ML20235B441

Ro:On 870811,spill of Reactor Bldg Closed Cooling Water Occurred.Caused by Failure to Follow Instructions on Switching & Tagging Request Form Re Valve Backseating. Procedures Will Be Revised & Training Provided

ML20235B441
Person / Time
Site:	Oyster Creek
Issue date:	09/22/1987
From:	Fiedler P GENERAL PUBLIC UTILITIES CORP.
To:	NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM)
References
0645Q, 645Q, NUDOCS 8709240105
Download: ML20235B441 (7)

Text

n. - _

I

?

• )

GPU Nuclear Corporation -l Nuclear  ::=:;388 Forked River, New Jersey 08731-0388 609 971-4000 Writer's Direct Dial Number:

Document Control Desk september 22, 1987 U.S. Nuclear Regulatory Commission Washington, DC 20555 l

Subject:

Oyster Creek Nuclear Generating Station Docket No. 50-219 Spill of Reactor Building Closed Cooling Water during Maintenance On September 11, 1987, a safety limit violation occurred at the Oyster Creek Nuclear Generating Station. An analysis of that event and GPUN's corrective action was contained in our submittal dated September 20, 1987. That letter stated that a detailed review of the maintenance activities leading to that violation would be submitted via a separate submittal. Attachment I to this letter provides the results of that review.

If any further information is required, please contact Mr. John Rogers of my staff at (609) 971 -4893.

VMtrul _

p y P te B. Fiedler V President and Director er Creek PBF/JR/dmd Attachment (0645Q) cc: Mr. William T. Russell, Administrator Mr. Lee H. Bettenhausen Region I Chief, Projects Branch No. I U.S. Nuclear Regulatory Commission US NRC 631 Park Avenue Region I King of Prussia, PA 19406 631 Park Avenue King of Prussia, PA 19406 Mr. Alexander W. Dromerick, Project Manager U.S. Nuclear Regulatory Commission Dr. Thomas E. Murley, Director Division of Reactor Projects I/II Division of Nuclear Reactor Reg.

7920 Norfolk Avenue, Phillips Bldg. US NRC Betheda, MD 20014 Washington, DC 20555 NRC Resident Inspector Oyster Creek Nuclear Generating Station 8709240105 9709g2 PDR ADOCK 05000219' S PDR GPU Nuclear Corporation is a subsidiary of the General Pubhc Utihties Corporation gh

ATTACHMENT I Title of Event:

Spill of Reactor Building Closed Cooling water containing very low levels of contamination while repacking the stem of an isolation valve in the Reactor Building Closed Cooling Water System.

Brief Description of the Event At about 0208 on September 11, 1987, a leak occurred while a mechanic from the Plant's Maintenance Department was.in the process of removing the packing from Reactor Building Closed Cooling Water (RBCCW) drywell isolation valve V-5-167. The mechanic directed the flow of water into a catch bag but the bag started to overflow and he called for help. A radiological control technician, who was stationed nearby, heard.the call for help and noticed that water was spraying down from the overhead. After checking the status of the person who was calling for help and confirming that he (the mechanic) was uninjured, the radiological control technician immediately notified the control room of the spill and requested their assistance. The mechanic who had been sprayed with water had come down from the scaffold platform and was waiting for further instructions. The Group Operating Supervisor (G0S) was immediately dispatched to the Reactor Building 23' elevation to investigate the source of the leak.

The G0S reported to the Control Room that it appeared a rupture of RBCCW piping had occurred and recommended that they isolate cooling water flow to the drywell. The RBCCW system was providing cooling water to the Drywell Equipment Drain Tank, the operating drywell recirculating fans, and the operating reactor recirculation pumps (B&C). The control room operator secured the operating reactor recirculation pumps in preparation for securing RBCCW flow to the drywell. (For discussion on the related safety limit violation, see GPUN letter, P.R. Clark to Dr. Murley dated September 20, 1987)

A few minutes later, the GOS was informed by the control room that the reactor recirculation pumps had been secured and that RBCCW flow to the drywell could be isolated. The G0S directed a nearby equipment operator to go to the reactor building 51' elevation and shut the manually operated RBCCW drywell isolation valve V-5-709. The GOS instructed the control room operator to close the motor operated drywell isolation valve (V-5-166) which is located inside the drywell. These valves were shut. The leak rate was reduced but not completely stopped. The GOS determined that the leak was from the packing on V-5-167.

About this time, the Maintenance, Construction, and Facilities (MCF) Area Supervisor arrived at the spill scene. He knew that an RBCCW valve was being repacked on the 23' elevation and thought that the spill was a result of this work. He discussed the position of the valve with the GOS and was informed that the valve had been electrically backseated. The MCF Area Supervisor thought that the leakage was too excessive for the valve to be on its backseat. The MCF area supervisor thought it would take too much time to get an electrician to verify the valve was on its backseat. He requested permission to try to manually backseat the valve. An MCF management representative, the MCF Area Supervisor, and the G0S discussed the situation

, and developed a plan by which the MCF Area Supervisor could attempt to manually backseat the valve in an attempt to isolate the leak. The MCF area

supervisor then checked with the radiological control personnel who were in the area and was directed to don appropriate waterproof protective clothing prior to entering the area. The MCF Area Supervisor suited up and entered the area to backseat the valve. He opened the valve one additional turn of the manual handwheel onto its backseat at which point the leak stopped.

The area was immediately secured and appropriate spill response actions were taken. Both the mechanic and the MCF Area Supervisor had been sprayed with water. The MCF Area Supervisor was determined to be uncontaminated and was released from the area. Radiological control personnel determined that the mechanic's clothing was contaminated and had him undress. The mechanic was then released from the area and subsequently given a whole body count as a precautionary measure. The RBCCW system uses a corrosion inhibitor. The mechanic required some minor first aid later in the morning as the solution had caused some minor irritation to his right eye. The spill recovery actions were closely supervised by the plant's safety department.

Plant Conditions Prior to the Event The reactor was shutdown with the Mode Selector Switch locked in the Shutdown position. Reactor water temperature was 140*F and the reactor was vented.

Reactor water level was approximately 156 inches above the Top of the Active Fuel (TAF). Condensate pump "C" was operating, providing water to the reactor as necessary, and Shutdown Cooling Pumps A and B were in service. Reactor recirculation pumps B and C were in operation.

The RBCCW system was operating with A and B pumps providing 82 psig of pressure. RBCCW water temperature was 89'F. A sample of RBCCW system water taken 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> after the event contained 5.4 x 10-4 uci/ml of activity, pH was 10.19, i

l l

Actions Leading to the Event On 8-6-87, Plant Engineering initiated Maintenance Short Form No. 46069 to repack V-5-167 with live loaded packing in accordance with Procedure 700.1.030. The valve was required to be repacked to correct the root cause of a previous surveillance test failure.

A work package was developed by the MCF planning section. The maintenance supervisor assigned to the job reviewed the work package. He submitted a Switching and Tagging Request on 9-10-87 which requested that the valve be isolated and the system depressurized. This condition was specified on the tagging request as it was a prerequisite in the referenced repacking procedure.

At approximately 10:00 AM the Group Shift Supervisor (GSS) advised the maintenance supervisor that the valve could not be isolated and that the tagging request would have to be revised to repack the valve on its backseat.

The maintenance supervisor did not believe that the repacking procedure permitted repacking the valve on its backseat and advised his management. MCF management contacted Plant Engineering to determine if the valve had a backseat (it did) and question the intent of the procedure. Plant Engineering stated that the procedure was not written to repack valves on their backseat because of the uncertainty involved with isolating the stuffing box using the backseat, but it was " technically possible" to repack the valve on its '

backseat. This was construed by MCF management to mean that it was permissible to repack this valve on its backseat.

A subsequent discussion held between Plant Engineering and the MCF Job Planner identified that a procedure revision to include additional precautions and instructions should be issued to the repacking procedure to specifically permit repacking a valve on its backseat. This procedure revision was not initiated.

The switching and tagging request was revised and resubmitted on the same day (9/10/87) at approximately 1:30 p.m. specifying that the valve was to be placed on its backseat and tagged out. The GSS attempted to place the valve on its backseat using the normal control switch by giving it an additional opening signal. (This did not result in any valve motion.) The GSS incorrectly thought that a second opening signal would bypass the limit switch and the valve would "go out on its torque switch" against the backseat. The maintenance supervisor was in the control room during this evolution and was advised by the GSS that the valve was just on or very close to its backseat.

The maintenance supervisor was cautioned by the GSS to look for leakage and if excessive leakage were noted, an operator would further backseat the valve manually. The Maintenance Supervisor questioned why there wasn't an isolation boundary tag on the hand wheel. The GSS told him that this way the valve could be manually backseated if required. The GSS was very concerned about backseating the valve as previous valve failures were attributable to improper backseating techniques. Further, he preferred electrically backseating valves to manually backseating valves, because he thought that a valve backseated

backseated electrically would be protected by a torque switch. The preference to electrically backseat, vice manually backseat, was based on information presented in the operator tr aining program. No approved station procedure or method was used to backseat this valve.

Maintenance personnel were prepared to repack the valve on the 4 - 12 shift but could not because the new packing follower required machining. The repacking was subsequently started on the 12 midnight to 8:00 a.m. shift on 9/11/87. The turnover between operating shifts and maintenance shif ts stated that the valve was electrically backseated. The mechanic performing the repacking was concerned that the valve might not be backseated and reviewed the copy of the switching and tagging request. He concluded that the valve was backseated based on the switching and tagging request. The mechanic proceeded to remove the packing from the valve one ring at a time. No leakage was observed until the fif th of 8 rings was in the process of being removed, when the remaining packing blew out, causing an RBCCW 1eak, of approximately 20 to 25 gallons per minute.

Consequences of the Event

1. Areas Contaminated:

The east side section of the reactor building on the 23' elevation had been sprayed with water. It was subsequently secured and posted. During the spill, the water had been substantially contained by diking the area with clean, unused protective clothing. After the leak was stopped, the water was pumped to a nearby floor drain.

Surveys of the spill residual showed contamination which was not evident on readings of the RBCCW water. Apparently, the water flow had flushed over pipes and conduits on its way to the floor. The area has been cleaned up and decontaminated, l

2. Personnel Contaminated: j The mechanic was not contaminated, but his clothing was contaminated. He frisked out clean (less than 100 cpm) when he was  !

unclothed and was released by the Radiological Control Department. J The mechanic was given a whole body count as he had been drenched and possibly ingested some liquid. His whole body count, while unclothed, read only normal levels of potassium 40.

The MCF area supervisor was not contaminated. As a result of wearing the waterproof suit the MCF area supervisor frisked out clean (less than 100 cpm). The MCF area supervisor was subsequently given a precautionary whole body count on 9/15/87. His whole body count read only normal levels of potassium 40.

l

3. Personnel Injured:

The mechanic came into physical contact with the RBCCW water and was instructed by the Safety and Health Engineer to shower for 15 minutes. He stated that some of this RBCCW water got into his eye.

He was instructed to flush his eye while showering. As a precautionary measure later in the morning, the mechanic was directed to report to the company doctor for an examination and any necessary treatment. His right eye had drops instilled prophylactically and he was released by the medical department.

The MCF area supervisor did not require any medical attention.

Cause of the Event Root Cause The root cause of this event was failing to properly execute the specified instructions on the switching and tagging request form by the operations department personnel.

Contributing Causes

1. Station administrative procedures are not clear in assigning responsibilities for Operations and Maintenance. The specific maintenance procedures used for this event were insufficient to perform the maintenance evolution as planned.

The operations and maintenance personnel involved either were not aware of the procedural limitations or did not address the inadequacies (e.g., issue a temporary change to the procedure) prior to implementation of the maintenance effort.

2. The operators did not comply with specific details of procedures controlling plant equipment, specifically as they relate to switching and tagging. Plant maintenance did not comply with procedural requirements to verify the safety and adequacy of the pressure boundaries established for this maintenance activity. Additionally, communication was noted to be weak during shift or personnel turnovers.

1

3. Training on Valve Motor Operators was misleading and incomplete in l some respects. Although the information presented was generally accurate, the emphasis placed on certain valve operator characteristics and the omission of the motor operator control l circuit opening logic contributed to the incorrect decisions reached by the GSS.

I 1

l l

Corrective Action The following corrective actions will be taken prior to restart:

1. Procedures for the operations / maintenance interface will be revised to more clearly assign responsibilities. Operators and maintenance personnel will have these responsibilities emphasized to them as part of a training session.

2. Operations and maintenance management will stress the importance of procedural compliance to their personnel. Training will be provided on switching and tagging requirements and approved valve backseating techniques to appropriate personnel. Additionally, maintenance management will issue a policy providing guidance to maintenance supervisors on proper job turnovers during shift changes.

3. Detailed information relating to control logic of the valve motor operator and a copy of the critique of this event will be placed in the required reading programs for appropriate personnel.

The following corrective actions will be completed prior to December 31, 1987:

1. Specific procedures relating to backseating and unbackseating of valves will be combined into a single procedure under the control of the Operations department. The maintenance procedure which was applicable to repacking the valve during this event, will be revised to identify prerequisites, precautions, and limitations which will allow safely repacking this valve on its backseat.

2. Maintenance department will issue a procedure to formalize its policy on proper job turnover during maintenance.

3. Formal training will be provided to appropriate personnel on specific details of valve motor opertor control logic.

4. A Management Oversight and Risk Tree (MORT) analysis will be performed on this event to ensure that no additional attributing causes were omitted.

1

__