ML19310A386

Ro:On 800520,while Performing Routine Pulses,One of Five Pulses Produced Peak Power & Measured Fuel Temp Beyond Expectation for 2.55 Dollar Reactivity Insertion. Apparently Caused by 2.75 Dollar Reactivity Insertation

ML19310A386
Person / Time
Site:	Oregon State University
Issue date:	06/03/1980
From:	Ringle J Oregon State University, CORVALLIS, OR
To:
References
NUDOCS 8006110183
Download: ML19310A386 (4)

Text

.

l 1

. r ,  ;

Radiation Center We".

University Corvallis, Oregon 97331  : sos) 754 234:

w J June 3,1980 U.S. Nuclear Regulatory Commission Region V Office of Inspection & Enforcement 1990 N California Blvd.

Walnut Creek Plaza, Suite 202 Walnut Creek, CA 94596

Reference:

Docket No. 50-243, License No. R-106 Gentlemen:

On May 20, 1980 we apparently inserted approximately $2.75 in a reactivity pulse, exceeding the limit of $2.55 in Part 3.3 of our Technical Specifications. This was reported to your office by telephone on May 21, 1980, as required by Part 6.7a of our Technical Specifications.

This written report, as required by Part 6.7b of our Technical Specifica-tions, provides more details of the incident and the corrective action

., we have taken to prevent, hopefully, this incident from happening again.

Incident On May 20, 1980 we were performing routine pulses for an experimentor.

All of the pulses involved reactivity insertions of $2.55, inserted by means of our transient rod. The first three pulses that day produced peak powers which varied from 3050 MW to 3200 MW, and measured fuel temperatures which varied from 397*C to 405'C. These were all within the normal range of variability we have observed for a $2.55 insertion.

The fourth pulse produced a peak power of 3900 MW and a measured fuel temperature of 440*C. These values were higher than any we have encountered for a $2.55 insertion. A fifth pulse was then performed to see if our pulsing instrumentation and circuitry were working properly.

This resulted in a 3040 MW pulse with a temperature of 390'C, again for a $2.55 insertion--well within the normally expected range.

The conclusion seemed to be that the fourth pulse that day was produced by a reactivity insertion greater than $2.55. If the reactivity insertion had been about $2.75 rather than $2.55, we would expect a peak power and a measured fuel temperature corresponding to the observed values.

0 D

Oregon State University is an Allitmative Action / Equal Opportunity Employer 80 06110/33 J . . _ _ _ . _ _ . . _ _ _ . . _ _

U.S. flRC Page Two June 3, 1980 Background Information and Safety Implications 4 Having concluded that the fourth pulse on May 20, 1980 was probably genuine and that it came about from a reactivity insertion of about-$2.75, we then attempted to ascertain how this reactivity could have been introduced. The transient rod has a mechanical stop which limits the pulse reactivity insertion of this rod to $2.55. This stop was in place and functioning properly.

Thus, at most we could have gotten $2.55 from the transient rod.

The additional $0.20 must have come from one of the other rods, and' the regulating (reg) rod was the most logical candidate since the servo system moves this rod automatically in the AUTOMATIC or SQUARE-WAVE mode.

By carefully examining the switching sequence for pulsing, we discovered the following sequence which could have given an extra $0.20 of reactivity to a pulse:

1. Bring reactor critical at a low power in the AUTOMATIC mode.

2. Switch the mode switch to SQUARE-WAVE (one position beyond STEADY-STATE). The reactor is now operating as in the AUTOMATIC mode with the servo system controlling the reg rod position.

3. Switch the range switch to 1 MW, in preparation for the pulse.

! The servo system now starts driving out the reg rod in an attempt to bring the reactor up to the new, higher power level that is demanded.

4. ' Switch the mode switch to PULSE (HIGH or LOW). The reg rod now i stops driving out. It -is further out, however, than its critical -

position of step #1, and some positive reactivity has been added to the previously critical reactor. The reactor is now on a positive period.

5. Fire the transient rod. The reactivity value of the transient rod is now added to the positive reactivity from the reg rod

withdrawal.

If the sequence had been followed, the reg rod would have only i .had to move out about 30 units to add an additional $0.20 of reactivity, and it would have taken .it less than 2 seconds to move those 30 units.

We feel this is' a plausible explanation of how the $2.75 reac.tfvity insertion might have occurred. The reactor operator on the console at .

the time this incident occurred agreed that this switching error could 1 ~certainly have happened. He could not verify that it did happen, only 3 tha: -it was a possibility and an explanation for the occurrence.

1

. 4 . .~. . -+ *_ . . . , . .

l 2

/ l l

U.S. NRC Page Three June 3, 1980 f

1 An operator performing a pulse using the sequence just described is not following the proper sequence in our Oregon State TRIGA Reactor j Operating Procedures (OSTROP), Chapter 4 (Reactor Operation Procedures)..

i This approved procedure is identical to the sequence just described

except that Step #2 says to switch 'the mode switch to STEADY-STATE, not

, SQUARE-WAVE. If the proper procedure is followed, the reg rod will f not drive out and the only reactivity inserted will be from the transient rod. Since the STEADY-STATE and SQUARE-WAVE positions are next to

~

each other on the mode switch, it is certainly possible that an error of this type could be made.

Such a switching error is apparently quite rare, however. We have performed 2,172 pulses in the past 13 years, and this is the first time we have observed that an error of this type might have happened. We are also personally not aware of such an incident happening at any other TRIGA reactor facility.

E The measured fuel temperature that occurred during this apparent reactivity insertion of $2.75 was 440*C, well below the 510*C temperature which is our Limiting Safety System Setting (LSSS) for

' fuel temperature. Hence, the maximum fuel temperature was also well

below the fuel temperature safety limit.

To examine the potential safety implications of such an incident,

one can postulate the worst situation

the reg rod drives out entirely, then the transient rod is fired. The reg rod is normally somewhat more than half out when the reactor is brought to a critical condition-at low power prior to a pulse. For our reactor, the reg rod has a total worth of $2.62, and in the critical configuration prior to

pulsing, about $1.18 of worth remains in the rod. Theoretically, then, one could add about $1.18 of positive reactivity ~from the reg rod in addition to the $2.55 from thn transient rod, making the total reactivity insertion about $3.73. This is not physically possible, however.

4 Our reactor has an interlock set at a power-Tevel of 1 kW; if the-reactor power exceeds 1 kW, this interlock prevents firing of the transient rod. Thus, the maximum reactivity that can be added from the reg rod during such an incident is that amount that can be added before the power -level reaches 1 kW and trips the interlock. - Our 4

analysis has shown that if the reg rod drives out at its maximum speed, it will take about 6.3-6.4 seconds for the reactor power to increase I to l' kW. In this time interval, the reg rod has driven out about 115 units and the corresponding reactivity insertion is about $0.63.

1 4

5

- W.. .---,+,,,,.m- ,,--.-w - - , - -M .r- &-- e---y y wN - vv -

U.S. NRC Page Four June 3, 1980 Thus, the maximum possible reactivity insertion that we could experience during such an incident is about $3.18 ($0.63 + $2.55).

The measured fuel temperature expected for a $3.18 insertion is about 550*C. This temperature is about 40 C above our LSSS temperature of 510 C, but the maximum fuel temperature would still be 150-160*C below the fuel temperature safety limit. Thus, no fuel melting or fuel cladding failures would be expected.

Corrective Action The Reactor Operations Committee (ROC) met on May 23, 1980 to discuss and review this incident. The ROC concurred with the reactor staff findings that safety of the reactor and integrity of the fuel had not been compromised. The ROC also agreed that had the " worst case" scenario described above happened, the fuel te.nperature safety limit would not have been exceeded or even approached very closely.

The ROC did express concern that such an incident could occur, however. They agreed that the existing OSTROP procedures for pulsing, if they are followed, should prevent such an incident. The ROC voted to amend these procedures, however, to include an additional step just prior to firing the transient rod for a pulse. This additional step calls for the operator to verify that the reg rod has not moved from its critical configuration. This step should provide a final check on the amount of reactivity to be added and verify that the operator has performed the switching sequence in the proper order. This change had been implemented on an interim basis by the operations staff on May 21, 1980. Upon approval by the R0C, it immediately became a part of the OSTROP pulsing procedures.

The ROC felt that no other corrective action was necessary or justified at this time.

If you have any questions or desire any further information about this incident, please contact us.

Sincerely, b

J hn C. Rin le sistant Reactor Administrator JCR/rk cc: USNRC Office of Inspection & Enforcement, Mshington, D.C.

USNRC Document Management Branch, Washington, D.C.

Oregon Department of Energy C.H. khng, Reactor Administrator. OSU A.G. Johnson, Senior Health physicist, OSU T.V. Anderson, Reactor Supervisor, OSU