ML20134Q104

Forwards Rept Re Problem W/One Kilowatt Interlock and Corrective Action Taken

ML20134Q104
Person / Time
Site:	University of Illinois
Issue date:	02/14/1997
From:	Holm R ILLINOIS, UNIV. OF, URBANA, IL
To:	Alexander Adams NRC
References
NUDOCS 9702260248
Download: ML20134Q104 (3)

Text

-. . -. . - - - - . - .

[d'IS'l Radiation Science Laboratorv

, University of Ilkno 's at Urbana-Champaign

,/ Department of Nuclear Ent gneenng / Conege of Engineering N 214 Nuclear Engineering Laboratory 217 333-7755/0866 e j? 22 103 South Goodwin Avenue 217-244-5916 fax  ;

&yy%K1 Urbana, IL 6180?-2984 _

m --

x /:

A 'l ' .

G~

~~ I* i /

Advrainiserefor Director of Research

• Richard L. Holm Dr. Sheldon Landsberger .*

\ l r-holm@usuc edu sheldon@uxh cso uiuc edu h

L ,

February 14,1997 l

Docket No. 50-151 J

Mr. Alexander Adams,Jr.

Sr. Project Manager United States Nuclear Regulatory Commission M.S.1 I-B-20 Washington, DC 20555

Dear Sir,

I have attached a copy of the report given to my Reactor Committee on the problem with our i kilowatt interlock and the corrective action taken. A copy of the wiring diagram for the corrective action is included.

As I stated during our conversation on Monday, February 10,1997 this is a violation of our Technical Specifications Section 3.5, but it is not defined as an Abnormal Occurrence.

1 If you have any questions please do not hesitate to contact me.

Sincerely, Richard L IIolm Reactor Administrator c: file 990 Y 9702260248 970214 [

PDR ADOCK 0500 P

__ _ _. . _ . _ _ _ . _ _ . - _ - __ _ _ _ -.- _ .- -.m_ ._. _ .___._ . .

I j .g' l

(

, g ,__g.. p _ , . . >

gg.. p_, ___ _ y Sabanitted to Remeter Connemidae 2/1237 l

In order to operate while troubleshooting a problem with the campbell region (operational when power level > 10%)

of the NM-1000 a deviation from normal procedures was written. The NM.1000 operates in two modes - count rate and ~W The count rate mode is still functioning normally and is the only mode that provides for any safety function or Icquired instrumentation. The count rate mode provides the source count interlock and the I kw interio.;k. The source count interlock prevents rod withdrawal if the fission chamber is not seeing a valid neutron count rate prior to reactor startup and the I kw interlock prevents firing of the transient rods at too high of a power level to prevent exceeding a limiting safety system setpoint (fuel temperature). As part of this deviation it was l

I decided to physically verify that the I kw interlock was operational. This interlock is normally tested by insertion of a i

test signal and verification that the interlock indicates ON. It is not normally tested in manual mode by physically l trying to fire the air rods. It was found that with the reactor shutdown and the manual rods on the bottom it was

, possible to fire the air rods with reactor power indicating greater than 1 kw and the interlock ON. This is a violation of Technical Specification 3.5 which states that this interlock must be present in ALL MODES OF OPERATION. i l

While this was a violation of our Technical Specifications, it was not an abnormal occurrence since the problem was discovered while the reactor was shutdown. It has been reported to our NRC project manager (2/10/97) and will be l followed up with a copy of this report.

A review of the documentation for the console and physical tracing of wiring confirmed that there is no hardwired l

relay that performs the function of blocking firing of the air rods for this interlock. The interlock is generated in the l

NM 1000 and a signal is sent to the CSC computer soAware which then deternunes whether or not to allow firing the l

sir rods based on other information. The interlock is fully functional in other modes of operation. Our determination is that the soAware is written in such a manner that the mode described above is not covered by the bterlock i General Atomics also believes this to be the case and it is further reinforced based on information from other facilities on this interlock. A review of the Acce= Test Procedure for the console performed in 1992 does not have a physical test of this interlock, again reinforcing our belief that this is simply the way the soAware is written. There is l

some variation between other facilities, but we are the only facility with a 100% sir operated transient rod which' makes this problem unique to us.

A $0.59 review has been wdtten to use the intedock directly off of the NM-1000 rather than through the CSC soAware thus bypassing the programming in the CSC. This interlock is wired directly in the firing circuit for the I

transient rods.

l i

l i

1

l. l L  ;

I

CSC -

SW3 (FIRE)

"oO Oco"" NM'- 1000 CONNECTION BOX (beh;nd CSC TB-4 TBb5 / ,

ocoococ<

blo ck / green /

/

twisted poir in ,

[_ nut ti-c onduc t or TB-5 TB3-6 dj",*1'[t,>

RELAY EDARD DI-22 l O O S7-05 57-06 A5A1 DIS-64 Modification of Transient Rod Fire Cicuit for Hardwired i KW N"

Interlock

-