ML20046A491

LER 93-003-00:on 930625,SG Tripped Due to Low Water Level. Caused by Insufficient Feedwater Addition Due to Inadequate Communication.Reemphasis on Improved Communication Stressed. W/930722 Ltr

ML20046A491
Person / Time
Site:	Calvert Cliffs
Issue date:	07/22/1993
From:	Cruse C, Muth D BALTIMORE GAS & ELECTRIC CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-93-003-01, LER-93-3-1, NUDOCS 9307280187
Download: ML20046A491 (7)

Text

,

d

. BALTIMORE GAS AND

. ELECTRIC CALVERT CLIFFS NUCLEAR POWER PLANT 1650 CALVERT CLIFFS PARKWAY

• LUSBY, MARYLAND 20657-4702 CHARLES H. CRUSE ,

PLANT GENERAL MANAGER cALVERT CUFFS July 22, 1993 U.S. Nuclear Regulatory Commission Washington, D.C. 20555 ATTENTION: Document Control Desk .j

SUBJECT:

Calvert Cliffs Nuclear Power Plant Unit No. 2; Docket No. 50-318; License No. DPR 69 Licensee Event Report 93-003 Trip on Low Steam Cenerator Level Due to Insufficient l Feedwater Addition H The attached report is being sent to you . as required under '10 CFR 50.73 .!

i guidelines. Should you have any questions regarding this report, we will be pleased to discuss them with you.

Very truly yours, l

l CHC/DWM/bj d Attachment g/ ./ Mfg /

,7 cc: D. A. Brune, Esquire  ;

J. E. Silberg, Esquire  ;

R. A. Capra, NRC I D. C, Mcdonald, Jr. , NRC ]

T. T. Martin, NRC  ;

P. R. Wilson, NRC =

R. I. McLean, DNR J. H. Walter, PSC Director, Office of Management Information and Program Control 2800as -

S

/19#8

'9307280187 930722 {

DR ADOCKOSOOg8

NRC FORM 366 U. S. NUCLEAR REGULATORY COMMISSION WW No MM M2) .

ESTIMATED BUF0EN PG KSFONSE TO COMFUTWITH TKB INFC3AAT10N

. CCLIEctlON KOUEST: 5a0 HH1 FOFWAFU COMMENTS KoAFDtG EMN ESTWE TO THE INFCFWON AND KCORDS MAPMGEMENT L1CENSEE EVENT REPORT (LER) BMNCH (MNDB T/14). U.S6 NUCLEAR KoVLATOFNtNme"N, WASHINGTON, DC 2m550001, AND TO THE PAPEFWORK KDUCTION PRCUECT p15)0104), OFTICE OF MAPMOEMENT AND BUDGET, WASHINGTON DC 20Sn (See reverse for reauired number of diaits/ characters for each block)

FACILITY NAME (1) DOCKET NUMBER (2) PAGE (3)

Calvert Clif f s, Unit 2 05000 318 1 OF 06 TITLE (4)

Trip on Low Steam Generator Level Due to insufficient Feedwater Addition EVENT DATE (5) LER NUMBER (6) REPORT DATE (7) OTHER FACILITIES INVOLVED (8)

MOMTH DAY YEAR YEAR MO*(rH DAY YEAR '

NUMBER NUMBER 06 25 93 93 - 003 -

00 07 22 93 05000 j OPERATING THit REPORT IS SUBMITTED PURSUANT TO THE REoVIREMENTS OF 10 CFR :(Check one or more) [11) 2 20.402(b) 20.405(c) )( 50.73(a)(2)(iv)

MODE (9) 73.7,(3) 20.405(a)(1)(i) 50.36(c)(1) 50.73(a)(2)(v) 73.71(c)

LEVEL 3 2n405(ani>00 sa36(cx2) Sa73(an2)(va g7gg,,

~

(10) 20.405(a)(1)(iii) 50.73(a)(2)(i) 50.73(a)(2)(viii)(A) (Specify in Abstract below and in Text, NRC Form 30GA) 20.405(a)(1)(iv) 50.73(a)(2)(ii) 50.73(a)(2)(viii)(B) 20.405(a)(1)(v) 50.73(a)(2)0ii) 50.73(a)(2)(x)

LICENSEE CONTACT FOR THIS LER (12)

NAME TELEPHONE NUMBER (include Area Code)

D. W. Muth, Compliance Engineer 410-260-3592 COMPLETE ONE LINE FOR EACH COMPONENT FAILURE DESCRIBEDIN THIS REPORT (13)

CAUSE SYSTEM COMPONENT

" #0 CAUSE SYSTEM COMPONENT

^

TURER NPRDS TURER NPROS l

l SUPPLEMENTAL REPORT EXPECTED (14) EXPECTED MONTH DAY YEAR ygg SUBMISSION (if yes complete EXPFCTED SURMISSION DATF) DATE (15)

ABST RACT (Limit to 1400 spaces Lo., approximately15 singlespace typewritten hnes) (16)

On Thursday, June 25, 1993 at 8:05 p.m., Calvert Cliffs Unit 2 automatically tripped on low level in the 21 Steam Generator (SG). Operators had noted divergent SG 1evel oscillations with the Full Range Digital Feedwater Control System (FCS) in automatic. They took manual control with SG level at

+25 inches but did not provide sufficient feedwater. The reactor tripped on low SG level.

The causes of this event include inadequate communication of pertinent information regarding the response of the FCS at low power, inadequate communication during shif t turnover regarding a just-completed power increase, the lack of Project Team involvement at the time of the incident, and the work practices of the operations personnel involved in this incident.

Operations management will reemphasize expectations for the improvement of communications between operating crews. An FCS Project Team representative will be available to the Control Room for future startups and planned shutdowns until FCS performance meets expectations. We will provide classroom training to all operating crews on the details of this event.

iitFormn(594

I NRC FORM 366A U.S. HUCLEAR REGULATORY COMMISSION APPROVED BY OMB NO. 3150 0104 002 EXPIRES 5/31/95

. ESTIMATED DURDEN PER RESPONSE To COMPLY WTTH THis HFoRMATON CoLLECTloN REoUEST: 50.0 HRS, FORWARD LICENSEE EVENT REPORT (LER) COMMENTS REGARDWG BURDEN ESTIMATE To THE MoRMATON AND RECORDS MANAGEMENT BRANCH (MNBB 7714), U.S. NUCLEAR TEXT CONTINUATION nEGutAToRv CouuissoN, wasswGToN. DC 20suact, ANoTo THE PAPERWORK REDUCTON PROJECT (31500104), oFFCE oF MANAGEMENT AND BUDGET, WASHINGTON, DC 20003.

FACIUTY NAME (1) DOCKET NUMBER (2) LIR NUMBER (3) PAGE (4)

Calvert Cliffs, Unit 2 05000 3 1 8 93 -

003 - 00 02 0F 06 TEXT pr move space b required, use addmonaJ copes of NRC Form 368A} {17)

I. DESCRIPTION OF EVENT On Thursday, June 25, 1993 at 8:05:09 p.m., Calvert Cliffs Unit 2 automatically tripped on low level in the 21 Steam Generator (SC). Operators had noted divergent level oscillations with the Feedwater Control System (FCS) in automatic. They took manual control with SG 1evel at +25 inches but did not provide sufficient feedwater. The underfeed condition lasted approximately 35 minutes at which time level reached -50 inches, tripping the reactor. The Unit was at 2.8 percent power in MODE 2 at the time of the event.

In response to previous problems controlling feedwater during startup, a new Full Range Digital FCS was installed in Unit 2 during the recently completed refueling outage. The FCS was first used during initial startup on June 12, 1993. It was noted to perform particularly well below 2 percent and above 8 percent power, with minor, gradual oscillations in SG level. However, significant oscillations were observed as power was increased between 2 and 8 percent power. These oscillations had a maximum amplitude of approximately 25 inches and period of about 30 minutes. The Operations crew performing this phase of the startup considered the oscillations acceptable when compared either with the old automatic system or with what was possible in manual and therefore did not communicate them to other Operations shifts.

On June 25, 1993 day shift had begun plant startup and had been controlling feedwater in automatic most of the day with no abnormalities noted. Power was increased from 0.1 percent to 2.8 percent just as day shif t was ending. The change in power was not adequately communicated during shift turnover.

The night shif t came on at about 6:00 p.m. , noting the FCS in automatic cind I reactor power at about 3 percent. They believed that the reactor had been at this power level for some time. At about 6:30 p.m., the Control Room Supervisor (CRS) noted level oscillations beginning in the 21 SG and began monitoring level l closely. At about 7:15 p.m., the CRS directed the Control Room Operator (CRO) I to take manual control. The CRS had previously been informed that the feedwater regulating bypass valve had been modulating in automatic between 18 and 26 percent open throughout most of the previous shift and assumed, since he  !

believed the plant had been at about 3 percent power during this time, that this was the proper-valve setting for this power level. At the time he ordered the bypass valve to be taken into manual, the CRS noted the valve to be at 32 percent. He therefore assumed that this was too far open and that the FCS was not controlling properly.

l NRC FORM 366A U.S. NUCLEAR REGULATORY COMMISSION APPROVED BY OMB NO.3150-0104 (Fea EXPIRES 5/31/95

. EEMATED BURDEN PER RESPONSE To COMPLY WITH THtS lNFoRMAT1oN CouECT1oN REoVEST: 50.0 HRS. FORWARD LICENSEE EVENT REPORT (LER) COMMENTS REGA5DNQ BURDEN ESDMATE To THE INFoRMATM j ANo RECORDS MANAGEMENT BRANCH (MNB8 7714), U.S. NUCLEAR TEXT CONTINUATION REGuuToRv Couuissa, wAsmNGioN. oC 20ssmot AND To THE PAPERWORK REDUCTION PROJECT (3150-0104), OFFICE oF l

MANAGEMENT AND BUDGET, WA9HINGToN, DC 20503. l l

F ACluTY NAME (1) DOCKET NUMBER (4 LER NUMBER (3) PAGE (4)

Calvert Cliffs, Unit 2 05000 3 1 8 93 -

003 - 00 03 0F 06 TEXT in o,. w. ,.w,.a. .aoiuoa.i cop, oe NRc For. 38e4) (i7)

The CRS did not realize that what had actually happened was that, with the recent increase in power to 2.8 percent, the FCS had entered the region in which increases in power produced significant SG oscillations. At the time the bypass valve was taken into manual, SG 1evel was peaking and the FCS had closed the valve to bring level back down. With the bypass valve at 32 percent, SG 21 was actually in an underfeed condition. 1 The CR0 took manual control with SG level still increasing and at +25 inches.

He used the manual push-button controller to slightly close the valve. He noted that SG 1evel began decreasing almost immediately and assumed that this was the result of his actions. Actually, as noted above, the FCS had already closed the valve sufficiently-and the level decrease was due to this prior automatic action. The CR0 began briefly depressing the control button to provide small open commands to the feedwater bypass valve with the intention of bringing level smoothly to zero. As SG 1evel continued to decrease rapidly (3-4 inches per minute), the CR0 continued to apply slight open commands to the valve that he thought more than compensate 3 for the closure signal with which he started.

The previous SG 1evel control system had used a knob to control valve position.

The new system uses a membrane push-button with a logarithmic response. The longer the button is depressed, the faster the valve opens. The CRO was aware of the functioning of the new controller but was not aware of how little  ;

response his short presses of the button actually produced. He did not use controller output indication to obtain feedback on the effectiveness of his j actions. He had previously been successful monitoring only SG 1evel indication and controller knob position when controlling SG 1evel with the old system.

1 Within 10 minutes after manual control was assumed, level dropped below zero and I was not responding appreciably to operator actions with the bypass valve. The CR0 continued single depressions of the controller in hopes of a gradual approach to zero inches. About five minutes later, with level approaching

-15 inches, the CR0 considered more aggressive feedwater injection but was I concerned with the effects of level shrink if large quantities of cold water were injected into the SG. When level reached about -20 inches, the CR0 took additional measures to increase level, including isolation of SG blowdown, placing the 22 feedwater bypass valve in manual and closing it slightly, and increasing 21 SG Feed Pump speed. The combination of these measures were effective in eventually terminating the drop in level at about.-45 inches.

However, minor level oscillations occurred, and low-level trip signals were j received by Safety Channels C and D. The reactor tripped at 8:05:09 p.m. The i total elapsed time for this event was approximately 45 minutes.

NRC FORM 366A ~ U.S. NUCLEAR REGULATORY COMMISSION APPROVED BY OMB NO. 3150-0104 pca EXPIRES 5/31/95

. ESinMA4D BURDEN PER RESPONSE To COMPLY WITH THis INFoRMATioN CoufCTION REQUEST: 50.0 HRS. FORWARD LICENSEE EVENT REPORT (LER) cowuENTs REoARDmo suRoEN EsTiuATE To THE mroRuAroN AND RECORDS MANAGEMENT BRANCH (MNilB T714), U S. NUCLEAR TEXT CONTINUATION REcutAToRY CouussoN, WASHm0 ton, DC 205554m, AND To THE PAPERWORK REDUCToN PROJECT (31540104), oFFoE oF MANAGEMENT ANo BUDGET, WASHINGTON, DC 20503.

F ACluTY NAME (1) DOCKET NUMBER (4 LER NUMBER (4 PAGE (4)

Calvert Cliffs, Unit 2 05000 3 1 8 93 -

003 - 00 04 0F 06 Ttna .v.c. % .m iwe aNRCr maca w)

The appropriate Emergency Operating Procedures were performed without incident.

II. CAUSE OF EVENT There are several causes of this event. The first is that the response of the FCS at low power was not known by Operations shifts other than the one that initially started up using the new system. This crew had observed the unanticipated oscillations, concluded that they were acceptable, and therefore did not communicate them to any other crews. The oscillations had not been discussed in training or modelled on the simulator as they had not been anticipated by the team developing the training. The crew involved in this event came on shift expecting no significant oscillations from the FCS.

A second cause is that the shift turnover did not address the increase in power that took place shortly before. This left the oncoming shift with no explanation for the SG level oscillations other than problems with the MS.

Knowledge of the power change, particularly if combined with information from the initial startup, may have helped the operators anticipate the level oscillations and either leave the FCS in automatic or understand better the need for increased feed flow in manual.

A third cause is the lack of Project Team involvement on this shif t. The new FCS was sufficiently complicated that not all plant response was anticipated. A similar system at another plant maintained a flat level trend at all power levels. The response of the FCS had been compensated for during initial startup and even during the day shift prior to this event by the presence in the Control Room of a member of the FCS Project Team. This individual had proven valuable during initial startup by discussing the oscillations with the FCS manufacturer and advising the shift crew on how to respond. A member of the team could have provided a similar service during this shift.

A fourth cause of thie event was the work practice of the CRO, who continued to underfeed the SG for about 35 minutes. He did not use controller output indication to verify the amount of movement the valve was making in response to his depressions of the control switch. A better understanding of the relationship of controller operation to valve movement might have resulted in more aggressive actions prior to the point where level shrinkage effects were significant. Crew supervision, including the SRO, failed to provide the CR0 sufficient coaching.

'NRC FORM 366A U.S. NUCLEAR REGULATORY COMMISSION APPROVED BY OMB NO. 3150-0104 sea EXPIRES 5/31/95

. ESTIMATED BURDEN PER RESPONSE To COMPLY W"H THIS INFoRMATloN Cou.ECTioN REouEST: 50.0 HRS. FORWARD LICENSEE EVENT REPORT (LER) COMMENTS REGARDING BURDEN ESTIMATE To THE INFoRMAThoN AND RECoRDG MANAoEMENT BRANCH (MNBB 7714), U.S. NUCl. EAR TEXT CONTINUATION REaut4 Tory CouuissoN, wAssNoToN. DC 20ss>ooot. AND To THE PAPERWORK REDUCTON PROJECT (31500104), OFFICE oF MANAGEMENT AND BUDGET, WASHINGTON, DC 20503.

f ACluTY NAME (1) DOCKET NUMBER (2) LER NUMBER (3) PACE (4)

Calvert Cliffs, Unit 2 05000 3 1 8 93 -

003 - 00 05 0F 06 TEXT (If more space is required, use addmonal copes of NRC Form 300A) (17)

We have since experienced problems with automatic FCS control causing Operators to place FCS in manual. This indicates a need to improve FCS performance to reduce challenges to SG level control.

III. ANALYSIS OF EVENT The worst-case loss of feedwater flow transient described in the Updated Final Safety Analysis Report assumes a total loss of feedwater flow at full power and concludes that no significant safety consequences will result from this event.

This analysis is bounding for this event. There are no significant safety consequences resulting from this event.

This item is reportable under the provisions of 10 CFR 50.73(A)(2)(iv) as a Reactor Protective System actuation.

IV. CORRECTIVE ACTIONS A. Operations managece t will reemphasize expectations for communications bett an operating crews through better use of available mechanist (i.e., operator logs, turnover sheets, and turnover briefings) Operations management will review and discuss situational leadersi p with Shift Supervisors.

B. An FCS Froject Team representative will be available to the Control Room for fdture startups and planned shutdowns until FCS performance l meets expectations, and Operations personnel have gained sufficient i experience with the system.

J C. We will provide classroom training to all operating crews on the details of this event. We will also give Operators additional i hands-on training on the use of the push-button controller in l conjunction with valve controller output indication for the control of steam generator level.

D. FCS will be modified to improve performance.

i I

s - _ _

i

. . )

NRC FORM 366A U.S. NUCLEAR REGULATORY COMMISSION APPROVED BY OMD NO. 3150-0104 pea EXPIRES 5/31/95

• EST1 MATED BURDEN PER RESPONSE TO COMPLY WITH THIS i INFORMATION COLLECTION REQUEST: 50.0 HRS FORWARD l LICENSEE EVENT REPORT (LER) cOuuEurS REGARolNG BuRoEN ESTiuATE To THe woRuAToN AND RECORDS MAfuGEMENT BRANCH (MNBB 7714), U S. NUCLEAR TEXT CONTINUATION REGULATORY COMMtSSON, WASHINGTON. DC 20$5S@01, AND TO j THE FAPERWORK REDUCTION PROJECT (3150-0104), OFFICE OF MANAGEMENT AND BUDGET, WASHINGTON, DC 20503.

FACILITY NAME (1) DOCKET NUMBER (2) LER NUMBER (3) PAGE (4) l Calvert Cliffs, Unit 2 05000 3 1 8 93 -

003 - 00 06 0F 06 L_

Tr.xT of more .p.co is requ6 red, use adchtional copies of NRC Form 306A) (17)

V. ADDITIONAL INFORMATION A. Affected Component Identification:

IEEE 803 IEEE 805 Component or System EIIS Funct System ID Steam Generator MX SJ Feedwater Control System TC SJ  ;

Feedwater Regulating Bypass Valve LCV SJ Manual Push-button Controller LCO SJ B. Previous Similar Events:

LER 50-317/85-009 described a trip on low SG water level from 19 percent power due to underfeeding the SG in manual. ,

j i

l