ML20045H882
| ML20045H882 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 07/16/1993 |
| From: | Simpkin T COMMONWEALTH EDISON CO. |
| To: | Murley T NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM), Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9307220067 | |
| Download: ML20045H882 (4) | |
Text
"
- Ccmminwtalth Edison g
^
O '/ L 1400 Opus Place Downero Grove, Illinois 60515 July 16, 1993 Dr. Thomas E. Murley, Director Office of. Nuclear Reactor Regulation
'U.S.
Nuclear Regulatory Commission Washington, D.C.
20555 Attention: Do'cument Control' Desk Subj ect :
Zion Station Unit 1 Eagle 21 Process Protection System Periodic System Performance Report NRC Docket No. 50-295
References:
(a)
April 10, 1992 letter from S.F. Stimac to T.E. Murley
Dear Dr. Murley:
Commonwealth Edison Company committed via reference (a) to provide NRC periodic performance reports related to the Zion Unit 1 Eagle 21 Process Protection System.
Pursuant to this commitment, please find enclosed o.a copy the subject report _for the interval of January 16, 1993 through May 15, 1993.
As described in reference (a), additional performance reports will be submitted periodically throughout the first Unit 1 operating cycle with Eagle 21.
Please direct any questions you may have to this office.
Respectfully, T'
T.W. Simp in Nuclear Licensing Administrator TWS/gp Enclosure cc:
Regional Administrator - RIII C.Y. Shiraki, Project Manager - NRR J.D. Smith, Senior Resident Inspector Zion Office of Nuclear. Facility Safety - IDNS 5
k t \\ni n \\11.2 - nre:. wp f d
l!
9307220067 930716
'\\~
ADDCK 05000295.{Ej i
PDR p
PDR gg
= ' *.
j i
ZION STATION UNIT 1 EAGLE 21 SYSTEM PERFORMANCE REPORT L
JANUARY 16,1993-MAY 15,1993 During the period from January 16,1993, to May 15,1993, Unit I has maintained mode 1 operation. Within this time period the following events related to Eagle 21 have occurred.
On February 11,1993, a momentary voltage degradation on th : 111 Instrument Inverter caused the Loop Calculation Processor (LCP) and Test Control Processor (TSP) single board computers in Protection Set I, rack 3, to halt processing. All indication provided by the rack froze in the position being indicated at when the LCP lock-up occurred, and all channels tripped to the fail safe state.
Action Taken: The Eagle 21 rack wiring was verified to not be causing a fault on the inverter.
since no failed components could be identified within the rack, the rack 3 LCP -
reset button was depressed. The rack boot-up cycle, test panel indication, power supply voltages, and dynamic information all appeared normal.
Operating performed a Periodic Test (PT)-0, Appendix L (channel cross check) for the affected chtnnels and returned the rack to service with no further problems. This problem rendered the rack inoperable for a short period of time. However, the minimum requirements for channel redundancy were not compromised.
Root Cause:
The unanticipated voltage degradation experier.ced on instrument inverter 111 caused the rack 3 Primary and Secondary power. supplies to degrade in I
voltage. The Eagle 21 LCP and TSP Multibus circuit boards are both Inte:
286/12 single board computers, which require a certain amount of voltage l
within a specific time frame to maintain processing. If supply voltage to these l
boards is not sufficient, they will seize up until supply voltage is above this threshold, and the boards are reset (reset button on front of Eagle 21 rack).
Therefore, the Eagle 21 equipment functioned as designed.
The processor lock-ups did not occur on racks 1 or 2, because the voltage levels are dependent on rack loading. Since rack loads are somewhat dynamic, j
this could have occurred on one or both of the other racks, depending on the length and severity of the inverter degradation.
1 i
i
{....
j s
On April 2,1993, the Protection Set 111, rack 12, Protection Channel Trouble alarm actuated.
An inspection of the rack revealed that the Secondary power supply 15Vdc Light Emitting Diode (LED) was not illuminating, which was indicative of a power supply failure.
Action Taken: A power supply failure was verined by taking voltage measurements, and by checking the error code sent to the Man Machine Interface (MMI); Once the failure was conGrmed, the power supply was replaced, and the appropriate power supply adjustments and voltage verincations were performed.
Root Cause: The cause for this power supply failure is presently unknown. The failed power supply was sent back to Westinghouse for repair and failure analysis.
Failure analysis results will be provided in the next Eagle 21 Performance Report after the results are available.
Since reaching power operation on Unit 1, spurious " Trouble" alarms have been received for various protection sets. These alarms are not concurrent with any type of plant transients or hardware failures. The alarms appear to be random and last for only 100ms.
Initially, the problem occurred on Protection Set II, on November 16,1992, and then again on December 2,1992, but no cause could be found. The alarms were too shon in duration and 7
occurred too infrequently to pinpoint the problem. After the December 2nd event the problem seemed to disappear. However, on March 1,1993 the spurious alarms recurred, but this time more frequently, and on multiple protection sets.
These " Trouble" alarms are controlled by the Tester Subsystem, therefore, the protective functions of Eagle 21 are unaffected by this problem.
Action Taken: Since Protection Set I appeared to be generating the spurious alarms more frequently, the investigation has thus far concentrated on this protection set.
The three Eagle ra:ks within Protection Set I were monitored using digital data acqu sition equipment. Monitoring took place for several weeks at different points in the racks. The results of this trouble-shooting showed that the alarms were being generated by rack 2, and confirmed that the Tester Subsystem was actually requesting an alarm. These results eliminated the possibility of the spurious alarms being generated by the Annunciator l
System or the Eagle Analog Output board.
The Eagle Tester Subsystem is designed to generate an error code whenever an alarm condition is sensed. This code is sent to the Man Machine Interface (MMI) to aid in diagnosing system problems. However, since the spurious alarms clear so rapidly in this case, the MMI cannot be used to read the error codes.
In order to capture this error code a minor char.ge was made to the TSP l
software by Westinghouse. Essentially, a write command was added to the 1
l
_. - ~ _ -. _ _ _,
=
j. * -
4 software to send any error codes generated to a spare output port on the TSP board. A serial printer was then interfaced with this port to print any error codes generated.
t Westinghouse installed the write command on spare TSP EPROM's anc' c
performed validation testing. Upon successful completion of validation testir.g the EPROM's were shipped to CECO for installation under an approved Temporary Alteration and 50.59 Safety Evaluation.
On April 23,1993, a spurious alarm occurred and an error code was captured.
The error code corresponded to an LCPffSP communications failure. The Data Link Handler was then replaced because it provides the com nunication link between the LCP and the TSP. As of 5/10/93 no further alarms have actuated on Protection Set I.
Similar trouble-shooting efforts are underway on alternate protection sets.
'j An update of this problem will be provided in the next Eagle 21 System i
Performance Report.
Root Cause:
The Data Link Handler board has been sent back to Westinghouse for a root cause analysis. An update will be pro.rided in_the next Eagle 21 System Performance Report after the board is returned from Westinghouse.
As discussed in the first Eagle 21 Performance Report (Sept. 22,1992, ZRAD-92-091, page 2 of 9), there were seven Eagle Partial Trip (EPT) board failures involving blown fuses in the EPT board output stage for which no failure analysis had been obtained from Westinghouse.
Failure analysis results have since been received from Westinghouse for these EPT boards.
The results of this analysis showed that all, but one board, had no problems found. This would indicate that the fuses blew by having the output shoted during test equipment hook-up at the factory or at the site during installation.
One EPT board was found to have a mechanical problem with the fuse clamp for one channel.'
i This fuse clamp had one end stretched out, not allowing the clamp to securely hold the fuse, and thus creating an unstable conduction path, Action Taken: The fuse clamp was realigned to hold the fuse securely. Subsequent factory testing showed acceptable performance. The board was shipped back to CECO to be used as a spare.
Root Cause:
The cause for the fuse clamp misalignment is unknown, but was most likely caused during board fabrication, installation, or shipping. This problem has not recurred since Eagle 21 installation activities have been completed.
. -