IR 05000289/1986003

Insp Rept 50-289/86-03 on 860303-27.Major Areas Inspected: Operational Readiness of Emergency Feedwater Sys.Seven Inspector Followup Items & 18 Potential Enforcement Findings Noted

ML20195D466
Person / Time
Site:	Three Mile Island
Issue date:	04/29/1986
From:	Beall J, Callan L, Caphton D, Isom J, Mary Johnson, Martin T, Mckee P, Saunders A, James Smith, Danielle Sullivan NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I), NRC OFFICE OF INSPECTION & ENFORCEMENT (IE)
To:
Shared Package
ML20195D446	List: ML20195D443 ML20195D466
References
50-289-86-03, 50-289-86-3, IEIN-83-64, NUDOCS 8606040217
Download: ML20195D466 (38)

Text

.

,

'

0FFICE OF INSPECTION AND ENFORCEMENT DIVISION OF INSPECTION PROGRAMS Report No.: 50-289/86-03 Licensee: General Public Utilities Nuclear Corporation P. O. Box 480 Middletown, Pennsylvania 17057 Docket No.: 50-289 License No.: DPR-50 Facility Name: Three Mile Island - Unit 1 Inspection Conducted M ch 3-27, 1986 Inspectors: I # F4 L. X Callan :hief, Performance Appraisal Section, I / Date Y Tfff E'. B 7, PrJject Engpeer, Region I /Date/

JE - 6 yAr/a

- L'. ton, enior Te nical Reviewer, Region I ' Oate/

v pf' 2ffM J. A. I o , Reactor Opdrations Engineer, IE 'Date

~

'

f/h M. R. J6Mson, Reactor Fperations Engineer, IE /Date Tn.D& In pection Specialist, IE aksk

'Datd T. 0.'W74artin L v/u/x A. agiers,Inspcti#nSpecialist,IE 'Aate/

) _ Mb6[Ab nspection Specialist, IE //Da ~

J.} 'i h

D.U. S llivan, Jr. , Inspec on Specialist, IE O/R

@ath Accompanying Personnel: *D. F. Humenansky, OCM Cont cto 2: *E. T. Du 1 p, *G. W. Morris, *G. J. Overbeck Approved by: Y/29/86 Phillip F. @ Kee, Chief Date Operating Reactor Programs Branch, IE

• Present during the exit interview on March 27, 198 PDR ADOCK 05000289 G PDR

_- -- _ _ . - . - -

.

1 O j SCOPE: This special, announced team inspection conducted an in-depth assessment of the operational readiness of the emergency feedwater syste RESULTS: The licensee's operational readiness and management controls were reviewed in six functional areas, primarily as they related to the emergency feedwater system. The functional areas reviewed were:

Design Changes and Modifications Maintenance Surveillance Testing Operations Quality Assurance Training Eighteen potential enforcement findings, identified in this report as Unresolved Items, and seven inspector followup items will be followed up by the NRC Region 1.

!

,

J

.

I i

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

.

. _ - .

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

,

.

!

s INSPECTION OBJECTIVE The objective of the team inspection at Three Mile Island-Unit I was to assess the operational readiness of the emergency feedwater (EFW) system by

~

determining whether:

o The system was capable of performing the safety functions required by its design basi o Testing was adequate to demonstrate that the system would perform all of the safety functions require o System maintenance (with emphasis on pumps and valves) was adequate to ensure system operability under postulated accident condition o Operator and maintenance technician training was adequate to ensure proper operations and maintenance of the syste o Human factors considerations relating to the EFW system (e.g.,

accessibility and labelling of valves) and the system's supporting procedures were adequate to ensure proper system operation under normal and accident condition II. SUMMARY OF SIGNIFICANT INSPECTION FINDINGS ,

-

This section summarizes the safety effects of the more significant findings on the operational readiness of the Three Mile Island (TMI)-Unit I safety systems.Section III provides the detailed findings pertaining to the major functional areas evaluate Safety Effects on the Emergency Feedwater (EFW) System The NRC inspection team identified the following design concerns in the EFW syste The two-hour backup supply air system that supplies the pneumatically operated EFW flow control valves did not meet the required single failure criteria. Specifically, the team determined that a single failure of either of two check valves, which were not part of a routine test program, could have caused the depressurization of both trains of the two-hour backup supply air system. In the event of such a depressurization, the EFW flow control valves are designed to fail full ope Certain remote shutdown panel instrumentation for the EFW system was imporperly designed in a design change package prepared by a licensee contractor, Gilbert Comonwealth. This design change was initiated as part of the licensee's program to comply with the requirements of 10 CFR 50, Appendix R, and it had been approved and released for implementation during the next TMI-1 refueling outage. Through review of construction drawings in the design change package the inspection team determined that the power supplied to the affected EFW instrumentation on the remote shutdown panel would not be isolated from the control room as required. Specifically, train A EFW flow remote shutdown panel indication and train B EFW flow and-1-

__ . . ._ -_ _ __ _ ___

! - -

, ,

- steam generator level remote shutdown panel indications were all powered from circuits that were to be electrically cross-connected with the control room. The team was concerned that, in the event of a need to evacuate the control room due to a fire, the potential

,

existed for the EFW instrumentation discussed above not to be l available for the operators at the remote shutdown panel. This

! same weakness was noted to exist in the current configuration of

'

the remote shutdown panel at the time of the inspection; however,

'

the licensee was not connitted to the NRC to have the affected EFW instrumentation in compliance with the requirements of 10 CFR. 50, Appendix R, until cycle 6 startup at the conclusion of the next refueling outag . Although the inspection team determined that the TMI-1 maintenance and surveillance testing programs were generally effective, weaknesses were identified regarding the manner in which certain components of the EFW and supporting systems were tested and maintaine , The two-hour backup supply air system relied on various check valves

- to seat to establish the integrity of the seismic /non-seismic boundary in the event of an earthquake. The team determined that these check valves were not routinely tested. As a consequence, the team was concerned that the system was susceptible to undetected

failure .

• The safety-related air system that accomplished the fail-safe posi-tioning of the EFW flow control valves (full open) and the steam

,

generator atmospheric dump valves (shut) was not tested to verify proper operation in the event of a loss of the two-hour backup supply air system. This " final positioning" air system relied on the proper

> operation of various automatic valves and check valves that were not

]

included in a routine test program, i The installation of the air cylinders in the two-hour backup supply air system was not consistent with that specified in the structural design analysis. Specifically, the seismic restraints (chains, turn-buckles, eye-bolts, etc.) designed for the air cylinders had not been adequately maintained after installation to ensure that l

original design requirements were met. The team found loose chains

'

with open S links, missing turnbuckles, and eye-bolts not securely fastened, which all contributed to the team's concern that vertical l

movement during a seismic event could potentially cause the failure j of connecting tubing.

' The procedure for replacing the EFW pump packing was determined to be not sufficiently detailed to ensure proper performance of the

,

task. The licensee had been performing this maintenance task using a i

generic pump packing procedure that did not address several critical steps that applied specifically to the EFW pump. The team considered this issue significant because errors in perfonning the

'

omitted critical steps had led to a previous EFW pump failure.

[ Certain aspects of the routine EFW pump surveillance tests were j determined to be weak because artificial initial conditions were establishe __. . . _ . - . _ - _ - _

.

..

.. 1) The steam supply lines to the turbine-driven EFW pump were routinely blown dry of collected condensate immediately prior to testing the pump. The licensee had not determined if the collected condensate could impact the performance of the pump's turbine driver in the event of an automatic start. This concern was mitigated by the fact that the licensee was in the practice of blowing down the steam supply lines each shift as a routine precautio ) The EFW pump discharge check valves on the EFW pumps were not tested in the reverse direction. During routine flow testing of the EFW pumps, the pump being tested was isolated from the idle EFW pumps. As a consequence, the ability of the discharge check valves on the idle EFW pumps to prevent back-flow was not routinely verifie Effects on Other Safety Systems In addition to the specific concerns discussed above that relate to the EFW system, the team also identified several general concerns that have the potential to affect the proper operation of other safety system . Several weaknesses were identified in the implementation of the TMI-1 design change progra .

~ Examples where the mini-mod process for accomplishing plant modifications of limited scope was not being implemented as intended by plant procedures, Examples where the design verification process was not implemented as required by plant procedure c Examples where design input was not always controlled as required by ANSI N45.2.11. Cases were identified in both the mechanical and electrical area Examples where the boundaries between seismic and non-seismic systems or components were not consistently shown on piping and flow diagram In some cases, boundaries were not shown, were incorrectly shown, or were contradictory between document . Weaknesses were identified in the program for control of permanent and temporary lead shielding. The team found that 10 CFR 50.59 evaluations for shielding installations were not being accomplished; that the engineering calculations for a standard table in the shielding control procedure did not consider the effects of seismic events, pipe configuration, types of anchors, or concentrated loads; that the shielding control procedure did not address installation techniques or requirements; that engineering calculations for shielding installations were not being verified in a timely manner; and that a permanent shielding installation did not conform to the technical analysi . It appeared that circuit breaker sizes had not been properly coordinated to ensure fault clearing for certain safety-related and non-safety-related power supply inverters. The team was concerned that a fault on a single-3-

-

'.. -

,' circuit could result in the loss of an inverter fed bus. Related to this concern was the additional observation that the licensee apparently had not considered the effects on safety-related load centers of the failure of loads that had not been qualified to operate in a harsh environment following a high-energy line brea . Environmental qualification records were not maintained in accordance with 10 CFR 50.49 with respect to the electrical cabling to the EFW discharge header cross-connect valve The team was concerned that additional environmental qualification record problems may exis III. DETAILED INSPECTION FINDINGS Design Changes and Modifications Mechanical Systems Design Change Review The inspection team examined the design aspects of restart modification task RM-13H in detail. This modification added the safety-related two-hour backup supply air system to provide compressed air for operation of components within the main steam (MS) and emergency feedwater (EFW)

systems for a period of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> without the availability of the plant instrument air compressor In addition, design analyses associated with adding cavitating venturis (long term modification task LM-5), locking ,

open EFW pump recirculation control valves (task LM-12), and reducing the

-

setpoints of EFW turbine steam supply relief valves (task RM-13H) were reviewed. The following observations were made: The team determined that the the two-hour backup supply air system did not meet the single failure criteria. Following a seismic event which may require initiation of the EFW system, a single failure of either of two check valves, IA-V-1451 and I A-V-1460, could cause the depressurization of both trains of the two-hour backup supply air syste Figure 1 on page 5 illustrates that portion of the two-hour backup supply air system containing check valves IA-V-1451 and IA-V-146 Following a seismic event and a subsequent loss of the non-seismic i

backup instrument air header, the failure to seat of either check valve IA-V-1451 or IA-V-1460 would depressurize train A of the two-hour backup supply air system. After sensing that train A was depre surized, proper functioning of switching valve IA-V-1632 would cause train B to be depressurized because the check valves were downstream of this switching valv The FSAR, Section 9.10.3.2, stated that the two-hour backup supply air system would mitigate the loss of instrument air as a result of design basis or seismic events and that the system design meets the single failure criteria. Similar wording appeared in the licensee's System Design Description (SDD) 424-C, " Division I System Design Description for the Two-Hour Air Supply for Main Steam And Emergency Feedwater System Controls," Revision 1. Division I system design descriptions were comprehensive criteria documents that defined the design, operation, maintenance, and testing requirements of system _. ..

_ _

. - _ - , - . _-

. _ _ - - _

.

,

-

DET BE m

'

"$R "$ =

mas as l sa am -

,

95 95 i 2 s ,, ,, IA-V1130 llA-V1449 mo 4 L 2 r ,

, x -

m o ci

> c 01 1 y-<m = ^ lA-V1450 elA-V1451

> -

_

- -

X [

E <

y a g 1A-V1644 1A-V1645 ,---i w -

X 's_ _ _i1 c  % % T i

' ' ' 1A-V1646 iA-V1647 T& ~ -

l'\

,

_ "a

-

X / f i l

m wV ,

hb

~

d- 1A-V1131 1A-V1458 cn thw A ,

~ r 7 f I y -

IA-V1459 'I A-V1460 I-

'iL__ y o kn g

gm@

$ cn I M /V1X < ij g)2 m

I b rC = ,j1 ,L 1 1A-V1648 ilA-V1649

/

g-<m ' '

\

ESM ES " ' /

• $m$ 2MG 2e 05' TE m

i "

@ C '

m R9 R9 MH ms ~

m l - 9,o H

"

<

CD

__

O

,

This design weakness was discussed with licensee management, and the licensee subsequently made an immediate notification to the NRC Operations Center as required by 10 CFR 50.72. The licensee also implemented corrective actions, which included shutting valves IA-V-1450 and IA-V-1459, to remove the susceptibility of the two-hour backup supply air system to loss by a single failure. This item will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-01). The team found that two-hour backup supply air system check valves were not periodically teste Division II SDD 424C, "Two-Hour Air Supply for Main Steam and Emergency Feedwater System Controls,"

Revision 0, required that these air system check valves be tested once every refueling cycle. The team noted that an undetected failure of any one check valve to seat in combination with a postu-lated single active failure within the opposite loop could cause the loss of the two-hour backup supply air system. This concern resulted from the fact that the seismic /non-seismic boundary between the two-hour backup supply air system and the instrument and backup instrument air systems was maintained by a single check valve at each air user, except for EFW flow control valves EF-V-30A and EF-V-30B where a tripping valve was used instead of a check valve. The two-hour backup supply air system depended upon the ability of the check and tripping valves at each seismic boundary to seat to ensure, proper system performance following a seismic event or events that

result in the loss of instrument and backup air compressors, such as a high-energy line break in the intermediate buildin Further discussion of the apparent failure to provide adequate testing for the two-hour backup supply air check valves is provided in Surveillance and Testing observation Post-modification functional testing of the two-hour backup supply air system was found to be inadequate. Low power natural circulation test TP 700/2, Revision STR-3, was performed, in part, to verify that the bottled air supply was capable of supplying air to valves EF-V-30A/B, MS-V-6, and MS-V-4A/B for a minimum of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> following the loss of normal and backup instrument air systems. However, the test was not structured to confirm that the design bases for the two-hour backup supply air system had been satisfied. For example, the test did not confirm that the system was adequately sized to supply sufficient air to cycle all associated valves 20 times as specified in the design or to observe a minimum of 10 strokes per valve as required by the 500. In addition, the test did not establish an initial condition to have the bottled air pressure be at its minimum pressure of 1500 psi .

Functional Test TP 248/2 was perfonned, in part, to demonstrate the operability of the two-hour backup supply air system. The team found little correlation between the design bases of the system and the testing performed as demonstrated by the following tabl .. __ .-.

. =. . . - . --- . -- - - . _ _ . _ - _

.

,

REQUIREMENT DESIGN TESTING

..

Period of operation 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> I hour i

Valves using air (excluding EF-V-30A/B EF-V-30A/B

EFW pump recirculation valves) MS-V-4A/B MS-V-4A/B MS-V-6

I Number of times valves 20 5 are cycled Minimum manifold pressure 1500 psig Not specified per

! initially test; however train

, A and B started at 1700 psig

Minimum manifold pressure 300 psig greater than 50 after two hours of use (Note 1) psig System leakage rate 0.03 scfm at Not addressed 2500 psi Note 1: The pressure regulators were designed to supply maximum ,

- flowrate at a minimum inlet pressure of 300 psi .

!- The team considered that a reduced number of cycles and a shortened l

test period was reasonable provided they were correlated to the design i

'

l bases and that acceptance criteria would reflect that correlatio ;

ANSI N18.7 requires that tests be performed following plant modifi-cations to confirm that the modifications produce expected results and do not reduce safety of operations, and that test procedures include appropriate quantitative or qualitative acceptance criteri The team considered the test performed on the two hatr-backup supply air system did not confirm that the modification proauced expected results per the design bases and did not have acceptance criteria consistent with these system design bases.

, This item was discussed with licensee management and will remain unresolved pending followup by the Region I Office (50-289/86-03-02). The team found the site installation of the air cylinders for the two-hour backup supply air system differed from that specified in

>

the structural design analysis. Calculation 609-0293, " Bottle Rack i

for RM-13h," Revision 0, provided the rack design for restraint

of these air cylinders during a seismic event. The design included chain restraints to preclude vertical movernent with turnbuckles attached to the chain to assure adequate tension. The chain and turnbuckle connections were to be made by open "S" chain links that were intended by design to be closed af ter installation. Inspection

.

of the installation by the team revealed that no turnbuckles were installed, the chain restraints were loose, "S" links were not a

closed, and eye-bolts were not securely fastened to the fram ; Although the existing arrangement differed slightly from the original

design by not having turnbuckles installed, the intent to restrain i the bottles from vertical movement may have been effective if the i'

- 7-

,

. . - - e .- .,_.---, y . . - ~ _ _ - . . - . - _ _ _ _ . _ . - - .,,_-._,_,._.,-..--,----_,-.e-r.--,,-c, ,.-_-,,..,---n . . - , - - - - - - -

.

-

,.

chains had been maintained snug. The team's concern regarding this installation was that vertical movement allowed by loose chains during a seismic event might cause the failure of conne.cting tubing and loss of the air supply. One of the design requirements of the two-hour backup supply air system was to provide a reliable source of air following design basis events, such as an earthquak This item has been discussed with licensee management and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-03).

e. Design input was not always controlled consistent with the require-ments of ANSI N45.2.1 The following examples were noted:

1) Design input for the two-hour backup supply air system was incorrectly selected and incorporated into the system desig GPU calculation C-1101-852-5360-001, "Two Hour Backup Instrument Air System Pressure Low limit," Revision 0, determined the minimum allowable manifold pressure to maintain an adequate stored air capacity to operate the EFW and MS valves for a period of 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. This calculation was based on system leak rates obtained in leak rate testing performed in May 1983 which essentially conducted a wld test for the high pressure portion of the system. This t;st preceded system functional testing accomplished by TP 248/2. The hold test on train A was .

terminated because of excessive leakage. The hold test on

'

train B was not representative because leaking air cylinders were isolated. An approved test procedure did not exist, and documentation consisted of handwritten observations by a test engineer. As such, the team did not consider the test results appropriate for use in the GPU design analysi During the team's walkdown of the system, air leakage was noted from various fittings. The team was concerned that normal leakage of both the high and low pressure portions of the system may have been much greater than assumed in the calculation. The team found no evidence that actual system leak rates were being determined, such as by trending air leak rates, or through periodic testing to determine actual leak rate ) Design input associated with sizing of regulating valves IA-V-1621A and IA-V-1621B was provided by GPU to an architect engineer but was subsequently changed by the licensee without notifying the affected design organization. Specifically, the minimum bottle pressure was reduced from 300 psig to 100 psig. As a con-sequence, the regulating valves would not provide maximum design flow at the reduced bottle pressure. .The team considered the technical significance of this concern to be minimal as long as air capacity was oversized and pressure was maintained high, but the item illustrated how an apparently minor design input change can have a significant effect on the design accomplished by an external design organizatio As discussed above, the use and control of verified design input was not consistently performed. In general, the weaknesses identified did not adversely affect the design of installed hardware but could have

~

-8-

-

-

..

affected the assumed operating / design margin. However, the identi-

,~

fied weaknesses contributed to the team's concern with respect to the overall control of the design process. Additional weaknesses in implementing the requirements of ANSI N45.2.11 regarding control of design inputs and design assumptions are discussed in Design Changes and Modifications observation The team considered that the classifications for major systems, components, and structures were not clearly identifiable by using the Quality Classification List (QCL). This conclusion was based upon the following observation ) The EFW system was not within a single classification. Portions of the system were indicated in the QCL as being either important-to-safety (ITS), nuclear safety-related (NSR), or benefits reactor shutdown (BRS). Reference to the subsystems and components for detailed classification identified the following:

a) The EFW pump suction from condensate storage tanks and hotwell was identified as IT b) The EFW pump control instrumentation was identified as IT c) The EFW control system was identified as IT ,

- Other than these items, no other classifications were identified for the EFW system. The team found this limited information inconsistent with that implied on the EFW system pipe and flow diagram. This drawing indicated that some portions of the system were NSR but did not contain flags to further identify what was NSR, ITS, or BRS. This lack of identification was inconsistent with the other licensee pipe and flow diagrams reviewed. Based on the above, the team could have concluded by examination of the drawing that the entire system was NS However, the team was informed by the licensee that the EFW system would not be fully safety-grade until the completion of long term modification task LM-13, and some components such as the turbine-driven EFW pump would not be safety-grade because of seismic consideration ) The two-hour backup supply air system was not identified in the QCL. The team determined that the system was classified as ITS based upon information in the Division I SDD and the pipe and flow diagra The licensee acknowledged that the system level QCL required users to obtain assistance from a QCL " interpreter" who was specifically trained to render these interpretations. The licensee also indicated that a component level QCL was currently under development. The progress of this effort will remain an inspector followup item (50-289/86-03-04). Electrical Systems Design Change Review The inspection team examined design analyses associated with electrical protection of the EFW pump motors and other large motors, battery

. '.. *

..

sizing, the integrated control system (ICS) and EFW instrumentation and control inverters, DC system power distribution, and motor-operated valves. In each case, the team detennined that portions of the calculations and design analyses reviewed were not consistent with the design control requirements of ANSI N45.2.11. Specifically, design inputs and assumptions often were not documented or verified, and some calculations were not sufficiently complete to pemit design verification without recourse to the originator. The following examples were identified:

1) Design analyses for EFW pump motor overcurrent protection were considered weak due to incorrect relay settings and the apparent lack of consideration for long-term thermal degradation of the motors. The EFW pumps and motors were originally not considered to be nuclear safety-related components. The EFW system control circuits and power supplies were upgraded in modifications RM-13E and LM-13, task 10, from non-safety-related to safety-related. The team found that the analyses performed around 1970 to detennine the setpoints for the overcurrent relays protecting the EFW pump motors (and other large safety-related motors) had not been update For example, review of records revealed that the safe stall time at rated locked-rotor current for the EFW pump motors was ,

5 seconds for hot restart when the motor had been running within

the previous hou The signoffs on the original setting notice in 1973 indicated that the overcurrent trip delay setting for the EFW pump motors was set at 5.8 seconds. The licensee was unable tc provide an analysis to support this relay settin The team was concerned that the overcurrent relay protection provided for safety-related motors had not been verified against actual motor data such as acceleration curves or motor thermal damage curve ) Control of design inputs associated with battery sizing was considered weak. The team reviewed the latest sizing calculation for the new battery to be installed during the March / April 1986 outage and determined that the battery was sized based upon a minimum battery temperature of 72* No reference was included in the calculation for the basis of this minimum temperatur The team reviewed the electrolyte temperature recorded weekly in accordance with procedure 1301-4.6, " Weekly Surveillance Check," to determine what temperatures the existing batteries were experiencing. The weekly surveillance for the first 10 weeks of 1986 for both safety-related batteries indicated that 17 of 20 readings were below the miniinum design temperature of 72* Some of the temperature readings were as low as 65'F, which could result in a battery capacity approximately 5% lower than assumed in the analysis. The team noted that the battery surveillance procedures did not include acceptance criteria for battery temperature. The team also found that other design input data used in the calculation, such as pump and valve starting and running currents, lacked sufficient references to permit complete verification of the calculation .

.

..

It appeared that the 25% margin included in the battery for aging effects would more than compensate for any near term capacity problems. However, the team wa:; concerned that an unacceptable loss of battery capacity could result if the battery room temperature were not maintained above the minimum design temperature as the battery reached mid-lif ) Design analyses were not available to demonstrate the capability of some of the 118Vac power panel circuits providing power to ICS and EFW instrumentation to clear electrical faults. The safety-related power panels were fed solely from inverters and did not have an alternate power source for fault clearing. The power panel schedules indicated circuit breaker ratings as high as 30 amperes on the safety-related panels and 50 to 70 amperes on the non-safety-related power panels. The non-safety-related power panels fed loads such as ICS and the backup manual con-trollers for the EFW flow control valves. The team was concerned that a fault on a single circuit could drag the associated inverter into the current limit, low voltage mode and result in loss of an entire inverter fed bu ) Preliminary, unverified design input was used as a basis for fuse changes in the de system power distribution panel These changes were accomplished by Job Ticket CC 219, 10/4/83, ,

to increase the interrupting capability of the fuse The

-

task was authorized by memorandum LAI 83-0037-8/24/83, which referenced unissued Technical Data Report No. 374. The team was concerned that the Technical Data Report referenced as the basis for this change had yet to be completed, verified, or issue ) Design analysis for determining minimum motor starting voltages for certain safety-related valves appeared to be inadequat An analysis for minimum motor starting voltage for safety-related valves had been perfomed in 1979. This analysis resulted in actuator modifications to 2 of the 31 valves analyzed so they would operate at the minimum required voltage. The EFW system Valves, and MS and Condensate valves supporting the EFW system, were not reviewed at that time because they were not considered safety-related. The team determined that analyses for these valves had still not been accomplished even though several were now considered safety-related valves. In addition, it did not appear that the actual minimum voltage available at any of the safety-related motor-operated valves had been determined to support the assumption in the original analysis that the voltage at the valve operators would not drop below 75% of the motor rated voltag .

The weaknesses discussed in the above five examples and the two examples discussed previously in observation 1.e appear to reflect inadequate implementation of the requirements of ANSI N45.2.11 regarding the documentation and verification of design inputs and assumptions, and the conditions of design calculations. This item was discussed with licensee management and will remain unresolved pending followup by the NRC Region 1 Office (50-209/86-03-05).

- 11 -

N -

' ~

l b. A proposed design change to the remote shutdown panel failed to meet

'

the requirements of 10 CFR 50, Appendix R, by not isolating the power supply for the EFW remote shutdown panel instrumentation from the control room. This oi. sign change had been prepared by a licensee contractor, Gilbert Commonwealth, and was approved and released for implementation during the next TMI-1 refueling outage. The design '

i change was initiated as part of the licensee's program to comply l with the requirements of 10 CFR 50, Appendix R. Appendix R, Sections

! III G2 and G3, require that alternate shutdown capability be independent of the circuits which may fail because of the postulated fire (such as circuits located in the control room). A review of the construction drawings in the design change package revealed the following examples involving EFW instrumentation where the require-ments of 10 CFR 50, Appendix R, did not appear to be met:

1) Power panel VBB, breaker 22, fed train B signal conditioning cabinet 81 via circuit EA 6824. Circuit EA 6856 fed the power supply from cabinet 81 to cabinet B2. The electronics in cabinet B2 developed train B EFW flow and OTSG 1evel instrumen-tation on the remote shutdown panel. Power for the electronics for this instrumentation on the remote shutdown panel came from the receptacle circuit in cabinet B2. However, this receptacle circuit was to be powered by the same circuit that provided power to three other circuits (EA 6858. EA 6860, and EA 6868) that all

went to the control roo ) Power panel V8C, breaker 20, fed the containment water level cabinet C. The electronics in this cabinet developed train A EFW flow indication on the remote shutdown panel. However, the same circuit that provided power to containment water level cabinet C also was to feed circuit EA 517, which went to the control roo The team also reviewed the existing installation for these remote shutdown circuits and found the same problems. The team noted, however, that the licensee was not committed to the NRC to have the affected EFW circuits in compliance with 10 CFR 50, Appendix R, l

'

until cycle 6 startup at the conclusion of the next refueling outag The weaknesses identified in the design change package for the remote shutdown panel were discussed with licensee management and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-06).

c. The TMI environmental qualification file failed to meet the require-ments of 10 CFR 50.49 by not including plant specific data on the cabling to the EFW discharge header cross-connect valves. The team identified that circuits CG 241 and CH 871 control EFW valves EF-V-2A and EF-V-2B, respectively. The team determined from the pull slips that both of these cables were type EK-9G and from purchase order 97099 determined that cable EK-9G was a KERITE Company Flame Retardant insulation / Flame Retardant Jacket (KERITE FR), However, the System Component Evaluation Worksheet (SCEW) for the KERITE cable used at TMI(SCEW-TI-770-005) identified the cable as a KERITE HTK There was no SCEW sheet available to support the use of KERITE FR insulated i

'

- 12 -

t l

'.

,.

cable at TMI-1. Although the TMI-1 environmental qualification records were detennined to be incomplete in this instance, the licensee did have a SCEW sheet (SCEW-0C-770-006) for KERITE FR cable used at their Oyster Creek plan Subsequent to the completion of the team's on-site inspection activities, the licensee identified additional potential concerns regarding the adequacy of the TMI-1 environmental qualification file for installed electrical cabling. This item was discussed with licensee management and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-07).

d. The team identified a number of loads connected to the Class IE power system that were either not safety-related, or were safety-related but not environmentally qualified for the ensironment in which they would operate. For example, the EFW pump room cooling units AH-E-24A and AH-E-24B were classified as nuclear safety-related and were connected to 480Vac control centers IA-ES and 18-ES, re:pectivel These loads were located in the intermediate building but had not been qualified to operate in a harsh environment that would exist following a high-energy line break in that area. Although cooling from these units was not required for operation of the EFW pumps, the units would continue to operate in the harsh environment. The team considered that a postulated electrical fault on these relatively small circuits probably would not affect the Class 1E power system;

however, the team was concerned that the licensee may not have reviewed the effect of failures of other larger loads on the breaker coordination of loaded motor control centers and unit substation It appeared to the team that the associated circuit overcurrent protection coordination review performed by the licensee for 10 CFR 50, Appendix R, may not have included the effect of the remaining load on a bus while comparing the breaker or fuse curves of the feeder with those of the largest circuit load. This concern was discussed with licensee management and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-08).

e. During review of the control circuit modification for the EFW control valves, the team noted that the cable shield was tied to the signal common wir The signal common wire was eventually connected to ground. The original system designer included an installation note on drawing 210-006, Revision 21, that cautioned that the (cable)

shields should be tied together and then connected to ground. It appeared that the intent of the original system designer was to have separate wires for signal comons and shield grounds and that the shields be individually connected to the panel ground bus. The team noted that such a grounding scheme would also be consistent with the guidance provided by IEEE 518-197 The as-installed condition, however, was such that noise on the shield drain would now be conducted to the ground bus via the signal common wir This grounding practice did not appear to present a problem for the specific control circuits reviewed because of their relatively slow response tim However, if this practice were used in higher speed or digital circuits, then spurious noise affecting the signal common voltage could result in improper control action '.

,. 3. Design Change Program Review Weaknesses were identified in the program for control of permanent and temporary lead shielding. The use of shielding for ALARA con-siderations was evaluated with regard to whether temporary or permanent design changes hid been made to plant systems without adequate design evaluation. The following five concerns were identified in this revie ) No documented 10 CFR 50.59 evaluations had been accomplished for lead shielding installations in the plant since procedure 9100-IMP-3282.01, "Use of Permanent and Temporary Shielding," was issued in 1983. Further, the team noted that procedure 9100-TMI-3282.01 did not address the subject of 10 CFR 50.59 evaluation IE Information Notice 83-64, " Lead Shielding Attached to Safety-Related Systems Without 10 CFR 50.59 Evaluations," dated September 29, 1983, addresses lead shielding installations and indicates that failure to analyze for possible seismic and structural effects (both dynamic and static) of lead shielding on safety-related systems potentially constitutes an unreviewed safety question. At the time of this inspection lead shielding installations were on safety-related decay heat system suction piping to pumps OH-P-1A and DH-P-1 ,

-

It was also noted that IE Information Notice 83-64 was reviewed by GPU in 1983 and the determination made that procedure 9100-IMP-3282.01 satisfied the 10 CFR 50.59 requirements identified in the notic ) Exhibit 2 in procedure 9100-IMP-3282.01 was a load table for use in determining allowable loads that could be installed on plant piping without further engineering review. Based on a review of the engineering calculation to support this table, four significant discrepancies were noted:

seismic ef fects were not considered;

the effects of pipe configuration between supports were not considered (other than straight pipe);

the effects of positive anchors at supports were not considered; and

• the effects of concentrated loads, such as valves, between supports were not considere ,

In effect, the table could have been used for all applications when, in fact, it only applied to straight pipe, simply supported, at specified maximum span ) Procedure 9100-lMP-3282.01 did not address installation require-ments or approved techniques to assure that shielding was safely

- 14 -

'.

C

.. and correctly installed with approved installation materials and procedures. This information was not available in other procedures as wel ) Calculations to support lead shielding installations were not verified in a timely manner. There had been approximately 10 shielding calculations accomplished and only one had been verified in accordance with EP-009, " Design Verification." The practice apparently had been to accomplish supportive cal-culations and to install the shielding prior to completing the design verifications. The team considered that lack of completion of design verification prior to installing temporary shielding to be significant since shielded systems are not necessarily isolated from plant use and no functional test can be accomplished to assure adequacy of the temporary design chang ) One permanent shielding installation reviewed involved the use of loose concrete blocks to shield drain lines for letdown prefilters MU-F-2A and MU-F-28. The criteria established by the licensee for installation of the blocks was that the center of gravity of the top blocks be no more than 12 inches off the floor, the blocks be no closer than 2 feet to important-to-safety (ITS) equipment due to seismic considerations, and .

that a warning sign be installed identifying the 2 foot require-ment. Site inspection of the as-installed blocks by the team

,

revealed that the top block center of gravity was 15 inches off the floor in some locations, ITS valves SF-V-77 and SF-V-71 were located within 6 inches and 19 inches of the blocks, respectively, and no warning sign was installe The above inadequacies in the program for control of temporary and permanent shielding were discussed with licensee management and will remain unresolved pending followup by the Region I Office (50-289/86-03-09). The mini-mod process of procedure EMP-002, " Mint-Mods," was reviewed in detail. This is an expedited process developed to provide rapid response capability for accomplishing plant modifications which meet the mini-mod criteria. The key criteria center around budgetary constraints, capability of the on-site group, and scope of the modi-fication. In addition, four mini-mod work packages were evaluated for procedure compliance to EMP-002 requirements as well as appropriate design change control requirements. The four mini-mods reviewed were:

BA 123164 "EFW Turbine Inlet Pressure Control Modifications";

installation documents were released for constructio *

BA 123170 " Removal of Instrument Air to Valves EF-V-8A, B, C";

installation documents were released for constructio *

BA 123166 "AH-VIC ESAS Test Group Modification"; installation documents were released for constructio .

BA 215504 " Fuel Handling Building Crane Modification";

,,

installation was complete and turned ove In general, the basic program and controls of EMP-002 and other associated procedures appeared to be adequate for the mini-mod process. However, two types of procedural implementation problems were noted:

1) 10 CFR 50.59 evaluations for two mini-mods were incorrectly marked as no change being required to the FSAR when, in fact, the text or drawings in the FSAR appeared to be affecte BA 215504 added a second fuel handling crane limit switch to improve reliability and safety. The FSAR had specific wording regarding the fuel handling crane limit switches in section 9.7.1.6 that this modification appeared to affec BA 123170 removed instrument air tubing to valves EF-V-8A, EF-V-88, and EF-V-8C. FSAR Figure 10.6-1 for EFW depicted air tubing to these valves and that figure would have to be revised once the air tubing was remove The team's concern, based upon examples such as those above, is, that a proper analysis for an unreviewed safety question may not

-

be perfomed if it is not recognized that the FSAR is affected by the modificatio ) The installation specifications issued to accomplish the four mini-mods reviewed did not address the attributes required by procedure EMP-002. Paragraph 4.0 of Exhibit 2 (Design Require-ments) to procedure EMP-002 specified two attributes to be addressed and paragraph 5 of Exhibit 2 (Design Description)

specified nine attributes to be addressed in installat'on specifications. None of the four mini-mods reviewed covered these entirely. The team considered that all of these attributes were relevant to design requirements and description and should be part of a controlled design change process for documentation of design input and outpu The above weaknesses regarding implementation of the procedure EMP-002, " Mini-Mods," were discussed with licensee management and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-10). Implementation of the design verification process as required by EP-009 was considered to be a weakness. The following problems were identified in the inspection:

1) Three engineering calculations reviewed had no design verifica-tion accomplished, calc # 1101X - 322F-165 Flowrates for two-hour backup air supply system

- 16 -

-.

".

" calc # 1101X-322F-424-1 - EFW system resistance

,,

calc # 1302X-5320-A50 - Shielding stress 2) Three design verifications reviewed had no checklists as required by EP-00 calc #1101X-322B-003 - Air consumption by EF-V-30 valves calc # 1101X-322F-157 - EFWP turbine relief valve setpoint calc # 1101X-3228-004 - Air consumption by MS-V-6 3) System Design Descriptions (SDDs) were not being design verified as required by EP-009. This was true for Division I and II of SDDs 474A, B, C, D and E. Further, discussions with licensee personnel revealed that design verification was not considered necessary for SDDs and that SDDs would not necessarily be updated to reflect changes. The team considered use of SDDs to be very beneficial and a practice that should be continued, but the SDDs need to be updated and verified since they are used for design and are considered to be a source of design input as well as a training input document. ANSI N45.2.11 requires that design inputs be verifie ,

- The team noted that Technical Functions Procedure EP-005, " Modi-fication and System Design Descriptions," recognized SDD Division I as a record of design inputs and indicated that for future modifications the design engineer must know the complete basis for the design. It was noted by the team while at the licensee's architect engineer's offices that the SDD Division I was considered to be design input for their design work. These inputs were not verified by their design veri-fication process but were considered to be fact. Team dis-cussions with licensee personnel revealed that it was assumed that the architect engineer was design verifying GPU SDD design input The design verfication concerns identified above were discussed with licensee management and will remain unresolved pending followup by NRC Region I Office (50-289/86-03-11). Design document updating procedures were considered weak. In this review 75 documents were identified that had 6 or more change documents posted against them. These documents were mostly drawings, but also included installation specifications, an instrument list (308001), an SDD Division II (232E), and a' master EQ equipment list (990-1429). Procedure EMP-015 stated that documents with more than five outstanding changes posted against them are subject to a mandatory updat The change document history for 10 of the documents identified with six or more changes was reviewed with the following results:

- 17 -

- - ._

'.

. Total Number Date of Document of Changes 6th Change drawing ID-662-18-002 8 5/83 drawing 641-074 7 9/83 drawing 311-842 7 10/83 drawing 215-021 10 6/83 drawing 215-051 9 7/83 drawing 304-641 8 1/84

.

drawing 224-503 18 3/82 SDD 232E 6 5/83 Instrument List 20 9/8F Master Environmental 9 4/85 Qualifications List Based on above examples, it appeared that the drawing updating criteria established by procedure EMP-015 was not being adequately implemente During the inspection it was noted that the TMI-1 quality assurance organization identified similar concerns in September 1985 in Audit No. S-TMI-85-10. The Engineering Services response to that audit in October 1985 indicated that:

,

The instrument list was being revised and the list is now subject to routine, timely maintenance. However, the NRC inspection team found that the instrument list still had 20 change documents posted against it. Some of these outstanding changes dated back to 198 *

The 215 series drawings were redundant to other data and would be voided. The NRC inspection team found that these drawings apparently had not been voided, as indicated by the fact that they were still listed as being in need of updatin The desirability of maintaining the updating requirement for certain drawings, installation specifications and SDDs was

< referred to the Manager-Engineering Projects, TMI-1. The NRC inspection team found that apparently no decision had been made, as these documents were still in need of updating and no other correspondence had been issue The licensee stated that a change to procedure EP-002, "GPUN Drawings,"

that addresses the concerns discussed above was being prepared. The progress of this effort will remain an inspector followup item (50-289/86-03-12), The team found that the Computer Assisted Records and Information Retrieval System (CARIRS) was the established means of controlling design documents that define or change the functional configuration of TMI-1. EMP-016, " Plant Configuration Control Lists," established the CARIRS method for maintaining plant configuration; however, the team was concerned that there was no procedure for use of CARIRS as

a tool for design, engineering, and general plant maintenance and i

- 18 -

.__ _ _ _ . . _ ._ , _ . , . _ _ _ . _ _ _ _ _ _ _ _ . _ _ , _ . . _ _ _ _ _ ___ . . _ _ . _ _ . _ ,

'.

.

operations use. Discussions with licensee management representatives revealed that use of and interpretation of CARIRS was n)t well under-stood by TMI-1 plant personne A sample of 20 as-built piping and instrumentation drawings (P& ids)

on file in the control room was checked for correct revision status, and two P& ids were found to be out of date. This was contrary to procedure EP-025, "As-Built Drawings," which stated that control room drawings were to be maintained current to reflect actual plant conditions. The affected drawings were:

P&ID 302-231 did not reflect the last three changes issued for it (dated 8/85,11/84,10/85),and

• P&ID 302-660 did not reflect the last change issued for it (dated 7/84).

The team was told that periodic audits were done by TMI-1 Design and Drafting to assure control room drawings were up to date with no change documents posted against them. However, the last audit was done in July 198 This apparent failure to properly implement the requirements of procedure EP-025 for control of as-built drawings in the control .

room will remain unresolved pending followup by the Region I Office

-

(50-289/-86-03-13). Inconsistencies and errors in drawings and System Design Descriptions were noted by the tean throughout this inspectio ) Some of the drawings affected were:

Drawing Discrepancy

1. 302-011 MS-V-13A & B shown as normally open, should be normally closed; MS-V-10A & B shown as normally open, should bc normally closed;

1. 600-520 Flow indicators were shown cross connecte # 600-347 EFW logic did not show manual loade # 600-340 Old E/P converters were shown for both EF-V-30A & B; backup t.ransfer switches were not shown; remote transfer switches were not shown; no reference was made to either manual loade # 210-707 E/P converters shown for EF-V-30A & B when E/I and I/P module combinations were actually installed; this dwg was in conflict with cable pull dwg 212-009-RF 126 for cables RF 126 and 128 because E/P was shown on this dw w - n--nx- s=..- .-.-_..w . - -. . . _ _ . - . _ ~ - - . . _ _, . _ _ - _ , . - ~ _ -

-

.

- # 660-42-017 Dwgs showed + 10 volts supply to HIC-849/850

1. 660-42-018 + 24 volts from power when it was supplies shown on 210 actually !959; dwg 017 showed l

,

alternate power from panel ATA, breakers 6 j and 16 when power actually came from ATB,

breaker 15; dwg 018 showed alternate power

~j from ATB, breakers 6 & 16 when it actually came from breaker 16; dwg 018 did not show output of HIC-850 connected to selector

,

switch; dwg also did not identify the transfer switches by component identification numbe # 600-435 Relays 86-CVI and 86-1/CVI, contacts 3-1-4,

,i 7-6-5, 9-11-8 and 3-1-4 were shcwn twice with different ITS designations in each case; these relay contacts also appeared on dwg 600-319

,

and none were marked IT t

1. 600-346 Remote manual loader not shown

1. 600-502 TB-4, connector 7 and 8 were identified with conflicting information, i.e., cable RF 143 was

,

shown connected to HS-003/HY-003, but according to 660-42-017, cable RF 147 connects to .

.

HS-001/HY-00 #302-273 Shows a solenoid enclosure on valves EF-V-30A and EF-V-308, but these enclosures were not [

actually installed. This as-installed

'

arrangement was shown correctly on dwg 308-416.

'

1. 302-272 Indicates air is supplied to SP-VSA and SP-VS However, these devices have been replaced with '

EQ I/P converters FY-849A and FY-8498.

p #302-271 Entire instrument air (IA) system was not depicted

! e392-273 on P& ids. Team walkdown found air users not shown

!

on dwgs and additional isolation valves, IA-V-98 and IA-V-99, between IA system and two-hour backup supply air system not shown on dwg. Dwg 302-273 indicated design pressures were 2500 psig u) to the regulating valves and 150 psig

!

from tiere on; but the FSAR indicated 2500 psig up to the regulating valve. 600 psig between the l regulating valve and the first switching valve,

and 150 psig from there on.

I

1. 302-271 Seismic I/ Seismic !!! boundary between the

1. 302-272 two-hour backup supply air system and the i #302-273 instrumentation air / backup instrument air systemn not shown

> r i 2) Errors identified in the Division !! SDD 424C, "TMI-1 Two-Hour l

Air Supply for Main Steam and Emergency Feedwater Controls,"

i were:

! - 20 -

L __ -

_ . - . -_ - - - _ __ . - . .- - - - - - -

.

i i

.

,

-

J

'

-

• For EF-V-30A, the wrong valve number was listed for the

'

manual isolation valve between the I/P converter and the

train A header of the two-hour backup supply air syste For EF-V-8C, the wrong valve number was listed for the manual isolation valve between the solenoid and train B

.

of two-hour backup supply air system. For EF-V-88, the l

' wrong valve number was listed for the manual isolation valve between the solenoid and trains A and B of the two-hour backup supply air system. The correct valve numbers were depicted on drawing 302-27 *

EF-V-30A E/P and EF-V-30B E/P were identified as instrument

,

air users. However, these E/P converters were replaced by

< environmental qualified I/P concerters by long term modification task LM-9. The correct instruments were i '

FY-894A and FY-850B as shown on drawing IA-424-42-1000, i

• The system relief valves were identified as CROSBY J05-15-A when the valves were actually CROSBY J05-15- The correct relief valve type was identified from i nameplate data and was correctly shown on the valve data sheet dated May 22, 1981. Likewise, the capacity of the relief valve was listed as 48,000 scfh, while the nameplate capacity was 87.660 scf ,

" * The 500 indicated that the high pressure instrument air within the high pressure air storage bottles had a dew l

point of less than or equal to -60*C. The team found that a prncurement document for truck air specified a dryness equivalent to -10*C. The actual dryness of the air supplied

,

was a -89'C. The Division I SDD required a dryness a equivalent to that of the instrument air system (-40*C).

t

The SDD stated that no more than one train of the two-hour backup supply air system can be out of service at any one time and that this time shall be kept to a

)

minimum. The maximum length of time consistent with the technical specification limit for one train of the EFW l

system was not specified.

While discrepancies such as the above will not necessarily lead to design problems, they do make the documentation trail hard to follow for determining actual design conditions. It appeared to the team that while some of the examples cited were drafting or typographical '

errors, some were the result of the lack of proper updating of design documentation. ANSI N45.2.11 requires tha,t personnel use proper and

current instructions, procedures, drawings, and design inputs. Design documents and changes to them are to be controlled to ensure that i

l correct and appropriate documents are available for use. The drawing deficiencies identified above and in Maintenance observation 5.b will remain an unresolved item pending followup by the Region ! Office

,

(50-289/86-03-14), Two other isolated areas of concern were identified by the team i in the design change proces ;

I I

- 21 -

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

.

_

1) A minor design change was accomplished by a maintenance job ticket when a section of the air supply line to valve EF-V-8A

_

operator wss removed. This modification was done without following design change control measures required by EMP-019,

" Plant Modifications Proposed by Plant Engineering." The Plant Engineering instructions attached to the job ticket made the statement that this was not considered a modification per EMP-019. The team considered this to be an incorrect judgmen ) The temporary modification process of AP-1013. " Bypass of Safety Functions and Jumper Control," was used for control of a jumper on MS-V13B to defeat the auto start capability. The 10 CFR 50.59 evaluation for this modification was marked that no cnange to the FSAR was required. However, Section 7.1.4.2.b of the FSAR indicated that the turbine pump will auto start for train 6 actuation. This temporary design change defeated the auto start capability and consequently affected the FSAR wordin The team acknowledges that this was a temporary modification, but one that definitely affected the FSA B. Maintenance 1. Several strengths were noted in the THI-1 preventive maintenance progra An extensive trending report was found to be routinely developed every 3 ,

months based on corrective maintenance activities performed over the

~

previous 12-month period. These reports identified problem areas on both a system and component basis. A review of the PM activities conducted on the non-safety-related integrated control system (ICS) revealed an ex-tensive coverage of the system on a circuit by circuit basis, including cleanings, calibrations, and refurbishment One weakness was found regarding preventive maintenanc Equipment history records revealed that, prior to this inspection, preventive maintenance had last been conducted on MU-V-16C, a high pressure injection discharge isolation valve, in August 1981. The licensee's program identi-fied this valve as requiring preventive maintenance every three year Licensee personnel advised the team that preventive maintenance was conducted on this valve on the last day of the inspection, March 27, 198 . A review was made of the licensee's program for the maintenance of motor-operated valves (MOVs). The licensee was found to be aggressively pur-suing the action items contained in IE Bulletin 85-03, " Motor-Operated Valve Common Mode Failures During Plant Transients Due to Improper Switch Settings." Design basis thrust values were being determined for MOVs and subsequent MOV testing was being conducted with a load cell to ensure that M0V torque switches were properly set to permit, the valves to achieve their design basis thrust. The' licensee had completed this state-of-the-art testing on 23 MOVs and planned to test approximately 70 more during an outage starting in March 1986. Although the licensee's overall program for maintenance of MOVs was determined to be adequate, several specific concerns, discussed below, were identifie The licensee had detailed procedures to cover the various aspects of MOV maintenance. A review of these procedures revealed several weaknesses.

l

- 22 -

.

.

(1) Procedure 1420-LTQ-2, "Limitorque Operator Limit Switch

'

Adjustment," Revision 8, described how to set the limit switch that prevents an MOV from backseating while opening. This procedure directed that this limit switch be set with the valve slightly closed to allow for coasting of moving parts. This was considered a weaknesses in that a more precise valve position for setting this limit switch was not provide (2) Procedure 1420-LTQ-2 also described how to set the limit switch that allows the valve to come off its shut seat without tripping on high torque. This procedure directed that this limit switch be set to remain closed for 3% to 10% of the valve stroke time. This was considered a weakness because 3% of the valve travel time may not be sufficient to overcome the unseating forces on the valve. The licensee advised the inspection team that this issue was currently under review and that they expected to change the setting of this limit switch to a more conservative 8% to 14% of valve trave (3) Preventive Maintenance Procedure E-13, "Limitorque Valves,"

Revision 12, directed the technician to jog the M0V to verify proper direction of motor rotation. No direction was provided as to how to jog the valve despite the fact that most MOVs have a seal-in feature that prevents intermittent operatio .

-

The procedural weaknesses identified above were discussed with the licensee and will remain an inspector followup item pending review of the licensee's corrective action (50-289/86-03-15).

b. Weaknesses were noted regarding the control of M0V torque switch se ttings . Equipment history records in some cases provided no explanation for apparent changes in torque switch settings and, in one case, indicated a change in a torque switch setting without foreman review or approval. The following examples pertai (1) Maintenance was conducted on MU-V-168, a high pressure injection isolation valve, on May 6, 1983. The data sheet for this activity indicated "N/A" for the equipment history torque values. The data sheet also indicated that the open torque switch setting had been adjusted from 3/4 to 13/4 and the close torque switch had been adjusted from 1/2 to 1 1/4. Despite the fact that the data sheet provided places for review and approval of these changes, neither was made. Additionally, the as-found torque switch values, both less than one, were below the re-commended values of 1 1/4 to 2 provided on the bill of materials for this valv (2) Maintenance was conducted on MU-V-16D, a high pressure injection isolation valve, on April 1, 1980. The as-found and as-left values for both the open and close torque switch settings were specified in the data sheet as 2 1/2. Maintenance was again conducted on this valve on August 12, 1981. The as-found and l

as-left torque switch settings were specified in the . data sheet

'

as open - 1 1/2 and close - 1 1/4. No records were available providing an explanation or basis for the apparent change in

- 23 -

,

<

!

,

torque switch settings for this valve. Additionally, no torque switch equipment history data was provided in the data sheets indicating the correct values to which the torque switches should have been se (3) Mainter,ance was conducted on MU-V-16C, a high pressure injection isolation valve, on April 1,1980. The as-found and as-left torque switch settings were specified in the data sheet as open - 3/4 and close - 1/2. Main;enance was again conducted on this valve on August 12, 1981. The as-found and as-Teft torque ,

switch settings were specified in the data sheet as open - 1 1/2 and close - 1 1/4. No records were available providing an explanation or basis for the apparent change in torque switch settings for this valve. Additionally, no torque switch equip-ment history data was provided in the data sheets indicating the proper values to which the torque switches should have been se Despite the weaknesses described above, the equipment history records indicated that current torque switch settings were within the manu-facturers recommended values. However, the team was concerned that the administrative system to control and document M0V torque switch settings may not be sufficient to maintain the new torque switch settings being established through the analysis and testing process described above. A licensee representative stated that MOV equipment history records would be improved and updated to clearly identify the

~

correct torque switch setting for each valve. This issue will remain an inspector followup item pending review of the licensee's cor-rective action (50-289/86-03-16).

3. A weakness was noted in the maintenance procedures for replacing the packing in the emergency feedwater (EFW) pumps. A review of maintenance activities since January 1985 revealed several recent occasions when corrective maintenance was conducted on EFW pump packing:

September 18, 1985 - Sone packing wa removed and adjusted on the outboard side of the turbine-driven EFW pum t December 9, 1985 - The outboard packing gland of turbine-driven EFW pump was repacked. The packing and lantern ring were found to be installed in such a way that the cooling water supply to the stuffing box was blocked. Additionally, the lantern ring was found to be warped, which may have caused *

increased heat to be generated in the stuffing box due to metal to metal contac April 4, 1985 - Someoftheinboardpickingwasremovedand adjusted on EFW pump 2 The activities described above were conducted using generic procedures 1410-P-1, " Repack Pump," and 1410-P-2, " Add Packing to Pumps and Adjust Packing Glands." These procedures were " generic" in the sense that they were written to be applicable to a variety of pumps. The team considered ,

these procedures weak for use on EFW pump packing because they failed to describe the method of installing the packing and lantern ring combination

.

- 24 -

.

., . , - - - - - - - - . . - - . . - . . . . - .

-

.

.

so as not to block cooling water flow into the stuffing box. This problem

~

had occurred, as identified above, during the December 9,1985, maintenance activity. The generic packing procedures also did not identify the specific combination of ribbon and braided packing to be used for the EFW pumps and did not provide tolerances for the EFW pump packing lantern rings, again related to a problem identified during the December 9,1985, maintenance activit A licensee representative stated that improvements would be.made to the procedural controls governing maintenance on EFW pump packing. This issue will remain an inspector followup item (50-289/86-03-17). The program for the use of vendor technical manuals as maintenance procedures was reviewed. In general, this program was considered acceptable. Numerous manuals had been reviewed and edited to provide assurance of their applicability to specific installed components at TMI-1. Procedure 1407-1, " Unit-1 General Corrective Maintenance Procedure," Revision 24, contained clear requirements regarding the use of technical manuals. This procedure allowed the use of a controlled technical manual for the conduct of maintenance without further en-gineering review of the manual. This procedure further required that, when a non-controlled manual is used, engineering review and concurrence be obtained to verify such things as torque values, dimensions and tolerances, and proper lubricant .

- A weakness was noted regarding the use of technical manual Instrument calibration procedures generally stated that, if as-found data are out of tolerance, the affected instrument is to be repaired using a specific vendor manual referenced in the calibration procedure. Such vendor technical manuals referenced in these procedures were found not to be controlled in 2 out of 10 cases checked. The two uncontrolled manuals were "Rosemont Model 1151 Level Transmitter," identified in Procedure 1302.5.15, " Core Flood Tanks Pressure and Level Channels," Revision 8; and " Bailey Instrument Manual E92-79 (Bailey Buffer Module)," identified in Procedure 1302.5.18, "High and Low Pressure Inspection Flow Channel,"

Revision 1 This issue will remain an inspector followup item pending licensee review and control of the two manuals identified above and further review of the use of potentially uncontrolled technical manuals in calibration pro-cedures (50-289/86-03-18). The inspection team conducted a detailed walkdown of the EFW system to verify that the system layout was as depicted in the system drawings, to ensure that the system was aligned as required by licensee procedures, to review component accessibility, and to evaluate the material condition and cleanliness of the system. The team observed that the licensee had expended considerable effort maintaining the general cleanliness and material condition of the plant and the EFW system in particula ,

Several weaknesses were noted: The washer on an installed concrete expansion anchor on pipe support EF-18 was loose and could easily be rotated by hand. This could indicate an improperly installed anchor bolt. This issue will remain unresolved pending followup by the NRC Region I (50-289/86-03-19). 1

- 25 -

_

.

.

- Actual component layout differed from the EFW system drawing, 5130 302-082, Revision 7, in several minor instances. Specifically:

(1) Pressure guages installed downstream of EF-V-48A and EF-V-50A, EFW pump packing cooling supply valves, were not shown on the system drawin (2) The location of EF-V-56, a drain valve on "A" EFW discharge header, was different from that indicated on the system drawin (3) Part of the section of 6-inch diameter piping between EF-V-2A and EF-V-30A was indicated on the drawing as 4-inch diameter pipin (4) The "B" condensate supply check valve, C0-V-168, was incorrectly labeled C0-V-16A. Similarly, the "A" condensate supply check valve C0-V-16A was incorrectly Iaoeled C0-V-16 (5) Check valve EF-V-198 was mislabeled EF-V-19 Additional drawing deficiencies were identified in Design Changes and Modifications observation EF-V-36A and B, EFW pump packing cooling valves, were identified in ,

the system valve lineup, Enclosure 1 to EFW Operating Procedure 1106-6,

~

as being throttled. These valves were not locked nor were their approximate positions identified on the valve lineup. The team was concerned that there appeared to be no way that the correct position of these valves could be verified without operating the syste The inspection team noted that the location and orientation of EF-V-54, a recently installed motor-operated block valve for EFW flow control valve EF-V-308, restricted manual valve operability due to the valve handwheel's distance above the floor and proximity to a wall. This observation was considered significant because the licensee had deenergized the motor-operator to EF-V-54 and intended to treat it as a manually operated valve. This will remain an inspector followup item pending determination that the valve can be manually operated (50-289/86-03-20).

. Operations Procedures and system drawings relating to nonnal and abnormal operations of the emergency feedwater (EFW) system and the integrated control system (ICS)

were reviewed in detail. The inspection team performed system walkdowns and verified procedural adequacy. Equipment was observe,d in operation, valve positions and equipment readiness were verified, and operator perfonnance was observe . Control room observations revealed the following strengths: Operators were considered to be both knowledgeable and professiona Control room activities were properly conducted. Conmunications were

!- both clear and concis .

E

' Shift turnovers were observed to be thorough. Similarly, the brief-ing given to the on-coming shift on in-progress and planned work activities was comprehensiv Interviews with operators and observations of operational and transient evolutions revealed they were knowledgeable concerning the ICS, including operating procedures, failure modes, and emergency procedures. The team noted that there were typically only four to six control room annunciator alanns lighted during plant operatio . Weaknesses were observed relating to out-of-specification operator log data entries. Specifically, the team identified examples where out-of-specificaiton log data entries were not circled as required by the instructions on the affected log sheets;

• explanatory notes were not made on the log sheets for out-of-specification data entries as required by procedure 1001 G,

" Procedure Utilization," Revision 11; and

• log sheet discrepancies, such as those noted above, went uncorrected and apparently unnoticed during shift turnover reviews of the logs by the shift foreman and the on-coming operato .

.

The most significant example noted by the team of recent problems with out-of-specification log data entries occurred on February 6 and 7,1986, when five consecutive operating shifts recorded decreasing EFW two-hour backup supply air system supply pressure readings (1100, 800, All 580, five 420, of and 400 psig) in the Secondary Auxiliary Operator's Lo these readings were below the specified acceptance criteria of 1700 psi Further, the operators had not circled these out-of-specificaiton readings and had made no explanatory log entries regarding the condition. The problem was corrected on February 7,1986, when an operations engineer reviewing the log sheets questioned the data entries, investigated, and found a closed truck supply valve. The operations engineer opened the valve to correct the decreasing pressure condition that had been indicated for the five shifts by the logged data entries. The team noted that during the five shifts in question the two-hour backup supply air

~

syctem's bottle pressure met procedural acceptance criteria required for system performanc In addition to the examples discussed above, the team noted several other isolated instances where out-of-specification data entries in the Primary and Secondary Auxiliary Operators' Logs were not circled during the period of March 1-9, 1986. Of particular concern to the team for each of the examples noted was that the discrepancies had not been identified and corrected during the shift turnover process as required by procedure 1012,

" Shift Relief and Log Entries," Revision 2 The weakness noted regarding the handling of out-of-specification log data entries were discussed with the licensee and will remain unresolved pending followup by the NRC Region I Office (50-289/86-03-21).

!

- 27 -

_ - - , -

-

.

• \

2 The inspection team identified deficiencies related to the procurement of make-up instrument air for the EFW two-hour backup supply air system which was being supplied continuously from a truck. This system was identified by the licensee as being important-to-safety, Air for this system was being procured under a purchase order (TP-035330) with a safety classification of not-important-to-safet Consequently site QA engineering did not review the purchase order, The purchase order specified " dry compressed air in bulk industrial grade with a dew point less that or equal to -10 Deg. C at 100 psig."

The TMI-1 FSAR (Sections 9.10.3.2 and 9.10.1.1) requires that this air to have a dew point of at least -40 F at 100 psig, filtered to 0.9 micron. The purchase order therefore specified an incorrect and nonconservative dew point (-10*C is equivalent to -14 F) and failed to invoke the filtration requiremen The Secondary Auxiliary Operator's log specified, for the instrument air system, a dew point reading range of -60 C to -10 C; and required notification of the Shift Foreman when the dewpoint exceeded -10 This upper limit of -10 C was not consistent with the upper limit of-40 F specified in the FSA Despite the lack of adequate administrative controls for the procurement of backup instrument air, the team noted that the air was supplied by the'

~

vendor at a dew point of -128 F at 100 psig filtered to 0.1 micron which exceeded the specification of the FSAR. Prior to the completion of the inspection, the licensee had requested a Certificate of Conformance from the vendor and was taking neasures to ensure that future air purchases were made under an important-to-safety purchase order. This issue will remain unresolved pending inspector review of the implementation of the licensee's proposed corrective actions (50-289/86-03-22). The team noted a minor weakness regarding the licensee's controls for lifted leads, jumpers, and temporary modifications that affect safety-related plent equipment. Procedure AP 1013, " Bypass of Safety Functions and Jumper Control," Revision dated 10/23/85, required that existing lif ted leads, jumpers, and temporary modifications be re-evaluated every 12 months using a fonn entitled, " Safety Evaluation / Design Review". This new safety evaluation was required to be included with the original safety-related evaluation in a log maintained in the control room. A review of the log revealed that the licensee's practice was not to re-submit a new Safety Evaluation / Design Review, but rather to initial or sign and date the existing safety evaluation, thereby indicating that the 12-month re-evaluation had been performed. This item will remain unresolved pending followup by NRC Region I (50-289/86-03,23). Surveillance and Testing The team reviewed the testing associated with assuring functionality of the emergency feedwater EFW system, the two-hour backup supply air system, and the integrated control system (ICS). In particular, the team sought to determine

! that system components had been adequately tested to demonstrate that they could perform their safety functions under all conditions.

- 28 -

.

.

.

'*

1. EFW system surveillance test (ST) procedures were found to be generally adequate for demonstrating system functionality. However, the team identified the following two weaknesses: The monthly surveillance test of the turbine-driven EFW pump (ST 1300-3G A/B, Revision 21) was performed only after ensuring that the steam supply lines were drained of condensate. This practice appeared to create an artificial initial condition for the sur-veillance test. The team noted, however, that the auxiliary operators were instructed to blow down once per shift all steam trap drains in the intermediate building where the turbine-driven pump is located. This once-per-shift blow down policy was incorporated into the auxiliary operators' logs and was intended to limit the amount of condensate in the steam supply lines. The team was concerned that the licensee was unaware of the effect that residual condensate in the steam supply lines would have on the turbine-driven EFW ptmp in the event of an automatic star The FSAR (Section 10.8.2.2.f) stated that the EFW system " flow test is conducted with the EFW system valves in their normal alignment."

Technical Specifications (Section 4.9.1.6) governing EFW system periodic testing placed no restriction on EFW valve alignment. In practice, surveillance procedure ST 1303-11.42 isolated the in-dividual pump being tested from other portions of the system not in .

the flow path to the steam generator being fed. Therefore, the EFW

-

pump discharge check valves (EF-V-11A, EF-V-11B, and EF-V-13) were verified to pass flor, but their ability to seat and prevent reverse flow was not periodically teste The team was concerned that isolating these check valves during pump flow surveillance testing created an artificial initial condition that would prevent reverse flow leakage past these valves from being considered. The team noted that these check valves were part of a mechanical maintenance task (>fi-000031) which required one EFW check valve to be disassembled and inspected annually in a 5-year cycle. Although this activity was not included in the licensee's inservice inspection and testing program, it would, if implemented as intended, provide some assurance that the discharge check valves would perform their intended functio Licensee records indicated that this maintenance activity had not yet been conducted on the EFW pump discharge check valve The two concerns listed above regarding EFW pump surveillance test procedures were discussed with the licensee and will remain an unresolved item pending followup by the NRC Region I (50-289/86-03-24).

2. The EFW two-hour backup supply air system was tested as part of its modification acceptance process, but discussion.s with licensee personnel and review of records revealed that no further testing had been performed and no future testing of this system was planned. Specific examples of testing weaknesses are discussed below: Proper operation of the EFW two-hour backup supply air system was found to depend on 10 identical isolation check valves that were located at various points along the interface with a non-seismic air system. These valves were not routinely tested and under normal operating conditions experien:e no differential pressure. Two of

- 29 -

.

.

s these untested valves (IA-V-1451 and IA-V-1460) were of particular

.

concern since they were located where a single failure following a seismic event could depressurize both trains of the EFW two-hour backup supply air system (See Design Changes and Modifications ob-servation 1.b for further details). The failure of any of the other eight untested valves could blow down the single train of air supply associated with the failed valv The control of each EFW flow-control valve (EF-V-30A and EF-V-308)

was found to depend on the repositioning of a three-way air valve (IA-V-1344 and IA-V-1440) to the EFW two-hour backup supply air system. These three-way valves were not routinely teste The ability of the EFW system flow-control valves to fail in a safe manner depended on a small pressurized air flask located at each flow-control valve. Each air flask was protected from depressurizing by a check valve, but those check valves were also not routinely teste In summary, the availability of the EFW two-hour backup supply air system and the fail-open feature of the EFW flow-control valves were dependent on valves which were not tested in the position required to fulfill their function. This weakness was discussed with the licensee and will remain an unresolved item pending followup by NRC Region I (50-289/86-03-25). .

~ TRAINING The team considered the management commitment to training at TMI-1 a strengt This commitment was evidenced by the corporate training policies and plant procedures establishing goals, priorities, resources, and authority regarding the implementation of training. However, the most notable evidence of the licensee's commitment to training was the quality of the various training activities observed by the team. Details are provided in the following observation . The quality of the licensed operator requalification training programs was considered a strength due to the following observations: The requalification classroom training program was well balanced between significant topics and review of plant and industry exper-ience. There appeared to be effective communications between the operations department and training staf The requalification program training materials generally consisted of high-quality lesson plans, video aids, and trainee handout The attendance at requalification training' sessions, including off-shift licensed personnel, was determined to be consistently goo The requalification program weekly quizzes were comprehensive, and a check of the grading against the examination key revealed no discrepancie _

,

O 2 Annual requalification examinations were reviewed for compre-

'

.,

hensiveness, difficulty, and grading. The examinations were determined to be challenging and had a good balance among various types of questions, such as true and false, fill-in-the-blanks, multiple choice, and essay questions. The grading of the examinations was found to be adequat The inspector observed portions of the annual requalification train-ing on the B&W Simulator by the licensed operators of one shift. The inspector observed the shift's response to five abnormal plant drill conditions and noted that the shift personnel displayed good team work, excellent communications, procedure utilization, and rapid identification of the problems. The drills were video taped with sound to assist the shift in critiquing their actions. Additionally, the inspector noted that senior plant management representatives routinely participated in portions of simulator training. Their primary function was to monitor and evaluate the perfonnance of the operators being traine The Basic Principles Trainer provided strong reinforcement for classroom and simulator training by demonstrating the effects of individual instrument or component failures. The Basic Principles Trainer appeared to be particularly effective in training operators to recognize and respond to Integrated Control System (ICS) failures '

and for technicians to diagnose ICS failure . The Operational Experience Feedback Program was comprehensive and readily provided information to the plant staff by means of required reading, training letters, and requalification lecture Lessons learned from reactor trips due to personnel errors and equipment failures were emphasized as part of this program, with particular emphasis placed on errors caused by failure to adhere to procedure . S' .ificant plant modifications were incorporated into the requalifi-cation training program and taught prior to plant startup. Informa-tion on modifications that had only minor impact on plant operations was provided to operating shifts by training letters, Drills to practice cooldown from outside of the control room were conducted annually. These drills appeared to be comprehensive and covered:

(1) immediate control room evacuation, with all shutdown and cooldown actions performed outside of the control room; and (2) the initiation of a reactor trip and emergency boration prior to leaving the control room. Personn'el performance during these drills was evaluated by operations and training supervisor . The team reviewed the emergency feedwater and the instrument air sections of the Operating Plant Manual (OPM). The OPM was used as a reference manual for training. No errors were found in the emergency feedwater section; however, the instrument air section contained several error These errnrs were not found in the instrument air lesson plan, and no case was found where the OPM was directly used for training presentation .

.

.. The instructor initial and advanced training courses were considered a s trength. The initial instructor development course, taught semiannually, was a comprehensive basic course conducted by the Training and Educational Section. All operator training instructors were required to attend this course. A review of records revealed that all operator training in-structors had completed the initial training and most had attended the advanced training course . Maintenance staff training was considered a strength due to the indoctrination training program, the frequency of training, subjects taught, vendor training, on-the-job training, and the selection and training of instructors. The following observations were made: The two-week indoctrination program appeared to be effective in providing maintenance personnel with an overview of plant systems, basic plant safety, and maintenance fundamental The continuing training program for maintenance personnel appeared to be effective and comprehensive. Maintenance personnel typically attended one week of classroom training during each six-week rotation cycle. The training topics included: industry experience, admin-istrative procedures, and craft-specific trainin The Instrument and Control technicians had implemented a compre-hensive entry level-to-journeyman qualification program, referred to'

'

as the Automatic Mode of Progression (AMP) Program. This program required a technician to pass written tests and practical examina-tions for advancement. In addition, a technician was required to satisfactorily complete a requalification program, including a written examination, every two year Failure to successfully complete this program could result in reassignmen Quality Assurance The Quality Assurance (QA) program was found to be generally strong and effective. The team determined that the QA program exceededQA the minimum auditors, requirements of 10 CFR 50, Appendix B and ANSI Standard inspectors, and monitors appeared to have the necessary training and experience to identify many of the design and design control problems found by the NRC inspection team. An example of a GPUN audit with tech-nical findings in the design area was a recent corporate audit of en-gineering design calculations (Audit 0-COM-85-08). This audit resulted in three findings and 14 recommendations regarding technical issue Other examples where audits produced in-depth technical findings were the corporate audits of the GPUN Architect Engineer, Gilbert Commonwealth In (Audits 0-TMI-86-01 and 0-TMI-84-01). Likewise, on-site audits and monitor reports uncovered some technical and de' sign problems similar to those identified by the NRC inspectors. Overall, it appeared that the GPUN QA organization was fully capable of identifying in-depth technical and design issue . Despite the demonstrated capability of the corporate audit group as discussed above, the NRC inspection team identified design problems that had not been previously discovered. Examples were the design and design control problems with emergency feedwater (EFW) upgrade modification task

- 32 -

__ ,

,

.

'= RM-13H (see Design Changes and Modifications observation 1). Engineering review by the corporate QA Design and Procurement Section apparently was not in sufficient technical depth to reveal the deficiencies identified by the NRC tea In addition, a corporate audit of design control (0-TMI-84-06) reviewed this modification as one of a sample of seven, but identified only programmatic and procedural issues rather than technical problems. Another example of a design review deficiency that had not been previously identified related to theTheNRC team finding tha on-site QA Design Changes and Modifications observation 3.b).

engineering section design review of mini-mods failed to identify this prograrunatic weaknes . The inspection team found efficient and effective systems in place at TMI-1 to track and provide management review of actions required to correct deficiencies identified by the QA organization and other source Corrective actions by site personnel were usually prompt and complet Corrective actions by corporate groups appeared to take more time and An did not seem to be as consistent in resolving the identified problem example of a deficiency where timely, effective corrective action did not appear to have been taken by the responsible The licensee corporate Changes and Modifications observations 3.d, 3.f, and 3.g).

was cited by the NRC in 1981 for eight instances of non-compliance due to -

improper drawing control. Corporate and site audits of design control,

,

drawing control, the construction and modification program, plant engineering, and other areas had repeatedly identified discrepancies inThe drawings and the drawing control process. verified that these drawing pro action had apparently been ineffectiv . The inspection team assessed the licensee's independent technical and safety review process with emphasis on the activities of the P Group (PRG)."GPU Nuclear Safety Review and Approval," which provided corpora and controls for independent safety reviews, and TMI Division Procedure 1034, " Plant Review Group." This review verified that these procedures satisfied the requirements of Technical Specifications with regard to plant safety reviews.and an interview with the Manager, Nuclear Safety (who was site safety reviews) revealed that the on-site review program apparently adhered to these procedure Additionally, the NRC inspection team did not identify evidence of deficiencies in the safety reviews perfonned by the Plant Review Group. The training of the GPUN safety reviewers during 1985 was examined and was found to be satisfactor IV. MANAGEMENT EXIT MEETING An exit meeting was conducted on March 27, 1986, at TM TheAn additional exit licensee's meeting was conducted on April 7, 1986, at Bethesda, M representatives at each of these meetings are identified in the Appendi Mr. James M. Taylor, Director, IE; Mr. James G. Partlow, Director, Division of Inspection Programs, IE; and Mr. H. B. Kister, Branch Chief, NRC Region I,

- 33 -

. _ _ . - . . - -

.

.

.

l :

, -,

attended the March 27, 1986 exit meeting. The scope of the inspection was j discussed, and the licensee was informed that the inspection would continue i

with further in-office data review and analysis by team members. The licensee was informed that some of the observations could become potential enforcement findings. The team merrbers presented their observations for each area in-spected and responded to questions from licensee's representatives.

.

i l

I

.

l .

I o

t

- 34 - l

l

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

,

,

.

APPENDIX

,

Persons Contacted The following is a list of persons contacted during this inspection. There were other technical and administrative personnel who also were contacte All personnel listed are GPUNC employees unless noted otherwis +* F. Wilson, Vice President, Technical Functions

+ K. Croneberger, Director, Engineering and Design

• D. Hukill , Director, TMI-1

• J. Toole, Operations and Maintenance Director

• T. Shalikashvili, Manager, Plant Training

• E. Ballard, dr. , Manager, TMI QA Modifications / Operations

• C. Kazanas, Director, QA

• A. Nelson, Manager, Nuclear Safety

• J. Chisholm, Manager, Electric Power and Instrumentation

• P. Barbieri, TMI-1 Secondary Plant Manager

• G. R. Capodanno, Fluid Systems Director

• Behrle, Director, Startup and Test

• T. M. Hawkins, Manager Startup and Test

• C. W. Smyth, TMI-1 Licensing Manager

,

• R. J. McGoey, Manager, PWR Licensing

• J. J. Colitz, TMI-1 Plant Engineering Direc.or

,

• E. Neidig, Communications

• V. Hassler, Licensing Engineer

• C, A. Shorts, Manager, Technical Functions

• J. Smith, Project Manager, Gilbert Commonwealth, In *J. H. Brendlen, Jr., Project Engineering Manager, Gilbert Commonwealth, In D. Shovlin, Manager, Plant Maintenance P. Snyder, Preventive Maintenance Manager R. Harper, Corrective Maintenance Manager G. Lawrence, Lead I&C Foreman J. Bowman, Lead Electrical Foreman R. Natale, Lead Mechanical Foreman C. Hartman, Manager, Plant Engineering B. P. Leonard, Operator Training R. W. Zechman, Technician Training M. J. Ross, Plant Operations Director H. B. Shipman, Senior Operations Engineer D. W. Atherholt, Operations Engineer L. L. Ritter, Plant Operations Administrator C. Incorvati, QA Audits Supervisor J. Fornicola, QA Systems Engineering Manager ,

J. Marsden, QA Engineering Manager L. Wickas, Operations QA Manager

• Attended exit meeting on March 27, 1986.

+ Attended exit meeting on April 7,1986

- 35 -

.

.

_,' G. Sadavskas, Technical Functions I&C Manager S. Divito, Design and Drafting Supervisor D. G. Slear, Engineering Services Director D. J. Shivas, Engineering Data and Configuration Control Manager R. L. Summers, Plant Engineering and Mechanical Engineer J. H. Horton, Engineering Mechanics, Engineer J. W. Schmidt, Radiological Engineer S. Ku. Technical Functions Mechenical Systems Engineer

.

.

- 36 -