ML20234C959

From kanterella
Jump to navigation Jump to search
Final Technical Assistance Rept for Mar-May 1987 for Fermi Nuclear Station Unit 2
ML20234C959
Person / Time
Site: Fermi DTE Energy icon.png
Issue date: 06/02/1987
From: Gunther W, Shier W
BROOKHAVEN NATIONAL LABORATORY
To:
NRC
Shared Package
ML20234C924 List:
References
CON-FIN-A-3550 NUDOCS 8707070040
Download: ML20234C959 (23)
v • d • e


Text

_-_.

j q

.. 1

~ ,

I a

BROOKHAVEN NATIONAL LABORATORY TECHNICAL ASSISTANCE REPORT - FINAL FIN A-3550, TASK. ORDER 007 DATE OF REVIEW: March - May 1987 i

LICENSEE: Detroit Edison Company PLANT: Fermi Unit #2 Nuclear Station BNL TECHNICAL SPECIALISTS: Mb[

W. E. Gunther Date 2 67 W NN b 3 5] l W. Shier Dhte i

i l

l BROOKHAVEN NATIONAL LABORATORY l} g)l ASSOCIATED UNIVER$1 TIES, INC.(E lll .

l hh o

A

1 I

. -111- L.

  • i CONTENTS Section Title Page No. i i
1. INTRODUCTION....................................... 1
2. o1ScoSS10N......................................... 1  ;

I

3.

SUMMARY

OF RESULTS................................. 2 1

4. PERSONNEL C0 NTACTE D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

(

5. DOCUMENTS REVIEWED................................. 3
6. AREAS REVIEWED..................................... 4 6.1 Reactor Water C1eanup......................... 4 6.1.1 Background............................. 4 6.1.2 Flow Calibration....................... 5 6.1.3 Temperature Switch Problems............ 8 6.1.4 Mechanical Problems.................... 11 6.1.5 Design Problem ........................ 12 62 Surveillance Procedure........,................ 13 t

6.3 Motor Operated Valve (MOV) Program Review..... 15 ,

k 6.3.1 Background............................. 15 1 6.3.2 Maintenance Procedures................. 15 6.3.3 Training............................... 16 6.3.4 Preventive Maintenance Program......... 16 i

J 6.4 Post Accident Hydrogen /0xygen Analyzers. . . . . . . 17 l 6.4.1 Background ............................ 17 6.5 Plant Inspection.............................. 20

-~m I. INTRODUCTION At the request of the NRC Region III, BNL technical assistance has been provided at the Fermi Unit #2 Nuclear Station to review.and assess licensee activities and programs as directed by the NRC onsite inspectors. This review has encompassed the following areas:

1. Licensee Event Reports (LERs) associated with the Reactor Water Cleanup System.
2. Instrumentation and Control (I&C) surveillance and calibration procedures.
3. Motor operated valve (MOV) maintenance program.

I

4. Post accident hydrogen and oxygen sensor failures.

Fermi Unit #2 is a single unit BWR located south of Detroit, Michigan, and is operated by the Detroit Edison Company. The unit is presently in its power I ascension program intended to verify proper design and operation of the plant's safety and power conversion systems prior to commercial operation.

This report describes the status of the reviewed areas, and offers an assessment of licensee activities in those areas. It also provides recommenda-tions for further inspections and makes conclusions regarding program adequacy.

'~

2. DISCUSSION A two-day visit to the Fermi site was made in early March to complete site badging requirements, meet with NRC site inspectors, and obtain required docu-mentation, drawings, and procedures to support the inspection. Initial contact with licensee representatives knowledgeable in the reviewed areas was also made.

After reviewing the materials at BNL, a one week on site inspection was conducted beginning on 3/23/87. At that time, the following inspection activi-ties were completed.

1. Reviewed no. tor operated valve maintenance records, i
2. Inspected equipment associated with the reactor water cleanup system tem-perature and flow monitoring, as well as the RPS instrumentation involved with the reviewed I&C surveillance procedures.
3. Inspected instrumentation and equipment associated with the hydrogen and oxygen sensors manufactured by Exosensor.
4. Discussed I&C surveillance and calibration program with licensee represen- 1 tatives. l l

I

5. Obtained additional background and support documentation.

On 5/20/87 a follow up trip to the Fermi site was made in order to assess recent licensee activities in the reviewed aress. Findings, conce rns , and recommendations were presented to plant representatives at a meeting held on 5/22/87..

J

3.

SUMMARY

OF RESULTS The results of the inspection at Fermi Unit #2 may be summarized as j

, follows:

i A. Reactor Water Cleanup (RWCU) LERs l Extensive problems have been experienced with the operation, testing, and I maintenance of the RWCU system. For the most part, the LERs have been related to the inadvertent isolation of the system due to instrumentation errors. A number of significant hardware and administrative changes have been made which ,

should improve system performance.

Based on the in-depth review of the experiences and the corrective actions taken, it is recommended that 16 of the 23 LERs reviewed be cloced. Further resolution is required on several of the LERs. The basis for LER closure and the corrective actions still to be implemented are described in Section 6.1.

B. I6C Surveillance and Calibration Procedures The review of selected I&C surveillance procedures revealed that, in general, industry standards and regulatory guidance are met so far as technical adequacy and format a re concerned. As directed by the senior resident inspector, minor technical comments on specific procedures were forwarded to the  !

appropriate licensee representative for resolution. These comments, along with i

some general observations regarding the programmatic changes being made to the j surveillance procedures, including the additions of an impact statement and a '

loop diagram, are described in detail in Section 6.2.

C. Motor Operated Valve (MOV) Maintenance  !

Maintenance instructions, training lesson plans, Licensee Event Reports (LERs), Deviation Event Reports (DERs), and response to IE Bulletins and Notices related to MOVs were reviewed. This material contained detailed technical information which adequately addressed many of the industry wide hardware pro-blems which have been experienced with lubricants, limit switches, and torque switches. Exceptions to this, along with several areas where insufficient information exists at this time to perform an evaluation, are described in Section 6.3.

D. Exosensor Hydrogen / Oxygen Analyzer Problems The containment hydrogen and oxygen post accident monitoring instruments-tion has experienced a high f ailure rate since its installation. Recently, a number of positive actions have been taken by the utility to reduce this failure rate including increasing the calibration frequency, pernerming routine sensor replacement every 6 months, and participating in a test prcgram with the manu-facturer. Due to the lack of experience since the implementation of these actions, a positive determination that the root cause of the problem has been eliminated is not possible. This report provides a history of the problems experienced along with an evaluation of the corrective actions taken. Finally, it recommends the followup activities necessary in order to resolve this issue. l

,4. PERSONNEL CONTACTED A number of Fermi Unit #2 personnel were contacted during the course of this inspection. The following personnel provided information which was evaluated as part of this inspection.

1. F. Sondgeroth 9. J. Thorpe 17. W. McNeil l I
2. L. Bregni 10. L. Esau* 18. S. Cashell*
3. M. Sierra
  • 11. D. Thomas 19. S. Frost
  • i
4. R. DeWulf 12. C. Banj anin 20. G. Trahey* l S. T. Moffett 13. L. Ferguson 21. R. Ballis
6. R. Assenmacher 14. E. Vinsko 22. R. Filipek I
7. C. Phillips 15. K. Faulk 23. D. Schultz
8. R. Rateick 16. R. May In addition, Messrs. B. Sylvia, Nuclear Group Vice President, R. Lenart, .

Plant Manager, and J. Leman, Director of Plant Safety attended the exit meeting j f on May 22, 1987.

01 l

5. DOCUMENTS REVIEWED
1. IE Information Notice No. 86-69: " Spurious System Isolations Caused by the 3 Panalarm Model 86 Thermocouple Monitor," 8/18/86.
2. AEOD Engineering Evaluation Report AEOD/E604, 3/14/86, " Spurious System Isolations Caused by the Panalarm Model 86 Thermocouple Monitor."
3. ISA Standard S67.06, " Response Time Testing of Nuclear Safety Related i Instrument Channels in Nuclear Power Plants," 1984.
4. ISA Standard S67.04, "Setpoints for Nuclear Safety-Related Instrumentation -

Used in Nuclear Power Plants," 1982.

5. { 1 ANSI /IEEE Std. 338-1977, " Periodic Testing of Nuclear Power Generat.ing Station Safety Systems."

6.

Nuclear Power Experience (NPE) Reports; Volume BWR-2, VII, RWCU, pps 35-64.

7. LER Data Base, Sequence Coding System, RWCU LERs, 1985-1986.
8. E. Fermi Unit 2 PSAR.
9. LER Data Base, Sequence Coding System, Motor Operated Valves, 1985-1986.
10. Enrico Fermi Functional System Description for Primary Containment Monitoring System, FSD No. T50-00-SD, November 1983.

I1.

I&E Information Operator," May 1,Notice 1984. No. 84-36, " Loosening of Locking Nut on Limitorque 0 Attended exit meeting

1 1

l

. 12. I&E Information Bulletin No. 85-03, " Motor Operated Valve Common Mode Failures During Plant Transients Due to Improper Switch Settings," November 15, 1985.

13. I&E Inf o rmation Notice No. 85-20, " Motor Operated Valve Failure Due to l Hammering Effect," March 12, 1985. i 1
14. I&E In f o rmation Notice No. 85-22, " Failure of Limitorque Motor Operated Valves REsulting From Incorrect Installation of Pinion Cear," March 21, 4 1985. I
15. I&E Information Notice No. 86-29, "Ef fects of Changing Valve Motor Operator Switch Settings," April 25, 1986.  :
16. I&E Information Notice No. 86-93, "IEB 85-03 Evaluation of Motor Operators Identifies Improper Torque Switch Settings," November 3, 1986. ,

l

17. I&E Information Notice No. 87-08, " Degraded Motor Leads in Limitorque DC Motor Operators," February 4, 1987.
18. Enrico Fermi Deviation Event Reports on Exo-Sensor Analyzer,, 1985-1987.
19. Detroit Edison, Nuclear Operators Training Lesson Guides, " Motorized Valve Actuators," Rotork, Rotork (Lab Exercise) Limitorque, Limitorque (Lab Exer- l cise). '-
20. Fermi 2 Procedure-Operations , " Primary Containment Monitoring System" Pro-cedure Number 23.408, February 1987.
21. MI-IC-3051, "H2/02 (Exo-Sensor) Monitor Maintenance Instruction," September 1986.
22. AEOD Engineering Evaluation Report AEOD/E705, 3/87, "RUCU System Automatic Isolation and Safety Considerations."
6. AREAS REVIERED i

6.1 Reactor Water Cleanup 6.1.1 Background From 6/1/85 to 3/1/87, 25 LERs involving the Reactor Water Cleanup (RWCU) system at Fermi Unit 2 have been written. All of these LERs are associated with the steam leak detection and isolation subsystem of RWCU. A review was made of 23 of these LERs (2 had been previously closed) and its associated documenta-tion, including Deviation Event Reports (DERs), procedures, modifications, and training, in an effort to determine the licensee's effectiveness in providing proper and timely corrective action to the problems experienced. In addition, a review was made of the nuclear industry experience with RWCU including a review of the LERs (173) reported in 1985 and 1986 from all BWRs, and a large sample of RWCU problem reports described in the Nuclear Power Experience (NPE) data base.

n

4 f

, Section 1.2 of the report summarizes the experiences of Fermi and provides the lead in to the sections which follow, culminating with BNL recommendations for LER closeout in Section 1.4 By way of background, analysis of the LER data base is included here to indicate how Fermi-2 compares with other BWRs in the country as well as other i BLTRs in Region 3. Generally speaking, the newer plants have a larger percentage l of reported failures indicating design and/or procedural inadequacies. A recent i l

AEOD report ( AEOD/ E705, 3/87, "RWCU System Automatic Isolation and Safety Con-siderations") was also reviewed to obtain ?.nsight into generic concerns and safety implications. This report indicated that nearly 75% of the isolations l could be attributed to incorrect spurious operation of the instrumentation.

l The 23 LERs reviewed can be categorized into four areas; temperature detec- l tion sensitivity problems, dif ferential flow inaccuracy, equipment failure, and design problems. Repetitive problems experienced with temperature and flow instrumentation are addressed as are several design modifications and procedure changes made to alleviate them. Corrective action for personnel errors made in these areas is in the form of training and procedural change, while equipment failures have been addressed by modifications and procedural changes.

6.1.2 Flow Calibration Six LERs have been written due to flow instrumentation problems which have resulted in RWCU isolation, including LERs85-024, 034,85-046, 85-061,85-063, and 85-065. These LERs may be summarized as follows:

A. LER 85-024 Event Date - 6/22/85 Related DERs 0317, 85-0133 Two RWCU isolations occurred on 6/22/85 due to the differential flow instrumentation (square root converter) being out of calibration. Because of the concern that condenser vacuum changes due to gland sealing steam problems may also have af fected the flow instrumentation, the LER referenced DER 84-433 written to resolve that concern.

B. LER 85-034 Event Date - 7/6/85 Related DERs 0347, 85-0433 A RWUU isolation occurred due to inadequate calibration of the instrument loop. Following the event, a complete loop calibration revealed that the summ-ing portion of the circuit had not been checked and was causing the loop to read high. Procedure 44.020.152 was revised to include a calibration of the summing device.  ;

i i

C. LER 85-046 Event Date - 8/13/85 DER 86-0433 Due to a change in condenser back pressure and the sensitivity of the blow-down flow instrumentation to this pressure change, the RWCU isolated due to indicated high delta flow. An administrative limit of 300 gpm system flow was imposed.

t

]

REACTOR WATER CLEANUP SYSTEM ANALYSIS OF 1985 & 1986 LERS l

Rimrbend 1: 17 Lirnerick 1: 23 Hatch 2: 16 -

Hatch 1: 11 :!:::.  ::: - FERMI 2: 17 i

!!!!!!!!:...  : gg WPPSS 2: 9 , p# l

,- vp Shoreharn 7 c' j LaSalle 1: 7 s s

l1

~~~

24 Others: 66 ALL BWAS-173 EVENTS Reactor Water Cleanup System  !

Analysis of 1985 & 1986 LERs '

FERMI 2: 17 i

LaSalle 1; 7 i

4%(/;Qg Perry 1: 1 Quad Oties 2: 1

  • . Quad Cties 1: 1

!!!!!: Dresden 2: 1

'!!!!! Clinton 1: 2 LaSalle 2: 5 Region 3 BWRs-45 Events

1 D. LER 85-061 Event Date 9/24/85 DER 85-0508  !

i The system isolated on high differential flow when returning the Filter / l Demineralized to service following a first isolation. This apparently resulted from perturbations in system flow and pressure. The alarm circuitry was changed on 9/29/85 so that the high differential flow alarm would be obtained as soon as flow exceeded the setpoint rather than after the 45 second time delay.

E. LER 85-063 Event Date 9/20/85 DERs 85-0501, 85-0433 h

The RWCU system isolated on high differential flow while proceeding to a I hot standby condition (MSIVs closed, no condenser vacuum). The flow transmit-ters were found to be out of calibration. New instrumentation is being con- [

sidered. j F. LER 85-065 Event Date - 9/28/85 DERs 85-0513, 85-0433 1

During reactor heatup, the system isolated on high differential flow due to {

the errors in flow measurement attributable to changing water densities. The i loop was calibrated to a standard reference temperature to achieve zero indi-cated leakage at that condition.

Corrective Action The corrective actions taken to resolve' the differential flow problems

~ '

include the following:

l. The GEMAC flow transmitters have been replaced with more reliable and accurate Rosemount transmitters (EDP-4343).
2. The entire instrument loop was recalibrates to a reference temperature of 126*F. This provides for accurate indicated leakage at low tem-perature conditions (when temperature isolation sensors may not be l

effective).

3. Modified the high differential flow alarm to be initiated when the condition is sensed rather than when the isolation occurs (44 second time delay) as originally installed.
4. Provided differential flow indication in the main control room (0-200 gpm).
5. The calibration procedure was revised to include all devices in the l instrument loop.
6. The calibration frequency for the flow measurement loop was increased from quarterly to monthly.

According to the system engineer, a complete flow calibration will be per-formed during a future plant heatup when RWCU can be removed from service. This is necessary since at rated conditions an indicated flow of approximately 20 gpm exists with no actual system leakage. It should also be noted that the systems

engineer attended an INPO workshop on 4/13-14/87 on the subject of RWCU. Flow measurement problems were discussed, and it appears that a longterm solution to compensate for density changes at different plant conditions is being pursued by the industry.

BNL Assessment A review of industry wide LERs and NPE problem reports revealed that a num-ber of common problems associated with differential flow measurement exist.

With flow instrumentation calibrated for rated temperature and pressure condi-tions, significant errors are introduced during startup and shutdown conditions ]

due to water density differences. Consideration has been given to modifying the '

j flow logic to more accuracely represent mass flow rates by providing temperature j compensation or by changing the trip setpoint for different plant conditions. '

Before LER closute can be recommended for LERs85-046 and 85-065, a calibration of the flow instrumentation during heetup ehould be. completed.

Based on the actions completed to date, it is recommended that the follow-ing LERs be closed:

LERs85-024, 85-034,85-061, 85-063 These LERs were properly investigated with the ultimate corrective actions of transmitter replacement and alarm logic changes being made to mitigate the problem. , ,

It should be noted that since 9/28/85, there has not been a RWCU isolation due to flow instrumentation errors. While it is recognized that significant improvement has been obtained, the operation of the system during transient (heatup/cooldown) conditions has not been adequately tested.

6.1.3 Temperature Switch Problems Eleven LERs related to temperature switches in the RWCU isolation subsystem have occurred to date. These LERs may be briefly summarized as follows:

a. LER 85-025-01 Event Date - 6/14/85 DER 85-0301 Two isolations occurred due to spurious electrical signals generated during the reinstallation of fuses in the steam leak detection system.
b. LER 85-027 Event Date - 6/17/85 DER 0303 The outboard isolation valve closed on a spurious signal.
c. LER 85-028 Event Date - 6/27/85 DER 85-0325 A spurious temperature detection system trip resulted in closure of the inboard isolation valve.
d. LER 85-031 Event Date - 7/3/85 DER 85-0340

A spurious signal from a high dif ferential temperature switch caused an outboard isolation valve to close.

e. LER 85-050 Event Date - 8/14/85 DER 85-0432 Spurious actuation of a temperature switch resulted in an inadvertent isolation.
f. LER 85-061 Event Date - 9/24/85 DER 85-0508 During a surveillance test, the technician lifted the incorrect leads resulting in a RWCU isolation.
g. LER 85-064 Event Date - 10/24/85  ;

Switching of battery chargers may have caused an electrical transient which initiated the temperature detection circuitry.

h. LER 86-003 Event Date - 3/11/86 DER 86-0110 1&C technician lifted the incorrect lead causing an RWCU isolation.
1. LER 86-034-01 Event Date - 10/7/86 DER 86-0141 Due to a transient induced by operating the read / set switch on the temperature monitors, a RWCU isolation occurred. This event occurred following the addition -f a .5 second time delay relay in the steam leak detection circuitry.
j. LER 86-046 Event Date - 12/19/86 DER 86-0222 f l

The differential temperature detection system was inoperable due to the thermocouple leads being landed on the incorrect terminals.

k. LER 87-001 Event Date - 1/26/87 DER 87-026 An auto isolation occurred during calibration of the differential  !

temperature switches in the steam leak detection system. Poor communication j between technicians and operators was the root cause of the problem.

As observed from the preceding items, the majority of the isolations caused .

by temperature monitors were associated with spurious signals, although person- I nel errors were not insignificant. It should be noted that system isolations caused by spurious signals from the temperature monitoring portion of the steam leak detection subsystem are not unique to Fermi 2. IE Information Notice 86-69 issued on 8/18/86 summarizes an AEOD report (AEOD/E604, 3/14/86). This report, which was reviewed by BNL, describes 31 LERs from various facilities between 1/84 and 11/85 which are associated with spurious system isolations caused by inadvertent operation of Riley Thermocouple Monitors, the same type used at Fermi 2. While the report expresses concern regarding the isolation of the HPCI and RCIC systems due to temperature monitoring problems, the recommendations apply to the RWCU system as well. The major report recommendation is to install

- l l

a time delay in the steam levk detection circuitry. This suggestion is reiter-ated in the Information Notice and has been implemented at Fermi 2 as described in the corrective actions section which follows.

Corrective Actions 1

A number of corrective actions have been taken to minimize the potential j for inadvertent RWCU isolations due to temperature switch circuitry problems.  ;

These include the following: I

1. EDP 1860 was completed on 8/27/85 to supply the RWCU, HPCI, and RCIC tem-perature modules f rom inverters (B21-K601 A&B). This change was based on .

G.E. FDDR KH1-1104 and recommended in G.E. SIL 416.

2. EDP 4154, which was completed on 11/15/85, added a 0.5 second time delay to j the temperature isolation logic. Agastat relays A71B-K64 & 65 were instal- j led in panels H11-P622 and P623. This important modification was recom- j mended by G.E. and by NRC-AEOD as a measure to minimize spurious trips; although it should be noted that a 1.0 second time delay value was recommended.
3. EDP 5561 was cocpleted on 3/26/87 and was associated with installing test and disconnect terminal blocks for the thermocouple extension terminations in H11-P614. This modification permits easier access to the temperature monitors for testing and repair requirements.
4. EDP 4641 was completed on 3/19/86 to verify proper temperature monitor grounding in panel H11-P614 and to install surge suppression resistors across the Read / Set switches per FDDR KH1-1154.
5. Administrative changes have been made to minimize RWCU isolations due to personnel errors associated with the temperature monitoring system.

In addition to these completed modifications, further action is still being considered including changeout of the equipment to an improved model, increasing the time delay to 1.0 second, and/or inspecting and modifying the internal moni-tor circuitry. This latter item is addressed by PDC-5468 dated 1/24/87 which references C.E. FDI WHLY. The addition of a Normal / Test Logic Switch identical to the HPCI and RCIC systems is also scheduled.

BNL Assessment Based on the detailed review of the temperature related LER9 and its asso-ciated documentation, it is recommended that the following LERs be closed: l 85-025 85-027 85-028 85-031 85-050 85-061 85-064 86-003 86-046 87001 l Closure of these LERs is based not only on the hardware improvements des-  !

cribed earlier, but also on important administrative cetions taken as well. As part of the temperature monitoring review, the following procedures were ac-ceased for adequacy and completeness.

{

i

i

]

1. 2 3.424, Rev. 5, 8/9/86, Reactor Building Steam Leak Detection System i
2. 43.707.01, RWCU Leakage Monitoring Test i
3. 44.020.156, Rev. 11, 1/27/87; NSSSS-RWCU Area, Area Ventilation Differen-tial, and NRHX Temperature Divisioni, Functional Test
4. 44.020.158, Rev. 6, 1/27/86, NSSSS-RWCU Area, Area Ventilation Differen-tial, and NRHX Discharge Temperature, Division 1, Calibration.

These procedures were found to adequately address the operation and testing of the system including updating the procedures to reflect design modifica-tions. It is recommended that LER 86-034 remain open at this time pending the j Licensee's evaluation of an increase in the time delay from .5 to 1.0 seconds. j It should be noted that a modification to the modules is also planned pending

]

the availability of parts. i I

6.1.4 Mechanical Problems Four LERs were generated due to mechanical failures experienced with RWCU which resulted in actuation of the containment isolation valves. These LERs are 86-021,86-049, 86-050, and 87-005, and may be summarized as follows:

LER 85-021 Event Date 7/20/86 DEk 85-005 During the performance'of a response time test, the RWCU system isolated due to '

a faulty test cable installed into the system logic.

LER 86-049 Event Date 12/27/86 DER 86-231 The RWCU system isolated on high differential flow due to a failure of the main flange gasket in the "A" Filter / Demineralized.

LER 86-050 Event Date 12/30/86 DER 87-007 I A manual isolation of the RWCU cystem was initiated when water leakage was detected from the service air header. This occurred due to leakage past the high/ low pressure interface associated with the Filter / Demineralized backwash subsystem.

LER 87-005 Event Date 2/12/87 DER 87-058 A manual isolation of the RWCU system was initiated due to high dif ferential flow experienced when the main gasket or the "B" Filter / Demineralized failed.

Corrective Actions To allevirte these problems, a nusber of corrective actions have been taken including the following:

1. Administrative action was taken to inspect and test any cables used for the performance of surveillance tests. In addition, the 16C Policy Letter (page 6 of DER 86-005) requires identification and special storage to pre-clude using the incorrect cable. (LER 86-021)

, 2. Use of an improved gasket design has been recommended by the manufacturer and has been implemented by the Fermi Unit #2 maintenance personnel.

Unfortunately, the gasket change-out was not employed on the "B" Filter /

Demineralized in a more timely manner following the "A" Filter /Deminerali-zer gasket failure.

3. Incorporated the correct gasket part number and description into mainte-nance instruction MI-M173 entitled Reactor Water Cleanup, Fuel Pool Cooling and Cleenup Systems Filter / Demineralized Element Replacement.

4 Modified procedure SP23.707 to add a CAUTION statement in Section 4.10 (Backwashing Filter / Demineralized) to require that the Filter / Demineralized pressure be less than the backwash air supply pressure to prevent possible water backflow into the Station Air System. (LER 86-050)

5. Initiated a potential design change (PDC-6949) to install manual isolation valves and check valves to isolate the Filter / Demineralized f rom the ser-vice air header. (LER 86-050)

BNL Assessment With the improved gasket design in the Filter / Demineralized and the revised procedure being employed to assure proper gasket alignment, it is expected that premature gasket failure will not occur. Therefore, it is recommended that LERs86-049 and 87-005 be closed. -

Proper corrective action was implemented regarding the defective test cable employed during response time testing which resulted in an inadvertent RWCU iso-lation. Therefore, it is recommended that LER 86-021 be closed.

While the procedural changes made to 23.707 will minimize the potential for leakage from the F/D to the service air header, it is recommended that LER 86-050 not be closed until ad/itional isolation capability (isolation and/or check valves as suggested in PDC-6949) is installed. While an approved PDC was avail-able on 5/20/87, a corresponding EDP, along with an implementation schedule, remains open at this time.

6.1.5 Design Problem LER 86-012 addressed a design deficiency concerning the ability of the isolation detection system to actuate in the assumed time given a single fail-ure. The history of the subject of this LER actually extends from 7/85 when DER 85-0404 was initiated to evaluate a potential deficiency in the RWCU high energy line break analysis, to 2/10/87 when the QA signof f of DER 86-0183 occurred.

DER 86-0183, issued on 5/8/86, determined that corrective action was necessary  ;

to alleviate this design problem and resulted in the issuance of LER 86-012 and -

EDP 5702. The design change added a thermocouple to each RWCU pump room, the heat exchanger room, and the phase separator tank room and was completed on 9/4/86.

3 i

In order to assess the safety implications of this design deficiency, it is necessary that additional information be provided. This could include an area l temperature profile assuming various isolation times, and the impact that the l break has on equipment operation.  !

6.2 Surveillance Procedure Review

Background

A review of selected surveillance and calibration procedures at Fermi-2 was conducted to determine the technical adequacy of the procedures as well as the administrative controls in effect. The sample of procedures included the fol-lowing:

l

1. 44.010.001 PPS-Reactor Steam Dome Pressure Functional Test
2. 44.010.05 RPS-Reactor Steam Dome Pressure Calibration
3. 44.010.009 RPS-Reactor Steam Dome High Pressure Response Time Test 4 44.010.013 RPS and NSSSS-Reactor Vessel ~ Low Water Level Functional Test
5. 44.010.17 RPS and NSSSS-Reactor Vessel Low Water Level Calibration .
6. 44.010.21 RPS and NSSSS-Reactor Vessel Low Water Level Respense Time Test )
7. 44.010.227 NSSSS-HPCI and RCIC Room Area Temperature Functional Test
8. 44.020.156 NSSSS-RWCU Area - Functional test
9. 44.020.158 NSSSS-RWCU Area - Calibration
10. 44.020.007 NSSSS-Reactor Vessel Low Water Level Calibration In addition to these procedures selected, the senior resident inspector requested that BNL review two recently issued procedures. These proceduree are a new foreat which is being applied in a systematic backfit program to all sur-veillance procedures. They are:
1. 44.030.082 Accident Monitoring Reactor Vessel Water Level Calibration, Div. 1
2. 44.030.083 Accident Monitoring Reactor Vessel Water Level Calibration, Div. 2 Discussions with a licensee representative revealed that over the next year all I&C surveillance procedures would be revised to reflect the format of these 2 procedures.

BNL Assessment l A number of changes in format were apparent from the review of the two accident monitoring vessel level calibration procedures, including the follow- 1 ing:

a. Data is filled in and steps are initialed adjacent to the steps where the work is performed. This is an improvement in that it minimizes the poten-tial for error in transferring information from the body of the procedure j to attachments which is presently done.  !

l

b. The steps necessary for calibrating the instrumentation being tested is {

included in the procedure. While this makes the procedure more voluminous, l it does provide the technician with all of the information required to test and adjust the instrumentation if necessary.  !

l

c. A loop diagram sketch has been added illustrating the relay logic being l tested by the procedure, such as interlocks and indication. This is a valuable aid to operations as well as I&C since it provides a quick graphic indication of the logic being tested.
d. An impact statement which summarizes the purpose of the procedure, the {

plant systems and annunciators affected, and possible technical specifica-tion LCOs has also been added. As with the loop diagram, this addition provides Operations and I&C with required information, and avoids errors that could occur if an evaluation had to be performed each time the proce-dure was performed.

The review of the RPS, RWCU, and NSSSS procedures resulted in a small num-ber of comments being generated. With the exception of the first comment on 44.030.082, all of these comments were resolved as of 5/22/87.

44.010.01, Rev. 5 6/25/86 Calibration Attachment 1: Trip and Reset setpoints for B21-N678A should be given in milliamps not volts.

44.010.013, Rev. 3 2/17/87 Functional Test Minor typos on pgs 2 & 4 of Attachment 1, "43" instead of "44".

10 minute warmup time not required for readout assembly.

44.010.17, Rev. 4 2/13/86 Step 6.2.16 - with trip current knob counter-clockwise, don't think trip LED will go out. Same with 6.2.20.

44.010.21, Rev. 3 8/6/86 This procedure tests logic associated with Ell and Gil - why isn't this logic tested in functional procedure (44.010.013). See Step 6.4.

i i

44.030.082, Rev. 6 2/21/87 Testing includes impact on K14A relay which af fects containment spray ;

operability - not mentioned in procedure (Step 3.0 in impact state- l ment?). 1 6.2.3.6 should also require notification of shift supervision if as found reading exceeds certain values.

Step 6.3.3.4 Typo-Enclosure A not enclosure 1.

l l

Recommendations One general feature of the present procedures which is not altered in the new format being implemented is the manner in which the calibration test is per-formed. The transmitter and trip unit, which are the primary components of the instrumentation loops, are calibrated separately in the surveillance procedures. .

It is recommended that the licensee consider the performance of a loop check at I the conclusion of the individual calibrations to verify that loop integrity and accuracy is intact. This could be accomplished by providing a known input to the transmitter and verifying that the trip function occurs at the desired set-point.

I l

In addition, the review and epproval process for the surveillance pro- {

cedures would be more meaningful if the data sheets include equivalent pressure, power, level, flow, or temperature limits rather than strictly the electrical parameters (millivolts /milliamps) which are present.ly listed. l l

6.3 Motor Operated valve (Mov) Program Review 6.3.1 Background Motor operated values (MOVs) are employed in safety related applications throughout the plant. The motor operators are fairly compirm devices which have had their share of problems as witnessed by the number 01 ,oformation notices published on the subject.

BNL conducted a review of the MOV maintenance program at Fermi in order to assess the technical content of the maintenance procedures, that proper training of maintenance personnel is conducted, that preventive maintenance has been established, and that lessons learned from IE Notices and Bdlletins have been applied. This section of the report discusses the status of each of these areas and makes recommendations for future inspection activities.

6.3.2 Maintenance Procedures Twelve maintenance procedures covering the inspection, cleaning, disassem-bly, lubricating, assembly, adj us ting , and testing of the seven types of Limi-torque and Rotork valve motor operators used at Fermi Unit 2 vere briefly reviewed for technical content and accuracy. These detailed procedures included

l maintenance activities associated with the gear assembly, limit and torque switches, and the motor. They contained complete and accurate information and, addressed the probleets that have been experienced in the industry. These procedures incorporate the descriptive information from the manufacturer's instruction and maintenance manuals. Therefore, no recommendations are made in this area.

6.3.3 Training The training instructors responsible for implementing MOV training were interviewed during the course of the inspection. They appeared knowledgeable of the operation and maintenance of the valve operators, and provided BNL with copies of the training lescon plans that had been developed. They also des-cribed the test facility being constructed to provide hands-on training.  ;

{

Four training lesson guides for the Limitorque and Rotork valve operators, I including laboratory exerciees for each type, were reviewed. These lesson guides were prepared using information supplied by the manufacturers and provide j a detailed description of actuator operation. The lesson guides refer to appro- i priate IE bulletins, Notices, and INPO documents, and properly incorporate industry wide experiences into the training.

While the program appears to contain the ingredients necessary to make it effective, it must be noted that at the time of the inspection, no training using the reviewed lesson guides had actually bee'n conducted. Training had been conducted on-site during the preoperational testing phase by a vendor, Power Safety International. An informa t ion copy of the electrical section of the training publication was found to contain a representative sampling of the imrortant information.

Followup in this area is recommended to assure that the planned training is implemented. According to the maintenance training supervisor, this should begin in August, 1987.

6.3.4 Preventive Maintenance (PM) Program i The MOV PM Program, including a sample of specific PM scheduling sheets, was reviewed with the licensee. The PM Program established for MOVs is still in the early implementation stages, however, the basic philosophy being followed is that all safety related MOVs will be inspected on a 2 year frequency with the exception of in-containment valves and local leak rate tested (LLRT) valves, which are maintained every 18 months and 12 months, respectively.

The goal of the PM coordinator is to schedule MOV maintenance during planned system or subsystem outages. While this should improve system availa-bilities and reduce operations work load, it appears f rom several of the exam-ples reviewed, that significant MOV inspection delays may be encountered. For instance, MOV Ell-F008 (RHR) was scheduled to be maintained on 4/30/87, however, maintenance has been deferred until 12/31/87 to coordinate with an LLRT outage.

Likewi s e, MOVs E21F015B ' Sre Spray) and G33F004 (RWCU) were scheduled for 3/31/87, but have been posto until 12/31/87 for this same outage.

Additional experience with the program implementation must occur before a thorough evaluation can be made. However, based on the maintenance procedural l

content, the PM program commitment (SP12.000.17, Preventive Maintenance Pro-gram), and the samples of the TM scheduling sheets reviewed, it is concluded' that the necessary ingredients for an effective MOV PM Program exist.

Two concerns presented at the exit meeting on 5/22/87 are that:

1. Technical justification be required whenever there is a deferral of PM (see-examples). Deferring PM for 8 or 9 months for schedule " convenience" ' is insufficient.
2. As found and as left conditions are properly documented. This allows trending of parameters as data is accumulated.

6.4 Post Accident Hydrogen /0xygen Analyzers 6.4.1 Background The hydrogen / oxygen monitoring system consists of two redundant sets of instrumentation and equipment that monitor the hydrogen and oxygen levels in the primary containment during normal operation and accident conditions. The system is required -to operate under environmental conditions varying from normal operating temperature and humidity to extreme conditions associated with loss of coolant accidents. Testing of this equipment installed at Fermi-2 has been in progress since 1985, and a number of failures have been encountered, particular-ly with the censing units that determine the hydrogen and oxygen concentrations.

These sensors, supplied by Exosensor, Inc. are el'ectrochemical cells that gener-ate an electrical current that is directly proportional to the partial pressure of hydrogen and oxygen in the atmosphere. A review of the Deviation / Event Report file at Fermi-2 indicated that at least eight failures have been identi-fled with these sensors over a period of less than 18 months. In certain cases, the failed equipment has been returned to the manufacturer for a failure analy-sis and repair. Four failure analycis reports, prepared by Exosensor, Inc.,

were obtained and reviewed by BNL. In each case, the sensor returned to Exo-sensor had no electrolyte present; however, two sensors had been opened at Fermi after the failure and, thus could not be analyzed by Exosensor. The remaining two sensors were tested in the "as received" condition and did not operate. In addition, one failure report indicated that there was evidence of electrolyte leakage from a censor seal. All four sensors were reconditioned, filled with electrolyte (25% sulfuric acid solution), tested satisfactory, and returned to Fermi. Exosensor has proposed that the main cause of sensor failure is the eva-poration of the electrolyte which was enhanced by relatively low humidity and high temperatures at the sensors location. The elevated temperatures were the result of operation of the heat trace system, included in the design of the analyzer, at higher than expected temperatures.

Corrective Action As documented in correspondence f rom Exosensor to Fermi-2, an action plan was developed to close this issue.

1. Identification of the primary path of electrolyte loss through a series of tests.

J

, 2. Investigation of possible design changes to reduce the permeability of the sensor.

3. Reduction of operating temperature of the Fermi heat trace system.
4. Inspection (by Exosensor) of the sensors installed at Fermi on a 6 to 9 month cycle.
5. Verification of the sensor electrolyte level prior to shipment from Exosensor to Fermi.

Item 5 has been included in the plan since the sensors are subjected to a prolonged high temperature acceptance test prior to shipment. It is uncertain whether this final test is destructive since electrolyte levels could easily be affected. With only 5 cc of electrolyte volume, any loss could shorten the expected sensor life.

Several of the other items identified above are being progressed as fol-lows:

Item 3 has been implemented at Fermi-2 as of 3/23/87. Initially, the heat trace temperature setpoint was 130*F, which was determined by DECO to be the setpoint necessary to assure no substantial moisture buildup in the sensing lines. Based on the possibility of electrolyte evaporation, the setpoint has been reduced to 110*F, although it should be noted that actual temperatures as witnessed by the BNL inspector on 3/25/87 varied from ll7*F to 120*F. The test engineer indicated that small adj ustments will continue to be made until the final setpoint is achieved.

As part of Item 4, Fermi-2 has committed to replacing the sensors every 6 I months and returning them to Exosensor for evaluation. In addition, Fermi-2 has increased the instrumentation calibration frequency from quarterly to monthly.

ENL Assessment )

i BNL has also investigated several other areas related to the application of l Exosensor equipment at Fermi and other nuclear plants.

1. Licensee Event Reports (LERs)

The computerized LER data base at Oak Ridge National Lab. (ORNL) was queried to determine if other failures of Exosensor equipment had been reported. The only LER related to Exosensor reported in the ORNL system (which includes LERs from 1981 to the present) involves failures of the oxygen analyzers at Fitzpatrick. These failures were both procedural and sof tware related and at least one was reported under 10CFR21. Additional queries of the data base resulted in a number of hydrogen and oxygen analy-zer failures but none of these were identified as being related to Exosen-sor equipment.

.2. Vendor Program Branch inspection at Exosensor In May 1985, the Vndor Program Branch of t.he NRC Of fice of Inspection and Enforcement conducted an inspection at the Exosensor facility. This i inspection resulted in eleven nonconformances that. were mainly directed at )

deficiencies in Exosensor's quality assurance program. In addition, one j violation was issued due to the lack of appropriate procedures for evaluat-  !

ing deficiencies for deportability under 10CFR Part 21 or reporting the l deficiencies to customers for evaluation. BNL has obtained and reviewed l the Exosensor's responses to this inspection report, and these indicate l that a number of corrective actions are being implemented. j SNL reviewed the Fermi-2 qualification documents for the sensors. These documents indicated that the sensors were required to remain operable for 100 days following the DBA.

3. The recurring problems with the H2 and 02 sensors may constitute a report-able defect under 10CFR21. Exosensor has apparently determined that this problem is plant specific and therefore not subj ect to a Part 21 notifica- I tion. At the time of the inspection, Fermi-2 was evaluating their respon-sibility as far as deportability is concerned. The licensee agreed to provide by 6/30/87 an evaluation for Part 21 applicability which addressed all sensor failures. An interim report describing the sensor problems would be available by 6/16/87.
4. 16C surveillance procedure 44.120.028, " Post Accident Monitoring-Drywell Hydrogen /0xygen Analyzer Calibrat. ion Division I" includes a calibration test. of t.he analyzer equipment. If discrepancies are noted in the measured gas concentrations, the procedure requires that the Operational Support Engineer be contacted to make modifications t.o the software. However, t.he procedure does not provide any instruction on how the software is to be modified or any guidance to assist in determining if the discrepant read-ings were sof tware related or caused by another system malfunction. This procedure should be expanded to insure thst. software modifications are not made to change the analyzer system out. puts when other parts of the equip-ment are not operating properly. In addition, procedural steps should be provided such that any software modifications are documented and reviewed.

The licensee agreed to incorporate t.he necessary steps for the above adjustments as well as guidance for modifying the sof tware to verify proper functioning of the associated annunciators.

5. A review was made of the H7 /02 analyzer surveillance procedures conducted in March, April, and May 1987. Procedures 44.120.028 and 44.120.029 Lest t.he division 1 and division 2 channels respectively. In general, these procedures demonstrat ed system operability, but were not performed in accordance with technical specification administrative requirements, sections 6.8.1 and 6.8.2. On two occasions, the procedure was modified j without an approved permanent or temporary procedure change. Instead, a note on the surveillance performance form (spf) for the test conducted on 5/3/87 indicated that for steps 6.15.16 to 6.15.27 the " software for the K4 l

l.

1

. value was changed so that the annunciator check could be performed." The spf for the surveillance conducted on 4/9/87 also had the same notation, again without an approved procedural change.

Recommended Followup 1

Considering the number of f ailures of the Exosensor equipment at Fermi, and I the concerns that are not completely resolved at this time, the operation of the hydrogen / oxygen monitoring system should be scrutinized carefully in the future.

The following items are considered important:

1. Failed Sensor Evaluation Reports: Exosensor has provided evaluations reports on four sensors that Fermt has returned. Two additional sensors have been returned and the evaluations of these failures should be reviewed. j i
2. 10CFR21 Reporting: The determination of the need to report these sensor failures under 10CFR21 should be reviewed upon completion of the Fermi evaluation.
3. Heat Trace Temperature Design Change: The effect of the heat trace temper-ature reduction on the performance of the sensors should be monitored.

This should be done in conjunction with the review of the results of the evaluations that Exosensor will perform following the sensor replacement every six months. If there is evidence that the heat trace temperature reduction improved the sensor performance, consideration should be given to continuously monitoring the temperature and humidity in the vicinity of the sensors.

4. 16C Surveillance Procedures 44.120.028 and 44.120.029 should be expanded to provide the necessary guidance and procedural steps to perform any software modifications that are required. The licensee also indicated that a "47" series procedure would be revised to include adj us tment s that may be necessary to the equipment. These steps are presently included in the vendor manual.
5. The results of the surveillance testing that Fermi has committed to perform )

monthly on the sensors should be monitored closely as part of the evalua-tion of the sensor performance following the heat trace system temperature reduction. '

6.5 Plant Inspection As part of the overall review process, the equipment associated with the ,

post accident monitoring instrumentation (Panels P376 and 377) the testability j package for surveillance procedures, and the RWCU system instrumentation were J inspected. Equipment was in working order and no deviations were noted with the ]

following exceptions:

1

1. On Panel 614, six of the Riley temperature monitoring modules (3 on each division) contain handwritten identification numbers which differ from the metal tags attached to the modules. The licensee indicated that the metal

tag referred to the thermocouple while the handwritten number was the o module referred to in t.he surveillance procedures. It is recommended that permanent labels be affixed to the modules to avoid potential confusion during testing and calibration. (See LER 85-061 as an example.) A work request (#003A) was initiated on 5/22/87 to install new labels.

2. Trip unit labels on panels H21-P080, 82, 84, and 86 should be completed.

Several RPS and ECCS trip units are identified by a laminated tag hung by Operations which includes the function of the trip unit. Consideration should be given to approving " Operator Aids" of this type for all of the I trip units on the RPS and ECCS panels. I

{

3. A deficiency tag (#11467) dated 10/31/85 on the division 2 RPS testability panel indicates that the calibration unit readout assembly was not working i properly. The licensee verified that on 4/21/87 this work request was can-celled based on the fact that the problem was no longer being experienced.

This item is therefore closed.

1 l

1

..

Retrieved from "https://kanterella.com/w/index.php?title=ML20234C959&oldid=3194478"