ML17262A737

Svc Water Sys Operational Performance Insp 50-244/91-201 on 911202-20.Weaknessess Identified.Major Areas Inspected: Mechanical Design Review of Svc Water Sys,Detailed Sys Walkdowns & Review of Sys Operation,Maint & Surveillance

ML17262A737
Person / Time
Site:	Ginna
Issue date:	01/27/1992
From:	Gramm R, Imbro E, Mark Miller Office of Nuclear Reactor Regulation
To:
Shared Package
ML17262A736	List: ML17262A735 ML17262A737
References
50-244-91-201, NUDOCS 9202050439
Download: ML17262A737 (89)

See also: IR 05000244/1991201

Text

i

U.S.

NUCLEAR RlM'UIATORY OCNNISSION

OFFICE OF NUCLEAR REACIOR REGULATION

Division of Reactor Inspection and Safeguards

NRC Inspection Report:

50-244/91-201

License No.:

DFR-18

Ekocket No.:

50-244

Licensee:

Rochester

Gas and Electric Corporation

Facility Name:

Ginna Nuclear Power Station

Inspection Conducted:

December

2 through 20,

1991

Prepared by:

Inspection Team:

Melanic A. Miller, Team leader,

NRR

Suresh K. Chaudhary,

Sr. Reactor Engineer,

RI

'Ihomas A. Moslak, Sr. Resident Inspector,

RI

Donna Skay, General Engineer,

NRR

James

E. Tatum, Sr. Reactor Engineer,

NRR

Jerzy Parkitny,

NRC Consultant,

AECL

44

g LLk

Melanic A. Miller, Team leader

Team Inspection Section

B

Special Inspection Branch

Division of Reactor Inspection

and Safeguards

0

ice of Nuclear Reactor Regulation

Reviewed by:

A.

Robert A. Gramm, Section Chief

Team Inspection Section

B

Special Inspection Branch

Division of Reactor Inspection

ard Safeguards

Office of Nuclear Reactor Regulation

~p

~by:

Eugene V. Imbro, Chief

Special Inspection Branch

Division of Reactor Inspection

and Safeguards

Office of Nuclear Reactor Regulation

te

Date

Date

9202050439

920130

PDR

ADDCK 05000244

I

ii

8

PDR

'Ihe Special Inspection Branch of the U.S. Nuclear ~atoxy Ocaanission

performed a pilat team inspectian,

the Service Rater System Op~tianal

Performance Inspection, at the Ginna Nuclear Poem Station, during the period

December

2 thrn)gh 20, 1991.

%he inspection included a me~nical design review of the sexvice water

system

(SNS); detailed system waMkams; review of system opexation,

maintenance,

and surveillance;

and assessment

of quality assurance

and

corrective actions related to the SNS.

Ihe team also assessed

the licensee's

implementation of actions required by Generic Letter 89-13, "Sexvice Rater

Sy.~n Problems Affecting Safety-Related

Equipnent," as well as system

unavailability to gain additianal insights for prababilistic risk assessment

application.

After inspectirg the

SNS in detail, the team identified several deficiencies

and we dresses.

%he current technical specification requiring that two of the

four sexvice water

(SW) puris be cperable is not sufficient to account for

single-failure assumptions

since two punps are required to cperate dump the

accident recirculation phase. ~ team also identified a failure to

adequately control engineering docunents

and a nunher of inaccuracies in the

Updated Final Safety Analysis Report.

'Ihe team found two exzcaples of

tive operational ~osophy:

(1) the practice of not entering a

limiting condition for operation when performing equi@nant testing and of

taking redundant equignent out of service simiLtaneously and (2) the practice

of performing long-term maintenance of SW panps durirg plant operation.

lhe

team identified a greater need of the licensee to understand

and adjust the

system hydraulic flow, in particular that flmr through the diese1 generator

coolers.

Further, the licensee

had not cansidered the single failure of a

pump discharge

check valve,

and its implementation of Generic Letter 89-13

actions was not entirely adequate, ~xaially that of Action IV related to

verification of the SNS licensing basis.

1

review and corrective actions are necks;ey:

(1) mamlysis of the SNS

hydraulic model and applicatian of its reacts to the system,

(2) assessment

of the sirgle failure of a pump discharge ~~ valve,

(3) establishment of

the appmpriate

system lear-pressure

setpoint,

(4) assess'.nt

of any single

failures that would impair system operation due to the cross~nnected

configuration and relly.ation as necessary,

(5) evaluation of preoperational

test remits,

and

(6) inclusion of the ~rapriate

number of operable

pumps in

the Technical Specifications.

%he abave issues give rise to sane uncertainty

as to system operability.

However,

we conclude that, with the interim actions

taken for items

(2) and (6), a presumptian of operability is warranted while

the issues

above are fully resolved.

%he licensee considered

item (2) to be autside its design basis.

In addition,

the licensee considered the need to maintain four SW purges operable outside

its design basis.

The NRC is continuing to evaluate bath of these issues with

respect to design basis considerations.

I

g

i ~

~

~

1

0

Strengths regarding the system were also identified.

Operators were

kncvledgeable of the system alignment, operation,

and cmponent location.

Operations pnxm3ures were detailed and agp~xpriate,

and operator training was

a strength.

In maintenarme,

procedures

were generally good, personnel

were

akilled and- kncvledgmble,

and the training program was very good.

9he

licensee's

develogtent

'of its Service Water Systxun Reliability Optimimtion

Program document represented

a positive ~roach to addressing

SNS issues.

k

~

~

g

~

~

TAKZ OF CNZ/%IS

EXECU'E SUMMARY

~

~

~

~

~

~

~

~

~

~

~

~

~

~

. ~

~

~

~

~

~

~

~

~

~

~

~

~

~

1. 0

INSPECTION SCOPE AND OBHCZZVES

2. 0

SYSTEM DIXCRIETION

3.0

DEZAIIZD INSPECTION FINDINGS

3. 1

Mechanical Design Review

3.1. 1

Document Control

3.l. 2

Syahnn %hernial-Hydraulic Design

3.1.2.1

3.1.2.2

3.1.2.3

3.1.2.4

3.1.2.5

Hydraulic Model

Opponent Cooling Water Heat E>ochangers

Spent Fue1 Pool Heat Exchanger.....

Diesel Generator Coolers........

5

6

6

8

9

3.1.3

3.1.4

3.1.5

3.1.6

3.1.7

3.1.8

Caamon-Mode

and Single-Failure Analysis

~tial for Service Water Pump Overheating

Heat Exchanger Heat Transfer Testing Program

Modification Review

UP.AR Discrepancies

Concluslons

~

o

~

~

~

~

~

~

~

~

~

~

~

~

~

~

10ll

11

12

12

13

3 ~ 2

Of%rat lons

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

14

3.2.1

Systxan Configuration WaDackams

3.2. 2 Ope~tions Procedures

3.2. 3

Operator WalMowns

3.2.4

Operations Training

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~ t

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

, ~

~

.

~ ,, ~

14

15

16

16

3.2.4.1

Training Programs

3.2.4.2

Training Materials

.

3.2.5

Conclusions

~

~

~

~

~

~

~

~

~

0

~

~

~

~

~

~

~

~

~

~

16

17

18

3. 3

Maintenance

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

18

3.3.1

3.3.2

3.3.3

3.3.4

3.3.5

3.3.6

Procedures

~

~

~

~

~

~

Activities Observed

Taxiing of Preventive

Tra 'n1ng

e

~

~

~

~

~

System Unavailability

Conclusions

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

and Corrective

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

18

19

19

20

21

21

g

A

s

g

J'1

3.4

Surveillance

anal Test~

22

3.4'.1

3.4.2

3.4.3

3.4.4

3.4.5

3.4.6

Preogeraticnal Test~

Surveillance Procedures

Inservice Testiag of Bmps and Valves

Piping arxl Oaqmnent Xnspectians

Biofouling Control and Surveillance During

Opera

'ion

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

Oonclusians

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

System

~

~

~

~

~

0

~

~

~

~

~

~

~

~

22

23

23

24

25

25

3. 5 Quality Assurance

and Oorrective Actions

~

~

~

1

~

1

~

>

~

~

~

26

3.5. 1

Root-Cause Evaluations

3.5.2

Occtparison of Firdi~

3.5.3

Oonclusians

~

~

~

~

~

~

~

~

~

~

~

~

~ t

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

~

0

~

~

~

~

~

~

~

~

~

~

~

~

~

~

26

26

27

4. 0

EXIT MEZFING

~

~

~

~

~

~

~

~

~

~

~

~

~

~

27

APPENDIX A

SUN%9K OF INSPECTION FINDINGS

APPENDIX B

EXZI'EETING ATZEKRNCE IZST - DECKER 20, 1991

APPENDIX C

ABHREV32KIONS

1 GINNA SERVICE WAR SYSTEM........ ~... ~ ~........... ~.....

~ .. ~..........

2

1.0 INS~GN SCOPE AND OB3ECTXVES

Fran December

2 through 20, 1991, the U.S. Nuclear Regulatory Occakssion

(NRC}

staff performed an announced pilot Service Water System Operational

Performance Inspection

(SNSOPI) at the Girna Nuclear Pc@mr Statian.

'Ihe

w

and surveillance of the service water systen

(SWS).

In additian, the team

evaluated quality assurance

and corrective action proc'ects

related to

the SWS.

lhe primary objectives of the SNSOPI were to:

(1)

Asm~ the operational perfornance of the

SNS thmugh an in-depth review

of mec~nical sy.~ns'unctianal

design and thermalWydraulic

f

i ~,,

1

p

their implementation;

and cinerator trainirg an the SNS;

(2)

Verify that the functional designs

and operational controls of the

SWS

P

" l

~ i

and that SNS cxapanents

are cyerated in a manner cansistent with their

design bases;

(3)

Assess the licensee's

planned or ccapleted actions in respanse to Generic

Ietter 89-13; and

t!W

surveillance,

and cxxnponent failures.

%he team has characterized its fincULngs within this report as deficiencies,

unresolved items, or observations.

Deficiencies are either

(1) the parent

H

failure of the licensee to satisfy a written cxmmitment or to conform to the

provisions of applicable codes,

standards,

guides, or other a~~ industry

practices

when the canmitment has not been made a legally bindirg requinanent.

Unresolved items involve a concern about which more information is requuxd to

~mctain whether it is acceptable or deficient.

Appropriate items willbe

reviewed by the NRC regional office for any enforcetmnt actions.

Observations

are items considered appropriate to call to licensee

management

attention, but

which have no ~rent dh~t regulatory basis.

2. 0

SYSTEM DESCRIPTION

'Ihe Ginna SWS, depicted in Figure 1, is an open-loop system that takes suction

from Lake Ontario via the screenhouse

ard ~lies cooling water to various

turbine plant loads as well as auxiliary reactor plant loads. ~ SWS

discharges

back into Lake Ontario via the plant discharge canal.

%he

SWS

consists of four SW pumps, two loop supply headers,

isolation valves,

a normal

discharge header,

ard an alternate discharge header.

Allportions of the

SWS

serving safety-related

equipnent are designed

as seismic Category I. All

other portions of the

SWS serving non-safety-related

loads are designated

as

nonseismic and are capable of being isolated fran the safety-related portion

of the system.

'Ihe SWS provides cooling water to the follcarirg equipnent:

two cmponent

cooling water heat exchangers

(Ot:RHXs), two diesel generator

(DG) lubricating

oil an2 jacket water cooler sets,

two spent fuel pool heat exchangers

SCACfCCWfLL

A'OOP CCfAOf4

'd'OOP CcfAOf4

'TI

CCDC

CD

Q

VI

CD

CD

PD

Vl

C

CD

3

c)

O

'P

3P

(5

IN(o

gOy

DPz)

c

$

c

c

-

c

C

I

I

I tI" I

I PP

I

I peal

I

I

p (o

m

OO

$ C

~

(

~

li!

V

$~

t )~

b

JW.f

air coolers

(CACs), the safety-related

pump motor roan coolers (residual heat

~,

tllb

1

other non-safety-related

SW loads.

Manual valves are provided to isolate

varims canponents

when necessary.

%be SNS is also the alternate water source

for the auxiliary feahrater

pmIIps and the primary scmxe for the stardby

auxiliary feaster pcs.

I

'

, two~ge, cerrtrifugal pumps with a rated fleer of 5300 ~.

Plant load

requirements dictate two or three

SW puttps for normal full load and, for

accident conditions,

one

SW pump during the injectictn ~e ard two SW puaps

during the recirculation phase.

%he SNS cooley water supply consists of two trains

(A ard B) with two SW

pumps supplying each train.

Ihe SNS trains are cross-tied

and cooperate

as a

single system with main cross~nnect

valves at the DGs (4760 and 4669) ard

CACs (4756 and 4639) normally open.

She only train separation is normally

closed cross-cxxnect valves

(4610 and 4779) located in the main supply loop.

%he SNS can be isolated into two systems by manually closing the above-

mentioned cross~neet

valves with the exa~ion of the SI pumps which can

only be isolated so that one train would provide their total fleer.

%he SNS cooling loads sham di.~ibutian piping that can be fed fran either

one of the SNS headers

by the cpen cross~neet

valves (with the mxeption of

the CCNHXs and the standby auxiliary feechmter supply lines which are only

effectively cross-connected

through normally closed valves 4610 ard 4779) .

In

the event of a malfunction in either subsystem, it is possible to manually

isolate the portion of the subsystan affected ard maintain service to all

other portions of the plant.

%he engineered safety features equiprent can be

split between the two subsystems with the exception of the SI punps.

Double valves in series

(one papered

by train A, the other bjy train B) are

safety injection signal

(SIS) coincident with an urdervoltage or lossmf-

offsite-pc~ cordition.

On an SIS only, non-safety-related

loads are not

isolated.

~ OCNHXs, SFRiX A, SI pump thnxst bearings,

standby auxiliary feedwater

pump

man coolers,

and the safety-related

pump man coolers located in the

auxiliary building have redurdant

SNS discharge lines that empty ttuaugh a

seismic Category I discharge stature to Deer Cretic and then to lake Ontario.

discharge is only used if the normal discharge

becanes

inoperable.

the

SW puatps are connected to the 480 V train A and train B buses

(buses

17

ard 18, mspx<ively) that can be ~lied by the DGs in the event of a loss

of offsite power.

One

SW pump per bus is autanatically started on bus

urdervoltage or SIS.

For an urdexvoltage ~tion with SIS, the SNS is

configured to ~ly cooling water only to the required emergency

systems.

1

~

I

~

3.0

KSHPQIED INSPECTICN FINDINGS

3.1

Mechanical Design Review

She mechanical design review of the Ginna SNS included detara6aaticn of

whether the system's design basis,

design assuaptices,

calaQaticns,

analyses,

boundary corditions,

and medels met licensing ccamitments

and regulatory

requirements

and were capable of meetirg the thermal ard hp3raulic performance

specificatices

during accident ar ather abnormal cxzxKtians.

9he team also

reviewed the system design vulnerabilities and a sample of implemented

edification packages.

During the ~~ion, the licensee provided the team with limited design

dcx~tatian supporting the design basis of the SNS.

She licensee did not

have available a list of design calculations or any formal listing of SNS

design documents.

%he lack of a donunented design basis could be a

shortcmung

when modifying the plant Mxlware, software, or pracxdures ard

represented

a weakness.

3.1.1

Document Control

During the review of design calculations, modification packages,

and answers

provided in response to questions,

the team noted that the licensee

was not

properly controlling, verifying, and a~inp design reports, calculations,

or analyses.

For example,

(1)

%he NUS Corporation calculation, 'Ryiraulic Analysis of the Service Water

System,>> dated February 1988, supporting Engineering Work Request

(ENR)

1594 was submitted to the licensee marked <<preliminary, for review ard

comment

in March 198 8 ~

As of DBcGlk)ez

1991 g the analysis

had nat beell

reviewed and accept

by the licensee,

but was being used for hydraulic

analysis

and balancing of the

SNS following installation of SFFSX B.

(2)

Bechtel-KHU Report,

"Heat Zoad Capacity/Design Margin Analysis for RED,

CCNHX, SF&K, Non-Regenerative

HX,>> Job No. 20031,

Rev. 1, dated

January

22,

1989, did nat have any indicatian that it. had been reviewed

and accepted

as mguired by Section 3, >>Configuration Control," of the

licensee's

Quality Assurance Manual, Rev. 13, dated Noveakar 1, 1986.

(3)

Rochester

Gas ~ Electric Corporation

(RG&E) design analysis document

entitled>>Insitu Motor Zaad Detenninatians>> for ENR 4232,

<<SNS

Pump Rotor

Studies," dated July 15, 1986, included inca~ assunptians

based

on a

law slip motor ar6 inappropriate equatians.

Consequently,

the results

were also incorrect.

Ihe licensee

had known this for mere than a year;

however, action to revise the calculation and conclusions

was not

initiated.

'Ihe incorrect assumptions

had not been identified by the

design verification process.

'Ihe licensee

s actions are discussed in

Section 3.1.4.

(4)

Two copies of the RG&E design analysis ~rting ERR 3689

<<Containment

Fan Cooler Air Flow,>> Rev. 0, dated June 4, 1984, were given to the team.

One contained handwritten notes and corrections,

the otter did not.

We

licensee stated that.the correct design analysis was the ane with

~

~

~

handwritten additians even though they had not been incozporated into a

new revisions

(5)

Rg&E design analysis su~rting XHR 4658~W09,

Minimun Diesel

!!

dated July 25, 1991, was poured to justify DG operas.ity with high

differential pressure

(dp) across the coolers due to zebra mussel tube

plugging.

The calculatians

were prepared based

an data fran the Gild~

and Associates Billof Material No. 216048 which did not use the

~licable design conditians, e.g., cooling water inlet tarperatuze

ard

heat load.

'Ihe incorrect design values resulted in incozzect assurptions

and results.

'Ihis calculation is further discussed in Section 3.1.2.4.

'Ihe licensee stated that all dcxxment deficiencies identified by the team

would be properly addressed

and corrected.

%he team identwfied items

(3) and

(5)

(above)

as instances in which the lack of ~~riate donunent control and

verification resulted in technically inadequate

analyses.

Ghe failure to

prcperly review and control engineering

documents is identified as Deficiency

No. 91-201-01 in 2gperxtu( A.

'Ihe absence of effective document control and verification for the above cases

indicated a lack of a~ion to detail in canplyirg with documentation

!!

!

situation,

based

on the licensee's

zeanalysis,

where these weaknesses

had

impaired the capability of the SHS to perform its safety function.

3.1.2

System 'Ihezmal-Hydraulic Design

3.1.2. 1 Hy:haulic Model

'Ihe licensee

had a fairly detailed hydraulic model of the SNS that was

developed by NUS Corporation to ~rt installation of a new SFERX

(EHR 1594).

Ihe team reviewed the hydraulic model and concluded that it was

relatively accurate for most of the process equi~t and piping.

9he major

limitation associated with the model was the limited anaunt of actual plant

data used to verify the fleer resistance.

For emcnple, the dp was!

re+At of a single resistance mxltiplication factor having been used for all

equipnent arxi piping which was too high a value for these coolers and other

heat exchangers

(see Section 3.1.2.4 for naze details) .

In response to these

findings, the licensee

informed the team that it was planning to upgrade the

hydraulic resistances

in the model associated with the major flee paths in the

safety-related portion of the SHS.

As a minizxnn, the licensee will detezmine

the actual f1m'nd pressure

drop for the flow paths thigh the Cps,

CCHHXs,

and

DG coolers,

and apply an appropriate resistance multiplication factor in

the model.

Me pressure

drop and flow data would then be used to adjust

actual system flaws by setting throttling system valves as a~ropriate. It is

the team's understanding that the licensee will submit a letter to NRC

outlining its plans and schedule for inpmvtng its flow model and ~lying

those zesults to actual system flow balancing.

this model and flow balancing

is Unresolved Item No. 91-201W2 in Agpendhc A.

lhe team reviewed the SNS flows with re@met to design-basis-accident

conditions.

The injection phase requires

one

SW pump and the recizculation

5

1

~

<

~

~

~

y

phase requiem two pumps.

Any sir@le

SW pump falling an SIS ccahined with

an undervoltage event is capable

(using anergency power) of supplying the

required cooling capacity to the CACs,

DG coolers,

and auxiliary f~ter

pumps during the injectian phase.

Two SW pzqps are required during the

recinulatian ~e because of the additional load of the (XRKs.

Par the two

phases,

the flaw di.~ibutian to the critical safety canpaoents

appeared to be

ackquate.

3.1.2.2

Ccapanent

Cooling Water Heat Zhzhangers

'Ihe team reviewed available design calculations,

a general assembly drawing of

the OCRHXs, and the proauenant

and verdor specificatians

and their

specification sheets.

'Xhe team noh'hat the proaument specification

(Westirghause Specification No. 676228) required the CORK to be designed

and

KY

'RW'ASME)

Boiler and Pressure

Vessel

Oode Section III. Instead, the heat

exchanger

was designed

and marmfactured to the ASME Code Sectian VIII

requiranents

(vendor specification ~mt 54-Z-70135).

93m licen ae possessed

no documentation to support this code change, but stated that based

on the

original design phil~y, ASME Code Sectian VIIIwas appropriate for the

design of the OZRXs.

The code difference had been previously reviewed and

accepted by the NRC staff within the framework of its review of Systematic

Evaluation Prcxpam

(SEP) Topic III-1, "Classification of Stxuctures,

~nents,

and System."

'Ihe team performed independent sizing calculations of the CCNECs and concluded

that they were adequately sized.

Hal~er, at the normal flaw rates, there was

a potential for flaw-irduced vibratian.

Qhe team requited the maxim flaw

rates for the shell side of the

OCRHX to make an assessment

of the potential

for vibration-related tube damage.

%he licensee

was unable to pravide this

information.

'Ihis is Observation

No. 91-20103 in Appendix A.

~ team also observed that Bechtel-MJ Report No. 20031 arrived at a similar

conclusion regarding potential for flaw-irx2uced tube vibration and resulting

potential for tube deterioration due to fretting.

Ghe team concluded that the

licensee

had not given this report adequate attention

(see Section 3.1.1) .

Instead,

the licensee relied an a later ABB-Impell report

(No. 095~66~01),

which used simplified manual calculation methods ard did not identify the

potential for flaw-irxhmd tube vibration.

'Ihe team also reviewed the COAX eddy-c~nt inspectian procedure,

tube

plugging criteria, and tube pluggixg procedures

and found them acceI~le

except for nat irdicatmp an eddy~rrent testing frequency.

'Xhe licensee

stated that the CGA will be eddy-current-tested

every 10 years. ~ team

indicated that a 10-year eddy-cd~ testing frequency with no ather

inspections or observations in the interim may not be sufficient to pravide

adequate

warning of tube thinning or fretting caused by flaw-induced

vibration.

3.1.2.3

Spent Fuel Pool Heat Exchanger

'Ihe team reviewed available design calculations for the newly installed

SFRiX B, a general

assenibly drawing of the SFHKs A and B, and the procurement

4

I

5

I

IL

'

j4

~

~

I

~

and vendor specificatices

and associated specification sheets for both heat

exchange.

'Ihe team noted the follmring:

(1)

the procurement specificatian for SFFRX B (like that for SFHiX A)

II ~

fg

fabrication of the tube side.

Ihe verxhr's data sheet far SFFHX B

(Joseph Oat Oorporatian Job No. 2427) indicated that the heat e)zharx3er

I

. g

licensee pmvided additional design doaments to support that bath the

shell and the tube sides of the SF'

were designed

and manufactured to

be corrected accordingly.

9he team reviewed the additional evidence

and

found it acceptable.

I ~ I

provisions to permit easy disassembly of the tube hurdle for inspection

and cleaniLng of the shell side

(SNS side) of the tubes. ~ SFFHX B

I~I

gW

hurdle rival.

'Ihe licensee ~xmded that althm~ the SFRiX B had

provisions to allow shell and tube burxQe disassembly,

the original

design

(welded nozzle cannectians,

adjacent equignent) did not

specifically consider disassembly for cleaning. ~ licensee also stated

that if disassanbly for cleaning were required it could be performed,

albeit at a greater cost and effort to grind and ~d pipe cannections

atxi move adjacent equi~ent.

'Ihis was a~~le to the team.

'Ihe team performed independent sizing calculatians of the SFHK B arxl

concluded that it was adequately sized.

Hn~er, at the design flow rates

there was a potential for flew-induced vibration. ~ team also observed that

Bechtel-KHU Report No. 20031 arrived at a similar conclusian regarding

potential for flow-induced tube vibratian ard resulting potential for tube

deterioration due to fretting.

%he team concluded that the licensee had not

given this report adequate attention

(see ~on 3.1.1) .

Instead, the

licensee relied on a later ABB-spell report

(No. 095~66~01),

which used

simplified manual calculation methods and did not identify the potential for

flaw-induced tube vibration.

'Ihe team also reviewed the SF'ddy-current

inspection procedure,

tube

plugging criteria, arxl tube plugging procedures

and found them acceptable

except for not indicatirg an eddy-current testirg frequency. ~ licensee

stated that the SFERX will be eddy-current-tested

every 10 years. ~ team

indicated that a 10-year eddy-current testirg frequency with no ather

oections or cbservations in the interim may not be sufficient to provide

adequate

warning of tube thinning or fretting caused by flaw-induced

vibration.

'Ihe team noted that SF'

was no longer classified as safety related

follcaring the installation of SFERX B.

However,

SFFHX A remained a part of

the Section IIIpressure

boundary and served as a backup heat exchanger to

SFPHX B.

As such, it appears that SFFHX A should retain its safety-related

classification and be subject to the applicable code ard inspection

I

I

in Appendix A.

'L

~

1

~

~

~

BI

0

l~

If

~

~

~

3.1.2.4

Diesel Generator Coolers

~

~

~

~

~

~

~

~

~

~

~

'Ihe team reviewed available design calculaticas,

a geneva

assenhly drawing of

the

DG lubricating oil and jacket water coolers,

and the vendor specification

ard its specificatian st~.

%he team made the follawing absezvatians:

(1)

'Ihe calculation,

Minimm> Diesel Generator Jacket Cooler & Zube Oil

y

~ztirg ERR 4658~-009

was perfozmed to assess

the impact of changes

in SHS flow due to zebra nussel clogging an jacket water cooler

performance.

these calculations were based

on input data fnxn the

original Gilbert and Associates Bill of Material for the

DG coolers,

which g as already mentioned in Section

3 ~ 1 . 1, did not reflect the design

corxlitions.

For instance,

the SWS inlet tenperature to the oil cooler

should have been 80'F, not 75'F as used in the calculation.

Also, the

licensee

assumed that zebra mussel blockage decreased

SNS flaw to all

tubes uniformly without plugging any tubes in the first tube pass of

these two tube pass ooolers.

By assuming sane ccaplete tube blockage,

as

appeared

reasonable

on the basis of pictures of the actual zebra mussel

blockage,

heat transfer capability would have been rerhxxd.

(2)

The team noticed that the

DG coolers did nat have tube thinrung criteria

to irxhcate when tubes shauld be played.

'Ihe licensee develcped the

tube thinning criteria and gave the team an acceptable

preliminazy

analysis.

%he licensee

intended to formalize the analysis in the near

future.

'Ihe licensee also stated that both

DG coolers would be eddy-

currmt inspected

on an annual basis.

Gxtpletion of this analysis is

part of Observation No. 91-201M5 in Appendix A.

Me team performed independent sizing calculatians of the lubricating oil arxl

jacket water coolers ard concluded that both coolers ~~ed to be sized

adequately for the original design carditians.

Hater, at these conditions,

the jacket water cooler did nat appear to have any design mazgin. ~~fore,

any tube plugging at the design canditians would not be peanitted because the

heat transfer capability would not be sufficient.

However, at aperating

conditions, the heat duty recp.'unnnents

are less.

For example, at the worst-

case opemting conditions, the jacket water cooler's heat duty requirenents

would be reduced

(per the vendor's heat balance)

fran the design value of

3.924 MHIU/hr to only about 3.0 MHIU/hr rejected to the jacket water cooler.

'Ihis would allow the jacket water coolers to cperate with a limited number of

tubes plugged, either mechanically or by zebra mussels.

'Ihe licensee repeated

the jacket water coolers'perability calculations using the co~ data ard

assuming that zebra rmssels would block apprcochnately

66% to 764 (first tube

pass only) of the tubes in the jacket water cooler.

%he team reviewed the

prelimlunary calculations

and concluded that they were based

an oozzect design

parameters

arxl an adequate level of conservatism for design assumptions (i.e.,

numb'f tubes plugged) .

'Ihe licensee will formalize this calculation ard

amerd

(by adding an explanatozy note) or invalidate the original calculation.

Completion of this analysis is part of Gbsezvation

No. 91-20105 in

Appendix A.

In addition to zeviewing the design donunents,

the team reviewed the

~

~

~

~

administrative requirements for the pressure

drop (i.e., differential

pressure)

across the

DG coolers.

The licensee established

an adm:inistrative

8

L

~

I

~

~

~

1'.

~

~

>

~

~

~

limitof 25 psid across the coolers,

above which action would be taken to

reduce dp.

During the discussian of the pressure drop criteria, the team

noticed that the dp during ncemal cyeraticn was a~mimately 15 psid and was

thus excessive

when cxmpared with the original dp of 5.5 psid that was

required during the precperatianal test.

'Ihe licensee's first xespanse,

that

the increase in dp was largely due to the equipment deterioratica and the

resulting increase in resistance,

was qu~ianed further.

She licensee then

flaw to the

DG coolers were performed; the results

starved a flee of 526 to 556

d.>>.p>>

drop at 320 gran SHS fleer would have been approximately 5.5 psid.

'Ihe test

results indicated that the actual

SNS flaw to the

DG coolers during normal

plant operation was mxh higher than expected ard that this excessive flaw

acoomted for the unusually high dp.

She excessive

fleer to the DG coolers may

prevent scae other equipnent fran getting adequate flmr.

The licensee will

evaluate flmr to the

DG coolers and ather equi~nent as part of the plant data

gathering to cptimize the existing SNS hydraulic model and will adjust

throttle valves as necessary

(see Section 3.1.2).

Due is part of Unresolved

Item No. 91-20102 in Appendix A.

As a result of its design review of the

DG coolers, the team questioned

whether the coolers shmQd be included in the plant's thermal perfonnance

testing prcxp~.

We licensee stated that it would test thermal performance

an both sets of DG oil and jacket water coolers.

3.1.2.5

Contairnnent Air Coolers

'Ihe team reviewed the design analysis ard heat renaval capability of the

'

single

CAC heat renaval capabilities between

50 MBIU/hr (to canply with the

main steamline break analysis)

and 75 MBIU/hr (to canply with the loss-of-

n ~y

47 MBIU/hr, the value currently contained in the UFSAR.

About 1984, the

licensee determined that the CACs were not supplied with the full design

SNS

fleer since part of that fleer was used to cool the CAC fan motors (this fleer

had been previmaly negl~) .

this corditian was not fullycorrec~ in the

Updated Final Safety Analysis Report

(UFSAR) and was identified again in 1991

an9 docune~ted in Potential Oondition Adverse to Quality (PCAQ) Report No.91-048.

In response,

the licensee performed a reanalysis to determine that

the containment pressure integrity would nat be impaired by this diverted

f1m'.

'The reanalysis indicated that, the contairnnent pressure limit of 60 psig

would still be met in all cases,

but identified a number of cases

where heat

removal capability of a CAC would be less than 47 MBIU/hr. ~ reanalysis,

hemmer,

indicated that the contairnent would be subjected to a slightly

longer ard higher tenperature transient than originally determined.

Because

the contairnnent teaqxzature profile changed in oanparison with the original

profile, the team advised the licensee to include consideration of this charge

as part of an updated containment analysis to NRC.

'%he licensee will do so as

part of a licensing autxnittal to support a propo.-.ed barm concentration

mdification due in apprc0tunately

6 months.

%his is Observation No. 91-201-06

in 2gpendix A.

During the CAC design review, the team determined that the licensee had not

calculated condensate

drainage capability.

'Ihe licensee prepared preliminary

9

~

~

~

~

~

4

~

calculatians to subst mtiate that the drainage capability of the ChQs was

satisfactory. ~ team reviewed the calculations and cancluded that the

drainage capability of the coolers was adaymte.

%he team urderstood that the

licensee would formalize these preliminary calculations.

Oaapletion of this

calculatian is part of Obsexvatian No. 91-201W5 in ~ndhc A.

3.1.3

and Single-Failure Analysis

'Ihe team reviewed the system design as patented in the licensee's

UFSAR and

CF W~

conditions.

%he team identified the follawing vulnerabilities:

(1)

Currently, the

SWS has been operated in a czoss~nected

configuration

since 1988.

Ihis aligrunent could render the entire

SWS incperable during

an SI actuation if the discharge cia~ valve for cne of the

SW parries

failed to close upon system realignment so that, if the associated

pump

was not operating, water could be pumped in the ogxxsite dation

throogh the eke~ valve into the

SW intake bay. Its scenario was

identified, but not appropriately analyzed, in a single-failure report

(Altran report).

'Ihe licensee cxxrnitted to more ccNnpletely evaluate this

item, provide campensatozy

me;mures in the interim, and to review the

entire Altran report to ensure similar oversights did not occur.

'Ihis is

Unresolved Item No. 91-201W7 in ~rdix A.

Also, the safety evaluation

supporting the configuration change did nat cansider this failure.

(2)

'Ihe Ginna Technical Specifications

(TS) recp.dred two SW pe~ to be

opezable

(one fran each power train) .

However, the UESAR stated that two

SW punps are required for the recirculation phase following SI and, as

such, the TS were not consistent with the licensing basis when

considering potential single failures.

In its safety evaluation report

(SER) related to SEP Tcpic ZX-3, "Station Service and Cooling Water

Systems," dated Novenkxa 3, 1981, the NRC staff specifically requested

that the licensee address this vulnmahility.

'Ihe licensee had not

licensing basis is identified as Deficiency No. 91-201-08 in A~ndix A.

By metmzandum to the opemtians staff dated December 5, 1991, the

licensee

implemented adminiLstzative controls requizirg at least three

SW

pumps to be curable while the cordition is being evaluated.

Although

this did not eliminate the sir@le-failure vulnerability, it provided an

increased level of confidence that the system would function during an

event.

Also, as a result of the team's concern, the licensee increased

its sensitivity to nuncving the fourth SW pump fran sexvice for extended

periods of time.

To resolve this issue, the licensee

was pursuing

denenstration that one-pump operation durirxy the recirculation phase

would be adequate.

If the licensee intends to support an~np operation

for the recirculation phase,

the updated containment analysis

(see

Section 3.1.2.5)

should also address

scenarios in which the CACs will not

be supplied with an adequate

design flow.

Determination of the

a~ropriate

number of pumps that should be operable as defined in the TS

will be addressed

by the staff in its assesanent

of SEP Topic IX-3.

Refer to Section 3.1.7 for further discussion.

10

(3) ~ SNS was originally aligned with the SW supply to the two DGs split

between the two SW loop hemhers.

Hammer, the alignment for the

SW

supply to the DGs has been clanged to aperate with the cross~nnect

valves bet~ the two locp he.~rs

open. ~ canfiguratian makes both

DGs susceptible to a pipe crack in the safety-related

14-inch locp A

header supply piping.

%he team did nat find that a safety evaluatian had

been performed to assess this canfiguratian change.

(4)

During an SI actuatian,

non-safety-related

beakers are not autanatically

isolated unless thexe is a concurrent urxh~ltage canditian.

9hus,

a

pipe break in the non-safety-related

SW piping could starve safety-

related

SW loads.

I

A gasket failure an the discharge side of any punp cauld potentially

spray dawn all of the

SW punp motors, rendering all of the

SW pumps

inop~~le.

The NRC staff reviewed arxl accept

the SNS for Ginna urger SEP %epic IX-3.

Items (1) and

(3) arise because the current system canfiguratian is different

fran the configuration discussed in the SEP SER.

Item (2) was not fully

adchessed

by the licensee.

Item (4) arises because of the inaocuxacy in the

UFSAR and SEP SER regarding isolation on an SIS. ~se items alter the design

basis of the SNS as stated in the NRC staff's

SER related to SEP Tcpic IX-3

dated November 3, 1981.

Refer to Sectian 3.1.7 for further dimxmsian.

3.1.4

Potential for Service Water Pump Overheatirg

%he team observed that a number of licensee reports and consultant analyses

(e.g., the p~mratianal test report,. BR 4232~

nSNS Punp Motor Studi

the NUS hydraulic report) indicated a potential for the

SW punps to overheat.

'Ihe licensee stated that this was nat a valid cancern because

(1) the

precperational report did nat identify the method of current measurement,

did

not address

motor efficiency, and did not indicate that motor voltage was a

measured value;

(2) the ENR 4232 analysis

was based

on an incorrect assunption

of lear slip for these actors;

and (3) the NUS hydraulic report repeated the

incorrect conclusions fern ERR 4232.

'Ihe licensee had been reviewing the

sizing requirements for an

SW pump/m~r assembly and was canctuctiag tests to

determine the maxhaun capacity of the motor (horsepaMm ard torque) .

Test

results reviewed durirg the inspection did not indicate a punp averheatmp

problem,

and the team concluded that the licensee

was addressing this issue

appropriately.

3.1.5

Heat Exchanger Heat Transfer Testing Prn ram

Action II of Generic letter

(GL) 89-13 requ~ that licensees

conduct a test

procp~ to verify the heat transfer capability of heat exchangers

cooled by

open-cycle

SNSs.

In its Service Water System Reliability Optimization Prawn

(STROP), the licensee canmitted to axplete heat transfer test~ in

accordance with the GL request

(includiLng frequency) for the CXRHXs, SFESX B,

the CACs, and the cooling units for the standby auxiliary f~ter pump zacan.

~

~

~

~

" During the ~mction, the licensee

added the

DG jacket water and lube oil

coolers to the heat transfer test program.

%he licensee had slightly delayed

11

I

~

'

t

al

't ~

~

~

~

4

its initiatian of the heat transfer test pnxgram and had nat satisfied the

schedule stated in the GL, but in sane instances,

the licensee had to install

tlat

1

b

team considered the delay a~Rable.

'lhe licensee's

SNSRDP stated that the ranaining heat exchamgrs that perform

a safety-related

functian would be opened ard inspected in accordance with

9 ~! i

performance testing.

Although the team agreed with this approach,

the

licensee had nat established

a sound basis for the preventive maintenance

frequencies for these heat em%angers. ~ is part of Observation

No. 91-20109 in Apperdix A.

3.1.6 Modification Review

%he team reviewed three modificatian padcages:

(1) MR 3316, 'Mditional

Analysis ard RxLification of Service Water Pumps";

(2)

ENR 2512, "Seianic

gq

%he doamantation

associated with these madificatians was satisfactory even

though:

(1) a number of the cansultant's

reports carrtained in the abave

modification packages

lacked proper review ard a~ance

by the licensee

Y

June

6,, 1984) ard

(2) the donnnent verificatian and control process

was

inadequate for ENR 3689 since the design analysis dated June 19, 1984, had not

been revised to include necessary

correctians.

Selected design documents

fram

ather modification packages,

e.g.,

ENR 4232 and MR 4658, were not adequate,

as discussed in Section 3.1.1.

3.1. 7

UFSAR Discrepancies

In evaluating the adequacy of sy ~n design, the team reviewed the UPSAR

description of the

SNS ard ax@ared it to the installed system. ~ following

discrepancies

were identified:

(1)

'Ihe UFSAR stated that the SNS locp is isolated by normally closed valves

to pravide two ardent sy.~ns with no sizable cross-connections.

We team fourd that, in fact, the SNS was normally cross-cannected

through the 14-inch supply header to the ChCs, ard that the system was

normally cross-connected

by the ~ly headers

(rxxunally 3-irish

diameter) for the equi@rent being supplied by the SNS.

(2)

'Ihe UFSAR stated that all of the engineered safety features equignent is

split between the two systenns

so that anly half of the equipnent would be

affected by a malfunction.

%he team fourd that the system was not being

operated

as a split, or divided, system because the crom-connect valves

were open.

Also, in the case of SW coolie to the thrust bearings for

the three SI pumps, it was nat possible to align the SNS so that ane loop

header pravided cooling to ane SI puap and the other locp beach'ravided

cooling to the ather two SI puris.

(3)

'Ihe UFSAR stated that the SNS is designed to isolate non-safety-related

loads on an SIS.

'Ihe team fourd that the non~fety-related

loads were

only isolated on an SIS coincident with an urdervoltage cordition.

12

(4)

UFSAR Table 9.2-2 did nat accurately reflect the total flow to the CACs

because it neglects the flaw to the CAC fan motors.

%he NRC staff reviewed the canfiguration ard the design of the SNS at Ginna

within the framework of the SEP urger Topic IX-3.

9he team ~uded that

staff aocx~ce of the design and canfiguratian of the Ginna SNS was baal,

in part, an the ability of the SNS to perfozm its function ard nat be

adv~y affected by a single active or passive failure.

Ihe following

system design features were recognized:

(1) nan-critical loads are

autamatically isolated following an SIS (did nat include isolation of the

CCNEM) and

(2) normal sy.~ alignment did not pexmit a single active or

passive failure to result in the loss of sexvice water flaw to redurdant

critical loads.

As discussed

above and in Section 3.1.3, the first corxKtion

was not satisfied by the SNS design and canfiguratian.

%he secard candition

has not been satisfied frcxn 1988 to the present

and also fran initial plant

cperation until sane undeterauned

time when the cross-tie valves were closed.

%he SEP review, ard the licensee after receiving the SEP SER, did not

~nize these ina~cies in the discussion of system design.

Ghe licensee

ccamnitted to svkxnit a letter to NRC identifying inaccuzacies in the staff's

SEP SER.

As needed,

the staff will initiate further review of SEP Topic ZX-3

to address the inacxuzacies.

9he licensee s letter is Observation No. 91-

201-10 ard the UFSAR discrepancies

are Deficiency No. 91-201-11; both are

fourd in Apperdix A.

lhe licensee

was vezy re~mnsive in resolvinp the UFSAR discrepancies

identified by the team, either correctirg them in its Dexmnber 1991 UFSAR

update or irdicatirg that they would be corzeetex2 in the 1992 update.

3.1.8

Conclusions

'Ihe team concluded that. the licensee

was maintaining the SNS design in

accordance with licensing canmitments ard regulatory requirements with the .

exceptions of the deficiencies

and cbsezvations identified in the inspection.

Identified weaknesses

include failure to maintain an accurate

UFSAR consistent

with plant configuration, the lack of adequate cperability calculations for

the

DG jacket water cooler, the lack of cmpmhensive single failure analysis

of the

SW pump discharge

check valve, the irxnnsistmcy of TS with the sir@le

failure licensing basis,

ard the lack of adequate

documentation control.

Action IV of GL 89-13 requ~ that licensees

canfirm that the SNS will

perform its intended function in accordance with the licensixxy basis of the

plant.

the team's cbservations irdicate that the licensee's

review did not

identify those licensing basis items identified by the team ard thus did not

fully satisfy GL 89-13, Action IV. Gris is part of Obsexvation No. 91-201-09

in ApperxUx A.

~noghout the course of the inspection, the team corducted in-depth reviews

of available documentation,

system walkdowns,

and alternate calculations to

detezmine whether the SNS would perform its interxM safety functian.

'Ihe

i

review and corrective actions are necessazy:

13

1

~

~

(1)

%he licensee will reanalyze the SNS flaw balance to ensure that all

safety-related

equipment would be supplied with adecpate

fleer and then

apply results to the system (Section 3.1.2.1) .

(2)

Ghe licensee will analyze the single failure of a punp discharge check

valve.

In the interim, axrpensatazy

maasures (i.e., cyeratars

made aware

of this potential event,

a means of prcapt detectian develcped,

and

operator cxxnpensatoxy action indicated) willbe effected (Sectian 3.1.3) .

(3)

'Ihe licensee willestablish

SW header minhmm pressure

based an minimum

(4)

%he licensee willprovide assurance that no additianal single failures

will affect system operation due to the cross-carmected

canfiguration

(Sectian 3.1.3) .

(5)

%he licensee will assess

the preopezatianal test results and perfoxm

te.~ as necmmary to canfixm adequate

system operation (~an 3.4.1),

alxl

(6)

'Ihe licensee will ensure that the appropriate

number of cyezable purtps

are included in the Mchnical Specifications

(Sectian 3.1.3) .

't

However,

we conclude that, with the interim actians taken for items

(2) and

(6), a presunption of operability is warranted while the issues

abave are

fully resolved.

3.2

Operations

Plant operations

were reviewed to assess

the knowledge of the cgerators

and

the accuracy and canpleteness

of procedures

and training they received

regarding the SNS.

In reviewing the operation of the SNS, the team performed

detailed system waDcdawns;

reviewed the procedures for normal, off-nozmal, and

emexgency corditions; assessed

the conduct of cpezations in the field and

control roam; and evaluated trainirg manuals

and lessan plans.

3.2.1

System Configuration Wal3cdcams

%he team conducted detailed walkdcams of the safety-related portions of the

i

(P&XCs)

(No. 33013-1250,

Sheets

1, 2, ard 3) ard the system lineup procedure,

T-36.1, "Station Service Cooling Watex-Re-der Valve Alignment for Two Zacp

Operation."

The safety-related portions were fourd to be aligned in

accordance with the procedure

and drawings. ~ averall system material

condition reflects its age of operation and its continuous exposure to a

fresh-water envirornnent ard surface condensation.

Unpainted, external

sections of piping systems were corroded,

having pitted or pocketed surfaces

and

metal loss.

Bodies of small, manually apemted valves, flanges, bolts, and

supports were rusting to varying decrees.

'Xhe extent of surface corrosion

reflected the lack of a dedicated preservation

pxogram.

Hmmver, due to the

lm energy nature of the SNS, the system did not appear to be degraded to the

extent that the averall operability would be affected.

One small, manually

opexated valve (V-12502, an SNS packing drain valve to the A and B circulating

14

~

I

~

~

r,

rr

f4

"gjf

r

~

s

~

water puaps)

was faund open with the valve stem ~nically frozen, contrary

to the positian specified in the lineup prcxxdure.

A flarxyxi joint in the

traveling scamn wash, system was leaking slightly.

'Ibese cancKtians have

minor safety significance and were previously identified by the lioensee.

Ghe drawings were detailed and accurate.

'Ihe alignment procedure was also

W

verification of valve positian was rectuired by signoffs thnmghout the

procedure

o

Individual caqmnents

were appropriately identified using calor~ded durable

plastic tags.

Six valve tags were missing,

and ane valve tag was wnzg

(V-4620 was tagged as V-4620A) .

Ihese dis~ancies

had nat been identified

by the licensee's staff during routine plant tours and inspections,

nor while

performinp surveillance or maintenance activities.

3.2.2 ~actions Procedures

'%he team reviewed varicvs procxxtures associated with the

SWS operations

including:

Overall Plant Opmzltians (0), Turbine Plant ~actions

(T),

Reactor Plant Systems Operatices

(S), Abnormal Procedures

(AP), Zhergency

yell

(

), ~ y~

(ECA), Functional Restoration

(FR), Equignent Re~ration

(ER), Site

Contingency

(SC), and Alarm Response

(AR) .

Overall, the majority of

y '~

.

In particular, definitive guidance was provided for locating valves whose

locations were not readily parent.

%he alarm-respanse

pzacechuas

accurately

agreed with annunciator tile labels and system setpoints.

Hcamver,

wealcnesses

were identified in certain procedures.

Specifically:

(1)

In step three of AP-SW.1, "Service Water Leak," operators

were directed

~ to initiate actions should

SWS homier pressure

be 40 psig or less.

'Shee

was no engineering basis for this setpoint; it was chosen as a best

estimate.

'Ihe licensee

was aware of this issue but had taken no action

until requested

by the team. ~ licensee planned to establish

an

a~ropriate 1m'-pressure setpoint based

on an assessnent

that system

loads would continue to be ~lied.

this item is Unresolved Item

No. 91-201-12 in Appendix A.

(2)

'Ihe required actions specified in ARM-9, "AuxiliaryFeed

Pump Cooling

Water Filter High Differential Pressure," in response to high dp on the

auxiliary feecheter

pump cuoling water filter did not identify by number

the valves to be operated.

'Ihe licensee prcxnptly changed the procedure

to identify the valves by number.

(3 )

AR I 17 g

Travel Screen High Differential Zavel 6"," and AR I 25 g "Travel

Screen Emergency High Differential Level 10"," were essentially the same

procedure in content and required actions.

Both procedures required only

that verifications be performed for the identified ~tion and relied

heavily on operator judgment for folio@up actions.

Ihe licensee

processed

changes to improve the p~ure,

inclucUing identifying

w>>* ~

.

Clw

the relevant procedure should additianal actions be necessary.

LL

l

p

L.

t~

< ~

Por ~tians that caqxxmise a reliable source of SH, the team determined

I

actian levels.

9he procedure identified the diverse methods of measuring

water levels and graded responses

based

on level deczeases

that oat affect

the capabilities of plant cooling equipnent.

Actices included evaluating and

~ ~

I

I

I

I I

Implementing Pmcxxhue

(EPIP) 1-0, <<Gima Station Event Evaluation and

Classification," inplementing the applicable ecmxgency procedures

rechx-lng

pm~, ~~ing circulating water pux~, and shutt~ dawn the plant, if

necessaxy

o

3.2.3

Operator

Walkdawns

She team formally intexviewed several licensed senior reactor cyeratoxs,

reactor operators,

and nonlicensed auxiliary aperators

(ADs) to determine the

depth arxl scape of their technical krmrledge xegardirg routine operation of

the SNS, its role in accident mitigatian, and ~licable technical

/

Personnel

interviewed were knowledgeable recpu~ their specificI

F.

alignment and reconfiguration of the SNS during accident canditians.

'Ihese

personnel

were vezy aware of potential flew re.~ictian or plugging problems

that cauld result fnm zebra mussel grcmth.

Every faur hours,

ADs logged

differential pressure

real:bags

across

44 separate plant oarponents that are

~I

management to identify possible flew reduction fran zebra nussel claggirg.

'lhe team had operators walk through pnxxxiural steps,

perform specific tasks,

and explain their response

when implementing procedures

T-36.1, "Station

Service Water-Header Valve Alignment for Two Zaop Operatian<<;I AP-SW.1,

<<Sexvice Water Teak<<;

and ER-DG.2, "Alternate Cooling for the Bnergency Diesel

Generators."

In performing these procedures,

the operators

demonstrated

strict procedur~

adherence,

a sound knmrledge of canponent locations,

and the

skills necess-~ to operate the

SNS equignent.

When challenged by the team to

locate infrequently aperated

SNS valves on the plant's secorx3ary side,

operators readily found the valves using available resources,

including the

plant valve index,

P&IDs, ard associated

procedures.

Khnmgh observing the

operators,

the team concluded that the procedures

could be performed in a

timely manner and that SNS cxmponents were accessible for normal and emexgency

apemtion.

For a limited rabber of elevated valves that were anly accessible

by ladder, there were ladders staged near the valves.

Aft~ the team observed

performance of ER-DG.2, the licensee prcaaptly respanded to a team

renxmendation that two spanner wrenches,

rather than ane,

be available in DG

roam hose cabinets to assist

an operator in more quickly coupling/uncoupling

hose co~ions

when implementing ER-DG.2.

3.2.4

Operations Training

3.2.4.1

Txainirg Paxpmns

I I

I~

I

I

nonlicensed operators

were de.xxibed in the '%6&E Nuclear Training Manual" ard

<<RQ6Z Training Pxograns.<<these

manuals described in detail the training

16

~ I

~

process

and cantents of specific trainirx~ programs.

'Ibe opmatians training

pmclrMn had been accredited by the Institute of Nuclear Pea~ Operations

(INPO) .

Operatians training an the SNS consisted of classxaan,

simulator

scenarios,

in-plant training assignnents,

and job performance measures

(JEAN) .

Separate

classman les.~ plans vere pzavtded far ADs and for licensed

operators.

Mesc lesson plans were sufficiently detailed, describing system

bJ,

p

Y

under emergency ccexhtians.

Licensed persannel also received classman

training on actians to be taken in the event of a loss of service water when

implementing AP-SW.1.

Qm simulator lesson plan dealt with the actians to be followed by control

roan personnel in the event of loss of sexvice water, inclu~ imxxHate and

elemental actions.

'Jhe simulator txaining allawed the operators to

experience

an SW leak and take actians to mitigate the event.

%he in-plant training assigrunent

allowed the operators to study the actions

necessary to pravide alternate cooling to the DGs (Pnxxdure ER-DG.2) .

would be necessaxy

should the

SW perps be unavailable to pzavide cooling to

the DGs when they were neared.

Performance of this task,

as athers,

was

measured using an ~zapriate JM.

'

Auxiliary Operator Qua1ification Stand ud.

'Xbe ADs

performance of actions

contained in various emergency cgerating and abnornel procxRures was measured

against six JMs.

Additionally, ADs were trained an the 'IS requiranents

applicable to the SNS.

Licensed cyeratozs were requalified yearly in

responding to a loss of sexvice water event and received refresher trains on

the

SNS every three years.

ADs received

SNS trainirg annually.

3.2.4.2

Training Materials

Lhe team reviewed

SNS training materials,

includ.'mg relevant lesson plans,

JEKs,

AO qualification standards,

and associated

procedures.

In additian to

describing the system operation and function for the various aperatianal

nedes,

the trainboy materials

emphasized

adherence to plant procedures

when

performing tasks affecting the SNS.

Accordingly, these procedures

weze

incorporated into the lesson plans.

Tzainirg materials were current,

adclzessing recent plant modifications and operational experiences.

Details on

zebra mussel infestation, installation of chlorination equipnmt,

and

replacement of the SNS discharge valve and elbmr for the

CCRHX during the 1991

refueling autage were presented to licen~nonlicensed

pezsonnel in a timely

manner.

Training materials were pnxnptly updated due, in part, to good cxxmunications

between plant departments

and the training organization.

In addition to

formal training change requests that cauld be submitted by plant departments

to improve an a~~ of training, curriculum canmittees,

such as the ane that

specifically addressed

the SNS, routinely met to evaluate and readjust

training content.

'Xhe SNS curriculum cxxnmittee was particulaxly effective in

ensuring that personnel

were receiving pex~ent subject matter with the

appropriate level of detail.

Involveae~ by the txaining ozganizatian in

daily plant planning meetings

and use of a cxmputerized networking system by

tzaining managers

and instre~zs to mmpmR to inquiries and suggestions

also

enhanced the timeliness of training materials.

17

~ ~

'0

)

A

~

t

t'

3.2.5

Canclusians

The team concluded that the aperaticas

axea was a strength.

Operators were

knowledgeable of normal and emergency

system operation,

cazpanent locatian,

Pp

WM i

ll

unresolved item was identified regarding subsitantMtian of an ~zyriate law-

pressure set~int for the system.

Action U of GL 89-13 asked licensees to canfixm that aperating prcredures

and

training involving the SNS is adequate to ensure that safety-related

equignent

cooled by SW will functian as intended and that aperators of this equipment

will perform effectively.

On the basis of its review of specific cperating

p~ures,

practices,

and training documentatian,

the team cancluded that

Action U regarding cperatians materials was agyrcpriately a~lished.

3. 3

Maintenance

p p

~p li

tl i~i

whether the SNS components

and piping were being adequately maintained to

ensure their operability and to detect system equi@Dent that required frequent

maintenance

personnel.

3.3.1

PronRures

l

procedures ~licable to the SHS ard associated

equipment.

'Xhese procedures

addressed

predictive, preventive,

and corrective maintenance activities; i.e.,

~ion and repair of pumps, motors,

and valves; lubrication and adjustrrent

of equipnent;

and periodic overhaul of active canponents.

Althoucg the

procedures generally were well written and sufficiently detailed,

mast did not

include clearly defined acceptance criteria.

The procedures

required that

after the work was finished; but, in most cases,

there were no acceptance

criteria included in the prceedure to detexrnine ifthe measuxed data fell

within the a~able

range.

For example,

Procedure N-11.34.1,

"Fan

Maintenance

and Inspection; Safety Injection Pump A," Rev. 7, did not include

quantitative data for a~ance criteria for bearing temperature

and

Inspection and Maintenance,"

Rev. 16, required that the final torque value of

the flanges be within +5 ft-lb.

Hcamver,

no final value of the torque was

given in the procedure

as an a~ance limit against which the specified

tolerance

was to be ~lied.

In Pnxxdure H-11.24, 'Main Feec~np ~mction

ard MainUMance," Rev. 8, the data sheet of Attachment

2 irdicated the factory

specifications for the pumps, but did not irdicate that these values were

mardatory for ~plding.

Procedure M-15.9, "Clean and Inspect

DG, lube Oil,

Jacket Water Coolers," Rev. 7, did nat contain any aoa~~ance criteria.

the licensee

had initiated a Maintenance

Procedure

Upgrade

Program in 1989 in

which such procedural

weaknesses

as the general lack of a~ance criteria

would be addressed.

The program was scheduled for ocapletian in 1994-1995.

18

3;3.2 Activities Observed

fl~

~

SNS Pump B - MicrIn~ion - Work Package

No. 914 1062

~

SNS Pump B - Breaker Inspection - Work Package

No. 914 0651

~

SNS Pump B Replace Packing Gland Studs - Work Package

No. 902 4081

~

SNS Pump D - Adjust Packing - Work Pa~

No. 910 2046

Overall, the tas)cs weze well planned, cantrolled, ard coozdinated. ~ work

p

.,

'~

hg~p

detailed aperatianal

review chec)Gist,

assuring that the release of specific

cxxnpranise plant safety; specific work instructians unique to the task

performed;

and a supporting department's

review check~eet.

'Ihe team noted

that for the activities obsezved

acceptance criteria were either contained in

P

NQ. tY

quality control department actively monitored and performed requisite

verifications thrnxpxut the obsexved tasks.

Pze-job and post-job briefings

t

as-found and as-left equipmnt condition were terse praviding limited

information to the maintenance analyst during subsequent

review.

Hold tags were properly enployed,

and good procedure adherence

was cbsezved

with technicians

signirx~ off~ as they wexe cxepleted.

Post~intenance

testing was a~zopriately performed when specified.

Technicians were

adequately trained and demonstrated

the skills required to perfozm the

maintenance task.

Me team did observe

a casual

approach to a safety practice

in which technicians

who were helping an electrician install a brealmr in a

480 V bus entered the cabinet without wearing protective glaves.

Mhen this

was pointed out, the licensee took agprcpriate corrective action.

and co~ive maintenance

(CH) actians for the SNS and related systems.

Khese were reviewed to determine the rationale and ahxpxacy of the bases of EK

requirements

and effectiveness of CR activities.

In its review of SNS

activities, the team did not identify any recurrent equi~ent failures or

prcblems due to inadequate

CN.

Hm~er, the team was unable to establish

and

verify the basis for required W frequencies

on pumps, valves,

and motors.

'Ihe team determined that there was no fozmal structured program to tzerd the

adequacy

and effectiveness of recpured

XKs ard CÃs in the area of mechanical

equipnent.

'lhe original M program, which was berg replaced by a

reliability-centered maintenance

(RCM) procpmn,

was developed

an the basis of

individual engineering judgment.

In the area of electrical equipnent,

a

formal tze~

pzocpmn existed to trend the performance of electrical

breakers

and correlate the perfozmance to maintenance activities.

Hm~er,

motors and other rotating equipment were not cavered by a formal trending

program.

%he electrical maintenance

analysts informally created

and

maintained the equipnent hi.~ries to identify any apparent trends;

hcamver,

there was no formal policy guidance in this area.

19

~

~

y

reference,

but detexmiLning an apparent trend fran this data base zequhxrl a

1

.I

!&K

q

analyzed,

and correlated

equ~lpnerrt histary data fran the beginning of service

to 1987.

No formal program had been established within the RCH program to

continue the trending of equi~ant performance to assess

the effectiveness

ard

9 i;,

I

had not been implemented at the time of the in~ian.

As an example of the KR program nat appropriately incorporating data

P

trending proc'or vibration analysis.

Hanover, the vibration masuranents

on the

SW pump motors were nat able to show a trend of vtbratian changes aver

a period of time because the cbtained data anly reflected whether or not

'

t

%he licensee represmtatives

stated that there was a plan to consider and

develop a ocxnprehensive trencting program in the area of machanical equignent

ard to enlarge the scca of the electrical equipnent trenching procj~n.

(~y )f

tt

I

3

for the

SW pumps (except pm@ D which recpured major motor repair) cansisted

mainly of inspections,

averhauls,

and packing adjustamts.

Host activities

for the

DG lube oil ard jacket water coolers cansisted of inspections ard

cleaning.

%he CACs ard their fan nestor coolers eqmrienced leaks whiW

required tube plug maintenance,

patckurg,

and scrne replacement of coils.

Ihe

licensee

had recxgnized the prcblem of leaks and instituted a

repair/replacement

program for these ocaponents until they could be replaced.

Valves experienced

expected age-related

decpadation,

and no sir@le valve had a

significant amount of maintenance.

Despite the lack of a formal maintenance

trending proctram, the work orders reviewed did nat indicate evidence of any

recurring problems or trends that were not being addressed

by the licensee.

3.3.4

Training

lhe training procJram for Ehanical mainbmance personnel

was described by

Procedure

TRC.9

s'Mechanical Maintenance Personnel Training Procgam,<< ~.

3

'Ihe training program

agape +ed rigorous and thorough.

Xhe ~noedure

was

intended for training a new employee with basic education and ability to

became

a journeyman craftsman through structured classrocxn lectures ard

practical on-the-job trainirg.

%he program was divided into ~es with

specific time periods,

learning objectives,

and skill attainment goals.

Achievement of these goals was measured

by objective euuninations ard

assessments

at predetermined milestones.

%he specific academic or skill

coume modules were established with adequate

course content and learning

objectives.

%he lesson plans, tests,

job performamx. measures,

and other

training materials were adequately used by instnmtors during classes.

In addition, a program of continu~ education ard skill maintermme was

implemented to refresh and strengthen the skill and expertise of journ~man-

level personnel.

%his program enghasized the developM.nt of atMitianal skills

specifically identified through plant equipnent history, irdustxy experiences,

20

~

~

I

Ill

,L

I

and new pmctuct develegnent in the industxy.

Ghe licensee's policies ensured

general refresher course annually and ~ved any special training on an as-

n~ basis.

q~li

I

4H

I

was adequate to ensure that safety-related

equignent cooled by sexvice water

W

program and other efforts in this area were vexy good and axe ccmsidered

a

Wt.BJ~

I

appropriately accanplished.

3.3. 5

System Unavailability

~ unavailability of major SWS caqxments

was calculated for use as

statistical data in prababilistic risk assesstnent

(PBA) reviews ard to assess

the safety effects of remcving equipnent fran service.

Because Ginna's

SHS

contained several cross-ties

between redundant equi~ent, entire headers

were

not taken aut of service for maintenance

on a canponent.

Kherefore, the team

focused on the unavailability of the

SW pumps.

Out of service toms were

collected frcm equipnent hold records,

maintenance

work reports,

and limiting-

conDition-for-c~ration

(?CO) control reports for the period August 1989 to

Septenher

1991.

'Ihe team also included punp surveillance te.~ in the aut-

of sexvice time since the test removed the system fran its normal accident

alignment.

'Ihe calculated unavailability of ane of the four pumps during

palmr operation was 27%.

KtMrefoxe,

73% of the time the sy ~n was operating

with four SW pumps available.

Ihe individual pmp unavailabilities were:

pump A

1

9% g pump B

13 ~ 9% ~ pump C

1 3% g and pump D

9

8%

Because

current TS only required that two of the four pushes be operable,

the licensee

performed the majority of maintenance

during power operation, including

reaaving a single pump for lcnp-texm maintenance.

'Ihe team considered this

policy unconservative in view of its finding ~rhine the need for more than

two pumps to be operable,

as discussed in Section 3.1.3, ard requested that

the licensee ~~luate its scheduling practices for larg-texm pump

maintenance

during plant cperation.

'Ihis is CR~exvation No. 91-201-13 in

AppencUx A.

Unavailability of the

DG coolers and CACs due to maintenance or

cleardzx~ on the

SW side of the heat mcchangers

was very law:

0.1W.34 for the

DG coolers and 0-0.14 for each of the CACs.

'Ihe licensee

was in the process of developing its ERA at the time of the

~ion.

In response to the team's concerns regarding equignent out of

service for surveillance ard the need for more than two pumps to be operable,

the licensee revised its HQ. to explicitly model the potential of pumps being

unavailable due to testing and to evaluate the impact of requiring three

SW

pumps to be cpmwble.

3.3. 6

Conclusions

Ihe team concluded that the mairrtenance

program related to the

SWS was

procedures

were well written and canprehensive.

%he documentatian of

completed work was available and clear,

and the maintenance

pexsanne1

were

knowledgeable of procedures

and equignent in the plant.

No ocncerns were

identified during the cbsexvation of maintenance activities.

9he maintenance

21

I

i

ll

p

~'~~

strength.

%here was no structured txendirg program to assess

the adequacy of

M fzequencies

and to deterndae asap~ event degradaticn;

hmmver,

no

unidentified trends were found. ~ team identified an umxnsezvative

~

E- ~~d

&pp

operation.

3.4

Surveillance and Testirg

Surveillance and testing were inspect@6 to detexmine if sufficient testirg had

Cf

t ~

'

periodic suzveillance

and insexvice testing

(IST) wexe adequate to maintain

continued cperability of the system.

In this regard, the team reviewed

p~mrational test procedures

and cmpleted surveillance procedures.

Ghe

team also reviewed the licensee's

IST pxogram and implementing pzacechues.

3.4.1

Precperational

Testing

Preopezational testing of the SNS was ccapleted at Ginna in accordance with

Procedure

SU-4. 17

"House Sexvice Water," dated November 1, 1968.

She team's

review of the campleted precperational test procedure

focused primarily an the

test scope,

methodology,

and results.

'Ihe Ginna

SWS preogexational test was ccxrpleted between May and August 1969,

and the quality of the p~uxe and documented test results reflected the

less-rigorous

standards in effect at that time.

For exanple,

documentation of

system alignment and test data was not rigozms,

a test dier's log was not

maintained,

and a fozmal record of review and a~ance

of the test zesults

was not maintained.

During its review of the Ginna precpmatianal testing, the team made the

follmring observations:

MF

l~ capacity would meet the min:imun fleer requQ~nts for safeguard

operation given the system configuration resulting fran an SIS.

%he

actual test was performed with non-safety-related

loads isolated,

which

was inconsistent with the system alignment for SI actuation.

(2)

Me test did not evaluate sy~ vulnerability to punp runout conchtions.

(3)

Lhe system f1m'alance

was est'ablished

an the basis of three pumps

operating, rather than an the worst-case candition of ane pump operating

and supplyirg all safety-related

and non-safety-related

loads.

(4)

Although two pumps were required to handle the postaccident

decay heat

load, various cambinations of two-pump operation were not tested.

'lhe team concluded that the pzecperatianal test did nat fully dananstzate

that

the

SWS at Ginna could perform its functian during all pa.~lated conditions.

dpW U~

h

pt

substantiate

one-pump cperation during the recirculation phase.

On the basis

of these ~sments,

additional testing may be neechd to support operation in

22

~

~

0

IH,

I*

required system canfiguratians.

%his is Deficiency No. 91-201-14 in

Appendix A.

3.4.2

Surveillance Procedures

4d

1

most part, were limited to valve alignment and punp operability verification,

and insexvice testing of purrps and valves (discussed in Secticn 3.4.3) .

The

W

iW

found that, in general, the procedures

were well written and a~lished

their stated cbjectives.

However, the team also found that intecjrated system

testing per BSSP-2.1,

"Safety Injection Functianal Test," and RSSP

2 2g

"Diesel Generator

Zaad and Safecpmxd Sequence Test," did nat fullytest the

contxol cirn6txy for the

SW pmrps.

Specifically:

{1)

Each

SW pump selector switch (one each for tx~ A and train B) se1ects

one of two pumps to start an an SIS (i.e., twa~iticn switch) .

Testing

was not coordinated to ensure that each of the

SW puap selector switches

.

was test in both switch positions.

(2)

Testing was not performed to ensure that the

SW punp selector switches

were not shorted between the two switch positions.

A short be~mn the

two switch positions would allow two SW pumps to be loaded an a single

DG

during an SI actuation coincident with an unck~~ltage condition.

The licensee planned to revise its procure to actress

item (1) and was

evaluating resolution to item (2) .

3.4.3

Insexvice Testing of Pumps and Valves

'?he team found the licensee's

insexvioe testing

(IST) pxogram for SNS pumps

and valves to be generally cxxnplete, in cxxrpliance with ASME Code Sectian XI,

and adecpmte for detezauning the operability of the systeun.

'Ihe team reviewed

the licensee's

program dated June 7, 1991, ard verified that all requixM SNS

valves and punys were included and that proper suxveillance intervcQ.s were

WW

and VR-17 pertaining to disassembly

and inspection of the pump discharge

check

valves,

had been approved on an interim basis.

Ihe team reviewed the

procedures for alternate testing and deterauned that VR-17 was acceptable

as

it ccmplied with the zecp.urenents

outlined in the staff's safety evaluation

approving the program.

'Bm staff granted interim relief for EB-7 based

an the ability of the licensee

to shee that reputable

and mearuLngful pump test data could be obtained by

measuring flew at the CACs rather than at the outlet of the pur~.

%he

procedure required obtaining a flow reading of 1100 gran in each

CAC line while

isolating all other loads except the

DG coolers.

%he team questioned whether

this method established

repeatable test conditions since

SW flow thxough the

DGs was not measured.

At the time of the inspection, the licensee

was

implementing a nedification which would allow it to measure the flow at the

outlet of each pump, providing more accurate test data and eliminating the

need to isolate loads.

The licensee also had available portable test

equipnent which, if used in a repeatable

manner,

would eliminate the question

regarding the current test method.

23

~

1

I

~'I

ip

X)1

(

~

(

~

'Ihe team reviewed the following prcma9ure. relating to SNS testing:

PZ-2.7,

nSexvt.ce Water System)i

~ ET 36Q

nStandby AuxiliaryF~ter Systems

~

PT 2 10 11 g

Sexvioe Water System RsQUndant Discharge Line BcBxcising

g aIxl

M-37.142, 'V-4601, 4602,

4603,

4604 Disassembly."

9'rocedures

cantained

clear ard sufficient preparatian

and realignment steps,

acceptance criteria,

ard specific responsibilities

and qualificatians of test perscenel.

'Ihe team reviewed records of test results and found that the licensee trended

results for each car@anent to det~ degradatian;

recenfirmed reference values

dp

verified that new reference values represented a~able

cperation.

When

test data exceeded limits, the licensee increased the test frequency.

%he

licensee

had established

a policy of returning to regular testing frequency

after entering an alert rarge whenever two consecutive

normal te.~ were run.

13m team was concerned that this practice cauld result in equignent

degradation going unchecked. ~ team observed a case in which vibration

me-cerements for an

SW pump were in the alert range for appvmimately two

months.

When the vibratian zetxuned to the normal range, the licensee

returned to the recpQar testing frequency per plant policy despite the fact

that no maintenance

had been performed or cause determined as directed by ASHE

Code Section XI, Article IWP-3230.

%his issue is scxrewhat mitigated because

the licensee teats these

pumps mere often than required by the oode.

'Ihe team found that the licensee did nat enter ICOs during IST since it did

nat consider the equignent being tested to be aut of service.

Nat entering an

ICO allied the licensee to take redundant equignent out of service

simultaneously with IST.

%he IST of the

SW ~m recpdred that the system be

aligned differently fran its normal accident lineup.

Since marMal actians

would be required to return the system to its co~ aligrunent upon an SIS,

the team could not conclude that the train being tested was available to

perform its safety function for the period of the test.

Ibis candition

existed for several surveillance tests,

inclucUiag that for the contairnnent

spray pump in which the test would clearly disable the system.

Ghe team

questioned this practice and was oancerned

about the lack of procedures to

ensure that redundant equipnent would not be taken out of service while the

companion equipnent is undergoing testing.

She licensee stated that it would

reevaluate its testirg policy. ~ is Unresolved Item No. 91-201-15 in

Appendix A.

3.4.4

Piping and Canponent Inspectians

Action IIIof GL 89-13 requested that licensees establish

a routine inspection

and maintenance

program for cyen-cycle

SNS piping and cmponents to ensure

that corrosion, erosion, protective coating failure, silting, and biofouling

cannot degrade the performance of the safety-related

systems supplied by SW.

'The team reviewed the licensee's

SHSROP and implementing procedures

and

concluded that the licensee's

program for routine inspection and maintenance

of piping and ccanponents did nat satisfy the action requested by the GL.

More

specifically, the team identified the following we Qmesses:

(1)

Although the licensee had procedures in place for disassembly

and

inspection of certain oarponents,

measures

were nat taken to ensure that

24

J

~

~

the population of ~xeents being inspected

was acaplete

arxl that the

frequency of inspection was adequate.

(2)

'Ihe licensee

had nat included in its program any periodic inspection

Itb

instrunant lines were nat degraded.

(3)

Program elarents

were not established

and tzaiz~ was nat provided to

I~

ard operating activities, ard that apprcpriate follawup actions would be

taken upon di:~vexy of zebra mussel debris.

In response to the team's obu~tions, the licensee agreed to review its

SNSROP and to make changes

as ~rcpriate to better address the GL actions.

'Ihis is part of Observation No. 91-201~ in Appendix A.

3.4. 5 Biofauling Ccrdxal and Ru~illance During Systen Operatian

Action I of GL 89-13 requested that licensees

implement and mairrtain an

ongoing program of surveillance ard control techniques to significantly reduce

the incidence of flaw blockage prcblens as a result of biofauling.

%he GL

also requested that licensees in~ their intake structures during each

refuelirg outage and remove any accumQations of macroscopic biological

gmwth.

%he team reviewed the licensee's

SNSRQP and implementing procedures

and

concluded that the licensee's

program for in~ing the intake!~~~

during each refueling outage was a~able.

%he team also found that the

licensee

was age ressively inpleaM nting chlorination practices to the extent

allowed by envirorunental restrictians.

Hm~er, ~ropriate actions were not

defined to ensure that infrequently used loops and stagnant areas,

such as the

alternate discharge to Deer Creek, the

SNS ~lies to the auxiliary feaster

and stardby auxiliary feedwater sy.~-, the SNS ~ly to the

DG cooling

water expansion tank, and isolated heMer piping, would not berne clogged.

areas.

Also, the licensee

s SNSROP did not specifically define when system

layup would be required and layup procedures

were not established per the GL

request.

%he licensee

agreed to review its SNSRDP and make changes

as

appropriate to better adams the acticns requested by the GL.

'this is part

of Olmervation No. 91-201~ in A~dix A.

3.4.6

Conclusions

In general,

the team concluded that the lioensee's

performance in the area of

surveillance

and testing was satisfactory.

However, the team found that

certain impravements cauld be made to surveillance procedures to help ensure

the operability of the

SW pumps.

'Ihe team also found that cmMin

impravements in the licensee's

SNAB were necessary to more adequately

a&bess Actions I and IIIof GL 89-13.

Also, on the basis of its review of

the p~xzational test data, the team faurd that system flaw balancing,

and

possibly additional testing, is warranted.

A concern was identified regarding

the practice of nat entering an ZCO and taking redm9ant equignent cut of

service simultaneously.

25

~

~

p

~

r'

3.5 Quality Assurance

and Oarrective Acticas

3.5.1

Root-Cause Evaluations

'Ihe team reviewed a list of entries into the Nuclear Plant Reliability Data

System

(NHKG) related to the

SWS at Ginna.

Ghese entries

reacceded

specific

cxmponent failures back to 1970.

~me were 26 SWS-related failure reports in

the list.

'Xhe zeviev of the root causes

contained in these reports appe~ued

to be adequate,

but thexe was no documentation available to indicate the

proces used for zoot~use detern~ticn or to establish the validity of the

zoot cause itself.

In many cases,

the root cause indicated was a simple statement that the

failure was due to normal service wear.

In the case of two sucomsive

pump

mator failures for SW pump D, the first cause

was listed as "unknown."

%he

SW

pump D actor repair was performed by Reliance Electric.

2gpzaximately three

months after repair, the motor failed a seoced time.

%he vendor determined

that the second mator failed because of service loading (avexvoltage)

an the

gator beyord its design ratings.

She licensee,

however, disagreed with this

conclusion and, in the fall of 1990, provided data to Reliance Electric to

support its contention that the puap motor failure had not been caused by

ovexvoltage.

However, na documentatian

was provided to shaw that the zoot

cause of the failure had been detaradned

by the end of the inspection.

In

addition to the lack of a donanented root cause detexrnination, the failure

analysis

and zepair of this mator was performed without ooxrective action

reports

(CARs) because

the licensee's

program required initiation of a CAR

only in the case of a "significant" cordition adverse to quality; the licensee

did not consider this failure significant.

%he licensee

was develcping a procedure, with a March 1992 implertentation

schedule, to formalize the root~use analysis process.

9he draft procedure

included a structured formal process for initiatirg arxl docunenting zoat-

cause analyses for "significant ~tians adverse to quality," but the

process would not be mandatoxy for any conchtion that would nat be deemed

significant.

'Ihis was a weakness in the prcpcmed pzogzam because

a condition

that may nat initiallyappar to be significant may be found to be significant

when the root-cause is detexmined.

In the area of co~ive actions, the team found that the licensee's

corrective actions for identified deficiencies

and problems had been generally

effective.

'lhe corrective action implemented for failure of circuit breaker

DB-25 was an mamtple of effective ard extensive corrective action.

3.5.2

Comparison of Findings

In ccmparing its firxhngs to the licensee's

awn assessment

of system status,

the team found areas in which the licensee

had nat identified or resolved

items the team had noted.

For example, the potential of a single failure of

an

SW pump discharge

chec3c valve should have been zecxxpu.zed and addressed

as

part of the licensee's

evaluation in zespanse to GL 89-13.

Qhe licensee

should also have been alerted to this condition by the Altxan single-failuxe

report.

She lack of thorough reviev of the Altran report was an illustration

of the team's finding regarding inadequate

document control.

A 1989 NRC

mction recagnized the lack of engineering assurance

as a weakness.

26

~

~

~

~

I

I'

Holmver,

a Decanber

1990 selfmudit of the document ocetml pmcess

("Audit of

Ginna Configuration and Document Control") did not.

Another finding of the

team involving discrepancies

bedpan the UPBEAR description ard actual system

design was not identified by the licensee until October 1991 in a PCAQ report

(No. 91047) which recognized that the UFSAR was incorrect in stating that the

train cd-cannect valves were closed.

'Ihe team's cbservation that the root-cause analysis

and maintenance trending

programs were weak was also identified by the licx~~a in its 1990 Quality

management

evaluate the co~ive actian system to eliminate ccanplexities,

unify all sources of trerding data,

and use the system to eliminate prcblems

rather than as a tracking system.

She recaanerdatians

had not been acted upon

to date.

Nalkdawns corducteD by the team revealed scaa missing or inaccurate

tags ard bent valve stems.

'Ihe fact that these were not faund by the licensee

irdicated that the routine plant tours and inspections by plant personnel

were

not fully effective.

3.5. 3

Conclusions

On the basis of the inspectian findings, the team concluded that the licensee

did not have an approved and structured root~use evaluatian program.

Although the corrective actions implemer~ for identified problem appeared

adequate,

there was a lack of objective docunentatian to establish the

evaluation process

arxl the depth of evaluatian.

%he licensee,

however,

was in

the process of developing a more statured root~use evaluatian prcxpam,

which may improve the process.

lhe team identified certain issues that the

licensee either had not identified, or if identified, had not acted upon.

4. 0

EXIT M&1'ING

On Decetnber 20,

1991, the team conducted

an exit meeting at the Ginna Nuclear

~ Station.

NRC and licensee personnel atterding this meeting are listed

in Apperdix B.

'Ihe licensee did not identify as proprietary any materials

given to the insIxation team.

During the exit meeting, the team sunanarized

the scope ard findings of the inspectian.

27

~'

APE%ÃDIX A

SUt%ARY OF INSPKCTICN FINDINGS

CZE8CY

.

- 0 Wl

(Section 3.1.1)

DescTi

ion of Conchtion:

During its review, the team noted that the licensee

was not properly

controlling, verifying, and a~ing design reports, calculations, or

analyses.

For example,

'Ihe NUS Corporation calculation supporting Engineering Work Ra~st

(ERR)

1594, "Hydraulic Analysis of the Sexvice Water System," dated February

1988 g was suhaLtted to the licensee marked "preliminaxy, for review and

caament" in March 1988.

As of December

1991, the analysis had not been

reviewed and accepted by the licensee,

but was being used for hydraulic

analysis

and balancing of the SNS follawiing installatian of SF'.

(2)

Bechtel-KHU Report, 'Beat Zaad Capacity/Design Margin Analysis for RHRHX,

CCRHX, SF%Kg Non-Regenerative

HX," Job No. 20031,

RBV

1g dated

January

22,

1989, did not have any indication that it had been reviewed

ard accepted

as required by the licensee's Quality Assurance Manual, Rev.

13, dated Navenher 1, 1986.

(3)

(4)

Rochester

Gas ard Electric Corporatian

(RG&E) design analysis document

entitled "Insitu Motor Zaad Determinations" for ERR 4232,

"SWS Pump Motor

Studies," dated July 15, 1986, included incorrect assumptians

based

on a

low slip motor and ~rcpriate equations.

Consequently,

the results

were also inca~.

'Ihe licensee had known this for moxa than a year:

however, action to revise the calculation and conclusions

was not

initiated.

'Ihe incorrect assumptions

had not been identified by the

design verification process.

Two copies of the RG&E design analysis for SR 3689,

"Containment Fan

Cooler Air Flaw," Rev. 0, dated June 4, 1984, were given to the team.

One contained handwritten notes and correctians,

the other did not.

licensee stated that the correct design analysis

was the ane with

handwritten additions even though they had not been incorporated into a

new revision.

(5)

RG&E design analysis for ERR 4658~-009,

"Minimm Diesel Generator

Cooler and tube Oil Cooler Water Flow Requirmants,"

Rev.

0, dated

July 25, 1991,

was prep-wed to justify DG ~~iilitywith high

differential pressure

(dp) across the coolers due to zebra musel tube

plugging.

%he calculatians

were prepare based an data fran the Gilbert

and Associates Bill of Material No. 216048 which did not use the

applicable design conditions, e.g., cooling water inlet tenyexature

arxl

A-1

g A

C

t

Y4

h,

heat load.

%he incorrect design values resulted in incorrect assumptions

and results.

the team identified items

(3) and

(5)

(abave)

as instances in erich the lack

of appropriate

document control and verification resulted in technically

inadequate

analyses.

BRQtRS

Ginna QMlityAssurance Hanual, Sectian 3, <<Configuration Control," Rev. 13/

dated Novemkmr 1, 1986, states in Section 3.4.3,

"Design Verification," <<The

design verification shall assure that the design outputs (i.e., drawirgs,

analysis arxl specifications)

including design outputs fran equi~M nt y~ y

apped. <<Section 3.4.4, <<Interface Gantrol," states that interface

procedures

between

RG&E engineering

and carrtractor eng~ing organizations

shall include <<instructians regarding the canted:s of docutnent transau.ttals

9

docenents shall identify the status of the docenents

and identify, where

necessary,

inccaplete items which require further evaluation,

review, or

~roval."

10 CFR 50, ApperxRuc B, Criterian III, "Design Control,<<requiresg

in partI

that '94easures

shall be established for the identification m9 control of

design interfaces

and for coordination

among participating design

organizations.

~me measures shall include the establistmant of p~ures

among participating design organizatians for the review, appraval,

release,

distribution, and revision of docunents involving design interfaces.

%he

design control measures shall provide for verifying or cheaking the adequacy

of design,

such as by the performance of design reviews."

10 CFR 50, Appendix B, Criterion VI, <<Docuaant Contmlg

requires

in part,

that 'measures

shall be established to control the issuance of documents,

such

as instructions, p~ures

and drawings, including changes thereto,

which

prescribe all activities affecting quality.

these measures shall ensure that

doaments,

including cbees,

are reviewed for adequacy

and appraved for

release by authorized personnel..."

~ferences:

None

A-2

s

ind

3.1.2.4)

i

ion of Condition:

of SNS Hphaulic Model (Sections 3.1.2.1 and

%he team reviewed the SNS hydraulic mxhl developed by NUS Oorporatian in

~rt of EWR 1594 and ~uded that it was relatively accurate for most of

the process

equipment

and piping.

%be major limitatian associated with the

model was the limited amount of actual plant data used to verify the flmr

resistance.

For example, the differential pressure

was av~imated

and the

fleer rate was underestimated for the

DG coolers as a result of using an

ezzoneously high zesistance miLtiplication factor for all egui~ent and

piping.

In response to these fixxUirgs, the licensee informed the team that it

was planning to upcpade the hydraulic zesistances

associated with the major

fleer paths in the safety-related portion of the SWS.

As a muumum, the

licensee caamitted to detezxaine the actual floe and pressure'drop for the flmr '"

paths through the CACs,

CCWHXs, ar6 DG coolers,

and to ~ly an apprcpriate

resistance multiplication factor in the model.

Ihe pressure

drop and flew

data would then be used to adjust actual system flows by setting throttling

system valves as appropriate.

'Ibe licensee caanitted to su?xnit a letter to

NRC outlining its plans and schedule for impzavirg its flmr mocha. and applying

those results to actual system flow balancing.

BRAGI

~

~

~

~

~

~

~

~

Ginna Technical Specification 3.3.4,

"Sezvice Water System," mcpxires, in

part, that at least two sezvice water puaps and one locp header be operable.

Reference:

1.

'lhe NUS Corporation calculation,

"Hydraulic Analysis of the Sezvice Water

System," dated Februazy

1988 ~zting ERR 1594.

A-3

I

4

I

0

~

~

~

~

~

~

3

Find'itle:

Oarponent Cooling Water Heat Exchanger Vibration Potential

(Section 3.1.2.2)

i

ian of

%he team identified that at normal flmr rates through its shell side, the

cxxnponent cooling water heat exchangers

(OCMHXs) were susceptible to flow-

induced vibration which could result in tube deterioratian

due to fretting.

Bechtel-KHU Report No. 20031 arrived at a similar canclusian. ~ team

requested the sexual

OCHHX shellmide flaw rate to assess

the flaw-induced

vibration potential at flaw rates

above normal.

this informatian was not

provided to the team,

and the licensee

had nat deteradumd the maxirum flaw

rates nor assessed

whether a greater potential for flow-irx'b.red vibration

ex1sts

~

ference:

1.

Bechtel-MJ Report No. 20031, %eat Load Capacity/Design Margin Analysis

for RHRHX,

OCMHX, SF&K," dated Januazy

22, 1989.

A-4

ii i

~-

W

Safety-Related

(Section 3.1.2. 3)

'on of~ti

'Ihe licenc~ installed a second spent fuel pool heat exctmnger

(SFBK B) and

in dairy so relegated

SFRK A to sexving a backup function.

ikrwever, this

heat mxhanger was not isolated and cantinued to perform as an ASME Cede

Section IIIpressure

boundary.

9herefore, it appears that SFBK A should

retain its safety-related classificatian

and be subject to applicable code and

W

P

re~mnents

are being met for SFHiX A.

irement:

10 CFR 50.55a(a) (1) ~dry; that "structures, sy.~ns,

and nx~onents shall

quality standards

caixnensurate with the importance of the safety function to

be performed."

References:

None

A-5

BoUI:

1*

I d

y

1 I

hN the

(Sectians 3.1.2.4 and 3.1.2.5)

i

ion of Co

't'n

the course of its in~ion, the team asked

a rnmnber of questians in

response to which the licensee generated

a preliminary calculatian.

Each of

these calculatians

neech to be finalized within the licensee's

review and

approval pm'~ for ergineering documents.

She calculatians include:

(1)

%he licensee provided a preliaLinmy calculation of the

DG jacket water

coolers'perability using correct design inputs and assuming that zebra

mussels would block an appropriate

number of jacket water cooler tubes

(66~ to 76% of first tube pass).

(2)

%he licensee prepared

a preliminary calculation to demonstrate that the

containment air coolers had adequate

drainage capability.

(3)

'Ihe licensee develcped

a preliminary calculation of DG cooler tube

thinning criteria to ahav at what point of tube thirsting that tubes

shauld be plugged.

%he team reviewed the prelinunary calculaticns

and found that they addressed

the issues raised by the team.

References:

None

A-6

~y

~

~

P

i

ion of Cordition:

Analysis (Section 3.1.2.5)

'%he licensee initiated potential

No.91-048 on November 18,

1991 identifying that the Updated Final Safety

i

p

gggg

response,

the licensee performed a majber of analyses to determine that the

contairunent pressure integrity would not be htpaixed by this diverted flow.

t

gg

g

t

g

g

still be met, but, at a heat removal rate lower than 47 MHIU/hr in sare cases.

(Ibe UFSAR curtly indicates a value of 47 MBIU/hr.) ~ reanalysis,

tt

'

>>'gg

longer and higher temperature transient than originally determined.

Because

the containment temperature profile changed in cxxparison with the original

profile, the team advised the licensee to include consideration of this change

as part of an updated containment analysis to NRC.

'Ihe licensee will do so as

part of a licensirg submittal to ~rt a propc~ boron concentration

modification due in approxunately

6 months.

References:

1.

Potential

Condition Mverse to Quality Report No. 91048 dated

tt~

g

I

~

~

November 18,

1991.

2.

Updated Final Safety Analysis Report Table 9.2-2, "Service Water System

Design Flows. "

A-7

T

P

P,

r

L

u

i

se

t!

Ne M

Valve (Sectian 3.1.3)

ion of ConH i

'Ihe SHS has been operated in a cross-connected

configuration since 1988.

'Ihis

alignment cauld render the entire SNS inoperable during a safety injection

actuation if the discharge

check valve for ane of the SH puaps failed to close

upon system realignment such that. if the associated

pump was nat operating,

water could be pumped in the cpposite du~tion through the check valve irrto

the

SW intake bay. Its scenario was identified, but nat appropriately

analyzed, in a single-failure report developed by Altran.

'Ihe licensee will

mare campletely evaluate this item to ensure that such a single failure would

not camprcanise

systen capabilities.

Bm licensee will also review the Altran

report thoroughly ta ensure that other single failures identified had been

~rcpriately analyzed.

Zn the inte~, the licensee institut~ measures to

alert the cyerations staff to this potential event,

develcped

a means of

prompt detection,

and identified mitigating acticas.

ts 0

10 CFR 50.46(d) states,

in part, that the criteria in 50.46(b) recpa3ilxf

cooling perfonnance calculated in accordance with an acceptable

evaluation

medel are in implementation of the general reguh~nts of ECCS performance

coolie design inclucUirg Criterion 35 of 10 CFR 50, Appendix A.

10 CFR 50, Apperdix A, Critexion 35 requires, in part, that the ECCS safety

function "can be accaqplished,

assuming

a sir@le failure."

Reference:

1.

Altran Report No. 90121.6,

"Single Active Failure Analysis of the R.E.

Ginna Station Sexvice Water System," Rev. 1, dated November 18, 1991.

A-8

IW

I

WW

'tV~

ion o

Existing 'IS 3.3.4 anly requires

ane locp header and two service water pumps to

Iv

not ensure the capability of the service water system to perform its safety

function in accordance with the SNS licensing basis as defined in the UFSAR.

According to the UFSAR, two service water pumps are rapxired for post-accident

heat rival during the recirculatian phase of the accident.

Hmmver, the TS

d

sufficient to ensure that two pmtps would be available during recirculation.

'Ihe staff's

SER dated November 3, 1981, regarding

SEP Topic IX-3, specifically

requested that the licensee address the single failure susceptibility when

anly having two pumps curable.

Dais issue had not been addressed

by the

licensee.

10 CFR 50.36 requires limiting carxhtions for cperation to be established that

are representative

of "...the l~ functional aq~ility or perforaence

levels of equignent rmyured for safe cgeration of the facility."

Refmwnam:

1.

2 ~

3.

Ginna UI'SAR Section 9.2.1,

"Service Water System."

Ginna Technical Specification

3 3 4~ nService Water System

NRC Safety Evaluation Report an SEP Tcpic IX-3, "Station Service and

Cooling Water Sy <<~as," dated November

3g 1981.

(

I

1

f

t

0

3.4.4,

and 3.4.5}

'on of Ceexht on:

Action I of GL 89-13 requested that licensees

implement and maintain an

ongoing program of surveillance

and control techniques to significantly reduce

the incidence of flow blockage p~lems as a result of biofouling. ~ GL

also requested that licensees

inspect their intake stnmtures during each

refueling outage

and reaave

any acnumlations of macroscopic biological

~h.

~rcpriate actions were not defined in the licensee's

Service Water

u'ev rye

'

~

~y

and stagnant ~, such as the alternate discharge to Deer Creek, the SNS

supplies to the auxiliary feedwater and standby auxiliary feedwater systems,

the SNS supply to the

DG cooling water expansion tank, and isolated header

piping, would not becane clogged.

For example,

inspection and flushirg

requirements

were not defined for these areas.

Also, the licensee's

SWSRDP

did not specifically define when system layup would be required and layup

pvmadures

were not established per the GL request.

Ghe licensee agreed to

review its SNSROP mxl make c9vmpes as ~ropriate to better address the

actions requested

by the GL.

Action II of GL 89-13 requested that licensees

conduct a test program to

verify the heat transfer capability of heat exchangers

cooled by open-cycle

service water systems.

%he licensee's

Service Water System Reliability

Optimization Program stated that the renaining safety-related heat exchangers

(besides the CCNHXs,

SFESX B, the CACs, the cooling units for the standby

auxiliary f~ter pump roan,

and the

DG coolers)

would be opened ard

w>>

p

lieu of heat transfer performamm testing.

Although the team agreed with this

approach,

the licensee

had not established

a sound basis for the preventive

maintenance

frequencies for these heat exchangers.

Action IIIof GL 89-13 request that licensees ~lish a routine inspection

arxl maintenance

program for open-cycle

SNS piping and catpanents to ensure

that corrosion, erosion, protective coating failure, silting, and biofculing

cannot degrade the perfoxamree of the safety-related

systems supplied by SW.

%he team concluded that the licensee's

program for routine inspection and

maintenance of piping and canponents did not satisfy the action requested

by

the GL.

'lhe licensee

agreed to review its SNSRDP and to make changes

as

appropriate to better address the GL actions.

Ihe team identified the

following weakness<+:

(1)

Although the licensee had procedures in place for disassembly

and

inspection of certain ccaqnnents,

measures

were not taken to ensure that

the ~ation of canponents

being inspected

was camplete and that the

frequency of inspection was ac@quate.

(2)

'Ihe licensee had not included in its pnxpam ~ periodic inspection

tt

W

lllIW

instruaent lines were not degraded.

(3)

Program elements were not established

and training was not provided to

l

tl

and operating activities,

and that ~rrpriate followup acticns would be

taken upon discovery of zebra mussel debris.

Action IV of GL 89-13 requested that licensees

confirm that the SHS will

perform its intended functian in accordance with the licensing basis of the

plant.

'Ihe team's ck~rvatians irdicated that the licensee's

review did not

fullysatisfy Action IV of GL 89-13 because,

for exanple,

UFSAR dis~ancies

were not identified, the single failure of the SW puap discharge

chec3c=valve

WW

l

ll, ~t

licensing basis ard TS for SW pump operability was not identified, adequate

operability calculations for the

DG )acket water cooler had not been

performed,

systzun hydraulic fidel inadequacies

wcisIted, and inacauacies in

the staff 's SER related to SEP Topic ZX-3 were not identified.

ferences:

1.

Generic Letter 89-13, "Service Water System Problems Affecting Safety-

Related Equitant," dated July 18, 1989.

2.

Rochester

Gas ard Electric, Ginna Station,

"Sexvice Water System

Reliability Optimization Program," dated November 19, 1991.

-0-0

i

ion of

2m NRC staff reviewed the canfiguratian and the design of the SNS at Ginna

during SEP under Tcpic IX-3, "Station Service and Cooling Water Systems,"

as

donunented in its SER dated November 3, 1981.

She team cancluded that staff

acceptance of the design and canfiguratian of the Ginna

SWS was based, in

part,

on the ability of the SNS to perform its function ard not be adversely

affected by a single active or passive failure.

She following system design

features were recognized:

(1) nan-critical loads are autcaatically isolated

follcaring an SIS (did not include isolation of the OCRHXs) and

(2) normal

system alignment did not permit a single active or passive failure to result

in the loss of service water fleer to redundant critical loads.

She first

condition was not satisfied by the

SNS design and canfiguratian because

non-

critical loads are only isolated follmrimp an SIS coincident with an

undervoltage condition and a critical load, the CCRHXs, are isolated during

the injection ph.me.

Also, the second condition has not been accurate since

1988 plus frcan initial plant operation until sane undetermined

time when the

crom-tie valves were closed.

Ghe SEP review, and the licensee after

receiving the SEP SER, did nat recxxp6ze these inaccuracies in the dimxxssion

of system design.

Also, the SEP SER specifically requested that the licensee

address the sir@le failures associated with the

SW pur~.

9%is issue had not

been addre~A by the licensee.

'Ihe licensee amnitted to m9mit a letter to

NRC clarifying the SEP SER inaccuracies.

As needed,

the staff will initiate

further review of SEP Topic ZX-3 to ahhxas the inaccuracies.

Reference:

1.

NRC Safety Evaluation Report on SEP Tcpic ZX-3, "Station Service and

Cooling Water c~~ns~ n dated November

3~ 1981

A-12

k

IL

.

~

E

J

l5;

i

ion of Ccexhtion:

The team identified a number of discrepancies with the Girna Updated Final

Safety Analysis Report regardirg the discussion of the SNS.

(1)

UFSAR Set&ion 9.2.1.3 stated "...the service water locy is isolated by

cross-connections."

'Ihe SHS is cross connected by the 14" ~ly header

for the contairttl.nt air coolers.

The system is also cross~meted

at

the three inch equignent cooler supply heelers, at the three inch SI pump

supply headers,

and at the four inch ~xodent coolizg water cross-

connect in the supply header.

(2)

UFSAR Sectian 9.2.1.3 stated "...Allengineered safety features equi~mt

is split, between the two systens

so that only half of the sy ~n would be

affected by a malfunction."

The systen is not designed with half of the

safety equipnent split between the two headers.

In fact, the c~~ is

designed with each loop header typically crom~nected at the various

supply headers for the safety equipment,

and if the system were operated

split the three safety injection pumps would all be an ane header since

they cannot be divided between headm~.

(3)

UFSAR Sections 9.2.1.2.1 arrl 9.2.1.2.2 stated that the service water

system is designed to isolate non-safety-related

leeds an an accident and

a safety injection signal, respectively.

She

SNS does not isolate non-

safety-related

loads on an SIS unless

an undmvoltage ccexhtian also

exists

o

(4)

UF.'AR Table 9.2-2 did nat a~tely reflect the total fleer to the CACs

because it neglected the flaw to the CAC fan motors.

In 1988, the licensee altered the system operating configuration by opening

the second containment air cooler isolation valve.

The UFSAR was not revised

to reflect this change.

tsar

10 CFR Part 50.34(b) states,

in part, '"Ihe final safety analysis report shall

include information that describes the facility, presents the design bases

and

the limits on its operation,

and presents

a safety analysis of the structures,

systems

and camponents...and

shall include...a description and analysis of the

...caqmnents

of the facility, with emphasis

upon perfonnance

requirements,

the bases,

with tet:9mical justification therefor,

upon which such requinanents

have been established,

and the evaluations required to shor that safety

functions will be acxxxrplished.

The chmcriptian shall be sufficient to peraut

understarding of the system designs

and their relationship to safety

evaluations."

1

~

ll

10 CFR Part 50.71(e) states,

in part, ~ person licensed to cperate

a

nuclear pm'eactor...shall

update periodically, as provided in paragraphs

(e) (3) and

(4) of this section, the final safety analysis report

(%SAR)

originally su}mitted as part of the application for cperating license, to

assure that the information included in the FBAR oarMins the latest material

developed...Ghe

updated

FSAR shall be revised to include the effects of all

changes

made in the facility or pracahuas

as described in the

FSAR...revisions shall be filed no less frequently than annually and shall

reflect all changes

up to maxixmm of 6 months prior to the date of filing..."

ferences:

I

i

ion of Oordition:

In the review of cperations p~ures,

the team noted that in step three of

AP-SW.1, >>Service Water Leak,>> operators were du~ted to initiate actions

should

SNS header pressure

be 40 psig or less.

%bmoc was no engineering basis

for this setpoint that was chosen as a best est&nate.

Wit2xmt sufficient

system pa~sure,

the pumps may be unable to supply sufficient fleer to critical

system components

such that the system would be cxesidm~ inoperable.

%he

licensee

was aware of this issue but had taken no action until recpmsted by

the team.

'Ihe licensee will establish

an ~rcpriate 1m'-pressure

setpoint

based

on an assesarant

that system loads would continue to be sqpplied.

i'ent:

Ginna Technical Specification 3.3.4, "Service Water Sy:Mm,>> requires, in

part, that at least two service water pur~ and one locp he.Wr be operable.

Reference:

l.

Rz)cedures

AP-SW.1, >>Service Water Leak."

A-15

~

I

Y

RW

i

ion of Oondition:

Practices

(Section 3.3.5)

Because current TS only axydxed that two of the fear pm>ps be operable,

the

>> ~

>>>>

>>

consi~ this policy urxoonservative in view of its finding regaling the

need for more than two pm~ to be operable,

as discussed in Section 3.1.3 and

Deficiency No. 91-201-08,

and ~mted that the licensee reevaluate its

scheduling practices for long-term punp maintenance during cperation.

References:

1.

Ginna Technical Specification

3 ~3.4, "Service Water System."

2. J~ A. Widay memorandum to operatices staff dated December 5, 1991,

>>>>>>

A-16

w

i

ion of Condition:

Startup testing of the SNS was performed in accordance with procedure SU-4.17.

'Xhe testing that was performed did not fullyverify the capability of the

sy ~.

'Ihe licensee had not reviewed the preoperatianal test results in

cxxnparison to current sy.~n cperation and configuration and determined the

need for additional testing to support operation in ret:px'uxd system

configurations.

Ibis had not been done despite

system cperating changes since

initial licensing involving the numtmr of pumps normally cperating and various

changes to the system valve alignment.

(1)

'Ihe non~fety-related

loads were isolated fnxn the 1oap hem3ers

when the

safety-related

performance of the system was tested

(single pump

operation) .

Current operation of the plant does not isolate the non-

safety-related

loads during an SIS.

(2)

We safety-related

performance of the system was not tested with two

pumps operating.

Two pumps are rmpured to handle the post-accident heat

load dur~ recirculation.

(3)

'lhe system flow balance was established

ban% on three pump operation,

not based

on the limiting case of one pump operating ~lying all

safety-related

and non-safety-related

loads.

(4)

Pump run-out conditions were not evaluated or considered.

ts 0

10 CFR 50, Appendix B, Criterion III( "Design Oanhx)1," requires that design

control measures

"...provide for verifying or checkirg the adequacy of design,

such as by the performance of design reviews, by the use of alternate or

simplified calculational methods, or by the perforaence of a suitable testing

procpmn."

10 CFR 50, Appendix B, Criterion XI, '~ Control," recp6ras, in part, that

test results be documented

and evaluated to ensure satisfactory cxxopletion of

test requirenents.

8

1.

Procedure

SU-4.17,

"House Service Water," dated November 1, 1968.

A-17

g 'I

~l~

t

L

i

ion o ~tion:

~

~

~

~

~

5

M:

p

1%M

Similtaneously Taking Ebxhndant P~prent Out of Sezvioe

(Section 3.4.3)

She team found that the licensee did not enid. Gxjs during IST since it did

not consider the equignent beirut tested to be aut of service.

Not entering an

UX) allured the licensee to take redundant equignent aut of sezvice

simultaneously with IST.

'Ihe IST of the

SW punps required that the system be

aligned differently freya its normal accident lineup.

Since manual actions

would be required to return the sy ~n to its correct alignment upon an SIS,

the team cauld not conclude that the train being tested

was available to

perform its safety function for the period of the test.

Quis condition

existed for several surveillance Umts, including that for the containment

spray pump in which the test disables the system by blodcinp the spray path to

the spray rirg.

The team questianed this practice ard was concerned about the

lack of procedures

to. ensure that rmhxndant ecp.CipMnt would not be taken aut

of service while the oanpanion ecpugrlent is urxkxgoing testing.

Bm licensee

stated that it would reevaluate its policy in this regard.

10 CFR 50, Appendix B, Criterion V, "Instructions, Procedures,

and Drawings,"

requires that "activities affecting quality shall be prescribed by doamented

instructions, p~ures,

or dzawirgs."

Reference:

1.

GL 91-18, "Information to Licensees

Regarding

Two HRC Inspection Manual

Sections

on Resolution of Degraded

and Nancmfozming ~tians

and on

Operability," dated November 7, 1991.

A-18

>i

~

~ g

4 (

~ ~ *g

~

t

1

EXIT MEETING ATHXQANCE LIST

HK%KKR 20,

1991

'ensee

Personnel

Steven Adams

Charles Anderson

Bernard Carrick

Matt Clark

John Cook

'Ihcanas Daniels

James

Dunne

Clair~

Mark Fitzsimmons

Brian Flynn

Charles Forkell

Paul Gorski

'Ihcaas Harding

Glenn Hermes

Gregg Joss

Michael Kennedy

Steven Iawlor

Russ Lingl

Richard Marchionda

'Ihamas Marlow

Robert Mecredy

Jack Metzger

Theodore Miller

John St. Martin

Gene Voci

J~ Widay

Paul Wilkens

George Wlubel

Technical Manager

Qh Manager

Team Manager

Team Leader

Reactor Engineer

PRA Project

Mechanical Engineer

Manager, Electrical/I&C

tural Engineer

Nuclear Safety and Licensing

Manager, Electrical Engineerirg

Modification Support Coordination

Nuclear Safety and Licensing

Results

and Test Supervisor

Mechanical Engineer

gmrations Shift Supervisor

Superintendent,

Sugmrt Services

Superintendent,

Ginna Production

Vice President,

Ginna Nuclear Production

Nuclear Safety and Licensing

Electrical Engineer

Co~ive Action Coordinator

Manager, Mechanical Engineering

Plant Manager

Nuclear Engineering Services

Deponent

Manager

Manager, Nuclear Safety and Licensirg

NRC Personnel

Suxesh Chaudhary

Robert

Grar((m

Eugene Imbro

Allen Johnson

James Linville

Melanic Miller

Khanas Moslak

Jerzy Parkitny

Donna Skay

James

Talcum

Senior Reactor Engineer,

Region I

Section Chief,

NRR

Branch Chief,

NRR

Project Manager,

NRR

Branch Chief, Region I

Team Leader,

NRR

Senior Resident Inspector,

Region I

Consultant,

AECL

General Engineer,

NRR

Senior Reactor Engineer,

NRR

AE%%NDIX C

ABHREVIATIGHS

auxiliary operator

American Society of Mechanical Engineers

containtmnt air cooler

corrective action report

cmponent cooling water heat exchanger

diesel generator

differential pressure

EPIP

ERR

emergency plan inplemeMIng procedure

engaging

work z&p36st

Final Safety Analysis Report

generic letter

Institute of Nuclear ~~ Operatians

inservice te~irg

job perfornance

me mure

limiting cordition for operation

U.S. Nuclear Regulatory Gxmlission

Nuclear Plant Reliability Data System

PCAQ

P&ID

EH

potential conchtion adverse to quality

preventive maintenance

prcbabilistic risk assessment

'tY

Rochester

Gas ard Electric Oorporatian

SEP

SER

SI

SIS

SFERX

SW

SWS

SWSOPI

SWSROP

Systematic Evaluation Pnxpmn

safety evaluation report

safety injectian

safety injection signal

spent fuel pool heat exchanger

service water

service water system

Service Water Systxua ~~~tianal Performance Inspection

Service Water System Reliability Optimization Pmgram

Technical Specifications

Updated Final Safety Analysis Report

C-1

$gsp