ML033580472

Appendix R Regulatory Conference with Comments

ML033580472
Person / Time
Site:	Arkansas Nuclear
Issue date:	07/10/2003
From:	Anderson C Entergy Operations
To:	Office of Nuclear Reactor Regulation
References
FOIA/PA-2003-0358
Download: ML033580472 (20)

Text

I ARKANSAS NUCLEAR ONE I

I APPENDIX R REGULATORY CONFERENCE July 10, 2003

.101p oz 6

2 OPENING REMARKS

• CiA

.I Craig Anderson Vice President, ANO

Risk Assessment Comparison

.I*

• ~

(4,.

NRC ANO

• 4250 F cable failure

• 7000 F cable failure temperature temperature

• Zone wide prompt

• Limited time phased da'mage damage

• Generic HRA

• Plant specific HRA

- Based on zone wide - Scenario specific operator prompt damage actions evaluated

- Included LOOP - No LOOP

• Greater than Green

• Green finding finding I

A_

- - o -

Unit 1 4KV Switchgear Room (fire zone 99M) 16

.j A,,-I . /l 99M - south view B6 Load center Dry-type N transformer

... '. 44

- a II' iAs Typical ANO switchgear cabinet wiring, control cubicle I,

Fire Scenario Selection:

General Approach 18 Three distinct fire scenario classifications:

'An electrical fire (non-energetic) in any of the electrical cabinets in the room Fire may spread in the cable trays, but requires considerable time

• Circuit damage/failures follow a time-phased sequence with first damage after 10 minutes

.er~

A high energy arcing fault switchgear fire that may initiate secondary fire

The event has an initial (immediate) pressure phase that causes damage to targets and ignites exposed cables in the vicinity a The fire may continue in the switchgear and grow within the ignited combustibles (e.g., cable trays) in the vicinity

• There is an initial/immediate circuit damage/failure followed by potential time-phased circuit damage/failures A transient fire that may spread into cable trays

• A transient fire between B55 and B56 was selected as the maximum expected scenario due to its potential for extent and timing of damage

• ircuit damage/failures follow a time-phased sequence with first damage after 10 minutes

Fire Characterization 21 Electrical cabinet fires "The heat release rate data profile is

',based on the best available fire test HRR Profile data 120 Sandia National Lab (NUREG/CR-4527, loo 87/88) and VTT (Valtion Teknillinen Tutkimuskeskus, 94/96) in Finland 80 Same test used in the NRC SDP analysis = 60

- The ANO HRR is based on the highest t 40-peak of ST5 (unqualified, open 110 21_

,' KBTU loading)-and-allqualified-vert I 20

/ _cabRinetsrxcluding PCT6 and test 23 )

( h1.5_MBTU loading) 0 5 10 15 20

• The NRC HRR is based on test 23 (qualified, open 1.47 MBTU loading) and Time [min]

test 24 (unqualified, open, 1.44 MBTU)

- Time-to-peak is based on the average

- Tests are based on control panels

- The switchgear, MCC's and load centers are enclosed with sealed penetrations

• Used for scenarios 1a, 2 -5

• J

" i M la la a Is In Oa l a ai - E - -

Fire Characterization (cont.) 24 ZOI of the High-energy Switchgear Arcing Fire Top View -0  %%

\- Oe l :l r oaYC. a3swus^tv? Ssu.u;^l; 4%,.

'a I

I Mi I a I-S 1:.0

.1 t>

0 L.

CD 0 4 CD CD. .1 0 4 a)

CD 16.

CDJ CD1 0 6n % CD CD (n

__ _ _ _.1 _ ___ ___ ___ ___ _ I ____ ___ .1 _ ____ __I k

, .,:x; ;,TIW,%E53i~e,'-'.S^. ^1 II I I I I I ~ ~ 61 E 2,40

%. - -.- 10

- ~ 0~S'C"IVc-_p Side View -

-P'+/-,,, 7,

,- Ir '! iWx - it , -i. I

-j _,'I-). . -.--4,_' ilf;i_' .!;. 7 7:1 I ., , -, -,- __

.; V 77_17_7 - --- _ _ -_,

, . ',I.

_., 1" 7:;:.,

1i,.,,,-A f--W EAQ2 1Va15foK- I~t>

9, vhq;v DAD noSP.,E2TE~til S^t I- I V I

I~~~~~~~~~~ I Relative Door I Cable trays are

4. J4 JJ4. 4 .L. S ) C O C I offset as Height 0 o~ 0 0 A

0

~ 0D U'

0) o (04 o - . W -.

indicated in top I _ _ _ _ __ _ _ _ _ _ _ _ _

N~~~~~

O' *.

-~~~~~~~~~~~~~~~~~.

_,' Initial ignited view U&&. If p area V9{%l

.. I ,  :, x

?--

II,,

_II i

I F

ti t

k-i iI I

i k

N .

.I ..

.I Results I.

loP CFAST Results Ci~ tr Scenario b, Open door 2500 600 I 500 2000 1500 D

in' ' 4~~~~ 400 o 300 a.

1000 -

I- I. - - - _~~~~~~~~~~

i 2 00 E 500 'n IUU 1A 0 I

.4 I l

-~~~~~~~~~~~~~~~~~~~~~~~~~~~~'

,- 0

.0 2 4 6 8 10 12 14 Time [Min]

- -Calculated HRR __ Input HRR Ipu UL. Temperature Tm

lesults (cont.) 31 The limiting fire scenario, one that can generate a damaging HGL, is not credible

- The non-suppression probability by.the brigade for very long duration cable fires (100 minutes for the high-energy switchgear event) is 0.01 (per EPRI Fire PRA Guide)

- Fuel depletion, cables ignited earlier have burned out

- Parts of the cable trays are coated with flamastics which both delays ignition and slows propagation of cable fires

- Continued growth of the non-piloted cable fire for a long time is not likely. (Tests reported in NUREG/CR-5387 state that cable fires, "spreading horizontally only as it progressed from level to level")

Maximum expected fire is a high-energy switchgear fire o No credible fire reaches 7000 F in this room (limiting fire scenario)

Results: 34 Frequency of Fire Scenarios in Fire Zon M 1 ~~ ~ ~ ~

IA ANO SDP Analysis Results /1 1I -

V I- _ _ ._

/_

1 -

/

0 WMs Ratio oF -- Pns by WFI Floor

.2 Generic (ignition event for a plant Pn by Source (location area ratio Sev U Frequency weighting source (transient severe personnel fire Results in F for weighting switchgear or fire brigade factor) fires) watch factor) fire la Fire In the A4 switchgear.

Nominal value. 100 KW fire 1.50E-02 2.50E-0 5 88E-Ot 1.00E00 1.20E-01 2.50E01 I OOE+OO 1.00E00 6.62E.05 lb High energy arcing fault In any of the A4 swilchgear breaker cubicles 2 Fire in the B55 MCC. Nominal 100 KW fire. Fires In Inverter Y28 are bounded by this scenario.

t___ ___

.50E02 2.50E-01 5 88-02 1.OOE*00 1.20E-01 1.00E400 1.00E00 1.004E00 2.65E-05 3 Fire In the B56 MCC. Nominal

_----- -. QKMiLfire 50E-02 2.50E-0 1 5 88E-02 1.OOE400 1.20E-01 1.OOE00 t.OOE+00 t1.OE+00 2.65E-05 4 Fire In the Y22 Inverter. Base case. 100 KW fire. Fires in Y24 and Y 25 are bounded by.this scenario. I.50E-02 2 50E-01 5 88E-02 1,00200 1.202-01 1.00E+00 1.00E+00 5 OE-01 1.32E-05 5 Fire in the Load Center 68.

_ a - I QW r fnmlRR t _ 1.50E-02 2.50E-01 5 88E-02 1.002400 t.20E-01 .OOE+00 10.00 2.OE-01 5.29E.06 6a Transient fire In areas o the room.

where cable trays are exposed to a loor-based fire. Nominal Value

• of 50KW . 3 60E-02 2 0OE.00 1.80E-02 1.00E-01 1.0012400 1.00+00 5.002E-O 1.0OE+00 6.48E-05 6b Cable lire caused by welding and cutting In areas of the room where cable trays are exposed to a floor-based fire. Nominal Value of 150KW. 1.30E43 2 OOE+00 2.00E-02 1.00E-01 1.002400 1.00E+00 5.00E02 1.00E+00 2.60E-07 NRC SDP Analysis Results (May 15, 2003 Supplemental Letter Page 25)

Sourco Frequency Electrical cabinets 2.3E-04 Transformers 1.6E-05 Ventilation Subsystems 4.4E-06 r

Key Systems Affected in the Risk- 39 Significance Determination (Fire Zone 99M)

The following systems/trains are directly failed due to fire induced power losses of A4 and B6

- One train and the swing pump of service water

- One train and the swing pump of HPI (makeup)

- The A4 associated diesel is no longer usable I;

SV~ke_ Ce<

Circuit Analysis 40

• Detailed circuit analysis performed on zone 99M

• Investigation of cables located in the trays and conduits associated with the target sets

• Analysis showed no loss of offsite power associated with zone 99M

- NRC evaluation did use loss of offsite power

• Analysis of associated failure modes for affected cables

• Failures unrelated to safe shutdown also examined to provide accurate portrayal of the risk caused by the fire

44 Summary of Procedural Guidance Key Action Previous Procedures ' New Procedure 'a The previous procedures discuss this in great Starting EFW P-7A detail. Spurious and false indicators are not Discussion in new procedure includes 1 manually and positioning mentioned which could delay operator functional indicators.

associated valves response.

Lack of adequate and correct indication is Controlling EFW (A or B) Previous procedures discuss this local or directly discussed in the new procedure to prevent overfill which makes this action more likely in the new procedure.

Local closing of bus A3 This action not explicitly discussed in the T n

!s~Witchgear for P-7B and normal operating procedures but is discussed T new he procedure explicitly addresses HPI A (e.g., inverter fires) injAftemate Shutdo-wn. l Discussed in previous procedures. The timing of this action depends on when letdown is The new procedure addresses the 4 Starting HPI Makeup isolated. possibility of starting the HPI pump locally.

k Isolation of letdown to Inboth the previous and new procedures, this In both the previous and new procedures, 5 avoid needing HPI action is discussed and can be performed in this action is discussed and can be (Makeup) sooner the control room. performed in the control room.

In both the previous and new procedures; this In both the previous and new procedures, 6 S action is discussed and can be performed in this action is discussed and can be long-term cooling the control room. performed in the control room.

I I 4. n.

I' It .-_

hj i.

Simulator Scenario for Zone 99M 45

• Fire damage chosen to provide HRA information for multiple operator actions

- Fire model beginning with an A4 switchgear fire

- Fire propagated throughout zone causing wide-spread cable damage

- Damage for scenario extends beyond credible fires

• Realistic control room communication challenges

- Fire brigade leader communication

• Timelines based on actual fire drill

• Included need to contact local fire department

- In plant auxiliary operator used for operator actions

• Radio and telephone communications used l/ , .I A r.

ka~~f,)I &cj flj g.s7v, &{c\r mk,1:nic~ Y Atc Q al

EPRI Calculator 53

• Industry sponsored method provides a process for book keeping HRA evaluations

• Addresses HRA requirements in ASME PRA Standard 2002

• Includes several methods for quantification

- Industry and NRC sponsored 2 iU

- Generic data quantitatively differentiate human error probabilities (HEP's) for key characteristics of procec ures and man machine interface 3 HRA analyst judgment is still required

. - I

• >SO~, ks s y

• OG 't{ vsL fVcAt ,?L, 4 4 ( i., -, ': C,; .Crw J or.1

Overall Summary

• Detailed'analysis of zone 99M

- Credible fires result in time-phased failures ithout zone-wide damage (7000F damage temperature for thermoset cables) (- -

- Detailed circuit analysis indicates there is not a loss of offsite power from any fire scenario

- Simulator scenarios provided realistic data for assessment of operator reliability in the use of previous and new procedures

- ACDF for 99M is 2.2E-07/yr

• Total Unit Risk

- Two additional zones considered risk significant for Unit 1

- Risk assessment of zone 99M conservative with respect to other zones

- Conservative estimate of total unit ACDF is < 6.6E-07/yr

• The significance of the use of manual actions to achieve safe shutdown has very low safety significance and should be characterized as GREEN

Overall Summary (cont.) 70 ANO fire protection program

- Defense in depth strategy to prevent and mitigate fires

- Explicit control of combustibles

- Fire brigade effectiveness

• Primarily rely on barriers or physical separation for-equipment required for safe shutdown

- Fire detection and suppression

- Limited use of manual actions utilized for Appendix R compliance

• Actions taken to further reduce risk

- Validated circuit analysis

- Feasibility evaluation of manual actions (E 71111.05)

- New procedures developed to enhance operator response

- Fire detection reliability improved

• ANO can successfully achieve safe shutdown in the event of a fire in any zone , _

I*_

V I