ML20205B199
| ML20205B199 | |
| Person / Time | |
|---|---|
| Site: | Paducah Gaseous Diffusion Plant |
| Issue date: | 03/25/1999 |
| From: | Pulley H UNITED STATES ENRICHMENT CORP. (USEC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GDP-99-1007, NUDOCS 9903310171 | |
| Download: ML20205B199 (7) | |
Text
.
.USEC A Global Energy Company March 25,1999 GDP 99-1007 United States Nuclear Regulatory Commission Attention: Document Control Desk Washington, D.C. 20555-0001 Paducah Gaseous Diffusion Plant (PGDP)
Docket No. 70-7001 Event Report ER-99-01 Pursuant to 10CFR76.120(d)(2), enclosed is the required 30-day written event report covering the safety system failure of the UF Release Detection System in C-333, Unit 4, Cell 10. The Nuclear 6
Regulatory Commission Headquarters (NRC-HQ) operations office was notified of the event on February 26,1999 (NRC No. 35409). Commitments contained in this submittal are identified in.
Any questions regarding this matter should be directed to Larry Jackson at (502) 441-6796.
Sincerely, owar uley General Manager Paducah Gaseous Diffusion Plant
Enclosure:
As Stated
'/
cc: NRC Region Ill Office NRC Resident inspector-PGDP
[$
]
~
~
9903310171 990325 PDR ADOCK 07007001 C
PDR,
~
P.O. Box 1410, Paducah, KY 42001 Telephone 502-441-5803 Fax 502-441-5801 http://www.usec.com Offices in Livermore, CA Paducah, KY Portsmouth, OH Tashington, DC
Docket No. 70-7001 GDP 99-1007 Page1of5 EVENT REPORT ER-99-01 BACKGROUND-The UF Release Detection System heads are required to be " test fired" twice each shift when 6
operating in Cascade Mode 2 (above atmospheric pressure) according to Technical Safety Requirement (TSR) surveillance SR 2.4.4.1-1. Detector heads are fired by supplying voltage to the head which results in an alarm in the Area Control Room (ACR). Firing the heads maintains sensitivity of the heads in elevated temperatures, such as that found in cell housings. Additionally, the UF Release Detection System heads are physically tested annually by applying test smoke near 6
the heads and verifying an alarm is received in the ACR (SR 2.4.4.1-2).
To manually activate the test firing, the operator depresses a test push-buttoa at a cell signal conditioner panel until alarm lights for all detectors are illuminated and verifies that the alann is mceived in the ACR. De operator then msets all the detectors for a cell by pressing the reset button -
on the signal conditioner until the alarm lights have cleared. Depressing the reset button momentarily removes the 200 VDC to the detectors, allowing the detectors to reset.
He signal conditioner has two different modes of operation, "nonnal" and " override." In " normal" mode, tests can either be initiated by an operator as described above or can be performed under computer softwam control. If testing is being performed by the computer, a yellow " computer test" light is illuminated.' During the testing by the computer, varying ramp voltages are applied to the detector head until the detector head fires. The ramping of all the detector heads in a cell normally
{
takes less than two minutes. While the computer is ramping the heads, the alarm in the ACR for the UF Release Detection System is disabled. Once mmping is complete and all detectors have fired,-
6 the computer resets the detectors. If the computer is ramping the detectors as indicated by the yellow " computer test" light being illuminated, switching from " normal" mode to " override" mode will remove computer control, thus disconnecting the ramp voltage from the detectors and reenabling the ACR annunciator.
. DESCRIPTION OF EVENT l
On Febmary 26,1999, at appmximately 0845 hours0.00978 days <br />0.235 hours <br />0.0014 weeks <br />3.215225e-4 months <br />, while performing TSR surveillance SR 2.4.4.1-1 on the UF Release Detection System in C-333 building, it was discovered that the detector heads 6
on Unit 4 Cell 10 would not fire. The operator observed that the yellow " computer test" light was
-illuminated on the signal conditioner. The operator performing the surveillance switched the operation mode of the signal conditioner fmm " normal" to " override" mode, as instructed by the UF.
i Release Detection System surveillance procedure (to be done when the computer test light is
4
(
Docket No. 70-7001 GDP 99-1007 Page 2 of 5
' illuminated for over three minutes). He then successfully test fired the heads. Based on a further review ofthe surveillance procedum with two of his managers, the operator was instructed to switch the mode of operation from the " override" mode back to the " normal" mode. This caused the UF.
' Release' Detection System for Unit 4 Cell 10 to be inoperable Approximately 90 minutes later based on further technical discussions related to the system configuration, the inoperability of the system was malized and was verified by a smoke test of a detector head. The mode of operation was then switched back to " override," retuming the system to an operable condition. Operability was verified by a smoke test and manually test firing the detector heads.
~ The UF Release Detection System is required to be operable when operating in Cascade Mode 2, 6
as requimd by TSR Limiting Conditions for Operation 2.4.4.1. At the time of the failure, C-333 Unit 4 Cell 10 was ~ operating in Mode 2. Pursuant to 10 CFR 76.120(c)(2), the Nuclear Regulatory Commission Headquarters operations oflice was notified of this event on February 26,1999, at 1400 hours0.0162 days <br />0.389 hours <br />0.00231 weeks <br />5.327e-4 months <br />. Event Notification Worksheet No. 35409 was assigned to this event.
CAUSES OF EVENT A.
Direct Cause The initial inoperability of the UF Release Detection System was discovered during the 6
operator surveillance when the Unit 4 Cell 10 detector heads would not fire. The length of time the system was inoperable is unknown; however, Unit 4 Cell 10 detector heads had passed their surveillance less than eight hours earlier. This inoperability was caused by a failure of the digital output card for the Cascade-ADP scanner located at the UF Release 6
Detection System signal conditioner. Troubleshooting revealed that the card failed resulting in closure of the contact which energizes the "K-6" relay. Energizing the t'K-6" relay. caused the yellow" computer test" light to be lit and the UF Release Detection System annunciator 6
in the ACR to be disabled. Dem was no actual computer test in progress. The failure of the card gave a false indication that a computer test was occurring. This is not easily detected
.because the indicator light is on the local panel and not observable from the ACR.
Additionally, the contact on the digital output card which allows current to the "K-4" relay also failed closed. With the "K-6" relay energized, this closed contact caused the "K-4" relay to energize. This put the signal conditioner in a continual reset condition, removing
' the 200 VDC fmm the detectors. Therefore, the combination of these failures on the digital output card while in " normal" mode disabled the ACR annuncia r alarm and removed the 200 VDC necessary for the detectors to perform their intended function.
He second period ofinoperability of the UF Release Detection System occurred when the operator switched the mode of operation frorn the " override" to the " normal" mode. While the signal conditioner was in the " override" mode, the failures on the digital output card had 4
t.
r.
i l
Docket No. 70-7001 GDP 99-1007 Page 3 of 5 no impact on system operability, because " override" mode switches out the Cascade-ADP scanner. Retuming the system to the " normal" mode reconnected the Cascade-ADP scanner, resulting in the Unit 4 Cell 10 UF. detectors end the associated ACR annunciators alarms
' being disabled, as well as the loss of the 200 VDC again.
l l
. B.
- Root Cause The root cause of the failure of the digital output card could not be determined. It is unknown how long this digital output card had been in service. %cussions with Operations and Maintenance indicate that this type of failure is not typical. There is no evidence of any transients or evolutions occurring prior to the event that could have led to the failure of the digital output card.
The second period ofinoperability, which occurred when the operator switched the mode of operation from " override" to " normal," was due to the Operations personnel not having enough system knowledge io realize that the system was inoperable when in " normal" mode with the yellow " computer test" light lit. The yellow " computer test" light being illuminated in " normal" mode typically indicates thM the computer is performing a test which ramps the detectors and temporarily disables the UF. Release Detection System annunciator in the ACR during computer testing.
= Although both the operators and management involved knew the heads did not fire earlier when the system was in " normal" mode with the yellow " computer test" light lit, they o returned the system to this mode, based on their review of the surveillance procedure. The procedure instructs the operator: "If the yellow ' comp test' light on the cell signal conditioner is illuminated for more than 3 minutes, place the mode of operation switch in
' override' and notify first-line manager." The next step of the procedure instructs the operator to verify that the mode of operations switch is in " normal," before testing the
' heads. The procedure provides no information that the computer test temporarily renders the system inoperable.
Although the procedure may have been misleading, the lack of system knowledge prevented the Operations personnel from overcoming the procedure's shortcomings. Once it was
)
~ known that the heads would not test fire in " normal" mode and would test fire in the "ovenide" mode, additional discussion and problem resolution should have taken place prior to placing the system back in a known failed mode, regardless of the procedure. Review of
- the Operations training module on UF. detectors indicated that the module did not discuss the inoperability of the system during the computer test.
The immediate response to the incident was to change the surveillance procedure to require the system to be routinely maintained in " override" mode, eliminating the potential problems
,o tu 4
I Docket No. 70-7001 GDP 99-1007 Page 4 of 5 that can be created by the computer interface. Additionally, all signal conditioners for the UF Release Detection System in cells operating above atmosphere were set to " override" 6
mode per the revised surveillance procedure and operators were briefed on the changes.
CORRECTIVE ACTIONS 4
- A.. Corrective Actions Taken i
1.
On February 26,1999, CP4-CO-CN6020t,"TSR Surveillance - Test Firing of Pyr-A-Larm, High Voltage UF Detection Systems in C-331/333/335/337," was revised to routinely 6
maintain the signal conditioner in "over:ide" mode.
2.
On February 26,1999, High Voltage.UF Detection Systems in C-333 operating above 6
atmosphere were set to " override" mode and operators and managers in C-333 were briefed of the changes via the tumover process.
3.
On March 4,1999, a " lessons learned" on this incident was placed in 30-day required reading for review by all cascade operations personnel.
4.
On March 4,1999, troubleshooting of the system was performed by Computer Maintenance and the failed digital output card was replaced. After switching the system to " normal" mode, a smoke test was performed and all detector heads were manually test fired. All results were satisfactory. The system was then placed in " override" mode, the heads were test fired, and the system was retumed to an operable status.
' B.
Corrective Actions Planned.
. 1.
By June 16,1999, Operations will revise CP4-CO-CN6020t, "TSR Surveillance - Test j
Firing of Pyr-A-Larm, High Voltage UF Detection Systems in C-331/333/335/337," to 6
add verbiage to indicate the impact of the computer test mode on system operability.
I Affected operators will review the revised procedure via required reading by the same j
date.
~ 2.
By July 15,1999, Training will revise training module, "UF Detectors," to incorporate 6
" lessons learned" from this event to ensure Operations personnel are aware the UF 6 Release Detection System is inoperable during the " computer test" mode and may be inoperable any time the yellow " comp test" light is illuminated.
3.
By November 17,1999, Operations will review the training modules / procedures for UF6
~ leak detection for the UF Handling Area (C-310, C-315, C-333-A, C-337-A, C-360) for i
6 a
(.
i i
I Docket No. 70-7001 GDP 99-1007 Page 5 of 5 adequacy in providing guidance to personnel for determining the operability of the system based on system indications (indicating lights, alanns, etc.). Based on reviews of the TSR, this was the only system in UF Handling identified as potentially having issues similar to 6
those identified in Event Report ER-99-01. Additional actions will be input into BPS for correction of any identified deficiencies.
4.
By November 17,1999, Operations will review the training modules / procedures for UF6 leak detection, criticality accident alarm system, and freezer /sublimers for adequacy in providing guidance to personnel for determining the operability of the system based on
)
system indications (indicating lights, alarms, etc.). Based on review of the TSR, these
{
were the only systems in Cascade identified as potentially having issues similar to those identified in Event Report ER-99-01. Additional actions will be input into BPS for correction of any identified deficiencies.
EXTENT OF EXPOSURE OF INDIVIDUALS TO RADIATION OR TO RADIOACTIVE MATERIALS No release of radioactive material occurred during the inoperability of the UF Release Detection System.
LESSONS LEARNED Personnel should be aware that the UF Release Detection System is inoperable durin.g the 6
" computer test" which ramps the detector heads and may be inoperable any time the yellow " comp test" is illuminated. Additionally, personnel should ensure they fully understand the operability considerations before placing a system in a mode that previously did not pass a surveillance.
4 1
Docket No. 70-7001 GDP 99-1007 Page1of1
. List of Commitments Event Report ER-99-01 1.
By June 16,1999, Operations will revise CP4-CO-CN6020t, "TSR Surveillance - Test Firing ofPyr-A-Imm, High Voltage UF Detection Systems in C-331/333/335/337," to add verbiage to indicate the impact of the computer test mode on system operability. Affected operators will review the revised procedure via required reading by the same date.
2.
By July 15, 1999, Training will revise training module, "UF Detectors," to incorporate
" lessons leamed" from this event to ensure Operations personnel are aware the UF Release Detection System is inoperable during the " computer test" mode and may be inoperable any time the yellow " comp test" light is illuminated.
3.
By November 17,1999, Operations will review the training modules / procedures for UF leak detection for the UF. Handling Area (C-310, C-315, C-333-A, C-337-A, C-360) for adequacy in providing guidance to personnel for determining the operability of the system based on.
system indications (indicating lights, alarms, etc.). Based on reviews of the TSR, this was the only system in UF. Handling identified as potentially having issues similar to those identified in Event Report ER-99-01. Additional actions will be input into BPS for correction of any identified deficiencies.
4.
By November 17,1999, Operations will review the training modules / procedures for UF. leak detection, criticality accident alarm system, and freezer /sublimers for adequacy in providing guidance to personnel for determining the operability of the system based on system indications (indicating lights, alarms, etc.). Based on review of the TSR, these were the only
- systems in Cascade identified as potentially having issues similar to those identified in Event Report ER-99-01. Additional actions will be input into BPS for correction of any identified deficiencies.
a