ML20126F962

Ceramic Fiber Blanket Wrap for Fire Protection of Cable Trays & Conduits

ML20126F962
Person / Time
Site:	Rancho Seco, Zimmer
Issue date:	03/17/1981
From:	Chaille C, Reiman R BABCOCK & WILCOX CO., CINCINNATI GAS & ELECTRIC CO.
To:
Shared Package
ML19343C516	List: ML20126F962
References
TAC-43320, NUDOCS 8103240419
Download: ML20126F962 (10)

Text

.

ATTACHMENT 2

}

e O

CERAMIC FIBER BLANKET WRAP FOR FIRE PROTECT'3N r F CABLE TRAYS AND CONDUlTS

. Charles E.ChaiMe Senior Project Engineer Richard J. Reiman Senior Electrical Power Plant Engineer Babcock And Wilcox Company

• Cincinnati Cas And Electric Company Refractories Dhision Cincinnati, Ohio 45201 Augusta, Georgia 30903 INTRODUCTION

• SM ALL SCALE TESTS With the recent fire at the Browns Ferry Nuclear Materials and Equipment Power Plant, (1,2), there has been a great deal of interest in protecting electt: cal cables m :ase of a The protective wraps for the cable trays and fire. In some areas of nuc! car power plants, cables conduit were made from ceramic fiber blankets of redundant electr;calsystems,which are necessary which are prepared from high temperature aluraino-for the safe shutdown of the reactor, are in close silicate glass fibers. These blankets contain no proxinity. If a fire sho.!d occur in one of these organic binders, have a very low therma! conducti-a'eas, integrity of redundant systems may be lost vity and can be used up to 1250 C.

before the fire is extinguished. Therefore, fire The fire tuts were performed in a catenary protection to maintain 90 minutes of integrity for type furnace. 0.90 m (36 in.) deep and 0.90 m redundant safety related systems is needed when (36 in.) wide. The furnace contained two natural they are m close proximity. At the W. H. Zimmer gas burners capable of delivering 1.3 billion Jou!cs Nuclear Power Station, the Nuclear Regulatory (1.25 million BTU's) per hour per burner. The O Commission (NRC) critena for fire protection between redundant safety related cable tray furnace heating rate was controlled and program-med in accordance with the heating rate spde:fted division is, where the redundant trays occupy in ASTM E119 and recorded throui;heut the test.

the same room or are within s proximity of 20 feet, all but one of the tray di isions tr.us be pro- Eight type E thermocouples, connected to a multi-tected by fire bamers navmg a 90 minute ;.re resis- point recorder, were used to noniter the temper-tance rating qualif.ed to an AST5! E119 type fire. atures in var.ous locations in the cable tray and As a g-r eral rule, wht e automati: sprinCer pro- conduit.

tection is used, the f.re resistar.ce rating can be A :ircuh breaker light display board was con-reduced to 30 rninutes. structed to mordtor the capability of each cab!e to carry current during the test. The display board Kaowool O " ceramic fiber blauet was was capable of monitoring 20 circuits m the cable evaluated as a fire protective wrao around cable tray and conduit. The display board would indicate trays and conduits. Two series of fire tests wer, a failure if a circuit should open or if a cable perfortned, or. ' small scale (indep endent of testing should short against another cable or agairist the for the Zimmer Plant) and one large scale, to deter.

mine the duration of exposure of cables in protect. cable tray or conduit.

ed cable trays and concuits to an ASTM E119 fire before failure of the cables. Fire Test Resul:s Four fire protection tests were performed.

The cabic tray in each test was approximate!y 3Ce

• This paper is in part a compilation of two full and only two lengths of cable werc in the con-previously presented papers at the Summer, 1979, and1Yinter.1950, Mertin;;s of the IEEE duit. The cables were laid in the tray and conduit Power Engineer.ag Soc i ety paper numbers and looped at one end so that the cables could be easily connected to the circuit breaker.li;ht display A 79 477-2 and A S0 093 5. boardi The cables were placed in the tray in an orderly fash on to that the approximate cable lo-Kaowool O ceramd fiber blanket.130 kg/M 3

Q "

(3 2bs./ft'),75 mr ' in.) thick, manufactur. cation corresponding to each circuit on the licht board would be known. A ;chematic diagram of a ed !>y Babcock and W1.rox Company, 8108240 % ,

i

?- .; .

L l thermocouples in the cable tray and conduit is 0

typical byout of the cables with respect to the circuit numcers is shown in Figure 1. The cables shown in Figure 2. The temperature increase in

- f:t circuits 15,16, and 17 were not qualified as various areas of the cable tray was quite erratic.

non propagating according to IEEE.383. The re- This was probably caused by the thermocouples snainder were qualified cables. being near burning cables which would produce ,

large temperature fluctuations. In addition, there were some oscillations in the furnace controller during the early stages of firing. The thermo-couples may also have come in contact with high

i. is is is is a is ,, n so ,, voltage during the test, which could produce very e io ,, , .3

3I.I@Jrh s.r a sb s 2 top of the cable tray increased m temperature

---

q\

' ' a 'd '" '" r- c " a i " *" * 't -

A.. \-- increased in temperature more slowly since they

,o.] m were protected by the cables on top. The thermo-

[\-is \ , couple in the conduit increased in temperature Figure I slowly at first but, after seven (7) minutes into the Schematie diagram of approximate cable locations in cable test, began to increase very rapidly.

tray with respect to circuit nurnbers on light board for ,_

small scale tests (Note: Circuit 17 is in condnit). * -

E 119 y

- J~

Test Number 1 In order to establish a basis F ** -

for comparison, no msulation was used to protect .co. f g f - #'5. .f' '

the cables in the cable tray or conduit in the first  !,,_ f,_-l,R'k ~

, ' ],

test. A solid bottom, steel, galvanized tray and steel conduit were ured in this test. The cable a

i'"-

U

' ' d'

/I

~ ~~

-A

%7 tray and conduit were installed in the furnace so  : **- . h_ ,,cf.' , ,, ,

that their centers were in the center o' the furnace. oo. /

The cable tray was raised approximately 50 mm , , _ / \. c a,,,,

(2 in.) so that the flames from the burners would -

be along the sides and bottom of the cable tray. o The conduit was suspended above the cable tray. * """

The ends of the furnace and conduit were sealed with ceramic fiber blanket. Figure 2 The sequence of cable failures which occurred Temperature increase in various locations in cable tray during Test No.1 is shown in Table !. The first and conduit during test No.1 unprotected steel. solid cable to fail was an IEEE 383 non. qualified cable bottom cable tray and steet conduit.

in the tray. It failed at eight (8) minutes into the test. The second cable failed at ten (10) minutes These results, of course, indicate that unpro-into the test. It was an IEEE.3S3 quahfied able tected IEEE.383 qualified cables cannot maintain located on the bottom of the tray near th side. integrity for 30 minutes in an E119 type fire.

The third cable to fail was the power cable Therefore, a definite need exists for a fire protect.

l carrying circuits 18,10, and 20. When this cable ive barrier around cables in certain areas of many l I I failed, there was a power surge through the circuit nuc! car power plants.

breatser panel and ligSt board. The power to all i

cables and the light toard was turned off for ap. Test Number 2 . In the second test,50 mm proximately one minute until the power to this (2 in.) of ceramic fiber blanket wrap was tested to cable was turned off. When the power was turned determine if it would provide one hour of protec. l tion. The same type of cables, cable tray and con- I back on, several cables had apparently failed. The cables which failed during this power outa;c are duit were used in this test as were used in Test No.

marked with an asterisk in the table. The test was terminated af ter thirty (30) minutes.

1. For the cable tray 25 mm (1 in.) of blanket was placed on top of the cables. Two 25 mm (1 in.)

n V

l '

The temperature rise as indicated by the thick blankets were then wrapped around the cab!c 2

- - - - - ,-.. ,s_,-,,

. , y  ; .

- .-  ;.-'4' L.

. .y a .

3 . . ... .,

. m > -- .. .

6 - .

-" ?. . .

~..

' ' . 7, ,g.%pe v r

tray with 75 mm (3 in.) overlap joints where the -a.,- '" 9,9  ;

blankets met. A schematic dir. gam nf the insula -

ed cable tray is shown in Figure 3. The extertar V E'

...7 7

}?; W n i blanket wrap was held on with steel brackets. A- .h,~ .C M .; N9 3

'.O@

'*",. . . T- 'e * . "

long the length of the cable tray, where one blanket 7 -/ I J. , / . *1' W

~ '

ended and ancther began, the blankets were bu :ed '" .J-s w*ly.

tog tther. The butt joints for the inner blankets g*] ');.P. 4 - .l and outer blankets were separsted by approxtmate- - i b , I ly 0.45 m (18 in.) and the brackets wer? spaced , N., -c'fy

'~

approximately 0.60 m (2 i in.). In order to moni- -

kh  %-/~W[.: dr P-(

. % . pg.- g%Q tot the tempcrature on the outside of the wrap, -Q one thermocouple was placed outside the cable .'-.-%,'~,  : .;;

~

tray on the top surface of the blanket insulat:en.

- 1;[p[

r' - f 'q d ".g, . ..

The conduit was also wrapped w:in two - 25 mm .

MN 7 {% #.'

(1 m.) thick ceramic fiber blankets. The blankets . a . ., ;- L+, = t were secured usir g banding material. .,.A wo bandeg materials were tested, one' was common carbon Figure 4 steel, the other was stainless steel. The bands were Insulated cable tray and conduit installation la ferr.aee placed 50 rum (2 in.) on either side of the butt for rtre protection Test No. 2 (be ore tes:). ,

joints of the exterior blanket. A photopaph of the ,

insulau:3 car.e tray and conduit after insertion in. -

to the furnace is shown in Figure 4. ,,.. . , , _,

.7~~~~~~

w ns srtzt Truu c <>ece sw.<tt, p "see .

. N am e te.axim gu:st? 7%awa il*)1*sCa .,

=$ \ ' -

0 N

- -=

-o+

-z .- m.e

= - I: -- --

/ " -'= " ~~'-

K.ti.__ .r.W-.-.:

_ , . _ - ,_ - n. y<_.=.:.~O..

m

. f .*_ .-

._s 2

-t

~

ys- M _.:.a.N

-;y i,, g 4gg, ',,.n.,,/*.'~.~.~.~

a. __- ,,,.

W[} -( 'l I ) jA

,1 is

?

m- /

/

/

A v r v3 .

,.,,~,.~ '.

,i; L I i 1. >

$ 7,i g - m.

h o,

0, i 6>. 9< . y/& s' ,,. ' ~~~~

Qb- - -. _-c -: - :.: - - r.~: -:.-df n' M m w w + =. - w : -: : + % - @ , , , ,, ,, ., , ,,  ;, ,,

Tunt mam2

~ [ctuur. s et = ame= tht.aexn ow somiri ega c4att ? au oo as's. uett is.O l

i . ex ._.

Firure 3 Figure 5 l Schematic d:: gram of insulated cable tray for Cre protect. Te cperature increase in various locations in cable tray I lon Tests Nos. 2 and 3. ..

and conduit dunng Test No. 2 steel. solid sottom tray ,

and steel conduit wrapped with 50 mm (2 in.) etramic >

l ,

riber blanket (loose butt ;oints).

l The sequence of cable failures in Test No. 2 is shown in Table II. With the 50 m:-- (2 in.) of . .

ceramic fiber bhnket wrap, the first cable fr.ilure The results indicate that the insulation is veryf l occurre'd at fif:y-one (51) minutes mto the test, effective in retarding the heat flow inte the cable 7 l The cabic was the IEEE-333 qualified cabic located tray and conduit. The control thermocouple and in the bottom cf I te cab:e tray next to the side. the thermocouple located outside the insulation The cable in the unduit, which we.s a non qualified indicate that the furnace temperature followed cable, failed at one (1) hour and five (5) minutes closely the AST.\1 E119 curve. The steel brackets into the tent. The first non. qualified cable in the around the cable tray and the re;;ular carbon steel tray to fx was .',o.15

• w hich f.:iled at (1) hour and and str.intess steel banding material around the

, ten (10) mmutes into the test. The test was term conduit insulation performca satisfactcrily during .

l

' g nated at:er one (1) hour and sewnteen (17) minutes.

The temrerature ti.se in the cable tray and the test.

Althougt this design perfcrmed fairly well, conduit dur:nc Test No. 2 is shown in Figure 5. It did not provide the anticipated one hour of pro-l l . 3

~ . ~

w , - , w r-- --

,-w-

.. -- 4 s. - .

.~ . .

s :. .. -. .

..n c, N:=; ; .

.. .r,

._ -m- ,. .

4

.;. -- . -;.=

e.

m- s --

.. . - . . a ._ . .

~

the test. This es. ole was the IEEE 383 qualified

tection. One possible fault with the desig- was cabie TocalecTTn7he T>ottorh or the citie tray near that the steel brackets were not properly-locatedc

'to adequately seal the butt joints on the cuter ~ the side. The first non-qualified cable failed at one bbnket wrap. It was believed that if the brackets (1) hour and two (2) mmutes int, the test. The were not located on each side of a butt joint, the test was terminated after one (1) ho.tr and thirteen tray may flex during testing and the butt jomt (13 minutes. The conduit cable did not fail be- - ' -

could open. This would permit the fhmes to fort de test was terminatec.

move in and behind the first layer of blanket in- The ternperature rtse in the cable tray for Test rulation and could lead to an early failure of the No. 3 is shown in tigure 7. Agrin the cer mic cables. Therefore, additional testing was done to fiber blanket wrap retarded the heat transfer into determine the effect of relocating the b sckets. the tray. This resulted in a slow temperature rise in the tray. The thermocouple which was placed Test Number 3 The third test was simihr to against the side of the tray indicated the highest Test No. 2 except tne steel brackets, used to hold temperature of all of the thermocouples in 2e tray.

...l. .

the cerarme fiber 5hnket insulat:en on tne cable -

' ~

J. ~ .

tray, we-e relocated. The brackets were placed 75 N - - -

mm (3 in.) en each side of ie butt joint on 2e - ..-r cuter blanket wrap A photeg-sph shewmg t.he - "*'

u.ma in.c rx , , ,..' ~ ,,,_ i location of the bra:kets for Test No. 3 is shown in Fig tre 6. In add. nan, an a!uminum, open hdder  ;; = - ,

-J ' L. - i m, , , .E ~ ~

cable tray and conduit were used, instead of de -

.Y" - . . .

{3 m. /

-,/'...-

steel cable tray and conduit. . .

' ~

~

i '"- ruv 7teunat

'P~*MT. N @ {# jt -P.x. _ .ih 3-

• - -5

. - b. s.g, . .,N si *ii. t ,4 { 7d "#q' I.

. .m.

1,

- ' t p .. " - .

roe g

1. ~~ '~ b hs$i.b.

j ,,

h.. . m !"

4 ls Q ,.cf

,-. - ss, f_

,.'. ;;, Q,Qd

-bM Figure 7

- e; . . .

~

. M ;,.,

kg" '

t j, ,' d Ternperature ineresse in ratious toe:tio .s n cable trry and cor.duit dunng Test No. 3. alunnnum, open 1:dder cable

. .. ,(

tray and aluminum eendo:t u r:p;ed =,tt f,0 mm (': m.) -

9.*'

w ,. * [,,,,,

,- " ,.. -:.:!c

- Kh. y 7. .. .'g%

. ' igt caramic fiber blanket (tight butt Join.s).

f; .  :. T.' w.M .d,;-";  ;.-Q%g. =W p%(.. q w g' I

. ._.,. w

. .; -'.~ ..

7

_ __ m. -~

~

Test Nurnber 4.The bst test in this series was Figure 6 simihr to Test No. 2 except 25 mm (1 in.) of cers.

Insulated cab's tray for Test No. 3. Note: relocation of mic fiber blanket w sp was used on the steel cab!e I holding brsekets spaced 15 mm (3 in.) on each side of tray rather than 50 mm (2 in.). This was dona to l

butt joints. determine if 25 mm (1 in.) of wrap wou!d qua!!fy The se,';uence of cable failures in Test No. 3 for 30 minutes of fire protection. Twenty five b shown in Table 111. The relocation of the holding millimeters (1 in.) of blanket was bid on top of the cablesin the cable tray. Then 25 mm (1 in.) of

• brackets to within 75 mm (3 in.) on each side of

' the butt joint improved the effectiveness of the bhnket was wrapped around the cabic tray with a bbnket wrap system. As shown in Tabic.!!I even 75 mm (" in.) overbp. Along the length of the with .2n open ladder, aluminum cable tray which cable tray where two blanket ends met,a 100 mm could flex considerably durin; a fire tent, the cera- (4 in.) wide strip of ceramic fiber blanket was mic fiber bl.mket insula. ion provided cr.cellent wrapped around the butt joint to (c.rm a tyod seali ~

protection fer the cabi s as long as the butt joints The steel bracket was pbced over the 120 mm N were kept lightly wahd. The fi st cable failure in.) wide strip. T':is wrarping cchnique is shown Q occurred at one (1) hour and one (1) nunute mto in the photo;r.;ph .n Figure S. In addition,in this h

• .* g

l

. (

4 test the conduit was rotected with 38 mm (1% in.) the tray was indicated by the iermoccuple which

_ of pipe insulat.on in;.tead of ceramic fiber blanket. was located on the side of the tr:y. The thermo-couple in the conduit was inds:ating somewhat

,~~- . . ~ ..., - - ~. .. ~ ,,p,.y,,c~ m ,

, higher temperatures than the thermocouples at:ach-

~~

J ' Y,' / , i ed to the cabies in the tray.

,. ,, _, -- ~) , ; . - t.*

. v. ,

- ,r- . _ . . - .y -

~

'l i: ' ? -

J:j LARGE SCALE TESTS

~ -

i , Large scale tests were performed specifica!!y

'p 'l 4

/4 for the W. H. Zimmer Nuclear Power Station I'mt

1. This PCwer station is located 25 miles southeast 4 , y - , ' w'

c. '.1 g . 3 -- < of Cincinna:i on the Ohio River, and is jointly own.

$ gp':: .. -C r.

?.i

.t.. . . .. ' J ." . 7. " . . . _4 ed by The Cincinnati Gas & Elec:nc Comp:ny,

~4

,4 L nil Columbus and Sou:nern Oh:o E.ec~.: Company.

p:; '" K .',. . .4 and The Day:en Power and Li;h: Company. The l ir y , ::'./ #- sg*':.g-

- f 7.- 243G megawatt Boiling Wate: Ructor is de f.rst

%'-;;Q' .' - 4.) ;.' , .' . .

-~.4 of the BWR.5 se-ies tea: tors to undergo licensmg.

".f e "

C ;-.. ~~'

Ml-~

The design employs three redundant dh .sions. '

_- Figure s ne Zimrner Nuclear Power Sta:!cn unli:es Insulated cab!e t sy for Tes: No. 4 100 mm (4 in.) wide solid bottom trays, and IEEE 353 quahfied cables cenmic Cber blar.iet s*np sad br2:ke: crer but: tott. . throughou , The redundan; safety div:sien tr:ys are separated in accordance with IEEE-354 separ.

The sequence of cable failt es for this test is ation enteria (3 foot horizontal, 5 foot verd:a!).

shown in Tab:e IV. The first cab.e failure occurred In some areas of the plant, where cables of redun. ,

g at fort, (40) minutes into the test Again this cable was the :EE' -333 qualified cable loca:cd in dant safety reinted divisions are in close proximity, the decision was made to avoid the use of sprinkler the bottorn of 2e tray near the side. The firs: syster s. The presence of water could cause short '

'non.quali!!ed :able in the tray failed at fif:y two c::cui: failu: s in the elecince.! equipmen: There-(52) minutes mio the test. The cab:e in de con- fore, the NEC requ. red a fire protection barrier duit failad at feny.e.ght (4 3) minutes into de test. around all but one of the divisions which would he test was termina d af:e one (1) hour. provide 90 minutes of protec icn. Large scale test, The temp-ra:ure rise in the ::sy and conduit were performed at thePor:!a::d Cemen: Assoc::non l Is shown in Figure 9. The highest :emperatu:e in to test the ceramic f:ber blanket wrap design fe a 90 minute fire dt.:stion.

rv=w: nemes g y .-: . n

~~ '

Blanket Install:rion Procedure 8 % ";* p k. cuesta ~ # g~

~ ~ ~ J: tnrmeer E* For the Zimmer Station, the design emp!oyed

- y ***- l ,, was to pin three 25 mm (1 in.) laye s of the fiber

! see. l blanket to the solid bottom tray.a procedure which 5

, ,,,, Y .,/ /, has previously been done to insulate walls of fcssil

[

m., l ,,,,

, f *, /

ca,,3vn k .

/ *[, ,

fired furnaces. Each layer w.s held by a steel chp.

The blankets were secured to the bottom, sides,

• J/ ,', -; .4 ' 3 '

, ,,/ -

and lip of the tray with specid care to produce

,../

tight butt joints. (See Fqu:es 10 :nd 11 for photo-graphs of method.) The c:ble tray cover wr.s inde-

.e A

k '

4 ' So 4 pendently covered with the fiber b!:nket, with the

, y p,,,,,,,.t u blanket e.ste: ding beyond each side of the cover, so that the bLmket on the cover would seal:.g.mst Figure 9 Temperature inen ase in vanous tocations in utie it:y and the blanket on :he tray. This desi;n was chu>cn to parmit e:3;ct :ccess to the tray mterior if addit:en-con 6!! during Test No. 4 s:cel. sor.d bottom tt:y w ra;p.

ed with 25 mm (1 in.) ectr.mic fit,er b anket and 5:cel al c:.bles nec kd to be inst.lled. The cover w:s conduit wnh 23 mm (W in.) ppe insu!. tion. secured to the t:2y using enbon steel battdir g

..~.-.- . , .  :. . ._ . .

5 ;,: . _

.~.---: u - . --- ...-- . .

= -. . . . . . _

~~* :==- ~

~"-:;-;'**: h~~ '

rnster.'d - -

'iF i1

~ ~

x- r  %

.~ -.--g.,-..-- __ . . -_ . _ . ,. w.

i N

(. .J w.;- a v . ,- p-p_-

]

• • • V #~ - _ _ _ __

p 74

• s , .:::y va .

p.are-s ta -

.1

- e, 1

'y'

., g ./ ~'P'-auf

h. ,. ^ ~ <.,.

\

. s.-, v,s. . ,,,.,y , .,

% " =

v ; ~ .;r-.-

a . .*.s* ,'

ess af*.": <;4 4' 4 -i j- *,+ ".;

, .:. -- ~ - - - : ; -;:.M ,

, y . ? . . .. 9-Wp. -*

,. ea e3

-s .g

. ;..--.< f...e r:ri.y..- ,O'I W '-%. M ",~."-fC ... 2. *

.. ~-

Y- e , . .

=Q -

1 e& . gs .

- ---n___ . .-=>u:

~

a

6. j. . .-.- *r., 1

' .n=* ' _ ="*yf,.,,r.,,. . _, :. [

.~~~~ "

tk;. .-_ :

w

-- t,\*3 f .,..

s.

.. . ~,ac.p. - , ]

,..-.g.

- - - Firare 10

~. - -

P.ns si::d seie:d o .to bo:::= and s; des cf :25te iny to tv :: sed to h:!d een ..: fiber buntet wrap en tray.-

g '. '

y s~.:5?Q,m..

z^( ..) Ql ~

. ~ . -

~---. ..

g; ' - 6 e

.~ -

E- M ,' --

. Figure I:

..-... . ,-/ .'. . O

, Arrangement of cable t sys in Portla d Ceraent Associati :

Beam Tur .act for 3r:e scale fee tes s.

. - , . 2 ._ -

- s.

i m-=

.. ~

- ~

wum, ,, m r' . . .

- t.y . .a. .

c.a .s. sua we.4

,e

~

se=' 1* e- - 7

-- O. C.

,,,. p

,,*r l- .:.,T.M .-

m V

,poa ;

_ m.l ---y axx' )

m e.l- * . :;; .s a' w

. =*. . .. p ,-

. ;. . .-=5

= - .

.=

sm , r -

W- . y,&.. .- .:-

. a .a w,,

u .; e

.m<. w w: m.: .m w.n ....i .

, p

~

..; .~,. . ... . -. ;

\-

c. s

'; y aa v

. = . . . . . . .... . w

• • /

A ' *p .. .:

o -~; *

,f 9 *- .

. F; rare 1 Botte= and s; des of c:b;e tr:y c rn;!etely wrapped with #-.,,!.

curs =k Cher bunket for ia p sca;e tes*.s. -

.- V 'r & ,~. < ~'

• .~~ .

Te5! Picred'Jte .,

f~~.u.c+1

.- 54 -

. ~

- r-.

Four tr:} s were tested in a ten fact section of c- e 2.ma o es.e us s us - w the Pct:1:nd Cement Assee:: tion Beam Furnsee. -

This furn:ce. with a r:ted input of 7.3 bill:on _. u.m . a*d Jcules (7 m!!! ion STil} *ae hour. could m: int:in u .. t e.c. > w t':e required hcstin;; rate desi;;nated in AST.'! E119. C wr.

,, , g m*u4

,, o, i, ,

'i" te four toys were fM*cd to 40~. with seven con. -, v y wwe e. - - ==== ,,;,3,=./p,.v ductor, EPP. insulated h> palon peketed c:bies. , y ,. .. . , , m, ,,, .,,, ,, '  ;..c, The four trays werc arr:n;;ed in the furnace as shown in F4ure 12. Two of the trays were sesled y,gure o A circu;try for monit:ri ; centinuity and s. hens in re; re. V at both ends (Trays No.1 :nd No. 4). The otLer two were seuled at one end only. This w:s done sent:tive c:t'.es dmn; tr.r:e se8e tests.

6. -

* ' ' ' **P- -w ...g. , ,,

i to determine if leaving the sea! out on one end tarding de heat flow into the tray.

would in:: ease the time before cable failure in the .

su o ,

" $TIyI*

• gg LArt eg To determine the time of failure, the continui. y) ties of 15 cables at the bottom, s. des. middle and 3.,

ecc ,, ,<

top of each tray were monitored dunng de test. ,. ,

intau (See Figure 13 for the monit:nng circuit:y). The location of these monitored cables in de tray is , [s,:.c . . , w r , , ,,,, s shown in Firae 14. In addit:en. represen stive cables were meggered dunng and after the tct to um wc m

• t*

obtain a =easure of the megch= va.lue. sies yet e W 1*J,T s,p;m

.- .I

• s HTC .M ' *C 6YC

.-- . _ - - .re . ,: . .or:

e "* e@# ,V n.- . I &

a4

~

g s*: .ms e . err c I t W asv=t: 4t

?

A# w'- [ r ^

,s wwa ,

top vre

~

%--W T v Y Yt] H ,'5: - - - - - --

re n Figure 1.s Ternpenture di:tn'bution thrcugh 5ta-ket byers and with.

iV Schecutie dbgram of approximate cable locations in bs Tra; No. 4 at 90 minutes into test. .

cable cays for meggenng and light panel momtont.g.

Tray No. 4 was ! eft in the hot furnace for Results and Discussion approximately 3% hours after the end of de test.

It was then removed, opened and inspected. In

. The sequence of cah!e failures for dis test is genera.. the condmon of the cables was good (See shown in Table V. Th; f::st cable fatlure occur:ed Firre 16). Some of the outer ja:kets en a few of at 94 r inutes into the test indicating that 75 mm 2e cables had softened. Howeser, c' e cf be (3 in.) of the b!a-ket wTap provi:: cd de 90 rninutes cables that had short '.ircuited during the test was of protection required by the NF.C. Note that the meggered :. gain and fc und to be free of short c:rc.uts.

three cable failu es occurred in tre lower trays, wtt3e no failures occurred in the upper trays. Ap- AMPACITY TESTS

, parently the lower trays had more heat applied since they were within the burner ih:nes and the Independent of the fire tests, ampacity tests apper trays were effectively shielded by the lower of de blanket wrapped cab:e bus ar.d cable ::ay trays. Also note that the cables that failed are all were run for the followin; parameters:

war the bottom corner of the lower trays. This

' " ~

puttern,.which occurred in the small scale tests. is Kaow ool pobably caused by the heat bemg applied from Service lead Insubtion turo directions (bottom and side). Smcc cable .

failures occurred at yproxima:cly the same time 1455 ampere 2 inches 750 MCM EP/Hypalon

. . in Tr:ys No. 3 7 4 No. 4, there is apparently no cable m cable bus for benefit derived Lj Icavmg out the seal at one end C.E.G. mainteed

- oggneg,37, The temperature distribution through the 3 inches g'

bkanket byersand within Tray No. 4 at 90 minutes Cab!c tray with power 12.6 wat s/

into the test is shown in Figure 15. These results cab!c (7">0.',1CM cable it.or ,

show how effective the bbniet insubt on is in re- used fer icading) 7 .

4

_ **v m wo- v *-.w. m. , . .

~

ever,'If a "very conservative ambient of 400C p,.:4 .-,.,g ..

.,g - Mj . .T . .M w.-t -

~

,;, ,". .. . ,.q:-' .

. . (1040F) is assumed, then the total temperature

-g.

*

,, , lyy - would be 910 C with the tray in the horizontal position. Since this exceeded the manufacturer's D4-, W y ." wr}< ? "', recommendation for cont nuous operation, the

~~

'.M ' . , .

k,"M[

i . , . -%,, i t 'I ' .'f \

.o r p

p decision was made at Zimmer to build a structural corridor around the cable bus rather than wrap it

[ ~ "* G -y ' di L]M]r,c. *M with ceramic fiber blanket.

4 'a The results for the cable tray are shown in i.bg "J,} *gg, . 4; . ' ' '

Table VII. The maximum steady state temperature

-: r d'.* : 4i $g  ; I achieved in this test was below 900 C, even if it is C' /C7 *: 9 4 L..',, assumed that ambient may rise to 400 C (1040 F).

r *3 f . l. .k,. Q ",P.T ..- '

kw Therefore, the ceramic fiber blanket wrap system r, 41

.: f. - , - is being used where required on the cable trays at

(..

. /,..m ',rikd.G kk p.

the Zimmer plant.

g? --

. COST OF M ATERIALS AND INSTALLATION u-w .u . .

. j Using the technique developed for the Zimmer b

f, '

f.

t t'

~ h Station, where the cable tray and lid are insulated

. g p. . Wd . l. Qe..h[.;

~

., . separately, it is estimated that the c >st including material and labor to wrap a 460 mm (15 in.) wide

. M 47" 'f-[t

• Q',W f if x 150 mm (6 in.) deep tray with ceramic fiber f.I7.? jrFfc..

, g.f. blanket is $14/ inch of blanket / linear foot of tray.

F. -r , .P i, j .

h- M ;' g

.h~! ,p"'mL ] ) . ','"

4,g, h~wu "' Qg ,

Th refore, to wrap a 3.05 rn (10 ft.) tray with 50 mm (2 in.) of blanket would cost S250. To wrap the same tray with 75 mm (3 m.) of blanket would be $420. The cost can be reduced by wrapping Figure 16 the lid and tray as a unit. However, this reduces Condit;ca of esbles in Tray No. 4 after test. the ease of access to the interior of the tray for inspection purposes or the addition of more cables.

For the cable bus, the actual feeder cable load of a diesel engine generator was used. For the cab!c CONCLUSIONS tray, the worse case loading for the safety related cables was utih:ed. Within the limits of this investigation, the Tests were performed by circulating the cut- results indicate the following conclusions:

rent provided by a current transformer, through 750 MCM EPl!!ypalon 3S3 grade cable, until '.hc 1.1EEE 383 qualified cables in an unprotect-surface temperature of the cable inside the en- ed steel, solid bottom cable tray (without closure reached steady state. The temperature lid) fail in a;' proximately ten (10j minutes measured included the ambient temperature which in an ASTM 2110 d"signated fire.

was also meuured separately. The tests were per-formed with the tray oriented horicontally and 2. Wrapping solid bottom cable trays with 25 vertically, and with both ends of the tray sealed. mm (1 in.) of ceramic fiber blanket (100 In addition tests were made with only one end mm [4 in.) overlap over butt joints) pro-scaled to determine the benefits of partial venti- vides forty (40) minutes of protection in lation of the mapped cable tray. The criterion for an E119 fire.

passing the test was to maintitin a steady state temperature in the wrapped tray below 90 C. 3. Wrappin;; solid bo* ,m or open ladder chich is the manufacturer's rated cab!c tempera- cable tmys and cone..it with 50 mm (2 in.) .

turc for continuous use of the cables. of cerande fiber bianket (,vith all butt ,

The results for the cab!: bus are shown in joints ti;;ht) provides sixty (GO minutes) of Tab!c VI. The tot for the desi;,n load passed, flow. protection in an E110 fire.

-8'

~

H .

1 .

[gr - -

1

4. Wrapping solid bottom cable trays (with (2) A.J. Pryor, " Browns Ferry Revisited," Fire lids) with 75 mm (3 in.) of cerarnic fiber Journal, pp. 85-86, 88 89,120121, May, blanket provides ninety (90) minutes of ,1977.

protection in an E119 fire.

5.The ceramic fit:r blanket wrap presides ,

protection for noth IEEE.353 qualified TA8t.E cable and cable not qualified as non propa.

gating accordmg to Ir E.383, O Fahr Wu 'm he No.1 - L' mW 5:M sot,4,o,i. cau, Tra3 and steel cone..:

6. Loose or open butt joints in the insulation rai'a may lead to early cable fa!!cre in engulf- Time ci,cou cable cabee ment fires. (N' W ' Number 1*c8na Inst.tanee a

0:00 Test Sun - -

7. Insulating cable trays with .,5

. mm (3 in.) 0.08 15 iny trEEcs3 sence.tx of ceramic fiber blanket does not merease 0:10 1 Tar Quaar.co the temperature of lEEE.333 qua!!fied 0:12 18, 19, 20 Tny Q wJed C'13, 3 Tny QWiried cables above the man _facturer.s ratmg for 0 13' 4 Tny qwgg continuous operation (90CC) when loaded 4:13' s Tny Qul.rw at 12.6 witts/f t. C:12' 6 Tar Quahr ed 3:13 Tny Qunr.ed 0:13' 8 Tny Q.sahred ACKNOWLEDGET!ENT 0.13' :s Tny son eaannee .

0:13' 17 t".ondut Non 4;a:.hed The authors wish to acknowledge the invalu-able assistance of:

$8 0.21 1l 2

Tny NrN Qwnfted Tny Q.aat.ed (47 0:22 10 1 Tny Quhred R.E. Cotta Sargent & Jundy, Zimmer Elect. Ns0 ,u , ,, g ric Project Engineer M.S. Abrams Director. Fire Research Depart.

• Faard danes pe=n fadun sad shutou=n ment, Port!and Cement Assoc:ation. Re-search and Development Construction Tech-cology Laboratories. Skokie !!1. TABLE !!

C.F. Es!dassarra - Schirmer Engmeerin; Ccrp.,

caw ra w se m a Tm % -

Zimmer Fire Protect:en Consultant steel sohd r.onom cme Tn, -

F.L Banta, T. Wall Husky Products Com- md Sted condan korped =un so em C rn )

cenu Fw un hose can Jsanto pany who aided in both fire and ampacity tests.

A.J. Staubit:- R.E. Kramig Co., who provided Fsned pinning of the blanket to the trays and ['1,

,$, "'$ , C*$',

,, gC',$,,

covers L.C. Albers CGLE E:ectrical Sponsor Engi. O co Test san - -

0.51 1 Tny tEEE.383 Quahf.ed UN 0:54 20 Tny QdifM T.t!. Mulhkin CC&E Technical Assistant 0 s; 11 T,3 y quanf#d R.A. Glsaby Bechtel Corpor: tion, Project 0:s9 3 Tay Quhfed 0:59 19 Tny Qsanfaco U* 3:01 18 Trey Qushfd R.M. Jenks - Esbcock & Wilcox, Ltb Manager 2:05 17 cocAn sone. ant.ed 1:01 5 Trey Qvahfed 1:10 4 Tny Quhrd 1:10 2 Tny Quahreed REFERTNCES 1:10 8 Tny Qushre 1:10 E Tny '

Qushf na I R.G. Sawyer and J.A. El net, " Cable Fire at 1:10 la hr = N'a4 *"*J (1) y Browns Ferry Nuclear Power Plant,,. fit.t 1:10 1:Is 12 7 Tny Tny Quahf.ed Qu.nrw JoitrmL pp. 510, July,1976. 1:17 'A at Term.nated - -

3 9

k TAntE Ill . TABLE V Cable Failun Sequence in Test No. 3 AJeminuan Open Ladder Cable Tray and Cable Failure Sequence in Large Scale Aluminum Conduit krapped with 50 mm t:in i Test Steel Solid 11ottom Cable Tray Cenmic Fiber Banket ITight Butt Jointst With Lid Wrapped With 75 mm (3 in.) ,

Faded Ceramic Fiber Blanket firne Circuit Cable Cable Otr/%M , Number Loca tion Imuistion FaBed ow Tat Stars - -

Trsy Cable (Hr/ Min.) Number Number Comments 1:01 1 Tray. IEEE 383 QualJied 1:02 16 nsy Non. Qualified 0:00 - . Test Started 1:05 19 Trey Quahned 1:34 3 15 Shcrt Circuit 1:06 2 Tny Qu,gx 1:36 .4 6 Short Circui:

1:os =0 Tny Qucfd 1:41 4 14 Short Circuit Beginning 1:09 5 Tny Quc'2.d 1:42 - - Test Terminated 2:o9 3 y,,, . q,med ,

1:10 4 Tray Quae d Id1 18 Tray QuaVaed 3:13 '"en Temur.ated - .-

e TAE'.E V1

~

7ABLE TV Results of Ampacity Testing of Cable Dus Cable FaDun Sequence in Test No. 4. Steel Solid Bartorr Tray Wrapped With 50 mm (1-in.) of Ceramic

-Wrapped ith 25 mm il m. Cersmic Fiber Blanket i100 mr, c4 ina Overup Sinp Oser Buti Jointi and Steel Conduit mich Fiber BEnket - Q 3a ram tib in.) Pipe Insulation Faned Time Cirevit Cable Cable Ampere Tray Ends Temperature (oC)

Otr '%n ) Nu mber tration Imut non Load Orien t a tinn Se Id A m bien t Ri<e Tots!

0@ Test $f.est - -

0:40 1 Trey IEEE.353 Quahfied 69 Qualif;,e 1500* Horizontal Both 18 51 c:4s :0 iny c:4s 19 Tr y quanfied 1200 Horizontal Both 18 75 93 i c:4s 17 Conduit t! n.Qua:1.d 1500 Vertical Both 24 14 38 0:50 9 Trey Quanfied 34 51 I

c:52 15 Trey Non.Quahr.ed 1800 Vertical Both 17 0:$a Is Tray Nonquahid 1800 Vertical Bottom 16 18 34 0 13

. 2 Tny Quhfied 0.54 18 Tny Quahtied c ss a Tay quanf4.d

• Load of 1500 ampere represents 500 ampere per l o:ss 4 Tny Qu.nfied phase in 3 phases while 1600 ampere represents

' 600 ampere per phase in 3 phases. Design load is IN 14 No ed 486 ampere per circuit or 1458 ampere for 3 phases.

1:00 Tec Ter:ninated - -

f i

e 10

~ - _ .

m --r