ML20217H841
| ML20217H841 | |
| Person / Time | |
|---|---|
| Site: | Paducah Gaseous Diffusion Plant |
| Issue date: | 07/31/1997 |
| From: | Polston S UNITED STATES ENRICHMENT CORP. (USEC) |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| GDP-97-1022, NUDOCS 9708130286 | |
| Download: ML20217H841 (8) | |
Text
...... _......
j e
United States Ennchment Corporati<m Paducah Site Offwe P.O Ikn 1410 Padwah KY 42001 Tel 502 441-5H03 ist 502 441-5801
-3 July 31,1997 United States Nuclear Regulatory Commission SERIAL: GDP 97-1022 Attention: Document Control Desk Washington, DC 20555-001 Paducah Gaseous Diffusion Plant (PGDP)- Docket No. 70-7001 - Event Report ER-97-13 Pursuant to 10CFR 76.120(c)(2), Enclosure 1 is the required 30-day written Event Report covering ER-97-13. This event invohu the interruption, on two occasions, of direct current I
(DC) power to Technical Safety Requirement (TSR) required equipment. The Nuclear Regulatory Commission (NRC) was notified of the event on July 1,1997, and July 13,1997 (NRC No. 32571 and 32622). Enclosure 2 is a list of commitments made in this report.
Should you require further information on this subject, please contact Bill Sykes at (502) 441-6796.
Sincerely, A
eve olston A
General Manager Paducah Gaseous Diffusion Plant SP:WES:JLH:mcl
\\
\\
Enclosures (2) cc:
NRC Region til NRC Senior Resident inspector, PGDP s
j [J t) h
~~
3 uv i
9708130286 970731
~
PDR ADOCK 07007001 C
r-p I
k Docket No. 70-7001 Page1of5 EVENT REPORT ER 9713 BACKGROUND The re ction of uranium hexaflouride (UF.) and water (free atmospheric humidity) in the case of
/
a UF. release produces uranyl fluoride (UO2F2) as particulates, and hydrogen fluoride (HF) as gas which will hydrate. The UO2F2 and HF*x(H2) are highly visible as " smoke". The UF.
release detection system detects the presence of" smoke" which generates an alarm to alert operating personnel to initiate corrective / mitigative actions. The UF release detection system heads are " test fired" twice each shift if operating in mode 2 or 3 (when UF pressure is above atmospheric pressure). Additionally, a functional test by actuating the UF. release detection system heads with " smoke" is performed quarterly. The UF. detection system is used to monitor condensers, accumulators, and piping heated housings.
On July 1,1997, and again on July 13,1997, UF /uection system alarm audibility for building C-310 was lost for approximately five minutes and one minute, respectively; however, all the design safety detection functions were not affected during the DC power loss. Only the audibility and indication functions in the area control room (ACR) were inoperable. The building C-331 ACR operator notified the building C-310 ACR operator that he (building C-331
-s operator) had received an annunciator direct current (DC) power failure in building C-310. The building C-310 ACR operator tested his ACR alarms and received no response.
The Paducah Gaseous Diffusion Plant (PGDP) Technical Safety Requirements (TSR) Limiting Conditions For Operations (LCO), Section 2.3.4.4 states "the high voltage (old) UF. release detection heads monitoring the subject equipment shall be operable."
The temporary loss of alarming to the C-310 UF. detection system is reportable in accordance with 10CFR_76.120(c)(2). On July 1,1997, and July 13,1997, the Nuclear Regulatory Commission (NRC) headquarters was notified of the events (see NRC Event 32571 and 32622).
On May 18,1997, a similar event occurred when power was lost to the C-310 Criticality Accident Alarm System (CAAS) and UF. leak detection system. For additional information relative to this event, please see Event Report ER-97-10 (NRC Event 32381).
ewe e
+
- * - ~
Docket No. 70 7001 Page 2 of 5 QESCRIPTION OF EVENT (Reference NRC No. 32571)
On July 1,1997, at approximately 1635 the Building C-331 area control room (ACR) operator notified the C 310 ACR operator that he (building C-331 operator) had received an annunciator DC power failure alarm. The C-310 operator attempted an alann lamp test; however, there was no alarm power, At approximately 1637 one operator was sent to test fire the UF detector heads and another operator was sent to the equipment control center to manually trip the rectifier breaker to disengage the normal power to the rectifier on the alternating current (AC) electrical distribution panel A--1-4. By disengaging normal power to the rectifiers this caused the transfer switch to tluowover to emergency (battery) restoring DC power. At approximately 1640, after i
tripping the breaker on the AC distribution panel A-4-4, the C-310 ACR, DC alarm power and i
C-331 DC alarm was restored. UF. detector heads were tested and audible and visual alarms were received.
DESCRIPTION OF EVENT (Reference NRC No. 32662)
On July 13,1997, at approximately 1935 the C-331 ACR operator notified the C-310 ACR that he (building C-331 operator) had received an annunciator DC power failure alarm. At approximately 1936 the C-331 operator manually tripped the rectifier area control room breaker to disengage the normal power to the rectifiers on the AC electriest distribution panel A-4-4. By disengaging normal power to the rectifiers this caused the transfer switch to throwaver to emergency (battery) restoring DC power. Atler tripping the breaker on the AC distribution panel A-4-4, the C-310 ACR, DC alarm power and C-331 DC alarm was restored. UF. detector heads were tested and audible and visual alarms were received.
There are a pair of rectifiers (north and south) located in building C-310. The loss of one rectifier results in the loss of one leg of 250 volt (V) DC capability. Without the transfer (throwover) switch transferring properly to battery supply there is only 125 VDC on one half of the electrical distribution panel. The transfer switch did not function properly because of the existence of 125 VDC on one side of the transfer relay coil and the existence of a ground on the other side of the coil. The existence of the 125 VDC across the coil is sufficient to prevent the 250 VDC transfer relay from sensing a loss of voltage to the rectifier. This 125 VDC power is necessary to operate the C-310 UF detector annunciators which were rendered inoperable due to the loss of DC power. If UF detection had occurred the alarm would have annunciated when DC power was restored.
The following was determined during trouble-shooting:
A. The main contactor/ overloads for the north rectifier tripped; a ground on the positive leg of the alarm power 125/250 VDC was discovered; and low DC voltage was supplied from the north rectifier.
1
)
Docket No. 70 7001 Page 3 of 5 B. Contactor/ overloads were tested and results indicate that the component performed as designed.
C. The ground was found to be a high resistance ground and one path of the ground was located in the alarm relays for an out of service piece of equipment. Wiring was lifted on the alarm relays to eliminate the source of the ground, The primary component failure was the original selenium rectifier bridge, which is obsolete. A i
new three-phased diode bridge, with new wiring and associated parts was installed to replace the obsolete and degraded components.
A concein which was developed during this investigation was an apparent lack of understanding L
relative to what functions, such as CAAS howlers and UF detection alarms, are affected by the 6
t inoperability of the north, or south, rectifiers. An operator's aid would eliminate any confusion and probably enhance efficiency.
I During the DC power interruption the CAAS remained operational because electrical power continued from the south rectifier.
None of the actions directed by TSR, LCO 2.3.4.4 were implemented because power was restcred within the one hour time limit as prescribed by the LCO.
On July 14,1997, Engineering Service Order (ESO) Z98990 was submitted requesting that an engineering design be performed to replace the existing 250 VDC transfer switch relay with two 125 VDC relays; one sensing each rectifier leg to neut al(- 125 V to neutral and + 125 V to neutral). The existing transfer relay is incapable of reliably sensing the loss of one rectifier from the pair; therefore, half of the DC distribution power may be lost. The transfer relay should be.
capable of sensing the loss of partial power and transfer to battery.
CAUSE OF THE EVENT A. Direct Cause (NRC No. 32571,32622)
The direct cause of both events was the inability of the transfer switch relay to sense the loss of voltage from the rectifier.
B. Root Cause (NRC No. 32571,32622)
The root cause of both events was inadequate transfer switch relay design specifications. The design specifications failed to anticipate, as currently configured, voltage sensing problems l
~ -.... _ _ _ _ -
=.
^
D.ocket No. 70 7001 Page 4 of 5 during the loss of a rectifier. The existing transfer switch relay is incapable of reliably sensing the loss of one rectifier from the pair. The transfer relay switch should be capable of sensing the loss of partial power and transfer to battery.
C. Contributing Cause (NRC No. 32571,32662)
A contributing cause of both events was the corrective action recommended in Event ER-97-10 relative to preventive maintenance (PM) has not been implemented. These latest events occurred prior to the scheduled date for implementing corrective action No. 4 contained in ER-97-10.
CORRECTIVE ACTIONS A. Completed Corrective Actions
- 1. On July 14,1997, Engineering Service Order (ESO) Z98990 was submitted for approval.
- 2. On July 25,1997, Maintenance replaced the existing rectifier bridge on the north and south rectifiers in building C-310 with a three-phase diode bridge rectifier.
B. Previously Approved Corrective Actions (NRC No. 32381)
The following commitments were previously approved (NRC No. 32381) and are applicable to these two events (NRC No. 32571,32662):
- 1. By July 31,1997, Engineering will inspect, with the assistance of Electrical Maintenance, all DC power transfer systems in C-331, -333, -335,-337. The inspection will determine the current condition of the transfer relay and associated fuses to the rectifier of the DC power transfer systems and will identify deficiencies. A copy of the findings will be provided to Reliability Engineering.
- 2. By August 15,1997, Operations will ensure a system ofcommunication of preventive maintenance (PM) status / clear ownership and approval of PM deferral exists.
- 3. By August 29,1997, Operations will develop a schedule to correct deficiencies, as required, identified in corrective action No.1, above.
- 4. By September 30,1997, Reliability Engineering will evaluate (taking into account the -
results of corrective action No.1, above) and determine the PM requirements for the transfer relays in C-310, -331, -333, -335, and -337. This evaluation will also include y
determining the adequacy of the frequency of the DC power switch test.
- 5. By September 30,1997, Operations will issue an off-normal procedure for loss of DC power in C-310.
l
Docket No. 70-7001 Page 5 of 5
- 6. By September 30,1997 Operations will issue off-normal procedures for loss of DC power in C-331 - 333, -335, and -337.
. C, Planned Corrective Actions (NRC No. 32571,32662)
- 1. ' By September 25,1997, Operations will develop and disseminate an operator aid for appropriate operators and front line managers which will clarify the functions of the rectifiers and what systems are affected by the inoperability of ACR rectifiers.
- 2. By November 14,1997, Engineering, in conjunction with Maintenance, will evaluate replacement of the existing 250 VDC transfer switch relay with two 125 VDC relays as
~ described in ESO Z98990 in C-310.
- 3. By February 20,1998, Engineering, in conjunction with Maintenance, will evaluate replacement of the existing 250 VDC transfer switch relay with two 125 VDC relays as described in ESO Z98990 in C-331, -333, 335, and -337, if a similar system exists as -
discovered in C-310.
i
systems supplying power directly to safety systems which may have similar automatic thowover characteristics that are subject to the common mode failure.
systems identified in corrective Action 4, above.
- 6. By March 5,1998, Reliability Engineering wUl evaluate results in corrective Action 5, above and determine preventive maintenance requirements for the transfer relays.
LESSONS LEARNED The repetitive loss of DC power is indicative of a failure to understand the depth of the problem and a failure to rigorously correct deficiencies.
EXTENT OF EXPOSURE OF INDIVIDUALS TO RADIOACTIVE MATERIALS
- None 4
gg g
-d
@ ++
'-O
4 Docket No. 70-7001 Page 1 of 2 ER-97-13 List of Commitments PREVIOUS COMMITMENTS (NRC No. 32381)
- 1. By July 31,1997, Engineering will inspect, with the assistance of Electrical Maintenance, all DC power transfer systems in C-331, -333, -335, -337. The inspection will determine the current condition of the transfer relay and associated fuses to the rectifier of the DC power transfer systems and will identify deficiencies. A copy of the fmdings will be provided to Reliability Engineering.
- 2. By August 15,1997, Operations will ensure a system of communication of preventive maintenance (PM) status / clear ownership and approval of PM deferral exista.
l
- 3. By August 29,1997, Operations will develop a schedule to correct deficiencies, as required, identified in corrective action No.1, above.
- 4. By September 30,1997, Reliability Engineering will evaluate (taking into account the results of corrective action No.1, above) and determine the PM requirements for the transfer relays in C-310, -331, -333, -335, and -337. This evaluation will also include determining the adequacy of the frequency of the DC power switch test.
- 5. By September 30,1997, Operations will issue an off-normal procedure for loss of DC i
power in C-310.
- 6. By September 30,1997, Operations will issue off normal procedures for loss ot DC power in C 331, - 333, -335, and -337.
CURRENT COMMITMENTS (NRC No. 32571, 32662)
- 1. By September 25,1997, Operatiore, will develop and disseminate an operator aid for appropriate operators and front-line managers which will clarify the functions of the rectifiers and what systems are affected by the inoperability of ACR rectifiers.
- 2. By November 14,1997, Engineering, in conjunction with Maintenance, will evaluate replacement of the existing 250 VDC transfer switch relay with two 125 VDC relays as described in ESO Z98990 in C-310.
- 3. By February 20,1998 Engineering, in conjunction with Maintenance, will evaluate replacement of the existing 250 VDC transfer switch relay with two 125 VDC relays as i
Docket No. 70-7001 Page 2 of 2 described in ESO Z98990 in C-331, -333,-335, and -337, if a similar system exists as discovered in C-310.
- 4. By October 15,1997, Engineering will identify any additional DC rectifier / transfer systems supplying power directly to safety systems which may have similar automatic thowever characteristics that are subject to the common mode failure.
- 5. By January 30,1998 Engineering willinspect any additional DC rectifier / transfer systems identified in corrective Action 4, above.
- 6. By March 5,1998, Reliability Engineering will evaluate results a corrective Action 5, abover
%termine preventive maintenance requirements for the transfer relays.
~ - -
..