IR 05000289/1988016

Insp Rept 50-289/88-16 on 880829-0902.No Violations or Deviations Noted.Major Areas Inspected:Licensee Actions on Previously Identified Insp Findings.Deficiencies Noted Re Failure to Implement & Evaluate Emergency Feedwater Upgrade

ML20205J714
Person / Time
Site:	Three Mile Island
Issue date:	10/18/1988
From:	Anderson C, Thomas Koshy, Roy Mathew NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20205J706	List: IR 05000289/1988016 ML20205J702 ML20205J714
References
50-289-88-16, NUDOCS 8810310465
Preceding documents:	ML20247M945 (27 March 1989)
Download: ML20205J714 (10)

Text

- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ _ - _ _

.- , ,

\

U.S. NUCLEAR REGULATORY COMMISSION REGICN I Report N /89-16 Docket N License N OPR-50 Licensee: GPU Nuclear Corporation i PTBox 480 Riddletown, Pennsylvania 17057 Facility Name: Three Mile Island, Unit 1 Inspection At: Parsippay _ New Jersey and Middletovn. Pennsylvania Ins;>ection Conducted: August 29 - September .',1983 Inspectors: / /O'N'

Inomas Koshy, $ Piir~hractor Engineer date

( /O /7 N

/M.hoy K. Mathew, actor Engineer date Approved by: b' ~

C. J. A. erson, Thief 7PTir,J.,:t'<mssection

/v/3' II date

-

Inspection Summary: Inspection on August 29-31,1988JCorporate_ Office)

}eptember 1-271983_(TNI-3 l' *j - Inspect 1on ReportTo. FO-2iW/88a15 Areas Inspected: This was an announced inspection to review the licensee's action on previously identified inspection finding Results: No violations or deviations were identifted. Fiwe unresolved items were closed. Two deficiencies were note The licenre has not fully implemented and evaluated the EFV system upgrad The adequacy of the diesel generator capacity was not well supported in the licensee's plant loading calculation PDR ADOCK 050002G9 0 FDC

- _

. - _ - -

- - - - - _ _ - - - - _ ,

,

• *

.

, ,

• DETAILS 1.0 Persons contacted 1.1 GPU Nuclear Corporation (GPUN)

• A. Agarwal, Instrumentation Manager-

• J. Anger, PWR Licensing Engineer

• Braulke, TMI Project -

T. G. Broughton, OTM Director, TMI-1

• L. Cavaliere, Equipment Qualification Engineer

• W. M. Drendall, Instrumentation & Controls Engineer

• R. Ezzo, Electrical Engineer

• B. Gan, Project Engineer C. E. Hartman, Manager, Plant Engineering

• J. Horton, Engineer-H. J. Hukill, Director, TMI-1

• Hull, Instrumentation & Controls Engineer C. Incorvati, TMI Audit Manager B. Knight, TMI-1, Licensing Engineer

• S. Y. Ku, Engineer

• J. Mancinelli, Manage'., Equipment Qualification D. J. McGettrick, Technical Function, EP&I R. J. McGoey, Licensing Manager M. A. Nelson, Manager, Nuclear Safety

• E. Pagan, Equipment Qualification Enginect

• H. Robinson, Electrical Power Manager

• J. Sadauskas, Manager, Electrical Power Instruments

• R. W. Wulf, Manager, TMI Projects U.S. Nuclear Regulatory Commission (NRC)

R. Conte, Senior Resident Inspector

• D. Johnson, Resident Inspector 1.3 Pennsylvania State Representativ A. K. Bhattacharyya, Nuclear Engineer

• Not present at the exit meetin _ _ _, . -_ . _ , _ _ . - _ , - _ . _ -_,

- _ _ _ _ - _ , - _ _ _________ _ ___________ _ __ __-__ ____ ______ _ - ___ _

.. .- ,

2.0 Purpose The purpose of this inspection was to review and verify the licensee's corrective actions for previously identified NRC finding '

3.0 Followup of Previous Inspection Findings 3.1 Closed (289/86-06-07. Item 2) Qualification of BIW Silicone Rubber  !

'

Cabl This type of cable is used in specific instrument applications inside containmont as follows:

-BIW 3/C #16AWG, 30 mils flame retardant Silicone Rubber (SR)

insulation, 45 mils flame retardant Silicone Rubber (SR) jacket with overall shield. This cable is used as an instrument cable for: RE-TE-1033 (Weed RTO with milli-volt and milli-amp circuit).  ;

-BIW 4/C #14AWG, 45 mils flame retardant Silicone Rubber (SR)

insulation, 45 mils flame retardant Silicone Rubber (SR) Jacket *

with overall shield. This cable is used as an instrument cable

,

for: WEL-LT-804, WDL-LT-805, WOL-LT-806, and WDL-LT-80 (Transamerica Delaval (Gem) Level Transmi6,ters with 115 VAC and 1/2 amp circuit).

No LOCA type test was conducted for this type cable. However, the EQ file did contain a BIW test report #B924 which included an oven ,

'

temperature test, a water immersion test, and a radiation tes To address the LOCA portion of the type test, the licensee used the LOCA test reports of the following five cable types. all silicone rubber insulated:

Test Report # Manufacture Cable Tested L i 1) Franklin F-C2946 Continental 7/C #12 AWG, 45 mil 2) Anaconda Report Continental 1/C #12 AWG, 45 mil No. 79118 on LOCA 3) Rockbestos A-708-86 Rockbestos 1/C #14 AWG, 30 mil 4) Anaconda-Ericson Anaconda 1/C #14 AWG, 45 mil report No. 80330-2 Ericson MSLB test 5) Anaconda-Ericson Anaconda - 2/C #16 AWG, 30 mil report No. 81028-2 Ericson SLB/LOCA test

,

The tested temperature / pressure profiles enveloped the TMI plant

profil The similarity between the installed cable and the tested cables was discussed on page 3 of the EQ supplemental sheet in TMI EQ file No. TI-163. The lowest IR measured during the LOCA condition

,. .-

,

7 was 1 X 10 chms for a length of 20 feet cable. Based on test data from the above test reports and the limited application of BIW cable at TMI, the inspector concluoed that the BIW cable at TMI is qualificd for its specific application. This item is considered close .2 (Closed) Unresolved Item (50-289/87-09-03) Condensate Storage Tank

~

(CST) Level Oscillations The licensee installed new safety grade level transmitters on the condensate storage tank (CST) suction lines to supply the necessary input signals for level indications and low-low level alarm as specified by Regulatory Guide (RG) 1.97 and Restart License Condition No. 3.a. Following the installation, the licensee noted large oscillations in level indications in the control room when the Auxiliary / Emergency feedwater pump is runnin An initial engineering evaluation of this problem indicated that the fluctuations in CST suction pressure were induced by fluid flow dynamics which caused the level signal oscillations. The licensee briefed the NRC staff on the problem and committed to upgrade the CST tank level indication / alarm by the cycle 7 start-up. The non-safety related CST level system was placed back in operation to complete the study on the safety grade transmitter The licensee performed the following evaluation of the new safety-grade level transmitted installation. GPUN Letter 5211-87-2128 dated June 29, 1967 and the Licensee's Safety Evaluation Report SE 412024-004 Revision No. 4 provided the results of an evaluation of the unstable CST level indications. Findings reported that the low-low level alarms on the condensate storage tanks which are set at a tank level of 5 feet are connected to pressure transmitters that are ,nounted in the condensate storage tank drain line. This configuration makes the transmitters sensitive to changes in flow causing an oscillatory input to the indication / alarm circuit During the starting of the emergency fksdwater pump and the initation of flow in the CST drain line, the output of the transmitters may dip the equivalent of a 3 to 4 foot tank level drop for a very short duration. The flow indication later stabilizes at a value of approxi-mately 0.75 feet below the actual CST tank level due to the Becrinol effect at the detecto The licensee concluded that they could reinstate the safety grade transmitters based on the following reasons. Operations normally maintains the CST Level at approximately 15-17 feet and the high level (20 feet), low level (11-5 feet) and low low level approxi-mately at 5 fee If the level droos below the low level alarm set point which is the minimum technical specification level of 11.2 feet, the operators must take corrective actions to restore the level within 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> The error in level signal caused by this oscillation and the steady state error is acceptable due to the followir.g reason *

. , .

a) The oscillation is momentary and has not caused the alarm. The level signal to the alarm is in the conservative direction causing, at worst, an early low-low level alarm b) The level error would not normally activate the low-low level alarm set point because the tank level must be maintained above the Technical Specification level of 11.2 feet (150,000 gallons)

c) If the actual tank level drops below 9 feet during abnormal plant conditions and a premature low-level alarm occurs, the operator would have sufficient time to switch to the secondary source. This early transfer to the secondary condensate source is acceptable since it does not create an undue risk to safe plant operation. Both of these transient and steady state errors are in the conservative direction. They would not cause any substantial operational problems nor any safety concern The inspectors verified the installation and concurs with the licensee's justification for utilizing the safety grade level transmitter This item is close .3 (Closed) Unresolved Item (289/86-12-17) Remote Shutdown Panel EFW Instrumentation Electrical Isolation from Control Room panels and Seismic Qualification of EFW Digital Indications During NRC Inspection 86-12 the licensee committed to provide electrical isolation between the power supplies to the EFW digital indicators on the remote shutdown panel and the control room panel This isolation was considered essential to prevent the loss of both indications in the event of a puwer supply problem in either of the locations for any reason including a seismic event. The inspector confirmed by a review of Gilbert Orawing 5130-B-600-509, revision 10-0 dated October 27, 1986 that the power supply isolation design modification provides the required isolation. A review of the licensee installation confirmed that this modification has been installed and is operationa During the NRC Inspection 86-12, the inspector noted that the electrical isolation would be of significant concern if the control room indications were not seismically qualified. A failure of the indicator due to a seismic event could affect the entire safety grade instrument loop. During the 86-12 inspection the licensee reported that the Weston Series 2470 indicators are seismically qualifie Their qualification was left as an unresolved item pending Region I review of the licensee qualification data package for these i n t,trument s .

The inspectors reviewed the Wyle Laboratories seismic qualifications test report 47430-1 Revision A dated October 3, 1984 for the Weston

.- '

., .

Series 2470 Digital Panel Meter. This report concludes that "It was demonstrated that the specimen possessed sufficient integrity to withstand, without compromise of structure of electrical functions, the prescribed simulated seismic environment." No discrepancies were observe This item is close .4 (Closed) Unresolved Item (50-289/86-13-06) Seismic Qualification of Breaker Modification The licensee modified the electro mechanical tripping device of Westinghouse 08-25 and 08-50 breakers with a Westinghouse Ampetector 1A solid state trip system. During a previous NRC Inspection, the inspector witnessed the breaker modificatio However, the seismic qualification of the modification was not available for revie During this inspection, the inspectors reviewed the seismic qualification report WCAP 10449 dated January 198 This is a generic qualification report applicable to the solid state modification of the DB series of Westinghouse breakers. Westinghouse letter dated September 5, 1985 states that the particular mounting configuration utilized at TMI-1 is a modified version of the <>riginal mounting and that Westinghouse has analyzed this configuration as presented in drawing 4378596. They concluded that it is seismically qualified for the specified application. This modification provides better breaker coordination and repeatability of trip characteristics. The licensee modified 44 breakers in safety related applications and 77 breakers in the balance of plant application This item is close .5 (Closed) Unresolved item (50-289/87-23-01) Evaluation of the

-

Voltage Dip at the 4160 Volt Bus On November 9, 1987, the output voltage of the "1B" auxiliary transformer (AXT) momentarily dippe The "1B" AXT normally supplies one-of-two vital 4160 kv buses in addition to other non-safety buses / loads, The voltage dropped down to 2400 volt However, the duration of the voltage drop was not long enough for the time delay relay to cause the associated emergency diesel generator to star Various plant equipment responded to the voltage transient, such as alternate d.c. powered equipment starting. The main turbine experiencert a runback of about 6MW (megawatts). As a result, reactor power dropped from about 99 percent to abot.t 98 percen The plant was restored to full power short'y thereaf te The licensee review determined th t the voltage dip was a result of one of the six circulating water pumps (CW-P-1F) for the secondary plant condenser experiencing an overcurrent situation. The circulating water pump motor is protected by instantaneous and time

.

. ,

overcurrent relays in each phase. The as found settings for the instantaneous relays correspond to primary currents of 3024A, 2840A, and 2992A. If the actual fault currents were lower than these values, the fault could exist for 4 1/2 seconds or more before the i relays pickup. But since the loss of voltage relays (set at 2400V)

actuated, the fault currents had to exceed 10,500 amps prior to flashing to ground. This is the minimum additional current necessary to cause the voltage to dip this low. Therefore, the fault flashed '

to ground almost immediately and the ground fault (50G) relay ,

responded faster. Typical response times at the maximum fault '

current would be approximately 1/2 cycle for the phase overcurrent relays and less than 1/4 cycle for the ground relay (50G).

Electrical faults of this nature are rare where a high fault between phases lowers the grid voltage substantially for a duration and then flashe, over to the ground. The licensee practices on the protection system was in accordance with Westinghouse Applied Protective Relaying handbook and 1EEE standard 242-1986 Recommended Practice For Protection And Coordination For Industrial And Commercial Power System The protection system responded as designed. Any prolonged voltage degradation on the safety bus would have lead to the starting of the Emergency Diesel generator and isolating the non-safety related buses which caused the fault. Even though such faults can influence plant operation, their effect will be limited to one safety trai This item is close .6 [Open) Unresolved Item (50-289/87-02-03) The Emergency Diesel Generator Load Scheme and The,Use of "0APPER" Computer Program In the course of respond.ng to action items in NUREG-0737, the licensee added several loads to the emergency bus which are required to be energized by the emergency diesel generator. By letter dated January 11, 1985 the licensee indicated that their review of the emergency power bus loadings confirmed that adequate bus capacity was available to accept the additional loads from the safety system modificatio The inspectors reviewed Technical Data Report (TOR) 836 "Evaluation of Loading for the Emergency Diesel Generator and Engineered Safeguards Buses" dated March 12, 190,7. The licensee evaluated the loading under various modes of plant operation, including simultaneous unlikely events such as loss of redundant power channel concurrent with a degraded bus voltage and loss of off-site power concurrent with loss of a redundant power channe In the loading calculation, the licensee relies on seasonal load requirements such as winter emergency loads and summer emergency loads on plant heating and cooling loads such as air conditioning, heat tracing, etc. The inspectors questioned this approach as the

_

g-.

e

r . .

. -

.

.

,

TDR did not contain sufficient bases to support this approach. There l was no verification data that these loads are prevented from starting when there is no seasonal demand. The heating and cooling systems could draw full power when abruptly called upon to operate. The licensee is taking actions to reduce the 1E Bus loads and has already relocated 90 kilowatts of load to a non-safety bus. The. licensee stated that it is unlikely that the worst case heating and cooling ,

seasonal loads would occur. They have concluded that their diesels -

are operable. Should the worst case seasonal loads occur, the diesels  ;

could be loaded to about 3000 kw, the continuous : Jty rating of the l diesels. This is below the maximum short term rating of the diesels .

of 3300 kw. In addition, plant procedures direct the operators to  !

monitor diesel loading when the diesels start to assure that the dierels are not overloaded. The licensee agreed to document their  ;

estimate of the worst case loading within a month. In addition, the  ;

licensee committed to develop a detailed calculation with sufficient '

bases to confirm the adequacy of the EDG loading oy June 198 This item remains unresolved pending further NRC revie .7 (Closed) Unresolved Item (50-289/86-19-02) EFW Back-Up I Instrument Air Banks orotection from Seismic Missiles ,

r

'

During a previous inspection, the NRC staff expressed concern about the scismic installation ef ducting, piping, and other components installed above the redundant two-hour backup instrument air banks in the "B" emergency diesel generator (EDG) room. The licensee  :

respondad by indicating that the EDG air intake supply ducting was  !

upgraded to meet seismic criteria in accordance with the FSAR commitments. No piping is above the air bank. The remaining cable and conduit, although not seismically mounted, by engineering i

judgement, would not fall and render the air banks inoperable. The licensee did not provide their basis for this engineering judgement.

, A review was made of an engineering analysis conducted by the i 1 licensee dated August 3, 1987 entitled Technical Assessment on l Scismicity of Deadweight Supported Domestic Water Piping in the TMI-1 '

DGB which is installed above the 2HBUIA. This analysis is made to evaluate whether a domestic 1/2 inch copper tubing water line

, installed above the air banks represents a seismic hazard to the air [

^

bank The analysis refers to NVREG.1061, Volume #2 Addendum Report  ;

'

of the US Nuclear Regulatory Commission piping review committee,  !

summary and evaluation of historical strong motion sarthquake seismic  !

response and damage to above ground piping, dated April 198 t Piping Seismic adequacy criteria recommendations based on performance during earthquakes, by G. S. Hardy, P. D. Smith and Y. K. Tono, presented at the symposium on current issues related to nuclear power plant structure, equipment and piping, North Carolina State *

University, December 12, 1986. The analysis made by the licensee  ;

discusses the various pipe f ailure and failure modes which were  !

reported in the two reference documents cited above and relates these f to the 1/2 inch copper tubing above the air banks.

,

1 5

--e -~ c__ -_ ~ ~ , . , ,_,,,_y_.-- ,em.w-_.__r._,__-,m____._.,,~__,,,,..,___y m_.m_.__..,__c_..m- - - , _ , ------m, , ,

r

,o. -

The analysis concludes that the domestic water line above the back-up instrument air banks in the EDG building can withstand an SSE without falling and damaging the back-up air supply. A review was made of the licensees seismic evaluation analyses of the EDG air ducts and air intake filter including its support This seismic evaluation was made by the licensee's engineering mechanics group under calcula-tion numbers 1101 X dated May 22, 1980 and calculation number 1101 X dated May 22, 1980 and calculation number 1101 X -322C-A27 dated May 26, 1981. As a result this analysis additional supports and bracing were added to the ducting and the air filter. The licensee evaluation concludes that with the additional support in place in accordance with the details provided by the analyses, the EDG ducting and air filter do not constitute a missile hazard to the air bands during SSE. The inspector confirmed the additional support and bracing t,y a visual inspectio This item is close .0 Emergency Feedwater System Upgrades Durtnc this intpection, the NRC inspectors reviewed certain areas of the licensee's modification to upgrade the emergency feedwater system to a safety grade system. The EFW is designed to initiate on any of the following signal . Low level in either OTSG High Containment pressure Main Feedwater Loss Loss of reactor coolant pumps The inspectors verified the installation of instruments, cable routing, trays, conduits for high containment pressure signal, a new signal and main feedwater loss signal, a previously existing signal to determine the adequacy of the cable routing and installatio The high containment pressure signal instruments PT1186, 1187, 1188 and 1189 and its respective conduits, trays, cables up to heat sink protection cabinets were verified and found to be color coded and installed per GPUs 500 772-A electrical cable and raceway routing criteri However, the existing main feed water loss signal instruments OPS 829, 542, 543 and 830 sensing lines, trays, conduits and cables were not upgraded. TTe licensee considers this to be a non safety related signal. At nstruments DPS 829 and 542 the inspectors observed that one if the two mounting U bolts of the instrument had missing nuts the tubing supports were missing, and some loose tubing was tied with loose wire to a conduit. The

-. _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . __ _ _

- -

.

, ,

inspectors reviewed the surveillance record in Procedure 1302-06.17 dated June 19, 1988. This record indicated a random drift of a 1 setpoint in the instrument The present condition of the instrument '

mounting and the cable routing for the loss of main feed flow signal i for the emergency feedwater actuation system could lead to undue challenges to the safety system. The inspectors relayed these concerns to the licensee management. The licensee committed to '

implement corrective action by October 30, 1988. This is an ,

unresolved item pending NRC review of the licensee action to improve l the reliability of the loss of main feed flow signal (50-289/88-16-01).

5.0 Unresolved Items Unresolved items are matters for which more information is required in order to ascertain whether they are acceptable, violations, or deviation One unresolved item is discussed in Section 4.0 of this repor .0 Exit Interview At the conclusion of the inspection on September 2, 1988, the inspectors met with the licensee representatives denoted in Section 1.0. The inspectors summarized the scope and findings of the inspection at that tim No written material was provided to the licensee by the inspectors.

j I

t I

i

,