ML20247G866

Safety Evaluation Re Silicone Rubber Insulated Cables. Anaconda & Rockbestos Cables at Plant Environmentally Qualified for Intended Function at Plant & Use Acceptable for 40 Yrs

ML20247G866
Person / Time
Site:	Sequoyah
Issue date:	07/21/1989
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20247G855	List: ML20247G852 ML20247G866
References
NUDOCS 8907280199
Download: ML20247G866 (4)

Text

- _

pn Mouq#0 4 g UNIT ED STATES d*

{

)e g y

NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20555

...~

ENCLOSURE l

1 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION SILICONE RUBBER INSULATED CABLES SEQUOYAH NUCLEAR PLANT UNITS 1 AND 2 i TENNESSEE VALLEY AUTHORITY

{

DOCKET NOS. 50-327 AND 50-328

1.0 INTRODUCTION

1 The NRC has received a number of employee concerns reldted to inadequate cable  ;

installation practices at the Watts Bar plant. These concerns were extended to I the Sequoyah plant because of the similar design and installation practices at ,

both plants. The staff and its consultant, Franklin Research Center (FRC),

evaluated these concerns by plant walkdowns, review of the installation pro-cedures at Sequoyah, and interviews with the craft personnel. A staff Safety Evaluation Report (SER) including the Technical Evaludtion Report prepared by FRC was transmitted to TVA by letter dated on March 9,1987. The SER identi-fied concerns related to cable damage caused by pullbys, jamming, and verti-cally run cable.s supported by 90 condulets. These concerns required resolu-tion prior to plant restart. Other concerns related to cabic bend radii, revision of genera'i construction specifications and cable installation proce-dures, support of long vertical runs of cable, and the cable monitoring program were scheduled for resolution post restart. By letter dated July 31, 1987, TVA submitted a cable test program to address the concerns stated above. The program called for testing of 15 worst-case conduits to determine cable damage 4 from pull.bys, jamming, and vertically run cables supported by 90 condulets.

This test program consisted of in-situ Hi-Pot testing of these cables (925 cables for pullby and jamming) in dry / wet conditions by subjecting them to a voltage level equivalent to 240 Vdc/ mil of insulation based on the minimum qualified insulation thickness of the cable. TVA completed testing for pullbys and jamming, and the results were submitted to NRC by letter dated November 20, !

1987.

The staff and its consultant witnessed part of the testing performed at the Sequoyah site. The test results demonstrated that no gross cable damage had resulted from pullby or jamming during installation. The majority of cables tested for pullby or jamming damage were not insulated with silicone rubber.

The few silicone rubber insulated cables included in this test met the dCCeptance criteria for this test.

A high rate of test failures resulted during Hi-Pot testing conducted to demonstrate the adequacy of cable installation for cables which are vertically supported by 90 condulets. All of these cables were silicone rubber insulated cables. The majority of these failures were attributed to cables manufactured 8907280199 890721 PDR P ADOCK 05000327 PDC

. i 4

j by the American Insulating Wire (AIW) Company ar.d to a lesser extent to cables j manufactured by the Rockbestos company. TVA informed the staff about these j failures in a meeting held on September 10, 1987. The staff's concern regarding l l the integrity of these cables is documented in the NRC letter to TVA dated j November 13, 1987. As a result of the staff's concerns, TVA conducted environ- "

mental qualification (EQ) testing of the silicone rubber insulated cables at Wyle Laboratory to demonstrate adequacy of the cables for a ten year life. . TVA submitted the test results to NRC by letter dated November 24, 1987. TVA also elected to remove all AIW cables from its Sequoyah plant as a result of the previous high rate of failures of these cables during tests.

By letter dated December 28, 1987, TVA documented their basis for concluding that the remaining silicone rubber insulated cables (Rockbestos and Anaconda) are adequate to perform their intended function. The staff reviewed the test results and concluded (1) that these cables were acceptable for interim use and (2) that TVA should qualify these c3 h les for the expected life of the plant at Sequoyah.

The NRC staff in a letter to TVA, dated May 25, 1988, provided the guidelines for a complete environmental qualification test program required to be con-ducted by TVA to ascertain the integrity of the remaining silicone rubber insulated cables instelled at Sequoyah and Watts Bar for the life of the plants. In this letter, the staff requested that TVA remove installed cable from Sequoyah or Watts Bar for testing. TVA, in letters dated July 6, 1988 and October 31, 1988, provided the details of their cable test program, the sample selection, sample size, cable removal from Watts Bar, and resolution of test anomalies and test failures. The staff's evaluation of the TVA program is documented in Section 3.12 of NURG-1232, Vol. 2, Supplement 1, which was issued in the staff's letter dated February 3,1989.

The staff also witnessed the cable removal at the Watts Bar plant on September 12-13, 1988, to ensure that the cable removal would not result in any damage to the cables.

2.0 EVALUATION In a letter dated March 13, 1989, TVA submitted for staff review the EQ test report No. 18056-1, " Qualification Test Program for Silicone Rubber Insulated Cables For Use In Tennessee Valley Authority's Sequoyah and Watts Bar Nuclear Plants," prepared by Wyle Laboratory. Cables tested at Wyle were placed in six different trays, each tray containing five cables from one manufacturer.

Four of the trays were designated for Sequoyah, two to hold Anaconda cables and two to the hold Rockbestos cables. The other two trays were designated for Watts Bar Rockbestos cables. There are no Aneconda cables installed at Watts Bar. One tray was aged for 15 years of life and another tray was aged for 40 years of life. Insulation resistance (IR) measurements were taken of each cable specimen and each cable was immersed in water and had 500 Vdc applied to it for one minute. Subsequently, the cables were subjected sequen-tially to the normal radiation dose, thermal aging, and accident radiation dose, i

1 wm._-._._.. _ . _ . _ . _ . .

Measurements of IR were taken after each step. At the end of the accident radiation dose, two of the five cable specimens tested for Watts Bar, for. ,

a 40-year life, recorded very low IR readings. The IR readings for the two cables recorded values in the order of 1 mega ohm, compared to other test specimens which measured four or five decades higher. IR readings of all cables tested for Sequoyah recorded acceptable values. It should be noted, however, thag the integrated radiation dose (normal + accident) for Wagts Bar was 2.585x10 rads while, for Sequoyah, the dose was lower at 1.527x10 rads. l Because of the low IR readings, TVA decided to discontinue the qualification testing for Watts Bar but continued testing the cables aged for Sequoyah. The staff inspected the cables at the Wyle Laboratory before the loss-of-coolant accident (LOCA) test chamber was sealed.

The test specimens were subjected to the LOCA environment and were energized continuously with the expected service voltage and current. Leakage current was measured throughout the test. All cables performed according to their functional requirement, 1.e., the cables carried the required current at the  ;

service voltages.  ;

After the LOCA test, IR measurements were taken for all test samples with the LOCAchambergryandflooded.STwo of the Rockbestos cables recorded an IR value of 5x10 ohms and 1.0x10 ohms at 109 Vdc with the LOCA chamber dry.

The remaining Rockbestos and Anaconda cables recorded an IR reading above 1 mega ohm. One of the Rockbestos cables shorted at 10 Vdc when the IR measurements were taken with the LOCA chamber flooded, which demonstrated that no margin is available for uncertainty in these cables. The acceptance criteria generally require a minimum IR value of 1 mega ohm. However, these smaller IR values.are acceptable because silicone rubber insulated cables are not used for the instrumentation circuits where high leakage currents due to low insulation resistance could cause instrument inaccuracies.

After the LOCA test, the test samples were taken out of the LOCA chamber, wrapped around metal mandrels, submerged in water, and subjected to DC Hi-Pot testing while the test voltage was increased at a constant rate to 6600 Vdc.

This post-LOCA Hi-Pot test was conducted to determine whether there is additional margin on these cables to resolve any uncertainties associated with the EQ test.

The staff witnessed this test and noted that all Rockbestos cables exhibited abnormal conditions, i.e., insulation breakdown, severe voltage fluctuation, arcing and smoking during this test whereas all Anaconda cables successfully passed this test.

i Based on the overall tests results, it is eviGent that although both Anaconda-and Rockbestos cables passed the required LOCA test, the Anaconda .

. cables appear to have additional margin, which is consistent with the Hi-Pot l testing conducted in prior test programs on these cables.

l

3.0 CONCLUSION

l Based on the staff's review of the Wyle Laboratory test report, the staff i concludes it is evident that both the Anaconda and Rockbestos cables at j L- _ - ___ - _ -

.l 1

> ..~ j

' 4-Sequoyah are environmentally qualified for their intended function at Sequoyah - 3 and their use at Sequoyah is acceptable for 40 years. ' The staff's conclusion .)

is based on the following:

'1. The test environment envelops the Sequoyah environmental conditions that would be present during a design basis accident, The test samples were selected from the worst case conduit installation, 2.

at Watts Bar, with the highest potential for installation damage, and

3. These cables are only used for low voltage power and ' control; therefore, the reduced insulation resistance during a LOCA would not affect the performance of these cables.

Principal Contributor: H. Garg Dated : July 21, 1989

- _ _ ___. . _ _ . _