ML20151U607

Forwards Copy of Submittal Made to NRC Staff on 980216 Re Integrated Safety Analysis of Recirculation Spray Sys,Iaw Discussion During Prehearing Conference Held on 980902

ML20151U607
Person / Time
Site:	Millstone
Issue date:	09/02/1998
From:	Repka D NORTHEAST NUCLEAR ENERGY CO., WINSTON & STRAWN
To:	Cole R, Kelber C, Moore T Atomic Safety and Licensing Board Panel
References
CON-#398-19503 B17050, LA, NUDOCS 9809110070
Download: ML20151U607 (72)

Text

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

t

(

g WINSTON & STRAdfhPfC 35 WEST WACKER OHIVE

- CHICAco, ILLINOIS 60601-9703 1400 L STREET, N W.

WASHtNGToN, o.o. 20005 3502 N 0 E : % "."a' A '"$ a' 200 PARK AVENUE NEW YORK, NY 10166 4193 (202) 371-5700 SULAYMAN!YAH CENTER OFF G OF P' ' ""'M""' **" ' ^"^" '

r4CsutE a02) 37i.mo RULE t/ t s 3 O DAVID A. REPKA "" **

<2023 37i.s728 ADJUDL iE ToCv'/EWITzERuNo September 2,1998 l

Thomas S. Moore Chairman Dr. Charles N. Kelber Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Atomic Safety and Licensing Board Washington, DC 20555-0001 U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 Dr. Richard F. Cole Administrative Judge Atomic Safety and Licensing Board U.S. Nuclear Regulatory Commission Washington, DC 20555-0001 ,

Subj: Northeast Nuclear Energy Company (Millstone Station, Unit 3),

Docket 50-4234 A

Dear Administrative Judges:

In accordance with the discussion during the prehearing conference call held toda enclosed is a copy of the submittal made to the NRC Staff on February 16,1998 (B17050). Th submittal is an integrated safety analysis of the Recirculation Spray System. {

Sincerely,

k. k David A. Repka N Counsel for Nonheast Nuclear Energy Company Enclosure i

cc.

Service List (w/ attachment) 9809110070 980902 <

PDR ADOCK 05000423 o eon

)go)

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

• . k ..

Northeast ""i % HCH""'c W1. h rr-r. term,w, Nuclear Energy y , n,,,,,, %,. i..,, ,,,,, ,,, s ,, .,, ,,,

Nnhcasi Leicar I:ncrgy C.unginny l'.O. Ibx 128 Waterfi.rd. CT iWl3?i.t}1211 (800) 447 1791 Fax (860) 444 42" "Ihe Nrtheast CoLcies System February,16,1998 ,

Docket No. 50-423 817050 U.S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555 Millstone Nuclear Power Station Unit No. 3 Response to Notice Reauest for Information On The Recirculation Sorav System By letter dated February 3,1998, the NRC Special Projects . Office (SPO) l transmitted a request to provide information relating to the Design and Licensing Bases of the Recirculation Spray System (RSS). This letter and its enclosures

provide the information requested.

An integrated Safety Analysis has been performed on the Recirculation Spray System using the Millstone Unit 3 Safety Evaluation Report (NUREG-1031),

I issued by the Nuclear Regulatory Commission, as the basis. Modifications to the system that have been implemented since the issuance of the Safety Evaluation Report have been evaluated individually and on an integrated basis as they relate to the current RSS configuration. The Integrated Safety Analysis identified an Unreviewed Safety Question (USQ) associated with the modification made in 1986 which eliminated direct injection to the RCS. Our assessment has also concluded that the configuration of the Recirculation Spray System, past and present, has historically been operable despite the 1986 change.

In accordance with the guidance contained within Generic Letter 91-18, Revision 1, a license amendment request to support unit operations with this USQ will be 4

provided prior to entry into Mode 4. The amendment request will be in the form

{ of a FSAR change for the Recirculation Spray System. No Technical

- Specification changes are required to support this modification.

a.

(.

l l

, U.S. Nuclear Regulatd, Commission B17050\Page 2 C

l l

A review of the preliminary findings from the Independent Corrective Action Verification Program (ICAVP) Tier 1 inspection conducted by the NRC and a review of the potential Discrepancy Reports provided by the ICAVP Contractor found no discrepant conditions, not yet addressed, that would call into question operability or functionality of the Recirculation Spray System. These items will be reviewed in accordance with the criteria provided in the response to question 3 of the NRC's April 16,1997 letter pursuant to 10 CFR 50.54(f).

Enclosure 1 provides the responses to Questions 1 through 5 of your request. A "roadmap" is provided to allow easy alignment of the questions and answers. A portion of question 4, concerning training on modifications made to the Recirculation Spray System, was discussed in a public meeting on January 29, 1998 at NRC Region 1 Headquarters.

Enclosure 2 provides the response to Question 6 which requests information on our program for compliance with Technical Specification 6.8.4. NNECO will be in compliance with this specification prior to entry into Mode 4.

The outstanding issues with respect to the Recirculation Spray System will be included as part of the overall, comprehensive, mode change assessment process. Each item will be reviewed to assure compliance with the Design and Licensing bases as it applies to the mode change being assessed.

Should you have any questions regarding the information contained herein, please contact Mr. David A. Smith at (860) 437-5840.

Very truly yours, NORTHEAST NUCLEAR ENERGY COMPANY Martin L. Bowling, Jr.t/

Millstone Unit No. 2 - Recovery Officer cc: see page 3 '

i, i

U.-S. Nuclear Regulatoi, Commission IA I l

B17050\Page 3 cc: H. J. Miller, Region i Administrator S. Dembek, NRC Project Manager, Millstone Unit No.1 D. G. Mcdonald, Jr., NRC Senior Project Manager, Millstone Unit No. 2 J. W. Andersen, NRC Project Manager, Millstone Unit No. 3 T. A. Eastick, Senior Resident inspector, Millstone Unit No.1 1 D. P. Beaulieu, Senior Resident. inspector, Millstone Unit No. 2 A. C. Cerne, Senior Resident inspector, Millstone Unit No. 3 )

W. D. Travers, PhD, Director, Special Projects Office P. F. McKee, Deputy Director of Licensing, Special Projects Office W. D. Lanning, Deputy Director of Inspections, Special Projects Office E. V. Imbro, Deputy Director, ICAVP Oversight, Special Projects Office

S. A. Reynolds, Branch Chief, ICAVP Oversight, Special Projects Office J. P. Durr, Chief, Division of Reactor Projects, SPO D. Schopfer, Sargen' and Lundy ICAVP Project Manager 1

3 l

l i

a

..'. i,

. . f- l l

l Docket No. 50-423 B17050 l

Attachment 1 Millstone Nuclear Power Station, Unit No. 3 Commitments I February 16,1998

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

t '

0 Attachment 1 1 Regulatory Commitment Enclosure List of Regulatory Commitments The follow lng table identifies those actions committed to by NNECO in this docum Any other actions discussed in the submittal represent intended or planne6 ac NNECO. The Director - Regulatory Affairs or Manager - Regulatory Compliance Unit 3 should be notified of any questions regarding this document or any associated regulatory commitments.

REGULATORY COMMITMENT COMMITTED DATE OR  !

OUTAGE B17050-01: RWST Back-leakage Verification Prior to the next refueling outage Test Procedures will be included in the Overall Leakage Reduction Program prior to the next '

refueling outage.

B17050-02: RWST Backleakage tests will be Mode 4 from current outage completed prior to entry into Mode 4.

B17050-03: Operator Crews will be tested to i Prior to the next refueling outage show they can accomplish transfer from injection mode to recirculation mode in less than 25

minutes.

i i

B17050-04: A Debris Transport Methodology will December 1,1998 be dev, eloped and submitted to the NRC for review by December 1,1998.

B17050-05: A FSAR Change containing a USQ Mode 4 from current outage I

l on RSS will be submitted to the NRC prior to 1 entry into Mode 4.

l l'

l l

i i

...e-.

.. i.

p p..

ENCLOSURE 1 '

MP3 - RSS SYSTEM;.

RESPONSE TO NRC INFORMATION REQUEST DATED FEBRUARY 3,1998

(

i OVERVIEW HISTORY DESIGN AND OPERATIONALISSUES PLANT LICENSING AND DESIGN BASIS REVIEW INTEGRATED SAFETY ASSESSMENT RELATIVE TO THE ORIGINAL NRC APPROVED DESIGN (CIRCA 1985) i -

f ^

11P3. ItSS SYSTEi! t' r+6 RI.SPONSt. 'l O NitC !\l't mit i t'lON Ri Qt 'I.Sl Dawd i A L 199X Table of Contents

1. INTRODUCTION AND HISTORICAL OVERVIEW.................................

1.1 In t rod u c tio n a n d s u m m a ry ..... . . . .. ..... ... . .............. .... ... ... ...... .... ...... ... ....... ... 1

1. 2 O ri gi n a l Des i gn.. .... .

. . ... .. . . ........ ................. . .. ... ... .. .. . .... .... .. .... I 1.3 Cu rrent (1998) Design ... ....... . . .. .. ... ...... .........-... .......................3 1.4 Ilisto rical Ch a n ges ....... . . . .. ... .

... .. .. . .. ..... .. ... ....... . .. ..... . .. 4 1.5 Su m m a ry of 10 CFR 50.59 Review. ... ..... ... ... ... ................ .................. . .. . ... . . .. 5 1.6 Su m m a ry of Res po n se t o N RC Question ....... . .... . . ...... .. ............ .... .. ...... ....... 6

2. COMPARISON OF RSS ORIGINAL DESIGN TO THE CURRENT DESIGN.. 8 2.1 O rigin a l Design (Ci re a 19 8 5) . .. ... ....... . ...... .... . ... ... ....... ... .......... ......... . ...... . . .. .. 8 2.1.1 Injection Phase..

2.1.2 Cold Leg Recirculation Phase.. .8

.8 2.1.3 Hot Leg Recirculation Phase (Two Path-Hot & Cold Leg).. .9 2.1.4 Basic Performance Parameters (Flow Rates, Timing, Operator Actions).. .9 2.2 Description of 1998 Dcsign...... . . .. . .......

2.2.1 Injection Phase..

............... ...........................................10

.10 2.2.2 Cold Leg Recirculation Phase..

. 10 2.2.3 Hot Leg Recirculation Phase (Two Path Hot & Cold Leg) . . 10 2.2.4 Use ci Direct Injection..

.10 7

2.2.5 Basic Performance Parameters (Flow Rates, Timing, Operator Actions).. . .I1

3. DESCRIPTION OF CHANGES WHICH AFFECTED RSS SYSTEM OPERATION..................................................................................................12 l 3.1 1986 Elimina tion of Direct Injection Design Chan ge .................. .... .. ...... . .. -.... .......... 12 3.2 1991 Containment Operating Prcssare Design Change....... .. . .. . . ..... . ... . . . ... .. 12 3.3 Current RSS Modifications.... .. .... .... ........................... ........................13 3.4 Use of Direct Injection for Limited Passive Failures and Multiple Failures.... .. . .... ...... ... . 14 3.5 Operator Action Time Change,10 to 25 minutes..... . . . . ...................14

r - -

t c

g%

I

• 1 page u_

hir3. RSS SYSTlaf kl.NI'ONSI'.10 NRC INI OR \lXilON Riot T.NT i

Dated lih 3.1998

4. ACCIDENT ANALYSIS - ASSESSMENT OF CHANGES SINCE NRC REVIEW AND APPROVAL (ClRCA 1 9 8 5 ) . .. . . . . . . . . . . . . . .

i

! 4.1 Ananysis Computer Code / Design inputs / Analytical Models/ Assumptions.. .........15 4.1.1 Computer Codes / Analytical Models Used in the Safety Analysis . . , . . . 15 4.1.2 Design inputs / Assumptions used in the Safety Analysis.. .. .

.. . . . 16 4.2 Impaet cf the Reduced RSS Flow on Accident Analysis...... ....... ...-. .. ..~.~.....-~~~.. 17 4.2.1 Large Break LOCA Assessment.. .

4.2.2 Long. term cooling..

.. .17 4.2.3 Baron Precipitation Control

. . . .17 l

. . 18

! 4.2.4 Small Break LOCA Assessment.. l

.18 4.2.5 Peak Pressure and Temperature.. ..

)

4.2.6 Containment Depressurization.. . .

.18

.19 l 4.2.7 Pump NPSH: Minimum Sump Level and Maximum Temperature. . .19 4.2.8 Containment Liner Temperature..

4.2.9 Equipment Environmental Qualification .

.. .20 4.2.10 Radiological Evaluation . .20 i

.20 4.2.11 Combustible Gas Generation.. . .

. .21 4.3 O t h e r Acc i dent s.....

4.3.1 Steam Line Breaks ...... . ... ......... ... . ......... . .......... . .. .. .................. . .. ... . 21

.21 i

4.3.2 Feedwater Line Breaks.. . .. .

i

.21 4.4 Assessment of the Effect on the Licensing Bases

. .......... ..... -. . ............. ....... ..... ...... 22 4.5 Summary ~~~.-~~......~.....;...........,,,,,,,,,,,,,,,,,,,,,,,,,_,,__________

5. SYS T E M M O DIFIC ATIO N S .......... . ... . .... ... ... .... .. . . . ..... .. . . .

6. PROBABILISTIC RISK ASSESSMENT - ASSESSMENT OF RSS SYSTE CHANGES SINCE NRC REVIEW AND APPROVAL (CIRCA 25 198 5)...

6.1 The Impact of Removing Direct Inject from Design Basis on Individual Plant Examination (IPE)........................................................................................25 6.2 Use of IPE insights during review of RSS design modifications made during mid-cycle 6 shutdown.....-.........................................................................26

7. O T H E R I S S U E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

7.1Tralning.....................-...........................................,.............................29 7.2 System Testin g Co nside ration s .... . .... ....... ..... .... ..... ..~......-... ........ .. . ..... . ... .. . 30 l 7.2.1 Test Program.. .

t . . . .30 7.2.2 Assessment of RSS Modification Effect On Previous Testing . .. .32 7.2.3 Post Modification Testing for RSS Pump Restriction Orifices..

.33 7.2.4 Post Modification Testing for RSS Miniflow Valves (3RSS*MOV38A,B). .35 7.3 Operational issues Resulting from the RSS Modifications...

7.3.1 Testing Results.

.. . . . . . . . . . .35 l ii

P 1-

' * \ll'1. RSS SYSTia ,

page is amuwr 10 NHC NUR.\l. MON RT:Qnst Daled Feb 3,195.'It

8. INTEG RATE D S AFETY AS S ES SM EN T....................................................... 37 8.1 10CFR50.59 Review Of RSS Modific.ations. . ...... . . ... .. . -... .... .... 37 8.2 Elimination of RSS Dircet injection .. ..... .. ...... .... .. .... ... . . ". .... . .... 38 8.3 Containment Pressure Change From Subatmospheric to Atmospheric. . .... ........... ... ... 39 8.4 Reduction in RSS Flow Due to Installation of Orifices.... . . ......... ... ..... .39 .

8.5 Credit tor RSS Direct Injection for Failure Mitigatton ............. . . . ..... .... ... ...... 39 8.6 Increase in Assumed Operator Action Time

. ... . ..... .. ..... . .. . .............. 40 8.7 Changes in Analysis Methads and In pu(s ......

... .. . .... . .... ..... ..... .. .... . 40

9. C O N C L U S i O N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 0. R E F E R E N C E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10.1 Design Ch an ge Docu me n t s .... ............... . ... ... ..... ........... ...

..................................42 10.2 S a fet y E va tu a t io n s . .. .. ....... ........ .. ... .. .......... ............. ...... .......... .. ..................... . ...... 4 2 10.3 Wes t i n gh ou s e Co rres po n de n e c....

. ............. ........... .. ... ....... ................ ................. . ...... 4 3  !

10.4 S ton e & We b st e r Co rres pon dclice .. .. ........ .................. ......... ....... ........ . .. .. .43 4

i 1 1. L I S T O F ATTAC H M E N,T S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. . . . . . . .. . . . . . . .. ...... . .

1 2. LI S T O F FI G U R E S . . .. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . .. . . ... . .. . .

l l

t l

bi l

i

l s ,

4 r t

i l

l 1.

INTRODUCTION AND HISTORICAL OVERVIEW This document prcvides a response to the request for information conceming the Containment Recirculation Soray System (RSS) at Millstone Unit 3, dated February 3, 1998. Specifically, this repc.t addresses the RSS and provides an overview of system history, design and operati;nalissues, plant licensing and design basis review, training, and integrated safety assessment.' relative to the original NRC approved design (NUREG 1031, July 1084).

1.1 Introduction and Summary The Containment RSS of Millstone Unit 3 is part of the Engineered Safety Features (ESF) which are designed to mitigate the consequences of a Loss of Coolant Accident (LOCA).

The RSS system uses four pumps to recirculate water from the containment sump and performs the following functions:

1 1.

Supplements the Quench Spray System (OSS) during the early part of the LOCA to depressurize the containment, and provides for long-term control of containment pressure and temperature after QSS has completed its safety function.

~

2.

Provides the Ultimate Heat Sink for the Emergency Core Cooling function after direct injection of the Refueling Water Storage Tank (RWST) has been terminated.

3.

Provides the long term iodine scavenging function from the containment l

atmosphere after QSS is terminated following a large break LOCA.

This system has undergone several modifications over the life of the plant. The purpose of this report is to describe these modifications and how they have affected system functionality and compliance with regulatory requirements. This introduction and Historical Overview section is intended to highlight the most significant issues and to

, provide an overview of the report.

i 1.2 OriginalDesign in the original design, the RSS assumed the following configurations during the

! three phases of a LOCA.

Injection Phase All 4 RSS pumps auto start approximately 11 minutes after a Gontainment Depressurization Actuation (CDA) signal and augment QSS to spray. and depressurize the containment.

1 l

1

, , - ,- - ..n-., - - - - ,

, . _ _ .. . - - - - ---- - ^^--- ~ ' ~ ~ ~

^~~~~~~T 4

• • *t *

(Q, e

.r

• e q,

i Cold Leg Recirculation Phase When the Low-Low levelin the Refueling Water Storage Tank is reached i

(-33 min.), the following manual alignment was performed as directed by the l EOPs:

1 1

2 RSS pumps remain aligned to the containment spray header The other 2 RSS pumps were isolated from the spray header and realigned to provide direct injection supply to the suction of the injection pumps as follows:

2 RSS pumps - > Supply:

- 2 Charging pumps ---- > 4 cold legs

- 2 Intermediate Pressure (SlH) pumps - > 4 cold legs

- 2 Cold Leg Direct injection Paths > 4 cold legs Hot Leg Recirculation Phase 3 .. .

Approximately 9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> into a LOCA, injection is manually realigned to the hot legs to prevent boron precipitation as follows:

2 RSS pumps remain aligned to Containment sprays 2 RSS pumps

> Supply:

- 2 Charging pumps > 4 cold legs

- 2 Intermediate Pressure (SlH) pumps - > 4 hot legs

- 2 Cold Leg Direct Injection Paths -

-> ISOLATED

- 1 Hot Leg Direct injection Path > 2 hot legs i

S e

2

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

,; , a,

, I l

1.3 Current (1998) Design In the 1998 current design, the RSS assumes the following configurations during the three phases of a LOCA l l

injection Phase j t l All 4 RSS pumps auto start' approximately 11 minutes after a containment Depressurization Actuation (CDA) signal and augment OSS to spray

! containment.

Cold Leg Recirculation Phase  !

l t

When the Low-Low levelin the refueling water storage tank (RWST) is reached l

(~33 min) the following manual alignment is performed in accordance with the EOPs: l L

2 RSS pumps remain aligned to the containment pray header.

1

' The other 2 RSS pumps remain aligned to the containment spray header and are also aligned to the suction of the injection pumps as follows:

2 RSS pumps --> Supply to:

- 2 Charging pumps > 4 cold legs

- 2 Intermediate pressure (SlH) pumps > 4 cold legs

- 2 Cold Leg Direct Injection paths > !solated Hot Leg Recirculation Phase i Approximately 9 hours1.041667e-4 days <br />0.0025 hours <br />1.488095e-5 weeks <br />3.4245e-6 months <br /> into the LOCA, injection is realigned to the hot legs to prevent boron precipitation as follows:

2 RSS pumps remain aligned to the containment spray header only l

2 RSS pumps -- --> Supply to:

- 2 Charging pumps > 4 cold legs

- 2 Intermediate pressure (SlH) pumps > 4 hot legs

- 2 Cold Leg Direct injection Paths - -

> lsolated

- 1 Hot Leg Direct injection Path > lsolated i

f.

3 l

l - .. . . ,

1

(' lm i 1 l 1.4 Historical Changes 1986 Elimination of Direct injection The issue driving this functional modification was observed vibration of the RSS heat exchangers due to excessive flow. This phenomenon was discussed during pre-ops testing in 1985. Westinghouse analysis demonstrated that the flow provided by the two safety injection pumps and two charging pumps in the recirculation phases was in excess of the flow required for core cooling and therefore, direct injection from the RSS pumps into the cold (and hot) legs was not required. Evaluation of injection flow after RSS direct injection elimination demonstrated accident acceptance criteria were being met. RSS direct injection was eliminated by revising Emergency Operating Procedures to require the operators to close the direct injection flowpaths. However, provisions in the Emergency Operating Procedures were retained to open the valves for direct cold leg injection as a contingency action if required.

1991 Change of containment from subatmosph[eric to atmospheric in order to allow for more expedient containment access during power operation, a change was processed and implemented in 1991 to change the containment design from subatmospheric to near atmospheric. This change did not alter system function as described in Section 1.3 above. However, the design basis of RSS was significantly changed.' The initial design basis, applicable to subatmospheric. containment, required that OSS and RSS bring the containment pressure to subatmospheric conditions within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after a LOCA. The revised design basis, applicable to an atmospheric containment, requires that OSS and RSS reduce containment pressure below 50% of peak accident pressure within 24 hrs, after a LOCA. This change was reviewed and approved by the NRC in Amendment No. 59 (TAC No. 76066), dated January 25,1991.

1996 - 1998 RSS Changes During the current plant shutdown, a number of conditions were identified which affect the RSS system and require corrective actions. The changes implemented to correct these conditions are described in Section 3 and 5 of this report and fall into two distinct categories:

Actions to restore systems and components to the level of reliability expected in the original design. These actions addressed various issues that were present in the original and subsequent designs, These issues include, piping qualifications, expansion joint qualification, sump restoration, ECCS loops seal issues and RSS pump seal and testing issues.

Changes which affect RSS system operation or per'ormance. There are three changes of this type. The first change insta!L, RSS pump restricting orifices. The second change once again recognizes the use of direct cold leg injection.

The third change increases the documented time to complete switchover to cold leg recirculation.

4

r n l

. . L l 1

1.5 Summary of 10 CFR 50.59 Renew This integrated safety assessment reviews the modifications, changes and analysis used to support the RSS design from 1985 until 1998. These changes  !

were evaluated as to whether they affected the functionality of the design or whether they restored or improved the original assumed reliability. The reviews of the functional changes are described below.

Elimination of direct injection changed an automatic safety function with no operator action to a contingent safety function requiring operater action.

Therefore this change could have increased the probability of malfunction of equipment important to safety and is an Unreviewed Safety Question.

The change in containment operating pressure required a Technical Specification change and received NRC review and approval prior to its implementation. This change resulted in a relaxation of the containment depressurization acceptance criteria.

RSS pump restriction orifices were addedlo the system to eliminate the possibility of suction line flashing and potential water hammer. Although, this change significantly affected the system performance, the evaluation of the impact of this change on malfunctions, accidents, and margin of safety i determined that there was not an Unreviewed Safety Question.

The FSAR and supporting documentation are being changed to clearly reflect the use of RSS direct injection as a means for mitigating certain assumed limited passive failures in the design bases. This change does not result in an Unreviewed Safety Question. '

i The increase in required time to complete switchover to cold leg recirculation documents a change to the minimum time available for the operators to complete switchover from the Injection Phase to the Cold Leg i

Recirculation Phase. 'There is no change to the operator actions or to the probability that the operators or the equipment will fail. This change does not result in an Unreviewed Safety Question.

9 S

5 l

V # 9 ,

e-e --

t ,m y

t 1

1.6 Summary of Response to NRC Question l

Question 1 Based on the 1986 operational change to the (RSS) and all the subsequent modifications made during the current outage, please provide an integrated safety assessment which compares the current system design to the original NRC approved RSS design-l This assessment should address significant changes to analytical models and inputs used to calculate thermal-hydraulic performance, containment pressure / temperature, and radiological dose consequences of postulated accidents.

Response: This document provides the requested integrated safety assessment of the RSS design and modifications to the current configuration. j Question 2 l Based on the limited staff review of the 1986 operational change to the RSS system, it appears that eliminating cold and hot leg direct injection resulted in a reduction to the margin of safety. Please address this concern.

Response: The review of the 1986 change which eliminated the RSS direct injection path concludes that this change should have been considered an Unreviewed Safety Question. Refer to Section 8.

Question 3 Provide a description of each major RSS system modification and your determination of whether an Unreviewed Safety Question exists.

Response: Sections 3 and 5 of this report describe each of the significant modifications to the Rj5S system. Attachment 2 contains a listing and summary of all design changes since the operating license was issued. Other than the elimination of direct injection in 1986, none of the other changes has resulted in an Unreviewed Safety Question.

Question 4 Discuss the training received by operators on the modifications made to the RSS system since plant was shutdown, as well as any new insights revealed by the IPE for the current system configuration / operation.

Response: The training which has been received by the operators is discussed in Section 7 of this report. In addition the information on training was previously supplied during a public meeting on January 29,1998 at NRC Region 1 HQ. Section 6 describes how IPE insights were used in evaluating and improving the redesign of the RSS. .

6

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

., .. . .n

( e e, !

Question 5 Provide an evaluation of new operational issues resulting from modifications to the RSS such as the vibration resulting from the installation of the flow restricting orifices.

Response: Section 7 of this report provides a discussion of the vibration issues and testing which is being performed to assure the reliability _of,the RSS.

No other operational issues were identified. ,

Question 6 With regard to the implementation of Technical Specification 6.8.4, describe the methods by which leakage from the RSS and associated systems outside contaiament will be controlled and monitored to ensure that the radiological dose consequences of postulated accident are within the plant's licensing bases.

Response: This question is addressed in Endlosure 2.

I 1

1

!. i l

i. I i-l i

+

y L

7 i

i

1 I'

2.

COMPARISON OF RSS ORIGINAL DESIGN TO THE CURRENT DESIGN 2.1 OriginalDesign (Circa 1985)

The original RSS design supported both a subatmospheric containment and an ECCS function. The RSS system performs a containment heat removal function during the initialinjection phase of an accident when ECCS pumps draw water from the RWST. The RSS system performs both a containment heat removal function and an ECCS function during the recirculation phase of an accident when the RWST is depleted and the containment sump is the only source of water.

2.1.1 Injection Phase The injection phase of operation is actuated automatically on high high containment pressure. After approximately an 11 minute time delay to allow the- containment sump to fill, all four RSS pumps begin to recirculate water from the sump and discharge through their respective coolers to the spray headers. All RSS flow is initially directed to the RSS spray headers in the containment dome.

In addition to containment spray, several pathways are available to provide core cooling and makeup through the Emergency Core Cooling System (ECCS) as depicted in Figure 1A. Upon receipt of a safety injection signal (SIS), the following pumps start and align to provide core cooling and makeup from the RWST: one centrifugal charging pump starts, one pump is already in operation, and both pumps inject,into the cold legs of all four Reactor Coolant System (RCS) loops. Two intermediate head safety injection (SlH) pumps start. Once the RCS pressure is below the shutoff head of the SlH pumps, they begin to take borated water from the RWST and deliver it to the cold legs in all four RCS loops. Two residual heat removal (SIL) pumps start. Once RCS pressure is below the SIL pump shutoff head, they begin to inject into the four~ RCS cold legs. When the RCS pressure drops below the pressure of the four safety injection accumulator tanks, they discharge their contents into the four RCS ccid legs.

The duration of the injection phase depends upon the nature and severity of the accident. When the RWST water level drops to a predetermined point, the injection phase is discontinued and the cold leg recirculation phase is initiated.

2.1.2 Cold Leg Recirculation Phase When the low-low level in the RWST is reached (approximately 33 minutes for LB LOCA maximum ESF in the original design) two~ SIL (RHR) pumps are tnpped two RSS spray header isolation valves are 6

o

. '. i

• Y l e manually closed and the flow from the two associated RSS pumps is directed to the two cold leg direct injection flowpaths, the two operating charging pumps, and the two operating SlH pumps. The remaining two RSS pumps continue to operate as before, with their discharge directed to the spray headers (see Figure 18). After the switchover to the recirculation phase of operation is accomplished, the system is in its basic configuration for long term operation.

2.1.3 Hot Leg Recirculation Phase (Two Path Hot & Cold Ceg')

in order to prevent boron precipitation following a LOCA, one additional realignment is performed at nine hours after the initiation of the accident.

The SlH pumps are realigned to provide flow to all four RCS hot legs.

The RSS pumps are aligned to one hot leg direct injection flow path, supplying flow to Loops 2 and 4. The charging pumps continue to provide flow to the four cold legs, as before. The RSS cold leg direct injection paths are isolated. This alignment is illustrated in Figure 1C.

2.1.4 Basic Perforrnance Parameters (Flob Rates, Timing, Operator Actions) 1985 RSS Minimurn ESF Flow Performance Data Mode Spray Flow, gpm ECCS Flow, gpm injection 7760 -

Cold Leg Recirculation 3880 395L Hot Leg Recirculation 3880 3950N i Notes: (1) Hot leg recirculation ECCS flow is estimated to be approximately equal to the cold leg recirculation ECCS flow.

1985 RSS System Timino / Operator Actions Summary The RSS pumps start approximately 11 minutes following the receipt of CDA (containment depressurization actuation) signal. During LBLOCA maximum ESF, switchover to cold leg recirculation occurs at approximately 33 minutes following the accident. The process to accomplish switchover to cold leg recirculation utilizes a combination of manual actions, interlocks and automatic actions, and is controlled by l the EOPs. The switchover to cold leg recirculation was required to be accomplished in ten minutes.

d l

9 l

r

.  ?

2.2 Description of 1998 Design 2.2.1 Injection Phase The injection Phase alignment of the ECCS and RSS system is unchanged from the onginal 1985 alignment.

2.2.2 Cold Leg Recirculation Phase -

The Cold Leg Recirculation Phase operation of the system in 1998 is similar to the original 1985 alignment with two substantial changes. First the RSS pumps are not isolated from the containment spray header when they are aligned to supply ECCS cooling. These pumps will provide a dual function: ECCS injection, as well as, containment spray function. Second, the direct injection paths from the RSS pumps to the RCS cold legs are isolated to prevent direct injection to prevent tube vibration.

The direct injection flow path remains available as a contingency action in the event that" charging and SlH become unavailable. This alignment is depicted in Figure 28 and is controlled by the EOPs.

2.2.3 Hot Leg Recirculation Phase (Two Path-Hot & Cold Leg)

The Hot Leg Recirculation Phase operation of the system in the 1998 configuration differs from the original 1985 alignment in that 1) the RSS pumps are not isolate'd from the containment spray header when they are aligned to supply ECCS cooling, and 2) the direct injection path from the RSS pumps to the RCS hot leg is not opened. The 1998 flow paths are shown in Figure 2C.

2.2.4 Use of Direct injection The direct injection paths of the RSS pumps to the cold legs are available to the operators as contingencies for mitigating situations in which at least one SlH and one charging pump are not available (Figure 2C). These situations could arise either as a result of multiple failures or from isolating certain limited passive failures in the ECCS. The use of these contingencies are controlled by the EOPs.

10

~

i (h

u 2.2.5 Basic Performance Parameters (Flow Rates, Timing, Operator

, Actions) 1998 RSS Minimum ESF Flow Performance Data Mode Spray Flow, gpm ECCS Flow, gpm injection 4030 ~

Cold Leg Recirculation 3150W 1100 Hot Leg Recirculation 3150N 1100 Notes:

(1) The total 1998 spray flow is based on one dedicated RSS pump plus the excess flow that is not diverted to the ECCS pump from the second RSS pump.

_1998 RSS System Timino / Operator Actions Summary The RSS pumps start approximately 11 minutes following the receipt of CDA signal.

During LBLOCA maximum ESF, switchover to cold leg recirculation occurs at approximately 33 minutes following the accident. The process to accomplish switchover to cold leg recirculation utilizes a combination cf manual actions, interlocks and automatic actions. The required time for switch.sver to cold leg recirculation is increased from ten minutes to twenty five minutes. The 1998 switchover requires the closure of the two RSS direct injection isolation valves, but it has eliminated the closure of the two spray headerisolation valves. Two RSS pumps are required to provide both RSS spray and ECCS injection. The switchoveris controlled by the EOP.

e 11

.. '~'

3.

DESCRIPTION OF CHANGES WHICH AFFECTED RSS SYSTEM OPERATION 3.1 1986 Elimination of Direct injection Design Change During startup testing (1985) of the RSS system, excessive RSS heat exchanger tube vibration was observed during certain modes of operation. The specific test which resulted in excess flow included operation of one RSS pump feeding the two Charging (CHS) pumps, the two SlH pumps and the hot leg direct injection path. It was determined that excessive RSS heat exchanger tube vibration could occur for heat exchanger flows in excess of 4600 gpm. The corrective action was to eliminate RSS direct injection to reduce RSS heat exchanger flow.

Emergency Operating Procedures EOP 35 ES-1.3, " Transfer to Cold Leg Recirculation" and EOP 35 ES-1.4, " Transfer to Hot Leg Recirculation," were revised to terminate flow from the RSS pumps directly to the RCS immediately after transfer to cold leg recirculation. As a result, the RSS pumps only supplied flow to the suction of the SlH pumps and the CHS pumps. The minimum ESF

( ECCS alignment for cold leg recirculation consisted of one CHS and one SlH pump injecting into the RCS cold legs. During hot leg recirculation, the SlH pumps were realigned to inject into the hot legs.

With these changes, the long term core cooling and containment heat removal design basis requirements were still met. These changes were evaluated in accordance with 10CFR50.59 and submitted to the NRC as FSAR updates in November, 1987.

Northeast Utilities has recognized that this change constituted an Unreviewed Safety Question.

3.2 1991 Containment Operating Pressure Design Change s

in 1991, the plant Technical Specification 3/4.6 was changed to increase containment pressure during normal plant operation so as to p( wit more expedient entry into the containment. With the nearly atmospheric normal operating containment pressure, the post LOCA one hour depressurization requirement to subatmospheric pressure no longer applied and was deleted from Technical Specification 3/4.6. The new design basis for containment pressure is to depressurize containment to less then 50 percent of the peak precsure within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the postulated accident. The new design basis significantly reduces the heat removal requirements for QSS and RSS since these systems are no longer required to remove containment sensible and latent heat in order to reach subatmospheric conditions within one hour. Peak containment pressure is a function of energy release and containment  ;

depressurization is a function of containment heat sinks. It is the internal I containment heat sinks and the 50'F QSS spray that effectively accomplishes containment depressurization in the first three (3) hours of a postulated LOCA.

RSS starts after the maximum containment pressure occurs and, together with ObS. initially functions to prevent the occurrence of E second pressure peak.

12

+. .'.'. .

The RSS is~ relied upon to provide recirculation water for long term ECCS core cooling and to maintain long term containment depressurization.

i The other significant aspect of the change in containment operating pressure was the addition of new safety functions to the QSS and RSS systems. For 4

original plant design, OSS and RSS were not considered to perform a fission

product removal function. As a result of the containment pressure change, fission product _removalis now a required function of both systems. QSS/RSS is

' now required to comply with the provisions of SRP 6.5.2, "ContalRmsnt Spray as a Fission Product Cleanup System", Rev 2, dated December 1988. Per the provisions of SRP 6.5.2, the QSS/RSS is. credited with a decontamination factor (DO of 200 based upon calculations performed using the methodology of ANSl/ANS-56.5. The RSS also performs a long term fission product retention i

function after QSS is secured. The design of both spray systems has been i

! determined to meet the applicable provisions of SRP 6.5.2. The NRC staff has 1

determined that the OSS/RSS is designed in accordance with requirements of l

~ NUREG 0800 SRP Section 6.5.2 and approved the Technical Specification ,

change per Amendment No. 59 (TAC No. 76066), dated January 25,1991.

e.

3.3 Current RSS Modifications In 1997, the following concerns were identified with the RSS design

potential  ;

t for RSS suction line flashing due to excessive frictional losses, potential for '

vortexing at the sump inlets due to inadequate water level in the sump during startup of the system, and potential for water hammer in the pump discharge piping following a stop and restart of an RSS pump. The suction line flashing and water hammer issues were resolved by installing restriction orifices on the j- discharge of each RSS pump. The issue of potential vortexing at the sump inlet 1

was resolved by lowering the vortex grating by 12 inches.

Because flow to the RSS spray headers is reduced by the RSS pump restriction orifices, fifty percent of the RSS spray nozzles were plugged in order to maintain adequate pressure drop across the spray nozzles. This was also required to maintain the required spray droplet size distribution for spray thermal effectipeness. Extensive analysis of the containment water level showed that the water levelin the sump might be too low to prevent vortexing at the sump suction inlets for a small break LOCA inside the reactor cavity. The vortex breakers located in the sump were lowered by one foot to resolve this potential concem, in order to assure that the qualification of electrical equipment inside containment could be supported, the containment spray isolation valves remain open when the RSS pumps are realigned to the ECCS. The excess RSS pump flow capacity above the flow required for the CHS and SIH pumps is used to supplement the containment spray cooling provided by the RSS pumps dedicated to containment spray.

12

f' O

' As a consequence of the reduction in RSS pump flow, the RSS pump mini-flow recirculation valves were changed from normally open to normally closed and i

the valve opening logic was interlocked with the RSS spray header isolation valves. This change provides added assurance that a large portion of the process flow (about 1000 gpm) from being recirculated and lost from the cooling process. Additionally, flow test loops were added around RSS pumps C and D l

to facilitate pump testing, similar to the mini flow test loops that exist for RSS pumps A and B.

l 3.4 Use of Direct injection for Limited Passive Failures and Multiple Failures During the 10 CFR 50.54f review it was realized that certain long term passive failures could result in actions that would reduce ECCS core injection flow below the values assumed in the accident analysis. It was also recognized that the original reason for eliminating RSS - direct injection (excessive RSS heat j exchanger flows) was no longer an issue because of the addition of the RSS orifices.  !

However, direct injection remains in the RSS design basis as a I contingency action in response to postulated long term passive failures.

3.5 Operator Action Time Change,10 to 25 minutes i

The transition between the injection phase of a LOCA and the cold leg i recirculation phase requires manual operator action to realign the suction of the l

charging and intermediate head SI pumps from the RWS'l to the discharge of the RSS pumps (see Figures 18 and 28). Prior to 1998 the FSAR has stated j

that the operators could complete this transfer within 10 minutes after receipt of the low level RWST alarm. This time was in the original plant design to ensure an adequate RWST inventory for Quench Spray operation and to meet the 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> subatmospheric requirement. This requirement was eliminated by the  !

containment design change (see Section 3.2).

As a result of changes in valve stroke times in the command and control communication protocol in the control room, the time for operators to complete the transfer from injection phase to recirculation phase has increased. The FSAR is being modified to state that the switchover will be completed within 25 minutes, Calculations have been performed which demonstrate that sufficient RWST inventory is available to support ECCS pump operation for a minimum of 25 minutes after reaching the Low Low RWST level.

i l

u

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

.. m t P 4.

ACCIDENT ANALYSIS - ASSESSMENT OF CHANGES SINCE NRC REVIEW AND APPROVAL (CIRCA 1985)

This section provides a summary of the RSS performance assumed in the safety analysis of the Millstone Unit 3 containment and long-term core cooling. The RSS system is an Engineered Safeguards System that supports both an ECCS function as l

well as the containment heat removal function. ~

l

' The current modifications to the RSS system affected its performance and required a complete re-analysis of the containment response to the design basis accidents. This summary is organized in the form of a general comparison of the system performance at the time when it was originally put into service to the performance established after the recent extensive modifications. The safety functions that the system needs to fulfill will also be a part of the comparison. (Refer to Figure 3 for a schematic showing the location of recent major RSS design modifications). The safety functions of the P.SS are: containment heat removal after an accident, long-term core cooling, reactor vessel inventory control of the boric acid concentration in the. vessel and long term iodine scrubbing.

4.1 Analysis Computer Code / Design inputs / Analytical Models/ Assumptions l

4.1.1 Computer Codes / Analytical Models Used in the Safety Analysis ECCS LOCA Evaluation Model The original 1985 ECCS analysis used the Westinghouse 1981 Large Break LOCA methodology. In 1990 the methodology was upgraded to include the modeling of the VANTAGE SH fuel assemblies.

Computer codes used for LOCA licensing analysis:

LBLOCA: SATAN-IV global modeling of the RCS and the secondary side WREFLOOD using the data from SATAN-IV, calculates the tirne to bottom of core recovery during refill)

BASH calculates the reflood phase of a Large Break LOCA BART calculates the entrainment rate during reflood LOCBART calculates core average conditions for BASH input COCO containment pressure transient used as a boundary condition in WREFLOOD I

15

.. . E SBLOCA: NOTRUMP calculates the transient depressurization of the RCS, mass and enthalpy of the break ,

flow '

LOCTA-IV core thermal analysis using the NOTRUMP data)

Containment Analysis Model

~

LOCTIC Verr. ion 23, Level 02S (1985)

LOCTIC Ve Tion 23, Level 03" (1998)

• LOCTIC Version 23, Level 02, permits modeling of: (1) two flow paths for the RSS during ECCS injection phase, (2) one flow path for the RSS during sump recirculation phase, (3) two Service Water flow paths to the RSS heat exchangers during ECCS injection phase, and (4) one Service Water flow path during sump recirculation phase. Service Water flow and heat exchanger overall heat transfer coefficient cannot be varied with time.

"LOCTIC Version 23, Level 03 permits modelirig'of: (1) four independent flow paths for the RSS during ECCS injection phase, (2) two independent flow paths for the RSS during sump recirculation phase, (3) four Service Water flow paths to the RSS heat exchangers during ECCS injection phase, and (4) two Service Water flow paths during sump recirculation phase. Service Water flow and heat exchanger overall heat transfer coefficient can be varied with time.

4.1.2 Design inputs /Assurnptions used in the Safety Analysis ECCS Performance Analysis:

The input data in the ECCS analysis has undergone relatively minor changes.

The changes to the input reflect mainly the change in the fuel composition, and the increased operating pressure in the containment These changes had a relatively small impact on the calculated peak cladding temperature. The modifications to the RSS did not impact the ECCS analysis since the RSS provides a long-term ECCS function and has no impact on the calculated maximum temperature in the core after an accident.

Fuel Type: Westinghouse STANDARD Assembly fuel (1985)

Westinghouse VANTAGE SH fuel (1997)

Break Boundary Conditions: Subatmospheric containment (1985)

Near-atmospheric containment (1997)

Containment Analysis:

The containment analysis has undergone a major revision as a result of the RSS modifications. The RSS affects containment response by providing means

, for the energy transfer from the sump to the ultimate heat sink, via the RSS heat exchangers. The RSS is also credited in the offsite dose reduction by reducing r

16

(

. . 9 the pressure in the containment and by scrubbing the fission products from the atmosphere. The changes to the input include:

l .

increase in the normal operating pressure (subatmospheric to near-atmospheric containment),

reduced RSS flow rate (3900 gpm to 2200 gpm, per pump),

e increased recirculation system fill time, reduction in the RSS spray efficiency, -

inclusion of 120 second RSS system timer tolerance, a new decay heat model incorporating the 24-month fuel cycle, no credit for the RSS spray (steam line breaks only),

new model of the debris transport to the sump, heat exchanger tube fouling on the SW side, l

increased containment mass inventory of heat sinks, and l

increased hydrogen generation after a LOCA due to increased zinc and l aluminum sources.

There is no substantive change in the modeling of the containment passive sinks, except for small additions of new heat sink inventory. l l

4.2 Impact of the Reduced RSS Flow on Accident Analysis i 4.2.1 Large Break LOCA Assessment The peak clad temperature is not impacted by the modifications to the RSS since it occurs prior to the RSS becoming active during the accident. By the time the RSS begins to spray the containment, the core temperature will have peaked over due to the cooling from the safety injection. The peak cladding i temperature in the current cycle is 2054' F for the large break LOCA.

4.2.2 Long term cooling For long-term recirculation from the sump, the RSS and ECCS systems provide flow to the vesselin excess of the minimum required to remove the decay heat and replenish the inventory lost to boiloff in the core. The vendor analysis (NSAL Letter NSAL 95-001, Minimum Cold Leg Recirculation Flow) shows that at 30 minutes after shutdown, at the time of the switchover to cold leg recirculation, the vessel make up requirements are 531 gpm to due to the decay heat, and 107 gpm due to extended boiling in the reactor vessel downcomer and the lower plenum, for a total of 638 gpm.

In the original design, one RSS pump delivered 3950 gpm, feeding one charging (CHS) pump, one high head safety injection (SlH) pump, and direct injection to the cold legs. The excess injection spilled out of the break.

In the modified design, one RSS pump delivers 2200 gpm, feeding one CHS pump, one SlH pump and an RSS spray header. The flow split is 1100 gpm to the spray header, and 1100 gpm combined to the CHS and SlH pumps.

Assuming that 275 gpm (25% of the 1100 gpm) is lost from the break and 531 gpm is needed for makeup, the minimum flow required for long term cooling is 17

_ _ _ _ ~. ___ _ ~ . __ . _ _ _

913 gpm. That stillleaves 187 gpm of excess flow which goes to spillage. The modified alignment will provide sufficient ECCS flow.

4.2.3 Boron Precipitation Control For boron precipitation control, the recirculation from the sump is split into the cold legs and into the hot legs. The vendor analysis shows that the. minimum flow required for boric acid flushing is 31 lb/sec. (225 gpm) into the cold legs and 36 lbs/sec. (260 gpm) to the hot legs, for a total of 485 gpm.

In the 1985 configuration, one RSS pump was capable of delivering l approximately 3950 gpm, which supplied direct injection to the hot legs, the.

suction side of the CHS pumps aligned to the cold legs, and the suction side of the SlH pumps aligned to the hot legs. This alignment provided adequate hot leg flow for boron precipitation control, in the modified design, the RSS pump still supplies the CHS and SlH pumps as in the original design 1100 gpm is injected by the SlH (590 gpm) and CHS (510 gpm) pumps. Given a worst case passive failure, one SlH pump is aligned for hot leg injection and one cold leg direct injection pathway is aligned. Hence, boron precipitation control is maintained.

4.2.4 Small Break LOCA Assessment The limiting small break LOCA is a 3-inch diameter rupture of the RCS cold leg.

Peak clad temperature is reached during the injection phase of the accident.

The RSS pumps do not contribute to the ECCS flow in the injection phase and therefore RSS changes do not impact the small break LOCA assessment.

1 4.2.5 Peak Pressure and Temperature The design basis accident for the maximum containment pressure is the double-ended rupture of a hot leg. Assuming minimum encjineered safeguards, the revised peak calculated pressure is 38.40 psig and occurs 18 seconds after the accident initiation. Therefore, the reduced RSS flow has no impact on the peak containment pressure since the RSS will not start operating until about 11 minutes after the accident.

A steam line break results in the highest containment temperature of any postulated accident. The limiting containment temperature is 336' F as a result of the double-ended rupture of a steam line at 75% power. The time of the peak is 14 seconds after the break. Therefore, the reduced RSS flow has no impact on the peak temperature.

18

e 4.2.6 Containment Depressurization Containment depressurization rate after a design basis accident is important for

'the containment leakage considerations, which have an irnpact on the ,

magnitude of radiological releases after the design basis accident. In the

original design, the RSS and QSS systems function to return the containment to subatmospheric pressure within one hour after an accident. This requirement is no longer necessary with the changes to the plant's Technical Specifications which allow the slightly sub-atmospheric pressure in the containment during normal operation.

The new design basis for containment pressure is to depressurize containment to less than 50 percent of the peak pressure within 24 hrs. The new design basis is met with the reduced RSS flow and the containment leakage rates are  !

within specified limits.

. 1 4.2.7 Pump NPSH: Minimum Sump Level and Ma'ximum Temperature The evaluation of the RSS pump operability issues has identified the limiting transient for the pump NPSH as a 4-inch break at a hot leg nozzle inside the l reactor cavity. This is because the break effluent can be trapped inside the cavity which does not drain to the containment sump. This assumption reduces the level of the water collected in the sump when the RSS pumps begin to operate, which could result in vortexing and subsequent air entrainment in the pump suction lines.

It was determined that the existing vortex suppresser was inadequate and that it needed to be lowered 12 inches (see Figure 6). The hew location does not invalidate the results of the tests performed by Alden Research Laboratory which originally qualified the device as capable of preventing vortices under the most adverse conditions expected in the sump. The basic design is unchanged.

In addition, the sump geometric layout with the new vortex suppressor is consistent with the design recommendations of NUREG 0897, Revision 1.

A loss of cooling to the RSS heat exchangers, due to a failure in the Service Water system, presents a special concem for recirculation piping temperature downstream of the RSS heat exchanger. The resultant piping temperature may reach 260' F. This affects the thermal qualification of the piping and components used in mitigating the consequences of the accident. An evaluation of this concern found that certain piping lines, used for recirculation from the sump, would operate outside their design limits during the transient. As a result of the evaluation, allinvolved systems have been requalified for the increased thermalloads.

19

(.

p 4.2.8 Containment Liner Temperature The mechanical loads on the liner affected by RSS modifications are primarily those that result from the thermcl expansion of the liner and containment pressure. The design basis accident is the transient that produces the largest l temperature differential, (AT), between accident initiation at time zero and the l

time when the containment depressurizes to O psig. An increase in the liner differential temperature above the analyzed AT will increase liner loads..

l The design basis accident for the liner temperature is the'O.40 ft2 main steam line break with failure of an emergency diesel generator. The maximum liner temperature was previously calculated to be 250.9 F and the design AT of 200*

F was used for stress calculation. The new steam line break analysis does not take credit for the RSS spray function and maximum liner temperature 'was recalculated to be 255.9a F. The liner stress analysis remains valid since the new liner parameters are enveloped by the design AT of 200 F. The new maximum AT was calculated to be 180.9 F.

e.

4.2.9 Equipment Environmental Qualification The Electrical Equipment Qualification (EEQ) requirements for electrical equipment inside the containment are based on the analysis of the containment response to the design basis accident. Newly generated environmental curves indicate that the containment pressure is within the current EEQ limit, but the new temperature profile exceeded the existing EEQ temperature envelope by as much as 10' F one hour after the accident and up to 5 days. To compensate for the increased temperature, the revised temperature envelope provides for a reduction in the profile in the period between 5 and 30 days after the accident, by ramping the profile, rather then reducing it in stepwise fashion (see Figures SA and 58). The ramping function follows the actual temperature profile more closely. This satisfies the EEQ requirements for electrical equipment inside the containment.

4.2.10 Radiological Evaluation The changes to the RSS System have been evaluated for their effect on the calculated radiological consequences of a LOCA. They do not affect the ,

consequences because the iodine removal coefficients and sprayed volume are a result of OSS. The radiological consequences also depend on the mixing rate

)

between spray and unsprayed regions and the containment pressure and its effect on leak rate assumptions. Each of these parameters has been evaluated to have no adverse consequences on the radiological consequence analysis.

The radiological calculation used conservative assumptions with regard to the effectiveness of containment spray. The radiological calculation used quench l spray iodine removal coefficients, quench spray volume and mixing rate values  !

which are less than recirculation spray parameters. RSS is only credited with i maintaining spray after QSS shuts down. The bounding containment pressure curve used in the analysis is unchanged because it conservatively assumes 45  ;

psig containment design pressure as the pressure for design containment j 20  !

f f

leakage; therefore it is not affected by the reduced flow rate of the RSS spray.

The radiological evaluation for offsite doses uses the Standard Review Plan's assumption of a containment leak rate at 0.5 L. after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

The radiological evaluation for control room doses takes exception to the SRP and assumes a containment leak rate at 0.5 L. after 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. This exception was approved in Amendment 59 (TAC No. 76066), dated January 25,1991.

Based on this assessment, it i.s concluded that previous radiologicafevaluations bound the consequences that are calculated using the revised recirculation spray characteristics.

4.2.11 Combustible Gas Generation The design changes in the RSS system and the addition of corrosive materials to the containment previous to the RSS changes have necessitated a complete revision of the post-DBA analysis for hydrogen generation in the containment.

The sources of hydrogen during an accident are fuel cladding reacting with steam during extended core uncovery, radiolysis, hydrogen released from the primary coolant, and corrosion of zinc and aluminum materials.

The new analysis indicates that the hydrogen generated during the postulated DBA will not exceed the 4% concentration limit. The largest volume of H2 is generated by the double-ended break at the reactor coolant pump suction, assuming a loss of motor control center MCC32-4T as a single failure. With a single recombiner starting 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the accident, the volume of hydrogen generated reaches a maximum concentration of 3.98% 19 days later. With two recombiners operating 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> after the accident, the concentration will not exceed 3%.

4.3 Other Accidents l 4.3.1 Steam Line Breaks l The main steam line break analysis has been revised. The new analysis does not take credit for the RSS spray function in the containment.

l The steam line breaks result in the highest containment temperature of any postulated accident. The limiting containment temperature was found to be 335.9* F at 14 seconds and is the result of the full double-ended rupture of a steam line at 75% reactor power. The peak temperature is within the EEQ temperature envelope for this time period. Therefore, the rt.suced RSS flow has no impact on the EEQ qualifications from the steam line breUKs.

4.3.2 Feedwater Line Breaks The steam line breaks result in the highest containment temperature of any postulated accident and bounds the feed line breaks.

21

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

. . .', a I

\

O l 4.4 Assessment of the Effect on the Licensing Bases The effect of the modifications with respect to the original safety analysis documented in the licensing basis of the plant is minimal. The modifications do not affect the safety functions of the equipment used for the protection of the core against the design basis accident, and do not create potential for a new malfunction. The current licensing analysis remains valid. Although there is some reduction in the ECCS flow, the modified performance remains significantly in excess of the minimum required to provide long-term core cooling and boron precipitation control.

The safety functions of the containment are not altered or challenged by the modifications. The RSS system will function in effectively mitigating the consequences of the accidents. The depressurization of the containment after an accident is affected by the reduced system performance, but the containment is operated near atmospheric conditions which requires far less robust spray function of the RSS to remove the energy after a LOCA. The changes to the Containment depressurization do not affect the radiological consequences since the dose calculation assumptions of pressure and leakage rates remain bounding. ,.

l 4.5 Summary The modifications do not increase the probability of an accident or a malfunction of the equipment important to safety. The RSS and SI equipment have been evaluated and meet standards for design and operation. The changes do not increase the consequences of previously evaluated accidents. No new accidents or malfunctions will l

result from the modifications. There is no impact on the margin of safety. The containment design pressure, temperature, and liner temperatu , limits are not l

exceeded. The core cooling function of the RSS post LOCA is maintained. The

changes are safe and do not result in an Unreviewed Safety Question, j

I l

I l

l l

i

'F 22

}

t C- \

l.

The following table provides the summary of the key core and containment parameters that are, or are not, affected by the 1998 configuration.

l t --

AFFECTED BY THE RSS MODIFICATION l YES NO CORE COOLING l Peak Clad Temperature: No SB ~ '

l Peak Clad Temperature: No LB l Long-term Core Cooling No Boron Precipitation No Control CONTAINMENT:LOCA Peak Pressure i No Peak Temperature -

No Post-LOCA Yes")

Depressurization Sump Temperature No Fission Product Removal No (')

Offsite Dose No EEQ Yes(3) l CONTAINMENT:MSLB Peak Pressure No Peak Temperature No Liner Temperature Yes")

EEQ No NOTES:

1, One-hour depressurization to subatmospheric after an accident no longer required.

2.

Both the QSS and the RSS are credited for iodine removal from the containment atmosphere.

3.

The current EEQ limits are exceeded between 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 5 days after the DBA by as much as 10' F. A revised profile follows the temperature more closely, l- rather than in a step-wise fashion. The equipment has been requalified to the j,

revised EEQ function (Sect.4.2.9).

i

' 4.

The increased liner temperature has no consequences on the analyzed structural loads (Sect. 4.2.8) 23

e

.P*.

5 'm *

5. System Modifications There have been several modifications to the RSS system ess -

which w deficiencies and potential failures. These changes were em made to reliability without increasing the probability of a malfunction, ,

adding each major RSS system modification and the c complete2. list of modifications made to the RSS system since 1986 is. p Attachment MP3-88-009:

Modification to the Cold Leg Recirculation Array The Cold Leg Recirculation Array on Main Control Board #2 was chan the operator reflected with an arrangement of control switches identical to the in EOP ES-1.3.

With this change, the operator can complete all actions location. This design change was determined to not be n.

MP3-96-054: RSS Component Temperature Rerates Given a single active failure in the Service Water supply to the RSS co The uncooled sump water temperature F.

can approach 260 This temperature exceeded the RSS design temperature in the flowpaths downstream of the coolers and required re-analysis and re-rating of the RSS system piping, and components inside containment. As a result, pipe support modifications required; the RSS cooler service water expansion joints and the containment penetrations be a USQ. were also analyzed and re-rated. These changes were determine MP3 96-056:

RSS Pipe Support Modifications Outside Containment Given the potential for uncdied sump temperatures of 260 F, RSS pip supports outside containment were re-rated and reanalyzed. The revised analysis resulted in no RSS pipe support modifications, .'

modifications were required. These changes were determined to not be MP3 96-063 RSS Support Modifications inside Containment Based on a revised containment temperature profile and postulation of a failure in the SWS supply to the RSS coolers, RSS piping was reanalyzed and pipe support loads were developed for normal / upset ASME Code Pipeconditions.

stress levels were qualified per Table 3.98-11 of the FSAR, Pipe support modifica ,

weld size, location of gussets) were necessary to loads.

and The insert revised plate pipe analysis. support loads were also used in a revised containmen Unreviewed Safety Question. The safety evaluation for this change resulted in no 24

C R MP3-97-042: Flow Test Line for RSS Pumps C & D This design change implements flow test lines for these pumps which allows for te during any mode of operation.

Additionally, this design change improves system availability during Modes 1. 2 and 3 for surveillance testing of these pumps.It also eliminates the need to station a dedicated operator at valve 3RHS*V43 during testin The change was determined to not be a USQ.

MP3-97-045:

RSS Pump Restriction Orifices to Prevent Suction Line Flas'hing This design change accomplishes three (3) objectives:

Installation of restriction orifice plates on the RSS pump discharge lines for 3RSS*P1 A-D to prevent suction line flashing (discussed in Section 3)

Lowering of the RSS sump vortex suppresser grating (discussed in Section 3)

Installation of vent lines on the RSS pump casing to eliminate air binding (see below)

Based upon an evaluation of the effect of trapped air on startup of the RSS pumps after a LOCA, a calculation determined that a 1-inch vent line needed to be installed on the RSS pump casing. The vent line was routed from the vent plug in the pump casing to approximately 5 feet above in the discharge line of the pump. A check valve is installed in the vent line to prevent any loss of flow from the RSS pump discharge once the pump has started. Judicious location of the check valve (i.e., close to the main run header) eliminated any significant water hammer loading in the vent line due to header fluid transient analysis. This change was determined to not be a USQ.

MP3-97-094: FillNent Lines ECCS Loop Seal Branch lines from the RSS pump discharge lines down stream of the RSS Containment Recirculation Coolers are non- self venting because the piping forms an expansion loop.

Air could accumulate in these lines during system filling after a LOCA in Containment. Upon switch over, this air could reach the Safety injection and Charging pumps. The design change installs a vent valve and test connection in each thermal expansion loop. These new connections will allow for compliance with Technical Specification requirements to keep the line full. This change was determined to not be a USQ.

6.

PROBABILISTIC RISK ASSESSMENT- ASSESSMENT OF RSS SYSTEM CHANGES SINCE NRC REVIEW AND APPROVAL (CIRCA 1985) 6.1 The Impact of Rernoving Direct inject from Design Basis on IndividualPlant Examination (IPE)

The direct injection function of RSS was removed from the plant's engineering design basis in 1986. However, EOP ES-1.3 retained the capability to reinstate direct injection if high head injection failed. Given the loss of high head injection, direct injection would not result in excessive RSS heat exchanger flow.

Therefore, the ability to credit the direct inject function within the IPE, for loss of all high head cooling event was retained.

25

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

n i

As part of a PRA model update performed in 1987, a 5% increase in core damage frequency (CDF) was calculated due to changes occurring to the legchanges:

two recirculation function. The overall increase was attributed to the

1. recirculation.

l_onger surveillance test intervals for MOVs used to establish cold leg

2. Modifying the cold leg recirculation alignment to reserve RSS dirict injection The CDF increase associated with these two changes was calculated to be 3E-06/yr. Since the PRA performed prior to the 1987 update preceded the finalized in-Service Test (IST) program, the surveillance intervals used in the model were all assumed. When the IST program was fully implemented, it was determined that some of the assumed test intervals in the original PRA model were not valid, resulting in an under prediction of the CDF.

This was the, . case for a number of valves needed to establish cold leg recirculation. The fraction of the 3E-06/yr. increase due to this under prediction is difficult to determine; however, the CDF increase due to the col & leg recirculation alignment modification is estimated to be 1E-06/yr.

6.2 Use ofIPE insights during review of RSS design modifications made during mid-cycle 6 shutdown The modifications implemented during the current mid-cycle shutdown result in an overall positive safety benefit to the RSS system reliability. Each individual modification has either a positive or no impact on system reliability. The following summarizes the major RSS modifications and their impact on either the IPE analysis or RSS system reliability:

Due to the possibility of air becoming trapped in the RSS pump casing, a vent line was installed on the casing of each RSS pump by DCR M3-97045 as discussed in Section 5. The addition of the vent line results in a positive benefit to RSS system reliability.

During redesign of RSS pumps A and B automatic miniflow valve logic IPE insights were used to change the valves to normally closed. This eliminated an unnecessary failure mode associated with diversion of RSS flow in the event the valve (s) fail to close. Having the valves normally closed does not impact pump reliability because the RSS spray header valves remain open, j ensuring that the pumps would not operate deadheaded. The redesign of the miniflow valves in conjunction with keeping the spray header valves open results in a positive benefit to RSS system reliability.

IPE insights identified that the testing method used during quarterly RSS pump surveillances incurred undue unavailability on an entire RSS pump train. As a result, test flow lines were installed for the C and D RSS pumps

, eliminating the time both pumps in one train are placed in pull-to. lock during the quarterly surveillance. This design also eliminates the need to station a dedicated operator at 3RHS*V43 to provide a flowpath back to the i

26

p RWST during quarterly testing. Installation of the test flow lines results in a positive benefit to RSS system availability.

• . Prior to the current outage, the RSS system was designed such that the A and 8 pumps were dedicated for core recirculation and the C and D pumps were dedicated for recirculation spray. Following installation of the RSS pump discharge orifices the system will be operated such that all four RSS pumps continually provide recirculation spray and a portion of the flow from the A and B pumps provide core recirculation. The success criteria used within the IPE study for the RSS system is 2 of 4 pumps, with 1 pump providing core recirculation and 1 pump providing recirculation spray. The design change has no impact' on the IPE success criteria for the RSS system. Two RSS pumps remain capable of adequately providing both the core cooling and containment depressurization functions.

The RSS piping for each pump train contains a loop seal downstream of each heat exchanger but upstream of the cross tie header supplying the high head pumps. These piping segments,are required to be verified full every month per Technical Specification. Du'e to the possibility of trapping air within these loop seals and potentially impacting the high head pumps, additional vent valves were installed and the loop seal drain line was capped the design change associated with the RSS piping loop seats results in a positive benefit to RSS and high head injection system reliability.

The FSAR stated that the operator can align containment recirculation within 10 minutes upon receipt of the low-low level RWST alarm. The I Training department collected simulator data from 6 operator crews between 9/96 and 10/96 revealing the average response time to be 15 minutes. The increase in operator response time results from increased valve stroke time and improved command and control communication when performing emergency operating procedures. Analysis shows that at least 25 minutes are available for the operators to complete the switchover to cold leg recirculation while assuring pump operation. The change in required operator response time from 10 to 25 minutes does not, in and of itself, directly impact the human error probability of switchover to containment sump recirculation as modeled in the (PE. The operator failure probability for switchover is a function of average opera'.or response time, variation in that time, and time available before adverse consequences can be realistically expected. Specifying a time limit in the FSAR does not impact the average operator response time, nor the variation in time.

These two times are functions of the emergency operating procedures, ]

operator training and experience. Since data have been collected for only one point in plant operating life, that being the 1996 simulator exercises, it can not be determined how these may have changed since plant start-up.

Three-way communication in the control room, as well as some increases in C MOV stroke time, have tended to increase the time it takes for the

, j operators to perform the switchover. However, increased communication

)

would also tend to minimize the error of commission.

l 27

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

-- - - - - ' - ~ ~ ~ r , C ~ .~

l l

1 th i

i i

The time available to the operators considers two potential adverse consequences. Prior to the 1991 change, the required operator action time of 10 minutes was limited by the necessity to preserve RWST inventory for l quench spray in order to ensure that the containment was rendered sub-

! atmospheric within 60 minutes. The 1991 change eliminated that design requirement. The 25 minute operator response time now cpacifies the minimum time available to the operators to avoid net positive suction head

, problems with the charging pumps. Available operator response times

! using realistic assumptions are substantially greater, and it is th5se' realistic times that are input to the'IPE human error probability, not the 10 or 25 minute limiting criteria.

Hence, within the limits of available data, and existing human reliability modeling capability, it is concluded that there have been no overall adverse

( impact on the iPE from plant hc anges related to sump switchover since 1986.

l.

7, l

l 1

l l

i l

e l

l l

}

28 r :- .

._ _ . . . . . . _ =_ _ . _ _ . _ ._ __ .

. . .s ,

." (A  !

Vl 7...OTHERISSUES 7.1 Training I

Due to the nature of the modifications to the Millstone 3 RSS system, significant training has been conducted over the past 2 months. Four I modifications have been presented to MP3 operators during two separate presentations. -

- DCR M3 96077: ECCS orifices and throttle valves, DCR M3-97045: Resolve RSS pump suction issues, DCR M3-97042: Testflowline RSS - C & D pumps, elimination of dedicated operator on 3RHS*V43, and

! DCR M3-97079: RSS cubicle flooding and containment structuralintegrity The second presentation was a review of the first classroom presentation and included dynamic simulator demonstrations after the MP3 plant specific simulator had been modified to replicate the changes made to the RSS system.

In addition, the following items have been presented to MP3 operators during either a classroom or simulator presentation:

{ - Design modification to maintain RSS thermal expansion loop piping full of l

water,

' ECCS surveillance requirement changes, Procedure changes (ES-1.3) addressing the RSS modifications,

' FSAR changes that increased operator action time from 10 to 25 minutes to cold leg recirculation, and Passive RSS failure post LOCA The changes to EOP procedure ES-1.3, " Transfer To Cold Leg Recirculation,"

were covered in significant depth in the classroom, including the basis behind each step. Additionally, each operating and administrative crew performed the l

l transfer to cold leg recirculation on the simulator twice. The first time was a normal transfer with full electrical power available, the second time included the t

failure of one emergency diesel generator.

l' l

The simulator has been updated to ensure that the plant modifications to the l plant are included in the Millstone model.

Operator training for the RSS modifications was discussed with the NRC on L January 29,1998 at the Region I headquarters.

a f

29

- ~ . ~__. . . . . _

t IS i

7.2 System Testing Considerations 7.2.1 Test Program 7.2.1.1 Initial RSS Preoperational Testing

[-

Initial RSS pre-operational testing was performed in 1985/86 using Test Procedure T3306P. This testing was performed to verify that the RSS met its safety functional requirements. The testing encompassed the following:

1) Containment Recirculation Spray Pump Suction isolation l

L Valves (3RSS*MOV23A,B,C,D), and the annunciation, l indication, logic and stroke times associated with these valves, were tested to verify that they can be opened remotely, closed remotely (with no CDA signal present),

and that they will open automatically in response to a simulated CDA initiation signal. Stroke time for each valve tested was satisfactory.

{ 2) Containment Recirculation Spray Water Spray Header l lsolation Valves (3RSS*MOV20A,B,C,D), and the annunciation, indication, logic and stroke times associated with these valves, were tested to verify that they can be opened remotely, closed remotely (with no CDA signal present), and that they will open automatically in response to a simulated CDA initiation signal. Stroke time for each valve tested was satisfactory.

! 3) Containment Recirculation Spray System to Residual Heat Removal System Cross Connect Valves l

(3RSS*MOV8837A,8 and 3RSS*MOV8838A,B), and the hnnunciation, indication and stroke times associated with l

L these valves, were tested to verify that they can be closed remotely, and that they can be opened remotely with the interlocks and logic associated with aligning the valves for the recirculation phase of RSS operation. Stroke time for i

each valve tested was satisfactory.

4) Containment Recirculation Spray Pump Miniflow Valves (3RSS*MOV38A,B), and the annunciation, indication and logic associated with these valves, were tested to verify manual operation and automatic operation in response to an RSS pump running and flow signal to maintain a flowrate greater than minimum flow requirements.

Containment Recircul:nion Pump Miniflow was measured

.)

at 1100 gpm.

e

~

[.

5) Containment Recirculation Spray Coolers (3RSS*E1 A,B,C,D) were tested to verify that the coolers can be filled and automatically vented via the discharge of the RSS pumps without damage to baffles or other internal components of the coolers.

6) Containment Recirculation Spray Dewatering Pumps (3RSS*P2A,B) were tested to verify manual operation and the pump's capability to dewater the RSS. The' test was performed in accordance with OP 3306, Section 7.3.

7) Containment Recirculation Spray Pumps (3RSS*P1 A,B,C,0), and the annunciation, indication, logic and interlocks associated with these pumps, were tested to verify manual operation and automatic response to a simulated CDA initiation signal. The test included RSS train A and B alignment for taking a suction from the containment sump and discharging for cold and hot leg recirculation, with cold and' hot leg RSS direct injection.

All pump logic, safeguard signals, blocking signals, permissives, interlocks were verified for remote and automatic pump operation using Procedure 3lNT-2004, EOP 351.3 and EOP 351.4.

Each RSS train was tested individually. Testing was performed by installing a temporary cofferdam in the containment sump with a capacity of 30,000 gallons of water for about 6 riinutes of pump operation. Makeup to the cofferdam was from a temporary connection from the RWST. Each RSS train was aligned to take a suction on the containment sump and discharge to the suction of a single CHS and SlH pump and to provide direct cold leg injection to 2 RCS loops. The test was initiated using Procedure 31NT-2004 and EOP 351.3 for transfer to cold leg recirculation using the recirculation changeover array at MB-2.

The RSS pumps ("A" and "B") were started on miniflow recirculation and was first aligned for RCS cold leg direct injection. Once the air was purged from the RSS into the RCS cold legs, each RSS pump was aligned to the suction of one CHS and SlH pump, which were realigned from RWST recirculation to RCS cold leg injection. The test was run for 5 minutes with total flow recorded every minute. The test acceptance criterion for RSS pump flow was to be less than 5000 gpm. The total flow from each pump was verified to be less than that.

Each RSS train was tested individually. Testing was initiated from the cold leg recirculation configuration, with the RSS pump discharging to the suction of both trains of CHS and SlH pump and to the cold leg direct injection path. The test was initiated using EOP 351.4 for transfer to hot leg recirculation using the Recirculation Changeover Array at M B-2. The RSS was 31

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

. D, realigned from cold leg to hot leg direct injection and continued to supply the suction of the CHS and SlH pumps. The SlH pumps were realigned from cold leg to hot leg injection. The test was run for about 5 minutes with total flow recorded every minute.

Restoration from hot leg recirculation was performed using EOP 351.4 for transfer to hot leg recirculation using the recirculation changeover array at MB-2.

This test accep,tance criterion for RSS pump flow wis to be less than 500 gpm. However, the total flow from RSS Train A was verified to be greater than 5000 gpm. The test therefore was unsatisfactory and led to the elimination of RSS hot and cold leg direct injection in 1986.

7.2.2 Assessment of RSS Modification Effect On Previous Testing Of the eighteen (18) modifications made since OL that affect the RSS,.

eight (8) were completed prior to the 199.6, MP3 shutdown and ten (10) are in-process for completion prior to restart.

The eight (8) modifications completed prior to the MP3 shutdown are as listed below:

MP3-85-004," Tubing Protection Barriers" MP3-85-014 " Supports Required for Seismic Interaction" MP3-86-094," Modifications to the ESF Status Panel (31HA-ANNMB2E)"

MP3-88-009," Modification to the Cold Leg Recirculation Array  ;

MP3-89-013,"MP3 Containment Design Pressure. Change" MP3-92-004, "Limitorque Torque Switch for Operators Material Change" MP3-93-015," Replacement of 3RSS*MOV23A through D" MP3-94-162," Installation of RSS Containment Sump Flanges" None of these modifications invalidated the RSS initial preoperational startup testing results.

The ten (10) modifications in-process for completion prior to restart include:

MP3 96054, "RSS Component Temperature Rerates" MP3-96056,"RSS Support Modifications Outside Containment" MP3-96063,"RSS Support Modifications inside Containment" MP3-97042," Mini-Flow Test Line for RSS Pumps C & D" MP3-97045, "RSS Pump Restriction Orifices to Prevent Suction Line Flashing" MP3-97063, "RSS Expansion Joint / Support Modifications" l 32

.s '

t P

l MP3-97094, "FiWent Lines ECCS Loop Sear i

l MP3-97102,' Reinstatement of RSS Cold Leg Direct injection" l MP3-97106, "Switchover Time Increased (10 min.-25 min.)"

MP3-98008, " Modification of RSS Pumps' Seal Water Coolers" '

Of tiiese, the cnly in process modification which potentially affects RSS preoperational testing results is MP3-97045, "RSS Pump Rerdnction Orifices to Prevent Suction Line Flashing. Post modification testing for each task under this modification is described below.

7.2.3 Post Modification Testing for RSS Pump Restriction Orifices i

j Containment Recirculation Spray Pumps (3RSS*P1 A.B.C.D) p Based upon the initial preoperational testing results, the RSS pump l

performance has been periodically tested for conformance with the degraded manufacturer's pump curve per the currently approved  ;

Technical Specification requirements. Post modification RSS pump and l restriction orifice testing will validate the initial preoperational testing by performing a full flow test for each RSS pump and orifice in accordance i with the Inservice Testing (ISI) Program for the Proposed Technical j Specification Change for this modification.

l l

Previously performed pump performance calculations and system hydraulic analysis have been modeled to produce results that are in reasonable agreement with the preoperational testing results. For the RSS Pump Restriction Orifice modification, the same computer modeling ll and calculational methodology has been applied in perfomling pump l

performance calculations and system hydraulic analyses, with the restriction orifice modeled.

The purpose of the post modification test is to verify that the i

performance of the RSS pump, with the restriction orifice installed, is as

~

predicted in the computer model of the RSS test loop. If the results of the test validate the computer model pump and orifice performance calculation and system hydraulic analyses predictions, then the pump performance test is appropriate to validate the plant safety analyses as l was done using the original preoperational test results.

Each RSS pump will be tested. The test will be performed by recirculating the flow to/from the Refueling Water Storage Tank (RWST) via the cross connect to the RHR pump test line. Flow measurements l

will be taken in the pump discharge and the differential pressure measured across the pumps, across the restriction orifices, and across

" the pumps and the orifice plates. Test results will be verified to be in conformance with the acceptance criteria for the revised degraded pump curve per Proposed Technical Specification Change (PTSCR) #3-35-97.

33

~

. q l l

Test results will also be compared to the calculated pressure drops and i

flow rates predicted by the computer model of the RSS test loop to verify '

that tne orifices are correctly sized and that the orifice loss coefficient (K value) is ccrrectly modeled. This will ensure that the restriction orifices i

will limii the maximum RSS pump flow to prevent suction line flashing and, at the same time, provide sufficient flow to support the plant safety analysis.

With this verification, the computer modeling, calculational methodology, and the calculated orifice flow characteristics used in the current performance calculations and hydraulic analyses wilfbe' validated as reasonably predicting RSS flow rates for each recirculation phase flow path combination.

Preliminary testing indicates that the acceptance limits for each orifice resistance can be satisfied. The flow rates achieved in the preliminary full flow tests are within 3 percent of the flow rate predicted by the hydraulic model of the RSS test alignment. Based on this result, it is concluded that the methodology for predicting RSS flow rates, including the modeling of system resistances, can be validated with the test data.

Therefore, the RSS flow calculation is considered to be sufficiently accurate to predict the flow rates after an accident and RSS system response and performance appropriate to support the plant safety analysis.

Additional Testino Additional testing for the other design changes made under DCR M3 045 include:

Vibration measurement of the piping as well as the pumps will be performed during the pump performance testing to identify any vibration issues with the pump or the piping system. See Section 7.3, for more detail.

The RSS pump vent lines will be tested by confirming that the check valve closes when the pump is operating and that the check valve is open when the pump is not running. This will be done by running the pumps and verifying that there is no back leakage through the valve and then with the pumps not running, verifying the valve is open by passing air through the valve to ensure the valve has not stuck shut.

The pump casing vent line will be verified to vent the air out of the pump suction casing by filling the RSS pump discharge with the vent line isolation valve closed. Once the discharge line is filled to an adequate level, the vent line vent valve will be opened and the l pump suction casing will be verified to be vented when the vent line vent valve discharges water. Since the vent lines are identical on all four RSS pumps, only one pump vent will be tested.

The new weld joints will be tested in accordance with ASME Ill Nondestructive Examination cnteria and ASME X1 N416-1 pressure 24 l

{

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

^ /.

t n s

. i.

test to be performed to the 1992 Edition of ASME. The RSS Spray Headers will be tested to verify that there are no leaks and no obstructions from the plugged spray nozzles per Surveillance Procedure SP31106. The electrical modifications will be tested in accordance with Test Procedure C-PT-1408.

7.2.4 Post Modification Testing for RSS Miniflow Valves (3RSS*MOV38A,B)

Containment Recirculation Spray Pump Miniflow Valves (3RSS*MOV38A,8), and the annunciation, indication and logic associated with these valves, will be tested to verify manual operation and automatic operation in response to an RSS pump running and flow signal to maintain a flowrate greater than minimum flow requirements.

During the RSS pump performance test, the flow will be redirected from the RWST to the miniflow and test recirculation lines. The flow and differential pressure across the pump and also across the pump and orifice plates will be measured through these lines. This data will form the baseline for future operational testing through the miniflow and test recirculation lines.

The miniflow valves control logic will be verified that the electrical

' scheme functions as designed. A visual check of cable termination's to ensure that the cabling agrees with the interconnection diagrams as well as a test of the circuits ability to function as shown on the all hpplicable diagrams will be performed. Finally, a functional test will be performed of all possible combinations of components in order to uncover possible sneak circuits.

7.3 OperationalIssues Resulting from the RSS Modifications With the exception of vibration there are no new operational issues resulting from modifications to the RSS system. Unlike the other characteristics of the

' system, vibration due to operation is evaluated during the testing phase of modification implementation, Other operational issues (e.g. seismic response, thermal expansion, waterhammer) can be adequately evaluated through the utilization of computer codes during the design phase of the modification process. Walkdowns and as-built drawings are then used to confirm the adequacy of the piping system layout.

e 35 1 .

g i 7.3.1 Testing Results l

7.3.1.1 Pumps and Motors I

f Initial vibration test results have been obtained for the B and D RSS pumps. (Refer to Figure 7 for RSS system flow test vibration monitoring points). The data indicates no significant changes in pump / motor vibration and overati vibration amplitudes remain well within acceptable limits. The pump shaft

! vibration amplitudes did not change, indicating that the pump mechanical seal and bearings are unaffected. Vibration levels i on the motor frame have increased due to flow induced vibration from the orifice plates. The flow induced vibration is seen in the i

l vibration spectrum as increased amplitudes at the motor / pump structural natural frequencies and an increase in the noise floor.

There have been no significant changes at the vibration frequencies which could impact the operation of the pumps or motors. The changes due to the flow induced vibration are of little consequence because of the low amplitudes, and the fact that the motor rotor and frame move together at the structural natural frequencies. This type of vibratory motion contributes very little to the limiting factors of bearing loading and rotor deflection relative to the stator.

The vibration data from the initial test of A RSS pump was lost as a result of the failure of a newly installed vent line. Further testing of the A and C pumps is expected to verify acceptable vibration performance.

7.3.1.2 Piping I

The one inch vent line associated with the A RSS pump failed during 2.itial testing of the A RSS pump / piping. System operating i

vibration and a weak structural design was determined to be the cause of the vent line failure. The vent lines have been redesigned to accommodate system operating vibration (e.g. A flex hose has been installed between the pump and the vent line.

A six way restraint has been located at the interface of the flex hose and the one inch vent line). Piping vibration data has been obtained for the B and D revised vent line piping arrangement l

(refer to Figure 7 for vibration monitoring location and the revised vent line piping arrangement). All piping component vibrations for the B and D pump / piping loops are confirmed to be within L acceptable limits with the exception of the two cantilevered pressure taps associated with each pump. Additional vibration monitoring and evaluations are required to determine if structural l

modification is required to meet ASME OM3 and EPRI guidelines for these taps. Similar results are anticipated for A and C piping.

l 3G i-

j

^ ,-n f ~, l The potential failure of the vent line check valve to seat was evaluated to determine the impact to the vent lines structural integrity, the revised vent line support arrangement and the flex hose ensure that the vent line system can accommodate the potential for reverse flow. The flex hose is designed for the reverse flow that would result from a check valve failure.

7.3.1.3 Expansion Joints The expansion joints located in the pump discharge lines for the "B" and "D" pumps were noted to have relatively high vibration amplitudes on their center rings in the axial direction. The center rings appear to vibrate at their natural frequency of 28 Hz. The maximum vibration amplitude was .12 inches peak to peak. The expansion joint vendor, Senior Flexonics, has specified a not to exceed limit of .24 inches, above which a closer evaluation is

' required. Since the recorded peak to peak displace.nent is less than the maximum allowable displacement, the expansion joint displacement is acceptable similar' results are expected for the "A" and "C" pump expansion joints.

8. INTEGRATED SAFETY ASSESSMENT l 8.1 10CFR50.59 Review Of RSS Modifications l

The modifications to the RSS from the original design approval in 1985 until the current proposed configuration have been summarized in Sections 3 and 5. j These changes can be categorized in two categories: i Changes to restore systems and components to the level of reliability expected in the original design Changes which affect RSS system operation or performance Each of the modifications to the system were evaluated against the criteria of l

10CFR50.59 to determine whether the change constituted an Unreviewed Safety l Question. 'A large number of the changes were implemented to restore the i functionality or reliability of the system as assumed in the original design. These were described in Section 5 and will not be addressed in detail.

l The changes which affect system operation or performance are:

1986 elimination of the use of RSS direct injection Containment pressure change from subatmospheric to atmospheric Reduction in RSS flow due to insta!Iatic.i of orifices Reestablishing the credited use of RSS direct injection to assure mitigation of an assumed limited passive failure e

increase in required operator action time to complete switchover to cold leg

, recirculation e

in addition, changes in analysis methods and inputs will be evaluated, t

37

• . a. .

9 #%

.. , t .

8.2 Elimination of RSS Direct Injection The 1986 change to the EOP's and the MP-3 design basis to eliminate the use of RSS direct injection during cold leg and hot leg recirculation was a change to system operation. This modification changed the RSS direct core cooling safety function from a redundant safety function and flow path to a contingency action in the event of failures to the other injection paths. The EOP's have maintained i steps to use the cold leg direct injection path in the event that an SlH and CHS path was not available. This guidance would support the core cooling function ,

for the design basis limited passive failure mitigation as well as the beyond l design basis multiple failure situation. it should be noted that the safety evaluation written to support this change in 1986 did not specifically aadress the limited passive failure condition.

With respect to the 10CFR50.59 questions, this change does not:

increase the consequences of a malfunctio'n of equipment important to I safety e

create the possibility of a malfunction of a different type e increase the probability of a previously evaluated accident e increase the consequences of a previously evaluated accident e

create the possibility of an accident of a different type The other two questions of 50,59 bear closer examination.

In the original design as described in the SAR, the direct inject valves remained open in the switch over to cold leg recirculation. In order to prevent excessive tube vibration due to high flow, the 1986 change required operator action to isolate the direct injection paths. Since the probability of failure of operator action is not zero, the probability of failure of the RSS heat exchanger increased relative to the facility described in the SAR.

The original design of the RSS system included direct injection during the recirculation phase, without operator action. After 1986, direct injection was available, but required operator action. Since the probability of failure of operator action is not zero, the probability of failure for using RSS direct injection increased.

l

, Based on the additional operator actions discussed above, elimination of RSS l direct injection was a Unreviewed Safety Question.

In considering the impact on the margin of safety as defined in the bases of the Technical Specifications, the impact of the change on design basis analyses must be evaluated. There was no change in the results of the design basis

j. analysis due to elimination of direct injection nor was there an impact on other margins of safety defined in the Technical Specifications. Therefore, there was ,

a no reduction in the margin of safety from this change.

W I

l- 38 1

. .t.

~

After the direct injection path was eliminated, an evaluation of the impact of this change on the core damage frequency was performed by the PRA section.

This ev.31uation is discussed in Section 6 and concluded that the change resulted in an increase in the CDF. The probability of core, damage in this evaluation is the result of multiple failures which are outside the bounds of the design bases.

8.3 Containment Pressure Change From Subatmospheric to Atmospheric The change in the allowable containment pressure from subatmospheric to atmospheric was submitted and approved as a License amendment in 1992.

8.4 Reduction ia RSS Flow Due to Installation of Orifices The evaluation of the reduction in RSS flow from the flow orifice installation included assessments of the impact of this flow reduction on all of the applicable safety analyses. The analyses yielded acceptable results and therefore no reduction in the margin of safety was identified. The impact of the reduced flow on various failure mechanisms and reliability standards was addressed in the safety evaluations. The conclusion of these safety evaluations was that the changes do not involve a Unreviewed Safety Question.

8.5 Credit for RSS Direct injection for Failure Mitigation The change to implement t'he direct injection for failure mitigation consists primarily of an FSAR change and providing guidance to the Station Emergency Response Organization to assist the operators in mitigating an assumed limited passive failure. In addition, physical plant changes are being made to enhance the ability to identify the location of passive failures. The EOP's currently have guidance which will initiate direct injection. These changes formalize the use of direct injection for the design bases and are consistent with the SAR description of the operator response to a limited passive failure. These changes do not result in an Unreviewed Safety Question.

39

1

(%

1 8.6 increase in Assumed Operator Action Time The current FSAR identifies a 10 minute allowance for the operators to i l

switchover from the injection mode to cold leg circulation after receipt of the RWST low level alarm which occurs at approximately mid-level (520,000 gallons). Timing of operatur crews have indicated that the 10 minute allowance is not bounding. Therefore the FSAR is being modified to change the time available for the switchover period.

OCR M3-9710S modifies these FSAR sections to document $n allowable l switchover time of 25 minutes. This 25 minutes is shown through conservative l calculations of the drawdown of the RWST to result in adequate NPSH for the ECCS pumps throughout the switchover period. The evaluation used higher pump flow rates, lower RWST inventory and longer operator response times than the original evaluation. In addition, the NPSH requirements for the charging pumps have increased from 18 ft to 30 ft. The 25 minutes, therefore, is considered to be a consentative minimum time available for the operators to complete the switchover.

The acceptance criteria for this calculation is that the equipment (pumps) are not in a degraded condition due to NPSH or vortexing during the drawdown period. This calculation sets the time limits which the operators must meet in simulations (i.e. the required response time). The changes being processed in DCR M3-97106 justifies an increase in the required response time for the l operators and therefore, is not a USQ.

8.7 Changes in Analysis Methods andinputs l

The changes in the LOCTIC code and inputs were discussed in Section 4. l These changes are minor in nature and do not invalidate the review and acceptance of the original analyses.

In evaluating the changes associated with the RSS orifice installation the calculations on sump vortex suppression, NPSH and suction line flashing were revisited. The original calculations assumed the large break LOCA to provide the most limiting conditions because of the amount of debris generated by that accident. However, the calculations supporting this change have concluded that small break LOCA's can result in lower sump inventory at the time of RSS initiation. Also, the initial start up transient for the RSS results in an initial high flow condition for the pump prior to filling the spray header. This start up period could potentially result in more adverse conditions at the pump suction than the steady state conditions evaluated originally. In order to calculate the start up transient, a methodology was used to examine the debris transport to the sump i

screens fo!!owing initiation of RSS. This methodology is generally consistent l

with the guidance provided in NUREG 0897 Rev.1. However, the new debris i

methodology was not used in the original evaluation. The major change resulting from these new calculations was the lowering of the vortex suppressio,n grate in the sump to assure this grate was covered during the start up transient. An additional assessment of the sump level and vortex grate location was performed and confirmed that using the original debns evaluation method did not result in a reduction in the margin of safety as defined by the 40

i .. .. .

,+ ,

)

(

D.. -

4 original methods. (The original method assumed all of the debris, which could be transported, was on the screens and used the steady state sump flow rate.)

l The change to the facility (vortex grate location) is not an Unreviewed Safety Question.

l

9. CONCLUSION The original design basis of the RSS was reviewed along with all of the changes up to the present time. The results of the review identified that the elimination of RSS direct injection was an Unreviewed Safety Question that was not identified at that time. All

. changes, however, were safe.

i

+.

I l

l 6 41

p

10. REFERENCES 10.1 Design Change Documents

1) MP3-85-004,03/23/88, Tubing Protection Barriers (DMR-481 and 482)

2) MP3-85-014,10/29/93, Supports Required for Seis aic interaction

3) MP3-86-094,04/17/86, Modifications to the ESF Status Panel (31HA-ANNMB2E) - -

4) MP3-88-009,07/17/89, Modification to the Cold Leg Recirculation Array

5) MP3-89-013,01/30/91, MP3 Containment Design Pressure Change 6)' MP3-92-004,02/02/93, Limitorque Torque Switch for 00 Operators Material Change

7) MP3-93-015,11/16/93, Replacement of 3RSS*MOV23A through D

8) MP3+94-162,12/21/94, Installation of RSS Containment Sump Flanges

9) PDCR-94-135: " Installation of TSP Baskets in Containment, Millstone Unit 3" _

10) M3-96054, RSS Component Temperature Rerates

11) M3-96056 RSS Support Modifications Outside Containment

12) M3-96063, RSS Support Modifications inside Containment

13) M3-97042, Mini-Flow Test Line for RSS Pumps C & D

14) M3-97045: RSS Pump Restriction Orifices To Prevent Suction Line Flashing.

15) M3-97063, in-Process, RSS Expansion Joint / Support Modifications

16) M3-97094, in-Process Fill / Vent Lines ECCS Loop Seal

17) M3-97102, in-Process Reinstatement of RSS Cold Leg Direct injection

18) M3-97106, in Process, Switchover Time Increased (10 Min., to 25 min.)

19) M3-98008, RSS Pump 3RSS*P1 A, B, C and D Seal Tank Tubing Modification 10.2 Safety Evaluations i

1. Safety Evaluation No. S3-EB-97-0293, Rev.1,"RSS Pump Restriction Orifices to Prevent Suction Line Flashing."

2. Integrated Safety Evaluation No. E3-EV-97-0040, Rev. O, " Addition of RSS*P1C, D Test Recirculation Lines and Ultrasonic Flowrate Transducers for 3RSS*P1 A, B, C, D Recirculation Lines," 10-28-97.

3. Integrated Safety Evaluation No. E3-EV-97-0014, Rev. O, " Millstone Unit 3 Modifications to the QSS, RSS and Safety injection Systems," 11 97.

4. Safety Evaluation 'No. SE-EV-97-0375, Rev. O, "RSS Expansion Joint / Piping Cupport Modifications," 10-21-97.

5. Safety Evaluation No. S3-EV-97-0014, Rev. 02, " Millstone Unit 3 ECCS Orifices and Throttle Valves."

6. Safety Evaluation No. S3-EV 97-0045, Rev. O, "MP3 Containment l Recirculation System Modifications."

l- 7. E3-EV-97-0043, ISE for M3-97-045, RSS Pump Restriction Orifices to stop flashing 42

v, ..

II l

- C l 10.3 Westinghouse Correspondence l 8. Westinghouse Letter NEU-6098, dated March 5,1986, Integrated Safety l Evaluation on EOP's ES-1,3 and 1.40.

10.4 Stone & Webster Correspondence

1. SWEC correspondenc'e No. NES-40346, dated January 11,1986, MP3-l Post LOCA containment Heat Removal Analysis Safety Evaluation

11. LIST OF ATTACHMENTS l l

l ATTACHMENT 1 REVIEW RSS HISTORICAL KEY CONTAINMENT / RSS PERFORMANCE PARAMETERS COMPARISON l

ATTACHMENT 2

SUMMARY

OF PLANT DESIGN CHANGES FOR RSS e

1

12. LIST OF FIGURES

. Figure 1 A,B,C " Original System Alignments (Injection, Cold Leg Recirculation, Hot  :

Leg Recirculation Phase)"

. Figure 2 A,B,C "1998 System Alignments (Injection, Cold Leg Recirculation, Hot Leg -

Recirculation Phase)"

= Figure 3 "RSS System Schematic Showing Location of Major Design Modifications

. Figure 4A " Containment EEQ Temperature Profile Overlay (1998 versus original 1985)

. Figure 48 Containment Temperature Profile Overlay (Transients versus 1998 EEQ Envelope)

. Figure SA Containment EEQ Pressure Profile Overlay (1998 versus original 1985)

. Figure 58 Containment Pressure Profile Overlay (Transients versus 1998 EEQ Envelope)

. Figure 6 " Vortex Suppression Plate Design and RSS Pump Layout"

. Figure 7 "RSS System Flow Test Vibration Monitoring Points" t

43

ATTACHMENT 1 REVIEW RSS HISTORICAL KEY CONTAINMENT / RSS PERFORMANCE PARAMETERS COMPARISON I Original 1985 1986 1992 1998 License / Original License Eliminated RSS Atmospheric PSS Pump C ,ign Basis Direct lajection Containment Restriction Orifice NUREG 1031 US(B)-273, Rev. 4 US(B)-337, Rev.1 US(B)-273, Rev. 5 US(B)-273, Rev. 6 CONTAINMENT: NORMAL Normal Operating Pressure Range 8.0-10.6 8.9 - 9.8 8.9 - 9.8 10.6 - 14.0 10.6 - 14.0 .

(psia) 120 120 120 120 120 )

Normal Operating Temperature Max ( F)

CONTAINMENT: POST LOCA LOCA Peak Pressure (psig) 45 design 36.09 36.09 38.49 38.40 LOCA Peak Temperature ( F) 280 design 263.3 263.3 26.1.81 262.0 LOCA Max. Depressurization Time (sec) 3600 design 2560 2560 NR NR

}OCA Sump Temperature ( F) -

257.1 257.1 256.9 256.9 LOCA Long Term Sump pH 7.0- 10.5 7.4 7.4 7.4 7.4 QSS/RSS Fission Product Removal (DF) No Credit No Credit No Credit Credited 12 Credited 200 Containment Leak Rate (vol%/ day) -

0.9 0.9 0.65 0.3 Secondary Bypass Leak Rate (vol%/ day) -

0.009 0.009 0.0278 0.0126 LOCA - EAB - Whole Body Dose (Rem) 25 16.8 . 16.8 19.5 8. 2 "* " '

LOCA - EAB - Thyroid Dose (Rem) 300 238 238 150 100 ** " '* '

LOCA - LPZ - Whole Body (Rem) 25 1.59 1.59 3.54 1. 3 "* " '

}-

LOCA - LPZ - Thyroid Dose (Rem) 300 16.1 16.1 31.6 9. 6 "' " '

• INADVERTENT QSS (psia) 8.0 8.02 8.02 8.02 8.02 CONTAINMENT: POST MSLB 2 MSI. B Peak Pressure (psig) 45 design 31.49 31.49 34.14 34.14 MSLB Peak Temperature ( F) 350 327.4 327.4 335.94 335.94 Containnient Liner Temperature 250.9 255.9 RECIRCULATION SPRAY SYSTEM:

44

Original 1985 1986 1992 1998 Licensel Original License Eliminated RSS Atmospheric RSS Pump Design Basis Direct injection Containment Restriction Orifice NUREG 1031 US(B)-273, Rev. 4 US(B)-337, Rev.1 US(B)-273, Rev. S US(B)-273. Rev. 6 TDH @ Rated Flow (ft) 342 342 342 300 300 Rated Flow (gpm) 3950 3950 3950 4130 2200 NPSH, / NPSH, (ft) w/ Debris Loading, As Applicable '

@RSS Pump Start -

13.27/11.0 13.27/11.0 13.27/11.0 19.1/4.0

@ Effective Spray Time 17.49/5.0 17.49/5.0 17.49/5.0 Note 2 21.5/<4.0 )

w Spray Nozzles (per header) 322 322 322 322 162 R

LOCA Spray Header Fill Time (sec) -

87.2 87.2 87.6 153.64 Sump Screen Approach Velocity (ft/sec)

LBLOCA 0% blockage bases .

Full submergence bases 0.24@724 sec 0.24 @ 724 sec 0.17 @ 740 0.12 0.12 sec SBLOCA <0.1 0% blockage bases N/A N/A N/A N/A Full submergence bases 0.3 @ 1200 sec N/A 3

CORE COOLING: 1 ECCS Minimum Safeguards Flow; -

3950 1100 1100 1100 Appropnate Gross Pathway Flow, (RSS Direct (CHS+SlH) (CHS+SlH) (CHS+SlH) 10CA Recirculation Phase (gpm) +CHS+SlH)

Note 1 - Reduction in offsite doses are not affected by RSS changes. Offsite doses were affected by a reduction in containment leak rate and change in spray assumptions per SRP 6.5.2, Revision 2.

Note 2 - 4" small break LOCA in hot leg is the limiting case .

45

l-i P-j l

ATTACHMENT 2 1

SUMMARY

OF PLANT DESIGN CHANGES FOR RSS l

l l CHANGE NO. TITLE DATE USQ DETERMINATION l MP3-85-004 Tubing Protection Barriers (DMR-481 and 03/23/88 - NO USO L 482) -

MP3-85-014 Supports Required for Seismic Interaction 10/29/93 NO USO MP3-86-094 Modifications to the ESF Status Panel 04/17/86 NO USQ l (31HA-ANNMB2E)

. MP3-87-059 RSS Pump Suction LLRT Plug Strongback Voided NO USQ MP3-88-009 Modification to the Cold Leg Recirculation 07/17/89 NO USQ Array .,

, MP3-89-013 MP3 Containment Design Pressure Change 01/30/91 NO USQ MP3 92-004 Limitorque Torque Switch for 00 Operators 02/02/93 NO USQ

, Material Change MP3 93-015 Replacement of 3RSS*MOV23A through D 11/16/93 NO USQ l MP3-94-162 Installation of RSS Containment Sump 12/21/94 NO USQ Flanges l MP3-96-054 RSS Component Temperature Rerates In-Process NO USQ MP3-96-056 RSS Support Modifications Outside In-Process NO USO  !

Containment MP3-96-063 RSS Support Modifications inside In-Process NO USQ Containment MP3-97-042 Mini-Flow Test Line for RSS Pumps C & D In-Process NO USQ MP3-97-045 RSS Pump Restriction Orifices to Prevent In-Process NO USQ .

Suction Line Flashing MP3-97-063 RSS Expansion Joint / Support Modifications in-Process NO USQ MP3-97-094 FillNent Lines ECCS Loop Seal in-Process NO USQ l MP3-97-102 Limited Passive Failure (Design Basis In-Process NO USQ Reinstatement of Direct Injection)

MP3-97-106 Switchover Time Increased (10 min. to 25 In-Process NO USQ min.)

MP3-98-008 Modification of RSS Pumps' Seal Water In-Process NO USQ Coolers I 46

~

l . I!

l ATTACHMENT 2, CONTINUED MP3-85-004 Tubing Protection Barriers (DMR-481 and 482)

Added protective barriers to a number of instrument tubing sections for the sensing line of 3RSS-FT400. upstream of root valve 3RSS*V925. Due to the proximity of l instrument tubing routing to aisle ways and personnel access ways, this modification of barriers was required in order to reduce the potential for tubing damage.. This design change had no safety significance to the RSS design.

1 MP3 85-014 Supports Required for Seismic interaction l Provided additional supports to limit sway of non-seismic Category i piping during a seismic event to preclude unacceptable interaction with seismic Category i RSS piping and equipment. No design changes were made to the RSS system itself.

l MP3-86-094 Modifications to the ESF Status Panel (3,1HA-ANNM82E)

Modified RSS Status Lights on the ESF Status Panel (31HA-ANNMB2E) for 3RSS*MOV20A, B, C, D to provide proper display information by correcting deficiencies identified by the NUSCO MP3 Control Room Review Team per NUREG 0700. This design change had no safety significance to the RSS design.

MP3-88-009 Modification to the Cold Leg Recirculation Array l Changed the Cold Leg Recirculation Array on Main Control Board #2 to provide the operator with an arrangement of control switches identical to the action steps reflected in EOP 35 ES-1.3. With this change, the operator can complete all actions required for switchover from cold leg injection to cold leg recirculation at the array location. Each train for the cold leg recirculation array was modified to hold three (3) additional control switches (for a total of 13 switches) for control of 3SIL*MV8809A,B,3CHS*MV8511 A,8 and 3CHS*MV8512A,B. This design change had no safety significance beyond the existing changes reflected in EOP 35ES1.3 (i.e. elimination of the cold leg direct injection flow paths).

MP3-89-013 MP3 Containment Design Pressure Change Changed the maximum allowable containment operating pressure during Modes 1-4 from 9.8 psia to 14.0 psia. This pressure change ~ incorporated a reduction in containment leak rate (La) from 0.9 to 0.65 weight percent per day and an increase in secondary containment bypass leakage from the containment from 0.01 to 0.042La-The new DBA LOCA maximum containment accident pressure (Pa) is 38.57 psig (53.27 psia). The change was implemented because the original sub-atmospheric j containment operating pressure increased the potential for personnel injury when entering containment for minor repairs that do not require the plant to be in cold shutdown. The change enabled operating personnel access into containment during all

, modes of power operation while reducing the potential for personnel injuries. This change takes credit for QSS and RSS fission product removal and thus constitutes a 47 l .

n gm design basis change; no system design changes were necessary. This containment design pressure change was accepted under Amendment No. NPF-49, dated 1/25/91.

MP3-92-004 Limitorque Torque Switch for 00 Operators Material Change This change allowed the use of Limitorque's modified 00 torque switch (Part Number PN 11500-158) in the replacement of either PN 11500-009 (non-QA) or PN 11500-010 (QA) torque switches. Replaced a 3/32" roll pin with a 1/8",416 stainless ,qteel groove pin. The new part number was the result of changes incorporated to correct a 10 CFR Part 21 condition (basic component deficiency), resulting from fatigue failures of the roll pin. This material change was applicable to Limitorque operators employed on RSS MOVs and did not change the form, fit or function of the original design. The safety evaluation for this change resulted in no Unreviewed Safety Question.

MP3-93-015 Replacement of 3RSS*MOV23A through D Replaced 3RSS*MOV23A through D containment isolation valves with new valves that had an improved seat design. The original valves were EPT (Ethylene Propylene Terpolymer) rubber lined Henry Pratt model N-MK-Il butterfly valves. The replacement valves were Henry Pratt model 1202 butterfly valves purchased under specification SP-ME-784. The change was implemented because of operating problems experienced where the EPT seat became detached from the valve body and prevented the valve from seating properly. Since the seat was bonded directly to the valve body, it was not field replaceable. This caused two failures of LLRTs during outages. The replacement valve material for body, seat, shaft and bearings is identical to that of the original design and utilized the original valve operators; however, the replacement valves now have field replaceable seats. This design change did not change the form, fit or function of the original design. The safety evaluation for this change resulted in no Unreviewed Safety Question.

MP3-94-162 installation of RSS Containment Sump Flanges Permanent 1" thick SA240 TP304 SS test rings with diilled and tapped holes were installed on the four RSS suction lines in the containment sump in order to provide a mounting location for a 14" 150# ANSI blind flange to be used in conducting LLRT Type C testing of the RSS pump inlet isolation valves 3RSS*MOV 23A-D. The test rings are the same diameter as the outside diameter of the RSS suction piping. The safety evaluation for this change resulted in no Unreviewed Safety Question. (Note:

This PDCR was the subject of review in a NRC Report on the Special Inspection of Engineering and Licensing Activities at Millstone Nuclear Power Plant, September 1996).

r A0

[

The following three (3) design changes (MP3-96054, 96056 and 96063) collectively implement design changes resulting from a revised containment temperature profile and a postulated single failure in the SWS supply to the RSS coolers coincident with a LOCA. These were identified as item 1 in the Significant items List. .

MP3 96-054 RSS Component Temperature Rerates Based on a revised containment temperature profile and postulation of a active single failure in the SWS supply to the RSS coolers coincident with a LOCA, uncooled containment sump water will be floveing through the RSS from the sump to the spray header. The uncooled sump water temperature approaches 260 F. This temperature exceeded the previous RSS design temperature and required re-rating and re-analysis of the RSS system piping, equipment and components inside containment and resulted in pipe support modifications; the RSS cooler service water expansion joints and the containment penetrations were also re-rated and analyzed. The safety evaluation for this change results in no Unreviewed Safety Question.

MP3-96-056 RSS Pipe Support Modifications Oytside Containment Based on a revised containment temperature profile and postulation of a active single failure in the SWS supply to the RSS coolers coincident with a LOCA, RSS piping and pipe supports outside containment were re-rated and reanalyzed; this revised analysis resulted in no RSS pipe support modifications. Further, revised movements of the RSS coolers resulted in revised analysis of the SWS piping and supports; this revised analysis resulted in minor SWS pipe support modifications (one (1) spring hanger adjustment and two (2) snubber replacements). The safety evaluation for this change results in no Unreviewed Safety Question.

MP3-96-063 RSS Support Modifications inside Containment Based on a revised containment temperature profile and postulation of a active single failure in the SWS supply to the RSS coolers, RSS piping was reanalyzed and revised pipe support loads were developed for normal / upset ASME Code conditions. Pipe stress levels were quahfied per Table 3.98-11 of the FSAR. Pipe support modifications to 15 RSS supports on the piping system risers and ring headers (increased support weld size, location of gussets) were necessary to support the change in pipe support loads. The revised pipe support loads were also used in a revised containment liner and insert plate analysis.

The safety evaluation for this change resulted in no Unreviewed Safety Question.

MP3-97-042 Flow Test Line for RSS Pumps C & D Previously,3RSS*P1C and 3RSS*P1D could not be tested during Mode 4 due to the need to use a RHR flow path from the RWST. This design change implements mini-flow test lines for these pumps, similar to the flow test lines currently installed for 3RSS*P1A and 3RSS*P18. This will allow testing of the pumps during any Mode of operation. Additionally, this design change eliminates a high risk condition which also existed when 3RHS*V43 was open during Modes 1-3 for ISI testing of these pumps, should a DBA occur during the testing. It also eliminates the need to station an -

operator at 3RHS*V43 during testing. The safety evaluation for this change results in no Unreviewed Safety Question.

49

( _ . . . . . . . . ... . -

t sy l

MP3-97-045 RSS Pump Restriction Orifices to Prevent Suction Line Flashing i

! This design change accomplishes three objectives:

l Installation of restriction orifice plates on the RSS pump discharge lines for 3RSS*P1 A-D to prevent suction line flashing {

l Installation of vent lines on the RSS pump casing to eliminate air bindirtg -

Lowering of the RSS sump vorte'x suppresser grading The installation of RSS suction line restriction orifice plates is necessary to prevent suction line flashing with containment sump water temperature as high as 260 F. This

design change reduces the peak RSS flow to 3000 gpm. With this design change, several other design changes are required. In conjunction with the installation of RSS j suction line restriction orifice plates, the number of spray nozzles will be reduced from l 322 to 162 per spray header to maintain the required droplet size for acceptable heat removal from Containment. The lower flow velocity due to the orifice plate installation has the added value of reducing the impact force of water hammer on the RSS heat exchanger baffle plate such that conformance with ASME code allowable is maintained. Additionally, with the installation of the RSS pump discharge orifice plates, operation of the RSS is changed such that flow to the spray headers will be increased by permitting flow from the pumps which supply core cooling to also continue to supply their respective spray headers.

Based upon an evaluation of the effect of trapped air on startup of the RSS Pumps after a LOCA, a 1 inch vent line will be installed on the RSS pump casing. The vent line will be routed from the vent plug in the pump casing to approximately 5 feet above in the discharge line of the pump. A check valve will be installed to prevent any loss of flow from the discharge line once the RSS pump has started. The check valve will be installed with an isolation valve and a test valve to allow inspection of the valve.

Judicious location of the check valve (i.e., close to the main run header) eliminates any significant water hammer loading in the vent line due to header fluid transient analysis.

The hydraulic analysis of the sump water level, determined that the sump water level could be as much as 4 inches below the bottom of the vortex suppresser grating at the time of RSS pump start in the event of a small break LOCA in the reactor cavity. This was outside the geometric dimension specified in NUREG 0897, Revision 1, published October 1985.

To correct this condition, the vortex grating has been lowered by approximately 12 inches to position the grating in accordance with the guidance provided in NUREG 0897. The new sump design is consistent with the geometric dimension specified in NUREG 0897 and therefore, a new model sump test is not required.

i Tests on these types of vortex suppressers at Alden Research Laboratory (ARL) have demonstrated their capability to reduce air ingestion to zero even under the most l adverse conditions simulated. .

4

By lowering the vortex grating by 12 inches, the Millstone Unit 3 sump design has been l

brought back into compliance with the geometric dimensions specified in NUREG 0897, 50

. , 1 l

n l .

Revision 1. Based on the testing performed by ARL and documented in the NUREG,

! the Millstone Unit 3 containment sump is determined to be qualified by the testing ,

performed by ARL. Additional testing is not warranted. The safety evaluation for these l changes results in no Unreviewed Safety Question. '

l MP3-97-063 RSS Expansion Joint / Support Modifications Defective design of the RSS expansion joint tie rod assembly was reported under LER 97-021-00. Subsequent analysis of'the expansion joints 3RSS*EJ1A/B/C and D and 3RSS*E2A/B/C and D determined that design changes to the RSS pump discharge l piping was required relative to the expansion joint allowable movement and stiffness. '

The resulting design change included: 1) RSS pump discharge piping support desigt; l change, 2) an additional pipe support added, 3) the expansion joint reorientation i relative to expansion joint tie rod orientation, and 4) expansion joint liner replacement l for expansion joints 3RSS*EJ2A/B/C and D only. These changes restore the expansion joints to within their original design parameters. The safety evaluation for this change results in no Unreviewed Safety Question.

1 l l 1 l MP3 97-094 FillNent Lines ECCS Loop Seal 1

Branch lines from the RSS pump discharge lines: down stream of the RSS l Containment Recirculation Coolers are non- self venting because the piping forms an I expansion loop. Air could accumulate in these lines during system filling after a LOCA in Containment. Upon switch over, this trapped air could reach the Safety injection and l Charging pumps and potentially cause pump performance degradation until the air is passed through the pump. l l

These modifications will allow for compliance with Technical Specification Section  :

4.5.2.b.1, modification to the RSS system will mitigate the possibility for air entrainment in the RSS thermal expansion loops.

l MP3-97-102 Limited Passive Failure (Design Basis Reinstatement of Direct l Injection)

Assuming limited passive failure in the ECCS system in the long term post design basis accident (DBA) is a licensing requirement. Discrepancies in MP3 conformance with this criterion were identified during the FSAR verification process as part of the MP3 CMP a Condition Report (CR) was initiated to address inconsistencies in the licensing and design basis documents for the issue and definition of ECCS limited passive failure with respect to the single failure criterion for fluid system design. The CR provides guidance on the proper application of the passive failure criterion. The corrective l action was to generate a DCR to implement changes as required to adequately establish the design basis for ECCS limited passive failure and implement any necessary physical modifications.

l 51 l

e I S y

! - 3.%

. f l

l MP3-97-106 Switchover Time increased (10 min. - 25 min.)

l The transition between the injection phase of a LOCA and the cold leg recirculation t

phase requires manual operator action to realign the suction of the charging and l intermediate head SI pumps from the RWST to the discharge of the RSS pumps (see l Figures 2-A and 2-B), Prior to 1998 the FSAR has stated that the operators could complete this transfer within 10 minutes.

As a result of changes in the valve. stroke times from re-gearing of the"M Vs and changes in the command and control communication protocol in the control room, the time for operators to complete these steps has increased. Therefore, the FSAR is being modified to state that the switchover will be completed within 25 minutes.

Calculations have been performed which demonstrate that the remaining inventory in the RWST after 25 minutes is sufficient to meet the requirements of the ECCS pumps.

MP3-98-008 Modification of RSS Pumps' Seal Water Coolers The Containment Recirculation Pumps 3RSS*P1A/B/C/D utilize a tandem seal to l preclude leakage of the process fluid from the pump to"the atmosphere. These seals are designed to ensure the seat leakage is pushed back to the process fluid stream.

This is accomplished by using two seats and maintaining a higher pressure in the l outboard seal cavity then the inboard seal cavity. The pressure differential between the inboard and outboard seal cavities is maintained at approximately 1 psi by a pressure chamber. It has been determined through calculation that due to line losses in the outboard seal cooling lines the pressure at the outboard seal cavity is not greater then

, the pressure at the inboard seal cavity. This means the inboard seal may leak to the l

outboard seal and subsequently to~ the atmosphere during pump operation. This modification will apply the output from the pressure chamber directly to the outboard seal cavity, thereby preventing sealleakage.

l l

l 52 i

L

, _ 7,,

l l

l l ENCLOSURE 2 1

1 MP3 - RSS SYSTEM; I l

RESPONSE TO NRC INFORMATION REQUEST DATED FEBRUARY 3,1998 l

TECHNICAL SPECIFICATION 6.8.4.a LEAKAGE REDUCTION PROGRAM i

l

~

r f.

1. Purpose l

The purpose of this report is to describe the Technica: Specification 6.8.4.a Leakage l Reduction Program for the integrity of systems outside containment likely to contain radioactive materials at Millstone Unit 3. This report addresses the request made by '

Question 6 of the NRC letter to NNECO of February 3,1997 conceming Miilstone, Unit 3 - Recirculation Spray System. ,

1

2. Background

The following is a summary of the background issues relevant to the current status of the Technical Specification 6.8.4.a Leakage Reduction Program.

A. NUREG 0737," Clarification of TMl Action plan Requirements" NUREG 0737 Section Ill.D.1.1 (Reference 1) disc 0sses systems located outside containment that will or may handle liquids or gases containing large radioactive inventories after a serious transient or accident. The NUREG position, in part, was to establish and implement a program of preventive maintenance to reduce leakage to as-low-as-practical levels. This program shall include periodic integrated leak tests at a frequency not to exceed refueling cycle intervals. NU has committed (FSAR Table 1.10-1) to establish these requirements as a continuing program.

The original 1985 leakage reduction program instituted at Millstone Unit 3 conformed to Technical Specification 6.8.4.a and was accepted by the NRC  ;

I (Reference,2) as appropriate in meeting the criteria of NUREG 0737 Ill.D.1.1.

B. NRC Information Notice (lN) 91-56 " Potential Radioactive Leakage to Tank Vented to Atmosphere"  ;

issued in September 1991, IN 91-56 (Reference 3) broadened the scope of NUREG 0737 to include leakage past valves which would not be externally visible.

IN 91-56 described potential situations involving radioactive water back-leakage in flowpaths, outside of containment, that were not previously recognized as falling under Technical Specification 6.8.4.a criteria. Specifically, it was identified that the potential existed for certain ECCS isolation valves located outside of containment to fail to fully isolate and be vulnerable to highly radioactive containment sump water back leakage. The highly radioactive back leakage could then be recirculated into tanks which are vented directly to atmosphere.

Based primarily on the credibility of valve failure in the leakage flow paths, results l of a NU review of IN 91-56 (Reference 4) concluded that the probability was low that the situation described could occur at MP3. Although the report focused on the leakage paths identified in IN 91-56, it did not expand into other areas that could contain potentialleakage paths. During a design basis seat leakage review -

under the Configuration Management Program (Reference 5), however, an issue i

^ ~"

,i e, 5 O -

l of potential post accident back-leakage from the Recirculation Spray System (RSG) to the Refueling Water Storage Tank (RWST) was identified (CR M3 97- ,

1936). Since the RWST is vented directly to atmosphere, there was a potential for I an inadvenent release of radioactivity not previously accounted for in offsite dose assumptions.

During the NRC ICAVP Gut of Scope inspection, the NRC also questioned (Question 321) how the utility accounted for post accident back-leakage to the RWST (CRs M3-97-3218 and M3-97-4482). This same issue, but in a mere programmatic sense, was subsequently identified during the NRC ICAVP Tier 2 Inspection, i.e., the Tec'n Spec 6.8.4.a Leakage Program was questioned (Question 135) regarding the lack of surveillances to address some possible leak paths such as heat exchangers and valves which may contain leakage that is not externally visible (CR M3 97-4588). This condition was determined to be reportable and was reported (LER 97-061) pursuant to 10CFR 50.73(a)(2)(i)(B), as any operation or condition prohibited by ine plant's Technical Specifications.

C. Recirculation Spray System (RSS) Issues Recently, the NRC has requested additional information (Reference 8) regarding changes made to RSS system design and operating modes. Included, was a request for additional information with regard to Technical Specification 6.8.4.a, describing the methods by which leakage from RSS and associated systems outside containment will be controlled and monitored to ensure that the radiological dose consequences of postulated accidents are within the plant's licensing basis.

The potential for RWST back-leakage from RSS is discussed above. In addition, the following RSS issues have been resolved as indicated:

1. The RSS Pump Seal Water sub-system has a tank and accumulator designed to provide a supply of clean water to the RSS pump seat for a period of approximately 7 days (Reference 9). The NRC inspection (UIR 1014; CR M3-97-4823) identified that the mechanical calculation for leakage from systems identified in Technical Specification 6.8.4.a (Reference 10) assumed a zero leakage rate for the RSS pump seals. The expected, worst case leakage rate from these pumps, however, is 21 cc/hr per pump (Reference 9) which is consistent with the leakage rates from the other ECCS pumps described in the calculation. The mechanical calculation (Reference 10) will be revised to include RSS pump seal leakage as a source for contaminated flJid Outside containment post accident. In order to maintain the calculation output value for total leakage constant, the assumption for " miscellaneous" (unidentified) leakage of 3,000 cc/hr will be reduced by the above RSS pump sealleakage values. This activity is considered a post restart activity since the calculation input to dose assessment calculations will not change.

2. The RSS heat exchangers have a potential for leakage across the tube sheet into the Service Water System. During NRC inspections, it was noted that measurement of RSS heat exchanger leakage was not included in component surveillances as required by the Technical Specification (CR M3-97-4827).

Leakage testing of RSS heat exchanger leakage has now been specified as desenbed in Section 3.D below.

I 2

,1 .".

Leakage across the tube sheets of the RHS heat exchangers into the Component Cooling Water (CCP) is not of concem with regard to the Technical Specification, since the CCP surge tank vent effluent is within the l Supplemental Leakage Collection and Release System boundary.

3. Implementation of the Technical Specification 6.8.4.a Leakage Reduction Program -

The following discussion provides the details of the existing program to satisfy the requirements of Technical Specification 6.8.4a, along with the program enhancements to address the issues identified above:

Technical Specification 6.8.4.a states:

"The following programs shall be established, implemented, and maintained:

n.

a. Primary Coolant Sources Outside Containment A program to reduce leakage from those portions of systems outside containment that could contain highly radioactive fluids during a serious transient or accident to as low as practical level. The systems include the Recirculation spray, Safety injection, charging portion of chemical and volume control, and hydrogen recombiners. The program shall include the following:

1) Preventive maintenance and periodic visual inspection requirements, and

2) Integrated leak test requirements for each system at refueling cycle intervals or less."

The Leakage Reduction Program comprises the preventive maintenance, periodic visual inspection and integrated leak testing elements of the specification for the indicated Millstone Unit 3 systems as follows:

A. Preventive Maintenance

1. Program Requirement - The systems identified in Technical Specification 6.8.4.a shall be included in preventative maintenance programs.

.2. Implementation - Each of the systems identified in Technical Specification 6.8.4.a is included in the Plant Preventative Maintenance Program. Since all equipment in these systems is either welded or considered passive (i.e.

i gaskets, packing, pump seals, etc.) and is subject to the visual inspections j described below, the normal Plant Preventive Maintenance Program is

~ deemed appropriate and satisfies the intent of the program. No special

, preventative maintenance activities which specifically reference Technical l

Specification 6.8.4.a is required.

l L

3 i i

I r l

l l

B. Periodic Visualinspections l

1. Program Requirement - Periodic visual inspections of the systems identified in i Technical Specification 6.8.4.a shall be conducted.

2. Implementation - Each of the systems identified in Technical Specification 6.8.4.a is included in the Visual Inspection Program, inspections are performed during such activities as ASME Section XI Pressure Tests, Operator Rounds, Health Physics Surveillances and Walkdowns, and, for systems not routinely in service, during Operability Surveillance Testing. Any leakage or residual boric acid buildup is identified using a trouble report and corrected in accordance with the Corrective Action Program.

C. Integrated Leak Test Requirements - Visible Leakage l

1. Program Requirement - Integrated, external leak tests for each of the systems identified in Technical Specification 6.8.4.a are performed at refueling cycle intervals or more frequently. .

2. Implementation - An integrated External Leak Test Program has been implemented. Testing is performed in Extemal Leaktightness Verification procedures for each of the applicable systems or sub-systems. In general, these procedures inspect the system or sub-system boundary, measure visible leakage from mechanical joints, and compare with acceptance criteria, sum the total leakage, and trend leakage results over time. Leakage totals are maintained by the Operations Department. The External Leaktightness Verification procedures use'd to perform these inspections are identified in Attachment 1.

A calculation (Reference 10) quantifies the potential total external leakage from system components such as piping, pump seals, flanges, and valve stem leakoffs, and miscellaneous (unidentified) external leakage. The calculation also establishes an assumed RSS heat exchanger tube leakage which would leak internally into Service Water. The total leak rate of 5,000 cc/hr derived from this calculation, is the maximum operational leak rate which, when doubled for conservatism, is the maximum post-LOCA equipment leakage within the filtered (SLRCS) boundary assumed in the radiological consequences analysis for the LOCA event (Reference 11). The current acceptance criteria for the total extemal leakage tests measured in these tests is 5,000 cc/hr with the exception of a zero leakage criteria for several piping segments which extend beyond the SLCRS bcundary. Since these tests do not measure the internal leakage across the RSS heat exchangers, the acceptance criteria for the external tests will be reduced by the assumed total intemal leakage across the RSS heat exchangers of 60 cc/hr per heat exchanger. (Reference 10). This activity is considered a post restart activity since the total leakage criteria remains at 5,000 cc/hr and the corresponding input to dose assessment calculations will therefore not change. Leakage l

testing of the RSS heat exchangers is discussed in Section D.2 below.

4 l

.t .'. ,

• . f.

Where practical, the inspection boundary for the external leakage inspections j is consistent with the ASME Section XI Class 1 pressure test requirements which states that the VT-2 examination (visual) shall extend to and include the second closed valve at the boundary extremity. A review of the current test boundaries has been performed. As a result of the review, some additional tests will be performed prior to Mode 4, and the test procedures will be revised to incorporate the additional tests prior to the next refueling.

D. Integrated Leak Test Requirements - System-To-System Leakage

1. Program Requirement - Integrated, system-to-system leak tests for each of the analyzed pathways among systems identified in Technical Specificatien 6.8.4.a are performed at refueling cycle intervals or more frequently.

2. Implementation - A System-To-System Leak Test Program has been developed and is being implemented. Testing will be performed to measure potentialinternalleakage from the RSS system back-leakage flowpaths to the RWST and from RSS into Service Water through the RSS heat exchangers.

l Leakage testing for each of the identified RWST back-leakage pathways is l currently in progress and tests are planned for completion prior to Mode 4.

The identified pathways comprise the pump suction and recirculation lines l from the Charging, Residual Heat Removal, and Safety injection Pumps, the  !

RHR Purification Cross-Connect, and the RSS Test Line and De-watering l Connections. In general, the testing will pressurize system piping segments l and inspect for leakage from vented piping segments or components on the RWST side of boundary isolation points. A mechanical calculation (Reference

7) provides the basis for the leakage acceptance criteria for each pathway. A radiological assessment has been completed using the back-leakage values assumed in the mechanical calculation. Based on these leakage i assumptions, the dose contribution from these pathways would be approximately 1% of the federallimit. If the actual leakage measured during testing exceeds the values assumed, attempts will be made to reduce the leakage or, the mechanical calculation will be revised and the radiological assessment re-performed. RWST Back-Leakage Verification Test procedures will be incorporated into the overall Leakage Reduction Program prior to next refueling.

Leak tests to quantify RSS leakage into Service Water through the RSS heat exchangers are in progress, and are also planned for completion prior to Mode 4. The acceptance criteria for this leakage is 60 cc/hr for each RSS heat exchanger (Reference 10). In general, the testing is performed by pressurizing one side of the RSS heat exchangers with the other side isolated and drained. Any leakage across the tube sheets is noted by monitoring the open drains. RSS Heat Exchanger Leakage Verification Test procedures will be incorporated into the overall Leakage Reduction Program prior to next refueling.

5

• a M

. 4 I

4. Summary I

( The Leakage Reduction Program described above satisfies the requirements of Technical Specification 6.8.4.a. The program addresses the issues raised in NRC Information Notice (lN) 9156 along with relevant NU and NRC identified issues associated with the RSS system, RWST back leakage, and other system-to-system i lea'Kage pathways. l l .

l S. References -

1) NUREG 0737," Clarification of TMI Action Plan Requirements -Integrity of Systems Outside Containment Likely to Contain Radioactive Materials for Pressurized-Water Reactors and Boiling Water Reactors".

l 2). NRC Letter " Issuance of Supplement No. 4 to NUREG 1031 - Millstone Nuclear Power Station, Unit No. 3", c161001, dated 12/6/85.

l 3) NRC Information Notice 91-56: Potential Radioactive Leakage To TankVented To Atmosphere, dated 9/19/91.

4) NU Memo MP3-92-260, "lEN 91-56

Potential Radioactive Leakage To Tank Vented To Atmosphere", C. H. Clement to T. J. Dente, dated 7/31/92.

5) Stone & Webster Memo MP-0148, " Refueling Water Storage Tank (RWST)

Radioactive Fluid Back-Leakage Concem - Millstone Nuclear Power Plant - Unit 3", J. T. Creamer to M. Etre, dated 6/23/97.

6) LER 97-061-00, " Technical Specification 6.8.4.a Leakage Reduction program Does Not Address All Leakage Paths", B16934, dated 1/17/98.

7) Calculation RWST-01543-03, "ECCS Back Leakage to the RWST During Post LOCA Sump Recirculation", Rev,0, dated 1/16/98.

8) NRC Letter E. V. Imbro to M. L. Bowling, " Millstone Nuclear Power Station, Unit 3

- Recirculation Spray System (Significant items List items 1 and 85", dated 2/3/98.

9) MPR Associates Calculations 282-025-jih-2, " Time That Pressure Chamber Can Provide Water to Seals", Rev.1 and 282-025-jih-3," Leak Rate to Atmosphere After Pressure Chamber Water Volume is Gone," Rev.1.

10) Calculation 1279-746P(R), ECCS System Leakage Outside Containment".

Rev. O, CCN 1, dated 5/12/97.

11) MNPS-3 FSAR, Table 15.6-9, " Assumptions Used for the Radiological Consequences of a LOCA Analysis", Amendment June 1996.

1 6

i

\

I p  !

ATTACHMENT 1 External Leakage Tests Performed i

l Proc. No. Title Freq. Response 3604A.1 2 Charging Pump A Leaktightness Verification Check RO Tech Spec.

3604A.1-3 Charging Pump Common Discharge Header RO Tech Spec.

Leaktightness Verification Check ~ '

3604A.2 2 Charging Pump A Leaktightness Verification Check RO Tech Spec.

3604A.3-2 Charging Pump B Leaktightness Verification Check RO Tech Spec.

3606.1-2 Leaktightness Verification of 3RSS*P1A Suction and RO Tech Spec.

Discharge Header. Common Piping and Valves 3606.2-2 Leaktightness Verification of 3RSS*P1B Suction and RO Tech Spec.

Discharge Header, Common Piping and Valves 3606.3-2 Leaktightness Verification of 3RSS*P1C Suction and RO Tech Spec.

Discharge Header, Common Piping and Valves 3606.4-2 Leaktightness Verification of 3RSS*P1D Suction and RO Tech Spec.

Discharge Header, Common Piping and Valves "

3608.1-3 Safety injection Pump A and Common Header RO Tech Spec.

Leaktightness Verification Check 3608.2-3 Safety injection Pump B Leak Tightness Verification RO Tech Spec.

Check 3613A.2-1 Hydrogen Recombiner Train A Leak Tightness RO Tech Spec Verification Check 3613A.2-2 Hydrogen Recombiner Train B Leak Tightness RO Tech Spec Verification Check 3613A.4-1 Hydrogen Recombiner Train A Leak Tightness As Req. Tech Spec Verification Following Pressure Boundary Maintenance 3613A.4-2 Hydrogen Recombiner Train B Leak Tightness As Req. Tech Spec Verification Following Pressure Boundary Maintenance ,

l i

7 l

-+

t f

RWST .

• 4 CoLo F 0 LEGS

[_ 1 r

S!H PUMP A RSS PUMP D CONTAINMENT

[ ..)

(  :

' $!H PUMP B RCS g g LEGS m ,

RSS PUMP C _

i 4 COLD RHR PUuP A 7

. (SIL)

.i  ;

RSS PUuP B

( RHR PUMP B (SIL ) *** r RSS PUMP A .]

( .

i CHS PUMP A

(  ;

i CHS PUMP B INJECTION PHASE  :

(1cj85 - ORIGINAL SYSTEM ALIGNMENT)

FIGURE IA ,

f

e

~t I

4 COLO '

LEGS r

i S!H PUMP A 3 /

( (

i RSS PUMP 0 CONTAINMENT $lH PUMP B RCS

') ,

t

( i I l 4 CRD i LEGS m RSS PUMP C '

i,

(

b4 COLD

• LEGS RSS PUMP B 0

( ( i RSS PUMP A CHS PUMP A

(

~

CHS PUMP B COLD LEG RECIRCULATION PHASE (1985 - ORIGINAL SYSTEM ALIGNMENT)

FIGURE IB

5

_ 4 HOT LEGS

' SIH PUMP A

. 3 r i

RSS PUMP D CONTAINuENT SIH PUMP 6 RCS },

[

r i

l l 4 COLD 2 HOT LEGS m M i RSS PUMP C r g

[ -_

f ~ %D RSS PUMP B f ,

8

( CHS PUuP A 8

8 RSS PUMP A , )

CHS PUMP B FA O i MULTIPLE ACTIVE FAILURES:

FULLY QUALIFIED)

- I

-l HOT LEG RECIRCULATION PHASE (TWO PATH RECIRCULATION)  ;

(1985 - ORIGINAL SYSTEM ALIGNMENT) ,

FIGURE 1C i

! l -  !: i!1 i 7 ;i : j

.N .-

7 t

S C

R

_ u y D O S D 5 LS _

OC E G

E R _

C L L C _

_ 4 4 _

A B A 8 .

P P P P _

u u M M _

U U U U P P P P )

H H S S _

H T _

( ( H

! I _

S S C C

N A B E _

P P M M

U M

U EN _

P l P I SG R l R L AI

_ I H H I RI S RI S H L P A A _

2 _

NME R T

OEU G S ITI F W

R T S C Y E

J S

)

T N-I N

E M

N 8

I A l 9 T

N 9 f O C l 1

(

O C B A P P P P M M M M U U U U P P P P S

S R

S S

R

( S S

R

[ S S

R

_ Y

~

~

4 COLD LEGS p

S!H PUMP A

[.

RSS PUMP D CONTAINMENT SIH PUMP B RCS }

l

[

l l 4 COLD i

LEGS RSS PUMP C

[ .

4 COLO LEGS RSS PUMP B I

I f $

1 RSS PUMP A CHS PUMP A )I

( l E

l CHS PUMP B l 1

~' ' ~

~ ~ ~ ~ (ON ~ ~((F SIII'OR CNhi t[NA[A[L BLk FULLY QUALIFIED)

COLD LEG RECIRCULATION PHASE (1998 - SYSTEM ALIGNMENT)

FIGURE 28

4 HOT LEGS r

8 SlH PUuP A

[

RSS PUuP O CONTAINuENT S!H PUMP B RCS

[ _

l I 4 COLD 2 HOT

• LEGS m LEGS i RSS PUuP C F 4 ---; j i l

4 COLD i*

' LEGS RSS PUup B e

[ CHS PUuP A ,

l RSS PUuP A ,

8 8

(AVAILABLE FOR PAS $!VE'8 s CHS PUMP 8 FAILURE OR i MULTIPLE ACTIVE ' 8 FAILURES:

FULLY QUALIFIEDI ,

,  ; I\

f!

i

, , i

~

HOT LEG RECIRCULATION PHASE l = rutor auittrian

'----~~~'~~~~~~' '

(TWO PATH RECIRCULATION)

(1998 - SYSTEM ALIGNMENT ) .

i FIGURE 2C

A .

a.

s la ' '.'tCATION SUMuARY As Or 2/12/5s8

' Rss sysTEu sCnOSUE 2HOWING LOCATION '# "'"'" " #W ""'"*"'" " " ' ' ~ " "

OF MAJOR DESIGN M001FICATIONs. g E O f ,"$ U$jl j @ "dOEI N I Sc'ONews ns D(" "$-S6062 ass 5*** **5 'N5'ot (m'='

ou oc Or roun Ass Pow /ux rLO. Panes see=N h DCR M3-97042 Test LpEs.PuuP C & D ONLY ORIF K / eeGZZtE REOuCIK>e /

ns 4 * "2" ;L*4M&a,E, suPPnssER "$

. @ OCa u3-s70s3 Ers i Per surPORis ns  %

$ cCA u3-s707s Esr stRe gE,f*E ns- -)

, 4 uuan in-siseo Catar <Nr sue REsr0RAi 0N vEs CONTAINMENT h DCR M3-9M94 FEL / vtNT LlWs ECCs LOOP SEAL us

$TRUCTURE E.S.F. BUILDING h ""*

E'Y"g'Y E '"'E'b' h CCA u3-97108 [y,7.E" Id ,*"I'W C

uig NO

" O .

, h DCft u3-98000 $pTygML M sistEu ,g g

' #  ; DB RE sTafEMNT p

m E,f"* NOT (FC 2 & 4 afC INJECTION 1" "

A C#5 "EO CL 96-06 WATER MauuElt h

EuPANsf0pt JOINT ND LEC 3 & 4 - TO LlutfE0 f NCY g

Cit? & CAP NALF Op W

ALW AVs OPEN eAFFLE PLAIED W SM3 WODELLING FtPE SUPPORT uaDst DIRECT INJECIION Passive f AILuRE .

SPRAY N0ZZLEs C&O FOR INJECit0N MX #4R EJ IIf4RI FOR LATERAL

• . DE FLICIt0Nsl

,a . . 10 s LER 57-28 EWematNTAL TEW &

N ' ' <

2f# CONst0ERED /

01 IOpetNG ANALTsts PERFORWD 20's NE W M NT/ FILL 10RICINAL ANALists ts0T S WALVEs gd flue PNAsED P!PE TEup uCC FATLLftE h ppg I hA P

]

a n gg - t' .

s!N PuuPs GNTERuEDI ATE NtCN 6( AD 8 8 " 8" '

70 ECCs 4'8 I a g *h g ho504 LDCER CMDif Rss WsL8 utTICAf!Cpk LOOP p

=u X-@ e?Ons"'210?.'JE: *"

m

>J PLMP WENT I'

sCMEN APE A

'g A

1. 4 R if I

t E

E ,ts t fMC LE R* 3b* 39 L

d V (*g {${ R(((RC $(g) (NOEOINb PIMPI suPPP(sst0N ^ C

• I LOOPS C & D ONLY E O I'#IU W REROUTE UIO"INJ"EC"T10N8 Pt Att LER 37-ts DEBRIS TRANsPEWtf e

"yl '! ][lC MEI LINC l.0 .

hRI Cl0ENT FL AsNM LER 97-C28 1,

Esr OLOC simp LER 37 44

+W

., OtsCat

/ 90* H *

• 4 b VNT D000 '4-E I2'

,0R ,E , _s su-

- *a t .. -

r tar sEn O , , .. P=i A.LE C0wAEss0%

d sterRtss A / -

273 I H Arm N

.,. M GrR Our wV .

.. i y

s PuuP Sf,f,.*gTuwAr

@g.

ss 1

- - - - - - ________.__j

. I. -

[ . ,!

.j b .

. t 2

M I I & I I l 8 1 i. p 375 -

, i 1

'I O ' l -

350 g i -

325 -

i

• 300 of 275 -

' ~

, .t .

250 - I i 8E '

4 -

-( 225 . I 200 '

i g r l  !

I 175 -

g t

. d. -

. s ,

e "% =; -

~

150 - . s 323 - .

e

(

• '*~**"*-*"* -

I 100 T .

l, ,

25 . . g

.. t i I I I I I l I 6

50 l'107 4 6 8 i 1 10 100 > l'80I l'10 l'IO S 1*l0 1*10 rmw(sec.)

- - 1998 Pro 61e

-*- 1985 Profile t  :

I ,

- Figure 4a - Containment EEQ Temperature Profile Overlay (1998 versus Original- 1985)  !

4 5  ;

3 i

I ,

i 400

MSI.B (Full DER, P02) '

350

~ \

Envelope

~

' N 300 #!

PSD

\

250 ^^h^

g' .

l ,

l 2 200

, - a\

j p \; _

E Musta(.4r 2 Split, P39) l , mN '

e 13, f // N 100

k

i. N )

MSLB (.84 ft Split,P51)I i

50 1 . . . . . . . . . . . . . . . . . . . . . . . . .

o 160 1000 10000 ~100000 1000000 100000E0 0.1 1 10~

I ' '

Time After Accident, sec .

• ; Figure 48

! Containment Temperature Profile Overlay I (Trans'ients versus 1998 EEQ Envelope)

, i

• e- ,

$; i . .

5 4

.; g

'l s ..

f 3

1 f

i i. .

N 3 is l I 3- 3 I 4 .

'8

+

N *~

, e 1

+ ,

35 - ') ~

a

',9-.

30 -

i

~

1s - . ,

1a . . ,e 1 E i 20 - .

l  ? i 1

,3 _ , g  ; .

J N

. f. N.

10 - t .! N.

~

} [ 'N ~

3 -

L 1 N. s i

is ==.~*'*~._

• y 0

i

.I i l

~

M- ', 'd - i, I f I I I I I 7

1 10 100 I l*l0 3

l'104 l*l0 5 l*10' l*10 l'108

, r- c )

-*- 1985 profile '

1998 profile ,

e 1

. Figure Sa - Containthent EEQ Pressure Profile Overlay (1998 versus Original- 1985)

? ,

I .

. 1 I

. h .

t l l s 3 8 4

t * <

~

i r  ; -

Y  :

• i t .

4 50 , ,

- ' i. ,

+ .

j* ,

4

! HLDER '

i

..- /

40 PSDER s n.

^ '/ '

)

35 Envelope

{'

o

. 25

/. T - .

[(

/ .

e k

D .

3 .PSDER

MSLB (#ull DE R, P12) .

. ~ .

l i 10 i ,  ; \I z

/ / X <,% .") '

3 ' " N_\

.l -

/ / \L H ER

gg,j p ,2 ,p,7 yg a r, y -

l

-5 1000 10000 100009 1000960 10000000 0.1 1 10' .100 .

! Time After Accident, sec .

' ' Figure 5B

. Containment Pressure Profile Overlay i

[ (Transients versus 1998 EEQ Envelope) l'

_ _ - . .- - ,, m' "

s . t .

$ .- I s

. 4 I.E c

85 p

$$Jm*

f_: NI **-, 1..%.. .

9 3 b 'b ,_

ad - -=

1:o

?.

E l

8 .

) )

...{ . .-._. ,._. ._ t _. .

d

s. , ., g s os 's ,,* .- X s g os 's s 3151 g  %\ I

% Si 2,. 4*2 h t \ og i *

• g '$

' s o' ,', .

g :n mi!! .

os s ,

m 3,

\

i

\

s g

\

+s 0% \\

$\ \\

t f.

w s,

a:

E6

{ \

t $\ \ \' ,, fl -

,b S

s , ss \s -

5;,5 g opg i s , ss ts

• j 0% \ \

g

(. 4 g3

- . _ 3. _ . _ .. 5g. _

3 g . _.__

_ _ . 3__ ___- ---- _-'

H*sc w & % ...v $)

t

.. Q } *i J 4 it

• gg u g T g

i 3 ___ w- --

p n,.--Q>>l

~-

,i i

i us ui i,

i,

,g,

,8 e- u .

v

=t ac Y Q '

ug \s

, n g.

nn'

~ s $is ' ,

• 55$ (NN f.-

i n' , ',

. ui i,, ,

s_m;E. . J..

-- . - ,NS . .I .g. .Ag # $-- 4 =. k' hk. .

E{

~

MW.

~

Q , ',','i

. cam ni i -

, anmetrcr Ed a

• p5dtp*

p"y3.= mc4 i un 't ,

, cu.uouow

_. 2 - - - - - -xi , ni is i

, , , ..=,., pggggs .

m ,

, , , , ii . .

. , g %% e h

, .- m.

~~

3 MNN' 2' h

. - V:n ? S 4 Y '

W m o ye

-- O

@*b

• w ES t wo . ~R --

c9 5: *- " >- o -

C*t g 4>-

• 8 tyr BA -

, . , . . p .gA.- . ...' . - . .

g 4;< 8.*. =s *yF g3j- OC " yd.8 . - {d-

, 8 o a,,

6 ~

-cE AM bI~ .4 .m.

• . L1.18 so, "

.I 411$ -Q:V"  ! YJ [5I g*3+a 9l3 t i-jd ~!l e 8d

'EMMMsK # 3 ya. {

Y

• !s am - c. m .

ttuniat Atrdhttd"$"$' tu"

• b n. Ed

$titr F' uter a i iE j

\

sI.I c'I StnigitV.igtinf *o" D

z Ie '""

Q3gt? g4 I& mtunnit?"'uti i-a i

e-

/s O INDICATES VIBRATION MEASUREMENT POINTS

/ s

/ \

g,. . .

,i

_f/

s '

s ,

~

/

i I SUPPORT l _ . . __ ... . .. PLAN VIEW . . _ . .

Q.-- -.

7\

l Z

l ...

I I l  ; A f

~

l i . .

i L.e - -- c

. r-- -.. p.. c. . . ..

q mz X O i

N

/.. . .

,._ _ x ,\

i __

/

T- -

l t x. s 8

SUPPORT X

EM~d as F-1

.\ '

j e

(

l

-l bk.- F d

i 3

L.. j

(- j f(<- .-------- ,-..---.

[. . . ..

y ,/

e gy'

, -i

[ 's F- ,

i I

//

W I PUMP SUPPORT I

i FIGURE 7 l RSS SYSTEM FLOW TEST VIBRATION '

SKM-CMC-PUMP MONITORING POINTS  ;

j r - '

g ,-

.Q -