ML20148J838

Provides Addl Info to Support Environ Qualification Per 10CFR50.49 of Bunker-Ramo,per 880322 Telcon.Attachment I Identifies Bunker Ramo Instrumentation Penetration Info.Addl Info Will Be Transmitted as Described

ML20148J838
Person / Time
Site:	Braidwood
Issue date:	03/23/1988
From:	Hunsader S COMMONWEALTH EDISON CO.
To:	Murley T NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM), Office of Nuclear Reactor Regulation
References
NUDOCS 8803300405
Download: ML20148J838 (40)

Text

. - . -

/~N Commonwealth Edison

• 9 ,,I One Fir:t Ndonal Plaza, Chicago, Ilkno:s kN O ~] Address Reply to: Post Offics Box 767

/ Chicago, Illinois 60690 - 0767 March 23, 1968 Mr. T. E. Murley Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Cumnis91on Washington, DC. 20555 Attn: Document Control Desk

Subject:

Braidwood Unit 2 Environmental Qualification Bunker Ramo penetration NRC Docket No. 50-457

Dear Mr. Murley:

The purpose of this letter is to provide the NRC staff with additional documentation to provide support for the environmental qualification, under 10CFR 50.49, of a Bunker Ramo penetration used at Braidwood Station Unit 2. This piece of equipment is a Bunker-Ramo manufactured instrument penetration used to provide access through the Unit 2 containment wall in four (4) locations for circuits that carry electrical signals from instrumentation inside the containment to main control room indicators and protective circuitry. This penetration provides this function while maintaining the integrity of the containment pressure boundary. This penetration is identified at the four (4) locations as 2SIo5E, 068, 078 and 088. Though substantial substative documentation exists to provide support for environmental qualification, additional documentation has been determined to be necessary by the NRC staff to make the documented basis for environmental qualification fully auditable.

Exhibit I shows the location of the penetration as installed in the containment wall.

As agreed upon during our telecon of 03-22-88, Attachment I identifies the preliminary Bunker Ramo Instrumentation penetration information being submitted to you for your review and acceptance.

Additional supportive information as described in Attachment I will be transmitted to you on Monday, March 28, 1988.

Please address any questions concerning this matter to this office.

Very truly yours,

' 8803300405 080323 DR ADOCK 0500 7 S. C. Hunsader i Nuclear Licensing Administrator l /klj cc: NRC Region III k Braidwood Resident '

l l S. Sands  ;

! Braidwood Unit 2 Penetration Location Exhibit I Relative ~to Chemical Spray Header .

Containment

[ nn Q Spray header n n n - ---

i g g n elevation-567 f t.

ee i /

I

/) tEt Support plate > - ..

4

1. 16AWG penetration pigtall (BlW) Penetrat n e g Raychem WCSF-N splice r ----

439 f t/418 f t.

Field catie > "

! Module

! Cable trayd

--> 4-Wall thickness 3 f t 6 in.

I C1595.040/M 03 88 J

'o '

4 s _ . _ _ __ _ _ _ _ _ _ _

. March 23, 1988 ATTACHMENT 1 Bunker Ramo Instrumentation Penetration Preliminary Information Submittal

1. Appendix A - Braidwood Unit 2 Environmental Qualification Evaluation for the Bunker Ramo Instrumentation Penetration Assemblies

• 2. Appendix B - Supporting Qualification Data for the Environ-mental Qualification of Braidwood Unit 2 Bunker Ramo Instrumentation Penetration Assemblies

3. Supplement to Appendix B to Answer NRC Questions of March 16, 1988

4. Midland II Test Curve of a Bunker Ramo Instrumentation Penetrations l

• • 5. Bunker Ramo Design Qualification Test Report 123-2201, l Rev. A, dated February 1979.

l

• Handed out at the March 16, 1988, NRC/ Ceco Washington meeting

• • Air mailed to the NRC (Washington) on 03-22-88

Appendix A BRAIDWOOD UNIT 2 ENVIRONMENTAL QUALIFICATION EVALUATION FOR THE BUNKER RAMO INSTRUMENTATION PENETRATION ASSEK3 LIES

1. PURPOSE The purpose of this evaluation is to demonstrate the accept-ability of the Bunker Ramo Environmental Qualification of the Braidwood Unit 2 instrumentation penetration assem' ies.

The Bunker Ramo Tect Report identified an anomaly regatding insulation resistance values. The anomaly has no affect on the pressure-retaining capability of the penetration ass,?m-blies. As explained below, the anomaly is only applicable to the test situation and not to the installed plant configur-ation.

2. IDENTIFICATION OF ANOMALY The Bunker Ramo Test Procedure (Reference A) indicates that two low voltage prototype penetration assenblies were tested.

! Table V of the Test Report (see Exhibit 1) summarizes the l insulation resistance values recorded for selected circuits l

l in these prototype assemblies, during and after the LOCA test.

l Exhibit 1 reveals that some of the insulation resistance val-i ues for the selected penetration circuits are low. The se-l t

lected circuits listed in Table V did not include insulation resistance measurements for a (16AWG penetration module pig-l tail assembly which is similar to that installed at the Braid-wood Unit 2 penetrations in question. However, the low insula-l

Appendix A tion resistance values recorded for the circuits listed in Table V (utilizing a similar module design) prompted the ques-tions by the NRC regarding the integrity of the penetration module.

Generally, the low insulation resistance values for the cir-cuits in Table V have been identified in the Test Reports as anomalies. These anomalies have been attributed /disposi-tioned as either the result of "shorting during the LOCA" or the occurrence of "a service interruption for periodic testing which may have resulted in an unusual voltage stress, i.e., I.R. was high (2.7 x 10 7 ) just prior to anomaly and 100 Ohms immediately after." Furthermore, the Test Reports stato that (a) "these circuits met the continuity and gas leak rate requirements", (b) "all insulating materials reflect the impact of the specified environment (c) "no significant deterioration occurred in the Amphenol module or seal mate-rial", and (d) the test results conservatively scope Class lE safety related requirements". (References A & B).

3. TEST SPECIMEN CONFIGURATION The Bunker Ramo Test Procedure (Reference A) states that the tested Low Voltage Penetration assemblies shall have addition-al junction box internals installed on the outboard and in-board side to qualify these items under LOCA/DBE condit. ions.

Appendix A These internals connect the penetration circuits to terminal blocks and connectors in addition to the hardware utilized to connect the required instrumentation for monitoring the insulation resistance during the LOCA test.

Exhibit 2 provides an illustration of the tested configuration and construction of a typical Bunker Ramo instrumentation penetration module and pigtail assembly utilizing terminal blocks. It should be concluded from this illustration and the anomaly discussion in Section 2 above, that the root cause of the low insulation resistance values experienced during the LOCA test can be attributed (a) to the shorting of the penetration pigtails at the terminal block connections and at the connectors within the junction boxes and (b) to the service interruption that occurred during the LOCA test.

The low insulation resistance values were not caused by a failure within the penetration module itself.

The above conclusion can be further substantiated by examining the installation and construction attributes of a typical Bunker Ramo Instrumentation Penetration utilizing the "post-crimp" module design (See Exhibits 2 & 3) . It can be seen from these exhibits that the module feedthrough conductors are insulated from one another and from the header plate via a glass reinf orced epc xy. The pigtail conductors are crimped directly onto the module conductors and again insulated with the glass reinforced epoxy.

A leak free assembly is achieved

- Appendix A by mounting seals on both ends of each module. Furthermore, the area between the seals is pressurized with dry nitorgen at all times (during shipment, storage, and operation) to assure a dry atmosphere and thus maintain the integrity of the module insulation resistance (i.e. the module feedthrough conductors between the seals are isolated from the LOCA envi-ronment). The penetration pigtail conductors were manufac-tured by Boston Insulated Wire (BIW). These cables have been independently qualified by tests and have exhibited negligible insulation resistance degradation. Therefore, we do not be-lieve that the low insulation resistance values should be attributed to the penetration module and pigtail assembly.

As a matter of information, we have reassessed IE Bulletin 82-04 for identified deficiencies that could be pertinent to these penetrations. All of the reported deficiencies have either been corrected or were determined as not applicable to the installed Braidwood Unit 2 penetration assemblies.

4. INSTALLED PENETRATION CONFIGURATION There is one major and distinct difference between the instru-mentation penetration assemblies tested configuration and installed configuration. As illustrated in Exhibit 2, the initalled termination method utilized within the containment at Braidwood Unit 2 consists of in-line butt splices. These splices are insulated with Raychem WCSF-N heat shrinkable

. Appendix A tubing, in place of the terminal blocks or connectors used in the test. Braidwood's Environmental Qualification (EQ)

Binder EQ-BB-120, documents that Raychem splices have been independently tested in the LOCA environment and hava exhibit-ed negligible insulation resistance degradation. As a result, we believe the insulation resistance values recorded in Table V of the Bunker Ramo 'fest Report are not applicable to the Braidwood installation of these electrical penetrations.

All other te5L results in the Bunker Ramo test report are acceptable and representative of the installed penetration configuration.

In view of the above, we had utilized the insulation resis-tance values it. the BIW and Raychem splice LOCA tests for our instrument loop accuracy calculations (rather than the insulation resistance values in the Bunker Ramo test report) since they are representative of the Braidwood installation.

These insulation resistance values provide the required instru-mentation accuracy with substantial margin. We believe that this qualification approach meets the guideline and intent of NUREG-0588 requirements.

s

' App 8ndix A

- . 5. CONCLUSION Dased on the above review and analysis of the qualification test information, Commonwealth Edison believes that the Bunker Ramo penetration assemblies, as installed, are qualified and meet 10CFR50.49.

r

_-.. - ,.,- =_ - ..-... . , - , _ . . , - . . . . , _ - . , . - . . . _ _ . . . . _ _ - . . . . . , - . - _ - _ - . - _ , - . . . . . - - - . - - _ . - _ - -----

Appendix A

6. REFERENCES A. Bunker Ramo Generic I Qualification Test Procedure 123-2159, Rev. 5A, dated 06/01/79 B. Bunker Ramo Design Qualification Test Report 123-2220, Rev.

4, dated 10/10/79 C. Dunker Ramo Loss of Coolant Accident (LOCA) Test Report 123-2159-18, Revision 1, dated 06/18/79

7. EXHIBITS Exhibit 1 - Table V from Bunker Ramo Test Report (Reference A) l Exhibit 2 - Tested / Installed Penetration Pigtail / Module Con-figuration Exhibit 3 - Top Asacably Drawing - Instrumentation Penetration Assembly l

l l

LXI;7JIT 1.

7 Av.t r_ *1 TS:

_ LOSS OF COJ_LA._NT AC__C_I

-( D_

L_O_CA.-)

_E_N_T_RI:5UL ----(OllM. S).-

leth 22nd 20th 30th 5th 7th Final Classi- 1st 2nd 4th Oy IM D_ ay, 13 Initial R.wp se a ,,p nyy 3 3 ty Circuat f i c a t a v_n 5

3 2.5x10 1.3x10" 1.5x10* 1.5x10" 1.8x10 2000 1.3x10 1.0x10 In t. 1.sx10 3.7x10' 8.5x10 Traax 1.2x10 2.0>10 4

1.0x10* 1.2x10 4.0x10" Inst. 3.6x10 3.4x10 1.9x10 220 1.2x10' 65 Co.a x , Tuc 4 #

8 4.'Ox10 1.4'410 3.4x10 3.2x10 3.0x10 LVP 1.2x10 3.4x10 6

1.4x10 2.8x10' l.5x10 8.0x10 3x350 #

2.1x10 Bon 100 3.bx10 2.8x10 7.0x10' 2.1x10 1.4x10 LVC 4.2x10 4.3x10 1.1x10 5 09 814 (T.D) 1.6x10" II 7 1200 1.0x10' 3.3x10* 4.5x10' 3.4x10" 1.9x10 a.0x10 2.4x10 1.2x10 600 69s14(PrP/Tu) 1.V C N/A N/A 4.1x10 N/A N/A N/A N/A MVP 5.4x10 N/A N/A N/A MV 1.5x10 2.5x10* 2.0x10' 5.0x10 1.0xlO* 1.2x10 2.1x10 3.7x10 1000 9.0x10* 3000 1wanam I s > <. t .

80 150 150 185 230 I 1000 9.5xlO 100 100 40 Coax, .<nt Just. 1.0x10 4.6x10 3.6x10 9.0x10 1.8x10' L.7x10" 2.3A10 2.5x10 1.4x10 2.7x10 8.0x10 6 9 14 (t.< :;t ) LVC 8.Ux10 Conelstzon.: Hoom 240 240 243 240 240 340 340 300 265 1canperature (* t ) Hoom 11 10 10 10 0 104 53 25 10 Pr essur e (psaq) 0 104 6.6 6.6 6.4 0 11-9.4 0.2 5.8 O >1l *11 *11 gN 1.F. values <1 x 10 are based on Simpson meter 9 22.5 VDC.

NO70:

l Exhibit 2 Typical Bunker Ramo Instrumentation Penetration Module and Pigtail Assembly Design

! Tested and Installed Configuration -

Terminal block inside containment Outside containment

]

]

l f 7-- n o ,

l i

i Tested configuration bM ((((((M ((((((Mb =

a a m y 6 Installed configuration ;j//

! Raychem WCSF-N Penetration I splice pigtail Module feedthrough l conductor conductors

! Raychem Glass tubing reinforced 4 Cabletray epoxy

/

Field cable C1595.006/M 03-88

' , r si _ . .

t-

p. t

, i: 2 E'$

b ("$,- $ -

, i j2

1I l:tt T.M. i 1  : i- *. tj

' i -

3 y pQ 's ;i ' ,i f; l'

~ ~ 'S 7; ,,

i k, !.

b

. f.j.

j j !I

7. .

. ,?

m, 4

7 e'$ b ' .. .

p'.j,; .

j c'. _ ;,

f2 '. 2 .. I P*-

.f.i8 i'c

'/

a l

~

.k g-m j g;, j ,4t f

I I)4 y/4< . .h j!

/ g.u/ / ,i -

Ul l a

.t L1 ' pp M q ,c h _ij -- ~ l ;i o,3 m ;I ( , '

v. 2

< ) d4n

.. v .

g, 14 1i .,

-.1 I I l w ,

.s l g

w e

• d ni a v --a v-] ,/ qt o,

[. .. ..

3, p .

@ $ 5 4. f

.c -

.f i il !, e. .

x

• 4 W , E ! 'fl'hif I

s, .7D A 1 f*

!;- ,s

% : I .

12 ill

~ !.

a I

' 3.I ---

.o*

+

'[l ' :.

• U ] I 8f f3 b* *I

gl vg

• a'g

!a .$.

8

{ s,.

4 4 t' 1 g .e

.~.

[. j w@g ga l-4.

i

-Q w ,,

i si el ad -~

l 'g L( . .

y 7i; 3: ,

%. M

~ ':

• 'I  :

T ...

.ll

>  : !i!

.x ' = Tl=s l E?Gi -

w _-) ,

N ,

lij

.,A N

tu - A

,, ,v 4 j l 3.'

\ '--

< - V m _

Mf, 0

.-- G [.:.

I

' W ,  ::;.

Wa s<%'  :.- .=. m- u -

11) l l :l .: L.u 4 r

. a

• % -%)s'

. gf.. /

. ~g 6

?

t i

J

t *

$C 5

' j.! -

41,

- ai3 _y f<4f.

l' i :s .

4ll l . . / ' **u O* t '!

l '

+ " < j. '/

9 ,

• /  ; t l .* ,2i

{

jt i ,l]4,'.'l;a *

. 3

': i {i: c;.

(.- ;L ..q ,} !, v' @F,; l a t

l ,

.' :r  : s I

s. j j i =a ==

l i 9

} .t.% riy e e i

'sa

}

= 1 l1  ;$ ,

1. }d J'_

t

• u 2

.a .. .

.- 4 e fd

.' e g o '

A __1_,

h d

i IlIf[sa

- , M., u_ i) 1*

K' I '*---

• G

-; j i

. - *: %l - 12 s~ ";

a ll!

a l

i l*

s/ @ *t

.$- , ' ,8 p ., r

%:g p gA -

9.J . N .

q. Ap sg

- -v i-,-

%i ' ; s '"

N a N., A ,g 3 -.9 f ~~ j, g

", *: , s . .

.c jt -

". 4, m , .I

'y ,P t*

/ a*) N ,

Q a

\.M g

/, }9.

.~'s l WU u ,

///

i  ;

d%'

~

W JI i - ,

a- m.

e .p$

)

/ r' 8 '.M }7 d 5

/,i ( 4 : */ i E F  ; u -

,: s h{-.4 1 'a i y

n'a'u . w . .. , .

' m9q:. ,.cs n, y 2- i. i

<e a .e ee

• 7, .

.% v. s A

~

'g, -p b' { f I l f

I I

Appendix B

. SUPPORTING QUALIFICATION DATA FOR THE ENVIRONMENTAL QUALIFICATION OF BRAIDWOOD UNIT 2 BUNKER RAMO INSTRUMENTATION PENETRATION ASSEMBLIES INTRODDCTION In addition to the environmental qualification (EQ) data presented for the Bunker Ramo instrumentation penetration assemblies used at Braidwood Unit 2, there have been other EQ test data which further support the adequacy of these penetrations. These tests were performed for:

a) Midland Station Unit 2 test of Bunker Ramo penetration assemblies by Engineering Analysis and Test Laboratory, (EATL) b) Viking Industries penetration test by Wyle Laboratories c) Amphenol penetration test by Conax

~

d) Bunker Ramo penetration test for Calvert Cliffs This appendix will provide the penetration test configurations (when applicable) and parameters and discuss how the test results apply to Braidwood Unit 2 installed configurations.

MIDLAND TEST REPORT Exhibit 1 shows the Midland 2 tested configurations (terminal blocks and Raychem splicea) and the Braidwood Unit 2 installed configuration. The Midland 2 configuration with the Raychem spli-ces and the Braidwood Unit 2 configuration are quite similar with two differences:

i l

t

e .

Appendix B a) Midland 2 pigtail material is Raychem flamtrol while Braid-wood Unit 2 is Boston Insulated Wire (BIW). However, both materials are qualified for this application.

b) The Midland 2 RFR Raychem splices do not provide an envi-ronmentally sealed connection. The Braidwood Unit 2 WCFS-N Raychem splices do provide an environmentally sealed con-nection.

Exhibit 2 shows a comparison between the Midland 2 EQ test para-meters and Braidwood Unit 2 committed EQ requirements. Exhibit 3 shows the Braidwood Unit 2 test, committed, actual Main Steam (MS) break and Reactor Coolant System (RCS) break profiles as well as the Midland 2 test profile overlapping each other. Exhi-bits 2 and 3 show that the Midland 2 test conditions are equal to the Braidwood Unit 2 requirements with the following exceptions:

a) The Midland 2 Insulation Resistance (IR) measurement volt-age is 500 VDC, while the maximum Braidwood Unit 2 circuit voltage i s 4 0 VDC . The IR values at 40 VIO would be much

higher that those at 500 VDC. Ther9 fore, using the Mid-land 2 test results is very conservative.

b) The concentration of the Boron in the M161and 2 test (13,000 ppm) exceeds the Braidwood Unit 2 requirements (2,000 ppm).

The Midland spray is more electrically con 6tetive, c) The pH value of the Midland 2 spray (7.0-7.H is about neutral. This value is not detrimental. Increasing the pH value to the Braidwood Unit 2 level ( 8. 5-10, 5) will

! not degrade the electrical performance of the circuit.

Moreover the spray period of the Midland 2 test was much longer than the Braidwood required duration.

Appendix B d) The peak temperature of Midland 2 test is 300 F while the committed peak temperature of Braidwood Unit 2 is 3200F. Exhibit 3 also shows the worst two actual accident profiles, from which the committed enveloped curve was derived. Only the Main Steam break has a peak te nperature exceeding 300 F for 60 seconds. The 320 F only occurs momentarily at the middle of this 60 second period. It is our opinion that the temperature at the penetration location would not exceed 3000 F in the actual accident condition based on thermal lag considerations, e) Exhibit 3 shows also that the transient temperatures of the Midland 2 test are slightly lower than the Braidwood Unit 2 committed values. Experience has shown that fail-ures generally occur at the peak temperature rather than the lower temperature range.

Exhibit 4 shows the IR values recorded during different phases of the Midland 2 test. These values are acceptable for Braidwood Unit 2 application.

Exhibit 5 shows the measured IR values for both terminal blocks and Epoxy End Seal configurations. This data is reported through a letter from ANCO Engineers, dated March 11, 1988. The config-uration utilizing terminal blocks has much lower IR values which supports our previous conclusions.

VIKING TEST REPORT Exhibits 6 and 7 show the test configuration and parameters as well as the available results. The configuration has some simi-t larity to the Braidwood Unit 2 configuration, with Braidwood Unit 2 having compatible or better materials. The Viking test parameters exceed Braidwood Unit 2 requirements and the test results show acceptable IR values.

l

Appandix B AMPHENOL PENETRATION TEST REPORT BY CONAX Exhibit 8 states the test conditions and describes the configura-tion. The configuration has soft Epoxy versus the hard Epoxy used in Braidwood Unit 2 which should perform as well or better.

The lowest IR value associated with the 224cP saturated steam is 3 x 10 7 ohms.

CALVERT CLIFF ANALYSIS OF AMPHENOL PENETRATION TEST REPORT Exhibit 9 shows the configuration of the Amphenol penetration used at Calvert Cliff. It also shows the test parameters as they are available in the Calver t Clif f analysis report. The leakage current during the test was reported to be less than 1 mA. This is equivalent to an acceptable IR value which supports the Braid-wood Unit 2 conclusions.

CONCLUSION The above analysis and the additional qualification test data of the Bunker Ramo penetration assemblies further confirm that the Braidwood Unit 2 installed configuration is qualified for the Braidwood Unit 2 EQ requirements in accordance with 10CFR50.49.

4 INTRODUCTION NRC AGREED WITH OUR POSITION WITH TWO CONCERNS:

a. NO ELECTRICAL MEASUREMENTS TAKEN FROM THE INSTALLED CONFIGURATION

b. ADDITIONAL TEST DATA SHOULD BE FOR A CONFIGURATION THAT IS SIMILAR TO THE INSTALLED CONFIGURATION CECO LOCATED ADDITIONAL INFORMATION TO RESPOND TO THESE CONCERNS IN THIS PRESENTATION WE WILL SHARE WITH YOU THIS INFORMATION AND SHOW HOW IT RESOLVES YOUR AB0VE CONCERNS.

l Exhibit 1 l

Midland II Tested Configurations Versus Braidwood Unit 2 Installed Configuration i

)

j' Terminal j block Inside containment Outside containment l "

i j '

3 (r-Raychem RFR #16AWG perietration l

l splice pigtail (Raychem Flamtrol) 1 Epoxy ,

l end seal

}

]

Midland tested configurations g .

! // . 6 Braidwood installed configuration g [

[' [ V/ggg Raychem WCSF-N splice d p n talon Module feedthrough pigtail (BlW) conductors W j h

Glass j tub g reinforced a m m m "**~ ~ -i epoxy 4 Cabletray Field cable l C1595.037/M 03-88

Exhibit 2 Midland Environmental Qualification ,

) Test Parameters Versus Braidwood l Committed EQ Requirements i Midland Braidwood i

Radiation 2 x 108rads. 2 x 10s rads.

! Humidity 100% R.H. 100% R.H.

Peak LOCA temperature 300 F max. for 190 secs. 320 F niax. for 170 secs.

i Voltage 500 Vdc-IR measurement 40V-max. circuit l

l Chemical spray 0.15 gpm/ft.2 0.15 gpm/ft 2 l 13,000 ppm boric acid 2,000 ppm boron i pH 7 - 7.5 pH 8.5 - 10.5 l 30 days 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> I

i l

i l

C1595.036/M 03-88 l

l

11 Exhibit 3 Environmental Qualification Profiles Braidwood committed EO profile Braidwood test profile

- Midland test profile i

~ Braidwood MS break profile j

' - Braidwood RCS break profile j 320*F at j 50 psig - T*

, ( i i

\

270*F at ,l '

1 50 psig ~ "PV 9 170*F a g .6 hr

- # 155*

~

+ '

155*F at - }

' 21 days -/'345 days

! 5 psig , N i

l N

~

f k I fab cde gh

!  ! l l l l i  !

17 days 23.5 days 30 days End of test 10 secs. 3.2 hrs. 8.5 hrs. l 4 days 2.1 mins. 3.5 hrs. 11.5 hrs.

5.5 hrs. 17.0 hrs.

C1595 007 3-15-88

Exhibit 4 j Midland ll Test Reported IR Values f

)

Inspection Thermal aging Radiation aging Post LOCA l

! (min.) (max.) (min.) (max.) (min.) (max.) (min.) (max.)

i 3 x 10' - 5 x 10 3 0 1.6 x 10 - 2 x 10 4 0 6 x 10' 0- 1 x 10 4 0 5.5 x 108 - 1.4 x 10' O Note: The above IR values represent the range (minimum and maximum) of insulation resistance values reported for the following l

J tested penetration modules:

! Module A: 69 # 16AWG

' Module B: 22 # 6 Module C: 3 x 350 l The above reported values are not attributed to a specific module

but they do bound the results for all modules C1595.039.M 03 88

i Exhibit 5 Midland ll Test Configurations and i LOCA IR Values From the Test Log Outboard 1 einboard RFR splice Epoxy end seal

. e n , ,

.g i i 3 Circuit No.1

o T  :

~1 s I

1. 16 AWG-TPS erminal Nock RFR splice

• ^

x j .  :  :

)

o  :  :  :>-c Y -

m g

\

I

1. 16 AWG-TPS Circuit No. 2 Recorded insulation resistance values (Ohms) during LOCA (#16AWG-TPS only)

Measurement 1 Measurement 2 Circuit No.1 1.5 x 106 conductor to conductor 4.7 x 10s conductor to conductor f

1.0 x 106 conductor to ground 3.6 x 108 conductor to groun'd Circuit No. 2 4.2 x 103 conductor to conductor 1.6 x 105 conductor to conductor 7.2 x 103 conductor to ground 1.8 x 10 5 conductor to ground C1505.038/M 03-88 i

CONCLUSIONS DRAWN FROM MIDLAND II TEST:

THE RESULTS OF THE MIDLAND TEST SUPPORT THE CONCLUSION ,

PREVIOUSLY DRAWN FROM THE BRAIDWOOD TEST THAT THE INSTALLED PENETRATIONS ARE QUALIFIED IN ACCORDANCE WITH 10CFR50.49.

MIDLAND II TEST MEETS BRAIDWOOD UNIT 2 TEST REQUIREMENTS CIRCUITS UTILIZING TERMINAL BLOCKS EXHIBIT LOW IR VALUES COMPARISON OF IR VALUES OBTAINED FROM SPLICES VS THOSE FROM TERMINAL BLOCKS DESONSTRATE THE INTEGRITY OF THE PENETRATION MODULE ITSELF BRAIDWOOD UNIT 2 WCSF-N SPLICES SHOULD YIELD BETTER IR VALUES THAN THE RFR SPLICES IN THE MIDLAND II TEST (i.e. RFR SPLICES DO NOT PROVIDE AN ENVIRONMENTAL SFAL THE WCSF-N SPLICES DO)

Exhibit 6 Containment Penetration

! Penetration type: Viking industries

- Tested configuration:

Inside containment l Outside containment Module feedthrough Raychem WCSF-N Polyurethane potting

' conductor splice

' ^

n. ,

p .

[ [

. 3 A mz e Rm A I

Field cable Penetratitan Polyurethane potting Polysulfane pigtall insulator conductor Heat shrink tubing i

. Test report: WYLE's NEO 46880-1 .

1 I

- Maximum LOCA test temperature 130 F l C1595.029 M 03-88

Exhibit 7 Containment Penetration 4

i j Penetration type: Viking Industries Highlights l

j Test configuration is the same as Braidwood installed configuration l Leakage current reported at t=200 sec in test at 375 F-2 mA at 50.47 V for 5 circuits in series (i.e.,1.26 x 105DIR for single circuit)

! - Leakage current reported at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> in test at 300 F-1 mA at 50.28 V for 5 circuits in series (i.e.,2.5 x 105DIR for single circuit) 1 Test conditions more severe than Braidwood Station l Braidwood Station electrical penetration material is the same as

or of superior quality than the tested one i

l i

l 1

C15%030,tA 03 88 I

1 i

I Exhibit 8 Containment Penetration .

j Penetration type: Amphenol module assembly J

Test report: Conax No. IPS-1077 1 - Test condition:

Steam and humidity environment Module assembly was installed in a hea' der plate test fixture j Raychem WCSF heat shrinkage sleeves were Installed on the inboard ends

', Soft epoxy was used in module Maximum saturated steam temperature-224 F i

) Highlights

- Lowest IR reported at 100 Vde-3.0 x 10h (conductor to shield)

Braidwood penetrations contain hard epoxy which should perform as well or better if protected from mechanical damage We have addressed mechanical damage in our response to IE Bullet.in 82-04 l

i cis s o33:u 03-88

1 l

Exhibit 9 Containment Penetration .

Penetration type: Amphenol

- Tested configuration:

Inside containment Outside containment

.t j Raychem WCSF-N 2

Modu!e feedthrough conductor splice V

i I 6 x

i Field Penetration Glass reinforced epoxy cable pigtall conductor Raychem heat shrink tubing

! - Test report: Amphenol No. 123-1252 Maximum LOCA test temperature-276 F for 10.5 minutes Temperatures between 276 F and 250 F for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> Highlights 1

i Test configuration is similar to Braidwood installed configuration i

l - Leakage current reported during the test is less than 1 mA C1595.031/M 03-88 O

9 03-22-88 SUPPLEMENT TO APPENDIX B In the March 16, 1988 meeting in Washington, several questions were posed by the NRC Staff. The following are respor.ses to specific NRC questions regarding'the acceptability of the Mid-land II test results presented in Appendix B and the extent of their applicability to the Braidwood Unit 2 instrumentation penetrations:

1. NRC QUESTION The Midland II test results presented in Appendix B indicate that the minimum Insulation Resistance (IR) recorded during the LOCA, for the #16 AWG instrumentation circuits utilizing 6

the Raychem RFR splices and Epoxy End Seal, was 1.0 x 10 Ohms. This IR value was measured "eight hours into the LOCA" (Reference Page 14 of the Midland II Test Report 123-2201). Based on the Midland II Test Report LOCA profile, eight hours into the LOCA corresponds to 100 F. However, the Braidwood Unit 2 peak LOCA temperature is 320 F. There-l fore, what assurance is there that the recorded IR value l of 1.0 x 10 6 Ohms represents the lowest value that may be

! encountered since it was not measured at the peak temper-ature of the LOCA profile?

l CECO RESPONSE To address the above NRC question we have separated our response into three sections. Section A discusses the activ-ities we went through and the documentation acquired from ANCO Engineers, Inc., to further substantiate the recorded 6

IR value of 1.0 x 10 Ohms; Section B discusses the analy-tical model of the time vs. temperature transients, at the penetration feedthrough modules' location; and Section C

'f 4

discusses the synergestic effects on the penetrations' IR values f rom LOCA parameters versus temperature alone. Based on the information that follows, Commonwealth Edison believes that the Bunker Ramo penetration assemblies, as installed, are qualified and meet 10CFR50.49 A. Subsequent to the March 16, 1988 Washington meeting, ANCO Engineers Inc., provided correspondence (dated March 21, 1988) that demonstrated when and at what tem-perature the Midland II instrumentation penetration assem-blies IR values were measured. ANCO Engineers, Inc.,

indicated a review was conducted of the temperature strip chart records for the Midland II penetration LOCA test. ANCO concludes from their review of the strip charts and the log book, that the initial IR readings were taken between 3:12 and 7:00 a.m., October 25, 1978.

ANCO states that during this time the temperature in the test chamber ranged between 200 F and 250 F. ANCO documented that the second IR reading was taken at 10:00 a.m. on October 27, 1978. At this time the temperature 0

in the test chamber was between 200 F and 250 F. ANCO further states that during October 25 and 27, 1978, the temperature was above 200 F except for four very short periods (less than 5 minutes) where the tempera-ture dropped to 175 F-220 F.

The ANCO information as provided in their March 21, D

1988 letter is the basis for our use of 200 F as the minimum temperature for the measured Midland IR value 6

of 10 Ohms. Section B below documents the use of the Bunker Ramo penetration assemblies based on the informa-tion from the Midland test. Section C below indepen-dently documents the qualification of the Bunker Ramo penetration assemblies via an analysis of the Braidwcod test report data, i

~ ~~-

B. To provide additional confidence, we have performed calculations indicating that the MSLB peak temperatures will not be seen by the penetrations prior to ESF actu-ation and during the time frame when the Post Accident Monitoring (PAM) instrumentation would be used. This evaluation documents that the temperatures at the pene-tration feed through modules will not exceed 200 F prior to initiation of the necessary instrument signals to trip the reactor and initiate safety injection. This evaluation also documents that the temperature at the feed through module will remain at a value well balow 200 F during the time frame when the Post Accident Moni-toring System (PAMS) is required. This evaluation docu-ments that the temperatures at which the IR values were takt. 3 the Midland II test envelope the anticipated feed through module temperatures during the accidents.

To evaluate the effect of the accident temperature pres-sure transient on the feed through module, a computer model of the containment and the penetration was pre-pared. The first node in the model was the area between the center support plate and the closure flange of the penetration. The feed through module is partially ex-posed to the environment in this node. The second node in the model was the portion of the penetration assembly between the inboard support plate and the center support l plate. The third node in this model was the containment

! volume. The nodes communicate with each other by means I

of the openings in the support plates. The model also accounts for heat transfer into the surrounding steel and concrete and into the Auxiliary Building Electrical Penetration area. A diagram of the model is given in l

l Exhibit B-1.

i i

The accident that was chosen for this evaluation was the Main Steam Line Break (MSLB). This accident was considered limiting for two reasons:

1. A comparison of the pressure temperature curves in the FSAR Chapter 6 shows chat the peak containment tempera-ture of 318 F for the main steam line break significant-ly exceeds the peak temperature of 267 F for the LOCA (Ref. Table 6. 2-1)

2. A comparison of the time to actuate for the pro ~tective functions for LOCA and MSLB shows that the protective function actuates in 10 seconds for MSLB (p. 15.1-19) and in one seccnd for LOCA (p. 15.6-30).

This computer analysis utilized the Westinghouse mass-energy release data and the containment pressure - temperature curves from FSAR Chapter 6 to establish the time dependent conditions in Node 3. Using this inpet, the time history of the pressure / temperature conditions in Nodes 1 and 2 was calculated. Exhibit B-2 shows the temperature on Node 1 (adjacent to the feed through module) plotted on the con-tainment temperature curve for the Main Steam Line Break (FSAR Figure 6.2-14). It should be noted that the maximum temperature at the feed through module prior to trip initi-ation (10 seconds) is 151 F. This temperature is well belo's the temperature at which the Midland IR value was measured 0

i of 200 F. Because the temperature at the Braidwood penetra-tion feed through module will never exceed the 200 F value

at which the Midland IR test data was taken, the MSLB com-l outer analysis is considered binding for the Main Feedwater

! Line Break, also. This is based on FSAR Section 6.2.1.4.

i t

o ,

e .

Finally, an evaluation of the long term penetration temperature was made. As can be seen from Exhibit B-2, the model predicted that the penetration would return to less than 130 F after approximately 400 secands.

A steady state evaluation is being performed using the longer term conditions to demonstrate that no signifi-cant heatup due to convective mixing between the vapors in the penetration and the vapors in the conto.inment was expected in the long term. This evaluation enve-lopes the long term LOCA and MSLB environments. No significant heatup of the penetration due to convection is expected.

In summary, calculations were performed that demonstrate that the temperature rise at the Braidwood feed through modules during the initiation of the MSLB, will not exceed the 200 F temperature where the Midland IR data was collected. This calculation envelops the LOCA and FWLB and demonstrates the actuation of reactor trip and safety injection prior to achieving an excessive temperature. An additional evaluation is being performed to confirm that a long term heat up of the penetration is not expected in the post-accident mode that would potentially affect the post accident monitoring instru-mentation. The purpose of this evaluation is to provide further support for the use of the Midland penetration qualification data. The completed evaluation will be summarized and available for NRC review on Monday, March 28, 1988.

Based on the above, it should be concluded that the 6

IR value of 1.0 x 10 Ohms measured during the Midland 0

j II LOCA test at 200 F represents the minimum IR value that may be encountered at the Braidwood Unit 2 penetra-tions. This conclusion is conservative because (a)

the maximum temperature at the installed Braidwood Unit i

2 penetrations has been calculated to be under the 200 F that was measured in the Midland II test (i.e. a higher IR value should be expected at the lower calculated temperatures at Braidwood Unit 2) and (b) the installed penetrations include environmentally sealed connections while the Midland II test specimens did not.

C. This section analyzes the Braidwood test report IR data collected during the LOCA and independently arrives at the same conclusion as stated above.

Material Insulation Resistance varies inversely with temperature (i.e., lower IR values should be expected at higher temperatures). However, when IR is measured during the LOCA test, one must consider the synergistic effects on IR from all LOCA parameters present (i.e.,

pressure, humidity, and chemical spray) rather than temperature alone. An examination of the IR values recorded during the Braidwood LOCA test (see Table V, Exhibit B-3) demonstrates that the overwhelming cause for the low IR values cannot be attributed to temper-ature. The IR values recorded are generally higher at the peak temperature (i.e., at 3400 F in the first and second LOCA ramps) than the IR values recorded at lower temperatures. This is not indicative of IR be-

., havior based on effects of temperature alone. In fact, this behavior is opposite of that expected. The lower IR values can, therefore, be attributed directly to leakage from the unsealed connections (i.e., terminal blocks, connectors, and splices) used in the test and caused by the presence of the humid / chemical spray en-vironment. The final Braidwood IR values recorded, f

indicate substantial recovery thus reflecting the dimin-ishing effects of the humid / chemical spray environment i

' ,b'r .

I c s

}, which provided a higher conductive medium for the leak-age current during the LOCA through the unsealed connec-tions. As previously stated, the IR values recorded for braidwood at the peak temperature of 340 0 F were substantially higher and include all of the expected IR drop due to temperature alone as well as the initial effects of the humidity / chemical spray.

4 Based on the above analysis of the test data and the fact that the installed Braidwood Unit 2 penetrations, which only include environmentally sealed connections, we expect to encounter the higher IR values as described in Appendix A and certainly not lower than the 1 x 106 Ohms recorded during the LOCA in the Midland II test.

2. NRC QUESTION.

The Midland II and Braidwood Unit 2 Test IR values were measured at 500 Vdc. However, the installed Braidwood Unit 2 instrumentation circuit voltage is 40 Vdc. Demonstrate that IR measurements at the installed circuit voltage of 40 Vdc would be more conservative than those taken during the test at 500 Vdc.

CECO RESPONSE Higher voltages produce higher corona effects and voltage stresses on cable conductors causing insulation breakdowns.

It is, therefore, expected that IR measurements at 40 Vdc will result in higher IR values than measurements at 500 Vdc. However, the improvement of IR at the lower circuit l voltage cannot be quantified. Therefore, we have utilized a conservative approach by utilizing the IR values measured at 500 Vdc for qualification of the penetrations.

O

3. NRC QUESTION The Midland II Test Report indicates that throughout the thirty day LOCA test, the low voltage power penetration modules were supplied with the rated voltage and current.

Demonstrate that the resulting heat from these energized circuits did not result in better IR measurements (i.e.,

the heat produced from the energized circuits may have provided a less conductive (dryer) atmosphere resulting in better IR values).

CECO RESPONSE The heat contribution from the energized circuits is negligible and would not result in a less conductive path around the conductors due to its potential drying effect. This heat contribution is calculated as follows:

Resistance of $16 AWG conductor = 0.523 Ohms /100 feet 500 Vdc

~4

= 5 X 10 Amperes 6

10 Ohms 2

Heat produced = I 3

= (5 x 10-4) 2 Amperes x 0.523 Ohms / foot = 1.3 x 10 -9 watts Even if the circuit was carrying rated current throughout l

the LOCA test, (e.g. 5 amperes) the resulting heat can be calculated as follows:

l 2

Heat produced = I R = (5) 2 Amperes x 0.0523 Ohms / foot = 0.13 watts The above produced heat is negligible when compared to the test I

LOCA temperatures and it would not result in a less conductive (dryer) atmosphere to cause higher IR measurements.

~

Braidwood Unit 2 Bunker-Ramo Exhibit B-1 Penetration Heat Transfer Model 7" 10" 10" 10" 5" 13.5" 1  ? O > 1 7 <  ? Oi < >

____7_ T '" - - '- T - - -'- T -'- --

11" O I

-1 _

) 7" 6

4 2 1 Penn3r3y3r37 _ _ - -_

1 7 6 5 4 16 !

4.__.___ _ ;.'.' 6 _

3 3*

~L_./-~

I* -l 'U* --

L

• • . -L I' Containment --- D5+ - 1 14* _L 3* 1 12 o M Haq building i

1 M T

'm #" A mT##

% K er-Boundary

<--@r -

node 3 17 ,,,_

F- --- ---- --

Active M 7

@ 6

@ 5

@ 4 16 l "l Oute' s i.... .E A *th pa l l g

. I g { gConductmn heat transfer path

' ~

b l Convection heat

• transfer path 3c g 21 l

g

' o l 3 o l 4

• Node 3 includes contelnment l

Pfus 7 inches of containment

-@l 15 83 . 14 1

@ 13

@ 12 C1595.0042/M 03-88 L _ _ _ _ _ _c o n c re t e j

1 I

~

Braidwood Unit 2 Bunker-Ramo Penetration Short-Term Heatup Curve

! Exhibit B-2 i Temperature ( F) 350 ~~-

i y Steam

• 300 i

i 1

250 ,

l ,

j ,-

/

! 200 /

i / Penetration feed

- nr through module

~

150 #

f 'N /

/w f

j j# -

,  %=_./ __,

1 l 100 i VVater*

50 l

l l 0 10 10' 10 2 10 10 4 Time (sec)

• Source: FSAR Figure 6.2- 14, "0.94 2 f2 t split upture at C 1595.0C8 3-23-88 30% power with s'eamline stop valve failure"

~ ' ' '

• EXHIBIT B-3

%th t. V .

LOSS OF COOLANT ACCIDENT (LOCA) RESULTS: (OHetS)

Classi- 1st 2nd 4th 5th 7th 14th 22nd 28th 30th

! Dav Dav Final Dav Circuit [ication InitialAI Ramp Ramp Ramo Rarnp 5

Dav 3

Day 3 4 4 1.5x10' l.8x10 Triax Inst. 1.8x10 3.7x10' 8.5x10 2000 1.3x10 1.0x10 2.5x10 1.3x10 1.5x10 4

Coax, TNC Inst. 3.6x10 12 3.4x10 1.9x10 220 1.2x10 65 1.2x10 2.0x10 1.0x10* 1.2x10 4.0x10 4 4 4 4 5

3x350 LVP 1.2x10 3.4x10 1.4x10 2.8x10 5 1.5x10 8.0x10' 4.0x10 1.4x10 3.4x10 3.2x10 3.0x10

$ # 3 $

69814(T.B) LVC 4.2x10 l

4.3x10 1.1x10 800 100 3.Ex10 2.8x10" 7.0x10 2.1x10 1.4x10 2.3x10 4 # 8 69814(RTR/TB) LVC 8.0x10 2.4x10 1.2x10 600 1200 1.0x10 3.3x10 4.5x10 3.4x10 1.9x10 1.6x10 MV MVP 5.4x10 0

N/A N/A N/A N/A N/A N/A N/A N/A N/A 4.1x10 4 # # 4 i 8 1.?x10 2.3x10 Twinax Inst. 3.7x10 1000 9.0x10* 3000 1.5x10 2.5x10 2.0x10 5.0x10 1.0x10 180 185 230 Coax, BNC Inst. 1.0x10 1 1000 9.5x10 100 100 40 80 150 4

11 6 8.0x10 4.6x10' 3.6x10' 9.0x10" 1.8x10 5.7x10 2.3x10 69 814 (WCST) LVC 8.0x10 2.5x10 1.4x10 2.7x10 conditions:

240 243 240 240 Poom Temperature (*T) Room 340 J40 300 265 240 Pressure (psig) 0 104 104 53 25 10 11 10 10 10 0 5.8 6.6 6.6 6.4 0 PH 0 >ll >ll >11 >l1-9.4 8.2 NOTE: I.R. values <1 x 107 are based on Simpson meter 9 22.5 VDC.

l l

l l

l ,

4 123-2201 REV. A l -

w: e. ,

i W

.A i NM ,p v  %

I i J e, .,.

m I. . g, ...g..,... . . . . . j ..

........i..... . ... .. .. .. .. i ./ l t. . . . . . . . .....3._.,..... . . . -

z a  : --__..s..__.__..

o C .A _ _ . g e. _

g . . s,i e . . i . . .. _ .-_ .. ! _....

- . _ . ~L_. .. . 1 _ _ ,i. . .,i...

.i P ~4 e i lii i I 5

.i i ..

.l. iiii.i

@l!.lIld!iS:lillil.l.jlInilul

,i..O .i;jt.ii _

y 'o , _y n pnn lu ~~..!,

.. ' ! l l ' l l l .l l e,d.

z I -- = 'I m

O o..=..=.__...:.:_-__=,...m.-._-._.=._=

4 ~; - .

- - _ . = = . . = . _ . ,. . - a I n. n - , p . .. . .

g i _. . -.- : ~ a -

p--

t g ,q -.

.M T /,. 1:- i 2 g .. i .e.

f . ... i i.

o H

[f) p4- b [

!Ii!iili, i:/elli.;ili's.lll:llill!!!:.il

/'

l k D i. $'

lIll illl}l!l l !l lIl .

e o z

w .

+. . .

..___.__...f,_

x . . . - ...-,]

_m_ .__ . ..a . . .. _. . _ . ._ ._e.

. i. _ . _ . _ _ .

I 3 . .. .. i. ,,, .

.. ..., ... 41 ,_

e d lY . 1- i . 1: U.4 ,!......li;!iliii!!'i'it:i4 ti!. li E

A b_-

.,1i '

i:.ll!i!Illlljl'lll.ilh8

!. I iI l

i mi 5. l.!'; i I W i.l, e e Iq l .

o -

TM.f EE :i/J--- -b -M -~-- -----Y.B <

g.g "-

- --/ F . "= h

.p b .,-l: .-j $

6 g=

3

~

= i i . i i .i.... .. ... ..

..v .i ._. ./. '.'.Tn. .

g 'C= , , uj II /,,  ! ' l 3 0 ! ' l I i l ! I ! . l t l i ' l l ! ! ! l A l E 3 3* g' 8 l l l ' ' ' I i 5 m m a pS , , -

c

! l,lil.!!!llilill!llh lp@, p..i 3..nl

<9;. .:.

i , ..l;-

; i (q '_

; !I so m -

i

! s

... . . u .-

c.!:'

-=,.e  :. a.

m n o.._.2_..:a

.s.__ __

_n _

< st o ~ t.W--

. . :,i......

b Y.  !

i lll!!:il'illllillli:lllillill'l!I'l'l 3 o 5 il F

i ..sj a;!,lllj;. ' '

c) h P '

it  ! l').'. llll

..!l l . l '. I I :l !

!l' llli i.

!i  ;'

i! _f e g

..:-3_ i I , ___ . .= . . ._r-

. ,._ d.; _ .- ' !g.,

o .. - - . . ,

i ..

I c.)

b i.

.,.--H_ ,.

O I (f) g. i si .- 's N- l: .l: !,! .'ve si e .ti i i. 6- g m

s 0.1 . _.- . -i. ..;;

• i.: [ *1' it!! ':l;;l lil: ..:.  ;

ll, p 1 !)!

It 8: 1881 i! :llt l li, e.

i P-

'.l

. .z _

d

-..-J....,...._

[ . ....

O 4 _.1 . ..-

!i.

a g 4 e.), ..._i..1.3........,._____.

x.r. . .

h g '

$ 4 .

.j ..i,..p...__.=..._ .q . . .l._ ' ; j_ ';, _ . ._ .

a .I .. ._i.

ijf.

H l .i'- i

. l i . . il : '-

x C. U

. q. . .;  ;  : i  ! -

$ Y c, o .a 'o> N .N % o n o o o

o n

m o.

e c

c. n ~

io g

I ul r- (J.) rganTum lle:!aa umJ 5- h 2l of 46

.i