ML20081D508

Forwards Analysis Summary Rept Re RCS Drain Down Event,Rept Summarizes Analyses to Determine RCS & ECCS Conditions During Actual Event

ML20081D508
Person / Time
Site:	Wolf Creek
Issue date:	03/01/1995
From:	Neises G WOLF CREEK NUCLEAR OPERATING CORP.
To:	Stevens L WOLF CREEK NUCLEAR OPERATING CORP.
Shared Package
ML20081D511	List: ML20081D508 ML20081D548
References
NE-95-0043, NE-95-43, NUDOCS 9503200273
Download: ML20081D508 (8)

Text

p. q O

'"'""o"'c" coaa"s"o~o""c" C

NUCLEAR OPERATING CORPORATION TO: L. W. Stevens (GB-NSE) NE 95-0043 TE 43343 - K04 FROM: G. J. Neises (GB-NE)

DATE: March 1,1995

SUBJECT:

RCS Drain Down Event - Analysis Summary Report This letter provides a summary of the analyses which were perfonned to support the evaluation of the RCS drain down event. The attached repon summarizes the analyses to determine the RCS and ECCS conditions during the actual event. It also discusses the results of several cases if the drain down would not have been isolated.

A calculation package documenting the analysis results is complete and is available upon request. If you have any questions or concerns regarding this summary report, please contact myself or the engineers involved, Dao Nguyen or Mike Howard.

s.

l GJN/DHN/pb ,

Attachments: WCNOC Executive Summary ,

Westinghouse Letter # SAP-95-120 Vendor Calculation Package Listing ,

l i

ec: R. E. Banh (GB-SYS), w/a W. B. Nonon (GB-NE), w/a R. D. Flannigan (SE LI), w/a L. L. Parmenter (AD-OP), w/a M. L. Howard (GB-NE), w/a D. R. Smith (GB-NSE), w/a l J. N. Hseu (GB-NE), w/a G. R. Smith (TR-TR), w/a D. H. Nguyen (GB-NE), w/a Records Management. (GB-DS), w/a l

9503200273 950301 PDR ADOCK 05000482 P PDR 160032 g(DI ' l .

" SAFETY FIRST, MAKES YOU LAST" fg #;

fW E  ;

1

.D

. Attachment I to NE 95-0043 Page1 of4 WCNOC EXECUTIVE

SUMMARY

Introduction This report summarizes the WCNOC, Westinghouse and Computer Simulation & Analysis, Inc. (CSA) analyses for the Wolf Creek Generating Station (WCGS) RCS drain down event. The Westinghouse analysis was performed using WGOTHIC computer code and the CSA analysis was perfonned utilizing the RETR AN computer code. Both analyses determine the fluid conditions in the RCS and ECCS piping following the discharge of RCS fluid to the RWST. Cases were analyzed assuming both operator action and no operator action to isolate the drain down flow path. The main objective of this analysis is to determine 1) core conditions following the event,2) ECCS pump operability during the event and 3) ECCS make-up c9 ability during the recovery period.

Modeling WGOTHIC Model The Westinghouse WGOTHIC model for the RCS drain down analysis was developed based on a generic 4-loop model used in the surge line flooding analyses (Reference 1). A RHR system model was developed using WCGS plant specific data and was merged with the RCS model. He RCS model contains two flow loops (hot legs, SG tubes and cold legs). The loop from which the RHR pump takes suction, was modeled as a single flow loop while the other three loops were lumped together into a second loop. The RHR system model simulates both the normal flow path through the RHR A back to the RCS and the drain down flow path to the RWST.

He RHR heat exchangers parameters, pipe volumes, elevations, areas, diameters, and loss coefficients for th:s model were supplied by WCNOC. The RHR system noding structure was developed based upon the data supplied by WCNOC and by review of the piping isometrics.

RETRAN Model The CSA RETRAN model was developed based on the WCGS specific 2-loop model (Reference 2), i.e., each loop of the model represents two loops in the plant. A RHR system model was developed using WCGS plant specific data and was merged with the RCS model. De RCS model contains two symmetrical flow loops (hot legs, SG tubes and cold legs). Therefore, the loop from which the RHR pump takes suction, was modeled as a double flow loop. This modeling was chosen so that only two running RCPs would be represented at the event initiation. He RHR system model simulates both the normal flow path through the RHR-A back to the RCS and the drain down flow path to the RWST. The RHR heat exchangers parameters, pipe volumes, elevations, areas, diameters, and loss coefficients for this model were supplied by WCNOC. De RHR system noding structure was developed based upon the data supplied by WCNOC and by review of the piping isometrics Analyses Analyses were performed utilizing both computer models for the following scenarios. These scenarios assume initial conditions consistent with either the actual event conditions or the conditions at the beginning of Mode 4 shutdown. Since the RETRAN model's results confirm the results from the WGOTHIC model, only one discussion will be presented for each scenario.

, Attachment I to NE 95-0043 Page 2 of 4 Benchmark Car.e ne actual RCS drain down event was modeled as a benchmark test case with the combined RCS+RHR system model. The analysis results from this benchmark case matched reasonably well with the data collected during the event at WCGS. The benchmark results can be seen in the table below. Based on these results, the combined RCS+RHR system model wasjudged to be acceptable for calculating the system thermal hydraulic behaviors.

The analysis for the benchmark case indicated that voiding level was insignificantly low (less than 1%)in the 24" common header connecting to the ECCS pumps suction during the event. Westinghouse indicated that the pump performance would not be significantly reduced with entrained gas level of up to 5% by volume. For the actual drain down event, the ECCS pumps would therefore operate satisfactorily to provide make-up and cooling as needed.

Benchmark Results WGOTHIC Model Plant Data Initial Conditions Final Conditions initial Conditions Final Conditions RCS Temperature 297*F 308*F 300*F 307*F RCS Pressure 345 psig 205 psig 340 psig 225 psig Pressurizer Level 720 in 288 in 743 in 361 in RWST Level 99% Full 99 % Full Integrated Flow 9625 gallons 9200 gallons Cases with Failure of Drain Dow , ~ solation The analysis was extended ta determine the system thermal hydraulic behaviors when the operator failed to isolate the drain down flow path. Two cases were used in the analysis assuming no operator actions other than tripping the reactor coolant pumps (RCPs) at an assumed 15 seconds. The first case started from the actual event initial conditions and the second case (worst case) started from Mode 4 initial conditions (350 *F and 400 psig).

For case using the actual event conditions, the reactor vessel began to void at approximately 3 minutes into the event and the RHR pump A failed (on an assumed 15% void fraction at the pump suction) about 30 seconds later. The RCS was drained to mid-loop level at about 5 minutes and core uncovery was estimated to begin at 30 minutes. The analysis indicated that at 340 seconds, highly voided steam / water mixture (90% void fraction) existed in the 24" common header connecting to the ECCS pumps suction. This highly voided steam / water mixture would jeopardize all ECCS pumps operation and consequently all ECCS make-up capability.

For the hypothetical worst case in Mode 4, the analysis indicated that the reactor vessel began to void at approximately 2 minutes following the event and the RHR pump failed at 2.5 minutes. The RCS was drained to mid-loop level at approximately 220 seconds and core uncovery was estimated to begin at 24 minutes. At approximately 120 seconds, the 24" common header was filled with highly voided steam / water mixture (90%

void fraction). At this time, starting the ECCS pumps would result in pump failure due to high voiding level in the pump suction.

Attachment I to NE 95-0043

, Page 3 of 4 Cases with Ooerator Actions from WCGS Simulator Run of November 1994 Additional cases were analyzed using the time line obtained from the simulator run at WCGS facility. He main objective of these cases was to determine the fluid conditions in the ECCS piping before and after flow path isolation such that the ECCS pumps operation could be determined. Using the time line obtained from the r simulator run, the analysis assumed that the RCP was tripped at 60 seconds into the event and the RHR pump A was secured at approximately 100 seconds. The CCP and SI pumps were started at approximately 5 and 13 minutes, respectively, ne drain down flow path was isolated at 14 minutes and the RHR pump B was started at 35 minutes. For the hypothetical worst case, the analysis indicated that highly voided steam / mixture exists in the common header before the flow path isolation. The CCP and SI pumps which take suction directly from the RHR rewn line would immediately fail due to high void fraction in the pump suction. Following the isolation of the drain down flow path, the common header was quickly refilled from the RWST and at the time the RHR pump B started (1 minute follewing the drain down isolation), sufficient liquid already existed in the common header to fill the RHR pump suction line which is connected to the bottom of the common header. The analysis  ;

indicated that the void fraction at the RHR pump B suction (less than 1%) was well below the acceptable level suggested by Westinghouse for pump operation. However, it indicated that at the time the RHR pump B was placed in operation, the RCS pressure was already above the pump shut off head and the RHR pump B will not be able to supply cooling to the RCS.

The analysis also indicated that once the drain down is isolated, it would take approximately 5 minutes to fill the common header to the point where a CCP or SI pump may be started.

Effects of RCP Trin Delav Due to the model limitation of both WGOTHIC and RETRAN, the effects of tim RCP trip time on the drain down event was not fully ascessed. Both analyses indicated that any RCP trip time greater than 1 mine- would cause the codes to fail and result in improper termination. De assessment of the effects of RCP trip time was therefore limited to two cases that assumed 15 and 60 seconds trip time. De analysis indicated that delaying the trip time resulted in faster RCS inventory loss and consequently faster core uncovery.

Conclusions in su.nmary, the analysis predicted that if the drain down can not be isolated, the RCS would be in saturated  !

conditions shortly following the drain down. De RCS continues to drain to the bottom of the hot legs. At this time, steam continues to be released to the RWST and for the hypothetical worst case, core uncovery would be expected at approximately 24 minutes. It is indicated that delaying RCPs trip time would result in faster RCS inventory loss and consequently faster core uncovery, ne RHR pump A would fait due to the existence of highly voided steam / mixture in the pump suction. ne common header would be filled with highly voided steam / mixture. His highly voided steam / water mixture wouldjeopardize all ECCS pumps operation and ,

consequently all ECCS make-up capability.

i l  ?

l I

Attachment I to NE 95-0043 Page 4 of 4 In the event the drain down can be isolated, the RCS conditions would depend on the time of the isolation. The analysis indicated the following for the hypothetical worst case, the RCS would remain in subcooled conditions if the drain down can be isolated within 30 seconds. The analysis indicated that the highly voided steam / water mixture did not exist in the 24" common header until approximately 100 seconds. This implies if the drain down can be isolated before this time then the ECCS pumps performance would not be affected. Using the time line from the WCGS simulator run in which the drain down was isolated at approximately 14 minutes, the RCS would be in saturated conditions within 2 minutes. Due to the release of the RCS steam / water mixture, the common header beco'nes highly voided (approximately 90% void fraction). The CCP and SI pumps which were started before the isolation would fail on high voiding in the pump suction. However, following the isolation, the voiding level would drop significantly due to refilling from the RWST. At 1 minute following the isolation, the analysis indicated that the remained RHR pump B would operate satisfactorily. However, it indicated that at this point, the RCS pressure might already be higher than the RHR pump shut off head and the RHR pump will not be able to deliver cooling the RCS. The analysis also indicated that once the drain down is isolated, it would take approximately 5 minutes to fill the common header to the point where an available CCP or Si pump may be started.

References

1. WCAP-14089 Rev.1," Analyses to Develop a Basis for Surge Line Flooding Response to Support Shutdown Operations".

2. NSAG - 006 Rev. O," Transient Analysis Methodology for the Wolf Creek Generating Station", January 1991.

I

Attacturt 2 to E 50043 ]

@!f l

, Westinghouse Erergy Systems acw nenneets awisen  !

J4 Electric Corporation ,,333 Pmswgh Pennsywane 15230-0355 NTD-NSRLA OPL-95483 SAP-95-120 February 23,1995 Ref.: TSM13104400, Rev. 2 Mr. Glenn Noises

~

Wolf Creek Nuclear Operating Corporation P.O. Box 411 Burlington,KS 66839 Wolf Creek Nuclear Operating Corporation Wolf Creek Nuclear Station i

RCS Drain Down Event Analvals

Dear Mr. Noises:

l This letter documents the final deliverables which complete the RCS Drain Down Event Analysis performed by Westinghouse. De following the documents have been previously supplied to l

WCNOC via FedEx on 1/27/95 and 2/23/95. An additional copy of the Summary Report is attached to this transmittal.

Wolf Creek RCS Drain Down Evaluation Summary Report, February 1995 (ATTACHED).

One 3.5' Computer Disk - Drain Down Data PRS Files - Pump suction pressure transient TMP Files - Pump suction temperature transient Westinghouse thth Note: CN-CRA 94-208, Revision 0 Modeling he Wolf Creek Mode 4 Loss Of RRR Event With WOOTHIC Westinghouse Calculation Note: CN CRA 94-208, Revision 1 Modeling he Wolf Creek Mode 4 Laas Of RHR Event With WOOTHIC l MI M M

)

i NTD-NSRLA OPL-95483  :

SAP-95-120 ,

February 23,1995 Page 2 of 3 At the request of WCNOC, Westingbouse included additional plots obtained from the analysis in the 2/23 FedEx delivery. These plots showed the pressure and temperature transients for various  ;

volumes down stream of RHR pump A.

If you have any questions please do not hesitate to contact us.

Very truly yours, w

I l  :- ,

M. C. bich Special Sales Representative TJK Power Systems Field Sales Attachment cc: J.N. Hseu (WC Site) IL  :

W.B. Norton (WC Site WC-AD) 1L ,

D. Nguyen IL ,

K.M. Harvey (WCNOC Doc. Control) IL, IA y e

\ p" i 4 asent amtf myrren5

~ ,

$ ( C / p t. -

i i

E7 5 4 o 3

P1GURE 2 FOR INTERNAL USE QNLY' DESIGN INFORMATION TRANSMITTAL (DIT) 1 1

l C0C'. MENT st; mary j i

Cocument

Title:

ECS D NQ D*Wd N OT kw ht 4St 5, Quality Assurance Record 8: SAP - W - /2 0 mb WRt.A - Of t.- R S-Og3 fource Documents /Reforences: WTb-WM-Cth- M -02,0 WTb - Lto1 - RAS - As - 6 3,3 mb - AL ol - ths . Ms-eu.

QUALITY REVIDi CHECKLIST

• Yes. No J/A  !

1.

Are all design to a Quality inputs fnfomatten, Assurance Racert? and/or results traceable 1.

Are all applicable and appropriate references identtfted? WA

3. Are all asseptions clearly identified?

V

1. h,,asa hA t

, at regu1re and1tfanal verifIcat1en/

5.

Are all Ifetts of asellcattlity citarly deltneated? Y- 1

6. '

Mas all technical and tant taput been accurately and V l correctly transcribed late the DIT? l

?. Mas all proprietary fafematten been identtfled and properly earned? M

• i COM4ENTS:

thw m w Pame te it te m - N - 306 4 0 + b \

I

' l y

Autner/5tgaat W 0 ate di .2/ahs '4L/5toum z/*As-Verif ter/5tgaatsre/Date ' '

' . To be assumed to due ranssuaal leuer as a blind carece copy.

. . . , - Page 5 of 6 * * * *

• I

- _ - . .