ML20059L798
| ML20059L798 | |
| Person / Time | |
|---|---|
| Site: | University of Missouri-Rolla |
| Issue date: | 11/09/1993 |
| From: | Freeman D MISSOURI, UNIV. OF, ROLLA, MO |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 9311170403 | |
| Download: ML20059L798 (10) | |
Text
,
I I. Nuclear Reactor Facdity j Nuclear Reactor II Rolla. MO 65401 0249 Telephone (314) 341 4236 UNIVERSITY OF MISSOURI-ROLLA Missouri's Technological University November 9, 1993 Director
, Office of Nuclear Reactor Regulation
! U.S. Nuclear Regulatory Commission Washington, D.C. 20555
Dear NRC:
Please find attached the written report on the October 26, 1993 reportable occurrence at the University of Missouri-Rolla Reactor facility as required by our Technical Specifications Section 6.2.2(2)(c).
Sincerely, I tsh Jf#5 vid W. Freeman UMR Reactor Manager DWF/lp Enclosure copies to: Regional Administrator, NRC, Region III Marvin Mendonca, NRC Project Manager William Vernetson, Previous TRTR Chairman ;
Wade Richards, TRTR Chairman '
UMR Radiation Safety Committee Dr. Albert E. Bolon, UMR Reactor Director Dr. Lee Saperstein, Dean, School of Mines & Met Dr. Walter Gajda, V. Chancellor, Academic Affairs 3 Signed before me this k:O day of November, 1993.
x4Yw Notary Public SUE HUFHAM NCFTARY PUBUC STATE OF MISSOURI PHELPS COUNTY MYCOMMISSICN EXP. AUG.30,1416 1 G fl O '1 Q 9311170403 931109 ',
M PD9 ADOCK 05000123 1 *"'*"'""""*"S**" '/ gr%
S PDR _; / s \-
REPORT ON THE OPERATION OF THE UNIVERSITY OF MISSOURI-ROLLA (UMR) REACTOR FACILITY WITH AN INOPERATIVE PERIOD CHANNEL November 9, 1993 .
t i
l l
I Prepared By: David W. Freeman UMR Reactor Manager l
t
E 1
1.0 Introduction On Tuesday, October 26, 1993, the UMR Reactor was started and inadvertently operated from 1350 to 1448 without the Period Channel functioning properly. Upon recognition of the problem, the reactor was promptly shut down. At no time was reactor safety compromised. The event was promptly -
reported to NRC. This report details the operation and subsequent actions taken.
2.0 Descriotion of Operational Events A routine NE 306, " Reactor Operations", class that provides students with " hands-on" operational training on the reactor controls was started at 1300, October 26, 1993. The instructor of the course was a licensed Reactor Operator who was functioning as both the course instructor and the licensed Operator-on-Duty in the control room.
The Pre-Startup Checklist (which incorporates our daily ,
checks) was performed between 1304 and 1350. The Pre-Startup Checklist contains four tests of the Period Channel.
The Period Channel responded properly to those tests, as was verified by reviewing the recorder strip charts.
The startup commenced at 1350. The reactor was leveled at a power of 2W at 1411 and logs were taken.
At 1420 a power increase to 20W was initiated. At 1424 power was leveled at 20W and logs were taken. At 1430 a power increase to 200W was initiated. At 1435 power was leveled at 200W and logs were taken.
At 1444 a power increase to 2kW was initiated. During this transient, the improper behavior of the Period Channel was identified, the Senior Operator-on-Duty (the Reactor Manager was the Senior Operator-on-Duty) was summoned, and an immediate shutdown was initiated at 1448.
l The shutdown was initiated during the up-power transient.
The peak power attained was 97% of 2kW.
Upon shutdown the Reactor Manager tagged out the console.
During the entire run, the Period Channel indicated an infinite period. It should be noted that the Period Channel does not begin to respond to neutrons until reactor power is approximately 1W. Thus, no signal other than infinity can be obtained from the Period Channel during most of the reactor startup to low power. During this stage of reactor startup, an infinite reading on the Period Channel is
2 appropriate. Therefore, the faulty signal could only have been identified by the operator starting at about 1411, the time power reached 2W. In this sense, the length of time that the reactor was operated after the first indication of malfunction sitould have been noted by the Reactor Operator was 1411 to 1448 (37 minutes). I During steady-state operation, the normal period indication is infinity. The channel only gave incorrect information during the power transients. The total time of transients !
above 2W was:
i
. Increase from 2W to 20W (1420 to 1424) 4 min.
Increase from 20W to 200W (1430 to 1435) 5 min.
. Increase from 200W to identitication of problem and shutdown (1444 to 1448) 4 min.
TOTAL 13 min.
This analysis shows that the operator failed to identify the [
problem that manifested itself over a total of 13 minutes of the 58 minute run.
The Reactor Director, who was out-of-town at the time of the occurrence, was contacted at home at approximately 1830 by the Reactor Manager. A determination that the event was a reportable occurrence was made at that tirae.
At approximately 9:00 am the next morning (October 27, 1993), Marvin Mendonca at NRC was notified of the problem.
Many phone calls with NRC Region III have since ensued.
The equipment failure that lead to the inoperable Period Channel has been determined to be the sticking of the " Trip Test" front panel control knob in the " test" position. The last of the four Period Channel tests performed on the Pre-Startup Checklist involves the use of the spring return
" Trip Test" control knob. By studying the Period Strip Chart, it has been determined that the channel passed the check. It has been determined that the " Trip Test" knob did not spring return to the operate position, but rather stuck in the test position thus rendering the Period Channel ,
NRC inspectors visited the UMR Reactor facility on November 3, 4, and 5.
3.0 Corrective Action Plan The management of the UMR Reactor Facility is committed to taking strong action designed to assure the occurrence of October 26, 1993 does not happen at this facility again.
l l
l 3
Further, the UMR Reactor Facility commits to informing the '
TRTR community of the event, in detail, so that other similar facilities can learn from our experience.
We believe the root cause of the event was equipment failure combined with operator error. Our proposed corrective -
actions fall into three categories; 1) Operator Training, 2)
Equipment Repair and Replacement, and 3) SOP Revisions.
Each category is discussed below. We firmly believe that ,
the proposed measures will preclude recurrence of this type of event at the facility in the future. We also are r committed to an ongoing assessment of our corrective actions to identify other areas for improvements.
l 1 3.1 Operator Trainina The UMR Reactor Facility management is committed to ,
l assure that all operators maintain a high level of
, proficiency with regard to reactor operations. This l
event may have been averted had the operator involved been given more training with emphasis on good control room practices, the importance of following procedures, and proper supervision of nonlicensed operators at the controls. To that end the following remedial training will be provided to the operator involved prior to resumption of unsupervised reactor operations:
- a. A lecture review of the startup, power change, and ,
shutdown SOP's will be performed with the operator. This review will stress the importance of following procedures and the importance of :
^
diligently monitoring all of the nuclear instruments both during transients and steady-state operation.
- b. Six hours of one-on-one console training of the operator by an SRO. This training will focus on the basics of operations including proper observance of the nuclear instrument channel outputs and on the expected proper response of each channel under both steady-state and transient conditions. This training will also focus on the proper techniques for supervising non-licensed personnel at the reactor controls.
This training will be successfully completed as determined by the Reactor Director and will be documented in writing prior to allowing the operator involved to operate the reactor unsupervised.
l 4
We feel certain that the above personal training will provide the heightened awareness necessary for ;
competent operations. It should be noted that the operator involved is highly capable and extremely well versed in the nuclear sciences. The operator has ;
previously held a Reactor Operators license at another l facility and has a PhD in Nuclear Engineering. -
Generic staff training that will be completed prior to ,
restarting the reactor consists of a staff meeting including all licensed operators that:
- a. Clearly communicates the details of the event followed by open staff discussion,
- b. Emphasizes the importance of following procedures. i
- c. Emphasizes the importance of diligently monitoring all of the nuclear instruments both during transients and steady-state operation.
- d. Emphasis on the licensed operators responsibility l when a non-licensed person is at the controls.
Generic staff training will be incorporated after the reactor has been restarted that will consist of Monthly l Training sessions for all licensed operators which will cover such topics as control room operations and the supervision of non-licensed personnel at the console.
The monthly training sessions will be performed for a six month period, at which time the effectiveness of i the training sessions will be evaluated. Based on that evaluation, the frequency of the training sessions may I be changed as deemed necessary by the Reactor Director.
3.2 Equipment Repair and Replacement i
i Corrective action to be performed on the reactor =
equipment prior to restarting the reactor includes:
I
- a. Identify the cause of the equipment failure.
The cause of the equipment problem is the failure i l of the period " Trip Test" knob which is used to l l interject a simulated period signal for testing l purposes. The knob is a combination spring return 3 l
push switch and potentiometer. The spring return l switch portion of the component switches the <
drawer between the test input signal and the real !
detector signal. The potentiometer is used to I
" dial" in different simulated periods.
I I
1 l
5 The failure occurred when the spring return switch failed to " return" to the operate mode due to wear. This caused the Period Channel to be '
" stuck" in test mode.
The stuck switch condition was reproduced during troubleshooting and found to display exactly the symptoms associated with the october 26 run.
- b. Repair failed components.
The component that failed is a specialty part that combines a double pole single-throw switch (S23) with a 1 k-ohm variable resister (R23).
Figure 1 shows the schematic representation of the parts.
We plan to replace the switch / potentiometer arrangement with two separate components; 1) a simple spring-return double pole single-throw toggle switch and 2) a simple 1 k-ohm potentiometer. This modification is necessary because we have been unable to locate a replacement part. This modification is also advantageous in that the position of the toggle switch will be clearly visible to the operator.
- c. Inspect all other similar console test switches for similar problems and replace as necessary.
One other similar test switch is involved in the .
Log N test circuitry. This test switch is not used; however, it has been checked and is found to be in good working order. Two other multiposition test switches on the Log N and Period drawer were checked and some signs of wear were identified.
These switches have been replaced. (These two multiposition switches are standard switches --
not a specialty component like the " Trip Test" knob that failed.) The only other multiposition ,
test switch on the console is the Source Range l Test switch. This switch was checked and found to be in good working order.
Long term corrective action to be implemented include first and foremost the replacement of the old console equipment with the new Gamma-Metrics equipment which is already in house.
1 i
. 1
. l 1
l rip est 5 G
)
J2 G 3' ~
5 CRAM REL A Y . ~
<>--l-c
. 52/A-5 o+l .
(2I S23 .
2c+ ; c _
s, o /. : :
3 9 o/
.72 5-7 I 5 - 3 n - t ,
l ?
s i
0- DS 2/ .
i J2 2- / 4 i ~ K I E / O .. . .
~
3 i 6 -
.. _5 .
khrj9stg&: ..-5h1%gg_ s. . .x.'
dhe;c4y 9;r 9. x 7__3 S: >i 3 -9 3
.. Q, g;nE2n:.
g;y%qg . . $. y;c,
. - . g_. .,y :49ygp.gy . . r . .-. . . ;.y. - .gg..y,
.,..._ . , ~a_x l_ --
. ' .y ..
- I .
]l.-:H*.x....T<E.
^
- ,7 .
3 _ 7.,
c-
-J~
~~ ~
r-l 8 . a M.
. . o# _ ,yj - g - _
70 0l(
- 24 b '
l
.. . R2!
7/:-[-[' f 3 o h^ JZ4-9
^ g a -
. . ... . . M .
o gl-- *K ^ -'
. lA T I , 13 7ppy .
1%
1 ,
i C 7 'P - A /D .
Figure 1. Schematic Representation of the Period Channel " Trip Test" Control Knob l l
1 i
7 3.3 SOP Changes There are certain SOP improvements that will be made that, if in place, would have lead to the early identification of the malfunctioning channel. These SOP revisions will be made prior to restarting the reactor:
l a. Revise the Pre-Startup Checklist (SOP 102): move the PuBe source " spike" of the instruments to the end of the checklist, after all front panel drawer i
tests have been completed. This will assure that the CICc and fission chamber are responding to neutrons at the time of reactor startup. Had this measure been in place, the malfunctioning test switch would have been identified prior to the reactor startup.
- b. Revise the Hourly Checklist SOP 104: incorporate a line item to check the Period stripchart history for reasonable response. Also add a line item to record the Startup Channel level. In this way, proper operation of all of the NI channels will be verified upon leveling at a particular power (and hourly after that).
- c. Add Statement to SOP 101, " General Operational Procedures": clearly define the licensed operators duties when a non-licensed operator (student or trainee) is at the console manipulating the controls.
A preliminary review of SOP 107, " Routine Stable Operational Procedures" and SOP 103, " Reactor Startup Procedure" has been conducted to identify what I revisions could help clarify that the operator is to monitor all NI channels during transients. It has been determined that some changes are in order, but that these changes are not necessary prior to restart, but should be complete in a timely fashion, prior to the end of the calendar year.
l i l
i
~.
8 i l I l 4.0 Safety Sionificance The actual safety significance of this event was negligible. ~
The core excess reactivity is 0.496% delta-k/k. Using the simple kinetics formula:
t = 0-P = 5.9 sec l AP i
l Assume: B = 0.0079 (UMR SAR) p = 0.00496 (UMR excess reactivity)
A = 0.1 sec-2 l This analysis shows that even if all control rods were l completely withdrawn from our core, it is doubtful that a 5 l second period could have ever been reached. '
1 i Further, the SAR Section 9.5, "Startup Accident" shows that if all rods were simultaneously and continuously pulled until all rods were completely withdrawn, that even with no safety functions, no damage to the reactor would occur.
Section 9-6 of the SAR, " Maximum Reactivity Insertion" shows that our reactor can safely withstand a prompt critical excursion with a step increase of 1.5% delta-k/k without fuel damage with no safety functions. Clearly, our above i calculated 5.9 second maximum period is well below the l prompt critical excursion analyzed in Section 9.6.
Thus no adverse consequences to the reactor were possible in terms of safety significance.
5.0 Conclusion .
1 The incident that occurred on October 26,1993 should never have happened at our facility. The operator on duty should have identified the malfunctioning Period Channel some 37 l minutes earlier than she did. Although there was no safety significance associated with this event, the fact that the operator failed to identify the equipment problem in a timely fashion is viewed as a serious flaw in the operator's performance. We believe that the above specified corrective actions which address training, equipment, and SOP issues i are appropriate and that they sufficiently deal with the i issues to assure that this type of incident does not happen again. !