ML20141F085

Responds to NRC Re Violations Noted in Insp Rept 50-289/96-201.Corrective Actions:Will Perform Analysis of Environ Effects of Letdown Line Break & Will Evaluate Effects of safety-related Equipment in Affected Areas

ML20141F085
Person / Time
Site:	Three Mile Island
Issue date:	06/23/1997
From:	Keaten R GENERAL PUBLIC UTILITIES CORP.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
50-289-96-201, 6710-97-2242, NUDOCS 9707020031
Download: ML20141F085 (46)

Text

a

{ G PU Nuclear, Inc,

( One Upper Pond Road NUCLEAR Parsippany. NJ 07054-1E5 Tel 201-316-7000 June 23,1997 6710-97-2242 U. S. Nuclear Regulatory Commission Attn.: Document Control Desk Washington,DC 20555

Subject:

Three Mile Island Nuclear Station, Unit 1 (TMI-1)

Operating License No. DPR-50 Docket No. 50-289 GPU Nuclear Response to inspection Report (IR)96-201 A design team inspection was performed at TMI-1 by the Special Inspection Branch of Nuclear Reactor Regulation (NRR) and its contractor Stone & Webster Engineering Corporation (SWEC) during the period November 12,1996 through January 10,1997. The inspection team performed a comprehensive, in-depth examination of the design and licensing basis documentation for the Makeup and Purification (MU&P) and Decay Heat Removal (DHR) Systems. The MU&P System includes high pressure injection (HPI) and the DHR System includes low pressure injection (LPI).

A public exit meeting was held at TMI on January 30,1997.

Byletterdated April 15,1997 NRC provided Inspection Report (IR)96-201, "Three Mile Island -

Unit 1, Design Inspection." The cover letter stated that the team noted that the design documents for the reviewed systems appropriately implemented the intent of the design and licensing basis except for the specific cases identified in the report. Appendix A of the report listed the open items, which were categorized as either unresolved items (URis) or inspection follow-up items (IFis). The letter requested that GPU Nuclear provide a schedule for completion ofour corrective actions for the open items within 60 days. GPU Nuclear's request for an additional day was granted by Mr.

Robert M. Gallo, Chief Special Inspection Branch by telephone on June 20,1997. Enclosed is the GPU Nuclear response to that request.

The NRC's letter of April 25,1997 identified those open items from IR 96-201 that were considered to be potential violations and needed to be discussed in a predecisional enforcement conference. During the conference which was held on May 22,1997, GPU Nuclear agreed with the i NRC's assessment ofmany of the issues that were raised. In our presentation we provided some additional perspective on the issues raised in the repon and related some of the important actions that are being taken to prevent a recurrence of those findings including: the implementation of a new process-based engineering organization, development of a new self assessment program, and the J l initiation of a new corrective action process (CAP), as well as many of the actions that are being t taken to address the specific findings in IR 96-201. y I g

.A y 9707020031 970623 ll ll llll11lllll l PDR ADOCK 05000289 llll lllll3)lslll,lmlQllill,lillliIl i

ip G PDR

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

t 4

6710-97-2242 Page 2 of 2 1

i I l

This letter provides an update on the corrective actions, including a current status and schedule for completion of the remaining items. The information provided in Attachment I does not significantly differ from the comments provided by GPU Nuclear at the predecisional enforcement conference except that after further discussion with the NRC staffour positions in response to the findings represented in Ols 96-201-14 and 96-201-18 have been revised. GPU Nuclear has the resources i and is committed to completing these actions within the current schedule or as close to these dates

as possible.

Sincerely, i

R. W. Keaten Vice President and Director, Engineering MRK Attachment cc: Administrator, NRC Region I TMI Senior NRC Resident inspector TMI Senior NRC Project Manager I

i 1

l

)

i l

~_ _. _ . _ . .- ...___.. _ .. _. _ ._ _ _. _

=

Attachment 1 GPU Nuclear Response to NRC Inspection Report (IR)96-201 IR %20I included 30 Open Items (Ols) listed in Appendix A (Ols %201-01 through %201-28 inchxiing the three part OI E201-17A, B and C). For cach of these Ols, this attaciunent gives the short caption for the finding from Appendix A, followed by A) a summary description of the finding, B) a i discussion in response to tic finding, C) tic planned corrective actions, and D) GPU Nuclear's current )

scledule for completion of the remaining actions. Note that the caption for cach of the Ols in Appeixlix A of tic IR includes (in parentheses) a reference to tic specific section in the IR wicre tic 01 is discussed.

l 1

I 1

l 1

1

6710-97-2242 Attachment 1 Page1of43 IFI %201-01 " Letdown line Break in the Auxiliary Building (Section El.2.2.2.a)"

A. Description of the fmding:

The inspection team noted that a properly approved analysis of the consequences of a letdown line break in the Auxiliary Building (AB) was not availabic. The team inferred this issue to the technical resiew branch in the Office of Nuclear Reactor Regulation (NRR) stalT for resiew regarding the extent to which TMI-l was required to consider the efTects of a letdown line break in the auxiliary building. The staff resiew concluded that the TMI-l licensing basis for pipe breaks includes tic postulation of full diameter breaks in the letdown line between the containment penetration and the breakdown orifice as described in Appendix 14 A to the FSAR. Therefore, the design of safety-related equipment in the alTected areas should consider the conditions resulting from these breaks.

B. Discussion:

GPU Nuclear has been unabic to retrieve the 1973 analysis, which is referred to in Updated FSAR Appendix 14A, Section 4.3, page 14 A-11. In the absena: of this analysis, GPU Nuclear is evaluating whether the consequences of this event are acceptable. Our evaluations to date indicate that the plant could be safely shut down if such an esent occurred because:

1. With respect to pipe w hip andjet impingement:

a. The postulated letdown line break does not produce suflicient energy to cause the letdown line to whip. MU-V3 and its associated operator, power and control signals would sunire the break ,

because of 1) the relative location of this equipment to the postulated breaks 2) the physical  ;

protection prosided for the operator by an intentning platfonn, and 3) the large si/c of the i pneumatic operator considering the break energy.

b. Postulated breaks are not in locations that could alTect other safety-related equipment by either pipe whip orjet impingement.

2. With respect to the emironmental conextuences of postulated breaks:

a. Safety relatd equipment in '.cca, Equipment Qualification (EQ) zonc 7. is qualified to sunive the post Loss d Coolant Ar.cident (LOCA) emironment inside containment.

b. The static pressuri/ation arnlysis perfonned by the architect engineer, Gilbert Associates, Inc.

(gal) in 1973, while simpic, yicids a valid conclusion that pressuri/ation in the area is very small as long as letdown is isolated in a short time (minutes) aller the break,

c. Based on operator actions, the esent would be tenninated mpidly enough to prevent an enviromnent requiring qualification under 10CFR 50.49 Evaluations to date show that the consequences of the letdown line break do not result in consequences that would cause fuel damage or oft-site releases that exceed 10 CFR 100 limits. The statements in the FSAR have been substantiated by our resicus to date. Manual actions, and possibly automatic isolation sigrmis used to protect the core during Small Break LOCAs would be efTective during this esent and the efTect of these i consequences on the Reactor Coolant System (RCS) would be bounded by presiously analyzed cold leg breaks.

l l

l 1

. _ _ . .- . _ . _ . _ _.___._._m.__- _ . _ _ . . . . . _ . . _ _ .. ._. . __. _ . . . . _ . _ . - _. .

6710-97-2242 Attachment 1 -

L Page 2 of 43 C. Corrective Actions: ,

I

'1, _ Perform analysis of tic cmiro unental cfTects of a letdown line break between the containment and the block orifice.

2, Detennine tic cfrects of this emironment on safety related equipment in tic affected arcas.

3. Resise tic FSAR to reference the new analysis. t D. Schedule for Completion of Corrective Actions: ,

1, . Analysis is scheduled for completion by October,1997, ,

2. Evaluation of tic cfTects on safety-related equipment will be completed by December 31,1997. We >

cxpect that the results will be acceptabic.

3. l'evision of the FSAR will follow in the next FSAR update following completion of tic analysis arx!

cvaluations.

t 1 l I

l i

l l

1 f

i s

1

l l, 'p -*

i'- .

)

6710-97-2242 l

- Attachment I-Page 3 of 43

]

i IFl %201-02 " Evaluation of Simultaneous Start of MU&P Pump MU-P-IC and Suction and Discharge Vahts (Section E l .2.2.2.b)"

A. ' Description of the Finding:

In the normal operating configuration ofIhe Makeup and Purification System, Makeup Pmnp MU-P-1C is isolated from tic Makeup Tank (MUT) by manual vahrs and isolated from the Borated Water Storage Tank j (BWST) by Emergency Safeguards (ES) moter operated vaht MU V-14B. Makeup Pump MU-P-1C is not i lined up to any water source until MU-V-14B opens. Tic inspection team was concerned that tic clTect on tic pump due to a slow opening suction valve combined with a rapid start of the pump and a fast opening j high pressure injection vahe had not been analyzed.

1 B. Discussion: I As described in tic inspection report, GPU Nuclear committed to reduce tic design basis vahe stroke time  !

for MU-V-14A/B in the Surveillance Procedure (SP) 13(X)-3H,"lST of MU Pumps and Vahrs," from 22 i seconds to 13 secomis and to analyze the efTect on system perfonnance ofconcurrent maximum suction valve i stroke time, minimum pump startup time and minimum discharge vahr stroke time. '  !

l 'l. The simultancous start of Makeup Pump MU-P-lC with opening of MU-V-14B was tested during initial startup tuing in 1974. MU-V-14B stroke time was recorded at 9 seconds during that test.

I

2. The valve stroke time of MU-V-14 A/B (motor operated stop check vahrs) has remained 9 seconds or less as verified by a resiew of test records.

I j

3. ' The initial startup test ecsults are consistent with a reasonabic cxpectation for the vahe and pump )

combination. The minimum flow rate for a TM1-1 Makeup Pump is 40 gpm which can be met at less

{

than full open valve positions. After 3 seconds of vahe travel, MU-V-148 pennits a flow area of .

I approximately 27 sq. in. Using an equivalent pipe size for this flow arca, Cranc Handbook 410 " Flow of l Fluids througinahes, fittings, and pipe," shows approximately 100 gpm flowrate with a pressure drop j l near 1 psid. Ticrefore, simultaneous operation will provide sufficient flow to meet the minimum pump 1 l- flow requirements u hich are significantly less than full ECCS flow.

)

l ' 4. MU-P-1C is presently operabic.  ;

l. 1 2

i 5. If MU-V148 stroke time increased by 45% (from 9 seconds to 13 seconds), tic pump would still have l sufficient flow to meet tic pump minimum flow requirements during a startup transient at less than full  ;

open valve position. i

C. Correcthe Actions

1. Suntillance procedure 1300-3H, "lST of MU Pumps and Valves " was revised in Resision 46, effecthe March 26,1997. Data slect D now identifics the design basis stroke time for MU-V-14 A/B as 13 seconds.'

D. Schedule for Completion of Corrective Action:

l No additional corrective action is required to resolve this issue.

~

i l

l 6710 97-2242 Attachment i Page 4 of 43 IFl 96-201-03 Evaluation of Gas Accumulation in Suction Piping for MU&P Pump MU-P-1C (Section El.2.2.2.b)"

A. Description of tle Finding:

j MU-P-1C suction piping is not normally lined up to a source of water. Tic inspection team identified the I potential for accumulation of non<ondensabics, such as hydrogen, released from tic stagnant water in tic )

suction line because of tic physical configuration of the line. During tic inspection. tic vent valve in tle -i high point of this pipe section was opened and it was verified that there was no gas accumulation in tic suction pipe. Tim inspection team considered that a positive pump suction pressure was not necessarily an indication of absence of gas accumulation in the piping. (El.2.2.2.b)

B. Discussion:.

As described in the inspection report, GPU Nuclear has instituted periodic checks of the MU-P lC suction pressure as an interim measure. An operator checks the MU-P lC suction pressure once per day as required on the Primary Auxiliary Operator (AO) Log sheet. If the suction piping pressure is  ;

maintained at greater than 30 psig, no mechanism has been identified which could cause the accumulation of non-condensabics in this pipe section.

C. ' Corrective Actions:

1. A long term resolution of this concern is being developed.

D. Schedule for Completion of Corrective Actions:

1. Action to determine a long tenn resolution is being tracked by Licensing Action Requc<:1(LAR) 97052.02 w hich is scheduled for completion prior to startup following the Cycle 12 Refueling Outage scleduled to begin in September 1997. j i

l

)

l

6710-97-2242 i Attachment 1 Page 5 of 43

URI %201-04 " Adequacy of Makeup Tank Pressurc/ Level Cunts (Section E1.2.2.2.c)"

1 A. Description of the Finding:

The NRC inspection team identified several nonsonservative assumptions in the calculations that sent as the basis for the Makeup Tank (MUT) Net Positive Suction Head (NPSH) and gas entrainment cunes.

B. Discussion:

The lowcr pressurc/ level cune on Figure I of Opemting Procedure (OP) 1104-2," Makeup and Purification System," is designed to ensure adequate Makeup Ptunp NPSil u hen taking suction from the MUT and is based on a limiting scenario w here a Reactor Coolant Sy stem (RCS) break occurs on the nonnat makeup line. The operator starts the second Makeup Pump and opens MU-V-217 following a reactor trip in accordance with Abnornul Transient Procedure (ATP) 1210-1," Reactor Trip." This scenario is limiting because of the high Makeup Pump flow rates and rapidly decreasing NPSH availabic. The ch:dlcnge is over upon Emergency Safeguards (ES) actuation or operator action to open MU-V-14. The primary concern raised with c:dculation C-1101-211-53604)03," Makeup Pump NPSH," invohrd the use of a 1600 psig backpressure versus a O psig tuckpressure at the break on the normal makeup line.

Since lower backpressure will increase the drawdown rate from the MUT, GPU Nucicar agreed with the inspectien team on makeup line back pressure and acted expeditiously to resiew these concerns. Preliminary I calculatio is and sensitivity analyscs were used as part of the operability resiew which concluded that the system was operable but degraded based on the short period of time with degraded NPSH. Procedure changes were promptly initiated to place operational limitations on makeup flow with MU-V-217 open to i preclude the potential for degraded NPSH. l Preliminary analy scs to support the operability review has demonstrated that the Makeup Pumps would be in l a degraded NPSH condition for less than thirty seconds. Makeup Pump perfonnance is not required for core i cooling during the period of degraded NPSH. It was detennined that the cfTcct of operation with the degraded l

NPSH for tic limited time period would not prevent the pump from performing its design function following l ES actuation. Therefore, when the suction valves from the BWST open, the system is restored to a fully functional condition and there is no adverse impact on nuclear safety.

1 The upper pressure /lesci cune on Figure I of OP 1104-2 is designed to prevent Makeup Pump gas entrainment uhen taking suction from the MUT during high pressure injection prior to the MU-V-14 valves opening. The inspection team raised concerns with sestral non-consenative design inputs in calculation C-1101-211-53104)47," Makeup Tank Drawdown During LOCA." These included the omission of some Makeup Pump suction pipe and fittings and undertstinuting Makeup Pump flow, minimum Borated Water Storage Tank (BWST) level, BWST water temperature, and MUT uscable volume.

GPU Nuclear agrecs that these concerns result in minor non conscnative conclusions in the calculation.

Howeser it should be noted that the BWST temperature and MUT volume concerns have minimal impact and that the other obsenations arc oscrconc by the overall conservatism of the c:dculation.

The gas entrainment analysis consenalively postulates a large break Loss of Coolant Acciaent (LOCA) and assumes full flow from both trains of Low Pressure injection (LPI) and Buildmg Spray (BS). In such a scenario the High Pressure inja: tion (HPI) System is not required for accident mitiga' ion. Small and l intennediate si/c breaks result in slower BWST/MUT dnmdown rates. The slower drawdown rates afTord I the operator ample time to make a successful switchoser of suction to the BWST. Gas entrainment of the l

Makeup Pumps is not a concern following switchover because of the large inventory of water available as a suction sourec. Therefore, the HP1 System will perform its intended safety function when required to do so and there is no acherse impact on nuclear safety.

. .- . - . . . . _ . . . . -. . - . _ . _ - . - - _ _ . ~ - . - . - . - . - _ .

d ,

6710-97-2242

. Attachment 1 )

Page 6 of 43 .

l C. Corrective Actions: -

l

1. Operating Procedure (OP) 1104-2. " Makeup & Purification System." Resision 107, c1Tective January 30,1997 and Abnormal Transient Procedure (ATP) 1210-01, " Reactor Trip," Resision 36, l cffective January 31,1997 were resised to limit makeup flow to less ilum 500 gpm when taking manual t action to open MU-V-217 following a reactor trip. This satisfics the vendor recommended NPSH '

requirenents throughout the postulated Makeup line becak scenario.

2. Calculations C-1101-211-5360-003," Makeup Pump NPSH." Resision I and C-1101-211-5310-047,

" Makeup Tank Drawdown During LOCA," Resision 0 are being revised to address the relevant non-  !

conservative assumptions identified during tic inspection.

3. Operating Procedure (OP) 1101-1, " Plant Limits and Precautions." will be resised to deletc the irrelevant infonnation about uscabic MUT volume.

4. The System Design Basis Document (SDBD) will be revised to clarify information on MUT volume.

5. Programmatic improvements in the calculation preparation and control processes are addressai in 01  !

, 96-201-28, Corrcctive Action C,4.

D. Schedule for Completion of Corrective Actions:

1. Revisions of calculations C-1101 211-5360-003," Makeup Pump NPSH." Revision I and - {

C-1101-211 5310-047," Makeup Tank Drawdown During LOCA," Resision 0 arc scheduled for completionl3 eptember30,1997; S

2. The System Design Basis Document (SDBD)is scheduled to be updated by August 30,1997. ,

3. Procedure 1101-1," Plant Limits and Precautions." is scleduled to be resised by Da:cmber 31,1997.  !

l l

l f

6710-97-2242 Attachment i Page 7 of 43 URI 96-201-05 " Design Basis Vahe Stroke Times in Surveillance Procedure (Section El.2.2.2.d)"

A. Description of the Finding: l FSAR Table 14.2-14 states that the Emergency Core Cooling System (ECCS) delay time assumed in the Loss l I

cf Coolant Accident (LOCA) accident analysis is 35 seconds. The licensec stated that this delay time is c>mposed of I second for instrumentation lag,10 seconds for start of the cmcrgency power source if offsite l power is not available, and 24 seconds for system response (pump acceleration and vahe stroke time). FSAR l Section 6.13. I states that the system is designed to be in full operation within 25 seconds after receiving an actuation signal, and Suncillance Procedure (SP) 13(X)-3H,"lST of MU Pumps and Valves," Resision 44, prosides a design basis stroke time of 25 seconds for the makeup pump recirculation isolation valves MU-V-36&37 and the injection isolation valves MU V-16A through D. The team observed that the 1 25 second startup delay time added to the 1 I second delay of the actuation signal results in a total delay time of i 36 scconds, which would be an unanalyzed condition as stated in the System Design Basis Document l

(SDBD). This clunge to the facility imd not apparently been resiewcd in accordance nith 10 CFR 50.59. The licensec initiated a work request to revise the design basis stroke times in procedure 1300-3H and initiated a l clunge to the FSAR. I B. Discussion:

l l

The inspection identified documentation discrepancies between the FSAR, SDBD and SPs associated with the  :

High Pressure injection (HPI) System response time. The inconsistencies between these documents were all l related to whether a i second delay for instnunentation lag was consistently applied. The FSAR states thm the Makeup pumps and vahes need to operate within 25 seconds. SP 1300-3H,"'.5 r of MU Pumps and Valves,"

identifies a design basis stroke time for HPI valves of 25 seconds.

A 35 second HPI Sy stem response time is assumed in the LOCA analy sis u hich verifics ECCS perfornumcc l in accordance nith 10CFR50.46. This time includes an instnunentation and actuation logic cystem delay l (l sec.), an emergency dicsci generator startup & load time delay (10 sec), and allowance for HPI components l to start and valves to stroke (24 seconds). l The vahe stroke times in question have always been well within the design basis. The slowest actual valve stroke time is approximately 15 seconds vs. the design basis allowable of 24 seconds. The discrepancies j identified Imd not resulted in any iruppropriate design changes or inadequate mainteitmcc or testing. The FSAR and SP are bcing conected. The inaccurate identification of the valve design tusis stroke time in the SP and FSAR had no adverse impact on nuclear safety.

C. Corrective Actions:

1. SP 1300-3H has a maximum stroke time of 18 seconds for MU-V-16A through D and a maximum stroke time of 19 seconds for MU-V-36 and MU-V-37. These maximum stroke times do not need to be changed.

Exceeding these maximum stroke times would love caused the alTected component to have been declared inoperabic and would have started the Technical Specification time clock in accordance with Administrative Procedure (AP) 1041,"lST Program Rcquirements." The alTected design basis stroke times in SP 13(X)-3H have been resised to 24 seconds to address the documentation discrepancies.

2. FSAR Section 6.1.3.1 states that the Makeup & Purification System is designed to be in full operation within 25 seconds aller receiving an actuation signat A Proposed FSAR Update (PFU) has been prepared (PFU-98-TI-124) to change this to read as follows

"The LOCA analysis assumes HPI operation 35 seconds aller actuation. This allows 10 seconds to energi/c the on site cmcrgency power source, I second for instrumentation response and 24 seconds for Makeup System component response time."

! D. Schedule for Completion of Corrective Actions:

l l 1. PFU-98-TI 124 will Se included with FSAR Update 14, which is currently duc in April 1998.

l I

l l 6710-97-2242 ~

Attachment 1 i Page 8 of 43

\ .

IFl %201-06 " Consequences of Failure of Auxiliary Steam Piping (Section El.2.2.e)" .

A. Description of tic Finding:

Tim consequences of failure of non safety-related Auxiliary Steam (AS) piping in the Auxiliary Building

- (AB). Failure of the auxiliary steam piping due to a scismic crent could result in degradation of safety-related equipment classified for a mild emirorunent.  ;

Discussion:

B.

Although GPU Nuclear has been unable to retrieve tle final report docunenting tic analysis of pipe breaks in tic AB, a draft of tic report has been located.

l Tim AS lines in de Fuci Handling and Auxiliary Buildings are low pressure and thus were imestigated for cracks only. The pressure is too low to achicyc critical flow, and the staunjet force from dic largest postulated steam line crack is small. No damage to electrical or Instrumentation and Control (l&C) equipment will result.

A preliminary calculation, which was completed before tic cnd of the inspection, shows timt leakage from a

. crack ( 1/2 pipe ID x t/2 thickness) is 580 lbm/hr at i 1 psig line pressure. This is equal to a leat load of approximately 700,000 blu/hr, which in our engineeringjudgment is not suflicient to cause a harsh j cm5nment.

Tic routing of the AS lines in ihe AB is shown in Updated FSAR Figures 14A-3 and 14 A-4. The conclusions  ;

in die Updated FSAR lunc becn substantiated by reconstituted analysis although this analysis has not yet 1 design verified.

Tic finding expressed in IR E201 raised de additional concern of failure of de AS line due to a scismic -

event. Based on tic preliminary results from a letdown line break in the AB, it does not appear timt tic failure of tic AS line would result in a harsh environment before the break could be detected and isolated.

C. Correcthe Actions:

1. Analysis will be performed to a) confirm our engineeringjudgment that the heat load from a crack break is insullicient to cause a harsh emironment, and b) the consequences of a scismically induced failure of tic AS line is bounded by the letdown line break.

2. The ellects of an AS line break in de AB on safety-related equipment will be evaluated.

3. Documentation will be completed to support statements in the Updated FSAR regarding the consequences of an AS System line break in Oc AB.

D, Schedule for Completion of Corrective Actions:

1. Tic analysis is scheduled for completion by October,1997.

2. Evaluation of the effects on safety-related equipment will be completed by December 31,1997. We do not expect de results will show any effects on safety related equipment.

3. Changes to the FSAR will be submitted in FSAR Update 14, which is currently due in April 1998.

1 p.

l.._...._.

6710-97-2242 l Attachment 1

- Page 9 of 43 "IFl %-201-07 " Loss of Pressure in MU&P Tank due to Letdown Line Break (Section El.2.2.c)"

l A. Description of tic Finding: l l

Evaluate Oc impact ofloss of pressure in the Makeup Tank (MUT) on tic liet Positive Suction licad (NPSH)

! for tic High Pressure injection (HPI) Pumps due to letdown line break or er ek combined with um failure of tic check valve in tic line to tic MUT.

B. Discussion:

1. A break upstream of tle block orifice large enough to depressurize tic lire to less than MUT pressure would res dt in Emergency Safeguards (ES) actuation and automatic closure of MU-V-3 & MU-V-2A l and MU V-2B. Check v.hc MU-V-115 would prewnt backflow from the MUT to tie break. If this cime:k vahe faki, da gas overpresuic in Oc MlJF would be lost. The loss of MUT pressure is not an  ;

imnediate threat to Makeup Pump operation. Adequate NPSH is available to tic operating Makeup  !

Pump without any gas ostrpressure and minimum operating water level. The pressure and level in tic MUT are continuously moniscrW a:n if tle required pressure and level in the MUT and Oc Reactor ,

Coolant Systern (RCS) pressurizer cruld not be maintained, a controlled phmt shutdown and cooldown )

would be initiated. l

'I

2. The letdown line downstream of tic block orifice is Scismic I piping at low pressure and temperature - {

Tiere is no licensing or design requirement to postulate a break in this line. i

)

C. Corrective Actions:

No acklitional action is required.

j

l 6710-97-2242 Attachment 1 Page 10 of 43

IFl 96-201-08 "M&UP Pump NPSH When Taking Suction From BWST or Makeup Tank (Section El.2.2.c)"

i A. Description of tic Finding:

GPU Nucicar was asked to resiew the available Net Positive Suction Head (NPSH) for the High Pressure injection (HPI) pumps when taking suction from tic Borated Water Storage Tank (BWST) at " Low-low Lest!" or during potential vacuum conditions in the nukeup tank.

B. Discussion:

Abnomul Transient Procedures (ATPs) 1210-6, "Small Break LOCA Cooldown," Resision 23; 1210-7 "Large Break LOCA Cooldown," Resision 23 " and 1210-10, "Abnornal Transient Rules, Guides, and Graphs," Resision 32, were effective on March 19,1997. These procedurcs ensure that the " piggy-back" mode is established prior to reaching the BWST " Low Lescl" alarm at 9 ft 6 in. Therefore, with 9 ft 6 in of level in the BWST, the Makeup Pumps invc adequate NPSH under maximum flow operating conditions.

With MUT lcsti and pressure nmintained in the normal opcmting band, there is no mechanism or specific sequence ofcsents that could result in a vacuum in the MUT during tank drawdowit The operator is required to open the suction vahe from the BWST if MUT level reaches 55 inches. Mlff indicated Icsci would lave to be less than 0 incies before sacuum conditions develop. Therefore, the potential for vacuum conditions in the MUT is precluded by system design and operational controls.

C. Corrective Actions:

No additional action is required.

.: I 6710-97-2242  !

! Attachment 1 Page11 of43 IFl 96-201-09 " Incomplete DC System Voltage Drop Calculations (Section E1.2.3 2.a)"

l A. Description of the Finding:

l L Calculation C-110l' -734-5350 004,"TMI-l DC System Calculation." Resision 1, prmides irxiividual DC l

circuit voltage drop analyses for the redundant station battery circuits. The circuits for Makeup System f l isolation valve MU-V-18 and numerous oller DC circuits were listed in tic calculation, but tic witage drop '

analyscs were not performed. The calculation also concluded that furtler investigation was required to I determine the adequacy of tenuinal voltages at various DC equipment.  ;

B. Discussion: '

l

(

i A preliminary voltage drop analysis was perfonned during tic inspection that demonstrated adequate voltage was awlable at the tenninals of MU-V 18. l l

C. Corrective Actions.

1. Calculation C-1101-734-53504)4 "TMI-l DC System Calculation." will be updated to address the availabic voltage at MU-V-18 and the remaining circuits. The updated calculation is expected to also continn adequate voltage is availabic at the tenninals of MU-V-18.

D. Schedule for Completion of tic Conectisc Actions:

1. Tic revision 10 calculation C-1101-734-53504)4 "TMI-l DC System Calculation," is scheduled to be completed by December 31,1997.

i l

, - , . . - - , . - - ~. - - - . - . = n , . - .-- -

6710-97-2242 Attachment 1 Page 12 of 43 URI %201 10 " Alignment of MU-V-18 DC Power Supply (Section El.2.3.2.b)"

A. Description of the Finding:

During the inspection GPU Nuclear told the inspection team that the loss of 1 A or iB 125 VDC power concunent with an High Pressure Injection (HPI) line break downstream of the last check vahc on the injection line would prevent the HPI system from perfonning its design Emergency Core Cooling System (ECCS) function if the power source lincup and the vah c positions in die cross-connect line between Makeup Pumps A and B were not controlled. If the system is not properly aligncxt, the loss of a 125 VDC power train could result in failure of MU-V-18 to close. thus, degrading one HPl train and the other HPl train would be inoperable due to the loss of the 125 VDC breaker control power from the same source. The team uns told that this problem had been discovered before the inspection team arrived at the site. But when the alignment was checked on January 8,1997, it was discovered diat the electrical powcr aligrunent was incorrect. The licensce issued temporary orders to control room stalT to pct fonn checks of the cir-connect vahc positions and DC power source iilignment every shift. The licensec issued PFU 98-Tl N resisc the FSAR and initiated changes to the operating procedures. T he team noted that the requirem0.mn 10 CFR 50, Appendix B, Criterion V, " Instructions, Procedures, and Dmwings" regarding prescribing activitics alTecting quality by documend instructions, procedures or drawings and accomplishing the activitics in accordance with these instructions, procedures, or drawings were not apptrently met.

B. Discussion:

The alignment of the power supply for MU-V-18 is not required to ensure nuclear safety. Further resiew ,

following the design inspection determined that regardless of the power alignment for 1M DC distribution panel and MU-V-18, there is no single failure w hich can reduce HPI capability to less than that assumed in the ECCS Loss of Coolant Accident (LOCA) analysis.

The concern at the time of the inspection was for a break in the nonnal makeup line at the Reactor Coolant System (RCS) noule and an assumed failure of the "A" or "B" DC distribution sy stem, that only one HPI Pump would be operating and that the minimum HPI flow would not be supplied to the RCS. This concern was based on de MU-V-18 vah c failing open concurrent with the redundimt train HPI Pump failing to start on loss of DC power.

A more detailed analysis concluded that these events (MU-V-18 "open" with HPI Pump failure) would not occur for the postulated single failure. Two scenarios were considered. One where the Emergency Safeguards (ES) actuation occurs prior to the loss of DC power and a second w here dic ES actuation occurs aner de loss of DC power.

In de first case, both trains of HPI are initiated and then MU-V-18 fails "open" w hen DC is lost. The running iIPI Pump does not trip w hen DC power is lost. With two HPl Pumps operating the flow through MU-V-18 does not prevent HPI from perfonning its nuclear safety function.

In the second case, the IM DC distribution pmel transfers to the ahernate pouct source w hen DC is lost and MU-V-18 will close upon ES although only one HPI pump will start. One HPI train is capable of providing die minimum HPI flow assumed in the ECCS LOCA analysis.

There was no adverse impact on nuclear safety from this finding. This concern was bart on an incomplete j analysis of the events at the time of the inspection.

C. Corrective Actions:

1. FSAR Section 6.1.3.1 will be revised.

I

. . . _ _ . . . ~ _ . _ _ . . _ _ . . . . . _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . _ . . . . . . . _ . . _ _ _ _ _ . . _ _ _ _ _ _ _ .

t 6710 97-2242

• i . Attachment 1 1

Page 13 of 43 ,

D. Schedule for Completion of Corrective Action: ,

1. A change to FSAR Section 6.1.3.1 has been prepared (PFU 98-TI-119) and will be incorporated into FSAR Update 14 which is cunently due in April 1998.

b e f

e

?

i 1

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

I 6710-97-2242 Attachment 1 Page 14 of 43 l URI 96-201-11 " Makeup Tank level Instrument Imop Tolemnces (Section El.2.4.2.a)"

l A. Description of the Finding- 1 The taun resiewed the instrument data sheets in Suncillance Procedure (SP) 1302-5.17, Resision i

17 "Make-up Tank lesel Instrumentation." The data shec' for instrument loop MU14-LT specified  ;

a loop crror (tolerance) of *1% Howescr, calculation C-1101-662-53504)49 "TMI-I Makeup Tank lxscl Error for Accident Conditions (LT-778 Loop)," Resisjon 0, estimated a loop crror of

• l.23%

Although the loop crror in the data sheet was more restrictive, the tann questioned the inconsistency in the two documents. Tic licensec stated that for all safety-related instrument loops a consen atist loop crror ofil% was assumed. Documentationjustifying this assumption could not be retricted.

After further res icw of suncillance data sheets and calculations, the tea n identified three other instances of inconsistency in the documents. The data sheets for control room indicator instnament loop MU-LI-778A and computer point A0498 instnunent loop MU14-LT specified a loop crror of 1% cach, but calculation C-1101-662-53504)49,"TMI-l Makeup Tank Lcycl Error for Accident Conditions (LT-778 LOOP),"

Resision 0, detennined that the loop crrors should be limited to i 0.64% and 0.73% respectively.

Caletdation C-1101-624-53504X)2, " Makeup Tank Level Error for Accident Conditions (MU-14-LT Loop),"

Resision I, detennined that the loop crror should be 0.57% for MUl4-LR in instrument loop MU 14-LT, but the data sheet specified a loop crror ofil.0. The calculated allowable loop crrors were more restrictive than the instrument loop data sheets. The licensee initiated action to resolve the team's concerns.

B. Discussion:

Multiple calculations of Makeup Tank (MUT) level inarument loop aauracy are on record. These calculations were identified with different system designations. Tolerances detennined by these calculations did not match the acceptance criteria in the SP 1302-5.17 l

Calculation C-1101-624-53504)02, " Makeup Tank Lesci Ermr for Accident Conditions (MU 14-LT Loop),"

is for MU-Vl4-LT loop crrors. Resision 2 of this calculation was to support a modification that would have 1 replaccxl the analog Integrated Control System /Non Nuclear Instrumentation (ICS/NNI) System with a Digital Control System, but the modification was not installed.

Calculation C-1101-662-53504149 "TMI-I Makeup Tank Level Error for Accident Conditions (LT-778 LOOP)," is for LT-778 loop crrors.

Calculation C-1101-211-5350-057, " Makeup Tank Level Instrument Drift," w hich was intended for MU-14-LR and MU-LI 778A Instrumentation Drift, was not found until after the inspection. It supports the 1% toicrance for some of the loop outputs.

The surveillance program has maintained the instnunent cdibration to a high degree of accuracy (allowable error was 1%). Calculation C-1101-211-53504)57, ' Makeup Tank Level Instrument Drift," supports the existing suncillance tolerance for the critical loop outputs.

In order to assure that any potentially similar problems with other process measurements and setpoints are addressed, plans for a setpoint basis update program was initiatoj in response to the NRC's 50.54(f) letter of October 9,1996. This program will preside a data base for reference to the design basis document (i.e.

setpoint calculation) for cach Nuclear Safety Related (NSR) and Regulatory Required (RR) setpoint. If a design basis docimient can not be found, an engineering task will be initiated to have the scpoint design basis established. Once a design basis document is established, the document w;11 be referenced in the setpoint program database as a source document. This program will ensure that design basis references for setpoints are captured and casily retricsuble.

_ n_ - _ . . . . - ~ ,- - - - , . . . . - . . .. _- - ~ -

l l .

6710-97-2242 l Attachment 1

! Page 15 of 43 l

i C. Corrective Actions:

l.

1. A new calculation which includes all the outputs from MU-14-LT and MU4.T-778 will be preptred.

2. SP 1302-5.17, "Make-up Tank lxvel Instrumentation," will be revised to agite with the tolcrances in the resised calculation.

3. Tic calculation process upgrades will allow quickcr micw and identification of calculations w hich support key plant parameters. (Sec 0196-201-28 Corrective Action C.4)

D, Scledule for Completion of Corrective Actions:

1. Tic new calculation for Makeup Tank level instrument loop accuracy is sclwduled to be completed by July 301997.

2. SP 1302-5.17. "Make-up Tank Lesel Instrumentation." is scheduled to be revised by August 30,1997.

i

! I l

6710-97-2242 Attachment 1 Page 16 of 43 URI 96-201-12 "FSAR Discreparcies (Section El.2.7)"

A. Description of tic Finding:

Tic team identified four discrepancies between de FSAR and oller plant docunentation. These four items wereidentified as follows:

1. FS AR Tabic 14.2 18, Sheet 1, indicated duit " total flow" for a High Pressure Injection (HPI) line break at a Reactor Coolant System (RCS) pressure of I800 psig is 347.5 gpm rather dian 397.5 gpm.

2. FSAR Section 14.2.2.4.3.a indicated that an open issue existed with regard to tripping the Reactor Coolant Pumps (RCPs) during a Small Break less of Coolant Accident (LOCA) but did not state duit a manual trip on a loss of subcooling margin was acceptabic.

3. FSAR Table 9.1-2 and tic GMS-2 database contain conflicting design data for MU&P system components.

4. FSAR Section 8.2.3.1.b describes tic Emergency Diesel Generator (EDG) continuous rating as 2GX) kw instead of the correct rating of 2750 kw stated in the vendor manual, VM-TM 0191, Resision 29.

B. Discussion:

The resolution of these discrepancies requires a revision of the FSAR. Tim GPU Nuclear process for FSAR updates requires that a Preliminary FSAR Update (PFU) be submitted. A PFU includes die recommended text resision and the required safety resiew. PFUs have been submitted for all of de discrepancies identified.

Each of dose discrepancies is discussed as follows:

1. Table 14.2-18 indicated HPI flow of 347.5 gpm rather than 397.5 gpm. No analysis or testing /suncillance activitics were identified where this error uns continued. Tic HPl flow test and LOCA analysis both used tic proper value of 397.5 gpm.

2. Section 14.2.2.4.3.a indicated there is an open issue with regard to tripping RCPs. This error had no impact on nucicar safety because the "open issue" had been resohrd. This omission in Oc FSAR could only tune caused sonconc to raisc Oc question as de inspection I;am has dore.

3. Tabic 9.12 contaited erroncous Makeup System comporent data. Tic crroncous component data were corrected to re' lect the equipment installed. A review of Omsc crrors did not identify any installed equipment Gat is not qualified for its senice.

4, Section 8.2.3.1.b indicated an EDG continuous rating of 2MX) kw rather than 2750 kw. Had tic crroncous value of 2NX) kw been used as the EDG rating in an evaluation, the conclusion reacted would have been conservative because tle actual EDG rating is 2750 kw.

In September 19'X>, a new cmphasis was placed on the quality and completeness of tic FSAR. FSAR section owners have been established as a step to improve the quality.

6710-97-2242 '

l Attachment 1 ,

i Page 17 of 43 <

\ '

i l  !

I C. Corrective Actions:

l-I j l. As stated in the inspection report, the following PFUs were prepared to correct tim four discrepancies l identified atxnc: PFU 98-TI-ll3; PFU 98-TI-123; PFU 98-TI-127; and PFU 98-TI-131 It is

! noteworthy that after a PFU has been submitted, the change is placed in a datatxtse for immediate use by reviewers in perfonning safety cvaluations. '

D. Schedule for Completion of Corrective Actions:

1. The above listed PFUs will be included in FSAR Update 14, w hich is currently due in April 1998.

1

+

6 1

I

6710-97-2242 Attachment 1 Page 18 of 43 URI %-201-13 " Adequacy of BWST Setpoint for DHRS Pump Switchoscr to RB Sump (Section El.3.2.2.a)"

A. Description of the Finding:

The design inspection team identified live concerns with a 1994 calculation provided to demonstrate the adequacy of the plant design for Borated Water Storage Tank (BWST) switchover The analysis costred the evolution where it is required to switch the Emergency Core Cooling System (ECCS) pump suction to the Reactor Building (RB) sump w hen the BWST is depleted in a Loss of Coolant Accident (LOCA) scenario.

The NRC concerns were: (1) failure to consider instrument error in the alann, (2) use of a non-consenative RB pressure assumption, (3) the validity of the assumed time for operator response, (4) disagreement between the valve stroke times used and the design basis time stated in the valve stroke test procedure, and (5) improper selection of a constant and invalid flow assumption in the application of methodology for vortex detennination.

B. Discussion:

The concerns of the inspection team were culuated. A brief assessment of the error and the signifiamcc of that crror in the anal 3sis resuhs is crahiated as follows:

1. The failure to consider instrument crror was caused by a misconception in that, since the "Lo Lo Lescl" alann setpoint was detennined using instrument crror, the error was already accounted for and would l not need to considered again in this analysis. The actual instrument crror was measured on December 21,19% and found to be very small.

2. The RB pressure asumptbn was the best available information and was felt to be an adequate assumption for the purpose of the analysis being perfonned. The infonnation diat confirmed that this 1 was not a conscnative assumption was not available until December 19% during Oc inspection. The change in RB pressure had a significant non-consenative effect on the anal3sis. .

3. The operator response times were taken from simulator perfonnance records. The appropriateness of data from plamied s ersus unplanned activitics in tic simulator was not considered. There were some misinterpretations of the data due the manner in which plant computer data is collected and presented.

The operator response times used were reasonable.

4. The valve stroke times used in the analysis were appropriate. However, the sun cillance procedures u hich test de stroke time of these valves had acceptance criteria greater than the assumed stroke time.

5. The calculation of de minimum level at a specified flow mte required to prevent air entrainment due to vortexing was based on TM1-1 stanup data and a methodology from an industry periodical. The startup test was not performed at a constant flowrate and this was not obvious in the presentation of the results of die test. Also, the empirical method described in the periodical was valid with only a limited range of constants. The TMI voncx analysis included errors associated with die flowrate from the TMI-I startup test and use of a constant outside the prescribed range for the methodology used. These errors produced a small non-conscnative change in the result.

The overall impact of these errors was enluated when the concerns were raised. The above items 1,3,4 & 5 together would not ha e changed the conclusion of the analysis. However, the new RB pessure assumption resulted in a signifiamt decrease in die time available to completc the evolution and meet de acceptance criteria. GPU Nuclear initiated work on this problem immediately on Dec. 20,19% w hen the preliminary drawdown analysis using the res iced RB pressure indicated a potential problem. At 4:00 am on December 21,19%, GPU Nuclear determined diat cunent analysis methods could not definitively conclude that some quantity of air would not be drawn into the suction of the Decay Heat (DH) and Building Spray (BS) pumps in the Large Break LOCA scenario with conservative assumptions for the plant conditions.

Therefore, the DH and BS pumps were declared inopembic.

6710-97-2242-Attachment 1 Page 19 of 43 A reportable event was declared and reportal to NRC by phone and by submittal oflicensec cent report (LER (X>4X)2).

Prior to expiration of the Technical Specification Allow Outage Time (AOT), an alternative procedure had been implenented which allowed GPU Nuclear to conclude that air entrainment in tic ECCS pumps would not occur during the postulated LOCA.

C. Corrective Actions:

1. A resised switchover sequence was developed which increased the time for operator action to completc l the transfer to the RB recirculation mode prior to reaching the point of vortexing. Temporary procedure changes were implemented on December 21,19'X) for the toised switchover sequence.

Abnornut Transient Pmcedurcs (ATPs) 1210-6,"Small Break LOCA Cooldown," Revision 23; 1210-7,"Large Break LOCA Cooldown," Resision 23,1210-8,"RCS Superheated," Resision 20; and 1210-10, "Abnonnal Transients Rules, Guides, and Graphs," Resision 32, were cfrective March D,1997 to replace the temporary changes.

2. All licensed operators were bricfed on the procedure clumges.

3. The new switchover sequence was successfully performed at the TMl-1 Replica Simulator using the raised procedural guidana:.

4. All persons u ho are certified to perfonn calculations or design verifications have received additional training on the calculation and design verification process with special emphasis on validation ofinputs and assmnptions.

5. Fonnal analysis will be completed to document the new procedure for switchoser.

D. Schedule for Completion of Corrective Actions:

1. The revised BWST switchover analysis will be completed by December 31,1997. Other actions listed above are complete.

l

6710-97-2242 l Attachment 1 Page 20 of 43 l URI 96-201-14 " Adequacy of Safety Evaluation of an FSAR Change (Section El.3.2.2.b)"

A. Description of die Finding:

Safety Evaluation (SE) No. I 15403-004, Rev. O, did not identify diat, because the required Net Positive Suction Head (NPSH) for the Decay Heat Removal (DHR) Pumps would not be met without taking credit for contaimnent overpressure, die probability of occurrence of malfunction of the DGR Pumps pres iously cvahiated in the safety analysis report may be increased, and thus, a potential Unresiewed Safety Question (USQ) as defined in 10CFR50.59 was involvec'.

B. Discussion:

Safety Evaluation SE-1154034X)4, Rev 0, considered that the assumption of no credit for contaimnent overpressure above the sump vapor pressure only applied to the licensing basis accident analysis condition of 3(xx) gpm low Pressure Injection (LPI) flow rate and 1500 gpm RB Spray flow rate, and adequate NPSH was demonstrated at these conditions. A flow rate limited to 3300 gpm by Abnormal Transient Procedure

( ATP) 1210-7,"Large Break LOCA Cooldou n," was considered to be beyond the licensing basis as presently defined by the TMI-I FSAR and the NRC Safety Evaluation Report (SER) dated July 11,1973 for TMI-1.

Since this condition was interpreted as beire beyond the licensing basis, it was determined that use of consenutive but more realistic assumptions in Penns of containment oscrpressure was acceptabic.

While applying containment overpressure could agit;mately be viewed as a design control concern under Appendix B to 10 CFR 50, at the preliminary enfc,'ecment conference on May 22,1997, GPU Nuclear questioned uhether SE No. I15403-004, Resision t should have detennined diat an unreviewed safety question existed by considering assumptions beyond the licensing basis. Safety Evaluation 115403-004, Resision 0, explicitly documented the rationale subr tantiating the safety evaluation conclusions, and conservatively incorporated instrument error not prmously considered in the licensing basis. GPU Nuclear requested further consideration and guid:mcc from Jic NRC Staff on this question. We suggested diat safety evahiations should be performed in a mamier cmsi ; tent with licensing basis assumptions because the objective of those evaluations is to detennim, w hether a change preserves the plant's licensing basis. Based on additional discussion with NRC Staff, GPU Nuclear understands diat it is the NRC's position diat use of containment overpressure for the 3300 gpm LPI flow rate is a desiation from the existing licensing basis.

GPU Nuclear understands that 10CFR50.59 and US? criteria interpretation is an evohing issue and is continuing to develop based on issuance of NRC dwument, SECY 97-035. We are participating in various industry groups iciated to this issue and contimm. to monitor these activitics to fully understand how to more cITectively address 10 CFR 50.59 criteria.

GPU Nuclear understands that this open item reflects the NRC staffs concern that the NPSH calculation for i die DHR Pumps should assume LPI flow of 3150 gpm (the LPl flow at 0 psig presented in FS AR Table 1 14.2-27 and derived from the flow cunes used in the large break LOCA analysis) and an appropriate allow 1mcc for instnunent error. In contrast, Section 6.4.2 of the FSAR reficcts a licensing basis calculation that assumes LPI flow of 3(XX) gpm (the nominal flow rate used in the accident analysis) with no allowance for instrument error.

1 We belicyc that this NPSH calculation, using the licensing basis assumptions, is not sufliciently consenalive.

We also believe that the NPSH for the DilR Pumps should be based on an LPI flow of 3300 gpm, based on the ATP 1210-7 procedure flowrate limit, coupled with an appropriate allowance for instnunent crror. This will bound dic 3150 gpm flowrate described in FSAR Table 14.2-27.

The existing licensing basis only addresses the accident analy sis assumed LPI flow rate of 3000 gpm. Safety Esuluation i15403-004, Resision 0, recegni/cd that ATP 1210-7 allowed a higher LPI flow mtc. This ,

higher LPI flow rate provides additional margin beyond the accident analy sis value in tenns of core cooling l (10 CFR 50 46) concerns. The suhic used for the assumption on contaimncut overpressure was conscnative l but reflected an expected Reactor Building response to the postulated design basis accident. Safety

l o

l 6710-97-2242 Attachment 1 Page 21 of 43 Evaluation i 154034X)4, Resision 0, detennined the NPSH available to tic DHR Pumps to be adequate at both the 3(XX) gpm accident im.Jysis value and the 3300 gpm procedurally limited flow value. ,

Tic calculated NPSH availabic for tic DHR Pumps at conditions of 3300 gpm LPI and 1500 gpm RB Spray Ilow was approximately 0.5 A. (0.22 psi) less than required with no credit for contaimnent overpressure based l

on a 1990 calculation and tic associated safety evaluation. Abnonnal Transient Procedure (ATP) 12104)7,

' "Large Break LOCA Cooldown," directs the operator to turn off tic RB Spray Pumps at a RB Pressure of '

4.0 psig. Under this condition there is an excess available NPSH of 2.9 ft. Thus additional NPSH margin is prosided w hen considering tic expected plant and operator response to tic postulated design basis accident.

No immediate corrective actions were required since this condition only involves a reduction in tic margin inchided in the NPSH determination and does not represent a safety or operability issue.

r i

! C. Corrective Actions:

1. GPU Nuclear will submit a license amendment application for NRC approval to resisc tic licensing basis so that the NPSH for the DHR Pumps is calculated based on an LPI flow rate of 3300 gpm and an appropriate allowance for instrument crror. Tic application will include use of containnent ,

overpressure in the calculation of available NPSH for the maximum pump flow rate of 33tX) gpm. Tic calculation will be resised accordingly. l Because this application will change the plant's licensing basis and NPSH calculation for the DliR Pump, and will be approved by license amendment, it will also climinate any USQ that may exist as a result of the revised profile evaluated in Safety Evaluation i154034X)4, Resision O.

D. Schedule for Completion of Corrective Actions:

1. The DHR Pump NPSH calculation using 3300 gpm LPI nmximum flow and an appropriate allowance for instnmient error will be completed to support submittal of a license amcminent application by August 1997 for NRC approval prior to startup following the Cycle 12 refueling (12R) Outage, which is scheduled to begin in September,1997.

l 6710-97-2242 i ~ Attachment 1

! Page 22 of 43 l

l IFl 96-201 15 " Technical Specification Discrepancy (Section El.3.2.2.c)"

A. Description of tic Finding:

l Technical Specification (TS) 3.3.1.1.f states: "The two reactor building sump isolation valves (Dii-V-6A/B) shall be citler nunually or remotely operabic." This requirenent was not appropriate because only remote operation of the vahts would suppurt the design txisis assumptions for the transition time of the Decay Heat Removal (DHR) Pump suction switclxnct from the Borated Water Storage Tank (BWST) to the Reactor Building (RB) Sump during post-accident conditions. In addition, numual operation of these vahts during an accident may not be possible because tic DHR System pump rooms may not be accessible due to high mdiation conditions.

B. Discussion:

Valve surstillance tests require these valves to be tested for remote operation. TMI-l las not been operated with these valves in a condition where they were only numually operabic.

C. Corrective Actions:

1. Technical Specification Clange Request (TSCR) No. 263 has been submitted for NRC resicw and approval. This TSCR deletes the words "cither nunually or" and includes other unrelated changes that were requested prior to startup for Cytic 12 operation.

C. Schedule for Completion of Corrective Actions:

1. This action will be complete upon issuance by the NRC of the amendment requested in TSCR No. 263.

l

6710-97-2242 Attachment 1 Page 23 of 43 URI 96-201 16 " Discrepancy Between FSAR and Technical Specifications Regarding DHRS Leakage (Section El.3.2.2.d)

A. Description of tic Finding:

FSAR Sections 6.4.3,6.4.4 and Tabic 6.4-3 stated tic design basis leakage in the Auxiliary Building (AB) from de Decay Heat Removal (DHR) System and Building Spray (BS) System as 2255 cm'/hr -

(0.6 gal /hr). FSAR Section 14.2.2.5 and Table 14.2-20 docunented a 2 leur dose for engineered safeguards leakage of 2255 ml/hr of 0.037 rem. i Technical Specifications (TS) 4.5.4, " Decay Heat Remosal System Leakage," allowed 6 gal /lu leakage from i tle DHRS and referred to FSAR Section 6.4.4 and Table 6.4-3. The TS bases stated that dose was 0.39 rem from tic 6 gal /hr leakage and used FSAR Section 14.2.2.5(d) as a reference for this dose. There were no Technical Specifications for BS System leakagc. The team noted that leakage control in the DHRS was more important Omn for the BS system since de BS system operates for a few hours after the accident while the DHR System operates over tle entire duration of the accident.

The discrepancy between the FSAR and TS was identifiert by tic licensce as inspection obsenation 212-61 in report TDR 1092 during a self-assessment of de DHR System performed in 1992. Tic pieliminary safety ,

significtmcc resiew, documented in memorandum C320-92 1287, stated that the obsenation was not safety significant because the dose of 0.39 rem "still represents a negligible portion of the total two-hour thyroid dose of I89 rem." No action was taken to resolve Oc discrepancy.

Tic licensce stated the FSAR would be resised as part oflicensing Action Rcqtest (LAR) 95076.01. The team noted that the LAR was initiated May 25,1996. Tic above discrepancies had not been corrected and the FSAR updated to assure that de information included in the FSAR contained tic latest material as

)

required by 10 CFR 50.7l(c).

B. Discussion:

GPU Nuclear acknowledges that tle discrepancy identified is valid and has initiated a corrective action item  ;

to evaluate this discrepancy and revise tic FSAR and oller documents as required. Tic discrepancy between tic FSAR and TS basis for Emergency Core Cooling System (ECCS) leakage las a very minor impact on de Maximum Hypothetical Accident (MHA) dose consequence result with minimal impact on nuclear safety as j assessed at tic time of tic law Pressure (LPI) Safety System Functional Inspection (SSFI).

C. Corrective Actions:

1. The MHA dose consequence analysis (Calculation C-1101 202-E260-329) is being resised and NRC resiew of a change will be requested to include a more conservatise leakage assumption consistent widi tic TS. '

2. Changes to TS 4.5.4 will be requested to include leakage from all " recirculation" systems.

3. Tic afTected FSAR sections will be resised.

D. Schedule for Completion of Corrective Actions:

1. Submittal of the revised assessment of dose consequences from the MHA using tic resised leakage assumptions is scheduled for submittal by carly July 1997.

2. The Technical Specification Change Rcquest to resisc tic TS Surveillance Specification Section 4.5.4 is scheduled for submittal by carly July 1997.

. - . ..... . .- - . - . .- - .. . - .- - . ~ .. . ... . .- . - . . . . . ~ . .- . .

6710 97-2242 #

Attachment 1 ~

Page 24 of 43

3. Tic affected FSAR sections will be resised in tic next FSAR update following completion of ticsc -

actions which require NRC approval. LAR 95076.01 will track tic completion of tle FSAR Update. l h

1 l

I 1

I I

l l

I l

t

! 6710-97-2242

[ . Attachment 1 i Page 25 of 43 URI 96-201 17A " leak Testing of DHRS Pump Suction Check Valves (Section E13.2.2.c)"

A. Description of the Finding:

Tic function of tle Deca / Heat Removal (DHR) System vahrs, DH-V-14 A and DH-V 148, had been evaluated as part of an Insenice Teing (IST) program scope resiew. A calculation was performed which

! accurately determined that there wac Mcquale pressure to force post accident water through this check . .

valve flowpath and become a potential source ofoffsite dose. However, the calculation was docimented only J in a memo. The inspection team questioned the Reactor Building (RB) pressure assumption in tic nemorandum. Tic Equipment Qtmlification (EQ) RB pressure profile had been revised aller the j memorandum had been prepared.

B. Discussion:

l-The resised RB pressure profile did result in higler RB pressure which would potentially challenge the DHR System check valves using the conservative assumptions typically assumed for such analysis. Additional RB .

pressure analysis was perfonned u hich demonstrated that with more limiting operating restrictions of RB pressure, Borated Water Storage Tank (BWST) temperature and RB fan cooler availability, it could again be assured that leakage through ticsc check valves could not reach the environment. Interim operating i restrictions were established to ensure that there would be sufficient margin pending Om destlopment and completion of appropriate IST.

l T1c potential safety issue is leakage past this clock valve in a post Loss of Coolant Accident (LOCA) condition where diat leakage then flows into the BWST which is vented to atmosphere. Although not formally proceduralized as a test, tic DH-V-14 A and DH V-14B check vahes have been functionally tested cach time the plant *.ransitioned to DHR operations. During this evolution, Reactor Coolant System (RCS) )

pressure (greater th m 200 psig) is applied across this check ulve and recent experience (tle last couple of l outages) has been 0 at no leakage was observed. Leakage is indicated by lining relief vahr DH-V-57 or l observation of BWS " level increasing. DH V-14A/B are disassembled and inspected as part of the IST program and significant wear would be detected and repaired.

C. Coritctive Actions:

1. Training has been prosided to the engineering stafTwho may perform design basis calculations. This training emphasiecd procedural compliance and specifically the use of tic EP-006 fonnat for all design calculations which includes the ie.cification of other affected calculations.

2, Tic operational restrictions put in placc December 20,1996 will remain in effect until Oc Cycle 12 .

Refueling (12R) Outage w hich is scleduled to begin in September,1997, Thereafter, beginning in the 12R outage, valves DH V-14A and DH-V 14B will be tested as part of the IST program. j 1

D. ScheduleofCorrective Actions

1. Leakage testing of the DH-V-14A/B will be implemented as part of the IST program during tle 12R Outage tests.

i I

• 6710-97-2242 Attachment 1 Page 26 of 43 URI 96-201 17B " Leak Testing of DHRS Pump Discharge Cleck Valves (Section El.3.2.2.c)"

A. Description of tic Finding:

i

(

In a design basis accident,if a Decay Heat (DH) Pump fails. tic discharge cross connect valves DH-V-38A and DH V-38B arc opened and flow is prosided through both low pressure injection paths. In this condition the discharge cleck valve (DH-V-16) on de failed pump would need to close to ensure adequate Low Pressure injection (LPI) flow. The Insenice Testing (IST) program did not recogniec and test this "Closc"

[ furstion of DH-V-16A and DH-V 168.  ;

B. Discussion:

The DH-V-16A/B valves are tested for their "open" function quarterly and during refteling outages. GPU Nuclear agrecs diat these valves have a "close" safety function. A test procedure was written and performed within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> ofdisco try of the condition to verify Oc "close" safety function.

i

. These vahts are periodically disassembled and inspected as part of the preventative maintenance program.

These inspections have shown that Oc valves are in good operating condition widi no significant signs of degradation. The "open" fimetion of thesc valves has been periodically tested as part of tic IST program.

Therefore, we believe that the vahrs would have functioned as required in the past. The IST program will continue to ensure Oc "close" functionality of dese vahrs in the future.

C. Corrective Actions:

1. A test on December 19,19% verified that vahrs DH V 16A and DH-V-16B scated in de " closed" position. Both vahes performed acceptably during the test.

2. IST procedures will be resised to include " closed" function testing of de DH V 16A and DH-V-16B valves.

3. "Close" testing of DH V 16A/B has been added to tle work scope for cold shutdown aethitics in tic event of an unscheduled cold shutdown prior to de IST procedure updates being completed.

D.. Schedule for Compiction of Corrective Action:

1. Changes to IST procedures to add " closed" function testing of the DH-V-16A and DH V 16B vahrs will be completed prior to de tests required during tic Cycle 12 Refueling (12R) Outage which is scleduled to begin in September 19(n.

i

6710-97 2242 Attachment i Page 27 of 43

URI 96 201-17C " Inspection of DHRS Pump Vault Floor Drain Check Valves (Section El.3.2.2.c)"

l A. Description of tic Finding:

$ As described in FSAR section 6.4.5 a leak in a Decay Heat (DH) vault will not affect more than one train of equipment because there are check valves in the floor drains to picvent water from backing up into the vaults.

11cre are two floor drains from each DH vault to the Auxiliary Building (AB) Sump. The inspection team identified that tic procedure which verifies tie operability of desc clock vahes did not include both valves.

B. Discussion:

Wien the team identified tic omission from tic procedure, action was begun immcdiately to disassembic and scrify the condition of Oc salves which were not being suncilled. Also, a procedure change was prepared to add these presiously omitted valves to tic procedure.

The procedure in question covered all of the floor drains in the auxiliary and fuct handling buildings. Each room in desc buildings has one of a typical floor drain ammgement with a swing check valve and special access for inspection and cicanout. Tic procedure costred all the drains of this t3pe. The Energency Core Cooling System (ECCS), Decay Heat, Building Spray and Makeup Pump Rooms cach have an additional drain with a level alarm. These drains have a ball float type check valve. Tlcsc ball float type check vahes were not being inspected. The purpose of tle procedure cicarly addresses tic IE Circular 7846 which identified Oc issue years ago. It also recogniecs that Oc concem is " potential flooding of ECCS rooms."

C. Corrective Actions:

1, These valves were inspected and found to be operabic.

2. Tic procedure U-17. "Zurn Floor Drain Inspection" was resised in Revision 5, cffecthe February 20,1997 to include testing Oc ball check vahes.

3. Each of tic ECCS pump room drain vahes will be added to the Insenice Testing (IST) program as augmented IST.

4. Flow Diagram 302 719, will be resised to identify Oc drains associated with tic Building Spray, Decay Heat and Makeup pump rooms.

D. Schedule for Completion of Corrective Actions:

1. The addition of ECCS pump room floor drain mhes to the Augmented IST program will be completed before stanup following Oc C 3cle 12 Refueling (12R) Outage which is scheduled to begin in September i997,

2. Flow diagram 302-719," Reactor & Auxiliary Bldg. Sump Pump and Drainage System" is scleduled to be resised by December 31,1997, t

I' l

l

. . . . - - . - - _ --. . - ~ . . - . - - - - . - - - . . - . . . . - - -

6710-97-2242 Attachment 1

- Page 28 of 43 URI 96-201-18 " Timeliness of Action on SSFI Open items (Section El.3.2.2.f)"

A. Desenption of de Finding:

This item deals with timeliness of action on Safety System Functional Inspection (SSFI) Open Items. As stated in de inspection report, GPU . Nuclear performed an SSFI of the Decay Heat Renxwal (DHR) System in 1992. Technical Data Report (TDR) No.1092, %w Pressurc Injection S 3stem - Safciy System -

Furstional inspection," was approved on January 12,1993. Tic team sclxted about 30 open items to verify tic status of corrective actions. In tic following instances, tic licensce delayed actions on SSF1 open items that had been pending for up to four years:

1. SSF1 Items 212-1 through 212-11 concem design basis piping analysis arx! related matters.

2. SSFI Items 212-12 and 212-13 relate to potentially non-conservative RB pressure and sump liquid temperatures used in the Low Pressure Injection (LPI) Pump Net Positive Suction Head (NPSH) calculation.

3. SSFI Item 212-42 questioned why de DHR Pump vent vahes, DH V-75A&B and DH V-76A&B, were not in the Equipment Qualification (EQ) program. ,

4. SSFI Item 212-61 deals with discrepancies between tic FSAR and the Technical Specifications on DHR System allowable leakage.

B. Discussion: -

Tic NRC has identified this URI as an example of potentially not satisfying 10 CFR 50 Apperulix B, Criterion XVI in that conditions adscrse to quality luxi not been corrected promptly. At the May 22,1997 Predecisional Enforcement Conference GPU Nuclear agreed that prompt action had not been taken on all LPI SSFI open items. GPU Nuclear is committed to prompt closcout of significant issues according to relative prioritics and availabic resources. GPU Nuclear agreed to resiew this issue and proside additional infonnation.

A broader assessnent of this URI has been conducted w hich agrees that the items identified during tic j design inspection and discussed in IR 96-201 are coixlitions adverse to quality. Tlcsc items should have been addressed in a more lincly marmer. Each are brictly desenbcd below. Although ticy arejudged to represent low safety significmcc, they are cumples ofcorxlitioits which need to be corrected.

GPU Nuclear offers the following additional infonnation regarding tic SSFI items listed in de inspection  ;

report:

l. SSFI Items 212-1 tluough 11: nese items represent problems associated with locating design infonnation, consistency between calculations and plant configuration, and die tieroughness of j doctmenting the basis of design decisions. Many of dcsc relate to work performcd in tic mid 1980s. j The initial review fourx! these to tune low safety significance, and as ticsc items are being completed, j we are finding that tle initial assessments in 1992 are being confirmed. The physical plant design has l proven to be conscnathe.

l 2. SSFI Items 212-12 and 13: Tlcsc items have been resched and closed with no safety impact or concerns. Delay in completing the NPSH calculation, once the required information was availabic, was dic to resource limitations.

3. SSF1 Item 212-42: This item resulted from not rescarcidng design infonnation sufliciently to

( understarxl the design basis of Decay Heat Pump icnt vahrs. An early assessment indicated dicsc should be in the EQ program, lxmeser later paperwork (not located during the Inspection) and assessment has indicated this was not warranted. Tlcsc vahes were put in Oc EQ Program to bc l l r

j

6710-97-2242 Attachment 1

- Page 29 of 43 conscivative arx! remain since it is beneficial, however they are not required to satisfy 10 CFR 50.49 criteria invoking clectrical equipment.

4. SSFI Item 212-61: This item is discussed as 01 %201-16. With regard to Criterion 16, this is an exampic ofinconsistency between the Technical Specifications and the FSAR and is a condition adverse to quality that was not closed in a timely manner.

GPU Nuclear has had an extensive program of scif-assessment through SSFis and Service Water System Operational Perfonnance inspections (SWSOPIs) as well as design basis reconstitution through the preparation of Design Basis Documents (DBDs). Ticsc cfTorts, fmm 1989 to 19%, identificd over 450 issues that imehui some form of follow-up action. Approximately 400 items have been closed with tic balance being managed through tic Corrective Action Process (CAP). GPU Nucicar places great value in finding and resohing issues as discussed at the Predecisional Enforcement Conference. Placing tic items in the new CAP Process was an action initiated prior to inspection 96-201. Imprming tic managenent of issues was one of tic reasons for initiating the major organi/alional changes in engineering in 1996. The change accomplished many objectives such as establishing a programs group to better focus attention and resources on Configuration Managenent and in following up activitics on various initiathes like DBDs arxl SSFis.

The results of work on the open items so far indicates that these represent issues of effective closcout tracking and documentation; there appears to be no substantial impact on nuclear safety associated with the delayed completion of these particular items. Based on the work completed to date, there is nothing to challenge the conclusions of tic original evaluations of the safety significarce of these issues.

To promote timely response to action items (such as those cited in URI %201 18), all open items from the System Design Basis Document (SDBD? prognun and the SSF1 and SWSOPI Inspections (a total of 65 items) Imc been entered into tic Correcthe Action Process (CAP) as conunitted to in the response to the NRC's 50.54(f) letter dated October 9,1996. All open items entered into the CAP and tic Engineering Division Task Tracking System will be tracked to completion.

C. Corrective Actions:

SSFI Items 212-12 and 212 13 are complete. SSF1 Item 212-42 is complete. Vahts Dil V-75A&B and DH-V-76A&B were stim to the EQ List as of February 5,1997.

D. Schedule for Completion of Corrective Actions:

1. SSFI Items 212-1 through 212-11. Those items that are not already closed are targeted for completion by December 31,1997. GPU Nuclear is in tic process of contracting out work on sevend of these items. Therefore, the target date of December 31,1997 could slip if external resources are unmailable.

2. SSFI Item 212-61 will be completed as described in response to 01 %201-16.

l

__ __ . .._ m . _ _. _ ... . _ .. _ .. _ _ . . . _ _ . _ . _ _ . _ . _ . _ . _ _ _ _.

1 6710-97-2242 Attachment i Page 30 of 43 -

IFl 96-201 19 " Evaluation of Reactor Building Sump Screen for Reverse Flow (Section El.3.2.2.h)"

- A. Description of tic Finding  !

The team questioned whether the Reactor Building (RB) sump screens had been anal >7cd to verify that they were designed to withstand pressures due to reverse flow wlen the Reactor Coolant Sysicm (RCS) drop line is opened during post-accident conditions to control boron precipitation in the reactor core.

B. Discussion:

This concern results from opening the Decay llcat drop line to the RE Scmp during a large break Loss of Coolant Accident (LOCA) as a means of avoiding core boron precipitation.

C. Corrective Actions:

1. Analysis is being perfonned to detennine tic maximum pressure tle RB Sump screens c m withstand if the RCS drop line were opened during post-accident conditions to control boron precipitation in the reactor core.

D.- Schedule for Completion of Corrective Actions:

1. The current phase of the analysis is expected to be completed by July 31,1997. Depending on the results. additional work could be required.

j 1

1 l

l l

l I

l

6710-97-2242 Attachment 1 Page 31 of 43 URI 96-201-20 " Testing of Molded Case Cixuit Breakers (Section El.3.3.2.a)"

A. Description of the Finding:

Molded case circuit breakers in two safety related Motor Control Centers (MCCs) for ventilation systems used during refueling had not been periodically tested since their installation in 1986 beciuse they were not included in the original nuintenance dataluse until 1993. They are now scheduled to be tested in 1997 and 1998. Also, in 1993, the testing frequency for a feeder circuit breaker in iB Emergency Safeguards (ES)

Valves MCC Unit 7A for the iS ES Engineered Safety Features (ESF) Vent was deleted from GMS-2. As of this inspection the circuit breaker in the IB ES Vahes Unit 7A was still within its four year maintenance cycle.

CAP Tl9974XX)6 was written to address the failure to include these breakers in the Preventhe Maintenance (PM) program. Tic 1 A & IB ESF Vent MCC feeder breakers and all breakers supplying ESF Ventilation system loads were tested during February 1997.

B. Discussion:

GPU Nuclear concurs that the ESF Vent MCC molded case circuit breakers ucrc omitted from the PM program. The 1 A and IB ESF Vent MCCs and associated molded case circuit breakers werc not added to the PM Program w hen they were installed in 1986. With the change from GMS-1 (a task oriented system) to GMS-2 (a component oriented sy stem) at the end of 1989 through the beginning of 1990, discrctc components (circuit breakers, motors, pumps, etc.) were loaded into the GMS-2 component database. The correct component-specific tasks were then established to test these individual circuit breakers. As part of the task sustructuring project, however, the task to t< 'hc breaker for IB ESF Vent MCC Fecdct brciker (IB ESV Unit 7A) was inadvertendy canceled.

The loads powered from the 1 A & IB ESF Vent MCCs are non-NSR loads. The ESF Ventilation System is classified as Regulatory Required (RR) and the system is shutdown by an ES signal. The Borated Water Storage Timk Heaters (BW-H4XX)1 & BW-H4X)02) are also feed from the ESF Vent MCCs. These heaters are iripped and locked out by an ES signal concurrent with a loss of olisite power. Components powered from these MCCs are not required for ES conditions. The molded case breakers feeding the MCCs were alway s uithin their test interval. The breakers wcrc operable to isolate Oc ESF Vent MCCs from either the i 1 A or IB ES Valves MCC in the event of a fault on the i A or IB ESF Vent MCC. Nuclear Safety was not impacted since an electrical fault on a non-Nuclear Safety-Related (non-NSR) component powered from cither ESF Vent MCC would have been isolated by a tested NSR breaker.

In 1988, the Maintenance Assessment group was fomed and this resiew was assigned as part of that group's responsibilitics. This resulted in the process becoming more formali/cd. Sr ecific individuals are now required to perfonn this resiew function, as outlinst i:i attachments to Administrative Procedures (APs) 102 I and 1043 for modification closcouts; omissions tv.h a ; this are less likely to occur with the process in place today.

C. Corrective Actions:

1. A resiew was conducted of the molded case breaker PM tasks to ensure that all NSR MCC molded case breaker are included in the test program. Twenty breaker tasks were updated as a result of this resiew  !

which was completed on June 3,1997.  ;

2. Tests have been completed on the three breakers (BW FH 1, BW-H4X)1 and BW-H4X)2) w hich arc  !

powered from the 1 A & 1B ESF Vent MCCs. 'I nese tests had not been completed by the close of the j design inspation. -

1

. - -~ . .. . . - . . . - . . . . . . . _ - . - . . - - -- . . . - --. . ~. . - . ... . - -... .-~. .. ...

6710-97-2242'

- Attachment 1 Page 32 of 43 r

3. A GMS-2 Task to test the molded case fecdct breaker to tic IB ESF Vent MCC has been re-establisted.

D. Schedule for Completion of Corrective Actionsi

1. Resiew of tie Nuclear Safety Related (NSR) MCC breaker PM tasks will tc completed by July 31,1997. Tic other actions described above ime been completed.

i l

l l

t I

i m.- m - .. e - + . w e .

6710-97-2242 Attachment 1 Page 33 of 43 '

URI 96-201-21 " Static Ikad Correction for Borated Water Storage Tank (BWST) Lesci Transmitter (Section El.3.4.2.a)" '

A. Description of tic Finding:

Suncillance Procedure (SP) 1302-5.19, " Borated Water Storage Tank Level Indicator," data shects shosv de head correction (due to clevation differences between de level transmitters and tic bottom of de BWST) arxi span correction (dic to tic specific gnnity of borated water) for Oc BWST level transmitters. No engineering documentation was found that details tic assumptions and references used to establish the

' calibration data.

B.~ Discussion:

Although Oc span and head correction values in tic procedure are correct, tic lack of a reference to a source docmnent for dose values is a weakness in tic procedure. No formal setpoint calculation exists for tic BWST level indication loops. A setpoint calculation would include the infonnation, references, and sketches necessary to define the basis for the setpoint. Without the setpoint calculation, alternate sources were used to establish the calibration data in the surveillance proccdure as discussed below.

Field sketcles show the proper head correction. Also, tic span and head correction infonnation used in the  !

suncillance procedure was obtained from engineering evaluations that show the proper head correction.

llowever, tic span correction value used in these evaluations did not include a source referencxr, and the head conection infonnation from the field sketches in the engineering evaluations was not incorporated into appropriate design documents, Piping and instrumentation Drawings (P&lDs).

l Tic SP did not contain any crrors related to lead or span correction. This finding invohes maintaining )

proper documentation of de basis for tic salues being used. .

In order to assure that any potentially similar probicms with otler process measurenents and setpoints are addressed, plans for a setpoint basis update program was initiated in response to the NRC's 50.54(f) letter of October 9,1996. This program will proside a data base for reference to de design basis document (i c.

setpoint calculation) for cach Nucicar Safety Related (NSR) and Regulatory Rcquired (RR) setpoint. If a design basis docunent can not be found, an engineering task will be initiated to hase tic setpoint design basis establisluxl. Once a design basis document is established, the docmnent will be referenced in the setpoint program database as a source document. This program will casure that design basis references for setpoints are captured and casily retrievabic.

C Corrective Actions:

1. Tasks were initiated during de inspection to track completion of the update documentation related to l Dil-LT-0808 and DH-LT 0809 including:  !

a. Drawings B-308-810 and D-308-920 were updated to show tic transmitter and BWST clesations. This prosides the basis for the head correction for cach transmitter.

b. The GMS2 database was updated to reference tic change document, correct drawing  ;

references, and update tic range for level transmitter, DH-LT-0809.

c. The vendor manual was added as a reference to the SP as the basis for the spam correction. The updated Piping and Instrurnentation Diagrams (P&lD), D-308-920 and B-308-810, were also included as references in de procedure.

2. SP 1302-5.19," Borated Water Storage Tank Level Indicator," was changed in Resision 19 on February 5,1997 to incorporate Oc appropriate changes.

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

1 6710-97-2242 Attachment i Page 34 of 43 D. Sclaiule for Completion of Correctist Actions:

Except for tic long term commitments in response to tic NRC's 50.54(0 letter oroctober 9.19<x> rercrred to abmt, tic oiler actions described abmt to address tic BWST level transmitters span and head correction ,

have been completed. .i i

i l

l l

i t

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

_ . ~ . . . - - -

i i .

6710-97-2242' Attachment 1 L ' Page 35 of 43 IFI 96-201-22 "BWST Level Instrunent DriA (Section E1.3.4.2.b)" )

- A. Description of tic Finding:

After de res icw of several instrument loop accuracy calculations, the team noted inconsistent treatment of drin errors in these calculations. For example: calculation C 1101-662-53504)59," LOOP Accuracy R.G 1 1.97 BWST latl Incorp. Suncillance LOOP Tolcrance (MTX)," Revision 0, which established tic loop l accuracy of Borated Water Storage Tank (BWST) level instnunentation stated an arbitrary value for drin as .

I one-half of Oc accuracy of the loop components; calculation C-1101-212-53504)51, "BWST lect l Instrunent Drill " Resisjon 0, empirically determined drift; GPU Nucicar Engineering Standard ES4X)2, l Resision 4, "Instrunent Error Calculation and Sctpoint Determination," addressed drill as a variabic to be considered while performing error analysis, but it did not olTer any guidance on how to calculate drift; and ]

calculation C87064)21 "R.G.1.97 RMT Transmitter leop Accuracy" did not address drift error in de loop

analysis, but formed the initial basis for the loop accuracy of de BWST level instnunents.

Licensce memo 5450-884X)23 in calculation C8706-021, stated that the BWST level instrument accuracy requirements were on hold pending completion of two other evaluations. The memo also stated that de basis for the loop accuracy requirements was included in Technical Data Report (TDR) No. 883, Resision 1. The team resiewed this TDR and did not find die loop accuracy bases. i Tic team noted instrument loop crror concerns similar to the ones discussed in 0196-201 11. Calculation C-1101-662-5350-59 stated that the existing surveillance toicrance for LT-808 and LT-809 was

• 2% A memo dated August 8,1988, attached to lle calculation stated that to be consistent with the usual approach which provided margin and allowed for additional drill the surveillance accuracy should be lowered to 1.5%  ;

Tic memo stated that supporting calculations were being developed. GPU Nucicar was unabic to proside these calculations. Tic discussion in the memo was not consistent with the present suncillance data sheets u hich specify a 1.0% loop accuracy which are more restrictive. Additionally, safety esaluation SE4XX)-2144X)l, Resision 2, Section 3.3.2.l(d) stated an allowance ofi 2% for level error for tic low-low alann generated by BWST lesel tmnsmitters DH-LT-808 and Dil-LT-809. These inconsistencies had not been resolved by the end of de inspection.

B. Discussion: 1 GPU Nuclear is pmcceding with tic actions described below.

C. Corrective Actions-l

1. GPU Nuclear Engineering Standard ES-002, Resision 4, " Instrument Error Calculation and Sctpomt -1 Detennination" will be used to develop a new calculation for BWST level instrument errors. Drift I will be determined from Oc manufacturer's specifications. The "as-left" calibration toicrance limit l will be established by tic calculation. The following explains tic disposition of documents upon issuance of Oc new BWST level calculation:

a. Calculation C-1101-662-5350-059, " Loop Accuracy RG 1.9) BWST Level hicorp.

Suncillance Loop Tolerance (MTX)," Revision 0, will be superceded by the new calculation.

b. Calculation C-1101-212-5350-051, "BWST Level Instnunent Dnft," Revision 0, will be an i input to de new calculation.

l l 2. Those portions of the architect engineer, Gilbert Associates Inc. (GAI), calculation C87064)21, "TMl 1 l New RG 1.97 RMT Transnutter Loop Accuracies," that are related to BWST level will be superceded  ;

l Note: TDR 883, Revision 1, "TMJ-l Equipment Qualification - Perfomiance Evahiation ofInstrument Loops," remains unaffected by de issuance of the new calculation since it does not address BWST level.

-i

O g 4 6710-97-2242 Attachment 1 Page 36 of 43 D. Scialule for Completion of Corrective Actions: '

l. Actions required to complete the new calculation are being tracked as task #2162 which is scialuted to be completed by July 30,1997.

2. GAI calculation C8706-021 is scinfuled to be revised by October 31,1997.

4

'k i

s e

r i

1 i

l t

6710 97-2242 Attachment 1 Page 37 of 43 URI %201-23 " Selection of BWST Low Level Alarm for Operator Action (Section El.3.4.2.c)"

l A. Description of the Finding: .

The initiation of tic switchover to sump recirculation was accomplished by a Borated Water Storage Tank (BWST) low level alarm from switch Dil-DPS-914. This action is a critical operation. The redundant action initiator is considered computer alarm A0486 from DH-LT4)M09. The NRC inspection team was concerned that critical operations were being initiatcxl utilizing non safety-related alarms.

B. Discussion:

Tic preparation and review process of the Temporary Change Noticcs (TCNs) that initiated the switchover l of Emergency Core Cooling System (ECCS) pump stk.: tion to the Reactor Building (RB) sump from the BWST utiliecd control room overhead annunciator alarms. Standard operator direction is to use redundant instrumentation uhere available which includes safety grade consolc indicators.

C. Corrective Actions: h

1. Abnormal Transient Proaxiores (ATPs) 1210-6, "Small Break LOCA Cooldown." Resision 13 and 1210-7 ' Large Break L(k A Cooldown." Resision 23 were completed on March 19,1997 to specifically include the use of consolc indication for the initiation of ECCS suction switchoser.

2. Operator training materials will be resised to reflect operator action based on the console Icsel indication.

D. Schedule for completion of Corrective Actions:

1. Operator Training lesson plan nutcrials will be resised and any additional training that may be necessary because of these changes will be completed by December 31,1997.

1

6710-97-2242 Attachment 1 Page 38 of 43 IFl 96-20124 " Undocumented Modification (Section El.3.6.2.b)"

A. Description of the Finding:

On December 19,19'X> dunng a walkdown of the Lav Pressure injection /High Pressure Injection (LPl/HPI)

Systems the inspection team obsentd a hue and metal tubing / fittings attached to the open end of discharge >

relief valve BS-V-63B. The team questioned w hether this was an acceptable configuration.

B. Discussion:

GPU Nuclear inunediately recognized and infonned the NRC inspection team that this was not an acceptable configuration (pwtubited by ASME code) and one timt was probably not resiewed or approved by site engineering. At tic conclusion of the walkdown, GPU Nuclear verified tint the configuration was not previously resiewed by engineering and removed the attachments to the relief valve. A corrective action task was initiated to develop any required actions to pres ent similar occurrences.

GPU Nuclear recogni/cd that this attachment to the relief valve is not allowed by General Design

'cquirements (USAS B31.7 Nuclear Power Piping Section I-722.6) and that this condition could cause a oackpressure on the relief valve, or fluid buildup on tic discharge that would affect the lift setpoint of the relief valve. There are no normal system conditions or analyzed accidents that create the condition where this vahe would be expected to lift.

C. Corrective Actions:

1. An Awareness Memorarclum was prepared u hich included the esent description, consequences. root cause, and corrective actions. The memorandum became required reading for all Operating Crews and Operations Management. As part of the corrective action, Shift Supenisors were required to discuss the following items with the crews:

a. If this type situation is know n or seen in any other area of the plant the una shori/cd equipment is to be removed.

b. If ticre are known problems with relief valves lifting carly or leaking by, this infonnation is to be addressed byjob ticket to correct the deficient condition.

c. If there are specific vahts u hich operators belicyc need a tail pipe, they are to notify the Shift Supen isor/ Operations Engincering so that an appropriate solution can be obtained.

l 2. Although not part of the official closcout of the ctrrective action task, the Awareness Memorandum was routed as required reading to all TMl Systcm Engineers u ho routinely perfonn system walkdowns.

D. Schedule for Completion of Corrective Actions:

All follow-up action is complete.

i

- ~ -.- - - - - - - . - . - . . - . . - - . . - . . - . - - - - - _ - - _ - - . .

L.,.-

!, 6710-97-2242

{

Attachment i .

Page 39 of 43 URI 96-201 25 "BWST level Transmitter Enclosure and Heat Tracing (Section El .3.6.2.c)"

i

' A. ? Description (,f tic Finding: '

Tic team noted that the cover plate of tic enclosure for Borated Water Storage Tank (BWST) lesci  !

instnunent Dil-LT-808 was open and tic fasteners for the cover were missing. No work was in progress that coeld explain Oc reason for 6e coser being open. The enclosure is a metal barrict that prmides separation i between redundant transmitters and process tubing. Tierefore, tic enclosure was installed in accordance -

with Section 3.3 of SP-9000-44-001, ." Instrument and Control Instrument installation," Resision 0, which t specifics tic use of prote tyc barricts w here separation criteria could not be met.  !

The team also observed that tic heat tracing for BWST safety-rciated level instrunent DH-LT-808 was Icfl i coiled witidn tic slect metal enclosure and not wrapped around the sensing lines or the transnutter. This configuration was not in accordance with tic vendor drawing (ET-30250, Revision 2) and inainterumcc

- procedure 1420-iffl, " Heat Trace Repair and Replacenent," Revision i1. Although tiere had been no history of freezing of tic transmitter or its associated tubing, tic team was concerned that tie transmitter luid  ;

tic potential to frec/c ifleft in the as-found condition.  !

B. Discussion: 4 Tic configuration was non-standard rciative to applicabic docuracnts: Ncison drawing ET-30250R2 and TMl heat tnice maintenance procedure 1420-HTI. Tic enclosure is prosided for physical separation and not l for Ocrmal insulation. Recent TMI heat trace design is governed by a sendor heat trace program. Tids program is designed to correspond to standard configurations (trace applied to pipe / tubing and with sonc minimmu amount ofinsulation) and thus may not produce correct results for tic subject application.

C. Corrective Actions:

1. Tic cmcr plate was reinstalled. l l

2. Suncillance Procedure (SP) 1302-5.19, " Borated Water Storage Tank Level Irxlicator," was resised in Resision 19 on February 5,1997 to ensure that Oc enclosure cover is closed after completion of the suncillance.

3. The leat trace installation luis been corrected.

D. Schedule for Compiction of Correctisc Actions:

No additional action is regtdred.

1 l

6710-97-2242 Attachmer't 1 Page 40 of 43 URI 96-201-26 "FSAR Discrepancies (Section EI.3.7)'

A. Description of the Finding:

Tic team resiewed the appropriate FSAR sections for the In ;y Heat Removal (DHR) System rad for the associated electrical and instrumentation and control systems.

Tic team identified the following discrep;mcies in the FSAR:

1. FSAR Section 6.1.2.1.b. I stated that the maximum flow through the DHR pump bypass line is 125 gpm * $ gpm w hen operating at a shutofT head of 425 feet. Also, tle system design basis document SDDD-TI-212, Revision 1, stated the same recirculation flow rata The flow as detennined in calculation C-1101-212-5360-008, Revision 0 was 150-155 gpm.

2. FSAR Tabic 9.5-2 contained several cr.trics of two design temperatm. and/c pressures for DHR pumps and coolers. The licensce rtated that these were incorrect.

3. FSAR Section 8.2.2.10.g stated tha non-tegregated, metal enclosed 4 f s V otWxts were used for major circuit runs from the unit auxiliay transfonners to 416n v ,md o90c, V '. ases. The equipment bill of materials showed a voltage of 7.2 KV for the bus ducts.

4. FSAR Tables 8.2.8 and 8.2.9 (Emergency Diesel Generator load ng) refcared to the Borated Water Storage Tank (BWST) heat tracing load as BS-T-2D, hov crer, this load designation was for the sodium hydroxide tank heat tracing.

B. Discussion:

Tic resolution of these discrepancies requires a resision to the FSAR. The GPU Nuclear process for FSAR updates icquires tint a Preliminary FSAR Update (PFU) be submitted. A PFU includes the recommended text resision and tic required safety resiew. PFUs have been submitted for all of the discrepancies identified.

Section 6.1.2. I b 1 indicated a Decay Heat (DH) Pump recirculation flow of 125 gpm vs.150 gpm. Analyses of Low Pressure injection (LPI) or DH System capability have all used an assumed recirculation flow of 150 gpm. Table 9.5-2 contained erroncous DH pump and cooler component data. The errors on this table can be categori/cd as conservative (cooler shcIl side dcagn temperature inaccurately shown as 200 vs. 250 F) or oln ious errors (two different values were indNated for DH pump design tv .perature; the table indicated timt nuclear services closed cooling cooled the DH coolers). Section 8.2.2.10 g indicated that 72(X)V bus were run in 4160V bus ducts. This error was as an otnious documentation error. Tables 8.2.8 & 8.2.9 crroneously referred to BWST heet trace as BS-T-2B heat trace. Tic tab!c was used to tabulate Emergency Diesel Generator (EDG) loads. The values used were correct. Tic error or'.y affected tic name associated with the load.

In September 19%, a new emphasis was placed on the quality and completeness of the FSAR. FSAR section owners were established as a step to improve the quality.  !

l C. Corrective Action:

l. As stated in the inspection report, PFU 98-TI-126 and PFU 98-TI-129 were prepared to resolve these discrepancies. PFU 98-TI-139 and PFU 98-TI-127 have also been prepared to resche these FSAR discrepancies. It is noteworthy that after a PFU has been submitted, the change is placed in a dambase )

for immediate use by resicuers in perfonning safety evaluations.  !

D. Schedule of Corrective Action:

1. The abost Ibted PPJs will be h thMed in FSAR Update 14, which is currently due in April 1998.

i l

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

I 1

..F I 6710-97-2242

! Attachment 1 -

Page 41 of 43 IFl %20127 " Definition of Single Active Failure (Section El.3.7)"

l A. Description of the Finding: l j

. Section 4.1.3.7, " Single Failure," of System Design Basis Document (SDBD) TI-212 states that the l definition of acthe component has been interpreted to exclude self-actuating components for which ticre is )

adcquate positive force to assure they function (i.e., check valves). This definition and its application in safety-related systems had not been resohtd by the end of the inspection.  !

B. Discussion:

For the design and licensing of TMI 1, Energency Core Cooling System (ECCS) check valves were treated as passive components. The ECCS Singic Failure Analysis in FSAR Section 6.7.3 states "the ECCS mcchanical design adequacy is dependent on the following assumptions: (1) flow check valves assume the ,

proper open or closed position wlen required;(2) pressure relief vahts assume tic proper open or closed position wlen required. Components u hich are assumed to operate and arc '.:xciated from " single failure consideration are specifically required to be operabic at all times without exception in tic Technical Specifications.

Tic Design Basis Document (DBD) discussion is specific to " single failure" application for ECCS analysis. .,

Check valves are tested as required by tic TMI-I Technical Specifications and ASME Boiler & Pressure l Vessel Code,Section XI.

C. Corrective Actions:

)

No corrective action is required.

l e

i l

l

.o.

6710-97-2242 Attachment i Page 42 of 43 URI W201 28 " Control of C;dculations (Section El.4)"

A. Description of the Finding:

Design Control - Control of Calculations - The inspection report cited two design control concerns related to control of calculations. The inspection team identified that memoranda, Technical Data Reports (ERs) and plant Engineering Evaluation Rcquests (EERs) were used to perform safety related calculations, which did not comply with GPU Nuclear proa: dure EP-006 and could result in the requirements for verifying and checking of design work not being met. The second concern was that several calculations and analyses resiewed by the inspection team were not the latest documents because subsequent calculations performed tic same or very similar analyses. The older analyses were not identified as superseded. Additionally, clunges to data from other sources that were used as input to calculations were not consistently incorporated into the completed calculations.

B. Discussion:

GPU Nuclear concurs with this fmding and has initiated a series of actions to address the concerns regarding the costroi af calculations as presented in the inspection report.

Following NRC identification of the issue, GPU Nuclear promptly issued a Quality Deficiency Repcrt (QDR).

GPU Nuclear inunediately provided supplemental training to the engineering staff that reinforced procedural requirements for the performance of calctdations. Additional actions were taken to detennine the root ciuscs of this condition and to evaluate and enhance the c6 sting process. GPU Nuclear has initiated an extensive rniew to identify design basis cdculations for key system parameters. GPU Nuclear has also undertaken an efTort to rnicw mcmos, EERs and TDRs to detennine the extent to w hich design basis calculations may have been documented outside of the established procedural requirements.

The initial root cause evaluation has been completed. The results indicate that both process and human perfonmmcc issues contributed to this concern. Additional contributing factors have also been identified.

The process enlumccments currently underway address these items. The impact of management factors on this process is also being evaluated. The TDRs have been screened and final rniews are in process. Sampic rniews of memoranda and EERs are also being performed at this time.

Pregious observations have, for the most part, indicated that GPU Nuclear calculations arc detailed and technically sound. Initial resiews are demonstrating that there is not a more widespread problem with the caletdation procedure than that u hich has been identified. The implementation of process enhancements, datatuse improvements and reinforcenwnt of procedural compliance that are planned will further strengthen j the GPU Nuclear calculation process. I l

C. Corrective Actions: )

1. Supplemental training on the current procedure has been provided for the engineering stafT.

2. GPU Nucicar has completed the initial identification of the root causes and other factors that contributed to calculation process issues.

3. An evaluation is being prepared which higidights the impact of management factors that may have contributed to the deficiencies noted in the inspection report uhich wcre related to the calculation process.

4. A new cdculation process is being implemented. This will include procedural improvements along with training on the changes.

5. The extent of non-compliance with the calculation procedure is currently being assessed aral the conditions of non-compliance that are identified will be corrected.

f

^ *

,me.

6710-97-2242

~ Attachment 1 Page 43 of 43 D. Sclxxiule for Completion of Corrcx:tive Action:

l l 1. An evaluation of management's impact on tic calculation process deficiencies is scleduled for l completion by August 1997.

l

2. Proccxiural changes, including training on tic changes, is scleduled for completion by September 1997.

3. Tic Assessment of non-compliance with tic calculation proccxture is scheduled for completion by June 30,1997. These will be coluated individually and brought into compliance in a time franc commensurate with their impact on safe plant operation.

4 i

i l

l i

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