ML20217A517
| ML20217A517 | |
| Person / Time | |
|---|---|
| Site: | Paducah Gaseous Diffusion Plant |
| Issue date: | 03/19/1998 |
| From: | Polston S UNITED STATES ENRICHMENT CORP. (USEC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GDP-98-1017, NUDOCS 9803250096 | |
| Download: ML20217A517 (6) | |
Text
4 t
w
.L USEC A Glot,al Energy Company March 19,1998 GDP 98-1017 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-98-03, Rev.1 Pursuant to 10 CFR 76.120(d)(2), enclosed is the final report for the C-333 CAAS inoperability event, which resulted from the inadvertent continuous sounding of the C-333 building howlers. This report provides supplemental information for Event Report 98-03, which was issued to the Nuclear Regulatory Commission (NRC) on February 20,1998.' The supplemental information is designated by a revision bar in the right margin of the report. Enclosure 2 is a list of commitments made in this report.
Any questions regarding this matter should be directed to Larry Jackson at (502) 441-6796.
Sincerely,
(
be Polst General Manage Paducah Gaseous Diffusion Plant Enclosures cc:
NRC Region 111 Office
. NRC Resident inspector - PGDP 9803250096 900319 PDR ADOCK 07007001i C
\\
\\
P.O. Box 1410, Paducah, KY 42001 Telephone 502-441-5803 Fax 502-441-5801 http://www.usec.com g
' Omccs in Livermore, CA Paducah, KY Portsmouth, OH. Washington, DC k
N
.I Docket No. 70-7001 GDP 98-1017 Page1of4 EVENT REPORT ER-98-03, Rev.1
' DESCRIPTION OF EVENT On January 23,1998, at 0900 the C-333 area control room (ACR) attempted to signal operators in the building to' call the ACR by sounding a short blast on the building howlers. This is a routine
. method of signaling building personnel to come to, or call the ACR. When the ACR operator pressed the howler actuator button, the button stuck in the "on" position and the howler sounded continuously. The Plant Shift Superintendent (PSS) was immediately notified and electrical
' maintenance was called to respond to the building to stop the howler's sounding. The PSS became concerned about the audibility of the building Criticality Accident Alarm System (CAAS) and
' declared the building CAAS inoperable at 0917. The general Limiting Conditions for Operation (LCO) actions under Technical Safety Requircments (TSR) 2.4.4.2b and 2.2.4.3b were taken, as required.-
At 0925, Electrical Maintenance personnel silenced the howlers and at 0942 the CAAS was declared operable by the PSS and the LCO actions lifled. The building howler system is an "AQ" system.
Its safety function is to provide notification ofincidents requiring evacuation of personnel.
At approximately 1100 on January 23,1998, the PSS requested the CAAS system engineer to initiate an operability evaluation (OE) on the howler failure. This evaluation (No. OE-C-822-98-001) concluded that the continuous sounding of the building howler system provided a significant increase in the background sound pressure levels on the C-333 cell floor for approximately 25 minutes. During this period, the ability of the CAAS horns located on the C-333 cell floor to meet requirement 4.4.1 of ANSI /ANS-8.3-1986, would be significantly degraded and thus, the ability of
~ the system to perform its intended safety function (to alert building personnel of a criticality) would be marginal.- The OE stated that the ability of the C-333 ground floor CAAS horn system was not adversely affected by the cell floor building howler sounding. Ability of the CAAS to detect a radiation event and sound local horns was not affected by this event.
. On January 24,1998, at 0241 CST the PSS notified the NRC Headquarters operation office of the event in accordance'with 10CFR76.120(c)(2) and NRC No. 33591 was assigned to the event.
The process building howlers are used to initiate evacuations of building personnel. In the event of a criticality accident, the building howlers would be actuated by the building ACR operator to augment the audibility function of the CAAS local homs to ensure that all building personnel are alerted and can evacuate in a timely manner. The continuous sounding of the howlers is the "alen
. signal," which directs pemonnel to evacuate the building immediately and to listen to the plant public
f Docket No. 70-7001 GDP 98-1017
- Page 2 of 4 address system for further instructions. The operability of the howlers is verified at least semi-
- annually. During the investigation of this event, a search of the problem reporting data base was conducted to determine if similar events had occurred in the past. A problem report (PR-CO : 5034) described the same situation in C-333 on September 8,1997, where the cell floor howlers stuck "on" for approximately three hours. The proper LCOs were entered due to concerns about CAAS audibility, but the event was not reported to NRC, as required. This discovery was discussed with the NRC Resident. This issue will be corrected through the implementation of SCAQ Action Plan No. PR-CO-97-6836. This SCAQ addresses previous failures to make notifications to NRC.
During trouble shooting of the howler electrical system, a light bulb on the evacuation system Jelectrical control panel (attack and alert signaling system) had come loose from its socket and was 7 laying such as to make contact with the socket base and ground. This ground by itself would not cause the failure, since the system is isolated from ground.
Subsequent investigation and electrical troubleshooting confirmed a second major ground on the l
250 VDC system. This ground was traced to ajunction box located on the ground floor at Unit 2, l
Cell 6. A wire in thisjunction box was found with deteriorated insulation which allowed the wire l
to contact thejunction box. The wiring in thisjunction box is original to the plant and was installed l
in the 1950's. The insulation is of a braided cloth type, sometimes called varnished cambric. The l
inspection of thisjunction box noted that the inside showed signs of being wet in the past and some l
rust was present. Cell floor water leaks have occurred in the past, which when coupled with the age l
of the wiring, is the most probable cause of the wire insulation deteriorations.-
l l
The wire making contact with the box grounded the negative side of the 250 VDC system. The 250 l
' VDC system is center tapped, providing two sources of 125 VDC. This second ground, coupled l
1 with the previously mentioned ground, allowed an increased current demand on the howler push l
button switch. The current load was beyond the design capabilities of the switch and it could not i
break this current. The switch arced and welded the switch contacts closed, causing the continuous
]
. howler sounding. The switch was replaced and the deteriorated portion of the wire was removed and l
reterminated. This cleared the ground. This major ground is thought to have existed for some time l
and was at least a contributing cause of the unplanned howler sounding on September 8,1997. Cell l
6 has been out,
" since October 26,1990.
l I
i i
1 i.
I
.....m
..,,m.r...
4
'a e f
Docket No. 70-7001 GDP 98-1017 Page 3 of 4 l
1 CAUSE OF EVENT l-
'A. Direct Cause The direct cause is attributed to the welded contacts in the howler switch which occured from the arcing during excessive current load from the electrical ground. This failure caused the.
howlers to sound continuously for 25 minutes, which degraded the audibility function of the building CAAS.
l-I B. Root Cause -
-l I
The root cause of the howler failure is the lack of a method to detect and correct DC electrical j
system grounds in a timely manner. Currently, the only way of detecting DC grounds is by l
switching the " alarm" DC system to the " control" DC system and monitoring the ground l
detection indicator on the control DC system battery charger. This is known as the DC l
throwover test. This test is currently required to be conducted once a year and does not require l
the monitoring of the DC ground indicator on the battery charger. This test was conducted in i
August 1996 and in February 1998. This testing, as stated above, did not include a step to l
monitor the battery charger ground detector. Thus, any grounds in existence at the time were not l
detected nor corrected. The August 1997 test was delayed until February 1998, due to battery l
room problems.
l l
The DC throwover test is scheduled by plant procedure CP4-CO-CM6001, " Routine Tests and l
Inspection." Due to the simple nature of the test, the DC throwaver test is not proceduralized.-
l-The completion of the test is documented on the building single line PM schedule and building l
- log. Corrective Actions No.1 & 2 will ensure timely detection and correction of DC electrical I
grounds.
I l
Due to the relatively low significance of this event, the commitments made in this report are not l
- designed to prevent recurrence, but to minimize the affect of similar problems in the ft.ture.
l I
l C. Contributing Cause.
l-l A contributing cause of this event relates to the lack of a procedure requirement to inspect l
L affected electrical equipment following major cell floor water leaks. Operations procedu e CP4-i et CO-ON3002," Process Recirculating Cooling Water Failure," does not require the inspection of i affected electric equipment for water damage. Although not proceduralized, general practice has I
( been to check major process electrical transformer, switchgear, etc. Corrective Action No. 2 will l
! address this issue.
l
j 4
4 I
l
- Docket No. 70-7001 l
GDP 98-1017 -l Page 4 of 4 l
l l
PLANNED CORRECTIVE ACTIONS l
l
- 1. By June 30,1998, Operations will develop and implement a check sheet for the completion and I
documentation of DC throwover tests. The sheet should include requirements for checking of l the battery charger ground indicators for grounds and when grounds are found, initiating work - l requests for troubleshooting and repairs.
l l
- 2. By June 30,1998, Operations will revise procedure CP4-CO-ON3002," Process RCW Failure,"
l to require the inspection of potentially affected electrical equipment for signs of water damage l
and correct as necessary.
l l
l EXTENT OF EXPOSURE OF INDIVIDUALS TO RADIATION OR RADIOACTIVE l
MATERIALS l
I There were no exposures to individuals from this incident.
l l
l LESSONS LEARNED l
l Electrical grounds must be detected and corrected in a timely manner to preclude operational upsets.
l l
l
w l
/
~.,
Docket No. 70-7001
.l GDP_98-1017 l~
Page1ofI l
l I
Event Report ER-98-03, Rev.1 l
List of Commitments l
l l
I c
- 1. By June 30,1998, Operations will develop and implement a check sheet for the completion and l
documentation of DC throwover tests.- The sheet should include requirements for checking of l the battery charger ground indicators for grounds and when grounds are found, initiating work l
requests for troubleshooting and repairs.
l l
- 2. By June 30,1998, Operations will revise procedure CP4-CO-ON3002," Process RCW Failure,"
l to require the inspection of potentially affected electrical equipment for signs of water damage l
and correct as necessary.
l
(
..