IR 05000445/1990034

Insp Repts 50-445/90-34 & 50-446/90-34 on 900731-0807. Violation Noted.Major Areas inspected:900730 Main Steam Line Low Pressure Safety Injection Event,Including Actions Taken Prior To,During & Subsequent to Event

ML20059D659
Person / Time
Site:	Comanche Peak
Issue date:	08/27/1990
From:	Chamberlain D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION IV)
To:
Shared Package
ML20059D654	List: IR 05000445/1990034 ML20059D651 ML20059D656
References
50-445-90-34, 50-446-90-34, NUDOCS 9009070135
Download: ML20059D659 (18)

Text

,

i

,. .

l l

APPENDIX B U.S. NUCLEAR REGULATORY COMMISSION

REGION IV

l NRC Inspect. ion Report: 50-445/90-34 l 50-446/90-34 j Dockets: 50-445 Unit 1 Operating License: NPF-87 1 50-446 Unit 2 Construction Perr. .t: CPPR-127 Expires: August 1, 1992 Licensee: TV 11cetric Skyway Tower 1 400 North Olive Street I Lock Box 81 Dallas, Texas 75201 F:cility Name: Comanche Peak Steam Electric Station (CPSES), bnits 1 and 2 Inspection At: Glen Rose, Texas Inspection Conducted: July 31 through August 7,1990 Team Leader: A. T. Howell, Resident Inspector Team Members: D. L. Kelley, Reactor Inspector, Division of Reactor Safety (DRS), Region IV R. D. Vickrey, Reactor Inspector, DRS, Region IV D. N. Graves, Resident Inspector Reviewed by: ,

k2f10 D. Df Chamberlain, Chief, Project Section B Date Division of Ceactor Project:

Inspection Summary Inspection Conducted Ju y 31 through August 7. 1990 (Report 50-445/90-34; 50-446/90-34)

Areas Inspected: Announted special team inspection of the July 30, 1990, main

, steam line now pressure svfety injection event. This includes the events which I

occurred and actions taker by TV Electric personnel prior to, during, and subsequent to the July 30, 1990, main steam line low pressure safety injection that occurred following the opening of the No. 3 steam generator (SG)

atmospheric 7elief valve (ARV). No inspection of Unit 2 was conducted.

l

'

9009070135 900829 DR ADOCK OMj4 5 j

- ..

,- ..

-2-

'

Results:

Overall, the ;1censee's investigation of the safety injection was both prompt and thorough, and the actions taken, or that will be taken, by the licensee should reduce the likelihood of similar events. Operator response to the event and use of emergency operating procedures was good. The licensee had developed a thorough understanding of the sequence of events associated with the SI, as well as plant and equipment response. Short-term and long-term ,

corrective actions were needed in order to ensure that AC essential lighting will be available in the area of the No. 3 ARV and other areas of the plan .

The licensee concluded that a main steam line low pressure safety injection occurred as a result of the erratic performance of the No. 3 ARV which was '

caused by the passing of both steam and water through the No. 3 ARV when a control room operator attempted to open the relief valve from the control roo The rate compensated main steam line low pressure SI actuation circuitry performed as designed to initiate the SI. The licensee concluded that the source of water on the upstream side of the No. 3 ARV was from condensed steam that had backed up to the upstream side of the No. 3 ARV. This occurred because the main steam isolation valve (Maiv) Anstream drippot (located on the upstream side of the closed No. 3 MSIV) was isolated for personnel protection during maintenance on the main condense .

One violation was identified in that the hot 'andb; sperating procedure was inadequate because it did not provide guidat . for the periodic draining of isolated MSIV upstream drippots under plant conditions that are similar to those that existed on July 30, 1990 (Section 2.4.4).

,

Six inspector followup items were identified as a result of this inspectio These include: (1) a design modification to remove local articles for essential lighti,$g, as appropriate (Section 2.2.3);.(2) licensee's evaluation of the need to automatically isolate letdown during a safety injection and reroute the reactor coolant pump seal return relief valve line directly to the pressure relief tank (PRT) (Section 2.2.4); (3) monitor light box indicatic" following a safety injection electrical equipment load shed (Section 2.2.6);

l (4) long-term corrective action to ensure the expeditious removal

of personnel from containment during events (Section 2.4.1); (5) staging of l operator aids to facilitate the manipulation of remotely operated

manual valves (Section 2.4.3), and (6) the pressurizer heatup transient that occurred as a result of the safety injection (Section 2.5).

l l

!

._

,. ..

,

-3-DETAILS

, PERSON,$ CONTACTED

• J. L. Barker, Manager, independent 3afety Engineering Group (ISEG)

K. L. Berrett, Quality Assurance (QA)

• M. W. Blevins, Manager of Nuclear Operations Support W. J. Boatwright, Reactor Engineer

• H. D. Bruner, Senior Vice President

• J. H. Buck, Independent Advisory Group (IAG)

• W. J. Cahill, Executive Vice President, Nuclear

• C B. Corbin, Licensing Engineer .

'.. G. Creamer, Plant Engineering R. Flores, Shift Operations Manager

• G. Guldemond, Manager of Site Licensing -

• J. C. Hicks, Unit 2 Licensing Manager

• C. B. Hogg, Chief Engineer

• A. riusain, Director, Reactor Engineering -

• J. J. Kelley, Plant Manager J. Linnard, Westinghouse

• D. M. McAfee, Manager, QA J. McInvale, Plant Operations staff

• J. F. McMahon, Manager Nuclear Training

• J. W. Muffett, Manager of Project Engineering

• Nyer, IAG

• E. F. Ottney, Project Manager, Citizens Association for Sound Energy (CASE)

• S. S. Palmer, Stipulation Manager ,

• F. S. Poppe11, Licensing Engineer

• O. Porter, Operations Support Engineering (OSE) t

• C. W. Rau, Unit 2 Project Manager '

• D. M. Reynerson, Director of Construction

• M. J. Riggs, Plant Evaluation Manager, Operations

• A. B. Scott, Vice President, Nuclear Operations B. J. Smith, Operations Support

• J. C. Smith, Plant Operations Staff W, A, Smith, OSE

• P. B. Stevens, Manager of OSE

• L. Terry, Director of QA

• 0. W. Thero, CASE

• T. G. Tyler, Director, Management Services B. Voltig, Westinghouse

• R. D. Walker, Manager of Nuclear Licensing

• J. R. Waters, Site Licensing

• A. West Project Engineer

The intpectors also 60ntacted other licensee employees during this inspection l perio * Denotes personnel present at the August 7, 1990, exit intervie !

_

.. ..

-4-NRC personnel present at the August 7 exit interview:

S. D. Bitter, Resident Inspector L. J. Callan, Director, Division of Reactor Safety (DRS), Region IV D. D. Chamberlain, Chief, Project Section B, Division oi Reactor Projects, Region IV D. N. Graves, Resident Inspector A. T. Howell, Resident Inspector W. D. Johnson, Senior Resident Inspector, Unit 1 D. L. Kelley, Reactor Inspector, DRS, Region IV R. M. Latta, Senior Resident Inspector, Unit 2 FOLLOWUP ON MAIN STEAM LINE LOW PRESSURE SAFETY INJECTION (93702)

2.1 Overview r On July 30, 1990, CPSES, Unit 1, was in Operational Mode 3 (hot standby), with i the reactor shut down, and reactor coolant system (RCS) temperature and pressure at approximately 557'F and 2250 psig, respectively. The MSIVs were shut, and the plant was in the 5th day of a planned maintenance outage that included work on the main condenser. The plar.t was being maintained in hot standby by the use of the auxiliary feedwater (AFW) system and the SG ARV At approximately 9:56 p.m. (CDT) on July 30, 1990, a reactor operator attentpted to manually open the No. 3 ARV (1-HV-2327) in order to reduce the No. 3 SG level. The operator increased controller demand to approximately 20 percent ope Initially, the No. 3 ARV failed to respond as expected, and then it rapidly opened to approximately midposition about 20 seconds late Following the rapid valve opening, a safety injection (Sil actuation signal occurred on rate compensated low main steam pressure. Additionally, a reactor trip signal

,

'

was received on two of four channels of No. 3 steam generator lo-lo leve Control room operators immediately recognized that the No. 3 ARV opened more than was demanded by the ARV controller, and they attempted to close the No. 3 ARV from the control room. After the attempt to close the No. 3 ARV from the control room failed, auxiliary operators were dispatched to lo ally close the No. 3 ARV upstream isolation valve (IMS-098) which is operated by a remote, manual operator just outside the ARV roo There was no apparent lighting in the No. 3 ARV room, and there was evidence of ventilation ducting and ARV vent insulation damage, as well as lighting fixture damage near the No. 3 ARV. This was apparently caused by the blowback of a steam and water mixture from the ARV vent stack into the ARV room. The auxiliary operators were initially unable to close the No. 3 ARV (which was positioned to approximately 50 percent open by its manual operator) because of the inadequate lighting and environmental cohditions in the ARV room. The l auxiliary operators were able to close the No. 3 ARV upstream isolation valve, which took approximately 8-10 minutes to close by its remote operator

'

, (approximately 760 turns on the manual operator handwheel). The operators then i positioned the No. 3 ARV manual operator to the neutral N ition, and the No. 3 ARV fully shut because contioller demand was previou y reduced to

'

O percent demand.

!

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

..

,. ..

-5-A Notification of Unusual Event (NOVE) was declared and notification to St.Ste, local officials, and the NRC was mad The SI signal and SI sequence were reset by procedure and running equipment was '

secured. Safety injection flow was terminated at 10:19 p.m. after injecting approximately 3000 gallons of water from the refueling water >corage tank (RWST). During the event, pressurizer pressure fell to 2210 psig, and the RCS Loop 3 cold leg temperature fell approximately 30'F before the No. 3 ARV was isolated at 10:13 p.m. Pressurizer level increased to 73 percen At the time of the event, there were three people inside containment performing a routine inspection for RCS leakage. They exited the containment at approximately 10:40 p.m., af ter power (which had been load shed as a result of the safety injection) was restored to the normal personnel air loc Normal letdown flow and reactor coolant pump seal leakoff were reestablished by 10:25 p.m. The NOVE was terminated at 12 midnight. Within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after the event, licensee management decided to cool the plant down to cold shut down in order to investigate the cause of the event. The purposes of the licensee's '

investigation were: (1) to evaluate the safety injection in order to determine the adequacy of equipment response, pt:rsonnel actions, and plant performance, as well as determine the root cause and prescribe corrective action; (2) to determine if the design and existing method of operation of Unit 1 in Modes 3 and 4 are compatible; and (3) to compare the results of the July 26 and July 30 saf:ty injection event reviews with industry experience to determine any common concerns, 2.2 Plant Systems and Equipment Response Prior to the event, the plant was shutdown and was being maintained at normal operating temperature and pressure in hot standby. All four MSIVs were shut while maintenance was being performed on the main condenser. The MSIV upstream drippots, which drain water from the piping upstream of the MSIVs to the condenser, were also isolated. Apparently, condensation in this section of i piping upstream of the No. 3 MSIV caused a significant amount of water to accumulat The AR/ piping is connected to the main steam piping upstream of the MSI All safety systems required for the mitigation of the event functioned properly; however, several systems and components displayed responses that were ,

determined by the inspectors to require further inspection followup. These systems and components included the No. 3 atmospheric relief valve, the letdown and reactor coolant pump (RCP) seal return relief valve pipin;, reactor trip i instrumentation, DC em rgency lighting, and monitor light box indication .2.1 Description of the No. 3 ARV (1-HV-2327) and Actuator fhe No. 3 main steam line ARV (1-HV-2327) is a Fisher 8 X 6 Design EWP body with a Size 80 Type 667 actuator with a side mounted manual handwhee .

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

.. ..

-6-The valve body is a single port globe design with a cage guided push-to-close, primary valve plug leaded and unloaded by an internal pilot plug. This design provides a single-seat, metal-to-metal shutoff of flow when close With the valve closed, the primary plug is held against its seat, which is an integral part of the valve cage, by the force of the valve actuator on the pilot plug and valve inlet pressure on top of the primary plu As the valve begins to stroke open, the pilot plug lifts off its seat on the primary plug and relieves the inlet pressure from the top of the primary plug through registration holer, in the primary plug to the valve outlet. The primary plug is held shut by the precompressed pilot p*iug springs until the pilot plug disc contacts the primary plug and lifts it off of its sea '

As the valve strokes to the closed position, the primary plug first contacts i its seating surface on the cage and shuts off flow through the valve at the primary plug seat. At this point, the pilot plug is still open and passing i flow through the registration holes in the primary plug. The pilot plug continues tn close, shuts off flow throup', the registration holes, and compresse. he pilot plug spring The valve actuator is a reverse-acting, spring-opposed, diaph*agm actuator and is used to position the valve fully op a, fully shut, or in throttled position Instrument air provides the pneumatics for opening the valve while the actuator spring provides the closing force. Increasing air pressure below the actuator diaphragm will cause the actuator stem, which is coupled to the valve stem, to move upward. As air pressure below the diaphragm decreases, the spring forces the actuator stem downward, closing the valve. On a loss of instrument air, the actuator moves the valve to the shut positio The air pressure supplied to the actuator is determined by the valve control system. An electrical control signal is generated at one of two auto / manual hand control stations at the main control board or at the remote shutdown panel. A pneumatic transducer (1/P) converts the electrical control signal to a corresponding 3-15 psig air signal, which is directed to the valve positioner located on the side of the actuator. The positioner utilizes the 3-15 psig signal and either increases or decreases the air directed to the actuator diaphragm. As the actuator stem moves, a positioner feedback arm, attached to the valve stem, feeds the actual valve position back to the positioner to stop i the change in the supplied pressure. If the valve does not reposition, the positioner will continue to increase or decrease the air loading under the actuator diaphragm to attempt to move the valve to the correct position. The ,

air pressure to the actuator diaphragm is regulated and limited to '

approximately 55 psi A handwheel mounted on the side of the actuator allows manual operation of the valve. The handwheel is continuously engaged via a worm gear arrangement, with a threaded sleeve that fits around the valve stem. A position indicator is mounted on the sleeve to indicate its position. When in the NEUTRAL position, the sleeve does not restrict movement of the valve by the air actuator. When positioned in the open direction, the sleeve will contact the travel stop and i

l

.

.

1

~7-lift the stem and the valve will open as previously described. If moved in the close direction, the sleeve will contact the stem connector and attempt to force the stem in the close direction. Once the sleeve is out of the NEUTRAL position, valve movement by the pneumatic actuator may be restricte .2.2 No. 3 Atmospheric Relief Valve Response During the Event At approximately 9:56 p.m. on July 30, 1990, the reactor operator attempted to open the No. 3 ARV by dialing in a 20 percent open demand signal from the control room. The valve did not appear to open as the operator expecte !

Approximately 20 seconds later, the ARV opened rapidly to at least the 50 percent open position. The resulting pressure transient initiated a SI on main steam line low pressure. Attempts to close the ARV from the remote controller were unsuccessful and the valve was subsequently isolated locally as described in paragraph 2.1. The initial observed response of the ARV is attributed to the accumulation o; water upstream of the ARV. The licensee concluded that the source of water on the upstream side of the No. 3 ARV was l from condensed steam that had backed up because the No. 3 MSIV upstream drippot l was isolated for main condenser maintenance (see Section 2.4.4 for further

'

details). As the pilot plug began to open, water began to flow onto the primary plug. This water could have flashed to steam, creating a higher pressure on the primary plug and keeping it on its shut seat. With the valve failing to reposition as demanded by the actuator, the positioner would continue to increase the pressure to the actuator diaphragm. At some actuator pressure, higher than the normal pressure for 20 percent open, the primary plug may have become unseated. As flow increased through the primary plug, this ,

water also flashed to steam, creating a large pressure surge on the underside i of the primary plug, forcing it in the open direction. This sudden force caused the valve to " pop" open and allowed a large surge of water and steam into the valve discharge pipin The discharge line & ain was found partially plugged following the event, thus the discharge line n.sy have also contained water. The other three ARV discharge drain lines were inspected and found to be plugged with rust. The i large surge of steam and water overcame the vent stack flow capacity and ,

allowed steam and/or water to flow back into the room. This flow back into the ;

room, in addition to causing damage to two light fixtures and insulation in the area, could have impinged on the manual actuator handwheel with sufficient force to rotate the handwheel and move the manual actuator sleeve into the 50 percent open position. This prevented the valve from closing when the operator attempted to shut the ARV from the controller in the control roo No specific evidence could be found to support the handwheel rotation theory, but the lack of any physical actuator damage indicates that mechanical impact did :

not reposition the manual actuator sleeve. Two auxiliary operators stated that 1 upon arrival at the scene, the ARV was approximately 50 percent open and the manual handwheel position indicator was in the "open" direction. After closing the manual upstrear. isolation valve, the operator repositioned the manual handwheel to " neutral" and the valve closed (the controller in the control room was still in the zero demand or closed position). The steam and water blowback into the ARV room caused damage to ventilation ducting, ARV vent insulation, and two lighting fixtures.

l

,

i

,. .

-8-As t result of the ARV's response, a number of actions were performed by the licensee in order to evaluate the cause:

The valve was operated throughout its full range of travel using the manual handwhee Valve response was observed to be norma *

The valve was operated throughout its full range of travel using the remote manual / auto statio Valve response was observed to be norma *

Valve stroke length was measured and determined to be correc *

The bench set of the valve was verified to be correc o

,

The locally-mounted control components were visually inspected externally and no damage was note * The valve actuator was disassembled and inspecte No indications of !

damage or unusual wear were note l

• The ARV was disassembled and inspected. No damage was observed that could have caused the ARV's observed performance, nor was any damage observed that would have been caused by the event. The valve cage was replaced due to a small defect in the seating surface that appeared to be caused when .

the cage was being machined to remove a small steam cut. The primary plug }

was replaced to correct for overtravel of the pilot plug assembly. These were determined to have had no effect on valve operation during this !

event, but could contribute to valve leakag *

The valve was reassembled and stroked nanual!v and with the remote controller. The plant was in cold shutdown at be time. No problems with valve operation were note ,

i

A loop calibration of the controller was performe *

The valve was stroked again while in hot standb Ne problems with valve operation were note !

The valve was stroked again once reactor power reached approximately !

35 percent in accordance with temporary Performance and Test Procedure PPT-TP-90A-031, Revision 0, " Partial Stroke of Atmospheric Relief Valve "

No problems with valve operation were note .2.3 Adequacy of Lighting in the No. 3 ARV Room i

During the July 30, 'E SI, lighting in the area of the No. 3 ARV.was reported ;

to be poo The inspectors questioned why the battery powered DC emergency '

lights in the area did not automatically energize during the event to provide adequate lighting. The inspectors determined that the DC emergency lights did not automatically energize following the safety injection because power to AC essential lighting in the area was not load-shed during the safety injectio The lighting system is designed in such a way that, if area lighting has lights

i

,. .

. that are powered f rom an AC essential power source, th' DC emergency lighting in that area will only energize on a loss of the AC issential power sourc For a safety injection, the essential lighting is not loed-shed, thus, in the No. 3 ARV area, the DC emergency lights did not energize bec use power was not lost to AC essential buses 1EB1 and IEB2, which power the AC essential lights through motor control centers and lighting distribution panel The licensee determined that the lighting in the No. 3 ARV room was inadequate because the essential light in the room was turned off at the local wall-mounted switc Discussions with licensee personnel revealed that the auxiliary operators who responded to the event did not attempt to turn on the AC essential lights by manipulating the local wall mounted switches. Had operators attempted to turn on the AC essential lights, they may have been confused because of the 3-way position of the wall switches (one at each end of the corridor) and the fact that the lights are sodium vapor type, which require about 2 minutes to come o The inspectors determined that the lighting subsystems (Class IE AC essential, Class 1E AC nonessential, and DC emergency lighting) in the area of the No. 3 ARV were installed in accordance with their design and that the design appeared to meet the applicable regulatory requirements. The inspectors were concerned, however, that the conditions that existed in the area of the No. 3 ARV on July 30, 1990, not only presented a potential personnel hazard, but also might have hampered operator actions to isolate the No. 3 ARV. As a result of these concerns, the licensee was asked to address, in writing, at a technical meeting conducted on August 2,1990, their assessment of the adequacy of emergency lighting. As a result of the licensee's investigation, licensee personnel performed a walkdown of the plant in orcer to ensure that all AC essential lights were energized. The licensee also instructed plant operators to assure that essential lighting is left on. These actions were completed prior to the restart of Unit The licensee committed in an August 4, ?990, lette (TXX-90283) to NRC Region IV that TV Electric would implement a acsign modification to remove the AC essential lighting light switches as appropriato in order to ensure that there will be sufficient lighting if a similar event occurs in the future. Therefore, pending the implementation of this design modification, this item is identified as an inspector followup item (IFI 445/9034-01).

2.2.4 Water in the Containment Sump Following the event, approximately 500 gallons of water was pumped out of the west containment sum A Phase ' A' containment isolation signal was received concurrent with the SI signal . This containment isolation signal shuts the normal return path for the RCP seal return flow and shuts the containment isolation valves for letdown flow. Relief Valves 1-8117, in the letdown system, and 1-8121, in the RCP seal return line (located upstream of the closed isolation valves), lif ted af ter the SI occurred. These two relief valves and two relief valves from the -esidual heat removal (RMR) system (1-8708A & B) tie into a common 6-inch header which empties into the pressure relief tank (PRT).

The letdown relief valve and the RHR relief valves normally relieve steam to the PRT. Each of the three relief valve discharge lines ccntain a steam trap

, .

,

-10-which is designed to drain water from its respective line to prevent waterhammer in the 6-inch header when they lift. The RCP seal return relief valve normally relieves water into the common 6-inch header. This water is passed by the three steam traps directly into the containment sump. This condition continued until the containment isolation valves were restored to their normal position and RCP seal return was restored to its normal lineu Any water contained in the letdown system relief valve discharge line would I also be directed into the containment sump via the steam trap l The emergency responso guidelines have been revised to more clearly state that letdown should be verified isolated upstream of Relief Valve 1-8117 to minimize the amo;nt of time that the letdown relief valve would be required to lif An ,

inspection of the RCS was performed in order to verify that no other leakage i into the sump occurre The licensee was also evaluating the need for design ,

modifications to automatically isolate letdown during a safety injection and redirecting the RCP seal return relief valve discharge directly into the PRT instead of into the common header with the letdown and RHR relief valves. The ;

inspectors will monitor future licensee actions in this are These inspections will be tracked by an inspector followup item (IFI 445/9034-02).

2.2.5 Reactor Trip Signal on Lo-Lo Steam Generator Level A No. 3 SG lo-lo level reactor trip signal was generated during the event on two of the four channels of narrow range SG 1eve The two flow transmitters used for indicating steam flow atiliza pressure taps that are connected to two SG 1evel transmitter reference legs. During the event, No. 3 main steam line pressure decreased sharply when the No. 3 ARV rapidly opened. The decrease in steam line pressure caused flashing of the water accumulated in the steam line upstream of the ARV resulting in a rapid pressure increase in the main steam line. This positive pressure spike was transmitted through the steam flow transmitters into the reference legs of their respective steam generator level transmitters, causing them to spike low !

and produce a reactor trip signal. Actual level never decreased below approximately 55. pet,:ent on the narrow range instruments. The level transmitters that were not connected to the steam flow transmitters indicated very little fluctuation during the transien An analog channel operations test (ACOT) was performed on one of the level transmitter's (1-LT-539) protection channels to ensure that instrument calibration was not the cause of the trip signal. The ACOT was satisfactor The calibration of the level transmitter was verified using the applicable portion of the channel calibration procedure and found to be satisfactor .2.6 FaStv &nitor Light Box Indication Dwing the pe formance of E0P-0.0, " Reactor Trip or Safety Injection," several i discrepancies were noted between monitor light box (MLBs) indications and actual equipment responses. The MLBs indicating that lead shedding was complete for Buses 1EB4-2 and XEB2-2 did not illuminate. MLBs 181-4.5,

.. .

-11-

4A2-2.7, 4A2-4.6, 4B2-2.4, and 4B2-4.5 did not illuminate to indicate that the I indicated components were in their actuated positions. All components were verified to be in the correct position, including the components required to be load-shed from 1EB4-2 and XEB2-2, by other available indications. The MLB is an operator aid only and does not replace the methods of operator verifications by procedur Troubleshooting was ic progress to investigate the cause(s) of the MLB malfunctions and future inspection of the licensee's actions will be tracked by an inspector followup item (IFI 445/9034-03).

2.3 Review of Main Steam Line Low Pressure Safety Injection (SI) Actuation Signal Prior to the event, No. 3 main steam line pressure was indicating approximately 1060 psig. When the No. 3 ARV opened, steam line pressure rapidly decreased to approximately 1025 psig, at which point the SI occurre Because the SI occurred at a relatively high main steam line pressure, there was concern that the main steam line pressure instrumentation may not be processing the signal properl The main steam line pressure signal is processed through a lead-lag circuit that rate-compensates the steam pressure signal. The . eam pressure signal is modified according to the rate at which it is decreasing. With a rapid decrease in steam pressure, such as the one that occurred when the No. 3 ARV opened, the steam pressure signal was adjusted downward and initiated the S The main sterm line low pressure SI setpoint is 605 psi The licensee attempted to verify tha*. the observed and recorded main steam line pressure decrease would Mye caused tne SI on rate-compensated low main steam line pressure. On the bt. sis of the recorded data points, it was not possible to analytically F ;. strate that the SI setpoint was reached for this even Because the data _.allection frequency is once per second, it is possible that tne initial pressure tra.nsient maximum and/or minimum may not have been recorded. Other values were analyzed that would have resulted in exceeding the safety injection setpoint. As a result of these analyses, and the close agreement between data recorded during the SI of July 26, 1990, and the analytical model, it was concluded by the licensee that the lead-lag circuitry performed as designe .4 Review of Operator Response and Procedural Adequacy On the basis of the inspectors' review, the operators' response throughout the event was timely and appropriate. The operators entered E0P-0.0A, " Reactor Trip and Safety Injection," immediately after determining that the event was a main steam line low pressure SI. Operators were dispatched to the No. 3 ARV upstream isolation valve once it was determined that the valve would not close in response to remote operation from the control room. The operators properly transitioned from E0P-0.0A to E05-1.1A, " Safety Injection Termination."

During this event, a number of observations revealed that some procedural enhancements might be appropriate and, in one case, the applicable procedure

2L

.

O-12-i was inadequate for the existing plant conditions. These observations listed pertain to operator involvement during the event. These are discussed in more detail belo ]

l

• Three people were in the containment building during the event and did not exit the containment until power (which had been load shed during the SI)

was rescored to the normal personnel air lock approximately 40 minutes after the S * It tt ok three operators approximately 8-10 minutes to close the No. 3 ARV upstream isolation valve because it took approximately 760 revolutions of the manually operated handwheel to close the valv * Environmental conditions in the No. 3 ARV room (e.g., inadequate lighting, loud steam flow noises from the No. 3 ARV, and water from the blowback of ;

steam and water around the ARV) significantly hampered the operator's ability to close the No. 3 ARV at the ARV itsel * The hot standby operating procedure did not have a requirement to periodically drain isolated MSIV upstream drippots when the MSIVs are shut. As a result, water backed up to the upstream side of the No. 3 ARV, thereby, cas' sing the erratic performance of the AR . Personnel in Containment Three people were inside containment at the time of the SI. They were unable to exit through the normal personnel air lock because power was lost to the air lock ar. a result of the SI load shed and the pump required for manual operation was not inside containment. These people were in communication with control room operators, and it was agreed that the people in containment would exit through the normal personnel air lock after the SI signal was reset and power was restored to the air lock. This occurred approximately 40 minutes after the initiation of the SI. The licensee noted that these people could have exited containment by use of the emergency air lock on the 905-foot elevation of the containment building if conditions would have warranted expeditious evacuation of containment. As a result of the licensee's investigation, however, it was determined that not all of these peuple were familiar with the operation of the emergency air loc The licensee identified the below listed concerns. The inspectors also had similar concerns relating to personnel in containmen *

No pump was available inside containment to manually operate the normal personnel air lock,

No instructions were available inside containment for either the normal personnel air lock or the emergency air lock (although operations personnel have been trained on manual operation).

- _ - .

.. ..

l -13-l

'

Emergency operating procedures did not address loss of/ restoration of power to the normal personnel air lock or the need to evacuate personnel from containmen *

Containment temperatures increased after the SI due to loss of containment recirc stion ceiling fans. This could have presented a personnel safety hazar During a technical meeting held between NRC and the licensee, NRC Region IV requested that TV Electric address, in writing, at least 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> prior to reentry into Mode 3, the potential personnel safety hazards associated with personnel in containment during plant events. The licensee responded on August 4, 1990. As a result of the licensee's investigation, the licensee placed job aids at both air locks to facilitate operations and revised the emergency operating procedures to provide guidance for the expeditious removal of personnel from containment. These actions were completed prior to the startup of Unit 1. The inspectors determined that there was no personnel safety hazard associated with potential heat stress because the wet bulb temperature was approximately 82 F which would not require a stay time under the conditions defined in STA-674, " Heat Stress Management." Actual i containment temperatures rose approximately 5-10*F (containment average I temperature) during the event, and this rise would not have significantly i increased the wet bulb temperatures, which containment stay times are based o The licensee was also considering longer-term actions, such as providing

,

'

additional training for nonoperations department personnel on manual air lock operations and evaluating whether vital power should be provided to the l hydraulic operating system for the normal personnel air iock. The inspectors l will monitor future licensee actions in this area. These inspections will be l tracked by an inspector followup item (IFI 445/9034-04).

2.4.2 Damage in No. 3 ARV Room From Steam and Water Blowback Another potential personnel safety hazard that the licensee was asked to aodress in writing during the August 2, 1990, technical meeting concerned the blowback of steam and water in the No. 3 ARV roo This blowback caused insulation damage to ventilation ducting and the No. 3 ARV vent, as well as two light fixtures. The inspectors were concerned that personnel might have been ',

injured had they been near the No. 3 ARV when this blowback occurred. The '

licensee responded to this concern in their August 4, 1990, letter and noted that the design of the ARV vent stack precludes blowback of steam into the room. However, on July 30, 1990, the presence of water restricted flow through l the vent, contributing to the blowback. In order to preclude the future presence of water, the licensee revised Initial Plant Operating Procedure IP0-007A, Revision 5, " Maintaining Hot Standby," to ensure that MSIV upstream drippot isolation valves are open or that the affected MSIV upstream drippot is i periodically dra's.d when the MSIVs are closed (see Section 2.4.4 for !

additional deta a). Additionally, the licensee noted that it was previously l recognized that manual closure of an ARV would expose the individual operating I the valve to adverse environmental conditions. As a result, the ARV upstream l

l

..

o' -

-14-isolation valves w re r 6sequently configured to allow manual operation of the valves from outside the ;RV rooms. At the end of the inspection, the licensee was modifying the appli able plant procedures to reinforce the use of the upstream isolation valve instead of trying to close the ARV at the ARV itsel .4.3 Operation of the No. 3 ARV Manual Upstream Isolation Valve The inspectors were concerned that it took approximately 10 minutes to close the No. 3 ARV Upstream Isolation Valve IMS-0098 which required approximately 760 turns on the remote manual operator. As a result of a previous NRC concern, the licensee conducted a study in 1989 to identify problems with local operation of valves with remote operators. Valves identified as difficult to operste as a result of this study were added as an attachment to Operations Work Instruction Procedure 0WI-206, " Guidelines for Operation of Manual and Power Operated Valves." However, this attachment did not include the steam generator ARV upstream isolation valves. Discussion' with licensee personnel and a review of licensee records revealed that the ,.i89 study was performed on the basis of a list of valves with remote operators provider in Specification MS-100. The revision of MS-100 used (1987) did not include the ARV upstream isolation valves since these valves were not configured with remote operators in 1987. As a result, the licensee plans to revise the 1989 study to address changes since the 1987 revision of MS-100. Additionally, the licensee was already conducting a study prior to the July 30, 1990, safety injection to identify accessibility problems with local valves. Finally, the engineering department will review the feasibility of using air wrenches on remote operated manual valves that are difficult or time consuming to operat The inspectors will monitor these activities during future inspections. These inspections will be tracked by an inspector followup item (IFI 445/9034-05).

2.4.4 Hot Standby Operating Procedure and Alarm Procedures for Upstream MSIV Drippots As a result of its investigation, the licensee concluded that the presence of water on the upstream side of the No. 3 ARV caused the erratic performance of !

the ARV. The source of this water was from condensed main steam that had backed up to the upstream side of the ARV as a result of the No. 3 MSIV (1-HV-2335A) being closed and its associated upstream drippot (MS-23) being isolated for personnel protection during main condenser maintenance. The licensee performed a survey of other Westinghouse plants to determine if its method of operation of the plant in hot standby was consistent with industry practius. The licensee also reviewed Initial Plant Operating Procedure IP0-007A, Revision 5, " Maintaining Hot Standby," in order to determine if this procedure provided adequate guidance for the plant conditions that existed at the time of the July 30 SI (i.e., Mode 3, MSIVs shut, steam generator blowdown system and upstream MSIV drippots isolated for condenser maintenance). The inspectors also reviewed IPO-007A for adequac The licensee dete" mined that IPO-007A was adequate for maintaining a normal hot standby configuration but was not adequate for the plant conditir.ns (operation with the main condenser isolated) that existed on July 30, 1990. TI.e licensee review of IPO-007A revealed that this procedure: (1) did not provide guidance l

.

i

I

.. !

-15-to locally drain main s.eam line drippots periodically when the normal dra h flow path is not available; (2) did not provide a caution addressing the l'-

of condenser vacuum and resulting loss of normal steam line drippot drain 9 path; (3) did not provide a caution addressing loss of the condensate system and the resulting loss of steam ger.erator blowdown system which affects SG 1evel control; and (4) did not provide additional direction to an existing step for SG level control to allow steaming of a steam generator as an option for lowering level. The inspectors concluded that had this procedural guidance been provided, the likelihood of the July 30, 1990, SI event would have been significantly reduced. Failure to provide an adequate general operating procedure is an apparent violation of Technical Specification 6.8. (445/9034-06). Licensee identification of this violation and prompt correction ,

of IPO-007A prior to the restart of Unit I would have met the criteria for a *

noncited violation as described in the NRC Enforcement Policy; however, this violation is being cited because of its potential safety significance in that the condition that existed on July 30, 1990, could have affected the operability of all four ARV The inspectors concurred with the licensee's conclusion that the indications associated with the high-level alarm for the MSIV upstream drippots as well as the guidance provided in Alarm Response Procedure ALM-0071A, " Alarm Procedure 1-ALB-7A," contributed to the event. As a result of the licensee's investigation, the licensee determined that the hi-hi level annunciator was in for several days prior to the cm.nt, but the hi-hi level lights (one for each steam line) were not believed to have been lit as should have been the case when the hi-hi level annunciator light was lit. Troubleshooting by the licensee d not reveal any problems with the MSIV upstream drippot level alarms. Even with the automatic drains for the MSIV upstream drippots open, the drippots could not drain because they were isolated downstream of the automatic drain valves. The alarm response procedure provided optional guidance to locally blow down the drippots to the floor drains, but it did not explicitly require that this be performed. The licensee concluded that the operators may have failed to implement this optional procedural guidance because operators failed to recognize the significance of water in the MSIV upstream drippots (due in part to the unusual mode of plant operations) and because there were numerous (approximately 150) alarms including various other drippot level alarms (approximately 20). As previously stated above, revision to Ip0-007A should preclude future occurrences of water (from inlated MSIV upstream drippots) backing up to the upstream side of the ARVs: aowever, the licensee is also evaluating the need to revise Alarm Response t'rocedure ALM-0071A, " Alarm Procedure 1-ALB-74."

2.5 Licensee Investigation of the Event Immediately following the SI event on July 30, 1990, the licensee issued Operations Notification and Evaluation (ONE) form FX 90-1976. The management assessment for this ONE form determined that an evaluation team would be required because a coordinated effort was needed to evaluate the incident and to recommend necessary corrective actions. Within I hour of the event, licensee senior management directed a plant cooldown end initiated action to develop a charter for the team. The charter for the team was to evaluate the

- - .

, __ ____ _ _ _

9* SI in order to determine the adequacy of equipment response, personnel actions, and plant performance in order to determine a root cause and prescribe corrective action and determine if the design and current method of operation of the plant in Modes 3 and 4 are compatible. Additionally, the plant evaluation manager was tasked with comparing the results of the July 26 and July 30 SI event reviews with industry experience to determine common concern These cetions were approved during a special Station Operations Review Committee meeting conducted on July 31, 1990. The licensee conducted their activities in accordance with Procedure STA-423, " Evaluation Team," which included provision for interface with NR A SORC meeting was conducted at 3 p.m. on August 3, 1990, where an update of the the evaluation team activities and findings was presented. TN plant 1 evaluation manager also presented a summary of his review which indicated that i there were no lessons learned resulting from the review of the July 26, 1990, SI event which cou~i have prevented the July 30, 1990, SI event. The plant evaluation manager tid indicate that a general comparison could be made with j the Vogtle loss of vital AC power event (NUREG-1410) insofar as the plant l appe&rs to be more susceptible to events (such as e safety injection event) !

when it is shutdown and extensive maintenance is being conducted. The licensee '

determined that the oot cause of the event was a less than adequate plant impact review for ope ations in hot standby with the condenser isolated.

<

!

A special NRC inspect.6 .3 Ge July 30,19F, SI event was initiated on July 31, 1990. The inspectors conducted an independent review of the event and i the licensee's investigation of the event. No significant observations were made by the inspectors that had not been or were not planned to be evaluated by the licensee's team. However, the inspectors did note that an evaluation of the pressurizer heatup was not documented in the licensee's initial draft evaluation report. Discussions with licensee personnel revealed, however, that an evaluation of the pressurizer heatup had been conducted but had not been i documente The licensee was still investigating this apparent oversight at the end of the inspection. The inspectors will monitor licensee action in this area. These inspections will be tracked by an inspector followup item (IFI 445/9034-07).

The licensee completed the following actions prior to entering Mode 3:

Initial post-emergency safety features actuation review (ODA-108).

Revised Integrated Plant Operating Procedure IP0-007A to provide additional guidance and actions for maintaining the plant in hot standb Tested, disassembled, inspected, and retested the No. 3 AR *

Ensured that all AC essential lighting throughout the plant was energize *

Revised emergency operating procedures to include explicit guidance for the expeditious removal of personnel from containmen l

,

'

Staged job aids at the normal personnel air lock and the emergency air !

lock to facilitate operatio !

l l

i

o.* (.

-17-

Repaired the broken light fixtures in the No. 3 ARV room and initiated Work Order C90-5056 to repair insulation damag *

Calibrated the narrow-range level channels for the No. 3 S *

Verified that the compensated main steam line low pressure SI actuation circuitry functioned as designe *

Cleaned the downstream vent line orifices for all ARV The licensee had not issued its final evaluation report at the conclusion of the inspection, but a review of planned actions by the inspectors determined that a thorough evaluation was being conducted by the licensee. The final report by the licensee's evaluation team was expected to be issued by August 17, 1990. One significant item to be addressed by the licensee is the implementation of a design modification to remove AC essential lighting light switches. All the other items were evaluated by the licensee end completed or planned actions appeared appropriat . CONCLUSIONS The following conclusions were reached as a result of the special team inspection of this event and licensee action i

The licensee's investigation of this event was prompt .1d thoroug l

'

Items identified by the licensee to be completed prior to reentry into Mode 3 and for further evaluation were appropriat *

Operator response to the event and use of emergency operating procedures

, was goo Three areas noted where training and procedure improvements

' were appropriate were providing guidance for the expeditious removal of I

personnel from containment, using ARV upstream isolation valves in order to isolate a stuck open ARV, and operation of manual valves that are *

difficult or time consuming to operat *

The licensee had developed a thorough understanding of the sequence of events associated with the SI, as well as plant and eqttpment respons One exception, however, was that there was no conclusive explanation as to how the No. 3 ARV was open to approximately midposition by its manual handwheel operato ,

Appropriate notifications were made to NRC regarding this even *

The plant and available equipment responded as expected during the event with the exception of equipment identified in Section *

Followup inspections should be performed to review the appropriateness of licensee ongoing actions from their evaluation team findings. In particular, this should include a review of the desing modification to l remove AC essential lighing local light switches.

_

os * .

-18-In addition, the inspection team concurred with the following licensee conclusion *

The initiating cause of this event was the initial pressure transient that occurred when the No. 3 ARV came off of its closed seat. This same pressure spike also induced a reactor trip signal on two of four SG 1evel transmitters on lo-lo leve * The compensated main steam line low pressure SI tctuation circuitry performed in accordance with its design. The inspectors concluded that, on the bases of this SI logic design, this event could o: cur at other Westinghouse facilities if a similar erratic operation of an ARV occurred with the plant in hot standby condition *

The erratic perform:nce of the No. 3 ARV could be attributed to the !

presence of water on the upstream side of the ARV, and this source or l water was from condensed main steam that had backed up because the MSIV I upstream drippot was isolate *

The damage in the No. 3 ARV room was caused by a steam and water mixture that blew back down the ARV vent stac *

The installed emergency lighting in the area of the No. 3 ARV was in accordance with the emergency lighting system design and that the design met applicable regulatory requirements. However, short-term and long-term corrective actions were needed in order to ensure that AC essential lighting will be available in the area of the No. 3 ARV and ot.ker areas of the plan *

The hot standby operating procedure was inadequate for *.he plant conditions that existed at the time of the even . EXIT MEETING (30703)

An exit meeting was conducted on August 7, 1990, with the licensee's representatives identified in paragraph 1 of this report. The licensee did not identify as proprietary any of the materials provided to, or reviewed by, the inspectors during this inspection. During this meeting, j the inspectors summarized the scope and findings of-the inspection.

l

!

i

. -