IR 05000423/1986028

Insp Rept 50-423/86-28 on 860812-1006.No Violations Noted. Major Areas Inspected:Plant Operations,Radiation Protection, Shutdown Planning,Physical Security,Fire Protection, Surveillance & Maint

ML20215M779
Person / Time
Site:	Millstone
Issue date:	10/28/1986
From:	Mccabe E NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:
Shared Package
ML20215M769	List: IR 05000423/1986028 ML20215M768
References
50-423-86-28, IEB-85-002, IEB-85-003, IEB-85-2, IEB-85-3, IEIN-80-22, NUDOCS 8611030285
Download: ML20215M779 (24)

Text

_ _ - _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ - - _ _

.

.

U.S. NUCLEAR REGULATORY COMMISSION

REGION I

Report N /86-28 Docket N License N NPF-49 Licensee: Northeast Nuclear Energy Company l

P.O. Box 270 l Hartford, CT 06101-0270 Facility Name: Millstone Nuclear Power Station, Unit 3 ,

Inspection At: Waterford, Connecticut Inspection Conducted: August 12 - October 6, 1986 Inspectors: J. T. Shedlosky, Senior Resident Inspector F. A. Casella, Resident Inspector E. L. Conner, Project Engineer M. C. Kray, Reactor Engineer Approved by: [2 E. C. McCabe, Chief, React #r Projects Section 3B b d /0[2 8-/r2l;

'Date Inspection Summary:

Areas Inspected: Routine on-site resident inspection (203 hours0.00235 days <br />0.0564 hours <br />3.356481e-4 weeks <br />7.72415e-5 months <br />) of shutdown

,

planning, plant operations, radiation protection, physical security, fire protec-l tion, surveillance and maintenanc Results: This inspection identified satisfactory performance in all area The following conditions were nonetheless noted as significan Plant trips occurred due to steam generator (SG) level control problems. Operator j response to the SG 1evel transients was considered excellent. The equipment in-adequacies which are producing ongoing problems with SG 1evel control are being pursued by the licensee and the system vendor. However, as is shown by the recur-rence of SG level oscillations, the problem persist The number of illuminated annunciators in the control room caused the inspector to question whether sufficient priority was being placed on correcting the condi-tions causing these annunciators. Future inspections will address this matter furthe PDR 0 ADOCK 05000423 ppg

_ - _ _ _ _ _ _ _ _ _

. _ _ _ __ .-

.

.

TABLE OF CONTENTS Page Summary of Facility Activities................................... ... 1 Review of Activities Occurring During the Inspection Period.......... 2 Non-Conservative Reactor Trip System Instrument Setpoint........ 2 Power Operated Relief Valve Repair.............................. 4 Reactor Coolant System Identified Leakage Rate High............. 5 Analysis of Failed Snubbers..................................... 6 Feedwater Leakage Within Containment............................ 6 Control Building Isolation on Chlorine Monitor Actuation. . . . . . . . 7

, Radiation Monitors.............................................. 8 Removal of Spent Resin.......................................... 8 Feedwater Transients............................................ 9

,

' Marathon Terminal Blocks........................................ 10 Control Board Annunciator Status;............................... 10 Licensee Actions on Previous Inspection Findings. . . . . . . . . . . . . . . . . . . . . 10 Breakdowns in Contamination Control Problems.................... 10 Undervoltage Trip Attachments of Westinghouse 08-50 Type Reactor Trip Breakers................................................... 11 Motor-Operated Valve Common Mode Failure due to Improper Switch Settings........................................................ 11 Allegation RI-86-A-94, Non-HP-Trained Fire Watches (unsubstantiated). 12 Review of Licensee Event Reports (LERs).............................. 12 On Site Safety Committee Meetings.................................... 15

, Management Meetings.................................................. 16 Attachment - Main Board Annunciators Illuminated As Of 9:00 a.m. September 25, 1986.

I I

i i

!

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

.

.

DETAILS Summary of Facility Activities This report period began with the plant recovering from a 24 day maintenance outage. Several problems occurred during that effort. Power operated relief valves (PORVs) which had been repaired during the outage failed stroke time tests and required modifications. There were difficulties in opening N Main Steam Isolation Valve when solenoids in the opening circuit stuck shu This was not a safety concern because valve closure was unaffected by this problem. The reactor was made critical at 0659 on August 16. At 2355, after synchronizing to the grid and loading the generator to 65 MWe, the "C" Feed-water Regulating By-Pass valve was placed in automatic and went shut. Control of the valve was returned to manual, but not in time to prevent a low steam

'

generator water level and an accompanying reactor trip at 0001 on August 1 The reactor was again critical at 0533 that day. However, while at 20% power at 1546, a level oscillation in the No.3 Steam Generator led to a Feedwater Isolation on high level followed by a reactor trip on low leve Criticality was again achie.ved at 1946 on August 1 On August 16, Steam Generator Safety Valve simmer tests were performed to verify lift setpoints after blowdown ring adjustments were made during the outage. Thirteen of the 20 Safety Valves had low setpoints; all were rese A hurricane threat was adequately anticipated and severe weather preparations were carried out on August 18. The storm track shifted east considerably south of the site. Some high winds of short duration were experienced but no significant problems occurre To help determine the cause of the Steam Generator Water Level Control System oscillations experienced since initial power ascension testing, an in-service test was performed on August 21. With the plant at 69% power, a step decrease of 60 MWe (5% of rated power) was initiated using the turbine load limit con-trolle Feedwater regulating valve position was recorded through use of linear variable differential transformers. Based on the data, the Nuclear Steam System Supply vendor and the licensee hypothesized that there may be a design problem with the instrumentation condensing pots. These pots estab-lish the common reference legs for both steam flow and steam generator level differential pressure transmitters for each steam generator. Further instru-mentation and testing is being planned to gather additional dat The plant reached full power on August 21 and remained at that level, except for minor changes for surveillance and preventive maintenance, until September 5, 198 On August 27, the license completed the initial resin transfer from a letdown ion exchanger to the spent resin hold tank without inciden i l

.

.

An Unusual Event was declared at 2310 on September 4 when Reactor Coolant System Identified Leakage Rate reached 14 gpm, exceeding the Technical Speci-fication limit of 10 gpm. Reactor power was reduced to 45% to allow entry into the loop areas within the containmen Stem packing leakage from a Residual Heat Removal isolation valve was located and correcte Leakage returned to an acceptable rate, and the Unusual Event was concluded at 0822 on September The reactor was returned to full power on September 5. It remained there until September 15, when power was reduced to 90% to start the motor driven feedwater pump. This was in response to a seal failure in a turbine driven feedwater pump. During the evolution, control of the "C" Feedwater Regulating Valve (FRV) was lost. The valve failed full open. Steam generator feedwater control was accomplished by manually throttling the FRV isolation valv Power was then slowly reduced to 20% to allow operation on the bypass valv During that power reduction, close coordination was maintained between the Control Room and the plant operators who were manually controlling the bypass valve. After repair of the FRV, power was raised and held at 50% while re-pairing a turbine driven feed pump seal and a minimum flow valve associated with the motor driven feed pump. This valve had failed at about the same time as the "C" FRV. The reactor was returned to full power on September 17 and remained at that level until the end of this report perio . Review of Activities Occurring During the Inspection period Non-Conservative Reactor Trip System Instrument Setpoint The licensee discovered two errors in the Reactor Protection System (RPS)

Trip Setpoint for Overtemperature Differential Temperature (OT-Delta T).

These errors resulted in that trip setpoint being less conservative than the setpoint allowed by the Technical Specification 2.2.1 Limiting Safety System Setting stated in Table 2.2-1 as Functional Unit 7.a. The errors within the RPS Solid State Protection System (SSPS) calibration procedure were found during a setpoint verification review. Both errors concerned incorrect parameters being entered into the SSPS for the calculation of OT-Delta T. That calculation applies a rate compensation to the measured Reactor Coolant System (RCS) loop differential temperature (Delta T) and compares it to a setpoint determined from the RCS loop Delta T at rated thermal power corrected by the Average RCS Temperature, Pressurizer Pressure and Axial Neutron Flu:. Difference. Calculation constants and time constants are included for rate correction The specific errors concerned the use of an incorrect constant for the setpoint calculation and penalty value used to reduce the setpoint in the event of an excessive Axial Flux Difference. Both of these errors resulted from the failure of the licensee to recognize changes from values published in Westinghouse Electric Corporation Setpoint and Startup documents and those stated in the Operating License Technical Specifications. Although these errors were non-conservative, both were minor and had no significant effect on the calculated 0T-Delta T setpoin e

.

.

On August 15, the licensee discovered that the base four loop constant value, K1, which is used in the OT-Delta T calculation was 1.10 vs. the required value of 1.08. That resulted in a 2 percent non-conservative error to the calculated trip valu The second error was discovered on September 5 and concerned the penalty factor which is applied to the trip setpoint when the Axial Flux Difference (Delta I) exceeds a value of +10 percen The Technical Specifications require that the trip setpoint be automatically reduced 2.0 percent of its value at rated thermal powe The protection channels were calibrated with a 1.16 percent penalty, 0.84 '

percent less than the correct on The licensee reviewed monthly surveillance data and determined that al-though the OT-Delta T trip setpoint was less conservative than the re-quired trip setpoint, it remained within the setpoint allowable valu In addition, the Axial Flux Difference (AFD) has not exceeded +10 percent and as a result, the penalty was not in effec The review of setpoint calculations was undertaken by the licensee after receiving an August 13, 1986 letter from Westinghouse about uncertainties in excess of those previously identified for certain pressure transmit-ters. (That problem was identified when Public Service of New Hampshire discovered an excessive change in transmitter accuracy as ambient tem-perature change Initial tests of Veritrak Level A transmitters were limited to 130 degrees Fahrenheit. Subsequent testing of the transmit-ters original calibration points at 280 and 320 degrees Fahrenheit de-monstrated significant errors.) In that letter, Westinghouse stated that, even with the increased instrument uncertainty, the current Final Safety Analyses Report conclusions in Chapter 15 remained valid and that a Sub-stantial Safety Hazard did not exist. After' receipt of this information, the licensee began a review of setpoints for OT-Delta T, high and low pressurizer pressure, and high T low steam generator level. The errors in calculated OT-Delta T trip setpoint discovered during that review are an entirely different problem from instrument uncertainty analysi The licensee established the instrument trip setpoints prior to the is-suance of the Operating License through the use of the Westinghouse Startup Manual and the Westinghouse " Precautions, limitations and Set-points (PLS) Document," which was amended through July 11, 1985. How-ever, the Operating License Technical Specifications used the Westing-house "Setpoint Methodology for Protection Systems," dated November 1985 as a source document for establishing the RPS Trip Setpoints and the Limiting Safety System Settings. The PLS was never updated and the cali-bration procedure setpoints remained unchanged. Although reviews were conducted to verify proper implementation of the Technical Specifications, those reviews did not identify these problem The licensee corrected the calibration procedures and re-calibrated the protective instruments after each of these errors was foun In addition, they committed to completing an independent review of all Reactor Trip and Emergency Safeguards Features Setpoints by November 15, 198 .

.

There was one other event similar to this in which a change to the Tech-nical Specifications was implemented in their original issue but went unnoticed. On March 15, 1986, the licensee discovered that Three Loop Protective Interlock P-8 reset at a thermal power higher than that al-lowed by Technical Specifications. Following that event, the licensee verified all protective interlock setpoints. Following notification of the Veritrak instrument problem, the setpoints of all associated instru-ments were reviewed. After discovery of the OT-Delta T K1 error, all calculated setpoints were reviewed. All setpoints were reviewed after the discovery of the second OT-Delta T proble The events are addressed in LERs 50-423/86-24-00 and 86-47-00. No un-acceptable conditions were identified by NRC inspection of this matte b. Power Operated Relief Valve (PORV) Repair Excessive leakage past the PORVs had resulted in block valve closure during operation to reduce Pressurizer Relief Tank pumping and cooling requirements. The licensee decided to repair both PORVs during the July 24 - August 13 outage. A valve vendor representative was called in to assist maintenance technicians in assessment of the cause of leakage and subsequent repai Both valves were removed from the relief system; work was performed in the refueling building. This was done principally for ALARA considera-tions but also assuaged scheduling conflict with the safety valve re-placement work (initially the critical path for the outage). Work was -

performed on one valve at a time to prevent mixing of parts. The valves were opened and inspected to determine the cause of leakage. Both were found to have scored seats. These were corrected by machining. The capscrews that fasten the seats to the valve body were found loose, re-sulting in poor contact of the grafoil gaskets between the seat and bod This was considered the most probable source of leakage. The vendor has been asked to reassess the design for a more positive method of securing the seats. However, the solenoid valves, another possible source of leakage, were not disassembled and inspected due to lack of spare part Subsequent to the repairs, stroke time testing showed the valves operat-ing much slower than the 2 seconds used in the analyses for Cold Over-pressure Protection System (COPPS) operability. Valve 455A initially operated in 11 seconds and valve 456 behaved similarly, initially closing in 6 second The cause of the high stroke time has not been positively determine It was thought that continued main seat and/or solenoid valve seat leakage could be reducing the differential pressure across the main valve, thereby increasing the stroke time. To reduce the stroke time, the NSSS vendor and the PORV vendor recommended removing the restricting orifices that were installed to slow stroke time for valves having water seals

. .-- _ -- -

.

.

on the inlet side to reduce the hydraulic effects in the tail piece Since Millstone 3 has no water seal, there was no reason to have the orifices installed. The licensee and the vendor analyzed the blowdown effect with a 0.1 second stroke time and found no overstress conditio The orifices were removed. Subsequent stroke times were 0.6 secon On August 19, PORV 455A was again blocked to isolate excessive leakag Thermocouples installed on the safety and relief valve tail pieces are being used through a remote data logger to monitor valve leakages. PORV 456 leakage has been increasing als The inspector continues to be concerned about PORV leakage for two reasons. The first is that blocking the PORVs removes their availability for preventing safety valve challenges. The second is that leakage, along with block valve leakage, heats up the tailpieces and pressurizer relief tank (PRT) so that displays on Main Board 4 that should be in reserve to positively indicate a safety valve lifting are already indi-cating that condition (see Attachment 1, items 48-51.) However, no NRC

! requirements are violated by the existing conditions, and other indica-tions of safety valve actuation (e.g., plant pressure and' pressurize level) are available to indicate such a condition, j c. Reactor Coolant System Identified Leakage Rate High

! Reactor Coolant System (RCS) leakage is required to be monitored and is l limited by Technical Specification 3.4.6.2 to insure system integrity.

l Values for identified and unidentified leakage are calculated periodic-l ally by mass balance routines in the plant process computer. Identified leakage rates were observed to have increased above the normal value of i 1 to 2 gpm late on September 4. At 2146, the identified leakage rate I

had increased to 6.53 gpm; unidentified leakage remained low at 0.77 gp Continued plant operations requires identified leakage of less than 10 i gp The 10 gpm Limiting Condition for Operation was exceeded and the plant entered Technical Specification Action Statement 3.4.6.2.b at 2259, September 4. Because this parameter also exceeded an Emergency Plan Emergency Action Level, an Unusual Event was declared at 2310. The Technical Specification Action Statement requires that the leakage rate l be reduced to within limits in four (4) hours or the reactor be in Hot i Standby within the next six (6) hours. Licensee investigation as to the l source of the leakage was conducted through a series of containment l entries. Identified leakage is collected from various sources such as l mechanical fittings and valve packing glands in a closed piping system.

l

'

Licensee personnel used flow indicators located on collection headers to locate the area with the high leakage rate. At 0129, September 4, the source was narrowed down to four (4) valves within the "A" RCS loop area. An entry into the RCS loop area was made after reducing reactor power to 45 percent, and located the "A" RCS loop hot leg supply to the Residual Heat Removal System Isolation Valve (3RHS*MV-87010) as the l source. This valve has a connection to the leakage collection system I

.

.

between its two sets of stem packing. The leakage was stopped at 0430 by shutting an isolation valve to the leakage collection system. This pressurized the second set of stem packin (Any leakage from that packing will be classified as unidentified leakage, which is limited by Specification 3.4.6.2.b to one (1) gpm.) Control room personnel then noted that leakage into the containment drain collection tanks was re-duced substantially. Calculations performed at 0628 resulted in finding Unidentified Leakage of 0.86 gpm and Identified Leakage of 1.68 gp Maintenance personnel entered the containment at 1135, September 5 and adjusted the valve stem packing of valve 3RHS*MV8701C. To safely perform the adjustment, the area between the packing was depressurized by allow-ing the leakage to flow to the collection system. The licensee was suc-cessful in placing the first set of packing back into servic Reactor power was then increased at 1401 on September d. Analysis of Failed Snubbers A snubber. specialist from the tontractor who performed snubber functional tests during the last outage was on site to disassemble and examine re-jected units. At a meeting on August 29, the specigjist presented his findings. From the 15 units inspected, he concluded that boric acid contamination was the predominant contributor to high drag forces. The one failed snubber, a PSA-1/4, was locked up due to significant internal boric acid. Mechanical deficiencies accounted for high drag forces in 3 units: one PSA-1/4 had bent tension rods, one PSA-3 had a bent tang on the capstan spring (possible due to a high dynamic loading), and one PSA-3 had a nicked ball scre The licensee has yet to make a final statement on the number of failure The inspector understands that there were 3 failures, 2 PSA-1/4s and 1 PSA-3. Technical Specification Surveillance Requirement 4.7.10.b re-quires that 2 failures of the same design and manufacture (PSA-1/4) re-sults in a reduced subsequent visual inspection period of 6 months (+/-25%) for all similar snubbers. The failed PSA-3 unit mandates a 12 month +/-25% period for that type snubber. With present scheduling, the spring outage should incorporate the 6 month PSA-1/4 inspection within the tolerance band. The inspector will review the licensee's conclusion on this matter when issue e. Feedwater Leakage Within the Containment

On September 23, the licensee noted an increase in containment floor drain sump inventory. This was calculated to be an increase of about five (5) gpm. There was no accompanying change in Reactor Coolant System leakage calculated by the plant process computer. In addition there was, based on expansion tank level, no leakage from closed cooling water sys-tems servicing components within the containment. The licensee confirmed that the leakage was from the steam generator feedwater system during

.

.

a containment entr The leak was from stem packing on the first isola-tion valve for a high point vent on the No.3 steam generator feedwater heade The leak was stopped on September 24 when maintenance technicians, in full anti-contamination clothing and bio pacs in the sub-atmospheric pressure containment, used special equipment to reach the leak 35 feet above the refueling floor. The leak was stopped by torquing the first isolation valve shut; this depressurized the valve stem packing. The sump pump out rate returned to normal. The licensee has elected to re-pack this and similarly inaccessible valves during the next outage, f. Control Building Isolation on Chlorine Monitor Actuation At approximately 11:00 a.m. on August 24, 1986, a Control Building Iso-lation (CBI) occurred due to a chlorine monitor actuation. The signal was determined to be spurious and was reset prior to control building pressurization. Approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> later, the chlorine detector was declared inoperable due to a defective probe. In accordance with Tech-nical Specifications, the Control Room Emergency Ventilation System was replaced in the filtered recirculation mode until the chlorine monitor was restored to operabilit Since tuel loading in December 1985, approximately 10 CBIs have occurred due to spurious actuations from the chlorine monitors. These actuations were not limited to either train. The licensee's corrective actions as discussed in LERs 86-31, 86-37, 86-39, and 86-40 include the following:

(1) bi-weekly preventive maintenance (PM) to add electrolyte to the probe to prevent dryout, (2) placing this PM on the Technical Specification surveillance schedule to ensure its implementation, (3) monthly rotation of an inservice detector with a cleaned and calibrated spare detector, and (4) weekly probe voltage measurements to check for degradation. Of these corrective actions, only the adding of electrolyte has been suc-cessful in controlling one of the multiple contributors to the monitors'

malfunctioning. Spurious actuations due to probe dryout have been cor-recte Based on recommendations from the detector vendor, a plant modification to relocate the detectors and to remount the probe to a vertical orientation will be implemented to increase detector reliabilit This plant modification is currently scheduled for completion by December 31, 1986. The licensee is considering whethar to pursue the elimination of the chlorine monitors on the basis that chlorine is no longer stored on sit In most instances of a spurious CBI, the signal was reset to prevent control building pressurization and the ventilation system was placed into filtered recirculatio There was an exception on August 24 when the signal was reset but the control building was not placed into the recirculation mode until the monitor was declared inoperable 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> late Technical Specification (TS) 3.3.2 requires that the Control Room Emer-gency Ventilation System be placed in the recirculation mode within 1

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

.

.

!.

.

hour of declaring one of the chlorine monitors inoperable. In this case, j the TS was literally met. The inspector noted, however, that current

! operating procedures require neither that monitor operability be verified

'

immediately following a spurious alarm nor that the ventilation system be placed in recirculation until monitor operability can be confirme l These procedures may need to be revised to address the consideratir;n that

! spurious alarms may indicate monitor inoperability. This concern will i be re examined during routine inspection, g. Radiation Monitors

,

In addition to the chlorine detector problems, the licensee is experi-l encing numerous spurious Control Building Isolations (CBIs) from the two j

'

radiation monitors. The Kaman monitors utilize beta-scintillation type

'

detectors located at the 66 feet elevation of the control buildin '

There are two additional monitors, not addressed by Technical Specifica- '

tions, with similar problem Since fuel loading in December 1985, these

, monitors have caused about 1-2 spurious alarms per day. The licensee I has implemented corrective actions such as grounding and shielding of I cables. An original contributor to the problem was the check source feature of the detector. This was corrected by increasing the delay time for reinstating the alarm following exposure of the probe to the check source, thus allowing the check source signal additional time to fad Recent troubleshooting indicates that there may be a problem with the l compatibility of the detector and the amplifie The problem appears to be generic. However, a software change that has been successful at

'

correcting this problem for other facilities is not in accordance with i the Millstone 3 plant specific control room habitability analysis. Lic-

-

'

ensee efforts are now directed toward procurement of a monitor for off-line troubleshooting. The inspector will follow this issue during l routine inspectio '

h. Removal of Spent Resin On August 27, the inspector observed the initial transfer of spent resin from lithiated ion exchanger CHS-1A to the spent resin hold tank. The i

initial briefing included a detailed operations department explanation of the procedure and a Health Physics (HP) department description of

. protective measures. Health Physics took adequate measures to maintain ;

exposure ALARA. A hot spot of 800 R/hr had been measured on contact with

,

the ion exchanger vessel. The unit had been isolated for 2 days prior j

'

to transfer to allow short lived contributors to decay off. Ten HP technicians were posted to guard exclusion areas and monitor general area

, dose rates during the transfe '

i Operations personnel acted conservatively during the line-up and actual

transfer. At one point all work was halted to resolve a minor typographi-

cal error in the procedure. The radwaste systems startup engineer was 1 present to assist in problem resolution as necessary. Total man-rem ac-cumulation was 70 mre The inspector had no questions.

1

!

f (i

.

l l

'

9 Feedwater Transients A leaking mechanical seal on the "A" turbine driven feed pump required pump shutdown for repair. On September 15, during a transfer of feed pumps from the "A" Turbine Driven unit to the motor driven unit, "C" feedwater regulating valve failed to respond to remote manual control, resulting in a loss of level control on the No.3 steam generator. Plant Equipment Operators (PE0s) were dispatched to the valve and attempted to take local control but were not aware of the necessity to vent off the valve operator dome. The Shift Supervisor contacted the load dis-patcher and began reducing power. Increasing level was finally arrested by manually throttling the next upstream vaive, a motor-operated gate valv As levels began to stabilize, the condensate polishing facility high differential pressure annunciator illuminated, followed by all four feed flow / steam flow mismatch annunciators. The Senior Control Operator's attention was brought to the condensate header flow indicator, which was pegged high. Compiling this information led to his quick determination that FV20, the motor driven feed pump minimum flow valve, was failed open, directing excessive feed flow to the main condenser hot well. Operators were able to take local control and power flow through on FV20, raising steam generator levels to normal and preventing a reactor trip. Manual local control of the No. 3 steam generator with a large gate valve re-quired close coordination and anticipation of level changes. At one point, a single channel of steam generator high level (82%) feed water isolation bistable tripped, but the level rise had been anticipated and the gate valve had been ordered shut sufficiently early in the level transient to preclude a second bistable from tripping. A feedwater isolation (2 out of 4 coincidence) and a reactor trip were thus prevente At 50% power, the motor driven feed pump was tripped; power was further reduced to 20% and level control transferred to the feed regulating valve bypasses. The "C" feed regulating valve was shut and the valve positioner was replaced. Power was raised to 50% while the seals on "A" turbine driven feed pump and valve FV20 were being repaired. After seal re-placement, power was restored to 100% on September 1 During this sequence of events, the Shift Supervisor stepped back and maintained overall controi of the plant. Moreover, the assistant opera-tions supervisors handled the decision process on valve repairs with maintenance and I&C personnel, ensuring the Shift Supervisor was not distracted from plant operation. While two control ooerators were closely involved with the feed station control manipulations, the Senior Control Operator monitored the remainder of the secondary plant. His performance, along with the abundance of annunciated plant parameters, led to rapid diagnosis of the failed open motor driven feed pump re-circulating valv .. . . -. . . . - . - -- - - ._- . - -

.

l

'

,

l i 10

!

(

i l The inspector noted the operators' and supervisors' competent performance l during this complex event. His only concern was operator uncertainty during the attempt to take manual control of the feed regulating valve.

-

Subsequently, a training note to the operations staff and a revision to i

the feed system operating procedure were being formulated to promulgate j the manual control method. The inspector will review this change during j routine inspection, t Marathon Terminal Blocks Following problems at another power plant with environmental qualifica-tion of Marathon 6000 series terminal blocks on Limitorque 480V valve operators, the inspector determined that a similar problem did not exist at Millstone Unit 3. As a result of industry concerns over Marathon environmental qualifications, a licensee review was previously performed, ;

documented by letter GEE-85-1435 (Nicosia to Orefice), and concluded that no Marathon Terminal blocks exist in areas requiring environmental quali- ;

fication equipment. Further, according to supervisory Production Test and Electr,1 cal personnel, there are no power lead terminal blocks on valve operators in containment, because Raychem splices were used ex-clusively. The inspector had no further questions in this are l

. Control Board Annunciator Status On September 25 at 9:00 a.m., the inspector reviewed the control board I annunciator status. The listing of the illuminatad annunciators is at- ;

'

tached to this repor The licensee's commitment, documented in Mr. Opeka's letter to Region I dated June 19, 1986, is to achieve a total r

" blackboard." The inspector questioned whether sufficient priority was being placed on important annunciators (e.g., alarms 51 through 55 in the attached list). This concern will continue to be reviewed under the ,

routine inspection program, t 3. Licensee Actions on Previous Inspection Findings (Closed) IE_Information Notice 80-22 Breakdowns in Contamination Control Programs The subject notice addresses three concerns: 1) disposal of supposedly '

clean trash in nearby county sanitary landfills; 2) licensee sale of scrap material to salvage dealers in the area; and 3) soil contaminated on site extending beyond the protected area fence. The inspector dis-cussed these issues with the licensee and reviewed related materia In response to an INPO finding closely related to Concern 1, all trash >

released from the controlled area since April 1986 has been and will continue to be monitored before released for disposal. To date, no radioactive material has been discovere ,

'

t

__

t l

.

!

SHP 4917, Unconditional Radiological Release of Material Off-Site, clearly specifies the criteria to be met before any item is given an unconditional radiological release (Concern 2). It includes equipment, parts, and supplies which have been in radiation or contamination areas or which have been installed in potentially contaminated system Concern 3 is resolved by the weekly soil surveys taken around all three units in accordance with OP 915/2915/3915 for Units 1/2/3, respectivel For the Unit 3 records reviewed, at least 40 soil surveys were taken and recorded on a plot plan each week. The survey points are random to pro-mote detection of radioactive material at any site locatio The inspector concluded that licensee's radiological controls should prevent any breakdown allowing release of radioactivity from the site via the paths mentioned in the Information Notic b. (Closed) IE Bulletin 85-02, Undervoltage Trip Attachments of Westinghouse 08-50 Type Reactor Trip Breakers Millstone' Unit 3 does not use the subject Westinghouse 08-50 Breaker DS-416 breakers are used for the reactor trip function and IE Bulletin 85-02 therefore is not applicabl T(0 pen) IE Dulletin o improper Switch 85-03, Settings Motor-Operated Valve Common Mode Failure Due By letter dated June 11, 1986, Northeast Utilities provided the response for their four plant The response describes implementation of a pro-l gram to ensure that switch settings on the specified safety-related motor-operated valves are selected, set and maintained correctly to accommodate the maximum differential pressures expected during normal and abnormal events within the design basis. Scheduled completion dates are provide The inspector found the licensee's response timely, in that the program i for implementation was submitted within the extension period requested on May 14, 1986. Discussions were held with corporate and site personr.el I

regarding their selection of motor-operated valves covered by the bul-letin, the three vendors being considered to supply test equipment, and the order of testing (Millstone-1 in May 1987 with Haddam Neck, Mill-stone-3, and M111 stone-2 following). The amount of testing to be per-

, formed at Millstone-3 will be determined following review of the vendor l testing program. No problems with the licensee's program were identified, l However, the NU response to the concerns expressed in Bulletin 85-03 will be reviewed at a later dat _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ . . _ - - _ _ _ - - _ _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ - _

.

.

4. Allegations Allegation RI-86-A-94, Non-HP-Trained Fire Watches An anonymous individual alleged that a licensee contractor was using non-HP-trained individuals as firewatches in the containment (a radiation area).

He named 3 individuals who allegedly fell into this category. The inspector reviewed contractor personnel information, Radiation Work Permits (RWPs), and containment entry records for the 4 days immediately preceding the date of the allegatio He found that the named individuals did work for the con-tractor and were not HP qualified. However, they were not listed on any RWPs for containment work and computer records for those 4 days showed they had not entered containment. The inspector interviewed HP personnel and observed the HP access control and security access control at the containment personnel air lock. It was concluded that personnel do check through both before entr In summary, no substantiation of this allegation has been foun . Review of Licensee Event Reports (LERs)

LERs submitted during this report period were reviewed. The inspector as-sessed LER accuracy, whether further information was required, if there were generic implications, adequacy of corrective actions, and compliance with the reporting requirements of 10 CFR 50.73 and Administrative Control Procedurc ACP-QA-10.09. Selected corrective actions were checked for thoroughness and implementation, as documented elsewhere in this repor Those LERs reviewed were:

86-038-00 Main Steam Valve Building Supplementary Leak Collection and Release System pressure boundary found with an opening to the outside atmospher At 0905, June 5, 1986, an operator discovered an unsealed cable penetration in a wall which is a ventilation system boundary. That opening was the result of work performed during November 198 The opening has been sealed and ad-ditional administrative controls have been implemented to control boundary wor Failure of the Control Room Ventilation Chlorine Detecto At 2045, June 25 and 1634, June 29, Control Building ventilation isolations occurred because of an accumulation of dirt on the detection probe. Probe operating voltage is monitored once a week and the probes are replaced monthly in an effort to correct the proble There is no longer any liquid Chlorine stored on site. The licensee has modified the ventilation system to improve system reliability.

.

86-040-00 Failure of the Control Room Ventilation Chlorine Detecto Control Building ventilation isolations occurred at 2155, July 21 and 1352, July 24 due to invalid Chlorine Detector actuatio The licensee has made modifications to improve reliabilit Reactor Trip Due to Low Steam Generator Leve A reactor trip occurred at 2302, July 24, 1986 from 20 percent power due to low level in the No. 4 steam generator. At the time, plant operators were transferring control of steam generator feedwater from the main regulating valves to the bypass valves. A level increase within the No. 3 steam genera-tor resulted in a feedwater isolation and turbine tri No. 4 steam generator level fell to the low level reactor trip poin An investigation subsequent to the trip discovered a problem with the four feedwater regulating bypass valve positioners. With a demand signal to close, the valves were found from 10 percent to 40 percent open. To correct this problem, both the valve positioners and position limit switches were adjuste Engineered Safety Features, Emergency Core Cooling System Safety injection At 1147, July 25, a train "A" ::afety injection occurred which injected water into the Reactor Coolant System for 1.5 minutes. The pressurizer low pressure safety injection signal had been blocked, but the block reset for no apparent reason. The licensee believes that a high resistance switch contact along with vibration resulting from the "A" Emergency Diesel Generator output breaker caused the circuit to reset. The licensee inspected all switches in both safeguards division transfer switch panels, and found no similar problem Incorrect Settings of Main Steam Safety Valve Blowdown Ring The licensee discovered that the upper and lower blowdown adjusting rings were not properly set by the vendor af ter functional testing. These rings control the amount of pressure blow-down for the Main Steam Safety Valves. Although the rings have been reset to the recommended settings, their as-found condi-tion cannat be evaluated. The amount of pressure blow down might not have been conservative with that assumed in the steam generator tube rupture analysis, if the safety valves had been required to actuate before correct setpoints were establishe Failure to Perform a Double Value Lineup Verification Prior to a Liquid Waste Discharge Which Bypassed the Effluent Radiation Monito A liquid radioactive waste discharge was initiated at 0025, July 31 with the process effluent radiation monitor out of service. The plant staff failed to make a double valve alignment verification in accordance with Technical Specification 3.3.3.9. The discharge was terminated when this was discovered, and the valve alignment was then verified to be correct. The double check

.

.

is intended to prevent a discharge from the wrong tank or from more than one tank at a time when the radiation monitor is not available. The licensee's corrective action was in the form of additional trainin Excessive Leakage through Containment Isolation Valves The sum of leakage through containment penetration isolation valves was found to exceed the Technical Specification limit on July 31. Local leak rate testing, conducted during an outage, found excessive leakage through Chilled Water System Containment Isolation valves (3CDS*CTV91A, and 91B). The licen-see suspects that excessive wear occurred during the preoperational test flush progra The "T" ring valve seats were replaced; this restored acceptable leak tight-ness to the valves. The licensee is committed to re-testing these valves dur%g the first refueling outag Trip of the "B" Emergency Diesel Generato The "B" Emergency Diesel Generator (EDG) tripped during surveillance testing on August 1, 1986. The EDG was started at 2318, was synchronized to the "B" 4160v safeguards electrical bus at 2327, and tripped at 2331 as load was being increased. At the time of the trip, load was approximately 80 percent of full capacity or 4000 KW, The cause of this trip has not been determined. A plant operator was monitoring the diesel generator gauge panel at the time of the tri There were no abnormal conditions observed. The licensee investigated this occurrence and verified all support control circuits were operating pro-perly. The EDG was returned to standby service at 1500 August 3 following additional and successful testin Overtemperature Differential Temperature Trip Setpoint Incorrectly Established above the Technical Specification Allowable Trip Setpoin The licensee discovered that two errors had been made in establishing the procedures for calibration of the Reactor Protection System (RPS) Overtempera-ture Differential Temperature Trip Setpoints. In the first error found on August 15, an incorrect value for constant K1 had been used. Selecting 1.10 vs. the required value of 1.08 resulted in a 2 percent nonconservative erro The second error resulted in selecting an incorrect penalty value for exces-sive axial flux difference. Where the Technical Specifications required a

.

penalty of 2 percent penalty per percent of Axial flux difference above +10 percent, the calibration procedure setting was 1.16 percent. These errors have been corrected. The licensee is conducting a third party audit of all Technical Specification RPS and Engineered Safety Features Setpoint Reactor Trip Due to Low Steam Generator Leve A Reactor Trip occurred at 0001 on August 17 from 11 percent reactor power due to Low Level No. 3 Steam Generator. The level transient occurred when shifting the feedwater Regulating Valves from manual into automatic contro .

.

No. 3 Steam Generator level was slightly above programmed level at that time; the other three Steam Generators were below programmed level. No. 3 Steam Generator level dropped rapidly while correcting for the other Steam Generator level The licensee readjusted the No. 3 Steam Generator level controller making it more responsive and with a gain similar to the other controller Feed Water Isolation and Reactor Trip due to Steam Generator Water Level Transien A Feed Water Isolation (FWI) occurred at 1545 on August 17 due to High level No. 3 Steam Generator while at 71 percent reactor power. Steam Generator Feed Water was being supplied through the main Feedwater Regulating Valves, which were in manual level control mode. The one turbine driven feedwater pump was also in manual speed control. Inadequate coordination between operators con-I trolling steam generator level and pump speed resulted in level oscillations and overfeeding No. 3 Steam Generator. The lack of feedwater following the FWI caused a low level reactor trip. The licensee's corrective actions in-cluded changes to the feedwater pump speed controller response time as well

,

as operator trainin A Portion of the Main Steam Valve Building exceeded the Specified Temperature Limi The top floor (elevation 71' 2") of the Main Steam Valve Building (MSVB) ex-ceeded the 120 degrees Fahrenheit (F) temperature limit at 0139 on September 2 for greater than eight hours. Since that time, the temperature in that area has varied between 119 and 125 degrees F. The licensee has initiated a plant modification to the MSVB ventilation system. In addition, the licensee has analyzed the effect on continued operability of components with sustained temperatures of 130 degrees F for six (6) months. The inspector will continue to monitor this issue under the normal inspection progra Reactor Trip due to Low Steam Generator Leve A reactor trip occurred at 0928, September 6, due to low level in the No. 4 Steam Generator because of a failure of the feedwater isolation valve, 3FWS*CTV41 The reactor had been operating at 80 percent power. An inves-tigation failed to determine the cause of the inadvertent valve closur There had been a valve stem packing leak which resulted in steam impinging on the operating package. This may have resulted in a transient electrical fault which shut the valv . On Site Safety Committee Meetings The inspector attended Plant Operations Review Committee (PORC) meetings on August 14, September 10 and 11 and also a Site Operations Review Committee (50RC) meeting on October 1. Technical Specification requirements for attend-ance were met. The meetings were characterized by frank discussions and

F

.

! -

l

,

questioning of causes and corrective actions. In particular, attention was

'

given to confirming plant conditions related to several Technical Specifica-tions LCOs and Action Statements. Individual members' opinions were encour-age No deficiencies in Committee performance was observe . Management Meetinas i During this inspection, periodic meeting were held with senior plant manage-

! ment to discuss the scope and findings. No proprietary information was iden-tified as being in the inspection coverage. No written material was provided to the licensee by the inspector.

l l

i c--__-_______________________ .

.

ATTACHMENT 1 MAIN BOARD ANNUNCIATORS ILLUMINATED AS OF 9:00 /..M. SEPTEMBER 25, 1986 Asterisked items (*) have been addressed under plant modification requests (PMRs)

3-86-281 through 3-86-304, which have generated 13 Plant Design Change Requests (PDCRs) that are in various stages of work. The intent of these PDCRs is to reduce the number of unnecessary alarms on e main boards. Additional actions to reduce the number of alarms are expecte ALARM REASON Fuel Pool Level Low Fuel Pool is draine . Fuel Pool Cooling System Trouble Fuel Pool is draine Control Instrument Air Compressor (CIAC) Control Instrument Air Compressor ) Annunciator remains lit when (CIAC) A Cooling Water Temperature ) CIAC breakers are ope CIAC High* ) A & B are not in servic ) CIAC B Cooling Water Temperature High*)

) CIAC A Lube Oil Pressure Low * )

) CIAC B Lube Oil Pressure Low * )

) CIAC A Discharge Temperature High* )

) CIAC B Discharge Temperature High* ) Chlorine System Trouble ) Chlorine system remove )

10. Chlorinator Room Chlorine High* )

11. Primary Drains Transfer Tank Low Actual alarm, 250 gallon low level alarm after pumpin . Control Instrunent Air Pressure Low * Actual alarm condition (a check valve separates the sensor from the header).

13. Post-Accident Sampling System Panel Door microswitch not functioning Door Open* properl . Miscellaneous Aux Ckt. Trouble 8 train MSIV test rack spurious trouble light on aux ckt panel (repeater).

. c

.

Attachment 1 2 ALARM REASON 15. Hydrogen Reserve Bank Pressure Low Actual alarm, reserve tank is isolated to minimize the pos-sibility of a hydrogen leak, the sensor is downstream of the manual isolation valv . Loose Parts Monitor * Actual alarm set off by control rod movements-threshold level for upper head transducer is sensitive to noise generated by rod control syste . Cable Spreading CO2 System Locked Out* Actual conditio Fire watches statione . Yard South Heat Trace Trouble Heat tracing secure . Yard N-E Heat Trace Trouble Heat tracing secure . Aux Bldg Heat Trace Trouble Heat tracing secure . Containment Recirc (RSS) Cooler Seawater RSS not operatin Flow Low 22. Chiller A Evaporator Outlet Temp High* Unit not operatin . Chiller A Chilled Water Flow Low Unit not operatin . Refueling Water Storage Tank Level High Actual high leve l 25. Hydrogen Recombiner Train A Trouble * Unit not operatin . Hydrogen Recombiner Train B Trouble * Unit not operatin . Hydrogen Analyzer Train A Trouble Unit not operatin . Hydrogen Analyzer Train B Trouble Unit not operatin ,

29. Containment Recirculating System (RSS)) Calibrated at atmospheric pres-A Seal Tank Level High ) sure, containment vacuum causes

) dp change, giving alar '

30. RSS C Seal Tank Level High )

31. Chiller B Condenser Seawater Flow Low Unit in standb . RPCCW Heat Exchanger Outlet Temp Low ) Low temperature and high flow

) due to Volume Control Tank In-33. RPCCW Header Supply Flow High ) Service Tes !

.

.

.

Attachment 1 3 ALARM REASON 34. Sequencer A Trouble Auto test sequence stops at step F19, giving trouble ligh . Sequencer A Door Open* Microswitch problem, door is close . Radiation Alert ) Radiation monitoring system

) alarms recurring frequently 37. RMS Trouble ) (See report detail 2g).

)

38. Radiation High )

39. Boron Recovery System Trouble System is shutdow . Charging Pump Auxiliary Oil Off* Pump of . Boron Addition Tank A Temp High* ) Alarms at 100 degrees, actual

) temperature was 108 degree . Boron Addition Tank B Temp High* ) There is no method to cool down

) down the tan . Boron Addition Tank A Level Low ) Low alarm 93.7%, Low Low alarm

) 92.7%. Actual alarm condition 44. Boron Addition Tank A Level Low Low ) but setpoint shoud be readjusted

) because overflow lines prevent 45. Boron Addition Tank 8 Level Low ) level from getting high enough

) to clear alarm . Boron Addition Tank B Level Low Low )

47. Pressurizer Relief Tank Temp ) All present due to PORV or safety High (118 Degrees) ) valve leakage. These, coupled

) with the constant open position 48. Pressurizer Relief Tank Pressure ) of flow switches (FS) 48 A High (29 psia)* ) through C negate the intent of

) NUREG-0737 Item II.D.C in this 49. Pressurizer Relief Isolation Valve ) are Trouble Train A )

)

50. Pressurizer Safety Valve Discharge )

Temperature High )

)

51. Pressurizer Safety Valve Discharge )

Flow * )

52. Saturation Trouble Train A* ) Alarm setpoint proble )

53. Saturation Trouble Train B* )

.

.

Attachment 1 4 ALARM REASON 54. Core Exit Temperature High Train A* ) Computer point has not been

) validated, SPDS shows normal 55. Core Exit Temperature High Train B* ) exit temperatur . Source Range Loss of Detector ) Expected alarms in power rang Voltage * )

)

57. Source Range Shutdown Flux High )

Blocked * )

58. Process & Protection and Central Equipment failure. Trouble Power Supply Failure report issue . Any B0P Instrument Rack Open* Actual conditio PERMISSIVES 60. Source Range Reactor Trip Blocked ) Reactor at power, trip is Train A ) blocke )

61. Source Range Reactor Trip Blocked )

Train B )

62. Source Range Trip Block Permissive Intermediate Range >10E 10m both channel . Intermediate Range Reactor Trip ) Reactor at power, trip is Blocked Train A ) blocke )

64. Intermediate Range Reactor Trip )

Blocked Train B )

65. Power Range Reactor Trip Blocked ) Reactor at power, low flux

) trip blocke . Power Range Reactor Trip Blocked )

67. Reactor at Power P-10 Reactor is >10% power on all channel . Condenser Available for Steam Dump C-9 Vacuum greater than 5 inches absolut .-73. Loop 1-4 Cold Leg d T Interlock Cold leg stop valve interlocks are me .-78. Loop 1-4 Hot Leg delta T Interlock Hot leg stop valve interlocks are me . - ---,- - - . . -- . _ - - . . - _ , - . _ . . . _ - . _ _ -

- - . . . _ . -

_

- , _

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

.

Attachment 1 5 ALARM REASON 79. Turbine Driven Feedwater Pump A Turbine Turbine is shut down, stop Trip valves are close . Aux. Feedwater Pump A lube oil pressure Pump not runnin low *

81. Aux. Feedwater Pump B lube oil pressure Pump not runnin l low * I 82. Demineralized Water Storage Tank Level Actual high level due to High* overfil . Hydrazine Feed Tank Level High ) Tanks are routinely filled to

) the high level alarm . Ammonium Hydroxide Tank A Level High )

85. Auxiliary Condensate Conductivity High* Actual alarm, a retransmitted point from gaseous or liquid waste or Boron Recovery System auxiliary condensate high conductivity 86. Condenser Pit Sump Level alarm level indicator proble Trouble report submitte . Amertap Cycle Failure ) System shut dow )

88. Amertap Emergency Backwash )

89. Turbine Plant Component Cooling Water Actual alarm, unit in fresh water l Heat Exchanger Service Water Outlet layu Logic assumes HX left l Pressure Low * with service water pressure on HX at all time . Dirty 011 Tank Level Low * Actual alar . Drain Pot Level High DIM-A0V21A ) Level switch problem ) Trouble report submitte . Drain Pot Level High DIM-A0V30A )

)

93. Drain Pot Level High DIM-A0V308 )

)

94. Drain Pot Level High DIM-A0V30C )

95. Control Building Chiller Condenser B Unit shutdow Service water flow low *

_ _ _ _ _ _

_ - -- _ _ _ _ _ _ _ _ _ _ _ _

_

.

Attachment l' 6 ALARM REASON

'l 96. Containment Air Recirculating Fan B Unit shutdow )

flow low *

97. Containment Air Recirculating Chiller Unit shutdow chilled water flow low *

l

..

..

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