ML20202E223
| ML20202E223 | |
| Person / Time | |
|---|---|
| Site: | Cooper  |
| Issue date: | 01/28/1999 |
| From: | Swailes J NEBRASKA PUBLIC POWER DISTRICT |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GL-88-20, NLS990008, NUDOCS 9902020340 | |
| Download: ML20202E223 (50) | |
Text
_
1 N
Nebraska Public Power District Nebraska's Energy Leader NLS990008 January 28,1999 U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555-0001
)
Gentlemen:
Subject:
Response to Request for Additional Information - Individual Plant Examination for External Events (IPEEE)
Cooper Nuclear Station, NRC Docket 50-298, DPR-46
Reference:
- 1. Generic Letter 88-20, Supplement 4," Individual Plant Examination for External Events (IPEEE) for Severe Accident Vulnerabilities" l
- 2. Letter (No. NLS960143) to USNRC Document Contro! Desk from G. R.
Horn (NPPD) dated October 30,1996, " Individual Plant Examination for External Events (IPEEE) Report - 10 CFR 50.54(f)"
- 3. Letter to G. R. Horn (NPPD) from James R. Hall (USNRC) dated June 3, 1998, " Request for Additional Information Related to the Individual Plant i
Examination of External Events (IPEEE) for the Cooper Nuclear Station (TAC No. M83611)"
The purpose of this letter is to submit to the Nuclear Regulatory Commission (NRC) the Nebraska Public Power District's (District's) response to the Request for Additional Information l (RAI) dated June 3,1998 (Reference 3). The RAI questions were based on the NRC's ongoing review of the Individual Phnt Examination of External Events (IPEEE) for the Cooper Nuclear f
Station (CNS), submitted Ocicber 30,1996 (Reference 2). The IPEEE report (Reference 2) was submitted to the NRC under the requirements of 10 CFR 50.54(f) in response to Generic Letter (GL) 88-20, Supplement 4 (Reference 1).
. Please find attached the District's response for the Cooper Nuclear Station (CNS) to the subject RAI. This response restates each of the RAI questions (identified in i:alics)in the order they were presented in Reference 3.
9902020340 PDR P
ADOCK 0500029 996138 [9 PDRr OD m .
l Cooper Nudear Station PO. Box 98/ Brownville, NE 68321D090 Telephone: (402) 82S 38H / Fax: (402) 825-52H http //www.nppd com t
' NLS990008 January 28,1999 Page 2 0f 2
- Should you have any questions concerning this matter, please contact me.
Sincerely,
'~ ..
- ; /
.Sw hs J
i Pre ' der t ofNuclear Energy -
/dnm Attachm ec: Regional Administrator
. USNRC - Region IV.
i
. Senior Project Manager .
USNRC - NRR Project Directorate IV-1 ;
Senior Resident Inspector USNRC NPG Distribution i i
J t
i j
i
Nebraska Public Power District's Response to NRC ' Request for AdditionalInformation regarding the Individual Plant Examination for External J
Events (IPEEE) at Cooper Nuclear Station Table of Contents Attachment 1 - Response to NRC RAI on Cooper Nuclear Station IPEEE (32 Pages Total)
Page1 'NRC Questions A.1 - A.13 and District's Response on Fire Events Page 23 NRC Questions B.1 - B.3 and District's Response on Seismic Events Page 27 NRC Questions C.1 - C.2 and District's Response on High Winds, Floods, and other External Events Page 31 References Attachment 2 - Figures Related to NRC Question B.2 and District's Response (13 Pages Total)
Attachment 3 - List of NRC Commitments 6
1
. . - . - . . - . - - . . .- .. ~.-
Attochment 1 to NLS990008 Page1of31 Response to Nuclear Regulatory Commission Request for Additionalinformation '
Concerning the Cooper Nuclear Station Individual Plant Examination for External Events The following is the Nebraska Public Power District's (District's) respense to each of the individual questions contaim:d in the Nuclear Regulatory Commission's (NRC) Request for Additional Information (RAi), dated June 3,1998 [ Reference 7], concerning the Individual Plant Examination for External Events (IPEEE) for Cooper Nuclear Station (CNS). The individual NRC questions (identified in italic) are presented below in the same order as identified in the RAI, and are followed by the District's response to each question. The references identified in the NRC questions and the District's response are listed on Pages 30 and 31 of this RAI response. The seismic response figures related to NRC Question 13.2 are located in Attacnment 2 of this RAI response. Attachment 3 identifies that there are no regulatory commitments contained in this response. i
)
A. Fire Events 1 NRC Question A.1: \
l The IPEEE submittalindicated that hot short (HS) failures were considered in the assessment; l
however it cannot be determined to what extent the licensee has considered HS as afailure mode for control or instrumentation cables. In particular, HS considerations should include the treatment ofconductor-to-conductor shorts within a system oftwo or more cables. HSs in control cables can simulate the closing ofcontrol switches leading to the repositioning of valves, spurious operation ofmotors andpumps, or the shutdown ofoperating equipment. These types offaults might lead to a loss-of-coolant accident (LOCA), diversion offlow within various plant systems, deadheading andfailure ofimportant pumps, premature or undesirable switching of pump suction sources, or undesirable equipment operations. For main control room (MCR) abandonment scenarios, such spurious operations and actions may not be indicated at the remote shutdown panel (s), may not be directly recoverablefrom remote shutdown locations, or may lead to the loss ofremote shutdown capability (e.g., through loss ofpower sources to the remote safe shutdownpanel.) In instrumentation circuits, HSs may cause misleadingplant readingspotentially leading to inappropriate control actions or generation ofactuation signals for emergency safeguardfeatures.
Pages 4-3 and 4-6 in the licensee 's submittal address the treatment ofHSs in the CNSfire risk assessment. From this briefdiscussion, it is not clear to what extent HSs were included in the fire analysis. In particular, the potentialforfire-induced LOCAs (e.g., through spurious opening ofsafety reliefvalves) and interfacing system LOCAs is not discussed in the submittal. In addition, the potentialfor HSs in control room abandonment scenarios was not specifically addressed
. Attachant 1 ts
'NLS990008 Page 2 of 31
. Discuss how the above HS issues (that is, the impact ofHS-Inducedfailures on safety systems or functions) have been consideredin the CNS1PEEE. Ifthey have not been considered, provide an
- , assessment ofhow the inclusion ofp'otential HSfailures wouldimpact the quantfication offire riskscenarios in the CNSIPEEE.
4 District Response to A.1:
These issues have been explicitly considered in the CNS IPEEE fire analysis. The fire analysis is based on the Appendix R documentation for the majority of the cable spatial vulnerabilities. In 4 the treatment of postulated cable failures induced by fire, all credible failure modes (open circuit, short circuit, and hot short) are considered, to the extent the cables were previously identified in ,
the Appendix R analysis. A key feature of the CNS analysis is that spurious equipment actuation is considered for all three of these failure modes.
4 For the postulated open circuit mode, interruption of current occurs in the affected cable. In this case, components would align themselves in their ' failed' state (e.g., " failed open,"" failed closed,"" failed in place"). Valves that require power to maintain a desired position, are assumed to change state. Relays that are normally energized, are assumed to become de-energized. The consequences of this relay action may include spurious actuation of mechanical system components. In all cases, fire induced failures are not postulated to assist components in :
achieving the desired state for this analysis. The "open circuit" failure mode was treated using a scenario specific conditional failure probability of P = 1.0.
Postulated short circuit conditions are defined as those fire induced failures wherein the conductors of an individual cable become ' connected' together in any combination. Again, this failure mode is applicable to power cables, control cables and instrumentation cables. The 4
failure modes considered include shorting of all conductors in power circuits, and the selected !
e shorting of conductors within individual control cables, to cause spurious equipment actuation. '
For example, a control cable between a motor control center and the control room was treated
. using failure modes including the shorting of conductors to generate a spurious valve-open or l
> close signal. As in the prior case, fire induced cable failures were not postulated to assist I components in achieving the desired state for this analysis. The "short circuit" failure mode was also treated with a conditional failure probability of P = 1.0.
A~ postulated hot short is a special case of the more general short circuit failure mode. This case involves'a very specific phenomenon wherein the energized conductor of a given cable becomes selectively connected to the de-energized conductor of another cable, thereby causing undesired spurious actuation of equipment associated with the second cable. This failure mode is very specific and unlikely, since it also requires that these ' shorted' conductors do not include certain other conductors'such as neutral or ground, and that they remain connected long enough to allow the affected component to change state. A conditional hot short probability of P = 1.0 was used for all fire zones except in Zone I A. The general statement [on page 4-3 of the IPEEE submittal] j
, "6. Fire-induced hot shorting of circuits is assumed to occur with a conditional probability at i
i l
Att: chm:nt 1 to NLS990008 Page 3 of 31 6.0E-2..." is not applicable to all cases analyzed. The conditional probability value P = 6.0E-2 was used in Fire Zone I A, only for the two Reactor Equipment Cooling (REC) heat exchanger valves identified below. In ull other cases P = 1.0 was used.
The conditional failure probability of P = 6.0E-2 for hot shorts was calculated, based on guidance L provided in the Electric Power Research Institute (EPRI) Fire Probabilistic Risk Assessment
! - (PRA) Implementation Guide. The application of this value has been validated by considering a case involving two cables, each consisting of two conductors. Cable 1 is assumed to be a power cable consisting of conductors 'P' (hot conductor) and 'N' (neutral conductor). Cable 2 is assumed to be connected to a de-energized solenoid valve whose desired post-fire status is to l remain de-energized. The cables are assumed to be routed in a grounded raceway system. This l
l configuration involves five conductors. The grounded raceway system is treated as a single
' virtual' conductor. l The possible cable shorting configurations involve pairs of conductors, groups of three, of four, and a single case involving all five conductors. There is a total of 26 combinations. Howeser, i only one of these combinations would result in the spurious energizing of the solenoid valve.
This would involve the connection of conductor 'P' of Cable I with the appropriate conductor in l Cable 2. All other combinations of conductor shorting would not result in spuriously energizing !
the solenoid valve. Analysis of configurations involving larger numbers of conductors results in l
l lower probabilities. Given the result of this calculation, the value of P = 6.0E-2, which is
! recommended in the EPRI Fire PRA Implementation Guide, has been selected for the REC heat ;
y exchanger outlet control valves SW-AOV-TCV451 A and SW-AOV-TCV451B, which are l
located in Fire Zone I A.:
L Fire Zone 1 A extends over a compartment consisting of two elevations in the Northeast corner l room of the Reactor Building. The walls, floor, and ceiling werejudged sufficient to meet the L FIVE boundary criteria. The scope of Appendix R equipment located in this area comprises:
Core Spray (CS) Pump 1 A (with associated valves and instrumentation) g RCIC System Equipment (turbine, pump, valves, starter racks, instrumentation) 125 V DC Starter Rack l L Area Cooler (HV-R-lF)
L L Other components / functions served by cables routed through this area are:
L ADS permissive signal from CS Loop A Remote level indication signal for ECST B (CM-LT-618B)
Torus temperature monitoring signal (PC-TE-1 A,1D,2B, and 2F) l HPCI torus suction valve - spurious operation (HPCI-MOV-MO58) l REC Heat Exchanger SW outlet valves - spurious operation (SW-AOV-TCV451 A & B) l l ,
L
i Att:chment I to NLS990008 >
Page 4 of 31 -
l-I !
In the Probabilistic Safety ' Analysis [PSA, Reference 8] quantification for this fire zone, the hot short failure mode was added to the potential failure modes for the REC heat exchanger SW.
' outlet valves.. Quantification of this scenario shows that simultaneous, independent hot shorts of the cables to these two valves would result in a total loss of REC cooling. The principal effect of i this failure is loss of room cooling and pump seal cooling for the inventory control and decay
~
- heat removal equipment (CS, HPCI, Residual Heat Removal (RHR), and RCIC) for this !
L particular event sequence. As indicated in the CNS Safe Shutdown Functional Requirements, I
{ room cooling capability is not required until the following points in time: l L RCIC - four'(4) hours after system start, HPCI - twenty (20) minutes after system start, and RHR & CS - one (1) hour after system start.
Thus, for safe shutdown during a postulated fire scenario, the room cooling is not required until l
after the initial plant response. Seal cooling is required only for the Core Spray pumps, and in
- engineering calculations which were performed in support of Appendix R analysis it was conclu'ded that core spray pump seal cooling is not required for the duration of the fire event mitigation.
l
'As described above, interruptions of the REC support function does not result in an immediate loss of the heat removal functions provided by the front line ECCS systems involved.
Additionally, the valves are air-operated, they are designed to fail open, and their failure is l l recoverable by operator action. In order to explore the uncertainty connected with the
! conditional hot short probability P = 6.0E-2, a sensitivity calculation was performed with this L value increased by a factor of 10, to P = 6.0E-1 Using this value, the resulting core damage
! frequency (CDF) was still below the screening criterion of 1.0E-6/yr. Thus, it is concluded that the hot short failure mode was appropriately analyzed for the Fire Zone 1 A scenario, with a resulting conditional core damage probability of CCDP = 2.75E-5.
! General Insights on Treatment of Hot Short Effects in the CNS IPEEE The CNS fire IPEEE does not contain specific reviews of the susceptibility to hot short induced l LOCAs, interfacing system LOCAs, or other hot short issues coincident with main control room evacuation scenarios.
i However, with regard to hot shorts causing LOCAs, all of the event trees used in the i lquantification of the conditional core damage frequency for the fire initiators include branches
- for single and multiple open relief valves. While the specific probability of the hot short was not included in the nodal probabilities, the phenomena associated with the loss ofinventory are
. explicitly considered. The conditional probabilities used for a single and for multiple stuck open
, relief valves are P = 1.0E-2 and P = 1.2E-3, respectively. These values are similar in magnitude j to the probabilities of single and multiple hot shorts. Therefore, no additional insights would be
- . gained by adjusting these values and re-performing the quantification. Consequently, LOCA i
l ~ - - . - _ _ _ - - - _ - ~ . -_ -
Att:chment I to NLS990008 ~
Page 5 of 31 scenarios caused by hot shorts are already covered by the analysis of stuck open relief valves.
1 I
With regard to hot shorts causing interfacing system LOCAs, the CNS IPE [ Reference 8]
l indicates six possible interface paths that could be affected by hot shorts. These are the HPCI,
)
RCIC, CS, and RHR injection paths. Each of these paths is also isolated by one or more check l
- valves between the system injection valve and the reactor vessel. These check valves are l i
normally in the closed position, and therefore are considered very reliable for providing isolation in the unlikely event that the system injection valve spuriously opens.
Automatic Depressurintion System (ADS) Actuation by Hot Shorts Another important LOCA scenario is a hot short induced actuation of the automatic l depressurization' system in an event sequence where the ability to provide the required makeup
- flow may be affected. The analyzed scenario for the control room involves Panel 9-3. This L panel was modeled as described in the IPEEE submittal [ Reference 9] and in the detailed i response to Question A.13. The configuration of Panel 9-3 in the CNS control room is an open j' back panel, with a detector located immediately over the panel back. This protective feature plus ,
l the fact that it is located in the control room provide high confidence that any fire would be 1 l- rapidly detected.
l In the event of a fire being detected, a fire brigade response would occur immediately. This L initial response is not dependent on the control room staffing needs to respond to the event. i l Thus,'while the fire would be responded to, the control room operators would be attending to the safe shutdown of the plant. A fire induced reactor trip or an operator initiated reactor trip is postulated in response to a diagnosis that the fire may impact safe shutdown. Because of past l analyses, which determined the importance of Panel 9-3, the District has implemented operator L training which stresses the importance of rapid response to fires in the control room and, ;
specifically to Panel 9-3. Thus, the operator would trip the reactor and begin immediately the d orderly safe shutdown sequence.' During the initial steps of this response the operator would set the ADS " inhibit" switch in order to preclude ADS operation.
l In the event where the control room supervisor had determined that positive control of the plant i could not be maintained due to environmental reasons (i.e., smoke) an evacuation to the alternate i L shutdown room would occur. At CNS it takes approximately four to six minutes to reach the
[ alternate shutdown panel, located in the alternate shutdown room. Another possible reason for L evacuation to the alternate shutdown room would be the detection of ADS blowdown after the
!~
ADS " inhibit" switch had been set. This would indicate that the fire had damaged the
[ components necessary to maintain positive control from the main control room, and therefore a transition to the alternate shutdown room would be initiated im. mediately. During the time of the evacuation to the alternate shutdown room, fire suppression activities would continue and may
. succeed in time for return to the main control room later in that same sequence. However, this
- l. recovery of the main control room during the same fire sequence was not modeled in the IPEEE
{ tire quantifications. This is a conservatism.
I I
m - _ . .m- .. _._ . - _ . . - _ _ . _ _ _ _ . _ ._. __ __ . _ _
Att:chment 1 ta NLS990008 Page 6 of 31.
The District has performed thermal hydraulic calculations of the blowdown rates that would occur when the ADS is actuated. These analytical results show that three ADS valves can lower the reactor pressure to a value at which the low pressure core cooling systems can inject, within approximately six minutes. If all eight ADS valves are actuated, depressurization occurs in about two minutes. The above sequence discussion is the limiting sequence of events that could occur, Hot Short Summarv-In summary, the IPEEE analysis of hot short effects in Panel 9-3 in conjunction with alternate shutdown scenarios covers all reasonably postulatable scenarios, if the multiple defense-in-depth ,
aspects of the bounding event sequence are considered. >
1 1
Main Control Room Evacuation With regard to main control room evacuation, the IPEEE submittal [ Reference 9] indicated on
- page 4-28 that a high screening value (conditional failure probability P = 1.0E-01) was used for i failure to shut down from the alternate shutdown panel.' Although not explicitly stated in the l submittal, the r,ssigned value was conservatively chosen because it was judged that failure to
- achieve plant shutdown from the alternate shutdown panel is dominated by human error. With this screening value it was determined if detailed modeling of the attemate shutdown process was ;
necessary. The analysis results showed that more detailed modeling was not necessary,
^ NRC Question A.2:
q' Fires in the MCR are potentially risk-sigmffcant because they can cause instrumentation and control (1& C) failures (e.g., loss ofsignals or spurious signals)for redundant division, and
' because they canforce MCR abandonment. Although datafrom two experiments concerning the timing ofsmoke-induced, forced MCR abandonment are available [ Reference 1], the data must
, be carefully interpreted, and the analysis mustproperly consider the differences in configuration .
between the experiments and the actual AfCR being evaluatedforfire risk. Inparticular, the experimental configuration includedplacement ofsmoke detectors inside the cabinet in which ,
l, thefire originated, as well as an open cabinet doorfor that cabinet. In one case, failure to V accountfor these configuration differences led to more than an order ofmagnitude l underestimate in the conditionalprobability offorced MCR abandonment [ Reference 2]. In
[' addition, another study raises questions about MCR habitability due to room air temperature i
concerns [ Reference 3]. :
V Provide the detailed assumptions (including the assumedfirefrequency, anyfrequency reduction factors, and theprobability ofMCR ebandonment) used in analyzing the MCR and the basisfor these assumptions: Inparticular, iftheprobability ofMCR abandonment is based on a probability distributionfor the time required to suppress thefire, provide ajustification ofthe
!. ' basisfor the selection ofthe parametricform ofthe probability distribution andspecify the data used in quantifying the distribution parameters. ,
4 1
! W f -
n
. -Att:chment I ts NLS990008 Page 7 of 31 District Response to A.2:
The fire analysis for the CNS IPEEE is not exclusively a Probabilistic Risk Assessment (PRA).
Instead, the FIVE methodology was used, which contains probabilistic analysis features where appropriate. Thus, the fire frequency estimates were based on the ignition source data sheets. l These were developed, based on the FIVE methodology and on consideration of the panel types in the main control room. As discussed below, the probability of main control room l abandonment is not based on a probability distribution.
l In the CNS fire analysis of the main control room, two basic scenarios are considered. In the first scenario the fire is successfully suppressed, and control room habitability is maintained. In the second scenario the fire suppression is not successful, and loss of control room function
- occurs. In the analysis for the second scenario it is assumed that a reactor trip is initiated prior to evacuation of the main control room. In the analysis for the first scenario, the determination of whether a reactor trip occurs is based on the controls and instrumentation that are present at the postulated fire location. Successful fire suppression is defined as extinguishment within 15 minutes.- Therefore, a reactor trip is postulated for all main control room fire scenarios with a fire duration in excess of 15 minutes.
Qualitative screening of control room scenarios was considered only for the successful fire suppression cases. In these cases, a postulated fire wasjudged to be oflimited duration (15 minutes or less) and would have consequences that are limited to loss of the controls and equipment at the postulated fire location. Fires with this limited duration were not considered to be a credible threat to control room habitability. Control room fire scenarios with fire durations
' of 15 minutes or less, and that met any one of the following three criteria were screened out.
- 1. Loss of equipment and wiring at the postulated fire location (panel) does not result in the loss of Appendix R safe shutdown systems and is not expected to cause or require a reactor trip, or
- 2. Loss of equipment and wiring at the postulated fire location (panel) does not result in the j loss of Appendix R safe shutdown systems, but may result or require a reactor trip, or
. 3. Loss of equipment and wiring at the postulated fire location (panel) results in the loss of I Appendix R safe shutdown systems, but is not expected to cause or require a reactor trip. l These criteria are based on the FIVE Phase I screening criteria. Criterion 1 is conservative since it requires that no effect on Appendix R safe shutdown systems and no reactor trip occur.
, Criterion 2 is valid and already considers the potential for a reactor trip. Criterion 3 was applied only after careful consideration of reactor trip likelihood. The determination of whether a reactor
[ trip occurs was based on an examination of the controls at the panel location and was not limited l
[' to Appendix R equipment and circuits.
i l l
l i
. . , . _ _ -. _ _ _ _ . _ . _ . _ . _ _ . _ . _ . - _ - . _ m . .- . . - _ - . . . . _ _ . .
Att:chmmt 1 to NLS990008 Page 8 of 31 The second screen was applied only to selected panel fire scenarios in the main control room
. where the postulated fire was successfully suppressed. Tb determination of whether a reactor trip occurred for these cases was based on an examination of the controls and instrumentation on
. the panel face. In this assessment, failure of all equipment with controls or related instrumentation at that panel location was postulated. In these cases, however, the effects of
' failures that occurred prior to successful fire ' suppression were included in the PSA quantifications.
The CNS IPEEE analysis ofpostulated main control room fires demonstrates that main control room evacuation is not a dominant contributor to core damage frequency. Even if the probability l of main control room abandonment were increased, for sensitivity purposes, by an order of ;
magnitude, no further insights would be identified. '
s~
NRC Question A.3:
NUREG-1407, Section 4.2 and Appendix C.3, and Generic Letter (GL) 88-20, Supplement 4, request that documentation be submitted with the JPEEE submittal with regard to the Fire Risk Scoping Study (FRSS) issues, including the basis and assumptions used to address these issues, and a discussion ofthefindings and conclusions. NUREG-1407 also requests that evaluation results andpotentialimprovements be specifically highlighted. Controlsystem interactions irwolving a combination offire-inducedfailures and high probability random equipmentfailures were identified in the FRSS aspotential contributors tofire risk.
The issue ofcontrol systems interactions is associatedprimarily with the potential that afire in the plant (e.g., the AfCR) might lead to potential controlsystems vulnerabilities. Given afire at the CNS, control systems interactions could occur between the MCR, the remote shutdown panel (RSP), and other systems requiredfor safe shutdown. Specific areas (sub issues) that have been identified as requiring attention in the resolution ofthis issue include:
3.1 Electrical independence ofthe remote controlsystems: The primary concern ofcontrol !
. systems interactions occurs atplants that do not provide independent remote (any location other than the MCR) control systems. The electricalindependence ofthe RSP and the evaluation ofthe level ofindication and control ofremote control and monitoring circuits (e.g., water level control, reactorpressure) need to be assessed.
' 3. 2 Loss ofcontrol equipment orpower before transfer: The potentialfor loss ofcontrol powerfor certain control circuits as a result ofHSs and/or blownfuses before !
transferring controlfrom the MCR to remote shutdown locations (the RSP or any other location) needs to be assessed. i 3.3 Spurious actuation ofcomponents leading to component damage, a LOCA, or an interfacing systems LOCA: The spurious actuation ofone or more safe-shutdown-related components as a result offire-induced cablefaults, HSs, or componentfailures leading to a
, - , --aa3- e a g-,,. -.--. y -y,, --- - . , , - m-
If l Att:chme;t I tm I E
NLS990008 l . Page 9 of 31 l component damage, LOC 1, or interfacing systems LOCA, prior to taking controlfrom l the RSP or any other location, needs to be assessed This assessment also needs to include the spurious starting and running ofpumps as well as the spurious repositioning ofvalves.
"3. 4 Totalloss ofsystemfimetion: The potentialfor totalloss ofsystemfunction as a result of ,
fire-induced redundant componentfailures or electrical distribution system (power source) failures needs to be addressed.
Provide an evaluation ofwhether loss ofcontrolpower due to HSs and/or blownfuses could occurprior to transferring control to the remote shutdown locations and identify the core damagefrequency (CDF) contribution ofthese types offailures. Ifthesefailures are screened
'for afire area, pleaseprovide the basisfor the screening In addition provide an evaluation of whether spurious actuation ofcomponents as a result offire-induced individual cablefaults, HS, or componentfailures could lead to loss ofsystemfunction, a LOCA and interfacing systems LOCA prior to taking controlfrom the remote shutdown locations (considering both spurious starting and running ofpumps as well as the spurious repositioning ofvalves).
District Response to A.3:
. The comprehensive treatment of hot shorts in the CNS IPEEE is described in the response to Question A.1 above. Additional detail, regarding fire analysis of Control Panel 9-3 and Control Board C is provided in the response to Question A.13 below.
In general, the District's response to the issues of the Fire Risk Scoping Study is provided in Section 4.8 of the submittal document [ Reference 9]. Taking the responses to Questions A.1 and I A.13, together with Section 4.8 of the submittal document, the issues discussed in the Question above are sufficiently addressed to satisfy the Fire Risk Scoping Study requirements identified by the NRC staffin the Safety Evaluation Report (SER) approving the FIVE methodology.
Thus, the FRSS issues are appropriately enveloped for CNS.
NRC Question A.4: l The JPEEE submittal indicated thatfires initiated in control cabinets wotdd be confined to the cabinet but evaluated the potentialfor damage to cabling above the cabinets by assuming afire located on top ofthe cabinet. The heat release rates (HRRs)for cabinetfires, an important heat transferparameter in determining the potentialfor damage, are not specified in the submittal.
In the Electric Power Research Institute (EPRI) Fire PRA implementation Guide, test resultsfor the control cabinet HRRs have been misinterpreted and have been inappropriately extrapolated l Cabinet HRRs as low as 65 BTU /sec are used in the Guide, in contrast, experimental work has
\ developed HRRs rangingfrom 23 BTU /see to 1171 BTU /sec.Considering the range ofHRRs that could be applicable to diferent control cabinetfires, and to ensure that cabinetfire areas are notprematurely screened out ofthefire risk analysis, a HRR value in the mid-range ofthe !
l 1
I
! ~ Attachment I ts l
, NLS990008 Page 10 of 31 currently available experimental data (e.g., 550 BTU /sec) is considered to be appropriate and should be usedfor the analysis. Discuss the HRRs usedin the CNSIPEEEfire assessment of control cabinetfires and the changes in thefire assessment results ifit is assumed that the HRR valuefor ^a cabinetfire is increased to 550 BTU /sec.
District Response to A.4:
i-The analytical approach used for evaluating fires initiated in control cabinets is described in detail in Section 4.6.9 of the IPEEE submittal [ Reference 9], within the discussions for fire areas 2A/2C (Reactor Building el. 903' 6", CRD units) and 3C/3D/3E (Reactor Building el. 932' 6",
i REC system).~ As described there, a phased approach was used. For clarification, the following expanded discussion of this phased approach is provided.
All motor control centers (MCCs) and control power panels analyzed are sealed and have sealed top penetrations. This precludes ventilation inside the cabinets and thus severely limits the L
oxygen supply for the postulated fire inside the cabinet.~ In addition, as described in the discussion on pages 4-13 and 4-16 of the IPEEE submittal, all of the power distribution j
. equipmem panels are installed using conduit entries only. For the purposes of the IPEEE fire i
_ analyses, conduit entries extending over at least 3 feet are considered sealed, in accordance with '
the guidance provided in the EPRI Fire PRA Implementation Guide. '
l Thus, any fire based on a burning'(o' xidation) process would be starved of oxygen and limited by the small amount of fire loading (cable insulation). Therefore the fire could not acquire a level of energy, which would enable it to propagate beyond the cabinet envelope. Electrical arcs,
!: initiated by locally increased resistance that may be caused by a degraded contact, are relatively unstable phenomena. They are typically self clearing, by welting offlocally a sufficient amount of metal to create unsustainable conditions for further arcing. Dr these reasons, fires in these 1
[ cabinets were modeled as being confined to the cabinet, and all of the components within the cabinet were postulated to be failed.
l However, for reasons of additional added conservatism, a fire at the top of the panel was L modeled, for the sole purpose of determining the sensitivity of the assumption regarding containment of the fire within the cabinet. A conservative assessment of the consequences of . ;
such a fire was performed using the FIVE "Inside of Plume" worksheet. For this sensitivity analysis, a 65 BTU /sec heat release rate was selected; higher values would not be appropriate for j the above reasons. :
l Given the use of conduits and sealed cabinets, in conjunction with a low amount of available combustibles in the configurations encountered in the cable spreading room at CNS, the use of a higher heat release rate as suggested by Question A.4 is not justified. As described in the IPEEE i submittal, a walkdown of these compartments verified that raceways, which do not terminate at L , the motor control panel (MCC) or panel in question, are located above the critical damage height.
I~ Consequently, it has been determined that the consequences of a postulated fire in an MCC or
Att:chment I to NLS990008 Page 11 of 31 l
power panel would be limited to failure of the affected MCC or panel alone.
NRC Question A.5:
The IPEEEfire assessment assumed thatfires initiated in power distribution panels located in the cable spreading room (CSR) would be retained in the cabinets. Although this assumption was usedfor all cabinetfires in the IPEEE submittal, the potentialfor damage to overhead cables was evaluated exceptfor the cabinets in the CSR. In addition, it is not clearfrom the submittal whether there is apotentialforfire propagation between adjacentpanels. The EPRI Fire PRA Implementation Guide assumes thatfire propagation to adjacent cabinets cannot occur ifthe cabinets are separated by a double wall with an air gap or ifthe cabinet in which the l fire originates has an open top. This can be an optimistic assumptionfor high-voltage cabinets since an explosive breakdown ofthe electrical conductors may breach the integrity ofthe cabinet and allowfire topropagate to combustibles located above the cabinet. For example; switchgear ;
fires at Yankee-Rowe in 1984 and Oconee Unit 1 in 1989 both resulted infire damage outside ;
the cubicles.
Provide the basisfor the assumption that allfires in the electrical cabinets in the CSR will be retained in the CSR cabinets and a quantitative assessment ofthe impact offire propagation fromfires in these cabinets.
l District Response to A.5:
The power distribution panels installed in the CNS cable spreading room are largely stand-alone cabinets, have enclosed, non-ventilated tops with conduit entries only. This is not common in most nuclear plants (e.g., open raceways), and it justifies the analytical treatment described in the CNS IPEEE submittal. The evaluation of these cabinets utilized the assumptions and guidelines cited in the EPRI Fire PRA Implementation Guide to reach conclusions regarding fire I propagation to other panels.
All MCCs and control power panels analyzed are sealed and have sealed top penetrations. This precludes ventilation inside the cabinets and thus severely limits the oxygen supply for the postulated fire inside the cabinet. In addition, as described in the discussion on pages 4-13 and 4-16 of the IPEEE submittal, all of the power distribution equipment panels are installed with conduit entries only.
Thus, any fire based on a burning (oxidation) process would be starved of oxygen and limited by the sr. ll amount of fire loading (cable insulation). Therefore the fire could not acquire a level of energy, which would enable it to propagate beyond the cabinet envelope. Electrical arcs,
! initiated by locally increased resistance, which may be caused by a degraded contact, are F relatively unstable phenomena. They are typically self clearing, after melting offlocally a
- sufficient amount of metal to create unsustainable conditions for further arcing. This melting-off j process always occurs, simply because electrical arcs impose temperatures which are above the e n, y e e- -- -
e , , , m --, - -,- q,,-- ,. w - -
,,--q
Attachment 1 to NLS990008 Page 12 of 31 melting temperatures of copper or brass.
Consequently, fires in these cabinets were modeled as being confined to the cabinet, and all of the components within the cabinet were postulated to be failed. This does not imply that, in case of an actual fire, there would be no visible symptoms outside of the cabinet and total absence of any secondary effects. However, propagation of a cabinet fire into adjacent cabinets was not postulated. The basis for this approach is the guidance provided in the EPRI Fire PRA Implementation Guide and in the FIVE methodology documentation. Based on the above, a quantitative assessment of fire propagation to neighboring cabinets would not be meaningful.
NRC Question f.6:
Fire severityfactors (SFs) were used in the analysis ofmany unscreenedfire compartments in the CNSIPEEE. In the case ofoil spills, valuesfor SFs were apparently obtained through a limited data analysis as recommended by the EPR1 Fire PRA Implementation Guide. The SFs for otherfire types appear to have been subjectively defined, but no basis wasprovided. The SFs were also used infire scenarios wherefIre suppression was credited. Since the potentialfor a largefire coidd be dependent upon one or morefire suppression agents, there appears to be a sigm'ficant possibility that the use ofafire SF whenfire suppression is modeledfor afire scenario (which determines afire type and afire si:e) may be inappropriate (a potential situation ofdouble credits).
For thefire scenarios where bothfire suppression andSFs were credited, ifany, discuss the appropriateness ofcrediting both. Also, various SFs used in the IPEEE arefound to be lower than typically observed SF estimates. Provide the basisfor the lower estimates ofSFs used in the JPEEE.
District Response to A.6:
In the CNS fire analysis, severity factors with values ofless than 1.0, together v.ith fire suppression capability, are invoked in the final quantification for Fire Zone 10B, Fire Zone 13B, and Fire Zone 20A.
In Fire Zone 10B [ Main Control Room], various control panel fires were analyzed. As described in detail in the submittal [ Reference 9], with additional detail provided in the response to Question A.13, some control panels required a specialized treatment methodology (overlapping zone approach), and propagation to neighboring panels was not postulated. Manual fire suppression capability was invoked. The basis for these scenario details is that the Main Control Room is occupied at all times by a full operating crew, with high level of attention.
In Zone 13B [ Noncritical Switchgear Room], a 10% propagation probability was postulated, and the Fire Zone was screened out. Additionally, a sensitivity calculation with a factor of four was performed, after which the Zone was still screened out. Manual fire suppression capability was
b Attichernt 1 to i NLS990008 Page 13 of 31
-invoked. The basis for this treatment is that the room is continuously manned with persons that are fire brigade trained.
In Zone 20A [ Service Water Pump Room], various large and small oil spill fire calculations were performed, as described in detail in the submittal. The basis for the treatment described in the submittal is that the room is equipped with special fire protection features: three independent i and diverse detection systems (flame sensors, smoke sensors, temperature sensors); spray shields !
' on the motors, to preclude oil from being sprayed in case ofleakage; a total flooding halon system; hand held fire extinguishers, etc. This pump room has received heightened attention and scrutiny in the past. The room presently contains one fire pump plus the four service water '
pumps, all electrically driven. '
. i Concerning the size of the fire severity factors used, it is not clear which fire zones might have ,
severity factors that are outside the industry norm. Therefore, a discussion of the severity factors
_ for the remaining zones listed in Section 4.6.9 of the submittal [ Reference 9] is presented.
+
Fire Zone 3C/3D/3E [ Reactor Equipment Cooling System] and Fire Zone 9A [ Cable Spreading Room] employ a severity factor for hot work that leads to a credible fire event. The value of this p factor is 0.1. Fire Zone 9B [ Cable Expansion Room] also uses a severity factor for hot work that leads to a credible fire event, but its value is 0.5. Since the contribution to CDF from hot work initiated fires in these zones is small, raising these factors to a value of 1.0 would have no etTect on the final screening results for these zones.
In the discussion for Fire Zone 3A [Switchgear Room 1F] and Fire Zone 3B [Switchgear Room 1G] a severity factor for fires that propagate from one switchgear to the other is presented, with a value of 0.1. However, as discussed in the submittal, this factor was not used to screen these zones. In Zone 2A/2C [ Reactor Building, CRD Units], a severity factor SF = 1.0 was used, and the zone was screened out.
In Fire Zones 7A [RHR-SW Booster Pump and Air Compressor Room],8A [ Auxiliary Relay Room],8B [RPS Room IB],8C [RPS Room 1 A],8D [ Seal Water Pump Area],8E [ Battery Room 1 A],8F [ Battery Room IB],8G [DC Switchgear Room IB],8H [DC Switchgear Room 1 A], and 12D [ Turbine Building North], a fire severity factor of 1.0 was used, and all of these
- zones were screened out.
NRC Question A.7:
The Fire-Induced Vulnerability Examination (FIVE) methodology requires that propagation throughfire barriers that do not meet thefire compartment interactions analysis (FCIA) criteria be consideredin the analysis. The IPEEEsubmittal does not always indicate affire spread through unscreened compartment boundaries was accountedfor in the IPEEE. Specifically, in the quantitative screening assessment ofcompartments with unscreened boundaries,
~ ...- . - - , . ~..- - , . . . . . . _ - - _.- -- ._.-. - . - ._- . . .-..~
l
-' Att:chment 1 to NLS990008 - I Page 14 of 31 it is unclear ifcomponents in the adjoining compartment (s) were damaged by thefire in a given
- compartment.
l
- Discuss the impact offire propagation through the unscreened boundaries on the quantitative screening assessment performedfor thefollowingfire compartments: IF, 2B, 2D, 4A/4C/4D, 5B, l ]
2E, and 13D. Ifpropagation' -induced component damage through these boundaries was not considered, discuss the impact ofinter-compartmentpropagation on thefire-induced CDFfor applicable compartments. ;
District Response to A.7: '
l L 1
L The fire zones that were not screened out, due to not meeting the FIVE boundary criteria, were evaluated for propagation to adjacent zones, as indicated in the FCIA, for each zone reviewed.-
']
The information requested in the above question for zones IF,2B,2D,4A/4C/4D,5B,2E, and '
13D is provided below. As discussed below, each of these zones has been individually N
- evaluated.
1
- l. J l
Fire Zone IF consists of the entire torus compartment. It is bounded on the top by the Reactor I Building floor (at elevation 903' 0"), and on the sides by the corner rooms and outside walls. All
{
of the boundaries, except for the ceiling, are adequate to satisfy the FIVE boundary criteria. The ;
FCIA for this compartment showed a potential for propagation of fires to Fire Zone 2A/2C (main area of the Reactor Building at elevation 903' 0"). In the quantification for this area all impacted '
systems were postulated to be failed, and a CDF value of 1.73E-07/ year was calculated.
However, this fire compartm' eut could not be screened out due to the need to consider 1
propagation to Fire Zone 2A/2C. The propagation to Zone 2A/2C was resolved by taking into ,
account the low amount of combustible loading in the area (48 BTU /sq ft), the lack of any
- significant ignition sources, and the limited access to the area. These considerations resulted in the conclusion that the boundary of this fire zone is sufficient to prevent the spread of any fire.
Fire Zone 2B consists of the Division 1 RHR heat exchanger room. The compartment is bounded at the bottom by the torus compartment, on top by the Reactor Building (elevation 958' 3"), and on the sides by other plant buildings (structures) and other Reactor Building walls. The majority.of the boundaries between this compartment and the other Reactor Building compartments satisfy the FIVE boundary criteria. The exceptions are the boundaries to 2A/2/C
' and 3C/3D/3E, which were not screened out due to unrated doors. Consistent with the approach taken in Zone IF, a bounding fire quantification was performed which yielded a CDF value of 1.29E-08/yr. This fire area has a low combustible loading (2,000 BTU /sq ft). This low loading and the lack of any_significant ignition sources resulted in the conclusion that the boundary of this zone is sufficient to prevent the spread of fire to other zones.
- Fire Zone 2D consists of the Division 2 RHR heat exchanger room. This compartment is bounded at the bottom by the torus compartment, on top by the Reactor Building (elevation 958' 3"), and on the sides by other plant buildings (structures) and other Reactor Building walls.
a e , - , , ,
l '
' Att:chment I to NLS990008 Page 15 of 31 The majority of the boundaries between this compartment and the other Reactor Building compartments satisfies the FIVE boundary criteria. The exceptions are the boundaries to 2A/2/C,3C/3D/3E, and 4A/4C/4D, which were not screened out due to unrated doors.
Consistent with the approach taken in Zone 2B, a bounding fire quantification was performed ,
l which yielded a CDF value of 1.30E-08/yr. This fire area has a low combustible loading (2,000 )
ll BTU /sq ft). This low loading and the lack of any significant ignition sources resulted in the conclusion that the boundary of this zone is sufficient to prevent the spread of any fire to j adjacent zones. '
Fire Zone 4A/4C/4D is in the Reactor Building at elevation 958' 3"- Fuel Pool Heat Exchanger / Lube Oil. The FCIA for this zone showed the potential for propagation to Zones 4B, 5A,3C/3D/3E, or 2D. The propagation into 4B or SA was determined to have no adverse impact due to.the lack of any Appendix R components in those areas. The remaining propagation paths were due to open stairways, piping penetrations, and the large opening at the southwest corner of this area. There is a potential that a fire could propagate to the upper elevations of the Reactor ,
l Building. This propagation was evaluated based on a detailed walkdown of this area, adjacent l areas, and the propagation pathways. The walkdown notes were then combinen with fire
! modeling of potential combustible materials (Reactor Recirculation Motor Generator (RR MG)
L set lube oil and cable insulation) and it was concluded that the credible fire events would not l propagate beyond the ' boundary' of this fire area.
l Fire Zone 5B consists of the main area of the Reactor Building west of column line N at elevation 976' 0". The area is bounded on the bottom by the building floor, on top by the Reactor Building refueling floor (elevation 1001' 0"), and on the sides by other plant buildings (structures) and outside walls. The fire compartment was not screened out due to a potential L . propagation path to 4A/4C/4D and 6. The propagation into fire Zone 6 was determined to have no adverse impact due to the lack of any Appendix R components in that area. Consistent with that described above for 4A/4C/4D, this propagation was evaluated based on a detailed walkdown of this area, adjacent areas and the propagation pathways. The walkdown notes were then combined with fire modeling of potential combustible materials (RR MG set lube oil and cable insulation), and it was concluded that the credible fire events would not propagate beyond
- l. the ' boundary' of this fire zone. It is noted that for this fire mo'deling, the large volume of the L Reactor Building at this elevation is considered sufficient to preclude the formation of a hot gas layer.
Fire Zone 2E consists of the steam tunnel. A propagation path was identified to Zone 12C, which would allow a postulated fire to involve a significant portion of the turbine building. This zone has a very low combustible loading (532 BTU /sq ft), only one potential ignition source (a single electrical cabinet - local instrument rack), and limited access during plant operations.
These considerations resulted in the conclusion that the boundary of this zone is sufficient to
- . prevent the spread of any fire to adjacent zones, d
j ' As an additional note, in an Appendix R evaluation performed in 1997, the fire barrier between h , - , , , _ - - - - - . , - - _ -
l 1
Att:chmext I to I NLS990008 - l 5 age 16 of 31 the steam tunnel and the turbine building was evaluated. It was concluded that this barrier meets
' the FIVE boundary criteria. This analysis was based on th < existing configuration of the plant.
Had the evaluation been perfonned prior to the CNS Fire IPEEE analyses, this area would have been screened out in the Phase I analysis.
l L- '
Fire Zone 13D consists of the I&C shop area and two adjacent fire zones. This fire zone was included in the quantification due to the potential for spread of the fire into area 12F,13A, and l! 13D. In the quantification of this fire zone reported in the submittal, all of these zones were L considered affected and the resulting calculated CDF value 4.36E-08/yr includes the propagation
- of this fire.-
~ NRC Question A.8:
j The unreliability estimatesfor automaticfire detection system (AFDS) and the suppression systems (SS) provided in the FIVE methodology were used in the JPEEE submittal. Section 4.5 ofthe JPEEE indicates that use ofthis data is appropriate because the AFDS and the SS at CNS are designed and maintained in accordance with appropriate industry standards, such as the 1 National Fire Protection Association (NFPA). However, an USNRCInspection [ Reference 4]
\5 has indicated that not all AFDS and the SS at CNS meet NFPA standards.
l l~
Discuss the impact ofthe above non-conformance to NFPA standards on the unreliability estimates ofthe AFDS and the SS and the resulting CDFfor thefire compartments in the Reactor Building containing the nonconforming AFDS and the SS. Alternatively, provide a basis
~for the lower unreliability estimates in the JPEEE ofthe AFDS and the SS which are based on the NFPA standards. 1 l District Response to A.8: l L j The statement [on Page 4-7 of the IPEEE submittal, Reference 9] "The fire detection and
. suppression systems at CNS are designed ... in accordance with appropriate industry standards, i such as NFPA '..."is in error. The following excerpt from the USAR is provided for background information, and to clarify the original design standards used at the Cooper Nuclear Station in the '
area of fire detection and suppression systems:
"The primary standard for the design criteria for the overall fire protection program at CNS was the ;
'NELPIA Guide to Basic Fire Protection for Nuclear Power Plants, dated April,1968'. This program consists of the following types of fire protection:
'l. Automatic and manual fire detection systems throughout areas of the main plant, I, with both local alarm and remote audible and visual annunciation in the Control
- Room.
h 2. Manual and automatic fire extinguishing systems located in the main plant, with L
L, __ _ _ _ _ __ ,_
- =
u
)
L i
! . Attichm:st 1 to l L NLS990008 l Page 17 of 31 l
L local alarm and remote audible and visual annunciation in the Control Room.
- 3. Appropriate Administrative Control Procedures developed to insure the y availability of these systems, should they be needed.
l
~
f- During the design and construction stages, plans and specifications for the numerous fire protection
~ l systems were submitted to NELPIA for their review, comment, and approval. Monthly site inspections L : were scheduled to permit the NELPIA field representative to survey the progress of the plant ;
- construction and fire p;otection systems installation, with a timely review of plant design modifications ;
i as they occurred with'the necessary modification of, or addition to, the fire protection program as deemed necessary."
CNS has been designed, built and licensed utilizing the guidance of NFPA standards for fire detection systems and fire suppression systems available at the time of construction, in addition i
to plant specific guidance provided by NRC reviewers. The NRC reviewed, and then either approved the systems or recommended changes. Changes to the automatic fire detection systems and the fire suppression systems were made with prior NRC approval. Concurrence of system adequacy was provided by the NRC in the late 1970s. Since that time, no changes have been l made to the systems in question without NRC approval. The Inspection Report 50-298/96-25 3
- cited above identifies the fact that the systems as found during the inspection do not conform with the then-current version of the NFPA standards.
Regarding the unreliability values provided in the documentation of the FIVE methodology, l these are generic, conservatively bounding values for pumped water systems, which can also be l obtained from other publications. As such, these values are not based on or related to L conformance of the system to any specific version of the fire protection standard. During the L - IPEEE walkdowns, which were performed for evaluation of the fire detection and suppression systems, no deficiencies or outliers were identified. Consequently, the unreliability values provided in the IPEEE submittal were judged appropriate.
NRC Question A.9:
In general, thefire risk associated with a given compartment is composed omntributionsfrom
, fixed and transient ignition sources. Neglect ofeither contribution can lud to an underestimate j ofthe con,partment 's risk and, in some cases, to improper screening offire scenarios. Further, I the presence oftransient combustibles can also impact the potentialforfire propagation and b component damage. The IPEEE appears to have eliminated transient ignition sources in some areas based on observations during the walkdowns. In addition, thefire risk assessment appears to have only considered the amounts, types, andlocations oftransient combustibles identified i during the plant walkdowns. The assessment ofthe transientfire risk based on a one-time
[ ' observation is questionable. i r
4 l
l - _ _ _ .. _-
Attichment I to NLS990008
,- Page 18 of 31 Provide an evaluation ofthefire risk in unscreenedfire compartmentsfrom transient combustibles and ignition sources taking into account the ignition sources that would be allowed in the compartments. This evaluation shouldinclude the amount oftransient combustibles that may be present in the compartments at various times, including thepotentialfor violation of transient combustible controls, which have occurred at the CNS [ Reference 4].
District Response to A.9:
In Inspection Report 50-298/96-25, referenced above, programmatic issues regarding 10CFR50, Appendix B " Quality Assurance Criteria...," Criterion XVI " Corrective Action" were identified.
These issues included, but were not limited to, revising Procedure 0.7.1 " Control of Combustibles," and interim plant walkdowns for identification of potential problems with control of transient combustibles {see Reference 10]. Specifically, the substance of the issues identified was of a general housekeeping nature, related to a generic lack of compliance with Procedure 0.7.1, rather than actual fire risk issues. The transient combustibles identified in this Inspection Report were evaluated against the Fire Hazard Analysis. It was determined [NLS970125, Reference 11] that in all cases they were enveloped by the limits imposed by the Fire Hazard Analysis. Thus, credited safe shutdown equipment was protected at all times. Procedure 0.7.1 has since been revised, and the programmatic issues have been addressed.
In the CNS IPEEE fire analysis, the contribution of transient combustibles to the total fire risk has been individually determined for each Fire Zone in which a calculation was performed. In some Fire Zones, the contribution from transient combustibles was enveloped by a postulated exposure fire. The amount of transient combustibles postulated in the calculations was based on data gathered from plant walkdowns, as described on pages 4-3 and 4-4 of the IPEEE submittal.
Also, where appropriate, combustible loading limits specified in the Fire Hazard Analysis were consulted. Additionally, in order to assure conservatism, a trash can fire was modeled in all screened and unscreened fire zones to determine if the screening was appropriate and to add to the fires considered in the unscreened zones.
Exceedances of administrative transient combustible limits are required to be properly evaluated and documented. Exceedances are administered and controlled by Plant Procedure 0.7.1
" Control of Combustibles." In case of an exceedance of an administrative limit, compensatory measures are invoked if appropriate, and if necessary, applicable portions of the Fire Hazard
. Analysis are examined to assure that potential fire challenges are minimized.
NRC Question A.10:
The JPEEE submittal does not indicate whether cables in the CNS are Institute ofElectrical and Electronic Engineers (IEEE)-383 quahfiedcables. The submittallists damage criteriafor both qualified and non-quahfied cables implying that both exist at the CNS; however, insufficient information was provided to determine ifunquahfied cablefires were included in the initiating event (IE) frequencies usedin the quantitative screening. Furthermore, thefire modeling did not
I Att:chmnt 1 (2 NLS990008 Page 19 of 31 include unquahfied cable-initiatedfires.
. Discuss ifquahfied1EEE-383 cables are currently used at the CNS and offires initiated by unquahfied cables were included in thefire 1Efrequencies used in the screeningfire assessment.
Ifunquahfied cables were excludedfrom thefire IEfrequencies, justify this exclusion and discuss the impact ofthe exclusion on thefire-induced CDF. Provide the t) pes ofcables treated as targets in the detailedfire assessments performedfor each ofthe unscreened compartments.
District Response to A.10:
Cooper Nuclear Station was constructed before the Standard "IEEE 383-1974" had become a consensus standard. In accordance with Branch Technical Position 9.5-1, Appendix A, CNS l submitted a letter on December 17,1976, which detailed the specifications and testing performed on cable types present in safety related areas of CNS, in order to demonstrate the technical basis for equivalency to IEEE 383-1974. This letter was accepted in the SER for Fire Protection, associated with Technical Specification Amendment 56, dated May 23,1979. Thus, the established equivalency to IEEE 383-1974 was the basis for utilizing the qualified cable data in the IPEEE fire analysis.
The above equivalency qualification applies to all of the cables used in safety related areas of CNS. Therefore, no ' unqualified' cables are present, and the CNS fire IPEEE analysis did not have to include any specific analytical treatment of unqualified cables. The only area where there might have been a concern in this direction is the Turbine Building (non-safety related cables). However, in the quantitative analysis for the Turbine Building, the total loss of all systems was conservatively postulated in that area. Additionally, cables were included in the ignition source calculations for their respective areas. In the CNS Fire IPEEE, this means that the ' equivalency qualified' cables andjunction boxes are treated as ignition sources which are factored into the fire frequency estimates.
NRC Question A.11:
The heat lossfactor (HLF) is defined as thefraction ofenergy released by afire that is transferred to the enclosure boundaries. This is a keyparameter in the prediction ofcomponent damage, as it determines the amount ofheat available to the hot gas layer (HGL). In FIVE the HLF is modeled as being inversely related to the amount ofheat required to cause a given l temperature rise. Thus,for example, a larger HLF means that a larger amount ofheat (due to a more severefire, a longer burning time, or both) is needed to cause a given temperature rise. It can be seen that ifthe value assumedfor the HLF is unrealistically high, fire scenarios can be l improperly screened out. Figure 1 ofReference 5 provides a representative example ofhow l HGL temperaturepredictions can change assuming different HLFs. Please note that: (1) the curves are computedfor a 1000 KWfire in a 10m x Sm x 4m compartment with aforced i
ventilation rate of1130 cfm; (2) the FIVE-recommended damage temperature is 700 Ffor quallfled cable an 450 Ffor unquahfied cable; and, (3) the Societyfor Fire Protection Engineers l
l
_. .. m . . _ _ _..._ . .__ _ _. ._ _ _ _ _ _ __. _ _ _____ ._ __ .__
~ Attrchmert 1 to
- NLS990008
. Page 20 of 31
, (SFPE) curve in Figure 1 is generatedfrom a correlation provided in the SFPE Handbook
[ Reference 5].: ,
Based on evidence provided by a 1982 paper by Cooper, et al. [ Reference 6], the EPRI Fire PRA Implementation Guide recommends an HLF of 0.94forfires with durations greater thanfive
- minutes and 0,85for "exposurefires awayfrom a wall and quickly developing HGLs. "
However, as a generalstatement, this appears to be a misinterpretation ofthe results. Reference
- 6, which documents the results ofmulti-compartmentfire experiments, statedthat the higher HLFs are associated with the movement ofthe HGLfrom the burning compartment to adjacent, cooler compartments. Earlier in the experiments, where the HGL is limited to the burning compartment, Reference 6 reports much lower HLFs (on the order of0.51 to 0. 74).
, These lower HLFs are more appropriate when analyzing a single compartmentfire. In summary, (a) HGL predictions are very sensitive to the assumed value ofthe HLF; and (b) large HLFs cannot bejustifiedfor single-room scenarios based on the information reference in the EPRIFire PRA Implementation Guide.
For eachfire scenario where the HGL temperature was calculated, specify the HLF value used in the analysis. In light ofthe preceding discussion, either: (a) justify the value used and discuss i its effect on the identification offire-included vulnerabilities, or (b) repeat the analysis using a morejustifiable value andprovide the resulting change in scenario contribution to CDF.
District Response to A.11:
The EPRI Fire PRA Implementation Guide presents a wide range of heat loss factors. However, in all cases where a heat loss factor was used in the various fire analyses for the CNS IPEEE, a value of 0.70 was utilized.
- This value is recommended in the guidance documentation for the FIVE methodology. It is appupriately conservative, and it is within the range of the experimental results described in the 1982 paper by Cooper et al. ["An experimental Study of Upper Hot Layer Stratification...," :
Reference 6]. Thus, the heat loss factor used in the fire analyses for the CNS IPEEE is based on conservative data and is experimentally justified.
- NRC Question A.12
i The JPEEE submittal indicated that non-Appendix R (Title 10 CFR, Part 50) systems were credited in protecting againstfires in the upper elevations ofthe Reactor Building (RB) without
} an explicit examination ofthe associated cabling. This was done based on discussions with
- j. plant personnel that the non-Appendix R equipment (identified as balance-of-plant equipment) is
- the nor
- -Appendix R systems credited in the CNS IPEEE consisted ofoffsite power system and the i main condenser system. Operation ofthe main condenser as a decay heat removal system i ,
1 y + . . - . e.. - ,m 4- . - ,.,m-- ,,.- y ..e-- -
. Att chme:t 1 to
' NLS990008 -
Page 21 of 31 - .
l requires operation ofmany support systems including cooling water systems, non-Class IE power source, and instrument air. In addition, successful operation also requires that the main steam isolation valves (MSIVs) remain open.
Provide the basis that applicable support systems requiredfor operation ofthe main condenser would not be impacted byfires in the upper RB. In particular, verify thatfires in the RB compartments would not impact systems whosefailure would result in an MSIV closure signal, and would not affect CNS operator ability to reopen the MSIVs, ifclosed, due tofires.
District Response to A.12:
l As further expanded on in the response to Question A.7 above, an evaluation of the vulnerabilities of the support systems was performed for those BOP functions, which were assumed to be available for the scenario being analyzed. In cases where the fire would impact the MSIV controls or any part of the required power conversion system (PCS), heat removal via the main condenser was postulated to be unavailable, and no recovery (i.e., reopening) was ,
postulated for the duration of the fire scenario. The resulting PSA model quantification for these areas reflects these assumptions.
Loss of MSIVs (PCS function) was postulated for fires in the following zones: l 3A - Switchgear 1F Equipment Room 3B - Switchgear 1G Equipment Room 4A/4C/4D - RB Elevation 958' 3" 7A - RHRSW Booster Pump / Air Compressor Room 8-1 - Turbine Building, General Area 12D - - Turbine Building, North 13B - Non-Essential Switchgear Room 13C - Electrical Shop Thus, fires in all areas where loss of MSIV function may occur have been properly treated in the CNS fire quantification.
NRC Question A.13:
The IPEEEfire assessment ofthe control room indicated that the analysis offires in two panels l were treated uniquely. One controlpanel (9-3) has no internal barriers, andfires that were l postulated were assumed to impact onlyportions (or sections) ofthe panel. Another panel (Board C) contains partial internal barriers that were assumed to be effective in preventingfire propagationfrom one section to another section in thepanel. Creditingpartial barriers and/or its equivalent aspart ofprevention offire growth within the panel is questionable in afire risk e analysis.
l l
l
. . . . ~ .4 . _ _ -. _ . _ . .- - - . _. _ _ _ , _ _ _ _ __. . _ _ . _ . ~
' Attschment 1 (2 NLS990008 -
Page 22 of 31 Provide an assessment ofthe CDFfromfires in these twopanels that couldpropagate
~ throughout the panel.
District Response to A.13:
2 In the case of Control Panel 9-3 and Control Board C it had been determined that, based on the -
geometric configuration (extreme ' width), alternative evaluation techniques can be used, as i x described in the following..
- As discussed in Section 4.6.9 of the IPEEE submittal [ Reference 9, page 4-27], Control Panel 9-3 contains the controls for essentially all of the ECCS components and the MSIVs. If a nonmechanistic, " horizontal" fire were postulated, covering the entire panel, a reactor trip with loss.of the power conversion system would occur, since the MSIV controls are located in this panel.'
The analysis performed of this panel consisted of consideration of the overlapping zones as ;
illustrated in the submittal. The panel has a geometry oflong width versus height. Therefore, in order to fully involve the cabinet, significant horizontal fire spread would have to occur. This .
, was not considered realistic. The overlapping zone approach allowed for any one fire to spread to ;
a damage impact zone of approximately 3/4 of the panel, given suppression within 15 minutes.
The quan'tification for this overlapping zone approach involved four separate PSA ,
quantifications. The most limiting one of these cases resulted in a total CDF value of 3.39E- 1 7/ year. l The approach described ab ve is considered appropriate, considering the continuous presence of the control room crew, the lack of combustibles, the lack o' fignition sources in the panel and the unique panel geometry. The conclusion is that fires would be slow to develop and would receive l rapid operator response and, thus, suppression prior to significant spreading would be reliable.
The analysis of hot shorts connected with Panel 9-3 is' discussed in the response to Question A.l.
Control Board C Control Board C was treated with the same calculational methodology (the overlapping zone approach) as Control Panel 9-3 above. As described on Page 4-27 of the submittal [ Reference 9],
this treatment of Board C is considered appropriate for fully meeting the intent and purpose of
' the IPEEE analysis. The quantification of Board C in the manner described above resulted in similar calculated CDF values.
L
' Fires resulting in a loss of all offsite power were recognized as an insight in the CNS IPEEE. As l
. part of the implementation of Severe Accident Management at CNS, this scenario (Fires in Control Panels C and F) was further investigated. Using a conditional probability of P = 1.0E-1
.' j j
- - = - ~ . - - - . . . . .- - . . - - . .
l Att:ch'mert I ta .
NLS990008 Page 23 0f 31 i
for fire spread throughout the panel and using the full PSA model, a CDF value of 2.18E-7/yr was calculated for this scenario. In the course of the Severe Accident Management (SAM) project it was determined that this figure is sufficiently small, that no changes to the plant are warranted to address these particular fires. l l ~ B. Seismic Events NRC Question B.1:
System analysisfor Safe Shutdown Equipment List (SSEL) development is discussed in Section 3.1.2 ofthe IPEEE submittal. For success path andsystem selection, EPRINP-6041-SL states \
\
that "in general, the selectedpathforperforming the safetyfunctions to shutdown the reactor will be the one consisting ofthefront-line systems (and their necessary support systems) that were provided as a 'first line ofdefense', and designed to respond automatically (at least in the short time during and after the seismic margin earthquake) to the types oftransients and/or ,
accidents that might be induced by a margin earthquake. " Based on this industry criterion, the l
high pressure injection systems (e.g., the high pressure coolant injection (HPCI) system and the '
reactor core isolation cooling (RCIC) system) seem to provide a choicefor coolant injection to the vessel. -However, the above listed highpressure injection systems are not included in the CNS SSEL, and the lowpressure coolant systems (e.g., the lowpressure coolant Injection (LPCI) system and the core spray (CS) system) are usedfor both successpaths. Afanual reactor pressure vessel (RPV) depressurization is therefore requiredfor both paths, and consequently, the demands on the depressurization system and operator actions are significant.
RPV depressurization is provided at CNS by the operation ofthe safety reliefvalves (SR Vs) usingpneumaticpowerprovided by accumulators. Since theplant nitrogen / instrument air systems, which supply the accumulators, are not included in the SSEL, accumulators are the only pneumaticpower sourcefor the SR h. Hence, their performance (applicable current design requirements and unreliabilityper demand considerations)for transient and small LOCA events under seismic margin earthquake (SME) conditionsfor a mission time of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> needs to be evaluatedand discussed.
According to the JPEE'E submittal, successful RPVdepressurization requires the operation of three SRVs, and theprobability of"operatorfailure to depressuri:e with 3 SRVs"is 4.2E-2.
Failure to depressurize the RPVin a timelyfashion may therefore be a sigmficant contributor to
, failing to place the plant under a safe shutdown conditionfollowing a seismic margin earthquake (SME) evem.
Basedon the above considerations:
I
- a. Discuss how the EPRINP-6041-SL criterion was appliedfor selecting appropriate l systemsfor the CNS SSEL, quoted above, in the CNSIPEEE discussion. Provide also the basisfor not including the HPCIsystem and the RCIC system in the CNS SSEL (other
- Att:chment 1 if R NLS990008 L Page 24 of 31 l :
- than the unreliability considerations ofthese systems based on industry experience)
- b. Based on CNSprocedures, describe the expected operator actionsfollowing an SME event. Discuss in detail the operator actions and theirfailure probabilitiesfollowing an SME eventfor RPVdepressurization, including the ability ofthe SR V accumulators to i provide pneumaticpower suf]Icientfor SR Voperationfor mission time of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
J
- c. ' List all major equipmentfor the HPCIsystem and the RCIC system and describe any
. weak links in these systemsfollowing an SME event.
L District Response to B.1.a:
The purpose of the Seismic Margin Methodology (SMM) evaluation in the IPEEE is to provide l . assurance that a seismically rugged success path for plant shutdown exists, commensurate with the postulated, extreme review level earthquake (RLE) conditions, which are beyond the design
- basis (Safe Shutdown Earthquake) conditions. The equipment included in the IPEEE SSEL has
- . undergone a special seismic ruggedness evaluation. However, it would be incorrect to assume that any equipment not included in the IPEEE SSEL is necessarily seismically insufficient or
" unqualified" to be utilized during a seismic event.
4 The statement from NP-6041-SL cited above is not a selection criterion; it is a general observation. There is no criterion or requirement which would favor the high pressure injection systems over the low pressure injection systems or vice versa, and both types of systems are fully qualified frontline systems for heat removal and coolant inventory control. For any piece of equipment, not being on the IPEEE SSEL merely means that it has not been selected for the RLE ,
SMM success path and therefore has not undergone the additional RLE ruggedness analysis.
Hence, having selected the low pressure injection systems for the SMM success path does not imply that the high pressure injection systems would be expected to malfunction during an RLE levent.
No basis exists for excluding specific systems (e.g., the high pressure injection systems) from the IPEEE SSEL. Instead, the rationale for selecting the low pressure injection systems for the .RLE
- success path was derived from the intent to have the ADS included in the success path analysis, based on the conservative reasoning that depressurization may be operationally desirable during a postulated RLE scenario, and therefore should be analyzed in that context.
' District Response to B.I.b:
It would be incorrect to assume that manual depressurization of the reactor vessel is required in order to enable the low pressure injection systems to function. Comparing high pressure injection with low pressure injection for the success path, both paths function in an automatic fashion, ifleft to themselves. If high pressure injection were to malfunction when required to D operate during a seismic event, the automatic' depressurization system (ADS) would act
i Attrchm:nt I ta NLS990008 Page 25 0f 31 automatically to depressurize the reactor vessel, in order to accommodate low pressure injection systems. The fact that the control room operators have the option ofinhibiting ADS action, for the case that they might prefer manual control of reactor pressure, does not affect existing automatic system features.
As indicated in the IPEEE submittal, a sensitivity calculation was performed with the CNS PSA model to assess the importance of post-seismic operator actions. This sensitivity calculation provides a general assessment of the importance of operator actions, with no regard to the type of coolant injection (high pressure versus low pressure) being used. Quantification of the important seismic event sequences is described in Section 3.1.2.1.8 of the IPEEE. Special screening values (human failure probabilities which are intentionally overstated by two orders of magnitude, i.e.
by a factor of 1.0E+2) were used for operator actions in order to elucidate the importance of these actions to post-seismic plant response and to obtain an upper bound value for their potential effect on the CDF value.
The result of this sensitivity study is that the calculated CDF increases very little (about 2 %).
Based on this evaluation, it has been concluded that post-seismic operator actions are not dominant for the seismic risk at CNS. The quantification provided in the CNS IPEEE submittal regarding the'effect of operator actions is conservative and bounding.
The difference in demand for operator action while using high pressure versus low pressure injection systems is not significant. Instead, operator actions will mainly be dependent on the steps required first by the Earthquake procedure, and then by either the Normal Shutdown procedure, or the Emergency Shutdown procedure, depending on the situation. All system alignments for safe plant shutdown are performed from the Main Control Room using existing plant procedures.
With regard to the number of relief valves required for depressurization of the reactor vessel, some clarification is required. The thermo-hydraulic transient calculations performed for the IPEEE have been reviewed, and they show that not three SRVs. but only one SRV is required for RPV depressurization. Although depressurization with one valve is not as rapid as it is with three valves, the criterion for system success is not affected.
When the IPEEE submittal was written, the prevalent approach was to utilize screening values wherever possible, for the purpose of analytical simplification and also for the purpose of retaining additional added conservatism in the analytical model, with the expectation of facilitating utilization of the model in an enveloping fashion for various additional analytical applications and/or scenarios, which might emerge at a later point in time. For this reason, the screening value of"three SRVs required" was used in the IPEEE model. providing an overstated probability for failure to depressurize.
Regarding the duration of accumulator operability for the six ADS valves and the two Low-low-set (LLS) valves, the latest test data set was reviewed. These data show that three of the six ADS
Att: chm::t 1 ta NLS990008 Page 26 of 31
~
valve accumulators and both accumulators for the LLS valves have an operability duration. i without the instrument air supply, in excess of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Consequently, five valves would be i
. available to preclude repressurization'of the RPV during the postulated 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> scenario, with a minimum of one valve required. With these data it can clearly be stated that sufficient redundancy is in place regarding accumulator capacity for assuring that repressurization of the RPV is precluded during the mission time of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
i District Response to B.I.c:
- Item B.I .c is requesting a seismic ruggedness evaluation of the HPCI and the RCIC systems.
However, this is unnecessary because considerable conservatisms are already contained in the selected safe shutdown path, as discussed in the responses to Items B.l.a and B.l.b above. '
Consequently, the District has not identified any need to perform further seismic evaluations of, additional systems at CNS.
i NRC Question B.2:
. Submit better copies ofFigures 3.1.2 and 3.1.3 ofthe JPEEE submittal, tofacilitate timely '
completion ofthe staffs review ofthe seismicportion ofthe IPEEE.
District Response to B.2:
The requested copies of the Figures are provided as Attachment 2. They are provided in an enlarged format, one graph per page, for enhanced legibility.
NRC Question B.3:
What are thefindings ofthe resolution of USI A-46, and the IPEEE-identified weak equipment, if any, which have a seismic capacity below the 0.3 gpeak ground acceleration (pga) review level earthquake (RLE)? Discuss the estimated overall high confidence oflowprobability offailure (HCLPF) capacity ofthe weakestplant equipment based on completion ofthe USIA-46 resolution program and the CNS IPEEEprogram District Response to B.3: !
The original USI-46 information relevant to this question has been provided to the NRC in the l
A-46 submittal. Please refer to correspondence NLS960076, submitted on June 13,1996 i
[ Reference 13]. Information on the A-46 resolution program is currently in the process of being l finalized for submittal. I
- The Table in Section 3.1.1.5 of the IPEEE submittal [ Reference 9, page 3-6] lists six items which
. were identified as " seismically weak", i.e., having a HCLPF value below the postulated maximum peak ground acceleration value. Of these, the first five are on the A-46 SSEL, and ;
1 l
Att:chmnt I t2 NLS990008 Page 27 of 31 outlier resolution on these has been completed. The sixth item, which is the fan coil unit in the SE and the NE quads, is not on the A-46 SSEL. However, since plant shutdown can be accomplished with one RHR pump, these fan coil units are actually not needed for safe plant shutdown. It has been determined that room cooling is not needed for running only one RHR pump.
C. High winds. flood L _and other external events (HFOs)
NRC Question C.1:
Section 5.4.1 ofthe CNSIPEEEprovided a discussion on aircraft ha:ardsfrom nearby airports and its impact on plant safety. It is the staff's understanding that, aspart ofthis evaluation, the licensee has made use ofold air aviation data (e.g.,1974 data). NUREG-1407 guidance requested that the JPEEE should make use ofthe recent historical aviation data and sigmficant changes in air traficproblems, ifany, as part ofthe aircraft crash eventfrequency estimation process. Discuss whether the IPEEE has made use ofrecent data, (i.e., datafor the period between the OL date of 01/18/1974 and 12/1990) on sigm*ficant changes in air trafficpattern (e.g., increase in number oflandings and takeoffsperyear at nearby airports.)
District Response to C.1:
Contrary to the above, the IPEEE has made use of recent aviation data. The IPEEE submittal did not make it clear that the nearest airports, Rock Port Municipal Airport and Auburn Municipal Airport (Farington Field), as well as the air routes near the site had been evaluated, specifically for the CNS IPEEE, using the most recent information obtained from the airports and from the Federal Aviation Administration (FAA).
There is no major commercial airport with a control tower within 80 Km (50 mi) of the plant site.
The nearest major airports are located in Omaha, Nebraska, in Lincoln, Nebraska and in Kansas City, Missouri. These commercial airports are more than 105 Km (65 mi) from the plant site. In addition to these, there are several small local airstrips within 16 Km (10 mi) of the Cooper Nuclear Station.
The Rock Port Municipal Airport is located in Rock Port, Missouri, approximately 8 Km (5 mi) from the plant. The airport has one paved runway but no building facilities or communications.
It is used only by small aircraft and has not experienced any activity since the flood in the summer of 1993. At the time of the IPEEE evaluation, the airport was scheduled to reopen sometime in 1995. It has subsequently been reopened.
The Aubum Municipal Airport is located near Auburn, Nebraska, approximately 13 Km (8 mi) from the plant. The airport has a grass runway and experiences very light small aircraft traffic (approximately three to four flights per week).
c
-. _ m _ . _ . _ . _ _ _ _ . - . _ - _ ._ . . ._ _ _ _ _
J Att:clim:nt 1 ts ,
NLS990008 Page 28 ef 31 Local air traffic near the plant site does not present a significant hazard for the following reasons:
There are no airports within 16 Km (10 mi) with projected operations greater than (500)D2,
' nor are there any airports outside.of a 16 Km (10 mi) radius having projected operations
. greater than (1000)D2 (where D equals the distance in miles from the plant); and
( There are no military facilities within 16 Km (10 mi) of the plant. I The CNS design analysis had not previously addressed the low altitude and high altitude
. commercial flights on the air routes proximate to the plant site. Therefore, these items may not be screened out per NUREG-1407. Accordingly, an evaluation of the aircraft hazard was performed for the IPEEE. This evaluation is documented in an engineering calculation, which !
, covers the four '
air routes. l l
The low altitude airways VR540/541 and V216 pass to the . North and South of the plant, ;
respectively. The high altitudejet routes J64 and J192 also pass to the North and South of the plant, respectively. Table 5.3-4 of the calculation provides a description of these airways o including their width, centerline distance from the site, and annual traffic volume. For the l purpose of this evaluation, the high altitudejet airways were also assumed to have a width of l 14.8 Km (8 nautical miles), although the FAA does not assign a specific width to these airways.
1 Results of this engineering calculation show that the annual frequency of an aircraft accident, l
. which could affect the site, is around f = 5.0E-8/yr, based on the methodology outlined in SRP Section 3.5.1.6: This value corresponds to an annual probability value of P = 1.0 - exp[5.0E-8] =
5.0E-8 which is, by a factor of 20, below the credibility value of 1.0E-6, as specified in National
' i
-- Standafd ANSI /ANS 52.1-1983 (" Nuclear Safety Criteria for the Design of Stationary Boiling )
Water Reactor Plants"). It is therefore concluded that aircraft traffic, based on latest available j
. data, does nat pose a significant hazard for the Cooper Nuclear Station.
NRC Questiers C2:
Section 5.5.Iprovides a discussion on safetyproblems related to lightning hazards at the CNS.
Discuss the roh and importance ofstation batteries and any other emergencypower equipment needed in shtetting down the CNS during severe lightning events. '
i
' District Response to C.2:
1
~ ' Although the lightning issue had previously been evaluated for CNS in NUREG/CR-4767,
" Shutdown Decay Heat Removal Analysis of a General Electric BWR 4/ Mark 1", Section 5.5.1 of the IPEEE [ Reference 9) states, in part, that "due to the fact that the lightning design basis for )
.the CNS control building is not currently well documented..., the District is considering a review
. of the potential vulnerability of the control building..." The intent of this statement was to identify the need for further evaluation of this issue. However, in revisiting this issue, the o ,
, ,, . . . . , . ~ , . , . - _ . . . , , , - - , ,. . --.
i Att:chment 1 to NLS990008 Page 29 0f 31 l
j determination was made that there is no basis for requiring lightning protection equipment on the Control Building.
In order to gain plant specific information on this issue, walkdowns were performed on every level in the Control Building and in the adjacent Multipurpose Facility, including the roofs o!
both buildings. Applicable structural drawings (4501 through 4506) were reviewed.
The Control Building houses the Service Water Booster Pumps and service air compressors at the basement level (floor elevation 882.5 ft). At the mezzanine level (floor elevation 903.5 ft) it contains the battery rooms and DC switchgear rooms. The East Battery Room (Room 1B) contains the Division 2 batteries (125 V and 250 V), and the West Battery Room (Room l A) ,
contains the Division 1 batteries (125 V and 250 V). On the next higher level (at floor elevation '
918.0 ft) is the Cable Spreading Room, and above it (floor elevation 932.5 ft) the Main Control Room. The roofis at elevation 949.13 ft. The parapet top rim is at elevation 954.2 ft.
l The Control Building has no external lightning protection equipment mounted on it. Given the structural design features and configuration of the plant, no such equipment is needed. The Control Building (upper parapet rim at 19.6 m above grade level) is protected by umbrella effects from the Turbine Building (upper parapet rim at 36.8 m above grade) and the Reactor Building (upper parapet rim at 49.6 m above grade level, more than twice the height of the Control Building), which in turn is protected by the umbrella effect from the much taller, nearby elevated release point (ERP) tower (about 99 m tall). Furthermore, because of an exceptionally dense embedded steel grid, the building is functionally a Fareday cage, as further described below. 1 The Control Building is a tornado proof structure, constructed of heavily reinforced concrete, with a dense grid of steel rebar, which is fully grounded. The flat roof of the building is 2 ft thick, designed to function as a tornado missile barrier. It is covered with a rubber mat, to protect the concrete against weather effects. All exterior walls have a minimum thickness of 1.5 ft. The interior walls and floors are generally about 1.0 ft thick. The floor of the Main Control Room is 0.67 ft thick, while the separation wall between the Division 1 and Division 2 Battery Rooms is 2.0 ft thick. The dense grid of grounded steel rebar provides a Faraday Cage effect, while the batteries, with the DC system being ungrounded, do not offer a viable target point for a lightning strike.
The exterior walls of the Control Building are not exposed to the outside. The North wall is covered by the Multipurpose Facility, the East wall by the Turbine Building, the South wall by the Reactor Building, and the West wall by the Radwaste Building. The adjacent Multipurpose Facility is a steel structure, built of heavy I-beams, metal walls and metal roof. Its entire structure is grounded.
The EPRI Ilandbook (" Grounding and Lightning Protection," December 1991) states that
- structures surrounded with steel sheets of at least 4.8 mm (3/16 inch) thickness do not need lightning protection equipment if the steel sheet is adequately grounded. For instance," grounded l
- Att:chmut I ta
~ NLS990008 '
Page 30 of 31
~
tanks containing flammable liquids or liquefied petroleum gas under pressure do not require
- lightning protection," if the tank is at least'4.8 mm thick and adequately grounded [EPRI Handbook, Page 5-17]. The upper 15 m of the Reactor Building are clad with corrugated steel siding which is adequately grounded. It is mounted on a grounded steel frame, and thus serves as ,
+-
a large-area receptor for any lightning bolt which is not captured by the ERP Tower. The Turbine Building also is clad with grounded metal siding (generally above an elevation of 8 m to 9m).
- All of the high voltage power lines coming into the 345 KV switchyard are equipped with a grounded shield wire, which runs above the three phase conductors, at an estimated 30 m ;
elevation above grade level. These shield wires, in combination with more than ten lightning rods, located on elevated switchyard structures, are very effective in draining offlocal atmospheric charges, thereby providing an electrically more balanced environment for the '
vicinity of the plant. Additional lightning rods are deployed in the plant switchyard, directly adjacent to the Reactor Building. In an exceptionally severe lightning event, possibly one or two of the offsite power lines may trip off. However, each one of the five high voltage lines is capable of exporting the entire powei production of the plant. Therefore, given the fault tolerant ring bus configuration in 'the 345 KV switchyard, this would have no effect on plant operations.
Utilizing upper bound data from NUREG/CR-4767, an informal probabilistic calculation has been performed, in order to compare the lightning hazard with the overall random loss probability of a DC power division. NUREG-0666 provides a random failure frequency for a vital DC bus at f = 6.0E-3/ year. This value is also provided in NUREG/CR-4550, and it has been used in the CNS PSA. Using the exponential distribution (valid for the Poisson process), this frequency corresponds to an annual probability value of P = 1.0 - exp[6.0E-3] = 5.98E-3. The .
lightning hazard calculation shows that the probability of a lightning bolt hitting on the roof of the Control building is about three orders of magnitude (i.e., a factor of 1.0E+3) below this probability for loss of a DC power division from all random causes. Furthermore, if a lightning were to hit on the roof of the Control Building, it could not penetrate past the dense steel grid of grounded rebar, which is embedded in the 2 ft thick ceiling. The physical plant description, in conjunction with the~results from this informal calculation demonstrate that lightning hazard at CNS is clearly not a significant risk.
. Thus, the role and functioning of the station batteries and other emergency power equipment needed for shutting down the plant are not affected by lightning storms.
References
- 1. J. Chavez, et al. "An Experimental Investigation ofInternally Ignited Fires in Nuclear Power Plant Cabinets, Part Il-Room Effect Tests," NUREG/CR-4257/V2, October 1988.
. 2. J. Lambright, et. al.,"A Review of Fire PRA Requantification Studies Reported in
. -~ _ - . - . . - - - ~ . - ..... - . - - . . . . - . . - . - . -.
Attachment 1 ts NLS990008 Page 31 of 31 NSAC/181," prepared for the United States Nuclear Regulatory Commission, April 1994.
~ 3. J. Usher and J. Boccio, " Fire Environment Determination in the LaSalle Nuclear Power Plant Control Room," NUREG/CR-5037, prepared for the United States Nuclear Regulatory Commission, October 1987.
- 4. Cooper Nuclear Station NRC Inspection Report 50-298/96-25," United States Nuclear
- Regulatory Commission, March 1997.
t
- 5. P.J. DiNenno, et. al., eds., "SFPE Handbook of Fire Protection Engineering," 2"d Edition, c National Fire Protection Association, p. 3-140,1995.
- 6. L.Y. Cooper, M. Harkleroad, J. Quintiere, W. Rinkinen, "An Experimental Study of
, -- Upper Hot Layer Stratification in Full-Scale Multi room Fire Scenarios," ASME Journal of Heat Transfer, ifL4,741-749, November 1982.
. 7. NRC Letter dated June 3.1998 from James R. Hall, Sr. Project Manager, Office of
. Nuclear Reactor Regulation, to G. R. Horn, Sr. Vice President of Energy Supply, NPPD. .
" Request for Additional Information Related to the Individual Plant Examination of External Events (IPEEE) for the Cooper Nuclear Station (TAC NO. M 83611)" .
- 8. CNS Probabilistic Safety Assessment (IPEL Rev.1996-A Probabilistic Safety Assessment of the Cooper Nuclear Power Station. Reliability Engineering Group, Cooper Nuclear Power Station,1996.
' 9.- CNS PSA - IPEEE CNS Probabilistic Safety Assessment -Individual Plant Examination for External Events. . Reliability Engineering Group, Cooper Nuclear Power Station.
y Sent to the NRC Document Control Desk under Cover Letter NLS960143, signed by G.
9 R. Horn, dated October 30,1996.
' 10. . NLS 970056 Letter to the US Nuclear Regulatory Commission, from P. D. Graham, dated April 17,1997, " Reply to a Notice of Violation, NRC Inspection Report No.
50-298/96-25, Cooper Nuclear Station, NRC Docket 50-298, DPR-46."
- 11. NLS 970125 Letter to the US Nuclear Regulatory Commission, from P. D. Graham,
. dated June 26,1997, " Reply to i. Notice of Violation, NRC Inspection Report No.
50-298/96-25, Cooper Nuclear Station, NRC Docket 50-298, DPR-46."
- 12. SCAO 97-0337. IR 96-25. Violation 96-26-07 Evaluation dated July 15,1997.
- 13. NLS960076 Letter to the US Nuclear Regulatory Commission, from G. R. Horn, dated June 13,1996," Submittal of Unresolved Safety Issue (USI) A-46 Summary Report."
- l Attichment 2 to NLS990008 13 Pages Total Nebraska Public Power District Cooper Nuclear Station Figures Related to NRC Question B.2 .
1 1
i I
,. . - . - . . ,, ._r -
l l
I 1.0 1
---~- 0.8g pse NP-4041 screening level,5% damping I
0.2g pga Destn Basts Earthquake (DBE),5% damping 0.3g pga Review Love 1Eatthquske,5% damping !
l D .8 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
1 1
i 0.8 9
C 3, P*%
1 /
\
% / \
/ \
3u / \ 1
<- / g i 0A / \- !
/ s- i
/ \ j
/ \ i
/ N
/ N s
/ s
/ 's s
/ ~~ '
g, / ~~ .
/
/
/
0 0.1 1 10 100 Frequency (Hz) l L
i-i I
1 1
0.8 RLE,C 877.EWMS Envelope 5% damping l DBE.C 873. EWMS Erwelope,6% damping !
l t' 3 I
I l
\
0.8 I l l,
I \ '
l \
1 \ l l L J '\ !
n I \ l 8 I \
a l g e I i 10.4 A
,I
\
/
0 / \
l \
N ; \
\
l \
g ;
,I \ !
I N i \ i i
I g -
I ~~___.
02 ,- g
/
/
/
/
/
/
,/
0 0.1 1 10 100 Frequency (Hz) i
l l
1.25 RLE.C903, EWMS Envelope. 5% damphg
DBE, C903. EW/NS Envelope. 5% dam ping l
' I i i '
/' i 1A0 f g
/ 1 j
/ I '
/ l !
/ l
/ I l l
/
I i l I l 1 0.75 I
^
e (h t 1 w il I l
I i E '
I i 3- 1 l I I ~
i 1 g i l l
< l I !
I i l 0.50 l l
/ I 1 l l 1 I I i
,. / \
l 1 \
l ~s, 1 I '
4 ___
0.25 f l'
/ l
,-J
/
/
/
,/
l 0
0.1 1 10 100 Frequency (Hz) i l
i 2.0 DBE, C918.EWMS Envelope,5% damping RLE,C918,EWMS Envelope,5% damphg 1.5
- 8. .
E
! 1.0
.a 3
4 I I t
l lL?\! \\
l \
cl \
ti I 0.5 ; I/\
\
..l t r
\f %
/,/, ""^~s ~~____
rd
,sJ
- '/
O 0.1 1 10 100 Frequency (Hz) l
I 2.5 DBE,C932,EW/NS Envelope,5% damping RLE.C932,EW/NS Envelope,5% damphg 2.fr I t l \c .,
>. I
/ I I I
/ g l i g
I u i i g I l
e '
I-j i
@ l y
g '
f I U I l 4 # (
th r-' I I
h I ,
I I I
' I g
I
\
I
\
,I \
as
,-s
/ \s_,~~~,-~~_
,/
/
5 I i n 1
l i I
I I \
I % ~~ --
)
,s~
/
I
/
/
l
,/
0 0.1 1 10 100 Frequency (Hz)
, - .. ...- ~.~. . .- . ~ . . ~ . ~ , . ~ - - . - , ~ ~ . . , . . . . . . - . . ~ ~.~-... .. ..--.u.
f I.
t
. d ;;
l- ~ - 1JS 1
I
DBE, R805. EW/NS Ewelope,5 % alem phg
- RLE.R881. EWMS Envelope.6% siemphg
.r-.
I \:
t \
.' 1.00 .; \
l i I I I l-
., I i l'. I 'I i
1; I> 1 0.75 1 -I
-- , g e-w i I T -
c- l \
I I
.- ,h. ' . I
'. M e / ,
r e
- i \
-<- I 0.50 L, *
/
I \
1
' / r - L -- *
\
/
I 0.25 l g
---~_.
1
-J
/
/
/
/
0 O.1 , 1. 10 100 Frequency (Hz) t'" ,
m t
l-4
+ w -v- -, + e - e -, - . * - -
l 2h l l
RLE R903 EWMS Envelope,6% damphg
- - - - - - DBE,R903,EWMS Envelope,5% damping I l
1.6
,l' ~ ',
I I I I I l i i l \
t \
v
\
\
\
\
c j o
\ \
3 tA \ \
It g i
g t l o \
< i i I l
I i l t I \
8 i
y I
/ I
- L O .6
\
/
\'\~~~ ~~
/
/
\ ~ - ~ ~ .
/~~/
/
g D.1 1 10 100 Frequency (Hz)
I i
l 2.6
- RLE,R931 EWMS Envelope,6% damphg
DBE,R931 EWMS Envelope,6% damping i i i \
l \
2b I I I I I \
l I l i I I I \
l l I i 1.6 l \ l
~ l \ '
\
- 2. l I e 1 E I \
e i \
.m I i l I u I \
- l \
13 i i l i I \
l i I i
/
/
/
I 0.6 ,/ ~_--._,N
/ s s-
/
/
,,<<~'
0 0.1 1 10 100 Frequency (Hz)
l l
i RLE.R958,EW/NS Envelope,5% damping DBE R958, EW/NS Envelope,5% damping r,
3 i l i \
l i I \
l \
l i I \
l I 8 I \
- I I E I I
- i. 2 l i
g l
I I i i
< I \
l l 1 I I l 1 1 l 1 1 \
l \
j i i
/
/
'\
j
' \
/
t' ~~~~~~'~~~_______
pM y-e s*
0.1 1 10 100 Frequency (Hz) l
l l
l l
4- RLE,R976 EWES Envelope,5% damphig l
DB E, R 976. EWMS Envelope,5% damping 1 r- - ,
I k I \
l 1 l l I l ,
I I l l l 3 i l i l i 1 5 \
/ l -
1 I l 1 i l
\ \ l I
El w I I l c I i A I I I
~$ 2 I i l m* i i I I o I i
< l I i i I t i l I i I \
l l 1 %
i \
I l l l
/ ,
/ t I
- %' --~~- -----.
/ '
l 1
~"" ..
0.1 1 10 100 Frequency (Hz)
l 1
1.0 l
I - - - - - - 1.3x DBE,R854 EWMS Envelope 5% damping RLE,R859 EWMS Envelope,5% damphg I
( l\
0.8 I I I
I I l, 1 1 I i i \
l l \- ' T l \
l \
0.6 i g
$ l
\
g E i 1- 1 l
I ^~ l g / t i
i
< ' l 0.4 j I L,
\
'\
l l \'~~~~-- __
/
, a n, ,i -
l
/
)
s'
/
/
0 0.1 1 10 100 Frequency (Hz)
l f l l
1 t
l 1 1.6 l 1.876x DBE Ground Response Spectra (6% demping)
..... --- .- Reecler Building. Envelope of RLE FRS,931' and below (6% demping)
Cerdtel Bu5 ding. Envelope of RLE FRS 932' and below (6% demping) l l
l 1.0 l r 11 1 il i1 l 9
l ft [ I l
\ \
c 4 t
s I l 3 i
{ l
} I ; ,
8 [ t'f. l I E l' :. !Ut\j'\
l l j ' 'j., .;lv' ' ' h\
0.6 I - \
I
(* =
- 1. '.'s
\
\
k' ..
g ,,
N f 'e sl1 '
\ .' \
- A k '; ,,,,,,, ,
s' l l l
0 0.1 1 10 100 Frequecny (Hz) i I
1.
.. _ . _ . . . . .__m, , _ .. .-_,__- . . - . . - ..m. _ .... _ .- _ . . . _ . _ _ _ _ _ _ . . . _ . . . _.. ~ . _. .
4 l
l 1
l ATTACHMENT 3 LIST OF NRC COMMITMENTS l Correspondence.No: NLS990008 i
(l
' The fo11owing table identifies those' actions committed to by the District in this i
l document. - Any'other' actions discussed in the submittal represent intended or
- . planned actions,by the District. They are described to t.he NRC for'the NRC's I information and are not regulatory commitments. Please notify the NL&S Manager at
- Cooper Nuclear Station of any questions regarding this document or any associated regulatory cormnitments.
COMMITTED DATE COMMITMENT OR OUTAGE None- "
t,
.I i
t:
I' 1
i i
. .l l
r- 1 i
l i
I l
1 i
'l a
1 j
J i
t
-) i i
B
. l- PROCEDURE NUMBER 0.42 l REVISION NUMBER 6 l PAGE 9 OF 13 l W t 4 N m T ---^-A e -'-u- 't r~- e ve o-- -t&,--w --&"d - -
vvrn -'