ML20034E961

Responds to to Chairman Selin Re NRC Response to Event at Plant.Insp Rept 50-443/93-01 Provided as Explanation of NRC Actions

ML20034E961
Person / Time
Site:	Seabrook
Issue date:	02/23/1993
From:	Martin T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
To:	Gavutis S CITIZENS WITHIN THE TEN MILE RADIUS
References
NUDOCS 9303020130
Download: ML20034E961 (3)

See also: IR 05000443/1993001

Text

.C

,

(

a nts

g

UMITED STATES

J'.

, , ,

.. y

([ g

NUCLEAR REGULATORY COMMISSION

1

C'

REGION I

54\\

/.

S.1

'[

475 AliLNDALE ROAD

% .' .' .J

.

a

8

KING OF PRUSSIA, PENNSYLVANIA 19406 1415

.+

,-

-

feS 2. 3

M 13

i

Ms. Sandra Gavutis

Citizens Within the Ten Mile Radius

P.O. Box 382

Amesbury,. Massachusetts 01913

Dear Ms. Gavutis:

On January 8,1993, you sent a le-tter to Chairman Selin that expressed your concerns

regarding the NRC response to an event at the Seabrook station.

Like you, I was disturbed by the press report of a statement by our Resident. Inspector,

Richard laura, that left the impression' that the NRC found fault v>ith reactor operators, who

had appropriately shut down the Seabrook plant in response to equipment failures. As you

correctly pointed out in your letter, the NRC's mission in the regulation of operating reactors

is to assure the public health and safety. We take this responsibility very seriously, and

clearly any implication that we encourage imprudent risk is unacceptable.

As you suggested, I talked to our Resident Inspector and his management to understand what

had happened. The Resident Inspector was equally disturbed by the implications of the

l

newspaper report and strongly disavowed any criticism of the reactor operators for tripping

the reactor. 'Ihe newspaper report is the result of a very long interview and is essentially

accurate in what it says, but is misleading in its brevity and implication.

.

The Resident Inspector was and. remains concerned that. the.opprating staff.had mad. a-series

.

~

e

'

of errors in responding to an equipment malfunction. The inspector knew at the time of his'

,

"

interview that plant staff had failed to properly restore automatically isolated feedwater heater

strings, resulting in a complete loss of feedwater flow to the steam generators. It was that

l

!

error that our resident inspector pointed out to the reporter. The inspector was also aware at

that time that the control room operators, who recognized the adverse condition, had acted

.{

!

swiftly and correctly to bring that adverse condition to an end by manually shutting down the

reactor. The Resident Inspector was not and is not concerned that the plant operators

decided to trip the plant, since such action was clearly called for by the imposed transient.

Further, the inspector believed the original problem, a leaking feedwater heater relief valve,

could have been isolated without a loss of feedwater flow, thus avoiding a condition that

required plant operators to trip the reactor; a belief that has since been borne out by our

review of the event. I encourage you to read the attached special ins).ection report that

provides a more detailed event description.

l

{M $$$ I>Nh3

[(d

p

u

'

.

,.

4

t

2

i

'

With this perspective of the inspector's concerns, I understand what the reporter chose to

.

say, but I am disappointed that we were not successful in getting a more complete description

of our concems and their basis to the public than was provided in the newspaper report. I

appreciate the sincerity of your concern and am pleased that I could provide you with this

explanation of our actions and trust that I have satisfied your concerns. I can assure you that

i

the NRC has a strong commitment to protect the public health and safety.

Sincerely,

,

&.

Thomas T. Manin

'

Regional Administrator

Encl:

-

_

NRC Special Safety Inspection

Report No. 50 443/93-01

cc: (w/ Encl.)

_

,

Public Document Room (PDR)

l

Local Public Document Room (LPDR)

Resident Inspector

!

i

e

t

1

i

.

p afc

oq[0,

UMITED STATES

o

NUCLEAR REGULATORY COMMISSION

y'* } 3 .,.g

-'

,

g

5 E

REGION I

.,

!

475 ALLENDALE ROAD

o,

5

KING OF PRUSSIA, PENNSYLVANIA 19406-1415

JM n g3

Docket No. 50443

Mr. Ted C. Feigenbaum

Senior Vice President and Chief Operating Officer

North Atlantic Energy Service Corporation

Post Office Box 300

Seabrook, New Hampshire 03874

Dear Mr. Feigenbaum:

SUBJECT: NRC SPECIAL SAFETY INSPECTION REPORT No. 50-443/93-01

This refers to the special safety inspection findings of the resident inspectors at Seabrook'

Station for the period of January 3-20,1993. The inspectors evaluated station response to

the reactor trip that occurred on January 3,1993.

The in;pection concluded that your staff operated the facility in a safe and conservative

manner following a loss of feedwater flow. The NRC review and your post trip review

identified several weaknesses which contributed to the loss of feedwater flow. Your short-

term and long-term corrective actions address these weakness and should enhance future

facility operations.

Thank you for your cooperation.

Sincerely,

Y

A. Randolph Blough, Chief

Projects Branph No.4

Division of Reactor Projects

-

Enclosure:

NRC Special Safety Inspection Report No. 50-443/93-01

,

E

>0>0 4 0

. ~ . 4 - North Atlantic Energy Service Corporation 2 - cc w/ encl: B. L. Drawbridge, Executive Director of Nuclear Production W. DiProfio, Station Manager R. Hallisey, Director, Dept. of Public Health, Commonwealth of Massachusetts S. Woodhouse, legislative Assistant T. Rapone, Massachusetts Executive Office of Public Safety D. Tefft, Administrator, Bureau of Radiological Health, State of New Hampshire Public Document Room (PDR) local Public Document Room (LPDR) Nuclear Safety Information Center (NSIC) K. Abraham, PAO (2) NRC Resident Inspector State of New Hampshire, SLO Commonwealth of Massachusetts, SLO Designee Seabrook Service List

, > a > ! ! k - t i s i

_ _ _ _ _ _ _ _ - _ m

~ . _ S.EABROOK SERVICE LIST ~ Thomas Dignan, Esquire Diane Curran, Esquire John A. Ritscher, Esquire Harmon and Weiss Ropes and Gray 2001 S street, N.W. One International Place Suite 430 Boston, Massachusetts 02110-2624 Washington, D.C. 20009 Mr. J. F. Opeka Regional Administrator, Region I Northeast Utilities U. S. Nuclear Regulatory Commission P. O. Box 270 475 Allendale Road Hartford, Connecticut 06141-0270 King of Prussia, Pennsylvania 19406 Mr. A. David Rodham, Director John P. Arnold, Attorney General Massachusetts Emergency Management Agency G. Dana Bisbee, Associate Attorney 400 Worchester Road, Box 1496 General Framingham, Massachusetts 01701-0134 Attorney General's Office ATTN: Mr. James B. Muckerheide 25 Capitol Street State Nuclear Engineer Concord, New Hampshire 03301 Robert Backus, Esquire Mr. R. M. Kacich Backus, Meyer and Solomon Northeast Utilities Service Company 116 Lowell Street P. O. Box 270 Manchester, New Hampshire 03106 Hartford, Connecticut 06141-0270 Mr. T. L. Harpster Office of the Attorney deneral , North Atlantic Energy Service Corporation One Ashburton Place Post Office Box 300 20th Floor Seabrook, New Hampshire 03874 Boston, Massachusetts 02108 Seacoast Anti-Pollution Imague Town of Exeter l 5 Market Street 10 Front Street Portsmouth, New Hampshire 03801 Exeter, New Hampshire 03823 Mr. David W. Graham Stephen G. Burns, Director Fuel Supply Planning Manager Office of Commission Appellate Massachusetts Municipal Wholesale Adjudication Electric Company U. S. Nuc1 car Regulatory Commission - , P. O. Box 426 Washington, D. C. 20555 Ludlow, Massachusetts 01056 !

' . , , . Gerald Garfield, Esquite Board of Selectmen Day, Berry and Howard Town of Amesbury City Place Town Hall Hartford, Connecticut 01603-3499 Amesbury, Massachusetts 01913 Resident Inspector Mr. Peter Brann , U.S. Nuclear Regulatory Commission Assistant Attorney General Seabrook Nuclear Power Station State House, Station #6 Post Office Box 1149 Augusta, Maine 04333 Seabrook, New Hampshire 03874 Jane Spector Adjudicatory File (2) ' Federal Energy Regulatory Commission Atomic Safety and Licensing Board Panel 825 North Capital Street, N.E. Docket Room 8105 U.S. Nuclear Regulatory Commission Washington, D.C. 2042.6 Washingtei, D.C. 20555 Mr. George L. Inverson, Director Mr. Jack Dolan New Hampshire Office of Emergency Federal Emergency Management Agency Management Region I State Office Park South J.W. McCormack Post Office and 107 Pleasant Street Courthouse Building, Room 442 Concord, New Hampshire 03301 Boston, Massachusetts 02109 , Atomic Safety and Licensing Board Congressman Nicholas Mavroules Panel U.S. House of Representatives , U.S. Nuclear Regulatory Commission 70 Washington Street Washington, D.C. 20555 Salem, Massachusetts 01970 P 5 f. A D 1 &

- - . . ~ - . . - , .. . - . . .. . U. S.~ NUC' LEAR REGULATORY' COMMISSION ' - , REGION I , ! Report Number: 93-01 1

Docket No.: 50-443 , l License No.: NPF-86 .. Licensee: North Atlantic Energy Service Corporation - l Post Office Box 300

Seabrook, New Hampshire 03874 .l 1 Facility: Seabrook Station Dates: January 3-20, 1993

Inspector: Noel Dudley, Senior Resident Inspector i $ Richard Laura, Resident Inspector - Fred Paulitz, Electrical Engineer 3 Approved By: 6fohn F. Rogge,Wief Date - Reactor Projects Section 4B . ! OVERVIEW '! ! The operators manually isolated a feedwater heater to repair the bellows on a leaking tube I side relief valve. A subsequent feedwater system transient isolated a string of feedwater. i heaters. Incorrect operator actions during restoration of the feedwater heaters resulted in . ! completely isolating feedwater to the steam generators. The control room operators manually , tripped the reactor and stabilized plant conditions. . The station staff completed a post-trin l review prior to starting up the reactor and conducted an event evaluation.

.I [ Operators performed well in manually isolating the leaking relief valve. Operator errors l made while restoring the feedwater heaters caused a loss of feedwater flow. Improper verbal 4 communications, the lack of written guidance for abnormal feedwater transients, weaknesses j in operator training, and poor human factors at a turbine building control panel contributed to the operator errors. Operator response to the loss of feedwater conservatively demonstrated i an appropriate safety attitude. ! Plant staff fully understood the secondary plant transient for a single heater isolation. However, the staff misunderstood the effects of multiple feedwater heater isolations on plant i operations, since they believed that loss of two E-21/22 heater strings would result m a a reactor trip. ] The licensee's assessment of the event was thorough and well documented. The . recommended corrective actions in the station information report addressed the causes of the- event. . 1 g m o f o vg. ' u ' . . .. .

- . . ., . TABLE OF CONTENTS . O V ER V Y EW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i TABLE OF LONTENTS .......................................ii 1.0 INS PECTION S COPE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 2.0 EVENT OVERVIEW (71707, 93702) . . . . . . . . . . . . . . . . . . . . . . . . . . . I 2.1 Basic Description of Heater Operation . . . . . . . . . . . . . . . . . . . . . . . I 2.2 Event S u m mary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3.0 OPERATIONS (71707,93702) 3 ............................... 3.1 Isolation of Feedwater Heater E-25A 3 ....................... 3.2 Manual Reactor Trip /less of Feedwater Flow . . . . . . . . . . . . . . . . . . 4 3.3 Con clu sions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4.0 TECHNICAL SUPPORT / ENGINEERING (71707) . . . . . . . . . . . . . . . . . . . 6 , 4.1 Feedwater Heating System Operating Experience . . . . . . . . . . . . . . . . 6 4.2 Abnormal Operating Configuration . . . . . . . . . . 7 ! .............. 4.3 Bypassing the EX-V1 Interlock . . . . . . . . . . . . . . . . . . . . . . . . . . 8 4.4 Relief Valve Failure Mode . . . . . . . . . . . . . . . . . . . . . . 8 ....... , 1 5.0 SAFETY ASSESSMENT / QUALITY VERIFICATION (71707) 8 ........... 6.0 SUMMARY........................................... 9 7.0 M EETING S (30702) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 A'ITACHMENT I - Sequence of Events , ATTACHMENT II - System Diagrams , b f 5i

, ,. .

. ' DETAILS 1.0 INSPECTION SCOPE The inspectors independently reviewed the events leading up to the loss of both main feedwater pumps and the subsequent manual trip of the reactor, and evaluated North Atlantic's assessment capabilities. The team observed portions of plant stabilization and reactor startup; attended management meetings; interviewed various operators involved with the event; reviewed the control room operating journal; attended the post trip review meeting; reviewed the main plant computer system time sequence log printout; reviewed the post trip review package including the operator statements; reviewed applicable operating procedures; reviewed the station information report; and held discussions with operations management, operator training, and technical support personnel. 2.0 EVENT OVERVIEW (71707,93702) 2.1 Basic Description of Heater Operation . Condensate and feedwater heaters preheat water before it is fed to the steam generators, thereby improving plant efficiency and reducing thermal stresses in the steam generator. The low pressure heaters are arranged into three parallel strings (A,B,C) in four series stages (E-21/22/23/24). Higher pressure heaters are arranged into two sets of parallel strings (E-25A/B, E-26A/B). Condensate and feedwater flows through the tube side of a heat exchanger and is heated by extraction steam on the shell side of a heat exchanger. Extraction steam is steam drawn from various points in the turbine steam system, including several stages of the turbine. The dynamics of heat transfer in the heaters can cause rather compicx interactions. As the extraction steam is condensed into liquid, the shell side of the heat exchanger drains (cascades) to a lower pressure heat exchanger, the main condenser, or the heater drain tank. Water collected in the heater drain tank is pumped directly into the feedwater by the heater drain pumps. Changes in the flow rate through a heat exchanger will affect the overall heat - transfer and balance of the system. For example, an increase in condensate flow through the tubes (as happened several times in the January 3 event to be described later) will increase the condensing of extraction steam, lower shell side pressure reducing the ability to drain, and possibly result in high water level conditions in the shell. Heaters are typically provided with level controls, high level alarms, and HI HI level isolations to prevent flooding the heaters, which could cause water to backup in to the turbine blade area. , The condensate and feedwater heater system, as well as its controls and indications, performs no reactor safety function and is therefore classified as non-safety-related. However, upsets in the system can cause plant transient conditions such as a loss of normal feedwater flow (i.e., the January 3 event), that would require safety systems to actuate. A loss of normal feedwater flow to the steam generators is a transient that was analyzed in the Final Safety Analysis Report and reviewed by NRC before plant licensing.

- - .

. 2 . Simplified one-line diagrams from North Atlantic systems descriptions for the condensate system, high pressure feedwater heaters, extraction steam system, heater drain system, extraction steam valve control logic, and feedwater (FW) heater E-26 level control logic are provided as attachment 2. 2.2 Event Summary On January 3, at approximately 3:36 a.m., with the plant operating at 100% power, the E- 26A heater extraction steam valve, EX-V1, shut on a HI HI level signal. Heater drain tank level and pressure decreased due to the loss of fluid flow from the E-26A heater.' The pressure at the suction of the main feedwater pump (MFP) decreased causing the third condensate pump to automatically start. The increased condensate flow through the heat exchanger tubes, resulting from the start of the third condensate pump, caused a HI HI level on the shell side of the E-21C heat exchanger. The E-21/22C heater string isolated and the bypass valve, CO-V-75, opened. The unit shift supervisor (USS) directed the auxiliary operator (AO), over the portable radio system, to restore the E-21/22C heater string. The AO opened the inlet valve to FW heater E-21C and closed the bypass valve, CO-V-75, from control panel CP-66 in the turbine building. However, the AO failed to open the discharge valve from the E-22C heater, and thus no flow was established in that string. With the bypass valve closed condensate flow further increased through the two remaining E- 21/22 heater strings and caused a HI HI level in FW heater E-21 A. The E-21/22A heater string isolated and the bypass valve, CO-V-75, opened. The USS again directed the secondary AO, over the radio, to restore the E-21/22A heater string. The AO repeated the valving sequence on this string, again failed to open the heater discharge valve, and closed the bypass valve, CO-V-75. Thus, all condensate flow was directed through the only remaining string, the E-21/22B string. The one remaining E-21/22 heater string isolated on HI HI level in FW heater E-21B. The E-21/22B heater string isolated and the bypass valve, CO-V-75, opened. Again, the secondary AO failed to open the discharge valve from FW heater E-22B during heater restoration. In each of the three attempts to restore the heater strings the USS improperly relied on a feedwater heater string isolation light in the main control room to confirm the restoration of the heater string and directed the AO to open the E-26A feedwater heater extraction steam valve, EX-V1. Upon closing the bypass valve, CO-V-75, for the third time all condensate flow was isolated. The 'A' MFP tripped on low suction pressure causing a turbine set back to 54% power. When the second MFP tripped on low suction pressure, the operators manually tripped the reactor and entered the emergency operating procedures. The operators stabilized the plant in operational mode four, hot standby.

.. . . ~ . . , . 4 > 3

, The licensee event evaluation team reviewed the events and presented their fm' dings and j recommendations to management. Management identified the corrective actions required to be completed prior to reactor startup and approved the mode change checklist. The operators brought the reactor critical at approximately 5:42 a.m. on January 4 and connected the turbine generator to the electrical grid. Additional details of the event are described in Section 3.2. A detailed :cquence of events is provided at attachment 1. 3.0 OPERATIONS (71707,93702) 3.1 Isolation of Feedwater IIcater E-25A , ! On January 2, operators identified and isolated the leaking E-25A feedwater (FW) heater tube side relief valve, MVD-V105. Operators isolated the tube and shell side of the E-25A FW heater using Procedure ON 1040.022, "Feedwater Heater Isolation and Return to Service." After consulting with operations management and technical support engineers, the shift superintendent decided to leave FW heater E-26A in service and to stay at full power. The shift supervisor did not consider isolating FW heater E-26A, but did consider reducing reactor power. The shift supervisor determined that reducing power with FW heater E-25

isolated would introduce steam plant transients and was not desirable. At approximately 5:36 p.m., a HI H1 water level isolation occurred in the E-26A FW heater which caused extraction steam admission valve EX-VI to shut. The resulting feedwater system perturbation caused the third condensate pump to automatically stan and FW heaters E-21/22C to isolate on high shell side water level. The transient occurred hear shift turnover when a number of auxiliary operators were in the control room. Operators overrode and opened the E-22 A,B,C FW heaters high level dump valves.

The operating crew successfully restored the E-21/22C FW heaters which required opening the inlet and outlet valves, and closing the bypass valve. The AOs performed valve manipulations at a local FW heater control panel (CP-66) located in the turbine building. The AOs then opened the power supply breaker for EX-VI, manually opened EX-VI to admit steam to the E-26A FW heater, and lowered shell side water level. Opening the supply breaker defeats the interlock which shuts EX-VI to prevent water intrusion to the , turbine blades. Defeating the interlock has been an operating practice for the last two years - even though no procedural guidance existed. The shift superintendent stationed a local watch at FW heater E-26-A to immediately close the EX-VI breaker if the shell side water level increased out of sight. The inspectors considered the decision to take compensatory actions when an interlock was overridden to be positive. , I . . .

. . 4 Operators returned the third condensate pump to the normal standby alignment. The FW heater E-26A water level continuously cycled around the HI HI level alarm for approximately four hours, until the operators lowered the level control setpoint on the shell side spill valvv. The operating crew closed and aligned FW heaters E-22 A,B,C high level dump valves for automatic operation, shut the breaker for EX-VI, and secured the compensatory watch. The shell side water level in FW heater E-26A continued to cycle around the HI level alarm setpoint for approximately six hours until the HI HI level alarm was reached. , e The inspectors assessed that the operators performed well by identifying the leaking relief valve and isolating FW heater E-25A. The inspector considered it positive that the operators successfully res=d the E-26A and E-21/22C FW heaters, that the shift supervisor compensated for the defeated EX-VI interlock, and that operations management participated

in the decision to operate in the abnormal FW heater line-up. 3.2 Manual Reactor Trip / Loss of Feedwater Flow

At approximately 3:30 a.m. on January 3, the operators changed the condenser air removal pump configuration which resulted in a minor condenser vacuum change. Shortly , afterwards, steam extraction valve, EX-V1, automatically shut due to HI HI water level in FW heater E-26A. The third condensate pump automatically started. The unit shift ' supervisor (USS) directed the auxiliary operator (AO), by radio, to restored FW heater E- 26A by opening EX-V1. The E-21/22C FW heaters isolated on high shell side water level due to the increased condensate flow from the third condensate pump. The reactor thermal power increased above 100% rated conditions. The USS directed the control room operator to reduce generator load. The USS directed the AO to override and open the high level dump valves on the E-22 A,B,C FW heaters and to restore the E-21/22C FW heaters. The AO opened the high level dump valves, went to CP-66, opened the inlet valve, closed the bypass valve, but left the outlet valve shut. The AO failed to verify proper operation of the automatic water level controller. At this point, the control room. operators incorrectly thought FW heaters 21/22C were restored to service based on the FW heater E-21/22C isolation bistable light on the main control board. The main control board light extinguishes when either the inlet or outlet valve opens and does not indicate that the heaters are in service. The control room operators did 4 not utilize a FW heating system computer graphic, which displayed the FW heating system - valve positions, to verify restoration of the E-21/22C heater string. Instead, the operators selected an extraction steam computer graphic which displayed steam pressures relevant to

restoring the E-26A FW heater. , 1

. . . ,

. ! ' 5 The AO returned to work on opening EX-V1. Minutes later, FW heaters E-21/22A isolated on high shell side water level. The USS directed the AO to restore FW heaters E-21/22A. The AO returned to CP-66, opened the inlet valve, closed the bypass valve, left the outlet valve shut, and retumed to work on opening EX-V1. Control room operators again utilized the extinguishing of the heater bistable isolation light to verify that the AO had properly restored the heaters. The AO and control room operators thought the E-21/22 FW strings were properly restored when in actuality both E-21/22C and A heater strings were isolated. f Minutes later FW heaters E-21/22B isolated on high shell side water level and the USS directed the AO to restore FW heaters E-21/22B. The AO retumed to CP-66, opened the inlet valve, shut the bypass valve, left the outlet valve shut, and returned to EX-V1. The control room operators again incorrectly verified the heater was in service. As a result of improper restoration of the three E21/22 heater strings, no condensate flow existed to the feed pumps. One main feed pump tripped on low suction pressure. Operators attempted to manually lower reactor power when the loss of a main feed pump turbine setback occurred which decreased reactor power to 54%. The second feed pump tripped and the operators manually tripped the reactor due to the loss of feedwater flow to the steam generators. Operators placed the plant in a safe condition following the reactor trip. The inspectors assessed operator response to the loss of feedwater flow as excellent. The event lasted approximately 15 minutes from the time FW heater E-26A isolated to the time the operators manually tripped the reactor. The AO stated that he spent only a brief time at CP-66 to restore the three strings of low pressure heaters. The inspectors later observed two AOs working together to restore a E-21/22 heater string after the string isolated during the plant startup. The AOs, working in an expeditious manner, required approximately 20 minutes to complete the restoration of the single string. The inspectors concluded that rapidly restoring three E-21/22 heater strings did not provide the auxiliary operator sufficient time to verify the results of his actions. North Atlantic concluded that the human factors of the labeling of the valve control switches at CP-66 were poor because the switch nameplates for the inlet and outlet valves both read " heater isolation valve." The inspector noted informal operator aids near a few of the switches indicating inlet or outlet valve controllers. After the reactor trip Nonh Atlantic posted a schematic of the E-21/22 FW heaters on CP-66 as a formal operator aid. The inspectors reviewed the experience level of the AO involved and determined that he had been qualified for approximately two years, had limited experience with feedwater heaters, and had never restored the E-21/22 FW heaters. Through discussions.with the AO, the inspector could not determine whether the AO did not know that the FW heater E-21/22 outlet valves needed to be opened or whether the AO rushed and forgot to open the valves. A total of five AOs were on shift, three of which were on rounds in other areas of the plant.

1 . - .

. 6 The inspectors determined that informal verbal communications between the USS in the control room and the AO in the turbine building contributed to this event. The USS did not utilize specific valve-by-valve directions to the AO but rather relied on the AO to know how to restore the heaters without a procedure. Due to a knowledge deficiency regarding a control room indicating light, the control room operators failed to effectively verify the AO actions for restoring the heaters. The inspectors assessed that the control room operators- appeared to be focused on decreasing reactor power and opening EX-V1. The inspectors reviewed the AO training material for the condensate and feedwater systems. The material contained a good level of detail, but the AOs would require verbal or procedural direction when performing non-routine evolutions. The licensed operator training material did not include a full discussion of the feedwater control logic and status lights. The inspectors determined that the errors made by the non-licensed AO and the licensed operators' performance indicated ineffective training in the area of feedwater heater controls and communications. 3.3 Conclusions The inspectors attributed the cause of the reactor trip to repetitive operator errors made when trying to restore feedwater heaters. Several secondary causes contributed to the operator errors such as weak operator training in the feedwater heating system details and indications; poor human factors at CP-66; improper verbal communications; and the lack of written guidance for abnormal operation of the feedwater heaters. The inspectors concluded that proper action by the operators could have restored the feedwater heaters and prevented the thermal transient and the reactor trip. The inspectors determined that the decision by the control room operators to trip the reactor was conservative and demonstrated an appropriate safety attitude. 4.0 TECHNICAL SUPPORT / ENGINEERING (71707) 4.1 Feedwater Heating System Operating Experience North Atlantic has experienced difficulty in sustaining smooth and accurate control of the feedwater (FW) heater shell side water levels. No previous reactor trips were attributed to FW heater problems. However, there have been previous unexpected heater isolations which resulted in secondary plant transients. Based on review of operational data and discussions with operators and technical support personnel, the inspectors noted that the operational difficulties with the FW heaters stem primarily from the sluggish performance of the pneumatically controlled shell side water level control system. North Atlantic recognized the need to enhance the performance of the FW heaters and implemented several major modifications. During the first refueling outage the licensee installed new digital level controls on the heater drain tank and redesigned the heater drain

- _____-__ _ _ i

.

.- . 1. ~. ' 7 tank to provide a vent path after a reactor trip. During the second refueling outage, as documented in NRC Inspection Report 50-443/92-27, North Atlantic installed new digital controls on the E-21, E-22, and E-24 FW heaters, and the moisture separator reheaters, and installed new level columns on the heater drain tank. North Atlantic has scheduled the installation of digital controls on the E-26 FW heaters. The licensee had prioritized these modifications on a risk-based approach which the inspector considered to be proactive. In summary, North Atlantic has made significant modifications to improve the performance of the shell side water level controls in the FW heaters, and further modifications are scheduled to be implemented. I l l 4.2 Abnormal Opemting Configuration l The inspectors held discussions with technical support engineers and operations personnel. The licensee staff stated that the secondary system transients wer; a well known phenomenon. During the event, operators isolated the tube and shell side of the FW heater E-25A to establish mechanical isolation for the leaking relie' valve, MVD-V105. The heater drains from FW heater E-26A, which normally cascades through the shell side of FW heater E-25A to the heater drain tank drained directly to the heater drain tank through the spill line. Operators adjusted the setpoint of the FW heater E-26A spill valve to lower and control the shell side water level below the HI H1 water level isolation setpoint. The operators performed the routine operational adjustments with the assistance of instrument and control technicians. No procedural guidance existed for establishing or operating in the resulting i ! abnormal hneup. 1 The isolation of FW heater E-26A on HI H1 water level led to the isolation of the FW heater E-21/22C. When EX-V1 shut, the resultant heater drain tank perturbation caused the automatic start of the third condensate pump. The resulting increase in condensate flow through the tubes of the FW heaters E-21 A,B,C caused increased condensation and lowered shell side pressure in the feedwater heaters. The lower shell side pressure reduced the flow of water to the condenser resulting in the increase in shelj side level. To mitigate the

transient, operators manually overrode and opened the FW heater 22 A,B,C high level dump valves. After evaluating the transient, the licensee event evaluation team determined that a plant trip should not result from the isolation of two E-21/22 feedwater strings. North Atlantic instructors had previously taught operators that an isolatic.; of two E-21/22 feedwater strings would result in a reactor trip. As a result, operators had stressed rapid restoration of any isolated E-21/22 heater string. The inspectors concluded that operators and technical support engineers fully understood the secondary plant transients caused by isolating the E-25 and E-26 heaters. However, the plant staff did not know that the secondary system could operate with two E21/22 feedwater strings isolated and, therefore, placed urgency on restoring a single heater string. __________________ _ _ _ -

. . ~ . 8 . 4.3 Bypassing the EX-VI Interlock The extraction steam supply isolation valve (EX-V1) to the E-26A FW heaters automatically , closes on a HI HI water level on the shell side of the heater. The automatic closure feature of the extraction steam valve protects the turbine blades from water impingement and is described in the Updated Final Safety Analysis Report. To restore E-26.FW heaters, the , shell side water level must be lowered below the Hi HI level setpoint in order to open and l reset the extraction steam valves. Lowering the water level is accomplished by opening the - power supply to the motor operators on the extraction steam valves and manually opening the valves. The pressure of the extraction steam forces water out of the heater. The team noted that North Atlantic had no procedure or administrative controls for overriding the extraction steam valve automatic closure signal during secondary plant aansients. North Atlantic committed to incorporate appropriate controls for overriding the interlock in a abnormal , procedure. ' 4.4 Relief Valve Failure Mode Maintenance personnel replaced the leaking FW heater E-25A relief valve, MVD-V105, which was manufactured by Dresser Valve Company. The bellows of the relief valve provided a pressure boundary and when the bellows ruptured steam issued from a telltale hole in the valve bonnet. North Atlantic is consulting with the valve vendor to establish a ' failure analysis. Subsequent to the reactor trip, North Atlantic replaced the bellows in the tube side relief valves on the E-25 and E-26 feedwater heaters with a newer model of the bellows that has enhanced material. . 5.0 SAFETY ASSESSMENT / QUALITY VERIFICATION (71707) The inspectors arrived on site approximately two hours after the reactor trip. The licensee event evaluation team was present and collecting the data needed to review the event. The operators were preparing written statements concerning the event and operations personnel , were discussing the plant response with the operators. The licensee event evaluation team completed an initial assessment of the event and presented their findings and recommendations to site management, including the senior vice president and chief nuclear officer at 4:00 p.m. on January 3. The root cause of the event was not explicitly documented; however, the sequence of events and plant response were fully discussed and understood. Recommendations included isolation of both the E-25A and E- 26A feedwater heaters, until the feedwater relief valve on E-25A was repaired, and training ' crews on restoring feedwater heaters. Management approved reactor startup contingent on the completion of the recommended short-term corrective actions.

North Atlantic initiated a station information report (SIR). The SIR included the licensee event evaluation team report, a root cause analysis which used change analysis and Kepner- Tregoe analysis techniques, and a human performance enhancement system (HPES) review. !

_ _ _ _ - _ . - .~ . . - - , 9 The licensee event evaluation team identified long-term corrective actions including design, procedural, and training enhancements. Two specific recommendations were to reinforce and evaluate communication practices, and to develop an abnormal procedure to respond to secondary plant transients. The SIR determined the root cause to be incomplete communications. The main control room and secondary auxiliary operator did not communicate the level of detail required and caused an open communicatior. *. sop. A secondary root cause was the lack of an abnormal procedure. Contributing factor. included poor communications equipment, human factors deficiencies at the turbine building control panel, CP-66, and system design problems. Report recommendations included continued training with emphasis on good communication practices, and development of an abnormal procedure to respond to feedwater induced transients. The root cause analysis report noted that stress placed on the operators by the belief that the plant would trip if two E-21/22 heater strings isolated complicated the response to the transient. The SIR documented 14 recommended corrective actions, and assigned responsibility and due dates for the completion of the corrective actions. The recommendations included operations, training, and engineering actions. The inspectors reviewed the SIR and held discussions with licensing and operations personnel. The inspectors determined that the SIR accurately documented the event and reached root cause determinations based on thorough and detailed evaluations. North Atlantic committed to establishing administrative controls for overriding protective feedwater heater interlocks and to establishing priorities for restoring isolated feedwater heaters as part of an abnormal procedure for secondary plant transients. The inspectors concluded that the corrective actions identified in the SIR addressed the identified root causes. 6.0 SUMMARY Repetitive o, 'rator errors resulted in a loss of normal feedwater flow requiring a manual reactor trip. The operators responded well to the loss of feedwater and stabilized plant conditions. The Nonh Atlantic event evaluation team reviewed the event and management - approved reactor startup pending completion of identified short term corrective actions. , l The inspectors concluded the cause of the event was operator error. Secondary causes included operator training on feedwater heater system details and indications; poor human , I factors at CP-66; improper verbal communications; and the lack of written guidance for abnormal operation of the feedwater heaters. Proper action by the operators could have restored the heaters and prevented the thermal transient and the reactor trip. The decision of the control room operators to manually trip the reactor was conservative and demonstrated an ! appropriate safety attitude. The inspectors concluded that the North Atlantic assessment of the event was thorough and well documented. The inspectors determined that the recommended corrective actions in the SIR addressed the identified root causes. 1 < _ _ . _ _ - _ . _ - _ _ _ _ - _ _ _ _ _ - - _ - - - - _

. . ~

. 10 4 7.0 MEETINGS (30702) The scope and findings of the inspection were dis.;ussed periodically throughout the inspection. An oral summary of the inspection findings was provided to the station manager and his staff at the conclusion of the inspection period on January 21,1993. .

,. .. - . - ATTACHMENT I SEQUENCE OF EVENTS January 2,1993 5:15 p.m. Auxiliary operator (AO) reports a steam leak from the relief valve on the tube side of feedwater (FW) heater E-25A. 5:35 p.m. AOs isolate FW heater E-25A. The AOs direct the heater drains off FW heater E-26A directly to the heater drain tank through the high level spill valve. 5:36 p.m. A HI HI water level in FW heater E-26A shuts the steam extraction valve, EX-VI, to the heater. The third condensate pump automatically starts. 5:37 p.m. A HI HI water level in FW heater E-21C isolates the E-21/22C heater string by shutting the inlet and outlet isolation valves, and opening the bypass valve, CO-V-75. , 5:45 p.m. AOs open the high level dump valves on the three E-22 feedwater heaters. AOs restore the E-21/22C heater string by opening the inlet and outlet valves, ' and closing CO-V-75. 6:32 p.m. Operators secured condensate pump CO-P-30B. The E-26A heat exchanger water level oscillations cause continuous HI HI level alarms. 9:17 p.m. AOs adjust the setpoint for FW heater E-26A level control to prevent HI HI level alarms. The AOs open the breaker to EX-VI, manually open the valve, and shut the breaker. January 3,1993 > 0:25 a.m. The AOs shut the high level dump valves on the three E-22 feedwater heaters and returned the valves to automatic. 3:30 a.m. The AOs change the air removal pump line up which causes a minor fluctuation in condenser vacuum. 3:36 a.m. A HI HI water level in FW heater E-26A isolates steam extraction valve, EX- ' VI. The 'A' heater drain pump discharge valve closes. 3:37 a.m. The secondary AO begins to open EX-V1. . P 4

. . , - . . 3:39 a.m. Low main feed pump suction pressure starts the third condensate pump, CO-P- 30B. A HI H1 water level in FW heater E-21C isolates the E-21/22C heater , ' string by shutting the inlet and outlet isolation valves, and opening the bypass valve, CO-V-75. 3:40 a.m. Reactor thermal power exceeds license 100% rated conditions. Operators

reduce power which adds additional perturbations to the feedwater heaters. ' ! 3:41 a.m. The secondary AO opens the high level dump valves on the three E-22 feedwater heaters. The AO attempts to restore the E-21/22C heater string by opening the inlet valve and closing CO-V-75, but fails to open th,e outlet valve. The AO returns to EX-V1. 3:44 a.m A HI HI water level in FW heater E-21 A isolates the E-21/22A heater string by shutting the inlet and outlet isolation valves, and opening the bypass valveCO-V-75. 3:44 a.m. Thermal power decreased below 100% rated conditions. 3:45 a.m. The . secondary AO attempts to restore the E-21/22A heater string by opening the inlet valve and closing CO-V-75, but fails to open the outlet valve. The AO returns to EX-V1. 3:48 a.m A HI H1 water level in FW heater E-21B isolates the E-21/22B heater string by shutting the inlet and outlet isolation valves, and opening the bypass valve, CO-V-75. , 3:49 a.m. The secondary AO attempts to restore the E-21/22B heater string by opening the inlet valve and closing CO-V-75, but fails to open the outlet valve, eliminating all condensate flow. The AO retums to EX-V1. 3:49 a.m. The 'A' main feedwater pump trips and initiates a turbine setback. 3:50 a.m. The 'B' main feedwater pump trips. The operators manually trip the reactor. 4:00 p.m. North Atlantic conducts a post trip review meeting and authorizes plant startup pending completion of required work items. I i 2 ' , i

_ _ _ _ - _ _ _ _ - _ _ _ _ _ - _ _ _ , .. ,. .

. January 4,1993 ' 4:20 a.m. Operators take the reactor critical. 10:40 a.m. Operators place the turbine generator on the electrical grid. 11:05 a.m. A HI H1 water level in FW heater E-21C isolates the E-21/22C heater string by shutting the inlet and outlet isolation valves, and opening the bypass valve, CO-V-75. 11:35 a.m. Two AOs correctly restore the E-21/22C heater string to the normal lineup. 1 !

4 3 ___ ___ _ ____ - __- -_- - ___ _ __ _ - _ -

ATTACHMENT II ~* . . , 6 9 + 1e iS~ - ' E h, , n n . \\ -- r . - - - v su,,., s e 8, i e t,js , " ' , , %,/\\ !* 11 [.

'Vk - .I; j; $ ' ' " A ,o n1 . . s i . - i . .}*, , 18

L

5 , ., [-l (

~ . - \\

e ,

-

ws / r8 , - - ',

, i. j p- I r- .; - gii 11 cE n n , - , ,0 - , , +

- vs 7 e s,fs !!

• !

s.js a r e [ , V t:

-

-

I . s- _n } c r: r n e , , g g ,. g o e I s I !! _ _ ft ^c [I li II E#: 15

! !?

m r, i

-

ou dh k - I , , @-

i , 8 *o 1

- - -

sW g 'm

• ';1..

j p.. ..

i i l-l 0; $.5 & Y - n B i _e n n .. . g + ,s w

, njs e e{ e{ k j e s

to; r; . ) ., _,n . e ' u . n .g s . , o

/,,i

!s!U( - it - o a

':

1 i:. li i ,e !!i .1: - 58 , , , . 2 , . ' ' ' * * j p_ l .. .. L, J l I , " .. .. j .i 4 -- -- . m ., i; m . ce i il li io ^

• ?

r

1

i ::t _ OK \\,4 8 e&, !!

E !! u... . gg:

E

!

.

-

-

-

L si : l r2 !;s i . e y '. + g. j n..: } LJ 2* IlT / \\ t i I i 3~ {{I i 03

g

. a3 5! - e v < to1 .:3 2 1 ,

.. _ ._ .,. _ .. . ._ . , .i , . ! . . , '

-; .. i ! Figure 3.2- t Condensate System .!

'? , ! 4 I! , , , a k~ Q$oj ]$oj~ ].. ha' ha'- - h- j . . .. . _. , _. . ... . .,3 . .... . 3 . . .. . _3 =. 3 m.- ) ._. 3 I It='r : I H='m XI I I

!

I luem - .... . .th _ k.... . - .") _ _ . i ' o .. -. c) m'- ) ='- )

',

;'- )

t eie. - [ Q i . .. . . . . . ,,, . : _ o, . < _ _ o ~. . _ . a LJ L .l~ LJ LJ LJ LJ Lt - ! . . . .

== ,. . . "' OK OK OK . , , . . . ... ., O. X X QXX QX X a - a . .. , l l -i , , i, =UT - -- ;y,;" J l =") - . , .

. . . . ,'- t

- ,

r

i X ". X ". o L ............... 1.. ...... ...L.......... i . ., i 3 j ,. ............. _........ i o ~ >J b

i

. . .

. p' t... .._f ... i a . h. O

' CDs.m.c C 1 i l t 0L w.&F.

.

a. . . . ..e x , . . _ . .., ,

• * * '

- Mt / ' v H L* J

r. . . , m

.! e - - .. .. , } - = I h .7.?' H 'a'rt;....) 1 Li ! i . i ! 'o,.,o.,moorj

I a 'a= =a. ._ ! ' I. N 'bC>"' ' H 2.~' ot ) ~~ o -M.

• -

< . ..c,.~. - -1 a ~... .. .) . . . .,_....,_..O.-- t. . , . ~ ., 6 =

co e. cam.~-

_ .. m ,=

- .. , I Wr, . c . .I O, o.,o .o .u. m.e

.., u,.,.,...,,.e. ,

eu .*s . co.o, , . t 9... , i ....... ...... _.. , , =>-o u. ) . ( .. <. i , .n ..i .

, . , __.2____., ., co ...._. .c e , .. ta. .y

. ;:::; ,,,

o-. j y ...rus y o

.. g..,.

,r - @ p . . . m- l'c = . _ . _

. t d J l est -T- 9 , , '

-l @ l s' ] <: - A _ t-r ,. _ . ~ , . _ , ._ t . co. 4.ot m.9 $

401 b t ' .!. . > C O . '- l . 1/92 i . .i ! . I . . . . . ... , , - . - - . . - . . .- . - . ,, , , . - .n.

1 ! , 4.f - m

S 9 4 , O O o ' W l 9 e i t W - -; -9 . ., - -. s - e , o it ., ' g -, -, -- y - -- - ,. _, . e-6 76 x A '- a,

• ir

> -

-:* e

e -: - 3 I Y I Y E f E l ., . i i i i i i . r1 1 gg. 3- . , Ea

b% 4 E c .;f r o- I X .E .E -C=o-#*->< @- o l w C

I w O I I $ 9 1 m i i M & ,, [ o ci ;g } i -; --s -9 . - ,4

• [

-;;;g - c .,

• :E

E I 9 ~ o ' l o: a2 -, .- -, -, - , n. .a - ir .E

7-6

56 e n L. v I i -A % 7 .!.! ., e o i = a v e r e r e r M I I I I i 1 1 I 4O t 1 i -l e I"Ir 1 o

• ':

i po-#w-c c

9 .. e n .! I , T o 1 .' l M:S >c -> --a -P - . - o 1 -, -se- 1 I 9 E 1 [ i - g- -q --; - , t. _s - - -

b t v

• h

.c.o.4- J L ~E XO---@ UV!A _] g"E I f 3RD ST AGE I EnfRACTPON STEAM vg M I.I ,, v W l SYS TEM / HEATER P - E5 .g h CLOSES OM RESPECTevE y',f MA HeGH-HeOH HEATER I I REMEATER LEVEL OR TUResNE TRep I. - m. - - - L DAAsN - TANM . t , W A*O h CLOSES ON HsonageGH HEATER LEVEL ( ) Xo-@ = Gy!A. l Y = = g,. . ~- . .2 . . . . . . . ..u .- . . . - . . . . . . . .

Figure 4.1 Isolation and Nonreturn Check Valve Control INSTRUMENT AIR FROM TURBlNE TRIP AIR RELAY DUMP VALVE , FEEDWATER HEATER - (10 INCHES > NWL)@ LEVEL HIGH-HIGH ENERGlZED - 3 CS: TEST %> MANUAL TEST VALVE AUTO CLOSE- '- 'A >.

• TUROINE TRIP, OR

k

• FEEDWATER HEATER LEVEL ~~~~~~~l

t ' HIGH-HIGH I y g- ,, N RAIN OV(S) EXTRACTION \\ F~----> WHEN NOT FULLY OPEN i STEAM f v y ISOLATION FC , , VALVE _ ' 11 NON-RETURN Cif ECK VALVE FEEDWATER HEATER _ e ,e

k HD - 36 l . 11/91 < ._a . m-- - - -' --- - - ' * ~ ' ' " - ~ '

4 % . e e J ' s i - E i _= e 5 > P ,. .................... t r .w.gs. . x v.pavmaa ww.= v e:.:/ E v 1 .- 6 a . , e - 8 w 2 2 g

s. 4 s. !- 1 -

I 1 $ E w$ M E s E e wII 4 I w2 2 h E "s Q * C '? > f g

j#* E h ii

j 5" cE ,! _ ls' cri E,is"! s) e: j'l

W i - e i N-3 .: EEz.

• E-

$ ' J ,s y . .E - 2 7 ...... e l $z $ 3 T 6 $$ . f li~ ~ ' ' ' t _ O <: - - K < , U s f.! h d$ , c

y o t o.f; , N. U $ $1 3 , , m>

, - - - wo

I e Y

~~ e i a s , N o C [

i >? t O" E* l l 9~ - .9 e 4, ,sm .mn . .. - m= 6~ . - .,. I .

, W

e i y E AC m._.....; E ~' \\'/ l A g V

e a v . C -6- <g .

• t

e . _fs n

L&. La j ... & O 'k EI t O{ oat, t X X gr . _o , , . 8 f 4 e , U<a

• E

• E

_ _ _ _ _ _ i 6 . I i 5 i }}