the addition of the following top events to model pipe breaks outside of containment and additional activities to bring the plant to a safe stable state:

• The BR, Pipe Break, node is added to the event tree - The BR node is meant to distinguish between pipe breaks caused by overpressurization of the low pressure piping in the normal RHR suction-lines and failures of the RHR pump seal package. This is described in the IPE Plant Response Tree and Success Criteria Notebook. It is estimated in section 6.3.1.5 that the conditional probability of a failure of the normal RHR suction piping is 0.01.

" The LPI, Low Pressure SI, node is added due to the addition of the-BR node. The break is modeled as a 12-inch break and therefore the large LOCA success criteria are applicable.

Success of LPI is required for -the larger break size paths that lead to success; LPI is not important-for small breaks since the RCS pressure remains above the LPI shutoff head until substantial RCS cooldown and depressurization are completed.

" The OSR, Operator Action to Minimize SI, node is added for the BR failure branch only. The branches representing the RHR line break are modeled as a 12 inch break. Therefore, the RCS pressure falls rapidly to a point below the opening setpoint for the relief valves; there is no discharge of RCS inventory back to the containment for possible ECC recirculation.

Thus, operator action to minimize SI flow to prevent emptying the RWST is required for success for these sequences. The success criteria for OSR for the pipe break case was determined to be 40 minutes in Reference 36 based on the flows from the SI pumps and the RWST inventory to maintain core cooling for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This success criteria is applicable for this case. The-success criteria could be extended if RWST refill is modeled for these event tree branches.

For the RHR pump seal LOCA branch, the PRA Success Criteria in Reference 35 concluded that this action is required for success for the RHR pump seal LOCA case and the success criteria forthe operator action was found to be 180 minutes if high pressure SI recirculation was successful. It was further concluded in Reference 17 that RCS cooldown and depressurization would also be required for success. The assessments in Reference 17 further concluded that success could be attained even if OSR failed, provided that RCS cooldown and depressurization was successful. Implicit in the RCS cooldown and depressurization steps is the operator action to control SI according to the plant EOPs for RCS cooldown and depressurization. This, in effect, has the same result as intentionally throttling SI to conserve RWST inventory. Since success of RCS cooldown and depressurization leads to a success path regardless of the outcome of the OSR action, the OSR action can be bypassed for the RHR pump seal LOCA branches.

• SGP node has been extended to all success paths - For an unisolable LOCA outside containment, RCS cooldown and depressurization to near atmospheric pressure is required

Document Control Desk RC-12-0142 Page 66 of 87 to terminate the break flow. This is modeled in the event tree as feeding and dumping steam from the SGs for the cooldown and depressurization of the RCS. Therefore, it encompasses both steam dump availability as well as the operator actions to depressurize the SGs to atmospheric pressure..

An alternate method considered but not modeled in the event tree is the use of the unaffected normal RHR train taking suction from the RCS and discharging to the charging pumps via the high pressure ECC pathway. This pathway is unaffected by the overpressurization of RHR system that lead to-the ISLOCA. However, it is not modeled for several reasons -

" The use of the pathway requires that the unaffected RHR pump not be flooded or otherwise impacted by the RHR pump seal failure on the affected pump,

• The use of the pathway requires the operators to diagnose the affected train, which is not a simple matter, and

" The availability of the unaffected train requires several check valves to function properly to prevent overpressurization of the-other train.

The suggested ISLOCA event tree can be described by the following top nodes branches (the parenthetical 'S' or 'F' indicates success or failure at that event tree node):

Event Tree Path Event Tree Endstate BR (S) / HPI (S) / HPR (S) / EFW (S) / SGP (S) Success BR (S) / HPI (S) / HPR (S) / EFW (S) / SGP (F) Core Damage BR (S) / HPI (S) / HPR (S) / EFW (F) Core Damage BR (S) / HPI (S) / HPR (F) Core Damage BR (S) / HPI (F) Core Damage BR (F) / LPI (S)/ EFW (S)/ OSR (S) / SGP (S) Success BR (F) / LPI (S) / EFW (S) / OSR (S) / SGP (F) Core Damage BR (F) / LPI (S) / EFW_ (S) / OSR (F) Core Damage BR (F) / LPI (S) / EFW (F) Core Damage BR (F) / LPI (F) Core Damage Success Criteria:

BR: No breaks in RHR System piping.

HPI: 1 of 3 CCPs delivering flow to 1 of 3 Cold Legs.

LPI: 1 of 1 LHSI pumps delivering flow to 1 of 3 Cold Legs (one pump disabled by break or leak).

HPR: 1 of 3 CCPs to 1 of 2 RHR pumps delivering flow to 1 of 3 Cold Legs.

EFW 1 EFW Pump available.

Document Control Desk RC-12-0142 Page 67 of 87 OSR: Operator Action to Minimize SI, node is added for the BR failure branch only as discussed in the paragraphs describing this node. The success criteria for OSR for the pipe break case was determined to be 40 minutes in Reference 36 based on the flows from the SI pumps and the RWST inventory to maintain core cooling for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

SGP: 1 SG PORV on 1 SG plus operator action for SG depressurization.

PRA RAI 63 Finding in Attachment U, for supporting requirement SY-07 identifies the CST capacity as not satisfying the 24-hour mission time, but no backup or alternate source being modeled. The disposition of this item only states that the licensee documented why this was not required. A further assessment identified (AS-01-GA of Attachment U-2 of the submittal) that the resolution was not sufficient. However, the disposition of this item does not appear to provide any further basis for this critical assumption. Provide additional justification for this item.

SCE&G Response The CST re-fill issue resolution was documented in DC00300-146, Rev. 1. The original issue was based on the reviewer referencing the minimum required level in the CST (179,848 gallons). However, much more water is expected to be in the tank prior to an event.

The minimum volume (179,848 gallons) in the CST ensures that sufficient feedwater is available to maintain the Reactor Coolant System in Hot Standby for 11 hours1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br /> while steam is vented to atmosphere and a total loss of offsite electrical power is in progress. However, based on the volume of water that could be expected to be in the CST at the beginning of a transient, it is assumed that there is no need to refill the-CST during an SBO or a transient.

The CST level is normally maintained by automatic demineralized water makeup when CST level decreases to 30 feet (404,000 gallons) to fill it to 35 feet (468,500 gallons), at which time demineralized makeup stops.

Thus, there are normally over 400,000 gallons in the CST.

Also, even if CST level is low there are several makeup sources that could be modeled and EFW pumps have an automatic swap to Service Water on low suction pressure. If these alternate sources were modeled, the probability of failure would be negligible compared to already modeled EFW failure modes. It is judged that there is no need to add this complexity to the PRA model to capture the very low probability that the CST would not support a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> mission time.

Document Control Desk RC-12-0142 Page 68 of 87 PRA RAI 64 Disposition in Attachment U, Table U-2 for supporting requirement IE-01 -GA indicates that the finding is not adequately addressed. The ISLOCA frequency calculated using the variance treatment resulted in a factor of twenty higher than the baseline; however, this updated frequency was not included in the model. Confirm that ISLOCA treatment, including frequency calculation and analysis, was revised following the RG 1.200 gap assessment and addresses the weaknesses for the ISLOCA model identified during the peer review.

SCE&G- Response The ISLOCA frequency was updated in 2009 in Attachment 4 to calculation DC00300-150. This is after the 2007 peer review.

The discussion of ISLOCA sensitivity from Attachment 4 to calculation DC00300-150 is as follows:

A variance method like the one used in Reference 4 is a calculation that takes error factors or variances into account. This type of method was called for in the Westinghouse Owner's Group peer review of the VCSNS PRA. A variance method to calculate the ISLOCA frequency results in an ISLOCA frequency of 4.02E-04 per year. This is over an order of magnitude higher than the calculated ISLOCA frequency of 1.1 8E-06/yr.

Table 24 below shows the results of calculating the ISLOCA frequency with the variance method. The main contributors using this method are the RHR suction and discharge lines.

CDF and LERF Impact of ISLOCA Frequency Using the Variance Method 5C Baseline Using ISLOCA Variance  % Change CDF 1.3471E-05 4.9949E-05 270.8%

LERF 2.5402E-07 3.6739E-05 >14,000%

The increase in LERF using the variance method is significant. This increase is dominated by the variance in the rupture failure rate of MOVs. The large variance in this parameter is possibly caused by using very old data (PRA Procedures Guide data from 1985). The error factor for MOV ruptures using this data is 53. If the check valve rupture data from the ALWR data base were used as a surrogate for MOV rupture, the ISLOCA frequency in the sensitivity study would only have increased to 1.57E-06 from 1.1 8E-06 versus the increase to 4.02E-04 using the older data. Based on this discussion, the main source of uncertainty in the ISLOCA analysis is the MOV rupture failure rate. The mean value of 1.1 8E-06/yr for ISLOCA will be used with an error factor of 10.

Document Control Desk RC-12-0142 Page 69 of 87 PRA RAI 65 Finding in Attachment U, Table U-2 for supporting requirement SY-01-GA indicates F&O TH-03 has not been resolved from the original peer review. Table U-1 does not list TH-03 as a finding. Provide the finding associated with TH-03 from the original peer review and the disposition including justification for why room heatup and credit for local operator actions are not modeled in the PRA for electrical equipment rooms IB-63-01 and AB 01. In addition, justify how the equipment in the room is deemed operational within the PRA mission time if the temperature in the room exceeds design conditions.

SCE&G Response The-original peer review F&O (TH-03)-was level "C". The 2007 review against the ASME standard resulted in SY-01-GA. The reviewer for SY-01-GA was satisfied that no operator actions were necessary because temperatures that could be reached within a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period following loss of HVAC in the areas examined could affect the long term reliability of equipment, but not the short term operability of equipment.

The original TH-03 is shown attached (Attachment 2 below).

Document Control Desk RC-12-0142 Page 70 of 87 Attachment 2:

OBSERVATION ID: Th - 03 / Element TH / Sub-element 8 (Related Sub-elements: )

A detailed room heatup calc (CN-CDBT-92-374) was performed for electrical equipment rooms SIB-63-01 and AB-63-01. This showed that it is necessary to recognize loss of room cooling and open room doors in order to prevent room temperature from exceeding an apparent 132 degF acceptance limit ("apparent" rather than "assumed," because it is not explicitly stated as such in the calc). More specifically, the calc showed that for one of the rooms (AB-63-01), if the action were completed within 30 minutes of loss of room cooling, the acceptance temperature would be exceeded (although only slightly) within the PRA mission time (132 deg at 21 hours2.430556e-4 days <br />0.00583 hours <br />3.472222e-5 weeks <br />7.9905e-6 months <br />, 133 deg

-at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />). If the action were not taken until 90 minutes, the room would reach 132 deg at 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> and 135 deg at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The fault tree logic for the equipment in this room includes a requirement for the operator action success. Two comments:

(1) If the 132 degF temperature is an acceptance limit, then the calc really does not demonstrate 24-hour success. Although this may be splitting hairs, it is at least a source of uncertainty in the model.

(2) If credit is taken for the 30-minute operator response, the implication is that operators will always be able to successfully respond. quickly to such an event even if it occurs while they are responding to an accident initiator. This may be optimistic. If credit is only taken for a longer response time, then the acceptance temperature will be exceeded earlier in the mission time. In either case, this is an additional contributor to uncertainty in the model.

Such cases of inconclusive success criteria should be avoided wherever possible.

LEVEL OF SIGNIFICANCE: C The impact of this on PRA results is not likely to be significant, assuming only one train of equipment would be affected in each room.

POSSIBLE RESOLUTION Consider either re-examining the room heatup analysis to determine if there is excess conservatism such that success could be demonstrated for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, or removing credit for the action. Consider reviewing the HRA calc to ensure that it addresses the possibility that the loss of room cooling could occur at a time when the operators are dealing with serious distractions in responding to an accident.

PLANT RESPONSE OR RESOLUTION

Document Control Desk RC-12-0142 Page 71 of 87 PRA RAI 66 FSS-B2-011: It is not clear from the MCR abandonment document what the criteria are for CR abandonment for-main control board (MCB) fires. Secondly, from the MCB scenarios the conditional core damage-probability (CCDP) does not appear correct for the specified number of panels as given in Appendix B of the Fire Risk Quantification Task 14 document. For example, for scenario CB-17.01 MCB 11-10-1, 11-9-1, and 11-8-1, 2,3, and 4 panels are failed respectively, yet-the CCDP from the quantification document specified that the CCDP is largest when only 2 panels are damaged. Also for scenarios CB17.01 MCB 18-18-1, 18-17-1, 18-16-1, and 18-15-1 the CCDP is the same for each scenario even though 1, 2, 3, and 4 panels are damaged respectively. Provide the criteria for CR evacuation for MCB fires, and justify the CCDPs provided for the various number of panels damaged. The- response should take into account a further examination of the CCDPs for MCB fires-than identified in this question to evaluate if the CCDP problem is more extensive than discussed in this question. Provide updated CDF, LERF, ACDF, and ALERF values for MCB scenarios.

SCE&G Response The criteria for MCR abandonment are discussed in the response to PRA RAI 08.

The CCDP issue was an error in the model that was due to the FRANX software treatment of mutually exclusive events. The top logic structure of the fault tree (shown below) includes NOT logic that removes cutsets associated with mutually exclusive events. When the FRANX model sets the failures associated with the fire, it sets all of the failed events to TRUE and then compresses the tree and solves it. If two events are mapped to a scenario that appear in the mutually exclusive logic gate (FIREMEX), the-gate @CDFALLF is set to FALSE. This means that only the MCR abandonment logic is considered for the scenario. This was the root cause of the issue with the odd CCDP results for the main control board. scenarios. The model has been updated to address this issue and the updated results for the MCB scenarios are listed in the table below.

Document Control Desk RC-12-0142 Page 72 of 87 Table 1. Main Control Board Fire Scenarios - Results "Scenario: -IGF CCDP CDF CLERP LERF CB17.01_MCB-10 5.0364E-06 1.77E-01 8.92E-07 6.62E-04 3.33E-09 1

CB17.01_MCB-10-6-1 4.9205E-07 8.76E-01 4.31E-07 1.79E-02 8.80E-09 CB17.01_MCB-10-7-1 4.9205E-07 8.76E-01 4.31E-07 1.79E-02 8.80E-09 CB17.01_MCB-10-8-1 8.2127E-07 8.76E-01 7.19E-07 1.79E-02 1.47E-08 CB17.01_MCB-10-9-1 1.6745E-06 7.68E-01 1.29E-06 2.02E-02 3.38E-08 CB17.01_MCB-I-I-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-11 1.6745E-06 1.77E-01 2.97E-07 6.62E-04 1.11E-09 1

CB17.01 MCB-11 5.0364E-06 2.82E-02 1.42E-07 1.19E-05 6.01E-11 1

CB17.01_MCB-11-8-1 4.9205E-07 8.76E-01 4.31E-07 1.79E-02 8.80E-09 CB17.01 MCB-11-9-1 8.2127E-07 7.68E-01 6.31E-07 2.02E-02 1.66E-08 CB17.01_MCB-12 8.2127E-07 1.77E-01 1.45E-07 6.62E-04 5.43E-10

Document Control Desk RC-12-0142 Page 73 of 87 Table 1. Main Control Board Fire Scenarios - Results Scenario -IGF CCDP CDF. CLERP LERF CB17.01 MCB-12-1 1- 1.6745E-06 2.82E-02 4.71 E-08 1. 19E-05 2.00E-1 1 1

CB17.01_MCB-12 5.0364E-06 2.81E-02 1.42E-07 1.19E-05 6.01E-11 1

CB17.01_MCB-12-9-1 4.9205E-07 7.68E-01 3.78E-07 2.02E-02 9.93E-09 CB17.01_MCB-13 4.9205E-07 1.77E-01 8.72E-08 6.62E-04 3.26E-10 1

CB17.01_MCB-13 8;2127E-07 2.84E-02 2.33E-08 1-.20E-05 9.89E-12 1

CB17.01_MCB-13 1.6745E-06 2.84E-02 4.76E-08 1.20E-05 2.02E-11 1

CB17.01_MCB-13 5.0364E-06 8.49E-03 4.27E-08 3.58E-06 1.80E-11 1

CB17.01 MCB-14 4.9205E-07 2.84E-02 1.40E-08 1.20E-05 5.93E-12 1-CB17.01_MCB-14 8.2127E-07 2.84E-02 2.33E-08 1.20E-05 9.89E-12 1

CB17.01_MCB-14 1.6745E-06 8.49E-03 1.42E-08 3.58E-06 5.99E-12 1

CB17.01_MCB-14 5.0364E-06 8.49E-03 4.27E-08 3.58E-06 1.80E-11 1

CB17.01_MCB-15 4.9205E-07 2.84E-02 1.40E-08 1.20E-05 5.93E-12 1

CB17.01_MCB-15 8.2127E-07 8.49E-03 6.97E-09 3.58E-06 2.94E-12 1

CB17.01_MCB-15 1.6745E-06 8.49E-03 1.42E-08 3.58E-06 5.99E-12 1

CB17.01_MCB-15 5.0364E-06 7.58E-03 3.82E-08 3.20E-06 1.61E-11 1

CB17.01_MCB-16 4.9205E-07 8.49E-03 4.18E-09 3.58E-06 1.76E-12 1

CB17.01_MCB-16 8.2127E-07 8.49E-03 6.97E-09 3.58E-06 2.94E-12 1

CB17.01_MCB-16 1.6745E-06 7.58E-03 1.27E-08 3.20E-06 5.35E-12 1

CB17.01_MCB-16 5.0364E-06 7.58E-03 3.82E-08 3.20E-06 1.61E-11 1

CB17.01_MCB-17 4.9205E-07 8.49E-03 4.18E-09 3.58E-06 1.76E-12 1

CB17.01_MCB-17 8.2127E-07 7.58E-03 6.22E-09 3.20E-06 2.62E-12 1

CB17.01 MCB-17 1.6745E-06 7.58E-03 1.27E-08 3.20E-06 5.35E-12

Document Control Desk RC-12-0142 Page 74 of 87 Table 1. Main Control Board Fire Scenarios - Results Scenario IGF C(DP CDF CLERPP LERF 1

CB17.01_MCB-17 5.0364E-06 4.82E-03 2.43E-08 2.03E-06 1.02E-11 1

CR137.01 MCB-18 4.9205E-07 7.58E-03 3.73E-09 3.20E-06 1.57E-12 1

CB17.01_MCB-18 8.2127E-07 7.58E-03 6.22E-09 3.20E-06 2.62E-12 1

CB17.01_MCB-18 1.6745E-06 4.82E-03 -8.08E-09 2.03E-06 3.40E-12 1

CB17.01_MCB-18 5.0364E-06 4.65E-03 2-34E-08 1.96E-06 9.85E-12 1

CBR-7.01_MCB-2-1-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-2-2-1 5.0364E-06 4.67E-03 2.35E-08 -1.97E-06 9.90E-12 CB17.01 MCB-3-1-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01_MCB-3-2-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01_MCB-3-3-I 5.0364E 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-4-1-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01 MCB-4-2-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01 MCB-4-3-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CR17.01 MCB-4-4-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-5-1-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01 MCB-5-2-1 1.6745E-06 4.67E-03 7.82E-09 1.97E-06 3.29E-12 CB17.01 MCB-5-3-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-5-4-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01_MCB-5-5-1 5.0364E-06 4.67E-03 2.35E-08 1.97E-06 9.90E-12 CB17.01 MCB-6-1-1 8.2127E-07 1.04E-01 8.56E-08 4.41E-05 3.62E-11 CB17.01 MCB-6-2-1 8.2127E-07 1.04E-01 8.56E-08 4.41E-05 3.62E-11 CB17.01 MCB-6-3-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01 MCB-6-4-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01 MCB-6-5-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01 MCB-6-6-1 5.0364E-06 1.04E-01 5.25E-07 4.41 E-05 2.22E-10 CB17.01_MCB-7-1-1 8.2127E-07 1.04E-01 8.56E-08 4.41E-05 3.62E--1.

CR17.01 MCB-7-2-1 8.2127E-07 1.04E-01 8.56E-08 4.41E-05 3.62E-11 CB17.01 MCB-7-3-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01_MCB-7-4-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01_MCB-7-5-1 1.6745E-06 1.04E-01 1.74E-07 4.41E-05 7.38E-11 CB17.01_MCB-7-6-1 5.0364E-06 1.04E-01 5.25E-07 4.41 E-05 2.22E-10 CB17.01_MCB-7-7-1 5.0364E-06 1.04E-01 5.25E-07 4.41E-05 2.22E-10

Document Control Desk RC-12-0142 Page 75 of 87 Table 1. Main Control Board Fire Scenarios - Results Scenario iGF CCDP'. CDF CLERPf. LERF.:*-

CB17.01 MCB-8-1-1 4.9205E-07 1.07E-01 5.25E-08 1.53E-04 7.51E-11 CB17.01 MCB-8-2-1 4.9205E-07 1.07E-01 5.25E-08 1.53E-04 7.51E-11 CB17.01 MCB-8-3-1 8.2127E-07 1.07E-01 8.77E-08 1.53E-04 1.25E-10 CB17.01 MCB-8-4-1 8.2127E-07 1.07E-01 8.77E-08 1.53E-04 1.25E-10 CB17.01_MCB-8-5-1 8.2127E 1.07E-01 8.77E-08 1.53E-04 1.25E-10 CB17.01 MCB-8-6-1 8.2127E-07 1.07E-01 8.77E-08 1.53E-04 1.25E-10 CB17.01 MCB-8-7-1 1.6745E-06 1.07E-01 1.79E-07 1.53E-04 2.56E-10 CB17.01 MCB-8-8-1 5.0364E-06 1.07E-01 5-38E-07 1-.53E-04 7.69E-10 CB17.01 MCB-9-3-1 4:9205E-07 8.69E-01 4.27E-07 1.70E-02 8.37E-09 CB17.01 MCB-9-4-1 4.9205E-07 8.69E-01 4.27E-07 1.70E-02 8.37E-09 CB17.01 MCB-9-5-1 4.9205E-07 8.69E-01 4.27E-07 1.70E-02 8.37E-09 CB17.01 MCB-9-6-1 8.2127E 8.69E-01 7.13E-07 1.70E-02 1.40E-08 CB17.01 MCB-9-7-1 8.2127E-07 8.69E-0-1 7.13E-07 1.70E-02 1.40E-08 CB17.01- MCB-9-8-1 1.6745E-06 8.69E-01 1.45E-06 1.70E-02 2.85E-08 CB17.01_MCB-9-9-1 5.0364E-06 7.50E-01 3.78E-06 1.90E-02 9.57E-08 PRA RAI 67 In Table W-3 of the LAR, Fire areas CB10, CB12, and IBll1 contain VFDRs, yet the ACDF and ALERF are listed as N/A. Since all VFDRs are supposed to have their deltas calculated so that they can be summed and compared to R.G. 1.174, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis," guidelines, explain theN/A.

SCE&G Response The VFDRs in these-fire areas were resolved using performance-based fire modeling, as allowed under NFPA-805 Section 4.2.4.1. Under the regulations, the licensee has two options available for resolving VFDRs, 4.2.4.1 (fire modeling) OR 4.2.4.2 (fire risk evaluation). This means that if VFDRs are resolved under 4.2.4.1, they do not need to have delta-risk considered in the FREs. To require such calculations would mean that VFDRs resolved by performance-based fire modeling would be required to show compliance under both 4.2.4.1 AND 4.2.4.2, rather than OR as permitted under the regulation. This would render 4.2.4.1 as completely useless, thus contravening the regulation. It should be noted that the basic premise of 4.2.4.1 is that if sufficient performance-based margin can be demonstrated, the VFDR in question is not credible (i.e., the failures that cause the VFDR will not happen), which by definition means that the residual risk (if any) is negligible.

Document Control Desk RC-12-0142 Page 76 of 87 PRA RAI 68 In the Generic Fire Methodology report, the transient zone describes the zone of damage from the fixed or transient ignition source. According to page 12 of 73 of this report, the transient zone bounds the effect of flame spread or propagation since an extended range of 4 feet beyond the transient zone boundaries was examined for PRA targets. Explain how the transient zone boundary takes into account fire growth over time, or propagation into adjacent transient zones via secondary combustible fires. Should the finding be that the 4 feet margin does not account adequately for these issues, the PRA should be updated accordingly.

SCE&G Response The Fire PRA modeling strategy is as follows: All the targets within the transient zone are assumed failed. The transient zone is selected to be large and with an overlap to ensure that no targets are missed. If the transient zone can produce a hot gas layer scenarios, then the entire zone is assumed failed.

PRA RAI 69 On page 34 of 73 of the Generic Fire Methodology report, it is assumed that the growth of electrical motor fires to peak is 6 minutes. Since it is acknowledged that no basis for this can be -found in the literature, a sensitivity study should be performed evaluating different growth rates and its effects on the risk. A justification should be provided that the full expected range of the growth time has been explored in this sensitivity study.

SCE&G Response A sensitivity study was not performed. The electrical motor fires have the lowest peak HRR of all the ignition sources listed in Table E-1 of NUREG/CR-6850 (98th percentile: 69 kW).

Several CFAST runs were performed involving motors (for example, ABO1.09: DC0780B-013 and AB101.29: DC0780B-035) and damaging temperatures in the hot gas layer were not generated. The only significant fires were those that spread to overhead cables (IB10:

DC0780B-1 72), in which case the contribution of the heat from the motor was small.

PRA RAI 71 On page 64 of 73, of the Generic Fire Methodology report for transient fires, a HRR of 15kW and a time of 5 minutes is assumed to damage all targets within the ZOI. Justify that the 15 kW bounds all scenarios, (i.e. that the 15kW is the minimum transient HRR necessary to cause damage for all scenarios). If not, then adjust the minimum HRR for those scenarios which experience damage at less than 15kW and update the FPRA.

Document Control Desk RC-12-0142 Page 77 of 87 SCE&G Response There is no specific configuration for this assumption. A relatively small representative heat release rate value was selected to ensure a relatively high severity factor. A five minute value for failure of all the targets within the full transient zone was also selected as a bounding timing considering the relatively large selection of transient zones.

PRA RAI 72 On pg 66 of 73 of the Generic Fire Methodology report, per incipient detection systems, the report indicates that credit for incipient detection systems apply to components in low voltage cabinets in selected areas, and to other components expected to demonstrate an incipient fire growth stage. It is further assumed that those low voltage cabinets to which credit is applied do not have fast acting components in them. FAQ 46 only allows credit for low voltage cabinets, providing those components inside are not a part of those fast acting components identified in he FAQ. A review of those cabinet components should be performed and-the model modified if any of those components in the cabinets are fast acting. Discuss any variations from FAQ 46 in your treatment of incipient detection credit in the FPRA.

SCE&G Response An initial review by station engineering failed to identify any fast acting components within the scope of panels planned for Incipient Detection, which would be consistent with FAQ 46. Since this is a basis for the scope of panels to be provided with Incipient Detection (Table S-1), an independent validation is planned as part of the ECR 50811 Modification package development

[Action].

PRA RAI 73 For the SUPP credit on the incipient detection system (E2), 3.5 minutes is allowed for suppression. The assumption is that this is the time for a fire to grow 2.5 feet above the cabinet. Describe the height of target cables above the cabinets in which incipient detection is installed. Describe whether 2.5 feet is the minimum height. Describe the assumed scenario, taking into account the assumed HRR and indicating its source.

SCE&G Response Incipient detection systems apply to CB15 and CB06. The 2.5 feet is an error in the report

[Action]. The text should read "The 3.5 min value is a representative conservative value applied to all the scenarios. It is the estimated time it takes the flames of a fire postulated outside an electrical cabinet to reach the cable damage temperature of approximately 2050C, using the Heskestad plume temperature correlation, assuming a fire growth to a peak of 1 MW in 12 min (fire growth rate recommended in Appendix G of NUREG/CR-6850)".

Document Control Desk RC-12-0142 Page 78 of 87 PRA RAI 74 In Step 3 on page F-2 of the LAR, it is stated that, "The inclusion of MSOs in the FPRA is still needed". Clarify that the LAR means that, although screening is not needed, MSOs are still included in the PRA as directed by FAQ 07-0038.

SCE&G Response This is correct. Wherever it says inclusion in the FPRA is needed, the intent is that the MSO would be passed to the FPRA task, and in each case the MSO has beenevaluated and included appropriately in the FPRA model. The assessment of each MSO passed to the FPRA is fully documented in Attachment 4 to DC00340-001 Fire PRA Plant Response Model Report, Task 5.5, under the discussion of Step 8. The specifics for each MSO are presented on pages 37-55.

PRA RAI 75 On page 9 of 60 in the CR Risk Calculation report,lhe HRR profile decays in 5 minutes.

Provide justification for this decay time.

SCE&G Response The control room report, DC0780B-1 00, Section 4.3.1 states that the decay time of 5 minutes is the CFAST default, as follows: "After 20 minutes, the fire decays for a period of 300 seconds; the default rate included in the CFAST model." The CFAST default rate of 300 s (5 minutes) was used since specific values are not recommended in NUREG/CR-6850. In addition, the default value from CFAST is based on the following text from NUREG/CR-6850, Attachment G, page G-3: "Once the fire starts to decay, temperatures in the room will decrease to ambient conditions. As a result, and depending on the objectives of the simulation, modeling the decay stage of the fire usually does not provide critical information in support of risk decisions."

PRA RAI 76 A fire brigade arrival time of 10 minutes is assumed in the non-suppression probability for CR abandonment in the CR Risk Calculation. Describe how the 10 minutes compares to the time identified in pre-fire plans for fire brigade arrival to the CR during a fire or during a fire brigade drill. If the 10 minutes is unable to be justified, provide a more realistic time and indicate the impact on the PRA results.

SCE&G Response While 10 minutes was chosen as a realistic time to arrival for the fire brigade, the arrival times of 5 minutes and 20 minutes were also simulated, as stated in the report (DC078OB-100, Section 5.6).

Document Control Desk RC-12-0142 Page 79 of 87 PRA RAI 77 The Feb 2011 follow-on fire peer review F&Os are identified in Table V-18 as "DRAFT".

Confirmthe final versions are consistent with what is summarized in Table V-18 and that there are no additional F&Os.

SCE&G Response The final versions have been confirmed to be consistent with the summary. There are 42 Findings in the final version of the combined peer review report, as stated in Attachment V.

PRA RAI 78 The LAR states that the results of the 2005 review were "the VCSNS PRA (internal event-PRA) was found to meet CC-Il or better for 211 of the_271 SRs from the ASME PRA Standard, but 45 of the elements were found to either not meet the requirement or to meet the requirements at-a CC-I level."- Describe whether all 60 of the less than CC-Il level SR's are included in the tables in Appendix U. If not, describe which ones are missing and why. Add any missing SRs to tables in Appendix U.

SCE&G Response Appendix U lists A and B level F&Os, not supporting requirements (SRs) or their classifications.

If the reason for not meeting a CC-Il level resulted in an A or B level F&O in either the original 2002 Westinghouse peer review or the later gap analysis peer review, they are listed in Appendix U, otherwise, they are not.

PRA RAI 79 Attachment U, table V-18 does not identify a status of the F&Os as was done in Table U-1 for the IEPRA peer review F&Os. In some instances the disposition states that the-FPRA has been revised but does not confirm if the results presented in the submittal are based on the revised model. In other instances, the disposition indicates that the tasks are not complete. Clarify the status of each item in Table V-18.

SCE&G Response The NRC has asked specific question on each of the items for which the status was unclear, and these have been addressed individually above. The results presented in the submittal are based on the revised model.

Document Control Desk RC-12-0142 Page 80 of 87 PRA RAI 80 Confirm that the revised internal events model which addresses the F&Os was used in the FPRA.

SCE&G Response VCSNS confirms that the revised internal event model that addresses the F&Os was used in the Fire PRA.

PRA RAI 81 Attachment U, Table U-1 supporting requirement SY-01 addresses modeling of component cooling water support systems. The licensee's disposition indicates that service water and AC/DC power were added to the model; however, does not address loss of instrument air. Clarify how instrument air supports component cooling water system and how this relationship is modeled in the PRA.

SCE&G Response None of the functions for which component cooling water (CCW) is modeled in the VCSNS PRA depends on instrument air, so instrument air is not modeled as a support system for CCW.

PRA RAI 82 In Fire Risk Evaluation A.1.2.33 in the analysis of Power to Bus XMC, it states that those scenarios which are qualitatively or quantitatively screened do not contribute to the ACDF or ALERF. It also states that those unscreened scenarios are treated in a bounding approach. Describe the meaning of quantitatively screened, and indicate if single or multi-compartment contributions to CDF, LERF, ACDF, ALERF are screened. If these contributions are screened, quantitatively justify that these screened portions are insignificant with respect to the CDF, LERF, ACDF and ALERF acceptance guidelines.

Otherwise, include the results in the baseline risk values and ACDF/ALERF. Perform a review of this issue for the full plant analysis.

SCE&G Response Quantitatively screened means that no model refinements of these scenarios were done in order to make the result more realistic. In particular, it was assumed that all targets in the area were damaged by the scenario, rather than refining the fire modeling for greater realism. No single compartment scenarios are actually removed from the quantification. They were left in with their bounding values. Therefore, in this case "quantitative screening" means "screened from further detailed analysis," not "removed from the model." There were some multi-compartment scenarios screened by removal from the model, but this was limited to mutli-compartment scenarios that had the same impact (i.e., CCDP) as the associated single

Document Control Desk RC-12-0142 Page 81 of 87 compartment scenarios, since they would have a much lower frequency. Any screened scenarios are included in the calculation of the overall CDF and LERF values.

PRA RAI 83 In compartment AI301.03 of Attachment A of the Fire Modeling Scoping Report, pump fires from oil spills are assigned a HRR of 767 kW, regardless if the oil spill is 10% or 100% capacity of the pump. Explain why the HRR does not distinguish between oil spill sizes, or provide an updated analysis adjusting the HRR. Examine the rest of the pump oil spills to verify that this distinction is made in the compartment analyses,and correct if necessary.

SCE&G Response Response: The values of HRR for 10% and 100% capacity of pumps were only used in those scenarios that underwent detailed fire modeling. In Attachment A of the Fire Modeling Scoping Report (TR0780B-001), all compartments which underwent-detailed analysis (indicated by "Yes" in the box "Analyze this room?" in the upper left-hand corner of each page) use HRR values of 1917 kW for the 100% oil fire (ZO1) and 383 kW for the 10% oil fire (Z02). For the other zones that did not undergo detailed analysis (indicated by "Whole Compartment lost" in the "Analyze this room?" box), the values of HRR are listed as 767 kW, regardless if the oil spill is 10% or 100% capacity of the-pump. But, these values have no impact on the results, because damage is assumed at time=0, with severity factor of 1.0. To avoid confusion, the values of HRR for the "Whole Compartment lost" zones will be corrected to 1917 kW for the 100% oil fire (ZO1) and 383 kW for the 10% oil fire (Z02) in Attachment A of the Fire Modeling Scoping Report.

PRA RAI 84 FSS-D7-01 specifies that outlier experience was not examined for fire detection and suppression systems. The plant disposition indicates that generic values are larger than plant specific values and no outlier behavior exists. Justify the extent of the review of plant specific behavior. Provide a discussion of the plant-specific results and a discussion of the comparison to the generic values.

SCE&G Response

Background

Compliance with Supporting Requirement FSS-D7 at Category II requires a demonstration that "the system has not experienced outlier behavior relative to system unavailability." As stated in the associated Note 7 of the Standard (ASME/ANS RA-Sa-2009):

The intent for Capability Category II is to additionally require a review of plant records to determine if the generic unavailability credit is consistent with actual system unavailability. Outlier experience would be any experience indicating that actual system is unavailable more frequently than would be indicated by the generic values.

Document Control Desk RC-12-0142 Page 82 of 87 Peer Review Comment FSS-D7-01 found, based on follow-on peer review, that Supporting Requirement FSS-D7 was met at Category I, for the reasons stated in the comment:

Fire Modeling: Generic Methodology Calculation Number DC0780B-001, Section 6.1.3.3.

Per discussion in the calculation, each credited system was reviewed to ensure the applicable codes and standards are met and that there is current surveillance testing to-ensure operability is maintained. Plant specific data was not reviewed for this task; outlier experience was not searched for either.

The present RAI response demonstrates compliance at Category II by reviewing plant records and showing that the generic unavailability is consistent with actual system unavailability.

The generic failure probabilities for automatic suppression as represented in the FPRA are as follows (NUREG/CR-6850 [p. P-6]):

Autosuppression Type Failure Probability Wet-pipe sprinkler 0.02 Carbon dioxide 0.04 Deluge sprinkler 0.05 Preaction sprinkler 0.05 Halon 0.05 The generic failure probability for automatic detectors (whether for heat or smoke) is 0.05 (NUREG/CR-6850 [p. P-6]).

Method VC Summer Operations Removal and Restoration (R&R) forms indicate periods of time when components of the active fire suppression systems were out of service due to equipment failures or were taken out of service for maintenance, for testing, to avoid the potential hazards of spurious activation, or for other reasons. Unavailabilities over the period 5/1/2008 through 5/1/2012, for which R&R data were readily available, were calculated for each component as the ratios of the time out of service to the reactor operating time. In cases for which failure of one or more components will cause the system to fail, the overall system unavailability is approximated using the rare event approximation as the sum of the availabilities of the components. Because the nuctear safety functions of the components are not required when the plant is not operating, time out of service during a plant outage was not counted as unavailable time.

Carbon Dioxide Fire Suppression The CO 2 delivery system is illustrated in the drawing D-302-232. The following components, which are required for operation of the system for the Relay Room (CB06), are routinely taken out of service as documented in the R&R forms:

1. XPN-83-FS (XPN0083, C02 FIRE SYSTEM CONTROL PANEL)

Document Control Desk RC-12-0142 Page 83 of 87

2. XTK-125-FS (XTKO125, FIRE SERVICE LOW PRESS C02 STORAGE TANK)

3. 14089-FS (XVA14089-FS, BOP RELAY ROOM PILOT CONT VLV ISOL VLV)

The unavailabilities in the table below are calculated based on the R&R forms as noted in the Method section above. The failure of any of the three components will cause the system to fail.

Therefore, the overall system unavailability is approximated as the sum of the availabilities of the components. Because the calculated unavailability based on plant specific data (2.7E-02) does not show that the actual system is unavailable more frequently than would be indicated by the generic value (4E-02), outlier behavior is not observed and FSS-D7 is met-at Capability-Category II with respect to C02 fire suppression.

FSEquipID Description Unavailability XPNO083 C02 FIRE SYSTEM CONTROL PANEL 1.7E-03 XTKO125 FIRE SERVICE LOW PRESS C02 STORAGE TANK 2.4E-02 XVA14089-FS BOP RELAYROOM PILOT CONT VLV ISOL VLV 7.9E-04 Sum 2.7E-02 The C02 system is disabled to protect personnel whenever work is required where personnel could be harmed by discharge of C02. This results in a relatively high unavailability for the C02 storage tank. A continuous fire watch must be established within one hour of the system becoming inoperable (Enclosure 6.1.5 of Fire Protection Procedure FPP-024). Backup fire suppression equipment consisting of a single C02 portable fire extinguisher is also required during -the fire watch. No credit for the requirement for a fire watch is taken in the unavailability calculation. The requirement is presented here as an indication of defense in depth.

Preaction Sprinkler Fire Suppression The CableSpreading Rooms (CB15 and CB04) and Cable Chases (portions of CB01.01) in the Control Building and portions of Elevation 412 of the Intermediate Building are served by two separate preaction fire-suppression systems. These two systems are representative of preaction systems elsewhere in the plant.

(1) For the Cable Spreading Rooms and Cable Chases, the two valves listed in the following table are required for system operation (drawing D-302-231 SHT 5) and are sometimes removed from service as recorded in the R&R forms. Unavailabilities shown in the table below were calculated based on the R&R forms as noted in the Method section above.

FSEquiplD Description Unavailability XVG04064-FS CABLE SPREAD RM SPR SYS DELUGE VLV IN 1.9E-03 XVM04065-FS CABLE SPREAD RM SPR SYSTEM DELUGE VALVE 7.2E-03 Sum 9.0E-03 (2) For the Intermediate Building, the two valves listed in the following table are required for system operation (drawing D-302-231 SHT 5) and are sometimes removed from service as recorded in the R&R forms. The unavailabilities shown in the table below were calculated based on the R&R forms as noted in the Method section above.

Document Control Desk Attachment 5 RC-12-0142 Page 84 of 87 FSEquipID Description Unavailability XVG06934-FS lB SPRINKLER SYS DELUGE VLV INLET VALVE 4.2E-04 XVM06935-FS INTERMEDIATE BLDG SPR SYS DELUGE VALVE 2.6E-03 Sum 3.OE-03 Because the calculated unavailabilities based on plant specific data (9.OE-03 and 3.OE-3) do not show that the actual system is unavailable more frequently than would be indicated by the generic value (5E-02), outlier behavior is not observed and FSS-D7 is met at Capability

-Category II with respect to preaction fire suppression.

Wet Pipe Sprinkler Suppression Portions of Elevation 412 of the Control Building (CB01.01 at floor level) are served by a wet-pipe sprinkler system. The valve listed in the following table is required for system operation-(drawing D-302-231 SHT 5) and is sometimes removed from service-as recorded in the R&R forms. Unavailabilities shown in the table below were calculated based on the R&R forms as noted in the Method section above.

FSEquipID Name Comment Unavailability WET PIPE SPR SYS ALM LOTO for System Modifications XVG04104-FS CHK VLV IN VLV ECR50481B 2.2E-02

.... LOTO for Valve Repair 1.9E-03 Sum 2.4E-02 Because the calculated unavailability based on plant specific data (2.4E-02) indicates that the actual system is unavailable with approximately the same probability as indicated by the generic value (2E-02), outlier behavior is not observed and FSS-D7 is met at Capability Category II with respect to wet-pipe sprinkler fire suppression.

Detection for the C02 Fire Suppression Systems According to the R&R forms, detectors that initiate the CO 2 system were removed from service during the period 5/1/2008 through 5/1/2012. However, the time out of service occurred during a plant outage. Therefore, there is no contribution to system unavailability due to removal of the detection system from service. Therefore, outlier behavior is not observed and FSS-D7 is met at Capability Category II with respect to detection for the CO 2 fire suppression system.

Detection for the Sorinkler Systems

Document Control Desk RC-12-0142 Page 85 of 87 The R&R forms indicate that detectors for sprinkler systems were frequently taken out of service during plant outages during the period 5/1/2008 through 5/1/2012. On some occasions, detectors were taken out of service during plant operation as noted in the table below.

W 0 "

0 0 rjo E r= -Cr

@6 0-0 actor Removed from Svc to Support "A" r Mod. Less than min req'd Detectors in 100415 IXA04982P Zone. 11323.01-12-12 2207:30 7.0E-02 100256 IXA04989J Disabled for Welding B25.01.02-12-02 105:15 3.3E-03 "IXA04989K Disabled for Welding B325.0_I.02-12ý02 105:15 3.3E-03 1I(A04989L Disabled for Welding B25.01.02-1.2-02 105:15 3.3E-03 IXA04989M Disabled for Welding 1125.01.02-12-02 105:15 3.3E-03 The detector listed in the first row of the table above has a calculated unavailability greater than the generic pro~bability (0.05). However, because the room served by the detector is treated as a full-compartment burnout in the FPRA, detection and suppression are not credited. Therefore, comparison with the generic probability is not required.

The remaining unavailabilities in the table above are much less than the generic probability for detectors (0.05) and therefore do not show outlier behavior. Therefore, FSS-D7 is met at

.Capability Category 1rwith respect to detection for sprinkler fire suppression systems.

Summary Unavailability calculations based on suppression equipment being out of service during plant operation for C2 and pre-action firfe'suppression systems show unavailabilities less than the generic failure probabilities that are credited in the FPRA. Unavailability calculations based on wet-pipe fire suppression equipment being out of service during plant operation show unavailabilities approximately equal to the generic failure probability that is credited in the FPRA. Similarly for detection equipment, unavailability calculationis for fire detectors associated with the Cr 2 or sprinkler suppression systems show detector unavailabilities less than the generic failure probability that is credited in the FPow. Therefore, outlier behavior is not observed for fire detection and suppression systems that are credited in the FPRA and Supporting Requirement FSS-D7 is met at Category for c fire adetection and suppression.

Document Control Desk RC-12-0142 Page 86 of 87 PRA RAI 85 The plant disposition for UNC-A2-01, 02, and 03 indicate that the peer review finding has been addressed. Provide a technical discussion or analysis, indicating how this issue has been addressed. In addition, provide a justification for those sensitivity studies that are not consistent with expected results as well as insights gained.

SCE&G Response The F&O notes that that the VCS FPRA addressed the sources of uncertainty through sensitivity analysis rather than parametric uncertainty, but felt that the sources of-uncertainty were not well documented. The peer review team suggested adding a table that discusses the sources of uncertainty considered for each technical element (TE) and how they were treated.

The task 15 report (ATTACHMENT 13 TO DC00340-001 FIRE PRA SENSITIVITY AND UNCERTAINTY REPORT, TASK 5.15) has added a bulleted list (rather than a table) to provide the sources of uncertainty by TE, as indicated in the following:

Each fire PRA task, -was reviewed for identification of its sources of uncertainty. This included relevant assumptions and approximations made in the development of the fire PRA model. Identified sources of uncertainty are documented below. Tasks not listed did not have any sources of uncertainty identified for sensitivity analysis.

Task 5.5 - Fire PRA Modeling Plant modification modeling - Proposed plant modifications were incorporated into the Fire PRA model. Some of the major changes include 115kV re-route in-the Turbine Building, RCP Seal replacement. There are also numerous cables that will be protected that are removed from the map sets for particular fire zones. Several sensitivity runs associated with plant modifications will be run.

See Table 2 for details. Additional details about the plant modifications can be found in the Task 5.5 report, Step 7c.

" Alternate Seal Injection System - A new alternate seal injection system will be installed in the plant. Currently, there is no design for the system, so a conservative failure probability was used to provide credit for that system.

Sensitivity runs will be performed to determine the impact of the alternate seal injection system on the CDF and LERF.

" MSO Modeling - Multiple Spurious Operation (MSO) events are an important part of the Fire PRA. Several MSOs were identified and included in the Fire PRA

Document Control Desk RC-12-0142 Page 87 of 87 model. The model will be run with theses MSOs excluded to determine the impact on risk.

- Task 5.6 - Ignition Frequencies. Ignition Frequencies for each scenario will be run using the NUREG/CR-6850 frequencies.

"Task 5.9 (Task 4.4) - Detailed Circuit Analysis.

" General Cable Mapping - Detailed circuit analysis was used to eliminate cables from target mappings for certain scenarios. Sensitivity analyses will be performed-to determine the impact of adding those cables back into the target set for each scenario.

" Disconnect Switch Failure - It was determined that disconnect switches will not fail during fire scenarios and will be available (via detailed circuit analysis).

Functional States (FSs) associated with these components-are currently mapped to a dummy event so they do not impact the fire model. Sensitivities will be performed as detailed in Table 2.

"Task 5.10 - Spurious Hot Short Probabilities. Hot short probabilities are used for certain spurious FSs in the analysis. This sensitivity will determine the impact of the use of those probabilities by failing those FSs for all scenarios.

- Task 5.11 and, as relevant, Task 5.8 - Detailed Fire Modeling. Several factors used to develop ignition frequencies in the FMDB will be analyzed for model sensitivity (See Table 2).

" Non-Suppression Credit

" Incipient Detection

" Administrative Controls for Transient Materials (also associated with Task 5.12)

- Task 5.12 - HRA. HEPs area source of uncertainty. Sensitivity analyses performed can be found in Table 2.

Table 2 of the subject report provides the specific details of the sensitivity analysis that was performed, including a description of the sensitivity, how it was modeled/run, and the results. This table also provides insights/lessons learned from each sensitivity analysis.

Document Control Desk RC-12-0142 Page 1 of 2 VIRGIL C. SUMMER NUCLEAR STATION (VCSNS) Unit I DOCKET NO. 50-395 OPERATING LICENSE NO. NPF-12 ATTACHMENT 6 Radioactive Release (RR) Request for Additional Information (RAI) Responses

Document Control Desk Attachment 6 RC-12-0142 Page 2 of 2 RR RAI 01 For areas where containment/confinement is relied upon:

a. Liquid

1. Describe how the qualitativelquantitative assessment addressed capacities of sumps, tanks, transfer pumps, etc., as appropriate.

2. Describe if there are plant features that may divert the effluent flow that were not taken into account (e.g., Aux. Bid. roll-up doors).

b. Gaseous

1. Describe if there are plant features that can bypass the planned filteredlmonitored ventilation pathway that have not been accounted for.

SCE&G Response

a. Liquid

1. The capacities of sumps/tanks/transfer pumps are analyzed in calculation-DC07810-033 Evaluation of Fire System Flooding Effects Outside the Reactor Building, Revision 0. This calculation determines the flooding scenarios for the fire system and the capacities of the drains, sumps, and tanks to contain the water flow from sprinkler systems and hose stations.

2. Precautions of plant features, such as exterior doors, which could potentially release contaminated smoke or liquid runoff to the environment are specified in Attachment III - Required Pre-Fire Plan Revisions of calculation TR07800-006.

b. Gaseous

1. DC00040-116, Rev. 1, EAB TEDEs - NFPA 805 Compliance utilized conservative assumptions and methodology in calculating radioactive release to the environment; and did not assume filtration prior to a gaseous radioactive release.

RR RAI 02 For areas where containment/confinement is, not available describe whether the assessment credits operator actions.

SCE&G Response Calculation DC00040-116 EAB TEDEs - NFPA 805 Compliance was performed to analyze the total estimated dose equivalent for liquid and gaseous release to the exclusion area boundary in accordance with NFPA 805. Calculation DC00040-116 does not credit operator actions in the analysis of gaseous and liquid release of radioactive effluents for areas where containment/confinement is not available.