ML20195J758

Tests on Silicone Rubber Insulated Control Cables

ML20195J758
Person / Time
Site:	Sequoyah
Issue date:	07/20/1987
From:	Eldstad E CONNECTICUT, UNIV. OF, STORRS, CT
To:
Shared Package
ML20195J745	List: ML20195J742 ML20195J758 ML20195J772 ML20195J803 ML20195J825
References
LB-751381, NUDOCS 8801290007
Download: ML20195J758 (43)

Text

-- - - - .,

p totC Carst Att l*

j a 1" oes

si"; . , , ,

a .

- p ,i p N

' - 's , ,7 s

' N

• a. .. s i,

{te ',/'

('.y ' 48*7, * ~

b . x.

IMS <")

['y' C 7 y\_

Institute of Materials Science I '

b

, . b-( vos.mor .t c m.

i.h ',

1 i

TESTS O'! SILICONE RUBBER INSULATED CONTROL CABLES TVA rei arence no. LB-751381 Project: Sequoyah Nuclear Plant r

1. _

l 3 l .

l 1

hb S

AOK O 27 PDR

... 4 j,

, ; , ..r , ;

>+- '6%;-

~

x;- p

}

l V \

i 1

Electrical Insulation Research Center

. Institute of Meterials Science University of Connecticut -

Storrs CT 06268 t

1 6

I TESTS ON SILICONE RUBBER INSULATED CONTROL CABLES i

l TVA reference no, LB-751381 l Project: Sequoyah Nuclear Plant i

i I

t i

i l-Prepared by: E. Ildstad

, July 20, 1987 HE4O OPte d'M Ef tf 4"(b (PMSP %%f994g & (P8htgQ ye

'\ ,

t

Project Team Principal Investigators M. S. Mashikian

- J. Groeger E. Ildstad

. Research Assistant P. Wronski 1

i (Wi+4 4 8%v e 4* Agee d's. c,~,. I M S u-4~, c mi

July 20, 1987

SUMMARY

Visual, dimensionsl, electrical, mechanical and spectroscopic measurements were made on four service aged specimens and on one reference

. specimen.

Dimensional, electrical and mechanical test results were comparable for all specimens and provided no indications for the cause of failures during TVA testing.

Results of both Infrared and X-ray Emission Spectroscopy showed the absence of contaminants in the polymer insulation. If there had been any evidence of the cause for these electrical failures in the direct breakdown channel, it would likely have been burned during the in-situ 10.8 kV DC test.

Visual inspection of the jackets of the previously installed specimens displayed the expected signs of abrasion.

The cable specimen which exhibitid high leakage current during the in-situ testing was found to have minor strface crazing or "chicken tracks" over a three inch section. These surface di3 continuities were later determined to be less than 5 mils in depth. Located underneath the crazing marks several deep cracks were noted extending upward ? rom the conductor surf ace. Some of these cracks were determined to have penetrated more than 75% of the insulation wall. Additionally, it was noted that in areas exhibiting both of

- the above phenomena that the insulatior did not tightly adhere to the conductor strands. All other portions of the specimen examined tended towards a much greater degree of adherence between insulation and conductor.

A single interior crack was noted on one of the faulted specimens near the discharge channel. Minor crazing was noted on this specimen in the vicinity of the fault and the insulation exhibited a lack of adherence similar to that noted above.

The remaining two f aulted specimens appeared free of observed interior cracks. One of the two had minor surface crazing in the vacinity of the fault and both exhibited a lack of adherence of insulation to conductor near the fault similar to that noted above.

I I

1 i

(me< a = 4 oa aewe1

• C e= e' l M b L - .e.u, cr Co--co,,,.

l l

4

1. Introduction

• Tennessee Valley Authority (TVA) submitted four specimens of 1/C, 14 AWG copper control cables purchased on TVA contract 77K5-822502 fromThe American cables Insulated Wire Corporation and used at the Sequoyah Nuclear Plant.

were rated at 600V and had silicone rubber insulation with asbestos braid jacketing. The specified nominal wall thickness of the silicone insulation a and the asbestos braid was 45 and 40 mils respectively.

One reference specimen, in an unaged condition and of identical construction,' from the same contract, was provided along with four specimens of service aged cables. Three of these had failed and one had exhibited Cable #4 high did leakage current during an in-situ hipot test performed by TVA.

not breakdown and the others had failed at various voltages as listed in Table 1.

Table 1. Inventory of received cables.

Cable Results from DC Hipot Test

• Identification TVA Label Reel No.

1 2V5SS7B, G1AN 9A-283-3 Broke down after 1 min at 10.8kV 2V5888B, GIAN 9A-283-3 Broke down upon reaching 7.5 kV 2_ Broke down upon reaching 10.0 kV 3 2V5S893, G1AN 9A-283-3 4 2V53593, G1A3 9A-283-3 High leakage current at 10.8 kV 5 Reference specimen 9A-338-6 Unaged, not in-situ tested

• Testing dor.e by TVA EIRC agreed to perform cable testing and provide a test report according to the detailed requirements given in Appendix I.

2. Visual Inspection of the Braided Cables l

l No significant physical damage, such as holes, tears, or crushing of the braids, could be found, f The four cable specimens removed from service showed signs of abrasion.

Some asbestos fibers had fractured and the end of others had opened up, resulting in protrusions. Comparing figures la and Ib it can be seen that the reference sample had f-- fewer protrusions.

The surface of the never installed reference sample was also darker in since it is color. This color change does not appear to be of significance, probably a result of the pulling lubricant used during installation.

A more detailed test report from the visual examination is given in Appendix II. The overall specimen length and the positions of the different surface marks were noted. The positions were measured as distances from the end marked "terminal end" by TVA.

2 w . > .. .. . . . , . c. ~ . 1 M s n... . , e .-.m .

1

. 1

.. )

l l

F, .

l. o U,

Figure 1 (a) - Appearance of a portion of service-aged cable No. 1.

l j

[

6

u

?,fD 2 s .

Q l

v t :

b

)

~

A .,

.. )

m . . ..-

,__:rk a..L.. a . - - . .

Figure 1 (b) - Appearance of reference cable surface -

Photographs taken by Nikon macroscope Model 69059 3

i .. , . .. . . . . . . . . , . c . . . i v s r . ... . . c o... . . < .

3. Fault Location Instrument:

A small hand held Tesla coil, with a low current, variable voltage output (10 to 50 kV), at a frequency of approximately 0.5 MHZ. was used to generate the corona.

Procedure The Tes'la coil was attached to the conductor of each specimen by means of an alligator clip and a grounded loop of wire was passed along the outside of each cable. The diameter of the wire loop was about 10mm. Since this was a corona effect test, the voltage across the cable insulation could not be well defined. The voltage was increased uncil audible corona occurred in the Electric faults, pinholes or airgap between the cable and the grounded loop.

weakened insulation in the cable were easily detected as a visible arc originating from these sites.

After the faults were detected with the Tesla coil, the jackets were closely examined at these points, again, for damage etc., but none were noted.

Results The experimental results are presented in Table 2.

Cable Fault location.

Identification distance from "terminal end" Remarks 1 54 3/4" One discharge site 2 27 1/2" One discharge site 3 109" One discharge site 4 103" -

106" Numerous discharge sites 5

No discharge sites e

4 i .s . .. . . . ..e c .. .. . I M S r . .. e . c . .. . . ..

4. Jacket Removal and Measurement of its Thickness

- It was extremely difficult to remove the braided asbestos jacket without damaging the silicone rubber insulation. Several techniques The were attempted techniques which before the final, effective method was discovered. ,

failed are as follows:

i (a) Use of miniature diagonal cutter to slowly cut through the fibers of the jacket. This was extremely tedious and, at times, ineffective, as damage to the insulation could not be prevented.

(b) Use of modified small surgical scissors. This method was slightly better than the previous one but, again, several longitudinal incisions were inadvertently made in the insulation.

The successful technique, which was finally adapted, consisted of pulling the jacket off the cable with the help of two teams, one holding the insulated conductor, and another pulling the jacket off. The operation was fast and did not harm the insulation. Table 3 summarizes the techniques used to remove the jackets off each of the samples.

Table 3. - Summary of Methods Employed to Remove Jacket Cable Jacket Removal Technique Identification I Cutting with miniature diagonal cutter 2 Pulling 3 Pulling, except for 15" of cutting 4 Pulling 5 50T. pulling and 50% scissor cutting Jacket Thickness Procedure 7

I The jacket thickness was measured by means of a micrometer (STAR 436 - 1 inch) at three different locations: in the vicinity of the terminal and cut ends and in the middle of each cable sample.

Results l

l The jacket thickness values are shown in Table 4. All of the measured thicknesses are larger than the specified nominal and minimum wall thicknesses l ,

of 40 and 32 mils, respectively.

5 i ..u . . ~ . . . ., . c . . . . I v 5 r.. . ., c ..,m .,

m

2 Table 4. Recorded Jacket Thickness Cable Jacket Thickness (mils)

Terminal End Middle Cut End

- Identification 1 42.8 43.4 42.6 2 43.4 48.'O 44.0 3 44.1 43.2 43.0 4

42.1 46.5 48.4 5* 40.9 43.2 43.2

• Since cable #5 is the reference cable, the relative references.

"terminal, middle and cut end" are not applicable.

5. and 6. Visual Inspection of Insulation and Tesla Coil Testino Following jacket removal it was noted that the insulation on four service aged cables was generally darker than the reference specimen. It was also noted that the service aged cables appeared "cross-hatched" with black lines.

Closer examination and light rubbing showed that both the general darkness and the cross-hatching were surface phenomena only and appeared to be remnamt of

• braid impregnant most likely transferred there during pulling.

To reveal the locations of faults in the insulation, the Tesla Coil test was repeated in the absence of a jacket. This test revealed faults at the same locations as those noticed before the jacket was stripped. Detailed results are given in Appendix III. However, several additional cuts and gouges were observed in sections of cable which were stripped by cutting through the jackets. No such defects were observed in the cables stripped by pulling the jacket off. It was, therefore, concluded that these additional defects were created during the stripping operation.

The f ault sites were closely examined to determine if the causes of failure could be understood. Photographs of the faults are shown in Figures 2 through 6.

If any potential causes of failure were evident in the fault channels, they would have been destroyed during the in-situ dielectric breakdown testing, except in the case of cable 4 which showed only tiny pinholes, but no large electrical fault channels. These pinholes may either have been induced l

at weak or degraded sites by the high voltage present during the use of the Tesla coil, or they were present before the corona testing.

l The surface of the cable insulation in the area of these pinholes was l covered by "folds" or "chicken tracks", as shown in Figure 5. These chicken tracks were in alignment and appeared on one side of the insulated conductor only. The picture shows the "chicken tracks" as bold and distinct. This is due to the presence of saturant in the cracks. Light rubbing of the insulation surface re noves the saturant and results in the cracks appearing less distinct but does not alter the actual discontinuity.

6

% . .... .. . . .s c . ,... m s s ..-, ,, c . -.m ..

l ._ -

• .k:

s , .,

t _t 4.

7

n. . . . s .

4' d t~ *

. ~

v 1

l 4

V =

a n.

gM s ip .

s . .

Figure 2. Electrical failure, cable #1 (2V58878 G1AN)

.w.

• g

.. e

, 'e' . T. . .

. :. . y;lvId .

.. c .

. g 4,. -

.a C :m

~:

Figure 3. Electrical failure, cable #2 (2V58S8B G1AN) 7 i ,. .. . . - i . .. ., . c . i u s s . . . . <.m. .

\

1 l

I a

• i 1

%.'f,

1. .

. tAni c

-( '

s. -

1 * #. . , *. ~

.. .fl '

l

• 4 .

g, w .

'Q - -

..- .. .~t.q,J.: , y. I' .

g. . .-Q;  ;- 7.

! . . ; ',. y, *n".w +,.,: .^ ', ; .

.s .

.2 I L -.A ..Ig. %,s.3 . : .

x' ,

• 4 .

. a

, 11 s'

rt

' w

.4

.. 94 > , - >.

s. . '

~

yr<*.w. ..

g .. ..

h Figure 4 Electrical failure, cable #3 (2V5889B G1AN)

L 8 ,

, , . . . . . . . .., i u 5 . . ee-. <.

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

}

i i

1 m 1

., e.- _.

2 g- l t..

6 u 4p ..

< ,.M;;6  ;

.. e >

'-  ;

Figure 5. High leakage current region, cable #4

(?V;RRQR re1 AU

-= p':.g w.-

a ..a.- - . .

fk ,I,M. ,' Y.S ** 't ,7 'I * "k. . '.h '

wc;.~% . ,. .

wy,. '.

~

.c . 4 . x, . .

h w :.~. \

s ;v,,,eQ. Og'*'

[ tu

1. fip...w.e. . p:-- %;in!

l

'y#yk w ; ., spa. .

2: . L r

<.+

.- ..;fi , . . s# .

r).

-;b.

. .g.

..- L i

r I

Figure 6. Reference specimen, cable #5 l No electrical failure sites  !

l 3

p l

l .

1 i 9 l i - .. , . . . .. .. . . .. . , . c . . . . I M 5 . . . . - c o -~ .

l

A total of 10 pinholes were noted following the Tesla coil test in the "chicken track" area of cable 4 This site was, therefore, of significant interest to the investigation and the insulation surface from this area was subjected to scanning electron microscopy (SEM). Samples were prepared by cutting out a small section of the defective area and evaporating carbon on its surface, in preparation for the exanination by AMRAY SEM Model 1200.

Figures 7(a) through 8(b) are scanning electron micro They represent magnified views of the "chicken tracks' adjacent graphs to the die of the sa marker of the cable. This examination revealed that the "chicken tracks" It is not possible to turned out to be rather shallow surface cracks.

determine their exact depth by this technique. The examination of the damaged area of cable 4 was therefore continued in the following manner:

A 5 mil thick, approximately 1 inch long sample was longitudinally cut from the surface area containing "chicken. tracks". (This was done by using a Bridgeport mill, a technique to be described later in this report). This sample .es stretched while examined with a stereo microscope and it was determined that the surface cracks did not penetrate this sample, thus indicating that the surface cracks were less than 5 mils in depth.

4 1

l l

I $

10 l . ,,.4... . ....s.c, ..I M 5 - ..-co-- - .

l i .

, .. 'i . (.. _:

,*s.- g. .s l .- -

,e

~

. g . . -

. -. , .es _

.., ' . ~- .

~

u. . .

~ '

1 .. ~ r ~ . is K '~ ,

-p*.- "

. .-??

h . s 'er - '.

- ' ~ . p *-

g g ,4 -

m

. -~

y,

.p .g - w ,,.. a

. , n. _

4, ,

. f ', ( .;

, , '.-s .

v- . . .

'.k f ' _ _ r. .

1-, ,. - ; .

_[ .

1. -

~, . e,_-

$. n.

a. - - -

• /. .e ,g;e '-

s-  ;

[. . . . -A[ w _

__% . t

'. , 9. -t . t. .~ . . . '. .

.+ i .,

- %. ..g ,

r A. .

'.g steps.*' Wate-~ . - .-38kV,>% l'V'f,

• g 4 a

a) Surface crack through the die marker m.- -

.l & Q Ys'g 4pr',,g.- .a . .c Y , . -

7

~ '*

i. ,.

. w- .

• • . .S ,

% 3

• p ?~

) *r s-*.e r' .. ,G-

,1

.o 3

. p .

y.. G.h.. %% __%/ 1

.y ;.. . '

,.<* .* .? .- .

J $. .

- C .

. .. . . . o f' ~ .

- . .'. '- \ , . .

y, b ,~? ,

,. j, -'

,. r 3 - , ./ , _a N =

^

-$A,. g . .-

J,..'-

e

. . . . }. ;,s *v: ,.s - . p-

_~ -

.s . .-

l SStpan m259 ;39tv

'. ,. ., f,. .

.. s. . * -;r . '

., e f,, g As - .

b) Surface crack higher magnification Figure 7. SEM pictures of surface cracks, cable #4 1

11 i e..<.............. .c ....t M s < ,,.,,. .c ... ... .

l l

, _ _ - , _ , . . , , , . - - . , -w. - ,,,.w7-----3 -w,r

i e

<t h .

'/ '

• i a .

', qs- .

e s .,

v .,

N .

. f~ . S ,

'~

o

.e - .

i

. g:.

a .

n

.. e t . . _ . . ,. ..

! 4

. 7.

~

-~ , . ' eg

^'s i

l ..- , .

. . r.

lk e

l toep maae 3skv _ _ a4  ;

a) Surface crack adjacent to die marker

~ .

t

~

M ~

\

i - * ,

f

- e i

6 e

.k b) Close up of the same cract Figure 8. SEM pictures of surface cracks, cable 4 t

i 12  :

,..,....,........<....as......... ,

l l .

I u_ _ __ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ , , _ _ _ __,_ _,_ _

It was observed during the Teslo ceil testing that the punctures were typically located between the surface cracks. In order to find the reason for

' this, we looked to see if the interior of the insulation could provide any clues.

Faulted Area Cable 4:

' Several 1/2 inch long sections of insulation were removed from the area containing "chicken tracks". This was done by making radial cuts with a scalpel, about 1/2 inch apart, and then a longitudinal cut on the side of the insulation opposite the "chicken tracks". The insulation was then opened and the conductor strands carefully removed. In the area with the "chicken tracks", the insulation did not stick to the conductor strands and the insulation was easily unrolled. Visual inspection of the conductor side of the insulation showed that no part of the insulation was missing, but flattening the insulation revealed several deep anomalitios or cracks. located just underneath the pinholes detected by the Tesla coil.

Pictures (taken with a NIKON Microscope Modci 19059), showing the cenductor surface before and after being flattened between microscope slides, are presented in Figur e 9.

In order to make an estimate of the depth of these "cracks", the thickness of the remaining insulation wall was measured with an optical microscope (NIKON/BIOPHOT). First, the samples were flattened and the inside of the cracks were stained by using a thin Indian ink felt pen. Then, microscopy samples were prepared by making a longitudinal cut through the cracks. The staining was found to give good contrast between cracked and non-cracked material and the thickness of the remaining undamaged insulation was measured at the site of 6 different cracks. The results are presented in Table 5. It can be seen that some of these cracks penetrate more than 75% of the insulatien wall.

One cpecimen removed from this "chicken track" area had no pinhoies revealed by the Tesla coil testing. As can be seen from Figure 10(a), no cracks were observed at the conductor side of this sample.

l t

13 s e ~ . . ,. m. . . ., c .~ . I M S r .. . ., o. c ..m .,

\

i 4 l i

. l 4

J

~

%- A..

l i

s -. ,

! i t

I.

1 a) Insulation section as removed frocn conductor strands i s

t I

I I

t L

-^~ . a . .. t

@f .

I 1 t 1 ., '

\ 7u  :

,. q, W.  % , ,., -

.) . ' ci . , _ %  ;

F i

' i l

b) Same insulation section but flatte,ed Figurc 9. Pictures sho'ing the conductor strands and an ,

unrolled insulation sample from faulted section of cable 4 ,

N l

i .

1 1

i 14 t u;,,, ..,c...-. .u.. 1 m s e , , a u . .. . . - \

- t r

l

. _ _ _ _ _ . _ _ . . _ . . . _ _ _ _ _ _ _ _ . . _ . ~ . _ . - ,

i r i ,

!.i .

,i I

I i

i .

t 3

a

~

> %;.;i.

i l 3.':(-

r l

I s ,

1 1

t I

a) From the "chicken track" area, without electrical pinholes l

e te-m I

m .s.  % .q - .

i i

ww -

~

I i

^

~w, _ _ . .

w' .

i 3

i 2 l

I i

l l b) 2 feet away from the faulted area ,

e d

Figure 10. Pictures showing flattened l Insulation samples from cable #4 .

1 a

il 1

,I

• p i

!. 15 I

i ~.s.. ,.. o-a, .,.c.-- 1 M S 6 .,,,,c...n... ,

i

A Table 5. Results from Measurements of Remaining Insulation Thickness at Pinhole Sites. Cable #4 Insulation Thickness Crack No. (mils) ,

1 12 2 10 3 8 4 10 5 10 6 20 Mean Value 12 Non-Faulted Area. Cable 4:

Insulation samples were taken 2 feet away from the area with the "chicken tracks" from the terminal end and from the cut end, and a total 1 ngth of 3 1/2" was ex6 mined during this random sampling. No cracks similar to those '

under the chicken tracking were observed in these sections of the cable.

Another difference between these areas was that in the area without "chicken tracks" the insulation at some points stuck to the conductor strands.

Although the insulation was slowly unrolled and care was taken to carefully '

unbond the strands, it was not possible to remove all the insulation. It can l

be seen fron Figure 10(b) that as a result, some part of the insulation is missing from this flattened sample. ,

Cable No. 1, 2. 3-Insulation sections from these specimens were examined as close to the electric f ault sites as the remaining cable specimen allowed and in random locations away from the faults. A total length of 4-6 inches of cable was examined and the insulation was found to be loose only about 1 inch adjacent to the fault. Away from the fault the insulation stuck to the conductor. As typically sho.n in Figure 11(a), it was not possible to remove the insulation in one piece. Except for a single crack immediately adjacent to the electrical fault in cable #3 (see Figure ll[b]). no cracks similar to those observed in cable #4 were noted.

S i .

I6 ,

.. . . ., . c . .. I M S r. . . c..,4.,  :

om . r 4

- - - - ,n - - , , - -

! e

>r *L & +

A- 1 %e v -

s, ,, _.

i 1

.wjp%.. .

n. - ...,p i

I a) Conductor strancs and unrolled insulation of cable #1 l

Away from the electric fault.

i t

.s . , .

j

- *

• e v g l -

a s

4.

A,N .J i .

• s . L . -,, ;. : r .:.- . c ,

~ .g. . 3 ,

} .3 p. .

m

-4 .

w --

g

, '^' % , -

-- w.

l 2 . ~%, ' % : ;  : :

' --' 34 . .

y. . .. . ,3 x . 6 ' ,  ;

{' ' [%-

• [ e g y ,. .  ;

. . ., a -s y ,y. , , _ -

c

,b .,

b) Flattened insulation immediately adiacent to the i

electric fault of cable #3.

I Figure 11. Pictures showing conductor strands and '

unrolled insulation of cable 1 and 3.

1 8

l i

17 i m. 4. . ,. . . ., . c ,-... i M s r ,,,, c m,,,4..

l 1

i

_W9 - ww--w=erve-war-rs-----=-=------ . _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _

~

Reference Specimen, Cable 5

- This specimen was studied in random locations and a total length of 3 1/2 inches was examineJ. As can be seen from Figure 12 the insulation stuck to the conductor strands, but no cracks were revealed on the interior surface of the insulation.

I 1

3 I

4 i

O f

I l

l t

l l

l l

5 18 i eim , . .. ...c,,,.I M s e....w, ,co m , u

'"-e t w w w -ew. -+se---. , , - , _ _ , , , _ , . _ _ . _ _ _ ,

a

s. '

. #+-

m .4 h.

, -7, ; :w L

'o f ELX .,

i I

I -

,w-.~. . n p.,a q~

a) Conductor strand and unrolled insulation sample l

~

l .

l

\

.yr"~r-ssg: w- m n V.w.m 9 y . . . ;, . . g.2 g. y d.<,s.- w l

ng,  :, w ,- - .

.2  :

k MF % ,

' ~M?.E,wJ i w-x.

,pc- .. __

,,s

?

.4

• t k'- +

L !.?My - l '

'WWM %W.; .-M?##

, . - .. ..ww i ...,~..r.v?>.gxra:4;.r;y.;., q t

i b) Same sample as in a) but flattened 1

1 Figure 12. Pictures showing typical conductor strand and insulation l

samples of reference cable #5, after unrolling of the insulation E

l .

s 19 m ~..... m .,,....c,,,.i M s < ...... <,.. ,,

-wy-www--,-

o

7. Measurement of the Insulation Thickness Experimental Procedure The insulation thickness was measured close to and remote from the electrical fault locations. By making radial cuts in the insulation, it A 1 mm thick became possible to remove 1 cm long sections of insulation.

radial sample was then cut from this section for microscopy examination. in each sample, the~ insulation thickness was measured in four positions 90' apart.

The light microscope used was a NIKON/BIOPHOT at 40X magnification and the calibration of the ocular scale was measured to be 1 div = 1 mil.

Results The results of these measurements are shown in Table 6.

Table 6. Insulation Thickness in mils.

Cable # Position Remarks 1 2 3 4 1 53 52 54 57 10 mm from electrical fault 62 57 52 56 28 inches from electrical fault 2 58 53 50 53 10 m from electrical fault 51 53 56 49 22 inches from electrical fault 3 50 52 57 52 10 mm from electrical fault 57 54 58 57 22 inches from electrical fault 4 58 52 51 57 10 mm from electrical fault 53 55 58 54 25 inches from electrical fault 5 55 56 57 55 Reference specimen 56 60 56 56 All insulation thicknesses were found to be well over the specified values of 45 mils cominal and 40.5 mils minimum.

~

n o

20 i -, < . .. . , ,,,, . o c , .. . I M s v ...,4, , c ,,, ,,,

_ _ _ _--__m_ _ _ _ ___ __

8. Microsample Testing for Contaminants and Mechanical Strength Some experimentation was required to make thin sections of the insulation for Infrared Spectroscopy and Tensile Testing. A standard microtome did not give enough support for cutting silicone rubber insulation. A Bridgeport mill equipped with a sharp knife was successfully used to produce the specimens needed.

Infrared Spectroscopy An Infrared Spectrum of a material is obtained by transmitting Infrared energy through e thin section of the material while monitoring Infrared absorbance as a function of wavelength. At characteristic frequencies the different chemical bonds in the sample will absorb the IR energy at specific wavelengths and less is transmitted.

The purpose of this study was to check whether the faulted area of the cable specimens had higher concentrations of contaminants than the areas away from the electric faults. Also by comparing the IR spectra of the installed cables with that of the reference cable, it would be possible to detect any chemical changes in the insulation due to aging.

Sample Precaration Specimens with a thickness of 5 mils were cut radially through the electric faults of cables No. I through 4 The IR spectra were taken in two locations for each cable: as close as possible to the fault channel and in the non-faulted insulation about 2 mm away from the channel. On the 5 mil thick sample radially cut from the reference cable (No. 5), the spectra we-e taken close to the conductor and close to the outer insulation surface.

Spectrometer

!1 - Nanometrics Model: Nanospec/20 IR Microspectrophotometer Sampling windos: 140 x 70 u=

2esults The IR spectra obtained are attached in Appendix IV. When evaluating the spectra, one has to keep in mind that the insulation close to the electric fr., ult was heavily burned due to the high voltage DC test and a white powder i

had developed in the fault channels. The infrared spectra taken close to the l f at,l t channel, therefore showed less transmittance than in the non-faulted insulatton. Tha spectrum of sample No. 2 also indicates a higher concentration of filler particles in this region. Unfortunately, sample No. 3 turr.ed out to be closer to 10 mils than 5 mils thick and very poor spectra were obtained.

The results can be summarized in the observation that there is no I discernible difference between the IR spectra taken in the non-faulted area of l ,

the 5 cable specimens. The IR spectra developed from the electric discharge channels do not reveal any evidence for the cause of the breakdowns.

\ .

21

. . .m . -,. .m. . . m . ... . I u s s . .. , . m. .m . m,

s X-Ray Spectroscopy The principle of X-ray spectroscopy is based upon measurement of characteristic X-rays, generated by the different elements in a sample that is bombarded by e beam of high energy electrons.

The X-ray spectroscopic examination was also intended to detect "tell-tale" contaminants in the fault channels and in areas removed from the 'ault.

Samples were prepared by making cuts through the faults and then evaporating a thin layer of carbon onto the exposed surfaces. This is a standard procedure and the purpose of the carbon layer is to conduct away the electrons while the sample is subjected to X-ray emission spectroscopy.

Unfortunately, the vacuum system of the SEM/EDX system at the University laboratory was not functioning when this work had to be done. The X-ray spectroscopic work was, therefore performed at the Connecticut State Police Forensic laboratory, which has the same equipment. The instrument was operated by a principal investigator from UCONN/EIRC, and prior to taking the spectra, the EDX system was calibrated using a NBS Inconel Standard.

Instrumentation: Amray SEM model 1200 E0AX 9100/40 Excitation potential: 30kV _

u Live integration time: 500s Results Detailed test results are given in Appendix V. The elements found in the non-faulted insulation were mainly aluminum (Al), silicon (Si) and some traces of copper (Cu). Except for higher concentration of copper, and some traces of tin (Sn), there were no discernable differences between the non-faulted insulation and the electric fault channel.

Tensile and Elencation Test For tensile and elegation testing a three-inch section was removed from each cable, adjacent to the fault, and used in sample preparation. The purpose of the test was to show if aging degradation, voids, or bond ruptures had altered the mechanical properties of the installed cable specimens.

Follo*ing the ASTM D-1708-79 procedure, several 8 to 10 mil thick samples were made by taking longitudinal cuts from 2 1/2 inch long cable sections.

Prior to testing, the instrument was calibrated by using 0.5 lb. and I lb. NBS weights, shosing that the linea-ity was within 5%.

Instrumentation: Instron TTC MI Serial No. 3031 Crosshead speed: 5 cm/ min t Chart speed: 5 cm/ min l Grip distance: 0.439" (0.411" for cable No. 1)

The detailed results are presented in Appendix VI. The average and standard deviation values are given in Table 7.

22

% , ..,. .. .. . ..,, e c , . ,,. I M s v. . . ., ,, , c .,< . 4 .

Table 7. Results from Tensile and Elongation Tests Total Number Tensile Strength (psi) Elongation (%)

Cable i of Specimens Mean Value i ST0 Mean Value i ST0 1 8 1731 1 174 520 1 50 2 . 7 1957 189 453

• 45 3 8 1933 i 219 485 i 49 4 8 1820 1 253 421 1 63 6 1743 i 316 482 79 5

It.is worth noting that direct comparisons with manc?acturers' tests are not appropriate due to the differences in the configJration of theThese test specimens, in crosshead speed and type of grips used during the tests.

tests are, however, valid for comparison between these five specimens.

9. Measurements of Insulation Resistance Measurement Techniques The remaining sections of good cables were immersed in grounded tap water and the insulation resistance was measured between the conductor and ground.

The applied voltage was 1000V DC and the resistance was measured one minute after applicaticn of the voltage.

Since the first reference specimen (cable No. 5) was rather badly damaged during the removal of the asbestos braid, a new reference sample (cable No. SA) was prepared for this and the following electrical tests. On this specimen, the braid was removed by pulling.

Instrumentation: HP - High Resistance meter 4329 A.

The results are shown in Table 8.

Table 8. Measured Insulation Resistance at 1000V DC Cable # Total Length of Resistance (0)

Good Cable (inch) 1 32.7 7 x 10 ll 2 97 7 x 10 ll l 3 105 5 x 10 ll 4 104 5 x 10 ll SA 100 2 x 10 ll l

23 i m,,< . , - .. .- i, m . c .-,, . I M s t . . .~. , c o e ,4 ,

10. DC Withstand Test at 10.8kV The purpose of this test was to check if the remaining section of the cable specimens would withstand the in-situ test voltage at 10.8 kV in grounded conditions.

Apparatus Spellman DC - high voltage supply 0-15Kv Series resistor 10 Ma High voltage probe 1000:1 Fluke 8024B multimeter (For calibration chart see Appendix VII).

A sketch of the experimental set-up is shown in Figure 13.

CACLE (N GROUNCED WATER B ATH R = 10 MOHMS HIGH - VOLTAGE 115 V DC - SUPPLY j e

/ R f

HIGH VOLTAGE PRCBE Figure 13. The experimental set-up Test Procedure The voltage was raised to 10.8kV within 45s and after 5 minutes, slowly j reduced to zero within 10s.

l

( 24 i ,<, . . .. . .. . . . . . . c .. . . i u s . . . .. . c o. .,m .,

e Results

. The results are summarized in Table 9.

Table 9. Results from 5 minute DC testing on sections of good cable at 10.8 kV Cable # Result 1 Passed 2 Passed 3 Passed 4 Passed 5A Passed

11. High Voltaae DC Breakdown Test Apparatus: Hipotronics charging unit V - 100kV Smoothing capacitor: 0.5 uf Hipotronics High voltage capacitor Current limiter: 200 0 resistor Voltmeter: OC voltmeter (built-in) Hipotronics control unit Water bath filled with grounded tap water.

A sketch of the experimental arrangement is given in Figure 14 CABLE SPECIMEN R = 200 0HMS HIGH VOLTAGE 220 V DC RECTIFIER o.5 m ero F N \sGROUNDED TAP WATER VOLDAETER Figure 14, Sketch of the experimental set up 0

25

. . . .. ~ . . .. . .. . c . .. . . i M s < .. . , ~, , c o.. . o . ..

t t

e Test Procedure kV The voltage was increased at a rate of 0.5 /s and the breakdown voltage was recorded within 12kV. In order to avoid electric flashover at the terminal ends, the maximum voltage level of the experimental set-up was limited to 80kV. After breakdown, the tesla coil was used to locate the small breakdown pinholes.

Results All the breakdowns occurred in the section of the cables submerged in the water bath and the breakdown values are presented in Table 10.

Table 10. Measured DC breakdown values of good cable Breakdown Value Remark Specimen No.

(kV) 1 34 Breakdown at a longitudinal cut that appeared to be made during removal of the braid.

2 76 3 62

- 4- 55 5A 80 Breakdown after 2 minutes at this voltage level.

e 26 i m m . ., . . .. ,. . c . ... . i u s s . . . . . ,. c ... .. .

t Appendix I

- Detailed Requirements The cable samples provided shall be inspected and tested as follows (for comparison, copies of the cable manufacturer's routine test report and results from TVA's DC high-voltage testing of this cable are attached):

1. Prior to testing, the vendor shall subnit a detailed test procedure to TVA for approval.

2. Visually inspect jacket and record any defectr. or damage.

3. Parform dielectric test to determine fault location.

4. Remove the jacket and measure and record the jacket thickness.

5. Visually inspect insulation and record any defects or damage.

6. Repeat dielectric test to determine fault location.

7. Measure and record insulation thickness at the fault location.

8. Inspect insulation for contaminants as follows and record results:

A. Perform infrared spectroscopy on a section of good previously installed cable, a section of new (never installed) cable, and a section of cable from the feult location.

B. Perform x-ray spectroscopy on a section of good previously installed cable, a section of new cable, and a section of cable from the fault location.

C. Perform microsample tensile and elongation tests on a section of good previously installed cable, a section of new cable, and a section of cable from the fault location.

9. Measure and record insulatio, cesistance for a section of good cable.

Repeat this measurement on a sample of new cable and record the results.

10. Perform DC high voltage test on a section of good cable at 10.8kV for five minutes and record results. Repeat this test on a sample of new cable and record results.

11. Perform DC breakdown tests on e section of good cable and record results.

Repeat this test on a sample of new cable and record results.

! 12. The vendor shall submit to the Technical Engineer a test report upon completion of the above tests. These tests will be witnessed by a TVA representative, and the impact o# these results will be evaluated by TVA.

l 27 i x.,< . . - ~ . s . c .~.. I M s -.~ . c.,,u

r l

Appendix II

• Visual Inspection of the Braided Cable

- Cable #1: - Total length 84"

- Asbestos braid stripped back 4" with red adhesive close to the terminal end

- Label #1 51.8" from terminal end Cable #2: - Total length 125"

- Asbestos braid stripped back 4" with red adhesive 1" from terminal end

- White / gray RTV seal 15" and 45" from terminal end Cable #3: - Total length 131" .

- Asbestos braid stripped back 3 1/2" with red adhesive 1" from terminal end

- White / gray RTV seal 15 1/2" and 21" from terminal end

- Label #1 and #2, 97" and 32" from the terminal end, respectively

- This cable had more of a grayish color than the other Cable 44: - Total length 133"

- Asbestos braid stripped back 4" with red adhesive 1" from terminal end

- White / gray RTV seal 18" and 24" from terminal end

~

- The first 45" f rom the terminal end was not as heavily abraded as the rest of the cable Cable 85: - Total length 134"

- Control sample

- This cable was darker and had less or no sign of abrasion and very few asbestos fiber protrusions e

28 i o ..< , .~ ,. .~ ,. . c,... . I M s u.. ,-, , c .. .,

r O

Appendix III Visual Inspection of Insulation Surfaces Table III.1 Summary of observations Specimen No. Distance from Remark terminal end 1 *15.5" 5 m long longitudinal cut, located by the tesla coil

• 18.5" 10 mm gauge

• 19.5" 2 mm gauge

• 20.7" l mm gauge 54.3" The electric fault large pinhole approximately 1 mm in diameter.

• 62.5" Figure 2, 2 m long cut 2 27.0" The electric fault large pinhole approximately 2 mm in diameter.

Figure 3 3 109.0" The electric fault with 1 mm pinhole

• - diamter. Large chunk missing from the insulation. 5x2 mm in width.

Figure 4

• 120.0" 10 mm long longitudinal cut, no

• 122.0", 124" visible pinholes, located by the Tesla coil

• 126.0". 128.0" Radial cuts observed made when removing the jacket 4 103.0" - 106.0" 10 discrete failure sites identified by the Tesla coil. Several radial surface marks in this region.

Figure 5 5

• 3.3" 6 m long icogitudinal slice

• 55.5", 56.5", 87" Longitudinal cuts, 4, 5 and 3 mm long, respectively

• 127.7" Radial cut 4 mm long. Prior to electrical testing, this specimen was replaced by a new reference specimen which, as stated in Section

9. had all the braid removed by pulling.

The

• means that this abnormality was most likely introduced during the removal of the jacket.

29 l .-<..,.....,.<,....msm.--<..,~

y - - -

t Appendix IV Micro Infrared Spectra

~^

MIC F OMETE R$( M) 6 7 m 9 to 11 12 11 to 16 1R 20 2%

30 35 40 5  :- = - . . , . . -

2.5 -

~ BASE LttiE r

i.

f I t

O i 5 -

]

5 - NON. FAULTED AREA

/

y {' X 1 .

l 5[ j er i rt 4

i FAULT

[ CHANNEL i

[ i

,e

(

,. ^ -m .

._m. ._.... . . ,,

i:r, incq ano son enn l .t 24no 2nno innn inna 1460 4000 3600 3100 2900 rarn icurv ices I f Figure IV 1. u!R - Spectra of Specie.en No.1 30 e e . o . -. . . . . ... . c . .. . i G s . . .. . c , .,o , ,

t MIC A 0 V E T E RS t. M p 6 7 8 -9 to 11 12 13 to 1A 1A 29 2 's 2$ 30 35 4 f2 5

. . . . . . . . . , . . . . . . . . . - . . - . . . . . . . . . . ......_.....q

'I z -

i 9

e r

• 2 e  :

j NON. FAULTED AREA l ,

a s- .

p _

i i.,

u- '

FAULT CHANNEL 0 '.

L I

1 i , 9

• =

L

.1 l L ~

l .

(

' .w..._,.a . ... ..

1ASA tino 1'"' ten' A So '^n 46' 4000 3600 3200 2A06 2464 2006 1866 e n r,9.'s .n e r .

Figure IV 2. Specin.n No. 2 t

l 31 I

l i . o. . . ... .. . . . . . . . . . . i u s . . . .. . . . . . ,,

l

--,.n+ . . - - ~ - , , , , - . . . , , ,,n... ,, , ~ - - _ , , - - ._._,. - - - _ - _ _

f. t MIC R O M ET E R S 6.M) 4o 5 6  ; 8 9 10 11 12 13 14 16 16 20 25 25 . 3o 35 . ,_ , ,,,, ,,,,,,,, - _ , _ , . _

E t

a b

L.

iE ,- j OI .

GL o

E i e

zI 1 l e _

,J l -

w z -

u, s

WI

&L l

j _ NON FAULTED AREA t

, 1

, FAULT

~

l CHANNEL -

t

( -

i

,m _ A ann L - . . . . . . m . u a ' -

,;n, inno pne AM 24no 20nn im g 1400 4000 3600 3200 26SC r a r onr N e v ie's i l

l Figure IV 3. u!R - Spectra of Specimen No. 3 32 i o ... . . ~ , . . ... . c , ... I v 5 r . ~ - c-n 4..

m,e~ ~ v e --w.,-en--e,--w-w,w ,,,m,,,wne--,,,,,-,v,, war-e-www--r-- sm,-,, - - - - e -ws-m-<---,w---m-p=--Y_--~,,

t.

a A

4 MIC A0M ETE RS (.Mi 40 5 6 7 8 9 to 11 12 13 14 1A 1A 20 25 30 3.5 L

L l--

r -

9  :

. UI 2I et z

( "

E

- NON. FAULTED AREA zi 0L e-w

&L I

L I N L

i

~ ,

l FAULT CHANNEL

^

1P00 18Ao 1400 1,'0 '

tona Pon 500 4000 3600 3200 2400 2400 2000 rnroprNr'vmv i

- Figure IV 4. u !R - Spectra of Specimen No. 4 33 ie - o ............ .c ,-,.I M S s,...... co..,0..

. _ _m. . . _ . . _ .. .. _ m.

't

4

-e i

P 1

t MICRoVETE AS( M6 40 6 .7 8 0 to 15 12 1' 14 16 18 20 25 2.5 30 35 5 77.,....,_.......

J f

i i I .J zL 9- CLOSE TO THE '

BRAID UI-si e, i 5L

&. /

Elr / i

'l 7 I

i gj.r .

' h.! .

b CLOSETOTHE

/ CONDUCTOR l _

l

~

I i

i .

l _

A i'

';'S i nr.3 anc son g n-2%1 24M 2non iaM tain 1460 l *

.4000 3600 3200 l rn ,s . . . . . . . . . .

Figure IV 5. ulR - Spectra of Reference (Specimen flo. 5)

I 34 i w . .. . .s. . . . . t , .... i M s s . . ... ., % ,,,4.,

l 1

t Appendix V.

• Results from X-ray Spectroscopy Cable # Location Magnification Counts pr. sec., Element Al Si Cu Sn 1 Fault-Channel 386 64.5 1256 trace trace Non-faulted 240 3.6 **

2 Fault-Channel 220 33.9 707

• 1.8 **

Non-faulted 300 9.2 **

3 Fault-Channel 324 56.5 1047 794.7 1.9 **

Non-faulted 334 29.2 342.6 11.6 **

4 Fault-Channel 384 97.1 723.9 2.3 **

Non-faulted 386 27.8 6.3 **

5 Non-faulted 350 91.3 21307

• No elemental counts were taken. Only visual comparison of the X-ray spectra was performed. This was done by overlaying the spectra of the non-faulted area onto the spectra of the fault channel. Within limits there were no discernable differences.

• • Below background noise Interpretation

- Silicone (Si) - From the silicone polymer Aluminum (A1) - From the filler Copper (Cu) and Tin (Sn) - The higher concentration of these elements in the electric fault channel is probably due to evaporation of metal from the conductor.

e 35 i .......:.. i,,.n.c.~..I M s r....., co ....

t Appendix VI

' Results from Tensile and Elocation Test Width Comments Tensile  % Elong.

Specimen Sample Thickness Strength (psi)

No.

, ~

.059" Good Break 1589 500

^

1 1 008" Broke at grips 1667 520 2 .008" .060"

.057" 8;oke at grips 1754 600

'3 .009" 440 4 .010" . 061" Broke at grips 1557 Broke at grips 1792 580 5 009" .062"

. 063" Broke at grips 1587 500 6 .009" 520 7 .009" .061" Broke at grips 1821 Good Break 2083 520 8 .008" .061" _ _

~

2 1 .008" .060" Run stopped- 20S3 Equip. Prob.

Broke at grip Broke at grip 1815 480 2 .008" .062" Good Break 1821 460 3 .009" .061" Good Break 1821 450 4 .007" .061" 2292 500 5 .008" .060" Broke at grip 2049 460 6 .008" .061" Broke at grip

.062" Broke at grip 1815 370

- 7 .008" -

.060" Good Break 1875 500 3 1 .008"

.061" Good Break 2049 500 2 .008" 3 .009" .061" Chart not in Drive 1639

.059" Broke at grips 1883 500 4 .009"

.061" Good Break 1639 375 5 .009"

.061" Broke at grips 2004 500 6 .009"

.059" Good Break 2119 520 7 .008"

.061" Good Break 2254 500 8 .008" 1333 29'O 4 1 .010" .050" Broke in grips

.062" Good Break 1792 430 2 .009"

.060" Broke at grips 2083 450 3 .008"

.060" Broke in grips 1639 370 4 .008" 440 5 .008" .060" Broke at grips 2083 Good Break 1821 430 6 .009" .061" Good Break 2016 490 7 .008" .062" Broke at grips 1792 470 8 .009" .062" 1293 330 5 1 .00h .058" Good Break 520 2 008" .059" Good Break 2119 Good Break 1916 540 3 .009" .058" Good Break 1457 460 i

4 .009" .061" l

Good Break 1724 520 1 5 .009" .058" 1950 520 6 .009" .057" Broke at grip l

• l 36 t

t **. < . i% w a, ..c . c,,,, . I M $ r , , m, , c ,. %, ,,

f .

L

t APPENDIX - VII CALIBRATION OF VOLTMETER

. THE INSTRUMENT HAS BEEN VERIFIED ACCURATE WITHIN THE TO'_ERAf!CES SPECIFIED: .

INSTRUMENT: FLUKE 8024b MULTIMETER S/N: 3785052 STANDARD: HEWLETT-PACKARD 740B D.C. VOLTAGE STANDARD

.% OF FULL SCALE

.' RANGE 200mVDC .003%

2VOC .015%

20VDC .005%

200VDC .02%

1000VDC .1% ,

VERIFIED BY:

ELECTRONICS LAB, Institute of Materials Science 4

l i

l t

8 O

37 l

i c,,< . - .. . s . c ,.... I M s u. ..,.. c-m, l

r e

INSTITUTE OF MATERIALS SCIENCE The Institute of Materials Science (IMS) was established at The University of Connecticut in 1966 in x order to promote academic .research programs in 8 materials science. To provide requisite research laboratories and equipment, the State of Connecticut has provided 56,000,000, which has been augmented by l over 57,500,000 in federal grants. To operate the in-stitute, the State Legislature appropriates over

$1,200,000 annually for faculty and staff salaries, sup-plies and commodities, and supporting f acilities such as an electronics shop, instrument shop, a reading room, etc. This core funding has enabled IMS to attract over

=

~

55,500,000 annually in direct grants from federal agen-cies and industrial sponsors.

IMS fosters interdisciplinary research programs in various areas of materials science with special emphasis on adhesion, composites, corrosion, electrical insula-l tion, interfaces, liquid crystals, metals, and polymers.

l These programs are directed toward training graduate students while advancing the frontiers of knowledge and l meeting current and long range needs of our state and our nation.

l I

l l

1 l

l l

l l

l o r

l

! e l

l

t 9

ATTACHMENT 2 MEMORANDUM FROM D. W. HILSON TO W. S. RAUGHLEY SEQUOYAH NUCLEAR PLANT - IMPACT TESTING OF AIH CABLE July 23, 1987 P

e e

B46 '87 072 2 001 UNITED STATES GOVERNMENT Memorandum TENNESSEE VALLEY AUTHORITY

{ To W. S. Raughley, Chief Electrical Engineer, W8 C126 C-K ynny  : D. W. Wilson, Chief Nuclear Engineer, W10 C126 C-K oarE .

JUL 231987 senJECT. SEQUOYAH NUCLEAR PLANT - IMPACT TESTDiG OF AIW CABLE We are trans:iitting tne Singleton Materials Engineering Laboratory report SME-MET-87-043 as requested by Brian Reagan on July 6,1987.

i 0WS 9 D. W. Wilson GLR:HHD Attachments cc (Attach =ents): '

RIMS, SL26 C-K W. H. Childres, SME-K This was prepared principally by Gary L. Riner, extension 2771.

i

( A27135 1 l

I l

l l

1' l

I m

"nv I

• S. %:ines B.>ntis R, unlarly on Ihe Pa3 roll Savin::1 Plan

-