IR 05000245/1993013
| ML20036C273 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 06/09/1993 |
| From: | Doerflein L NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| Shared Package | |
| ML20036C269 | List: |
| References | |
| 50-245-93-13, 50-336-93-09, 50-336-93-9, 50-423-93-10, NUDOCS 9306160010 | |
| Download: ML20036C273 (28) | |
Text
.
.
-
!
U.S. NUCLEAR REGULATORY COMMISSION
REGION I
Report / Docket Nos.: 50-245/93-13 50-336/93-09 50-423/93-10 License Nos.:
,
DPR-65 NPF-49
'
Licensee:
Northeast Nuclear Energy Company P. O. Box 270 Hartford, CT 06141-0270 Facility:
Millstone Nuclear Power Station, Units 1, 2, and 3
,
Inspection at:
Waterford, CT Dates:
April 4,1993 - May 18,1993 Inspectors:
P. D. Swetland, Senior Resident Inspector K. S. Kolaczyk, Resident Inspector, Unit 1 D. A. Dempsey, Resident Inspector, Unit 2
,
R. J. Arrighi, Resident Inspector, Unit 3
]
I
,
kh h-k I 93 Approved by:
Cowh
Lawrence T. Doerflein, Chief
~
Date Reactor Projects Section 4A, DRP'
Scope: NRC resident inspection of core activities in the areas of plant operations, radiological controls, maintenance, surveillance, security, outage activities, licensee self-assessment, and periodic reports.
The inspectors reviewed plant operations during periods of backshifts (evening shifts) and deep backshifts (weekends, holidays, and midnight shifts). Coverage was provided for 113 hours0.00131 days <br />0.0314 hours <br />1.868386e-4 weeks <br />4.29965e-5 months <br /> during evening backshifts and 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> during deep backshifts.
Resnlis: See Executive Summary
.
9306260010 930609 PDR ADOCK 05000245 O
.
.
EXECUTIVE SUMMARY Millstone Nuclear Power Station Combined Inspection 245/93-13; 336/93-09; 423/93-10 Plant Operations Unit 1 operated at full power for the majority of the report period. A 4-day shutdown occurred from April 26 - May 1 due to excessive leakage found on two containment isolation valves.
Unit 2 operated at full power for most of the report period. Several power reductions were taken to repair feedwater system components.
Unit 3 started up on April 7 after unsuccessful troubleshooting for the cause of a March 31
plant trip due to turbine electro-hydraulic control anomalies. A power supply problem was identified and corrected when the phenomena recurred during the turbine warm up. The plant reached 100 percent power on April 13 and continued at essentially fuD power for the remainder of the report period.
Maintenance / Surveillance l
Generally good maintenance and surveillance practices were observed on each unit.
At Unit 1, a thorough engineermg walkdown of the drywell identified a low oil level and incomplete attachment on a recirculation pump motor snubber which led to further inspections and corrective action prior to plant startup on May 1. Incomplete attachment of this snubber was due to poor maintenance practices installing clevis pin fasteners. The cause of the snubber oil leakage and tha -N,t of these problems on equipment operability remained unresolved. A weaknes
.1 identified in the Unit I surveillance program, in that, the master surveillance sched. 'e was st updated at the required frequency so that the correct
,
completion of some new sun.
quirements are not being verified.
"-
-.
,
Unit 2 experienced a momentary loss of feedwater control when a work order intended for a training demonstration was mistakenly processed. During the equipment tagging process, control of a feed regulating valve was lost. The consequences of this tag isolation were not recognized by operators because previous modifications had not been correctly translated into system isolation procedures. Unit 2 also erroneously reset degraded and under-voltage protection bistables as a result of using incompatible test equipment during system calibration.
'
This issue remained unresolved pending licensee determination of the root cause and corrective actions for the problem.
t
,
i ii i
_.
.
.
Inspector review of Unit 3 containment isolation valve (CIV) testing revealed an incorrect i
interpretation of technical specification (TS) requirements for CIVs. No actual violations of TS requirements had occurred. Unit 3 has also implemented an appropriate power supply refurbishment program to address aging of electrolytic capacitors.
Engineering and Technical Support
.
Unit 1 engineers found that disassembled seismic supports on main steam flow sensing lines jeopardized the design function of these pipes and allowed fretting wear of the pipes due to i
vibration. This observation indicated good initiative and questioning attitude on the part of the engineers. Generally good corrective actions were implemented, except that an evaluation i
was not initiated of fatigue induced cracking due to the vibration of these unrestrained sensing lines.
>
!
Unit I also verified the satisfactory performance of their modifications to prevent reactor vessel level indication problems due to entrained gases in the reference legs. Temporary
Instruction 2515/119 was closed.
t Radiological Controls The inspectors reviewed the qualifications and training of personnel operating whole body
>
counters at the site. These technicians were found to be adequately trained for the limited
scope of their assigned duties.
-
Safety Assessment / Quality Verification Licensee actions regarding four open inspection items from previous inspections were found
acceptable for closure of these items.
,
i
!
!
!
.
-
,
f
-.
.
..
i
<
l
-
l
.
SUMMARY OF FACILITY ACTIVITIES Millstone Unit 1 entered the repon period at 100 percent power. Minor power reductions were performed on a planned basis to test the turbine stop valves, turbine control valves, and main steam isolation valves. On April 26, Unit I declared an Unusual Event and commenced
a plant shutdown when isolation valves on a atmosphere control system containment penetration failed a local leak rate test. Cold shutdown was reached the following day. The unit remained offline for the rest of the week to facilitate repair and retesting of the penetration isolation valves, repacking of an isolation condenser valve and the repair of steam
leaks in the turbine building. On May 1, the reactor was taken critical and power operations resumed. On May 11, reactor power was reduced to approximately 65 percent when the exciter on the "B" recirculation pump motor generator set failed. After extensive
,
troubleshooting activities, the failure was traced to a failed varistor in the exciter. The defective components were subsequently replaced and the recirculation pump was restarted on
'
May 12. Full reactor power was reached later that day. Unit I remained at full power for the rest of the repon period.
I Millstone Unit 2 was operating at full power at the start of the inspection period. On April 3, extraction steam was removed from several low pressure feedwater heaters when high water level alarms were received, causing a minor (less than 2%) power excursion. On April 6, power was reduced to 50 percent to repair tube leaks subsequently discovered in the 4A low pressure feedwater heater. Following repairs on April 8, full power operations
,
resumed. The plant operated at full power for the remainder of the inspection period except for a brief power reduction to 98 percent on April 28 to resolve a level control problem on the 2A low pressure feedwater heater.
Millstone Unit 3 was in hot shutdown (Mode 4) at the start of the inspection period. The licensee started up the reactor on April 7, and brought the turbine on line on April 8, when the electro-hydraulic control system transient which caused the March 31, plant trip repeated i
itself. The reactor remained critical in a stable condition while the failed component was identified and replaced. Power ascension continued on April 9, while repairs to the 'A'
'
moisture separator drain pump recirculation valve was completed and the end of life moderator temperature coefficient surveillance test was performed. The plant attained 100 percent of rated power on April 13. On May 2 and May 14, power was reduced to 90 percent to thermal backwash the condenser bays. At the end of the inspection period, the plant was at 100 percent power.
2.0 PLANT OPERATIONS (IP 71707, 93702)
2.1 Operational Safety Verification (All Units)
The inspectors performed selective inspections of control room activities, operability of engineered safety features systems, plant equipment conditions, and problem identification
systems. These reviews included attendance at periodic plant meetings and plant tours.
.
The inspectors made frequent tours of the control room to verify sufficient staffing, operator procedural adherence, operator cognizance of equipment status and control room alarms, conformance with technical specifications, and maintenance of control room logs. The inspectors observed control room operators response to alarms and off-normal conditions.
The inspectors verified safety system operability through independent reviews of: system configuration, outstanding trouble reports and incident reports, and surveillance test results.
During system walkdowns, the inspectors made note of equipment condition, tagging, and the existence of installed jumpers, bypasses, and lifted leads.
The accessible portions of plant areas were toured on a regular basis. The inspectors observed plant housekeeping conditions, general equipment conditions, and fire prevention practices. The inspectors also verified proper posting of contaminated, airborne, and radiation areas with respect to boundary identification and locking requirements. Selected aspects of security plan implementation were observed including site access controls, integrity of security barriers, implementation of compensatory measures, and guard fome response to
,
'
alarms and degraded conditions.
The inspectors determined that these operational activities were adequately implemented.
Specific observations are discussed in Section 2.2 to 2.4 below.
2.2 Plant Shutdown because of Excessive Containment Isolation Valve Irakage
-
Unit 1 On April 26,1993, at 6:30 p.m., Unit I declared an Unusual Event and commenced a plant shutdown when the primary containment was declared inoperable because of excessive leakage through a containment penetration. The leakage was detected during a containment local leak rate test (LLRT) of penetration X-25/202D, which is located in the atmospheric control system and is used to vent the containment drywell and torus areas. When tested, the penetration leaked approximately 180 standard cubic feet per hour (SCFH) or 0.4Wt%/ day.
Unit 1 Technical Specification 4.7.A.3.e.(1), Primary Containment, limits the maximum leakage through an individual penetration to 18.8 SCFH. Additionally, specification 4.7. A.3.e.(1) limits the maximum leakage through all penetrations that are subject to LLRT to 300.3 SCFH or 0.6Wt%/ day. When the leakage through penetration X-25/202D was added to the existing containment leak rate of approximately 0.4Wt%/ day, both TS limits were exceeded. Therefore, a plant shutdown was commenced ia accordance with plant TS 3.7.A.3. which requires the plant to be placed in Cold Shutdown within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
The plant shutdown was orderly; however, when turbine load was reduced to approximately 5-15 megawatts, a turbine trip occurred because of a high water level in a moisture separator reheater. In response to the trip, operators opened the main condenser bypass valves to maintain a heat sink for the reactor, and the plant shutdown was continued. The Unusual Event was terminated on April 27, at 6:35 a.m., when Cold Shutdown was reached.
l
.
i
Prior to commencement of the shutdown, the inspector noted that the incident reponing sheet which is filled out by a Shift Supervisor Staff Assistant and used by the licensee to inform offsite agencies of reponable occurrences, had noted the Unusual Event as occurring at 6:30 p.m., yet a radiopager announcement of the event to state and local officials was not made until 6:48 p.m. NRC requirements in 10 CFR 50, Appendix E state, in part, that the licensee should be able to notify state and local officials of an emergency within 15 minutes.
The inspector discussed the apparent late notification with the operations manager who stated that since the shift supervisor had logged the declaration of the Unusual Event at 6:35 p.m.,
the notification was not late since it was within the 15 minute time requirement. Further, the operations manager stated that actual power reduction was not commenced until 6:35 p.m.
.
Therefore, the manager stated that the 6:30 p.m., entry contained on the reporting form was premature.
NRC requirements for timely notification of reportable events allow about 15 minutes to notify offsite officials from the time that the responsible licensee official classifies the event.
Barring an individual performance problem with correct and timely classification of the event, the notification of state and local officials must be within about 15 minutes of the time that the responsible licensee official has the information needed to classify the event. Therefore, j
the inspector concluded that the licensee's position was not correct that the timeliness of reporting the April 26, Unusual Event should be measured from the time a shutdown was initiated and the event was logged. On May 13, 1993, the licensee reemphasized to its staff that the initiating time for reponing is the time at which operators recognize that events have occurred which make declaration of an emergency class appropriate.
The inspector concluded that the licensee's corrective action was appropriate ar.d that the 18
'
minutes required to report the April 26, event was not a significant departure from the required performance.
2.3 Shutdown Cooling System Walkdown - Unit 1 On May 5,1993, the inspector performed a walkdown of the accessible portions of the shutdown cooling system while it was in a " standby condition." All valves in the system were positioned as listed on the shutdown cooling system piping and instrument diagram.
Imcked valves in the system were as identified on the operations department locked valve list.
j The plant drawings of the shutdown cooling system accurately displayed the system configuration. During the walkdown, three discrepancies were identified. Valve 1-SD-23A which is a system drain valve was not labeled. Another valve 1-SD-35, which connects the fuel pool cooling system to the shutdown cooling system was incorrectly labeled as 1-FWA-
'
V49. These deficiencies were turned over to the Unit 1 operations assistant for
~
dispositioning. Additionally, the radiation survey for the shutdown cooling system heat i
'
exchanger room that was posted in the reactor building was out of date. Specifically, the
,
s k
.
i survey that was posted in the reactor building was performed on April 27, yet a more recent survey which was posted outside of the health physics office, was dated May 4. Two other surveys of the refuel floor and incore flux detector room that were posted in the reactor building were also superseded by trnre recent surveys located outside of the health physics
'
office.
The inspector reviewed the Unit I radiation survey program with the health physics manager.
The inspector noted that although the radiation surveys were conducted within the licensee's
,
administrative guidelines, health physics technicians were not posting the most recent results, i
The inspector informed the health physics manager that timely posting of survey results is needed to enable personnel in the field to make informed assessments of radiological l
conditions. The surveys were subsequently updated in the reactor building. The inspector concluded that the out-of-date surveys were not safety significant in this instance since any changes in radiation levels which had occurred in the out-of-date areas during the interval period had been adequately posted with local warning notices. To ensure the reactor buildmg surveys are updated when required, the health physics manager reinstructed the Unit I health physics technicians to update the survey sheets in the reactor building when surveys are revised. The inspector considered the corrective action to be appropriate.
2.4 Reactor Start-up - Unit 3 On April 7,1993, following troubleshooting of the electro-hydraulic control (EHC) circuits the plant started-up. The objective of the troubleshooting effort had been to identify the cause of the Unit 3 plant trip experienced on March 31 (reference Inspection Report 50-423/93-07). The root cause of the event could not be determined and following review of the event by the plant operations review committee, permission was granted to restart the unit with provisions for continued diagnostic monitoring of the EHC cabinet. Bypass jumper (BJ)
3-93-043 was initiated and implemented to allow for continuous monitoring of various EHC
,
system parameters to detect unusual perturbations during any future plant transients.
After reactor start-up with the plant at approximately 16 percent power, the turbine was brought on line. Prior to the turbine reaching 1000 revolutions per minute the March 31 event repeated itself; the number 2 stop valve and one intercept valve went closed. The plant remained at 16 percent power in a stable condition with the fault locked in. Subsequent investigation revealed an AC noise signal to the servo valves for the number 2 stop valve and i
the control and intercept valves. This noise was localized to the +22 volt power supply of the permanent magnet generator (PMG). Earlier efforts to troubleshoot the problem had been unsuccess41 because the PMG power supply does not come on line until the turbine reaches approximately 75 percent of rated speed; until this time, the non-vital 120 VAC power supply provides the source of x)wer for the EHC cabinet.
I
<
\\
.
_
_
__
l
\\
.
'
f l
The licensee replaced the +22 volt power supply on April 9, and commenced increasing l
reactor power; full power operations resumed on April 13. The licensee attributed the failure l
of the of the +22 volt EHC power supply to the aging of an electrolytic capacitor.
Replacements for the remaining EHC power supplies were not available; however, they were tested satisfactorily. The bypass jumper remamed in place as a precaution for continued monitoring of the EHC system. The remaining EHC power supplies are scheduled to be changed out during refueling outage 4, scheduled to begin on July 31,1993.
'
The inspector observed the licensee's troubleshooting efforts, attended the post trip review meeting, and reviewed the BJ for instrumenting the EHC cabinet. Although the licensee's initial troubleshooting efforts were unable to identify the cause of the transient, the inspector determined the review to be thorough.
3.0 MAINTENANCE (IP 62703)
.
The inspectors observed and reviewed selected portions of preventive and corrective maintenance to verify adherence to regulations, administrative control procedures and appropriate maintenance procedures; adherence to codes and standards; proper QA/QC involvement; proper use of bypass jumpers and safety tags; adequate personnel protection; and, appropriate equipment alignment and retest. The inspectors reviewed portions of the following work activities:
M1-93-04107 Inspection and Cleaning of 480v Contacts
M1-93-00175 Repair of Instrument Line Deficiencies
M2-93-04519 Diesel generator fuel oil system preventive maintenance
M2-93-06322 Connect recorder to 'B' diesel generator governor
M2-93-06135 Preventive maintenance on 'A' diesel generator
M2-93-06476 Sample diesel generator governor oil
M3-93-08290 Install blank downstream of containment sample isolation valve CTU33
M3-93-08034 Reposition ABFS for summer mode of operation
M3-93-08100 Ultrasonic test inspection of charging pump cooler 3CCE*ClB piping
M3-93-17057 Cut out SSR*CTV33 and weld in new valve, containment sample isolation
valve Except as noted below, the inspectors determined maintenance activities observed were performed adequately. Details of the inspector's observations are provided in Section 3.1 -,
6 3.1 Incorrect Installation of Snubbers Identified - Unit 1 While conducting routine inspections of equipment located in the drywell, licensee personnel noted that a clevis pin on a hydraulic snubber for the "A" recirculation pump motor had become partially dislodged. Additionally, the oil level in the snubber had decreased below the recommended minimum value for operability. Accordingly, the snubber was replaced with a spare assembly.
When the degraded snubber was tested, no changes in lockup or bleed rates were identified.
Therefore, the low oil level had not yet affected snubber operability. Examination of other hydraulic snubbers located in the drywell revealed that they were properly filled with hydraulic oil. The snubber of concern was manufactured by Bergen-Paterson and was installed in the 1989 refuel outage. No leakage was observed from the snubber when it was inspected in 1991 and 1992, respectively. At the close of the report period, the licensee had not disassembled the degraded snubber. Therefore, the cause of the low oil level had not been definitively identified. The licensee currently replaces the snubber seals every 5-7 years. If the licensee determines that the snubber leakage was caused by a seal failure, the aforementioned replacement schedule will be reexamined.
The licensee postulated that the snubber clevis pin became dislodged when an improperly installed cotter pin located at the end of the clevis pin fell out. Examination of the cotter pin which had been installed in the snubber revealed that the ends of the pin were not sufficiently bent back to prevent free movement. A 100 percent inspection of other hydraulic snubbers located in the drywell revealed that other cotter pins were also inadequately installed. To ensure the clevis pins on other snubbers would remain in position the cotter pins on those snubbers were bent back properly. At the close of the report period the licensee had not determined if the partially inserted clevis pin had affected the operability of the snubber.
The inspector examined the condition of the hydraulic snubbers located in the drywell. The inspector also noted that on several snubbers the cotter pins were inserted into the clevis pin with no crimping. The inspector considered those installations to be an example of poor workmanship. To ensure the cotter pins are installed correctly in the future, the licensee will modify the snubber installation procedure to require the cotter pins to be fully bent once the clevis pin is installed into the snubber.
The inspector considered the initial corrective action taken by the licensee to ensure continued operability of the drywell snubbers to be appropriate. However, the inspector considered the incorrect installation of the snubber cotter pins to be a maintenance weakness. This item will remain unresolved pending the licensee's evaluation of the effect of the loose clevis pin on system operability and the cause of the low oil level; and completion of the corrective actions (50-245/93-13-01).
_
!
.
3.2 Partial Loss of Steam Generator Feedwater Due to Inadequate Safety Tagout -
Unit 2 On March 22, 1993, with the plant operating at full power, feedwater flow to the #1 steam generator was lost momentarily when feedwater regulating valve (FRV) 2-FW-51 A shut unexpectedly. Operators had been implementing a safety tagout in preparation for corrective maintenance on feed regulating bypass valve (BPV) 2-FW-41 A pursuant to an authorized automated work order (AWO). When the control power fuseblock specified by the tagout to isolate the BPV controller was removed, power also was lost to the FRV causing it to lock up (fail "as-is"). The fuseblock was reinstalled and immediately caused the FRV to shut.
Operators correctly responded to the main control board indications and alarms associated with loss of FRV control, and restored steam generator level to the normal control band within 30 seconds. The minimum steam generator level reached during the event was 67 percent, well above the automatic reactor trip setpoint of 38 percent. In order to determine the cause of the transient and to assess licensee corrective actions, the inspector reviewed system drawings, administrative procedures, and plant modification documents, and conducted interviews with licensee personnel.
At Millstone, work orders are produced by the computerized planned maintenance management system (PMMS). On March 11, the Unit 2 instrumentation and controls department PMMS planner generated AWO M2-92-03699 at his work station for a training film on the upgraded work control process being implemented at the station. The training work activity was to clean, inspect, and stroke test valve 2-FW-41 A. The planner scheduled the AWO for April 30 in order to assure that the work order would not be processed routinely prior to the filming and could be canceled thereafter. However, the AWO did not contain any indication that it was for training only. The training film was produced on March 17, during which time several copies of the AWO were printed in the I&C shop.
Personnel involved in the filming believed that all of the copies were destroyed as they were printed. However, one copy apparently was left at the work control station among other legitimate AWOs which had been routinely printed for processing by the first-line supervisor
on March 18. The PMMS planner canceled the BPV AWO from the computer on March 18.
On March 19, the normally scheduled PMMS planners' meeting was not held due to personnel schedule conflicts. Also, during the afternoon management meeting at which planned work is discussed, no one who could have identified the training nature of the AWO was present. The work order was forwarded to the operations department work control center. The inspector noted that neither the use of the PMMS for training purposes nor the scheduling of PMMS planning meetings were addressed by licensee procedures or practices.
The inspector concluded that the precautions taken by the I&C planner had been ineffective, and that the lack of a policy regarding annotation of the AWO as a training aid contributed to the event.
.
_
The AWO was processed by the operations work control center on March 22, at which time
,
'
the safety tagout originally proposed by the I&C supervisor was amended to electrically disable the valve. The operating shift accepted the work order and initiated action to tagout the BPV. Since the operators were aware of existing material discrepancies with the BPV, the work activity did not appear to be abnormal. Administrative Control Pmcedure (ACP)
QA-2.06A, " Station Tagging," requires that all equipment that receives power from fuses that
!
will be pulled must be reviewed, including independent review by electrical engineering if the impact cannot be easily determined. Using Operating Procedure (OP) 2388N, "Fuseblock Ground Isolation Electrical Distribution," as a guide, the operator identified fuseblock FB-DFD-C05F (DFD) from vital AC circuit 1008 as the power source for the BPV. The inspector noted that the procedure explicitly is intended for use in preparing tagouts, and that fuseblock DFD was associated only with the BPV. The procedure did not provide, nor was i
I it intended to provide, information regarding the consequences of pulling the fuse.
Additional tagout guidance was contained in OP-2388E, "120 Volt Vital Instrument AC: VA-10, 20, 30, and 40, Ground Isolation Electrical Distribution." That procedure stated that deenergizing circuit 1008 would result in loss of control power to the FRV, which would fail
"as-is." No information regarding the BPV was provided. The inspector concluded that additional review by the operator or engineering might have identified that fuseblock DFD would affect the FRV, but that the operator had consulted the appropriate procedure, and selected the proper fuseblock to deenergize the BPV. The inspector also noted that there
'
were no warning labels or tags located at the fuseblocks to warn operators of the consequences of pulling the fuseblocks, but that there was no requirement to do so. Finally, the inspector concluded that OP-2388N was technically inadequate in that fuseblock DFD is j
one of two fuseblocks which provide control power to the FRV and the BPV, and that it did i
not provide sufficient detail to permit operations personnel to evaluate fully the consequences of pulling the fuseblock.
Through a walkdown of the main control board and review of electrical drawings, the inspector determined that fuseblock DFD had been added to the feedwater control system during the last refueling outage under plant design change record (PDCR) 2-114-92. The i
modification repowered the control functions of the FRV and BPV from the non-vital to the vital 120 Volt distribution system and added an automatic isolation function for a main steam line break accident. In its investigation of the event, the licensee found that the engineer in charge of the PDCR had marked up the affected procedures and forms to reflect the circuit
'
changes, but that the operations procedure writer's group had misunderstocx! the information and had implemented the recommended changes inadequately. Final review by the plant operations review committee also failed to identify the inadequacies.
In response to the March 22 event, the licensee counseled the personnel involved in generating the AWO, and imposed management controls on the training use of the PMMS.
Procedure OP-2388N was changed to show the correct fuseblock listing for the valves and explaining system response to loss of power. In addition, Unit 2 engineering reviewed 23 electrical PDCRs (out of approximately 200 total PDCRs) which were implemented in the last two years to identify similar procedure problems. The licensee found that several PDCRs
_
_
.
i
.
.
had not identified OP-2388N as needing revision, but that the necessary changes had been made. The inspector verified the licensee's results through independent review of a sample of the design packages. Finally, the licensee issued a memorandum, dated April 12, 1993, requiring engineers to review and verify the adequacy of procedure changes associated with plant modifications for which they are responsible. The inspector verified that the new requirement was discussed in department meetings.
Procedure ACP-QA-3.10, " Plant Design Change Records," which implements the requirements of 10 CFR 50, Appendix B, Criterion III, " Design Control," and V,
" Instructions, Procedures, and Drawings," requires the licensee to determine that PDCR administrative items, including procedure updates, are completed. The inspector reviewed PDCR 2-114-92 and noted that neither OP-2388E nor OP-2388N had been identified as requiring change. In addition, the responsible engineer did not perform a review to verify that the modification had been incorporated into these procedures. Failure to translate adequately the feedwater control system design changes into procedures was a violation of NRC requirements. However, this violation was an additional example of a violation previously documented in NRC Inspection Report 50-336/92-36. This violation involved the inadequate control of PDCR administrative requirements during the last refueling outage.
The inspector considered the licensee's corrective actions, in conjunction with the actions taken in response to the previous violation, to be acceptable and had no further questions regarding this event.
3.3 Containment Sample Isolation Valve Maintenance - Unit 3 On May 7 and May 11, the inspector observed portions of maintenance activities M3-93-08290 and M3-93-17057. The maintenance objectives were to establish the conditions necessary for and the replacement of accumulator containment sample isolation valve SSR*CTV33. The activity was performed by maintenance personnel, and supported by operations, health physics, and engineering personnel.
The inspector attended the pre-job and containment entry briefings, reviewed the tagout isolation for accuracy, and witnessed portions of the maintenance activities. The inspector noted that the shift supervisor demonstrated a good questioning attitude and understanding of the job scope prior to releasing the work order. Engineering provided an outline of the maintenance activity plan including a recommendation for valve manipulation for draining the maintenance area and technical specification (TS) action statement entries. Health physics provided photographs of the drain valve and its location within containment. The maintenance personnel replacing the isolation valve also demonstrated a good questioning attitude regarding the adequacy of the isolation and whether the work area had been properly drained.
.
--
- - -
.
i i
.
The inspector concluded that the technical specification requirements had been satisfied, proper radiological controls were implemented, and the maintenance activity was very well coordinated and supported by the organizations involved. The work was completed properly and the system was restored and retested satisfactorily.
4.0 SURVEILLANCE (IP 61726)
The inspectors observed and reviewed selected portio.a of surveillance tests, and reviewed test data, to verify adherence to procedures and technical s xcification limiting conditions for operation; proper removal and restoration of equipment; a ad, appropriate review and resolution of test deficiencies. The inspector reviewed po:tions of the following tests:
EN31026 New Fuel Receipt and Inspection
SP36(MC.1 Borated Water Source and Flow Path Verification
SP-2601D Power Range Channel and Delta-T Power Channel Calibration
SP-2403B ESAS Undervoltage Bistables and Sequencer Calibration / Functional Test
SP-2613A Diesel Generator Operability Test, Facility 1
SP-2604C Low Pressure Coolant Injection Pump Operability Test, Facility 1
MP-270lJ-19 Operating Cycle Preventive Maintenance - Diesel Generator
Except as noted below, the inspectors determined that the surveillance activities observed were performed adequately. Details of the inspector's observations are provided in report Sections 4.1 - 4.5.
!
4.1 Surveillance Testing of Containment Penetrations - Unit 1 The inspector observed the performance of local leak rate test testing (LLRT) on containment penetration X-25/202D. This penetration is located in the atmospheric control system and is utilized to vent the drywell and torus areas. The penetration was tested as part of a corrective action program that the licensee had implemented to improve overall containment leak tightness through the accelerated testing and if necessary the subsequent repair of penetrations which had previously failed numerous local leak rate tests. The containment corrective action program which the licensee had implemented was endorsed by the NRC in a January 11, 1993, letter to the licensee which approved an exemption to the integrated containment leak test requirements for Unit 1.
The inspector reviewed the LLRT procedure, SP 93-1-3, entitled, " Penetration X-25/202D On line Leak Rate Test," and verified that it established appropriate isolation of the test area and provided adequate provisions to ensure the valves were tested in the as-found condition.
During the test, the inspector verified that the procedure was appropriately implemented and personnel were knowledgeable of the test goals and objectives, and the requirements contained in the test procedure.
I
.
When the penetration was tested, an excessive leak rate was detected. Subsequently, licensee personnel tried to improve the seating of the valves through adjustment of the valve disc but were unable to reduce the leakage significantly from 180 standard cubic feet per hour (SCFH). The test acceptance criterion was 18 SCF3. Consequently, the penetration was declared inoperable and the plant was shutdown. To determine the source of the leakage in the atmospheric control system, acoustic monitoring devices were placed downstream of the six individual containment isolation valves which are tested during this LLRT. Two valves, the inboard torus vent valve (AC-11) and the inboard torus bypass vent valve (AC-12), were suspected to be leaking and they were removed from the system. Examination of valve AC-11, a 20 inch butterfly disc valve revealed that rust buildup on the valve disc may have prevented the formation of a leak tight sealing surface against the valve seat. Examination of valve AC-12, a two inch plug valve, revealed that incomplete rotation of the valve stem prevented complete seating of the valve plug. Tight stem packing appeared to inhibit complete rotation of the valve disc. Prior to reinstallation of the two valves, the penetration was retested with blank flanges installed over the valve openings. During this test, the leakage through the remaining valves was determined to be approximately 2 SCFH. When the refurbished valves were installed and adjusted, total leakage through the penetration was determined to be approximately 5.6 SCFH.
The inspector noted that the testing which was conducted following the plant shutdown
'
proved that the LLRT failure of valves AC-11 and AC-12 did not adversely affect actual containment integrity. This was demonstrated by the fact that absent a failure of a redundant i
containment isolation valve in this penetration, the minimum path leakage through this penetration would have been approximately 2 SCFH. That leak rate when added to the existing "as left" containment leakage of approximately 0.4077 weight percent per day was
,
still less than the TS limit for overall containment leakage of 1.2 weight percent per day.
To ensure the 2 inch plug valve seats are returning to the closed position following valve j
operation, the licensee has marked the valve disc and body. Following valve operation i
during ASME Section XI testing, the licensee intends to check the position of the valve markings to ensure the valve seat returns to the fully closed position. During the next refuel outage, the licensee will replace the steel valve disc on AC-11 with a stainless steel disc.
Further, if the plant shuts down for an extended period of time (i.e., greater than two weeks), the licensee will consider retesting the penetration.
The inspector observed portions of the troubleshooting operations that were conducted for j
penetration X-25/205 and determined that they were well controlled. Overall, the inspector i
concluded that the surveillance testing and repairs on the atmospheric control system valves were conducted appropriately.
j
_
. _ _
.
4.2 Wide Range Stack Gas Monitor Surveillance - Unit 1 The inspector observed the performance of Suiveillance Procedure SP 406BB, " Stack Gas High Range Radiation Monitor Functional Check." The performance of this test is required by Technical Specification (TS) 4.8.D.7, "High Range Stack Noble Gas Monitor," and is performed monthly. Since the performance of the functional check rendered the radiation monitor inoperable, the inspector verified that the applicable TS limiting condition for operation was entered prior to performance of the test. The instrumentation and controls (l&C) technicians who performed the test were knowledgeable and followed the surveillance procedure.
The inspector reviewed the vendor manaal for the radiation monitor and discussed the performance of the procedure with an I&C technician. Through this review, the inspector determined that the surveillance procedure performed an adequate channel check of the monitor as required by plant TS. The inspector noted that this surveillance test was not contained in Procedure ACP-QA-9.02A, " Surveillance Master Test Control List." This is a list of surveillances that TS's require to be performed at Unit 1. The licensee uses this list for scheduling required testing and verifying that surveillances are not missed. Procedure ACP-QA-9.02A states that the surveillance list is to be updated annually. The surveillance of the high range radiation monitor became a TS requirement in January 1992. Therefore, the licensee's administrative controls to assure that this new requirement was correctly implemented would not have been effective, had the surveillance not been properly conducted through another mechanism. According to the licensee, Procedure ACP-QA-9.02 was in the process of being revised to include the surveillance of the high range radiation monitor and other surveillances that are now requirements because of changes to the TS.
The licensee was testing the high range radiation monitor in accordance with TS. Therefore, the inspector noted the incorrect implementation of administrative controls as a surveillance program weakness.
4.3 Monthly Surveillance Test of ESAS Undervoltage Bistabks - Unit 2 During performance of monthly Surveillance Procedure SP-2403B, "ESAS Undervoltage Bistable and Sequencer Calibration / Functional Test," on April 14,1993, all of the degraded voltage and loss of normal power bistable setpoints were found to be 3 to 4 Volts AC (Vac)
below the nominal procedure acceptance criteria of 106 Vac and 85.5 Vac, respectively.
Each bistabic was reset to the required value during its test. The calibration was conducted using a Fluke model 45 multimeter for the first time. After resetting all the setpoints on channel "A", the instrumentation and controls technician verified the accuracy of the multimeter before proceeding further. A plant incident report (PIR) was initiated at the completion of the surveillance to investigate the cause of the discrepancie __
_
_ __
!
,
Millstone 2 has two levels of loss of power protection associated with each of the two 4160 Vac safety-related buses. There are four channels per bus for each protection level. The technical specification setpoints are 2912 Vac or greater for complete loss of normal power (LNP), and 3700 Vac or greater for sustained degraded voltage from the reserve station service transformer (RSST). These values correspond to the bistable voltage setpoints listed above. An LNP or sustained degraded voltage condition on either emergency bus will initiate an automat c transfer from the RSST to its respective emergency diesel generator. A time i
delay and bistable reset function for the degraded voltage protection permits recovery from a short term (less than 8 seconds) voltage transient without transfer to the emergency diesel generators. Proper calibration of the bistable setpoints is required to assure that sufficient voltage is maintained on the buses to permit startup and operation of emergency electrical equipment while preventing unnecessary transfers due to minor offsite voltage perturbations.
On April 15, the inspector observed licensee troubleshooting activities concerning the bistable setpoints. Under automated work order M2-93-05452, the licensee reperformed SP-2403B l
using Fluke model 8050A and model 45 multimeters and found that in each case the meter readings differed by 3 to 4 Vac. The licensee determined that the discrepancy was caused by the difference in input impedance between the Fluke 45 (1 Megohm) and the Fluke 8050A (10 Megohms) which had been historically used for this surveillance. The licensee concluded that the bistables had been erroneously reset during the calibrations on April 14. Another PIR was initiated to document the finding. The bistables immediately were reset using the l
Fluke 8050A. Since the setpoints at all times had remained above the minimum values i
required by the technical specifications, the inspector concluded that the actual safety significance of the incident was low. However, the root cause might just have easily resulted in misadjustment of the bistables in the nonconservative direction. Also, the inspector was concerned that safety systems could have been challenged unnecessarily by the overly conservative setpoints.
j The inspector reviewed the PIRs for identification of root cause and action to prevent
{
recurrence. The licensee considered the cause of the incident to have been " procedure deficiency - technical error," and its corrective action consisted of review of ESAS surveillance procedures to assure that no other erroneous readings or adjustments had been made because of incompatible test equipment. The inspector also learned that the incompatibility of the Fluke 45 meter for this application came about as a result of circuit modifications which were implemented during the previous refuel outage.
The inspector concluded that the PIR root cause was too narrowly focused in that it did not address the programmatic issues of design change control and inadequate technical review of test instrument suitability, nor the practice of completing the bistable readjustments on all four ESAS channels prior to consulting management and obtaining resolution of the suspected discrepancy. The licensee committed to reevaluate the cause of the problem and the actions needed to prevent recurrence. This concern remains unresolved pending completion of that reevaluation, and NRC review of the programmatic concerns noted above (50-336/93-09-02).
,
m
.
.
4.4 Containment Sample Isolation Valve Surveillance - Unit 3 On April 17, 1993, while performing the biennial containment sample line valve position indication test, the licensee was unable to establish flow to the sample sink through two penetrations. Specifically, no flow was observed through the Pressurizer R.elief Tank (PRT)
gas sample line (3/4 inch) or the pressurizer vapor space sample line (3/4 inch) with the containment isolation valves indicating open. This precluded completion of the surveillance test for both the inside and outside containment isolation valves because the test procedure includes verification of process flow stoppage, as well as valve position indication. This technical specification required surveillance had to be completed by May 12,1993. The valves were subsequently shut and a plant incident report (PIR) generated. The PIR stated that the surveillance test failed but the valves were operable so long as they were closed.
Operations department personnel danger tagged the valves closed per Operating Procedure (OP) 3273, " Technical Requirements - Supplementary Technical Specifications," which stated that a containment isolation valve shall be considered operable as long as it is in its required safety condition and is not required to change positions in an accident situation.
Technical Specification (TS) 3.6.3.b requires that with a containment isolation valve inoperable, another automatic isolation valve must remain operable in the affected penetration if that penetration is open, and the inoperable valve must be repaired or the penetration must be closed by a deenergized automatic isolation valve, a closed manual valve or a blank flange within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />. The inspectors questioned the licensees interpretation of valve operability based on the failure to meet the surveillance acceptance criteria. When the licensee concluded on April 17, that the test acceptance criteria had not been met, action needed to be taken to close the penetration by one of the three TS approved methods. The inspectors also noted that Generic Letter 91-18 states that a system, structure, or component (SSC) is either operable or inoperable at all times. It further states that when the operability of an SSC is in question, an operability determination is to be promptly performed. The timeliness of the determination must be commensurate with the potential safety significance of the issue.
On April 23, after discussion with the inspector, the licensee deactivated the containment sample isolation valves in the closed position in accordance with TS. Further licensee investigation of the event had concluded that a deficiency existed in the Procedure OP 3273 which allowed continued use of penetrations with inoperable containment isolation valves without complying with the compensatory actions of TS. On April 30, the licensee logged into the TS action statement after a plant operations review committee meeting was held and a decision was made to delete the TS clarification in Procedure OP 3273.
The licensee completed an operability determination on May 6. This determination concluded that the inability to establish the proper test conditions did not in and of itself constitute a test failure if reasonable assurance existed that the valves were functioning properly. The licensee stated that the design of the sample containment valves is such that a change of state in the j
indication (i.e., lights), is a positive verification of stem travel. This positive indication of stem movement exists because the reed switches, which are rigidly attached to the stem i
_ _ _
.
housing, can only change state with a change of stem position. Although a reed switch can drift or fail these would result in either dual or no valve position indication. Based on the test infonnation showing proper position indication, the licensee concluded that the valves may be considered operable until the expiration of the test surveillance interval. On May 7, after verifying no line blockage or valve mispositioning, the surveillance was rerun and
,
completed satisfactorily, demonstrating valve operability. The licensee attributed the cause of the initial problem to either a communications or personnel error which resulted in the installation of the test equipment on the wrong sample path.
As further corrective action, the licensee performed a review of past PIRs which documented containment valve failures to determine if any inoperable valves had not been properly compensated due to the incorrect TS clarificatien in Procedure OP 3273. No other problems were identified. In addition, the licensee is considering reviewing other TS clarifications using non Unit 3 personnel to determine if other non-conservative TS interpretations exist.
!
The inspector determined that the containment isolation valves had been operable. The
,
licensee conservatively isolated the penetration due to poor conduct of surveillance activities.
Inappropriate interpretation of TS requirements also contributed to the problems identified during this activity. The inspector concluded that the inability to establish initial test conditions with proper valve position indication did not immediately render the valves inoperable; however, the time required to perform the engineering evaluation (20 days) to support this conclusic n was not prompt. The inspector considered the licensee corrective action to review the remaining TS clarifications to be a good initiative.
4.5 Failure of Ele-trolytic Capacitors - Unit 3 On April 7,1993, Unit 3 experienced a plant transient which resulted in a reactor trip.
Troubleshooting efforts identified the root cause as a failure of an electrolytic capacitor in the
electro-hydraulic control (EHC) system. The inspector reviewed previous plant incident
!
reports and noted that electrolytic capacitors had failed on January 7,1991, and October 16, 1992. These two failures resulted in the failure of the 15 vde 'B' and 'A'
emergency generator load sequencers (EGLSs), respectively.
The purpose of the solid state EGIS is to prevent an overload of the diesel generator from simultaneous restart of loads on the emergency bus following a loss of normal power to the emergency bus. Loads are stripped from the bus then restarted in a predetermined sequence according to the accident condition accompanying the loss of power. The EGLS has a 15 vde power source to supply the logic circuitry, a 28 vde for control panel indication lights, and a 48 vde to supply the output relays. The 15 volt power supplies failed to maintain rated i
voltage under load.
,
p
-
--
l
.
,
16 During each of these events, the licensee entered the applicable technical specification action
>
statement and replaced the faulty power supply. After the January 7,1991 occurrence, the j
licensee had reviewed the nuclear plant reliability data system (NPRDS) for common mode failures of the EGLS. No complete failures of the sequencer power supplies (manufactured by 12mbda) were found. Subsequent to the second failure in the EGLS, the licensee reviewed NPRDS and identified seven failures of Lambda power supplies over the last 12 year period. Five failures for the 15 volt EGLS power supply and two failures for the 28 volt EGLS power supply, none were of the same model number as those installed in Unit 3's EGIE.
.
l Licensee investigation of the October 16, 1992, failure determined that the root cause of the event was the electrolytic capacitor reaching its expected "end of life." This 15 volt power supply had been in service for nine years. Industry experience has shown that the reliability
>
of power supplies containing electrolytic capacitors decreases over the life span of the power supply. The life expectancy of capacitors varies depending upon capacitor grade. The rate of electrolyte loss by means of vapor transmission through the end seal and the operating or storage temperature appear to be controlling factors.
As a result of the two 15 vde EGLS power supply failures, the licensee developed a surveillance procedure to test the reliability of the 48 vde EGLS power supplies. The procedure required testing the power supplies to 110 percent of design load. A plant i
operations review committee (PORC) meeting was held on October 30,1992, to approve the test procedure and its performance. A decision was made not to perform the test at that time since no prior failure of the 48 vde power supplies had been identified, these power supplies are not constantly loaded like the 15 vde power supplies (less aged), and because the test loaded the component above the design load. In addition, the power supplies installed in the EGLS were obsolete and no longer manufactured, and there were no spares available. The i
sequencer had been fully loaded and verified operable in January 1992, during the performance of Procedures SP 3646A.17 and 18, " Train 'A' and 'B' ESF Actuation with l
LOP Test."
In-service tests (ISTs) 3-92-032 and -033, "EGLS Train 'A' and 'B' Power Supply Data Acquisition Procedures," were developed to provide data collection on EGLS power supply performance during quarterly surveillance test (Procedures SP 3646C.1 and 2, EGLS Output Relay Test - Train 'A' and Train 'B') that apply a partial (20 percent) load to the 48 vdc power supplies. In January 1993, the quarterly surveillance of the EGLS was performed.
The engineering evaluation concluded that the three EGLS power supplies were operable and well within the desired levels required for proper EGLS operation.
The inspector noted that Unit 2 also experienced failure of power supplies caused by aging of electrolytic capacitors during the loss of annunciator event on July 26,1991, (reference Inspection Report 50-336\\91-18). As a result of the event, Unit 2 implemented a power supply replacement and performance program. A controlled routing was issued to the other Millstone units after that event recommending that they review the ~ Unit 2 event for lessons l
1
?
i l
learned in evaluation of lifetime considerations of important power supplies. In response to the controlled routing and the January 1991 EGLS failure, Unit 3 proposed to identify all power supplies, their make and model number, and contact the various vendors for recommended test methodology to determine power supply degradation due to age. In addition, the licensee would contact Westinghouse to determine the feasibility and cost associated to upgrade and/or refurbish the various power supplies in use at the unit. These
,
activities were in process when the second Unit 3 EGLS failure occurred.
The Unit 3 instrument and control department is developing a power supply preventive
'7 maintenance program to monitor capacitor performance (regulation) and to refurbish or replace the power supplies / electrolytic capacitors based on plant and industry experience.
Prior to these events there was no licensee preventive maintenance (PM) program for power
supplies. The replacement of the power supplies would be predicated on deficiencies noted during the performance of the periodic surveillances or on-line failures. The licensee has requested the vendor to supply replacement power supplies for the EGLS and plans to replace all the EGLS power supplies during the next refueling outage, scheduled for July 1993.
Additionally, based on the other power supply problems noted above, the licensee is
-
scheduled to replace / refurbish the electro-hydraulic control, nuclear instrument, and Foxboro
Spec 200 process control system power supplies during the upcoming refueling outage. Other power supply changeouts will be performed during subsequent refueling outages.
The inspector determined that the licensee responded appropriately to the knowledge that the life expectancy of electrolytic capacitors vary with age. The PORC decision to delay performance of the 110 percent load test was acceptable based upon the information available to the inspector. The inspector noted that although a failure in the EGLS could result in the inability of safety related equipment to automatically start upon a loss of power event, it does not disable manual starting of equipment. Operators are trained and the emergency operating procedures direct operators to take manual action to start required equipment that fails to start automatically. The inspector concluded that the development of the power supply replacement / refurbishment and enhanced PM programs was needed and demonstrates an appropriate licensee response for continued safe operation of the unit.
5.0 ENGINEERING /TECIINICAL SUPPORT (IP 37700, 37828)
5.1 Missing Supports Identified - Unit 1 During the containment isolation valve repair outage, a licensee engineer noted that the instrument lines for the "B" main steam line flow sensor had rubbed together wearing the outside of one instrument line. Examination of the instrument piping on the "A" main steam line flow transmitter revealed similar fretting induced wear. The fretting occurred because several seismic supports designed to restrain movement of the instrument lines were
inoperable (disassembled). The licensee theorized that instrument line vibration caused pipe restraining straps that are located on the supports to loosen, then disassemble, during plant operation. Since the separated restraining straps could not be located inside the drywell, the
,
'
.
J
licensee believes that the vibration induced degradation had not occurred recently. Although missing seismic supports were also identified on the "C" main steam line instrumentation line, fretting wear was not evident. The instrument piping of concern is 1-inch stainless steel, Class 1 piping and is exposed to full reactor coolant system pressure during plant operation. Licensee analyses of the piping response spectra during a design basis safe shutdown seismic event without the supports installed, revealed that the instrument piping may sever at high stress areas. Accordingly, the license reported the discovery of the missing supports on April 31,1993, to the NRC per 10 CFR 50.72 (b)(2)(i) as a condition outside of the design basis of the facility. To correct the instrument line deficiencies, the licensee repaired the supports prior to reactor plant startup.
,
The licensee performed a walkdown of similar seismic supports in the drywell and turbine
building; however, no other loose supports were identified. Ultrasonic testing of both instrument pipes revealed that the fretting had reduced the diameter of the instrument lines on
,
the "A" and "C" main steam lines to approximately 0.141 mills from a nominal wall thickness of 0.179. However, since the pipe diameter was above the minimum wall thickness
'
of 0.055 mils, the decrease in pipe diameter was not safety significant. Chaffing gear was
installed in areas that were subject to high vibration to reduce the possibility of future fretting
'
induced wear.
The inspector considered the identification of the degraded instrument piping to be a good observation by the licensee. The inspector reviewed the non-conformance reports which evaluated the degraded instrument piping and the work orders which repaired the seismic supports. No inadequacies were identified. One weakness was identified in the licensee's immediate followup to the event. Following restoration of the supports and reactor startup, the licensee had not evaluated the potential for fatigue crackina on the unsupported instrument piping until the inspector asked if such degradation could hav ccurred. After conducting
further evaluation, the licensee determined that inspection of tr.e piping during the next refuel i
outage would be prudent. The inspector considered the failure of the licensee to promptly evaluate the effects of cyclic loading of the instrument piping to be a weakness in the engineering followup to this issue.
'
,
During the February 1994 refuel outage, the licensee will reinspect the supports to ensure
they are securely fastened. Liquid penetrant testing of the high stress areas in the instrument lines will also be performed to ensure fatigue cracking of the instrument lines did not occur.
i
'
To ensure that degraded equipment in the drywell is identified earlier, the licensee is
,
.
evaluating the need for performance of a detailed walkdown of the drywell area by personnel
i before releasing the drywell for unrestricted access. The licensee believes that if thorough i
walkdowns of the drywell had been performed following previous plant shutdowns, the degraded supports may had been identified earlier. The inspector had no further questions regarding the issue at this time.
k
. _ - -. -.
i
-
,
I
.
5.2 Reactor Vessel Water Ixvel Instrumentation Inaccuracies (Temporary Instruetkm l
2515/119) - Unit 1 j
' Itis Temporary Instruction (TI) was issued by the NRC to verify the implementation of i
operator guidance and training to ensure appropriate operator actions will be taken concerning reactor vessel water level inaccuracies following rapid depressurization transients.
Recommended guidance to operators was contained in NRC Generic Ixtter 92-04, dated August 19, 1992, entitled, Resolution of the Issues Related to Reactor Vessel Water Vessel
'
Instrumentation in BWRs Pursuant to 10 CFR 50.54(f). The recommended guidance included j
l assessing the potential for reactor vessel water level uncertainties during a reactor plant j
l depressurization and providing guidance to operators on how to respond to the level oscillations should they occur.
Reactor vessel level oscillations were first noted at Unit I during a July 4,1992, reactor plant i
shutdown. The licensee attributed the level oscillations to the offgassing of noncondensable gasses from the reference leg of the level indicating system. To prevent the buildup of the noncondensable gases, the licensee installed a modification in August 1992, which l
continuously purges the reference leg with water from the control rod drive system. Prior to plant startup, operators were briefed on the level oscillations which were observed during the July 1992, reactor shutdown. Operators were also trained on the operation of the reference leg purge system. The installation, testing, and the training of operators on the use of the l
modification was reviewed by the NRC and documented in NRC Inspection Report 50-l 245/92-22. Based upon this review, the NRC concluded that the installation of the m'.,oFication would not adversely affect other reactor plant systems, and that operators were propenly trained on the operation of the modification.
The licensee elected not to provide supplemental training to assure that potential level errors wJi cot re*, ult in improper operator actions. This determination was predicated on the belief that the installation of the constant purge system would prevent level oscillations. This position was outlined in a September 28,1992, licensee response to NRC Generic Ixtter 92-04. The NRC accep:ed the licensee position in a February 22,1993, letter to the
,
!
licensee.
During a reactor plant shutdown on April 27,1993, the licensee monitored the performance of the wide range level indicating system. No level oscillations were observed. Prior to the reactor plant shutdown, the plant had been on line for approximately 144 days. Based upon the lack of level oscillations, the licensee concluded that the backfill modification successfully prevented the buildup of noncondensable gases in the reference leg of the level indicating system. This determination was outlined in an April 29,1993, letter to the NRC.
Based upon the acceptable performance of the reactor vessel level indicating system to date, the satisfactory system response during the April 27,1993, plant shutdown, and NRC acceptance of the licensee's position regarding the supplemental training that was recommended by Generic letter 9244; TI 2515/119 is closed.
j
.
6.0 RADIOLOGICAL CONTROLS GP 83723)
6.1 Whole Body Counting The inspector reviewed select areas of the licensee's whole body counting program. The review consisted of an examination of the licensee's training program for personnel who operate a whole body counter, interviews with personnel who operate the whole body counter, discussions with supervisory and training personnel and observation of a whole body
counter calibration. Other areas of the program were reviewed in NRC Inspection Report 50-245/92-19. The following observations were made.
-
The training program that is provided to personnel who operate the whole body counting devices is limited in scope. The program provides only enough training for personnel to calibrate and operate the machine. Knowledge of the principles and theory of equipment operation are not provided. Technicians are not trained to interpret abnormal whole body count results.
j i
-
A technician who operates the whole body counters was able to calibrate the counter.
The technician stated that if he had any questions concerning the whole body counter results, he would contact health physics supervision. This action is consistent with the licensee's procedures for operating the whole body counter.
-
Licensee supervision does not expect the technicians who operate the whole body counters to be knowledgeable of all facets of the counter operation. Rather, the i
licensee expects the technicians to contact health physics supervision if any problems arise in the operation of a whole body counter. This expectation is clearly defined in station procedures for operating the whole body counters.
Based upon the review, the inspector concluded that although the licensee's training program for personnel who operate the whole body counters is limited in scope, it is adequate to meet the operational requirements that the licensee has established.
I l
7.0 SAFETY ASSESSMENT / QUALITY VERIFICATION GP 40500,90712,92799)
i I
7.1 Review of Written Reports The inspector reviewed periodic reports, special reports, and licensee event reports (LERs)
for accuracy, root cause and safety significance determinations, and adequacy of corrective
.
action. The inspectors determined whether further information was required and verified that the reporting requirements of 10 CFR 50.73, station administrative and operating procedures, and Technical Specifications 6.6 and 6.9 had been met. The following reports and LER's were reviewed:
l
.
>
Unit 1 Monthly Operating Report for March 1993, dated April 8,1993.
Unit 2 Monthly Operating Report for March 1993, dated April 8,1993.
Unit 2 Monthly Operating Report for April 1993, dated May 10, 1993 Unit 3 Monthly Operating Report for March 1993, dated April 8,1993.
Unit 3 Monthly Operating Report for April 1993, dated May 7,1993.
LER 50-245/93-02 reported the loss of secondary containment at Unit 1 due to high winds
,
which damaged a reactor building blowout ventilation duct. This event was reviewed previously in NRC Inspection Report 50-245/93-07.
LER 50-336/93-005 discussed an automatic reactor protection system trip actuation which occurred during an unplanned reactor coolant system cooldown event while the plant was in operating Mode 3. The event was reviewed previously in Section 2.4 of NRC Inspection Report T<'330/93-06.
LER 50-336/93-006 involved a determination by the inspector that the service water piping to certain vital switchgear room coolers did not meet the structural integrity requirements of the technical specifications. The event was reviewed in Section 3.1 of NRC Inspection Report 50-336/93-03.
LER 50-336/93-007 discussed a potential loss of safety function caused by the practice of operating an emergency diesel generator (EDG) in parallel to the grid while the remaining EDG is inoperable, or operating both EDGs in parallel with the grid. The event was reviewed in Section 4.1 of NRC Inspection Report 50-336/93-06 and Section 4.2.1 of NRC Inspection Report 50-336/93-81.
Safeguards Event Report 93-001 discussed the impact of the March 13,1993, blizzard on site security systems. The inspector discussed with security management the preparations which were made in advance of the storm based on previous experiences with severe weather conditions and the actions taken during the storm. Additional reviews of this event may be conducted durir.g future security program inspections.
Millstone 2 Startup Test Report for Cycle 12, dated April 19, 1993 Millstone Nuclear Power Station Annual Radiological Environmental Operating Report, January 1,1992 to December 31,1992, dated April 23,1993.
7.2 Review of Previously Identified Issues
'
7.2.1 Main Steam Isolation Valve Vent Valve Unlocked - Unit 2
'
This item (Unresolved 50-336/91-28-01) involved a weakness in licensee administrative controls for post-maintenance restoration of safety-tagged equipment. In this case, a drain valve (2-MS-255) on a main steam line isolation valve bypass line, was not locked following
.
-
.
.
repair. The equipment boundary protection for the work consisted of tagging the reactor coolant pump motors to prevent heat addition to the reactor coolant system and swell in the steam generators. For this reason, the tagout sheet for the work did not include re-installation of the valve locking device. As corrective action the licensee added a step to maintenance retest matrix forms (Procedure MP-270lX) to verify that worked components are aligned in accordance with applicable Operations Department procedures. The inspector verified that the procedure change had been made and observed through review of plant incident reports for 1992 and 1993, that no similar incidents had occurred. The inspector concluded that the corrective action had been effective. This item is closed.
7.2.2 Failure To Declare Service Water Valve Inoperable During In-Service Testing -
Unit 2 This item (Violation 50-336/91-30-01) involved failure of an operator to declare service water valve 2-SW-8.lC inoperable following the performance of an in-service surveillance test in which the maximum allowable stroke time was exceeded. The inspector found the root cause of the incident to have been lack of operator awareness of in-service test program requirements for equipment operability. Licensee corrective actions included procedure changes to clarify program requirements and operator training on those changes. The inspector noted that the applicable surveillance procedures had been changed, and verified through interviews with operators that adequate training on the changes had occurred. Also, the inspector reviewed plant incident reports for 1992 and 1993, and observed that no similar incidents had occurred. The inspector concluded that the licensee's corrective actions had addressed the cause of the violation effectively. This item is closed.
i 7.2.3 Operability of Electrical Equipment in West Vital Switchgear Room - Unit 2 This item (Unresolved 50-336/93-03-002) involved plant operation from January 6 to 28,1993, with the west 480 Volt vital switchgear room coolers out of service. The coolers are not included in Unit 2 technical specifications, but are credited in the Final Safety Analysis Report to maintain room temperature less than 104 degrees Fahrenheit in order to assure the operability of safety-related electrical equipment following a design basis accident.
In order to maintain room temperature within acceptable limits, the licensee had opened a door to an adjoining switchgear room. However, the inspector found that the compensatory measures were not controlled by procedures, and the licensee had not performed a technical evaluation to show that the measures would be effective under post-accident conditions.
On March 23, the licensee changed Operating Procedure (OP) 2315D, " Vital Electrical Switchgear Cooling," to establish formal requirements for temporary cooling of the west switchgear room. The inspector reviewed the procedure and noted that it establishes a suitable room temperature limit and provides for establishment of temporary ventilation, periodic temperature monitoring, and opertor review of equipment operability requirement _
_ __ _
,
,
The licensee also is evaluating the need for additional procedural controls and monitoring requirements should the coolers need to be taken out of service for maintenance. The inspector concluded that the procedure war adequate.
The room temperature limit of 104 degrees Fahrenheit is based on aging considerations over the design life (typically 40 years) of the equipment when operated at rated load. On April 6 the licensee met with the inspector to present its technical evaluation of switchgear room ventilation requirements based on calculation 92-FFP-934ES, " West 480 Volt Loadcenter Heat Gains and Maximum Room Temperature." The calculation concluded that under worst-case accident conditions (loss of coolant accident with off-site power available), room l
temperature would peak at 114 degrees Fahrenheit. The calculations and operating experience indicated that the compensatory cooling methods would lower room temperature by approximately 10 to 25 degrees. The inspector noted that the calculation was conservative in that it assumed that all of the equipment served by load center 22E would be running continuously following an accident. Based on the brief time (i.e., several hours) during which the load center would experience maximum room temperature, the inspector concluded that the equipment would remain capable of performing its safety function, and that the licensee's compensatory measures were acceptable. This item is closed.
7.2.4 Temporary Sample Station Installation - Unit 2 This item (Unresolved 50-336/91-09-01) involved a discrepancy in the control of a temporary modification to the containment combustible gas control system. The inspector found a temporary rig installed consisting, in part, of tygon tubing; and used for remote sampling of the containment atmosphere prior to venting and purging, and routine entry. The sample rig was not identified on system drawings, and had not been controlled by the temporary modification procedure. The safety concern was that a release of radioactive material from the containment into the enclosure building could occur if the vent and drain valves were left open and the tygon tubing failed during post-accident operation of the hydrogen monitoring system.
The licensee removed the sample rig and initiated plant incident report 91-36 to establish the root cause of the discrepancy. The licensee subsequently determined that personnel had misunderstood the configuration control requirements needed to maintain design pressure boundary integrity, and addressed the requirements in a memorandum to plant staff, dated July 30,1991. This memorandum emphasized that such temporary installations must be consistent with system design parameters / specifications, controlled by approved procedures, and removed when the temporary process is complete. In particular, the policy separates continuously monitored equipment (testing equipment) from unmonitored installations (temporary modifications). The inspector emphasized that the latter installations require safety evaluations as 10 CFR 50.59 changes whether through a design change, temporary modification or procedure review process.
,
-
-
. _ _ _ _ _ _ _ _ _ _. _.. _ _ _ _ _ _ _
.
>
The licensee assigned responsibility for installation of the sample rig to the operations department by controlling manipulation of the vent and drain valves through the station tagging system and revised operations and chemistry procedures. The inspector reviewed Operating Procedure OP-2383A, " Process Radiation Monitors Operation," and Chemistry Procedure CP-2806X, " Containment Purge." The procedures require that the vent and drain valves be danger-tagged when the sample rig is not connected, and that operations personnel install and remove the rig. The inspector walked down containment radiation monitors 8123A/B and 8262A/B, verified that the required tags were installed, and verified through a comparison of tag clearance 2-5186-92 and operating logs that the controls were being implemented. This item is closed.
8.0 MANAGEMENT MEETLNGS Periodic meetings were held with various managers to discuss the inspection findings during the inspection period. Following the inspection, an exit meeting was held on June 4,1993, to discuss the inspection findings and observations with station management. Licensee comments concerning the issues in this report were documented in the applicable report section. No proprietary information was covered within the scope of the inspection. No written material regarding the inspection findings was given to the licensee during the inspection.
&