IR 05000344/1985026

Mgt Meeting Rept 50-344/85-26 on 850808.Major Areas Discussed:Auxiliary Feedwater Sys Past History & Reliability

ML20137L263
Person / Time
Site:	Trojan File:Portland General Electric icon.png
Issue date:	08/21/1985
From:	Dodds R, Pereira D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V)
To:
Shared Package
ML20137L224	List: IR 05000344/1985026 ML20137L218
References
50-344-85-26-MM, TAC-44654, TAC-60783, NUDOCS 8509120365
Download: ML20137L263 (19)

Text

r

. .

'

'

!

U.S. NUCLEAR REGULATORY COMMISSION

]

REGIO ;

l

.i Report No. 50-344/85-26

.

l Docket No. 50-344.

l

. License No. NPF- i ,

,

Licensee: Portland General Electric Company j 121 S. W. Salmon Stree :

Portland, Oregon 97204 t

Facility Name: Trojan Nuclear Plant l

Meeting Location: Region V Office, Walnut Creek, California

.

Meeting Conducted

August 8,' 1985 Prepared by: a f 2i g- 7

'

D. B. Perefra, Project' Inspector < Date Signed

.

Approved by: 9/a//fY

~

R. T. Dodds, Chief .

Date' Signed l Reactor Projects:Section No. 1

-

Summary: *

j Management meeting on August 8, 1985, to'exsmine the auxiliary feedwater

] system past history and reliability. The meeting was. held pursuant to i j MC3070 ,

e

,

i 1

i i

i

}-

-

'

0 '

'

-

-

.

.

l

e

.

.

DETAILS 1. Meeting Participants Nuclear Regulatory Commission J. B. Martin, Regional Administrator, Region V J. L. Crews, Senior Reactor Engineer D. F. Kirsch, Acting Director, Division of Reactor Safety and Projects A. E. Chaffee, Branch Chief, Reactor Projects B. H. Faulkenberry, Deputy Regional Administrator, Region V D. B. Pereira, Project Inspector S. Richards, Senior Resident Inspector, Trojan J. Singh, NRR Auxiliary Systems Branch E. Merschoff, I.E., Vendor Inspection Branch Portland General Electric B. Withers, Vice President, Nuclear C. Yundt, General Manager, Technical Functions A. S. Cohlmeyer, Supervisor, Technical Services C. A.Olmstead, Manager, Nuclear Quality Assurance J. Lentsch, Manager, Nuclear Safety and Regulation Department R. Susee, Operations Supervisor, Trojan

.

Bechtel Power Corporation H. Friend, Vice President R. Fosse, Project Engineer 2. Management Discussion A management meeting was held on August 8, 1985, at the U. S. Nuclear Regulatory Commission's (NRC) Region V. offices in Walnut Creek, California, to discuss several items of concern on Trojan's Auxiliary Feedwater (AFW) problems / reliabilit Mr. Martin opened the meeting by suggesting that the discussion follow the outline of the meeting agenda memorandum letter dated August 7, 1985, and agreed upon by the licensee and the staff. The five items agreed upon were as follows: Operational History and Related Problems of AFW Quality Assurance Assessment of AFW Reliability of AFW System Long Term Actions with Regards to AFW System Safety Significance of Current Configurations and Features of AFW System

.

7%

e-

.

The discussion of Operational history and related problems of AFW systems was led by B. Withers and P. Yundt. The licensee presented diagrams of the AFW system, the electric auxiliary feedwater pump, and the steam supplies to the Terry Turbine (see Attachment A). The licensee discussed the AFW failure history with the statement that since plant startup in 1975 there have been 29 avents where the AFW pumps failed to either start or run, out of approximately 356 system demands. In addition, there were 11 other reportable deficiencies for the auxiliary feedwater syste Eighteen of the 29 failures to start or run occurred during 1975-1976 and were associated with startup and initial design deficiencies. Three failures occurred in 1980 following an extensive outage which added TMI modifications to the plant. Of the five failure events since 1980, two were attributable to TMI modifications added to the auxiliary feedwater system without receiving adequate testing prior to placement into service. Attachement B reflects the AFW failure histor Mr. Crews expressed concern on whether post-maintenance / modification testing covers, in a substantial manner, all the essential design features that ensure the AFW system works correctly pric- to placement into service. Mr. Martin discussed the equipe.ent operators role in the AFW reliability problems and whether their training and experience were sufficient to foresee potential problems or have the ability to correct problem A discussion of the Quality Assurance (QA) assessment of the AFW system was presented by C. A. Olmstead (Attachment C). Several recommendations were presented as a result of this evaluation of modifications involving AFW pumps and safety-related equipment. Some of the more important recommended actions were as follows: Conduct a simultaneous auto start test of both AFW pumps following extended outages at minimum CWST leve Strengthen the design change process to ensure that all necessary and relevant post installation testing have been prescribe Increase emphasis on management follow-up and coordination of identified potential problems, Have the Nuclear Safety and Regulations department analyze the cause(s) for the main feed system isolation following a reactor trip and low average temperatur A discussion of reliability of the AFW System with consideration of probability risk assessment (PRA) was presented by J. Lentsch. The three studies presented were as follows: TMI Requirement, October 3, 1979 Report ITI Reliability Assessment, Dec. 28, 1979, and .INPO Reports, October 19, 1984, December 28,1979, and March 15, 198 Based upon these prior studies, it was decided by PGE that further PRA studies would not serve a useful purpose at this tim e

A discussion of the long term actions with regards to the AFW system was presented by P. Yundt. Bechtel Power Corporation has been selected to study the long term actions, and their scope of this work will be to review the AFW system component by component and provide recommendations for long term modifications to increase reliability. The schedule calls for Bechtel to complete the reviews of Operations and Maintenance history logs, review hardware modifications, procedural or administrative changes and provide possible hardware or procedural modifications by the end of September 1985. PGE will review and assess the Bechtel recommendations by the end of October, 1985. To the extend possible, PGE will implement the hardware and procedural modifications agreed upon during the May 1986 refueling outag A discussion of the safety significance of current configurations and features of the AFW system (10CFR 50.59) considerations was presented by the licensee. They stated that the 60% level limit on the condensate water storage tank (CWST) was controlled by administrative procedure Seventeen administrative procedures were changed which included Operating Instructions, Functional Restorational Instructions, and Off-Normal Instructions. Mr. Crews commented that it would be diserable for PGE to formally notify the NRC in writing of the AFW system problems, address the matter of Technical Specifications changes, long term solution of reliability of the automatic start feature of the AFW pumps and the safety significance of 60% CWST level. Mr. Crews stated that the final fix of the AFW system and its acceptance would be based on a review and assessment by NR The meeting closed by giving a synopsis of the meeting and the results and actions agreed upon. The licensee ' agreed to provide NRR a report concerning the safety evaluation of the CWST and administrative controls in place to maintain the CWST at 60% within 10 days. By November 1985, PGE will notify NRC of the long term modifications and procedural changes planned to be implemented to improve the reliability and operational performance of the AFW system, t

The attached hand out material was provided by the licensee during the meetin . ,

-- . - - -

.

.

. m %. ,

4 g

, ..

. ;

,

W' AUXILIARY FEEDWATER SYSTEM

'

- - - -

E!cctri S/G 3/G 3/G Driven S/G ,

Auxiliary A D C B b Feedwater :

'

i fV - V/ -

V - Q Pump Flow M A g M

' M i , 9 FW 2497A j

Control /* FW / 7 FW FW

Valves N -

Y' r4 l

l CV 3004 A1 ,

CV 3004 A2

,

g nkgns' pi pQ Recirculation from 24978 Auxiliary Feedwater M l enf

' ' ' ~ ~ ~

4 <}- T g Punnis 7 B _

1. WN, l h Service >

'

l-+ ! k '

Diesel Driven Pump P-1020, K- 102B

\

k Supply ( 30600 m Water Service Pl-f*-l p4 Cooling AL PI ,

l I

@ 3045A hN 3 fj 1 f

\ Servico ) j .-l [M 4 Turbine Driven Pump To Condon Water g' 3060A

• P-102 A, K-102 A ,,

Supply 1- To Turbine atora00 -

_>4y 4 9 Tank

.

d o m 1: m a m . c., L.O. Coolin0 s Fi pre 1

-

. <

_ ,

.

- .

.

'

,

W ELECTRIC AUXILIARY FEEDWATER PUMP PT-535 S/G-C Sicam Header a C04 To Pressure > '"" gy-- 1

/Ilh Condensate gjp CO4l V Storage Tank PY 2497A Dischargo gg Local

__

s

,-

h ] ll Steam ruru,,ne Driven

- \ PT Aux. FW PS Pump Condensato y p4 i > <}--

( 1 Storage )

Tank M '

O ' (@

'

Dischargo

,"'A > Diesei h / 4g7B Driven j4 Aj~7 Aux. FW

,

i k

n Pump A

%

x A

...

..

Nk %

tu P A **O*,10 204 g .

Figure 2

-

. n- ) .n  ;.3  ?. ( 3

-_

.. __

_ _ _ _ . _

-

. .

.

,

.

-

.

.. .

l

.

f STEAM SUPPLIES TO TERRY TURBINE

,

I SV-1451 ,',' ,'/ W

'" ~

From MSlV

~] Air Supp!y

,

jg ;a L , gJe

_AC + , -l i gef os l // e

= I r>ggg T iL/

ggg l .

,

ys

,

.

.

,

CV - 1452 g Aux FW Pmp Start S,gnal i T166A

^

l -gb  : sw

' ' '

"*"",*'^

L- 3I

,

Atmos

_

j l To C 100 Sentinel E MSIV dlk CV - 1453 g C 05 Valve j l

f[![T N i @-O I

I ,

Safety I - To J Valves CV - 1454 Pi 4.*1

'

31 71 3T

] d To Main Steam Isciation - M Drain - '

h 5 Steam Valve (MSIV)

% Generalors . , ,

..

i

'3>

\- N n

3-Oy I inf are li D

, D w ,

'

<

I . ,, ... A

<

_ _ _ _ _ _ _ _ . _ _ . . . - . - _ ___ _ _ _ _ _ _ ,

._ _

. ..

s' , .

Auxiliary Feedwater Failure History The attached failure history time line shows that since plant startup a pumps in 1975 there have been 29 events where the auxiliary feedw'ter failed to either start or run, out of approximately 356 system demands. In addition, there were 11 other reportable deficiencies for the auxiliary feedwater syste Eighteen of the 29 failures to start or run occurred during 1975-1976, associated with startup and initial design deficiencies. Three failures occurred in 1980 following an extensive outage which added TMI modifications to the plant. Of the 5 failure events since 1980, 2 are attributable to TM1 modifications added to the auxiliary feedwater system without receiving adequate testing prior to placement into servic While the reliability of the auxiliary feedwater system can probably be improved further following a comprehensive design review and improved test program, the overall system reliability that has been demonstrated since 1976 is not unreasonable.

i

$mpaa ~ b l /ge / ef!7

. .

'. T .

y

.)

,

1d i j

. .

o

.

_

.

4

. 7

$

fo

.

0 S

, o s

. r PP rP r

_

. . .

EA

.

r

._ f f

M o E E

.

T S M Y U _

.

S o E 77 ; 9 S e

7 e 8 v M 9 9 9 9 9 R I 1 1 1 1

E T

,

o T o A Y . .

,

W R D o T

o 72

/ f

/

/

E

.

/'c

.

E 5 v 0 F i H Y Y -

R o Y E h A R i

s 3 L u .

.

t f m

t r v

i t X e e o U A F O . v e

v e

b r A a

. T l v R h A u

.

T R o

.

S d e

f

. o

+

o a o o i i c

o

.

,

l

\ s Y

.

i s

_

o

& Va A w .

4 0 2 o o 8 v

_

7 7 8 8 8 w 9 9 9 9 i

o I 1 1 1 '

1_,- ;

b \Wk2 Q

-

_

_

y N h' %

• '

l

.,,. .

,

I

' \

AUXILIARY FEEDWATER PUMPS l

FAIL TO START EVENTS EVENT DATE LER PUMP DESCRIPTION 08/02/85 85-10 . Diesel Pump failed to start due to faulty potentio-meter in the speed control circui /20/84 84-16 Diesel Following a reactor trip, the turbine-driven Turbine pump started but tripped on low suction pressure af ter 7 minutes. The diesel pump failed to start. Low lube oil pressure time delay was set at 0 vice 25 second /03/84 83-22 Turbine Following scheduled maintenance, the pump failed to start remotely due to tripped overload contacts in the trip and throttle valve motor operator circui /22/83 83-02 Diesel AFW pumps were manually shut down due to high steam generator levels. When restart was attempted, the diesel-driven pump would not start because it had locked out on low lube oil pressur Turbine The steam-driven pump started but tripped on overspee /03/80 80-20 Both Both pumps failed to respond to an auto start signal from lo-lo steam generator level following a reactor trip. The auto y start from low-low level was inadvertently disabled due to a wiring erro /03/80 80-18 Diesel Pump would not start during periodic testing *

due to a bad cell in the starting batter /20/80 80-15 Turbine Pump would not start on a manual start signal because the trip and throttle valve did not open. The motor operator torque switch was improperly set following maintenanc /19/76 76-65 Turbine Pump failed to auto start when the main feed i

'

pumps were secured during a normal reactor i

shutdown. The trip and throttle valve was not properly reset following the previous

- shutdow /03/76 76-57 Diesel Pump failed to start on low-low steam l generator level after a reactor trip due to a blown fuse in the automatic start circuit.

i 1 l

l

/ffm aa B

/ < .7oS'7

.

o(. .

.

EVENT DATE LER PUMP DESCRIPTION 02/29/76 76-17 Both Following an auto start signal after a reactor trip, turbine trip and steam generator low-low level signal, both pumps apparently started and immediately tripped l on overspee ~*

..

.

02/27/76 76-16 Turbine Pump started but immediately tripped on J overspeed in three attempts. Apparent l design deficiency in not providing control I oil to the governor during the initial phases of cold start /17/76 76-15 Turbine Pump failed to start during testing because of stem binding in the motor-operated trip and throttle valv /09/76 76-19 Diesel When started automatically in auto speed control, the pump tripped on overspeed due to changes in engine conditions (oil and !

" ster temperatures) since the previous governor alignmen /23/76 76-12 Diesel Pump failed to start on first attempt due to a low lube oil pressure before the pump reached operating speed. A 24-second time delay was added to the low lube oil pressure circui /16/76 76-06 Both A lifted lead to allow steam generator blowdown with the main feed pumps tripped disabled the main feed pump trip auto start signal to the AFW pumps. One pump was manually starte /24/75 76-04 Diesel Following a planned reactor trip, the diesel pump failed to auto start after the main feed pumps were tripped. The fuel racks had not been reset following the previous shutdow /19/75 76-01 Diesel Operator attempted to start the diesel-driven pump after experiencing difficulty in controlling the speed of the turbine-driven pump. The diesel pump failed to start due to improper sequence of switch operations

~~

following the previous shutdown, e han b a, w

. _ . _ _ .. . . .

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

'

.

. .

• .

.

/

AUKILIARY FEEDWATER PUMPS FAIL TO RUN EV2NTS EVENT DATE LER PUMP DESCRIPTION 07/20/85 85-09 Both Pumps tripped on low suction pressu,r /20/83 83-12 Diesel Following a reactor trip and AFW pump auto start, the pump tripped on overspee /28/78 78-28 Diesel Pump was running to maintain steam generator level in Mode 3 when a fitting on the 3/8-inch fuel line rupture /17/77 77-45 Diesel The pump was manually started in Mode 3 but immediately tripped on overspeed due to an out of adjustment microswitch in the speed sensing circui /24/77 77-06 Diesel Both pumps were running in Mode 2 when the diesel pump tripped due to a broken crankshaf '

10/05/76 76-60 Turbine While being tested after maintenance (bearing inspection), the pump tripped on overspeed due to a failure in the governor speed sensor circui /02/76 76-56 Diesel In Mode 1 on September 2, and in Mode 2 on and September 9, the pump tripped on high jacket

, 09/09/76 water temperature. A loose spring in the temperature switch caused a faulty setpoint

1. 1. calibratio <

03/28/76 76-24 Diesel Both pumps were started manually. The diesel engine ran for approximately 30 seconds but tripped on overspeed when the idle hold was removed. The cause was faulty I

governor design resulting in a variable overspeed governor setpoin /14/76 76-14 Turbine The pump was idling on recire in Mode 2 with feedwater supplied by the diesel pump. The

, turbine-driven pump tripped on overspeed due to circuit failures in the governo /14/76 76-13 Diesel In Mode 2, the diesel driver tripped on high cooling water temperature. The water temperature switch was improperly set at 154*F.

!

pg $$

l (?ers.f'?

-- -

C .

o '.' ' .

.

EVENT DATE LER PUMP DESCRIPTION 01/04/76 76-05 Diesel Pump tripped on overspeed due to an electrical connection in the speed governor circuit that had vibrated loos /19/75 76-02 Turbine Pump was secured due to severely pseillating discharge pressure and impulse pr'ssur e Turbine journal and thrust bearings were damaged due to a broken shaft in the AFP lube oil pum .

{ k$ Y

,

( 0:2

/e

F

. , . .

. .

ASSOCIATED LERS (NOT INVOLVING FAIL TO START /RUN)

EVENT DATE LER PUMP DESCRIPTION 11/09/84 85-01 Both Incomplete seismic qualification of AFW control valve control elements. ,,..

03/20/84 84-05 Both Auto start signals disabled while running PICT; B train out of service for maintenanc /15/81 81-32 Diesel QA audit determined that diesel AFW pump day tank level verification had been deleted from a PO /12/80 81-01 Turbine Discovered a blown fuse in the control power circuit for the vent supply fan to the pump roo /11/80 80-24 Diesel While troubleshooting a steam generator sample valves problem, discovered a blown fuse in the automatic start circui /01/80 80-12 Turbine Engineering walkdown identified three missing seismic pipe restraints on a steam drain lin /07/80 80-02 Diesel Routine inspection identified a broken flexible jacket cooling water hos /15/79 79-06 Diesel Routine inspection identified a crack in a fuel line fitting, apparently caused by vibratio /26/78 78-16 Turbine Routine inspection identified a missing seismic restraint on the steam supply lin /17/77 77-46 Turbine The pump was manually tripped when it was no longer needed in Mode 3. A faulty limit switch in the overspeed circuit prevented placing the pump in standby to be ready for auto star /27/77 77-43 Diesel The feedwater isolation valve between the diesel pump and the B steam generator failed in the open positio ..

ARA /3kal f 1188G.885 tte HC+t prs s?'

,,

. ..

,

. . ...

'

• '

INTER OFFICE COMMUNICATION PORTLAND GENERAL ELFLTRIC COMPANY Date August 7, 1985 CAO-30-85 To W. S. Orser From C. A. Olmstead Subj et Results of QA Evaluation of Testing Requirements Associated with AW Pumps and Safety-Related Modifications Completed during the 1985 Refueling Outage As a result of previous low suction pressure problems with the safety-related AW pumps and a recent trip of both punps on July 20,1985, you requested the QA organization to independently evaluate the adequacy of testing requirements in the following areas:

- Modifications involving the AW pumps;

- hodifications involving safety-related equipment completed during the 1985 refueling outage;

- History of AW pump low suction pressure problem The following summarizes the results of our evaluatio To perform our evaluation, completed as-built documentation was reviewed to verify that adequate functional tests were specified to assure satis-factory system operatio It was also verified that adequate time was available for personnel to review the Detailed Construction Package (DCP) for implementing plant modifications and provide adequate test requirements. The average time between DCP issuance by Nuclear Plant Engineering (NPE) and Plant Super-visory review of the associated Maintenance Request was 1-1/2 month Modifications Involving AW Pumps and Safety-Related Equipment Although testing appeared to be adequate, a number of items were identi-fied as weaknesses in the overall test program. These items are listed below for consideration by PGE management: Nuclear Division and plant procedures do not adequately assign the responsibility for providing test requirements to one spe-cific group or organization within PGE. During our review of Maintenance Requests associated with modifications, it was noted 'that test procedures were specified by NPE and subse-quently changed by plant engineers and the cognizant work supervisor involved in the modification proces /t t $

p ore

I ,

.  :

' *

, * . }

CAO-30-85 August 7,1985 Page 2 Nuclear Division and plant procedures do not describe adequate test controls: A formalized method for determining test requirements is not provided for those individuals tasked with this responsibili t Test procedures were noted as being changed on Mainte-nance Requests without documented evidence that the originator agreed with the chang j Numerous field changes were noted as being made to modi-fications without documented evidence of formal evalua-tions of their effects on test requirement . Plant procedure A0-3-9 (Maintenance Requests) does not differen- i tiate between the type of test responsibilities and controls required for normal maintenance or system modifications. A0-3-9 describes test responsibilities and controls for the cognizant work group supe visor but does not address engineering responsi-bilities in this are On the basis of these identified weaknesses, the following recommendations are made by the QA organization to improve controls within the test program: NDP 200-1 (Design Change Control) should be revised to clearly identify the specific organization responsible for establishing test requirement . NDP 200-1 should provide a more definitive / formal test program, including detailed instructions for determining test . require-ments for modifications and maintenance activitie . A0-3-9 (Nkintenance Request) should be revised to dif ferentiate test responsibilities and controls for normal maintencnce and modification l

. .

1 Plant personnel should be reminded of the importance of identi-fying potential plant problems by using the " Request for Eval-uation" (RFE) System. Through discussions with Operations and Maintenance personnel it was implied that various_ problems were previously identified regarding the AFW system over a long period for which no corrective actions were taken. From our review of " Plant Problem Reports" and RFE reports, there was no evidence that this was the cas ' The proposed Maintenance and Reliability Program which estab-lishes the system engineering concept is encourage It is believed that this would be a positive step toward improving the existing process of identifying functional test require-ments for repaired or ' modified equipment / systems. In the majority of cases it appears that only two persons from plant fiYpc aw 0 A ,zcfr

I .

' '

. . .

CAO-30-85 August 7, 1985 Page 3-engineering are involved in the review of modifications for establishing adequate test requirements. These two individuals must have knowledge of all plant disciplines as well as adequate time to perform a satisfactory revie History of AFW Pump Low Suction Pressure Problems The final portion of this report details a history of AFW pump low suction pressure trip problems that occurred since 1979. From the results of our review in this area, it was determined that NPE and plant personnel were aware of the need for testing of both AFW pumps on autostart in 1980. At that time, it was decided to not test both AFW pumps simultaneously. This appeared to be a conscientious decision made by management based on engi-neering judgement and economic restraint A history of the AFW events and our recommendations are noted below for management information and consideration: /30/79 - The RDC 79-063 Design Review Report identified a deficiency. It stated that measurements should be taken at the suction of both AFW pumps during different modes of operatio . 4/21/80 - A meeting was held between Trojan and Generation Engineering staff members. Various solutions to prevent pumps tripping on low suction pressure during auto start were discussed. Difficulties in testing during auto start also were discusse . 5/7/80 - Memorandum SRC-995-80M from Generation Engineering to Trojan General Manager stated that testing by auto starting both AFW pumps would not be don Reasons given were "the difficult plant conditions required and the severe transient which could be imposed upon the plant".

Translation - It's too har . 8/7/80 - A memorandum from PRB Chairman (and approved by plant General thnager) to Generation Engineering requested evaluation of either replacing the existing line with a larger Seismic I qualified pipe or installing a second parallel condensate feed lin . 9/10/80 - In memorandum JWL-383-80M, the Manager, GLAD, responded to the PRB Chairman's memorandum. The memorandum stated that the low pressure suction trip and the throttling of each AFW pump discharge to 650 GPM upon startup was adequate. The memo-randum also stated that the " Operational Test Pro- j gram should demonstrate that the trip system installed functions properly under the full range f/

/ff/dv#^*f{$

of anticipated plant conditions." y [f

/,

f ]

.

.- .' .

CAO-30-85 August' 7, 1985-Page 4 /23/80 - An NRC letter provided the Safety Evaluation Review for the Trojan AFW System. It contained a require-ment to " verify the ability of each safety grade AFW pump to feed all four steam generators follow-ing an extended plant outage." PGE did not interpret this to require a simultaneous auto start of both pump . 9/23/84 - During a FRB meeting, the Plant QA Supervisor asked if _both AFW pumps should be started simultaneously for testing the functioning of the low suction pressure trips. The answer was n The emphasis was placed on testing the operation of the low suc-tion pressure instrumentation. No action or. follow up to this question was evident in subsequent data reviewe Summary The low suction pressure trips were installed to protect the AFW pumps in case of a seismic caused rupture of the suction line or the Condensate Storage Tank or any blockage of the suction lin The net result has been: A degradation in the reliability of the pump A requirement for operator action following an auto start to adjust the AFW flo . Thi AFW pumps were never tested under simultaneous auto start conditions and varying CST levels until July 198 Recommendatio ns Conduct a simultaneous auto start test of both AFW pumps follow-

-

ing extended outages at minimum CST leve . Strengthen the design change process to ensure that all necessary and relevant post installation testing is prescribed. A specific management review of the testing requirements contained in the DCP should be performe . Increase emphasis on management follow-up and coordination of identified potential problems. As noted above, there was ample evidence prior to the July 20, 1985 event that testing of the AFW sys tem may have been inadequate. Additionally, there is evidence that management reviews sometimes took the most expeditious or economical approach to solving problems without sound safety basis,

ey ar

- _ _ _ _ _

.

,1 . . ,

l CAO-30-85 August 7, 1985 Page 5 In conjunction with our evaluation, a concern was raised regarding the rea-son for Main Feed System isolation following a reactor trip and Lo T' av The basis for our concern is described below with a recommendation for NSRD action in this are Main Feed System Isolation Isolation of the main feed system occurs during all reactor trips. The iso-lation happens when Tavg reaches 564* F as the plant cools due to control rod insertion and supply of feed water to the steam generators. The primary operational concern during this time is to restore level indication in at least one steam generator to the narrow range. The plant temperature is expected to stabilize at the no-load Tavg of 557' F. In practice, the steam generator fill is started with the main feed pumps and completed with AFW pumps. During the plant trip on July 20, 1985, the P-2500 computer data shows that plant temperature stabilized at 554' F about seven minutes af ter the trip. This sequence of events produces the following results: Reactor plant temperature drop is not impeded in its progress from full load to no load Tav . The AFW system is challenged by each reactor trip - regardless of reason for the tri . The feed and condensate systems experience a severe water hamme Operators routinely depart the area upon announcement of a reactor tri Recommendation NSRD review the reasons for the main feed system isolation following a reactor trip and Lo Tav It should be determined if the reasons are still valid in light of current plant operating practices and the undesirable consequences of the feed system isolatio Consideration should also be given to reducing challenges to the ESF equipment when the non-ESF AFW pump is availabl b -c. $(LD/

J CHB:mih '

c: TNOB Members R. W. Griebe F. H. Lamoureaux J. W. Lentsch W. S. Orser R. L. Steele T. D. Walt B. D. Withers C. P. Yundt P. A. Mc}dllan NQAD File (Review, 8/85) / ! do'f"d"

e [0 6 QARID N .3.2.15 T

. .