Text

Provides Supplemental Info to Outlining Actions to Be Taken as Result of Investigation Into 940407 Unit 1 Reactor Trip & Safety Injections
	ML18100B083
Person / Time
Site:	Salem
Issue date:	04/29/1994
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 9405230268
	Download: ML18100B083 (11)
	v • d • e

(

Public Service Electric and Gas Company Stewen E. MIHenberger Public Service Electric and Gas Company P.O. Box 236, Hancocks Bridge, NJ 08038 609-339-11 oo Vice President and Chief Nuclear Officer Mr. T. Timothy Martin Regional Administrator APR 2 9 1994 NLR-N94084 U.S. Nuclear Requlatory Commission Region I 475 Allendale Road King of Prussia, PA 19406-1415

Dear Mr. Martin:

REQUEST FOR SUPPLEMENTAL INFORMATION SALEM GENERATING STATION UNIT NO. 1 DOCKET NO. 50-272 on April 20,.1994, PSE&G issued a letter outlining actions to be taken as*a result of the investigation into the April 7 Unit 1 reactor trip and safety injections. This letter supplements the information contained in that letter. Subsequent correspondence will address Power Operated Relief Valve.issues, Pressurizer Safety Valve issues, and a request for aqreement to restart. to this letter addresses the hardware issues, corrective actions, and the status of those items.

provides a summary of our root cause analysis.

addresses some of the items that PSE&G is planning to consider as part of our investigation into methods to mitigate the impact of marsh grass on the Circulating Water Intake Structure. includes a summary of the enhancements to operating procedures and licensed operator training that have been made as well as those that have not yet been incorporated.

includes a j"ustif ication to delay the installation of desiqn modifications for Unit 2 unt.tl the next refueling outage.

9405230268 940514

~~R ADOCK 05000272

~

PDR Sincerely,

(

Mr. T. Timothy Martin NLR-N94084 2

c Mr. J. c. Stone, Licensing Project Manager - Salem

u. s. Nuclear Regulatory Commission.

One White Flint North 11555 Rockville Pike Rockville, MD 20852 Mr. c. Marschall (S09)

USNRC Senior Resident Inspector Mr. K. Tosch, Manager, IV NJ Department *of Environmental Protection Division of Environmental Quality Bureau of Nuclear Engineering CN 415 Trenton, NJ 08625 APR 2 9 1994

(

i '

ATTACHMENT 1 HARDWARE RELATED ISSUES SOLID STATE PROTECTION SYSTEM (SSPS) ACTUATION The SSPS train 'A' and train 'B' responded differently due to High Steam Flow Input Relays having slightly different actuation time characteristics to an initiating pulse of short duration.

The pulse troubleshooting detected variances in the input relay actuation times.

Extensive troubleshooting has determined the sctual actuation times for both trains when subjected to short duration initiation signals.

These times varied from 16 to 35 milliseconds.

These vari-ances are expected for pulsed signals and are well within design limits.

Because of the short duration of the signal, only train 'A' responded to the signal.

As a result, only those components associated with train 'A' operated.

No component failures were identified and equipment time response tests were found satisfactory and did not indicate any degradation.

Operators were trained on being c~nsitive to the potential for the trains to actuate at slightly different times.

They were also given guidance to manually initiate the second train in similar future situations.

HIGH STEAM FLOW INPUT RELAYS The High Steam Flow Input Relays were replaced as a conservative measure.

During visual inspection, discoloration was noted in some of the relays.

Although the relays had different actuation time characteristics to an initiating pulse, both channels were within overall time response technical specifications and showed no indication of degradation.

The relays were satisfactorily replaced.

Subsequent pulse testing showed reduced time response differences between the train 'A' and 'B' input relays.

(

ATTACHMENT 1 HARDWARE* REIATED ISSUES PRESSURIZER BISTABLE PERFORMANCE When the Pressurizer Relief Tank (PRT) rupture disk functioned, as de~iqned, pressure went from approximately 91 psi to 0.5 psi in the PRT, affectinq the backpressure as sensed by the Pressure Relief Valves (PRVs), which resulted in a short duration pressurizer pressure pulse.

TWo of the four press~rizer pressure channels did not trip.

All four pressurizer pressure channels contain lead laq derivative amplifiers for pressure rate compensation.

Because of the short duration of the pulse and the tolerances of the

~nstalled equipment, it is not unexpected that only two of the four channels would provide an output.

This is consistent with the design.

Channel functional tests were performed to ensure equipment functioned as designed.

All equipment was found to be within specification. Operability of the amplifiers was verified via field calibration.

ROD SPEED PERFORMANCE The Rod Speed Control Circuit was oriqinally thought t*o have malfunctioned during the power reduction on April 7, 1994.

The operator thought that the Rod Control System did not respond appropriately and switched back to manual.

Sl.IC-CC.RCS-OOOl(Q), Rod Control System Automatic Speed Verification, was performed to verify proper operation of the Rod Control System.

Durinq the load reduction, the operator was monitoring the T error recorder on the panel to make a determination of what actual Rod Speed should be while in Auto.

With only a 5 degree temperature error (between Tave and Tref), Rod Speed should be 72 steps/minute.

However, the overall temperature error that Rod Speed Control will react to is determined by a summation of Nuclear power and Turbine power mismatch, combined with Traf and Tave.

This power mismatch value is unknown for the exact time of the incident, althouqh durinq testinq it was shown that a power mismatch can cancel out the Tave and Tref error.

Based on the testinq performed on the Rod Speed circuitry, the system worked as desiqned.

The T error recorder should not be read as an indicator of required Rod Speed durinq power changes.

This has been communicated to the licensed operators and will be reinforced. in operator traininq.

(

ATTACHMENT 1 HARDWARE.RELATED ISSUES PS-1 AND PS-3 LIMIT SWITCH OPERATION After the incident, an operator reported that be thought that one of the pressurizer spray valves (PS-1 and/or PS-3) may not have indicated closed, even though the c~ntrol air system had been -isolated tQ the containment (the valves fail closed).

The limit switch operation was verified to be satisfactory.

MAIN STEAM SAFETY VALVE OPERATION During the incident, several main steam safety valves operated due to the increase of secondary pressure. This pressure increase was the*result of the increase in RCS temperature.

The safety valves operated as per design.

As a conservative measure, during the upcoming start-up and while in Mode 3, the valves that lifted will be tested to verify that they remain within the proper settings.

MS-10 CONTROLLER PERFORMANCE The Atmospheric Relief Valves (MS-lOs) have a delay in opening due to the valve controller being below its setpoint for an extended period of time.

The design of the controller allows the controller output to saturate low when the process is below the control setpoint (reset wind-up).

This results in a need for manual action, which is procedurally controlled.

DCP lEC-3325 was issued to address the slow operation of the MS-10 controller.

The DCP installed a clamping circuit, changed the gain of the controller, and decreased the reset time.

These changes will improve the controller's time response to a rapidly increasing steam pressure signal and are expected to prevent MSSVs from opening.

CONDENSER VACUUM ALARM SETPOINT During the incident the control room operators noticed that the condenser vacuum low alarm did not come in.

Work orders were initiated to re-calibrate the pressure switches and to verify that the alarm was operable.

One switch required recalibration, the other switch and the alarm were found to be satisfactory.

(

ATTACHMENT 1 HARDWARE RELATED :ISSUES STEAM FLOW TRANSMITTER PERFORMANCE The shutting of the Turbine Stop Valves causes a pressure wave of a sufficient magnitude and duration to initiate a High Steam Flow Signal*.

This High Steam Flow Signal, together with a Low Low Tave ~ignal met the coincidence for a Safety

Injection (SI) signal.

Due to the short duration of the High Steam Flow Pulse (milliseconds), the S:I signal cleared before some plant equipment could latch and operate, allowing completion of all component actions.

Train 1 8 1 did not respond due to the short duration of the spike1 but operated within design specifications.

Ho equipment failures were noted.

A design modification has been installed to filter the pressure wave pulse.

RVLI:S Reactor Vessel Level Indication ~ystem (RVLIS) is required by Tech Specs to be operable in Modes 1, 2, and 3.

Although RVLIS indication was available to -the operators, it was not initially included in the Mode 4, 5, and 6 Control Room log.

Therefore, the Control Room operators.were not conditioned to monitor this indication since it is not considered an operable instrument in modes 4, 5, and 6.

The logs are being revised to require RVLIS indication to be logged in Modes 4, 5, and 6 and to provide procedural guidance to be taken if the level is below the specified limit.

In addition, control Room operators have been instructed to monitor all Control Room instrumentation, regardless of Tech Spec Mode applicability.

When anomalies are discovered by the.control operator, they will be reported to the Shift Supervisor.

The RVLIS indicated less than full scale due to the formation of a nitroqen bubble.

The source of the nitrogen, which accumulated on the reactor vessel head, was from the Volume Control Tank (VCT).

Nitrogen is used in the VCT as a cover gas.

The VCT was being maintained at 34 psi, with the Reactor Coolant System open to the atmosphere.

At this pressure, the nitrOCJen vent into solution and migrated into the reactor vessel, which is expected in this mode prior to fill and vent.... __

operations.

Since the Reactor Coolant System pressure is very close to atmospheric pressure,. the nitrogen came out of solution upon entry into the reactor vessel and accumulated in the reactor bead.

The operators have been provided guidance to maintain nitrogen cover gas in the VCT between 15 and 20 psi in order to minimize the effect of nitrogen going into solution.

(

.ATTACHMENT 2

SUMMARY

OF ROOT CAUSE ANALYS:tS AND SAFETY SIGNIFICANCE ROOT CAUSE A number of root causes can be categ9rized within the three phases of this e~ent. The first phase includes the rapid power reduction to low power operations up to the reactor trip. *The root cauae of this phase was *poor crew performance and inadequate communication betweml the crew members.

The second phase involves the reactor trip to the first safety injection.

The root causes of this phase are operator error in withdrawing the rods and a design problem with the steam flow transmitters.

The third phase includes tha period of the high steam flow on 11 Steam Generator until the recovery from.the entire event.

The root cause of this phase is a design problem (MSlOs) with a contribution from the control operator (did not take manual control in time).

A contributing factor is the lack of operator action to mitigate the primary temperature and secon~ary_pressure increase.

Another contributinq factor to this phase of the event was poor communications.

See the other Attachments for the Corrective Actions related to these root causes.

SAFETY SIGNIFICAHCE This incident was reviewed with respect to Condition II safety analysis limits as well as the impact on the plant component fatigue analysis, fuel integrity, and minimum average temperature. All-Condition II safety limits were met and the plant component fatigue analyses conclusions continue to be valid.

The reduction in the minimum average temperature below the Tech Specs was* not significant enough to have had safety implications.

No event-induced fuel failures resulted.

Therefore, this event was not a significant safety issue and the conclusion* of the UFSAR remain valid.

During the incident, both block valves and both PORVs were available and operable for pressure relief. Thus, the water

-~

filled pressurizer did not cause the event to degrade to a more serious condition.

(

ATTACHMENT 3 MARSH GRASS MITIGATION PLAN The followinq are some examples of actions that are beinq considered to improve the ability of the Circulating Water System to mitigate marsh grass:

1.

Increase in travelling screen speed

2.

Reorientation/addition of screen spray nozzles

3.

Addition of screen spray capacity/pressure

4.

Physic~l barriers in the river

s.

ML18100B083

Text

Dear Mr. Martin:

SUMMARY

Navigation menu

Search