ML20116D828

Forwards Info Which Updates NRC on Matter Re Magnet Amplifiers for Osu Research Reactor

ML20116D828
Person / Time
Site:	Ohio State University
Issue date:	01/06/1995
From:	Cruz J OHIO STATE UNIV., COLUMBUS, OH
To:	Caniano R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
References
NUDOCS 9608020340
Download: ML20116D828 (3)

Text

..

• ^

T-HE Engineering Experiment Station 156 Hitchcock Hall j

W 2070Neil Avenue Columbus, OH 43210-1275 Phonc 614-292-4903 FAX # 614-292-9021 UNIVERSITY g_

O"O )\\

January 6,1995 G

i

\\

Roy J. Caniano, Chief r\\

Reactor Support Programs Branch US NRC Region III 801 Warrenville Rd.

l Lisle,IL 60532-43S1 l

l

Subject:

Update of Magnet Amplifiers for the OSU Research Reactor, Docket 50-150 License P.-75 i

Dear Mr. Caniano,

In March of 1993 the OSURR reported a failure of a control rod to drop _ following a slow l

scram. The status of our progress in correcting this problem was reviewed by Mr. Tim l

Reidinger during his inspection of May 23-26,1994. Subsequent to his inspection new magnet control amplifiers have been installed, tested, and approved by the OSU Reactor i

L Operations Committee at its meeting of September 22,1994. The Reactor Staff agreed i

durmg this meeting to update the NRC on this matter which is the purpose of this letter i

and attachments. The Reactor staff realizes this item may not be closed out until the next on site inspection. -If you have questions on this matter please contact Mr. Richard Myser at 614/292-6755.

Sincerely, Qbyb Csu l

Jose B. Cruz, Jr., Dean l

Enclosures

c. Richard Myser i

Don W. Miller i

l Ted Michaels 1

i

)

i t

50 P

PDR l

(.

M IIM College of Engineering

i A

To : ROC Members From : Michael J. Davis Date : 10 August 1994 Subject : New Control Rod Magnet Amplifiers

[

/

In response to the failure of Shim Safety Rod # 1 to drop to the bottom following a manual scram, a new magnet control system has been designed, built, and tested. Failure of the previous system was determined to be due to a short occurring between the engage limit switch assembly and the magnet coil assembly, allowing Shim Safety Rod #1 to remain energized despite the insertion of a scram signal. As a result, the NRL staff agreed to design and mstall a magnet current amplifier that " electrically floats"; and to design a system to monitor the magnet coil cable to continuously check the integrity of the lines and provide an alarm to notify the operator if a problem develops.

The new system has a floating power supply to provide the " electrical float". Current from the power supply is regulated by an LM-317 current regulator. which is adjustable via a ten turn potentiometer. Technical Specification 3.2.3 requires that current be limited to less than or equal to 100 mA. Limitation is provided by the combination of an LM-339 comparator, and a 74C74 dual flip flop. The comparator has adjustable high and low current set points, that allow a maximum magnet current of 91 mA (<= 100 mA). It checks to ensure that magnet current is between the two set points, and sends its output to the flip flop. The flip flop has two outputs, one of which goes to the " Current OK" LED, j

and the other of which goes to the fast scram FET. When current is outside of the high I

and low set points, the flip flop sends a signal to the FET, causing it to open the current path to the magnet. This is in essence the effect of any fast scram signal. The slow scram system functions independently from the fast scram system. It is comprised of a 171 DIP 25 double pole relay that opens on both sides of the magnet coil when the relay coil ground path is opened due to the receipt of any slow scram signal. The fact that the current path is broken on both sides of the magnet coil, rather than just one side, is an.

improvement in the system that precludes a failure of the type that initiated the redesign of the system.

The " Current OK" LED indicates that the FET is closed and that there is a current path through the fast scram system to the magnet coil. The " System OK" LED indicates that the slow scram relay is closed, again allowing current flow to the magnet coil. There is an inherent continuous system check to ensure the integrity of the lines, in that if a short develops across the magnet coil while it is energized, the current will bypass the coil and i

the rod will drop. The system also measures voltage drop across the magnet coil. In the event of such a short, current will stay constant, but voltage will drop to zero, to alert the j

operator that a stort has developed.

The design of the system was reviewed with Dr. Richard Bailey and Howard Dyke on 28 Feb 1994. A prototype was built and bench tested on 29 Jun 1994. Bench testing included fast and slow scram tests, high and low current drop-out tests, determination of maximum failure current and power to the magnet coil, and determination of magnet current decay time. Fast scram, slow scram, and high / low drop-out tests functioned as expected. Maximum failure current (maximum current that can be delivered by the power supply without regulation and / or limitation) was measured to be 121 mA, and j

maximum power delivered (based on the maximum failure current) to the coil was then calculated to be 4 W. Maximum magnet coil decay time was measured to be 2-3 ms.

i The prototype was then installed and tested in the control room on 5 Jul 1994. Fast and i

slow scrams were tested IAW approved procedure IM-03. High and low current drop-out tests were also performed. All test results were satisfactory.

l l

~ - _.. _.

i o

To : ROC Members From : Michael J. Davis Date : 21 Sep 1994 Subject : Long Term Preventive Maintenance Items In Response To Shim Safety Rod #1 Failure Subsequent to the failure of Shim Safety Rod #1 to drop following a manual scram on 5 Mar 1993, the NRL agreed to perform a number of preventive maintenance items as part of our long term plan to prevent the recurrence of this type of event.

1) Examine magnet coil connectors for corrosion on a quarterly basis. Following l

these inspections, a determination will be made on the necessity of providing auxiliary ventilation to remove condensation.

Inspections showed no noticeable increase in corrosion or condensation. With the new magnet control system, a build-up of corrosion leading to a short to the outer magnet assembly will not adversely affect the safety system. In the two cases in which this short develops and the outer magnet assembly is at ground, as it should be, the current source l-to the magnet will still be the current amplifier in the control room, and thus, will be l

subject to safety system designs. If the current source side develops this short under this l

condition, current will be shunted around the magnet. If the return side develops this

- short, the magnet will continue to operate as normal. If this type of short develops and l

l the outer magnet assembly is not at ground (which was the case leading to our previous

)

l failure), then one of two things happens. If the short is on the current source side of the magnet, then the current loop is blocked by the circuitry in the magnet control system. A

'1 test of the blocking capability of the system was performed by inserting a 120 VAC signal to the magnet BNC center (current source side) of the control system. As expected, the system shut down. If the short occurs on the return side, it will be shunted l

to ground through the BNC shield ground, which will have no effect on the magnet current or the control system.

In light of all of this, the NRL staff has agreed that shim safety magnets shall be inspected 4

l on an annual basis and during magnet replacement operations. Furthermore, it has been decided that it is im-=y to modify the magnet coil housings to provide auxiliary i

ventilation to remove condensation.

2) Create a procedure to monitor the magnet coil cable on a monthly basis to check for electrical grounds.

j 1

This item was performed as part of our Pre-Start Checkout procedure (IM-03). As part of that procedure, a scram check is performed on each of the shim safety rods. With the old magnet current control system in place, if a short similar to the one that caused the rod drop failure had occurred, causing the expected electrical ground of the system to no longer be at ground on any of the shim safety rods, then the rod would fail to drop during this test. If all of the rods dropped as expected, then this indicated that the electrical grounds were still at ground. This system check was part of Procedure IM-03 prior to the l

rod drop failure incident. As a result, no new procedure needed to be developed. This check will continue as part of the Pre-Start Checkout procedure to be performed prior to each daily reactor operation.

f f

-