FVY-88-028, Forwards Addl Info in Support of Proposed Change 142 Re Logic Sys Functional Test Intervals,Per NRC 880315 Request. Info Re Equipment Relay Reliability Encl

Forwards Addl Info in Support of Proposed Change 142 Re Logic Sys Functional Test Intervals,Per NRC 880315 Request. Info Re Equipment Relay Reliability Encl

ML20151R073
Person / Time
Site:	Vermont Yankee
Issue date:	04/13/1988
From:	Capstick R VERMONT YANKEE NUCLEAR POWER CORP.
To:	NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM)
References
FVY-88-028, FVY-88-28, NUDOCS 8804270232
Download: ML20151R073 (11)

Text

. .

VERMONT YANKEE NUCLEAR POWER CORPORATION

,. RD 5. Box 16L Ferry Road. Brattleboro, VT 05301 , , , , , ,

y ENGINEERING OFFICE 1671 WORCESTER ROAD

• FR AMmHAM. M ASSACHUSETTS 01701 April 13, 1988 * * "'"* "

FVY 88-028 United States Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555

References:

(a) License No. DPR-28 (Docket No. 50-271)

(b) Letter, VYNPC to USNRC, FVY 87-107, "Proposed Change to the Vermont Yankee Technical Specifications - Logic System Functional Test Intervals," dated November 30, 1987 (c) Letter, VYNPC to USNRC, FVY 88-04, "Clarification to Vermont Yankee Proposed Change No. 142 - Logic System Functional Test Intervals," dated January 20, 1988 (d) Letter, USNRC to VYNPC, NVY 88-041, "Meeting Summary,"

dated March 17, 1988

Subject:

Additional Information in Support of Vermont Yankee Proposed Change No. 142 - Logic System Functional Test Intervals

Dear Sir:

By letter dated November 30, 1987 [ Reference (b)], Vermont Yankee submitted the subject proposed change to revise the Technical Specifications for trip system logic functional testing intervals as a result of the expanded testing methodology incorporated during the 1987 refueling outage. Pursuant to a recent discussion with the NRC staff associated with the review of the subject amendment request, Vermont Yankee has been requested to provide the j information supporting our technical presentation during the March 15, 1988 '

meeting with the staff [ Reference (d)]. Specifically, the staff has requested additional information regarding equipment (relay) reliability.

In accordance with the staff's request, we herewith provide, as Enc..osure 1 to this letter, the information supporting Vermont Yankee's technical presentation at the March 15, 1988 meeting concerning relay l reliability.

I \

8804270232 0 $7 PDR ADOCK O ppg P

]

United States Nuclear Regulatory Commission April 13, 1988 Attention: Document Control Desk Page 2 Should you have any questions or require further information concerning this matter, please contact this office.

Very truly yours, VERMONT YANKEE NUCLEAR POWER CORPORATION l /

R. W. Capstick Licensing Engineer RWC/25.525 Enclosure cc: USNRC - Office of NRR Mr. Vern Rooney, Senior Project Manager Project Directorate I-3 USNRC Region I USNRC Resident Inspector ASLB Service List l

ENCLOSURE 1 VERMONT YANKEE REIAY RELIABILITY

SUMMARY

During the March 15, 1988 presentation at NRC offices concerning the Vermont Yankee (VY) proposed Technical Specification change for Logic System functional test intervals, VY discussed the reliability of the relays utilized for these logic operations. The following supplemental information is provided in response to NRC's March 31, 1983 request:

Review of Vermont Yankee-Specific Relay Failures In an attempt to determine the reliability of relays utilized at VY within safety systems, a documentation review was performed to identify the number of relay failures experienced at VY. The following documents which identify relay failures were reviewed:

o Past Results of Logic System Functional Tests o License Event Reports (LERs) )

o Potential Reportable Occurrences (PR0s)

)

J Review of LERs and PR0s extended as far back as 1980. All three of the l above documents were reviewed for the years 1983 through 1987. The attached table (Table 1) summarizes the results of these reviews for the past five l years (1983 through 1987).

Although a combination of General Electric HFA, and HGA relays make up the majority of the relays installed in safety systems, the documentation review addressed failures of any type of relay utilized at VY. A total of eight failures over the five-year period were identified for all relays (approximately 500) installed in safety systems at VY, reflecting the excellent reliability of this equipment. These failures included only i HFA relay and 0 HGA relays. Further, the following considerations apply:

6463R/20.519

I o Five out of the eight failures involved timing relays. This type of failure does not necessarily mean that the required functions would have failed. It could mean that required functions would be performed, but either earlier or later than administrative or Technical Specification limits.

o Three out of the eight failures were relays which were normally

~

energized. These relays are designed to fail in the "safe direction." In all three cases, relays failed as-designed. In addition, these relays are the only relays out of the eight failures which were not timing relays.

o Two of the failed relays are not within the set of relajs for which the proposed amendment would alter testing requirements.

o All the relay failures were determined not to have adverse safety implications.

Review of Generic Relay Reliability To summarize, our review concerning the generic reliability of relays as discussed on Much b,1988 involved reviewing the following documents:

o "Nuclear Plant Reliability Data System (NPRDS) Search for HFA and HGA Relays."

l l

o NEDC-30851P, "Tecnnical Specification Analysis for BWR Reactor  !

Protection System," by General Electric, dated May 1985 (HFA relays). )

1 l

Results of both generic reviews indicated that subject relays have an excellent reliability.

1 The attached description (Attachment I) of the design and reliability of HFA and HGA relays was provided by General Electric (manufatturer of the relays) in response to recent questions regarding the reliability of these relays. The identification of 1 HFA failure and 0 HGA failures, as described 6463R/20.519

1 in the VY-specific documentation review, verifies the close correlation between the reliability of relays utilized at VY, and the generic published reliability (Attachment I) of these relays. It is significant to note that the single HFA relay failure was due to the old-style HFA coil and not'due to the newer Century Series HFA coil presently installed in VY safety systems.

Other BWRs As discussed during the March 15, 1988 presentation, VY would like to reiterate that the utilization of similar relays to those used at VY by the majority of other BWRs provides a large population of equipment (several hundred relays per plant) upon which reliability conclusions can be drawn.

Twenty-three of thirty other BWRs perform the Logic System Functional tests during refueling intervals or every 18 months. Furthernere, equipment at the older plants has been in service longer than the relays at VY with no decrease in reliability due to any aging effects. Past'and continuing experience with this equipment provides assurance that the equipment is indeed reliable over long periods of operation, and is capable of providing reliable service over its 40-year design life.

Vermont Yankee Commitment to Reliable and Safe Operation The reliability of relays or of any of the equipment utilized at VY is strengthened by VY's attention to maintenance as an important contributor to safe operation. This dedication was evidenced by VY's program, implemented over the years 1983 through 1986, to replace safety class HFA relay coils when the industry identified a generic problem with these coils back in 1982. The application and use of industry generic information regarding equipment problems has often precluded or eliminated the same problem from occurring at a specific plant. The real reliability of any component, therefore, should take into account that anticipatory and preventative maintenance programs are well organized and share a huge "database" of industry experience for just about any component used in a nuclear power plant. In ad/ition, the numerous detailed maintenance, operation, and testing procedures existing at VY ensure that commitments to reliable and safe operation are continuously implemented.

6463R/20.519

Conclusion Based upon the above discussions and the results of all subject reviews, it'is concluded that the relays utilized at VY have proven, on the basis of_ active service and vendor testing, to be highly reliable components whose reliability is insensitive-to testing frequency.

6463R/20.519

TABLE 1 Page 1 of 2 VERMONT YANKEE SPECIFIC RELAY FAILURES - 1983 TO 1987 Type How Was Would Logic How Was Of Normal Type Of Failure Testing Have Failure Failure Relay Condition Failure Discovered? Detected Failure? Documented? Year Comments No.

GE CR120 DE-EN Timer Logic Yes 1) Past Logic 1983 10A-K45A 1 Malfunction, Functional Surveillance TDPU MR Required Surveillance Test Result LER has Test INCOR ID I

2) LER 83-17/3L AGASTAT DE-EN Timer Logic Yes 1) Past Logic 1983 13A-K42 2 Model 2412 Malfunction - Functional Surveillance Time Delay Timer Not Surveillance Test Result Setpoint out of Adjustable, Test Administration MR Required 2) PRO-38 Limits. Not out of T.S.

AGASTAT DE-EN Timer Logic Yes 1) Past Logic 1984 13A-K42 3 Model 2412 Malfunction - Functional Surveillance Time Delay MR Required Surveillance Test Result Test AGASTAT DE-EN Timer Out Logic Yes 1) Past Logic 1985 13A-K42 -

Model 2412 Of Tolerance Functional Surveillance Time Delay

- Required Surveillance Test Result Not Considered Adjustment Test to be a Failure GE CR120 Energized Coil 1/2 SCRAM Not Applicable 1) LER 87-01 1987 Proposed Change 4 (Control Relay Received RPS in Testing is in the RPS MG Not Applicable Set) to This System GE HFA Energized Coil, 1/2 ISOL Yes 1) PRO-4 (1983) 1983 Fail Safe 5 i Relay Replaced Received Old Style Coil 16A-K3C

i TABLE 1 Page 2 of 2-VERMONT YANKEE SPECIFIC RELAY FAILURES - 1983 TO 1987 l

Type How Was Would Logic How Was l

Of Normal Type Of Failure Testing Have Failure Failure

Relay Condition Failure Discovered? Detected Failure? Documented? Year Comments No. ,

i l

l AGASTAT DE-EN Timer RCIC SIM Yes 1) PRO-56 (1983) 1983 13A-K7 6 Model E7014 Malfunction, Line High TD Timing Flow Mechanism Func./Calib.

Failed OP-4364 GE CR120 Energized Not Specified, During No, This Relay 1) PRO-18 (1985) 1985 16A-K16,. 7 MR Required Ground Check is Exempt From This Failure Logic Testing Would Have Been per Vermont Detected During Yankee Once/ Operating Technical Cycle Testing Specification AGASTAT DE-EN Timer Logic Yes 1) PRO-87-46 1987 10A-K50A 8 ETR Malfunction, Functional Timer Out Of Timer Reset Surveillance 2) Past Logic T.S. Limit Test Surveillance Test Result

f'APR- 1-GO FRI 15: 26 CE MCDD PURCHASING P.O2

• OEZG47 l . . . . ., , , ,

l 4

l O. .

New GE CEN"U3Y Series Auxi'iory 3e obs Types RFA,FGR,and HmA

.S .

..3 - - . . .

GEN ER Alh ELECTRIC ATTACHt4ENT I PAGE 10F3

) ASIB.c... 1, --& S . . F R .I 15: 27 GE MCDD PURCHASING P.O3 GEZ-6847 . l

.. . .'. i o .v. .L u..NEW GE CENTURY SERIES NUs LIARY RELAYS

' 's .v. E.; -TYPES HPA, HGA and HMA I

General Electric auxiliary relays such as UFA, HGA and HMA types, have a fine service record with very few failures. The service life of these rugged relays has been in the order of 30 to 4 0 years at 20'c average temperature, With this even when design, thecontinuously elapsed ener-time to gized at rated voltage.

first failure (that is, the time when it of all such relays have failed) is expected to be 10 to 12 years. Service experience of continuously energized HFA relays with ac coils has confirmed that expected life.

However, for nuclear stations, the Nuclear Regulatory Commission (NRC) is challenging the industry to design a cew criteria--not 40-year life, but 40 years with less than 1% failure. This is roughly four times longer than the present design which has an expected life of 10-12 years to 1% failure. Thus, 4 0-year life with less than 1% failure became the objective for a new GE auxiliary relay coil design.

The new design involves a change in the entire insulation struc-ture, Relays with ac coils are the greatest challenge. These e

relays contain a shading ring on the pole piece to prevent chatter. Eddy currents flowing in the shading ring create localized heating. When continuously energized, the area of the coil spool near the shading ring runsFor even' hotter this reason, l

than coil temperature rise would suggest. l the spool material is the finest high temperature polymer that could be found to obtain long-term strength at elevated temperatures. Under accelerated life testing, it did not crack or exhibit brittleness, e The wire insulation has been changing to polyamide-imide film. Here the requirements were to retain insulation integrity and mechanical strength at continuous elevated temperatures, and also to be non-hydroscopic and fungus resistant. .

e These polyamide-imide insulated coils, wound on high-temperature spools, are pre-baked to drive off all vclatile materials, vacuum-pressure impregnated with a solventless varnish, and then post-baked. The impregnation material is also non-hydroscopic and has temperature expandion coef ficients compatible with the spool and with the wire, so that stresses do not develop under temperature cycling. _

{c

PAGE 2 CF 3

• APR- 1-08 FRI 15: 27 GE MCDD PURCHASING P. 04 Y- :' . _ _ _

.i -

. GEz-681q

'J C' Ace'eier'a t'e'd . life!.tc)fts' hhvci be,en. compl.eted on..th, eye. new . coil s--tests ,

~ '

and. maximum. vol tage. Ds'ing" Arrh'enius" 7

u. mef.png 's t el~eVa ted > tbinpeya'threr .

"" 2 plo,hs, * ~ 'thh"n'6Wde'slin's ,haVe no t only met, but .have ,qxc.eeded', .' tbe i

" dasigh objecti'06".~ Th'e 'new coils have a lif e of 40 years to 1%  !

failure not just at 20'c but at 55'C . . . that is 12'4 *C continuously.

This predicted life is not just at rated voltage, but at 110% 1 rated voltage. At nominal conditions (that is at an ambient temper- i ature averaging 20*C year round, day to night, Winter to summer) when j energized continuously at 100% rated voltage, we can expect 100 years--

that's right--100 years average life even for ac coils !

The basic differences in the CENTURY auxiliary relays are as follows:

e Spool - High thermal strength polymer.

e Wire insulation - Polyamide~ imide wire coating (180*C rating)

Tef zel insulation where required, such as leads.

o Encapsulation - Polybutadiene solventless impregnant.

e Model No. - New but easy to determine. Simply add 100 to the old relay model number. Thus, HFA51A becomes HFA151A and HGAllJ becomes HGAlllJ.

e Nameplate - Green, for easy visual dif ferentiation m from standard life relays.

O Retrofit kits are now available for all prior design relays. All auxiliary relays now in service can be upgraded to the design life of the CENTURY series. . . 100-year average life under nominal conditions. -- .-

.If'GE

. . ..Type'HF

:.; . y;;'X relays are now installed, just replace the coil, magnet assembly and nameplate with a CENTURY design modification kit. ,

If an HGA relay is now installed, just replace the coil and nameplate.

The entire' relay need not be replaced.

1 1

In the case of the Type HMA relay, it is recommended that the entire )

relay be replaced with a CENTURY series HMA, since this relay cannot I be readily disassembled. j 1

In all cases, new relays or retrofit, the green nameplate will serve l as a reminder that this relay is a GE CENTURY series auxiliary relay.

• An established method for translating accelerated life tests at elevated temperature to service-life predictions.

, General Electric Company

~ Power Systems Management Philadelphia, PA 19142 December 16, 1977 3 p,gg 3 g,3