ML19305E476

Responds to NRC 800321 Ltr Re Violations Noted in IE Insp Rept 50-160/80-01.Corrective Actions:First Portion of Current Audit Conducted on 800324 & Schedule for Remaining Portion Will Be Developed in Apr

ML19305E476
Person / Time
Site:	Neely Research Reactor
Issue date:	04/09/1980
From:	Russel J Neely Research Reactor, ATLANTA, GA
To:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
Shared Package
ML19305E473	List: ML19305E476 ML19305E478
References
NUDOCS 8005120500
Download: ML19305E476 (4)

Text

. - - ..

80051207 N ,

4 n

V

, GEORGI A INSTITUTE OF TECHNOLOGY

' SCHOOL OF NUCLEAR ENGINEERING ATLANTA. GEORGl A 30332 FR AN K H. NEELY G NUCLEAR RESE ARCH CENTER  %,

T3LEPHONE: (404)094 3600 April 9, 1980

]'

E U.S. Nuclear Regulatory Comission, Region II *

• 101 Marietta Stmet, N.W. Suite 3100 "

Atlanta, Georgia "

30303 Gentlem n:

Reference:

RII:CJ; 50-160/80-1 Eis letter is in response to your report dated March 21, 1980, regarding an inspection of our facility by your staff.

A. Nuclear Safeguards Comittee Audits Imediate steps have been taken to correct this situation. At a meeting of the Nuclear Safeguards Comittee on March 17, 1980, ne Comittee reaffired its comittment to annual audits of the operation of the Georgia Tech Research Reactor. 'Ihe nature of an effective audit is such that it will require several man-days to complete. herefore, our audits will be conducted as a scheduled series of smaller units, performed by those members of the Nuclear Safeguards Committee who am not on the operating support staff for the reactor, as as-sisted by such technical specialists as they may designate. %e first portion of the current audit was conducted on March 24, 1980 and is documented in the attached memo to Dr. M.V. Davis, frm Drs. J. Russell and R. MacDonald, dated March 24, 1980. The schedule for the remaining portions will be developed at the April,1980 meeting of the Nuclear Safeguards Comittee.

B. Irradiation of Fissionable Materials f 'Ihere is some confusion, mostly created by our own procedure 3101, " operation of Expermental Facilities" about this problem. While the definition of a minor experiment indicates the material will be non-fissionable, specification H on page 2 of procedure 3101 reads in part: "me radioactive material content, in-cluding fission products of any doubly encapsulated or vsnted experiment..."

herefore the exact intent is not quite clear.

As noted by your staff, the three irradiations that were made involved only miligram amounts of uranium. While in partial conflict with our own procedure, I

the irradiations themselves caused no detectable reactivity effect, created no

{ radiological handling problem and thus were done, we believe, within the intent i of the minor experiment envelope. Because many materials contain trace amounts of uranium, additional confusion is caused by using the tem "non-fissionable" materials. Since we feel the conflict lies within our own procedure, the Nuclear Safeguards Comittee will be asked t'o specifically address this infraction. Un-til the Comittee msolves the issue, no known fissionable material except samples OFFICIAprO'M -

(- . _

t U.S. Nuclear Regulatory Comission, Region II Page 2 April 9, 1980 that might contain trace amounts of uranium or other fissionable isotopes will be irradiated in the Georgia Tech reactor as minor e meriments.

C. Quality Assurance Program for Experim nts Each major experiment to be done in the reactor requires a review by our Nuclear Safeguards Conmittee (NSC). 'Ihis review includes a knowlegable evaluation of the safety of the experiment, including identification of significant safety features and necessary equipment safety tests if any.

Reactor Operations and the Office of Radiological Safety are responsible for verifying that the experinent conforms to the NSC requirements.

Each minor experiment to be done in the reactor requires completion of our Fom II (3-68) (copy enclosed) entitled " Georgia Tech Research Reactor, Re-quest for Minor Experiment Approval". When properly completed, the form will identify the material and its weight, the estimated radioactivity to be pro-duced and the encapsulation of the material.

The fonn is validated by signatures from at least Reactor Operations and the Office of Radiological Safety. This document, rather than our procedure 3101 as identified on page 3 of your Detail section, constitutes our QA confirmation for these experiments. 'Ihis position was discussed and agreed upon by ourselves and Mr. Monte Conner, NRC Project Manager for our facility, during the develop-ment of the Technical Specifications as issued on June 6, 1974. However, as a part of the review of the procedure " Operation of Experimental Facilities" (B. above), the problem of QA procedures and verification will be addressed.

It is the intent to broaden the scope of our QA procedures for both major and minor experiments to assure they adequately document verification of the signifi-cant safety aspects of all reactor experinents.

D. Monthly Flow Test, ECCS After investigating this problem we have reached the following conclusions.

During the time period 10/9/79 to 12/5/79, the ECCS system was capable of delivering D Since 12/5/79, the monthly urveillance s,0 at a flowchecks rate ofof 8.0 the to 8.5rate flow gallons have per beenminute.

satisfactory 1.e.

8.5 1 0.2 spm. We believe, however, that in the period from 12/5/79 to 3/5/80 the minimum flow rate could, at times, have been as low as 8.0 gpm.

The operatienal lower limit for this flow rate is 8.5-0.2 or 8.3 gpm. 'Ihe re-quired minimum flow rate is 8.0 gpm as documented in our Safety Analysis Report (section 4.4.8.3, p. 79) and in a response letter to USAEC - Reactor Licensing dated July 13, 1971, question No.'6 (copy enclosed).

In our investigation, we found that while attenpting to detennine the ECCS flow rate on October 9, 1979, the reactor operator observed a flow rate of 8 gpm.

'Ihis was less that our operational limit of 8.510.2 gpn as stated on the job plan for this monthly surveillance. He further determined that the flow rate could be varied between 8.0 and 8.5 gpn by the manner in which ball-valve No.

94 was opened-to initiate flow fmm the 300 gallon tank TD-2. A mechanical stop is provided for valve 94 to mstrict its open position and thereby limit the total flow rate frm the tank. 'Ihe suspected cause of the variation in flow rate was thought to be due to a loosening of the lock nuts fixing the

' ~~,____ _7'--

'- *" y ,_. U'

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

U.S. Nuclear Regulatory Cormission, Region II Page 3 April 9,1980 handle to the ball operating shaft. he locknuts were tightened. A mis-understanding then arose resulting in the Reactor Supervisor believing the problem to be corrected. Further investigation revealed that some looseness still existed and, in March 1980, an additional attempt was made to correct the problem by inserting shim stock between the handle and the " flats" of the ball operating shaft. Complete verification of the fix has not been possible because the reactor is operating in Itde 1 as defined by our Tech-nical Specifications and the ECCS tank TD-2 is not filled with D 0.

2 We propose the following action to prevent this problem from occurring in the future. When additional D 0 is available, TD-2 will be filled and the operation of valve 94 complete verified. Prior to this, the reactor will not be operated at power levels greater than 1 f&l. Additionally, to strengthen our surveillance program, we will establish a second review of the surveillance documentation to be performed by a person of at least the level of plant know-ledge of a s6nior reactor operator. 'Ihis last action (second review) has al-ready been put into effect.

E. ECCS Report The details supplied in item D above are applicable to this item. A licensee Event Report IER 80-1, has been issued.

F. Measurement of Primary Coolant pH he measurement and interpretation of a pH value for a heavy water system presents some ambiguities. Because our cooling water is essentially formed of oxygen and the heavy hydrogen isotope deuterium, the term pH is probably nore correctly termed a pD. Attempts to measure a value of "pH" using chemical electrodes inserted into a sample of moderator generally results in false acid readings as the unbuffered pure D2 0 absorbs C0 3 from the air. To correct the problem stated in your inspection, we have re-instituted a weekly sampling of the D3 0 and will log a value for "pH". 'Ihis will be done in conjunction with our Weekly Precritical Startup Checklist, procedure 2002.

G. Kanne ChPJnber Inoperable he Technical Specification in question, 3.2a states, in its entirety, that the reactor shall not be made critical unless "'Ihe reactor safety systems and related instrumentation are operable in accordance with Tables 3.1 and 3.2 in-cluding the minimum number of channels and the indicated maximum or minimum set points." Table 3.2 clearly permits a channel such as. the Kanne to be by-passed for a period not to exceed 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> for test repair or calibration. It has, until now, been our understanding that this specification allowed multiple, daily startups consistent with our mode of operation as a research reactor. Be-cause of redundancy, if an instrument such as the Kanne is taken out of service for repair, there is a backup instrument that is operable. 'Ihe inspectorb in-terpretation of the limit is to allow only steady state operation in progress to continue. We have modi.fied our administrative controls to conform to this interpretation.

U.S. Nucl ar Hegulatory Commission, Region II Page 4

, April 9, 1980 H. Natural Convection Limiting Safety System Setting Once or twice a year for Nuclear Engineering laboratory course work, the reactor is operated at very low powers (< 1 kW) in a natural circulation mode to minimize the effect of flow turbluance on the reactor kinetic parameters being measured. The reactor power level is limited by adminis-trative measures to values of 1 kW or -less and it was not recognized that a 7bchnical Specification required the actual adjustment of the flux nonitor channels. Effective inmediately, operation of the reactor in the natural convection mode will not De permitted unless the flux monitor trip points have been mset to value of 1.1 kW or less.

Sincerely urs, Jo n L. Russell, Jr.

Dimetor JLR:lnn ec: Members of the Nuclear Safeguards Conmittee theiosures:

?

. GEORGI A INSTITUTE OF TECHNOLOGY school OF NUCLEAR ENGINEERING ATLANTA, GEORGI A 30332 March 24, 1980 PR ANM H. NEELY NUCLE AR RESE ARCH CENTER TE L E PH O N E: (404)894-3600 MEMO TO: Dr. M. V. Davig [# '

FROM: Dr. J. R 11, Dr. R. MacDonald f

SUBJECI: GTRR udit by NSC At the direction of the Nuclear Safeguard Comunittee, Dr. J. Russell and Dr. R. MacDonald audited GTRR operations:

Audited: Verification Filing System and Operator Log n

Problems:

1. Some system work sheets were un-numbered. This led to apparent noting of problems for which no action was taken.

2. Files not in order

3. Blank work sheets, i.e., dates and initials but no coassents on work done. .

4. Operations file does not retain record of H.P. compliance with calibration procedures.

Consnendatio' n :

1. Recognition of flow problem from very small decrease in normal

, flow rate, and took immediate corrective action.

Reconunendation: i

1. All system work sheets should have I.D. number,

2. Any statement of a problem should either state resolution or refer to another SWS I.D. number.

. 3. Filee in front office, console cards and maintainence lists should l be reconciled and put in order. Master list may need up-dating.

l

4. Since H. P. retains calibration records and their nsanbered SWS, I operations office should retain the note H. P. gives to Linda indicating the date of compliance. The calibration data can then be recalled via the date. . Blank work sheets should not be put in operations office files. j cc: Consmittee Kirkland Ja N

________________1-_