ML18100B086

Provides Addl Info Per 940511 NRC Request,Re Actions That Had to Be Taken or Would Be Taken as Result of Investigation Into 940407 Salem Unit 1 Reactor Trip & Safety Injection

ML18100B086
Person / Time
Site:	Salem
Issue date:	05/13/1994
From:	Miltenberger S Public Service Enterprise Group
To:	Martin T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
Shared Package
ML18100B081	List: ML18100B079 ML18100B082 ML18100B083 ML18100B085 ML18100B086 ML18100B087
References
NUDOCS 9405230278
Download: ML18100B086 (55)

Text

..

/~

Public St YiCl!I Eleerric and Gas Company Stewen E. Miltenberger Public Service Electric and Gas Compar1y P.O. Box 236, Ham:ocks Bridge, NJ 08038 609-339-4199 Vice µ,C!lidmd uml Chief Nuc:IMr oilicer MAY 131994 NLR-994094 Regional Administratc~r u.s. Nuclear Regulatory Commission Region I 475 Allendale Road Kinq of Prussia, PA 19406-1415

Dear Mr. Martin:

REQUEST FOR SUPPLEMENTAL INFORMA't!ON SALEM GENERATXNG STATION UNIT HO. 1 DOCKET NO. 50-272 On April 25, April 29, and May 10, 1994 PSE&G issued letters to the NRC and identified actions that had been taken or would be taken as a result of the investigation into the April 7, 1994 Salem unit 1 reactor trip and safety injections.

on May 11 and 12, 1994 the NRC requested additional information concerning Main Steam Flow Transmitters, Power Operated Relief Valves, Shift Co*poeition, Management Effectiveness, Marsh Grass, Work Practices and Unit 2 Desiqn Modifications.

The additional requested information is provided in the following attachments to this letter:

Attachment l Attachment J Attachment S Main Steam Flow Transmitters Power Operated Relief Valves Shift Composition Management Effectiveness Marsh Grass Work Practices Unit 2 Design Modifications PSE&G will submit a separate letter to request your agreement for restart and lifting of the confirmatory Action Letter.

Should you have any questions regardinq this submittal, please do not hesitate to contact us.

Attachments (4) 9405230278 9405~

-*~-"

PDR ADOCK 05000272 CA PDR

Mr. T. T. Hartin NLR-N94094 c

Hr. J. c. Stone, Licensing Project Manager - Salem U.S. Nuclear Regulatory Commission one White Flint North 11555 Rockville Pike Rockville, MD 20852 United States Nuclear Regu1atory Cot11Dlis5ion Document Control Desk Washington, DC 20555 Mr. c. Marschall (S09)

USNRC Senior Resident In~pector Mr. K. Tosch, Manager, IV NJ Department of Environmental Protection Division of Environmental Quality Bureau of Nuclear Engineering CN 415 Trenton, NJ 08625

SENT BY:

5-13-84 12:47 FSE&G LIC & REfr-.

~01 504 2162;, 4135 ATTACHMENT l STUM FU>W XHSTBUllENTS Alm U)GIC 'ftRI,!.,YS on May 11, 1994, NRC Reqion I requested additional information relative to Salem Unit Nos. 1 and 2 main steam flow instrumentation and reactor protection system logic relays.

The tollowinq information is provided in response to the NRC request.

STEAM FIDW INSTRUMENT DRIFT.AND CALIBRATION The steam flow in each main steam line is determined by continuously measurinq the pressure dirference across the steam line flow restrictor.

The flow restrictor is a venturi type flow meter with an overall pressure drop of approximately 5.0 psid at 100% rated flow.

steam r1ow measurement dritt has occurred since initial plant operations.

This phenomenon was QFesented to the NRC at Region I on May 15, 1989. It is believed that the steam flow calibration was changing (increasing) with time due to entrained gases collactinq in the sensing lines, which is caused by insufficien~ sensinq line slope and the use of insulated condensate pots.

TWo independent consultants have supported this determination.

Technical Specif !cations require a qualitative channel check to be performed every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />.

Additionally, the System Enqineer routinely trends the steam flow channels while at full power.

Recalibrations are performed whenever the indicated steam flows exceed station administrative limits of +/- Ji of rated steam flow as compared to the on line calorimetric reactor power while at 100% rated thermal power.

The administrative limits are contained in Salem Operatinq procedure SC.OP-DD~ZZ-0023 (Z).

Recalibrations have been performed to correct drifts of +/- 3% to st.

See the attached Cnit 1 Steam Flow - Cycle 11 graphs.

Decreases in indicated steam flow have been observed after reactor trips or plant shutdown because of qas entrainment in the sensor instrument tubing on either the high or low side of the sensor/

transmitter (see the attached Unit 1 Steam Flow - cycle 11 graphs) *. With changes in pressure in the steam lines, non-condensible gases will either go into or out of solution.

H~rizontal or negative slopes in the sensing line can resUlt in the collection of these entrained gases and add to sensor differential pressure error.

EFFORTS TO CORRECT IMSTBUMEN'l' DRIFT To eliminate gas entrainment and resulting instrument drift, a modification has been installed on Salem Unit No. 1 that included insta1lation of un-insulated larger condensing pots and larqer diameter, properly routed tubing with greater than 1.0 inch/ft slope.

SENT BY:

5-13-94 12:48 PSE&G LIC & RE~

301 504 2162;# 5135 It is believed that the presently installed, modified steam flow sGnsinq lines will effectively reduce the drift concern.

There has been linlited operatinq time with this newly insta11ed modi!ication.

A similar DCP for Unit No. 2 has been prepared, SORC approved, and is currently scheduled on the active work list of the Fall 1994 outage.

In the event that recalibration& are required during the-current Unit No. l fuel cycle, further analysis of the effectiveness of the modifications will be performed.

Lessons learned trom Unit No. 1 will be applied as practical to the Unit 2 modi~ications scheduled for the t'all 1994 refueling outage.

CYCLING or STEAM PLOW INPUT RELAYS PUE TO PRIFT AND NOISE Gradual upward driftinq of the steam flow siqnals, coinbinad with the low signal/noise ratio of the process resulted in the ralay chattering that has been identified.

The modifications to correct steam flow drift and the installation of a dampinq circuit to de-sensitize the transmitters* time respon~e will significantly reduce reiay chatter.

The main steam flow drift evaluation has determined that the most

.probable cause of the drift is instrument line configuration. as opposed to transmitter or sensor drift. Electrical noise associated with the steam flow signal is not considered to be a contributer to the cycling concern.

Positive steam flow drift of 3% to 5% coupled with sensed process noise, not electrical noise, allows the instrument loop comparator to exceed the setpoint. rt is the process noise that causes the multiple trip and reset comparator functions.

The resat value for the setpoint is 1% steam flow, 40 mvdc, of the four volt loop span~ The administrative limit of +/- 3%

discussed above has minimized this cycling concern.

A design modification to install a damping circuit to desensitize the transmitter in order to dampen process noise, has been implemented at Unit i. This modification will help prevent relay cyclinq as described above.

The installed Unit l modification is beinq evaluated to ansure adequate resolution of the process noise issue, and to support evaluation of a future design change for Unit 2.

STEAM PRESSUBE PULSE DURING AERIL 7. 1994 EVENT The turbine stop valves are reverse check ~alves held open by a hydraulic actuator.

The quick closing ot the turbine stop valves generates ~ compressive pressure wave which travels up and down the pipe at the sonic wave speed.

The errects of these shock waves on the sensing lines and ditterential transmitters were recently analyzed as-'-*

a result of this event and determined to be a contributor to the spurious high steam flow signals.

The efforts to reduce sensed process noise discussed above will help eliminate false steam flow signals resultinq from pressure pu1ses.

2 -

SE:'JT BY:

5-13-94 PSE&G LI C & REG-t

• 301 504 2162;# 6/35 U>GIC TBAINS' USPONSE DURING APRIL 7. 1994 EV'Eli'E There is currently insuff iciant data to support a cause/effect relationship between relay chatter and the difference in the logic trains* response durinq the April 7, 1994 event.

Note that Train A and B relays share inputs from each of the Channel I and II transmittars.

All relays performed well within the Technical Specification requirements based on response time testinq.

The as-found relay drop out ti.mas were tested and found to be reasonable for relays of this type, based on manufacturers* in~ormation on the replacement relays. Testing indicated that Train B responded slightly slower (15 msac) th~n.Train A due to ~purious short duration pulsed inputs.

PSE&G attributes the root cause of the difference in the loqic trains* re5ponse to the short duration of the protection siqnal combined with normal variations in relay pickup and dropout times.

A visual inspection of the high steilll tlow input relays indicated discoloration of the relay cases ana some apparent carbon deposits.

These relays will be sent to PSE&G's laboratory for analysis in an attempt to determine the effect oq actual relay performance.

All testinq showed that both logic trains performed within the Technical Specifications requirements for actuation and timQ response.

ASSVRANCE OF MAINTAINING STEAM FLOW CHANNEL PERFOBMANCE WITHIN TECHNICAL SPECIFIGATION LIMITS Each shift, channel Checks are performed for steam flow indications.

comparisons between reactor calorimetric power and steam flow indication are also performed on a regular basis.

Durinq channel checks, if one steam flow channei differs from the other channel by st at >15% power, Technical Specification 3.3.1.1 is entered.

System engineering is notified when the steam generator steam flow/power level channel check +/- 3% administ~ative limit is exceeded, per operations procedure sc.oP-DD.ZZ-OD23(Z).

Upon such notification, calorimetric and steam flow differential pressures are evaluated and recalibration is performed if necessary.

The recalibration of the transmitter (either directly or by adjustment of the summator) re-establishes the o-12ot steam flow corresponding to o-lOOt transmitter output.

This relationship is the basis for the setpoint calculation, and the scaling of th* steam flow channel.

As long as the transmitter relationship is maintained, the setpoint analysis (SC-CN007-01) is valid and the current Technicai Specifications setpoint ensures that the trips will occur when required to remain within the safety analyses.

3 -

CD 0

10-07-92 10-21*92 11-04-9"2 11-18-92 12-02-92 12-18-92 12*30-92 01-13-93 4

01-27-93 02~10-93

~

~

02-24-93 03-10-93

~ 03-24-93

-4 m 04-07-93 04-21-93.

05-05-93 05-19-93 06-02-93 06-16-93 06-30-93 07-14-93

~:><

07-28-93

u~ {IJ m

o-u::

08-11-93

~:;!!......

08-25-93

!~-;

09-0g...93 m

m

~

~

S~IL #~69T6 tOS TO~

~3M' '.Jfl ~Sci PERCENT I

DEGREEF

~~ 1 "I "

lri 0 a

I 1

.f 1

9 iit

.II 6t:6T 0

~

rt~t_t_

-~

~

~

~

-~

~

~

-I 0

I\\)

c tS-~T-£ c z -.... -

~

Ii:

"II

-~.....

-o

'" =e i*

• ~

..... 0

-S!.-

.... m co*

w~

8 I n

c

,l,.9 oos

• -~..-

8 10-07~92 10-21-92 11-04--92 11-18-92 12-02-92....

01-27 -

02-10 02-24-93 03-10-93

~ 03-24-93

~ 04-07-93 04-21 05-05 05-19 06-02-93 06-16-93 06-30-93 07-14-93 07-28-93 08-11-93 OS.25-93 09-08 X@

~a>

5i:c

~~

~

~

~

)(

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

~r?.e~...... ~

..6IVf\\)... iO>l i!!~ !~

3;

!!J

~ I~ #~Z9TZ tOS toe i-D~ ~ '.J 11 !WdSd PERCENT 8

~

m OEGREEF ii i IRI 0

~ A

i. J i

I

D to:~t B 8 g

0 Q

t*~ fit~ t~

f'B-CT-8 c z

=t rn

~

ii:

~s (ft :e -

s ;,

I

-'.(

-o

.e,..

~m co....

w.. -

6 i -

• n

c N

J.9 oos

PERCENT 8

~ !5 C) 0 0

0 w

UI 0

10-07-92 10.21-92 11-04-92 11-18-~

12--02-92 12-16-92 12-30-92 01-13-93 c

i z

01-27-93

-I

'fl.)

02-10.93 tn 02-24-93

-I

~

03-10-93 i:

§i? 03-24-93 "II

~b nf 04-07-93 i :e 04-21-9!!. -

.Q 05-05-93

.... 0

.e r-05-19-93

.... m

-~

co,.A 06-02-93 w~

8 06-16-93 06-30-93

--~.

"V 07-14-93 N

I 07-28-93

~ ><-@

0

D~~

c m c N

08-11-93

!~~

08-25-93 09-08-93

! m m

O>

.Co.

a:i 0

DEGREEF 11 i u ti t~ t~

fii?

~

~":

I\\)

~..

~

r;i or1~ ~'70T7 rno rnr

~

~ ~Tl ~~~~

• no*7T t~-RT-G

,1.g l\\BS

PERCENT

-L

<D U) to

0.

0 8

-L 0

UI 41,,,1 0

Co) 0

~-,,-*

10-07-92 H

10-21-92 11-04 11-18-92 12-02-92 12-16-92 12-30-92 01-13-93 c

z -

01-27-93

-t 02-10-93 m

02-24-93 03-10-93

~

'11

~ 03-24-93

.e 6 rrl 04-07-93 iii E 04-21-93 ID I

~~

05-05-93

.... 0

~~

05-19-93 co :::

w-06-02-93 r-0 06-16-93 0

'1J 06-~93 c..>>

07-14-93 07-28-93

i~

0 :c

':Jl ~O')

()8..11-93 mX:t:

~-'It:

08-25-93

,...~a 0

00-08-93..

! ~

m CJ>

CD C)

C)

0) 0 I\\:)

0) 00 0

DEGREEF

. !f 1 r f i t~ t~

Io

~

!!l-~ :

aa *

~..

i S~/01#!6916 tOS 10~

~

' Jll DW3Sd

! OS:61 t6-E:l-£

A9 lN3S

co 0

~ !5 r

10-07-92 PERCENT

...&. g 0

~ -

0 UI 0

10-21-92 *-

11-04-92 11-18-92 12-02-92 12-16-92 12..S0-92 01-13-93 01-27-93 02-10-93 02-24-93 03-10-93

~ 03-24-93 IT1 04-07-93 04-21-93 05-05-93 I

~

N s fl) 05-19-93 06-02-93 06-16-93 06-30-93

-t*

'¥id.a~.

07-14-93

.~.@

07-28-93

~~

08-11-93

~s

~o 08-25-93 j ~

09-08-93 n1 m

~

~

m

~

-~

o ro

~

~

~

o DECREEF 11 l 1 o t~ t-t

• =~ 9 ~~

~

~

~

-oi'l 4

~

~

C..>

~

I

-f m

~

~

I

0 c: z

=

~

~

'Tl

• GI r-

';:i: 0 1l! =e

~.

I

(")

- (j

~Iii co.....

w_....

6 0

1J w

I n

c N

SCITT#~?.QT?. tog Tor.

t-0::1}! ~~II ~s~

~ T9:?.T

~ tS-CT-S

AH l\\'3S

12-02-92 12-16-92 12-30-92 01~13-93 01-27-93 02-10-93 02-24-93 03-10-93

~ 03-24-93 rr1 04-07-93 04-21~93 CD Q.

0~5-93.

05-19-93 *-

06-02-93 06-16-93 06-30-93 07-14-93 07-28-93 08-11-93 *-

08-25-93 09-08-93 en 0

9C/7.Tf~?.OT?. fnQ Tnr.

,*.x

OJJtn mxI

~

<D "'CO

~ ~

O>

N

~~'.111~~..t PERCENT 8

8

~

.... i I\\)

~

(J) en DEGREE F i

iS

(,)

CJ')

00 0

0

~~:

[r~ r it~ t~

1~~

-I I o*

~ 1

0

"!I 0 i TQ: 7.T t~-r.T-Q c z -

-f

£?

!IC.,.

~~

~

C,0 I

NO

~'

1S ~

~

g

'll

• n

,1.g l~13S

PERCENT

<O co C&)

c 0

~

0 UI c

.w 0

10-07-92 10-21-92 11-04-92 *-

11-18-92 12-02-92 12-16-92 12"'30-92 01-13-93 c

01-27-93 N

z iU 02-10-93

~

~

i 02-24-93 03-10-93 j\\)

i 0

)> 03-24-93 "II

-f CD r"'

rn 04-07-93 *-

~~

04-21-93 fg I

05-05-93 I

I !J 05a19-93 i

.. 0

~ r-06-02-93

~ rn w

'°...

06-1(1.93 w _ _.

6 06-3().-93 0

07-14-93 1J

~

07-28-93 iR~~..

I

~

n 08-11-93

~_,t:

Ct) c:!?~

c 08-25-93 1D "D N

09-08-93

~ ~

~

m

)

CJ)

~

a>

O>

O>

~

0

0) o:i DEGREEF

"' t 0 t~ t~ t~

ii !

m

~ ~

~

d 0

.. );!

i i

J

-t m

D

D 9~/~l#!G9TG t09 TO~

t-03'8 ' Jll ~Sd ts:6t tl3-E!T-S

,i.g.LNdS

SENT BY:

5-13-94 12:52 PSE&G UC & RECJ-.

• 301 504 2162;#14135 ATTACHMENT 2 EYALOATION OP mg;T 1 AND 2

• POWER. OPERATED RELIEF VALVE CPQRVl ACCIPTABILI'l'Y Each PORV used at Salem is an air operated Copes-Vulcan 2-inch globe valve, with 3-inch inlet and outlet connections.

Ths valve is a sinqle seat, unbalanced plug, cage-guided, qlobe valve.

This design is typically used in both the colDlilercial and nuclear industries in a wide range of applications, includinq stea~, water and flashinq fluia.

The cage-guided, single s;gat unbalanced plug trim relies on actuator force to close the valve.

The cage-guided plug is the focal point o~ the design and provides a number of advantages, such as:

quick change trim capability, concentric alignment assurance for seatinq surfaces, even distribution of fluid thus limiting side-loads, restriction of lateral ~tem and plug deflection due to side-loads.

SAiiEM UNIT 1 PORY INTERNALS REPLACEMENT The internals. (stem,, cage and, pluq) of the PORVs, 1PR1 AND 1PR2 on the Salem Unit 1 pressurizer experienced material deg~adation followinq the.

reactor trip on April 7, 1994, and were subsequently replaced with new internal~. The design change replaced the stem, plug and cage assemblies with new asaemb1ies as shown below:

ITEM EXISTING MATERIAL NEW MATERIAL Stem ASTM A276, Type 316 ASTM A276. Type 316 cond. B. Chrome Plated Cond. B, Chrome Plated Plug ASTM A276, Type 420 ASTM A479, Type 316 Full Stellite except top surface cage ASTM A276, Type 420 AS'I'M A564, Type 630 (17-4 PH)

The above new materials selectsd for this application provide wear resistance and were recommended by the valve supplier (Copes-Vulcan) for this :modification.

A new plug design was implemented which eliminates the boss used in the existing design and provides a more riqid stem/pluq interface.

The stem is now pinned to the plug instead of through the boss.

In the new design, the pluq height haG been increased to account ~or the elimination of the boss, thus providing the same stroka length as before.

This modirication, therefore, will not affect valve opening or closing.time.

SENT BY:

5-13-l:J4 To decrease the likelihood of stem and pluq wear, PSE&G has modified its installation procedures to reduce the possibility for misaliqnment.

Tha ravi~ed procedures provide greater assurance of smoothness of movelllent and improved stem assembly centering.

This is accomplished by hand strokinq the valve periodically durinq the valve assembly process.

The tight clearance and tolerance are important for wear characteristics and functionality or the valve.

Sliding contact of the valve is not unexpected due to the intentionally tight tolerances of the valve design.

The replacement of the internals of the PORVs changes neither their functional nor their performance characteristics, nor will this replacement affect the ability of these valves to perform their safety functions durinq any dasiqn or licen5inq design basis event.

The actions taken, as descibed above, are appropriata to enGure the continued operability of these valves.

SAI.EM UNIT 2 PQRVS Durinq the 1993 Salemm Uni~ 2 Seventh Refueling outage, pluq, stem-and cage trim assemblies composed of 17-4 PH material were inadvertantly installed in the Salem Unit 2 PORVs.

17-4 PH stainless steel trim assemblies were previously used in the Salem PORVs ana the acceptability of their perf ormarice is supported by testinq conducted by EPRI and documented in EPRI NP-2628-SR, "EPRI PWR Safety and Relief Valve Test Program, Safety and Relief Valve Test Report", dated December, 1982.

A summary of the valve trim currently installed as new components is qiven in the following table:

.ITEM EXISTING MATERIAL HEAT TREATMENT Stem ASTM A564, Type 630 Hl100 (17-4 PH)

Plug ASTM A!!i64, Type 630 H900 (17-4 PH)

Cage ASTM AS64, Type 630 HllOO (17-4 PH) 17-4 PH stainless steel is a material which has been used for many years, and which continues to be used, in valve applications (including PORVs) at other nuclear plants as wall as* in non-nuclear industry service.

EPRI conducted a series of tests on safety and relief valves at two test sites, Wyle and Marsha1l.

These included tests of Copes-Vulcan valves similar to the Salem PORVs with 17-4 PH stem and pluq materials.

At the

(*

Quotes on the Wyle and Marshall test site information are taken from Reference jl) 2 -

SEi'JT BY:

wyle test site, "a total of eight (8) tests were performed on this valvG design.

During all tests, the valve fully opened and fully closed on demand.

Followinq completion of testing, the valve was disassembled and inspected by the Copes-Vulcan representative.

The cage to body gasket bad partially *washed out* during the testing.

No damage ~as ob~erved that would affect future valve performance.**

At the Marshall test site, "the valve fully opened on damand and closed on demand for each or eleven (11) evaluation test cycles."* Additional testing performed after*the successful testing produced the followinq results (specific information relative to the additional testinq parameters/conditions is not detailed in the raport).

• A new set ot the same design cage and plug parts were then installed and the valve was cycied to inYesti9ate the cage to body gasket performance and to support other Marshall Stea~ Station test functions.

The valve fully opened on demand and closed on demand for the nGxt 43 cycles. six (6) or these cycles were performed under full pressure/flow conditions.

The remaining cycles were either dry, unpressurized actuations or openings/closinga in conjunction with other valve testing.

During the next five full pressure/flow tests performed, the valve did not rully close on demand.

However, the valve always closed to within 13% of the full closed position.

Disassembly showed galling of the cage and plug guiding surfaces."*

As a result of the unknown parameters/conditions of the additional tests, results of.these tests are inconclusive. It is noted that other plants are currently oparatinq with similar 17-4 PH trim assemblies.

(*

Quotes on the Wyle and Marshall test site information are taken from Reference #1)

A large bOdy of data on wear is reported in an early report on the Naval Nuclear Program (Reterence 2).

In this research, wear is reported in terms of weight loss (milligrams) per pound load for a million cycles of wear travel (mg/lb-million).

A low value is associated with poor wear resistance.

The data in Table 7-3 of the reference involved both piston-cylinder and journal-sleeve tests.

The wear for the 420 aqainat 420 *ia reported as ranging trom 130-185 mq/lb-million, while the wear for 17-4 PH against 17-4 PH is reported a5 460 mg/lb-miilion.

For comparison, the wear for 304 stainless ataQl against 304 stainless steel, a combination known to be suscepticle to qallinq, is 3,200 *g/lb-million.

It is therefore concluded that, although 17-4 PH with 17-4 PH is more susceptible to wear than 420 with 420, the 17-4 PH plug and 17-4 PH caqe combination installed in 2PR1 and 2PR2 is expected to be satisfactory for the current fuel cycle.

The wear of 17-4 PH is si9ni~icantiy dependent on the hardness of the -~

material, which can vary substantially as a function of the heat treatment.

The Salem Unit 2 stem and cage were aged for four hours at 1105 F, then air cooled, with a reported averaqe hardness of 35.00 Rockwell c.

The plug material was pre-heated at 860 F for one half hour, aged at 900 F for 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />, then air cooled, with a reported average hardness of 43.50 Rockwell c.

The purpose of heat treating the plug and the cage to different hardnesses is to im~ro~e the wear characteristics in service.

In order to reduce wear/galling, a Copes-Vulcan practice ia to maintain a Rockwell c hardness dit!erence of 8 points via the heat 3 -

SENT BY:

5-13-54 FSE.&G LIC & REG-t

• 301 504 2162;#17/35

~

treatment process in like material.

In the Naval Nuclear Proqram woar tost~ reported above, tbere was no mention of the 17-4 PH pairs of materials havinq had diffQrent aging heat treatments to improve their wear characteristics.

The 17-4 PH pairs af materia1s now in Salem unit 2, with the two different hardnesses, are expected to perform better than 17-4 PH pairs with the same hardnesses.

In additio~, the EPRI tests made no mention of the pluq and cage material hardnesses.

Without an appropriate differential in pluq and cage hardnesses, wear would be expected to be qreater than in the case where the plug and oage do have an appropriate. differential in hardnesses.

The Salem 2 pluq and caga ~o have a hardness dirterential of s.s Rockwell c, which is the des~rable combination.

The valve supplier, copes-Vulcan, has stated that 17-4 PH can be used for this application, and that it is an ASHE code listed material for pressure retaining parts (Reference 3).

In summary, the 17-4 PH stainless steel stem, pluq and caqe installed in the Salem 2 PORV valves* (with the plug and cage havinq a differential in hardness) is reqarded as beinq a satisfactory materials selection for this application for the period of one fuel cycle.

PORV's with 17-4 PH internals were tested at operatinq conditions at the Wyle and Marshall test faci1ities.

No adverse performance was recorded at Wyle Lab *. The first set of trim :materials followinq the testinq at Marshall Test also indicated no adverse findinqs.

The second set of trim, toliowinq a total of 48 cycles, reported qallinq of the caqe and plug guiding surfaces.

The valve closed to within 13% of the full closed position.

No findinqs related to the valves ability to open were reported.

In the event of a PORV's failure to close, the block valves are capable of isolating the PORV to maintain reactor coolant presQura boundary integrity.

These valves have been verified to be capable of performing their function unde~ design basis conditions in accordance with the Generic Letter 89-10 MOV program.

A 10 CFR 50.59 safety evaluation has been performed which concluded that the 17-4 PH trim assemblies, installed in the Salem Unit 2 PORVs, will perform as desiqned with reasonable assurance and reliability and will remain capable of perfor.minq their specified functions tor the current fuel cycle.

4 -

.SENT BY:*

REFERENCES

l.

EPRI NP-2628-SR, "EPRI PWR Safety and.Rel~et Valve Test Proqram, Safety and Relief Valve Test Report*, December 1982

2.

TID-7006, Corrosion and Wear Hancll;>ook for Water cooled Reactors, D.J. DePaul, Ed., USAEC, March 1957

3.

Letter and attachment of May 2, 1994 from T. gunkle of Copes-Vulcan to c. Lambert of PSE&G

- s -

SENT BY:

5-13-94 12:55 PSE&G LI C & REQ-4 301 504 2162;#19/35 ATTACHMENT 3.

BASIS lQR MDUMt!H SHXPl CREW COMPQsxnoH PSE&G*s operation~l experience has established our confidence in the ability of the praQently required shitt crew to successfully operate the facility.

our experience has indicated that licensed operator simulator training with the minimum shift composition, as required by Technical specifications, is appropriate to propArly mitigata a11 Design basis Accidents and operational transients.

PSE&G identified a number of root causes and cau~al factors fer the Apri1 7th event.

One contributing factor was the utilization of available resources by the Nuclear Shift Supervisor (NSS).

'l'he resources available to the NSS were appropriate to successfully control plant operation, if they had baon more efffficiently utilized.

The prioritization direction and sequence of actions is referred to as resource :management in our corrective actions as shown on our April 29, 1994 letter (Ref:

~LR-N94084).

A number of other corrective actions have been taken.

Some of the corrective actions are in the form of procedural/training enhancements.

Other actions involve discussions of lessons learned with all shift pe~sonnal, and issuance ct an operations Manaqement Information Directive to all licensed and non-licensed operators cown'"unicating the lessons learned and manaqement expectations from the April 7th event.

Some of* the topics discussed with shift personnel durinq traininq sessions include:

(l}

temperature control associated with rapid down-power maneuversa, (2) resource manaqement, (l) prioritization or tasks and re-enforcement of proper colUlDunications, (4) minimum temperature for criticality and associated corrective actions.

Procedure changes associatad with this event.i.ncl.ude:

(1)

(2)

(3)

(4)

S1(2).0P-AB.COND-000l(Q), Loss of Condenser Vacuum s1(2).0P-AB.cw-0001(Q), Circulatinq Water system Malfunction.

S1(2).0P-AB.TRB-000l(Q), TUrhine Trip Below P-9 Sl(2).0P-IO.ZZ-0004(Q), Power Operation

SENT BY:

i2:aa i~t.M.I LIL & Kt.Lr The procedure changes and traininq enhancements coupled with the re-en~orcement of management's expectations will provide less challenges for the operators, better management of resources and proper prioritization of tasks.

In summary, the corrective actions to enhance our procedures and the ability of our personnel to manaqe transients will result in improved control room resource :management.

In conjunction with the licensed operator simulator training and demonstratecl *ability to mitigate all Design Basis Accidents, these items form PSE*G's basis for concludinq that the Technical Specification minimwn shift composition is appropriate.

2 -.

SENT BY:

5-13-94 12:55 r~E&G LIC & KEG-+

JOI 504 ~lb~;;21/35 PSE&G has taken a number of actions to ensure adequate supervisory and management oversight or plant operations.

Many of these actions have been completed and plans are in-place to implement the remaining actions.

The primary focus areas are:

Salem Performance Improvement t

Quality Assurance/Nuclear Safety Review Oversight t

Augmented Independent Oversight These areas are discussed in more detail below.

SAI.EM PERFORMANCE IMPROVEMENT Comprehensive Performance Assessment PSE&G Manaqement has implemented signiricant materiel condition upgrades at Salem, including many design changes that directly improved control room operations.

A signiticant procedural upgrade program was completed in 1993.

While aame improvement in personnel performance has been achieved (e.q., reduced number of personnel error LERs), we recognized the need for continued improvement in this area.

PSE&G recently performed a comprehensive performance assessment of deficiencies observed over the last few years, to ensure we fully understood the* and that appropriate actions were in-place t_o improve Salem performance.

PSE&G has incorporated the results of this assessment into a Nuclear Department implementation plan.

This implementation plan will be our major focus tor 1994 and beyond.

some of the actions identified in this plan are completed and others are underway.

Many of the actions resul.t in culto..J:"al changes which take time and nurturinq.

Salem Enhanced Superyisory oversight In order to make an immediate impact on the organization, the followinq near term actions have been initiated. to improve supervisory management oversiqht of plant operations.

our primary areas of focus are leadership improvements at all leve1s

.~ti~l l:H:

within the Salem organization, to assure the proper sense of ownerAh.ip from all Salem employees, and to improve overall efficiency through teamwork.

rmprovaments in these key areas are required to ensure across the.board improveme~t.

Improvements in ieadership, ownership, and teamwork will primarily result from improved management focus.

The improved focus will result from organization structural changes, putting the "riqht people" into key positions, investing sufficient supervisory time in the field, and continued maturing of the supervisory monitoring program.

Personnel and structural changes ara in-progres5.

The intent is to unitize Planning, Operations and Maintenance below the manager level.

AS part of this re-organization, we are increasinq the Salam staff ~y approximately (80) individuals.

About 2si of that total is additional supervisors.

The additional supervision will ensure sufficient oversight, increase time in the field, and enhance confidence that expectations are being met.

A supervisory model was developed to align and clarify standards and expectations for supervisory personnel.

This model has been communicated to supervisory pe~onnel. 'l'he overall objective of these changes is to improve teamwork, ownership1 and Sale*

personnel. focus.

A new team of department manaqera ia in-place at Salem.

In addition, we have established Station Planning as a separate department with its own manager.

We have added a second Maintenance Manager and separated the mechanical and controls departments under their own *anager.

This results in increased management focus for both the mechanical and controls areas.

Two new individuals have recently been assigned as unit l and Unit 2 operating Engineers.

A third Operating Engineer has been established on ~n interim basis to provide in-field oversiqht and direct monitoring and asseeement of supervisory personnel until such time as our standards are institutionalieed.

We have assigned a Unit 1 and Unit 2 Senior level supervisor to the operations work control Center, to provide a direct link to station planning and ensure timely notification and resolution of equipment deficiencies that affect operations.

The erfort to place the riqht people into key positions i~

oontinuinq *. As part or the planning and maintenance restructuring, maintenance senior supervisors and department engineers, and planninq outage manaqara, departnlent engineers and senior supervisors had-to rea~ply for their positionsa Selection wili be ~ased on panel interviews and professional assessment provided by a consulting firm.

ouring April, the department engineer positions in maihtenance and station planning were filled.

These individuals have assumed their duties.

'l'hese changes will provide the opportunity to improve teamwork, ownership, and focus ~or all station personnel.

The target date for completion of the senior supervisor seiections is the end of June.

2 -

• SENT BY:

5-13-94 12:57 PSE&G LIC & ~

301 504 2162;#23/35 A middle manaqement level review of troubleshooting plans was implemented to improve th* quality and results of troubleshooting.

These reviews have determined that the quality of troubleshootinq activities has improved.

Weekly focus and ownership maetinqs have been initiated by the Salem management team with their people.

This includes individuals rrom department bead positions to the custodians, These meetings focus on what is important to produce everit-free operation and what each individual can do to improve Salam performance.

Stronger leadership is in-place to focus individual and team efforts on event free Salem operation.

Improvements in the hardware, design, and procedures will continue.

As discussed above, many changes have been implemented and others are in-progress to improve performance.

Additionally, there is on-qoing re-analysis of existing projects to ensure that the

• current Salem management team concurs with the project scope and priorities, and that* any operabion issues are properly addressed.

The near term emphasis will be on leadership and making more effective use of our people.

OQALITX ASSUBANCE/NQ'CI&AR SAFETY REVIEW OY.EBSIGHT PSE&G has taken a number of steps to improve the QA/NSR oversight.

and ensure this function is effective in identifying problems to appropriate plant management.

Specific completed actions and plans are listed below.

Replaced the General Manager - Quality Assurance/Nuclear Safety Review.

Replaced the Manager

- Nuclear Safety Review (HSR).

They were provided

. direction from senior management to focus attention on improvinq oversight effectiveness.

Codified our behavioral expectations for Quality Assurance/Nuclear Safety Review personnel with the Quality Assuranc* and Nuclear safety Review Philosophies.

Conducted a third party independent effectiveness evaluation of the NSR organization, ll\\cluding assigned individuals.

Evaluation results helped to focus our efforts to redirect the department, ensure that the department is providinq an effective oversight function, and that properly qualified individuals are in-place.

Replaced the Salem SRG Enqineer (qroup supervisor) in April of 1994.

3 -

SENT BY:

One-year performance appraisals have been approved for all NSR supervisors, and are currently beinq reviewed with them.

These appraiaals are substantiall.y more detailed than has been the norm. They clearly convey management expectations and define required behavioral changes.

Rotated the Salem QA Manager in February, and moved the Hope Creek QA Manager to Salem.

Since takinq his position in February, the Salem QA Kanaqer meets one-on-one with the General Manager - Salem Ope~ations approximately weekly to discuss findings and observations.

Salem QA Manaqer is meeting with the operating shifts during the requalif ication training cycle to explain quality, and define QA's role and how it supports plant operations.

These meeting last approximately 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />.

The sessions are even1,.y split between presentations and questions and ~nswers. Feedback from the first 9roup was extremely positive.

The next group is scheduled for May 19.

Several actions have been taken to upqrade surveillances6 surveillance Reports now discuss the Four Levels ot Defense of Quality Model, and cit*

findinqs in terms of the Four Elements of Quality and indicate which levels were less than effective wh~n findings are identified.

We define the four levels as follows:

l)

Individuals and work qroups

2) supervision and manaqement

3)

Independent assessment

4)

External observation we have initiated semi-annual QA assessments of each Salem station department.

Tha first set are due in July.

General Manager - QA/NSR is personally workinq with the Audits G~oup to enhance audit effectiveness.

In the last year we have made substantially qreater use of technical specialists.

We have moved from a completely hlocked audit schedule to one that supports approximately two person-years of discretionary audits and evaluations per year.

At the beqinninq of this year, we initiated periodic Issues Meetings.

QA/NSR Managers attend thesa meetings to discuss items requiring mutual support and coordination, areas for QA/NSR improvement, and ways to improve Nuclear Department support.

4 -

301 504 2162i#25/35 wa further categorize surveillance tindinqs in t.:arms of which eleinent(s) or quality were.deficient.

We have strengthened surveillances on critical evolutions.

Following the daily Plan-of-the-Day meetings, Salem QA meets to define emerqent surveillances on critical evolutions that day.

We established 24-hour coverage at Salem Unit 1 in early April.

Durinq the upcoming Unit l startup, QA personnel who were previously licensed operators will monitor startup activities in the control room on a 24-hour a day, 7 days a week basis until reaching 100% power, we believe that the actions taken to date and those planned will siqnificantly improve our self-assesSJlent capability at Salem.

Deficient areas will be identified and brought to ~a.nagement*s attention, and escalated as appropriate

• AUGMENTED INDEPENDENT OYEBSIGHT For an interim period, we will supplement the current HSR oversight with 5 additional people.

They will report diractly to the Manager -

NSR and initially provide 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, 7 days a week coverage of for oversight of plant activities and evolutions.

Evolutions typical of those we would expect to overview include:

Reactor startups and shutdowns Low power operations Special tests (e.q., turbine valve testing)

Selected surveillances Selected major system evolutions safety taqging and work control Shift turnovers and plan-of-the-day meetings Key naaintenance evolutions Mate~ie1 condition walkdowns Control room demeanor and conduct Daily feedback will be provided to the Salem mana9e111ent team by the Manager -

NSR.

Weekly feedback will be provided to the vice President and Chief Nuclear Officer, and documented in the monthly report.

Items requiring immediate attention will be escalated as appropriate.

5 -

SENT BY:

The need to continue the augmented NSR oversight function will be evaluated periodical1y.

The decision to terminate the auqmented oversight will be based upon station performance, nature and significance of observations, and implementation of NSR organizational changes to improve NSR oversight.

The VP & CNO must concur with this decision prior to removal of the oversight function.

The augmented oversight will address both Salem Units and will be in-place prior to MODE 2 on Salem Unit 1.

SYMMARY AND CONCLQSIONS our comprehensive performance a~sessment defined specific problem statements.

We have assigned responsibility for resolution af identified weaknesses, prepared actions plans and associated schadula~, and established appropriate performance indicators to measure our progress.

We believe this effort wi11 establish long-term cultural improvements.

In the interim period, we have made chanqes a~d expanded our line management structure, particularly in the supervisory area.

We have strengthened and refocused our QA/NSR oversight.

An augmented independent oversight function is being ilnplemented to provide real time assessment of ongoing station activitiesf and to ensure prompt management attention to any noted deficiencies.

we are confident that the structural and personnel chanqes discussed above will provide the impetus and management attention required for significant and lasting improvements.

6 -

. I

...-'-~

SENT BY:

ATTACHMENT 5 JgR.SB GBASS JaTIGATIOH ~.

PSE&G's response to the increased marsh grass loading is divided into short-term and long-term actions and.additional studies of the grass issues.

Prior to the April 7 plant trip, PSE&G was aware* that the marsh qrass loading was aiqnif icantly denser than normal from prior years and had taken actions to mitigate the impact of the grass influx.

The amount ot qrass seen this year at the circulating water Intake structure is much more than anticipated based on past experience.

This years marsh grass situation appears to be due to the severe winter, which was characterized by eiqnificant ice storms.

~he severe winter was followed by exceptionally high tides that resulted in the deterioration and uprootinq of the marsh grass, which eventually made its way into the Delaware Bay.

As a result, operations and mainteoance personnel were assigned to the Circulating Water Intake Structure durinq.periods of time when a larqe amount of qrass influx was anticipated.

The purpose of this assiqnment was to ensure that personnel were available to clean the screens ~y spraying them and to maintain the screens in an operational condition by performing minor repairs, such as shear pin replacements.

'l'his practice will be re-instated if warranted by grass loading conditions.

'l'his decision will be baaed on the trending and assessments discussed below.

PSR&G has completed so~e equipment upqrades that will improve the Circulating Water System.

These ware done prior to the end of 1993 and include:

Slowdown valves have been installed on the Unit 1 and 2 travelling screen low pressure headers to clear siltation and improve.spray nozzle effectiveness.

Screen wash control panels and instrwaentat1on have been replaced/

refurbished on Unit 1 to improve system performance and reliability.

Similar improvements will be made to Unit 2 during refueling outaqe 2R8.

Subsequent to the trip, procedural enhancements were implemented, whioh includerl the addition of minimum condenser vacut:m and circulators in-service criteria for initiating a manual reactor and/or turbine trip.

~his provides c;JUidance to the operators to assure thnt their response to ~n influx at marsh qrass is appropriate and to belp to preclude a future similar event.

The procedural enhancements were reinforced to the operators through traininq.

_,._,_A

~.

301 504 2162;#28/35

• SENT BY; 5-13-84 12=58 FSE&G Ll C & REfr-t *

Further, longer-term, enhancements are being planned for the Circulating Water System.

They include the following:

Circulatinq water screen modifications (smoother screens, travellinq at a faster speed, resulting in hiqher capacity and more efficient removal of grass and debris)

It is planned to modify the traveling screens to. penRit hiqhar operational speeds.

By increasing screen travel speed, the volume of river water that any qiven screen must filter between cleanings is reduced, thus increasing the level of detritus density in the water which can ba accommodated without plugginq.

The raw river screens currently operate at multiple speeds, the highest of which is 17 feet per minute.

The*new design is expected to permit operation at a top speed of approximately double the current design.

The design details which will permit this higher speed include replacing the current metallic scr~an baskets with baskets constructed ot fiber rein!orced composite :material (thus lowerinq weight and lowerinq inertial loadings on screen motive components), and replacing the screen drive motors/

gearing and controls for the higher speeds.

It is also planned to replace the current wire mesh screen fabric with a new; smooth weave wire mesh sp$cifically desiqned to reduce grass "stapling" and thus permit more efficient detritus removal from the screens by spray water.

The Circulating Water screen modifications are expected to be completed by June 1995 for one of the Salem units and by June 1996 for the other unit.

The schedule for this modification is constrained by the basket manufacturer.

Efforts are underway to improve the installation date.

Upqraded trash rakes tc improve trash rack cleaning effectiveness and levelize intake velocity profiles It is planned to replace the existing trash rakes with an improved "clamshell" design to enhance trash rack cleaning effectiveness and clevelize intake velocity profiles.

This wil1 assist in precluding trash rack occlusion due to larqa debria and also assist in precludinq sudden screen detritus loadings dua to release or accumulated debris.

This modification is expected to be completed ~y the end of 2RS.

2 -

SEi~T BY~

+

upgraded screen wash pumps It is planned to replace two Screen Wash PUmps with pumps of upgraded materials and improved ability to accommodate traveling screen detritus (carryover/carrythrouqh).

'l'his will increase the reliability of these pumps, which are crucial to traveling screen effectiveness. It is further planned to.

model the screen wash system to determine optimal pump operating range, and to monitor the system for e!fectivenaas

-of improvements made.

'l'he remaininq six pumps will be replaced contingent on succeesfu1 operation of the first two.

The two pump modification is expected to be completed by the end of 2R8.

Redesigned spray wash headers It is planned to make several miscellaneous enhnnceJnents to the traveling screens to illlprove detritus handling capabilities, which include spray nozzle additions and re-orientations, internal-piping modifications, and new desiqn flap seals between tha stationary and mcvinq screen components.

Th.ase modifications are expected to be completed by June 1995 for one of the Salem units and by June 1996 ~or the other unit.

The implementation of these enhancements is not aonstrained by the pending New Jersey Pollutant Discharqe Elimination System permit.

A task force has been initiated to analyze the Circulating Water system from a holistic perspective.

Their charter includes a raviaw of all intertaces with the circulating Water system, includinq other systems, people, and the environDent.

They will also review additional engineered solutions, such as physical barriers in the river, to ensure that there are various approaches beinq considered ta improve the ability of the circulatinq water System to mitiqate marsh grass.

Additionally, the density of grass loading at the Circulating Water rntake is showing a 4ecreasinq trend.

The most recent data shows.a substantially lower density than was recorded at the tillie of the plant trip.

The decreasing trend can be attributed to the end of the seasonal cycle, which is marked by a chanqe in transport mechanisms (ice, snow, rain, wind) and tidal patterns as well as by the establishment of new growth in the marshes.

Therefore, the major impact is expected to be over for 1994.

In addition to the planned anhancemente, PSE&G is also conducti119"a number of studies to quantify and charaateriza the marsh grass in the river.

PSE&G is currently performing a quantificatio" study to trend the movement of grass in the water column in front of the intake structure.

This study consists of a full-scale bathymetric survey (sounding) that extended JOO feet in front of the intake structure, as well as daily nearfield surveys.

The daily surveys consist of 3 -

-~-

SENT BY:

5-13-94 13:01 FSE&G LIC & REG-+

301 so4 2162;#30/35 I soWldinqs at transects of 5, 10, and 25 feet at the various tidal cycles to determine the impact of plant operation and tides on qrasa movement.

Based on these daily surveys, PSE&G w111 determine the need to remove the marsh grass via dredging.

'l'he scope of this study wiil be reduced as grass loadings continue to diminish.

In addition, PSE&G is conducting a study to icienti!y the factors that influence the occurrence of high 9rass loadings.

The goal of these studies is to enable us to develop a predictive model to*forecast periods of critical loa~ing.

PSE&G is also reviewinq the original hydrological studies prepared for the intake structure.

Based on this review, further studies will be considered to identify and mitigate the marsh grass occurrences.

The combination of the upgrades that heve been made ta the Circulatinq water System, the availability of operations and maintenance personnel to be assigned to the circulating Water Intake structure, the procedural enhancements and training, and the end of the **asona1*cycle provide a measure of confidence in the Circulating Water Systam*s ability to operete reliably until~e longer term desiqn enhancements are implemented.

4 -

SENT BY:

5-13-94 13:01 PSE.&G LI C & REG-t 301 504 2162;#31135 ATTACHMENT 6.

PSE&G recoqnizes the vita1 rola which properly per~ormed work plays in the safe and reliable operation of its nuclear units.

This document describes the work control processes in**ffact or under development at PSE&G's nuclear facilities.

These processes provide assurance that our work activities presently meet or exceed industry standards and will continue to improve.

PSE&G work control processes are defined in Nuclear Department Administrative Procedures (NAP) NC.NA-AP.ZZ-0009(Q) Nork control Process, NC.NA-AP.ZZ-0069(Q) work contrgl Coor4ination, and NC.NA-AP.ZZ-0015(Q) Safetv Tagging Proara**

These procedures identify the actions to be followed to address aquipmant problems.

The procefilJ includes steps to perform the following:

Problem identification

~

Operability determination P1anninq and scheduJ.inq Work package development Removal of equipment from service Return of equipment to service Operations approval to start maintenance

• work performance Supervisor verificatio~

Equipment retest In order to continually improve the work practices at Salem, PSE&G Management is increasing its focus on the work control process.

The following corrective actions have been or are beinq taken:

~he Vice President - Nuclear Operations directed station managers and supervisors to increase in-plant and face-to-face supervisory contact time.

The increased supervisory

• presence will improve work monitorinq and assessment, availability and ac;cessibility of work direction and timely application of appropriate corrac:tive actions, if needed.

• PSE&G is reinforcing field observation skills vith all supervisors via an established work observation program.

+

The onsite Safety Review Group has been assigned raaponsibility for reviewing outage schedules and performing qualitative Risk Assessment aqainst procedurally-defined criteria.

_SENT BY:

As an interim enhancement, PSE&G has assiqned middle-level management representative5 with specialized technical skilla to review and approve controls troUbleshootinq.

These personnel, possessing proven technical ability in the controls area, review I&C troubleshooting plans before implementation.

These reviews have resuited in improving the quality and results of troubleshooting activities.

+

PSE&G is carefully reviewing the scope of fUture. outages to ensure that the station infrastructure required to support outage scope is suff ioient to preclude schedule-induced preaeuras and to ensure adequate lllill\\agement oversight.

t PSE&G intsnds to reduce the number of Plant Betterment &

Maintenance contractor firms from three to two.

This will provide stronger oversight.

In addition, PSE'G will diract craft supervisory personnel, responsible for complex installation packaqes, to arrive on-site prior to the outaqe.

This will ensure that appropriate pre-job reviews are performed and orqanizational interfaces are defined.

All craft personnel wil1 '2-eoeive additional traininq in PSE&C 1s safety taqqinq program.

+

PSE&G is improving the.focus of the station Planning organization by establishing separate groups for Unit 1 and Onit 2.

Further improvement will be achieved by establishinq separate Work Control Centers for eaC?h unit.

The work control Centers are expected to be in placa by the end of 1996.

PSE&G has established a Work Control Hiqh Impact Team for Outaqe support.

This team*s function is to:

1.

Perform pre-outaqe work process reviews.

These reviews encompass work package assembly and reviews, safety tagginq, and equipment staging.

2.

Review the work control process.

Specifically, they provide input to the process which identif iac and controls the issuance of emarqent work, work package close out, safety taqqinq, and status of scheduling Updates.

J.

Reviaw previous outaqe incidents ~or 1essons learned from ~vents related to the work standards, contractor control tind work control process.

.SENT BY:

The actions stated above provide for a better focused organization to oversee planned activities and emergent work.

In addition, PSE~C Management wil1 assess the er~ectiveness of the corrective actions via tracking and trendin~ of personnel error incident reports related-to work control and field observation results.

A review of the performance indicators at Salem show an illproving trend.

Beqinninq with 1990 and continuing through 1993, the following Salem 1 and 2 combined indicators confirm this performance improvement:

1.

2.

3.

4.

5.

6.

Corrective Work Order Backlog decreased by approximately 1000 work orders (50t reduction).

Preventive Maintenance Overdue Work orders decreased from 610 to 37 (94t reduction).

Total leaks at Salem decreased from 760 to 81 (89t reduction).

Unplanned Reactor trips decreased from 18 to 4 (78%

reduction).

Licensee Event Reports decreased from 84 to 32 (62i reduction).

Personnel Error Licensee Event Reports decreased from 21 to 7 {67i reduction).

PSE'G believes these indicato~s confirm an improving trend in performance.

The less-than-expected results for the last Salem Unit l refueling outage were due to the large scope of work and are considered an anomalous deviation from projected results.

CONCLtlSIQN PSE&G has established a well-defined work control process.

PSE&G has developed programs for performance trending and review and for management oversight to continually upqrade the work contro1 process.

we believe continued improvement will result from communicatinq c1ear expectations to our workers and affective aonitorin9/assess*ent by management personnel in the tield to provide reenforcement.

~t.l't l tH ;

1... n.... u.u '-

• w I.JC.

'\\.L.U

• AT'.rAalllBllT 7

'[JHXT 2 DESIG.H llQDXPXCA1XONS AS a result of lessons learned from the Apri1 7, 1994 event at unit 1, operator retraining and procedural enhancements were implemented at Sa1em Unit 1 and 2.

Design modifications were implemented at unit l, and are planned ror unit 2 durinq an outage of sufficient duration (but not later than the *next refUelin9 outage, currently scheduled for October 15, 1994).

Thesa changes ~re: the modirication of the Main Steam Atmospheric Relief Valve (MSlO) circuitry to minimize challenges to the Main Steam Safety Valves (MSSVs), and the dampenin~ of tbe steam flow signal to reduce spurious Engineered Safety Feature (ESF) actuations caused by compressive pressure waves qenerated by rapid closure of the turbine stop valves.

These actions are considered enhancements to reduce challenges to the plant safety systems.

During the normal re-qualification cycle, operators are trained in th* appropriate responses to-al1 of these events, including the proper.operation ot the MSlO's.

As result-ot: this event, operators were retrained in the proper response to possible delays associated with the MS10 controllers.

Since there is no automatic operation of the xsio*s credited in the safety analysis, there is adequate time for the operators to respond appropr.iately.

From a historical perspective, a safety injection signal is not normally generated fcllowinq a reactor trip.

A safety injection is not expected to be generated since the coincidence of the

• required. RCS temperature (Tave), remains above the 543*F setpoint.

one of the root causes of safety injection experienced on April 1, 1994 was susceptibility of the main steam flow transmitters to pressure puJ.ses generated by turbine valve closure.

Personnel*performance was a1so a major contributor to the safety injection siqnal experienced Ol'.l.April 7, 1994.

Tha..

procedural changes and aanagement direction provided to the operators via tocused simulator traininq will provide appropriate compensatory measures as related to downpower operations and RCS temperature control.

To further enhance shift crew response to potent.ial p1ant transien.ts similar to the Unit 1 event, operator*~ were given additional simulator traininq and written guidance in the followin9 areas: low power operations, control room resource management, and proper actions to be taken for Solid State Protection System (SSPS) train disaqreement.

The MS10 desiqn modification only *ffects automatic operation.

• Automatic operation of the MSlO's is not credited in the safety analyses, which assumes the Main Steam Safety Valves (MSSV)

SENT BY:

r~

LlL & KEG--

~Ul 504 2162;#35/35 operate for overpressure protection.

Manual MSlO operation is discussed as part o~ the recovery actions from a number of events, such as: loss of feedwater, 1oas of offsite power, feeedline break, steamline break, loss-of load, and shutdoWn frOll outside the control room.

Additionally, the inadverten~ opening o! an MSlO is bounded by the analysis of Main steam Depressurization accidents, and the Steam Generator (S/G) pressure is assumed to be controlled by the MSSV's during a Tube Rupture event.

The steam flow siqnal design modification reduces the effects ~f process noise, thereby reducinq the potential for inadvertent ESF syatem actuations.

The operational enhancements described above ensure that the effects of any such actuations are minimized such that transients remain within the limits ot the plant safety analyses.

Althouqh these design Cbanqes could be performed at powe~, PSE~G would not realize a.net satety.sain by assuming the risk associated with work within the Solid State Protection system racks.

Therefore, delaying tho implementation to an outage of sufficient duration (but not later than the next refuelinq outaqe, currently scheduled for October 15, 1994) ia appropriate.

Enclosure Status of Major Issues Affecting Restart Activities at Salem-1 The following issues have been evaluated by NRC staff including (1) assessment of licensee submittals dated April 25, April 29, May 10 and May 13, 1994, (2) independent inspection of licensee activities and (3) discussion with appropriate licensee representatives.

A.

Equipment

1.

Pressurizer Power Operated Relief Valve (PORV) Operability Issue:

PSE&G Response:

As a result of the initial safety injection on April 7, the reactor coolant system (RCS) filled with water.

Without the normal pressurizer steam space to dampen pressure excursions, the continued injection from the first and second automatic safety injection actuations resulted in repeated actuations of the PORVs to limit RCS pressure. As a result of the challenge to the PORVs, the NRC AIT questioned whether any damage to the valves had occurred.

The licensee removed the PORV internals for inspection. The results of the licensee investigation showed that excessive wear was exhibited on the internals of one PORV and slight cracking on the internals of both PORVs. The licensee identified the source of the cracking at the boss used for the stem to plug interface in the valves to be intergranular stress corrosion cracking (IGSCC),

compounded by the stress induced from the different thermal expansion characteristics of the valve internal materials.

The cracking occurred where the stem of the valve, which was made of a 300-series stainless steel, was pinned through the boss to the plug of the valve, which was made of a 400-series stainless steel.

PSE&G replaced the internal parts of the Unit 1 pressurizer power-operated relief valves (PORVs),- lPR-1 and lPR-2, with new internals: a valve stem and plug made of 300-series stainless steel c.:nd a valve cage made of 17-4 pH stainless steel. The new stem and plug have essentially the same thermal expansion characteristics, which will relieve the stresses which contributed to the ob~rved cracking.

Furth.er, a new design of the valve eliminates the boss used in the previous design and provides a more rigid stem to plug interface. Other factors that promote the IGSCC include the preload stresses that are applied when the valve internals are assembled by the manufacturer. In fact, similar

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

cracking, though not as prominent, was observed on other valve internals that the licensee maintained as new spares.

Consequently, the licensee has initiated action to report this apparent equipment defect in accordance with 10 CFR 21.

Th.e licensee also modified the procedures used to assemble and ilistall the PORVs in order to prevent potential valve internal misalignment. PSE&G believed the misalignment, which was due to valve installation technique, contributed to the scuffing and galling observed on the valve internals after the event.

The NRC reviewed and discussed with licensee engineering the results of vendor analysis of the affected PORVs. The inspectors subsequently reviewed the PSE&G design change package and accompanying 10CFRs0.59 safety evaluation for the installation of the new valve internals. The inspectors determined that the new material combination, which has been used in this application before, and the new installation procedure adequately resolve the PORV operability concerns.

2.

Pressurizer Safety Relief Valves Issue:

As a result of the challenge to the PORVs discussed above, the NRC AIT also questioned whether any damage to the safety valves had occurred.

PSE&G Response:

PSE&G took steps to assure the operability of the pressurizer safety relief valves (lPR-3, lPR-4 and lPR-5).

These steps included visual inspection and* non-destructive examination of the valves and lift setpoint and seat leakage testing by a vendor, Wyle Laboratories.

lPR-3 and lPR-5 tested satisfactorily.

lPR-4 exhibited some seat leakage at 90% of the setpoint and lifted at a slightly higher setpoint. Wyle lightly lapped the seat of the lPR-4, adjusted the setpoint, and the valve retested satisfactorily.

NRC Followup:

The NRC discussed the licensee test plan with PSE&G engineering, reviewed the test results aehieved by Wyle Labs, and compared the performance of the lPR-3, lPR-4 and lPR-5 with other comparable industry results. The inspectors determined that PSE&G' s actions had been appropriate to assure that the pressurizer safety relief valves were operable prior to restart of Unit 1.

2

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

3.

Pressurizer PORV and Safety Relief Valve Piping and Supports Issue:

Following the Unit 1 trip, the pressurizer filled to a water solid condition, which resulted in operation of the PORVs and subsequent discharge of fluid from the press~rizer to the.

pn~ssurizer relief tank. The repeated cycling of the PORVs, and the associated repeated discharge of fluid, prompted the NRC to question the structural integrity of the affected PORV piping and supports.

PSE&G Response:

To assess the structural integrity of the PORV piping and supports, the licensee performed an engineering evaluation (S-1-RC-MEE-0898) and several system walkdowns. The engineering evaluation referenced numerous calculations, assessments, and additional engineering evaluations performed both prior to and following the event. The licensee's engineering analysis enveloped the effects on the system caused by the event. Based on system walkdown observations, the licensee concluded that there was no observable damage to piping or their supports due to the repeated discharge of fluid through the PORVs.

NRC Followup:

The NRC reviewed the details of the system walkdown, and the engineering evaluation (S-1-RC-MEE-0898).

Based on these reviews, the NRC concluded that the questions on the structural integrity of the affected PORV piping and supports had been adequately resolved.

4.

Steam Flow Transmitter Response to Turbine Trip Issue:

The initial Solid State Protection System (SSPS) actuation resulted from the coincidence of low RCS temperature (due to operator error) and a spurious high steam flow signal. Spurious high steam flow signals were previously identified by the licensee, but their cause had been attributed to a combination of the SSPS logic (a reactor trip automatically reduces the high steam flow setpoint from 110% to 40% of rated steam flow) and the actual decay in steam flo~ following a reactor-turbine trip.

3

-.l Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

PSE&G Response:

Upon closer analysis following the event, PSE&G identified that the actual cause of the indicated high steam flow signal following a turbine trip corresponded to the pressure wave initiated by the closure of the turbine stop valves, that appeared to the main steam flow transmitter as a short duration high steam flow condition.

'fh:e licensee subsequently installed a resistive-capacitive network to decrease steam flow instrument sensitivity to short-duration steam flow signals, while not preventing the instrument from properly sensing a true high steam flow condition.

NRC Followup:

The NRC reviewed the licensee modification package and concluded that the transmitter time delay circuit is an appropriate means of resolving the spurious steam signal phenomenon without compromising the safety function of the steam flow transmitter.

5.

Steam Flow Instrument Drift Issue:

PSE&G Response:

Steam flow instrument calibration at Salem station has been known to change with time [drift] since initial plant operation. As a result, indicated steam flow, for the same power level, increases with time at power and decreases with time after* a plant trip or shutdown.

Periodic re-calibration had been required to make indicated steam flow equal 100% at 100% power.

This phenomenon had caused, along with process noise, spurious frequent tripping of steam flow bistables and logic input relays.

Although this phenomena did not appear to play a direct role in the event, probably due to recent Unit 1 modifications, the historic frequent tripping of the bistable Iriay have contributed to premature deterioration of the safety injection logic relays and the different responses of the safety injection logic experienced during the event.

The licensee stated that the cause of the instrument drift was entrained gases in sensing lines leading to the instruments. In order to correct this problem they have replaced the instrument sensing lines with larger tubing, larger condensing pots, reoriented the lines to a consistent downward slope and' have removed insulation from sensing lines and* condensing pots to promote condensation and facilitate escape of noncondensible gasses. This modification was installed in Unit 1 during the last outage [Nov

'93-Feb '94] and will be installed at Unit 2 during the next outage

[Oct '94]. Results from operation at Unit 1 since startup have 4

.I Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

been inconclusive, since the unit has not been maintained at full power in any period sufficient to verify the effectiveness of the modifications. However, no re-calibrations have been required since the modification was installed. Additional plant operating time at full power will be needed to determine if the modification h~ been effective in reducing or eliminating the "drift".

The licensee has a surveillance procedure in place to monitor steam flow instrument calibration at both units. The procedure includes acceptance criteria for identifying unacceptable drift and identifies when recalibration should be accomplished.

The addition of the resistive-capacitive network to resolve the reaction to short duration pressure pulses will also reduce the sensitivity to process noise signals as discussed in item 4 above.

Licensee calculations show that calibration adjustments have not violated any technical specification requirements.

The licensee acknowledges the frequent tripping of the bistables, but believes there is insufficient data to support _a cause/ effect relationship between spurious frequent tripping (chatter) of logic relays and the difference in the logic trains' response during the event.

NRC staff has reviewed the licensee response concerning steam flow instrument drift. The licensee provided detailed information on their monitoring program and associated calibration adjustments that have been made to ensure-steam flow set point values remain within technical specification required values.

The NRC staff concluded that the steam flow instrument drift.

should be minimized by the condensing pot and sensing line modifications installed at Unit I and planned for Unit 2. The procedure for monitoring steam flow instrument calibration has been reviewed and found to be acceptable.

5

./

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

There is not a preponderance of evidence to prove that there is a nexus between steam flow instrument drift and associated input relay chatter and apparent differences in steam flow safety injection logic relays. The NRC staff has also concluded that the different responses of the "A" and "B" safety'injection logic relays arC? explainable as normal variations in time response of these relays.

Installation of a resistance-capacitance circuit in the steam flow instrument measuring circuit should minimize the steam flow instrument's sensitivity to short duration steam pressure pulses as well as process noise. This action appears acceptable.

Based on the licensee monitoring program in place to ensure instrument drift ltoes not result in the violation of technical specification limits, the safety function of the instrumentation will be assured.

6.

Solid State Protection System/High Steam Flow Input Relays Issue:

PSE&G Response:

Following the reactor trip and initial automatic safety injection (SI) of April 7, operators recognized that only train A of the solid state protection system (SSPS) had actuated. Several actions controlled by SSPS train A also failed to go to completion resulting in several components not operating as expected. The apparent disagreement between the SI logic trains was not provided for in the EOPs, and operator response to the event was delayed as they manually aligned the two trains and the affected components.

Due to the different responses of train A and train B of the solid state protection system (SSPS) to the event, PSE&G conducted further examination and testing of SSPS components. The licensee concluded that the very short duration of the high steam flow signal explained why only train A of SSPS initiated. Also, the various components within a SSPS train are operated by different latching and seal-in relays, that also have different response times.

This fact, along with the short duration high steam flow signal, explains why not all actions of train A (main steam and feedwater isolation) went to completion. While the licensee testing showed a difference between the time response of the two SSPS trains and found discoloration in some SSPS relays, the licensee determined that both channels operated within the SSPS design and Technical 6

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

Specification requirements. Further testing results confirmed that had an accident condition existed, both SSPS trains would have actuated and all actions would have gone to completion.

The licensee nonetheless replaced the high steam flow input relays, and subsequent testing showed the differences between th~ channel time re~ponses had been reduced. PSE&G provided additional guidance to plant operators on manual actions to be taken in the event the two trains of SSPS respond differently.

The NRC staff monitored the licensee investigation, reviewed the initial test data, and observed portions of the licensee follow-up testing of the SSPS relays. The inspectors determined that the licensee's root cause was acceptable. The staff also determined that the replacement of certain relays was prudent, and that the guidance providea to the operators was appropriate.

7.

Main Steam Atmospheric Relief Valve (MS-10) Controller Issue:

PSE&G Response:

The MS-lOs did not automatically respond to and control high steam generator pressure on April 7, 1994. Following the plant trip and initial safety injection, the reactor coolant system (RCS) temperature increased as a result of core decay heat and reactor coolant pump heat.

This RCS heatup, and the corresponding increase in steam generator pressures were not recognized by the Salem operators. Steam generator pressures increased above the setpoint of the atmospheric relief valves, because of a failure of the MS-10 controllers to promptly respond. Consequently, the steam generator code safety valve lifted.

The steam release through the safety valve caused a cooldown of the reactor coolant system. The cooldown of the RCS resulted in a rapid pressure decrease that initiated the second automatic safety injection due to an actual low pressurizer pressure condition.

During normal pfant operation the MS-10 controllers provide a constant close signal to the valves since normal steam pressure is much lower than the valve opening setpoint. This results in the saturation of the controller circuitry. As a result, the automatic opening of the valves is delayed during actual conditions of high steam generator pressure by an amount of time it takes to clear the saturated condition. The controllers were modified shortly after initial startup of the Salem Unit to prevent inadvertent opening of MS-10. PSE&G has now implemented a design change to install 7

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

8.

NRC Followup:

a discharge path for the capacitor in the control circuit which was susceptible to the saturation phenomenon. This design change re-installed the part of the circuit which the licensee had previously removed.

The controller gain and reset times have also been changed to further l.mprove the controller performance.

The NRC reviewed the design change package which implemented the changes in the MS-10 controller circuit, discussed the modification with licensee engineering, and concluded that the re-installation of the capacitor discharge path would provide better automatic control of steam generator pressure during transient plant conditions. Resident inspectors will observe licensee testing of this modification during plant heat-up.

Rod Control System Operation Issue:

The rod control system was being operated in the manual mode during the event due to ongoing system troubleshooting and operator uncertainty with regard to the system operability in the

. automatic mode. If the system had been operated in the automatic mode the excessive reactor coolant system cooldown may have been minimized or avoided.

PSE&G Response:

At the time of the event, the rod control system deficiencies had been resolved with the exception of monitoring a system isolator to determine if a drifting problem had been corrected.

Final system testing was scheduled the day of the event. Following the event, troubleshooting determined that the automatic mode was fully operable.

NRC Followup:

The AIT reviewed the results of the troubleshooting and testing of the rod control system and determined that PSE&G had adequately corrected the system deficiencies to permit operation of the rod control system in the automatic mode.

9.

Circulating Water Intake Issue:

Marsh grass accumulates in the Delaware River and is drawn into the circulating water system by the circulating water pumps.

When the grass quantities become large, it challenges the traveling screens' ability to remove the grass as fast as it accumulates, clogs the intake flow path and causes loss of cooling to the main 8

Status of Major Issues Affecting Restart Activities at Salem-1 (continued) condenser. Loss of cooling to the condenser requires reduction of plant load, or plant shutdown.

PSE&G Response:

The licensee's response is divided into short and long term actions.

In the short term the licensee has assigned maintenance and o~rations personnel to the circulating water intake structure to maintain and clean the screens. Prior to the last refueling outage the licensee installed low pressure headers to clear siltation and improve screen wash spray nozzle effectiveness.

Screen wash control panels and instrumentation were replaced or refurbished.

Procedural enhancements have been made since the event to give operators more guidance on responses to an influx of marsh grass.

Criteria for initiating a manual reactor and/or turbine trip have been included. ~e density of grass loading is currently showing a decreasing trend. The major impact of marsh grass is expected to be over for 1994.

Long term enhancements include modifications to the traveling screens to permit higher speeds. The higher screen speed will increase the grass removal capability of the screens and lessen the probability of loss of circulating water flow due to grass intrusion.

Higlfor speeds will be achieve4 by replacing.the screen baskets with lighter material and replacing the drive motors/gearing and

• controls for higher speeds. These modifications are expected to be completed by June 1995 for one Salem unit and by June 1996 for the other unit.

In addition, the existing trash rakes, which are positioned in front of the screens, will be replaced to enhance trash rack cleaning and levelize intake velocity profiles. This modification is expected to be completed in October 1994.

The licensee plans to replace two* screen wash pumps [there are 4 pen.1nit] with pumps of upgraded materials and lower maintenance requirements. The licensee then intends to evaluate the screen wash system to determine optimal pump operating range, and to monitor the system effectiveness. This modification is expected to be completed in October 1994. Pending the results of the experience with these two pumps, the remaining 6 pumps may be replaced with the new design.

9

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

PSE&G plans to make other modifications, including spray nozzle additions and re-orientations, internal piping modifications and new designed seals between stationary and moving screen components to improve grass handling capabilities. The implementation schedule for these modifications has not been established.

The licensee is also reviewing the circulating water system, the grass movements and loadings, and will consider various approaches, such as physical barriers in the river to improve the ability to mitigate marsh grass and removal of grass by dredging.

No schedule for completion of these studies has been provided.

Long and short term plans for coping with the grass problem have been reviewed by the staff and discussed with the licensee. Long term plans appearto be aimed at coping with potentially severe grass intrusions. Each of the licensee's.proposals appears to have merit.

The effectiveness of these modifications remain to be demonstrated. The NRC has reviewed the licensee's procedures and training of operators for coping with grass intrusions.

Evaluation of these procedures is discussed below. Plant design and the procedures that the licensee now has in place assure that the loss of circulating water to the main condenser will not challenge the safety of the nuclear plant.

B.

Procedure Improvements

1.

SC.OP-DD.ZZ-OD22(Z), "Control Room Reading Sheet Mode 5 Through 6" Issue:

PSE&G Response:

Following the plant cooldown-subsequent to the event, the NRC identified the Salem Unit 1 reactor vessel level indication system (RVLIS) indicated reactor vessel water level at 93%.

When questioned, the Salem control room operators could not explain the significance of the indication, nor were they required to monitor this indication in the current plant operating mode.

RVLIS values are now logged when a unit is in Mode 5 (Cold Shutdown) or Mode 6 (Refueling), and the procedure requires response' actions when the indicated level is below the minimum value specified in the proCedure.

10

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

The NRC staff reviewed the procedure change, discussed the change with Operations management, interviewed operators to assess their knowledge of the new requirements, and observed operator training in the Salem simulator. The inspectors concluded that action addressed the NRC-identified deficiency in Salem co~trol room operator use and application of RVLIS indication when the plant is in Mode 5 or 6.

2.

Sl(2),0P-AB.COND-0001(Q), "Loss of Condenser Vacuum" Issue:

During the rapid downpower conducted by Salem Unit 1 operators immediately preceding the reactor trip, the operators took extraordinary steps to attempt to keep the unit on line while dealing with the loss of circulating water pumps and main condenser cooling':' The NRC determined that a lack of procedural guidance existed for operators on when to trip the turbine and/or reactor during low power operation.

PSE&G Response:

The procedure now specifies actions to trip the reactor and/or turbine as a specific function of primary coolant temperature, condenser vacuum, condenser back pressure, reactor power, and turbine power conditions.

NRC Followup:

NRC reviewed the procedure change and noted that the specific guidance provided in the procedure now adequately directs operators on what the necessary plant conditions are to remove certain components from service.

The inspectors confirmed operator awareness of the new requirements through operator interviews and through observation of simulator training on the new procedure.

3.

Sl(2).0P-AB.CW-0001(Q), "Circulating Water System Malfunction" S1(2).0P-SO.CW-000l(Z), "Circulating Water Pump Operation!'

Issue:

The rapid downpower maneuver performed by Salem Unit 1 operators was necessitated by the rapid loss of the unit circulating water pumps due to river grass accumulation and the resultant loss of main* condenser cooling.

The NRC determined that the operators lacked procedural guidance on what specific actions were required when dealing with the effects of river grass on circulating water pumps.

11

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

PSE&G Response:

NRC Followup:

These procedures now specify operator actions for the condition when two or more circulating water pumps are out of service and identify actions for operators to take in the case of abnormal condenser vacuum situations.

~e NRC reviewed the procedure change, assessed operator knowledge of the new instructions, and observed their practice in the Salem simulator.

The inspectors determined that the new procedures provide the proper guidance tc; the plant operators for the loss of circulating water pumps.

4.

S1(2).0P-AB.TRB-0001{Q), "Turbine Trip Below P-9" Issue:

During the April 7 downpower maneuver, Salem operators reduced reactor and turbine power at different rates. The resulting power mismatch resulted in the overcooling of the primary coolant system and the subsequent operator action to withdraw control rods, which led to the reactor trip. The operators did not have guidance to manually trip the turbine off-line to restore primary coolant temperature.

PSE&G Response:

The turbine trip procedure now incorporates guidance for operator response to inadvertent or excessive primary coolant cooldown conditions when reactor power is below the P-9 setpoint. The procedure revision now includes specific direction to the operator to go to a new procedure attachment if at any time primary coolant temperature reaches 543 degrees F or less; the Technical Specification minimum temperature for criticality is 541 degrees F. The attachment provides direction to the operator to recover primary temperature, and if temperature can not be maintained above the minimum temperature for criticality, to manually trip the reactor.

12

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

The NRC reviewed the procedure change and noted that the guidance for operator action relative to a manual trip of the turbine was appropriate and properly addressed the concerns of the event.

The inspectors subsequently verified, through interviews, adequate operator knowledge of the new guidance and observed satisfactory

~rformance of the new procedure at the Salem simulator.

5.

Sl(2).0P-IO.ZZ-0004(Q), "Power Operation" Issue:

The power mismatch between the Salem Unit 1 reactor and turbine resulted in the overcooling of the primary coolant system to the point where coolant temperature. went below the minimum temperature for criticality as specified in the unit Technical Specifications. The operators did not have adequate procedure guidance for required action when plant operation did not meet the Technical Specification requirement for minimum temperature for criticality.

PSE&G Response:

The procedure for power operation of the Salem units now includes specific directions for maintaining primary coolant temperature above the Technical Specification minimum temperature for critical operations while performing a plant power reduction. The body of the procedure directs the operator to a new procedure attachment if at any time during the power reduction primary coolant temperature reaches or goes below 543 degrees F; the allowed minimum temperature for critical operations is 541 degrees F. The attachment provides direction to the operator to recover primary temperature, and if temperature can not be. maintained above the minimum temperature for criticality, to manually trip the reactor.

NRC Followup:

The NRC reviewed the new guidance and specific direction provided in the procedure change for maintaining primary coolant temperature above the Technical Specification limit.

The inspectors conducted operator interviews and observed operator simulator training and concluded that the procedure change and operator_ training adequately addressed the issue.

13

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

6.

Emergency Operating Procedures (EOPs)

Issue:

PSE&G Response:

During the operator response to the reactor trip and multiple safety injections, the operators encountered situations where the EOPs did not provide specific guidance or direction.

These situations in~luded:

Resolution of solid state protection system logic train disagreement, Manual operation of the steam generator atmospheric relief valves to control steam generator pressure and primary coolant system (RCS) heatup, and Prevention-of solid RCS conditions and, if they do occur, a plant cooldown under those conditions.

PSE&G is pursuing long term changes affecting the EOPs and Critical Safety Function Status Trees (CSFSTs), working in conjunction with the Westinghouse Owners Group. In the interim, the licensee has provided additional guidance concerning these situations to operators in an Operations Department Information Directive (ID) and in a simulator training lesson plan which addresses the entire event. In response to the above situations, the ID provides guidance to operators on: when a safety injection train disagreement is noted, to manually initiate a safety injection actuation for the train that did not automatically actuate; following a r~ctor trip, to take manual control of the MS-lOs at any time steam generator pressure is at or above the valve setpoint with no apparent valve motion; and, during EOP use after initiation of CSFSTs, and if no higher path conditions exist, the Shift Technical Advisor is to refer to Yellow Path Restoration Procedures to monitor RCS parameters and other indications in order to detect or prevent unexpected plant conditions, such as solid RCS conditions.

Reading, discussing and understanding the ID, and instruction using the simulator lesson plan were required of all licensed and non-lice~sed operators prior to their assuming a watch.

14

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

The NRC discussed the considered EOP changes with Salem Operations Department management, reviewed the guidance provided in the department's ID and the simulator training lesson plan, and observed the training of operators using the lesson plan at the simulator. The inspectors verified operator.knowledge of th~ new guidance through interviews of several operators from different shift crews. The inspectors concluded that the guidance provided in the ID and the training provided at the simulator were an effective means of resolving the evidenced EOP concerns.

C.

Salem Operating Crew *Shift Management Responsibilities Issue:

PSE&G Response:

In addition to the above identified equipment and procedure issues, the NRC identified several areas in which Salem control room operator perfomiance and resource management affected the response to the event. These areas included:

Control room crew communications, Prioritization of personnel assignments and use of additional licensed operating personnel, and Scope of Senior Nuclear Shift Supervisor involvement in Emergency Operating Procedure (BOP) operations.

Event Notification and Communication The licensee responded to the above identified concerns by the issuance of a Salem Operations Department Information Directive (ID) and simulator training lesson plan. Specifically, (1) operators received guidance relative to management's expectations on the quality of communications as to clarity and directness, and the avoidance of vague or imprecise instructions or responses; (2) formal training and guidance were provided relative to the management and control of operating personnel resources to assure that conservative actions are taken to either stabilize plant conditions in a safe and controlled manner or manually trip the reactor or turbine; the ID included guidance on where to assign personnel when the rod control system is in "manual", and the acquisition of additional personnel for significant off-normal events; and (3) the Senior Nuclear Shift Supervisors role relative 15

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

D.

Unit 2 Consideration Issue:

PSE&G Response:

to plant events was clarified to maintain supervisory overview and not become engrossed or involved in assisting the crew with EOP implementation.

All operating crews received the simulator training. on the lesson plcµi derived from the event, and all shifts were required to read and understand the directions provided in the ID prior to resuming a watch.

Before entering mode 2, the licensee will establish interim guidance for all communicators and shift supervisors relative to providing to the NRC fuller detail and explanation on significant events.

Action to modify the emergency plan relative to procedures on event notification and communication will be initiated with all mvolved agencies within the next seven days.

The NRC reviewed and inspected the above procedure changes and training enhancements.

The review included interviews with licensed operators, discussions with Operations Management, and observation of crew training at the Salem simulator.

The inspectors concluded that the changes made to the noted procedures, the additional training supplied to licensed operators, and the guidance provided or planned by management to the operators would effectively address the personnel performance issues identified as a result of the event.

Considering the procedure changes, training and hardware modifications identified from the event for implementation at Unit 1, the NRC questioned what short and long term corrective actions were planned or being implemented at Unit 2.

As a result ~f the event at Salem Unit 1, operator retraining and procedural enhancements were implemented at Unit 1 and 2.

Design modifications were performed at Unit 1 and are planned for Unit 2 no later than the next refueling outage, that begins October 15, 1994.

Operators were given additional training and written guidance on response to marsh grass, downpower and low power operations, RCS temperature control, control room resource management and 16

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

proper actions to be taken for solid state protection system train disagreement. Operators have been trained, prior to this event, on how to cope with MS-10 controller malfunctions and how to operate the system in manual. They were given additional training on use of MS-10 valves to control main steam press_ure following th~ event.

The Unit 2 PORV internals are of a different material, 17-4 pH stainless steel, than those at Unit 1. The 17-4 pH internals are approved for this use by the vendor and are similar to those which were installed in both Unit 1 and Unit 2 at the time of. initial operation. Finally, the licensee has not experienced any problems with this material to date, and believes continued use until the next refueling outage is justified.

The licensee believes that delaying implementation of the hardware fixes to an outage of sufficient duration, but not later than the next refueling outage, currently scheduled for October 15, 1994, is appropriate.

The NRC reviewed the 10CFR50.59 safety evaluation for continued operation with the Unit 2 PORVs in the as-is condition.

The NRC verified that the internals of the Unit 2 PORVs were replaced with components made from 17-4 pH stainless steel. In addition, the NRC confirmed that the material changes for the internals were approved by the PORV vendor. The PORVs will be inspec~ and a design change considered during the next refueling outage. The inspectors concluded that the Unit 2 PORVs are acceptable for continued operation of that unit.

The NRC staff has reviewed the planned modifications (MS-10 control circuit, and steam flow instrumentation configuration and circuit time delay) at Unit 2 and concluded that compensatory me;i..sures provided by improved procedures and operator training are acceptable until the next outage of sufficient duration to install the modifications.

The inspectors have reviewed procedures and training related to coping with rapid power reductions, use of reactor vessel level instrumentation, manual operation of MS-lOs, RCS temperature control, logic train disagreement, control of noncondensible gasses in the vessel and cooldown of a solid RCS. With these procedures 17

Status of Major Issues Affecting Restart Activities at Salem-1 (continued) in place and the associated training completed, operation of Unit 2 until October 15, 1994 is considered acceptable.

E.

Management Effectiveness in Resolving Long-Standing Problems Affecting Performance at Salem Issue:

. Since the November 1991 Turbine-Generator failure event, which resulted in review by an Augmented Inspection Team, PSE&G has continued to experience recurring operational, design, and maintenance-related problems.

Contributing causes to these occurrences have been weaknesses in management and oversight of activities, inadequate root cause analysis, failure to follow procedures, personnel error, ineffective approach to resolution of problems, and insufficient corrective actions. While none of the events have adversely affected public health and safety, the licensee's apparent inability to demonstrate improving performance has been a continuing concern to the NRC.

PSE&G Response:

In their May 13, 1993, letter, PSE&G noted that they have established plans and completed actions relative to: (1) Salem Performance Improvement; (2) Quality Assurance/Nuclear Safety Review Oversight; and (3) Augmented Independent Oversight.

Prior to the event PSE&G management had already implemented significant material condition upgrades at Salem, including design changes that directly improved. control room operations.

Additionally, a Procedural Upgrade Program was completed in 1993.

Although improving performance was indicated by the reduced number of events caused by personnel error, the licensee recognized that satisfactory performance had not yet been achieved.

Consequently, the licensee commissioned a special Comprehensive Assessment of Performance Team (CPAT) in the summer of 1993 to review and assess PSE&G's performance as indicated by the assessment of several deficient conditions and situations over the last few years. The CPAT activities are now completed and the results have been factored into the Nuclear Department Tactical Plan (Plan). The Plan identifies the program for implementing a comprehensive series of measures designed to effect and assure performance improvement.

18

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

NRC Followup:

Actions were also taken prior to the event relative to leadership improvement, including organizational structure

changes, reconstitution of the organization with more capable supervisors, and establishing requirements for increased supervisory oversight activities in the plant. An additional operating engineer has been as~igned to provide direct monitoring of the performance of supervisory personnel until all management enhancements are completed.

The management of Quality Assurance and Nuclear Safety Review Oversight Groups has recently been changed to improve oversight effectiveness. Other supervisory changes have been accomplished to effect better overall performance. An independent consultant has provided an evaluation to assure the selection of properly qualified personnel for this area.

Enhanced procedures and policies for safety

reviews, audits, assessments, and communications of findings were established prior to the event.

Subsequent to the event and until the results of the CP AT effort are established and the planned enhancements in organization, personnel, and policy are completed, an Augmented Independent Oversight group was selected to maintain full oversight coverage on all shifts, 7 days per week. The group has been directed to monitor activities such as reactor startups and shutdowns, low power operations, special tests and surveillances, major system and maintenance evolutions, work control performance and control room conduct, and shift tum-overs and planning meetings. The individuals will provide daily feedback to the Manager of Nuclear Safety Review, and weekly feedback to the Vice President and Chief Nuclear Officer. The Augmented Independent Oversight coverage will be maintained until significant improvement are noted in station performance and in the quality of the Nuclear Safety Review function.

Finally, the licensee has expressed confidence that these structural and personnel changes will provide the impetus and management attention, necessary for significant and lasting improvement.

Previously, the NRC has reviewed and assessed the licensee's CPAT effort.

The CPAT was thorough and developed a comprehensive list of problems and weaknesses that appear to be causal to the recurrent failures noted in the licensee's performance.

19

.._,_d

~

Status of Major Issues Affecting Restart Activities at Salem-1 (continued)

The NRC has also reviewed the Nuclear Department Tactical Plan which identifies the action and performance schedule to resolve each generic problem or weakness identified. The Plan appears thorough in the approach to resolution of* the weaknesses. The schedule, while extending into 1995 for some of the more difficult m~tters appears timely in view of the scoi)e of the effort. NRC has already noted aggressive action to re-evaluate the quality and performance of managers and supervisors in the Salem organization.

Several -replacements have already occurred, including the replacement of the previous General Manager-Salem Operations with the current Vice President-Operations for PSE&G.

NRC has reviewed the credentials of the individuals assigned to the Augmented Independent Oversight group. Their background, experience, and ability seem to be appropriate for the task at hand.

It is the expectation that the group will be successful in its endeavor to monitor the quality of performance and provide the necessary feedback to the right level of management to assure effectiveness and management cogniz.ance of the quality of operations.

While a positive trend has not yet been demonstrated in Salem performance, the near-term and long-term actions initiated by the licensee appear to be sufficient to cause improvement if management maintains their commitment to the program.

20

--"-~