ML20215J253

Forwards Suppl 1 to EGG-NTA-7296, Supplemental Technical Evaluation Rept,Pump & Valve Inservice Testing Program, TMI-1. Proposed Program in Accordance W/Established NRC Requirements W/Noted Exceptions

ML20215J253
Person / Time
Site:	Three Mile Island
Issue date:	02/17/1987
From:	Obenchain C EG&G IDAHO, INC., IDAHO NATIONAL ENGINEERING & ENVIRONMENTAL LABORATORY
To:	Carrington M Office of Nuclear Reactor Regulation
References
CON-FIN-A-6812 EGG-NTA-7296-S01, EGG-NTA-7296-S1, OBEN-23-87, NUDOCS 8705070316
Download: ML20215J253 (2)

Text

Po/L

/NEA idaho National Engineering laboratory February 17,1987 Mayo Carrington, Project Manager U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Technical Assistance Management Branch Planning and Program Analysis Staff 7920 Norfolk Avenue Bethesda, MD 20014 TRANSMITTAL 0F FINAL REPORT, " SUPPLEMENTAL TECHNICAL EVALUATION REPORT, PUMP AND VALVE INSERVICE TESTING PROGRAM, THREE MILE ISLAND NUCLEAR STATION, UNIT 1" (FIN A6812), EGG-NTA-7296, FEBRUARY 1987 - Oben-23-87 Ref: (a) NRC Form 189, " Review of Pump and Valve Inservice Testing Programs for Operating Plants (III)", (FIN A6812)

(b) C. F. Obenchain ltr to Mayo Carrington, Oben-109-86. Transmittal of Final Report, " Technical Evaluation Report, Pump and Valve '

Inservice Testing Program, Three Mile Island Nuclear Station,  !

Unit 1" (FIN A6812), EGG-EA-7296, September 1986, September 25, 1986

Dear Mr. Carrington:

The subject report is transmitted in fulfillment of the Three Mile Island, Unit 1 (TMI-1), IST program review identified in the referniced 189. The report is a supplement to the report transmitted by reference (b) and l addresses the response of TMI-1 to the Required Action Items identified in I the original report. This work is being conducted under the technical 1 direction of Mr. C. G. Hammer of the Office of Nuclear Reactor Regulation, Engineering Branch, Mechanical Engineering Section of the Division of PWR Licensing B.

This report documents our review of the Inservice Testing Program submitted by General Public Utilities Nuclear Corporation for Three Mile G705070316 870217 PDR ADOCK 05000289 P PDR hEGcGon.... P.o. sox tas idaho Faits, to estis

.s

[ Mayo Carrington February 17, 1987 Oben-23-87 "

Page 2 Island, Unit 1, and concludes that, with the exceptions noted therein, the

proposed program is in accordance with the established Nuclear Regulatory Commission requirements.

l Very truly yours,

. C. F0(sv4*E1 C. F. Obenchain, Manager A5  ;

NRR and I&E Support  !

TLC:ggo i

i

Enclosure:

As Stated l cc: C. G. Hammer, NRR/PBEB P. T. Kuo, NRR/PBEB ,

2 G. L. Jones, DOE-ID j J. O. Zane, EG&G Idaho (w/o Enclosure) l l

I i

i 5

1 i t

)

I l

i 1 I ._ - _ _ _ - _ _ ______ _ ___ _ _____ _ - _ _ _ _ . _ _ _ _ _ _ _ _ _ - _ _

EGG-NTA-7296 Supplement 1 February 1987 INFORMAL REPORT

/daho SUPPLEMENTAL TECHNICAL EVALUATION REPORT, PUMP Notional AND VALVE INSERVICE TESTING PROGRAM, THREE MILE DyinOOr/ng ISLAND NUCLEAR STATION, UNIT 1 thoratory Managed T. L. Cook by tho U.S. H. C. Rockhold ppartment ofEnergy l

i l

l 4D # "**

Prepared for the we- o,,a,, U. S, NUCLEAR REGULATORY COMMISSION DOE Contract No. DE-AC07 76tD01570

e 0

DISCLAIMER This book was prepared as an accourit of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees. makes any warranty. express or implied, or assumes any legal haMty or resconsibihty for the accuracy, completeness, or usefuhms of any information, apparatus, product or process disclosed, or represents that its use would not ininnge pnvately owned nghts. References herein to any specihc commercal product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessanty constitute or imply its endorsement, recommendation, or favonng by the United States Govemment or any agency thereof. The views and opinions of authors empressed herein do not necessanty state or reflect those of the United States Government or any agency thereof.

O 9

EGG-NTA-7296 Supplement 1 p

SUPPLEMENTAL TECHNICAL EVALUATION REPORT PUMP AND VALVE INSERVICE TESTING PROGRAM THREE MILE ISLAND NUCLEAR STATION, UNIT 1 Docket No. 50-289 T. L. Cook H. C. Rockhold Published February 1987 Idaho National Engineering Laboratory EG&G Idaho, Inc.

Idaho Falls, Idaho 83415 Prepared for the U.S. Nuclear Regulatory Commission Washington D.C. 20555 under 00E Contract No. DE-AC07-761001570 FIN No. A6812

O s A8STRACT This EG&G Idaho, Inc., report presents the results of our evaluation of the licensee's response to the " Required Action Items" summarized in the NRC staff's Inservice Testing Safety Evaluation Report, Appendix D, dated October 3, 1986, for the Three Mile Island Nuclear Station, Unit 1 -

Inservice Testing Program for safety-related pumps and valves.

n FOREWORD This report is supplied as part of the " Review of Pump and Valve Inservice Testing Programs for Operating Plants" Program being conducted for the U.S. Nuclear Regulatory Commission, Office of Nuclear Reactor Regulation 01 vision of Engineering, by EG&G Idaho, Inc., NRR and I&E i Support.

4 i

j The U.S. Nuclear Regulatory Commission funded the work under the i authorization 8&R 20-19 40-41-2, FIN No. A6812.

I s

i r

Docket No. 50-289 1

Il

CONTENTS 1

A8STRACT .............................................................. 11 FOREWORD .............................................................. 11 (S) 1. INTRODUCTION .................................................. 1 l

(S) 2. PUMP TESTING PROGRAM .......................................... 3 (S) 2.1 Control Building Chilled Water System ................... 3 4

(S) 2.2 Boric Acid System ....................................... 4 (S) 2.9 Screen House Ventilation System ......................... 8 J

! (S) 2.10 Boric Acid Recycle System ............................... 9 (S) 2.11 Makeup and Purification System .......................... 12 (S) 3. VALVE TESTING PROGRAM ......................................... 15 (S) 3.4 Decay Heat Removal System ............................... 15 j (S) 3.4.3 Category C Valves ............................... 15 (S) 3.6 Feedwater System ........................................ 17 (S) 3.6.1 Category C Valves ............................... 17 j (S) 3.8 Main Steam System ....................................... 18 (S) 3.8.2 Category C Valves ............................... 18 l.

1 i (S) 3.9 Makeup System ........................................... 21 i

(S) 3.9.2 Category B/C Valves ............................. 21 (S) 3.9.3 Category C Valves ............................... 23 f

(S) 3.11 Reactor Building Emergency Cooling System .............. 33 (S) 3.11.1 Category B Valves .............................. 33

, (S) 3.12 Reactor Building Spray System .......................... 33 (S) 3.12.1 Category C Valves .............................. 33 i *

(S) APPENDIX 0 ........................................................ 37 tit

SUPPLEMENTAL TECHNICAL EVALUATION REPORT PUMP AND VALVE INSERVICE TESTING PROGRAM THREE MILE ISLAND NUCLEAR STATION, UNIT 1 l 1. INTRODUCTION A meeting was held in Bethesda, Maryland on October 27, 1986, with

. representatives of the NRC staff, Three Mile Island, Unit 1, and EG&G, Idaho, to discuss the licensee's " Required Action Items" summarized in the NRC staff's Inservice Testing Safety Evaluation Report, Appendix 0, dated October 3, 1986. The licensee's responses to the Action Items were i transmitted to the staff by letter dated December 24, 1986, and are evaluated here using the same numerical identification system used in the SER dated October 3, 1986, except that each item number is preceded by an "S" to indicate the supplemental TER. The evaluation and discussions in i

this supplement are in the order taken from the licensee's letter of December 24, 1986. All additional information taken from that letter has been typed with single spacing in this supplement for ease of reference.

I i

i l

l <

r e

i e

1

. - . , . , . - _ - _ - _..-, . __ .m. _ _ _ _ __ _

(S) 2. PUMP TESTING PROGRAM (S) 2.1 Control Buildina Chilled Water System (S) 2.1.1 Relief Request The licensee has requested relief from measuring flow rate on the control building chilled water pumps, AH-P3A and -P38, in accordance with

. the requirements of Section XI, Paragraph IWP-3100, and proposed to install flow instrumentation prior to startup for Cycle 7.

(S) 2.1.1.1 Licensee's Basis for Reauestina Relief. GPUN accepts the staff's position on AH-P3A/8 as discussed in the October 1986 meeting. The modification to install flow instrumentation for AH-P3A/B is scheduled for completion prior to Cycle 7 operation. This is in accordance with the relief provided by the NRC's letter of October 23, 1984 (Item A-1). The SER also grants interim relief until the startup for Cycle 7 operation, but it is not clear from the wording in the SER that modifications are not being required prior to startup for Cycle 6. During the October 1986 meeting, NRC clarified that it was not their intent to require these modifications prior to Cycle 7. For documentation purposes, GPUN reauests that this be reconfirmed in a supplement to the SER.

(S) 2.1.1.2 Evaluation. The staff has agreed with the licensee's proposal to install flow instrumentation during the refueling outage prior to startup for Cycle 7. Clarification of the staff's position concerning when these modifications will be required to be completed will be provided in the cover letter transmitting this supplement.

(S) 2.1.1.3 Conclusion. The staff concludes that, for the balance of the period of the present fuel cycle, interim reitef is granted to test these pumps as proposed by the licensee. Requiring the licensee to make these modifications prior to the next refueling outage would impose unnecessary hardship on the licensee without a compensating increase in the level of safety.

0 3

b (S) 2.2 Boric, Acid System (S) 2.2.1 Relief Recoest The licensee has requested relief from quarterly testing and measuring flow rate, inlet pressure, differential pressure, bearing temperature, and -

the five minute run time on the boric acid pumps, CA-PIA and -P18, in accordance with the requirements of Section XI, Paragraphs IWP-3100 and -3500, and proposed to test these pumps during refueling outages and to calculate flow rate and inlet pressure at that time.

(S)2.2.1.1 Licensee's Basis for Requestina Relief. For CA-P1A/B, GPUN requests relief from quarterly testing and the measurement of Q, Pg, AP, and the five minute run time. The following testing will be conducted.

Testing CA-P1A/B will be conducted aach refueling interval. CA-PIA /B does not have a recirculation flow path. Without recirculation capability, the only method of testing these pumps is to inject into the reactor coolant makeup system. The resulting reactivity changes would affect plant operations adversely and would result in significant volumes of radioactive waste. .During normal operation, l testing one pump would generate 1,300 gallons of radwaste at the beginning of core life, 3,000 gallons at middle of core life, and 32,000 gallons at end of core life. This much radwaste is generated because injecting concentrated boric acid requires dilution of the RCS. For these reasons, it is impractical to test the subject pumps during operation. The appropriate test interval is each refueling due to the large quar.tities of liquid radwaste generated. ,

Flow rate (Q) will be calculated using makeup tank level change over

• time, since installed flow measuring instruments do not exist. Inlet

~

pressure (P ) is not available. There are no existing pressure gage g

taps. Since AP is not calculated as stated below, there is no benefit to calculating Pg based on suction tank level.

4

For CA-P1A/8, pump differential pressure (AP) will not be calculated since CA-PIA /B are positive displacement pumps. Relief is requested from the ASME Section XI, IWP-3110, requirement to calculate AP for CA-PlA/8. The calculation of AP for CA-PlA/B would not be meaningful since the flow rate is fixed solely by the displacement.of

, the cylinder and the speed of the pump, both of which are held-

~

constant, while AP is only a function of system resistance

. (backpressure). Therefore, such a calculation would be an unnecessary exercise for the operator.

ASME Section XI, IWP-3500, states that pumps under test should be run for at least five minutes under conditions as stable as the system

permits prior to taking data. To minimize radioactive waste, the subject pumps will be run until the system has stabilized and then 4

data will be recorded. GPUN believes this meets the intent of IWP-3500.

In conclusion, the CA-PlA/B test each refueling verifies the required flow rate for CA-PIA /B while pumping to the makeup tank. This ,

demonstrates that CA-PlA/B can perform its safety function by providing the required boric acid capacity at normal makeup tank ,

e backpressures. For this test, the makeup tank will be pressurized to its normal operating range. Pump vibration will beimeasured while pumping to the nakeup tank. ,

It is GPUN's position that the plant modifications to install recirculation piping and instrumentation which would be required in order'to test CA-PlA/B in accordance with Section XI on a quarterly frequency would not result in a substantial increase in the overall protection of the public health and safety over that afforded by the refueling interval test program. The primary source of emergency boration is from the borated water storage tank (BWST). The safety related function of CA-P1A/B is to provide one of the alternate sources of emergency boration. That is, to provide borated water from the boric acid mix tank (BAMT) to the RCS via the makeup' tank (MUT).

During the October 1986 meeting, the NRC suggested the GPUN investigate the feasibility of using other flow paths to test CA-PIA /8 as an alternative to plant modifications. Specifically, NRC suggested that batch volumes pumped 1

o 5

(

from the BAMT to the 4% tank might allow quarterly testing of the CA-P1A/8. GPUN has determined that pumping to some alternate tank (other than the makeup tank) is not practical for the following reasons:

1. Pumping to the spent fuel pool or BWST would require large i volumes to detect a level change for determining flowrate. Also,  ;

moving large volumes from the BAMT would, by level decrease, disqualify this tank as the T.S backup emergency boration source tank. -

2. Pumping to the 4% tank presents a high potential for disabling ~

the design function of this tank. Because the 4% tank is designed for core flood tank makeup, the boric acid concentration l 1s much less than the concentration of the 8AMT. Also, a test-that pumps the BAMT to the 4% tank may cause' crystallization of the solution in the 4% tank and disable our ability to provide makeup to the- core flood tanks because the 4% tank, piping, and

pump are not heat traced.

3. Batch pumping from the BAMT into some other container was also considered. However, there are no drain valves close to the pump, and any selected drain. remote from the pumps could provide misleading information since the observed flow could be more of a result of system draining rather than indicative of pump

. performance. Also, whatever batch size is required to perform the test, this quantity would have to be discarded as radwaste.

Therefore, we find that there is no practical flow path that will allow the measurement of CA-P1A/8 flow rate on a quarterly test frequency.

In a memorandum to J. F. Stolz dated April 8, 1982 MEB addressed the relief that had been requested for CA-P1A/8. .The memorandum stated that relief should be granted for Pg, bT , and refueling interval testing, but the licensee should calculate AP and flowrate. 'This memorandum was distributed as an attachment to the Meeting Notice of our May 15, 1984 meeting on IST. The position described in this memorandum was presented as the position of the NRC staff during our March 1984 meeting and was-transmitted to GPUN on May 29, 1984 as an enclosure to the NRC's Meeting Summary along with the transcripts. At the time of the meeting _in March 1984, GPUN was of the opinion that CA-P1A/8 could possibly be removed i from the IST program. Subsequent to the meeting in March, 1984, GPUN has accepted the inclusion of CA-P1A/B in the IST program. This meets the NRC

staff's position in 1984 except for Pg and AP will not be calculated.-

See Footnote 12 of Table A-2 for the reason for not calculating Pg and-AP. The IST program description (Attachment 2) includes'all of the meaningful testing which can be performed using the existing system design .

as requested by the NRC in the meeting of September 4-5, 1985. This is consistent with the agreement reached during the September 4-5, 1985 meeting and recorded in the transcripts on pp. 232-252. We request that -

the NRC reconsider.its current position that modifications be required and provide a Supplement SER which accepts the 1984 NRC Staff Position and.the position given in the September 4-5, 1985 meeting' transcripts. As stated i

i L

6

-in the cover letter, GPUN cannot possibly comply with the SER modification schedule as currently written. GPUN believes that the test program provided is adequate to assure the operational readiness of CA-PlA/S.

.(S) 2.2.1.2 Evaluation. The staff does not agree with the-licensee's basis and will require that system modifications be performed to permit 1 testing these pumps quarterly in accordance with the requirements of Section XI, therefore, permanent relief should not be granted from the requirements of Section XI. It is the staff's opinion that the licensee cannot assure the reasonable operational readiness of these pumps by utilizing a test program conducted at refueling outages due to ~ the. time interval between tests. The licensee will be required to install instrumentation to measure flow rate and pump discharge pressure so those parameters can be utilized to monitor pump performance and degradation.

The licensee has stated that these are positive displacement pumps, therefore, the staff feels that inlet'and differential pressure measurements are unnecessary because inlet pressure variations have no effect on positive displacement pump operation while discharge pressure and

flow rate measurements provide the necessary information to monitor the l hydraulic performance of the pump being tested. The staff also agrees that deletion of the bearing temperature measurement will not affect the licensee's monitoring program for these pumps.

(S) 2.2.1.3 Conclusion. The staff concludes that the licensee is

~

required to perform system modifications prior to the end of the next-refueling outage to allow testing the boric acid pumps in accordance with Section XI. For the balance of the period of the current fuel cycle, J

interim relief is granted to test the pumps as proposed by the licensee. l Requiring the licensee to make these modifications prior to the next

. refueling outage would impose unnecessary hardship on the licensee without l 1

a compensating increase in the level of safety.

I 4

)

7

. _ _ - _ . - . ~

(S) 2.9 Screen House Ventilation System (S) 2.9.1 Relief Request The licensee has requested relief from measuring flow rate on the-screen house ventilation equipment pumps, SW-P2A and -P28, in accordance ,

with the requirements of Section XI, Paragraph IWP-3100, and proposed to install flow instrumentation prior to startup for Cycle 7.

(S) 2.9.1.1 Licensee's Basis for Reauestina Relief. GPUN accepts the staff's position on SW-P2A/8 as discussed in the October 1986 meeting. The modification to install flow instrumentation for SW-P2A/8 is scheduled for

- completion prior to Cycle 7 operation. This is in accordance with the relief provided by the NRC's letter of October 23, 1984 (Item A-1). The SER grants interim relief until the startup for Cycle 7 operation, but it is not clear from the wording in the SER that modifications are not being required prior to startup for Cycle 6. During the October'1986 meeting, the NRC stated that it was not their intent to require these modifications prior to Cycle 7. For documentation purposes, GPUN requests that this be confirmed in a supplement to the SER.

(S) 2.9.1.2 Evaluation. The staff has agreed with the licensee's proposal to install flow instrumentation during the refueling outage prior to startup for Cycle 7. Clarification of the staff's position concerning when these modifications will be required to be completed will be provided in the cover letter transmitting this supplement.

(S) 2.9.1.3 Conclusion. The staff concludes that, for the balance of the period of the present fuel cycle, interim relief is granted to test these pumps as proposed by the licensee. Requiring the licensee to make these-modifications prior to the next refueling outage would impose unnecessary hardship on the licensee without a compensating increase in the level of safety.

W.

t i

8 I

(S) 2.10 Boric Acid Recycle System (S) 2.10.1 Relief Reauest i

The licensee has requested relief from measuring bearing temperature,

, flow rate, inlet and differential pressure, and the five minute run requirement on the boric acid recycle pumps, WDL-P13A and -P138, in

- accordance with Section XI, Paragraphs IWP-3100 and -3500, and proposed to ,

calculate inlet and differential pressure quarterly and to calculate flow rate during refueling outages.

(S) 2.10.1.1 Licensee's Basis for Requestino Relief. Pump bearing temperature cannot be measured on this pump since the bearings are located deep inside the pump casing and are surrounded by an oil reservoir.- An exception is requested per 10 CFR 50.55a(g)(5)(iii) in that measurement of parameter T s not practical within the limits of the design of this b

pump.

ASME Section XI, IWP-1100, states that emergency powered pumps should be included in the IST Program. WOL-P13A/B are not emergency powered, but l are included in the IST Program as requested by NRC. The addition of WDL-P13A/B in the program provides further assurance of the capability to supply concentrated boric acid to the reactor coolant makeup system.

For WDL-P13A/B, GPUN requests relief from the measurement of Q, Pg ,

AP, and the five minute run time. Testing will be conducted each refueling interval for-the same radwaste concerns expressed for CA-P1A/8.

Flow rate (Q) will be calculated using makeup tank level change over 4 time since installed flow measuring instruments do not exist. This test will verify the accident design flowrate.

Inlet pressure is not available (there are no existing pressure gage taps). Since AP is not calculated as stated below, there is no benefit to calculating Pg based on suction tank level.

9

I WOL-P13A/B's discharge valve is-a diaphragm operated valve that is either open or closed. There is no handwheel on the valve. Instead of throttling to a AP reference value, WDL-P13A/B will be tested unthrottled while pumping to the pressurized makeup tank. Since '

WOL-P13A/B is tested while pumping to the pressurized makeup tank, pump AP is variable (it increases over pumping time). This makes it- .

impractical to establish a AP reference value.

ASME Section XI, IWP-3500, states that pumps under test should be run for at least five minutes under con'dttions as stable as the system permits prior to taking data. To minimize radioactive waste, the refueling test will be run until the system has stabilized and then data will be recorded. GPUN believes this meets the intent of IWP-3500.

In addition to the refueling interval tests, when WDL-P13A/B is required to be operable by Technical Specifications, the following tests will be performed on recirculation quarterly:

1. P will be calculated using suction tank level.

3

2. AP will be calculated.

It is GPUN's position that plant modifications to install flow instrumentation which would be required in order to test WDL-P13A/B on a quarterly frequency in accordance with Section XI would not result in a substantial increase in the overall protection of the public health and safety over that afforded by GPUN's test program which includes certain quarterly tests in addition to the refueling interval tests. This is essentially the same issue as 0.1.2. The safety function of WDL-P13A/B is to provide another emergency boration source as backup to the BWST. That is , to provide borated water from the reclaimed boric acid tank (RBAT) to the RCS via the makeup tank (MUT).

During the October 1986 meeting, the NRC suggested that pumping from one RBAT to another would provide test capabilities to satisfy the quarterly test requirements. However, if a RBAT is being used as the -

T.S. backup boration source, the balance of concentration and volume will be adversely affected by pumping the T.S. RBAT to the other RBAT. Since this testing would disqualify the tank as a backup boration source, this test method is not considered.to be a practical alternative.

10

As discussed.in the response to 0.1.2, ME8 recommended granting the relief from quarterly test requirements for CA-PIA /B in a docketed memorandum. The NRC's position-at that time did not'specifically address WOL-P13A/8 because WOL-P13A/8 has been added to the IST program subsequent to the March 1984 meeting. The IST program description ( Attachment 2) reflects a test program for WOL-P13A/B which meets the NRC's position from 1984 son testing CA-P1A/8. The.

test program described in Attachment 2 is consistent with the

. agreement reached and recorded in the' transcripts from the September 4-5, 1985 meeting (pp. 232-252), that is, to perform all i meaningful tests that can be performed using the installed equipment.

We request that the NRC reconsider its current position that modifications be required and provide a supplement SER which accepts the 1984 staff position for pumps performing a similar function. As stated in the cover letter, GPUN cannot possibly comply with the SER modification schedule as currently written. GPUN believes that the test program we are providing is adequate to assure the operational j readiness of WOL-P13A/8.

! (S) 2.10.1.2 Evaluation. The staff agrees with the licensee's basis in part and, therefore, relief should be granted from the requirements of-Section XI to measure bearing temperature and the minimum five minute run time on the boric acid recycle pumps. Bearing temperature measurement is an unreliable method of bearing failure detection and the staff feels that deletion of the measurement will not affect the licensee's pump monitoring program. The staff agrees that the minimum five minute run time is unnecessary if the system can be stable in less time because that requirement of Section XI is to insure that system parameters will be recorded after system conditions have stabilized. The staff also agrees that calculating pump inlet pressure is an acceptable alternative to the direct pressure measurement required by Section XI.

The staff does not agree with the licensee's proposal to calculate flow rate of these pumps during refueling outages only because the licensee cannot adequately monitor pump readiness due to the time interval between tests. The staff will require that the licensee perform system

~

modifications to install flow instrumentation so pump flow rate can be recorded during the quarterly pump tests. The licensee will also be required to demonstrate operability of these pumps immediately following their being placed in their safety-related configuration in addition to quarterly testing while they are in that configuration.

, 11 4

,-r .,- , , - - _ . ~ -e,-..-,------ m , , ne, w- ,-,n, s, ,,- - - - --p , ~.n-

4 (S) 2.1'0.1.3 Conclusion. The staff -concludes that the licensee is required to perform system modifications prior to the end of the next refueling outage to allow obtaining flow rate measurements during quarterly pump testing. For the balance of the period of the current fuel cycle, interim relief is granted to test the pumps as proposed by the licensee.

Requiring the licensee to make these modifications prior to the next .

refueling outage would impose unnecessary hardship on the licensee without a compensating increase in the level of safety. - '

(S) 2.11 Makeup and Purification System (S) 2.11.1 Relief Request The licensee has requested relief from measuring flow rate on the makeup and purification pumps, MU-PIA, -P18, and -Plc, in accordance with the requirements of Section XI, Paragraph IWP-3100.

(S) 2.11.1.1 Licensee's Basis for Reauestina Relief. MU-P1A/8/C are tested using a combination of fixed' flow rate and measured flow rate. Flow through RC pump seal injection and normal makeup to the RCS is measured but flow through the minimum recirculation orifices (MU-F0-1/2/3) is held constant. MU-F0-1/2/3 are. multiple plate orifices.and the flow is held at a constant design value for the head at which the MU pumps operate.

]

Testing is conducted by throttling to a reference seal injection flow rate and a reference normal makeup rate to the RCS. Test flow rate is determined by the addition of the seal injection flow rate and normal makeup flow rate. If erosion or cavitation causes the orifice holes or plates to wear, MU-P1A/8/C's flow rate will increase, delta P will decrease, and the pump test data will fall in the alert or required action range of ASME Section XI Table IWP 3100-2. The deviation will be determined and the condition corrected per ASME Section XI, IWP-3230. This -

is conservative since work will be performed to determine the cause and

~

solution to either the pump problem or the orifice wear problem. If necessary, the orifice will be radiographed to verify degradation.

4 i

12

In addition, Surveillance Procedure 1303-11.8 verifies MU-P1A/8/C's accident _ design flow rate each refueling.

f-,

1 The SER states that the licensee has not adequately explained the-use of a reference value flow rate. The IST program description has_been revised to provide a better description of the use of the reference value flow rate for the test of MU-P1A/8/C. This confirms that the recirculation flow test of MU-P1A/8/C is truly a fixed reference flow test, as requested by the NRC'in the October 1986 meeting. The NRC

. also suggested that GPUN consider measuring stem positions for MU-V32 4

and MU-V17 to assure that the flow resistance will' remain constant from test to test. Due to the inaccuracies involved in stem position measurements, GPUN does not believe that the additional time and exposure of the. operators is warranted in attempting this technique for MU-V32 and MU-V17.

(S) 2.11.1.2 Evaluation. The staff agrees with the licensee's-basis and, therefore, relief should be granted from the requirement of Section XI-to measure total flow on the makeup and purification pumps. The licensee's i use of flow orifices in the pump minimum flow line that are fitted with multiple orifice plates should greatly reduce or completely eliminate the erosion caused when a large pressure is reduced across a single orifice plate. Because of this design feature, the flow rate in the minimum flow test path should remain constant and the summation of the three flow values

should provide repeatable data to utilize in the pump testing program.

(S) 2.11.1.3 Conclusion. The staff concludes that calculating the

flow rate of these pumps while they are operating in the test configuration l should provide sufficiently accurate information to utilize to monitor pump degradation. The staff concludes that the alternate testing proposed will give reasonable assurance of pump operability required as by the Code.

f j

13

O e

'l

{

- i f

l c

t i

14

l l

l l

(S) 3. VALVE TESTING PROGRAM

\

(S) 3.4 Decay Heat Removal System l (S) 3.4.3 Cateaory C Valves (S) 3.4.3.2 Relief Request. The licensee has requested relief from

. exercising valve DH-V14A/B, decay heat removal and reactor building spray pumps borated water storage tank suction checks, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to partial-stroke exercise these valves quarterly and to exercise them utilizing the low pressure injection flow rate during refueling outages.

(S) 3.4.3.2.1 Licensee's Basis for Reauestina Relief--Check valve DH-V14A/B supplies borated water from the borated water storage tank (BWST) to the decay heat removal pumps (DH-PlA/B) and the reactor building spray pumps (BS-P1A/B). DH-P1A/B are TMI-l's low pressure injection (LPI) pumps. Each refueling, Surveillance Procedure 1303-11.54 verifies LPI flow rate of 23000 gpm through DH-V14A/B. The accident design flow rate through DH-P1A/B is 3000 gpm and the accident design flow rate for BS-P1A/B is 1500 gpm. Therefore, SP 1303-11.54 verifies that DH-V14A/B can pass 2/3 - 3000 gpm/4500 gpm of total DH and BS flow rate. It is impractical to structure a test where 1.) a DH pump is injecting 3000 gpm from the BWST through DH-V14A/B to the reactor vessel, 2.) a BS pump is recirculating the BWST through DH-V14A/B at 1500 gpm, and 3.) the second DH pump is providing its decay heat removal function by recirculating the reactor vessel. The complexity of this evolution, time, and effort do not warrant this type of testing. Conceivably, this testing could be accomplished by filling the transfer canal part way then switching to the next loop to be tested. The risk of incorrect valve lineup leading to pump damage and hours spent

- switching loops and verifying valve lineups do not make such testing desirable. There are ALARA concerns when switching loops. An additional concern is refueling water clarity / cleanup due to crud that will be broken loose during the 3000 gpm injection into the reactor vessel. GPUN believes l

l l

15 l

i the burden of such testing is not warranted and that such testing is-impractical. The refueling (2/3 flow rate test demonstrates that DH-V14A/B 4

has the capability to open fully,-if needed.

These valves will be tested each refueling interval per Surveillance Procedure No. 1303-11.54 in accordance with Technical .

Specification 4.5.2.2b.

The refueling test frequency assures that the LPI System can supply equal or greater flow that the flow assumed in the Safety Analysis. This.

Technical Specification test frequency should be adequate for ASME Section XI.

The current maximum stroke test of DH-V14A/B demonstrates that these valves will allow 2/3 of the design basis flow rate as. described in i the IST program description. GPUN has agreed to a valve disassembly program for DH-V14A/S. However, we feel that disassembly of one valve every refueling interval is too frequent due to the risk of defeating.

decay heat removal capability. Disassembly of one of the DH-V14 valves would take a train of decay heat removal (DHR) out of service during the refueling outage when DHR capability needs to be maintained.

in accordance with T.S. 3.4.2.1. GPUN feels that because of the risk i

to decay heat removal capability and the exposures which would result from disassembly, the appropriate frequency of disassembly should be one valve each ten years. GPUN intends to disassemble both of these valves during the current refueling outage and report the results to the NRC following the outage. These valves see little service, so the possibility of service wear as exhibited in recent industry experience with check valve failure is minimal. Results of DH-V14 disassembly.

during this outage should progide further information to support

, GPUN's proposed ten year disa?sembly inspection frequency.

1 (S) 3.4.3.2.2 Evaluation--The staff does not agree that the licensee's praposed partial-stroke exercise testing during refueling

outages adequarely demonstrates the full-stroke capability of-valves DH-V14A/B because this test may not move the check valve disk to the full flow position and does not assure that it will move there when -

required by system demands. The licensee has stated that full-stroke exercising these valves with system flow is too complicated and time ~

consuming to be performed during refueling outages, therefore, the staff )

, will require that a valve sampling disassembly / inspection utilizing a i 16 l

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

manual full-stroke of the disk be utilized to verify the check valves'-

full-stroke capability. The sampling. technique requires that each valve in the group must be of the same design (manufacturer, size,.model number, and materials of construction) and must have the same service conditions.

Additionally, at each disassembly it must be verified that the disassembled valve is capable of full-stroking and that its internals are structurally sound (no loose or corroded parts).

A different valve of each group is required to be disassembled, inspected, and manually full-stroked at each refueling until the entire t group has been tested. If it is found that the disassembled valve's full-stroke capability is in question, the remainder of the-valves in that group must also be disassembled, inspected, and manually full-stroked at the same outage.

(S) 3.4.3.2.3 Conclusion--The staff concludes that a disassembly / inspection utilizing a manual full-stroke is the only method available to full-stroke exercise these valves. The licensee is-required to perform these tests at each refueling outage on a sampling basis.

l (S) 3.6 Feedwater System

(S) 3.6.1 Category C Valves (S) 3.6.1.1 Relief Request. The licensee has requested relief from' exercising valves FW-V12A/B, main feedwater header checks, in accordance with the requirements of Section XI, Paragraph IWV-3520, and proposed to leak test these valves to verify closure during refueling outages.

(S) 3.6.1.1.1 Licensee's Basis for Requestina Relief--For short

. cold shutdowns, condenser vacuum is maintained, if at all possible.

Testing of FW-V12A/B however, requires main condenser vacuum to be broken. i After acceptable chemistry samples are obtained, the secondary side of the j OTSG must be fed to the wet layup condition. Several operators'are l required for at least one shift. In addition, the main steam hangers must l l

i l

17 i

l l

l be blocked and pinned. This requires approximately 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> for 4 to 5' '

men. Then, for the closed test of FW-V12A/B, approximately 150 psig of nitrogen pressure is applied to the secondary side of the OTSGs. It is not- ,

practical to perform this evolution on a cold shutdown frequency. The l appropriate frequency is each refueling outage.

GPUN's commitment to develop and implement a test method to verify operability of FW-V12A/B has been met. The IST program description ,

has been revised to include a full closure test'of FW-V12A/B on a

. refueling interval basis. The test of FW-V12A/B is adequate to verify the valves closure capability and provides assurance of their operability.

(S) 3.6.1.1.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the exercising requirements of Section XI for valves FW-V12A/8. The staff agrees that these valves cannot be verified to shut quarter.ly during power operation because a reactor trip would result. The licensee has addressed exercising these valves during cold shutdowns and the staff agrees that the alternate test method, leak testing, may delay reactor startup due to the time required to perform the test.

(S) 3.6.1.1.3 Conclusion--The staff concludes that leak testing these valves during refueling outages is an acceptable alternate method to verify valve closure and should be sufficient to demonstrate proper valve

operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve operability as required by the Code.

(S) 3.8 Main Steam System (S) 3.8.2 Cateaory C Valves (S) 3.8.2.1 Relief Request. The licensee has requested relief from -

exercising valves MS-V9A/B, emergency feedwater pump turbine steam supply checks, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to partial-stroke exercise these valves quarterly and to disassemble and inspect one valve or the other at ten year = intervals.

i

, 18 i

(S) 3.8.2.1.1 Licensee's Basis for Requestino Relief--These check valves supply steam from the OTSGs to EF-U1. EF-Ul is the turbine drive unit for the steam driven emergency feedwater pump (EF-P1). During cold shutdown, it is impractical to stroke test these valves (full-stroke or partial-stroke) since the steam which would be needed to operate these valves is not available during cold shutdown conditions.

. Full-stroke testing of MS-V9A/B is also impractical due to other limitations during plant conditions when steam is available. EF-P1 must be tested using the recirculation line to the condensate storage tank bypassing the OTSG. This is to prevent degradation of the OTSGs by excessive thermal stress cycling of the emergency feedwater nozzles. The number of thermal cycles on the emergency feedwater nozzles is limited to (40) cycles over the life of the plant. Due to the small size of the recirculation line, EF-P1 cannot be tested at full capacity (MS-V9A/B will not open fully). Under these restrictions it is only possible to obtain approximately 48% flow which corresponds to about 36% opening of MS-V9A.'8.

Plant modificati,ons which would be required to perform full-stroke tests of MS-V9A/B either by piping in auxiliary steam or by replacing the recirculation piping with larger piping capable of recirculating the full-EFW pump capacity would introduce exorbitant cost. GPUN has not fully examined the cost and safety impact of modifications which would be required to full-stroke MS-V9A/8, however, does not feel that such modifications would be beneficial.

MS-V9B was disassembled for IST examination purposes in late 1984 and found to be in excellent condition. Since no indication of potential degradation was found, this provides additional assurance of the continued capability of MS-V9A/8 to open fully when needed. In addition, MS-V9A will

. be disassembled and inspected at the next refueling outage (end of cycle 5 refueling outage). Thereafter, either MS-V9A or B will be  !

disassembled / inspected on a 10 year basis (next inspection due ~1994) if the inspection of MS-V9A proves satisfactory. If degradation is found which would make the valve's full-stroke capability questionable, then the 19

other valve (MS-V9A or B) would be disassembled and inspected during that same outage. It is impractical to test MS-V9A/B when steam is not available and it is also impractical to perform a full-stroke test on MS-V9A/8. GPUN concludes that quarterly testing of MS-V9A/B at 48% flow (36% open) when steam is available and disassembly each 10 years meets the intent of the ASME Code Section XI and the relief which is being requested .

is therefore justified.

MS-V9A was disassembled during the current 6R refueling outage. A report of the disassembly results will be sent to the NRC by separate letter.

I (S) 3.8.2.1.2 Evaluation--The staff does not agree with the licensee's proposed alternate testing frequency and, therefore, relief should not be granted from the exercising requirements of Section XI for valves MS-V9A/8. The staff agrees that these valves cannot be full-stroke exercised during power operation because the only full flow path is into the steam generators and the colder water being pumped could damage the emergency feedwater. nozzles. Also, the pump recirculation line cannot accommodate the flow required to full-stroke exercise the steam supply checks. However, the staff does not agree that-the licensee's proposed quarterly partial-stroke exercise testing adequately demonstrates the

full-stroke capability of these valves because this testing does not move the check valve disk to the full flow position and does not assure that it will move there when required ny system demands. The staff _ agrees that valve disassembly / inspection on a sampling basis.is an acceptable alternate method to utilize to demonstrate operability but does not agree with the licensee's proposed ten year inspection frequency and will require that the

sample disassembly / inspection be performed at a refueling outage frequency.

1 (S) 3.8.2.1.3 Conclusion--The staff concludes that a disassembly / inspection utilizing a manual full-stroke is the only method -

, available to full-stroke exercise these valves. .The licensee is required

~

to perform these tests at each refueling outage on a sampling basis.

20

.-,n..

.- ,~,-m.- , , . - - - . , - - , - - , ,,. e - . _ . , , - , , , ,e--. , , . , -

r- ------ - - -,

l (S) 3.9 Makeuo System (S) 3.9.2 Category B/C Valves (S) 3.9.2.1 Relief Request. The licensee has requested relief from full-stroke exercising valves MU-V14A/B, makeup pump borated water storage tank suction stop checks, in accordance with'the requirements of

. Section XI, Paragraphs IWV-3400 and -3522, and proposed to full-stroke exercise these valves during refueling outages.

(S) 3.9.2.1.1 Licensee's Basis for Requestino Relief--When the MU pumps (HPI pumps) are racked in, Technical Specification 3.1.12.3 requires the RCS average temperature to be >275*F prior to opening the HPI block valves (MU-V16A/B/C/D). This is to prevent overpressurizing the RCS. In order to test the above valves, the piping configuration requires pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the RCS cold legs. Therefore, the only testing method allowable by Technical 4 Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is

, >215'F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present, TMI-1 has had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing frequency is refueling. To date, no testing failures have occurred: TMI-1 has always been able to deliver the accident design flow rate through the above listed valves.

Design flow testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This test results in a full-stroke of each of the HPI check valves listed. The test involves an injection I to the RCS through the HPI injection nozzles. Each injection through  !

an HPI nozzle results in a thermal cycle to the nozzle. Currently, )

only 48 thermal cycles of these nozzles are allowed. Quarterly

- testing of the HPI check valves would not be appropriate because testing at power would result in additional thermal cycles to the injection nozzles. Nor would testing of the HPI check valves on a cold shutdown frequency be appropriate because this would not eliminate the imposition of additional thermal cycles to the injection nozzles due to Technical Specification limitations. In order to prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the 'l

HPI injection valves (MU-V16A/B/C/0) be closed unless the reactor 21 I

i

vessel head is removed or the RCS temperature is >275'F. T.S.

4.5.2.1.c further limits the ability to test the HPI check valves without resulting in additional thermal cycles of the HPI injection nozzles by requiring RCS temperature to be >320*F for any test that

requires flow through MU-V16A/B/C/D unless the reactor vessel head is removed. Therefore, HPI check valve testing can only be performed

during refueling outages.

(S) 3.9.2.1.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the exercising requirements of Section XI for valves MU-V14A/B. The licensee is presently full-stroke exercising the operators on these valves quarterly, however;

since these valves are stop-check valves, this test does not exercise the valve disk and a flow test must be conducted. The staff agrees that these valves should not be partial- or full-stroke exercised during power operation because boron addition from the borated water storage tank could upset primary system boron control and could result in a reactor shutdown i in addition to reducing the service life of the high pressure injection l nozzles due to the thermal cycling. Exercising these valves during power operation could also result in a reactor trip due to loss of pressurizer level control. During cold shutdowns, the high pressure injection function is disabled to prevent a low-temperature overpressurization of the reactor coolant system and, with the reactor vessel head installed, system temperature must be raised above the minimum temperature for pressurization if any flow is to be introduced through the injection nozzles which, in turn, results in an additional thermal cycle to the nozzles.

(S) 3.9.2.1.3 Conclusion--The staff concludes that the licensee should continue to full-stroke exercise the operators on these valves quarterly. The staff also concludes that the proposed alternate testing of full-stroke exercising these valves during refueling outages when.the reactor vessel head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles to the injection ,

nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of. valve ,

operability as required by the Code.

4 22

. _ _ = _ _ .- __ _ ___

i 4

. (S) 3.9.3 Category C Valves (S) 3.9.3.1 Relief Reauest. The licensee has requested relief from exercising valves MU-V73A/B/C, makeup pump discharge checks, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to partial-stroke exercise them quarterly and to full-stroke exercise them during refueling outages.

(S) 3.9.3.1.1 Licensee's Basis for Requestino Relief--When the MU pumps (HPI pumps) are racked in Technical Specification 3.1.12.3 requires the RCS average temperature to be >275'F prior to opening the HPI block valves (MU-V16A/B/C/D). This is to prevent overpressurizing the RCS. In order to test the above valves, the piping configuration requires pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the RCS cold legs. Therefore, the only testing method allowable by Technical Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is

>275'F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present TMI-1 has

! had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing frequency is refueling. To date, no testing failures have~ occurred: TMI-1 has always been able to deliver the accident design flow rate through the above listed valves.

Design flow testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This-test results in a full-stroke of each of the HPI check valves listed. The test involves an injection to the RCS through the HPI injection nozzles. Each injection through an HPI nozzle results in a thermal cycle to the nozzle. Currently, only 48 thermal cycles of these nozzles are allowed. Quarterly testing of the HPI check valves would not be appropriate because testing at power would result in additional thermal cycles to the injection nozzles. Nor would testing of the HPI check valves on a

. cold shutdown frequency be appropriate because this would not eliminate the imposition of additional thermal cycles to-the injection nozzles due to Technical Specification limitations. In order to l prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the ,

HPI injection valves (MU-V16A/B/C/D) be closed unless the reactor l vessel head is removed or the RCS temperature is >275*F.

23 l

l

T.S. 4.5.2.1.c further limits the ability to test the HPI check valves without resulting in additional thermal cycles of the HPI injection nozzles by requiring RCS temperature to be >320*F for any test that requires flow through MU-V16A/8/C/D unless the reactor vessel head is removed. Therefore, HPI check valve testing can only be performed during refueling outages.

(S) 3.9.3.1.2 Evaluation--The staff agrees with the licensee's .

basis and, therefore, relief should be granted from the exercising requirements of Section XI for valves MU-V73A/8/C. The staff agrees that -

these valves should not be full-stroke exercised during power operation because boron addition from the borated water storage tank could upset primary system boron control and could result in a reactor shutdown in addition to reducing the service life of the high pressure injection nozzles due to the thermal cycling. Full-stroke exercising these valves during power operation could also result in a reactor trip due to loss of pressurizer level control. During cold shutdowns, the high pressure injection function is disabled to prevent a low-temperature overpressurization of the reactor coolant system and, with the reactor vessel head installed, system temperature must be raised above the minimum temperature for pressurization if any flow is to be introduced through the injection nozzles which, in turn, results in an additional thermal cycle to the nozzles.

(S) 3.9.3.1.3 Conclusion--The staff concludes that the proposed alternate testing of partial-stroke exercising these valves quarterly and full-stroke exercising them during refueling outages when the reactor vessel head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles to the injection nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve operability as required by the Code.

(S) 3.9.3.2 Relief Recuest. The licensee has requested relief from exercising valves MU-V86A/8, high pressure injection loop B cold leg

• 24

8 injection checks, in accordance with the: requirements of Section XI, Paragraph IWV-3522, and proposed to full-stroke exercise these valves j during refueling outages.

(S) 3.9.3.2.1- Licensee's Basis for Reauestina Relief--When the MU pumps (HPI pumps) are racked in Technical Specification 3.1.12.3 requires the RCS average temperature to be >275'F prior to opening the HPI block valves (MU-V16A/B/C/0). This is to prevent overpressurizing the RCS. In order to test the above valves, the piping configuration requires

{

pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the RCS cold legs. Therefore, the only testing method allowable by Technical

Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is

>275*F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present TMI-1 has had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing

frequency is refueling. To date, no testing failures have occurred

TMI-1 has always been able to deliver the accident design flow rate through the i above listed valves.

i Design flow. testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This test results in a full-stroke of <

each of the HPI check valves listed. The test involves an injection i to the RCS through the HPI injection nozzles. Each injection through

an HPI nozzle results in a thermal cycle to the nozzle. Currently,

! only 48 thermal cycles of these nozzles are allowed. Quarterly testing of the HPI check valves would not be appropriate because j testing at power would result in additional thermal cycles to the i injection nozzles. Nor would testing of the HPI check valves on a j cold shutdown frequency be appropriate because this would not eliminate the imposition of additional thermal cycles to the injection nozzles due to Technical Specification limitations. In order to prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the HPI injection valves (MU-V16A/B/C/D) be closed unless the reactor

vessel head is removed or the RCS temperature is >275'F.

j .

T.S. 4.5.2.1.c further limits the ability to test the HPI check valves

without resulting in additional thermal cycles of the HPI injection

, nozzles by requiring RCS temperature to be >320*F for any test that

- requires flow through MU-V16A/B/C/D unless the reactor vessel head is i removed. Therefore, HPI check valve testing can only be performed during refueling outages.

l l

i 25

(S) 3.9.3.2.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the exercising requirements of Section XI for valves MU-V86A/B. The staff agrees that these valves should not be partial- or full-stroke exercised during power operation due to the reduction in service life of the high pressure injection nozzles due to the thermal cycling. Also, exercising these .

valves during power operation could result in a reactor trip due to loss of pressurizer level control. During cold shutdowns, the high pressure -

injection function is disabled to prevent a low-temperature overpressurization of the reactor coolant system and, with the reactor vessel head installed, system temperature must be raised above the minimum 4

temperature for pressurization if any flow is to be introduced through the injection nozzles which, in turn, results in an additional thermal cycle to the nozzles.

~

(S) 3.9.3.2.3 Conclusion--The staff concludes that the proposed j alternate testing of full-stroke exercising these valves during refueling i outages when the reactor vessel head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles 1

to the injection nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve operability as required by the Code.

(S) 3.9.3.3 Relief Reauest. The itcensee has requested relief from exercising valve MU-V94, normal makeup header injection check, in j accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to partial-stroke exercise it quarterly and to full-stroke exercise it during refueling outages.

(S) 3.9.3.3.1 Licensee's Basis for Requestino Relief--When the l MU pumps (HPI pumps) are racked in. Technical Specification 3.1.12.3 -

, requires the RCS average temperature to be >275'F prior to opening the HPI block valves (MU-V16A/B/C/0). This is to prevent overpressurizing the

• l RCS. In order to test the above valves, the piping configuration requires l pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the i

9 26 i

RCS cold legs. Therefore, the only testing method allowable by Technical Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is

>275'F . This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present, TMI-1 has had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing

] frequency is refueling. To date, no testing failures have occurred: TMI-l j . has always been able to deliver the accident design flow rate through the j above listed valves.

4 MU-V94.cannot be full-stroke exercised each quarter because this would

be disruptive to normal plant operation and would create the potential for an unsafe pressure transient during power operation. Normal RCS makeup through MU-V94 is-40 gpm. Increasing this to accident design 4 flow rate may create an overpressure transient and challenge the a RPS high pressure trip setpoint or the PORV. In addition, if MU-V94 were to be tested each quarter, an additional letdown cooler would be placed into service. This would cause a thermal cycle on the cooler.

Therefore, in accordance with ASME Section XI IWV-3412, MU-V94 will be

part-stroked each quartar.

l (S) 3.9.3.3.2 Evaluation--The staff agrees with the licensee's j basis and, therefore, relief should be granted from the exercising-

requirements of Section XI for valve MU-V94. The staff agrees that this j valve should not be full-stroke exercised during power operation because the flow required could result in a reactor trip due to loss of pressurizer level control in addition to reducing the service life of the injection

{ nozzles and the letdown cooler. During cold shutdowns, the high pressure j injection function is disabled to present a low-temperature overpressurization of the reactor coolant system and, with the reactor j vessel head installed, system temperature must be raised above the minimum temperature for pressurization if any flow other than normal makeup is to be introduced through the injection nozzles which, in turn, results in an additional thermal cycle to the nozzles.

, (S) 3.9.3.3.3 Conclusion--The staff concludes that the proposed l alternate testing of partial-stroke exercising this valve quarterly and

} full-stroke exercising it during refueling outages when the reactor vessel l

21 i.

7-.-- a w m. ,c-- ,  %-.-=. r-,- 9 4,-pi- -w.m_,,,g. , , - , , . , , . , - -

- - - - - - - - wmy-w m - -es.-w.m--

4 head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles to the injection nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve operability as required by the Code.

(S) 3.9.3.4 Relief Request. The licensee has requested relief from exercising valve MU-V95, high pressure injection loop A' cold leg. injection .

check, in accordance with the requirements of Section XI, Paragraph j IWV-3522,'and proposed to full-stroke it during refueling outages.

(S) 3.9.3.4.1 Licensee's 8 asis for Reauestina Relief--When the

MU pumps (HPI pumps) are racked in Technical Specification 3.1.12.3 l requires the RCS average temperature to be >275'F prior to opening the. .

HPI block valves (MU-V16A/8/C/D). This is to prevent overpressurizing the i"

RCS. In order to test the above valves, the piping configuration requires i pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the l RCS cold legs. Therefore, the only testing method allowable by Technical Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is I

>275'F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present, TMI-1 has

]

i had more than 20 thermal cycles, therefore, it is not advisable to test.the f above valves at a cold shutdown frequency. The appropriate testing i

j frequency is refueling. To date, no testing failures have occurred: TMI-l l

has always been able to deliver the accident design flow rate through the i above listed valves.

i Design flow testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This test results in a full-stroke of each of the HPI check valves listed. The test involves an injection to the RCS through the HPI injection nozzles. Each injection-through an HPI nozzle results in a thermal cycle to the nozzle. Currently, only 48 thermal cycles of these nozzles are allowed. Quarterly testing of the HPI check valves would not be appropriate because testing at power would result in additional thermal cycles to the .

! injection nozzles. Nor would testing of the HPI check valves on a cold shutdown frequency be appropriate because this would not eliminate the imposition of additional thermal cycles to the injection i nozzles due to Technical Specification limitations. In order to prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the i

28 i

1 HPI injection valves (MU-V16A/8/C/D) be closed unless the reactor vessel head is removed or the RCS temperature is >275'F.

T.S. 4.5.2.1.c further limits the ability to test the HPI check valves  !

without resulting in additional thermal cycles of the HPI injection nozzles by requiring RCS temperature to be >320*F for any test that requires flow through MU-V16A/B/C/D unless the reactor vessel head is ,

removed. . Therefore, HPI check valve testing can only be performed during refueling outages. j

(S) 3.9.3.4.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the exercising i requirements of Section XI for valve MU-V95. The staff agrees that this  !

valve should not be partial- or full-stroke exercised during power operation due to the reduction in service life of the high pressure injection nozzles due to the thermal cycling. Also, exercising this valve during power operation could result in a reactor trip due to loss of pressurizer level control. During cold shutdowns, the high pressure injection function is disabled to prevent a low-temperature i overpressurization of the reactor coolant system and, with the reactor vessel head installed, system temperature must .5a raised above the minimum temperature for pressurization if any flow is to be introduced through the injection nozzles which, in turn, results in an additional thermal cycle to 4

the nozzles.

(S) 3.9.3.4.3 Conclusion--The staff concludes that-the proposed alternate testing of full-stroke exercising this valve during refueling outages when the reactor vessel head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles l to the injection nozzles should demonstrate proper valve operability. The  !

staff concludes that the alternate testing proposed will give reasonable i assurance of valve operability as required by the Code.

(S) 3.9.3.5 Relief Request. The licensee has requested relief from '

exercising valves MU-V107A/8/C/D, high pressure injection loop A and 8 cold leg injection checks, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to full-stroke exercise these valves during refueling outages.

29

1 l

(S) 3.9.3.5.1 Licensee's Basis for Recuesting Relief--When the MU pumps (HPI pumps) are racked in, Technical Specification 3.1.12.3 requires the RCS average temperature to be >275*f prior to opening the HPI block valves (MU-V16A/B/C/0). This is to prevent overpressurizing the RCS. In order to test the above valves, the piping configuration requires pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the .

RCS cold legs. Therefore, the only testing method allowable by Technical Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is -

>275'F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present, TMI-l has had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing frequency is refueling. To date, no testing failures have occurred: TMI-1 has always been able to deliver the accident design flow rate through the above listed valves. '

Oesign flow testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This test results in a full-stroke of each of the HPI check valves listed. The test involves an injection to the RCS through the HPI injection nozzles. Each injection through an HPI nozzle results in a thermal cycle to the nozzle. Currently, only 48 thermal cycles of these nozzles are allowed. Quarterly testing of the HPI check valves would not be appropriate because testing at power would result in additional thermal cycles to the injection nozzles. Nor would testing of the HPI check valves on a cold shutdown frequency be appropriate because this would not eliminate the imposition of additional thermal cycles to the injection nozzles due to Technical Specification limitations. In order to prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the HPI injection valves (MU-V16A/B/C/0) be closed unless the reactor vessel head is removed or the RCS temperature is >275'F.

T.S. 4.5.2.1.c further limits the ability to test the HPI check valves without resulting in additional thermal cycles of the HPI injection nozzles by requiring RCS temperature to be >320*F for any test that requires flow through MU-V16A/8/C/D unless the reactor vessel head is removed. Therefore, HPI check valve testing can only be performed during refueling outages.

(S) 3.9.3.5.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the exercising requirements of Section XI for valves MU-V107A/B/C/D. The staff agrees that these valves should not be partial- or full-stroke exercised during power operation due to the reduction in service life of the high pressure .

30 6

/

l j injection nozzles due to the thermal cycling. Also, exercising these j valves during power operation could result in a reactor trip due to loss of I pressurizer level control. During cold shutdowns, the high pressure injection function is disabled to prevent a low-temperature i overpressurization of the reactor coolant system and, with the reactor

, vessel head installed, system temperature must be raised above the minimum j temperature for pressurization if any flow is to be introduced through the

. injection nozzles which, in turn, results in an additional thermal cycle to the nozzles.

1 a

(S) 3.9.3.5.3 Conclusion--The staff concludes that the proposed alternate testing of full-stroke exercising these valves during refueling outages when the reactor vessel head is removed to accommodate the flow required and the system is cooled sufficiently to preclude thermal cycles to the injection nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve operability as required by the Code.

(S) 3.9.3.6 Relief Request. The licensee has requested relief from exercising valve MU-V220, high pressure injection loop A cold leg injection check, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to full-stroke exercise it during refueling outages.

(S) 3.9.3.6.1 Licensee's Basis for Requestina Relief--When the MU pumps (HPI pumps) are racked in, Technical Specification 3.1.12.3 j requires the RCS average temperature to be >275'F prior to opening the j HPI block valves (MU-V16A/B/C/D). This is to prevent overpressurizing the RCS. In order to test the above valves, the piping configuration requires I

pumping through the MU pumps to the MU-V16s and to the HPI nozzles at the RCS cold legs. Therefore, the only testing method allowable by Technical

. Specification 3.1.12.3 is to pump to the RCS when the RCS temperature is

>275'F. This uses one or more thermal cycles on the HPI nozzles. Recent B&W guidance allows only 48 cycles on these nozzles. At present, TMI-1 has

! had more than 20 thermal cycles, therefore, it is not advisable to test the above valves at a cold shutdown frequency. The appropriate testing i

1 31

frequency is refueling. To date, no' testing failures have occurred: TMI-1 has always been able to deliver the accident design flow rate through the

~

above listed valves.

Design _ flow testing of the HPI system is performed each refueling in accordance with T.S. 4.5.2.1. This test results in a full-stroke of each of the HPI check valves listed. The test involves an injection -

to the RCS through the HPI injection nozzles. Each injection through an HPI. nozzle results in a thermal cycle to the nozzle. Currently, ~

only 48 thermal cycles of these nozzles are allowed. Quarterly testing of the HPI check valves would not be appropriate because testing at power _would' result in additional thermal cycles to the injection nozzles. Nor would. testing of.the HPI check valves on a cold shutdown frequency be appropriate because this would not __

eliminate the imposition of additional thermal cycles to the injection ,

nozzles due to Technical Specification limitations. In order to prevent overpressurization of the RCS, T.S. 3.1.12.3 requires that the HPI injection valves _(MU-V16A/8/C/0)'be closed unless the reactor vessel head is removed or the RCS temperature is >275'F.

T.S. 4.5.2.1.c further limits the ability to test the HPI check valves without resulting in additional thermal cycles of the HPI injection nozzles by requiring RCS temperature to be >320*F for any test that-requires flow through MU-V16A/8/C/0 unless-the reactor vessel head is removed. Therefore, HPI check valve testing can only be performed during refueling outages.

(S) 3.9.3.6.2 Evaluation--The staff agrees with the licensee's <

basis and, therefore, relief should be granted from the exercising requirements of Section XI for valve MU-V220. The staff agrees that this valve should not be partial- or full-stroke exercised during power operation due to the reduction in service life of the high_ pressure injection nozzies due to the thermal cycling. Also, exercising this valve l during power operation could result in a reactor trip due to loss of l

pressurizer level control. During cold shutdowns, the high pressure l

! injection function is disabled to prevent a low-temperature overpressurization of the reactor coolant system and, with the reactor f vessel head installed, system temperature must be raised above the minimum temperature for pressurization if any flow is to be introduced through the -

l

! injection nozzles which, in turn, results in an additional thernel cycle to I

the nozzles.

(S) 3.9.3.6.3 Conclusion--The staff concludes that the proposed alternate testing of full-stroke exercising this valve during refueling outages when the reactor vessel head is removed to accommodate the flow 32 I

~.

required and the system is cooled sufficiently to preclude thermal cycles to the injection nozzles should demonstrate proper valve operability. The staff concludes that the alternate testing proposed will give reasonable assurance of valve' operability as required by the Code.

(S) 3.11 Reactor Bu11dina Emeraency Coolina System

. (S) 3.11.1 Category B Valves (S) 3.11.1.1 Relief Reauest. This relief request has been deleted.

4 (S) 3.11.1.2 Licensee's Basis for Deletina Relief. GPUN has accepted the staff's position on this issue. The IST program description has been revised to include trending of stroke time measurements of RR-V6 and our request for relief has been withdrawn.

(S) 3.12 Reactor Buildina Soray System (S) 3.12.1 Category C Valvti (S) 3.12.1.1 Relief Request. The licensee has requested relief from exercising valves BS-V30A/8, reactor building spray header checks, in accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to partial-stroke exercise these valves quarterly and to utilize a sample disassembly / inspection at ten year intervals to demonstrate operability.

(S) 3.12.1.1.1 Licensee's Basis for Requestino Relief--The

full-stroke of BS-V30A/B would require initiation of reactor building spray. This would entail spraying the reactor building with borated water, i

On 6-20-84 (Job Ticket CE-154), 8S-V30A was' disassembled for visual examination purposes. The examination was satisfactory (no unusual degradation and valve disk was free to open). 8S-V30A has approximately 10 years of in-service time and.this is the first time that the valve has j been opened. GPUN will continue to disassemble either of these static valves (alternating between A and B) during each ten year inspection interval. BS-V30A/8 are stainless steel and, therefore, not subject to 33 l

l l _ , .-. . . .. _- -

-l corrosive attack. GPUN is of the opinion that quarterly partial-stroke testing and disassembly, as described, will provide reasonable assurance that BS-V30A/B would open if needed. In addition, if disassembly / inspection reveals that the full-stroke capability of the j disassembled valve may be in question, the other valve will be disassembled  ;

and inspected at the same outage. .

1 GPUN intends to disassemble one of these valves during the current - l 6R refueling outage and submit a report on the results following the outage. These valves see very little service. Therefore, the susceptibility of BS-V30A/B to service related wear, as exhibited in recent industry experience with check valve failures, is minimal.

GPUN believes that the results of disassembly of one of the BS-V30 valves during 6R will pr7 vide additional information sufficient to justify less frequent disassembly than each refueling outage on a sampling basis as required by the SER.

(S) 3.12.1.1.2 Evaluation--The staff does not agree with the licensee's proposed alternata testing frequency and, therefore, relief should not be granted from the exercising requirements of Section XI for valves BS-V30A/8. The staff agrees that these valves cannot be full-stroke exercised using flow because no test flow path exists and testing would spray the equipment in the containment which could cause extensive water damage. On this basis, the staff agrees with the licensee's proposal to partial-stroke exercise these valves quarterly utilizing an air test and to disassemble each valve on a sampling basis to demonstrate operability, however, the staff does not agree with the proposed ten year inspection frequency and will require that the sample disassembly / inspection be performed at a refueling outage frequency.

l (S) 3.12.1.1.3 Conclusion--The staff concludes that a disassembly / inspection utilizing a manual full-stroke is the only method available to full-stroke exercise these valves. The licensee is required to perform these tests at each refueling outage on a sampling basis. -

l

' I (S) 3.12.1.2 Relief Request. The licensee has requested relief from exercising valves BS-V52A/8, sodium hydroxide tank outlet checks, in I l

i i

34 l l

$ accordance with the requirements of Section XI, Paragraph IWV-3522, and proposed to utilize a sample disassembly / inspection at ten year intervals to demonstrate operability.

(S) 3.12.1.2.1 Licensee's Basis for Reauestina Relief--Given the pipe geometry, there is no method of part-stroking or full-stroking'these check valves without contaminating the RCS with sodium hydroxide and other

. chemical contaminants in the dead leg of piping from these valves to the sodium hydroxide tank. During January. 1984 (Surveillance Procedure No. 1300-3P and Job Ticket Nos. CC818 and CC819) these valves were i disassembled and visually ~ examined with satisfactory results. The valve internals were found to be in "like new" condition. Disassembling each of these valves at or near the end of the ten year ISI interval provides l reasonable assurance of the operational readiness of BS-V52A/B.

1 GPUN intends to disassemble one of these valves during the current 6R refueling outage and submit a report on the results following the outage. These valves see very little service. Therefore, the

l. susceptibility of BS-V52A/B to service related wear, as exhibited in recent

industry experience with check valve failures, is minimal. GPUN believes-that-the results of disassembly of one of the BS-V52 valves during 6R will provide additional information sufficient to justify less frequent disassembly than one valve each refueling as required by the SER.

(S) 3.12.1.2.2. Evaluation--The staff does not agree with-the licensee's proposed alternate testing frequency and, therefore, relief should r.ot be granted from the exercising requirements of Section XI for valves BS-V52A/8. The staff does agree that these valves cannot be exercised with flow without spreading sodium hyd. oxide to the borated water storage tank and from there into the reactor coolant system which could:

upset the chemistry control of the entire system. Because of this chemical contamination, the staff agrees with the licensee's proposal to disassemble each valve on a sampling basis to demonstrate operability, however, the-

~

staff does not agree with the proposed ten year inspection frequency _and will require that the sample disassembly / inspection be performed at a i refueling outage frequency.

35

(S) 3.12.1.2.3 Conclusion--The staff concludes that a disassembly / inspection utilizing a manual full-stroke is the only method available to full-stroke exercise these valves. The licensee is required

- to perform these tests at each refueling outage on a sampling basis.

t s

4 36 l

._ , . ~ - . . - ,, , _ --

i l

(S) APPENDIX D (S) D.1 Licensee's Required Action Items (S) D.1.19 The licensee will be required to test the PORV, RC-RV2, in accordance with the requirements of Item A.9 of the SER at least ance each refueling cycle.

GPUN has accepted the staff's position to test the PORV as discussed with the NRC during the October 1986 meeting.. The IST program description has been revised to include the additional testing.

The PORV will be tested once per fuel cycle not to exceed a period of 24 months. During each refueling outage, one of the following three types of testing will be performed:

1) Actuate during shutdown conditions, or

2) Remove and bench test, or

3) Remove and replace with a spare valve previously bench tested within the last 3 years.

In addition, as-found and as-left visual examination will be performed. For in-place testing, the stroke time will be determined by the

use of the acoustic monitors. For bench testing, stroke time will be determined by observation. Fail-safe testing is to be determined by valve closure upon removal of the open actuation signal.

This exercising program is in agreement with the staff's revised test 4

requirements. This action item may be deleted when the licensee revises

, the IST program to categorize the PORV as Category B to agree with the j staff's position that it is a power operated valve.

(S) D.l.20 Spent Fuel Coolina System-As discussed with the NRC during the October 1986 meeting and a conference call on November 10, 1986, GPUN believes that sufficient reliable, redundant and diverse means are available for makeup to the 37

. . _ _ _ _ _ _ _ _ _ _ . _ _- , _ . _ _ . _ _ _ . - . . . . , _ . . . _ _ . , _ , . . -._ ~ ,_

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

e

.l 1 spent fuel pools to eliminate the need for adding the spent fuel I cooling system valves to the IST program or. installing an additional  !

safety grade' source of makeup water.

The following reliable sources of water are available at the spent fuel pools for makeup to compensate for boil-off which might result from a sustained total failure of both spent fuel cooling system trains:

1. The reclaimed water storage system is piped to the spent fuel pool. This system is relied upon to provide makeup to the spent fuel pools under normal conditions of-evaporative

• losses. Because of its other system support functions (i.e., RCP No. 3 seal purge and radwaste precoat filter processes) for continued power operation, this system should be available for makeup to the spent fuel pools.

~

2. As a backup to the reclaimed water system, the three reactor coolant bleed tanks-(RCBTs) normally contain collectively at l 1 east 80,000 gallons of water suitable for makeup to the spent fuel pools using a waste transfer pump. Although the liquid waste disposal system is not included in the IST program, the operability of this system is necessary to process radwaste and therefore necessary-to assure continued power operation.

3. The fire service system is available at the spent fuel pools through FS-V143 and the hose reel station as the ultimate backup source of makeup water. The fire service system, including this hose reel station, are required to be operable during all operating conditions (T.S. 3.18).

As discussed in the FSAR, section 9.4.7, in the event both spent fuel cooling loops were out of service simultaneously, commencing with the pool temperature initially at 147'F, approximately 12.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> would elapse before the temperature would heat up to 212*F. This is a conservative' estimate and normally the time available would be mi.ch longer, but 12.5 l hours would provide enough time to restore one of the loops before any boil-off could occur. Considerably more time would then need to elapse before a significant quantity of water could boil away.

In conclusion, GPUN believes that the spent fuel pool makeup capability from the existing reclaimed water system, reactor coolant

! bleed tanks, and the fire service system provides reliable, redundant, ,

and diverse sources of makeup capability in the event of boiling.

! Therefore, the spent fuel system valves need not be included in the

IST program. -

i Section XI requires that all valves which perform a-function in mitigating the consequences of an accident be ileservke tested. In the absence of a safety grade make up system whose components would be 38

4

, -inservice tested, the staff will require-that the spent fuel cooling system valves be included in the'IST program and tested in accordance with Section XI.

(S) D.1.21 The licensee will be required to include valves MU-V10, acid return from deborating demineralizers, and MU-Vil A/B, makeup and purification filters inlets, in the IST program and test them in accordance

. with Section XI because these valves are in an emergency boration flow path and the appropriate valves upstream of these three have been included in the program.

GPUN accepts the staff's position on MU-V10. The IST program description has been revised to include MU-V10. MU-VilA/B, however,

, are not required for emergency boration since one MU-V11 is open for normal operation. MU-VilA/B have no safety-related function, i MU-VilA/B are used for operating convenience to switch between-

! filters, MU-FIA/B. The first choice for emergency boration would be j the BWST, not using the MU filter flow path. Under emergency boration

from the backup source tanks, the makeup filters can be bypassed.

Procedural guidance on bypassing these filters for emergency boration is provided in EP-1202-37.

?

< Valve MU-V10 has been included in the IST program and categorized B.

The staf f agrees that MU-VilA/B do not perform a safety-related function and need not be included in the IST program. This' action item may be deleted.

(S) 0.1.22 The licensee has failed to include nuclear service closed cooling water valves NS-V108A/B, control building air conditioning cooling ,

water outlets, in the IST program. These valves should be included in the program.

i GPUN accepts the staff's position on control valve testing as discussed in the October 1986 meeting. NS-V108A/B, and AH-VilA/B have

, been added to the IST program (see Attachment 2). Also, GPUN has developed a test method for testing NS-V54A/B, NS-V55A/B, NS-V56A/B, and SW-V24A/B and our request for relief from IWV-3413 stroke timing 1 -

of these valves has been withdrawn.

Valves NS-V108A/B and AH-VilA/B have been included in the IST program and categorized B, therefore, this action item may be deleted.

2 39 i

, -----s- , , - _ . - _ _ , ,,-,_,,,.,.._,,.,.-_e y ,,_, - ,, ,, m. ,m, ,.._.,,.% ,. _ r,... , , , y ,.m. . . , _ . ,..,

(S) D.1.23 The licensee has failed to include the emergency diesel generator fuel oil transfer pumps, and appropriate valves, and the air start system, from the air receivers to the engine, in the IST program.

The staff position is that these auxiliary systems should be included in the IST program and tested in accordance with Section XI. Engine mounted 4 pumps are considered to be part of the diesel and need not be tested .

separately.

1 GPUN believes that the current test program for the emergency diesel generators is adequate to assure high reliability of the emergency diesel generators. Also, we do not believe that the ASME Section XI Code was intended to be applied to these diesel auxiliary systems.

However, as discussed with NRC during the meeting on October 27, 1986 and in a conference call on November 10, 1986, GPUN would be willing to add the following component tests to the IST program if the NRC would consider this program as described below to be a satisfactory resolution of this issue and grant the relief which would be needed.

1. The fuel oil transfer pumps [DF-P1A(AC-powered),

DF-P1B(DC-powered), DF-PIC( AC-powered), and DF-P1D(DC-powered)] and associated discharge check valves would be added to the IST Program.

Inlet pressure (Pg) would not be measured since.there are no existing pressure gages.

- Differential pressure (A P) would not be measured since there are no installed discharge pressure gages.

Flow rate (Q) would not be measured since there are no

, installed flow meters. During the monthly diesel l generator surveillance test, the-1 hour run which

! verifies design power output, the flow rate would be verified by confirmation of the ability of'the pump to maintain the day tank within acceptable level limits.

- Pump lubricant level would not be observed since the-i pump is lubricated by the pumped fuel oil and there is j no external lubrication mechanism.

I - Allowing 5 minutes run time prior to taking data may not be possible since the pump is controlled by a level -

I switch in the day tank.

- Pump inboard vibration readings would not be taken since there are no pump bearings. i

2. The diesel generator air start valves (EG-V16A/A, B/A, A/B,  !

and B/B) would be added to the IST program. l l

40

_ __ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ . _ _._. _.__ _ _ ._ ._ _ _ . _ . _ _ _ . _._ ~__

These valves would be verified to open by placing a hand on the valve and feeling it actuate and by

, acknowledging that the diesel rolls over.

. GPUN wishes to point out the TMI-1 diesel generators have demonstrated a high reliability. The reliability of TMI-1 diesel generators, as reported to the NRC in response to Generic Letter 84-15, documents this performance.

The licensee has added only the diesel generator solenoid air start valves to the IST program, therefore, this remains an action item for the l licensee to include the emergency diesel generator fuel oil pumps, and appropriate valves, and the air start system from the air receivers to the solenoid air start valves in the IST program and cannot be deleted.

(S) 0.1.24 The licensee has stated in the IST program that the internals will be removed from vc;ve EF-V3, emergency feedwater. pumps river water i supply check, during the March 1986 outage and they will then delete this valve from the IST program. This item will require verification because no information has been received that addresses this revision. (See IST Program Table 8-1, page 10 of 37).

The internals of EF-V3 have been removed during the Eddy Current Outage in March 1986. The IST program description has been revised.to reflect this change.

This item has been completed and may be deleted.

l (S) 0.2 Omissions and Errors Observed Durina Review of The Licensee's program (S) 0.2.1 The check valves CF-V4A/B do not isolate the reactor coolant system from the decay heat removal system as stated by the licensee in Table 8-1, Footnote 3, page 2 of 37. The staff has previously approved a sample disassembly program for these valves. (See Item 3.1.1.2)

The valve numbers which were given in note 3 on page 2 of 37 of the IST program description submitted to NRC on March 3,1986, were 41

..---,e--,. . , , - - ,_-m.,..,m __.wr , _ , . . , - . , ,___,,-_,_.,,,-p.- _ _ . , .r_.. .,,, , . . , . . . , ,g-m. ---,a ,__,-.--,--,%,

! incorrect. This has been corrected in the revised IST. program description and also a note 4 has been added in order to clarify the intent.

This item has been corrected and may be deleted.

OTHER (S) 3.4 Decay Heat Removal System ,

(S) 3.4.3 Category C Valves .

(S) 3.4.3.1 Relief Request. The licensee has requested relief from exercising valves DH-V16A/8, decay heat removal pump discharge checks, in accordance with Section XI, Paragraph IWV-3522.Jand proposed to

- partial-stroke exercise them quarterly and to full-stroke exercise than at.

least as often as refueling outages.

(S) 3.4.3.1.1 Licensee's Basis for Requestina Relief--Both DH-V16A and 8 will be full-stroke tested each refueling interval. During power operation the DHR System Operating Procedure requires DH-V128 to be locked open and DH-V12A to be locked closed. This is for protection

^

against boron precipitation in the event of a LOCA and is part of the resolution of NUREG 0703, Item II.B.2 shielding issues as described in the NRC's SER dated December 24, 1984. This results in the "B" DHR loop being f

favored for shutdowns. Therefore, DH-V168 will be full-stroke tested each cold shutdown by the normal operation of providing decay heat removal flow of >3000 gpm. See Table B-2.1 for a tabular presentation of testing for DH-V16A/B.

I The full-stroke test for the idle DH-V16 would require a deliberate shift of decay heat removal (DHR) trains. For cold shutdowns of short duration, it is impractical to require a complete system realignment to the opposite DHR system for the sole purpose of cycling this single check valve- .

at design system flow rate. The basis for this conclusion is that:

(1) The system realignment requires manipulation of several large nunual valves (DH-V12A and B) inside radiation areas where dose rate could be l

42 i

-. . - - - - - - . - . - - - - - _ _ ~ - --.--...-- -. - - _ - - . . . . .. ,. ..-

as high as 500 mR/hr. Stay time in the radiation area to cycle one DH-V12 is 15 minutes for three men. The extra radiation exposure and rad waste generation to cycle this one check valve (DH-V16A or B) is not. consistent with ALARA principles.

(2) The system realignment may be critical path time on a recovery from a forced outage. The cost of delaying startup in order to stroke this

. . one check valve makes testing at each cold shutdown impractical.

(3) A requirement to full-stroke this check valve each cold shutdown (if not performed within the last 3 months) would require the start of a

, large engineered safeguards (E.S.) pump and motor for the sole purpose of full-stroking this one check valve. The extra wear on this E.S.

system merely to full-stroke one check valve is undesirable.

(4) For normal planned cold shutdowns, routine maintenance activities have and will require that each DHR system be operated while the other is down for maintenance. When this occurs, both DH-V16A and B will be full-stroked.

(S) 3.4.3.1.2 Evaluation--The staff agrees with the licensee's basis and, therefore, relief should be granted from the Section XI requirement to full-stroke exercise both DH-V16A/B each cold shutdown. Due to previously approved procedural restrictions, each of these valves may or may not be full-stroke exercised each cold shutdown because operation of the decay heat remn %1 systems is not routinely alternated each cold shutdown. The staff agrees that, due to system design, procedural restrictions, and the manual operations required to sht 6 systems, reactor startup could be delayed if each valve were to be full-stroke exercised each cold shutdown. Both of these valves will be full-stroke exercised

, during refueling outages when ample time is available to realign the necessary manual valves.

43

l (S) 3.4.3.1.3 Conclusion--The staff concludes that the proposed alternate testing of partial-stroke exercising these valves quarterly'and full-stroke exercising each valve during refueling outages should be sufficient to demonstrate valve operability. The staff concludes that the' alternate testing proposed will give reasonable assurance of valve '

operability ~as required by the Code. .

s O

44

W 5, muCLEA 1 kE IWLATORY COassessSIOas e > 8P047 Numet A #A_ ;- r ey rsOC, aar stor se,,e anys 8umC PO#as 35

" EGG-NTA-7296

$*' BIBUOGRAPHIC DATA SHEET Supplement 1 ses ~struCriC~so~ r uevias.

UEPYE51YNTAL TECHNICAL EVALUATION REPORT, PUMP AND VALVE INSERVICE TESTING PROGRAM, THREE MILE ISLAND

• =^r " 'acarec= u'o NUCLEAR STATION, UNIT 1 oo~r. v..

g February 1987 i .o r-o.< s, T. L. Cook, H. C. Rockhold oomr. Y...

February l 1987 a enOJECf /T A&meWORE uNir NumetR

7. *t POmuiNG onGAmi2Ar ON NAME ANO M AsumG AOCRE53 (tarmer ** Cs ,

NRR & I&E

. ~ oa ca ~ r ~u a EG8G Idaho, Inc.

P. O. Box 1625 A6812 Idaho Falls, ID 83415 11e rYPEOsatPomr

10. $PoN5CRemG omO4*si2ArlON NAME 440 MetubG ADORESS #farmeele cases Division of PWR Licensing B Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission . .ia coCoviano,, .,

Washington, DC 20555 a

s2 $4P9(SWEN r4Av %QYt3 1

2 .. r..c t ,m ,

This EG&G Idaho, Inc. report presents the results of our evaluation of the licensee's responses to the " Required Action Items" summarized in the NRC staff's Inservice Testing Safety Evaluation Report, Appendix D, dated October 3,1986, for the Three Mile Island Nuclear Station, Unit 1, Inservice Testing Program for pumps and valves that perform a safety-related function.

a 4

,,,C.

in y ,* g rv i . ooCao ~r .~.s . ... . . . Y C us.o :

Unlimited i is SECbmifv CLAS51stCAftCN

' tThr sepop

.,o ~r,.....,o,.~i~cesriaws Unelassified a r,, ,

Unclassified l i , ~ u... . o , . .o u 18 P4tCE l

?