ML20128D464

Ro:On 921124,control Rod Interlock Sys Removed from Reactor Control Sys.Caused by Wiring Error Made During Mod That Disabled Interlock That Drops Reactor Out of Automatic Control.Control Rod Withdrawn to 23 Inches

ML20128D464
Person / Time
Site:	University of Michigan
Issue date:	11/25/1992
From:	Fleming R MICHIGAN, UNIV. OF, ANN ARBOR, MI
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9212070285
Download: ML20128D464 (4)

Text

l

• Nucicar !!cactor Laboratory The University of Michigan Ford Nuclear Reactor Phoenix Memorial Laboratory 2301 Bonisteel Boulevard November 25, 1992 Docket 50-2 License R-28 United Staten Nuclear Regulatory Commission A t t.n : Document Control Desk Wauhington, D.C. 20555 .

Reportable Occurrence No. 16 ~

Reactor Startup wi t.h Power Level Deviation Interlock Out. of .

Comminaion (Technical Specifications, Table 3.2)

Gent 1emon:

This letter is Reportable occurrence No. 16, startup of the Ford Nuclear Reactor with the power level deviation interlock as required b3 Technical specifications, Table 3.2, out of commission.

The event occurred on the afternoon of November 24, 1992, during a startup to cheth core reactivity that followed the performance of reactor pre-startup Checklist A. The fact that the interlock was out of commission was discovered during a subsequent performance of Checklist A on the midnight shift, November 25, 1992.

Sequence of Events on Tuesday, hvember 24, 1992, during a routine maintenance period, the shim nmge - control rod interlock system was removed from the reactor control system. The modification hc.d been reviewed and _

approved by t.h + facility Safety Review Committee on October 6, 1992.

Following the wi. ring modifications, a reactor pre-startup Checklist A was conducted to test the modification and to confirm the operation of the reactor control system. A wiring error had been raade during the modification that disabled the interlock that drops the reactor out of automatic control if the Linear Level neutron detection system indicated power is five percent below the automatic control setpoint, The fact that. this interlock was disabled was not discovered during the performance of Checklist A. A reconstruction of the performance of the check 1ist revealed the possible reason. A copy of step 33 of Checklist A is enclosed for reference.

1. Step 33 is the d rop-ou t -o f-au to test.

2. A senior reactor operator was at the console; a trainee was assisting with the checkout.

3. In step 33.a, the Linear Level selector switch is set to CALIBRATE which gives a power indication of 100%.

E ]

- - ~ ~ , , . . - . - , ~ - . . . . ~

E

-4 Reportable Occurrence No.'_16--

• i' United States NuclearfRegulatory. Commission-November 25, 1992 9

4. In step 33.b, the servo control _setpoint is set to 97%:

which permits the control rod to be put in automatic control. Indicated power must be at least 2% above the automatic control setpointito go--into automatic.

5. In Step 33.c. the control rod is manually withdrawn to 17-i_ inches. Zero to 24 inches is the full stroke of theirod.

l 6. In step 33.d, the reactor is placed in._ automatic control._ ~-c '

o Since the indication (100%)is above-the setpoint(97%),_the rod will begin to insert.

x,

7. In step 33.e, when the control rod is at 15 inches-and-going in, the operator increases the setpoint. _W hen the setpoint reaches 100%, control rod n.otion- stops.- As the setpoint is increased above 100%, the control' rod  ;

withdraws. At a setpoint of 105%, the control rod drops out of automatic, satisfying the interlock. ,

8. In this case -the trainee must have increased the setpoint too-slowly. It-takes approximately'28 seconds for the~ -

control rod to travel from 15. inches to 0 inches. If she '

did not reach a 100% setpoint within that 28. seconds, the control rod would reach the lower.1'imit (0-inches)-and thef control-rod would drop out of automatic.= Reaching'the lower limit is accompanied =by a rundown 1(automatic insertion) of the' shim-safety rods.

9. In -f act the- drop-out-of-auto 1was ' observed by the E senior reactor operator, as wasca-cundown._ He misinterpreted-the rundown and-believed tho' drop-out-of-auto was due-to-ther Linear Level-indication being-five_ percent-below the ~

r setpoint, thus satisfying. step 32.- ,

Following the " successful" completion of Checklist A,_the_: reactor:

was started up to ~approximately~5_kw to perform.asshutdown margin and excess reactivity 1 check. The control rod'was withdrawn:toJ24

- inches-and the-shim. rods to crit'icality. The'reactorewastnever .

placed in automaticEcontrol. Following the reactivity measurement, the reactor was: shutdown and secured.

-r The midnight shift. November 25, 1992, performed a Checklist A--in

- preparation for. power-operation. LThe crew was'not'able1to .

successfully complete Step.33.s -The Assistant ManagerLand'the, F Electrical. Engineer were? notified 4 - Between 2 a.m.-and 8 a.m.., the procedure was retried, the wiring. mistake was discovered-and-corrected, and-Checklist A was satisfactorily completed..

The personnel who_had conducted the' original Checklist A were '

-interviewed and the.seriesTof-errors.-and misinterpretations _were- ,

- resolved. At_9 a.m.', November-25, 1992, the: Reactor Manager gave-

permission to

startup the reactor for power operation.

L .

Page-2 4 e - , , .,w-. , ,"

F T T1 Y t

L Reportable Occurrence No. 16

. United States Nuclear Regulatory Commission November 25, 1992 l

Consequences of Operating Wi th t.he pc..ar Level Deviation Interlock Out. of Commission The worst possible consequence of operating with the power-level deviation interlock out of commission would be failure of the automatic control system followed by continuous withdrawal of the contro.1 rod. If the control rod were withdrawn from zero to 24 inches, 0.00475 delta K/K reactivity would be inserted. The-resultant nLable period would be t. seconds. At a period of 10 3 secotid s , an automatic rundown of t.he shim-safety rods occurs- The period scram setpoint is 5 seconds.

s Corrective Action The following changes to Checklist A will be made to ensure proper performance of Step 33.

33.c Withdraw the control rod to 23 inches.

33.e As the control rod begins to insert, increase the setpoint t. o 106%.

33.g Verify final control rod position above 5 inches.

Step 33.c will provide more time to adjust the setpoint. Step 33.g will ensure that a control rod lower limit d rop- o u t.-o f -au t o is not misinterpreted as a powe r level deviat ion drop-out-o f-nu to.

As a more general corrective action, when a modification is made, the functional changes related to the modification will be tested and verifled by the quality assurance. team members directly responsible for the modification. The Electrical Engineer was present during the initial performance of Checklist A, but neither he nor any other member of the quality assurance team was observing the pe r f o rmarm e of the checklist step-by-step.

Sincerely, Ronald F. Fleming Director xc: i> r . Wil'11am C. Kelly, Vice president. for Research Safety Review Committee Members Reactor Operations Staff United States Nuclear Regulatory Commission Attn: Charles Norelius 799 Roosevelt Road Glen Ellyn, .Ill i nc i s 60137 i.

Page 3

. .- -~_ . .

3

_,p  :-

. _1 .

l '

\

l r- FNR CONTROL SYSTEM STARTUP CHECELIST 'A' (continued)

Initial

30. Energize Linear Level automatic control unit.

31. Perform Log N channel tests:

a. Turn LOG N test switch to positions 1, 2, and 3 to check calibration of Log N meter and recorder,

b. Return Log N test switch to OPERATE.

c. Turn Log N period test switch from OPERATE to CALIBRATE and check the calibration of the period meter and recorder,

d. Press SCRAM RESET button. .

e. Switch period recorder 0FF while withdrawing control rod and verify control rod withdrawal inhibit. _

f. Return period recorder switch to the ON position. s 8 Turn Log N TRIP TEST ADJUST slowly to an indicated 30 second period and verif y control rod withdrawa! inhibit,

h. Check auto rundown at 10 seconds.

1. Return Log N test circuitry to OPERATE.

32. Perform Linear Level rundown test,

a. Simultaneously press TEST button and adjust potentiometer to read 115X.

b. Note auto rundown,

c. Adjust potentiometer to read 95X and release TEST button.

d. Depress auto rundown RESET button.

33. Perform drop-out-of-auto test,

a. Set Linear Level range selector switch to calibrate, j21 b, Set servo control setpoint at 97%.

c. Withdraw control rod to 17.0 inches.

d. Depress AUTO ON push button on console,

e. When control rod inserts to 15.0 inches move setpoint to 106. _

f. Note drop out of-auto control.

34. Perform control rod lower limit rundown test.

3. Set auto-control setpoint at 97%.

b. Depress AUTO ON button, l16

c. Note auto rundown when control rod reaches lower limit,

d. Depress auto rundown RESET button.

e. Set Linear Level range selector switch to lowest possible operating range.

35. Perform Linear Level abnormal test.

a. Press and hold LINEAR LEVEL ABHORMAL TEST button.

b. Note LIN LEV ABN alarm.

c Note 110% reading on console Linear Level meter.

d. Release TEST button.

36. Return magnet contact simulator switches to 0FF.

37. Return shim range bypass switch to NONE BYPASSED.

38. Unbypass SHIM RANGE DEFEATED scram function.

39. Verify D20 LOW LEVEL alarm unbypassed.

40. Check po61 lev i functions.

a. Test alarm at 5 inches.

b. Test auto rundown at 12 inches.

Page S m-____m _____.__.___m