ML20010B487

NRC Replication Tests for Fire Protection Sys. Related Info Encl

ML20010B487
Person / Time
Site:	Browns Ferry
Issue date:	01/13/1981
From:	NRC
To:
Shared Package
ML20010B468	List: ML20010B467 ML20010B471 ML20010B479 ML20010B486 ML20010B487
References
FOIA-81-204 NUDOCS 8108170070
Download: ML20010B487 (7)

Text

,

A C

%s.

January 13, 1981 NRC' REPLICATION TESTS FOR.-FIRE

. PROTECTION SYSTEMS Test Outline l.-

Overview

..In December-1978, the NRC staf f decided that full-scale.

replication testing of actual plant configucations and fire protectionLsystems should be implemented.

These plans.were reported to the Commission on August 8, 1979, SECY-79-478. The recent Commission' Order (CLI-80-21, May 27, 1980) emphasizes the importance that the Commission places on tests and requests that they be expedited and that a firm schedule be established.

Planning' for these tests was initiated in early 1979, culminating in a user regeest memorandum to the Division of Reactor Safety Research in July 1979.

Detailed plans were made for a full-scale replication test of the fire protection system for the Rancho Seco Make-Up Pump Room.

Prior to a final inspection and meeting with the utility in September 1979, it was learned that the fire protection system had been modified and that the planned test was no longer relevant.

Acco rd ingly,

it was-decided to alter the test plan (Schroeder. and Eisenhut-to Murley, September 21, 1979) and to begin planning for a

- full-scale replication test of the fire protection system for the Arkansas Auxiliary Building Corridor.

A plant visit was conducted in November 1979, but final plans we're delayed by problems in obtaining information from'the utility needed to replicate the test configuration.

As a result of the recent Commission Order (May 27,-1980),

which requested the staff to test first the configuration of

/

greatest concern to the staff, it was decided to test the Browns Ferry Reactor Building first, and the Brunswick intake structure basement next (Vollmer to Murley, June 26, 1980).

2.

Scope Two full-scale replication tests will be performed using a laboratory replica of a selected area of the Browns Ferry Reactor Building.

Also, separate effects tests of fire detector and sprinkler head response will be conducted.

3.

Test Obicctives The goal is to provide information on performance of a specific fire p'rotection configur-ation designed in accordance with NRC fire trotect, ion guidelines and found acceptable by i

the NRC staf f. ' In addition, without inter fering with the primary objective above, the program is to provide experimental data for analytical evaluation of fire suppression phenomenology.

kD fo{j70810604I WATKINS81-204 PDR

_~

4.

Test-Pian For Browns Ferry Reactor Building (Elevation 593', Area p to q and R6 to R7)

Separate effects tests will be conducted first to determine' smoke detector, heat detector, and sprinkler. head response times,-

and their variability, for the design basis fire.

Replication Test No. I will be conducted with the suppression systems operabic and manually activated at the-time determined,

by the separate effects tests.

Replication. Test No. 2 will be conducted with the suppression systems inoperable.

Fire fighting by a fire brigade will be simulated by supn ession with a water spray from an automatic

~

hose rack 5 min.tfter ignition.

-5.

Test Details 5.1 Separate Effects Tests 5.1.1 Configurations - The test configuration will be as shown in ILLS.-1 and 2.

-Cable will not be installed in the trays and conduits.

Ceraform boards will be placed in each tray to simulate loaded cable trays.

5.1.2 Materials 5.1.2.1 Smoke Detectors - Five ionization type (Penwall Type CPD-1201) with the corresponding control box

( Fenwall Type 20ZC).

t 5.1.2.2 Heat Detecto: 1 - Two sensor cables (Walter Kidde Type TE-30-45A (30 ft)

'and TE-39-4 5B ( 50 f t) ), control box and all associated equipment.

5.1.2.3 Cable Treys - Steel, open ladder type, 18 in. wide, (manufactured by P-W Industr ies).

5.1.2.4 Conduit - Three 3 in. aluminum conduits.

5.1.2.5 Sprinklers - Pendant type (Firematic Type TP-39) will be used.

The test system will be designed to obtain the i

same flow as that calculated for each head in the plant area.

.3 -

5.1.2.6 Water. Spray - Three types of nozzles

( 9-UNIFET - NO 1/4 4TT9 515, ll-STEINEN SOLID JET-CAT Il09, 5-3" TEES) will be used.

The-test system will be designed to obtain the same flow as 4

that calculated for each head in the plant area.

5.1.3 Procedure - Five gallons of heptane will be spilled within the curbing which surrounds the base of the vertical trays.

Response time of the heat and smoke detectors and the sprinkler head activation time will be monitored. No water will be discharged through the sprinkler heads or spray nozzles during these separate-ef fects tests.

5.2 Replication Test No. 1 Configurati~ n - The test configuration will be 5.2.1 o

as shown in ILLS. 1 and 2.

Cables will be in-stalled in the four vertical trays (TE, VE, KT, K E-ESII), the two lowest horizontal trays (TE, TK-ESII), and the three conduits.

Ceraform boards will be placed in.the other horizontal trays to simulate loaded cable trays.

5.2.2 Materials 5.2.2.1 Cables -

Tray KS-ESII 61 ea.

3C/12 AWG ( PE/PVC-PVC)

Tray KT 101 ea.

29C ( Hatfield ' Havinal) 48 ea.

3C/12 AWG (PE/PVC-PVC)

Tray VE 206 ea.

2C/16 AWG ( Hi-Rad by t

Times Wire)

Tray TE V&H 131 ea.

2C/16 AWG (PE/PE)

Tray TK-ESII 61 ea.

3C/12 AWG ( PE/PVC-PVC) 1 Conduit 3A-lPP285 3 ea.

2/0 AWG (XLPE/PVC)

Conduit 3A-ES2501 -

II 3 ea.

2/0 AWG ( XLPE/PVC)

Conduit 3A-ES50-I 3 ea.

2/0 AWG (XLPE/PVC) 5.2.2.2 Sprinklers - Same as for separate effects tests.

5.2.2.3 Water Spray - Same as for separate ef fects. tests.

5. 2. 2'J4 Cable Trays and Conduits - Same as for separate effects tests.

5.2.2.5 Coating - Flamemastic Type 77, 1/4 in.

. thick.on trays and cables except for 10 cables in-Tray VE.

1.*..

5.2.3 Procedure - Five gallons of heptane will be spilled within the curbing which surrounds the base of the vertical trays.

The heptane will be ignited to start the test.

Suppression syst ems will be manually activated at a pre-determined time based.upon the response times-from the separate effects tests.

Instrumenta-tion anticipated will be 100 Type K th'rmocouples, 5 calorimeters, and two probes for air velocity, measurement.

Instruuentation will be connected to a DDAS* system and recorded on magnetic tape for computer assisted analysis of data.

The.

test will be recorded in color on video tape.

Short circuit and open circuit current will be detected for each cable.

Failure criteria will be a short or open circuit in both tray KS-ESII and Conduit 3A-ES50-I or Conduits 3A-ES2501-II-and 3A-ES50-I.

5.3 Replication Test No. 2 5.3.1 Configuration - The test: configuration will'be the same as for Test No.1.

5.3.2 Materials - The materials will be the same-as for Test No.

1.

5.3.3 Procedure - The same procedure as Test No. I will be followed except the suppression systems will not be used.

The fire will be allowed to burn for 5 min.

A fire brigade will'then respond to extinguish the fire with a water spray (1-1/2 in. hose).

_uality Assurance - The tests will be conducted in accordance 6.

Q with the specifications of Laboratory Service Control and Assurance Program, US83-2.

7.

Report One report will be prepared describing this program.

8.

Schedule Procure Equipment January 15 - April 1 Construct Test Configuration January 15 - April 30 Separate Ef fects Tests May 1 - May 15 Install Cables and Prepare Test No. 11 May 15 - July 16 Conduct Test No. 1 July 17 Prepare for Test No. 2 July 20 August 26

5 --

" Conduct Test No.

2*

August 28

- Prepare Report Submit Quick-Look Repor t*

Tuly September 20

. September 30

- Submit Final Technical Report November 30, 1981 4

9 1

l i

i i

J 1

i

• The timing of Test #2 is based upon minimal replacement of cables, conduits and-coating material.

If significant re-

-construction is required, Tes t 2 and the final' report will need to be delayed'accordingly.

e M

e 8

0 7,

e' zj I _ ;_ m m.s_ m. m r

,Im a

,o.,g_________ _ -.'__ a

=

s s

.I s

?

s s

s h

i s

,b s

,N i

s

• i-(

.!h

'N

%_ s 8 I l

s i

t 3

i %N f

s

'e Z

'y y. :-

]

3 9

l

,.- l #,i

!l I.

u i~~l s

N

.?

g>

m S 8

s f

s 4.

b l

s w

1._:

l

- }-[ s v %

l 4

s l

.?! h.

s.4 l

w q

j l

se e

fI a.9

)

e s

ll g

I N

',e i=1 s

l E

l s

. i,i s-2 1 a

i s

' 's f _bF 4 -l i_,N x s c,3 s

s 7 'F s n,-,

a l

1

\\ sl l.)

ll I -.,

l s

M H ii f_

'8 l

f e

E s,

l h

y t_ p s _mm i

a s_,; s r _. w

-4,

.Y i

a p

s l

8 s

g b

s s

t s

s 6

I_ --I-I I'

_d L3

..6.-

,3

.e.

,C ; 41

~

D s

..___m

~,-.;..

.L,..~.

e Z:si g

.G w

\\

_J w

s

.s

\\

d x x x x _i ' x 6_ y1 ^inuus x xx x s _x xxxxx.xxxxxx x s xxxx x

.x s

ts i

i a

N

• ^

e

-N.

~

a 4

S N

.o

.s N

.?

7 N

I N

7 v :

1

.s i

?

'h

)^^

f' N

3 s

y t

1 go-

~

\\

o 3

.o.r y-3 q

J i,j i

~;.-

.e, -

.e.s

.se -

s i

s e

y 2,

7..

i

f - s N

p-

'q g ;; n a p3 A +l

(

vi s b

ts i

0A 5

c:: : : : : : : : : :

\\N cc :: : : : : : : : ::

4 d.:I h.

EF:: : T: : : : :

2r

'\\

f N

}

h

'ol '

.N Pt*1A%

'.o. N

.,?.

20

.N y

'h A _;

s a.

. ~

b*

)

l K,*

l3_;

.N A

n

._i.__,,_-__.I N

y N

-c E'

f\\

Q

(\\

\\

. - {

q m-. mt t-

==.m.m m p,

___1 s

=-,

i i

I

.v.h tj t.4..s s

)

)

i E5

=-

pe E

'N

,0;O1

• = ?s -

=

+

, s lu N

i v

a x

s:

h

.\\

N s

e 3

s N

s g g

~

_ w.. r.q.e.g.+.-

--4-._,...

4

,,o - a t

O

4,*

l=

y UNDERWRITERS LABORATORIES INC.

333 ITi%KTL% IstMD

• MetTHBhf M E. BLLIMM% 68W862 I

an independent,not ;0r-profit organization testingfor public safety May 8, 1981 US112

~

BlNK5678 Sandia Laboratories Albuquerque, New Mexic7 87115 Attention:

Mr. Leo Klamerus Systems Safety Technology Division Building 4442

Subject:

Status of Program Browns Ferry Replication Tests

Dear Mr. Klamerus:

Attached is the current test plan, please review and send us any comments you may have pertaining to this plan.

We would also appreciate your coifirmation that this plan is acceptable.

The tour of the test configuration by TVA persorinel prior to Test No. 1 is being arranged, but a date has not yet been set.

As soon as we have a date we will let you know.

Very truly yogrs, GLORIA SCHRAND Technical Correspondent rire Protection Department Reviewed by:

An LEON PRZYBYLA Engineering Team Leader Fire Protection Department GS:LJ:gs vcc:

U.

S. Nuclear Regulatory Commission Washington, DC 20555 Attn: Mr.

D. Motley Pheas 312 272 8800 f.we: 72 4334 Ce6+ UlikC h0ATM84004 at

i NRC REPLICATION TESTS I

FOR FIRE PROTECTION SYSTEMS 1

TEST OUTLINE j

ISSUED:

12.-16-80 REVISED:

5-8-81 O

s s-71--, w w w

e

/ S:.wG b t t.2 -/6-SO NRC REPLICATION TESTS FOR FIRE PROTECTION SYSTEMS Test Outline 1.

Overview In _ecember 1978, the NRC staff decided that full-scale replication testing of actual plant configurations and fire protection systems should be implememted.

These plans were reported to the Commission on August 8,

1979, SECY-79-478.

The recent Commission Order (CLI-80-21, May 27, 1980) emphasizes the importance that the Commission places on tests and requests that they be expedited and that a firm schedule be established.

Planning for these tests was initiated in early 1979, culminating in a user request memorandum to the Division of Reactor Safety Research in July 1979.

Detailed plans were made for a full-scale replication test of the fire protection system for the Rancho Seco Make-Up Pump Room.

Prior to a final inspection and meeting with the utility in September 19 79, it was learned that the fire protection system had been modified and that the planned test was no longer relevant.

Accordingly, it was decided to alter the test plan (Schroeder and Eisenhut to Murley, September 21, 1979) and to begin planning for~ a full-scale replication test of the fire protection system for the Arkansas Auxiliary Building Corridor.

A plant visit was conducted in Novenhor 19 79, but final pl ant. were delayed by problems in obtaining information from it? utility needed to replicate the tos c configuration.

As a result or the recent Commission Order (May 27, 19 80), which requested the sta f f to tes': first the con-figurations of greatest concern to the staff, it was decided to test the Browns Ferry Reactor Building first, and the Brunswick intake structure basement next (Vollmer to Murley, June 26, 1980).

2.

Scope Two full-scale replication tests will be performed using a laboratory replica of a selected area of the Browns Ferry Reactor Building.

Also, separate effects tests of fire detector and sprinkler head response will be conducted.

3.

Test Objectives To pro *de information on performance of a specific fire protection configuration designed in accordance with NRC fire protection guidelines and found acceptable by the NRC starf.

In addition, without interfering with the primary objective above, the program is to provide experimental data for analvtical evaluation of fire suppression phenomenology.

/

m.

Issued:

12-16-80 Revised: 4 81

' 4.

Test Plan For Browns Ferry Reactor puilding (Elevation 593', Area p to q and Rf to R7)

Separate effects tests will be conducted first to determine smoke detector and heat detector response times, and their variability, for the design basis fire.

Replication Test No. 1 will be conducted with the suppression systems operable and manually activated at the time determined by the separate effects tests.

Replication Test No. 2 will be conducted with the suppression systems inoperable.

Fire fighting by a fire brigade will be simulated by suppression with a water spray from an c.itomatic hose rack 5 min after ignition.

5.

Test Details 5.1 Separate Effects Tests 5.1.1 Configurations - rhe test configuration will be as shown in ILLS. 1 and 2.

Cable will not be installed in the trays and conduits.

Boards or ceramic blankets will be placed in appropriate trays'to simulate loaded cable trays.

5.1.2 Materials 5.1.2.1 Smoke Detectors - Four ioni n

type (Fenwall Type CPD-1212) 5.1.2.2 Heat Detectors - Two sensor cables (Walter Kidde-Eutectic Salt Typc) 5.1.2.3 Cable Trays - Steel, open ladder type, 18 in, wide, (manufactured by P-W Industries).

5.1.2.4 Conduit - Three 3 in. rigid aluminum conduits.

5.1.2.5 Sprinklers - None b

I

?

I

Issued:

12-16-80 Revised: 4-21-81 5.1.2.6 Water Spray - Spraying Systems 1/4" TT9515 and 3/4" HH71WSQ nozzles will be used.

The test system will be designed to obtain the same nozzle pressures as that obtained in the field check at the plant.

5.1.3 Procedure - Ten uncoated cables will be installed in tray KE-ESII up to the bottom of the horizontal trays (18 ft). Five gallons of heptane will be spilled within the curbing which surrounds the base of the vertical trays and ignited.

Response time of the heat and smoke detectors will be monitored.

No water will be discharged thro'agh the spray nozzles during these separate-effects tests.

5.2 Replication Tast No. 1 5.2.1 Configuration - The test configuration will be as shown in ILLS. 1 and 2.

Cables will be installed in the four vertical trays (TE, VE, KT, KE-ESII),

the two lowest horizontal trays (TE, TK-ESI'.), and the three conduits.

Ceraform boards will be placed in the other horizontal trays to simulate loaded cable trays.

2.2.2 Materials 5.2.2.1 Cables -

Tray KS-ESII 6'

- 3C/12 AWG (PE/PVC-PVC)

Tray KT 103 - 29C (Hatfield Havinal) 48 - 3C/12 AWG (PE/PVC-PVC)

Tray VE 206 - 2C/16 AWG (Hi-Rad by Times Wire)

Tray TE 131 - 2C/16 AWG (PE/PE)

Tray TK-ESII 61 - 3C/12 AWG (PE/PVC-PVC)

Conduit 3A-lPP285 3 - 2/0 AWG (XLPE/PVC)

Conduit 3A-ES2501 -

II 3 - 2/0 AWG (XLPE/PVC)

Conduit 3A-ES50-I 3 - 2/0 AWG (XLPE/PVC) 5.2.2.2 Sprinklers - None 5.2.2.3 Water Spray - Same as for separate effects tests 5.'2.2.4 Cable Trays,and Conduits - Same as for separate effects tests.

5.2.2.5 Coating - Flamemastic Type 77, 1/4 in.

thick on trays and cables except for 10 cables in Tray KS-ESII.

Issveo: 12-ts -Sb O

5.2. 3 Procedure Five gallons of heptane will be spilled within the curbing which surrounds the base of the vertical trays.

The heptane will be ignited to start the test.

Suppression systems will be manually activated at a predetermined time based upon the response times from the separate effects tests.

Instrumentation anticipated will be 100 Type K thermocouples, 5 calorimeters, and two probes for air velocity measurement.

Instrumentation will be connected to a DDAS system and recorded on magnetic tape for computer assisted analysis of data.

The test will be recorded in color on video tape.

5.3 Replication Test No. 2 5.3.1 Configuration - The test configuration will be the same as for Test No. 1.

5. 3.2 Materials - The materials will be the same as for Test No.

1.

5.3.3 Procedure - The same procedure as Test No, l vill be followed except the suppression systems will not be used.

The fire will be allowed to burn for 5 min.

A fire brigade will then respond.to extinguish the fire with a water spray (1-1/2 in. hose).

6.

Quality Assurance - The tests will be conducted in accordance with the specifications of Laboratory Service Control and Assurance Program, US83-2.

7.

Report One report will be prepared describing this program.

8.

Schedule Procure Equipment January 15 - April 1 Construct Test Configuration January 15 - April 30 Separate Effects Tests May 1 - May 15 Install Cables and Prepare Test No. 1 May 15 - July 16 Condupt Test No. 1 July 17 Prepare for Test No. 2,

July 20 - August 26 Condu'ct' Test No.

2*

Tugust 28 Prepare Report suly 20 - September 20 Submit Quick-Look Report

• September 30 Submit Final Technical Report November 30, 1981

• The timing of Test #2 is based upon minimal replacement of cables, conduits and coating material.

If significant reconstruction is required, Test 2 and the final Report will need to be delayed accordingly.

~.,.- g..-

ISGoGD ' ID-16-80 I? E ul s e O : 4-pl.gl

• ~9' s

e Q

k w

s b

4

\\

w sN g

s h\\\\\\\\\\\\\\\\\\\\

\\\\\\\\\\,,\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

U zz[.

u.

s N

s

\\Nk.

nn s

s 4

h I

l-H i

s h-k sA T

4 CDC0 s

1 I

I

\\g b

i nSo N

a "ll i,

ofy I

- \\g y

o o

EG M

t X L

u o

.I I

),

g l

pX g

h ggg s

of

'g

,,.,I I ;I, Nk 4 s c._.

C 535 353535.33 iM N

^

\\

)

c :: : :::.: :d:

sx 1

s (j N

\\

2 i

U$

i N

O s

h-u s

Ltj

',\\

_J F_i LLl<

I

~

N w

\\

\\'

=,

. \\

N

.\\

\\

s

\\

\\

N

- \\

.i"

\\

\\

s bl

\\

s s, i

g

\\

Ag

\\

~

x

\\

s l

P l

l e

g..,

=.

1 SSOED: I2-Ib-33

.,:.'. l' ;.

RGVIseD : 4-21-L'l e,

e e

e e

-u z

sxmxxmmxmxmx3 s

N N

- \\

\\

\\

N

\\

N N

N

\\

N

\\

1. 1

\\

\\

N a<

N

\\

..'s N

4

\\

N s

N

\\

\\

\\

\\

\\

z

\\

\\

N T,

\\

J

\\

N o_

\\

N

>o I

N N

\\

N N

y k=I j

h

~

V i:-

\\

N

\\ ;;;

\\

i, n, il e

NN i

m x

,t b

y N(y n

N h % m xxxy u

\\ x I

N N lll v

N N

i,,

'j e

=

s\\

k Ji'

\\

g s

\\

s k

N(

N

\\

N N

\\

i x

4 h

l c

i l

s ll i!

\\

I

$~

5..

Filo cc: Writer 'y Leon Przybyla R.

Parks Dr. Christian.

May 14, 1981 3

J Sandia Laboratories Albuquerque, NM B7135 Attention:

Mr.' Loo Klamerus Systems Safety Tcchnology Div.

Subject:

Soparate Effects Experiments No. 1, 2, And 3, For The NRC Replication Tests For Fire Protection Systems

Dear Leo:

This is a brief summary of the Separate Effects Experiments do. 1, 2, and 3.

These experiments were conducted on May 5, 8 and 11, 1981.

The objective of these experiments was to determine the response times of the smoke detectors, heat sensors and general characteristics of the heptane fire.

Attached is a table of the response times for the smoke detectors, No. 1 through 4 and heat sensors, uo. 1, Zone lE and No.

2, Zono 1F.

Listed also is the duration timos for the heptano and cable fires and the maximum temperatures of the junction box and aluminum conduit.

The junction box temperature was measured by thermocouple No. 67, located on the west side of the box at the midpoint, 1.2 meters above the floor.

The aluminum conduit tenperatures were measured by thormocouples pinged into the center of the conduits' exterior surface, approxiraatoly 2.5 meters above the floor.

shone thermocouples were identified as follows:

TC No. 68 (3A-ES50-I), TC No. 69 ( 3A-ES 2501-II) and TC No. 70 (3A-IPP285).

If you have any questions or comments concerning thosc l

cxperiments, please contact Leon Przybyla or myself.

Very truly yours, Reviewed by:

j T. PLENS L. J. PRZYBYLA Laboratory Technician Engineering Team Leader Firo Protection Department Fire Protection Department CC:

Mr._ David Notley

~dL Mail Stop U.S. Muclear Regulatory (ammission Washington, DC 20555

-~,

oo p

Sandia Laboratories Pago 2 May 14, 1981 Experiment 1 Experincnt 2 Experiment 3 Timo Time Time

~

Smoko Detector 01 10 sec 11' scc 10 sec 62 9 sec 9 sec 8 soc 63 8 sec 8 sec 7 soc E4 9 sec 10 sec 8 sec Hoat Sensor

• Zono lE-01 37 sec 45 sec UA

• Zonc 1F-82 NA NA Did not activato Duration of Ucptanc Firo 2 rdn,10 sec 3 min,15 sec 2 min, 30 sec Duration of Cable Firo 4 min,15 sec 4 min, 15 sec 5 min, 15 sec Maximum Temperature 4 min,15 sec 4 min, 30 soc 6 min, 00 sec conduit 3A-ES50-I TC 668/210 F TC 068/222 F TC 968/158 F 4 min, 15 sec 3 min,35 sec 4 min, 35 sec Conduit 3A-E52501-II TC #69/159 F TC #69/139 F TC #69/109 F 4 min,15 sec 2 min, 10 sec 4 min, 40 sec Conduit 3A-IPP285 TC 670/148 F TC 470/121 F TC 970/100 F Junction Box 2 min,00 sec 2 min,10 sec 2 min, 00 sec TC #G7/563 F 'TC 067/968 F TC #G7/529 F

• Note: Zonc lE-01 heat sensor was connected for experiments 91 and 2 only and Zonc lE-62 was only connected for experiment #3.

GEI!ERAL OBSERVATIOliS:

The tost duration for the 3 experiments was G min.

In all of the experiments heptano leaked out of the barrior area.

Flarecs from the heptanc fire impinged on the ceiling, approximately 7.7 meters abovo the floor.

This flaning wns restricted mainly to the contor of the fire pan area, between the 4. vertical cable trays.

The peak of the flamc3 outsido of the abovo contioned arca reached an approximato height of 3 motors.

1

Rationale for Exposure Fire Used for Testing We do not have a single design basis exposure fire. When a significant exposure fire hazard is identified in a particular area, this fire is used as the design basis and is also logically the correct fire to use for test-

~

ing of this configuration.

In many areas.there is no obvious exposure fire hazard. Yet experience tells us that exposure fires occur in such areas. We cannot predict in a particular area which fire will occur, but experience has shown the kinds of fires that occur in such a'reas. They relate to kinds of combustibles that could be present and the kinds of ignition sources that occur in fires that have been experienced.

For the initial testing series of plant cor.-

figurations, we have chosen two gallons of heptane in a shallow fan.

If an

~

exposure fire typical in heat content, burning rate, and burning time.

Of the kind of fires that should be prctected against in these areas nominally free of exposure fire hazards.

Whether two gellons of heptane should also be used as.a safety design basis

~

is a question with ra.ther different considerations than those governing

~

the use of this fire as a testing tool.

In general for such areas, the emphasis has been on operation of redundant components or cables rather than a detailed analysic using a specific design basis fire. An existing plant where adequate operation was not available, other protection was required. Thus it,seems that establishment of a design basis exposure fire forsuchareasisdot'requiredeventhoughwedoneedastandardtestfire for the_ experimental program.

1 i

---