ML18151A427

Assembly of rebuilt motor

02-CP-PM0-15C

Motor bearing replacement & test

01-CR-CRN-8

Inspection & test of trolley

hoist

01-AS-PMD-lA

Pump motor leads reconnect

Ol-EE-P-100

Calibration of level switch

Ol-EE-P-101

Calibration of level switch

.Ql-81-RCDR-SDE

Troubleshooting recorder inputs

02-W-PM0-18

Inspection, lubrication, and

service of pump motor

Ol-HS-PM0-3B

Inspection, lubrication, and

service of pump motor

01-BB-REL-PRRXB

Replacement of relay

02-EPL-BKR-181

Troubleshoot and return breaker

to service

The team examined, in detail, 36 completed work orders, and in addition,

reviewed post maintenance testi~g on a~other 23 work orders for switchgear

sent to an outside vendor for refurbishment.

The team reviewed 32 electrical CM and PM procedures.

Eleven procedures were

for the emergency diesel generators. Eleven (9 PMs & 2 CMs) were for MCCs,

circuit breakers, and switchgear. Three of these procedures were proposed,

bu.t not approved.

The other procedures were for various types of CM and PM

tasks for motors, time relays, batteries, battery chargers, cable

terminations, and electrical .calibrations.

findings

The walkdown inspections, observations, discussions with licensee personnel,

and document (work orders and procedures) reviews revealed the following:

In general, the material condition and housekeeping for the SR 480V and 4160V

and the NSR 4160V AC MCCs and switchgear panels were in good condition with

the following exceptions:

0

Iri the majority of the SR 4160 V panels, spare wires were not properly

identified and capped (taped back).

In several instances the tape was

loose or there was exposed bare copper. In four panels, Unit 1 - 15Jl0

and 15Jll and Unit 2 - 25J3 and 25Jll, the wrong type of tape was used.

This orange and purple plastic tape was extremely loose.

The licensee

took corrective action to fix these problems.

In the NSR 480 V MCCs, the cubicles were dirty. In NSR MCCs 2C2-l-53 and

lGl-1-lC numerous lifted leads were not identified or capped.

0

11

In the circulating water {CW) building, both batteries had loose power

connections at the positive and negative terminals. These batteries

provide the 125 VDC control power for the CW 4160 VAC pump motors switch-

gear.

The licensee had identified these conditions but did not take

corrective action until the team questioned the safety of the loose

terminals.

After the completion of the 480 VAC walkdown inspection, the team reviewed

the PM program for motor starters in the SR 480V MCCs.

The PM requirements

for inspection and testing of MCC thermal overload devices, motor starters .

(contactors) and molded case circuit breakers (MCCBs) specified in procedure

EPL-MCC-E/Rl were deferred (cancelled) for the last two scheduled periods due

to lack of manpower.

Licensee personnel stated the deferrals were approved

by the Site Nuclear Operations Committee.

The licensee also stated that

EPL-MCC-E/Rl will be replaced with an upgraded procedure, and the associated

components will be placed in the PM program and the required PMS will not be

deferred due to lack of manpower.

During the review of VPAP-0806, Power

Circuit Breaker and Switchgear Program, the team determined that MCCBs were not

included.

The licensee stated that MCCBs will be included for the next outage.

SR MCCBs will be tested on a five year basis and NSR MCCBs on a ten year

schedule.

With the inclusion of the molded case circuit breakers, the team

considered the new Power Circuit Breaker Program would be very good *

During the observation of work activities, the team found the electrical main-

. tenance personnel were knowledgeable and performed their work in a satisfactory

manner.

The foreman verified that electricians are task qualified to perform

the assigned work.

However, the team observed the foreman and electricians

spend considerable time performing planning duties instead of working in the

field. This subject is further discussed as a issue in Sections 3.g and 3.r.

From the review of the electrical maintenance procedures, the team identified

several weaknesses.

The licensee agreed to upgrade the procedures to specify

specific tolerances, test equipment, and disallow marking "N/A" (not required)

in the functional testing procedural step when operators are not available.

In addition, the licensee agreed to include the vendors recommendations for

motor starters (contactors) in PM procedures.

The team found the electrical

maintenance procedures, for safety-related equipment, had sufficient QC hold

points which is considered a strength.

During the review of completed work orders, the team identified the following

problems:

Work instructions are not adequately detailed

PMT is not specified (See Section 3.h)

Functional testing is not adequately specified

The craft do not specify in adequate detail the work performed and

problems identified in their "Work Performed

11 write up

Craft foremen are required to provide additional planning since the work

orders are very brief

..

12

Summary of Electrical Maintenance

Strengths

Personnel are knowledgeable of work and perform their duties in a

satisfactory manner.

Work is task oriented requiring craft to be qualified for each task.

Personnel use initiative to perform work and functional testing even

when not specifically specified in work order [This refers to PMT for

switchgear sent to outside vendor for refurbishment]

Personnel have a good attitude in the performance of their duties and try

to do things correctly

Weaknesses

Craft and foremen spend too much time in shop

Foremen have added burden of doing supplemental planning

Electrical maintenance procedures need to be upgraded

The size of the staff is marginal considering additional temporary duty

assignments and training

The work backlog is excessive and not being reduced

I

PM have been cancelled due to lack of manpower

Note:

These items are weaknesses for performing electrical maintenance,

but not necessarily the fault of the Maintenance Department.

The small

planning staff and poor coordination with operations together with

lack of adherence to the schedule work plan contribute significantly to

less effective electrical maintenance *

The team inspected the training program and facilities and found it to be

quite satisfactory with one exception.* The licensee has not completed

obtaining and installing all the necessary laboratory equipment to support

classroom training as planned.

Observations

The team observed the close working relationship of the Maintenance Department

with the various system engineers.

The team also had discussions with

operations personnel who considered the system engineers as an asset to the

plant.

The team also observed that the station engineering management has a

policy of walking down a different system with each system engineer on a

weekly basis.

The team considered this type of management involvement and the

system engineer as a strength.

I .

'.

..

13

2.e. 120 VOLT AC BUS AND 125 VOLT DC SYSTEMS

Background

The function of the 120 volt AC and DC systems are to provide reliable, unin-

terruptable vital power to instrumentation and control systems and components

during 911 modes of plant operation.

Each system includes two independent

trains of distribution panels and associated cabling while two independent

station batteries for each Unit provided emergency power to the vital AC busses

through a combination of uninterruptable power supply/battery charger static

devices on loss of normal AC distribution.

Inspection

The walkdown inspection included the majority of the Unit 1 distribution

system and panels and the station batteries were also inspected.

The team observed the following maintenance activities related to the AC and

DC distribution systems, and observed selected work activities of the I&C

technicians.

WO# 88416

WO# 89582

1-PT-28.8

Cal-630

1-PT-8.1

2-PT-8.5

Search for grounds on 18 DC Bus

C6rrect indication on Emergency Boration Line

Power Range Nuclear Inst. Calibration

Rescaling of Power Range Drawer

Reactor Protection System Logic (Periodic Test)

Consequence Limiting Safeguards Logic (Hi-Hi

Train, Periodic Test)

The team also reviewed numerous work orders for accuracy and completeness as

detailed below.

Findings

System walkdowns:

Each of the station

1s four vital batteries had been recently replaced. Cells

that were in acceptable material condition from the replaced vital batteries

were utilized to make up a new

11black battery

11 for each unit.* Critical loads,

such as the main turbine generator shutdown lube oil system, were removed from

the

11station

11 battery load, and transferred to the black battery load list.

Installation of this modification was perceived as a licensee strength in that

the modification significantly improved the station battery performance due to

load reduction. Deficient conditions were noted as follows:

o

Design Change Package 8532 (U-1) was not installed in accordance with

specifications. Problems such as ungrounded cable conduit, loose

intercell connector bolts, and loose bolts in the cell platform were

noted.

These conditions indicated a weakness in the control of electrical

work practices and is discussed further in Section 3.o.

' '.

'.

14

o

Cell 52, Vital Battery 2A, was noted to have a low electrolyte level

(the top of the meniscus was below the low level line, but above the

cell plates).

o

Cell 51, Vital Battery 28, was noted to have a thermometer left sitting

in cell. The thermometer was held captive in a rubber stopper, but-the

rubber stopper was simply resting in a cell lid cavity, i.e., cell over-

pressure (from gas) would not have been directed through the cell flame

arrester.

o

Cells 4 and 56, Unit 2 black battery, had significant verdigris growth on

the terminals and intercell connectors. Cells 5 and 11 had significant

chemical deposition on the flame arresters.

During the walkdown of the 125 volt DC (vital) and 120 volt AC (vital)

distribution systems, several items of concern to the team were noted.

In general the material condition and housekeeping of low voltage distribution

panels were marginally adequate.

Poor conditions and practices noted during

the walkdown are detailed below.

0

Most distribution panels contained schedules of breaker assignments that

were in error due to informal changes, such as hand-written corrections,

strike-overs, and white-outs.

When components listed on the schedules

.were compared to the applicable drawing (e.g., D.C. distribution panel 1-2

against Loading Table Bus Distribution Panels DC 1-1 and DC 1-2, dwg

11448-FE-llAE), it was noted that several errors existed.

o

Not all wires were labeled and/or labeled correctly (e.g., Vital Bus

1-IIIA, breaker #6 feeder cable was labeled 1VBS15-B; the wiring diagram

and loading table.drawing called for the cable to be labeled 1VBS15.

Vital Bus 1-IA, breaker #3 cable feed to Process Rack 3, was not labeled

as cable 1VSB44).

Spares were frequently not labeled and/or not properly

terminated (125 volt DC Panel 1-2).

Not all breakers were labeled (e.g., breaker #13, DC panel 1- 2).

Not all distribution panels were labeled on the outside to permit

ready identification (e.g., main 125 volt DC distribution panel lB).

Breaker amperage capacity installed in the distribution panels was

frequently different than plan/drawin.g requirements. This was

considered a weakness in the licensee's configuration control program

and is discussed further in Section 3.b.

Vital AC and DC distribution panels, and 480 volt and 4160 volt

breaker test panels (fed from vital DC), were very dirty,. and

contained trash, including loose metallic material. Cable termina-

tions at feeder breakers in the vital panels were occasionally

improper, e.g., not all wires captured under breaker clamp device,

insulation cut back too far, wires splayed in individual feed cables,

e

15

and bend radius too ~harp. Material conditions of panels were

deficient, including e.g., missing knockout plugs, improper hold-down

fasteners, and detached hold-downs for panel wireways. *These*

conditions indicated a weakness in the control of electrical work

practices and is discussed further in Section 3.o.

The system engineer accompanying the craft and inspection team took prompt and

substantive corrective action to correct the discrepancies noted, including

the preparation of work requests, station deviation reports, and drawing

change requests. The effective support to the craft, detailed system

knowledge and expertise, and high motivation of the system engineers was

perceived by the team to be a ncensee strength.

Review of Work Orders revealed the following discrepancies.:

Work Order 78160, dated 13 June 1989, required, Clean & Remove Trash"

from the Safety Related (SR) 125 Volt DC Distribution Cabinet IA (Mark

01-EPD-BC-lA-1). * The work actually performed included the disconnecting

and removal of eight ventilation fans from the cabinet. The team was of the

opini~n that the scope of work in the work order should have specifically

. addressed the subject of fan removal with a specific work step authorizing the

removal; a specific step should also have verified the operability of the fans

(PMT) after electrical reconnection.

Work Order 86936, dated 20 October 1989, required, "1. Troubleshoot/Repair"

and "2. Verify Operability" of a ground indication on the SR battery bus (125

Volt DC-Distribution Cabinet 28, Mark #02-EPD-B-2B).

Work actually performed

indicated that, "Found under voltage relay J Box half full of water near

lighting PNL 2Tl.

Need to replace relay and wire lugs. 10/28/89." A further

entry stated, "Replaced relay and lugs. Verified operability and returned to

service. There is still no ground on the DC bus at this time. 10/28/89"

The undervoltage relay replaced was not identified in the work order.

Purchase order or requisition documentation for the relay was not

included with the work order.

Since the relay did not appear to correct the ground, replacement of the

relay was clearly outside the scope of the work order. A work order

revision was not issued to change the.scope of work.

The Mark# for the relay was not listed in the work order, thus no

effective material history was generated as a result of this maintenance

activity.

No indication of what was accomplished to "Verify operability" was

included in the work order, thus the adequacy of the "PMT" was

questionable.

~----

16

No root cause evaluation was indicated as performed to determine why a

relay J Box was half filled with water, or if the fundamental problem

(water source) had been corrected to prevent recurrence of the failure.

Work Order 62675, dated 16 March 1988, required the replacement of motor

bearings on #2 Rod Control MG Set, Mark# 02-EPD-M-MG-2.

The work instructions of the work order were very sketchy, but did

reference the procedure EMP-C-EPL-117 regarding corrective maintenance

on the MG set. This procedure was unused during the work activity for

reasons unknown and not listed in the work order. A superseding procedure

EMP-C-EPCR-08, dtd. 4 Dec. 1986, Rod Drive Synchronous Alternators,

accompanied the work order and was the document used for the work

performed without formal revision of the original work order.

The scope of work in the work order was to

11replace motor bearings". The

actual work accomplished included motor and alternator bearing

replacements, with no revision to the work order.

M&TE used during the repair procedure were not properly listed on the

work order or in the procedure.

Although pre-work vibration analysis accomplished under WO #62597

indicated that the motor to generator coupling "had badly damaged teeth

and coupling grease had turned to powder.", WO# 62675 did riot reflect

that the motor coupling had been replaced_ or repaired. Steps at

paragraph 5.17 of procedure EMP-C-EPCR-08 simply installed the shaft

coupling with no indication of what was actually accomplished.

WO #62597 (vibration analysis) indicated the machine as safety related

(SR), while WO #62675 indicated the machine as non safety related (NSR).

The machine was NSR.

Work Order #66468, dated 16 June 1988, was prepared to

111. Troubleshoot and

Repair" and

112. Verify operability." of a grounded condition on both DC battery

busses when containment DC lighting leads were landed.

The leads had been

lifted and tagged when the condition occurred.

The work order Mark# was listed as Ol-EPD-BKR-14.

The actual grounds

were detected in the 1-ERCl Panel, breakers 6 and 8 feeds.

The Mark#

was not changed.

The scope of work increased significantly after troubleshooting, but no

modifications were made to the work order, i.e.,

8 breakers were replaced in (presumably) the 1-ERCl Panel,

2 light fixtures were replaced in the Blighting loop, 3

fixtures in the C loop, and an unknown number in the A loop,

17

lights in the "RCP cubes" were also repaired,

A work order "Repair or Replacement Follower" specifically required,

"Verify any leads lifted to isolate ground are landed with proper polarity

in accordance with applicable design drawings.

Functionally test circuit

after landing."

No evidence was included with the work package that any

of the lifted leads had been properly landed in accordance with the work

iflstruction.

Work Order #74377, dtd 8 December 1988, required "1. Troubleshoot & Rep~ir."

(and

112. Verify ground is cleared.i

1

) of Ckt. #2 in the DC Distribution Panel 1

- 2, Mark# Ol-EPD-BKR-43, the supply to SI "8

11 accumulator solenoid operated

valves {S0Vs) *. The electricians assigned to perform the task prepared their

own work instructions, including the requirement for tagout, notwithstanding

the inadequate work order received from planning.

The "work instructions"

prepared by the electricians comprehensively.addressed all steps expected by

the team; the instructions also referenced the subsequent Work Request#

submitted after the craft identified the problem SOV.

The work order was stamped, "Determined to be minor maintenance per

SUADM-M-16

11 , thus a "procedure" was not required. Contrary to this work

order stamp, work order preparation instructions specifically listed

maintenance requiring equipment tagouts as not qualifying as "minor

maintenance."

Thus this work order should not have been stamped "minor

maintenance."

The drawings listed by planning for performing the work were "FE-lG,

-lOA, and llAE".

The craft had to use drawing "FE3BK" to perform the

work.

The work instructions prepared by the craft that recorded the work

accomplished were perceived by the team as a job well done.

Work Order #86476, dtd 7 October 198Q, required the electrical craft to,

11 1.

Troubleshoot and repair breaker.", Reactor'Trip Breaker "B" Normal, Mark

01-RP-BKR-BNORM.

The problem was reported as the breaker not appearing to be

closed while the I&C craft were performing periodic reactor protection system

tests (PT - 8.1).

"PROCEDURE NOT REQUIRED

11 was conspicuously stamped on the work order,

notwithstanding the work scope authdrizing "troubleshoot" and "repair" of

the breaker.

The work accomplished reflected that the breaker was racked out, manually

closed, continuity checked satisfactory on all three phases (no work

performed), and the breaker was racked in~

PT - 8.1 was continued and

presumably performed without further difficulty.

No further evaluation

or deviation reporting was reflected by the work order *

18

The work was performed on 8 October 1989.

The

11Equipment Returned to

Service

11 date was listed as 6 December 1989, after the plant had returned

to full power.

This equipment should have been

11feturned to service

11

before power operations.

Based upon the above types of examples noted by the team in its review of

closed work orders, the team concluded that work order instructions prepared

during the planning phase of maintenance activities lacked adequate detail.

(Also see section 3.g.)

.. '-,

~,~

Review of maintenance in progress - I&C Group

"--

..

The Instrumentation and Control (I&C) group, re-organized into the Maintenance

Department in late 1989, was responsible for the planning, scheduling, calibra-

tion and maintenance of plant installed, primary and secondary, instrumentation

and controls, and the measuring and test equipment (M&TE) program.

Organiza~

tional and assignment of responsibilities procedures such as SUADM-ADM-47, dtd

18 September 1989, Operation of the Instrument Department, had not been revised

to reflect the re-organization. There was no visible evidence of effective

integration of the I&C group with the Maintenance Department since the group

was fundamentally doing business as before, e.g., work order planning was

performed at the craft level.

Staffing of the I&C group was noted to be a

11 numbers

11 problem, i.e., thirty

five technicians were authorized (no break-down between trainees and

technicians), but only twenty were actually staffed. Of a large group of

about fifteen contractors retained in late 1989, only two remained due to

factors related to the permissible overtime they could work.

Only thirteen

of twenty staff were

11technician

11 grade, the balance were

11trainees

11 in the

apprentice program.

As a consequence, I&C supervision expressed concern that

although they were currently maintaining reasonable control over I&C backlog,

any further increase in maintenance activities could not be adequately

supported by staff on-hand.

I&C Supervisors stated that their attempts at hiring had been halting at best;

resumes received through the personnel department were frequently aged to the

point that calls to prospective hirees found them unavailable. Further, from

point of interview to an offer of employment (while background checks and

clearances were obtained) was also a long enough period to result in the same

answer on calling potential hirees. Staffing was noted to be potential problem

in several areas, e.g., not 100% shift coverage, one person operating entire

calibration program, and ineffective work order planning.

Turnover, to the credit of the licensee, was noted to be a relatively low

number of approximately two per year. Supervision was not advised of the

loss mechanism, although it was understood that termination interviews were

conducted.

The team noted that the I&C group had the lowest level of mainte-

nance backlog of approximately 500 work orders of the three major craft *

' ..

19

The team was concerned about the level of planning performed in preparation of

work orders.

As discussed in Section 3.g., planning for the I&C group was

performed at the craft level.

Work Orders, in general, included work instruc-

tions of the form, Investigate and repair". For example, WO #89582, dated

5 March 1990, was written to "Correct Indication" on the emergency boration

line fl ow i ndi ca tor (Mark #Ol-CH-FT-1110). The job steps stated, "Invest/

Correct Indication", although the actual steps required were:

determining

the installed flow converter was faulty; obtaining and bench calibrating a new

converter; removing installed device (rad area) and installing new calibrated

device; and setting zero and span adjustments in place with a calibration

device.

The craft accomplished the preparation of "work instructions" to

support the work order by revising an existing calibration procedure

1-CAL-311, Boric Acid Bypass Flow (Emergency Borate Flow) F-1-110, dated

3 October 1988.

The revision of the existing procedure was accomplished in

accordance with appropriate procedures. This methodology was noted to be the

normal method of preparing I&C work order instructions on several occasions

during the inspection. It was also noted to be extremely cumbersome because

calibration procedures and periodic test procedures rarely fit the maintenance

activities. This methodology also resulted in work orders that had no overall

coordinated sequence of steps. Interviews with the craft indicated that they

thought they could not extract portions of approved procedures or technical

manuals to prepare supplemental-work instructions.

SUADM-ADM-47, dtd 18

September 1989, Operation of the Instrument Department, supported th.is

perception at paragraph 4.3.8.1 concerning preparation of work orders,

"Identify work procedures that are required.

Pay particular attention and

identify if the procedure will need to be deviated and/or pre-approval is

required." Based upon the above observations, it was the team's opinion that

program and implementation improvement was required in the area of preparation

of I&C supplemental instructions for work orders.

Clearance Program and Implementation Improvements Required:

The team reviewed the clearance log for both units to determine if adequate

clearances were being set to maintain equipment control and provide safety to

the maintenance technicians as well as the equipment itself.

The team noted that the 480 volt feeder breaker for the U-1 w*ater Chi 11 er for

the Air Conditioning Unit (switch gear 1B2-12B) was red tagged.with two red

tags (red tag #1246755 was associated with clearance #62476; red tag #1246062

was associated with clearance 2059274).

The "Remarks" block of the red tag was

annotated "OFF/OFF" on tag #1246755, and "OFF/ON" on tag #1246062.

The team

learned that the two positions referred to the proper tagged position listed

first, and the restoration position listed second.

The listing of these posi-

tions on the tags has been implemented by persons responsible for hanging tags,

but was not required nor described by SUADM-0-13 (dtd 1 Feb. 90), Operations

Department - Operation, Maintenance, and Tagging.

Only the Tagging Record

Form (#888.6A/7A) required the tagged and*restoration positions to be listed

in accordance with SUADM-0-13.

It was the team's opinion that the listing

of two different restoration positions for the same equipment was an implemen-

tation weakness in the clearance methodology.

.

'*

20

The team noted that a large number (91) of station deviation reports (DRs) had

been generated during 1989 related to tagged out components, including:

o

34 DRs for improperly tagged components

o

12 DRs for missing (abused) tags

o

15 DRs involved breakers out of position

o

22 DRs involved valves found out of position

o

9 DRs involved wrong components tagged

The team perceived this as a very large number that potentially affected

personnel safety.

Interviews with Operations personnel indicated that a large

portion of the problems were noted during the termination of outages, where

devices had been mis-positioned for reasons unknown.

The team noted that

paragraph 5.11.3, SUADM-0-13 specifically waived performance of the quarterly

tagout audit (field check of tags) during outages, and that only an Administra-

tive Review of the log was performed prior to startup. Manpower was stated as

one of the limiting factors in not performing field audits during outages,

even though there was a high probability for error during the intensely active

maintenance period. It was the team's opinion, based on the large number of

tagout discrepancies, especially associated with outages, that the field

audits should not be waived to provide reasonable assurance that components.

tagged for long periods of time would be periodically checked in their proper

position.

The team reviewed the tagout log to determine if program requirements were

being adhered to.

Many very old tagouts were in an active status, with no

apparent action currently underway to correct the condition.

(See section

3.m.) Unit #1 tagout #59551 was an Operations tagout on the chilled water (CD)

system. It was noted that tag number 863558 was listed as partially cleared on

a Tagout Partial Clearance form Figure 11; the Tagging Record Block 11 (Tag

Removed) for the tag was not properly signed.

Tag number 801005 was listed as

removed on the Tagging Record, Block 11; the tag was not authorized for

partial clearance on the appropriate Tagout Partial Clearance form.

Procedural Weaknesses Identified During Periodic Testing:

The team observed the performance of Periodic Test 1-PT-8.1 (Rev. 1), Reactor

Protection System Logic (For Normal Operations), on U-1.

The test ensured the

continued proper operability of the reactor trip portion of the reactor

protection system. Several strengths of the I&C technicians were noted during

the testing, including a good pre-job briefing that addressed previous

"lessons learned", good procedural adherence, good command and control of the

evolution by the team leader, and good communications between the three, two

man groups of technicians performing the test *

.

' .

21

Several procedural inadequacies were also noted as follows.

The craft

technicians were confused about the extent of completion of an Engineering

Work Request (EWR) affecting Extraction Steam Motor Operated Valves (MOVs),

and therefore convinced the Shift Supervisor (S/S) to unnecessarily

de-energize the breaker to an Extraction Steam valve (MDV ES-lOOA) as a part

of establishing initial conditions. Although the procedure did not address

the subject, the lead I&C technician, based on his understanding of the EWR,

requested the S/S to de-energize the.MDV to prevent an inadvertent actuation

of the valve during the logic testing.

In fact, the EWR had only been

partially completed on U-2, and then was canceled; the EWR had no

applicability to Unit 1 testing. It was the team's opinion that the procedure

should be unit-specific to the extent that a subject affecting only one unit

should be addressed in each unit procedure.

,

Precautions and Limitations paragraph 4.2 stated, "Bypass breakers will ti~ly

be closed long enough to perform required testing on associated trains.

11

The

team perceived this wording as vague when compared to the requirements of.

Technical Specifications (TS), Table 3.7-1, Action Statement 11. which stated,

" ****. one channel may be bypassed for up to 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months for surveillance testing

per Specification 4.1 ****

11

It was the team's opinion that the TS

requirements should be included in the Precautions and Limitations, and that

clock times should be entered at paragraphs 5.15 and 5.68 (time breaker

closed/opened) to assure specific awareness and compliance with TS

requirements.

The technicians demonstrated such awareness.

During performance of the procedure (at Step 5.83), a relay failure occurred

(promptly detected by the technicians by smell).

No provision had been .

provided in the procedure (or any higher tier procedure) for "backing out

11

of the sequence safely, thus a one-time change was executed to back out and

restore the plant to safe conditions. It was during this failure that the use

of a

11 For Reference Only

11 drawing occurred as discussed below. The fail~re

occurred relatively early in performance of the procedure, thus adequate time

was available to permit the time- consuming procedural change, and not violate

the two-hour requirement addressed above. It was the team's opinion that

instrument test procedures could be enhanced by consideration of the situation

of plant restoration if a system failure occurred, especially for time

sensitive procedures such as this.

A "For Reference Only" drawing (as opposed to a controlled drawing), Drawing

113E244, Reactor Protection System, was used by the technicians and their

supervisor for troubleshooting and procedure revision preparation, and

operational considerations by the Shift Supervisor to establish safe plant

conditions, without assuring the drawing was correct by comparison to a

controlled drawing.

The drawing was a logic wiring diagram for the reactor

protection circuitry. The team perceived the use of.uncontrolled drawings in

this manner as an implementation weakness since plant modifications could have

been made that would not have necessarily been recorded on the uncontrolled

drawing.

The use of the "For Reference Only

11 drawing was in violation of the

.

' .

22

licensee's procedure for use of station drawings, SUAD~-ADM-11, dated 29 Nov.

1989, Station Drawing Revision and Distribution, which stated at paragraph

4.7, "Individuals using drawings or aperture cards shall be responsible for

ensuring that the item used is the latest revision." This failure to follow

procedure is an example of violation 50~280,281/90-07-01.

Revision 1-16-90 to 1-PT-8.1 was a permanent change executed on 7 February

1990 (added a cautionary note to the procedure that had been omitted at the

last procedure update).

The team noted that the included 10 CFR 50.59

Screening Checklist on the Procedure Action Request (Form No. 730682) was not

completed correctly; i.e., the form required the listing of Sections of the

UFSAR that had been reviewed in performing the 50.59 evaluation, but none were

listed. The team perceived this failure to follow procedure as an

implementation weakness in executing procedure changes.

2.f EMERGENCY DIESEL GENERATOR (EDG) SYSTEM

Background

The purpose of the emergency diesel generator (EDG) system is to provide a

dependable source of onsite power capable of automatically starting and

supplying electrical power to loads necessary for safe shut down and

maintenance of safe shutdown conditions under all design basis conditions.

Major equipment and auxiliary systems included in the system walkdown were

three 20 cylinder diesel engines, generators, starting air systems, engine

cooling, fuel system, lubricating oil, and governors.

Inspection

The walkdown inspection included the majority of the above systems.

In

addition, the team witnessed monthly performance tests of an EOG, an engine-

driven compressor in the starting air system, the replacement of a starting

air system compressor, and reviewed closed and open work orders.* The team

also reviewed selected PM procedures for EDG components against vendor manual

requirements for EDG system equipment.

Findings

At the time of the inspection, the team found that the licensee did not have

documentation verifying that the EOG fuel oil transfer lines {approximately two

inches OD) were seismically qualified. These lines are installed between the

day tanks and the EDGs.

In addition, the teamfound one of the in-line flex

hoses replaced with a rigid section of pipe.

The licensee could neither

adequately determine the reason the flex hose was replaced nor provide

..

>

'

,

23

documentation which controlled or recorded the work involved. These flex hoses

serve to isolate the day tanks and associated piping from vibration while

the EOG is running.

During the inspection, the licensee performed a seismic

evaluation of the fuel lines and concluded that the existing configuration

was acceptable. However, the licensee plans to add supports, and replace the

rigid section of pipe with braided hose.

Need for further NRC review of the

calculations involved (SEQ-1517) will be identified as unresolved item

50-280/90-07-02 "EOG Day Tank Fuel Transfer Line Analysis."

Material condition of equipment in the EOG spaces was, in general good, with

exceptions noted below.

Several problems associated with a work order are

also listed below.

The team found a significant air leak at an air fitting on the pressure switch

which controls the motor-driven air compressor of the engine and motor-driven

compressor set in the air start system for EOG 3. A WO was written to correct

the problem.

The team observed work activity on WO 93103 for replacement of an air

compressor (one of six being replaced).* The compressors were being replaced*

because of moisture-induced deterioration of valves, and unavailability of

replacement valves. Since there were no air driers or coalescing filters at

the discharge of the air compressors, the new compressors are subject to the

same damage as those which were replaced. Details of this issue are discussed

in paragraphs 3.c and 3.f

In addition, on WO 93103, the team found that the vendor technical manual for

the compressor specified light oil or anti-seize compound for thread lubricant,

with 20 percent reduction in applied torque if anti-seize is used.

The work

package, however, only specified the threads be

11lubricated". The craft

applied anti-seize to the threads without questioning what type of lubricant

to use, and proceeded to torque the fasteners to the values indicated on the

work order. The specified torque values were based on the use of anti-seize,

however, had the craft used light oil and the torque values furnished with the

work order the fasteners would have been undertorqued.

The team noted that a valve which admits air to two air starting motors on one

of two trains in the air start system was not functioning properly.

The

licensee indicated the root cause was due to poor air quality. This issue is

discussed in paragraph 3.c.*

Local indications and remote sensors for vital EOG system conditions, such as

engine temperature, and starting air system pressure were not calibrated. The

remote sensors provide input which triggers the

11Diesel Trouble

11 control room

annunciators.

The result is a 11 sources of EOG system condition may not be

accurate. This issue is discussed in paragraph 3.1 *

..

24

2.g HEATING VENTILATING AND AIR CONDITIONING (HVAC) SYSTEM

The functions of the HVAC system is to provide for contamination control by

ensuring that air is not recirculated in areas of potential contamination, and

provide adequate seasonal ventilation and/or temperature control in occupied

machinery spaces including the control and relay room area. The system

consists of ductwork, fans, filters, dampers and associated controls.

Inspection

The team performed a walkdown inspection of the subject system in such areas

as the auxiliary building, mechanical equipment rooms -1, -2 and -3, control

room, and emergency switch gear room.

A selected sample of (20) completed work orders and the following Engineering

Work Requests (EWRs) were reviewed ..

EWR

Title 89-540

90-078 89-687

87-170 89-336

Evaluate VS Pressure Switch Calibration, Units 1 and 2

Evaluate VS MGR.Chiller Existing Capabilities (E4A, B, C)

Surry 1 and 2

Evaluate VS Fan Duct (l-VS-F-2, 12A and -128)

Repair of Containment Recirculation Fan (2-VS-F-lA)

Evaluate VS Repairs for Startup, Units 1 and 2

The team observed maintenance work in progress on three nonsafety-related

chillers listed below:

Component

01-CD-REF-lA

02-CD-REF-1

01-CD-REF-18

Findings

Work Order 089694

089693

089691

Activity Description

Weld Repair Tube Sheet and Gasket

Seating Surface and Recoax

Overhaul Compressor and Rod-Out Tubes

Overhaul Compreisor and Rod-Out Tubes

The above inspections, observationsl interviews and record/document reviews

revealed the following:

Rather than focussing on permanent repairs on the HVAC system, the licensee is

performing temporary fixes which include making extensive use of Foster's Duct

Sealant and red duct tape to plug holes and leaks in the ductwork throughout

the system.

The team observed an unusually large number of work request

tags on the vast majority of rotating components i.e., fan motors, sheaves,

dampers, actuators, metering devices, fan belts, etc. By-and-large, the

reason for these tags was for components requiring corrective maintenance or

replacement *

.

,

' .

>

25

Under the current licensee program, work requests are evaluated, and based

on their merit they are either rejected or accepted. If accepted, they are

assigned a work order number and a priority number.

According to figures

provided by the licensee there are pres~ntly approximately 220 outstanding job

orders for the VS waiting disposition.

In addition, there are 15 others, also

outstanding, waiting on parts non-traceable to equipment manufacturers.

All of the above work orders have been assigned a priority ranging from number

1 through 3 as required by SUADM-M-11, Attachment 2.

By this procedure, any

work order given a priority range of 1 through 3 must have the work begin from

48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months up to a maximum of 4 weeks, for priority 3.

Because of the this observation, the team has determined that this procedural

requirement is not being enforced since the scheduled start date on all of the

above and many other WRs has passed without work initiation. This failure to

comply by established procedural requirements is considered a programmatic

. weakness.

This matter is discussed further in Section 3.m.

2.h

MISCELLANEOUS MAINTENANCE ACTIVITIES

Inspection

The team observed the below indicated maintenance work activities related to

other systems:

Maintenance Work

Work Document ID

Description of Activity Observed

WO 067521 &

Alignment of Condensate Polishing (CP) Pump

EWR 86448

02-CP-P-15C

WO 93028

Alignment of CP Pump 02-CP-P-430

JN 087715

Preventative Maintenance

WO - Work Order

EWR - Engineering Work Request

JN - Job Number

Findings

The above inspection/observations revealed the following:

During observation of pump 02-0P-P-430 alignment using WO 93028, the team

noted that the only instruction on the WO was "Align Pump" with no reference

to any procedure, vendor manual, or other document.

Maintenance personnel

stated that this was in the "skill of the craft" and no procedures were *

needed.

Further review revealed that a detailed procedure was available for

pump alignment using the dial indicator method.

The procedure had numerous

sign-off steps, acceptance criteria, and the job had been started using the

procedure.

However, after the job started, the decision was made by the craft

to use the optical alignment method.

Since no procedure was available for

this equipment, the vendor manual for the equipment was used.

When q~estioned

.

'*

>

26 *

by the. inspector, the foreman stated that based on training, they knew what

the acceptance criteria were and therefore did not need a procedure.* The team

questioned the Mechanical Maintenance planner and supervisor responsible for

the WO and the response was that the WO probably should have referenced the

vendor manual for the alignment.

However, they further indicated that they

were busy trying to keep up with the more important jobs and did not have

adequate resources to do detailed planning of non-safety-related jobs and had

to depend on the "skill of the craft." This is an example of weaknesses in

planning and adequate resources discussed further in sections 3.f. and 3.g.

In general, maintenance mechanics, technicians, foremen and supervision

appeared to be well qualified, knowledgeable, and work was performed in a

professional manner.

In review of activities related to erosion/corrosion heater drain pipe

failure, the team found that the licensee has a well defined program for

inspection 'of pipe for erosion/corrosion thinning. The program is defined in

Engineering Standard SDT-GN-0033 which was in response to NRC Generic Letter 89-08 and was implemented January 1, 1990.

Prior to that date, procedure

SUADM-M-33 defined the program.

The pipe section that failed (at the

discharge from flow control throttling valve LCV-122B) was not in the program,

however, the elbow next to the pipe section was in the program.

There had

been no reason to suspect the pipe section prior to the failure. Although the

similar pipe section in Unit 2 had been replaced with CR-MO steel, the

replacement was a convenience replacement while replacing the elbow and not

because of excessive thinning.

The licensee

1s actions and planned actions

after the pipe failure appeared to be conservative and aggressive. Similar

piping at the discharge throttling valves for the other Unit 1 train and both

trains on Unit 1 were inspected for thinning.

The other Unit 1 train had

significant thinning and required pipe replacement.

Piping in both Unit 2

trains was acceptable.

The licensee was compiling a list of all throttling

valve configurations in the systems covered by the program.

This list was to

be examined for other similar piping that should be inspected.

The team also reviewed several completed work orders (WOs) related to work

performed on Motor Operated Valves (MOVs), replacement of electric solenoid

operated valves (SOVs) to extend the longevity of the qualified lives, and

replacement of components in NAMCO type limit switches, also to extend the

longevity of the qualified lives. Problems related to incomplete summary

descriptions on the cover sheet, partially complete information on the

model/serial numbers, and using the wrong illustration were identified. These

findings are discussed further in paragraph 3s.

The team noted an example of failure to follow procedures associated with

initallation of radiation monitor RM-SW-107 which involved failure to complete

sign-off steps in sequence; i.e., precondition step 2.4 was not signed off

even though work had proceeded to installation step 4.18. This constituted a

nonconformance to upper tier procedure STD-GN-0001 which mandates that sign-off

steps be completed in sequence barring specific notes to the contrary. The

licensee initiated DR Sl-90-321 to begin corrective action.

.

.

,.,

"

27

2.i

HEALTH PHYSICS

The team determined that the HP group was included in the planning and

scheduling of maintenance activities through representation at all daily

planning and briefing meetings. Interviews with maintenance and HP personnel

indicated that good lines of communication existed between the groups.

In addition to the licensee's general employee training for radiological

protection, the licensee had provided most maintenance workers advanced

radiation worker training. The training*provides maintenance workers with

additional instruction and practice in radiological survey and work practices.

The graduates are allowed to perform limited radiological monitoring functions

at their work sites as directed by health physics personnel. The majority of

maintenance workers interviewed believed that the training had helped thew

understand how radiological protection activities could be integrated into the

maintenance activities. The advanced radiation worker training had been

provided to about 60 percent of the maintenance staff and the licensee planned

to provide the training to all maintenance personnel.

Maintenance workers reported that the licensee's on-going decontamination

program was very beneficial to the maintenance process. Workers reported that

ready access to non-contaminated areas, that were contaminated in previous

years, had improved maintenance efficiency. Workers reported that there was

more effort by maintenance personnel to keep systems .and components from

leaking contaminated fluids and that housekeeping during and following

maintenance activities was an important element of their job. The licensee's

area of contaminated floor spaces has steadily declined in recent years *. The

licensee's floor.area contaminated in 1989 declined from 20,500 square feet

(ft2) in January to 14,500 ft2 in December. The licensee's goal for 1990 was

to reduce the area contaminated to 11,500 by December, 1990.

The,licensee's collective personnel exposures were 792 and 420 person~rem

per unit in 1988 and 1989. The licerisee's collective personnel exposure goal

for 1990 was 303 person-rem per unit. In interviews with maintenance workers

the team determined that worker awareness of ALARA goals and objectives were

high. The interviewed maintenance workers could adequately describe methods

for keeping collective radiation exposures ALARA and knew their lifetime,

quarterly, and annual radiological exposures. Workers reported that ALARA

activities were strongly supported by management. The licensee strengthened

its ALARA program during 1989 providing additional resources and management

attention to implement various source term reductions and ALARA program

initiatives. To increase facility staff involv~ment in the ALARA program,

various departments, including maintenance, were required to develop and.

implement department action plans to minimize personnel dose. Worker awareness

of the ALARA program was a program strength *

'

28

3.

ISSUES

3.a. SUITABILITY ANALYSIS FOR REPLACEMENTS

American Society For Mechanical Engineers Boiler and Pressure Vessel (ASME

B&PV) Code Section XI, Paragraph IWA 7220 requires the Owner to conduct an

evaluation of the suitability of replacements, prior to authorizing the

installation of those replacements. This requirement is implemented, by the

licensee in Procedure SUADM-M-08, dated February 27,1990,

11ASME Section XI

Repairs and Replacement Programs"

The lic~nsee informed the team that

SUADM-M-08 is applicable to pressure retaining components and their supports

only (items covered by ASME B & PV Code Subsections IWB, IWC, IWD, and IWF).

The licensee's program does not address the IWA 7220 suitability analysis

requirements for non pressure retaining replacement parts, such as bearings,

bushings, springs, stems, disks and shafts (items covered by ASME B & PV Code

Subsections IWP and IWV).

The lic~nsee was unable to provide a single example

where there was objective quality evidence attesting to the fact that a

responsible individual had made a conscious decision that replacements

11 in

kind", of non pressure retaining parts for Section XI components, were

suitable for the intended service.

The team concluded that a weakness exists in the licensee's program related

to the implementation of the suitability analysis requirements of ASME B & PV

Code Paragraph IWA 7220 for non pressure retaining components and for the

documentation of suitability analysis for all first time replacements "in

kind.

11

3.b. DOCUMENT CONTROL/CONFIGURATION MANAGEMENT _SYSTEM FOR MAINTENANCE

Drawing Revisions Not Issued Following Plant Modifications

During walkdowns of the 120 volt AC and DC vital and semi-vital distribution

system, the team noted several occasions where breaker amperage capacity listed

on drawings differed from the installed breakers as follows.

o

Drawings 11448-FE-llAE (Rev. 3), Loading Table Bus Dist. Panels DC 1-1 &

DC 1-2, and 11448-FE-lBJ, Wiring Details, Misc CKTS, Sheet 2, reflected

all breakers as being 15 ampere capacity. The installed breakers were 20 *

_ amp.

o

Drawing 11448-FE-llAA (Rev. 7), Loading Table, Vital Bus Distribution,

Panels 1-1, 1-III, reflected feeder breakers 16 and 19 to be 15 amp and 30

amp respectively. The actual breaker sizes installed were 20 and 15 amp

respectively.

o

Drawing 11448-FE-llAC (Rev. 6), Loading Table Semi-Vital Bus Distribution

Panel lSVBl, reflected breakers 32 and 33 as 20 amp and 30 amp respec-

tively. The installed breakers were 50 amp and 20 amp respectively.

..

29

Based on the large number of the above types of errors noted during the

walkdowns, the licensee initiated a check of all 120 volt AC and DC vital and

semi-vital panel breaker installations against all affected drawings, and

noted numerous errors (randomly dispersed within 16 additional drawings) of

similar nature i.e. labeling and fuse sizes in other panels and their drawings.

During the time of the inspection, it could not be determined why the errors

occurred.

The majority of incorrect drawings were considered to have occurred

following plant modifications; e.g., occurred as a result of performing a work

order to correct a tripping breaker.

In any case, drawings were not properly

updated in accordance with plant procedure (as it now exists) SUADM-ADM-11,

dtd. 29 November 1989, Station Drawing Revision and Distribution, which

required drawing revisions be issued when plant modifications were made or

when as- built conditions different than drawings were discovered. The

licensee verified that the proper sized breakers and cables were installed in

accordance with modification requirements. A total of eleven drawings required

~

correction. The licensee advised the inspection team that a program existed

for drawing changes upon completion of field design changes; based on the

examples noted by the team, it was concluded that the program for control of

drawing changes had not been effectively implemented.

This failure to follow

procedure is an example of violation 50-280,281/90-07-0l.

Vendor Supplied Information Not Incorporated in Plant Documents in Accordance

with Procedure

The team was concerned that vendor supplied information that necessitated

changes to plant procedures and documents was not being properly processed and

incorporated into station requirements.

The team reviewed a specific example

of vendor supplied information to determine if the licensee's process for

handling of vendor information was being properly incorporated.

Limitorque Corporation published Maintenance Update 89 - 1 in December 1989.

This important bulletin addressed several maintenance topics applicable to

Limitorque actuators, such as:

0

actuator pinion gear fit-up, orientation, and location

gear to shaft key material, fit-up, and retention (staking)

set screw spot drilling and retention (lockwiring/staking).

Although this bulletin had been received by both corporate and site personnel

responsible for the Motor Operated Valve (MDV) maintenance program, the team

noted that:

o

MOV maintenance procedures had not been updated to reflect the

requirements of the bulletin,

o

the TSC library, controlled document Vendor File and Vendor Manual

did not contain the bulletin or any reference to its contents, and

o

the licensee's Commitment Tracking System (CTS) did not contain

reference to the contents of the bulletin .

30

The team was able to.determine that a memorandum had been prepared by the

station MDV Coordinator that called the attention of Maintenance Engineering

to the contents of the bulletin, but no further action had been taken.

The

team determined that neither corporate nor site personnel directly associated

with the MOV program were aware of their responsibilities for processing

vendor supplied information such as the Limitorque Bulletin.

Station requirements for processing vendor supplied information were found in

SUADM-ADM-31, Vendor Interface/ Control of Vendor Documents, dated 5 Dec.

1985, at paragraph 8.0, Vendor Supplied Information, which stated,

8.1.1

11All technical correspondence from any vendor *.* shall be reviewed

by the appropriate department.

Each Department Head is responsible for

insuring that any of this correspondence received in his department is

forwarded promptly to the Supervisor Records Management.

8.1.2 If it is determined that corrective actions are necessary the item

shall be placed on the Commitment Tracking System.

8.1.3 If the completed item causes a change to the Vendor Manual or to the

Vendor File, the attached form

11Vendor

1s Manual/File Revision

11 (Attachment 3)

shall be completed by the Licensing Coordinator and forwarded to the TSC

Library for a controlled distribution and filing .***

11

The team noted that

Attachment 3 was the checklist that would cause, among other items, required

revisions to station procedures to be implemented.

Based upon the above example, the team concluded that the program for control

and incorporation of vendor supplied information was not being effectively

implemented in accordance with required station procedures. This failure to

follow procedure is an example of violation 50-280,281/90-07-0l and is

discussed further in section 3.k.

3.c. AIR SYSTEMS

The team found numerous examples of end-use devices which vibrate during

normal operation and are connected to stationary IA root valves by lengths of

small-diameter copper tubing. This was considered significant in light of

much industry experience with trips and transients due to vibration-induced

air line failures, including a trip at Surry

1s sister plant, North Anna

(February, 1989) in which an IA line on a feedwater regulating valve failed

due to vibration; a steam generator tube plug failure and tube rupture were

associated with this large transient.

Compressed air is supplied to containment by four (two per unit) rotary water

seal ring compressors which take a suction on the containment (typically

99 percent relative humidity at 118 degrees F.) then discharge into refrigera-

tion air driers. These air driers are not capable of attaining dew points of

less than 35 degrees, even under optimum conditions. The licensee

1s response

of February, 1989, to NRC GL-88-14 committed to conformance with ISA S7.3 which

states, in part, that at no time shall IA dewpoint exceed 35 degrees.

31

An annual PM procedure, IA-C-M/A2, written in 1985, covers the containment IA

air compressors and discharge air filters, inside the refrigeration air

driers at the discharge of the air compressors.

The team found that the most

recent performance of this PM procedure for the refrigeration air driers was

in mid-1989, by a contract maintenance firm.

At that time, the contractor

indicated that two of the discharge filters were so rotted and corroded they

could not be left in place; discharge air from these air driers to containment

IA loads was not being filtered at the time of the inspection. The remaining

two filters were dirty, however, they were left in place because spare filters

were not on hand.

This was still the case at the time of the inspection, per

cognizant licensee personnel. Further, the team found that the only time

prior to 1989, this procedure was ever performed was January, 1987 during

which time, th~ steps for inspection of the discharg~ filters was checked off

as "Not Applicable". There are no other records, per the licensee, that these

filters were ever changed since installation of the air driers, or about 7

years.

In addition, the team reviewed station deviation reports (DRs)

relating to water found in end-use devices and high dewpoints in the

containment IA system.

These DRs document ~ater squirting from solenoid

operated valves, flow gauges full of water, and dewpoint readings greater than

60 degrees F, among other things.

Interviews with statinn personnel also

indicated that air regulators at end-use devices were typically full of water

when blown down during outage PM work.

The lack of attention and timely

commitment of resources by management towards needed upgrades in the contain-

ment IA system, despite documentation of numerous problems, as well as the

general material condition of the containment IA system was considered a

weakness in the licensee's maintenance program.

The team found that check valves which ensure operability of backup accumula-

tors for air-operated valves required for safe-shutdown were not in the

licensee's inservice test (1ST) program.

These check valves were added to

the program late in 1989, and are scheduled for testing, however, the failure

to recognize the requirement for inclusion of these valves _in the IST program

was considered a weakness by the team.

In the EDG air start system, the licensee experienced chronic problems with

leaking check valves and compressors which have required frequent in-head

valve replacement; these valves are no longer available for this vintage of

compressor. Recently, the licensee replaced all six discharge check valves,

and has been in the process of replacing all six compressors under various

engineering work requests (EWRs).

These problems were primarily due to poor

air quality, per licensee correspondence, and cognizant licensee personnel.

Degradation of the air compressors and the check valves was due to accumulation

of water on top of the check valves, and *passage of the water into the air

compressors. The adequacy of the EDG air start system was also addressed in a

type 2 request for engineering and construction assistance, dated June 30,

1989, in which station engineering personnel identified two concerns which

were: the absence of a program for monitoring or controlling EDG starting air

quality, and the hi9h likelihood _that the air start receivers (18 total, 20

cubic feet per tank) are full of rust and scale from years of wet service.

,.

32

The team then witnessed the routine blowdown of the air start system for

EOG 1, and a significant quantity of water was discharged. Also note, that

the licensee recently removed over 20 pounds of rust from inside the service

air receiver in the turbine building. Air quality in the EOG air start system

can also affect the performance of the solenoid operated valves which admit

air to the air starting motors. Sluggish performance of one of these valves

was documented in August, 1989, for EOG 2 in which the licensee produced

-traces of engine RPM versus time. Cognizant licensee personnel indicated to

the team that the valve's degraded condition was primarily due to poor air

quality.

To alleviate these problems, and improve reliability of the diesel

air start system, station engineering personnel recommended installation

of air driers and filters, the goal being conformance with ISA S7.3.

The

licensee's response to the problems cited above so far has been a lack of

commitment to long term corrective action, with more emphasis on short-term

solutions with respect to the material condition of the EOG air start system.

This was considered a weakness in the licensee's maintenance program.

3.d. DEFICIENCY IDENTIFICATION AND TAGGING

During walkdown inspections and observation of work, the team evaluated the

licensee's program for deficiency identification and tagging.

The following

problems were identified with the program:

During walkdown of the SI system, a number of small leaks and minor deficien-

cies that had not been identified by the licensee were identified (see section

2.a for details).

In addition, many deficiencies identified in the walkdown

had been identified by the licensee during their walkdown in late 1989.

However, deficiency tags written by the system engineer had not been hung nor

had WRs been submitted.

A number of deficiencies were identified that had been previously identified

by the licensee, WRs had been issued, yet the deficiencies had not been tagged

or the tags had been removed.

Health Physics (HP) personnel responsible for

identifying leaks and sources of contamination estimated that 10% of the tags

they hang get torn off or removed for some reason before the corrective work

is accomplished.

During the SI system walkdown by the team (in the first week of the inspection)

deficiencies identified in systems other than the SI system had not been tagged

or WRs initiated by the close of the inspection. The SI system engineer, who

accompanied the team on the walkdown inspection, did not feel obligated to tag

the deficiencies in other systems *

.Inconsistencies i.n tagging known deficiencies detracts from the overall

process of identifying deficiencies and initiating corrective action in that

people are less likely to initiate a tag since they cannot be sure whether a

problem has already been identified and corrective action initiated. The above

problems appear to be the result of a weak procedure for identification and

tagging of deficiencies. The procedure (SUADM-M-11) does not clearly specify

..

33

that anyone identifying a deficiency is responsible for initiating a WR card

(tag).

In addition, the word "should is used throughout the procedure,.

.

detracting from the effectiveness of the procedure. Procedures could be

strengthene~ by requiring that all personnel who are in the plant on a regular

basis routinely take WR cards with them and tag any deficiency noted.

3.e. TIMELINESS OF CORRECTIVE ACTION

During the inspection, a number of cases were identified where actions to

correct known problems were not taken in a timely manner.

These items,

discussed in detail in this section or other sections of the report,are

summarized as follows:

In late 1989, the licensee performed a walkdown of the SI system and identified

a number of areas where the piping configuration was not like the drawing:

Five areas where the Unit 2 SI piping configuration. did not agree with flow

diagram 11548-FM-089A, Sheet 2 were identified. By the end of the inspection,

a drawing change request still had not been issued to correct the drawings.

Although the configuration problems do not affect plant or system operability,

corrective action has been slow. Also, a temporary support was found installed

for valve 2-SI-MOV-2885C in the same late 1989 walkdown.

As of the close

of this inspection, an Engineering Work Request (EWR) had not been issued to

evalu~te the support. The above problems indicates weakness in taking timely

corrective action for known drawing errors and deficient conditions.

During the review, the team found that adequate corrective action had not been

completed for QA audit finding S87-22-03 issued in October, 1987.

The .

finding identified numerous discrepancies i~ completed WOs (the team also

found problems with completed WOs - see section 3.s). This audit finding

relative to adequate corrective action by maintenance to improve the quality

of WOs has been escalated by QA to step 2 (August 15, 1989) of a 3 step

escalation process. This is another example of problems with timeliness of

corrective action.

HP was tracking 688 primary sources of contamination. Forty-three new sources

were identified in February, 1990, and only one was fixed indicating lack of

attention to the problem.

Necessary modifications to the radiation monitoring system, such as elimination

of high background readings in monitor locations, has existed for years - see

section 3.f.

The need to replace containment instrument air.filters was identified in

mid-1989.

The filters still have not been replaced (see section 3.c).

Necessary modifications to the HVAC system have existed for years - see

section 3.f *

.

'

,.

34

3~f ALLOCATION OF RESOURCES

In an effort to determine the root-cause for delays in corrective maintenance,

the team interviewed cognizant engineers, foremen and craft.

From these

interviews, the team has ascertained that there are several contributing

factors responsible for this lack of maintenance and the resulting HVAC system

degradation~ These are as follows:

1.

The licensee's resources are committed on a priority basis, first on

those systems which are important to nuclear safety, second, on systems

necessary for power generation and third, on balance of plant systems/

noncritical to safety. Because most of the HVAC system is nonsafety-

related, manpower -dedicated for system maintenance is limited to a crew -

of four, with no allowance for overtime. System maintenance is for the

most part, limited to the safety-related section of the system and its

associated components.

Management's support for maintenance on the

balance of the system is very limited and, not program driven i.e., a

component operates until it breaks down.

When it (component) no longer

functions, it is tagged out and repaired or replaced depending on avail-

ability of funds, parts and/or manpower.

2.

The team ascertained that the licensee maintains little or no inventory

of replacement parts i.e., fan belts, sheaves, motors for quick

maintenance/repairs are not on hand. Therefore, simple off the shelf

replacement,parts are purchased through a cumbersome procurement system

which treats all replacements as though they are safety-related causing

long delays and extended down time.

Other examples of system degradation and the licensee's failure to respond in

a positive manner are as follows:

On July 31, 1989, the station experienced an ESF actuation at a time when

the ventilation system VS was being aligned, i.e., repositioning of the

HVAC dampers, for testing purposes. -An engineering work request EWR 89-540, was issued to investigate the root cause of the problem and report

the findings to management.

A review of the subject reports entitled,

Systems Engineers Review of Ventilation ESF Actuation Surry Power Station,_

-dated August 23, 1989, identified the root-cause as VS leakage and

actuator blow-by combined with reduced Instrument Air Header pressure.

_ The reports stated that actions in progress taken to correct the problem

included: Service all actuators to eliminate blow-by, walkdown system to

identify and repair all air leaks, verify pressure switch settings for

compliance with design requirements.-

The team determined that as of March

29, 1990, no substantive corrective action(s) had been taken on subject

actuators and sixteen related dampers. This was attributed in part, to a

lack of replacement parts for reasons discussed earlier. A detailed

walkdown of the system to identify and repair all leaks has not been

performed, hence the system continues to leak a-ta rate of approximately

47 cubic feet per minute and lastly, pressure switch verification setting

is only partially completed.

~ ---

--

-

,.

35

Also, by document review,.i.e. EWR 87-170, 3/12/87, 87-170A, 5/8/87 and 89-336 with Field Changes A-Y dated 5/19/89 through 2/14/90, and through

discussions held with cognizant personnel, the team has determined that

the containment air cooling recirculation and air cooling control rod

drive mechanism (CROM), systems in both units are worn-out and unable to

function as originally designed. These documents show that as of May 1987

fans, dampers, louvers, actuators and linkages in the containments

recirculation system were no longer able to function as designed and

required extensive structural/welding repairs or replacement.

Bids for

replacement were solicited in December 1987 but the licensee rejected the

two responses submitted for consideration.

More recently the licensee

issued station commitments CTS 89-7836~001, 89~7460-001 to replace the

containment recirculation fans and dampers by June 4, 1992.

In reference to EWR 89-336, the team ascertained that the air cooling

CROM, system is in a similar degraded condition in that doors, dampers,

linkage brackets, louvers, fans and even some of the ductwork no longer

function as designed in that it required 22 field changed to the subject

EWR to help remedy existing field conditions.

For example, certain

dampers had to be either blocked or wired open for continued operation,

ductwork from fan 1-VS~F-4B to the main duct trunk line required extensive

structural/welding repairs, control rod vent shroud cooler access cover

seats required extensive repairs, damper linkage brackets on control rod

shroud cooling fans 2-VS-F-60A, -60C and -60D were found broken and were

wired open for continued operation. Similar temporary fixes were imposed

for the dampers on recirculating fan 1-VS-F-lB. These conditions help to

demonstrate further the systems degradation, resulting from inadequate

maintenance and a lack of the necessary resources to keep the system

functioning as designed.

During review of annual calibration data for various detectors in the radiation

monitoring system (RMS), the team noted that Procedure No. CAL-001, dated

17 Aug. 1989, Log Ratemeter Scintillation Detector Source Calibration, Initial

Conditions paragraph 3.3 required, "Background/process count rate must be at

least one decade below the calculated calibration source count rate." A review

of the latest calibration data performed for RM-CC-105 and 106 (Component

Cooling Water [effluent] A & B respectively) in January and February 1990

respectively showed the background counts to be approximately 1.1 E+5 cpm,

and the calibration source strength to be about 7.0 E+4 cpm.

That is, the

background was higher than the source strength, opposite that of the initial

conditions requirements.

Based on interviews with the craft, the team learned

that this condition had existed for several years, and that on each occasion

of performing th~ procedure, a procedure deviation was prepared to permit

omitting the paragraph 3.3 initial condition requirement.

The problem was

described as being caused by contaminated sediment on the walls of the CC

piping; thus the monitors were not observing actual fluid activity levels.

The team noted that a Technical Report (No. NE-697, Rev. 0), entitled "Changing

RM-CC-105/106 to Off-line Monitors, Type 1 Final Report, Surry Power Station"

was issued in March 1989, and recommended the off-line monitoring system as the

most practical and only solution expected to work.

A Request for Engineering

'.

36

and Construction Assistance (Type 3) was initiated by the station on 14 July

1989, with a required start date of 1 September 1989 and required completion

date of 1 June 1990. Capital Project Type 3 IR-6369 was issued on 1 August

for "Replacement of CC Radiation Monitors".

Since much of the RMS was not operating in accordance with UFSAR commitments,

a special SNSOC Radiation Monitor Subcommittee had been reviewing the overall

status of the Surry RMS for the past several months.

In their report of 9

March 1990, endorsed by SNSOC on 15 March 1990, a "Long Term Action Item" (#3)

was listed as:

113. A Type 3 study will be initiated for replacement of RM-CC-105, 106 if

necessary.

a. Track progress of CC task team.

If fixed background is found to be

significant, proceed with Type 3 IR 6369 (on hold) to replace monitors.

-ENG/NSS

11

No information was provided about the "on-hold" status of the

plant modification.

The team perceived the above lack of substantive action as an inappropriate,

continued delay in performing required plant modifications.

An appropriate

engineering study had been completed, and there was no evidence that any

conditions had signtficantly changed during the past year. Therefore, it

appeared to the team that it was time to get o~ with correcting the problem,

rather than continuing to study the problem ad infinitum.

In addition to the

monitoring problems caused by the high background, the team was concerned

about the mentality that continued deviations of procedures tended to foster

in the craft - i.e., simply revise the procedure if it is not possible to

perform in accordance with the procedure. It was the team's opinion that

adequate resource commitment should be made by management to avoid forcing

craft personnel into deviating procedures because equipment was not

functioning according to design.

Other long-standing problems with permanent solutions proposed but not yet

implemented due to lack of resource allocations are associated with the EOG

air start system and containment IA system.

Six compressors were being replaced on the EOG air start system. These

compressors required constant maintenance and eventual replacement due to

moisture backflow (past in-line check valves) during compressor idle stage and

resultant rusting valves. The proposed permanent solution to the problem

(addition of air dryers and filters at the compressors* discharge) was deleted

from the 1990 station budget.

Containment IA was known to be below industry standards by the licensee for

many months.

Problems include lack of filters, high-dewpoints and water in

end-use devices. Solutions and needed repairs have been proposed.

However,

no improvements or repairs had been completed at the end of this inspection *

The above problems indicate a significant weakness in the allocation of

resources and in addition in the timeliness of corrective action for known

problems.

..

. ...

37

3.g JOB PLANNING

Supplemental Work Instructions Accompanying Work Orders Were Frequently

Inadequate

The team reviewed two types of procedures in use by the licensee: formal

procedures that were used on a repetitive basis to perform recurring

maintenance activities, and supplemental work instructions" that were

prepared as a part of the planning process for unique maintenance activities.

"Procedures are di~cussed in Section 3.n of this report.

The latter

"instructions", considered by the team to have the force of procedures in

implementation, were reviewed as a part of the planning process, since they

should typically be prepared during the planning phase of Work Order

preparation.

The team noted that approximately two-thirds of all work orders reviewed

utilized work instructions that had the intent and complexity of "investigate

and repair".

Review of these types of work orders led the team to the

conclusion that many functions such as utilization of proper procedures,

technical manuals, and performance of appropriate post maintenance testing

(PMT), may not have been accomplished.

For maintenance activities observed by

the team, it was noted that "planning" typically occurred at the craft level,

when the assigned craft was ready to go into the field to accomplish the work.

Investigate and repair" activities accomplished just that, i.e., no revision

to the details or scope of work in the work order was accomplished after

"investigat1ng", prior to proceeding with the "repair. In an effort to

determine the reasons for observed implementation problems, the team focused

on the planning and work order preparation program.

SUADM-M-12, dtd 20 April 1989, Work Order Planning, was the governing document

for the preparation of Work Orders by the planning department. This procedure

was noted to be weak because details of an acceptable program were absent from

the procedure.

The procedure did not contain a reference list, thus such references as

ANSI N18.7 - 1976, Administrative Controls and Quality Assurance for the

Operational Phase of Nuclear Power Plants, and INPO 85-038, Guidelines

for Conduct of Maintenance at Nuclear Power Stations (Nov. '85) were not

included. Station procedures necessary for the implementation of work

orders, such as clearances, housekeeping, system cleanness, rigging, and

calibration, were also not included.

Section 2.0, Work Order Planning addressed the subject of providing

details to accomplish work with the following:

"2.3 The planner will check the work order for completeness, clarity and

accuracy and add any* additional required information." Since this

instruction appeared to apply to the original Work Request, the team

observed that details listed in the Work Order were usually no more than

what had appeared in the Work Request.

' .

'*

!I'

38

112.4 The_ planner will plan the job and estimate resources required.

2.4.6 Work procedures required will be identified and obtained.

11

For most Work Orders reviewed by the team, it was noted that the

11originator

11

of the Work Request performed the

11job planning/procedure writing" in his

preparation of the document that identified the problem, i.e., the work

request.

For maintenance activities that were observed and reviewed by the

team during the inspection, the work request would state,

11 Item Xis failed.

Troubleshoot and repair.

11

This was translated to the Work Order, and in most

cases became the

11procedure

11 for repair of the component.

In general, no

amplifying "procedural steps" or supplemental work instructions were prepared

except for complex repair activities.

11Planning

11 was performed at the craft

level, at the time the repqir activity was started.

Many maintenance.activities were not addressed by a pre-prepared procedure,

thus it was frequently necessary and appropriate to provide supplemental work

instructions in the Work Order. If procedures were available, work

instructions should be prepared to sequence the work, or direct the use of

~ections of pre- approved maintenance procedures or technical manuals.

In

cases where pre-approved procedures were available and accompanied the Work

Order, the team observed the use of one or two pages of the procedure that was

forty or fifty pages in length - the unused pages were marked

11 N/A 111 as not

appropriate to the Work Order being performed ..

o

No approval process for the content of the Work Order or the adequacy of

the planning process was included in the SUADM-M-12 procedure. Craft

concurrence was not included in the process. Quality review of the

work orders was not specified.

No check lists of necessary work order

attributes were available to assist the planners in preparing comprehen-

sive work orders.

o

Paragraph 2.4.1 stated, "Required plant status and initial equipment

conditions will be identified~"

No further details were provided in

the SUADM-M-12 procedure about preparatory steps for the setting of

clearances, establishing necessary plant conditions, or other meaning

the paragraph may have had.

Guidance on special considerations was not included in the procedure.

As examples, safety considerations such as confined space or fire

protection work permits were omitted from the planning procedure.

11Interfaces with other crafts or departments will be identified and

necessary work requests submitted.

11 was the extent of consideration

given to items such as RWPs, scaffolding, and special chemical

requirements. No guidance on the extent or level of detail of work

instructions was provided, i.e., word by word instructions for some

types of activity, but limited guidance,

11skill-of-the-craft

11 for

other activities. Guidance concerning pre-planning walkdowns of the

expected maintenance activity was not included.

' *.

' .

- *::.

39

Based upon the above types of inadequacies in the work order planning and

preparation instructions, the team concluded that the work order planning

program needs significant improvements.

During interviews with the craft and planning department personnel, the team

learr~ed that

11 planners

11 did not function as a centralized group responsible

for preparation of detailed work instructions. Planners were forced to

function as schedulers and material procurement personnel.

By craft, the

following was determined to be the planner staffing strength:

Staff

Contractor

On loan from

craft resources

Electrical

1

2

2

Mechanical

9

4

1

The planning department staffing did not permit the detailed work instruction

preparation that was typical 9f the industry, and the burden of this segment

of the planning effort defaulted to the craft. Since the craft felt they knew

what they were going to do, pre-job details. were frequently not added to the

work orders for the electrical and ~echinical crafts.

I&C technicians

generally attempted to revise existing procedures to

11make them fit

11 the

intended maintenance activity.

Based on the extent of the observed problems with the procedure governing the

maintenance planning effort, the inadequacy of work control documents, and the

sparse planning staff, the team concluded that the program for preparation of

maintenance authorization documents was a significant weakness.

3.h. POST MAINTENANCE TESTING (PMT) PROGRAM WEAKNESSES

The PMT program is governed by the following two procedures:

SUADM-M-27, Revision 1, Requirements for A Post Maintenance Testing

Follower

SUADM-M-47, Revision O, Post-Maintenance Test/Verification program

The following problems were identified with the program:

The current PMT program, procedure M-27, is very limited for equipment other

than ASME Section XI equipment.

The procedure was written -to cover Section XI

testing and has been revised to cover other safety-related equipment and, on a

very limited basis, non-safety-related equipment.

However, the procedure is

.primarily a Section XI procedure and provides very little detail for other

. '

.

'-.,,

40

equipment, e.g., electrical. Although, the PMT Follower is used for all

safety-related WRs, the lack of procedural detail results in inconsistences in

specifying PMT for equipment other than Section XI.

Personnel responsible for specifying PMT on the Follower have experience

primarily in Section XI testing and are not qualified to specify testing in

other areas such as electrical. Again, this results in inconsistences in

specifying test requirements.

On some WRs reviewed, the PMT follower

indicated that engineering was contacted for the required PMT.

However,

without procedural guidance, desired results cannot be assured.

In the electrical and I&C areas, the PMT Follower is attached to the work

order. The electrical and I&C sections stated that they are only required to

perform the PMT specified on the Follower. The planners stated they are not

responsible for PMT.

Fortunately, most of the electrical and I&C procedures

have sufficient testing and calibration requirements that are equivalent to

PMT.

However, two weaknesses were identified. One item is that when switch-

gear is sent to an outside vendor for refurbishment, PMT was not adequately

specified. Another item is that the maintenance procedural step for functional

testing can be marked

11 NA

11 (not required) if Operations is not available to

perform the step. The licensee stated that the procedures will be revised

disallowing NA

1d steps for functional testing by Operations.

The licensee

stated that temporary corrective action will be immediately implemented by

requiring that qualified electrical system engineers will specify all PMT on

the Followers until the new program is in place. The team considered these

responses by the licensee as acceptable.

Based on previous assessments (INPO and Licensee), the weaknesses in the PMT

program had been recognized and corrective actions initiated.

In October,

1989, a task team was assigned to develop a new PMT program.

At the time of

the inspection, a new comprehensive PMT was under development.

The program is

detailed in procedure M-47 and consists of a series of matrices for each

component under the program.

At the time of the inspection, the only matrix

that had been issued was for the Auxiliary Feedwater Pumps.

The scope of

equipment to be included in the program was still being developed.

By the

close of the inspection, the scope had been defined and was to include all

equipment on the mechanical and electrical

11Q

11 list {approximately 15,000

items). Licensee personnel indicated that tentative plans are to have

matrices completed for ISI/TS (Inservice Inspection/Tech Spec) items in 1990,

safety-related items in 1991, and BOP (balance of plant) items in 1992.

To

date, matrices have been developed for a significant portion of all ISi/TS

check valves, motor operated valves (MOVs), and safety-related 4160v and 480v

breakers. However, none of these matrices have been through the review

process and added to the procedure.

3.i. TESTING OF AUXILIARY FEEDWATER TURBINE

The auxiliary steam turbine-driven pump unit consists of a Terry turbine

coupled to an Ingersol (Ingersol/Dresser) pump equipped with a Woodward

governor and a trip throttle valve manufactured by Gimpel Machine Works

Incorporated (Gimpel), Philadelphia, Pennsylvania. Industry wide problems

,.

41

including overspeed trip failures were encountered with the Terry turbine

and IN 88-67 described two instances where the Terry turbines failed to trip

on overspeed.

INPO issued at least three SOERs on this subject. Major.

maintenance was performed on the Terry turbines installed at Surry Units 1

and 2 during 1988 (WO 58211) and 1986 (WO 41408), respectively. Attachment 8

to WO 58211 indicates that the Terry turbine overspeed trip was set at 6280 rpm

and verified to trip at the set point.

WO 41408 records did not indicate that

the overspeed trip mechanism operated. The Terry Corporation, the manufacturer

of the Terry turbine, in Instruction Section 7 of the Instruction Manual states

in part,

11It is most important that every overspeed device and trip mechanism

be tested regularly, preferably once monthly. This will insure that the

tripping mechanism is operating freely.* The test can be made manually or by

overspeeding the Unit. The mechanism, when tripped, should respond instantly

and reduce the speed of the Unit, or hold it from overspeeding with the

throttle valve wide open.

Record trip set point and date of initiation.

Verification of proper functioning and setting of the overspeed trip device

during initial startup is mandatory. This should be accomplished with the

turbine disconnected from the driven equipment.

11

The licensee did not

translate the vendor recommendations into a procedure to conduct an overspeed

trip testing of the turbine.

The team reviewed the action taken by licensee relative to IN 88-67 entitled,

11PWR Auxiliary Feedwater Pump Turbine Overspeed Trip Failure.

11

Station

Commitment Assignment/Response Form (SCARF) 88-6067-003, which was initiated

to track the actions relative to IN 88-67 stated that the Woodward governors*

would be replaced for Units 1 and 2 during the next refueling outages. It

also stated that the Terry turbine for Units 1 and 2 were overhau.led in 1988

and 1986, respectively, during which the tappet balls were replaced and

overspeed tests were performed.

In view of the recent overhaul and subsequent

overspeed trip test, the SCARF requested an extension of time to develop a

procedure to conduct an overspeed trip test with the new governor.

The

licensee was unable to produce any records to substantiate the statement in

the SCARF that the overspeed trip test was conducted on the Unit 2 Terry

turbine.

The Industry Operating Experience Review (lOER) Committee reviewed the "Testing

of Steam Turbine/Pump Overspeed Trip Devices" and provided a recommended action

plan. The document referenced:

'

(INPO) SOER 8113:

Current Loss of High Pressure Core Cooling Systems.

(INPO) SOER 86-13:

Reliability of PWR Auxiliary Feedwater Systems.

(INPO) SEN 55:

Failure of Woodward Governors results *in Auxiliary

Feedwater Pump Turbine Overspeed Trip Failure.

NRC IN 88-67:

PWR Auxiliary Feedwater Pump Turbine Overspeed Trip

Failure.

NRC Case Study Report AEOD/C602:

Operational Experience Involving

Turbine Overspeed Trip, dated August 1986.

' ...

..

42 .

The action plan for Surry stated three concerns and outlined the actions to be

taken for each concern. However, the team determined that attributes such as

the following were not considered:

Recommendations of the manufacturer of the Terry turbine

Research of the overhaul and maintenance records to determine if

overspeed trip tests were indeed performed.*

Engineering evaluation to determine the consequences of operating the

Terry turbines without knowing if overspeed trip device fails to operate

and overpressurizes the piping downstream of the turbine.

The frequency at which the overspeed test should be performed considering

attributes such as, the vendors recommendation to test the overspeed trip

mechanism monthly, the complexity of uncoupling the pump before the test,

the fact that the Terry turbine is operated infrequently, and that there

are no pressure relieving devices downstream of the AFW turbine.

IN 88-67 dated August 27, 1988, described a failure of the AFW pump turbine

overspeed mechanism in July 1988 at the San Onofre Station. A failure of the

overspeed trip mechanism occurred at the Ranch Seco Station during January

1989 during which the AFW System was pressurized. The team determined that

contrary to paragraph 6.2.1 of Procedure VPAP-0504, research and evaluation of

the Vendor Technical Manuals was not performed and the recommendations of the

vendor to test the overtrip mechanism was not incorporated in a suitable

procedure.

Station requirements for processing vendor supplied information were found in

SUADM-ADM-31, Vendor Interface/ Control of Vendor Documents, dated 5 Dec.

1985, at paragraph Nos. 8;o, 8.1.1, 8.1.2,and 8.1.3 (see section 3.b)

Based upon the above example, the team concluded that the program for control

and incorporation of vendor supplied information was not being effectively

implemented in accordance with required station procedures. This failure to

. follow procedure is an example of violation 50-280,281/90-07-0l *

3.j. PERSONNEL SECURITY TRAINING

Training in security access control was noted to be lax in at least one case

by the inspection team.

One maintenance technician was observed to

11tail-gate

11 another technician into the Emergency Switchgear and Relay Room.

Confusion between the concept of

11accountability

11 and

11controlled

accessibility

11 appeared to be the cause of the problem.

The licensee agreed

to take appropriate action in the matter .

.

...

.,

43

3.k. CONTROL AND CALIBRATlON OF MEASURING AND TEST EQUIPMENT (M&TE)

Procedure SUADM-M-39 (dtd 13 Dec. 1988), Control of Measuring and Test

Equipment, established the facility M&TE program.

The procedure detailed the

issue, recall, storage, and segregation of special, limited-use, or expired

M&TE.

The procedure also provided instructions for placing damaged equipment

out-of- service, and for resolution of out-of-tolerance equipment used in the

field. Although the calibration laboratory was small, it was found to be

clean and organized, and adequate in its function.

Each piece of equipment was assigned a unique

11SQC

11 number by which the

equipment

1s calibration and use history was tracked. All tracking and recall

was performed manually on cards associated with the equipment by the SQC

.

number.

No formal

11check-out

11 system was employed by the licensee for pieces

of M&TE under control of the I&C and Operations group; two way traceability on

equipment used in quality work was maintained by, 1) recording the equipment

and its SQC number on the work order itself, and 2} recording the date and

procedure that the M&TE was used on a Test Equipment Record card (Form#

MTM06) attached to the equipment itself (instructions for use of the MTM06

form were not included in the SUADM-M-39 procedure, but appeared to be

properly implemented).

Electrical and mechanical craft similarly maintained

storage of limited M&TE under their control, but most electrical and mechanical

items were checked _in and out of tool issue points. Recall of M&TE approaching

the calibration due date was controlled by manual review of history cards

segregated into month-due categories. The M&TE.technician generated a

Certification Due Notice, and forwarded it to the cognizant supervisor having

possession of the instrument.

Although the program appeared fundamentally adequate, several areas of concern

were noted by the team. A high level of responsibility was placed on the

various disciplines for proper control and handling of M&TE due to the

decentralized control of the equipment.

All disciplines except Operations

readily complied with requirements of SUADM-M-39 related to use, storage, and

recall. The team concluded that many of the following types of problems were

the result of no single point of contact in the Operations department taking

responsibility for the program.

Several areas of procedural non-compliance in M&TE control were noted in the

Operations department.

For ~xample, Section 8.0 of SUADM-M-39 addressed the

subject of storage of M&TE, including segregation;environment, and handling.

M&TE under the control of Operations was stored in two lockers behind the

control room adjacent to an air-conditioning unit. The lockers included all

kinds of paraphernalia, including some very heavy test gear such as connecting

piping, discarded radios and their associated chargers, etc. The lockers were

in total disarray, thus storage conditions could have (and had) damaged or

disrupted calibration of the instruments. As an example, an Eagle Eye Flow

Meter (3 - 9,000 GPM range) had a calibration sticker reflecting calibration

of 08/22/89, and due date of 8/90 (SQC #3708).

The meter face had been broken

out of the meter for a long period of time, but reflected a last used date of

2/18/90 on the Test Equipment Record Card (i.e., the equipment had not been

.

....

.>

r

44

removed from the system). The test for which the equipment was used was

illegible, and the card was not annotated with the SQC # of its associated

equipment (Card loose in the meter container).

Two 1,500 - 4,500 GPM flow

meters were also stored in the locker, one with a broken meter face.

Neither

meter was calibrated because the calibration facility did* not possess equipment

accurate enough for the 2

11 H20 range of the meter.

Two pressure gages were also found in the locker.

One O - 5000 PSIG pressure

gage was

11tested

11 12/17/87 (not in M&TE system), and one O - 1000 gage was

"calibrated" 1/29/90, and

11due

11 2/29/90 (i.e., past due - # SQC.3655).

Two dual stage (0 - 4,000, 0 - 400 PSIG), pressure regulators were also found

in the locker that were used to perform containment penetration leakage tests.

These regulators were un- tested and un-calibrated. Paragraph 6.2 of

SUADM-M-39 specifically required all M&TE to be calibrated prior to use on

safety-related systems, but specifically excepted devices that were contin~

uously monitored during use with other certified equipment (in this case, the

regulators were used with high accuracy, Heise gages).

In that instance,

paragraph 6.2 required application of a "No Calibration Required" sticker to

certify the acceptable use of the device with other calibrated devices.

Similar situations were noted with several power supplies in use by I&C

technicians.

In all cases (regulators, power supplies) observed by the team,

"No Calibration Required" stickers were not affixed to the devices which was

contrary to procedural requirements.

Improper storage conditions were also noted by the team for laboratory type,

six foot high, roll around instrument racks that were under the cognizance of

the I&C group, not Operations. * To meet seismic restraint requirements, the

licensee had "temporarily" (Summer of 1989) wrapped a large chain around the

upper cabinets in the rack.

The upper cabinets included power supplies,

counters, digital voltmeters, etc. The chain was then wrapped around an

adjacent stanchion to keep the entire assembly from damaging adjacent reactor

protection instrument racks. Th.is storage method was perceived by the team as

a poor practice since a bump into the rack resulted in the chain wrenching on

the instrument cabinets.

The licensee should correct this condition of

improper storage storage of M&TE.

The team reviewed instrument history sheets and recall lists, and noted that

several instruments under the control of Operations were overdue for calibra-

tion for several years (e.g., SQC # 3653 - 3/87, 3589 - 10/88, 3566 - 10/88,

3562 - 7/85, 3544 - 6/86). Several Notice(s) of Instrument Restriction had

been generated because the device(s) had gone beyond their certification

due date; there was no evidence that compliance with paragraph 7.5 of

SUADM-M-39 had occurred regarding application ~f Certification Overdue

stickers, and return of restriction notice copies to the calibration facility.

The team was especially concerned that the inventory of required test equipment

to support the plant was being "lost" due to such events as breakage and

contamination, but not being recognized due to lack of control of calibration

status. The team perceived this lack of control as a significant implementa-

tion weakness.

'*

.

....

_.,

45

The above examples in addition to those discussed in sections 2.e, 3.b and 3.i

indicate, that the licensee failed to follow procedures for maintenance. This

will be identified as violation 50-280,281/90-07-0l :"Failure To Follow

Maintenance Procedures

11

In exploring potential

11 loss

11 mechanisms of test equipment, the team learned

that the licensee has no facility for calibration or comparison checking of

contaminated equipment. If an M&TE device becomes contaminated (Heise gages

were a good example), the device was simply stored, and used in contaminated

applications only, until the expiration of its normal calibration period. At

this time, the device was permanently retired, with no check of accuracy

before retirement.

Thus compliance with paragraph 7.6 of SUADM-M-39 regarding.

evaluation of out-of- calibration conditions when a device was presented for

calibration was not possible. The licensee advised the team that the

,.

situation of contaminating an instrument had not occurred for a long period of

time.

The team noted areas of strengths and concerns for devices requiring off-site

calibration due to cal lab limitations.

For example, the qualified vendors

list was up-to-date, and a sample .check of devices recently sent

11out

11 for

calibration showed the items to have been sent to a qualified vendor.

The team

noted an effort on the part of the instrument technician to anticipate the

lead time required to obtain authorization to send devices off-site but also

noted a few cases where the ~evice was over-looked until due/past due.

However, on occasion, the attempts to anticipate calibration requirements were

thwarted by extremely variable processing time (months) on purchase orders to

requisition the calibration service. The net result was several items of M&TE

were out of calibration before required vendor support could be authorized and

obtained. Multiple pieces of the same equipment prevented this from becoming

a significant problem.

Related to sending equipment off-site, the team noted that infrequently,

quality control personnel performing receipt inspection of returned M&TE would

erroneously retain and/or separate an instrument

1s pedigree papers from its

associated device. Again, this was not a significant problem because

ultimately, the papers could be traced and recovered.

However, the situation

was of concern to the team because the licensee currently utilized only one

technician to manage the calibration program, and manage and operate the

calibration laboratory.

No persons-in-training were assigned to the

laboratory. It was the team's opinion that a program improvement in this

area should be considered by the licensee.

Another team concern was related to availability of M&TE device accuracy to

the technicians performing maintenance activities. This was of concern to the

team because procedures in use by technicians were noted to cover the entire

spectrum of required accuracies and required instruments to suppo~t the

procedure *

'*

46

For example:

(1)

Procedure l-PT-2.lA, dated 17 July 1987, Reactor Coolant Wide Range

Temperature (T-1-410), stated at Initial Conditions paragraph 3.4, "Insure

that the test equipment to be used has adequate precision and range to

measure the desired parameter and has been calibrated against standards

traceable *** ".

(2)

Procedure Cal-001, dated 17 August 1989, Log Ratemeter Scintillation

Detector Source Calibration, stated at Initial Conditions paragraph 3.6,

"The following test equipment or equivalent is available, calibrated and

meets accuracy requirements as specified.

3.6.1 Frequency counter or Scaler Timer.

3.6.2 Digital multimeter."

No accuracy requirements were specified.

(3)

Procedure Cal-044, dated January 12, 1990, stated at Initial Conditions

paragraph 3.2, "The following test equipment or equivalent is available

and calibrated. *

3.2.1 Function generator, HP 3310A.

3.2.2 Pulse generator, Rutherford 816.

3.2.3 Digit~l multimeter, Fluke 8110A.

II

The team observed that the first two examples relied on the technicians to

perform instrument acceptability determinations, and did not provide accuracy

values for the instruments that could be used in the performance of the

procedure.

The team concluded that preparation of

11upgraded procedures"

was tending towards overcoming inadequacies in accuracy requirements versus

accuracy of instruments available. It was the team

1s opinion that as an

interim methodology, the licensee should consider making readily available *to

the craft all M&TE instrument accuracies such that if accuracy criterion are

specified by a procedure, the craft will be able to readily determine

acceptability of instruments in use.

3.1. DEFINITION OF MAINTENANCE REQUIREMENTS

The team reviewed instruments and control devices to determine if adequate

calibration activities were being accomplished on control room indications

that operators would use to implement normal, off-normal, and emergency

operating procedures.

The team examined three categories of devices: 1) those

that fulfilled Technical Specification requirements, 2) those that fulfilled

regulatory guidance or operating procedure requirements, and 3) those that

provided personnel and/or equipment safety functions.

The team learned.that

instruments in the first category (some safety-related) were subjected to

formal, proceduralized Periodic Tests (PTs) that accomplished instrument

47

calibrations and checks consistent with Technicai Specification (TS) require-

ments. A sample check of the PT index, that listed the applicable PT procedure

number, the related TS table requirement, and the frequency, showed that all

instruments were properly addressed.

An informal checklist was in use by the

licensee that distributed the many calibrations (by month) over an 18-month

cycle. The team concluded that, based on the sampling conducted, an adequate

periodic calibration program for TS required instruments had been implemented.

It was noted, however, that periodic calibrations were in the beginning stages

of integration with the station preventive maintenance program.

In the second category of instruments, the team selected as a sample base,

Reg. Guide 1.97 instruments and learned that these instruments were equally

well addressed in periodic calibrations as the Technical Specification

required instruments.

However, of concern to the team was the fact that a

Reg. Guide 1.97 instrument did not have a formally approved procedure for

accomplishing appropriate calibrations.

For example, Unit 1 pressurizer

relief tank (PRT) temperature (T-1-471} was calibrated using a (apparently

startup) procedure that was not dated and not approved.

The last calibration

performed using the informal procedure was September 1989.

The team noted,

however, that Procedure Number 2-Cal-333 had been approved for the Unit 2 PRT

temperature device on 11 March 1986. Several other U-1 devices had no

currently approved calibration procedure, such as:

o

Containment Vacuum Pump Discharge Flow (F-CV-150)

o

Throttle Pressure (P-MS-102}

o

SW Flow to Control Room Chillers (PDI-SW-130A, B, C)

o

SW Flow to Component Cooling Heat Exchangers (PDI-SW-132A,

B, C, D)

These instruments were non-safety, but used by control room operators to

implement operational procedures.

The last category or group of instruments the team was particularly concerned

about were instruments that provided control room operators direct indication

of the integrity of operating safety-related equipment, e.g., the EDGs.

Several local alarm windows at the EOG local control panel were actuated by

pressure and temperature switches, such as:

Oil Pressure (low)

Crankcase Pressure

(high}

Cooling Water Press.

(low)

Starting Air Press.

(low)

Engine Temperature

(high}

20 psig (engine speed 125 - 870 rpm)

44 psig (engine speed above 870 rpm)

1 - 1.7" H20

20 psig

165 psig

190 deg. F.

.. .

48

Fuel and lube oil parameters were also monitored. These devices, in addition

to local alarms, provided contacts in a supervisory alarm circuit for the main

control room annunciator 1J-H7, "EOG #1 Trouble (similar for all diesels).

Control room personnel initiated off-normal response procedures on receipt of

the alarm(s).

None of the pressure or temperature devices that initiated

alarms were included in a periodic test/calibration program.

In addition, the

team noted that local reading instruments, e.g., the many pressure gages on

the air start receivers, were similarly not included in any periodic test or

calibration program.

Thus an auxiliary operator, when dispatched to the EOG

local panel on receipt of a control room alarm, could not be assured of valid

indication for engine parameters.

The lack of periodic calibration of

critical engine instrumentation was perceived by the team as a program and

implementation weakness.

The team noted, however, that another important

system, instrument air, did have many of its pressure control switches (e.g.,

those associated with the compressor operation) in a periodic test/calibration

program.

Coverage was found to be system dependent.

The licensee did not

have an instrument list, thus covered instruments were not able to be readily

determined.

The licensee advised the team that the "PM Upgrade Program** would

address this problem on a system by system basis as the program analysis is

completed during the next several years. It was the team's opinion that the

I&C group should provide Operations with a list of those instruments providing

direct or indirect indication (alarms) to the control room of system

integrity. Operations should evaluate those instruments that are not

calibrated for potential safety of equipment impact such that a decision on

the necessity for the instruments to be calibrated can be made in a more

timely manner.

The team concluded that the PM Upgrade Program when

implemented as described should tend to correct currently observed problems.

3.m. PRIORITIZATION AND BACKLOG CONTROLS

The team reviewed the licensee's records, schedules, interviewed maintenance

department personnel, and attended planning and scheduling meetings to

determine the effectiveness of the licensee's prioritization scheme, and the

extent and control of the maintenance backlog. Specific areas examined

included prioritization, deferred preventive and corrective maintenance,

and

measurements of past and current backlog *

During their review of work orders, the team observed numerous high priority

work orders that were very old:

WO#

Priority

WO Approval Date

076790

1

89/01/12

077997

1

89/02/03

086587

1

89/10/10

088023

1

89/11/17

The team reviewed the program for work order prioritization to determine the

adequacy of the program, and whether it was being properly implemented.

Procedure SUADM-M-11, dtd. 14 April 1989, Work Request System, was the

49

governing procedure for assignment of corrective maintenance (CM) priorities.

(Preventive maintenance activities were always assigned Priority l; this issue

is discussed below.)

Paragraph 3.2.1 stated, "The OMC completes Blocks 10 -

12, Blocks 14 - 20, reviews the WRC for accuracy and completeness." The

Operations Maintenance Coordinator (OMC) was the SRO-qualified person making

the prioritization decision for entry in Block 11 of the Work Request Card

{WRCJ.

Attachment 2, Work Request Card Completion, provided the guidance for

priorities as follows:

Priority 1

Priority 2

Priority 3

Priority 4

Priority 5

Priority 6

Priority 7

Urgent work, scheduled to start in 24 - 48 hrs.

Priority work, scheduled to start one week after approval

Used to build a backlog of work, ordinarily scheduled to

start 4 weeks after approval

Work to be done at time specified by originator, such as

outage

Priority work done during a trip

Work on equipment without a redundant system - treated as

priority as it could cause a shutdown

Work on equipment that could cause a shutdown if the redundant

system failed.

The team noted in practice that only priorities 1 - 4 were used in work

orders.

The team also noted what was considered a significant program

weakness in the above prioritization scheme:

o

The scheme did not consider Technical Specification (TS) LCO limitations

o

The scheme did not consider safety of personnel

o

The scheme did not consider safety of equipment

o

The scheme did not consider impact on unit generation capacity, etc.

The team discussed the above considerations, both the written program and

team's opinion of an adequate scheme, with the station personnel responsible

for prioritization assignment and learned the following.

Priorities 1 - 3

were assigned based on consideration of Technical Specifications, safety of

equipment and personnel, and unit generation capacity. Equipment redundancy,

extent of degradation, and assessment of risk of challenge to safety systems

were considered in the priority assignment.

As discussed below, the team also

learned that high priorities were conservatively assigned in an attempt to

overcome production sluggishness caused by other factors. Priority 4 appeared

to be the only priority that was assigned in accordance with the program

requirements - i.e., a Priority 4 Work Order was planned for a normal outage.

It was the team's opinion that the current program for priority assignment

constituted a weakness, however, the manner in which. the ineffective program

was being circumvented and implemented was adequate.

The team learned that once a priority had been assigned to a work order,

limitations on WPTS access and system difficulties made the change of the

priority very difficult. This was considered a program weakness because

several aged, Priority 1 work orders were being carried in the system that

were not actually high priority any longer. For e~ample, WO# 088023 required

the fabrication and installation of missing generator shroud bolts on #3 EOG.

>

50

It was a Priority 1 work order when written because the shroud was rubbing

on the engine to generator shaft; now that the shroud was temporarily,

correctly positioned by temporary bolts, it was appropriate to change the work

order to a lesser priority. The priority revision had not been made because

the change was too difficult. It was the team's opinion that a program change

should be made so that assigned priorities have meaning.

In the area of preventive maintenance deferrals, the team noted that the

corporate reported performance indicators for 1989 averaged approximately

1.1%, closely approximating the licensee's goal.

"Deferral" was defined as

the percentage of PM during the monthly period not completed within _the

assigned grace period (lX periodicity for EQ, 1.25X periodicity for non-EQ).

During the team's review of individual deferral approval sheets, the team

noted that the actual deferral rate was much higher during 1989, somewhere

approximating 15 - 20% average for the year (as high as 43% in June 1989).

The discrepancy between "reported" and actual deferrals related to the method

of record keeping - only if a PM activity were scheduled, then not completed,

did the PM become a reported "deferral

11

The team concluded that this m~thod

of record keeping constituted a poor practice in that it did not give a clear

picture of the amount of deferred preventive maintenance that contributed to a

high backlog.

Changes in record keeping and aggressive attention to completion

of PMs have recently been initiated that should cause reported "deferrals" to

track actual deferred preventive maintenance. Actual deferred PMs for February

1990 were approximately 1%.

A review of recently deferred PMs showed that some very important PMs were

being deferred for illogical reasons.

For example, PM EE-EDG-M/Al, Emergency

Diesel 1 Year Service and Inspection, stated the "Reason for Deferral" as,

"Parts" although the PM was not new.

Engineering review of the deferral had

rejected "parts

11 as not being technically justifiable.

PM EE-C-M/Ml, a

Monthly Compressor Check of the EDG air start compressor (Mark# 01-EGl-

C-QXl), stated the reason for deferral as,

11PM missed due to parts - new PM

11

Again, engineering review of the deferral had rejected "parts** as not being a

technically justifiable reason, and noted that the PM was the same as the

quarterly scheduled PM since 1987.

Manpower was also frequently listed as a

reason for deferral. It was the team's opinion that parts unavailability and

lack of manpower for long-standing PMs constituted poor planning.

The PM

Deferral Review Sheet did not provide for listing the last time the PM activity

was accomplished, i.e., how far beyond end-of-grace, although that information

was available in the WPTS by performing a special report run-off. If a PM

activity was indefinitely deferred, it was simply noted as "missed

11 at its

next due date in the WPTS to permit clearing the previously generated work

order number from the system.

The team also noted that all PMs were assigned

a priority of 1. This tended to defeat the purpose of a prioritization scheme,

and prevented effective management of backlogged work.

It was the team's

opinion that a significant program improvement would accrue from a two-fold

scheme that related PM priority to importance of operation to the plant and

importance of operation to the component itself *

.

f

51

For each deferred PM, the discipline supervisor was required to review the

deferral, then engineering reviewed and approved/ rejected the deferral. The

responsible superintendent (e.g., maintenance) reviewed and approved all

deferrals; SNSOC approval was required for all EQ PM deferrals. It was the

team's opinion that a program improvement would accrue from an increasing

level of management review of deferred PMs for increasing amounts of time

beyond end-of-grace period.

The licensee's program for backlog control was largely not proceduralized.

Passing mention of backlog control was made in procedure SUADM-M-14, Work

Order Scheduling, dtd 19 Dec. 1989, para. 2.3, that stated,

11A preliminary POD

will be established to reflect carry-over work, new work scheduled from the

backlog schedule, and new work scheduled as a result of station needs.

11

Backlog control was observed to be a de-centralized function of the craft~

schedulers reminding the craft foreman of outstanding work orders, and the

need to continue to work off backlog as

11filler

11 items when manpower

permitted. The centralized planning and scheduling function dealt primarily

with the high priority work items on a day-to-day basis. The above type of

backlog control was largely a function of planning, i.e., craft-level planning

for the I&C shop, shared craft-level and centralized planning for the

electrical craft, and mostly centralized planning for the mechanical craft.

In addition to craft "lists

11 of backlog items, the Operations_ Department had

recently begun preparing prioritized lists (watch station basis weekly) of

plant equipment that requi,red increased operator action because of equipment

faulty operation.

For example, numerous work orders were outstanding on

service water temperature control valves (Mark #s SW-TCV-108/208); since their

operation was unreliable, operator action was routinely required to monitor

and adjust. These lists of approximately 200 items affecting all crafts were

used to help set craft work priorities.

Even though the control of backlog-was not proceduralized, management tracked

and was acutely aware of backlog through performance indicators. A review of

the performance indicators showed an overall increasing trend of outage and

non-outage backlog from June 1989 (when data was first available) to the

present.

As of 28 March 1990, the total outstanding work items numbered 4727,

which included corrective maintenance {CM) work orders, PMs, and EWRs.

By the

end of 1989, the average age of non-outage CM had increased from approximately

200 to 300 days, but then dropped back to about 200 days in January 1990. The

total non-outage CM at the end of March 1990 was 2027 items broken down as

follows by craft and readiness to work:

Mech

Elect

Weld

Labor

Const. Other

In planning 447

374

183

59

7

16

698

Working

71

65

62

3

12

29

1

..

>

52

In addition to the above, approximately 500 PM activities were performed each

month for the three major crafts as follows:

Mech

Elect

240

200

60

Approximately 50 of the 2,000 CM items were recorded as Priority 1 work items.

Each of these work items was reviewed by the team with the licensee for its

safety impact on the operational status of the plant.

As discussed above

concerning work order prioritization, Priority 1 was applied to work orders in

a very conservative manner.

That is, only approximately five of the fifty

work orders. were actually high priority. The ba 1 ance of the work orders were

being treated as high priority because of their long term, potential impact on

equipment or on generating capability if further, significant degradation

occurred.

The five work orders were receiving detailed attention consistent

with their impact on plant operation. The team's review of other lesser

priority, outstanding work orders showed many of them to be 100 1s of days

old.

In a related issue, the team reviewed the open clearance (equipment

tag-out) records and noted the following figures:

Number of *

Days Active

Tagging Reports

Ave

Maximum

Unit 1

146

215

1,598

Unit 2

72

148

1,013

The team noted that there was a direct correlation between very aged equipment

clearances and work orders, i.e., most of the old clearances continued to

exist due to incomplete very old work orders.

The licensee advised the team that recent efforts at increasing the coordina-

tion between the craft and the support interfaces were having a positive

effect in stopping the increasing trend.

As noted in Section 3.r of this

report, it was the team's opinion that substantive action should be taken

by the licensee to improve the effectiveness of coordination and improve the

support interfaces to positively reverse the adverse trend of backlog.

3.n. PROVIDE MAINTENANCE PROCEDURES

The team reviewed procedures as a part of work observed in progress, and as a

part of work packages already accomplished. Comments on procedures generated

as a result of observation of the work in progress are discussed in Section

2.h.

Comments derived from review of work packages are included below.

Procedure ECM-1503-1, Rev. l,*MOV Motor Repair and Replacement, ,provided

instructions for replacement and repair of MOV motors. This procedure was

approved for use on 15 June 1989, and represented the

11upgraded

11 format.

This

procedure was also one of those in the licensee's special MOV program and

therefore received special scrutiny during its preparation phase.

...

..

53

The team carefully reviewed this procedure to determine the adequacy of the

licensee's maintenance procedure upgrade program. Split responsibility

between the electrical and mechanical craft existed in the MOV actuators - the

electrical craft were responsible for the actuator motor, and limit and torque

switches.

The mechanical craft had responsibility for the balance of the

actuator. Notwithstanding the special attention that the procedure had

received, the team noted the following areas for improvement.

Although portions of this procedure addressed many electrical operations that

had to be performed with the valve energized, paragraph 4.0 Precautions and

Limitations only listed "None". Paragraph 6.2.8 included a Caution that

stated, "Possible equipment damage.

Incorrect pinion gear installation may

not be detected in testing.", and then required, "Install pinion gear IAW

Attachment 1." Attachment 1 was found to be an elementary drawing of the

pinion gear configuration (gear shoulder related to the motor), with no

detailed information concerning the gear key (when applicable), set screw, nor

lockwire. A review of the Limitorque Instruction and Maintenance Manual

(SMBI-180D), SMB-0 to SMB-4 & SMB-4T, Reassembly, determined that Step 5.

required, "When reinstalling the motor pinion, pc #40, insure it is a tight

fit on the motor shaft (preferably a light press fit). Note that the SMB-0

motor pinion is installed with the set screw lockwire between the gear teeth

and the motor flange.

On the SMB-1 through 4, the gear teeth are between the

set screw/lockwire and the flange.

11

Limitorque drawing 08-408-0001-4 showed

the gear configuration and the lockwire installation, but was not included in

the subject procedure.

It was the team's opinion that inadequate detail was included in the upgraded

procedure in that the procedure did not include the fundamental repair

information of the Limitorque maintenance manual. Since the repair manuals or

portions thereof were not noted to accompany work orders, and the above

information could not necessarily be relied on to be skill of the craft, it

was also the team's opinion that such information should be included int.he

appropriate procedure.

Similar increased attention to detail was requi_red in mechanical MOV

procedures.

As an example, procedure MMP-C-MOV-178, dtd 20 September 1988,

Removal and Overhaul of Limitorque Model SMB-000 through SMB-00 and SB-00, was

used to overhaul an SMB-00 type operator (WO 565294, Charging Pump to Regen Hx

Stop Valve, MK #02-CH-2289A).

Page 36 of the procedure included in the work

order contained an illustration of the disassembly of a SMB-000 shaft (vs

. SMB-00), and the part numbers identified in the illustration were specific

only to the SMB-000 operator. Additionally, the part nomenclature and

assembly sequence was different from the SMB-00 operator. Although the

procedure was revised on 21 October 1988 (during the performance of the

maintenance), the illustration of the SMB-000 actuator remained unchanged on

page 37.

Both revisions of the procedure were used in WO 565294 to perform

the actuator overhaul, including replacement of the hypoid gear which was most

affected by the procedure error.

The above procedural problems were perceived by the team to be examples of

weaknesses in the procedure preparation program.

..

54

3.o

CONTROL OF ELECTRICAL WORK PRACTICES

Electrical work practices:*

Design Change Packages 8532 and 8534 (U-1 and U-2 respectively) installed new

station batteries (vital 120 volt DC power distribution system).

The team

noted during system walkdown inspections that several conditions fn the

batteries were not consistent with installation specifications. For example,

vital batteries lA and lB each required two conduit installations for

intercell cabling between battery halves. Neither cable conduit was grounded

in either battery.

The team reviewed NUS-2030, Specification for Electrical

Installation for Surry Units 1 & 2, Category Safety Related, and noted that at

paragraph 6.9, '.'Metal conduit systems shall be grounded by copper cable

connections to the ground grid, grounded cable tray system, or to building

steel using tinned copper lugs."

Vital battery lA, cell 30 to 31 intercell connector, exhibited two 1/4"

terminal bolts that did not not appear to have been properly tightened. The

lockwasher under the nut was not compressed (1/32" gap) against the Belleville

washer.

The support platform (battery rack) for vital battery lA was noted to have two

loose bolts (lockwasher not compressed and locking) on the end piece at cell

18.

The same condition was observed at cell 43.

Based upon the above observations, the team concluded that electrical work

practices and control, and the attendant quality control inspections, were an.

implementation weakness and required strengthening.

In each of the above

cases, the licensee immediately initiated the appropriate station deviation

report and work request to correct the adverse condition.

Material conditions of panels:

During system walkdown inspections of 120 volt AC and DC vital panels, the

inspection team noted occasions of dirty panels and poor material conditions

as follows *

While observing troubleshooting for ground faults in the vital 125 volt DC

distribution system, the team noted that the U-1 480 volt and 4160 volt

Breaker Test Panels (fed from 120 volt vital DC via cable 1B63) contained

significant amounts of trash and dirt. Numerous parts bags, several screws,

cut tie wraps, and cut wire ends were found in.the 480 volt test panel.

The

insulation on one wire run was badly frayed, not tie-wrapped, and not properly

stood off the panel frame to prevent chaffing on the door edge (after

repositioning the wire run, the ground improved significantly). One of two

hinge pins in the panel door was missing.

An un-numbered "Rod Control Signal Circuit" connection box, immediately

adjacent to the U-1 480 volt Breaker Test Panel, was missing one of four cover

screws in the gasketed cover plate, thus the box was not properly sealed

against moisture intrusion (immediately corrected).

55

U-1 125 volt DC distribution '.panel lB was extremely dirty and contained cut

tie wraps, discarded danger ~ag pieces, and numerous pieces of conduit Sealant.

Several improper nuts were used (acorn type nuts apparently lost) to hold the

cover panel captive. Panel 1-2 (125 volt DC.fed from lB) was also dirty,

containing screws, cut cable ties, tape, cut wires, and a case (panel) nipple

on top of the panel.

Similar conditions were observed in the 120 volt AC vital panels. Vital Bus

Distribution Panels 1-IA and 1-IIIA were found dusty (from concrete), dirty

with metal shavings from drilling, and contained cut tie wraps and wire ends.

In addition to dirt, Vital Bus 1-1 contained numerous pieces of excess

expandable foam (fire barrier material) in the panel bottom.

Panel 1-1 was

also missing two knockout plugs (one top, one side); Semi-Vital Bus Dist Panel

lSVBl was missing a knockout plug from the panel top.

Based upon the above observed material conditions, the team concluded that

the licensee's program and implementation for control of housekeeping during

electrical maintenance in 120 volt AC and DC panels constituted a program

weakness.

Improper terminations:

In addition to the adverse material conditions observed, the team noted several

instances of improper terminations as follows.

The U-1 480 volt Breaker Test

Panel (fed from vital 120 volt DC) had two wires in one wire run with lugs

bent over more than 90 degrees of crimp angle.

U-1 125 volt DC distribution

panel 1-2 had several unlabeled spares that were not correctly terminated

(bent over, loose wrap of electrician's tape). Several examples were noted of

excessive insulation cut back at the breaker feed cable (1/4

11 vice required

1/32

11

) and not all wire strands captured under breaker terminating clamp

(e.g., 125 volt vital DC panel 1-2, breakers 11 & 15, and 120 volt vital AC

panel 1-IIIA, breakers 3 & 5). Excessive insulation cut-back in 125 volt

vital DC panel lB had permitted the wire strands on one feed cable to splay

apart, potentially permitting strand breakage. Although the team recognized

that some of the above conditions may have existed since the plant was

constructed, there were also occasions of modifications that could have

permitted correction of the deficient conditions. Current practices were

clearly defined by the requirements of NUS-2030, Specification for Electrical

Installation for Surry Units 1 & 2, Category Safety Related in Secti,on 5, that

addressed termination requirements.

Based upon the above adverse conditions,

the team concluded that the licensee's ~rogram and implementation for control

of standard practices during electrical maintenance in 120 volt AC and DC

panels constituted a program weakness.

3.p. PERFORMANCE OF MAINTENANCE TRENDING

The team noted that the primary vehicle for reporting conditions adverse to

quality was the Station Deviation Report. Procedure SUADM-LR-13, dtd 29 Dec.

1989, Station Deviation Reports, provided detailed instructions for initiating

and processing Deviation Reports.

The team noted that the system was applic-

able to both safety related (SR) and non-safety related (NSR) components,

-*

56

systems, materials and services. The team's review of program criteria and

actual deviation reports (DRs) indicated that the threshold upon which station

personnel based their decision to prepare deviation reports was low.

During

1989, approximately 4,000 reports were initiated and evaluated.

Root Cause

Evaluations (RCE) were initiated based upon the significance of the deviation

report.

The licensee had recently implemented the Component Failure Evaluation (CFE)

Program (SUADM-M-48, dated 29 December 1989) as a direct adjunct to the

station deviation reporting system. A station deviation would normally set

the CFE program into motion for all SR component failures (except MOVs which

were handled separately under the MOV program).

NSR component failures were

evaluated under the CFE program when directed by management.

The CFE program

was designed to perform evaluations that did not require the extensive degree

of investigation necessary in the RCE program.

Since the CFE program had just been initiated, approximately twenty

evaluations had been completed.

The evaluati~ns were comprehensive, and

appeared to provide an adequate basis upon which a meaningful trending program

of component failures could be based.

Trending of station deviations has become more effective during the latter

half of 1989 with increased attention and staffing by the responsible station

group.

Monthly trend reports on station deviations have been issued that

depict major categories, such as procedure errors, personnel errors, EDGs,

pumps, and valves.

The level of detail in the trend graphs required

interested parties to review specific categories of deviations to formulate

plans of action.

Each monthly report also focused on the details of a

particular category; e.g., the December 1989 report, published in February

1990, focused on types of personnel errors for the whole 1989.

In the

maintenance area, 187 deviations were written, of which:

54 DRs involved valve maintenance

37 DRs involved pump maintenance

57 DRs involved procedures not processed correctly

82 DRs involved inadequate maintenance (e.g., wrong part, wiring not per

specification, not correctly installed)

Insufficient data history had been collected to permit trending of the above

types of data, although the licensee was moving in that direction. * Programs

had been implemented on a case basis in response to the initial DR(s) to

_correct the above types of problems - trending of the data could be used as a

tool to determine program effectiveness.

Based upon the above observations, the team concluded that the licensee had

implemented an adequate program for trending of maintenance problems.

Continued data collection and evaluation was required to determine the

program's effectiveness.

57

3.q. TOOL AND EQUIPMENT CONTROL

The team found three tool issue points in use; the primary tool issue point was

in the upper level of the condensate polishing building, and the contaminated

tool issue point (Auxiliary Tool Room) was located on the 45' level of the

Auxiliary Building. A third issue point was located outside the power block,

and was used as a construction issue point.

Tool inventory control was adequately coordinated between the issue points.

Cognizant division supervisors were responsible for notifying the tool room of

tool requirements.

The team observed evidence of supervisors properly fore-

casting tool requirements for forthcoming maintenance activities. Positive

inventory control resulted in approximately one or two demands per day for

devices that were not available at the issue point. A strength worthy of,.note

was the issue room's control of special tools and devices.

On each occasion

of acquiring, through manufacture or purchase, an unusual device or tool for a

special application, a photograph has been taken of the device, and its special

storage location annotated on the photograph.

Positive control of these

devices has been maintained, and workers seeking special tools simply have to

"thumb through

11 the book of photos until the desired device was located.

Rotating tools such as air grinders and electric drills were on a comprehensive

preventive maintenance program that included items such as lubrication checks

and ground fault checks.

Evidence of the checks having been done was attached

to the tool. The team noted that slings, cables, chain falls, etc. were also

subjected to periodic testing; distinctive colored tape was used to mark the

device when next due for test. A unique serial number for each device was also

attached by tag that permitted test tracking.

One nylon sling was noted by the

team to be in a ready-to-issue bin with no evidence of having been properly

tested.

Electrical and mechanical M&TE devices were found stored in a segregated area.

Positive control of issue and return was maintained, and included tracking of

the work order on which the M&TE device was used.

No effort was made by the

licensee to assure proper calibration of any returned device unless problems

were reported by the user. A complete engineering evaluation was performed

for the situation of a device failing calibration when tested at its normal

cycle. The team noted several micrometer adjustment type torque wrenches that

were stored in ready to issue bins at high torque values. The licensee advised

the team that manufacturer recommendations required setting the wrenches at 20%

of full scale, but these requirements had not been incorporated in procedures

governing operation of the tool room, in torque wrench calibration procedures,

nor posted in conspicuous locations where wrenches were stored. Manufacturer's

recommendations concerning exercising torque wrenches prior to calibration were

also not included in procedures.

Based on the above observations, the team

concluded that although the overall program for tool issue and control was

adequate, several program and implementation improvements should be considered *

'.

58

3.r. SUPPORT INTERFACES

One of the most detrimental conditions concerning the performance of mainte-

nance is the lack of coordination between the support interfaces. Maintenance

support interface problems are predominately with the following departments:

Planning, Scheduling, and Operations.

In addition, the material control for

parts is an area that significantly contributes to the Maintenance Department's

backlog.

In these areas, the team did not find fault with any individuals or

. managers.

The personnel were cooperative with each other and recognized the

limitations of the programs within the plant.

Planning

The team concluded that the Planning Department was not capable of performing

all planning functions independently because it was not adequately staffed.

Consequently maihtenance personnel, foreman and craft, have the added burden

of supporting planning.

For example, there is one permanent electrical

planner, one or two contractors, and two borrowed electricians to perform all

electrical planning.

The permanent _electrical planner spends considerable

time coordinating with scheduling and at the plan of the day (POD) meetings.

Since maintenance personnel support detail planning, this contributes to the

problems of maintenance. Maintenance personnel are required to interface more*

and are not performing their primary function of doing corrective and

preventive maintenance.

In addition, the team does not consider the

Maintenance Department adequately staffed to perform their main function,

reduce the backlog, do detail planning,*and provide craft to the Planning

Department.

Other planning issues are discussed in Section 3.g. and 3.m.

Schedulina

The team did not identify any specific problems with the Scheduling Department

or the schedule.

However, it was recognized that the "plant wide single train

method" is not used for scheduling.

The plant wide single train method is:

only work on equipment in a specific single train is scheduled for a specific

time period (day, etc.).

No other work in the other train(s) is allowed.

Since the licensee does not use the single train method, Operations has more

of a burden in controlling and approving work.

The main problem is schedule adherence.

The Scheduling Department provides

30, seven, four, and one day schedules.

Every day there is a POD meeting for

the next day's scheduled work.

During the periods when the team was onsite,

approxi'mately 40 percent of the POD scheduled work was actually started or

performed. * The licensee's statistics for POD effectiveness is as follows:

Month *

Department

Percent Effectiveness

--

Dec.

14%

Dec.

Electrical

64%

Jan.

22%

Jan.

Electrical

68%

,.

Month

Jan.

Feb.

Department

Mechanical

Electrical

Mechanical

59

Percent Effectiveness

61%

55%

74%

41%

Maintenance personnel stated that they were frustrated with the large number

of last minute cancellations of work already scheduled and approved at the POD.

During the inspection period, the team did not find any evidence that this

situation would be corrected.

Operations

The Operations Department is responsible for approving maintenance activities.

Through observation and discussions with maintenance personnel, the team

identified the method used to implement this approval as the major "bottleneck'

1

in the coordination of planning and scheduling maintenance. A shift supervisor

located in the control room annex reviews and approves work activities that

have already been scheduled during the POD.

Included in these duties are

reviewing and approving revisions to work orders and procedures. These* duties

of the shift supervisor are necessary for safe plant operation. However, the

method and time of implementation was found to be inefficient. The plant

work schedule for maintenance craft is from 7:00 a.m. to 3:30 p.m. daily

except for a skeleton staff on back shifts. Operations, including the shift

supervisor, isolate themselves during the period of 7:00-8:30 a.m. and

11:00 a.m to .1:00 p.m. daily. During this "quiet time" (3-1/2 hours) daily,

all maintenance activities requiring support and/or approval from Operation

before proceeding is placed on hold.

During the 7:00-8:30 a.m. "quiet time,"

POD scheduled items may be "deferred" and maintenance is not made aware of this

until after 8:30 a.m.

The same holds true for the 11:00 a.m.- 1:00 p.m. "quiet

time" period. Operations has assigned a SRO representative to coordinate with

planning and scheduling (POD).

This SRO interface does not appear to function

very well since the shift supervisor at the control room annex can override the

POD.

Discussions were held with Operations concerning the coordination problems

already discussed. Operations is aware of these problems, wants to correct

them, and plans to assign more li~ensed operators to each shift as soon as

they become available (late 1990).

Conclusion

The team concluded that coordination problems between the departments must be

resolved for the licensee to implement an effective maintenance program and

reduce the large backlog. Material control problems have been previously

identified by the NRC and INPO in other reports.

The team was told that lack

of spare parts continues to be a significant problem area *

..

60

3.s. LICENSEE REVIEW OF COMPLETED WORK RECORDS

To-evaluate the licensee's review of completed work documents the team

examined four completed WOs, evaluated the QA review process and reviewed the

the procedure for the QA review.

The team reviewed four completed WOs (79205-01-CH-MOV-12750,

535128-02-CH-MOV-2275A, 83923-02-SI-MOV-2867C, 65954-01-RS-MOV-1558 and

72503-02-CH-MOV-2289A) related to the work performed on Motor Operators.

WO 72503 documents the work activities performed on 92-CH-M0-2289 (charging

pump to regenerative heat exchanger stop valve).

Work performed was documented

on Procedures MMP-C-MOV-178 dated September 20, 1988 and October 21, 1988.

One

of items replaced was the

11Hypoid Gear.

11

The Hypoid Gear is listed as part

number 12 in the Limitorque vendor manual for SMBOO.

However, the above

procedures, on page 36, illustrated the assembly of SMPOOO of drive shaft and,

therefore, the replacement of the Hypoid Gear was not adequately captured i.n

the records.

The cognizant maintenance support engineer gave sufficient expla-

nation to the inspection team and produced evidence to support that the Hypoid

Gear was in fact replaced and the completed WO inadequately documents the

replacement.

QA reviewed this WO and determined it complete and acceptable.

The team reviewed the completed WOs related to the replacements of solenoid

operated valves (SOVs) to extend the longevity of the qualified lives of the

equipment.

The useful life of the SOVs were calculated to be 42 years taking

into consideration the environment (temperature and radiation) in which it *

(the SOV) is located. The following packages were reviewed: WO 71612;

WO 71613; WO 71564; and WO 71567.

WO 71612:

Replaced the 02-MS-S0V202A - In Attachment 7, the WO# was left

blank.

The serial number of the replaced valve is documented as 77826T-2.

The

same number is documented on Page 12. This may be erroneous, because the team

observed that this was the serial number of the SOV installed on 02-MS-SOV-202B

during the performance of periodic testing of the Auxiliary feedwater turbine

on March 16, 1990.

Copies of Attachment 7 are circulated to corporate and

other management officials to denote the completion of EQ related work *

W0-71613:

Replaced the 02-MS-SOV-202B.

On Page 12 of the WO and in

Attachment 7, the serial number of the SOV is indicated as 778262t-l. The

team observed that the serial number on the SOV was 778262t-2 during the

observation of the periodic test of the Auxiliary Feedwater Turbine on

March 14, 1990.

WO 71564:

Replaced the 02-CC-SOV-205A.

On Attachment 7 of model number of

the SOV was not documented.

61

WO 71567:

Replaced the 02-CC-S0V-205C.

On Attachment 7 of the Model number

of the SOV was not documented._ A comment on the cover sheet stated that the

actua 1 stroke of the va 1 ve is l/8

11 1 anger than the design.

EWR wi 11 be

written. However, a EWR number was not identified.

The team reviewed three completed WOs documenting the work performed on NAMCO

type limit switches to extend the longevity of the qualified lives.

WOs

02-MS-2S2-201A, -201B, -ZS1-201C documented the work performed on the limit

switches on the mainsteam trip containment isolation valves.

No adverse

findings were identified in this area other than the observations in the

following paragraphs.

Evaluation of the QA Review of WOs

QA reviews 20% of the completed WOs utilizing procedure QADI 35C, Revision 4,

dated February 1, 1990.

The rest are stamped

11Noted

11 and filed. This review

is in addition to QC in-process verifications. Review of this procedure

including Attachment 6.1 indicates that QA is not required to scrutinize a

completed WO to ascertain if it is technically complete in all respects.

The

team discussed with the QA Manager and his staff the above findings. The QA

Manager was responsive to the team concerns that technically incomplete and

incorrect WOs may be erroneously accepted as acceptable documents.

The QA

Manager assured the inspection team that he will consider implementing the

following measures:

Develop an adequate checklist which will reflect the salient steps

verified.

Provide sufficient input to the

11Procedure

11 writing group to retain only

the relevant information in a given procedure which is applicable to the

specific work being performed

Train QA reviewers to review the first few completed WOs prior to

resorting to sample reviews.

Examination of QA Procedure for the QA Review.

The team reviewed several completed WOs including those identified in above

paragraph in this section and observed the following weaknesses:

The procedures contained more general

11nice to have" than specific "need

to have

11 information, thus making the WO n10re voluminous.

The description of the work performed (on the cover sheet) did not

adequately describe the accomplishment *

62

If cable splicing was required, the WO should document that the cables

were crimped, bolted or terminated.

The engineers should determi.ne the

size of the conductors and specify the connection method.

If* replacing a SOV, the WO should state if the existing CONAX connector

was retained, or a new CONAX connector was used.

Attachment 7 for SOV WOs are used by licensee management to demonstrate

that EQ is current.

As such, the information on Attachment 7 should be

complete and accurate.

Procedures MMP-C-MOV-178, dated September 8, 1988 and October 21, 1988,

used to reflect work performed on SMPOO type Limitorque operators should

have a note on pages 3b and 37, respectively, to indicate-that the

illustration (SMPOOO) provided on those pages were incorrect.

The summary in WOs for limit switch (LS) related activities should

precisely state if the LS was replaced or if only the gasket covers were

replaced.

The exact torque valve to which the bolts (screws) were

torqued to should be stated.

4.

Evaluation of Maintenance

Summary Rating of Maintenance Process

Program:

SATISFACTORY

Implementation:

SATISFACTORY

The evaluation completed by the NRC team rated both the Surry maintenance

program and its implementation SATISFACTORY

The Surry maintenance rating was obtained by collating and assessing the

maintenance team inspection findings iri a special maintenance inspection logic

tree. The tree completed for Surry maintenance inspection is depicted in

Appendix 3.

The tree divides maintenance evaluation into three "parts" (I,

II, and III). The parts are divided into eight "areas" (1.0 through 8.0) and

the areas into individual maintenance topics or "elements" (1.1, 1.2, 2.1,

etc.). Based on their inspection findings (negative and positive), the team

established ratings for most of the elements. Subsequently, area ratings were

determined based on associated element ratings; part ratings based on the

associated area ratings; and, finally, a total_maintenance rating was

determined from the ratings for the parts. The team did not weight all

findings or ratings equally. *

.

)

63

Four rating categories were used and a color was assigned to each to aide in

displaying the ratings on the maintenance inspectiori tree. The rating

categories were as follows:

11Good

11 Performance (Green)

11Satisfactory

11 or

11Adequate

11

Performance (Yellow)

11Poor

11 Performance (Red)

(Blue)

Overall, better than

adequate, shows more than

minimal effort; can have a

few minor areas that need

improvement

Adequate, weaknesses approximately

offset by strengths

Inadequate or missing

Not evaluated or Insufficient

Information to Evaluate

Each part, area and element, as well as overall maintenance, is represented by

a block on the tree. Most of the blocks are split into two parts with the

upper portion representing program or process and the lower half representing

implementation.

The exception is for the part I blocks which are not

considered to have separate programs or implementation.

The parts and areas of the maintenance inspection tree are described below.

The inspection findings that contributed to the ratings are also given.

Individual element ratings are not described but are shown in Appendix 3.

4.a. Overall Plant Performance Related to Maintenance (Tree Part I)

Rating:

Satisfactory

This part of the tree is an overall assess~ent and rating of maintenance

through direct measures:

Its rating was based on the SATISFACTORY rating

determined for

11direct

11 measures below *

.. 4.a(l)

_Direct Measures (Tree Area 1.0)

Rating:

SATISFACTORY

The direct measures used to assess this area are the plant historic data

(tree element 1.1) on the performance of the operating units and observations

of housekeeping and material conditions observed in walkdown inspections

(element 1.2).

Historical Data Related to Maintenance

64

The team examined historical data for licensee's maintenance and maintenance-

related activities. The data covered the period from the beginning of 1989 to

present, plus some 5 year trends. The data reviewed was included in the

. Virginia Power Nuclear Performance documents issued on a monthly basis. The

data reviewed does not provide a totally accurate picture of the plants today

since both units returned to service from extended outages in the period

(Unit 1 - 7/89 and Unit 2 - 11/89). The results are mixed (some good and some

poor). The equivalent availability was poor at the beginning of 1989 and good

at the end of 1989 and to date in 1990 with a declining 5 year trend. The

forced outage rate was poor for 1989 and good to date for 1990 with an upward 5

year trend. The number of unplanned reactor trips was good to date for 1990

with a mixed (declining for Unit 1 and upward for Unit 2) 5 year trend. The

number of ESF actuations was mixed (good .for Unit 2 and poor for Unit 1) for

1989 and good for both units to date for 1990.

Radiation exposure was poor for

1989 and good for the beginning of 1990 with an improving 5 year trend. The

corrective maintenance backlog was excessively high for 1989 and appeared

to continue trending upward for the beginning of 1990.

In review of licensee

reports, i.e., LERs and r~ctor trip reports, for indicators such a unplanned

trips and ESF actuations, there was no conclusive evidence indicating

maintenance problems contributing to the events.

The findings of the team's evaluation of the historical data and the walkdown

inspections consisted of numerous minor discrepancies.

Examples are noted in

Section 2.a,b,c,etc. above.

On the basis of the number and significance of

the discrepancies observed the team rated the area SATISFACTORY

4.b Management Support of Maintenance (Tree Part II)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This part of the tree is an assessment and rating of areas of the tree which

represent management's support of maintenance through their commitment and

involvement, organization and administration, and provision of technical

support for the maintenance process.

4.b(l)

Management Commitment and Involvement (Tree

Area 2.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

65

This area consists of upper management's direct encouragement and promotion of

improvements in maintenance.

The rating of the program and implementation was

based on the following findings of strengths and weaknesses:

The STRENGTHS noted in this area included:

Management commitment to industry initiatives is good

MOV program is good,

System engineer program very strong (2.e)

Manager of engineering performed weekly walkdowns with a system engineer

(2.d)

The WEAKNESSES noted in this area included:

Problems identified in NRG Information Notices and industry communications

not available to working level people

Goal achievement not good (performance indicators)

Recognition and response to adverse performance indicators slow

Response to QA findings not always timely (2.c & 3.e)

Observations in this area included:

Reliability centered maintenance program not embraced

Self assessment performed late

189

Plant participating in EPRI plant aging program

Plant PM upgrade program

Procedure Upgrade Program

MOV program upgrades

Training Program is INPO accredited

4.b(2)

Management Organization and Administration

(Tree Area 3.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This area consists of management's support of and involvement in the control

of maintenance through developing and implementing a maintenance plan setting

goals and policies; allocating resources, defining maintenance requirements,

monitoring performance, providing document control and determining the need

for improvements to plant material condition. Its rating was based on the

following strengths and weaknesses:

The STRENGTHS noted in this area included:

Instructors from NTD required in plant

Supervisory span of control is good (2.a & 2.h)

NO sacrifices of safety are made to meet production goals

'

,.,

..

66

Predictive maintenance, oil and vibration analysis is good (2.a)

Post maintenance debriefs addressed possible additional requirements

Post maintenance debriefs for I&C group effective in identifying problems

and carrying forward to subsequent maintenance.(2.e)

Verbatim compliance by I&C technicians absolute - many changes had to be

initiated to make procedures fit the tasks. (2.e)

Corporate driving force in replacing SOVs to maintain EQ currency, and

causing existing MOV re-inspection

The WEAKNESSES noted in this area included:

Some systems such as IA and HVAC are not goal oriented for required

system improv_ements - evidenced by lack of commitment (3. f)

No visfble evidence of special maintenance department goals and

objectives.

Elect/Mech planning does not have adequate staffing, planning a function

of the craft (2.e, 3.g, 3.r)

Maintenance manpower inadequate (especial HVAC) (3.f)

Periods of no shift coverage by I&C - call in (2.e)

Maintenance often delayed due to non-availability of material (3.r)

Corrective maintenance backlog trend is upward (3.m)

Mairitenance often delayed, not supported, due to "quiet time" in control

room (3.r)

I&C manning not able to support full implementation of PM program

Current PMT program very limited for equipment other than ASME Section XI

(3.h)

Procedure for Predictive Maintenance is outdated (2.a)

Some electrical PMs deferred due to lack of manpower (2.d)

Appropriate Vendor PMT not accomplished (3.i)

Vendor tech manual requirements not implemented (AFW, motor starters PMs,

containment instrument air dryer filter) (2.d & 3.c)

PM program for HVAC limited (2.g & 3.f)

PM program for instruments other than Tech Spec required items was not

prepared or performed (e.g., EOG safety functions for alarm in CR,

pressure gage on N2 back-up to IA for AFW MS inlet FCV (3.1)

Persons specifying electrical PMT are not qualified (3.h)

Persons specifying electrical PMT inconsistently require varying PMT (or

none at all) for the same equipment. (3.h)

ASME requirements related to check valve performance have not been

incorporated in IST program (MS to AFW & IA)

PM deferral rate for

189 was not correctly counted to. provide

representative information. (3.m)

Drawings not updated following plant modifications (3.b)

Drawings not equal to plant as-built (2.a & 3.b)

"For ref only" drawings used in Trouble Shooting, preparation of procedure

revisions, and plant operations decisions (Rx protection system relay

fa i 1 ure) * ( 2. e)

2 occasions of Procedure Action Request (procedure change) not completed

accurately (1 time full,y signed off/1 time error detected by Shift Sup

prior to implementation). (3.b)

..

~---

-

67

Mods to plant procedures and equipment are not considered in a timely

manner (3.f &3.1)

Procedure for preparation and completion of Work Orders*does not contain

adequate detail.(3.g)

Resistance to dedicating adequate funds to replace/adequately repair

system *components (use band-aid repairs too frequently) (3.f)

OPs (operations) control of maintenance shifts work priorities too easily

Observations in this area included:

Maintenance plan has been prepared and in effect

The utility has been forced to utilize contractor support for maintenance

dept and system eng support because of inability to maintain permanent

staff

Thermography predictive analysis program not initiated

PM program undergoing Task Force review

Source documents that could change maintenance requirements are

problematically covered .

. Trend INPO performance indicators

Evaluating each SR work order for root cause problems (Component Failure

Evaluation Program - CFE), initiated Jan 90. Should be an effective

program.

Bi-annual review of maintenance procedures was accomplished (occasionally

not effective, inst. cal proc)

Computer based (WPTS) document control system is used for work orders.

1,7,30 day look ahead attempting to coordinate dept activities.

4.b(3)

Technical Support (Tree Area 4.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY.

This area encompasses the various elements of technical support that are

needed for maintenance to function effectively (e.g., engineering support,

health physics, QC, risk assessment, etc.). Its rating was based on the

following strengths and weaknesses:

The STRENGTHS noted in this area included:

Close liaison existed between system eng and craft.(2.d & 2.e)

Internal/corp. communication channel provided timely exchange of info

Strong involvement from system engineers in their respective

systems (2.a & 2.e)

68

EQ program well supported by corporate and engineering

MOV/Check valves/PM programs good eng support

Procurement and maintenance engineering must approve each part

procurement action.

Elect procedures include many hold points for QC sign-off (2.d)

Performance inspections by QA people (Quality Performance Group)

effectively augmented the maintenance process

Worker awareness of ALARA concerns was high

Advanced Rad Worker quals for QMT workers provided the ability to

perform work that otherwise would have required continuous HP coverage

Recovery of contaminated areas strong program

HP aggressive in identification of sources of contamination

Craft exhibited strong safety consciousness.

Overall HP program continues to improve

Licensing group was strong

Commitment tracking system was good

The WEAKNESSES noted in this area included:

Communications between craft (maint. dept.) and operations were not

optimum for completion of scheduled maintenance activities. (3.m)

Risk assessment was not considered in prioritization of work orders,

however at the implementation level, consideration was given

to safety risk for not accomplishing work. (3.m)

QC procedure for review of completed work order packages is inadequate

Generic letter commitment on Instrument Air poorly handled

Observations in this area included:

Current PMT program is written exclusive of most tests for non-ASME

Section XI equipment.(3.h)

PM program for HVAC limited

PMT not well defined (3.h)

Persons specifying electrical PMT are not qualified (3.h)

Engineering has not caused all PM requirements to be identified

(e.g., I&C) (3.1)

Many contractors to support engineering

System engineers located outside PA

Engineering analysis via Component Failure Analysis initiated Jan '90.

QMT members provided a high degree of quality control awareness, buf

quality control was frequently exercised by peers.

QC notified at start of all SR work orders

Guidance available for when QC inspection required.

Audit of maintenance dept on bi-annual basis

A

69

4.c. Maintenance Implementation (Tree Part III)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This part of the tree is an assessment and rating of the work, organizational,

hardware and personnel controls necessary to proper implementation of

maintenance.

4.c(l)

Work Control (Tree Area 5.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This area encompasses assessment of important elements of work control through

evaluation of maintenance in progress, work order control, planning,

scheduling, prioritizing, etc. Its rating was based on the following

strengths and weaknesses:

The STRENGTHS noted in this area included:

Good technician skills, knowledge, and abilities (2.e)

Good command and control of Periodic Te~t evolutions (2.d & 2.e)

Craft complied with procedures, and were attentive to necessity for

Procedural PARs

All craft assured task qualification proper prior to job start (2d)

Equipment history was readily available/accessible in the WPTS system

for post

184 events

Upgrade program for procedures was generally effective in improving

procedures when compared to pre-upgrade

PMT on periodic tests conducted by the Ops group were good.

The WEAKNESSES noted in this area included:

Clearance program could be improved in tagged position/restoration

position

Coordination with OPs for work accomplishment was poor (3.r.)

Completed work orders for replacing S0Vs to maintain EQ exhibited errors,

omissions and were cumbersome

Work orders occasionally lacked detail, relied heavily on skill of the

craft. (2.a, 2.d & 2.e)

Definition of special tools, special equipment, special people quals

was a function of procedural adequacy, not the work orders. (3.k., 3.g.,

3.n.

..

70

Mech/Elect/I&C planning did not have resources to perform detailed

planning - left up to the craft (2.e & 3.g}

I&C planning not integrated with the maintehance dept. - work order inst

were prepared by revising existing alignment procedures. (2.e)

Work instructions associated with the work package were in general very

limited for E and M craft.disciplines.

Programmatic prioritization scheme was time to complete dependent, not

safety of people, safety of equipment, or Tech Spec related. However,

implementation of priorities appeared to take the above types of

considerations into account.

Notwithstanding, many high priority work

orders were years old. Bottom line - no effective prioritization scheme

in effect.

Provision for modification of work item prioritization difficult once

assigned in WPTS, thus assigned priority may have been in error, but not

corrected (3.m)

Risk assessment was not considered in prioritization. (3.m)

Numerous Priority 1 work items were very old - driven by the

program ( 3 .m.)

Backlog of CM work orders has continued to increase in all categories, and

therefore backlog control must be ineffective (3.m)

Backlog CM average age is very old (3.m.)

Deferred PMs only recently brought under control (2.d)

Many (old) procedures did not specify required test equipment, or

tolerances, or acceptance criteria, or special tools.

Maint procedures (WO instructions) for replacing SOVs and MOV related

work were voluminous and cumbersome - not tailored to the specific

application and contained errors.

AFW turbine driven pump PT referenced the wrong Tech Spec requirements

(wrong para.)

Several fundamental instrument alignment procedures had not been initially

prepared and approved (PRT Temperature).(3.1)

.

Elect. Procedures frequently asked Ops to perform a step - if not

possible, N/A the step, (2.d)

I&C PT procedures did not adequately establish I/Cs, reference Tech Specs,

nor provide for exiting procedure in event of equipment failure (2.e)

Procedure for preparation and completion of Work Orders does not contain

adequate detail. (3.q}

Elect PMT was cancelled due to lack of manpower and the PMT was not

carried forward as overdue (2.d)

Current PMT program very limited for equip other than ASME Section XI

(3.h)

Persoris specifying electrical PMT are not qualified (2.d & 3.h)

Persons specifying electrical PMT inconsistently require varying PMT (or

none at all) for the same equipment.(2.d & 3.h)

QC review of completed work orders were not effective in that the reviews

focused only on detecting missing information, not incorrect information.

In some cases, WOs with incomplete information were not detected.(2.e)

Incomplete and erroneous work orders on EQ SOV replacements

Work accomplished lacked adequate description (2.d)

,,.

71

Observations in this area included:

Foreman on the job presence good

Craft were attentive to resolution of problems encountered, but resolution

took inordinate amounts of time

Lifted lead/jumper program and implementation adequate.

Clear provisions for accomplishing emergency work

WPTS computer system provides adequate equip history for items of less

than 5 years age. *

Licensee had to go to off-site facilities for documents related to ASME

Code vessels (i.e, State of Virginia)

Coordinated scheduling of functionally related equipments is not

effectively performed for all disciplines.

1, 7, and 30 day look aheads attempt to perform Work Order coordination

Procedural upgrade program was improving quality of formal maintenance

procedures.

Numerous, minor administrative documentation errors in Work Orders were

detected that were not of safety significance.

4.c(2)

Plant Maintenance Organization

(Tree Area 6.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This area encompasses the processes used by the maintenance organization to

control, support and direct maintenance activities. Its rating was based on

the following strengths and weaknesses:

The STRENGTHS noted in this area incl~ded:

All craft and foreman demonstrated good job skills (2.a, 2.d & 2.e)

Procedural adherence was good (2.e)

Unanticipated work problems promptly resulted in work stoppage, and

resolution initiated. (2.e)

Contractors in I&C area were required to meet same qualification

requirements as Surry people, including JPMs to perform plant work.

Contractors were under direct supervision and control of Surry

supervisors.

Contractors were required to operate under all applicable Surry

procedures.

Work request program was an easy, efficient means for any plant person to

identify and report deficient conditions, and was effectively used by most

plant personnel.

Station Deviation reporting system had a low threshold for reporting of

deviant conditions.

Review process provided for substantive evaluations.

(3.p)

System engineers were a visible support to the craft.(2.e)

,-

..

72

The WEAKNESSES noted in this area included:

120 v AC, 125 v DC vital panels contained trash, exhibited adverse

material conditions, and had improper cable terminations. (2.e & 3.o)

Vital station batteries exhibited some poor electrical work practices

after mods. (2.d, 2.e & 3.o)

PM program for HVAC _limited (2~g)

NSR 480 v panels contained excessive dirt and trash. (3.o)

Electrical foreman required to perform work order planning, thus reducing

- job site supervision and ability to initiate productive effort. (2.d)

Deficiencies such as leaks and missing fasteners not identified by

licensee.

(2.a & 3.d)

Deficiencies identified, but not tagged. (2.a & 3.d)

Tagging procedure is weak because of non-definitive requirements (3.d)

Operations support of work order initiation/commencement was inefficient

for at least three hours per day, including from 1100 - 1300. (3.r)

Coordination with OPs for work accomplishment was poor (3.r)

Operations group did not support all attributes of the M&TE program. (3.k)

Observations in this area included:

Contractors were utilized as system engineers.

Necessary clearances and work qualifications were not effectively

planned

I&C contractors met few job qualification requirements, therefore they

were on site for several months before they became an effective

contribution to the I&C group - matter of fact, they were a burden

(e.g., escort requirements)

Program is in its infancy for most maintenance trending programs (3.p)

CFE program initiated in Jan 1990 (3.p)

4.c(3)

Maintenance Facilities, Equipment and

Material Controls (Tree Area 7.0)

Rating:

Program:

SATISFACTORY

Implementation:

SATISFACTORY

This area encompasses the plant maintenance facilities, equipment and material

controls with regard to the part they play in supporting the maintenance

process. Its rating was based on the following strengths and weaknesses:

The STRENGTHS noted in this area included:

I&C, Mechanical supervision co-located with the craft.

Procedures, technical manuals, and M&TE were co-located with the I&C craft

Individual, lockable work space was provided for each I&C technician.

73

The M&TE lab was located in the I&C spaces - was clean and well

maintained, well organized,,but small. (3.k)

NTD laboratory facilities very effective, but not all planned equipment

installed.

Material storage was observed as a strength

The WEAKNESSES noted in this area included:

Material control program was informally implemented by checklist.

Increased material usage scrutiny (material dedication program) caused

excessive delays in initiating maintenance activities, e.g., HVAC. (2.g &

3. f)

,

.

Material procurement system was extremely cumbersome.

Inventory of spare parts in some cases (HVAC) was zero. (3.f)

Licensee unable to perform calibration of contaminated M&TE. (3.k)

Storage and recall program/implementation very weak for Ops. (3.k)

Range and characteristics of M&TE not readily available to technicians

(problem because the procedures do not contain required instruments, and

the techs have to perform the evaluation). (3.n)

Test instrument racks secured by large chains to meet seismic restraint

requirements could damage delicate instrumentation. (3.k)

Observations in this area included:

Welding and ma~hine shop areas were found to be small.

Welding of consequente was farmed out to contract personnel.

Mechanical hot shop not available, temporary tents set up for that kind of

work.

Tracking of M&TE used documented in work orders.

Many items of M&TE had a record accompanying the device showing its use.

(OPS and I&C)

Evaluation program in place for M&TE failing calibration *

. M&TE Entire program run manually by one person with no --trainee.

4.c(4)

Personnel Control (Tree Area 8.0)

Rating:

Program: GOOD

Implementation: SATISFACTORY

This area encompasses staffing controls (.personnel policies, turnover

minimization, shift coverage, etc.), training, testing and qualification and

the overall current status of personnel (actual turnover rate, extent of

.,

74

personnel trained and qualified~ drug problems, etc.). Its rating was based

on the following strengths and weaknesses:

The STRENGTHS noted in this area included:

Range and span of control of supervision was good.(2.a & 2.e)

Crews are maintained stable (roll together on watch)

Training program INPO accredited.

Strong apprentice training program.

Strong feedback program on training effectiveness.

Required quarterly hours for instructors in plant was high.

Apprentice motivation was high to complete training steps (increase

in pay, peer approval).

The WEAKNESSES noted in this area included:

I&C tech coverage not available 24hrs/day (2~e)

Staff not at full complement for I&C and electrical (2.e)

Backlog is high for mechanical and electrical (2.d & 3.m)

Many craft members pulled for Temporary Assignment (2.d)

I&C not totally integrated with the Maint Dept (2.e)

Craft not always in compliance with security measures (tailgating). (3.j}

Observations in this area included:

Craft foreman could readily (and did) determine individual qualification

status for craft assignment to particular job tasks. (2.d)

Use of contractors in the I&C area to cover staff shortfall was partially

effectiv1=.

10% of I&C production time was devoted to recurrent training.

Qualification process included oral, written, and performance

examinations.

5.

Exit Interview

The inspection scope and results were summarized on April 12, 1990, with those

persons indicated in Appendix 1.

The team leader described the areas inspected

and discussed in detail the inspection results. Proprietary information is

not contained in this report. Dissenting comments were not received from the

licensee.

(Open) Violation 50-280,281/90-07-0l:

"Failure to Follow Procedures for

Maintenance" paragraphs 2.e., 3.b., 3.i., and 3.k.

(Open) Unresolved item 50-280,281/90-07-02:

11 EDG Day Tank Fuel Transfer Line

Analysis" paragraph 2.f *

y

Licensee Employees

APPENDIX 1

PERSONS CONTACTED

R. Allen, Operations Maintenance Coordinator

R. Benthall, Supervisor Licensing

M. Boling, Asst. Plant Manager North Anna

D. Christian, Asst. Plant Manager for Operations and Maintenance

J. Downs, Power Planning

D. Erickson, RP Superintendent

A. Friednman, Superintendant of Nuclear Training

E. Grecheck, Asst Plant Manager for Licensing

R. Gwaltney, Superintendent of Maintenance

M. Hadduck, Supervisor Electrical Maintenance

D. Hanson, Supervisor .of Maintanance Support

E. Harrell, Vice President for Nuclear Operations

D. Hart, Supervisor QA

M. Kansler, Station Manager Surry

A. Keagy, Superintendent Materials

H. Miller, Director Maintenance Support - Corporate

F. Mone, Planning

R. Saunders, Manager Licensing

R. Scanlan, Licensing Engineer

E. Smith, Manager QA

D. Snoddy, Supervisor Mechanical Maintenance

R. Thornsberry, Planning

G. Tompson, Suprervisor Maintenance Engineering

J. Winebrenner, Supervisor Procurement Engineering

F. Walking, Nuclear Operations Support

NRC Personnel

W. Holland, Senior Resident Inspector

C. Julian, Chief Engineering Branch

S. Tingen, Resident Inspector

Attended Exit Interview on April 12, 1990

.....

ADM

ANSI

B & PV

BKR

cal

CAS

CD

CFE

CFR