IR 05000423/1986002

Insp Rept 50-423/86-02 on 860107-0224.Major Areas Inspected: Plant Events,Nonroutine repts,NUREG-0737 items,post-core Hot Functional Testing & Approach to Criticality.Number of Operational Events Identified as Noncompliances

ML20140J531
Person / Time
Site:	Millstone
Issue date:	03/21/1986
From:	Mccabe E NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20140J463	List: IR 05000423/1986002 ML20140J458
References
RTR-NUREG-0737, RTR-NUREG-737, TASK-1.C.1, TASK-1.G.1, TASK-TM 50-423-86-02, 50-423-86-2, NUDOCS 8604040350
Download: ML20140J531 (22)

Text

.-

.

U.S. NUCLEAR REGULATORY COMMISSION

REGION I

Report No.

50-423/86-02 Docket No.

50-42.*;

License No.

NPF-49 Licensee:

Northeast Nuclear Energy Company P.O. Box 270 Hartford, CT 06101 Facility Name: Millstone Nuclear Power Station, Unit 3 Inspection At: Waterford, Connecticut Inspection Conducted:

January 7,1986-February 24, 1986 Inspectors:

T. A. Rebelowski, Senior Resident Inspector,' Millstone 3

~F. A. Casella,~ Resident Inspector, Millstone 3 J. T. Shedlosky, Senior Resident Inspector, Millstone 1/2 R. J. Summers,. Project Engineer P. D. Swetland, Senior Resident Inspector,.Haddam Neck

~ Approved by:

&k llti 886 E. C. McCabe, Chief, Reactor Projects Section 3B date Inspection Summary:

Inspection 50-423/86-02, 1/7/86-2/24/86 Areas Inspected:

Routine onsite inspection by resident and project inspectors (470 hours0.00544 days <br />0.131 hours <br />7.771164e-4 weeks <br />1.78835e-4 months <br />).

Areas inspected included plant events and non-routine reports; NUREG 0737 items, Post Core Hot Functional Testing, Approach to Criticality, and Low Power Physics Testing.

Findings:

The licensee had several operational events, as documented in Detail Paragraph 8.

A number of these events were licensee identified non-compliances.

A Regional meeting was scheduled to discuss these events.

06gOhC PD

[

.

.

TABLE OF CONTENTS

<

Page 1.

Persons Contacted.....................................................

2.

S umma ry o f Faci l i ty Ac ti v i ti e s........................................

3.

Licensee Actions on Previous Inspection Findings......................

4.

Review of Licensee's Committments for Mode Changes and Verification mof License Conditions.................................................

i S.

Licensee Actions taken as a Result of TMI Action Plan Re Specified in NUREG 0737................................quirements

..............

6.

INP0 Accreditation of Training........................................

7.

Test Observation......................................................

8.

Review of Plant Events and Non-Routine Reports........................

9.

Review of Licensee Event Reports.....................................

10.

Review o f Pl ant Operations...........................................

11.

Containment High Range Radiation Monitor..............................

12. Management Meetings..................................................

13.

Unresolved and Other Items Involving Licensee Action.................

4 I

!

i (

..

l

-

.

.

DETAILS 1.

Persons Contacted J. Ferland, President J. Opeka, Senior Vice President', Nuclear Engineering and Operation R. Werner, Vice President, Generation and Construction Engineering W. Romberg, Station Superintendent J. Crockett, Unit 3 Superintendent F. Rothen, Construction Superintendent The inspector also contacted other licensee employees during the inspection, including members of the Operations, Radiation Protection, Chemistry, Instru-ment and Control, Maintenance, Reactor Engineering, Security and Training Departments.

2.

Summary of Facility Activities At the beginning of this report period, Millstone 3 was in Mode 5 at 170F and 380 PSIA, with a solid pressurizer. Containment was closed out on January 12 and the plant was brought to Mode 4 at 4:17 a.m. on January 13. Heatup on four reactor coolant pumps continued and the plant entered Mode 3 at 8:40 p.m. on January 15. Post-Core Hot Functional Testing was in progress as the plant was heated-up to normal operating temperature and pressure (557F, 2250 PSIA).

Initial criticality was achieved at 10 p.m. on January 23, at which time Low Power Physics Testing commenced.

The utility received a full power operating license (NPF-49) on January 31.

Three days were spent eliminating turbine vibration problems and reducing oxygen levels in the secondary systems.

Mode 1 (greater than 5% power) was attained on February 3.

From February 3 to February 24, power ascension testing progressed slowly due to a multitude of delays.

Among these were reactor trips and feedwater iso-lations while grooming the Steam Generator. Level Control System.

In addition, licensee-imposed secondary water chemistry sulfate requirements limited the plant to 30% power from the completion of 30% fuel conditioning on February 17 until shut down on February 22.

At the completion of this report period, the plant was in Mode 5, solid,_at 145F, in the process of draining and re-filling each steam generator to reduce sulfate' concentrations after secondary demineralizer resin intrusion.

3.

Licensee Action on Previous Inspection Findings a.

(Closed) Inspector Follow Item 85-62-07, Solid State Protection System (SSPS) Test Witnessing, Orange Bus Testing Incomplete.

The Orange Bus portions of SSPS monthly and refueling surveillances SP3446B11 Rev 0 and SP3446F31 Rev 0 have been completed.

The inspector had no further question.

'

b.

(Closed) Inspector Follow Item 86-05-08, Demonstrate the Filling of S Hety Injection Accumulators Using the Safety Injection Pumps.

On D ober 10, 1985, the licensee demonstrated the filling of each accumu-lotor, per OP3310B Section 7.1, with the Safety Injection pumps.

This item is closed.

c.

(Clos'ed) 83-PT-10 10 CFR 21 Report on GE AK-25, -30, -50 Breakers.

The Part 21 report addressed the low voltage circuit breaker electro / mech-anical trip.

This item was addressed in NRC report 84-09 page 4.

Addi-tional documentation reviewed included letter F0 432 dated April 27, 1984 NRC/NUSCO.

No unacceptable conditions were identified.

This item is administratively closed.

d.

(Closed) Inspector Follow Item 84-14-05, Three Concerns on Diesel Fuel Storage Tanks.

(1) The diesel oil transfer pumps were determined to be explosion proof.

(2) The tank saddle bolts were determined to require only one nut per foundation bolt.

The licensee exceeds the requirement by the placement of two nuts per bolt.

(3) The licensee has the ability to sample diesel fuel oil at the indi-vidual diesel generator fuel piping on Diesel Enclosures.

Permanent sample connections are addressed under a Design Deficiency Report (DDR 971) which recommends installation of fuel oil sampling piping.

No unacceptable conditions were identified.

e.

(Closed) Inspector Follow Item 84-02-05, Retest Recuirement for Electrical Tests.

A review of licensee's electrical tests incicated that a retest in these areas was required.

Satisfactory retests were performed as follows:

--

T 3415 P 003, Rev 0 Isolation Cabinets Group Rev 1-completed 3/12/84 T 3345 D 1101, Rev 0 Battery 6 H2 Detector Rev 1-completed 7/18/84

--

T 3345 CP 006, Rev 0 Battery Duty Cycle Test Sg-completed 1/10/86

--

This item is closed.

4.

Review of Licensee's Committments for Mode Changes and Verification of License Conditions Prior to proceeding to the next higher operational mode, the licensee actions to complete outstanding construction and license conditions identified in Operating Licenses NPF-44 and NPF-49 were reviewed by the inspector to insure that:

..

.

Surveillances necessary for operational entry into mode changes were

--

satisfactorily performed.

Co'nstruction work items were completed.

--

--

License items were addressed or completed.

Plant operational staffs met the operational experience levels.

--

Operating procedures were available and in use.

--

The following items were verified or observed.

a.

Electrical Items The licensee's review.of the exception. tracking system identified a num-ber of electrical items-during construction or on final building turn-overs. At the time of the approach to criticality, 15 items remained in this area.

The inspectors observed the placement of tray covers, cable wrapping, separation of cabling'in control building areas, and the use of Kellam grips to comply with cable support criteria.

One area of cable tray overfill was discussed with the licensee.

Engi-

-neering calculations by the licensee determined the weights of the cables, which were found acceptable.

Observations by.the inspector noted that trays were filled but properly protected by either bottom or top covers and that non-safety-related tray' were always~ separated from safety-s related trays.

Additional observations of cleanliness and of the carbon dioxide spray system availability to perform its function if called upon were satisfactory.

No deficiencies were identified.

b.

High Environmental Temperatures The licensee has determined that a number of areas in the Containment, Auxiliary Building, Engineered Safeguards Building, and Main Steam Valve Building have ambient temperatures greater than expected (90F).

As a result of these elevated temperatures, the licensee has identified a number of safety-related electrical items that could be subjected to accelerated degradation, thus effecting the Environmental Qualification of the electrical equipment.

The licensee stated that a temperature monitoring system will be installed to record the suspect high temperature areas.

That record will be used to determine any reduction in electrical lifetime of effected equipment (ie, position switches, transmitters, Barton dp. instruments, etc.) This item was included as a license condition.

The inspector observed the twelve RTDs (Resistance Temperature Detectors)

in containment and found two mounting problems that were corrected by the licensee.

Additional observations of the temperature detectors

.

L

.

.

readouts on a monitoring computer were made.

The temperature readouts were in agreement with measurements in the areas being monitored.

At the time, the containment temperatures were indicating 80*F.

The moni-tored areas of temperature were:

'3ECS-TE071-INSIDE CRANE WALL-80.8F 3ECS-TE072-INSIDE CRANE WALL-80.2F 3ECS-TE073-INSIDE CRANE WALL-83.5F 3ECS-TE074-INSIDE CRANE WALL-82.9F 3ECS-TE075-0VTSIDE CRANE. WALL-81.1F 3ECS-TE076-0VTSIDE CRANE WALL-80.9F 3ECS-TE077-INSIDE CRANE WALL-83.5F 3ECS-TE078-INSIDE CRANE WALL-82.1F 3ECS-TE079-0VTSIDE CRANE WALL-80.4F 3ECS-TE080-0VTSIDE CRANE WALL-81.1F 3ECS-TE081-INSIDE CRANE WALL-84.4F 3ECS-TE082-00TSIDE CRANE WALL-81.0F

No deficiencies were focnd in the present status and licensee plans for

this item, which will be reviewed again incident to routine inspect ton.

_

c.

-Emergency Communications During testing of emergency communications, the licensee was unable to communicate by the speaker system with personnel in certain areas of containment due to high noise levels.

The licensee had completed and satisfactorily tested all other areas.

The licensee decided that,' at the next refueling outage, visual warning

'

lights will be installed in containment to afford additional notification.

of personnel of emergency conditions.

In the interim, the licensee's containment entry-procedure 3312A (Rev 0) requires personnel to be

~;

equipped with radio communications for subatmospheric entries, with back-up teams available at the personnel' access hatch if assistance is necessary.

The procedure meets the need to continuously monitor person -

,

j nel to permit their evacuation in the event of an emergency.

The in-spectors verified that personnel in containment did receive.and transmit

instructions with the radio communication equipment during periods of'

high noise levels in containment.

No unacceptable conditions were identified.

'

d.

Control Circuit Protection to Fire Water Stop Valve The licensee, upon completion of fire protection review, identified the need to protect an electrical conduit that actuated the fire line stop valve to Containment.

The protection was supplied by a Thermo Lag-330

,

fire barrier system.

This method of protection is described in Technical Note 20684, which describes acceptable fabrication methods.

.

.

The methods necessary to achieve fire protection involved binding a-Thermo Lag pre-shaped conduit sleeve to the conduit and applying a coat-ing of fire protection subliming material to the sleeve.-

The method by which protection is afforded to conduit is that the Sub-liming Material, Thermo Lag-330, is a water based, subliming, thermal-activated ~ fire resistant material which volatilizes at a fixed tempera-ture, exhibits a volume i1 crease through the formation of a multi-cellu-lar material which blocks heat and thus protects the substrate material to which it is applied.

The inspectors witnessed the application of preshaped conduit protection and subliming material.

Final licensee inspection was' satisfactory.

As.an acditional measure, to insure fire protection and loss prevention,.

the licensee has committed to NRR (0peka to Grimes 1/30/86 letter B11975)

that the licensee will engage an independent audit team on a tri-annual.

basis.

The inspector had no further questions on these matters.

e.

Additional Commitments Additional items reviewed were the licensee's response to NRR in the areas of 1) Inservice Inspection Program, 2) Containment Sump Water Temperature Monitoring, 3) Moi.sture in Air Start System, 4) Operating Staff Experience Requirements, 5) Initial Test Program, 6) SPDS Commit-ments.

The review consisted of verification that documentation has been submitted to NRR and that hardware and personnel requirements were in place when the licensee exceeded five percent power.

No inadequacies were identified.

5.

Licensee Actions taken as a Result of TMI Action Plan Requirements The NRR Staff has reviewed the licensee's submittals associated with the NUREG 0737 items and has documented the results of this review in the Safety Evalu-ation Report (SER) related to the operation of Millstone Nuclear Power Station Unit 3.

During this inspection, the resident inspector verified that the following licensee actions have been taken as described in the SER.

a.

I.C.1.3.b (85-TM-29) Additional Review by Licensee of Accident Procedures gpen)

Item I.C.l.3.b. addresses revising short-term accident procedures.

Re-view by NRR has been completed, with three outstanding open items ad-dressed in Safety Evaluation Report Supplement.

.

The first item addressed the procedures for three-loop operation which requires additional review on the part of the licensee and remains open.

Further information is to be submitted to NRR.

In the meantime, three-loop operation is not permitted.

Two procedural changes were committed to for the. Reactor Vessel Level Indication System.

The inspector verified that changes to procedures in E0P-35, FRC-2 (Degraded Core Cooling Guidelines) and E0P-35, ERI-3 (Response to Voids) have included the Reactor Vessel Level Indication System parameters.

These items are closed.

b.

I.G.I. (85-TM-31) Preoperational and Low Power Testing Program (Closed)

and I.G.l.2 Review of Test Program vs. Reg. Guide 1.68 and FSAR Chapter 14 (Closed)

The licensee has completed the preoperational and low power physics testing.

The test program met the post-TMI requirements outlined in NUREG 660, NUREG 0694, and NUREG 0737.

NRC Region I Division of Reactor Safety inspectors and the resident-inspect. ors have monitored, by frequent-inspections including total coverage of initial criticality, preopera-tional and low power testing.

The results are documented in NRC Reports 85-03, 06, 10, 14, 18, 24, 25, 27, 43, 51, 55, 59, 75, and 76, and 86-01 and 86-07.

Minor discrepancies identified in the testing were acceptably corrected by licensee.

These TMI. items are closed.

c.

I.G.1.1 (85-TM-31) Training for each shift Experience levels of the intended shifts were reviewed by NRR and found to need additional experienced personnel during the first six months after exceeding the 5% power plateau for three of the six shifts.

Experience level upgreding for these shifts was met by the addition of-three experienced 11a nsed SR0s.

A training program in plant familiarity,-

Technical Specifications, and emergency response procedures was insti-tuted.

Successful completion of these training elements was documented for all three individuals.

A review by the inspector of the licensee's written examination was performed.

No deficiencies were noted.

Also, the following were noted:

(1) The licensee has trained operating shifts using the simulator on the various accident scenarios.

(2) The shifts have worked together, in part, during the preoperational test program.

(3) Each shift does have an individual trained in core thermal hydraulics.

The above items meet the NUREG 0737 criteria on this ite.

"

6.

INP0 Accreditation of Training

~

The licensee appointed (Fall-85) a new manager of Corporate Training.

Pre-viously,.they submitted an incomplete schedule of self-evaluation of their training to INP0 on January 13, 1984.

The templete schedule was finally sub-mitted on January 10, 1986 following a reanalysis of the overall training schedule.

The three Millstone plants are to be considered under one training program with core training and specific supplemental training modules for both license and non-license personnel.

Further NRC review of this item will be accom-plished during routine NRC review of licensee training.

No unacceptable con-ditions were identified in this area.

7.

Test Observation The inspectors witnessed portions of various Post Core Hot Functional Tests, Low Power Physics Tests, and Power Ascension. Tests.

These included INT-5000

. Appendix 5031, Full Flow Rod Drop Testing;" Appendix 5007, " Pressurizer Spray Bypass Flow;" Ap'pendix 5015, " Digital Individual Rod Position Indication;"

Appendix 5018, Rod Control;" and Appendix 5031 " Volume Control Tank Degas."

Various low power physics tests and the natural circulation test were wit-nessed with the help of regional specialists.

Details will be included in reports 50-423/86-01, /86-07.

Test performance was monitored for conformance to test procedures, operation of equipment in accordance with plant operating procedures, and good engineering practices.

No unacceptable conditions were identified during this inspection.

8.

Review of Plant Events and Non-Routine Reports a.

Overview of Events The following is a chronological summary of noteworthy events occurring within this report period.

DATE EVENT NRC IDENTIFIED ROOT CAUSE 1/15 Condensate S_torage Tank Rupture due Design deficiency to frozen shut relief valve 1/16 Safety Injection (SI) and Reactor Trip Operator error on Steam Line low pressure, Mode 3 1/18 Reactor Trip on high source range flux Management error due to welding cable induction, Mode 3 1/19 SI and Reactor Trip on steam line Operator error low pressure, Mode 3 1/23 Mode change while prohibited by Techni-Operator error cal Specification (TS) action statement

.

..

DATE EVENT NRC IDENTIFIED ROOT CAUSE 1/25&

Operations unaware of breached pressure Management eor 1/27 boundaries.outside containment due to inadequate work order clarity 2/2 Operations failure to comply with Management error TS action statements 2/4~

Feedwater Isolation, Mode 1 Equipment malfunction 2/4 Reactor Trip on low S/G level, Mode 1

' Operator error 2/5 Rod Drop, Mode 2 Equipment malfunction 2/5 ESF Actuation, Control Building Equipment malfunction Isolation 2/6 Feedwater Isolation, Mode 1 Personnel error 2/6 Primary System Leak (0.5gpm), Mode 1 Equipment malfunction.

2/7 Reactor Trip, Mode 1 Equipment malfuncticn 2/10 Reactor Trip, Mode 1 Personnel error-2/12 Reactor Trip, Mode 1 Operator error 2/13 Reactor Trip, Mode 1 Equipment malfunction 2/14 Feedwater Isolation, Mode 3 Operator error 2/22 Unscheduled plant shutdown to correct Equipment Malfunction Steam Generator (SG) Chemistry b.

Details of Events (1) Condensate Storage Tank Rupture On January 15, the 300,000 gallon Condensate Storage Tank (CST)

ruptured from overpressurization which resulted from filling the tank while its relief valve was frozen shut.

At the time of fail-ure, the' auxiliary feed pumps were drawing off about 60 gpm from

!

the CST, with their minimum flow recirculation path going to the Demineralized Water Storage Tank (DWST). Water treatment was making up about 120 gpm, for a net 60 gpm in-flow.

The tank was about half full of water with a' slight nitrogen overpressure.

Ambient tem-perature was well below freezing.

The tank relief, a dual acting relief and vacuum breaker, was not heat-traced and was located under

a protective hood which condensed moisture from the wet nitrogen i

.

.

escaping from the valve.

Ice formed externally and, as the valve internals cooled, disc icing closed off the relief path.

Tank pressure at the time of failure was calculated to be between-2.5 and 6.0 psig.

The relief was set at 0.75 psig.

Damage was extensive.

The side wall to roof joint was ripped open around 2/3 of the tank circumference, allowing a large portion of the roof to settle down inside the tank, below the top of the tank wall.

The upper half of the tank wall was buckled and the eight anchor bolts and chairs were plastically deformed.

The CST was originally designed and fabricated to ANSI B96.1 as an atmospheric storage tank with a 10-inch vent.

On a Steam Generator Owners Group recommendation,'a nitrogen blanketing system was added to the CST.

This modification blank-flanged the 10-inch vent and added the dual-acting relief / vacuum breaker in its. stead.

As corrective action, the CST is being rebuilt and the 10-inch vent line will be re-utilized with a rupture disc as backup relief pro-tection.

The relief valve is temporarily housed in a heated en-closure, pending replacement with a new configuration which can be heat-traced.

The same overpressure protection modifications are being made to the Condensate Surge Tank, the only other tank on site to be similarly de~ signed.

As an interim measure, the DWST was used as a source of steam generator feedwater while the secondary plant was offline, and the Condensate Surge Tank was backed up by 2 temporary storage tanks for secondary makeup during turbine operations.

This utilization of the DWST 16,000 gallon reserve volume (between the 334,000 gallon low limit and the 350,000 gallon capacity) re-quired fit-up of temporary fire hoses to the DWST overflow line and the use of a vent line for overfilling indication and protection.

The inspectors closely monitored licensee use of the DWST during this interim period, and found no cause for concern. When the tank was being filled, a person was on watch at the vent line to monitor an overfill condition.

There was a Heise gage, calibrated in inches of water, attached to the tank piping to provide accurate level in-dication.

Further NRC follow-up on this item will be conducted in-cident to routine inspection.

(2) Safety Injection and Reactor Trip Due to Steam Line Low Pressure On January 16, the plant was in Mode 3 at 2230 psia and 553F with 2 reactor coolant pumps running.

The Main Steam Isolation Valves (MSIVs) were shut and primary temperature was being controlled with the steam generator atmospheric steam dumps.

A control operator opened the "A" steam generator atmospheric dump too quickly, re-sulting in a step change of about 80 psi in steam line pressur.

.

10-This-caused a Safety Injection (SI) with a Reactor Trip.

Both reactor trip breakers opened. All control rods were fully inserted prior to the. event.

All Engineered Safety Feature-(ESF) Systems actuated properly.

The Millstone Unit 3 ESF logic for low steam line pressure contains-a lead / lag circuit as an anticipatory feature.

When the P11 inter-lock is satisfied (Pressurizer pressure above 1985 psia), the low pressure logic circuitry is in effect.

This circuit contains a bistable with a voltage setpoint of 5.0662V, equivalent to 658.6 psig. At the input to this bistable is a lead / lag unit with a 10.

second time constant.

The time responses for the lead / lag unit includes an output ~ signal exceeding the bistable trip point if there exists a rate of change of steam line pressure of -10 psi per second for 8 seconds, or a -44 psi step change.

The latter time response initiated the SI and Reactor Trip in this case.

Restoration from the Safety Injection was delayed by a fault in the B train Emergency Diesel Generator (EDG) Sequencer.

As a result of the fault, the Reactor Plant Component Cooling Water (RPCCW)

non-safety supply header could not be restored, nor could the "B" EDG be reset.

The "B" EDG was stopped by using the Emergency Shut-down feature.

The sequencer fault was traced to a burned out diode.

The diode failure led to high voltage (48VDC) failure of four cards within the sequencer.

After replacing the diode and cards, the sequencer tested satisfactorily and the RPCCW non-safety supply header was restored.

"A" EDG experienced an air leak on the air < tart distributor.

That leak was severe enough to prevent recharging the air start bottles to normal pressure. 'The licensee was in a limiting condition for operation action statement for loss of both EDGs for approximately 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> while the "A" EDG air start system was being repaired.

'The licensee is evaluating whether a design improvement is appro-priate for the lead / lag ESF logic.

The inspector had no further questions on this item or the EDG problems.

(3) Reactor Trip on High Source Range Flux On January 18, while shut down in Mode 3 with all rods bottomed, a reactor trip occurred when Channel "A" of source range (SR) nuc-lear instruments went above 10E5 counts per second (CPS).

Reactor trip breakers had been closed for Rod Drop Testing in accordance with 8000 series procedures.

Both' reactor trip breakers were veri-fled to be open after the trip.

The unexpectedly high source range level was found to be caused by electromagnetic interference from welding in the Containment.

The Shift Supervisor ordered welding stopped.

Source range counts returned to less than 3 r

.

.

To prevent recurrence, plant startup procedures 0P3201 Ch. 14 and OP3202 Ch. 6 were revised with warnings to restrict welding in the vicinity of NItcables and devices from the time of closing the trip breakers until power is above the source range P6 permissive actu-

'ation point.

In a related matter, on February 6, the inspector was in the control room when, with both trip breakers open, both saurce range channels were indicating greater than 2000 counts per second.

The problem was determined to be electromagnetic interference (EMI) from high frequency welding leads running close to SR instrument cables in the cable tunnels.

Prior to having made this determination, opera-tors veri.fied that reactor coolant system (RCS) baron concentration was above the 1.6% dk shutdown margin, that both reactor trip breakers were actually open, and that no change in RCS. temperature had occurred.

The amount of welding in the plant has decreased and this problem has not been repeated.

The inspector had no further concerns.

Routine inspection will address the adequacy of licensee controls on activities which produce EMI.

(4) Safety Injection and Reactor Trip Due to Steam Line Low Pressure On January 19, the plant was again in Mode 3 at 2230 psia with Main Steam Isolation Valves shut.

Rod control was in manual and steam generator atmospheric dumps were-in manual with pressure setpoints set high to prevent a possible recurrence of the then suspected but unconfirmed cause of the SI on January 16. Operators allowed RCS temperature to rise to 567F, increasing secondary steam pressure until two B steam generator safety valves opened.

The resulting 50 to 100 psi step drop in steam header pressure initiated a Safety Injection with Reactor Trip. All ESF systems actuated properly and restoration was completed without incident.

Followup investigation showed that the two safeties opened slightly below their set pressures.

A temperature of 567F corresponds to a saturation pressure of 1160 psig.

Safety valve RV228 setpoint was 1185 +/-1% (1173 to 1197 psig) and safety valve RV238 setpoint was 1195 +/-1% (1185 to 1209 psig), yet both valves lifted.

Sub-sequent surveillances performed by simmer testing the valves using a hydraulic test unit provided by the vendor revealed that RV22B was lifting at 1141 psig and RV23B was lifting at 1162 psig.

Both valves were reset to within their setpoint tolerances and satisfac-torily retested.

These surveillances were last performed on October 29, 1985 during hot functional testing.

At that time, all 20 valves were set and tested within their setpoint tolerance.

.

Although this reduction in setpoints was in the safe direction, the inspector was concerned that setpoint drift could occur in the op-posite, higher setpoint direction.

In discussions with licensee engineers and NRR staff, further information was obtained.

Testing performed by an independent test laboratory under contract by the safety valve vendor has revealed a very probable mechanism for set-point drift.

The valve springs are subject to moderately elevated temperatures, approximately 150F, and are under constant compression.

These_two conditions lead to low level time dependent creep, a de-formation of the crystalline structure that causes spring relaxation over time.

Since the spring force on the pilot disc determines the-valve lift pressure, a relaxation of that spring force will result

,

'

in a decreased lift pressure, which occurred in this case.

The inspector found no probable mechanism that could be responsible for set point drift in both directions.

He therefore concluded there was.no safety concern.

Adequacy of licensee corrective actions to minimize operator-induced plant transients will be reinspected incident to routine inspection.

(5) Mode Change While in an Action Statement On January 23, the plant was brought from Mode 3 to Mode 2 whi1e

~

in a Technical Specification (TS) limiting condition for operation (LCO) action statement which prohibited Mode changes.

One of two Control Building inlet ventilation radiation monitors (3RMS*RE16A)

was out of service, placing the plant simultaneously in two LC0 ac-tion statements.

The first TS, 3.3.2.C (Table 3.3-3.7.e), deals with ESF actuating sensors.

No mode changes are allowed while within this action statement (TS 3.0.4 is applicable).

The second TS, 3.3.3.1.b (Table 3.3-6.3.a), deals with radiation monitors for plant operation and allows mode changes while within this action statement (TS 3.0.4 is not applicable).

The shift supervisor was aware that a mode change was permissible by the second LC0 action statement and surmised this to also be true for the first action statement.

He thereby allowed a mode change in violation of plant technical specifications.

This was a viola-tion of NRC requirements.

The root cause of this occurrence was classed as operator error by the inspector.

This problem was discovered by the plant staff and identified to the inspector by the licensee.

The control room ventilation system was on filtered recirculation as required by the action statements; the inspector had no safety concern.

To prevent recurrence, this event was discussed by the licensee with all senior licensed operators.

The inspector had no further questions on this ite.

.

(6) Operators Unaware of Breached Pressure Boundaries On January 25, with the plant in Mode 2, the Control Room pressure boundary was-discovered to have been breached without the knowledge of the Shift Supervisor.

On January 27,-with the plant again in Mode 2,'the Supplementary Leak Collection Removal System (SLCRS)

boundary was discovered to have been breached, again without the knowledge of the Shift Supervisor.

In-both cases, the licensee entered'the apppropriate action statements for LCOs and secured the respective penetrations as soon as possible.

The safety concern here was the possibility of plant operations personnel being unaware of the inability to maintain either Control Room pressur.ization or SLCRS negative pressure during an accident.

These were also viola-tions of NRC requirements.

In both cases, approved work orders were in effect to remove pres-sure envelope electrical penetration wall seals which were not ade-quately identified to operations personnel during their review.

The seals were identified by penetration number and construction map, with which operators had little familiarity.

As corrective action, a form that clearly identifies all boundary penetrations and utilizes identification by building, elevation and room has been added to the work order package format.

The inspector is currently

,

'

following the. licensee implementation of this corrective action (UNR 423/86-02-01).

(7) Operations Failure to Comply with Technical Specification Action Statement On February 2, the licensee made a general interest ENS notification on a failure to comply with 2 LC0 action statements.

The first was a failure to implement 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> grab sampling of the plant vent as required by LC0 3.3.3.10 with the reactor plant ventilation dis-

,

charge radiation monitors out of service.

The monitors were taken out of service on January 31, at 9:40 a.m.

This oversight was

'

noticed at 10:10 a.m. on February 2 and the sampling program was then satisfactorily implemented.

In the past, the inspector has noticed a difference in shift supervisor grab sample log keeping methods.

Some supervisors routinely recorded 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> grab sample

.

results while others depended on the chemistry department to record the results.

Previous checks by the inspector of both chemistry

.

department records and shift logs have found no omitted survell-lances.

It was noted, however, that a consistent requirement for log keeping may have prevented this omission.

The second LC0 was 3.8.4.1:

with containment penetration overcur-rent protection devices inoperable, the backup breaker must be

.

tripped or the inoperative breaker racked out, and they must be so

'

verified every 7 days.

Or February 2, at 2:20 p.m., the shift supervisor noticed that these checks had not been ccmpleted since

,

t s

--

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

-

.

.

January 13. These are both violations of NRC requirements.

The problems were self-identified, immediately reported and corrective action was quickly implemented.

The inspector is following licensee corrective action for properly monitoring surveillance requirements. (UNR 423/86-02-02)

(8) Feedwater Isolation, Mode 1 On February 4, a Main Feedwater Flow Isolation (FWI) occurred during a changeover from main feedwater by pass flow operation to control on main feedwater regulating valves. Feed water leakage from "B" main feedwater regulator raised the level in B steam generator to 82% which caused the FWI.

The unit was at 6% power on the motor driven feed pump.

Operators restored the FWI, returned to main feed bypass regulator control and restored all steam generator levels.

The feedwater system has since been adjusted to prevent recurrence.

The inspector had no further questions on this item, which occurred during planned grooming of the feedsater system.

(9) Reactor Trip on Low Steam Generator Level A reactor trip occurred on February 4 due to low-low level 'on the

"B" Steam Generator while the plant was at 6% power.

"B" Steam Generator level control was in manual at the time, the control room operator was using the "B" Steam Generator chart recorder to monitor level.

After the trip, the low-low level trip point was checked and found to be 23.5% (2 out of 4 coincidence for trip is required on any Steam Generator).

The chart recorder was reading 36% level, trending down, at the time of the trip.

Subsequent investigation Showed that, at the time of the trip, 2 of 4 channels feeding Solid State Protection were at or below the trip setpoint, while the one channel feeding the recordcr was at 26%.

Preventive maintenance checks were performed on all 4 of the HAGAN OPTIMAC 3 pen Steam Generator recorders with no deficiencies noted.

The 10%. difference between the channel signal and the chart recorder pen location has been attributen to ink tube interference with pen movement.

One of the tubes in the B recorder was free from a restraint, and that may have led to the interference.

During the post trip investigation, other operators stated that the analog 1cvel meters appeared to be reading correctly.

The Plant Operations Review Committee (PORC) concluded that the root cause of the trip was the operator monitoring only one level indicator while that indicator was reading incorrectly.

Subsequent corrective action has included operator reinstruction and inspection of all chart recorders for problems which would impede pen movement.

No such problems ware found.

The inspector had no further questions on this item.

Corrective actions to preveht operator-induced plant transients will be evalu-ated for effectiveness during routine inspections.

.:

___.

....

.

,

.

-

'

(10) Rod Drop

-

On February 5 the licensee was pulling Shutdown Bank A to the full out position when Shutdown Rod B-4 dropped into the core.

Plant

parameters at the time of event were 200 counts per second on the p

source range, 555F and 2145 psia.

Prior to the rod dropping, B-4

had a -12 step deviation at 222 steps out.

Bank A was driven all

}

the way in, B-4 remained at -12 deviation until the bank was bot-

-

tomed.

Rod B-4 then dropped while the bank was being withdrawn.

.

I Following the rod drop, the licensee completed 18 hours2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> of trouble-shooting on rod B-4 control circuitry.

No indication of the cause

_

of the drop was identified.

The rod was exercised between 6 and

_

48 steps and fully withdrawn (228 steps) while monitoring rod drive power supply output and taking sound traces of mechanical movement.

!

These traces matched those taken during phase 1 and rod drop tests.

The traces have been sent to the vendor for further analysis.

The

{

inspector had no further questions.

(11) ESF Actuation, Control Building Isolation

.i At 3:20 p.m., on February 5, 1986, the plant experienced a Control Building Isolation due to a high signal spike in the Control Build-h ing Radiation Monitoring System (RMS).

The signal tripped control room ventilation and caused the dampers to shut, but a 60 second time delay allowed operators sufficient time to determine that an

-

actual high radiation condition did not exist and reset the isola-

-

tion prior to control room pressurization.

Electronic noise is an g

ongoing problem in the RMS system.

Millstone 3 experienced approxi-E mately 90 Control Building Isolations during February 1986 due to r

spurious signals on the "A" channel.

The normal range of these E

scintillation detectors is 50Hz to 55KHz, with the isolation trip ik at the upper end.

Detector saturation occurs at 166KHz causing a saturation alarm but no ventilation trip.

The engineer assigned

"

ll;.

to correcting this problem believes that incomplete coupling is b

driving the scintillations above the trip setpoint (55KHz) but not P

all the way to the saturation alarm setpoint (166KHz).

Investiga-L tion is ongoing. The licensee made a 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> general interest noti-p fication of this one acteation via the ENS.

Corrective action of

RMS noise problems will be reviewed incident to routine inspection.

_

b (12) Feedwater Isolation A main feedwater isolation occurred at 10:25 a.m. on February 6 dur-ing power ascension testing.

The plant was at 14% power with steam

-

generator levels being maintained on the bypass valves while steam generator level control system adjustments were in progress.

Plant temperature was being controlled by the steam dumps in the steam E

pressure mode with a no load setpoint of 1097 psi (557F); three

-

-

m-m

.

.

steam dumps were partially open.

"B" and "C" steam generator at-mospheric dumps were unisolated and set to open at 1125 psi.

The next power ascension test scheduled was the steam dump control test.

An instrument technician, setting up for the next test, was con-necting a chart recorder to the main steam pressure transmitter-(PT507).

This instrument.tas providing the sensed parameter control signal to the steam dump controller.

He shorted two pins, causing a voltage decrease which caused the steam dumps to shut.

The re-sulting shrink led the feed station operator to quickly raise generator levels to keep above the low level alarm point.

Tavg went to 564F before control rods were manually driven in.

The two at-mospheric dumps began opening, causing levels to swell; the feed operator began reducing the feed rate.

Simultaneously, the instru-ment technician removed the test leads from the pressure transmitter, causing the true steam pressure signal to pass through to the steam dump controller (a proportional integral type), which opened all 9 dump valves.

An operator quickly blocked the dump signal, but not before C steam generator swelled above the feedwater isolation setpoint.

The feedwater isolation circuit functioned as designed.

The opera-tors manually started the motor driven auxiliary feedwater pumps, restored generator levels, and recovered the plant. The entire scenario lasted about 60 seconds.

Operator action to block the steam dump signal was noteworthy.

Testing was performed to determine if PT 507 was damaged by the short.

It was not.

The inspector had no further questions on this item.

Effectiveness of licensee corrective actions for personnel-induced plant transients will be addressed during routine inspec-tions.

(13) Primary System Leak At about 10:30 a.m., on February 6, a health physics technician in containment noted leakage (0.5gpm) from a compression fitting on a loop flow transmitter.

The transmitter was isolated and the plant was taken below 5% power at 1:00 p.m. to effect repair.

Repairs were completed in about 1/2 hour and testing (SG 1evel control system grooming) was then resumed at about 15% power.

The inspector had no further questions on this item.

(14) Reactor Trip On February 7, with the plant at 15% power for steam generator water level control grooming, an auxiliary steam relief (ASS-RV21) lifted and stuck open.

Rod control was in manual, moderator temperature coefficient was just slightly negative; the increased steam flow rapidly lowered temperature.

Operator response was to pull rods to restore Tavg to 557F.

Steam generator water level was being

.

.

controlled by the feed regulating bypass valves in automatic.

Their control circuits compare actual steam generator level to auc-tioneered high neutron. flux to establish valve position.

As rods were being pulled, flux was increasing, causing the bypass valves to open to about 80%. When steam generator levels reached 65%, the operator took manual control of the bypass valves to reduce feed flow, but the coastdown and thermal expansion of the non preheated feedwater caused steam generator levels to exceed the 80% mark, resulting in a feedwater isolation.

Steam flow remained at 15% of rated flow. With the feedwater iso-lation in effect, levels began lowering quickly.

The operators started both motor-driven auxiliary feed pumps and reset the FWI signal, but were unable to get control of the level swing in two of the steam generators. A reactor trip was actuated by two steam generators at 23% level.

The turbine-driven auxiliary feedwater (TDAFW) pump automatically started.

Plant response and safety system performance was as per design.

The licensee reduced the gain on the flux input signal for the feed bypass valve control circuit to prevent recurrence.

The inspector had no further questions.

(15) Reactor Trip On February 10, the reactor was at 15% power steaming for Steam Generator Level Control (SGLC) grooming; feed regulating bypass valves were in automatic.

A test engineer, using long makeshift test leads, plugged a brush recorder into the front jack of a set of double stacked test jacks on a SGLC bypass valve control card.

The test lead tip passed through the front jack into the back jack and shorted the steam generator level set program for all four con-trol cards (level set.is bussed).

On-Main Board 5, the feed station operator saw 4 high level deviation alarms.

All 4 bypass valves went shut.

Levels swelled about 5% initially on the transient but rapidly lowered with 15% steam flow, The operator switched all 4 s

bypass valves to manual and began to feed, but the levels were dropping too fast.

"B" Steam Generator low-low level caused a reactor trip and motor-driven auxiliary feedwater pump start.

The three remaining levels reached the low-low setpoint immediately thereafter; the turbine-driven auxiliary feedwater pump automatically started.

The plant was stabilized at normal operating temperature and pressure in Mode 3 about 15 minutes after the trip.

Plant re-

,

sponse and safety systems performance was as per design.

Since the licensee had instructed personnel to use short test leads, this event is _ classed as due to personnel error.

As a measure to prevent recurrence, the licensee modified all cards having the troublesome stacked jack arrangement by inserting a high dielectric material between the front and back jacks.

The inspector had no further question.

.

'(16) Reactor Trip, Mode 1 On February 12, the reactor tripped from 30% power on low-low steam generator levels due to a loss of feed. The operators were shifting feed from the turbine-driven main feed pump to the motor-driven pump while a single condensate pump was running.

The lone condensate pump was unable to maintain suction pressure to both feed pumps

-

which resulted in a low suction pressure trip on the motor-driven main feed pump when all feed was shifted to it. The operators were unable to restore the turbine driven pump in time to prevent the reactor trip. The procedure (OP 3321) governing feed pump transfer did not adequately address the transfer from turbine-driven to motor-driven pump. As a result, the operators failed to start a second condensate pump. The licensee has revised OP 3321 to prevent recurrence and has reinstructed the operators.

The inspector-had no further questions.

(17) Reactor Trip On February 13, with the plant in Mode 1 at 15% reactor power, a reactor trip occurred on Solid State Protection System (SSPS) "A" and "B" train General Warning Alarms.

"A" train was being prepared for monthly surveillance testing; the "A" reactor trip bypass breaker was shut, giving the "A" General Warning Alarm.

A spurious

"B" train SSPS General Warning alarm occurred, giving a reactor trip.

The "B" train of SSPS has been instrumented since that time to de-termine the cause of the spurious signal.

There has been no repeat activity.

Licensee monitoring of the SSPS system will be examined incident to routine inspection.

-(18) Feedwater Isolation

<

On February 14, with the plant in Mode 3, operators were equalizing around and opening MSIVs in prepararion for a reactor startup.

"A" Steam Generator level swelled from 55% to 93% when "A" MSIV was opened at about 35 psid, resulting in a feedwater isolation (Fhl).

All equipment functioned as designed.

The FWI signal was reset and the feed system configuration restored.

Effectiveness of licensee corrective actions in preventing such events will be examined in-cident to routine inspection.

(19) Unscheduled Plant Shutdown

.On February 21, because steam generator sulfate concentration could not be brought below the steam generator owner's group guidelines by blowdown, the plant was shut down in order to lower the sulfate levels by draining and refilling the steam generators. Mode 5 was attained at 9:00 a.m., February 2,'

and the pressurizer was brought i

solid at 120F RCS temperature. Conaensate demineralizer resin in-trusion was suspected as the cause of the elevated sulfate concen-

%

,,

.-

- -, - -, -

,, -

g

,.- -. - - - -, - - -

.

.

tration. At the end of this reporting period, all condensate de-mineralizer resin retenticn filters were being replaced after wear induced holes were found in some of them. Further review of outage activities will be performed incident to routine inspection.

9.

Review of Licensee Event Reports (LERs)

LERs submitted during this report period were reviewed.

The inspector as-sessed whether further information was required, if'there were generic im-plications, LER accuracy, adequacy of corrective actions, and compliance with the reporting requirements of 10 CFR 50.73 and Administrative Control Proce-dure ACP-QA-10.09.

Selected corrective actions were checked for thoroughness and implementation.

Those LERs reviewed were:

--85-002-00 Reactor Trip due to Misapplication of Bypass Jumper Control Procedures

--85-003-00 Violation of Plant Technical Specifications-480VAC Emergency Bus Tagout

--86-004-00 Violation of Plant Technical Specifications-Mode Change with Action Statement In Effect (See Section 8, Part 5 of this report)

The inspector had no further questions on these LERs.

10.

Review of Plant Operations The inspector observed plant operations during regular and back shift tours of the following plant areas:

Control Room Fence Line (Protected Area)

Au.viliary Building Yard Areas Diesel Generator Room Turbine Building Intake Structure Vital Switchgear Areas Main Steam Valve Building The tours of the control room included observation of instrument parameters that are necessary for conformance to the Technical Specification requirements.

The alarm conditions which were in effect and those alarms received at the control room during the period of observation were reviewed.

The operators were cognizant of board conditions.

Shift manning was in conformance with Technical Specifications, with the shif t advisors in place prior to exceeding five percent power.

Plant housekeeping controls were observed.

Comments re-lated to improvements were expressed to the Shift Supervisor on duty, and the minor cleanups involved were accomplished promptly.

Also, during plant tours, the various logs in the Control Room, Chemistry department and Health Physic department were reviewed to determine if entries were made of events and that

-

-

_

_

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _.. _ _ _.

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

_ _ _ _ _ _ _ - _ _

,

20 records of equipment failures were recorded.

In addition, the inspector ob-served selected actions concerning site security including personnel monitor-ing, access control and placement of physical barriers.

No deficiencies were identified.

11.

Containment High Range Radiation Monitor The wide range containment radiation monitor consists of two channels and is designed to operate in a post-accident containment environment.

Prior to plant heatup and initial core criticality, a series of failures occurred in both channels.

These failures occurred as detector cable insulation broke down under instrument channel high voltage.

The licensee attempted to restore the two channels through a process of vacuum drying and heat tracing.

These attempts were not successful. Although a spare cable was provided in the system design for each channel, the licensee was not successful in restoring.

at least two cables for each detector.

Technical Specification 3.3.3.6 requires that a minimum of two channels be operable for Modes' 1, 2, and 3.

To meet this requirement, the licensee em-barked on a modification to replace the two inoperable channels with detectors, cable and electronics of a different design.

The detectors for the new system were to replace the detectors of the original system in the same location.

The licensee also restored one of the original channels to take advantage of the availability of two acceptable short lengths of cable.

However, this de-tector was relocated in the containment 24 foot elevation because of cable length restrictions.

The licensee has also completed the installation of one of the two new channels.

These were both operable for heatup and initial criticality.

The licensee intends to install'the detector and channel elec-tronics as soon as possible and will then have three operating channels.

There were no unacceptable conditions identified.

Containment radiation monitor operability will be re examined periodically incident to routine inspection.

12. Management Meetings At periodic intervals during the course of this inspection, meetings were held with senior plant management to discuss the scope and findings of this in-spection.

No proprietary information was identified as being in the inspec-tion coverage.

At no time during the inspection was written material provided to the licensee by the inspector.

13.

Unresolved and Other Items Involving Licensee Action Unresolved items on which additional information is required to determine whether the licensee is to be issued a violation are contained in Details 4.b.(6) and 4.b.(7) of this report.

Other items on which licensee actions have been found acceptable but whR n will be re-examined on a sampling basis are described in Report Details 4.b, 4.c, 4.d, 6, 8.b.(1), 8.b.(2), 8.b.(11),

8.b.(17), 8.b.(18), and 11.