HL-1967, Special Rept 1-91-010:on 911121,discovered That Blue Pen of post-accident Effluent Monitoring Sys Recorder 1D11-R631 Spiking Upscale at Regular Interval.Caused by Failed Flow Transducer.Flow Transducer Replaced

Special Rept 1-91-010:on 911121,discovered That Blue Pen of post-accident Effluent Monitoring Sys Recorder 1D11-R631 Spiking Upscale at Regular Interval.Caused by Failed Flow Transducer.Flow Transducer Replaced

ML20086J015
Person / Time
Site:	Hatch
Issue date:	12/05/1991
From:	Beckham J GEORGIA POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
1-91-010, 1-91-10, HL-1967, NUDOCS 9112100247
Download: ML20086J015 (5)

Text

cwga Nw cre; any 40 lrN0f fWSS Ci'"tOf PMh'VW/

! n,st ow su u>n o . Batmr r.;*em. Nahyrue 3'; et i

1.,pi.,# cm m m9 m

a u a m..m Georgia Power vo rmsew Nam , , , , ,

Hatn Pacgt '

HL-1967 002648 December 5, 1991 U.S. Nuclear Regulatory Commission ATIN: Document Control Desk Washington, D.C. 2t.,555

, PLANT HATCH - UNIT 1 NRC DOCKET 50-321 OPERATING LICENSE DPR-57 SPECIAL REPORT 1-91-010 MAIN STACK POST-ACCIDENT RANGE MONITORING SYSTEM TN0PERABLE FOR A PERIOD GREATER THAN 7 D M Gentlemen:

In accordance with Plant i;atch Unit 1 Technical Specifications Table 3.2-11 and Unit 2 Technical SP ecificat uns Table 3.3.6.4-1, Georgia Power Company is submitting the enclused Spe:iai Report concerning tt, main stack post accident effluent monitoring syste... '. hich was inoperable for a period greater than 7 days.

Sincerely,

[

g'fd. T. Beckham, Jr, OCV/cr

Enclosure:

Special Report 1-91-010 cc: .0fornia Power Compa_y Mr. H. L. Sumner, General Manager - Nuclear Plant NORMS U.S. Nuclear Reaulatcry Commission. Washinator. D.C.

Mr. K. Jabbour, licensing Project Manager - Hatch ll.S. Nuclear Regulatory Commission. Realon II I . S. D Ebneter, Regional Administrator Mr. L. D. Wert, Senior Resident Inspector - Hatch 9112100247 911205 / 2.

ges Anoct oscop,y1 I

g. -

/ .

i L. l

.. l ENCLOSURE PLANT HATCH - UNIT 1 NRC DOCKET 50-321 OPERATING LICENSE OPR-57 SPECIAL REPORT l-91-010 MAIN STACK POST-ACCIDENT RANGE MONITORING SYSTEM IN0PERABLE FOR GREATER THAN 7 DAYS A. REQUIREMENT FOR REPORI This_ report is required ner Unit 1 Technical Specifications Table 3.2-11,

-Footnote (g) and Unit 2 Technical Specifications Table 3.3.6.4-1, footnote (b). Specifically, the noted specifications require that in the event the Main Stack Post-Accident Effluent Monitoring System is inoperable for greater than seven days a Special Report is required to be submitted within 14 days of the_ event. In this event, the Main Stack Post-Accident Effluent Monitoring System wes inoperable for greater than seven days, thus, this Special Report is required.

B. UNIT STATUS Al TIME =OF EVENT On 11/21/91, at_2215 CST, Unit I was in Cold Shutdown and Unit 2 was in the

_Run mde at 2436 CHWT (100 percent of rated thermal power).

C. DESCRIPTION OF EVENT On 11/21/91, at 2215 CST, during a _ surveillance of Main Control Regn instrument panels, a licensed operator noted _that the blue pen of Post-Accident Effluent Monitoring System recorder 1Dll-R631 was spiking upscale ' at a _ regular interval. The blue pen provides - indication and recording capability for the Main Stack radioactive gaseous effluents. The Main - Stack itself_ provides an elevated release path for both Unit 1 and Unit 2 radioactive gaseous effluents.

-The Main' Stack Post-Accident Effluent Monitoring System was subsequently declared inoperabic. The aforementioned Unit 1 and Unit 2 Technical Specifications foo_tnotes require that the ystem'be returned to_ service within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> or the pre-planned alternate method for monio:. ring the Main Stack post-accident gaseous effluents be initiated end that a Srecial Report be submitted to the NRC if the system is out of service for greater than seven days. Limiting Cor.ditions for Operation (LCO) 1-91-783 and 2-91-762

were init:ated to ensure compliance with the specifications.

Instrument _ and Controls (l&C) technicians investigated the condition and found two problems': 1) a failed flow transducer which monitors sample flow to the system detectors and alarms on a low flow condition, and 2) the system computer being unresponsive to keyboard input. The computer was shutdown and then restarted. Several of the alarms cleared and within a i.

002648 HL-1967 E-1

s ENCLOSURE (Continued)

SPECIAL REPORT 1-91-010 MAIN STACK POST-ACCIDENT RANGE MONITORING SYSTEM IN0PERABLE FOR GREATER THAN 7 DAYS short period of time the computer anomaly cleared itself and responded to keyboard input. The computer was subsequently reprogrammed and returned to service on.11/23/91. The spiking phenomena did not recur when the syrtem was placed into service. The system was still considered inoperable because s repair of the flow transducer was awaiting replacement parts. Consequently, it was decided to observe the system over the next several days to ensure the spiking phenomena did not recur.

Seventy-two hours into the event, the system had not been returned to operable status. Consequently, the pre-planned alternate method for ,

monitoring the Main Stack effluents was initiated. The pre-planned alternato method for monitoring gaseous releases via the Main Stack consists of using the Main Stack Normal Range Monitoring System in conjunction with performing grab samples. Specifically, the normal range monitoring system is used until it indicates offscale at which time periodic grab samples are to be taken and analyzed for activity. Thus, at 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> into the event, the normal range monitoring system had been verified as being operable.

A replacement flow transducer was not readily available since the particular model in use was no longer being manufactured. A spare transducer was located at anather utility and provisions were made to send it overnight to the site via a commerciel shipping company. Unfortunately, the part was lost in shipment and, thus, a second transducer had to be obtained ftom the same utility. The part arrived on site the morning of 11/28/91. The transducer was replaced and satisfactorily furctionally tested per procedure 57SV-Dil-O'8 OS, "Kaman Accident Range Monitor FT&c." The system was then returned to operable status at 2205 . CST, on 11/28/91 - ten minutes before '

the_seven day LC0 expired. The LCOs were subsequently terminated. Since it was apparent that the system had not been inoperable for greater than seven days, a Special Report was not required.

Approximately three hours after the system was returned to operable status, however, the recorder began spiking again. The system again was removed from service and declared inoperable. Two new LCOs (LCOs 1-91-797 and 2-914-762) were initiated. Since the spiking problem which initially occurred on 11/21/91 apparently had not been repaired, the LCOs were bacedated to 11/21/91 which meant the system had been inoperable for greater than seven days, requiring the issuance of a Special Report.

I&C technicians again investigated the cor.dition and'found a bad computer central processing-unit _(CPU) board. The CPU board was replaced and the system was returned to service. A functional test was satisfactorily

, completed at which time the system was declared operable. The LCOs were subsequently terminated at 1600 CST, on 11/30/91.

002648 HL-1957 E-2

.t-ENCLOSURE (Continued)

SPECIAL REPORT 1-91-010 MAIN STACK POST-ACCIDENT RANGE MONITORING SYSTEM IMfERABLE FOR GREATER THAN 7 DAYS D. CAUSE OF EVENT The cause of the event was failed components. Specifically, the system flow transducer -and the system computer CPU board failed. The flow transducer failure resulted in the system low flow alarm being inoperable. The CPU board failure was intermittent resulting in the spiking condition randomly appearing and disappearing. As a consequence of this failure made, plant personnel were led to believe that the system was functioning properly since the condition did not recur between 11/23/91 and 11/28/91. However, after the system was declared operable, on 11/29/91, the condition reappeared.

E. ANALYSIS OF EVENT The Main Stack Post-Accident Effluent Monitoring Sy:. tem provides information regarding radioactive gaseous release rates via the plant Main Stack in a post-accident Otuation, The system consists of two Geiger-Mueller detectors: one for measuring mid-range noble gas activity and one for measuring high range noble gas activity. The system is normally on standby and is automatically initiated when a prescribed gaseous radioactivity limit is exceeded as measured by the Main Stack Normal Range Monitoring System.

Upon- initiation, a- sample pump- starts and a continuous sample of Main Stack effluent . is routed to the post-accident system detectors. The gaseous radioactivity, as measured by the detectors, is then displayed on recorder IDll-R631 in the Main Control Room. The pre-pianned alternate method for monitoring gaseous release via the Main Stack consists of using the normal range system in conjunction with performing grab samples. Specifically, the normal range monitoring system is used until-it indicates offscale at which time periodic grab samples are to be taken and samples-analyzed for activity.

In this event, the Main Stack Post-Accident Effluent Monitoring System was

-inoperable for nine days. During this time, the Main Stack Normal Range Mnnitoring System was operable -and showed no unusual radioactive gaseous releases via the Main Stack. Had a Design Basis Event occurred and resulted in radioactive gaseous releases exceeding the range of the Main Stack Normal Range Monitoring System, grab sampling and analysis in accordance with procedure 64CH-SAM-005-0S, " Gaseous Ef fluents: Sampling," would have been initiated providing Main Control Room personnel with the necessary information regarding radioactive gaseous releases.

002648 HL-1567 E-3

ENCLOSURE (Continued)

SPECIAL REPORT l-91-010 MAIN STACK POST-ACCIDENT RANGE MONITORING SYSTEM INOPERABLE LOR GREATER THAN 7 DAYS Based on this information, it is concluded that this event had no adverse impact on nucleir safety. This analysis applies to all operating conditions, F. CORRECTIVE ACTIONS The pre-planned alternate method for monitoring the Main Stack ef fluent releases was initiated within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> of the monitoring system becoming inoperable.

The computer CPU board and the flow transducer were replaced.

lhe system was satisfactorily functionally tested per procedure 575V-Dil-018-0S and returned to operable status by 1600 CST, on 11/30/91.

I h

002648 HL-1967 E-4 ,

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