ML17258A134

Analysis & Rept on Safety-Related Electric Penetration for Ginna Plant WX32714

ML17258A134
Person / Time
Site:	Ginna
Issue date:	08/07/1981
From:	Korner R WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP.
To:
Shared Package
ML17258A133	List: ML17258A132 ML17258A134 ML17258A135
References
546-SG-435907-S, PEN-TR-81-45, NUDOCS 8109100183
Download: ML17258A134 (133)

Text

Nestinghouse Electric Corporation Electronic Components Divisions Westinghouse Circle Horseheads New YorK 14B45 PEN"TR-81"45 August 7, 1981 ANALYS[S 'AND 'REPORT ON 'HE SAFETY RELATED ELECTRIC PENETRATION FOR THE GINNA PLANT WX32714 Ref.

W IGTD P.O.

f546-SG"435907"SN R.

L. Korner Project Engineer

<~<klIJNgag g Empt 8109i00183 Si0904 PDR ADOCK 05000244 P

PDR ~

I

INDEX 1,0 Identification of Equipment and Materials 2.0 Purpose of this 'Report 3.0 Qualification Test Plan 4.0 Required Environmental Conditions 5.0 Qualification Data Supplement 6,0 Identification of Materials (Pressure Retaining)

7. 0 Ident ifi ca t i on of Mater ia 1 s (Modu 1 e As semb 1 y) 8,0 Statement on Qualified Life 9,0 Cable Qualification 10, 0 Report Summa ry

\\

1.0 IDENTIFICATION OF EQUIPHENT AND HATERIALS Penet.

Nozzle Number Number of Cables T

e W

WX Number AE"12 (27)

(12)

(4)

¹16 TSP

¹16 TSQ

¹10 AWG COAX WX32714 This penetration consists of a 5" module inserted in a flange attached to the mating nozzle flange.

This report will be directed at the module and its cables.

This penetration is dedicated to instrumentation service;accordingly the effects of ohmic heating are minimal.

Short circuit condition cannot occur in any of the circuits.

The operating temperature will follow the ambient temperature of the containment.

2.0 PURPOSE OF THIS REPORT To provide additional information to supplement W Report AB"11/12/73 "Qualification Tests For A Hodular Penetration 5" Dia.

(Prototype B-1)".

3.0 QUALIFICATION TEST PLAN Additional test reports and analysis will be provided for both the penetration module and its cables.

1

4.0 REQU I RED ENV I RONMENTAL COND IT IONS "2-

2:05 Desi Re uirements 2:05.1 Design Parameters:

The design parameters for the penetration are as follows:

1.

Service life 40 years 2.

Design basis temperature (max)

3. 'esign basis pressure (max) 4.

Design basis relative humidity 268 F 60 psig 100X 5.

Ambient temperature (max) 120 F

6.

Ambient pressure (max) 7.

Integrated 40 year dose (max) 8.

Dose Rate (max) 0.3 psig 1.5 x 10 Rads 2.1 x 10 Rad/hr

~0 I~

g0 Ml

~

I Sl I

. II'~)

EJ

5.0 QUALIFICATION DATA TO SUPPLEMENT REPORT gAB-Il/12/73 At the time of manufacture, the aging data for the proprietary Q-I epoxy was not available.

Data is now "provided in the form of the Arrhenius curve for the Q-1 epoxy used in the manufacture of the Ginna Penetration.

See Research Report 75-7B5"BIGAL"R2 which follows.

Several explanatory remarks will be made to clarifp this report.

1.

The failure point was defined as 10 std. cc/sec; which is the same as the allowable leak rate for the penetration as installed as defined by IEEE 317"76.

2.

Although only three or four samples were used for each temperature point, each sample contained 39 seals of various diameters

(.041" to.437").

Therefore, the crite'ria that a large number of samples be used was complied with, 3.

Electrical loading was not included in the test because the investigation was intended to establish the effects of heat caused by I

R on the seal integrity.

As reported in the Appendix g2 to 75"7B5-BIGAL-R2 dielectric strength and insulation resistance were not affected aFter seal failures occurred.

To be sure no synergistic effects occur due to the combination of high temperature and applied voltage, a module was successfully run for 84 days 9 125'C at 480 volts without seal or electrical degradation.

(See enclosed report PEN-TR-79-73).

Research Report 75-7B5-BIGAL-R2 May 27, 1975 PREDICTING THE THERMAL LIFE OF MODULAR PENETRATIONS J.

F. Quirk Insulation Chemis try ABSTRACT Accelerated thermal endurance tests',

in compliance with IEEE standards, have been conducted using modular penetration specimens of a universal design.

These data have been treated by the Arrhenius technique to predict thermal endurance or thermal index (leak rate) of the modular penetration specimens over the temperature range of 70'o 200'C.

INTRODUCTION Procedures for estimating the thermal life of insulation materials at room temperature require life tests at several temperatures, above the. expected normal operating temperature.

By the selection (1,2,3) of relatively high temperatures for the tests, life of the insulation samples is terminated, according to some selected failure criterion, within relatively short times (i.e.,

one week to one year).

The result of these thermally accelerated life tests is a set of data of life times for a corresponding set of temperatures.

This set of life times can be used.to establish the mean life values at each temperature and'he functional dependence of life on temperature, as well as the statistical consistency, the confidence to be attributed to the mean life values and the functional life temperature dependence.*

• IEEE Guide 101.

Accelerated thermal endurance tests,. in compliance with IEEE standards 98 and 101, have been conducted using modular penetration specimens of a universaJ design.

'l'h<<. results ol these endurance Lusts are the subject of, this report.

(UNCT.lJS TONS Using the Arrhenius relationship shown in Fig. i, a lower 95% confidence limit or mean thermal life of the modular penetration seal is 350,000 h (40 yr) at 105'C.

RECOMMENDATION We recommend that modular penetrations of the design described

'herein be used in nuclear power plants at an operating temperature range of 70'C to 105'C.

They can be operated at higher temperatures but the life ctime is shortened significantly as shown in Fig. l.

EXPERIMENTAL A.

The Modular Penetration S ecimen The test specimens, provided by the Westinghouse Industrial and Government Tube Division, were of a universal design and were produced in a pilot production facility.

The conductor sizes and spacing are shown in Fig. 2.

All specimens were subjected to five cycles from -30'o 100'C prior to the start of elevated temperature aging.

B.

Selection of Test Tem erature and Failure Criterion IEEE 98 was used as a guide in selecting the exposure temperatures and times, both of which are given in Table I.

According to the guide, these temperatures and times per cycle are suitable for a material with a temperature

index in the range of 130'o 154'C.

The epoxy resin used to manufacture the test specimens has a temperature index of 135'C.

After each high temperature

exposure, specimens were subjected to helium leak testing to determine the degree of deterioration.

High temperature cycles were continued until the specimen displayed a leak rate

-2 of 1 x 10 std cc/sec.

Leak rates were determined by a mass spe'ctrometer

-11 leak detector with a sensitivity of 2 x 10 std cc/sec.

A specimen and the detector are shown in Fig.

3.

C.

Ovens The ovens used to provide the temperatures required were of the forced-air circulating type.

These ovens were thermally mapped to locate zones within which the required temperature was being maintained.

The test specimens were then located in that zone.

D.

Failure Data Accelerated thermal endurance aging was continued until the test

-2 specimens displayed a leak rate of" >

1 x 10 std cc/sec.

The test tempera-ture and time required to produce the failure criterion are given in Table II.

These data were analyzed according to the procedure specified in IEEE standard 101 using a computerized standard linear regression analysis program.

A summary of the results of that analysis

follows, E.

Life Test Values The following tabulation gives the mean and standard deviation of the life data.

Test Tem

. 'C 200 187.5 175

,150 Mean life h 129 343 1142 7709 Std.

Dev. - h 48 162 61 481

F.

The Fitted Arrhenius Model The Arrhenius model which best fits this set of data is LN h = A +

B (1/273 + temp.

'C)

I where LN h = natural log of life in hours at temp. 'C.

Using linear regression, we estimate A = -30.13 and B = 16562.8.

The Arrhenius model used to predict the mean life of the test soecimens at the test t'emperatures gave the following predictions:

~Tem

. 'C 200 187. 5 175 150 Mean Life Predicted <<h 132 342 935 8317 Mean life predictions are shown in Fig.

1 as the solid line drawn between the coordinates 200'C, 130 h; and 150'C, 8,300 h.

This solid line was extrapolated to temperatures below 150'C (broken line Fig. 1) to form an estimate of the mean life predictions at temperatures less than those selected for test.

The correlation between 'the observed and predicted life times, r, is 0.96 indicating the model fits the data 2

quite well.

G.

Confidence Limits The lower and upper 95% confidence limits, determined using the predicted mean values of life represented by the solid and broken lines in Fig. 1, are given below:

~Tem

'.C Mean Life Predicted h

Life h Lower 95%

~Ver 95%

200 187.5 175 150 105 133 343 936 8, 318 879, 866 99 277 782 6, 097 390, 314 176 424 1,119 11,347 1,983,439

These upper and lower 95% confidence limits are represented, graphically, in Fig. 1 as the envelope about the generated Arrhenius model.

The computerized program predicts, with 95% confidence, that zm

-2 the specimens will maintain a leak rate of

< 1 x 10 std cc/sec for 390>000 h

40 yry at 105 C ~

Since the expected range of 70-105'C, it can be the specimen will maintain a

in-service operating temperature is in the predicted, with at least 95% confidence, that

-2

<1 x 10 std cc/sec leak rate for 350,000 h

~ or 40 yr in that temperature range.

H.

Thermal Endurance of "0" Rin Seals The seal between the modular penetration and the containment wall is provided by a series of silicone "0" rings.

Mock-ups of these external seals were tested in the same manner as the penetration specimens.

The thermal endurance of the external silicone "0" seals compared to that of the modular test specimens are tabulated below:

~Tem

. 'C Penetration Specimen Mean Life h "0" Ring Mock-up Total h Ex osed "0" Ring Mock-up Leak Rate 8 Total h 200 175 150 129 1142 7709 535 1758 9000

<1 x

<1 x

<1 x

-9 10 std cc/sec

-9 10 std cc/sec

-9 10 std cc/sec It appears from these results that the thermal endurance of the external silicone "0" ring seals exceeds the estimated thermal endurance of the modular penetration test specimens; however,, this is not based on a statistical test.

Such a relationship eliminates the external seals from consideration in estimating useful thermal life (leak rate) of the design.

I.

Estimation of E uivalent Thermal Life The Arrhenius estimate of seal life, Fig. 1, can be used to approximate the equivalent thermal degradation which could be expected during 40 yr of service at a normal plant ambient of 70'C.

The method used to'ake this approximation is as follows.

The relationship between seal life and temperature is such.that an increase or decrease of <8'C in the test temperature results in a cor-responding doubling or halving of the specimen life.

Using this 8'C rule, an approximation of the equivalent thermal degradation expected during 40 yr of service at 70'C can be estimated.

A point representing 40 yr (360,000 h) at 70'C is located on Fig. l.

A line is drawn such that it passes through that point and is parallel to the line representing the Arrhenius estimate.

The relation-ship between seal life and temperature represented by this second line is such that for every increase or decrease of ~8'C in temperature results in a corresponding doubling or halving of the equivalent thermal degradation which could be expected.

Estimates of the equivalent thermal, degradation which could be expected in 40 yr at 70'C are ~400 h at 125'C or <40 h at 150 C.

Therefore, exposure of a specimen to >40 h at 150'C would produce the equivalent thermal degradation that could be expected at 260,000 h

at 70'C.

ACKNOWLEDGEMENT The author wishes to acknowledge the assistance rendered by T.

W. Dakin through the use of his computerized Arrhenius program and

. his guidance in interpreting the results of that analysis.

REFERENCES 1.

Berberich, L. J.,

T.

W. Dakin, "Part I:

Guiding Principles in the Thermal Evaluation of Electrical Insulation," Insulation Magazine,

p. 21-27, Feb 1956.

2.

Moses, G. L.,

et. al.,

IEEE Guide for the Pre aration of Test Procedures for the Thermal Evaluation of Insulation S stems for Electric E ui ment, No. 99.

3.

Berberich, L. J.,

T.

W. Dakin, "Part II:

Guiding Principles in the Thermal Evaluation of Electrical Insulation," Insulation Magazine,

p. 21-26, March 1956.

PERMANENT RECORD BOOK ENTRIES Fig.

Book No. 205631...pp.

65-67 and 71.

J.

F Quirk Insu ation Chemistry APPROVED:

lP+I'.

H. Runk, Manager Insulation Chemis try

TABLE I TEST TEMPERATURES AND EXPOSURE TIMES Test Temp.

'C Continuous, Exposure h

Followed by h/cycle 200

+ 4 187.5 + 2

'175' 2

150

+ 2 72 72 166 528 24 24 168 504 TABLE II TEST TEMPERATURE AND LIFE VALUES

':Test Temp.

'C No. of Specimens on test Life Values h

200 187. 5 175 150 3

4 4

96, 108, 184

163, 390, 477

1051, 1173,

1173, 1173

7334, 7334,

7834, 8340

2. 17 I

-3

x 10 T, 'K 1, 000, 000 500,000 2.1

2. 2 2.36 2-5 2.9 40 Years 100, 000 Arrhenius Estimate of Seal Life 95% Confiden ce Limits 50, 000 10, 000 5000 50 10 200 175 150 125 105 Temperature,

'C Fig. 1-Estimate of penetration life 70

> IIFSTINGHOUSE ELECTRIC CORPORATION ELECTRONIC TUGC DIYISICN SPECIFICATIONS 8a STANDARDS DEPT.

ELMIRA, iN. Y.

//dD 2 Vr", /."'?/// "

'P.8 lid'2'/7=~.~"///G '

/5'Q'USPECT'Pj 7 '-: S'/ ~-'rPA.,r,"

~ '8//. A/7 "

DATE Sl.'ZCIPIC/i.I."'~l t<O.

~ - g '79+

5"i'Lil E I"'D DATI."

> WOWS T P //

Co p7./'/.)

/8 GYP) e-',

e

/

'Q

/oo 24>. "

j,g4 Z fry.:Sir:=.

r I

~ 'pa> I

~r n

r COis/ g~

+.DO SS'.0

- /88

~ '-: 7 B.97 mlH.

I.[

r8 ri.cArNrcs uNa.rss OTNriiaisr. Si ccI ~Iso rOTC ~

Also see STO.

ET IO thru 36 OCC.

~

~

~ ~,000 0

.00 ~

,00 I

01

~'NAC.........

0 lrss CONCCNTNICITY 000 T.l ~II.

At,u OIH IN INCH% ~

NrNOVC Auu lluNAS i O'CNOTCS CHANOC Aul TN'OS. CI TA Oil ) ~

r,A o.

~ss Fig. 2.

Conductor siaes and spacings

'ig. 3.

Test specimen and leak detector NP 33 RM-63923

Westinghouse Electronic Components Electric Corporation Divisions Electronic Tube Oivisions Westinghouse Circle Horseheads New York 14845 May 12, 1975 Test Report tntPEN TR 75-10 Appendix ¹'2 Report 075-7B5>>BIGAL-R

SUBJECT:

Dielectric Strength and Insulation Resistance Test Results of ~-,

Modules which Were Subjected to Accelerated Heat Ageing at 150 C.

PURPOSE OF TEST, To demonstrate that the dielectric strength and insulation resistance remains at satistactory levels after accelerated life heat ageing.

TEST PROCEDURE Four modules were run at 150 C to the point where the leak rate reached 0

lx10 2 std cc/sec.

Dielectric Strength and 'Insulation Resistance tests were then performed; conductor to all other conductors and to ground.

TEST RESULTS Failure point Module ¹ 7

8 9

10 (based on Zxlo std. cc/sec.

leak rate)

Hours 7332 7332 7836 8336 After the above tests all modules. withstood 2,650 volts A.C. for five seconds.

Insulation resistance t'eadings with megohmmeter at 500VAC were lx10 ohms or higher.

Hay 12, 1975 page 2

CONCLUS I ON The results indicate that the dielectric strength and insulation resistance values remain at adequate levels after accelerated heat ageing.

Apparently the Arrhenius curve for the sealant material based on the electrical qualities is at a higher temperature level than the one based on leakage.

R. L. Korner Project Engineer Penetrations Dept.

/m

Westinghouse Electric Corporation Electronic Components Divisions Wesringhouse Circle Horseheads New York 14845 PEN-TR-81-46 August 7, 1981 EXPLANATiON OF ACCELERATED AGiNG R.

L. Korner Project Engineer

Attached Figure 2 shows how values of 150'C for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> were obtained as representative of a life of 40 years at a 70'C normal temperature, A parallel line was drawn to the 954 lower confidence limit starting at the intersection of 40 years and 70'C.

150'C was chosen as a reasonable temperature to produce accelerated aging.

The line starting from 70'-40 years is intersected by 150'C at 85 hours9.837963e-4 days <br />0.0236 hours <br />1.405423e-4 weeks <br />3.23425e-5 months <br />, These two values 150'C-85 hours are the equivalent aging of 70'C~40 years, One hundred hours was chosen to add some margin to the required time,

Accelerated aging at 150'C 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> aged the sample to 49 years at 70'C.

At the time of the test for the Ginna penetration reported in AB 11/12/73, the Arrhenius curve for the Q-I epoxy was not firmly established, there-

fore, a more conservative 504 hours0.00583 days <br />0.14 hours <br />8.333333e-4 weeks <br />1.91772e-4 months <br /> at 150'C was used to pre-age prior to LOCA.

, 1,000,000 500,000 I

-3

x 10 T, 'K

'.17 2.1

2. 2 2.36 2.5 2.9 100,000 50, 000 Arrhenius Estimate of Seal Life 40 Years 95% Confidence Limits M, 000 II 5000 C) 200 175 150 125 105 TempeFature,

'C 70 50

'Fig. 2 -Estimate of penetration life (Fi'gure l is part of report entitled "Predicting the Ther~i Life of Nodular Penetrations,"Research Report 75-785-BI GAL-R2).

0

Westinghouse Electric Corporation Electronic Components Oivisions V(estinghouse CIrcle Horseheads IIew YorK 14845 PEN"TR-79-73 Sept..18, 1979 ELECTRICAL PERFORHANCE OF AN

'ELECTRICAL PENETRATION t<ODULE

'UNDER ACCELERATED HEAT AGING CONDITIONS Approved by:

Test performed by:

R. L. Korner Project Engineer Vito Liotino NUclear Electrical Penetration Department

Pur ose oF Test To examine the insulation resistance characteristics of a modular electrical penetration filled with W Proprietary Epoxy "Q" and 3M product XR5237 in a 125'C accelerated aging ambient with operation voltages appl'ied.

The normal operating temperature of this device is 70'C.

, Test Specimen Descri tion The'test specimen is an electrical penetration module fitted with a mixture of cables and potted with a W proprietary epoxy designated as "Q" to form the primary seal of the module and a two part filled semi-flexible, flame retardant liquid epoxy XR5237 manufactured by 3M to form the secondary and finish seal.

A cross-section of the test module is shown in Figure 2.

A mixture of cables has been included to qualify the module for various service requirements.

The physical configuration of the cable electrodes in the module is shown on Drawing gE-2794.

The B-2 type (Hanford) module was assigned PSN66.

Histor of Hodule PSN66 The module was preconditioned in a previous qualification test following the requirements of the IEEE standard f317-1972 and as reported in report OPEN-TR-75-6.

The module was subjected to 90 hours0.00104 days <br />0.025 hours <br />1.488095e-4 weeks <br />3.4245e-5 months <br /> of cobalt 60 gamma radiation for a total integrated dose of 4.45 X 107 rads.

The assembly was also subjected to 340'F " 52 psig steam -

1000 humidity for six hours followed by eighteen hours at 280'F and 34 psig - 1004 humidity.

Insulation resistance readings of the subject cables, after manufacture taken at 500 Vdc,were in the 109

" to 10 ohm range.

After irradiation and LOCA, wi th rated voltage and current applied, IR measurements were stable and in some cases greatly improved.

Test Procedure and Discussion The module was placed in an air circulating oven and electrical power leads were attached to their respective cables.

A voltage of 120-127 Vac was impressed on the

¹14AWG cabl'es.

The center electrodes were spaced

.133" apart.

The remaining cables were hooked up to 480 Vac.

The center electrodes of the

¹10AWG cables were spaced

.176 apart.

The ¹6's were.303 apart and the ¹4's were spaced

.348 apart.

The module insulation resistance measurements at 500 Vdc were taken prior to the elevated temperature test.

The IR (ohms) values of ail subject cables fell within a 1.2 - 1.8 X 1011 range.

Periodic Insulation Resistance measurements were taken throughout the elevated temperature test.

See the Insulation Resistance (ohms) versus Time (day) graph in Figure 1 for.the IR plot.

Post test IR readings at room temperature were taken and all fell within a range of 1. 1 - 1.5 X 10 ohms.

A problem developed when the oven controller developed an amplifier tube failure resulting in a rise in oven temperatuie.

The temperature rose to 225 C.

It is estimated that the oven was at this temperature for a period of approximately one day.

The module temperature fell to room 'temperature during the oven controller repair period.

The module insulation resistance measurements at room temperature were within the range of 1. I X

10 1 5 X 1010.

The controller was repaired and the test continued.

Test Results Test Duration Insulation Resistance (ohm)

Days Start (RT) 1 ~ 2E11

~Tefe

'0

¹14AWG

¹10AWG 1.8E11 5 500 Vdc

¹6AWG 1,6E11 g4AWG Cab le 1 ~ 5E11 125 C

125 C

1.2E7 1-7E7 125 C'.6E6 125'C 1 ~ 1E7 (a)225.C (b) SE5 6.2E6 7.2E6 (b) 2.5E5 1.4E7 1.2E7 4.3E6 3.0E6 (b) 1.0E5 2.8E6 4.2E6

2. SE6 2.8E6 (b) S.OE4

2. 1E6 2.3E6 21 28 35 42 49 56 63 70 77 84 125'C 125 C

125 C

125'C 125 C

125'C 125'C 125 C

125 C

125 C

(RT) 2.0E7 1.0E7 1.0E7 1.6E7

1. 3E7 1;4E7 2 OE7 5

OE7

8. OE7 8.5E7 1.5E10

7. OE6

8. 6E6

8. SE6 7.0E6

1. 1E7 1.2E7 6.8E6 1-3E7 4.0E7

)

3-5E7 1

lE10

3. OE6

3. 4E6 3.5E6 3.5E6

4. OE6 3.7E6 2.5E6 4.0E6 5.0E6 6.0E6

'i. 4E10 2.7E6 2.3E6

2. 5E6 2.6E6 3.3E6 3.2E6 2.8E6 5.0E6 S.OE6 9 ~ OE6 1.0E10 (a)

(b)

Oven controller problem caused temperature to rise to 225 C

for a duration of approximately 1 day.

Insulation resistance reading taken at 100 Vdc.

(Reading not attainable at 500 Vdc).

Discussion of Results The first day of heating resulted in a drop of insulation resistance as expected due to the normal decrease of resistance of insulating materials with increase of temperature.

The resistance stabilized for 63 days and then took an unexplainable rise of about one half to one decade in the next 21 days.

After the accelerated temperature test was completed, the module on return to room temperature exhibited IR readings approximately one decade lower than the initial room temperature IR readings.

The accidental transient to 225'C (437'F) did not damage the penetration and the insulation resistance returned to normal when the temperature was set at 125'C again.

Conclusion l.

The upward trend in the last 21 days of the test indicate that I

the accelerated electrical life of the epoxy is well beyond the 84 days tested at 125'C.

20 A one day accidental transient to 225'C did not cause electrical fa i 1 ure.

4-

1011 Ql - XR5237 Epoxy-Potted Hodular Electrical Penetration (Hodule S/N ¹66) Voltage e

Elevated Temperature Test 1010 109 tnsulatton Resistance Readings Thru-out a 125'C Bake Cycle Voltage impressed on cables during the elevated temperature test are as follauts:

¹14AWG C0 4J lA 107

¹14 120Vac

¹10..-

480vac

¹ 6 -

48OVac

¹ 4 -

48ovac

~k rr r

x~

ll o

Q

¹6AWG l

~

~~pc O

l Cl

¹10AWG

.o~

~~

gy

¹4AWG 10 14 21 28 35 42 Time (days) 56 63 84

3H Potting - Typical, Both Sides W Potting IIQII Hodul ar E lectri ca l Penetra t ions See Drawing gE-2764 For Cab)e Layout Figure 2

~

~

~

~

~

r

~

J Q

0 0 0 0 C

0 r 0 0 o

I e

6.0 IDENTIFICATION OF MATERIALS USED IN THE PRESSURE RETAINING PORTION OF THE PENETRATION MODULE

,'IHGHQUSE Ef FCTRfC CORPORATfGH ELECTRI.NfC TUBE MVISlCT'l PECIFlCATlCNS 8, STANDARDS DEPT:

ELMlRA, N. Y.

/Q STir"8" r87 Y~D HSSEFi'.

DATE/-+j-g+

SPECiFlCATlON NO. 8& 8 8 88-/

SUPERSEDED DATE GHOUSE PROPRlETARY LA TE BEFORE OlgC<<olNG 4

W e

Z 9

3.c2 3', dk C7 j.gg. - ~ oc 8',oa I

4 I

~ v v

Iv X

h, IX vv Iv 0

I/7

~

)VLCC'1$

Q ~<<f'4rlI1 E SP CCIV'Ifg vCTP1 DC1 S COCCI.vt'LICIT'T 0N T I 8

~ i

1. OIV. IV IHCVCS

'I

'I )i(vi' I. *'

) '~h+

a g

~ )I'E0 CvAH C

I I

s

~

Ta

\\

)11 1

tD UN ONTHOLLEDCQP't Alsc see StI!.

ET 10 thI u 36 SCAI a

~

~'A

~

QD Vli r'/

I.~ ~ q g v

~ I I a c ~

a E

(.",r-,'-

IDENTIFICATION OF'MATERIALS DWG. 35-8388 I tern No.

Hater i a 1 Stainless Steel " Header Epoxy Glass Laminate 54

¹14AWG Copper

¹10AWG Copper 74

¹18AWG Copper 8-Epoxy - Q-I -

W PROPRIETARy Silicone Varnish ¹991 Only Items marked with an asterisk

(-) have pressure retaining function.

7.0 IDENTIFICATION OF MATERIALS USED IN THE PENETRATION MODULE ASSEMBLY

~

~

a

~

r 0

~

~

~ '

~

~

~

~

~ r

~

~

~

~ ~

~

~

~

0

~ ~

~

~

I

~

~ ~

~

j 0

~

I

~

~ ~

~

~

i o

All materials except the silicone "0" rings were qualified as assembled in the test specimen of W Report gAB - ll/12/73.

The neoprene "0" rings were not qualified as they are included as a third and fourth sealant to the two "0" ring seals which form the primary and secondary seal.

'he silicone "0" rings were separately qualified.

Separate.qualification data is provided for the Class I-E cable.

8.0 STATEMENT ON QUALIFIED LIFE Based on the attached analysis by T.

W. Dakin, the qualified

~ life of the penetration is in excess of 50 years at 70'C.

This particular penetration however operates at the containment ambient of 120'F, (49'C) and therefore can be expected to 'have an even longer 1 ife.

from NN Oare Subiecr RESEARCH LABORATORIES 236-3424, 5134 July 18, 1975 LIFE PREDICTION OF CAST EPOXY PENETRATION SEALS fo Research Laboratories J. F. Quirk cc:

Research Laboratories R. H. Runk Research Laboratories J.

Swiss The results of your life tests on the cast epoxy penetration seals have been graphed and a log vs 1/T (Arrhenius) linear regression line, and its 95% confidence limits, calculated according to IEEE Standard 101.

This calculation was accomplished with a computer using a program for this specific purpose.

The actual life tests were made at four temperatures from 150'o 200'C.

At the lowest life test temperature, failure times averaged a little more than 8000 hrs, approximately one year, which is about as long as is usually ever used in an accelerated life test of this sort.

The test values at all of the four test temperatures fit the calculated linear regression line, showing no tendency to deviate from a good fit to the theoretical linear dependence of log life vs reciprocal Kelvin tempera-ture.

This gives very good assurance, that the linear extrapolation of this line is probably valid.

The regression line extrapolates to predict an average life of 70 million hrs or about 8000 yrs at the expected use temperature, 70'C.

The lower 95% confidence limit on the regression line is about 684 yrs at the expected use temperature, 70'C.

This is more than 10X the required life of 40 yrs at 70'C.

The data and this extrapolation from it also predict that the penetration seals would have an average life of 44 yrs at 105'C with greater than 95% confidence..

This is 35'bove the expected operating temperature life.

The above extrapolation applies only to the average life.

Another factor to be considered is the statistical deviation of individual values from the average.

The spread or range of test values here is rather moderate at each temperature.

The calculated average o, standard deviation, in terms of percent of the life values at each temperature is 16.7%.

If we subtract 3 a, or 3 x 16.7%

~ 50.1X from the lower 95% confidence on the extrapolated average life of 684 yrs, this predicts (assuming a

normal distribution) that 99.7X of the specimens should have a life of greater than 342 yrs at 70'C.

This is obviously a very high estimated life expectancy.

Both the possible statistical variability of the extrapolated average life and the statistical variation of individual

J. F. Quirk July 18, 1975 Page 2

specimens from this average have been considered in making this estimate, which is based on a rational extrapolation from well behaved thermally accelerated life test data.

While the extrapolation of bu'own to 70 C

from the lowest test temperature of 150'C is large, the extremely high extrapolated value of life at 70'nd the high degree of statistical confidence which can be placed on it, together with the long actual life of about 1 yr at 150'C, I believe justifies the long extrapolation and warrants very great confidence in much more than 50 yrs of life without failure due to thermal aging. It is not practical or possible to make a

better prediction than this without many more years of testing.

T. W. Dakin, Manager Electrical Performance of Insulating Materials jas

9.0 CABLE QUALIFICATION DATA FOR THE PIOAWG - CLASS I "E OKONITE CABLE USED IN THE PENETRATION NODULE

QUALIFICATION TESTS Or ELECTRICAL CABLES THROUGH SEQUEN IAL EXPOSURE TO HEAT GAMMA RADIATION, LOCA, AND POST LOCA SIMULA'LIONS Engineering Report No.

141 appended herewith is based on test programs performed by Franklin Institute Research Laboratories and reported under cover of their documents F-C 3094 of July 1971 and F<<C 3171 of September 1971.

The date of ER No.

141 is February 29, 1972 and, therefore, it predates IEEE Standard 383-1974.

Although the environmental simulations were more severe than called out in Standard

383, the note at the bottom of Table 5 indicates compliance with the final withstand test requirements.

The specimens listed in ER No.

141 successfully withstood both the PWR and BWR life, accident, and post accident simulations and are discussed in IEEE paper T 74-044-4, copy included.

These represent a more severe simulation than contemplated in ff383, and is in harmony with the broader IEEE Standard f323-1974.

The foregoing is ample evidence that the cables and splice listed therein are suitable for the designed service.

EEM/row E. E. McIlveen November 22, 1974

February 29, 1972 THE OKONITE COMPANY

Ramsey, New Jersey ENGINEERING REPORT NO.

141

SUBJECT:

Aging, Exposure to 200 Megarads of Gamma Radiation and Accident Condition Qualification Testing of Power Cables, Control. Cables and Splice OBJECTIVE:

The. purpose of the program was to determine if control cables, power cables and splices would function properly under the environ-mental conditions expected to be present within the containment of a nuclear-fueled electrical power plant, both during and following a design-basis event (loss of coolant accident).

The program included subjecting the samples to thermal aging, radiation to 2 x 10 rads and to Pressurized Vi'ater Reactor (PWR) simulated accident, followed by exposure to a Boiling V,'ater Reactor (BWR) simulated accident while carrying rated voltage and current; CONC I.USIONS:

Alf samples successfully wi'thstood both the PWR and the BWR accident conditions after heat aging at 121 C for 168 hrs. followed by exposure to total dose of 200 megarad of gamma radiation.

During both the PWR and BWR accident conditions the cables were subjected to rated voltage and current.

In addition, the cable samples and splice were subjected to and withstood high voltage proof tests of two times rated voltage plus 1000 volts for a period of five minutes at the peak temperature and pressure conditions as shown in Figures 1 and 2.

1 CABLE SAMPLES TESTED:

One sample of each of the following types of cable and splice was tested.

The samples were designated as shown in Table I.

Samole No.

Table I Cable Designation A-4 D-4 E-4 B,-4 F-4 C-4 1/C /jl4 0. 030" Okonite,

0. 015" Okoprene 4/C gl4 0. Q30" Okonite, 0. Q15" Okoprene,

0. 045" Okoprene 7/C gl4 0. 030" Okonite,

0. 015" Okoprene,

0. 045" Okoprene 1/C 4/0 0. 055" Okonite,

0. 045" Okoprene 4/C $ 12 0. 047" Okonite,

0. 015" Okolon,

0. 045" 'Okolon 1/C 4/0 0. 140" Okoguard,

0. 065" Okolon with T-95 splice and T-35 jacketing tape

Engineering Report No.

141 February 29, 1972 TEST PROGRAM:

(a)

I Thermal A~in'he cable samples and splice were initiallyaged in an air oven for 168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> at 121 C.

(b)

Radiation Ex osure The samples were then exposed to a dose rate of 1 Mrad per hour for one hour, then p'laced at positions to receive 300, 000 rad/hour until a total dose of 2x 108 rads was reached.

(c)

Irradiation was by cobalt-60 in air at ambient temperature (68-70 F) and a slight negative pressure (-1/2" water).

Cables were rotated and turned. at intervals to achieve a better dose uniformity, This, together with the distance from the source (24") yielded a dose rate variation of up to t10~i'o.

This is, all portions of the cables received at least the minimum dose requested, and some portions (i. e., outer circumference) received up to an additional 10~~'o of the specified dose.

PAYR Steam/Chemical-SDrav Exnosure and Electrical Measurements The cables were then installed in a test chamber for exposure to a PlYR simulated accident'consisting of a 7-1/2 day steam/chemical spray environment while being electrica1ly energized.

The cables were looped on a shelf of perforated sheet metal which simulated a cable tray.

Prior to initiating the environmental exposure in the test chamber, insulation resistance and high voltage (ac) tests were conducted.

The insulation resistance was measured following the application of 500V dc for a period of 1 minute.

The high voltage tests con-sisted of applying 2. 2 kV ac for a period of 5 minutes to the 600 V cables and applying 6. 8 kV ac for a period of 5 minutes to the 5 kV cable.

The IR measurements were performed on the multi-

'onductor cables by grouping the terminal connections as shown in

'Tables 2 and 4.

The results of these insulation resistance measurements, as well as those taken periodically during the remaining portion of the PAR test, are also given in Table 2.

Subsequent to conducting the aforementioned electrical measure-

ments, the steam/chemical,-spray exposure was'initiated with the cables carrying the current and voltage loadings as shown in Table 3.

Table 2

d Daily blaaaarontrnta at td paigitdt F~

Cable Sample

'Xo.

Terminal Connections I

II (Grnd. I Cables in Test Chamber-Prior to Initiating Steam/Chemical Ex losulc During 0

80 psig-3Z4 F Divcll First Second ThIrd

~Da

~Da

~Da At End of Steam/Chemical Fourth Fifth Sixth Seventh Exposure IAt

~Da

~Da

~Da

~Da Ambient Cond.

1 Grnd.

1110

.58 5.8

5. 8

7. 0 9.3

10. 6 1Z. 8

14. 0 700 D-4 E-4 1,3 2,4 Z.4,6 1,3,5,7

71. 0

4. 60

~ 70

~ 0345 I. 45

. 066

1. 58

l. 7Z

l. 84

1. 64

2. 90

2. Z5

3. 56

2. 50

3. 04

2. 78

.o6

3. 45

'.45

~ lal9 Grnd.

235

. 306 I. 74

2. 04

l. 93

2. 35

2. 46 Z. 56 Z. 88 3o00

. F-4 1.3 2,4 SZ. 0

.83

4. 95

5. 92 5.2

5. 92 6.1 6.1 oo O 530 C-4 Grnd.

Z. 09

17. 2 21,2

18. 4

19. 8

19. 8

19. 8

22. 0 o500 Measured after the application of 500 V dc for a period of 1 minute.

80 C5 g) 60 Ot l

td td 40 3 HR 2OM IO SEC RISE TIMF 80 PSIG/324 F (SATURATED CONDITIONS) 40 MIN DECAY

~

~

~

).R. 8 Hl-POT.

ONCE DURING 8OPSIG DWELL; DAlLY DURING THE SEVEN DAY DWELL.

CHEIAlCAL SPRAY MAINTAINED DURING ENTIRE TEST IO,OOO PPM BORIC ACID BUFFERED VSITH SODIUM HYDROXIDE TO A PH OF IO.5 20 I6 PSIG/252F (SATURATED CONDITIONS)

I I

I I

I I

I I

0 4 HRS (MIN.)

7 DAYS FIGURE I TEST PROF ILE OF PV/R EXPOSURE

Engineering Report No.

141 w3<<

February ?9, 1972 Table 3

Cable Current and Voltage Loadinas During PWR and BWR Exposures Sam le No.

A-4 D-4 E-4 B<<4 C-4 F 4 Current Loadin (amp)

18. 0

18. 0

12. 5 280 280

21. 0 Volta e Loadin (volt)

'00 600

'00 600 2900 600 The test profile is illustrated in Figure 1.

It consisted of a 10 second rise time from normal room ambient conditions to a pressure and temperature of 80 psig/324 F.

These conditions were maintained for a period of 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, at which time the pressure and temperature were gradually reduced to 16 psig/

252 F and maintained for a period of seven days.

Throughout the test, the temperature was maintained as nearly as possible at values corresponding to saturated cond'.tions, 100 percent re-lative humidity. while the cable samples were subjected to a chemical spray consisting of 10, 000 ppm boric acid buffered with sodium hydroxide to a pH of 10.

5.'nsulation resistance mea'surements were taken in a manner identical to those described above during the dwell at 80 psig/

324 F, and daily thereafter.

The results of these measurements are given in Table 2.

At the end of the steam/chemical

exposure, at normal room am-bient conditions and with the cable samples remaining in the test

chamber, the aforementioned electrical measurements were again taken with the results given in Table 2.

(d)

BWR Steam Exposure and Electrical Measurements The same cables and splice that had been exposed to the FWR

. accident conditions were then exposed to a BWR simulated accident consisting of an environment of saturated steam while being electrically energized, with the same current and voltage loading as in the PWR exposure (Table 3).

The" test profile is illustrated in Figure'2.

It consisted of a series of transient cycles, each consisting of a rise to a specified prcssure and temperature conditions, a hold at these conditions for a specified time, and a gradual return to the initial condition

Engineering Report No.

141 February 29, 1972 or to atmospheric pressure.

Following the above transient cycles the samples were subjected to 100 days exposure to live steam 0 psig,'212 F.

The temperatures were maintained as nearly as possible at values corresponding to saturated conditions, 100/o relative humidity.

During the 2nd, 3rd, and 5th constant pressure/

temperature period, insulation resistance was:neasured.

During these same periods and at the end of the 100 day test, the. samples were subjected to the high potential ac tests.

The tests were con>>

ducted as described in section (c) for the P%VR exposure, at times indicated by broad lines in Figure 2.

Table 4 gives the results of the, insulation resistance tests for the BVTR accident conditions.

Tables 5 and 6 present the electrical and physical properties after the postulated loss of coolant accidents.

These tests establish the ability of the cables to perform their functions after LOCA. As a matter of information the initial values are also displayed.

FMM/row F. M. McAvoy, D rector Cable Evaluation 've lopment Sworn and subscribed to before me orC this day of ~l(M ie'/<

1972.

Notary Public of N. J.

My Cnnimission Expires August 4, 1976

Table 4

Insulation Resistance Measurements BNR Ex osure "c MA /1000 ft.

Cable Sample No.

Terminal Connections II Grnd.

Gnd.

During 104 psig-345 F Period

5. 8 During 75 psig-320oF Period 1.0 During 25 psig-272 F Peribd 5.7't gnd of Exposure (At Ambient Tem.)

160 D

4N>w 1,

3 2, 4

~ 0093

. 053

. 0069 7340 E-4 B-4 2, 4, 6

), 3, 5, 7 Gnd.

362

~ 092

.61

. 204

2. 24

4. 90 5500 2550 C-4 1,3 4

Gnd.

~ 47

.98

.14 1 ~ 98

l. 96

14. 8 3460 Measured after the application of 500 V dc for a period of 1 minute Sample D-4 was mechanically damaged at penetration resulting in lower reading.

p

120 l5M BHR 20M -""l04 PSIG/345 F Hl-POT TEST ONLY 80 4HR 27M75 PSIG/320 F K

O fC)

<. 60 0)

I L~J

<4O 0) l>jK C3 O

td CO 20 PSIG 258 F l5 PSIG 256 F

>,0 LQ 'g V)

LQ SHR 38M--25 PSIG/272 F IO HR 20 I9 HR 56M 0 PSIG/2I2 F'

-2 PSIG IOO DAYS TEST PROFILE OF BV/R EXPOSURE Fl GORE 2

~~ BROAD LINES INDICATE PERlODS %HEN MEASUREMENTS OF INSULATION RES IS" TANCE AND Hl"POT TESTS Yi/ERE MADE.

~

~ g Engineering Report No.

141 Table 5

Effect of Postulated LOCA - Electrical Following exposure to 2x10 rads of gamma radiation and PWR and BVfR accident conditions electrical measurements were made on Sample C -4, I/C 4/0

0. 104" Okoguard caole with splice removed.

Initial Temnerature C

0 Stress V/ Mil 40 80 SIC 20o 75o

4. 24

4. 64

4. 24

4. 64 P. F.

20 75 I. 19 I 74 I~ 21 1,79 After 2x 108 rads PlVR L BAVR accident 40 conditions 80

3. 79

4. 24

3. 80

4. 29

l. 24~

I. 84

1. 36 I ~ 99 The following electrical measurements were made on singles of sample F-4, 4/C;"',12 0. 047" Okonite, 0;015" Okolon, in 8" diameter coils Initial 40 80 t

3. 41 3 ~ 96

3. 44

3. 98

0. 5I

~ 37

0. 59 1 ~ 40 After 2x 10 rads 8

PVTR L BWR accident conditions 40 80

3. 21

3. 84 3

61 4.19 2

12 2.92

2. 17 3,01 It should be noted that initial measurements were composite measure-ments of insulation and jacket, and the final measurements in the case of the Okoguard was without the jacket while the Okonite was with a deteriorated jacket..This obviously distorts these measurements.

The impo'rtant fact is that the cable willcarry rated current at better than two times operating voltage plus 1000 volts and willallow an orderly shutdown of the station if an LOCA occurs.

Final withstand test:

passed 3600 volts ac (80 V/mil)for 5 n~utes.

~

~

~

~

Engineering Report No.

141 Table 6

Effect of Postulated LOCA - Ph sical Following exposure to 2 v 10 rads of gamma radiation and PAR and BlVR accident conditions, physical measurements were made on the following samples:

~

~

Initial Tensile Strength 200;o Modulus

~go Elongation 912 816 390 1275 956 325 Sample C-4 Sample F<<4 1/C 4/0 0.140" 4/C >12 0.047" Okoeuard Okonite-0.015" Okolon Sample E-4 7/C f14

0. 030" Okonite

0. 015" Okoorene 1310 894 36o After Radiation 2x 10 and PlVR, HVAR Accident Conditions Tensile Strength fo Elongation 900 100 6oo 90 578 30 Physical properties of the C-4 sample were on the insulation only Physical properties on F-4 and E-4 were composite of jacket and insulation

ABSTRACT With the publication of IEEE Guide P 383 For Qualification Testing.

it is appropriate to present typical data relating to this document,and to briefly discuss its significance.

Designed life perforznance for nuclear stations must be predicated for the most part on test data obtained from cable systezns under simulated environznental conditions which are peculiar to this application.

This paper presents data in the areas of (1) long time exposure to moisture at ele-vated temperature, (2) air oven heat aging,(3) sim-ulated reactor radiation during normal operation as well as during and after a design basis event.

and (4) flame testing of cables.

INTRODUCTION With the advent of nuclear power plants in the early 1960's, a new set of operating conditions for electrical equipment had to be recognized by the de-sign engineers.

As demonstrated by the work in the Nuclear Power Engineering Comxnittee, the evolution of design criteria has continued.

The imminent pub-lication of the IEEE Guide for Type Test to Qualify Electrical Cables and Connections, I is evidence of this activity, but it should be regarded as an interixn document which willbe up-dated from time-to-time.

This paper will present typical design and qualification data together with some explanation. of its significance.

Cable systezn designed life perfor-

znance, when based on long service experience can be quite reliable, but significantly different conditions and materials have made simple projections question-able.

Simulated service test data on representative cable constructions

znust, therefore,.be relied upon; and it can provide a reasonable basis for power plant cable systezn design and qualification.

The areas to be studied for qualification through performance testing arer (a) moisture resistance, (b) long terzn physical aging properties, (c) normal radiation exposure. and design basis events postulated on loss of coolant accident (LOCA) and, (d) cable tray fires.

MOISTURE RESISTANCE Moisture resistance is a major factox in deter-mining the normal life of a solid dielectric insulated conductor.

It has become traditional to gain assur-ance of long life performance by totally iznmersing a

812 or 14 conductor insulated with a 45 mil wall of dielectric in water at an elevated teznperature to ac-celerate the deteriorating effects of moisture.

Moni-toring the electrical properties then provide an indi-cation of long term behavior.

In the 1950-57 era with service gained experience that negative dc potential presented the znost severe condition. IPCEA develop-ed2 a 16 week test procedure along these lines based on

• continuous immersion at 50 C while under 600 volts dc. At this time, more than sixteen years later, new generation moisture resisting insulations of sim-ilar geoxnetry can be continuously immersed at 75 C

while under the same dc potential, and survive from l-l/2 to 2 years, or more.

This is at least 5 tiznes longer and at an effective temperature acceleration rate of 6 times greater than anticipated by that IPCEA procedure.

Since insulated conductors of the 1957 vintage dielectrics installed at Shippingsport, Indian Point and Peach Bottom, among others, have not ex-perienced distress due to moisture. it can be reason-ed that control cable insulations now specified which have the capability of withstanding total Immersion at 75 C under 600V dc as discussed herein should de-velop the designed life of the cable plant.

Fig. 1 pre-sents data for a 45 znil wall of an ethylene-propylene base insulation conductor, and Fig. 2 illustrates the electrical behavior of a composite wall coznposed of 30 mils EP base plus 15 mils neoprene compound.

Reference to Table I discloses siznilar data for an ethylene-propylene base dielectric and also a flame resistant cross-linked polyethylene compound (FR-CLPE), but at 90 C continuous water immersion while under 600V ac potential except when percent power factor (e'o PF) and the specific inductive capacity (SIC) are being zneasured at 40 and 80 V/mil'ac.

Following each test measureznent the specimens were subjected to a 5 minute withstand test at 110 V/mil. The spe-cific insulation resistance (SIR) were xnade at 500V dc while at 90 C.

The difficulty of predicting long term performance based on the customary 2 week test data is obvious.

It may be of interest that the tizne to failure for a particular specimen is a complex func-tion of several variables, one of which is the degree of mechanical perfection of the dielectric wall.

Fail-ure is often sudden with little or no forewarning, and occurs when the cable is undergoing 60 cycle power factor and capacity zneasurements, or during the sub-sequent withstand at 110 V/mil.

CLASS IE CABLES FOR NUCLEAR POWER GENERATING STATIONS E

E McIlveen V L. Garrison G. T. Dobrowolski The Okonite Company

Raxnsey, New Jersey Paper T 74 0444, recommended and approved by the IEEE Nuclear Power Engineering Committee of the IEEE Power Engineering Society for presentation at the IEEE PES Winter Sleeting, New York, N.Y.. January 37 February l. l974.

Manuscript submiued August 3l, l973: made avadable for printing November l6, l973.

Fig.

3 not only shows the SIC values for an ethylene-propylene base insulation during a-long term continuous water immersion study, but also the accel-erating effect of temperature as manifested by a

chango in the 60 cycle capacity.

The 142 C/42 psig steam autoclave exposure further accelerates the in-crease in the SIC value but could change the reaction mechaniszn.

In any event, if plotted on Fig.3 the end point is still some two years out on the tizne scale.

1121

Qs

$.2 S.I SO IO

.5J

.tS

.o ra r/

aaa o

o o

g 8

ei n

n n

n n

n

~o

/r 4

Matc

/r O

O 50002 AC SVOL WITHSTlHO AT EACH VAASVREVOIT IOAWO..OOS'WALL EP SASE lISIAATICN Io g

O 5

0 5

20 25

$0 SS o0 o5 50 SS CO 6$

WATER IVVERSCN TIVE WEEIIS AT TS/C

~

~

~

~

o

~

% PF IV ra 4

n A0 8

8 Fig. l. Water immersion Test of EP Dielectric Under 600V Negative DC IO >WC

~ 0$0'P SASE IHSVLATIOH

.00 HEOPREHE P

OL a

p -r

~r aL /

Af r

~r P

5SOOV. AC 5 VIV.WITHSTAIO

/

AFTER ElCH VEASIAIEVEHT cAPF I200 5

IO I$

20 2$

$0

$5 OO 45

$0 5S 50 5$

IO I$

50 55

$0 WATER HIVERSEO TIVE IH WEEXS AT TSC Fig. 2.

Water immersion Test of EP/Neoprene Under 600V Negative DC Table I Time Stress PF SIC SIR Period V/mil

+o C4 90 x 103 1 day 7

II 14 II 28 2 mos.

II 40 80 40 80 40 80 40 80 80 80 80 80 80 2,84 3.09 1.3 2.89 3,09 1 ~ 52 3.05 1.8

l. 55

3. 05 1 36 3 07 2

1 1.36 3.07 1.13 3.08 2.3 1 ~ 16

3. 08 1.10 3.09

3. 1 0.87

3. 17 3,5 0.79 3.20 4 3 0.70 3.26 4.7 0.70 3 ~ 30

5. 1 PF SIC SIR

'Yo

@90C x 103 1 ~ 06 2.88 2.0

1. 09

2. 88 1 ~ 07 294 30

1. 08

2. 94 1 ~ 09 2.95 2,9

l. 11

2. 95 1.24 2.96 2.8

1. 25

2. 96

1. 51 3 ~ 11 3 ~ 5 2.37 3 ~ 17 4.5 3,31 3.28 4.7 3 ~ 43 3 ~ 36 54 continuing 90 C Water Immersion-'-600V AC

. 045" Wall EP Base FR - CLPE I

ILI I-I2 (7I CO 2In IA IIS O

@0

.045 EP BASE DIELECTRIC 6OCN.

CONTINUOUS S.I.C MEASURED AT BOV/MIL.

5 YRS.

4 YRS.

3'RS.

3 YRS.

2 YRS.

I inYRS.

I YR.

6MOS.

2MOS.

1122 3.I 3.2 3 3 S.I.C.

Fig.3, Accelerating Effect of Temperature on SIC Values During Water Immersion

LONG TERM PHYSICAL AGEING PROPERTIES Data on electrical behavior in a combined heat and znoisture environznent, as covered in the previous

section, may be a better guide to service related per-

~

~

~

~

formance than that developed frozn simple physical aging in a dry air oven at elevated teznperatures such as 136,

150, 165 and 180 C.

The curves presented in Fig.4, 5,

6 and 7 are based on such data obtained through standard procedures.

3 These can be ana-lyzed by the Arrhenius technique

and, by analogy the useful life znay be predicted.

It should be. recognized that the Arrhenius equa-tion is valid only if the data represents a

single discreel cheznical reaction and the activation energy of that single reaction is within the teznperature limits of the data.

This equation can be derived frozn colli-sion theory and has been experimentally verified.

It serves to define the temperature coefficient of a dis-creet cheznical reaction and the activation energy of that reaction only within the teznperature limits of the experiznental data.

The equation is:

C ZQ X

120 IOO 20 OiEP BASE - COAKO CU. UNAGED EL 350 Q~EP BASE ON BARE CU.UNAGEO EL290 Q~CLPE 814 AWG

~ 047" Wall

/

~lz~~

0 o

0 I

0 0

2 4

6 8

IO I2

-DE k = Ae k

=

specific rate constant, A

=

frequency factor or collision frequency, activation energy -

the difference in the energy of a chemical species in the ground state and its activated state.

The activated state is not isolable and has a very short life time (In the order of nano or pico sec-onds) and collapses either to the original ground state of reactants or to the ground.

state of the products.

CO 00 a

g WEEKS AT I

C Fig, 4.

Air Oven Aging at 135oC BLI~ZB>~'iL 0> ctMc cc4Tzs QL vwco CL ~ N0%

o+ clMC cN wc cii wkccO cL+c&%

s'CLK 0

R T

gas content, absolute temperature 0

0 The specific reaction rate constant k represents a

single discreet chemical reation.

In the case of a simple uni-molecu)ar first order reaction A~ B, the following describes the rate where C = concentra-tion:

dCA kCA dt o

that is -- the change in concentration of reactant A with time is proportional to the initial concentration of A.

The differential equation is solved and k deter-mined frozn experimental measurements of concentra-tion vs. tizne. Distinct values of k must be determined at various teznperatures and znust be constant over a considerable range of conversion in the reaction, say from 20 to 80/o, for the data to be considered valid.

It can be used correctly only when there are discreet chemical reacrions whose rate can be precisely

measured, and described by a solvable differential equation.

A straight line will result from a plot of the logarithcm of the reaction cate k vs. 1/T provided there is no change in the reaction mechanism.

Fig. 5.

Air Oven Aging at 150 C ICI l4 Jell 047 wlU.

4 ~ ce IAsc ~ Iw4cD ceo iso ~

o ~ cr sec oc sac cu wasco g. ~ c~

s ~ ciec 0

0

~

a s'0 0

a 4

0 ICO 200 20

~.6JQS Fig. 6.

Air Oven Aging at 165 C

1123

t

wlu +ca ooo'

~ cooooc

\\Nkcco rL~ Mo%

I YR Q txoooox Ql ooox oo +wing Q,oooo'5 4

s ~ cloc 0

0 o

/ct P

~o oo H(Vtt 4T lOOC E

Fig, 7, Air Oven Aging at 180 C

5 90'c IZl x35 150 l65 175 In applying this Arrhenius analysis to the aging

data, the time to 40% loss of elongation is plotted on semi-log paper against the reciprocal of the absolute temperature (T) in degrees Kelvin. This is presented in Fig. 8.

In examining the validity of this treatment of the data obtained frozn Figs.

4 through 7

of this

paper, note there are at least four simultaneous re-actions: - (I) oxidative cleavage, (2) oxidative cross-linking, (3) therznal cleavage, (4) thermal cross-link-ing.

The first two of these reactions are at least second order in their rate law axxd must depend at a

minimum on the concentration of oxygen and the con-centration of reacting chemical bonds.

Since vulcan-ixed rubber is a complex xnixture of many chemical bond species, there are a multitude of individual'rate constants that must be measured.

This is an impos-ibility.

From the above, it could be argued that 'the occurrence of a linear plot in Fig. 8 in an Arrhenius treatment of aging data is a fortuitous event, but it does provide a means of comparison within the tern.

eratuz'e range of the data.

Nevertheless, the significance of loss in elonga-tion is related to the ability of the insulation to with-stand bending without physical cracking and ultimate electrical failure when moisture enters.

A 40/o loss still leaves 60% retention which probably represents an elongation on the order of 180% whereas 50%ulti-mate is usually reached before serious cracking de-velops.

Since thermosetting insulations of the 1957 type vintage have performed well in the nuclear plants cited herein and since Fig. 8 shows that the ethylene-propylene base and the cross-linked polyethylene in-sulations take 6 times as long to reach 40% loss of elongation as does butyl, it is safe to predict that these new insulations 5 will provide superior aging performance in service as far as this property is concerned.

During the development this aging data, it was

observed, as shown in Figs.

4 through 7 that at these elevated ternperaturcs an alloy coated copper con-ductor specimen out-perforzned a non-coated copper conductor.

This is "due to the catalytic effect of the copper on the degradation of organic dielectrics.

J n

n oo

+XXX N

oo

+

oo ol oo AI Fig. 8.

Time to 40'/o Loss of Elongation RADIATION EXPOSURE In 1968 studies were made on the performance 6

of 13 elastomer-based insulation/jacket combinations after'gamma irradiation in air.

Based on the effect upon the physical aging, electrical properties, znois-ture and steazxx resistance, and flame resistance, specific ethylene-propylene and cross-linked poly-ethylene insulations were found to be suitable for nuclear power plant service.

By mid-1970 qualification testing of specimens prc-aged prior to garnrna radiation to 3.5x 107 rads followed by a

siznulated loss of coolant accident (LOCA) bccazne necessary for design acceptance.

Again a specific ethylene-propylene insulation proved to be suitable.

The amount of aging had no adverse effect on the electrical properties of the ethylene-propylene base insu)ations,

. Table II presents the data.

Further reference to Table IIwilldisclose a

number of interesting points.

Sample B, a Hypalon base dielectric had good electrical and physical prop-erties following the simulated

aging, that is"before" radition.

"After" the LOCA simulation in the auto-clave the electrical properties fe)1 of significantly.

The physical appearance and resilience,

however, were good.

Saznple C represents two specimens which con-tained a hand-wrapped splice made with an ethylene-propylene base tape.

Although not pre-aged, there was little or no deformation from the cables laid on top of these two splices, and they rexnained resilient and firzn at all szagcs of the test.

1124

TABLE II Nuclear Simulation of October 1970 S~em i e:

A - 1/C 14 AWG coated copper,.060" EP Base insulation B - 1/C 14 AWG (7X) coated copper,.030" Hypalon Base insulation C - Hand wrapped EP tape splice 0

irradiated.

Radiation:

e To total of 3.5x10 rads at rate of 5.2x10 rads/hr.

LOCA:

12 hrs. steam lm 305 F/60 psig + 168 hrs.

230 F/5 psig with PWR chemical srpay while energized at rated voltage, Saznple A also received 10 hrs.

steam at 350oF/120 psig.

Pro erties P, F,

/d 9 80 V/mil Days

~Ae d (1) 6 (2) 10 (3) 14

~

Before

.43'36

.31 Adme

.63

.61

~ 63 Sam les A 1-3 Sam les B 1-3 Samaies C 1-3 Before 5,76

3. 52

4. 02

.42 1.81 (no pre-aging)

After Before After S,I.C. 9 80 V/mil (1) 6 (2) 10 (3) 14

3. 12

3. 12

3. 13 3 26 3,31

3. 30

6. 16

6. 32

6. 33 S, I.R, (x 10

)

(megs 1000')

(1)

(2) 10 (3) 14 20 20 20 16 16 16 2.5

2. 9 2.2

'DL DL DL Withstand, 5kV ac h dc

Tensile,

% Loss (1)

(2) 10 (3) 14 6

2 6

+5 9

4

+23 B

B

+10

+ 5

+

1 passed Elongation,

% Loss (1)

(2) 10 (3) 14 ll 14 51 57 54 46 B

B 60 72 68 Physical Condition E

= Excellent, resilient and firzn DL = Dead Leak G -"Good B

= Bonded so tight it could not be removed from the strand Within a year, reassessznent of design para-meters dictated higher level of radiation exposure and new LOCA simulation-profiles for qualification.

Table IIIexhibits the details for a

total disage of 1 x 108 rads plus a LOCA for a boiling water reactor (BWR). 8 Tho steam exposure at 212 F was continued for 100 days at atmospheric pressure with a "kicker" to 20 psi for the last 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.

This was followed by a 500 volt insulation resistance measurement and a

5 rninuto ac withstand tost.

The specimens success-fullywithstood the simulations and proof test.

Table IV and Fig. 9 present the LOCA siraula-tion for speciznens radiated to a total dosage of 2 x 10 rads.

This simulation is particularly severe because these specimens were loaded to rated cur-rent and voltage and tested to the pressurized water reactor (PWR) incident profile, and then these same specimens were subjected to a BWR profile.

It may be noted these profiles include oae 'peak" at 324 F/80 psig, two peaks at about 342 F/104 psig, and one at 320oF/75 psig plus the 100 day soak.

These multiple transients demoastrate that these specimens have significant margin, a performance characteristic requested in the Guide.

Additional qualification tests covering several differcat cable constructions are in progress at this writing.

The Appendix to the Guide P 383 suggests relative short term procedures in Section 2.4 which have evolved from the type of studies aad testing described above.

Eveatually, new data obtained in environzneats more closely approaching long terzn operational conditions may show that sequentially synthesized effects of temperature, radiatioa, atznosphere and movomeat are unnecessarilysevere.

1125

TABLE III Nuclear Siznulation of Jul 1971

~Sam les:

A - 1/C 14AWG.030" EP Base insulation+.015" Neoprene Base Cover.

B - 7/C 14 AWG sazne as A, cabled,

+ tape and.060" Neoprene Jacket.

C -

1/C 4/0 5kV EP Base

+ Hypalon with EP/Neoprene Splice.

Radiation:

To total of 1 x 10 rads at rate of 1 x 106 rads for the first hour followed by 3 x 10 rads per hr. for 330 hrs.

LOCA Periods:

BWR sequence of 55 minutes steazn Q 304 F/104 psig

+ 3 hrs.

20 min. 5 346 F/104 psig +

4 hrs.

27 min. N 320 F/75 psig

+

1 day at 256 F/15 psig minimuzn + 100 days at 212 F/0 psig, with last 10 hrs.

259 F/20 psig.

Electricals Durin LOCA:

~Sam le A

B C

Loadin

~Am esca Volts 600 600 2900

@ 346 165 525 950 Insulation Resistance 0

@ 320o 350 420 1800 0 SIS 2700 2600 12000 Periodic ac Withstand 1.3 kv OK 1.3 kv OK 6.8 kv OK NOTE:

e znegohms-foot, ss 5 minutes at end each period.

TABLE IV Nuclear Simulation of September 1971

~Sam I s:

Same as in Table III Radiation:

To total of 2 x 10 rads LOCA Periods:

Current and voltage loadings as jn Table III during 4 hrs. N 324 F/80 psig

+ 7 days Q 252oF/16 psig in addition and prior to the sazne BWR sequence detailed in Table III.

Fig, 9 shows this profile.

Electricals Durin LOCA:

~Sam le A

B C

500 270 800 850 4500 160 460 1700 5500 1600 12000 3460 Insulation Resistance ss Megs - 1000 ft.

@ 346

@ 320

@ 272 Final Q 70o Periodic ac Wzthstand AI>

2.2 kv OK 2.2 kV OK 6.8 kv OK NOTE:

As megohms-foot, sze 5 minutes at end each period.

80 60 ONCVtOALSIIIATVAINTAINCO 80PSIG/324F I 00SIPIO PN II TEST 10.000 CSATQRATEO PPv OOIIIO AOOOOFFEACO 2QM CQNQITICNS) s 0IITH SOOIIAS IOOIIOSIOC TO A SA OF IOS 40MOI DECAY l04 PSIG/ 345 20F

- - 75PSIG/3 OIIOAO UNCS LOCATE ICIIIISSSI WNOS VCASISICVCNTS OF OOIAATSSI RESISTANCE ANO

'S POT TESTS SICAE VAOE 20 IOSEC RISE TIME 4 HzS (MOQ (SAT(HATEO CONDITIONS)

I I

I I

I I

I 7 OATS PWR EXPORRE

-2 PSIG 3HR 38M 256F

- 25PSIG/272F 0 PSIG/2I2F BWR EXPOSURE Fig. 9.

Combined LOCA Simulations 1126

CABLE TRAY FIRES h TESTS Table V While not confined to nuclear plants, the losses incurred through cable fires in such installations has focused attention on not only the flame resistance of the

trays, both with and without covers oz baffles, the degree of separation, the matter of circuitredun-

dancy, and the effectiveness of various fire extin-guishing systems.

Studies on cable behavior, and the effect oE physical arrangeznent and separation have been covered in an earlier paper. 10 Redundancy and fire extinguishing systezns are subjects not relevant to type testing to qualify cables and connections other than to note that the redundancy concept among other

things, has znade obsolete the requirement of "time to short-circuit".

Also note that thermoset insulated and jacketed cables are zelatively unaffected by CO2 discharges from protective systexns.

Oil/Burla vs. Gas Flame Tra Test 3~cairn:

7/C 12 AWG Cu,. 030"/.015" EP base

+

neoprene, cabled,

. 060" thermoset jacket, OD M

P 7PIt Commercial Gas Oil/Burla Texnp.

Flame o F Height Max.

tccb s

Test Time Minutes Temp.

Flame oF Height Max.

tccbe 0

1260 1190 1190 1190 1190 1180 1170 1140 930 820 0

30 30 24 22 20 15 12 10 6

5 0

2 4

5 6

7 8

10 12 16 20 1450 16 1490 20 1500 24 1500 24 1500 28 1480 30 1460 36 1450 36 1450 30 1450 22 1450 16 Three in 7 to 16 minutes Short Circuit Range Allx in 5 to 9 minutes a

1 znin.:20 sec.-

40 inches 44 inches No After Burn 2 min.x30 sec.

Core Damage 15 inches Jacket Char 26 inches Propagated No Table VI Commercial vs.

Propane Gas

~Sclme:

7/C 12 AWG Cu,.030"/.015" EP base

+ neoprene,

cabled,

.060" thermosct jacket, OD =

0, 73" and 0, 71",

(From two different production runs)

Commercial Gas Pro ane Gas Test Txme Minutes Temp.

Flame oF Height Max.

Iacbes Temp.

Flame F

Height Max.

Inches 16 28 28 30 32 46 54 42 38 29 16 1450 1480 1480 1500 1500 1490 1480 1480 1470 1460 1450 0

2 4

5 6

7 8

10 12 16 20 16 24 28 30 32 36 42 48 40 28 16 1450 1480 1490 1490 1480 1480 1480 1480 1460 1460 1450 The type test data in Table Vprovides a compar-ison of the two different flame sources described in paragraphs 2.5.4,4 and 2.5.4.5 of the Guide, Part II.

Table VI gives test data which establishes the sixni-larity of results between specimens subjected to the Natural Grade propane gas flazne and those tested with the commercial gas flame.

Table VII presents data developed on small diameter cable constructions with different insulations.

Allx in 6 to 8 rninutcs Alit in 6 to 8 xninutes Short Circuit Range The "time to short-circuit" range is noted in Tables V, VI and VII to permit cross-comparisons to illustrate the variations which, even for ostensibly identical cables, range from 5 zninute's to 9 zninutes with mavericks (unknown) in some test runs as low as 4 zninutes in the gas fired test, and up to about 12 minutes with only one cable out of six failing in 0 znin. x 20 sec.

43 inches 51 inches No 1 min. x 25 sec.

41 inches 45 inches No After Burn Core Damage Jacket Char Propagated It follows that cable performance during a tray fire should be postulated as a design basis event (DBE), and a method of qualifying cable for tray sys-tems had to be developed.

The "Philadel'phia" Tray Cable Fire Propagation Test, which was devised in

196511, simulates a Eire in a vertical tray, asitua-tion which is more severe than with a horizontal tray.

The flame source in the original procedure was cruznpled oil-soaked burlap rag.

In the following years much efEort was expended in trying to eliminate some uncontrollable variables.

While a definite bur-lap folding sequence and oil dipping procedure were developed, and these did improve reproducibility

somewhat, the replacement of the oil-soaked burlap with

• large ribbon type gas burner together within-strumented control of the air/gas flow resulted in a

coznpletely reproducible flame environment.

This in-advertently resulted in a znore severe test because (a) the gas flame provides a constant heat input for the entire test period whereas the oil-soaked burlap flame reaches a znaxirnum within several minutes and then tapers off until it finally flickers out.

Further-

more, (b) the gas flame projects right through the cables thereby completely surrounding them whereas the oil-soaked burlap flame is vertical and in Eront of the cables so that the majority of the heat reaches about half the circumference.

The procedures for both of these methods may be found in Section 2.5 oE P 383.

1127

the oil/burlap fired test.

This "tizne" incidentally, is established when a 120/240 volt znonitoring circuit lights a small laznp.

It has already been deznon-strated10 that this "time" is also a function of the nuznber of conductors in the cable and/or their size.

Obviously it is the amount of heat sink provided by the metal that governs such results.

Other indepen-dent variables are the type and volume of insulation and cover.

Eillers, cable tapes, jacket znaterials, and metallic arznors.

This complex array of vari-ables alone precludes the establishment ofameaning-ful perforznance level denoted as "time to short-circuit" and is another reason why this was dropped from Section 5 of the original draft of the Guide Appendix.

Table VII Effect of Construction Tra Test S~ee'mete:

A - 7/C 16 AWG Cu,

~ 031" silicone + glass braid

cabled, tape, glass braid OD = 0.045" B - 7/C 16 AWG Cu,.030" flame resistant CLPE,

cabled, tape, neoprene jackeC, OD = 0. 50" C - 7/C 12 AWG Cu,.020"/.010" PE

+ PVCI

cabled, tape,.060" PVC jacket, OD = 0. 58" Test Time Semel A

~Saln 1

B

~Sam le C Again referring to these variables, the flazne height distances are subject to some error since they are visual and estiznated aC a stand-off distance.

The core damage also involves judgment, as does the overall jacket char distance measurement.

The only pracCical clear cut observation that can be made is that of propagation.

Failure by this znode is defined in paragraph 5.5 of the Appendix as occurring when the fire burns all the way up to the top of the tray, a distance of about 6 feet above the flazne source center.

ANone of the cables reported in Tables V and VI siiows this distress, but a review of the data in Table VII willdisclose a typical failure with Sample C.

1400o Gas Burner Minutes 2

4 6

8 10 12 16 20 Short Circuit AEtcr Burn Core Dam.

Jacket Char Propagate Flame Height Inches 18 24 18 16 16 16 16 16 None None 12" Z2" No Flame Height Inches 20 30 38 24 24 16 16 16 3'o 5'one 2 811 3311 No Flame Height Inches 28 30 39 40 60 72 74 74+

3'o 4'ontinued 7411 +

~ 74" +

Yes In addition to vertical tray flame tests de-cribed herein, a specimen of each type oE instruznent cable or the individually insulated or insulated and jacketed conductors removed from each rnulticon-Table VIII Unarmored vs. Arznored Cable ductor control cable which is type tested should pass Soecimensx 19/C //14 AWG Cu,. 030"/.015" EP/Neo.

a flame resistance test in accordance with ASTM D 2220-68, Section 5.

This precaution is taken to not only insure that the small single conductor com-ponents or units are flaxne resistant, so as to not con-tribute to a tray fire, but to also zninimize propaga-tion in control cabinets where the outer flame resis-tant jacket, armor, or other covering had been re-xnoved to permit spreading these individual conduc-tors or units for connecting to equipznent.

DISCUSSION It should be noted that the data presentedherein was developed on cables with specific constructions, insulations, and coverings.

It would be neither fair nor correct to assume that all insulations tagged as EP.

or FR-CLPE, or silicone, for example, werc identical to each other and would, therefore, perform the sazne.

This is also true Eor jackeCs compounded with neoprene.

Hypalon. or PVC.

Samnle A Temp.

Flame oF Height Max.

Inches Gas Burner Time Minutes Samole B Temp.

Flazne oF Height Max.

Inches 1530 1600 1620 1600 1600 1630 1600 1590 1560 16 36 38 38 40 60 7Z 72 +

72 0

1530 2

1560 4

1580 6

1600 8

1630 10 1650 12 1670 16 1640 20 1650 16 16 16 36 46 42 42 Z4 16 All: 5'.30" to 81 3011 Short Circuit Allx 5' Z5" to

61. 10>>

A - cabled,

tape,

~ 060" general purpose thermoset B - cable A plus.020" steel interlock armor 11 "20" 66 inches 72+ inches Yes After Burn

- 1'z30" Core Damage

- 28 inches Cover Burned 18 inches Propagate

- No While the moisture and aging data can be repzo-duced to a good degree of accuracywitha spread less than 10/o, the flame test data is largely based on qualitative observations

which, together with sample variations, result in considerable swings from one data sheet Co the next.

The consequence of the latter is a "go" or "no-go" criteria for propagation.

Fortu-nately this is the only performance characteristic which is really necessary to check-out. It is practical and was developed for a 7/C 1Z or 14 construction.

Note: - During the flazne testing of the interlock ar-mored cable, spasmodic gas bursts were observed up to the ZO inch level. After the test the steel was slightly rust colored and sooty.

1128

It follows that numerical values should not be lifted out of context for use in purchase specifications.

In general, the data presented herein is indicative of the type of Information that could be useful in estab-lishing long term behavior of a particular product to qualify a "line" for use in a nuclear power plant.

CONC LUSIONS l.

Based on the type of data presented herein, it can be logically shown that there are insulated cables which should survive the designed life events in areas of (a) moisture and steam, (b) heat, (c) radi-ation and LOCA, and (d) fire.

2.

There are constructions whose performance excels in several of these four environments, but very few will do well in all of the situations cited in 1.

3.

In view of the time element, namely I-I/2 to 2

years for water immersion testing, the moisture re-sistance data must be handled by certificdtest re-ports.

Note there is no proven znethod bywhicha two week test period can establish long term perform-ance+

4.

It is logical that ii a "new" insulation which perforzns say 6 tiznes better than a known dielectric that has a least a ten year established service record in a similar environznent, the new insulation could be expected to achieve the cable system designed life in the sazne environment.

5.

The designed life performance under radiation a'nd LOCAshould be considered acceptable if, assum-ing the two previous znentioned areas have been satisfied, the cable survives two or znore LOCA peaks.

This demonstrates extra znargin.

6.

While the cable tray flazne test does not require more than a few hours to perform, the logistics dictate it be called-out as a type qualifica-tion test -- not a production test.

7, Lack of space precludes presentation of addi-tional daCa upon which to base cable system design.

The data presented

does, however, indicate the acceptability of the EP base/neoprene cable with a

flazne resistant thermoset jacket.

The addition of an interlock or corrugated armor improves the flame resistance and further reduces the possibility of tray fire propagation regardless of the cable size.

ACKNOWLEDGE MENTS This paper presents the results of work perfor-med in The Okonite Company's Engineering and Research Laboratories.

The authors wish to espe-cially acknowledge the work on air oven aging and the Arrhenius equation analysis by Dr. J.

S.Lasky, Vice President-Reseazch and his associates. Mr.W.H.

Steigelmann and Dr. S.

Caz fagno of Franklin Institute Research Laboratories deserve special men-tion for their work in the development of qualification testing.

REFERENCES

1. Guide for Type Test of Class IE Electric

Cables, Field Spflces, and Connections for Nuclear Power Generating

Stations, IEEE/ICC WG 12-32 and NPEC S/C 2.4, IEEE P383, 1973.

2. Method for Determining the Suitability of Insula-tion Compounds for Use on D.C. Circuits in Wet Locations, IPCEA Handbook Procedure T-22-294 dated May 16, 1957.

3. Aging by the Air Oven Method, ASTM D 573-67.

4. Siznplified Methods of Calculating, Insulazion Life Characteristics, L. C. Whitmann, A'IEE Trans-

actions, October 1961, Part III, Power Apparatus and Systems, Vol. 80, pp 683-685.

5. A New Corona and Heat Resistant Cable Insulation Based on Ethylene-Propylene

Rubber, R.

B.

Blodgett and R. G. Fisher, AIEE Transactions Paper 63-162, New York Winter General Meet-ing, January 1963.

6. Insulation and Jackets for Control and Power Cables in Therznal Reactor Nuclear Generating

Stations, R. B. Blodgett and R. G. Fisher, IEEE Transactions, Vol. PAS-88, No.

5, May 1969, pp 529-541.

7. The Franklin Institute Research Report F-C 2830, October 1970.

8. The Franklin Institute Research Report F-C 3094, July 1971.

9. The Franklin Institute Research Laboratories Re-port F-C 3171, September 1971.

10. Flame Tests -- A Systems Appzoach for Power and Control Cable, F. M. McAvoy, IEEEConfer-ence Paper C 72 121-7, presented at the Winter Power Meeting, January 1972.

Note:

Since P383 has not been published at the date of this u riting. the final Section number references may be different.

11. Report on Control Cable Flammability, R.

H.

Logue, Insulated Conductor Committee Min-utes of Noveznber 17,

1965, Meeting in Philadelphia, Appendix F-4.

1129

Discussion T. H. Ling (Anaconda Wire and Cable Company, Marion, Indiana): I have the following two comments:

First. One of the author's major conclusion was that "a new insula-tion which performs say 6 times better than a known dielectric that has at least a ten year established service life in a similar environment, the new insulation could be expected to achieve the cable systems designed life in the same environment". As stated on page 4 of this paper. the known dielectric is Butyl. We wonder whether the authors have either physical or electrical test results on the specific butyl insulated wire which has served ten years inside nuclear containment.

We all know that butyl's aging and radiation resistance is far poorer than EPR's.

Does this offer us some indication that the current testing method for Class IE cable is rather unnecessarily severe?

Second. In this paper, the authors showed that the flame resistant cross-linked polyethylene insulated wire possesses acceptable moisture resistance, good aging and flame resistance.

Unfortunately there is no LQCA testing data available.

We suggest that LOCA simulation test results on flame resistant cross-linked polyethylene insulated wire be included in such a presentation in order to give a whole story of flame resistant cross-linked polyethylene insulated wire forClass IE application.

Manuscript received February 6, 1974.

E J. McGowan and F. E. LaFetra (Raychem Corporation, Menio Park, Ca.): The authors have presented a very timely paper since the P383 Guide will soon be issued. The data presented should offer considerable encouragement to the users of electric cable because it points out the work being done by manufacturers to provide reliable products for Manuscript received February 19, 1974.

nuclear power generating stations.

We would like to review several important aspects of the paper.

The authors present long-time immersion data at 75'C with 600 volt negative dw voltage applied to the specimens in Figures I and 2. In Table I, additional data is presented using a 90'C water immersion temperature with a 600 V. aw voltage potential applied to the speci-mens. Has any correlation been found between the effects of the ae and dw on these specimens?

~

In Tables II, III, and IV, the results of tests during and after "LOCA"simulation are presented.

The physical condition of the speci-mens after test are described only after the 3.5 x l07 rads exposure and not after the I.O and 2.0 x l08 rads exposure. Are the specimens able to be straightened and recoiled around a mandrel with a diameter of ap-proximately 40 times the overall cable diameter as described in para-graph 2.4.4 of the P383 Guide?

The theoretical discussion ol'hermal aging states that, "Distinct values of k must be determined at various temper-atures and must bevbpstant over a considerable range ol'con-version in the reaction, say from 20 to.8¹, for the data to be considered valid. It can be used correctly only when there are discreet chemical reactions whose rate can be precisely

measured, and described by a solvable differential equation.A straight line will result from a plot of the logarithm of the reaction rate k vs I/T provided there is no change in thc reaction mechanism."

The next paragraph goes on to say that the data of Figure 8 represents at least four different simultaneous reactions and that in vulcanized rubber there is a multitude of individual rate constants. In spite of these complicating mechanisms Arrhenius plots arc straight lines and this is considered by the authors to be a fortuitous event. In point of fact, careful examination of their data points shows that the curves for CLPE and EP are not straight lines but in each case can be represented more accurately by two straight lines. indicating the presence of different rate controlling mechanisms over thc pertinent rO rrr rz) lcs I YR 4>> ~

~

rvrO cn ccr CCr IO Author's lines C<

Vp0 I MO.

O Crr 0) tD lO 75 90'C l2I

I35, I50 I65 I7518 I

Fig. I Rewvaluation of Author's Figure 8 showing original data points and the occurrence oi'wo distinct slopes in EP and CLPE insulations.

1130

QUALIFICATION CERTIFIChTION 6 F - C 3694-I CLASS IE ELECTRIC CABLES FOR NUCLEAR POV'ER GENERATING STATIONS The cables listed below were subjected to a combined aging, radiation and LOCA, life simulation test by The Franklin Institute Research Laboratories.

The test was performed in accordance with IEEE Standard f383-1974, but r

with simultaneous exposures to simulated life aging and a thirty day loss of coolant accident (LOCA) including multiple temperature/pressure

peaks, all while exposed to radiation, the cumulative dose amounting to 2x 10 rads.

8 This was followed by a 100 day post LOCA steam exposure at 100 C, after which these insulated conductors withstood a 80 V/mil ac proof test as called out in 8383.

Qi This is a more severe simulation than contemplated in f383, and is in harmony with the broader IEEE Standard f323-1974.

The foregoing is ample evidence that the cables listed below are suitable for the designed service:

1/C f12 (7X) coated copper,

.045" Okonite insulation 7/C

$ 12 (7X) coated copper,

. 047" Okonite+. 015" Okolon, Okolon jkt. overall 1/C gl2 (7X)'coated copper,

. 030" Okonite+. 015" Okoprene 1/C II 6 (7X) 5kV Okoguard, copper tape shielded cable.

EEM/row Sworn and Subscribed to before me of November 1974.

< d. Pr7/i:

E. E.

McIlveen Notary Public of New Jersey My Commission Expires August 3, 1976

e 5-'. ('

tl '.!l'Ile@ II t e

e &~ eee 'Mise Final Repor t F-C3694 Report TYPE TEST CABLE EQUALIFI CATION PROGRAH AND DATA FOR NUCLEAR PLANT DESIGNED LIFE S IHULATION THROUGH S It1ULTANEOUS EXPOSURE e

January 1 974 Pvepaz ed foe The Okoni te Company

Ramsey, New Jersey 07446 A

~

~

~

~

~

~

~ ~

~

l1IIDTIIF FRAXKLI'NINSTITI.TF RFSFARCII LARORATIIRIFS

~eg I~ I eepklJ

~ ee

~ ellee ~ I<ee I AA4Hrke I eee

~ JOI gIeee

~ 4 I I eeeeee

~ ~ Aee

PREFACE Having qualified Okonite nuclear plant cables in previous tests, FIRL g F-C 2830, F-C 3094, and F-C 3171, through sequential expo-sures, it was felt that a behavioral study of various cable constructions under simultaneous radiation/aging and radiation/LOCA would be more r

meaningful and of value to engineers since these events would more closely approach a postulated service life than in prior studies.

The temperature-time profile, Figure dl, graphically shows, normal heat aging, simulated reactor life, and a 30 day LOCA, both while under radiation, and then a 100 day post LOCA steam exposure.

As a result of this broad study, relative performance ratings can be given for basic cable constructions, This investigation follows the guidance provided in IEEE Standard f383, and goes beyond it, embracing the concepts called out in P323.

It should be noted tnat these various simulated events were purposely designed to find "end points" rather than to simply reach a qualification level.

As a result, even though the basic constructions did not finish in a "dead-heat",

each one would provide the desired service.

F-C3694 1.

INTROOVCTIOH Nine electrical cable types manufactured by The Okonite Company weze subjected to qualification tests to determine their acceptab'ility for service within the containment of a nuclear power generating station.

The environmental test program consisted of the following:

a)

Nine cable types were thermally pre-aged by The Okonite Company at 250'F (121'C) for fourteen days.

b)

The thermally aged cables plus seven non-aged cables were subjected to combined gamma radiation (for a total dose of 50 megarads) and thermal aging at 240'F for seven-days.

c)

All of the above cables were then subjected to a simultaneous exposure to steam, chemical-spray, and ganesa radiation (S/C/R)

(for an additional dose of 150 megarads) to simulate a loss-of-

.coolant accident and post-accident conditions.

The temperature profile included two temperature/pressure transients to 346'F/

113 psig and a 31-day post-LOCA simulation.

The electrical integrity of the cables was evaluated by means of insulation resistance measurements, ability to maintain electrical loading during the ther~1/radiation aging and the S/C/R exposure and by high-potential withstand tests.

.The thermal aging/radiation exposure and the steam/chemical-spray/

radiation exposure were conducted by the Franklin Institute Research Laboratories (PIRL) from November 1973 through January 1974, using the services of a subcontractor for the radiation exposure.

F-C3694 2.

IDENTIFICATION OF TEST SPECIMENS CABLE NO.*

DESCRIPTION Grou I Cables Eth lene-Pro lene (EPR)

Base Product Line 1B 7/C 812 AWG (7X) coated

copper, 0.047" Okonite plus 0.015" Okolon, cabled, no fillers, 11 mil asbestos-Mylar tape, 0.060" Okolon Jacket, pre-aged 336 hours0.00389 days <br />0.0933 hours <br />5.555556e-4 weeks <br />1.27848e-4 months <br /> at 121 C in an air oven.**

lc 2B Same as 1B except without pre-aging.

7/C 812 AWG (7X) coated

copper, 0.030" Okonite plus 0.015" Okoprene,

cabled, 6 mil asbestos - Mylar tape, 0.060" experimental thermoset jacket, pre-aged 336 hours0.00389 days <br />0.0933 hours <br />5.555556e-4 weeks <br />1.27848e-4 months <br /> at 121'C in an ai'r oven.**

2C Same as 2B except without pre-aging.

9B 1/C 812 AWG (7X) coated

copper, 0.045" Okonite pre-aged 336 hours0.00389 days <br />0.0933 hours <br />5.555556e-4 weeks <br />1.27848e-4 months <br /> at 121'C in an air oven.+*

9C Same as 9B except without pre-aging.

llB 1/C 116 AWG (7X) bare copper, Semicon tape, 0.090" Okoguard, Semicon tape, 0.003" bare copper tape, pre-aged 336 hours0.00389 days <br />0.0933 hours <br />5.555556e-4 weeks <br />1.27848e-4 months <br /> at 121'C in an air oven.**

• Throughout this report, the cables are identified as indicated in this column, except that the number and letter are reversed in the original data sheets (e.g.,

1B and Bl refer to the same cable).

• • Information on pre-aging was provided by The Okonite Company.

2-.1

I+

\\

F-C3694 3.

TEST PROGRAM The test program involving simultaneous exposures was designed to more closely simulate actual service conditions than a sequegtial-r exposure does.

The procedures were in accord with IEEE Std 3S3-1974+

in so far as practical.

3.1 PRETEST ELECTRICAL MEASUREMENTS Prior to the simultaneous radiation/thermal aging and the simultaneous steam/chemical-spray/radiation

exposure, the cables vere ih subjected to insulation resistance (IR) measurements at 500 Vdc and high potential vithstand tests at 2200 Vac.

3.2 SIMULTANEOUS RADIATION/THERMALAGING EXPOSURE Whil~ electrically energized, the cables vere exposed for 7 days to an air-equivalent dose** of 50 megarads of gamma radiation.

During this exposure, the cables were thermally aged at 240'F and ambient chamber humidity.

3.3 LOSS-OF-COOLANT ACCIDENT (LOCA) ENVIRONMENT EXPOSURE Following the simultaneous radiation/thermal aging, vhile electrical energized, the cables were simultaneously exposed to steam, chemical. spray and gaana radiation (S/C/R) as illustrated in Figure 1 (Phase II).

A chemical spray consisting of 2000 ppm boron as boric acid, buffered with

• ZEEE Std 383-1974, IEEE Standard for Tv e Test of Class IE Electric Cables Field Splices, and Connect ions for Nuclear Power Genera tin

Stations, The Institute of Elec trical and Elec tronics Engineers, Inc.

Nev York, N.Y., 1974.

• +An air eauivaEent dose means that the volume occupied by the. cables receives an isotropic flux of gamma radiation such that this radiation dose would result if the volume contained air.

3-1'

340 320 NONE 50M RAO TOTAL TIME UNDER RADIATION I'50 M RAOS 346 F/ll3psig/IOO'/~RH WITHIN 3 TOSMIN.

I I

I L

I STANDARD LOCA SPRAY EXPOSURE A

A ~ 335'F/95 psig /100'/i RH

~ 3IS~F/69 psig /IOO /, RH A

NONE LEGEND A INSULATION RESISTANCE MEASUREMENT 250

~200 W

l40 INFORMATION ON SIMULATEOI NORMAL AGING REPORTED BY THE OKONITE COMPANY I

I I

I I

I 2400 F A

A 280'F/70psig (min)

WITHIN IO SEC

~265'F/28psig/IOO'/+RHi I

I 2I2'F/Opsig/100'/i RH

~

2I2'F STEAM ONLY INFORMATION ON IOO.DAY I

POST-LOCA SIMULATION A TWO TIMES~

REPORTED BY THE PER WEEK I

OKONITE COMPANY I

I l4 DAYS SIMULATED NORMAL AGING 7 DAYS S IMULATEO RADIATION AGING iO 3

S B

ii IS SEC.

HR. HR.

HR.

HR.

HR.

4 DA'YS 27 DAYS TEMPERATURE/PRESSURE PROFILE FOR SIMULATION OF LOSS-OF-COOLANT ACCIDENT IOO DAYS POST-LOCA SIMULATION NOTE: SEE TEST RESULTS FOR ACTUAL ENVIRONMENT Figure 1.

Cable gualification Test Profile for Life, LOCA and Post-LOCA Simulation n

I CA Ol lD

F-C3694 j

NaOH to a pH of 9-11, was applied to the cables within one minute after reaching 346'F.

The required rate of spray application was 0.15 gpm per square foot of spray area.

The gamma radiation dose rate was approximately 0.2 megarads per hour to arrive at an accumulated dose of 150 megarads for the 31-day

exposure, yielding a total of 200 megarads for the test program.

During the test program, the cable IR was measured at the times indicated in Figure 1.

3.4 POST LOCA TESTS After the 31"day LOCA environmental

exposure, the cables were subjected to IR measurements and high potential withstand tests.

The cable mandrels with cables installed were returned to The Okonite

Company, which exposed the cables for an additional 100 days at 212'F with steam while under rated voltage.*

There was no radiation during this period.

At the conclusion of the additional 100-day exposure, the.

cables were subjected to IR measurements and high potential withstand tests, after which they were bent around a mandrel not greater than 40 times the overall diameter and then subjected

'to a final ac withstand test at a potential of 80 V/mil while immersed in tap water at room temperature.

• The conditions for the 100-day additional exposure were-reported by The Okonite Company.

3-3

F-C3694 TC 6 I6 IN. BELOW TOP OF FLANGE CYLIN DR IGAL STEAM BAFFLE PERFORATED PIPE FOR

'TEAM INLET I IN. ABOVE MANDREL TC 2 I IN. ABOVE MANDREL CHAMBER HEAD CHAMBER FLANGE 4oo 0

Oo OO 9 IN 0

0 TC 589 l9 '/y IN.

0 0

0 0

0 TC 387 B SPRAY NOZZLES 0

0 0

0 0

TEST CABLES SUPPORTED ON MANDRELS 0

0 0

0 0

INNER 8 OUTER MANDRELS 0

0 SOLUTION TC488 IN SOLUTION Figure 3.'ketch of Pressure Yessel Showing Salient Features and Location of Thermocouples.

4-3'

F-C3694 Ld 3 Cll a.

tO OO 0

2 r

2 AUTO-TRANSFORMERS V)

V V

V>> V>, 8 V> = 600 Vac IIO Vac 2 SINGLE "PHASE POTENTIAL TRANSFORMERS OPEN " DELTA ARRANGEMENT CONDUCTOR ENDS OF CABLES WITHIN CHAMBER AMMETER lg EVEN-NUMBERED CONDUCTORS 3

7/C CABLE NO. I Ip ODD-NUMBERED CONDUCTORS AUTO-TRANSFORMERS TRANSFORMERS FOR CURRENT LOAD 600 Vac 600 Vac 6

o 2

3 7/C CABLE NO. 2 AND SO ON 600 Vac CROSS SECTION OF 7/C CABLE SHOWING CONDUCTOR NUMBERS AND POTENTIAL APPLICATION NOTE: ALL CONNECTIONS MADE OUTSIDE OF SPRAY CHAMBER Figure 5.

Diagram of Typical Energizing Circuit for Multiconductor Cables.

4-7

0

~

I

~

I

~

~

I I

~

~

~

~

~,

~

I

~

~

~

~

I

~

~

'e I

~ I

~

~ I I

~ ~

~

II

~

~

~

~

~

~

~

~ ~

~

I I

~ ~

~ t ~

~ ~

~

~

~

t ~

~ ~

~

~

~

~

~

~

~

~

~

~

~ ~

I I

I

~

~

~

I

~

I 0 ~

~

~

~

I 1

~ ~

~

~

~

~

~

~

i

~ ~

~

~

~ ~

~

~

4

~

~

~

o l

~ ~

~

~ I I I

~ ~

I I

~

~

~

~

~

~

~

~

~

~

~

~

R

F-C3694 6.

CONCLUSION Sixteen samples of nine types of electrical cables manufactured by The Okonite Company were exposed simultaneously to thermal aging and gamma radiation, followed by simultaneous exposure to steam, chemical spray and gamma radiation in accordance with a program designed to simulate normal service, a loss-of-coolant accident

{LOCA) and a 30-day cooldown following the LOCA.

The cumulative dose of gamma radiation was 2 x 10 rads.

Throughout the exposures, the cables were energized 8

(except cables which failed during the program) with potentials and currents simulating field service use.

At the conclusion of the above sequence of exposures, the cables were subjected to a <<igh potential withstand test.

Suaanary statements on the performance of these cables are given below.

~

Grou I Cables:

The EPR base product line demonstrated satis-factory electrical performance during the exposure simulating conditions of normal service, LOCA and 31-day cooldown following LOCA, and withstood high potential tests conducted at the end of the 31-day cooldown.

The experimental outer jackets on two of the Group I cables (2B and 2C) underwent severe physical degradation, but the single conductors survived.

P portion of'the Group I cables withstood additional high potential tests after an additional 100-day exposure to steam (only) at 212'F.*

The surviving cables were 9B, llB, the 7 conductors of cable 1C, 4 counductors of cable 2B, and 1

conductor of cable 2C.

The jackets of cables 1C, 2B and 2C were removed prior to the final electrical tests so that the individual conductors could be tested as single conductors.

• The environmental conditions were reported by The Okonite Company.

FIRL witnessed the final band and high potential withstand tests conducted in accordance with Section 2.4.4 of IEEE Std 383-1974.

(See footnote on page 3-1.)

6-1

F-C3694 7.

CERT IF1CAT 1 ON The undersigned certify that this report constitutes a true account of the test conducted and results obtained.

D. V. Paulson Project Leader

-L. E. Witcher Test Engineer APPROVED BY:

enons

Zudans, Director Engineering Department S.

P. Carfagn, ager Performance Qualifica tion Labora tory 7-1

,t ebruary 4e, L $ 75 Engine e ring Report No. ? 64 THE OKONITE COMPANY

Ramsey, New Jersey

SUBJECT:

Qualification of Cables and Splices for Nuclear Plan s Through Designed Life Simulation Testing.

OBJECT:

To present summary data with back-up references that support this certification of suitability for The Okonite Company's ethylene-propylene base insulated cable line, and compliance with IEEE Std. Nos.

323 and 383-1974.

CONTENTS:

Gene ral Dis cus sion and IEEE T 74-044-4 Certified Water Immersion and Aging Test Data Sequential Exposures to Heat, Radiation, LOCA Simultaneous Exposures and Life Simulations Vertical Tray Flame Te s ts Section 1

Section 2 Section 3 Section 4 Section 5

CONC LUSION:

The simulated service data and references presented herein provide substantial evidence that the ethylene-propylene insulated cables listed below are suitable for service in nuclear plants with a designed life objective of forty years, and do exceed the re-quirements of IEEE Std. No. 323 and 383, 1974.

1/C 2000V Okonite insulation 1/C 1000V Okonite/Okoprene or Okolon 1/C 5-8 kV Okoguard Okolon non-shielded 1/C 5 kV Okoguard shielded M/C 1000V Okonite/Okoprene, Okop ene M/C 1000V Okonite/Okolon, Okolon 1/C T95 and T35 hand-wrapped tape splice c.r. "

I EEM/row E. E. Mcllveen Vice President-Engineering Sworn and Su'oscribed to before me Notary Puolic of New Jersey My Commission Expire s

THE OKONITE COMPANY

Ramsey, New Jersey

.ENGPfEERING NOTE A/74-I Section:3 Date:

Feb.

8, 1974 Revised - February 24, 1975

SUBJECT:

CABLES FOR NUCLEAR POWER GENERATING STATIONS At the IEEE Winter Power Meeting in New York on January 28, 1974,. we presented a paper T74 044-4 on the above subject.

Our final copy is at-tached for your information.

Our verbal presentation is repeated below for your information.

INTRODUCTION:

While this paper is concerned with Cables for iNuclear Power Generating Stations, it applies equally well to cables for any elec-tric power facility except, of course, the need for resistance to radiation would be superfulous in non-nuclear applications.

Interestingly enough

though, even the loss of coolant accident simulation, better known as LOCA, happens to emerge in the laboratory as a steam autoclave exposure which had been used for many years to simulate long term exposure to water or mois ture.

With the advent of nuclear power plants at Shippingsport, Indian Point and Peach Bottom to mention a few, a whole new set of operating conditions had to be recognized.

As demonstrated by the work in the ICC and NPEC, the evolution of design criteria has continued.

The irraninent publication of the IEEE Guide F383 is evidence of only the present state of design.

In this paper we examine typical'design and qualification data together with some explanation of its significance.

Designed life performance, when based on long service can be quite reliable, but significantly different con-ditions and materials have come into use which made simple projections questionable.

Simulated service test data must, therefore, be relied upon.

The areas to be studied for qualification through performance testing are:

Moisture Resistance Thermal Aging Radiation and LOCA Cable/Tray Flame Resis tance

. MOISTURE RESISTANCE It has become traditional to gain assurance of long life performance by to-tally immersing a 912 or 14 conductor insulated witn a 45 mil wall of dielectric in water at an elevated temperature to accelerate the deterioat-ing effects of moisture, as shown in Fig. 1 on Okonite referred to herein as "EP".

In the 1950-57 era with service experience that suggested negative dc potential presented the most severe condition, IPCEA developed a 16 week test procedure based on a continuous immersion at 50 C while 'under 600 volt dc.

Now more than sixteen years later, new generation insulations can be

ENGPTEERING NOTE f74-.1 Page 2

continuously immersed at 75 or 90 C while under the same dc potential and survive 78 weeks or more.

This is at least 5 times at an effective accelera-tion rate of 6 times greater than anticipated by that IPCEA procedure.

The composite insulation wall composed of an EP dielectric, Okonite plus a flame resistant thermoset cover, OKOPRENE, still only totaling 45 mils,

'lso exhibits durability as shown in Fig. 2.

A comparison of the EP with a FR CLPE insulation at 90 C may be noted in Table I, and the accelerating effect of temperature on the, moisture absorption in Fig. 3.

THERMAL AGOG Data on electrical behavior in a combined heat and moisture environment may be a better guide to service performance than that developed from aging in a dry air oven at elevated temperatures.

However, such data as shown in Fig. 4 thru 7 can be analyzed by the Ar henius technique and, by analogy the useful life may be predicted.

In applying this Arrhenius analysis to the aging data, the time to 40~o loss of elongation is plotted on semi-log paper against the reciprocal of the absolute temperature in degrees Kelvin.

This is presented in Fig. 8.

In examining the validity of this treatment, there are at least four simultaneous reactions.

From this', it could be argued that the occurrence of a linear plot is a fortu>>

itous event, but it does provide a means of comparison within the temperature range of the data.

The significance of loss in elongation is related to the ability of the insulation to withstand bending without physical cracking and ultimate electrical failure when moisture enters.

A 40~o loss still 'eaves 60~y retention wheras 50'/o ultimate is usually reached before cracking de relops upon flexure.

Since thermosetting insulations of the 1957 type vintage have performed well in early nuclear plants and the ethylene-propylene base and the cross-linked polyethylene insulations take 6 times as long to reach,40~>'o loss of elongation as does butyl, it is safe to predict that new insulations such as Okonite and Okoguard willprovide a superior tnermal aging performance.

RADIATION AND LOCA In 1968 the first of many increasingly searching studies were made of the effect of gamma rad ation on the electrical and physical properties of 13 different insulations.

By the criteria then existant tne ethylene-propylene and the cross-linked polyethylene dielectrics were found to be'suitable.

ENGINEERING NOTE f74-1 Page 3

By the mid-1970 qualification testing of specimens pre-aged prior to gamma radiation to 3. 5 x 10 rads followed by a LOCA became necessary for design acceptance.

This work was done at FIRL.

Again specific ethylene-propylene insulations, i. e. Okonite and Okoguard, proved to be suitable, the amount of aging having no adverse effect on the properties of this insulation..

The Hypalon compound in Table IIhad good electrical and physical properties following the simulated aging, that is, "before" radiation.

"After" the LOCA simulation in the autoclave the electrical properties fell off significantly.

The physical appearance and resilience,

however, was good.

Two specimens which contained a hand-wrapped splice made'ith an ethylene-propylene base tape remained resilient and firm at all stages of the test.

Within a year, July 1971, reassessment of design parameters dictated a higher level of radiation exposure and new BAVR LOCA, simulation profile for qualificat:on.

Again'FIRL was called in.

The details may be found in Tables II, III, and IV.

Fig.

9 presents double LOCA simulation for specimens radiated to a total dosage of 2 x 10 rads.

This simulation is particularly severe because these specimens were loaded to rated current and voltage and tested to the pres-surized water reactor (FWR) incident profile, and then these same specimens were subjected to a BWR profile.at FIRL. lt may be noted in Fig.

9 these profiles include one "peak" at 324 F/80 psig, two peaks at about 342 F/104 psig, and one at 320 F/75 psig plus the 100 day soak.

These multiple tran-sients demonstrate that these Okonite and Okoguard specimens have significant

margin, a performance characteristic requested in the Guide.

Additional qualification tests covering several different cable constructions in simultaneous radiation/life tests have now been completed.

The Appendix to the IEEE Std. No. 383-74 suggests relative short term procedures which have evolved from the type of studies and testing described herein.

New data obtained in Okonite's latest test at FIRL in environments which more closely approach long term operational conditions indicates that sequential effects of moisture, thermal aging, radiation, LOCA and movement are no no less severe than simultaneous occurrences.

In other words.,

the simul-taneous simulation does not result in a synergistic effect.

The cable specimens were removed from the radiation chamber on January 28, 1974 after receiving 2 x 108 rads of gamma radiation and FIRL Report F-C 3694 covers this, CABLE TR AY F LAME TESTS The losses incurred througn cable fires has focused attention on not only the flame resistance in the trays, both with and without covers or baffles, but also the degree of separation, the matter of circuit redundancy, and the effectiveness of various fire extinguishing systems.

Studies on, cable behav-

I

'ENGINEERING NOTE 974-'I Page 4 ior, and the effect of physical arrangement and separation have been covered in earlier papers.

Redundancy and fire extinguishing systems are subjects not relevant to type testing to qualify cables and connections other than to note that the redundancy concept among other things, has made obsolete the requirement of "time to short-circuit".

l

'Cable performance during a tray fire should be postulated as a design basis

event, and a method of qualifying cable tray systems had to be developed.

The "Philadelphia" Tray Cable Fire Propagation Test, devised in 1965, simulates a fire in a vertical. tray, a situation which is more sever'e than in a horizontal tray.

The Qame source in the original procedure was crumpled oil-soaked burlap rag.

In the following years much effort was expended in trying to eliminate some uncontrollable variables.

AVhile a definite burlap folding sequence and oil dipping procedure were developed, the replacement of the oil-soaked burlap with a large ribbon type gas burner together with in-strumented control of the air/gas Qow resulted in a completely reproducible flame environment.

This inadvertently resulted in a more severe test be-cause (a) the ga's flame provides a constant heat input for the entire test peri'od whereas the oil-soaked burlap fame reaches a maximum within several minutes and then tapers off until it finally Qickers out.

Furthermore,

~

(b) the gas Qame projects right through the cables thereby completely sur-rounding them whereas the oil-soaked burlap flame is vertical and in front of the cables so that the majority of the heat reaches about half the circumfer-ence.

Table V provides the comparative data, as do Tables VI, VIIand VIII.

It was agreed by the committees that the only practical clear cut observation than can be made is that of propagation.

Failure by this mode is defined as occurring when the fire burns all the way. up to the top of the tray, a distanc of about 6 feet above the Qame source center.

SUMMARY

MOISTURE:

In the 1950-57 era with service experience that suggested negative dc potential presented the most severe condition, IPCEA developed a 16 week moisture test procedure based on a continuous water immersion at 50 C while under 600 volts dc.

Now more than sixteen years later, new 0

0 generation insulations can be continuously immersed at 75 or 90 C while under the same dc potential and survive 78 weeks or more.

This is at least 5 times longer and at an effective temperature acceleration rate of 6 times greater than anticipated by that IPCEA procedure.

AIR OVEN AGING:

Since thermosetting insulations of the 1957 vintage have performed well in early nuclear plants and the ethylene-propylene base and the cross-linked polyethylene insulations take 6-8 times as long to reach 40<~

loss of elongation as does butyl one of the bette" dielectrics which has seen at least ten years service, it is safe to predict that these new insulations will

l 0

l

~

ENGINEERRlG NOTE f74-1 Page 5

provide a superior thermal aging performance and should achieve the designed life of the plant.

RADIATION AND LOCA:

In 1968 the first of many increasingly searching studies were made of the effect of gamma radiation on the electrical and physical properties of 13 different insulations.

By the mid-1970 qualification testing of specimens pre-aged prior to gamma

~

radiation to 3. 5 x 10 rads followed by a LOCA became necessary for design acceptance.

This work was done at FIRL.

%'ithin a year, July 1971, reassessment of design parameters dictated a higher level of radiation exposure, namely 10 rads and new BWR LOCA simulation profile for qualification.

Again FIRL was called in.

The details may, of

course, be found in the paper.

p In September

1971, a double LOCA simulation for specimens radiated to a total dosage of 2 x 10 rads was completed.

This simulation is particularly severe because these specimens were loaded to rated current and voltage and tested to the PWR incident profile, and then these same specimens were subjected to a BWR profile by FIRL.

Now a fifth series of cables have been subjected to 2 x 10

rads, but simultaneously with life aging and LOCA.

This work is also covered by an FIRL report.

CABLE TRAY Fl 4hL~ TESTS:

Redundancy has made obsolete the require-ment of "time to short-circuit'.

This is fortunate indeed because it is not technically feasible to establish a singLe time requirement for all types and sizes even if someone could identify what period is operationally necessary.

It was agreed the only practical clear cut observation that can be made is that of propagation.

Failure by this mode is defined as occurring when the fire burns all the way up to the top of th tray, a distance of about 6 feet above the flame source center if the Standard P383 procedure is followed.

CONC LUSIONS

{1)

Based on the type of data presented in this paper, it can be logically shown that Okonite/Okoprene or Okolon and Okoguard insulated cables willsurvive designed life events in areas of (a) moisture and steam, (b) heat, (c) radiation and LOCA, and (d) fire all with a comfortable margin.

s P ENGINEERING NOTE f74-1 Page 6

There are constructions whose performance excels in several of these four environments, but only Okonite will do well in all of the situations cited herein.

(3)

In view'f the time element, long term water immersion data and Arrhenius charts must be handled by certified test reports.

The Okonite Company is prepared to do this on our premium station cable designs

~

NOTE:

All samples used in the tests described in this paper were manufactured by The Okonite Company ATTACHMENT:

T 74 044-4 s5 BY:

E. p~ iveen, V. L. Garrison and G.

T. Dobrowolski

~ ~

LONG TERM 90 C Vt'ATER IMh~&RSION TEST Satttole:

1/C f/12 AWG (7X) Coated Copper,

. 030" Okooite,

. 013" Okoloa 600V DC-Continuous 600V AC-Continuous Time Period Stress V/mil PF SIC

'10 Q 00oC SIR x 103 SIC SIR 190 C

x 103 1 day 7 days 14 days 1 month 2 months 4 months 6 months 40 80

,40 80 40 80 40 80 40 80 40'0 40 80

3. 14

3. 19 1,91

l. 98

1. 59

1. 63

l. 40 I. 45

l. 38

1. 42

l. 28

l. 32

l. 24

1. 31

3. 30

3. 30

3. 30 3 30

3. 34

3. 34

3. 42

3. 42

3. 53

3. 53

3. 65

3. 65

3. 73 3.73 0

1.3 1.6

l. 9 1.6

l. 6 2.2

3. 16

3. 18

1. 88

l. 91

l. 57

1. 60

1. 26

l. 30 1.'0

. 1.43

1. 27

1. 30

1. 20 1.,25,

'.32

3. 32

3. 34

3. 34

3. 38

3. 38

3. 46

.. 3.46

3. 56.

3. 56 3.70

=

3. 70

3. 81

3. 81 0.9 1.4 1.4 1.3 1.3 1.4 Increase in SIC Increase in SIC Stability Factor, Stability Factor, 1 - 14 days 7 - 14 days 14 days 6 months

l. 2%

1 ~ 2%

0. 04

0. 07

l. 8%

1. 2%

~ 0. 03 0,05 Sworn and Subscribed to before me m7."

this day of 7.: z...;.

11/26/74 Notary Public of New Zersey Ss

~

My Commission Expires

-".-~ ~ ~

ts

DISCUSSION OF THERMAL AGING/LIFE SIMULATIONS Arrhenius Plots:

The procedures for developing an Arrhenius plot may be found in a recent IEEE paper T 74 044-4 entitled "Class IE Cables for Nuclear Power Generating Stations" by E. E. McIlveen, V. L. Garrison and G. T. Dobrowolski, pages 3 and 4, as well as a discussion of the fallacies and limitations of such a plot.

Th

~

• 1 i.

• ii h

, as previously published, but the interpretation supporting a 40-year life based on air oven aging has been added to insure proper applica-tion.

For example, it may be noted that the time spread between our Okonex butyl base insulation and the Okoguard and the Okonite now being offered involves a factor of about 6. 7 times.

Recognizing that Okonex insulation has been widely used since its introduction in 1946, and that it has been in Peach Bottom Nuclear Plant No.

1 since 1967, a period of seven years, it is logical that the designed life willexceed 40 years (7x 6. 7

= 46 years).

.It might also be noted that in an oven aging test, the entire wall thick-ness is at the same temperature whereas in actual practice there is a gradient from the maximum temperature at the conductor down to the outside ambient.

Actual service is less severe than this simulation.

Accelerated Aging:

Since the publication of IEZE Std. 383-1974, it0 has been stated by various engineers that aging for 168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br /> at 150 C

is significant and infer that it projects to a 40-year life.

We have con-sistently objected to this method of forecasting.

We have used a pre-irradiation aging of 2 weeks at 121 C prior to subjecting the completed cables to another week at 121 C which included 5x107 rads of gamma radiation.

This more closely simulates service conditions.

It is of interest that IEEE Std. 383-1974, Section 2. 3. 1 states that a cable that has been manufactured and tested and passed the provisions of one or more of 8 listed industry standards, qualifies for normal life-time operation.

These same standards reference aging at only 121 C

0 for 1 week, UL included.

Another reason for pre-irradiation aging at 121 C was to chose a tem-perature which mould not unduly penalize the outer Neoprene and Hypalon coverings which do not age as well as the higher temperature Okonite FP base insulations.

Incidentally, these outer coverings are designed to provide flame resistance to the EP insulated single conductors.

0 QYVLtU X 10 0>>Vlt)IL>>NG Isa>>% INCH

~ ( IIIlIlIer V )Il'O

>>s>>I 4

oieo

~

~

p rserr

)hh>>

e e'r Q

Q

)

)I ad DO

~

~

9 l P 4 MDOI>>1

~e>>

u Q

1 ODMDOO u

a aors

~ OO

~

)3 4 ~ I

)! ~ ~

)I ~

s l ~"

!)i

~

~

~ ~ ~ ~

~ ~ I

))).

f el

) II,

~ ~

I:I I

~

l

~

l I

.,9

<<1>>

l

.I

.I, ).'I

, ~ (l I(s ~

~ e I,Ii lj',I I la(

")I

f. >>I I):I II e

5:j

~ ~ l ~

~ ~

~

~ l >> ~

)ijl 0!

I' I I )

I II II'I

~l

~ II 0,

1

]

"l

',J7 I

I e

Is

l~.

~

~

~ ~ ~

~ ~

~ ~

~ sl ~

) )

~ ~

'I'I,er" j)i

~

1 I all>>ll;

'ill I(. I l ~

~

~ )I>>II

)

s

~

li

~ ~ Is

~ ~

~

( l ~

~

s ~

llj:

III.

(1,'ill

.':. I e

~ ~ e

~ e ~

'.Ili

(/ee

~ ~ ~ ~

ll'!

I'I:

'll!

~ I

~

is)

jl)

~ '

e s ~ ~

e"~

Nl

~

s e

I jl:

I".

~'

e e"

el"

~ e

~ >> ~

e ~

eejl e

'I s

~ e)

~ s lls

~ l

~

I la l,.

~

~

s,

"~

'I'I ":

"I'.

e e

o e

~

e

~

~'

e ~ a

I:.

"I'a

'I" e>>)'

I):

a>>i js

'I'el's" I'

s

)

~

~ ~ l

>> I.

ll:.

eel e

~

e ~

I:

,I

)>ld h't o

'I-1 (hir I'>>

I'o e

lh o'I o Nl I,I s

~ ~

r;. IIl fl.II

(

) I').. I..

I lel

.I.

'I;

~ s e

~ ~

~

'I I

~ ~

( ).s W 'J

,.f, b>>:.

L.

~ e

~ I I

I s

I,I, l ~

e'a.o I.'~

'l'"

~

~

o

~ )

' I

~ ~ ~ I I

t>>'.I e'II;

'.I olla

'Is",I' I) '.

o'I',

rh>>' '

~

I.

I

)

I'I)

Il III g x I

)I ill

)ll'Ils ala

!i:

s ~ s s)

~

/44 Ii I

II, II>>!(: )

)

~ )I IIII>> "

I

~.

IIIIII j')li ('!

l)l ll y

(I

~

le

) I js es

~

~

~ e I

~,

. hei e

~ s o.l

.I, ee l

>>l. ~

~ s

~ e e ~

s' "I

~p e

e.

I e

'e e

lo

'I es

~ I

>>I s '.I)

'I I

e a) e

~

l

~ s ell

. )l.

~ ~

' I s

l.

e s'>>>>

>> e P>>

~

t)g s r) vq '

18:

.!A

~ ~

)E::l ~

'..l'i

I. I'l.

~ ~ '

III

)h I

'ol e

e I'

I

. I,Li' I; -,I I

I I

I I

I I

II I;

II I:

I I

I I

~ I e

I ll I'.

I; ll I ~ I

I e

e, s'

l>>I

I

,)I' I)

,jl:,)

Ij/j ice

~ ~ e e

e e

'I

'I.

~ '

I I

~

e e

e I';

J lie

I!

~

~

.I I.

o

'>>'I

>>;si es>>

"ll

':I:

1" e),

I) il:

l's

~

e)'I

)NI" I le

'i/:

e;l J

li s

>>e I

l ~.

l>>o l'lj

~

s,

~

e e ~

le Ill '

"I

~ ~ ))

~ e ~

I s.l, I..

sla e ~I le "

~ll~

,Isj II)l :::

~

o

~

~

all I

,II'l e

I ~

l ~ ~

I )

ll Iles I I ".

)I(

'.II

'I! '.li i/I.'I);.

er l.e

';I',,'

~:s

'I"

'e"

~ e

~ ~ Io

~

e lo,

,I

~

sl e ~

il sl

)I l ~

')o I

~ e

~ ~

,s

~ I cQ sl" II.

'I'l';

l I

. hei

.il:

~ ~

~ '

~I.

e ~ e l.

~ ~ I~

~ e ssl I)lj I!!i I I I ))

III'.

,':I:

jl:i

~ ~ ~

I'".

'[i.'

~ ~

o!q I;I s ~

I s

~

~ ~

i sl I ".

flml I'.

IN..

)-

~

~

~

sl

~e'

~

~ e ~ e le>>I'

~

~

~ ~

ee

)

s

~ ~

e e ~

~ )s o

I";I,

'.I.':

)I el ~

"I

>>)'el

~ ss

~ I

)Is'>>'.s

))>>o e ~

lli

.I:)

e lr

~

l~

~ e ls I

'I s'

e

Il

'Il I)I s ~

sl ~

~ ~

'll

~ )sI';I' e ~

II I

s e ~

~ ~ ~

~

~ ~ l.l e

~ ~ I

~ ~

l)I'.

I ~ ~ ~

I:.'.".'

I

.':; f.

4(

s e

l..

'e

~ e>>

~

~ e>>

~

I" Ila.

le ~

e

,e

'I::":

!li

~

S

.)I:"

jlis I,,'

',s..l.;

e

~Il e ~

~ s ~

'.(I'>>a II; le>>

~

s I I

'w

~ ~

'e);

~ )

>>l

';!I'.!ji)

'e ~ o e

bete s

I'-

e ~

s

~

~ e

~

'll' le

~ Iso

~ el

'e ae

~,,

}'>>

>>~

A~

0

~

~

L

~ ~

~ l I

. I.

~

~

l I )I

~ ~ ~

~ll I ~

~ ~

),>>

IIjl I',

I'I I)

I

~ (

!Ill

'l".i!~'t"

~ I:

~ ~ sl I

IsI:

e 1:

ll'

~ ~

1'Cl j

l

)

e

II; i)l

ace

.,II

))s ~

)

~ ~ ~

Il!I

~)ili I:II

'I I

I I'l s ~ I e>>

'I' s ~

la) all l.s

~ osl e

~ I s

~

e ale s ~ l hC

10. 0 REPORT

SUMMARY

demonstrate a qualified life in excess of 50 years.

Radiation " All materials used will readily withstand 1.5 X 10 Rads gamma as all tests were carried out at values exceeding 2 X 10

Rads, 8

LOCA " The required LOCA profile having a maximum of 265'F was enveloped by a 340'F temperature profile in the qualification testing performed, Ample margin was demonstrated.

Qualified Life - A qualified life in excess of 50 years from start of plant operation has been demonstrated by test and analysis.