ML18092B056

Maint & Surveillance Program Plan, Site Survey Rept

ML18092B056
Person / Time
Site:	Salem
Issue date:	10/31/1985
From:	Kootz J, Scott W, Scott W Battelle Memorial Institute, PACIFIC NORTHWEST NATION, NRC OFFICE OF INSPECTION & ENFORCEMENT (IE), Office of Nuclear Reactor Regulation
To:	NRC
Shared Package
ML18092B055	List: ML18092B054 ML18092B056
References
CON-FIN-B-2984 NUDOCS 8603250060
Download: ML18092B056 (41)

Text

.

I

• .h I

I e

TECHNICAL LETTER REPORT SALEM NUCLEAR POWER PLANT SITE SURVEY MAINTENANCE PROGRAM AND PRACTICES FIN 82984 February 13, 1986 8603250060 860318 PDR ADOCK 05000272 p

PDR Enclosure

.-.~.

EXECUTIVE.

SUMMARY

. Approach As part of the U.S. Nuclear Regulatory Comfssion's (NRC's") Survey and Evaluatfon*of Maintenance Effectiveness Project, a site survey of the Preventive Maintenance (PM) pr.ogram and practices was.conducted at the Salem nuclear power plant *. The purpose of the visit was* to collect descriptive data on the*

effects and costs of the changes in t.he Sa 1 em PM program mandated by NRC as a result of the anticipated transien*t*without scram (ATWS) events at Salem in 1983. The site *survey was conducted th-e week of October 28, 1985 by a team of two NRC and three Pacific Northwest Laboratory (PNL) staff.

The survey team collected infonnation on the Salem PM Program, its changes and the costs and benefits of those changes using a fonnalized PM review protocol. The protocol aided the site visit team in collecting infonnation in seven major categories:

PM Organization and Administration

• Facil1ties and Equipment PM Technical Procedures PM Personnel PM Work Control PM Cost-Benefit PM Program Performance.

The protocol contains the detailed information collected during the site visit; this report presents selected observations and summaries extracted from that protoco 1.

Summary of Sjgnjfjcant fjndjngs The Salem site visit confinned that PSE&G has focused attention on and committed substantial resources to maintenance at the Salem plant.

In this effort, PSE&G has changed the PM program significantly since the February_ 1983 ATWS events. Management believes that all requirements mandated by NRC have been met; however, they are still examining the PM program, identifying areas for improvement, and making major changes.

These changes include reorganizations, reevaluations of PM tasks, incorporation of ALARA considerations and specific component reliability data into PM task planning, and acquisition of an automated infonnation management system.

Because of this state of flux in the Salem PM program, ft is difficult to quantitatively describe the costs and effects of the PM program modifications.

The results of the site visit contained f n this report characterize the current state of PM at Salem and indicate the plans and expectations that PSE&G has for additional changes to the program.

i

In spite of the inadequacies of the IQ system, a review of *violations, operating events and Licensee Event Reports gave no indication of adverse safety effects duri*ng the restructur1 ng of the PM-program.

Changes in the PM Program.

PSE&G has already implemented many changes in the PM program and intends to implement more in the near future. Irrmediately following the ATWS events, PSE&G contracted with Westinghouse to conduct a thorough review of all manufacturers' and vendors* recorrmendations regarding PM on plant equipment. This effort resulted in the creation of a large number of new PM ta~ks that were entered into the IO system. Since this effort was completed only a short t1me before the site visit, the effects of many of those changes had not been realized.

At the time of the site visit, PSE&G management was in the midst of carefully

• evaluating the changes in the PM program and what some of the effects of those changes were. Their results indicated that:

• manufacturer's PM recommendations had been overemphasized

• more participation of station personnel in formulation of PM requirements and tasks was required

• the spare parts inventory should have been adjusted along with the identification of new PM tasks

• computerized information management support is required.

To address these areas for improvement, PSE&G was in the process of implementing an extensive reorganization of plant staff that had, as one of its objectives, the improvement of the maintenance department's capability to conduct efficient maintenance.

Some of the pertinent features of the reorganization are the placement of I&C in the maintenance department, the creation of a centralized planning and scheduling unit, and a restructuring of the procurement department to address problems with spare parts availability.

This new organization has, as one of *its first objectives, a review of the PM program by station personnel to evaluate the manufacturer's PM recommendations in light of equipment operating experience and reliability. This process may not be completed for some time, since much of the reliability informatfon on plant equipment has never been collected before. This review of the PM program will also factor in considerations for maintaining station personneJ doses as low as reasonably achievable (ALARA), an 1tem that was not considered in the program formulation to date.

1i

PSE&G fs in the process of install f ng a large, automated 1nfonnat1on management system that will include the features of work order control, job status, equipment hf story, and trending.

Impact of Changes.

PSE&G believes that the PM program is effectively.

maintaining pl~rit equipment and meets the requirements established by NRC following the ATWS events.

They also believe that the PM program 1s now beginning to improve the reliab111ty, availability, and materfal condition of safety-related equipment in the plant, even though ft 1s too early in the process for the results to be quantif~ed.

The changes at Salem that are presently being made and that remain to be made are those PSE&G believes will fmprove the cost-effectiveness of the PM program, reduce the radiation exposure of staff, and allow the PM program to adapt to changing plant and equipment conditions over the remainder of plant life.

iii

e

. TABLE OF CONTENTS *.

• EXECUTIVE

SUMMARY

111.

1.0 GENERAL SITE INFORMATION 1

2.0 SURVEY METHODOLOGY 1

3.0 DESCRIPTIVE DATA 2

3.1 PM ORGANIZATION AND ADMINISTRATION 2

3.1.1 Fonnal Structure 2

3.1.2 Admfn1strat1ve Goals, Pol1cfes, and Procedures 4

. 3.2 FACILITIES AND EQUIPMENT 7

3.2.l General Description 7

3.2.2 Specific Observations 8

3.3 TECHNICAL PROCEDURES FOR PREVENTIVE MAINTENANCE 10 3.3.1 General Description 10 3.3.2 Specific Observations

• • 11 3.4 PREVENTIVE MAINTENANCE PERSONNEL 12 3.4.1 General Description 12 3.4.2 Specific Observations 13.

'.~

3.5 WORK CONTROL 13 3.6 COSTS AND BENEFITS OF PM PROGRAM MODIFICATION 17 3.6.1 General Description 17 3.6.2 Specific Observations 18 3.7 PM PROGRAM PERFORMANCE EVALUATIONS 21 3.7.1 General Description 18 3.7.2 Specific Observations 22

4.0 CONCLUSION

S 23 iv

APPENDIX A.-.PLANT STAFF INTE~VIEWEO.DURING SJTE.VISIT APPENDIX B ~ATTENDANCE AT ENT~NCE AND.EXIT MEETINGS APPENDIX C ~ INTERVIEW SCHEDULE.

APPENDIX D - SALEM PLANT DATA APPENDIX E - SALEM EVENTS v

A:-1

.. e.1

• C.1 D.1 E.1

~ '

Docket No.:

Licensee:

e e

~.S. NUCLEAR REGULATORY.COMMISSION DIVISION OF HUMAN F~~TORS'TfCH~OLOGX (D~FT)

.SALEM.NUCLEAR POWER PLANT MSPP - SITE SURVEY REPORT 1.0 GENERAL INFORMATION 50-272/311 Public Service Electric & Gas Co.

P.O. Box 236 Hancocks Bridge, New Jersey 03038 License No.: DPR-70/75 Survey Conducted:

October 28 through October 31, 1985 Team Members: J. L. Koontz, NRC/DHFT Team Leader W. E. Scott, Jr., NRC/I&E W. B. Scott, PNL Team Leader J. t. Huenfeld, PNL R. W. Vallario, PNL 2.0 SURVEY METHODOLOGY The NRC has undertaken a program to investigate maintenance of U.S. nuclear power plants and, if necessary, instigate measures for its improvement. A multi-year Maintenance and Surveillance Program Plan (MSPP) (SECY 85-129) has been prepared to document this program.

The MSPP has two purposes: 1) to provide direction for NRC efforts to ensure effective maintenance and surveillance, and 2) to propose any necessary alternative regulatory approaches with respect to maintenance and surveillance activities. The MSPP identifies the technical and regulatory issues to be addressed and directs the integration and planning of the NRC's activities to accomplish these objectives.

One of the major projects of Phase I, Survey and Evaluation of Maintenance

• Effectiveness, is entitled, *Technical Assistance in Support of the Maintenance and Surveillance Program (Phase I) - FIN 82984." A major objective of this project is to obtain information on and assess the current practices of nuclear power plant maintenance and surveillance programs. A supporting objective of this effort is to evaluate the costs and benefits associated with the changes made in the Salem preventive maintenance (PM) program as a result of the anticipated transient without scram (ATWS) events at the plant 1n 1983. A PM protocol (formalized data gathering tool) was developed for the Salem site visit based on the protocol developed for other site visits undertaken for this project. This protocol focused on the aspects of plant maintenance programs that affect the conduct of PM:. The Salem site visit was the first 1

~

use of the PM protocol for data collection~. It is divided into seven mafn -.

sections:

PM Program Organ1iat1on and Administration

° Fae111ties and Equipment PM Technical Procedures PM Personne 1 PM Work Control PM Cost-Benefits PM Program Perfonnance.

The dimension of cost-benefit was not present in the other site visit protocols, but was added to the PM protocol to characterize the effects of the changes since* the ATWS events. This cost-benefit information was intended to be used in evaluating the cost-effectiveness of NRC's requirement for restructuring the Salem PM program.

Appendices A, 8, C, D, and E contain a list of the Salem staff who were interviewed in the data gathering phase, a list of attendees at the entrance and exit briefings, the schedule of site interviews, a sununary of the descriptive data collected, and a listing of significant events that occurred at the plant during the modification of the PM program, respectively. A completed PM protocol, and the materials and references obtained from the site, are part of the MSPP file which will be maintained for data background.

These materials have been cleared by PSE&G with respect to 10 CFR 2.790 (Public Inspections, Exemptions, Requests for Withholding).

3.0 DESCRIPTIVE DATA During the Salem site visit, data were collected on the seven items enumerated above.

These data are discussed here.

3.1 ORGANIZATION AND ADMINISTRATION 3.1.1 Formal Structure The Salem site organization begins with the Vice President--Nuclear, with some corporate offices located near the protected area surrounding the Salem site (see Figure 3.1). The quality assurance and quality control organizations are separate from the station organization and report offsite as well.

The senior management position inside the protected area is the General Manager--Salem Operations. The Maintenance Manager reports to the Assistant General Manager--Salem Operations along with the head of the opera~ions department (Operations Manager).

The remaining three department heads (the Technical Manager, the Radiation Protection/Chemistry Manager, and the Station Planning Manager) report directly to the General Manager--Salem Operations.

2

\\

\\..

(

Outside Protected Area (Off-Site}

Inside Protected Area (On-Site)

Assistant General ~anager Salem Operations

.J*

Vice President Nuclear General Manager Salem Operations Technical Manager Radiation Protection/

Chemistry Manager FIGURE 3.1. Salem Site Organization Station Planning Manager The station planning manager is currently responsible for scheduling outage maintenance projects. The technical manager heads the engineering support group, which includes the functions of reactor engineering and core perfonnance, reliability and assessment (including vibration monitoring, signature analysis, trending and other types of predictive maintenance), system engineering (including Nuclear Plant Reliability Data System input), and preparation of administrative procedures. Ultimately the technical department will have about 25 systems engineers who will become, in a sense, system experts.

The radiation pr:-otection/chemistry department, as the name implies, is responsible for controlling both the primary and secondary plant chemistry and monitoring, and cont!'Olling the amount of radiation received by station personnel.

Since the maintenance department at Salem was in the midst of a significant reorganization at the time of our visit, we will report on the organization as it existed at that time (see Figure 3.2). Later in this report we will indicate further organizational changes that are either incomplete or anticipated. The administrative procedure for station organization in place at the time of the site vf sft did not reflect the organization that was being implemented, but served as a good benchmark for comparing the present organization with that fn place prior to the ATWS event.

The maintenance manager is in charge of maintenance for both Salem units.

Reporting directly to the maintenance manager are three maintenance engineers (degreed), one for each unit and one directing a common support group.

Each unit's maintenance group is headed by two senior.maintenance supervisors_

(non-degreed)i one for electrical and one for mechanical, who report directly to the unit maintenance supervisor. Under the senior maintenance s~pervisors are the mechanical and electrical departments. The mechanical department is comprised of both machinists and those skilled in boiler repair. The electrical 3

\\ \\.

Assistant General -Manager Salem Operations I

I Operations Manager I I Maintenance Manager I

Maintenance Maintenance Maintenance Engineer Unit 1 Engineer Support Engineer Unit 2 I

I Senior Senior Staff of Senior Senior Supervisor Supervisor About 10 Supervisor Supervisor Mechanical Electrical Mechanical Electrical I

I 5 First Line 5 First Line 5 First Line 4 First Line I Supervisors Supervisors Supervisors Supervisors I Supervisor Supervisor Planning Planning Staff of 3 Staff of 4 FIGURE 3.2. Salem Maintenance Organization 4

department has recently been reorganized to include *both the electr_1cians and the instrument and control (I&C) t~chnician*s. All staff below the first 11ne supervhors are union (IBEW)".

There are about eight to nine workers (IBEW}

under each first line maintenance supervisor. Also under each of the two maintenance engineers 1s a separate planning organization. The planning organizations for I&C, electrical; and mechanical have only recently been combined.

There is no PM organization at Salem per se: rather, PM is one of the responsibilities of the maintenance dep*artment and, as will be discussed later, was incorporated into the preexisti"ng inspection order (IO) tracking system.

Whether an item is considered preventive maintenance or corrective maintenance (CM) is not clearly traceable, and in fact, completely loses its distinction at the worker level. Preventive maintenance is defined in the Salem Administrative Procedure for the Maintenance Program, AP-9, as *maintenance perfonned to prevent or reduce the need for corrective maintenance.* The working definition of PM is, however, somewhat broader, in that surveilla~ce tests.and in-~ervice inspections are included as part of Salem's PM program.

The plant organization has undergone significant changes since the ATWS event.

The previous organization and some of the major changes are discussed below.

• Previously, the assistant general manager had direct line responsibility for all departments including radiation protection and technical services.

• The chemistry department and the I&C department (including I&C planning) were previously part of the.technical services department. Under the current organization the chemistry department was combined with the radiological controls department, and the l&C department was separated and combined with unit electrical maintenance departments.

The l&C planning function was combined with the two-unit planning organizations.

It is noteworthy that some of the I&C personnel became part of the ra~iation protection/chemistry department, and have become a specialized, independent maintenance organization not under the direction of the maintenance engineers for either unit.

• The maintenance department was separated by unit and a connnon support group was formed.

This reorganization occurred just weeks before the site visit.

• Prior to 1983, station outage planning was accomplished offsite.

Further reorganization is under way to consolidate all planning into a central planning organization.

Furthermore, PSE&G was 1n the process of implementing a change 1n the station planning organization during our visit in an effort to improve the integration of preventive and corrective maintenance activities. Specifically, the reorganization will result in a more centralized planning organization that will have responsibility for preventive and corrective maintenance planning and scheduling for both the outage phase and the operational phase *. The planning organizations currently residing in each unit's maintenance department 5

~...

I under the maintenance er:ig1_neers will become. pa.rt of this centralized station planning ~rgan1zat1on.

• An add1.t1onal. ~hang~ taking ~la~e. is* the separation of yard a~(f.bu1Tdf ng *.

• services from the unit maintenance departments..Yard ~nd b~11dtng services w111 now be provided by a separate site services group.

A result of the ATWS event has been increased emphasis upon management involvement in plant activities.

• Management is expected to make tours of the plant regularly. The General Manager--Salem Operations indicates that plant management 1s expected to spend an average of 25% of its time in the plant on tours, observations, or special inspections.

3.1.2 Administrative Goals, Policies, and Procedures The PM program at Salem has been evolving since 1ts 1ncept1on in July 1982, when PSE&G implemented the Managed Maintenance Program (MMP).

This effor~

~as motivated by some INPO and industry findings at that time.

The scope was to include bdth balance-of-plant (BOP) and safety-related components. After the ATWS event, implementation of the MMP became part of NRC's required corrective actions, and the increased priority forced a postponement of BOP activities.

During the inception of the MMP, Westinghouse was selected as the contractor for development of PM requirements and was given responsibility for identifying the initial PM requirements. Oversight respons1bil1ty was maintained by PSE&G.

PM program development 1s now done entirely in-house.

In 1974 Salem instituted computerized tracking with the Inspection Order (IO) system. It was basically a tickler file for keeping track of recurring requirements.

The convenience of the IO system, combined with the tight time schedule for meeting NRC's ATWS corrective actions, resulted in the selection of the IO system as the vehicle for administration of the PM program at Salem.

The IO system is discussed 1n more detail in Section 3.5.

Since the 1nit1al 1dent1f1cation of PM requirements by Westinghouse, the PM program has continued to grow.

In fact, the number of !Os for both mechanical and electrfcal components has nearly doubled since completion of the Westinghouse contract. The number of mechanical and electrical IOs at the time of the site visit was approximately 17,500.

As the program evolved, PSE&G decided to make a concerted effort to optimize the PM program.

To do this, members of the maintenance department made several visits to other utilities to investigate the compatibility of other administrative strategies with the PM program at Salem.

The result of these efforts was the selection of the Plant Infonnation Management System (PIMS) developed for the Oiablo Canyon nuclear power plant by the TERA Corporation.

There is presently a major effort under way to adapt the PIMS system to the Salem PM program. Ultimately the IO system will be combined wfth the work order (WO} tracking system 1nto the new TERA-produced system, which.Salem will call the Managed Maintenance Infonnat1on System (MMIS).

The specifications of this system were unavailable at the time of the site vf s1t.

6

I In the fn1t1al _structurtng of the PM program by Westinghouse, PM requirements

~ere pr1orit1zed as follows:

1) manufacturer's reconmendat1ons :(e~g., vendor techn1ca1 manuals)e 2) engineering recmmiendations, and 3) s1te*specific operating experience.

PSE&G management believe that this has resulted in a less than optional PM program.

In order to optimize the progr8', they want to base PM requirements on station~specif1c infonnat1ono This continues to be difficult for PSE&G, 111ainly due to a lack of station-specific l"eliability and

. engineering data to modify the results of Westinghouse's initial work.

In many cases, PSE&G feels that the manufacturer's rec011111endations are overly conservative or too generic, with 11ttle regard for their actual application.

As an example, diesel generator PMs reconmended by the manufacturer are based on a continuously running engine, not on a standby engine that operates infrequently for short periods. The net result is PSE&G's belief that overly conservative PM requirements have been implemented.

Since completion of the *ATWS commitments in June 1984, efforts are being directed at fine-tuning the existing PM program to better match PM requirements with actual applications. To accomplish this objective, PSE&G places its*

hopes* in the 'MMIS.

It 1s planned that the MMIS be capable of handling much more data related to machinery history, spare parts, and management information needs than currently reside in the IO system. This new information will be useful for making engineering judgments related to the appropriateness and adequacy of the PM program.

As a result of the implementation of the PM program, PSE&G has identified some early lessons learned:

Spare Parts All levels of plant staff, from corporate to bargaining unit levels, indicated that the spare parts program has not been able to keep pace with the changes in the MMP.

Plant staff indicated that there have been instances in which the plant has oper.ated under a limiting condition for operation *(LCO) because spare parts have not been available.

PSE&G is currently implementing a component data base system that by 1986 will yield a complete inventory of parts on-hand.

By April 1987 it is planned to have in place a yet-to~be-specified statistically based inventory control system.

The new MMIS should interface with the spare parts system. Additionally, Maintenance and Plant Betterment is reorganizing procurement and material control, revising procedures, and conducting training to correct this spare parts problem.

Use of Manufacturer's Recommendations Manufacturer's recommendations have been overemphasized, largely due to a lack of documentation at the site to support alternative recommendations.

Resolution of this technically complex problem will result from the improved data gathering capabilities of the new MMIS.

7

Coordination of Actfyit1es Many plant activities have not been coordinated. For ~xample, the maf ntenance department may remove a valve from service to perform mechanical PM, only to have the same valve removed from service again a few weeks later to perform PM on the actuator.

These types of coordination problems have led to the 1q:>lementation of several corrective measures, the most s1gnf f1cant of which has been the centralization of station planning, as discussed in Section 3.1.1.

• Add1t1.onal1y,. comb-1n1ng I&C wHh the maintenance department; holding daily managers**and engineers' coordination meetings, and establishing the use of master system tagout schedules (see Section 3.5) will enhance the coordination function.

Automated Infonnatton Management

. Better computer system support is required and is expected to be

• implemented by adoption of the TERA MMIS.

ALARA Considerations As low as reasonably achievable (ALARA) principles were not given adequate consideration in the initial fonnulation of PM requirements. For example, PSE&G recognized that the number of containment entries had increased dramatically after implementation of the changes in the PM program.

Several entries were being made daily to conduct relatively unimportant (when compared against the man-rem impact) PM on fan bearing housings.

In the future, ALARA considerations wil1 be considered in defining PM tasks, and improved tracking of radiation doses will be available after implementation of MMIS.

Management Involvement One of the lessons learned by PSE&G is the need for increased management attention to daily plant operatiqn. All levels of management make regular tours of the plant.

On the average, the QA supervisor tours the plant once a week looking for leaks, tags, and calibration stickers; observing work in progress; and discussing plant status with the shift supervisor.

The maintenance manager spends 1.5 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> daily touring the plant (about 20% of his time). The general manager tours the plant daily.

The assistant general manager tours the plant about 15 times per month (prior to the ATWS, he toured it weekly). The operations manager tours the BOP portion of the plant daily and the auxiliary building and out-buildings twice a week, and makes frequent visits to the control room.

The Vice President--Nuclear tours the plant monthly.

The corporate goal for the PM program is to improve reliability and availability, achieving equipment availability of 80% by 1990, and 85% by 1995. Another goal is to achieve a PM:CM ratio of 75%:25%, or even 80%:20%

(present PM:CM ratio is about 1:5). A final goal is to reach a poi~t at which no corrective maintenance on critical safety equipment is required; i.e., all corrective maintenance will be eliminated by proper preventive maintenance.

8

• .Now that the co111111tme~t to. the NRC.has be~h. met_,* the Vke.Pres~dent--Nucl.ear._

would 11ke to see t~e PM program extend~d.to *eop ~qu1pmento *For example, he. *

. wants some r~ 11 ab111 ty cons 1 derat f ons app 11 ed to :cri ti ~a 1 BOP sys~ems to reduce economic risk* and to reduce the number of challenges to safety systems. Plant goals for 1986 are to complete 95% of all CM within 6 months: of work order issue~ to complete 95% of all PM w1thfn 1 week of scheduled c~letion, and to reduce reactor trf p frequency to 2 trips per year.

PSE&G also indicated that they desire to implement a predictive maintenance program to support the corporate goals. Presently, PSE&G takes some data on

.pumps and va-1 ves.

Corporate management believes that the plants, as built, will last considerably beyond design lifetime, and that an important benefit of a good PM program wf 11 be plant life extension. These and other anticipated benefits from a comprehensive maintenance program form the basis for PSE&G's firm commitment to improving maintenance at Salem.

3.2 FACILITIES AND EQUIPMENT 3.2.1 General Descr1ptfon The site visit teams were given a tour of the maintenance work areas and fac1litiese including the I&C shops, test instrument storage area, machine shop, electrical shop, in-plant spare parts storage area, and warehouses.

All the areas are well furnished and uncluttered. Working space appears adequate.

Interviews with staff substantiated this impression. Furthermore, bargaining unit staff reported that locker and tool space is generally adequate *

. A possible exception, suggested by plant staff, is the need for additional space for working on electrical (4 kV) breakers. This space is being provided by renovation of an out-building. Tours of the turbine building and maintenance shops found the housekeeping to be excellent. Lighting fs also adequate.

The shops are capable of performing almost all fabrication onsite. The only major function routinely accomplished offsite is the calibration of I&C test equipment. It is interesting to note that the training facility (located about 15 miles from the plant site) is well furnished with heavy machining equipment and does in fact perform some of the special fabrications required by the plant.

Labeling of valves and components in the opinion of the operations manager, is sufficient enough to minimize the *wrong train* problem. Continuous monitoring of valve labeling is accomplished during the conduct of valve lineups. The *wrong unit" problem is being addressed by color coding one unit blue and the other unit yellow.

As an additional level of assurance, symbols have been installed for.unit differentiation, a triangle for one unit and a hexagon for the other. These labels have not entirely eliminated the wrong unit problem, as evi.denced *by an event in November 1985.

The only suggestion for 1mprovement _ _made by the staff was to post names of rooms outside the rooms, especially those rooms that contain high radiation areas.

9

3.2.2 S,gec1f1c Qbseryat1ons The I&C shop and the electrical shop are be-1ng physically combined.-

The electrical staff's impression is that this will result in a better layout, w1th more spac~. The earlier layout had little room for spreading out, so jobs sometimes had to be ooved to other locations. The new shop does_ not, however, have good access for large equipment because 1t does not have crane or elevator access.

The priority, where possible, 1s to work on electrical equipment 1n place, rather than bring 1t to the shop.

The *staff feels that work bench space 1n the combined shop for both electrical and I&C w111 be adequate. Tools are issued permanently to electricians and I&C technicians. Calibrated electrical equipment is issued from the tool crib as ft is needed on a job.

The mechanics and electricians use small but adequate cages inside the radiation controlled area to store contaminated hand tools. These areas are not used for storage of contaminated calibrated electrical instruments, which must be conta1n~d (bagged) and taken in and out of the radiation controlled areas.

I&C maintains a *hot* equipment area for storage of calibrated gauges and other equipment.

The decision of whether to calibrate or dispose of contaminated

• tools is made by comparing replacement cost to decontamination plus calibration costs. Calibration of contaminated equipment fs performed in a dedicated area and controlled through the *hot* equipment area. Equipment in th1s storage area is inventoried weekly. Additional l&C storage areas are located inside both the Un1t 1 and Un1t 2 containments.

There ar thr~e major storage locations: Warehouse 2, the central warehouse, and the centralized storeroom. Implementation of the PM program, in combination with the construction of the Hope Creek plant adjacent to the Salem units, led to the construction of the central warehouse. While.large equipment is currently stored in both Warehouse 2 and the_ central warehouse, ft is planned that larg~ materi~l ~ill eventually be stored exclusively in the central warehous~. Other stor~ge locations include satellite warehouses and various mater1a1 drop=crf f po1 nts.

Plant staff be11~ve that hoisting and rigging problems do not adversely affect the plant's PM program.

The guidance given in NUREG-0612, Control of Heavy Loads, has been put into practice, w1th good results. It was indicated that general plant maintaina-b111ty 1n terms of space and removal/r1gg1ng paths is good. A possible exception 1s the area around the pressurizer, where space 1s very limited because of add1t1onal insulation and restraints. Where problems arise (e.g., finding spots to place rigging anchors), field questionnaires (FQs, a fonnal program governed by the adm1n1strat1ve procedures) are routed to Engineering for design and installation of the needed features. Plant staff i nd1 cate that Engineering has been very respon*s 1 ve.

According to one of the senior maintenance supervisors, the plant is moving toward eliminating the need for all outside contractors, and estab11shing an onsite calibration center that will do all calibration of test/measurement equipment. At the time of the visit, however, all I&C calibration ~as being accomplished offsite by a contractor. Other departments do most of their own 10

\\

.calibration. Calibration does have some impact upon perfonnance. of PM (e.g.,

a gauge may need to be calibrated every 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> of use): however, there was no indication of facility-related impacts.

In the I&C/Electrical equipment storage room a comprehensive status board is maintained to track the status of every piece of test equipment.

Re"cords are kept of calibration status and test equipment use by work order number.

Tool issue 1s governed by procedure. Personal tools are issued pennanently,

• Defective tools are handled differently by each department. Defective mechanical tools are identified by the mechanic and returned to a dedicated area in the storeroom. Defective instrumentation is identified by tagging.

Tagged equipment is sent to the calibration facili"ty by equipment room

• personnel *

• The communications system used at the facility consists of telephones, telephone

paging, a gray page system (operators only, and emergencies), and radios.

Radios are used mainly for remote jobs, high priority jobs, and inclement weather. Plant staff believe that plant communicattons systems are very good and that the systems do not adversely affect the perfonnance of PM.

• psE&G staff of many levels indicated that spare parts control (see Section*

3.1.2) 1s a problem area. Initial spare parts estimates were based on educated

• 11uesses and vendor recommendations, modified by plant experience. There was an admitted lack of fonnalization. At the time of the visit, no systematic method existed for predicting requirements for long lead time spares, nor was there any use of spare parts data for perfonning equipment failure analysis *

.One of the s~nior maintenance supervisors indicated that Salem may participate

~n one of the spare parts pooling systems. for major components.

One system

~sed for trying to identify what spare parts are necessary is the Mark Number Library, an inventory of all plant valves. The I&C department has developed a successful internal system (personal computer based) for keeping track of spare parts.

The site v1sft team observed that the training center is a very modern and

-comprehensive facility. Plant staff believe that the facilities are more than adequate to support maintenance training, and the site vi.sit team agreed. The saintenance training laboratories appeared to be equipped as well as, if not better thanw the onsite shop. The I&C shops contain numerous training

. simulators for instruction in both generic and specific applications. The i

training center maintains an essentially complete set of reactor protection instrumentation that 1s a duplicate (but reduced in scope, i.e., only one

• .representathte channel) of the system used at the plant.

PSE&G plans to have

.a function1~g reactor protection and control equipment down to and including

. an operat1ng control Y'Od drive mechanism.

So extensive 1s this operational

.setup that the training department accomplishes a significant portion of tands-on training at the training center rather than at the plant. Extensive

• use is made of mock-ups; the site visit team was shown a detailed mock-up of

,a reactor coolant pump seal used for maintenance training. The site v1sit team observed a group of technicians verifying the adequacy of a test procedure using the I&C equipment at the training center. According to plant staff, this 1s becoming standard practice.

11 I

I I

t I

i I '.

• c*

• The n~w or revised procedure is initiated by the station writer based upon departmental priorities governed by the Response Tracking Program. _

• The first stage of review is conducted by an individual designated as a stat1on*qualified reviewer (SQR).

The SQR is selected to review specific procedures based upon his or her education and experience._ A list of SQRs fs maintained by the Station Operations Review Co11111ittee (SORC).

The SQR conducts peer technical reviews as necessary, including cross-disciplinary review, if appropriate. The review fncludes a first=line determination of safety significance, a check of format, appropriateness of references, editorial/human factors, procedure verification (i.e.w can the procedure be used as written?), technical accuracy~ and othe*r items spec.ified in Salem Administrative Procedure AP-32.

• Nuclear Qualfty Assurance reviews only those procedures that involve new or changed inspection hold pointso Inspection hold point criteria are specified in AP-32.

• The maintenance department manager reviews the procedure for significant safety issues. If none exist, approval for implementation is given.

• If there is a significant safety issue, the procedure is routed for parallel review to the Offsite Safety Review Group and the Station Operations Review Co11111ittee.

The Offsite Review Group determines whether the change involves an Unreviewed Safety Question. If it does, the matter 1s given regulatory consideration. -If not, the procedure 1s approved for implementation by the SORC and General Manager, Salem Operations.

In addition to the above points, there is a provf sfon for on-the-spot changes*

to procedures that meet specific criteria. On-the-spot changes are routed through a review flow path similar to the one above that must be completed within 14 days of implementation.

3.3.2 Specific Observations Plant $1:aff 1ndicated that' all mandatory procedures have been instituted; however, development of new procedures (which we understood to be discretionary and not required procedures) is second in priority to completing the 2-year procedure reviews required by the license. The-workload required to keep up with these reviews limits the number of new procedures developed.

Since the training center maintains a growing collection of operating I&C equipment, I&C personnel may actually test a new (not yet approved) procedure on functional equipment at the training center. Where possible, PSE&G uses this method for verification and validation of procedures~ otherwise the SQR review serves as the validation and verification function.

Those procedures that do not receive full scope verification and validation are approved using desk=top reviewsr with full verification and validation occurring fn the field, when appropriate. According to the staff interviewed, although the training center's resources are used for validation and verification, there is no formal feedback from the training department to procedure writers.

13

Elimination of obsolete or superceded maintenance procedures requires the same review as that provided for a new or revised procedure *.

There is no difference in emphasis between procedures used for safety-related PMs and procedures used for BOP PMs.

Generic procedures (genera 1 and broadly app l1cab1 e to many maintenance situations) were still in use to a limited extent at the time of ~he site visit. These are gradually being replaced with specific procedures that contain detailed steps that can be used either for PM or CM work. These specific procedures remove much of the discretion allowed to the foreman by the generic procedures.

The master copy of procedures is kept on file by the procedure writer at all times. A working file of procedure copies is established by a dedicated plant clerk.

When a procedure is required for a work package, the procedure is pulled from the working file and stays with the work package until the job is completed.

PSE&G plans to include the procedure with the *pre-stagedu material (discussed in. Section 3.5) for the job: However, there is currently no.

mechanism, conceptual or otherwise, to.ensure that a procedure included in a work package of extended duration will be updated accordingly if a change to the procedure occurs while the job is active.

Salem does have a written policy (AP-9, Maintenance Program) regarding the use of maintenance procedures during the performance of work:

• The procedure shall be present when performing the task unless the job is of a simple or repetitive nature, and followed step-by-step while the task is being performed. Step-by-step adherence is required where the procedure has been developed for extensive jobs and reliance on memory cannot be trusted, or for tasks in which every step must be performed in a specified sequence.*

To properly interface normal operating pY'Ocedures with PM procedures, the plant maintains a set of integrated operating procedures that.specify in detail the PM items-(primarily surveill~nces) that*must be completed prior to changing technical specification operating modes.

Furthennore, PM procedures assume that initial system configuration is 1n accordance with normal operating requirements, consequently, PM tasks leave the system lined up in accordance with normal operating requirements upon completion.

3.4 fRE,YENIIVE MAINTENANCE PERSONNEL 3.4.1 General Description There is no distinction between PM personnel and maintenance personnel. The mechanical, electrical, and I&C departments are staffed to accomplish both corrective and preventive maintenance.

New staff are hired at entry levels and promoted to higher levels. Initial employee selection is the responsibility of the personnel department, including selection testing of applicants.

On 14

the other hand, the personnel department is not involved in selection of contractor personnel.

A new employee is given a certain amount of rotational experience to provide the opportunity to decide where in the maintenance organization he or she would like work.

New maintenance employees are sent through a basic technical skills training program that lasts approximately 26 weeks. Advanced skills training is conducted on a recurring basis. Approximately 651 of all training is station training (hands-on). Trainees ultimately are expected *to enter into the qualification program for the skill that they have chosen. Their progress and performance are monitored *by the "qualifications card" process.

3.-4.2 Specific Observations The training department meets with the operations and maintenance departments monthly at a minimum. This is a training review group meeting in which plant and industry operating experiences are addressed to evaluate training implications.

At the time of the site visit, the I&C training program had been accredited by INPO, with the electrical and mechanical programs** accreditation pending shortly. Accreditation of these training programs was subsequently granted November 14, 1983.

The training department makes extensive use of the inspection order (IO) system and station procedures to create a job task analysis (JTA) for maintenance positions.

3.5 WQRK CONTROL 3.5.1 General Descr1pt1on The heart of the current PM program is the IO system. Basically, it fs a computerized tickler file for keeping track of recurring tests and inspections.

It was in place almost a decade before the ATWS event and proved to be invaluable for rapidly implementing the new PM program.

The IO system serves two fundamental purposes:

1) 1t infonns the responsible department of scheduled tasks by providing a computer card (blue for Unit 1 and yellow for Unit 2) that suill'i1arizes details for ~ach particular task due (i.e., procedure number, schedule tolerance for completion, and other sW1111ary data), and 2) it provides station management with a list of outstanding and overdue IOs.

An additional system for tracking PM is the work order (WO) tracking system, nonnally used for corrective maintenance work control. This system, which has been automated for about 2 years, provides a means *of documenting approval signatures, work boundary isolation, work perfonned, parts used, post-maintenance testing, and operability testing requirements. Typically, all mechanical and electrical PM is controlled using the WO system, since a work order is written for each IO received by mechanical and electrical maintenance. The I&C department, on the other hand, controls their PMs using the IO system exclusively. In the I&C department, work orders are initiated for corrective maintenance only. Ultimately, the WO system and the.IO system 15

~. ~

will be combined into the new MMIS, with mechanical, electrical, and I&C all following the same procedures.

Once a WO 1s issued, the maintenance supervisor fn*charge of the work takes administrative charge. He.makes the initial designation of safety classffication; determines the need for hot work authorization,-checks for ff re protection impainnent, determines a recommended work boundary isolation, and mak~s other determinations required by administrative procedures. The WO is then reviewed by QA for verification of the safety-related or non-safety-related classification. The operations department authorizes and provides a tagout boundary for the work, with the shift supervisor providing the ff na 1 permfssf-on for work-commencement.

The mechanism by ~hich the operations department keeps track of system status is called the tagging request and inquiry system (TRIS)i established after the Salem ATWS evento TRIS fs used for generating and keeping track.of tags, as well as component position. The TRIS system fs essentially a data base of system status and can produce reports of various data. For example, the TRIS syst~m produces a Shift Turnover Log consisting of tagging requests confirmed, tagging reque'~ts released, and (on_ the midnight shift) a lfst of unavailable equipment.

An additional report compiled on each shift is the Components in Off~Normal Position Report.

In addit_ion to lfsting all components that are fn a position other than normal, this report gives the reason for the off-normal lineup.

The report can be generated for each system or can be generated for all systems.

To save time, the operations department maintains some selected

• standard" tagouts for components that are routinely removed and require complex isolation boundaries.

Upon completion of maintenance, the responsibility for retests and verification of operability generally resides with the operations department.

As a direct result of the ATWS event, the procedures controlling retests have been formalized.

The removal and restoration of safety-related equipment to operable status f s the subject of an operations directive. Retests either consist of running the operational surveillance test for that component, performing a

?~t~st as specified on the WO, or performing a retest using general guidelines 1n t~e operations directive. Status of operating systems!throughout removal and r~storation is continuously maintained by the TRIS system.

3o5.2 Specific Observations At the time of the visit, about 60% of the open work orders were for PM tasks.

Thf s number does not reflect the number of I&C preventive maintenance items initiated since I&C controls and tracks PM tasks directly from the IO cards.

The fac111ty has a large steady-state *backlog* of IOs {approximately 1000).

Facility staff warned against drawing conclusions based on backlog size because of the manner in which backlog fs defined. Because of IO system characteristics, tasks that have not even been fssued for work may yet be 11sted as backlog. Staff cited an example fn which IOs are issued from the system for preplanning and staging in advance of the completion date. Once these IOs are 1ssued, the system treats them as backlogged even though the completion date has not been exceeded. This situation has a large effect on 16

the number*of backlogged IOs 1mmed1ately before an outager when many tasks are prestaged and prepared in advance to eJiminate planning del~ys during an outage.

Surveillance testing is included 1n the IO system and is considered as PM.

The source document for delineation of respons1b111ty of individual surveillances fs Salem Administrative Procedure AP-12, *Technical Specification Surveillance Program.*

Currently jobs are scheduled by plann1~g organizations within the unit planning departments.

The unit plannin~ department staff includes formerly licensed senior reactor operators (SROsJ.

The pianning departments prepare the paperwork (e.g., WOs, radiation exposure permits, hot work authorizations, and drawings) for upcoming tasks, estimate the man-hours involved, and coordinate with the maintenance department to get the job done.

The planning department makes the final determinat1on of safety classification of a particular component (electrical or mechanical) from the machinery equipment lfst (MEL).

The planning department has begun to *prestage* jobs--that is, to assemble everything n~~ded for the job, including an approved tagout with tags. All that the individual assigned to the job needs to do fs have operations clear the equipment and hang the tags, and obtain a recent revision of the maintenance

• procedure for the applicable task. Ultimately, the station planning organization will be responsible for planning (with the exception of that accomplished by site services) and scheduling all PM and CM.

The station planning manager anticipates that all jobs will be prestaged with work packages that include the current revision of procedures.

In the pastr staff observed that during outages large numbers of systems were removed from service for essentially the entire outage with little or no coordination of maintenance tasks. The operations department was required at the end of the outage to complete all system reconfigurations, valve lineups, and retests in a short time in order to meet the schedule for restartc To remedy this situation, the station planning manager (whor incidentally, was previously the operations manager) w111 coordinat.e with all.departments to establish master system group.tagouts, where an *entire system w111 be tagged out for all the scheduled jobs. Work will be planned and sequenced such that th~ systems will be returned to operations 1n a staggered manner for performance of valve lineups and retests with less *end-of-schedulea pressure. Coordination of work 1n this manner 1s expected to greatly increase worker product1v1ty:

i.e., with this planning effort, 1t is expected that the next outage will be accomplished in 60 days.

In the past, scheduling was done around manpower:

in the future the station planning manager will actively schedule manpower to meet PM rquirements.

The un1t planning departments send out a worker feedback form w1th every PM task planned. Th1s is an avenue for the worker to provide ideas for improving or modifying each task. Completed feedback forms are given supervisory review.

Although the station planning manager is not responsible for planning jobs for the site services group, he identified some significant fnterfaces where he w111 be involved in scheduling. For example, the site servfces group will be responsible for the auxiliary boilers. Because of their potential 17

e operational impact upon the entire facility, the station planning manager will be responsible for scheduling (but not planning) that maintenance *.

The chemistry department, under the radiation protection/chemistry manager, perfonns maintenance on systems under its control. For instance, the condensate demineral1zers are operated (actual valve manipulations) by indjviduals in the radf ation protection/chemistry department, as opposed to the operations departmento Valves, instruments, and components operated by chemistry personnel are also maintained by chemistry personnel. The ultimate control system for PM tasks is still the IO systemi and will be the MMIS.

The evolution of this semi~autonomous group evidently results from the time when the technical department was responsible for chemistry and I&C.

Additional systems that fall under the control of the chemistry department are the hydrogen analyzing system, the waste gas analyzer, chemical feed valves in the chemical and volume control system (CVCS)p and some of the sampling valves outside containment.

The chemistry supervisor indicated that although there are no real conflicts, the definition of work boundaries fs not as formal as he would like it to be.

The autonomous nature of the chemistry group as 1t relates t6 PM is particularly significant 1n light of the fact that the Salem units have been credited with having the best chemistry control in the U.S.

Operation of all mechanical systems, with the exception of those under the control _of the chemistry department, is accomplished by the operations department.

The boundary of control for I&C PM tasks is up to, but not 1nclud1ng, instrument root valves--that is, the instrument must be removed from service by the operations department, and then the I&C technician may operate the isolated instrument. as necessary to conduct a calibration chec~.

Facility staff identified the following as the most PM-intensive components and tasks:

• 230V and 4160V ITE breakers--every breaker is removed and inspected.

-~

0 The boric acid transfer pumps~-substantial CM required because of harsh operating environment.

~

• Diaphragm valves~-frequent PM due to salt water corrosion.

• General valves--PM tasks are required, including packing and internal inspections. Solenoid valves in particular require considerable time (4 to 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> per component).

• Reactor coolant pump motor and seals~-various PM tasks required.

• Diesel generator--surveillance testing requirements.

• Steam Generators--inservice inspection (ISI).

PM tasks under WO control are reviewed by the shift supervisor for completeness of isolation and compatibility with plant conditions. Control of PMs not under 18

• r WO control (i.e., I&C) appears to be accomplished by control room operator interface with those technicians performing the test(s).

The repair and maintenance procedure system (RAMPS) was an old system, establ1shed several years.before the ATWS event, that was to become a PM system.

This was a corporate effort, as opposed to a site effort, and,111 be abandoned in f~vor of the more modern and capable MMIS.

One of the actions taken by the operations department as a result. of the ATWS event was to assign surveillance testing and equipment retesting to specific operating cy-ews, Y"ather than just scheduling the tests for a specific day.

The problem encountered was that a_ll sen$ it i ve or di f fi cu 1 t PM was deJ ayed at the discretion of the day and swing shifts, thus requiring the night shift to perform the task. Assignment of specific tasks to specific shifts ensures a better planning *of work load and plant configuration changes throughout a given day, rather than having a pile-up on one shift.

In general, contractors are not used to perfonn preventive or corrective maintenance at Salem. Contractors will continue to be used for specialty 1tems such as refueling. When this is done, quality control (QC) of contractor work will be under the direct control of the Salem QC organization. The upcoming 1986 outage will be the first time station QC will be used to control contractor work.

3.6 COSTS AND BENEFITS OF PM PROGRAM MODIFICATIONS 3.6.1 General Description Although the costs and benefits of the PM program.modifications have not been fonnally evaluated or quantified by Salem staff, the consensus is that overall improvements in equipment reliability have resulted, with concomitant improvements in plant reliability and public safety. However, plant staff at various l~vels fn the organization also believe that oyerimplementation of the PM pr@gr~m has ~~sulted fn h1gh~r operation costs, higher levels of personnel ~xposurei and lower overall cost~effectiveness of the PM program than is desfrableo Near-term efforts by Salem staff will focus on fine-tuning the PM program to 1mprove its cost-effectiveness based on assessments of the need for and impacts of each PM task.

Salem staff have not attempted to quantify the costs and benefits of the PM program modifications for the following reasons. First, as mentioned earlier, there is no distinct PM organization at Salem; rather, PM responsibilities are distributed across several departments. Additionally, the determination of whether an action 1s preventive maintenance or corrective maintenance is not clearly traceable, and in fact, loses its distinction at the worker level.

The result 1s a somewhat decentralized approach to PM that makes i~ difficult to develop aggregate estimates of costs and benefits.

A second and possibly more important reason that costs and benefits have not been quantified is that the information systems data base at the plant is not yet adequate for performing such cost-benefit evaluations. Detailed information 19

.-~

c.

necessary for cost-benefit evaluations will not be available until well after 1mplementation of the MMIS.

For example, information needed for estimating the operation costs, such as the actual staff time spent 1ndividually or collectively in PM tasks (before and after program modificat1ons), is not input to or available from the present data base systems (i.e., the present work order/inspection order systems or the machinery history file). Also, because there 1s no separate PM organization, Salem staff cannot estimate the effect that the PM program modifications have on overall staffing levels. As another example, information needed for estimating the benef1ts of an improved PM program, such as changes in equipment reliability, cannot be estimated because detailed equipment Mstory and fe11ab11ity data are not yet available.

It should b~ noted that plant management recogntzes the usefulness of having access to such cost-benefit information, and extensive efforts are under way to expand and upgrade the data base and 1nformation systems throughout the plant. Some of these planned upgrades have already been discussed (see Section 3.1.2).

Costs, benefits, and other effects attributed to the PM program modifications were*identifi~d through discussions with plant staffe These 1tems are listed and categorized in Table 3.1. Specific items are not ordered relative to the magnitude of cost, benefit, or effect (i.e., the items are not listed in any specific order). The table also 1ndicates the expected trend over time for each cost, benefit and other effect. This trend is a qualitative estimate based on 1nfonnat1on supplied by Salem staff of the expected change in each 1tem from the present to the t.ime when the PM program is fully optiinized.

3.6.2 Specific Observations Following 1s a discussion of some of the more specific insights gained through interviews of Salem staff regarding the costs, benefits, and other effects of the PM program modifications.

Accident and Public Exposure Risks.

Expanded PM has been implemented for safety-related equipment and limited balance-of-plant (BOP) equipment. Although 1t 1s too soon to have statistically sig*n1ficant perfonJJance data, Salem staff perceive that individual components and equipment*.are now more reliable. This 1ncreased reliability results in overall improvements 1n plant safety and reduced risks of accidents and subsequent public exposure.

For example, electrical supervisors already believe that.they are witnessing fewer circuit breaker problems. Additionally, motor control center and valve actuator problems appear to have decreased.

The staff feel that ult1mate safety benefits derive from two sources: 1) more reliable safety equipment, and 2) fewer challenges to plant safety systems facilitated by more reliable equipment.

Routine Occupational Exposure.

Accord1ng to Salem staff, there has been a tradeoff between 1mproved plant safety and increased levels of routine radiation exposure for at least some employees. For example, ma1ntenance staff are now required to enter radiat1on areas more frequently to conduct PM.

Although the radiation exposure levels for these individuals have increased somewhat, the levels are still within industry limits. Salem staff believe that some of the 1ncreases are due to PM tasks that may not contribute sufficient benefits to offset the increased radiation exposure. This suggests that ongoing 20

t.

~.

e TABLE 3.1. Costs, Benefits, and Other Effects of the PM Program Modifications Revi8't of manufacturer1s PM recC111111endations and dsvelopment of plant~spacific procedures and guidance Incorporation of PM tasks into IO system Training of personnel Cadditfonal>

Staff time to perform PM tasks (additional)

Demand for Replacement parts Ongoing review of cost--effect1veness of specific PM tasks PrQC:essin9 of Add1t1ona1 IOs and maintenance personnel feedback f onns Evaluation of data information needs, review of

~vailable data base systems, and planned acqufsftfon of MMIS Management and miscellaneous administration Benefits Equipment condftfon and reliability Plant Safety Overall plant av~ilabf11ty Corrective ~aintenance cost savings Equipment life extension Availability of equipment sel'"Vice information and reliability data More efficient and better trained personnel Other Effects Occupational exposure levels Scheduled outage length Backlog of PM and CM items Ca>

Trend 1s relative to ex1st1ng (present> condition.

21 Expected Futufl>

Trend 1n Cost Long-Tel"'lll O.cr~se Cimplem~ntatfon Cost>

Long-Term Decrease Cimplementatfon Cost>

Gradual Decrease Decrease Decrease Near-Term Increase then Gradual Decrease Decrease Signfffcant Increase 1n Near Term Slight Near-Term Increase Expected Future Trend 1n Demonstf!~le Benefit Sfgn1f1cant Long-Tenn Increase Signfffcant Long-Tenn Increase :

Long-Term Increase S1gn1ficant Increase Increase Significant Increase Increase Expected Trf n9 Over Time 1 Decrease Decrease Decrease

fine-tuning of the PM program and systematic weeding out of those unnecessary PM activities will not only improve the cost-effectiveness of the PM program, but will enable Salem staff to better satisfy ALARA objectives as well.

The health physics staff is presently involved in identifying PM tasks and other activities that result in high exposure levels: these tasks are scheduled for early review.

Plant Implementation Effects

• Implementation of the PM program required major efforts to review manufacturer's recommendations and develop PM procedures and guidance, as well as to input the PM requirements into the present IO system. Salem management relied heav1"1y on contractor support for this effort.

Although specific estimates were not available, staff indicated that significant implementation costs were incurred. Implementation costs will continue to accrue as Salem strives to optimize the PM program.

The cost of the MMIS is an implementation cost.

Plant Operation Effects. Several important costs and benefits have resulted that impact plant operation. Overall, the net economic effect on the utility of the imple~entation of the PM program modifications is viewed as a positive one. Staff believe that improvements in equipment reliability will result in improved reactor availability. With high replacement power costs, reliability improvements could result 1n a significant economic payback. Additional economic payback is expected through extended life of the equipment.

Plant staff noted that some unnecessary PM tasks are being performed and these reduce the cost-effectiveness of the PM program. Also, these tasks contribute to the backlog of PM tasks and to longer scheduled outages. Salem staff have instituted a program to systematically assess failure modes and effects to help identify components and activities that have the greatest influence on plant availability. To date, two assessments have been completed.

The first assessment was aimed at identifying the reasons for steam generator low-low level reactor trips. Staff expect that the results can be used to help prioritize PM activities. The second.assessment used similar analytic methods to identify the root caus~(s) for performance df fferences between Units 1 and

2. These efforts are expeet~d to resul~ 1n ~ mo~~ cost-effective and streamlined PM program, requiring less Staff t*ime and lower consumption of replacement parts,. also resulting in lower levels of occupational exposure, shorter outages~ and reduced backlogs of PM and CM items over time.

Staff observed that there are some tradeoffs between improving equipment reliability (while in service) and decreasing equipment availability by increasing the time a given piece of equipment is out of service because of PM tasks.

To minimize the impact of this tradeoff, they have 1mplemented a program to improve the schedul1ng and coordination between departments in removing equipment from service for PM to improve system availability. Also, because the expanded PM program places a greater demand on replacement parts, some delays have been experienced due to parts unavailability. Staff indicated that they are taking steps to resolve inadequacies in the parts supply inventory and thereby improve.the ava1lab111ty of serviced* equipment.

Although staff time allocated to PM activities has increased considerably, the overall staffing levels in the departments has been held constant by 22

management. Also, cetlings on overtime have been established. The cost 1mplications of these observations are not clear.

It was also noted that improved and more extensive tra1ning of maintenance personnel was one aspect of the PM program modification. Salem management felt that this improved training has resulted in more effic1ent ~intenance personnel and more consistently reliable work from these staff.

• Some PM activ1ties are performed so infrequently that they have not yet completed the f1rst, full cycle. Salem staff 1ndicated that the conclusions drawn regarding equipment reliability are preliminary and that final conclusions may not be available for several years.

The staff believe the PM program modifications have resulted fn a reduced need for corrective maintenance based upon early, 11m1ted experience. However, the magnitude of the change is presently unknown.and will rema1n unknown until additional operational experience and data are obtained. Data collected 1n the future will be used by Salem management to assess performance relative to newly formulated perfonnance goals (see Section 3.1.2).

3.7 PM PROGRAM PERFORMANCE EVALUATIQNS 3.7.1 General Description At the corporate level, Salem staff have developed and 1mplemented a performance monitor1ng system to track significant aspects of performance in plant operations. A monthly perfonnance report is generated for PSE&G's Vice President~-Nuclear and 1s discussed in the monthly meeting of Salem's nuclear department managers.

Some of the performance data, charts, and figures are posted in the plant for staff to review.

The plant performance mon1tor1ng system does not treat PM as a separate performance issue. Consequently there 1s no performance 1ndicator devoted exclusively to PM.

Rather, PM 1s 1ncorporated into the following two performance indicators:

• Salem Maintenance Work Order Backlog

• Salem I&C Inspect1on and Work Order Backlog.

No data for these 1ndicators were available.

At the department level, there 1s no more specific performance 1nfonnation ava1lable other than at the corporate level. Because of weaknesses 1n the present data base and 1nformation systems (see Sect1on 3.6), 1t*1s not possible to obtain basic 1nformat1on that would be useful 1n assess1ng PM program performance (e.g., labor hours devoted to PM).

Furthermore, the lack of baseline data on equipment performance and prior problems has made 1t difficult to assess the effects of the PM program.

Finally~ the expanded PM program has been 1mplemented for only a year and 1t 1s too early to observe the long-term anticipated benefits. Major efforts under way to develop the MMIS (see 23

. *~.

-~ '

Section 3.1.2) should provide the types *of infonnation useful in perfonning PM program eval uatf ens.

The fmplementatfon of the Managed Maintenance Program has resulted in sweeping changes f n the administration of maintenance at the Salem nuclear power plants.

As previously described 1n this reportr the Salem fac111ty fs -still undergoing transition. The continuation of extensive changes in the way fn which maintenance fs controlled, combined with considerable reorganization of the maintenance organization raises the concern of possible interim decreases in safety during the trans1t1on phase. Data is currently being collected on facility perfonnance in support of other NRC projects, specifically a ma 1 ntenance data base 1s be1-ng comp11 ed 1-n support of NRC

• s maintenance initiatives, and overall facility perfonnance measures are being established.

Although these initiatives are not yet complete, the data relative to the Salem plants was reviewed resulting in the following conclusions.

• The total number of violations/noncompliances at Salem 1 & 2 actually decreased during 1984 *

.

• During 1984 while the number of low severity violations remained relatively constant, the number of high severity violations (severity code less than 3) went from ~6 in 1983 to zero in 1984.

• The percentage of violations that were maintenance-related increased from O~ to 181 at Unit 1 and increased from 4% to 11.1 S at Unit 2 during

• 1984.

• The maintenance-related violations recorded during 1984 were all classified as low severity.

• The cause of the increase fn maintenance-related violations in 1984 is not readily apparent.

In order to give an indication of the operational history at the Salem plants, Append1x E presents a su11111ary review of operational occurrences at Salem 1 &

2 that have occurred since the February 1983 ATWS event. This summary was abstracted from Nuclear Power Experiences, and includes all available events through the December 1985 issuance. Occurrences categorized under Turbine Cycle Systems were not included in thf s list. Although the list given in Appendix E 1s perhaps not all inclusive, ft certainly can be considered to be representative of the types of occurrences that have been happening over the past three years. After reviewing those occurrences we cannot conclude that any adverse affect on safety has occurred as a result of restructuring the PM program at the facility.

3.7.2 Specific Observations Salem management is following guidance provided by the Institute of Nuclear Power Operations (INPO) in selecting and developing ~lant performance indicators for its own use. Additionally, management has developed systems for assigning cause codes to deficiency reports to help identify trends in probl~ms and thereby improve its ability to take corrective action.

24

APPENDIX A - PLANT STAFF INTERVIEWED DURING SITE VISIT Vice President, Nuclear General Manager~-Salem Operations Assistant General Manager~-Salem Operations Technical Manager,

Radiation Protection/Chemistry Manager Station Planning Manager Operations Manager Maintenance Manager Maintenance Engineer Unit 1 Maintenance Engineer Unit 2 Maintenance Engineer Support Supervisor, I&C Sr. Supervisor0 Boiler Repair Supervisor, Maintenance Machinist Electrician l&C Technician Manager, Station Quality Assurance Quality Control Supervisor Principal Engineer--QA Auditing Acting Manager, Procurement Warehouse Supervisor Principal Engineer, Engineering & Plant Bettennent Assistant Training Manager Supervisor, l&C Personnel Affairs Manager Lead Engineer Staff Engineer Reliability Engineer Principal Re11abi11ty Engineer General Manager-Licensing & Reliab111ty Techn1cal Supervisor-Chemistry NPRDS Coord1nator NPRDS Coordinator Planning Coordinator~Mechan1ca1 Ma1ntenance I&C.Plann1ng Engineer Shift Superv1sor-Operations Reactor Operator A.1 Corban McNef 11 John M. Zupko, Jr L. Mi 1 ler John Ronafalvy J. Trejo L. Fry J. Gueller Pell Wh1te John Morrison Robert VanderDecker 8i11 Lowry Tom Spencer Lou Noc1t1 Craig Short 8i11 George Doug Thomas Dave Perkins Dennis Tauber Paul Benini Mike Rosenzweig 8i11 Burke Bob Baird Bob Edmonds Bill Nieczpiel Stan Kosi erowski Ro.b Antonow Ray Theriault Frank Thompson Dick Murray Larry Ka11111erze11,

. *Robert Watson Jim Soletsk1 Frank Kaminski Jerry Bonzella Warren Downey Steve Sauer Bob Olsen

APPENDIX B - ATTENDANCE AT ENTRANCE AND EXIT MEETINGS Entraru;e Meeting J. L. Koontz, NRC/DHFS J. M. Zupko, PSE&G Jo W. Morrison, PSE&G W. E. Scott, NRC/I&E T. J. Kenny, NRC/Sr. Resident R. J. Potter, PSE&G W. H. Burke, PSE&G D. A. Perkins, PSE&G J. C. Gueller, PS£&G J. P. Ronafalvy, PSE&G R. WG Vallario, PNL J. C.*Huenfeld, PNL J. ~. Leachg PSE&G R. F. Murray0 PSE&G

• J. L~ Rupp, Interfacts R. Yanderdecker, PSE&G R. Borchardt, NRC/Res1dent L. Kammerzell, PSE&G W. B. Scott, PNL P. White, PSE&G L. Fry, PSE&G L. Ko Miller, PSE&G Exit Meeting J. L. Koontz, NRC/DHFT H. R. Booher, NRC/DHFT L. Kafililerzell, PSE&G L. K. Miller, PSE&G P. White, PSE&G J. Gueller, PSE&G J. T, Boettger, PSE&G J. Po Ronafalvy, PSE&G W. B. Scott, PNL J. W. Morrison, PSE&G D. A. Perkins, PSE&G B. Hall, PSE&G L. Fry, PSE&G D. C. Fischer, NRC/NRR J. L. Rupp, PSE&G R. W. Borchardt, NRC/Res1dent B. Eu111110nd; PSE&G B.1

I

---

~----

APPENDIX C ~ INTERVIEW SCHEDULE October 28 (Monday)

Morning Arrive at Salem plant, watch indoctrination film and take exam for unescorted security access, obtain badges and parking p~nnits.

Attend entrance briefing with plant staff to discuss purpose of site vf s1t and receive an overview of the Salem PM program.

Afternoon Df scuss organization and administration questions wfth general manager--Salem Operatfons and assistant general manager--Salem Operationso Take plant familfarfzat1on tour.

October 29 (Tuesday}

Mornfng Tour the trafnfng facilfty and discuss personnel trafnfng questions wfth assistant training manager Discuss questf ons regarding PM program implementation and operation costs wf th maintenance engineers.

Discuss organfzatfon and administration, facf lfties, procedures and work control questfons with maintenance manager.

Dfscuss survefllance testing, work control, co11111unfcatfons and personnel questions wfth operations managero Aft~rnoon Df scuss PM progr~ perfonnanceu procedures; and work contrel questions with t~chnical manager.

Df scuss PM procedures questions with maintenance engf neers.

Discuss work control and occupational exposure questions with radiation protection/health physics manager.

Discuss PM program reliability analysfs questions with engineering and plant bettennent staff.

Dfscuss maintenance work control, facflfties, procedures, and personnel questions with senior supervisor and supervisor of maintenance.

C.1

- -~

October 30 (Wednesday)

Morning Discuss PM goals, objectives and management co11111ittment issues with vice president--Nuclear.

Discuss facilitiesw work control, procedures, equfpment and tools questions with a machinist, an electrician and an l&C ~echnician.

Discuss PM program fonnulation questions with maintenance engineers.

Afternoon Discuss administrative controls, coordination and communications questions with quality assurance and quality control managers.

Discuss personnel and recruiting questions with personnel affairs manager.

Discuss schedule and outage questions w1th station planning manager.

Tour warehouses and discuss spare parts questions with warehouse manager.

Discuss clearance request, tagging and system restoration questions with operations manager, operations shift supervisor and an operator.

October 31 {Thursday)

Morning Review work control, work order accounting and scheduling questions for clarification.

Discuss procurement, spare parts specifications questions with engineering and plant bettennent engineer.

Discuss lubrication ~nd surveillanc* testing questioris with operations manager.

Brief resident inspectors, NRC Headquarters management on material to be presented in the exit briefing.

Afternoon Attend the ex1t briefing to present to the licensee the preliminary results of the site visit.

Leave the site.

C.2

• • • -**-*--*-----*-**-~-*-**-*--

APPENDIX D - SALEM PLANT DATA Type PWR L1censed Thennal Power (MWt) (Ul/U2)

Cool1ng 3338/3411 Delaware River Westf nghouse Westinghouse Reactor Suppl1er Turb1ne-Generator Manufacturer Eng1neer PSE & G Constructor UE & C 9-25-68/9-25-68 4-6-77/8-18=81 12-11-76/8-2-80 6-30-77/10-31-81 Construct1on Pennits (Ul/U2)

Operating L1censea (U1/U2)

Critical First Time (Ul/U2)

Coll'lllercial Operation (Ul/U2)

Most Recent SALP Ratings

}'

For Period 10-1-83 to 8-31-84:

Maintenance 2

Improved Surveillance 2

Improved SALP categories definitions Category 1:

Reduced NRC attention may be appropriate.

Licensee management attention and involvement are aggressive and oriented toward nuclear safety; licensee resources are ample and effectively used so that a high level of performance with respect to operational safety or construction 1s being achieved.

Category 2:

NRC attention should be maintained at normal levels. Licensee management attention and involvement are evident and concerned with nuclear safety; licensee resources are adequate and reasonably effective such that satisfactory performance with respect to operational safety or construction 1s being achieved.

Category 3:

Both NRC and licensee attention should be increased. Licensee management attention or involvement is acceptable and considers nuclear safety, but weaknesses are evident; litensee resources appear to be strained or not effectively used so that minimally satisfactory performance with respect to operational safety of construction is being achieved.

D.1

I

'\\

APPENDIX E SALEM EYENTS This Appendix presents a sunmary review of operational occurrences at Salem 1 and 2 that have occurred since the February 1983 ATWS event. This summary was abstracted from Nuclear Power Experiences, and includes all available events through the December 1985 issue.

SALEM 1 May 1983 - Contaminated water released during pumping down a SW system for maintenance (cold shutdown). Contractor personnel were involved.

May 1983 = V1ta1 instrument inverter erroneously deenergized - inadvertent operation of s1mi1ar equipment in opposite unit.

May.1983 -

~TD bypass 11ne spring hangers found pinned - procedural controls, personnel oversight. The lines had been pinned so that they could be used to support lead shielding. When the shielding was removed, the pins were_ not, and were not discovered for about 8 months.

May &

- Security guards accidentally damaged a DG relay, and actuated a DG Jul 1983 trip.pushbutton.

Two different guards involved. Temporary security post was established due to door modification.

Jun 1983 - Pressurizer control, and backup heaters deenergized - control relay coil overheated. Tech Spec LCO entered.

Jul 1983 - DCRs and procedure modification to counteract BIT inlet valve boron buildup.

Jul 1983 = #11 BAT pump b~~aker_trip declared inoperable **O so11d1fied boron.

Tags ~~re clear~d on 112 that was down for maintenance.

Aug 1983 = Charging pump $peed increaser lube oil cooler leak - erosion and corrosion due to sf lt laden SW.

Aug 1983 ~ AFW failed to start following a low-low SG levele Following maintenance and testing activities one week earlier, the trip valve had been left in the tripped position, and the limit switch for this was out of adjustment so that the control room had no indication of this condition.

Sep 1983 - Process computer failure - high ambient temperature. Engineering evaluation conducted on ventilation problem.

Oct 1983 ~ FM in SI pump oil reservoir - pump l~ak off drain plugged and overflowed.

Pump inoperable for 1-2 days.

E.1

Nov 1983 ~ Conta1nMent sump in-leakage - condensation from CFCU configuration change.

Tech specs LCO entered.

NOTE:

There is a history of CFCU silting problems in both units.

Nov 1983 = Containment H2 analyzer inoperable ~ detector failed.

Nov 1983 ~ Pressure transient ~ steam dump pressure controller servo unit failed

- setpoint station output unstable.

Nov 1983 - Radiation monitor channels failed to actuate containment ventilation isolation upon high radiation signal. Pressure relief was performed without containment auto-isolation function.

Dec 1983 ~ SG handhole cover leakage.

Dec 1983 = CFCU 114 fan inlet flexible connector deterioration - temporary repair and design change.

Dec *1983 = Leak discovered on diesel f1re* suppression pump ~ cooling water valve seal failure - normal wear.

Tech spec entered.

Dec 1983 = EOG Jacket water cooling leak from pump elbow - design changed to counteract vibrational stresses.

Feb*l984 - CFCU #15 SW supply header vent line leak - corrosion - excessive weld damaged protective cement coating.

Feb 1984 ~ SW intake structure max water level calculations incorrect. Design discrepancy discovered by NRC.

Mar to Excessive containment isolation valve leakage during LLRT - seat Aug 1984 - damage and packing leaks.

Apr 1984 = RCS loop RTD bypass line valve discs separated from stems ~

backs~at1ng force, thermal effects *oo Rockwell globe valves.

May 1984 ~ HPSI throttle valve stem-disc separations - generic implications.

May 1984 - Fire pUii1p discharge valve failed to open. Stem and disc separation when valve shut for surveillance.

Jun 1984 - Tech Specs and FSAR inconsistent on No *. of operating RCPs needed in mode 3&

Inconsistency was discovered by W.

Jun 1984 ~ Vital bus blackout actuation, DG run without cooling water - admin, personnel error. The reactor was defueled, Tech Specs no longer prevented removing more than one vital bus from service.

Jul 1984 - Charging pump 112 seizure~ Debris in.pump casing and suction lines.

Discovered during surveillance. Resin found in RWST.

Charging pump Ill also worn.

E.2

l

. Oct 1984 - CCW valve return from RCP thennal barrier failed closed. Pressure transient caused by flexing plate phenomena when starting another SW pump.

Lfmitorque operator. failed.

Oct 1984 - CIVs inoperable ~ redundant stop valve failed leak rate testing, used as CIV during stop valve repair - operations not aware that valve had failed leak rate test.

Oct 1984 - CIVs inoperable - stop valve failed leak rate test - all valves were operable - test results documentation error. Original test had not included boundary valve leakage.

Nov 1984 - Two CFCUs w~re declared inoperable ~ vent 11-ne weld leaks caused by condensation. Tech spec LCO entered.

Nov to Reactor trip - OT Delta T setpoint reduced - switching error -

Dec 1984 - faulty power range detector.

M~r 1985 - Unidentified RCS leakage exceeded limits - source was #13 CFCU

.motor cooler head gasket damage *** SW weld repair had just been done, leak occurred during retest.

SALEM-2 Jan to Aug 1983 - POPs and PORV valve modifications Mar 1983 - R/A monitor sample pump 1nadvertenly deenergized - deficient load list. The pump was powered from the breaker supplying R-41.

R-41 was deenergized to allow a design change.

The iodine and particulate samples for this monitor were not representative of actual values for a period of about one week.

Apr 1983 - Oysters plugged SW system coolers.

Apr 1983 - Pacific Scientific mechanical snubber problems.

Apr 1983 - Unplanned releases of low-level liquid radwaste to the owner controlled area - valve failures, and procedural error.

Apr 1983 - R/A monitor sensitfvfty low - sample line leaked.

Apr 1983 - Many personnel were contaminated by airborne R/A during SG nozzle dam maintenance - temporary ventilation was 1nplace, but not started

- monitoring procedures were not adequate. About 70 individuals were contaminated.

Apr 1983 - Cardox system pilot valve left closed following recharging - no procedural guidance.

E.3

'I - '

Apr 1983 - RCS vent path lost ~ blank flange installed - personnel error. The SS understood that approval of the work order (WO) was requf red for shop fabrication of the flange, and that operations would be informed prior to its installation. The boiler repair supervisor, however, understood the approval of the WO to be for all required work.

Apr to Invalid waste gas decay samples used - inadequate sample flow path Jul 1983 ~ - personnel error. Procedures were approved prior to verification

• of the flowpath.

What was thought to be decay tank gas samples was really instrument a1r. Th1s went on for two months.

May 1983 - V1ta1 instrument bus inverter damaged - containment integrity not est~bl1shed within 8 HR Tech Spec 11m1t - maintenance problems.

Work on an inverter took longer that expected, resulting in not being able to establish containment integrity withfn the required 8 hrs. Problem wfth procedure for reenergization of vital inverter was suspected.

May 1983 -1>fesel fuel ofl tank overflowed to storm sewer - operator error, no procedural guf dance.

Jun 1983 - Transformer fire - fault fn secondary windings - violation of procedures in shutting faulted breaker without detailed investigation.

Jun 1983 - SW Bay flooded, valve gasket failed due to improper gasket installation after recent header cleanfng. Tech Spec action statement entered, unusual event was declared (all SW was inoperable).

Jul 1983 - Control Rod Dropped - CROM cable connector pins pushed back into insulator block.

Jul 1983 ~ Core enthalpy hot channel factor exceeded low power limits ~ control rod insertion changed.

Aug 1983 ~ RCS pressure transients~ inadvertent SI ~ pressure controller replaced - had caused spray valves to fail open when different pressure instrument was selected.

Aug 1983 - RCS flow transmitter failure - Oscillator failed.

OCR submitted.

Sep 1983 - CIV leak rate excessive *** possible design problem.

Sep 1983 - Containment compressed air supply valve was left open with unit in hot shutdown - operator error. The valve being 1n an out of normal position was fnfact shown on the TRIS ~out of normal* report, but was over looked by the individuals performing the integrated procedures on two occasions, gofng into mode 3 and 4.

E.4

I

. Oct 1983 - During routine operation ft was discovered that the trip' fndfcatfon light for the turbine-driven AFW pump was illuminated.

Th~ trip valve had been left f n the tripped position following testing 1 hr earlier.

Oct 1983 - Aux FW pump and turbine trips - personnel error trip valve latching

- procedure deficiency - warmup line drain valves not checked open

- condensate caused water ha11111er.

Nov 1983 - DG inoperable - fuel 011 pump coupling set screws loose, manifold to injector gasket leaking.

Dec 1983 - DG voltage regylator control power transfonner-grounded.

DG tripped on over current due to high reactive load.

Jan 1984 - RCS loop RTD bypass line isolation valve stem to d1sc separations.

Feb 1984 - PORV actuation - POPs *ini*tiatfon - RCP start induced pressure transient.

Feb 1984 - Shutdown cooling lost - RHR suction valve closed, flow lost - breaker status wrongly tagged - procedural, personnel error. Bezel covers for the RHR valves indicated that they were tagged open for a previous job, so the same clearance was used for POPs testing.

Unknown to the SS, the clearance on the valves had already been cleared, but the bezel covers had not been removed.

Feb 1984 - DG output breaker trip - overcurrent relay did not reset after RCP start. Attributed to isolated case of relay failure. Planned to replace relay when a suitable replacement becomes available.

Mar 1984 - Unit shutdown - vital bus deenergized, IRPI lost, CIVs made inoperable during an attempted DG paralleling - paralleling out of phase was suspected.

Apr 1984 - Waste gas decay tank radf~actfve effluent released without proper sampling ~ procedural error. Afr to the sample isolation valve had been tagged out about one year earlier. Although the tag had been cleared, the tag had not been removed.

The contents of the tank had been released on two occasions without a representative

. sample.

May 1984 - Stem - Disc separations in Rockwell Int globe valves - generic _

problem. Manufacturing flaw.

May 1984 - Reactor Trip from 1001 pwr - High SG level signal injected during troubleshooting - lack of procedural guidance, personnel error.

Lack of control by supervision.

Jun 1984 - RHR inoperable - all 4kV group and vital buses lost - wrong switchgear opened - operator error. SS instructed EO to call Unit 1 control room to make out tags, and then inform the Unit 2 control E.5

q,

- i -,

room to open the breaker.

The EO told the Unit 1 control room to

.make out the tags and open the breaker. The Unit 1 operate~ opened the respective breaker for Unit 1, causing a loss of power (4kV) to both units.

Jul 1984 - Charging system main header piping leaked - vibrati9n 1nduced fatigue

- unit shutdown 0 ECCS subsystems declared inoperable and NRC infonned.

Jul 1984 - Rx trip, SI initiation - pressure transient - POPs relief valve block valve failure. Occurred during POPS functional test.

Jul 1984 - Both containment spray pumps out of service - personnel error.

Violation of Tech Specs for 3 hrs.

Aug 1984 - Containment airlock design deficiency - unqualified D/P gages discovered during OCR review.

Sep.1984 - Vent effluent sample pump sw1tched off - iodine composite sampling missed. Charcoal cartr1dges were monitored to detennine environmental release. During one release, charcoal cartridge had already been changed and the old one discarded.

Dec 1984 - Liquid radwaste releases not recorded - computer malfunction.

Discharge was monitored.

OCR submitted to require strip chart.

Jan 1985 - DG Trip - High jacket water temp signal - SW control valve actuator installed incorrectly - procedures not followed.

The root cause of the event was attributed to a lack of directions to the repair personnel before the job was perfonned and the failure of repair personnel to follow the procedure as written. The I&C department was not contacted to remove the air lines as specified in the procedure, and the valve was not installed per instructions contained 1n the vendor manualo Feb 1985 ~ Pump bear1ng damage - pump lube oil burnt, darkened - FM in EXXON Terrest1c T-68 lube oil. Many pumps involved.

Mar 1985 - Reactor tr1p ~ SG steam flow/feed flow mismatch - unit synchronized with SG steam flow b1stables tripped - personnel error.

Mar 1985 - POPs initiation during RCS fill and vent - reduced margin between RCS pressure and POPs channel setpoints.

Mar 1985 - DG failed to start during tests - Overspeed trip device shaft and fuel rack shaft were binding. Not discovered until the third similar failure - test frequency was boosted to every 3 days rather than every 14.

Apr 1985 - Reactor, Turbine trip - low turbine lube oil pressure - procedural deficiency.

E.6

• .. _ J_ \\ J

~ -.

~

• -,. ~

May 1985 - Dropped Control Rod - Grfppers faf led to energize, poor contact between CROM cable connector pf ns.

Jul 1985 ~ Un1t Shutdown - Hfgh RCS leakage ~ packfng gland leaks on pressurizer safety valve loop seal drafn valve.

Jul 1985 = DG inoperable fuel of l leak - cylfnder fuel fnJector fftting loose.

Aug 1985 - Low CCW Hx SW flow - outlet valve faf led closed, valve actuator stem separated - caused by vibratfon induced by operation with cavitrol tube bundle removed *** new tube bundle on order.

LCO entered and S/D began.

Sep 1985 - Leaking CFCU *23.* Declared inoperable. Caused by corrosion of pfpe nipple weld. Protective cement coating damaged during installatfon of nipple.

ABBREVIATIONS CROM -

Control Rod Drf ve Mechanisms RCS Reactor Coolant Systems Rx Reactor SI Safety Injection RCP Reactor Coolant Pump POPs -

Prfmary Overpressure Protectf on System PORV -

Power Operated Relief Valve CIV Contafnment Isolation Valve ECCS -

Emergency Core Coolfng System CCW Component Coolfng Water Hx Heat exchanger SW Service Water LCO Lfmitfng Condftion for Operation SID Shutdown RTD Resistance Temperature Devf ce OCR Desf gn Change Request DG Df esel Generator R/A Radiation/Alarm DIP Df fferentfal Pressure RHR Resf dual Heat Removal SS Shf ft Supervfsor EO Equipment Operator I&C Instrumentatf on and Control AFW Auxiliary Feedwater SG Steam Generator IRPI -

Indf vfdual Rod Posf tion Indicatf on FM Forefgn Material

• BIT Boric Acfd InJectfon Tank HPSI -

Hf gh Pressure Safety Injection CFCU -

Containment Fan Cooling Unit LLRT -

Load Leak Rate Test BAT Boric Acid Addftion Tank E.7