ML050750145

From kanterella
Jump to navigation Jump to search
VYC-1959, Rev. 1, Exhibit 1, Vermont Yankee - Proposed Technical Specification Change No. 263 - Supplement No. 24 Extended Power Uprate Response to RAI SPSB-C-35
ML050750145
Person / Time
Site: Vermont Yankee File:NorthStar Vermont Yankee icon.png
Issue date: 03/10/2005
From:
Entergy Nuclear Northeast, Entergy Nuclear Operations
To:
Office of Nuclear Reactor Regulation
References
BVY 05-024 VYC-1959, Rev 1
Download: ML050750145 (40)
v • d • e


Text

BVY 05-024 Docket No. 50-271 Exhibit I Vermont Yankee Nuclear Power Station Proposed Technical Specification Change No. 263 - Supplement No. 24 Extended Power Uprate Response to RAI SPSB-C-35 VYC-1959, Rev. 1 Total number of pages In Exhibit I (excluding this cover sheet) is 39

I~

VY CALCULATION TITLE PAGE VYC-1959 I

DE&S Rem. lTS/VY98-01 Rev 2 VY Calculation Number Revision Number Vendor Calculation Number Revision Number Tidle: DE&S REPORT:ITSIVY-98-01.-Analysis of Tests for Investigation the Effect of Coatings Debris on ECCS Strainer Performance for Vermont Yarnee' QA Status:

D SC [ NNS Q OQA Operating Cycle Nubcrn

' NA The Operating Cycle Number should only be entered here if the results of the calculation only apply during a specific operating cycle otherwise enter 'N/A'.

Calculation Supports A Design Change/Specification? [0 Yes -

No EDCR 97-423. VYS-049 R3 VYDCJMM/TM/Spec No.

Implementation Required7 0 Yes D] No Calculation Done as a Study Only? [ Yes 0 No Safety Evaluation Number: NA (Arilicabiliti determination Performed)

Superseded Calculation Number, Title and Revision VYC-1959. DE&S REPORT:ITS/VY-98Ml."Analvsis of Tests for Investization the Effect of Coatings Debris on ECCS Strainer Performance for Vernont Yankee'. Rev 0.

For Revisions: List CCNs, ills, or SAs incorporatedlsuperseded by this revision:

None Computer Code(s):

HLOSS -

Are there open tiems in this calculationlrevlsion?

O Yes 0 No Review and Approval: (Print and Sign Name)

Preparer:

Prepared and reviewed by DE&S Under Approved OA Prozram EJ.Beui (VY Cognizant & oreparer of 0017 oackaee) t Date:.

/9-1-0/

lnterdisciplinc Revicwer(s):

Not Reauired Date:

Independent Reviewers(s):

Date.

-0C Approved: S.D. Goodwin

-- Pn4y Date: kS-t_.r Final Turnover to DCC (Section 2):

1)

All open iterns, if any, have been closed.

2)

Implementation Confinratioo (Section 2.3.4)

Total No. Pages in Packa (including all attachmen 0 Calculation accurately reflects existing plant configuration.

9g7 (confimation method indicated below)

Walkdown 0 As-Built input review Q Discussion with age ts)

(Print Naame)

OR C NIA, calculation does not reflect existing plant configuration

3)

Resolution of documents identified in the Design Output Documents Section of VYAPF 0017.07 has been initiated as required (Section 2.3.6, 2.3.7) The changes made In revision 1 do not impact the output from Rev 0 used as inout to VYC-1924. RI.

Fnrico J. Betti I

2I Iz-K 0 Primted Name Page I of 6 Pages*

Date I For cacatalons performed using A? 0017Uds ltthe n ba ofpages inthe boy of the calcrdaion.

For vaedor calculations, thIs ithe number of pges of AP 001 7forms aded.

12tke page, reviewforms, data sheets. 50.59, et.)

VYAPF 0017.01 AP 0017 Rev. 8 Page I of 1 I

VY CALCULATION DATABASE INPUT FORM Place thisform in the calculation package immediatelyfollowing the 2itle page or CCNformi.

I WVt. 1"QS I

nggq R Tpnt-ITTSNVQLAIl Rev 2 VY CalculationlCCN Number Revision Number Vendor Calculation Number Revision Number Vendor Name: DIMrS PO Number 97-58141-00 Originating Department:

Mechanical/ Structural Desii_

Critical References Impacted: Q UFSAR Q DBD Q Reload. 'Check' the appropriate box if any critical document Is Identified in the tables below.

EMPAC Asset/Equipment ID Number(s):

S-X-224A. S-224B. ST-X-226A. SR-X-226B EMPAC Asset/System ID Number(s):

RHR. CS. PrimaryContainment Keywords:

ECCS strainer coatins ten flow beadloss de s HLOS For Revislon/CCN only: Are deletions to General References, Design Input.Documents or Design Output Documents required? Q Yest G3 No I

Design Input Documents and General References - The following documents provide design input or supporting infonmation to this calculation. (Refer to Appendix A, sections 3.2.7 and section 4)

Slgnifcanta Critical Difference Affected Reference

  • Reference I
  • DOC i REV ff
      • Document Title (including Date, if applicable)

RevIew tt Program (V)

ARL: "deo tapes of the L and C Tests Performed at Alden Research Laboratory, Inc, Janwry4Februnary 1998.

VYS-049 3

Bettl, EJ, Speclficatfon forRHR and CS Suction Straners, VYS-049: Rev. 2. Change 1, Vermont Yankee Electric Cormpany, November 12,1997.

NEDO.

0 BWROG, Utfilty Resolution Guidance for ECCS Suction Stralner8lockage, Boiling Water 32886 Owners Group, NEDO-32686. Rev. 0, November 1996.

Carborundum Company, Flterfrar Blanket and Met Products - Pm duct Specflcatons, The Carborundum Company, Fibers Dlvislon, Martch, 1990.

Copus, E. R., -Test Plan for Tests For investigating The Effect Of Coatings DebrIs On ECCS Strainer Performance For Vermont Yankee. ITSC Letter Report, January 15.

1998.

Copus, E. R. 'VY L-Serles Pretest Calculations uslng HLOSS, ITSC Letter Report, January 19. 1999..

Copus, E. P., VY C-Series Pretest CalculatIons using HLOSS-, ITSC Letter Report.

February 5, 1998.

VYC.1960 0

Johnson, A. B. and Padmanabhan. M., 'Hfead Lo=ssof Fibrou Insuato Debris, Paint Chips. Sludg. and Turbalenco Effects for Vermont Yankee BUR Suction Sfrahhers, ARL 6b-981M208F,Apri 1998 Mast% P.C. and Souto, F.J., HLOSS 1.O: A Code for the Prodiction of ECCS Strainer Head I Loss, Innovative Technology Solutions Corporatlon. ITS/TS-97.01, Rev. 0, May 15, 1997. 1 I

?P-V9-2 1

VYAPF 0017.07 AP 0017 Rev. 8,,>

Page I of 3

,.M 11.

VY CALCULATION DATABASE INPUT FORM (Continued)

VY CALCULATION DATABASE INPUT FORM (Continued)

Place this form in the calculation package immediately following the Title page or CCN form.

Vv 1959 oQ flPAL-RePnt. TT/IVYQR-O1 1Rev 2 VY Calculation/CCN Number Revision Nurnber Vendor Calculation Number Revision Number Significant Critical Difference Affected Reference

  • Reference #
    • DOC N REV N
    • Document Title (including Date, if applicabic)

Review tt Program

(/)

Shaffer, C.J., W. BemahL, J. Bildeau and DV. Rao, BLOCKAGE Z5 Reference manual.

NUREGICRM-371, U.S. Nudear Regulatory Commission. December 1 996.

VYC-1953 0

Rlpp, fR. TesIPlanforthe VennontYankeeECCSSuction StrnerDesign, Chugging T~ests, Cseffes, Duke EngtineerfngrSSesk TR-A346004S, Janurary t998.

VYC^1955

_Ripp,_R.,

_estPcdureforfheVentYarkeeECCSSucibnStiaieroesign Chugnng Tests, C sefies, Duke Engineering & SenfcesA TR.A34800-07. February 1998.

Souito, F. J. & Rao. D. V., Experimental investlgatfon of Sedimentatlon of LOCA-Generated Flbnt w Debrs and Sludge In BWR Suppression Pools, NUREGCR-6368, U.S. Nudear Regtiatory Corwninsslon, December1.

Zigler, G., J. Brldeau, D.V. Rao, C. Shaffer, F. Souto, and W. Thomas, Parametlc Study of th Potential for BWlR ECCS Strehier Blockage Due to LOCA Generated Debds, NUREGICR-6224. U.S. Nuclear Regulatory Commission, October 1995.

7~14 1.

VYAPF 0017.07 AP 0017 Rev. 8 Page 2 of 3

...'." -- 1... - I..

VY CALCULATION DATABASE INPUT FORM (Continued)

Design Output Documents - This calculation provides output to the following documents. (Refer to Appendix A, section 5)

.ttCritica Affected Reference

  • Reference I
  • . DOC #

REV 1 Document Title (including Date, if applicable)

Program

(/)

VYC-1924 I

DE&S Cal. DC-A34600 006,Vennont Yankee ECCS Suction Strainer Head Loss Performance Assessmentl.RHR and CS Debris Read Loss Calculationse.

BVY-99-164 0

LetterVYNPC to NRC, BVY 99-164, Notieatlion of the Installation of Larger Emergency Core Cooling System (ECCS) Suction Strainers In Accordance with Bulletin 86-03. 12129199.

  • Reference N -
    • Doc I -

4'* Document Title -

        • Affected Program -

Assigned by preparer to identify the reference In ihe body of the calculation.

Identifying number on the document, If any (e.g., 5920-0264, 0191172, VYC-1286)

List the specific documentation in this column. 'See attached list' Is not acceptable. Design Input/Output Documents should Identify the specific design Input document used in the calculation or the specific document affected by the calculation and not simply reference the document (e.g., VYDC, MM) that the calculation was written to suppor f a DBD is used as a general reference, include the most current Interim change number after the title.

List the affected program or the program that reference is related to or part of.

I tt If "yes,' attach a copy of VY Calculation Data' marked-up to rcflcct deletion (See Section 3.1.8 for Revision and 5.2.3.18 for CCNs).

If the listed Input Is a calculation listed in the calculation database that is not a calculaton of record (see definition), place a check mark In this space to indicate completion of the required significantdifferezce review. (seeAppendixA, section 4.1.4.4.3).

Otheriwse, enter N/A. '

If the reference is UFSAR. DBD or Reload (IASD or OPL), check Critical Reference column and check UFSAR, DBD or Reload, as appropriate, on this form (above).

ttt I

74",` Y/y'6 VYAPP 0017.07 AP 0017 Rev. 8 Page 3 of 3

DOCUMENTATION OF COMPUTER RESOURCE USE CALCULATION NO.:VYC-1959 REVISION NO.:-

CCN No.:I-V Computer Used (include manufacturer, CPU Type, and operating system version and level):

Vender Calculation That used HLOSS Under Vendor's OA Program Computer Input Attached*? a Yes 0 No Location/Identifier:

Computer Output Attached*?

s Yes a No Location/ldentifier: Pago 31 and 32of Calculation

  • Large volume input/output should be provided on CD. See Appendix E for format requirements.

List the computer codes used, and complete the following:

Approved per Appropriateness Outstanding SPRs PP 7800 I

Verified or Code Errors'

[Code NamefVersion and/or Script File Yes No I Yes I

No IYes2 No HLOSS.

X X

NA BLOCKAGE X

X NA

.,.. _.I

' Software Problem Report (SPR), does not exist as a reporting method in PP 7800 and AP 6030. Contact the Code sponsor and review any outstanding SPRs or Code errors. [ER2000805]

2 If yes, fill out information below.

3 If yes, include the Code name on the Computer Code line of the title page, VYAPF 0017.01.

If a computer code was not verified In accordance with PP 7800 and AP 6030, or if there are outstanding SPRs, state below why it is appropriate.

Code Name/Script File I

Appropriateness HLOSS Version 1.0 This code was used in a DE&S report. The code is qualified for use in this calculation under DE&S's QA program. The DE&S program was approved by VY.

BLOCKAGE Version 2.5 This code was used in a DE&S report. The code is qualified for use in this calculation under DE&S's QA program. The DE&S program was approved by VY.

VYAPF 0017.06 AP 0017 Rev. 8 5 t 6 Page 1 of 1 I

Calculation/Analysis Review, DE&S Report ITSNVY-98-01 Rev 2 Calc No.

VYC-1959 Revision No. I The following calculation (or Technical report) was done by a Vendor as part of a Contract to Vermont Yankee. All work on this document was done under the Vendors approved QA program. The purpose of using this VY calculation number and cover sheet Is for entering the document into the VY Information Management System. Reviews, Approvals, and Software control are handled under the Vendor's QA program.

Revision 1 to VYC-1 959 Incorporates three changes:

1) Incorporate Rev 2 of the DE&S Proprietary Report ITSNY-98-O1. This report was submitted to the NRC in BVY-99-164 (12-29-1999) to support VY response to Bulletin 96-03. The revision 2 report Included a number of text changes. There was no impact on the numbers generated for use in other calculations.

2 The DE&S Report indicated that the paint chips tested were K&L 7475 material. This was a typographical error. The paint tested at ARLwas K&L E-1-7475. As documented in the Enclosure A purchase documentation, the paint ordered by DE&S was K&L 7475. This is the epoxy topcoat originally used In the VY drywell. K&L supplied E-1-7475 as the equivalent paint for ARL testing.

3) Enclosure A was added to in corporate the Paint Chip Purchasing documentation in the calculation.

This provides source documentation for the correction to the tested paint material.

The applicability determination VYAPF 6002.03 determined that a 50.59 screen was not required.

i Company's Name:_

W Contract Number:

DE&S VY-1 097-08 Cover Page to DE&S Report ITSNY-98-01 rev 2, 1 DE&S Report ITSNY-98-01 rev 2, 32 Cover Page to Non PrprIetary Reportfor NRC Submital ITSIVY-98 NP. Rev 2 1 Non Prprietary Report for NRC Submital ITS/VY-98-01-NP, Rev 2 32 Affidavit of Dennis W. Murdock 2 Transmittal Letter, March 19, 1999 2 Enclosure A: Paint Chips Purchase Information for for ARL Testing 13 Enclosure B: Applicability Determination VYAPF 6002.03 3 Enclosure C: DE&S and ITS Rev 2 Cover Sheets 5 Total Paces 97 PAGE E d

r'

. X

  • Documentto..rTTfV-9-01-NP, Revision 2 I

ANA4LYS OFTST^S FOR 1NVFSTIGATING THIEEFFECT OF COAITNGS DEBRIS ONECCS SIRMNI PERlFOR MANGE FOR

  • ERONT YA.

NON-PROPRIETARY ISSUE

DocaientA~

7-9801-M Revlon 2 Page 1 of32 DUK ENGIN =RIG & SERVICES COMPANY DISCLAIM STATEMENT PlesseRead Carefully D M&S makes no waranty or representation (expressed or implied) with the respect to this document, and assumes no liabilty as to the completeness, accuracy, or useffilness of the information contained herein, or that its use may not infringe privately owned rights: nor does DE&S assume any responsibility fr ailty or damage of any kind which may result from the use of any of the information contained in this report This report is also an unpublished work protected by the copyright laws ofthe United States of America.

I

AnalysL of Tekstsforlnvesfgafing the FecrofCoaDDebris on ECCSStrabzerPerformzncefor Vermont Yankee DocmentZo.XlTSY-98-014PM, Rev~slon 2 Table of Contents Executive Summary............

3 1.0 Introduction.............

S 2.0 L-Series T ests 7

2.1 Qualitative Results from theL Tests............................

8 2.2 L Series Tests - Quantitative Results................

9 2.21 L SeriesData 9

2.2 Densty13timateforiberbedintheLTests 10 2.2.3 FiberDiameterEstimate forthe L tests.

10 3.0 C-Series Tests:.

12 3.1 Qualitative Analysis ofthe C Tests 15 3.2. QuantitativeAnalysisfortheCTests 19 3.2.1 C tests -DensityEstimateforfiberbed inthe C Tests...9 3.2.2 FiberDiameterEsinateforthe Cte sts.

21 3.2.3 DebrisDeposidonAnalysisusingtheA1RLCtests.................................................. 21 3.2.3.1 Test C2a..06 (C2a)-Lo low andwthelowest chugging energy. Fiberpluspaint but very little paint on strainer.............................................................

21 3.2.3.2 TestClc-.06-OM (Clc)-Uwfilowandzero chugging. Piberdebis............... 21 3.2.3.3 Test C2c-0.06-O (C2f)-Low flow and zero chugging. Fiber debris plus paint but very little paint on strainer..

22 3.2.4 Comparison to Blockage Calculations

............................ 23 4.0 Summary and Conclusions.

25 5.0 References....

.2.......

28 AtachmentA................

.30 List of Tables Table 1. VYPaint Chip Size Distrbution.

7 Table 2. L Series TestMatrix.

8 Table 3. L SeriesfData Sets & irividualpoints 9

Table 4. L SerieseadLoss results...

.........................10 Table 5.

aloss vs. L Senes Test Values...........................

11 Table 6. C Series Testaix....

14 Table 7. C Series TestResults

.5 Table S. Comparison of test results and HlLOSS calculations for different fiber diameters.,.

21 Table 9. Blockage Resus vs. C Series Test Vauis..................................

23 list oFigures Figure 1A. Head loss results for test C2a-0.06..18 Figure 1B. Head lossresultsfortest C2b-0.06...........................

18 Figure 2. Photo inage of front side of strainer, Test Cla...........................................

19 Figure 3. Photo image of back side of straier, Test Cla..

20 Figure 4. Comparison ofLowFlowRateResuts for the C Series Tests.22 2 cf32

Anagyss of estsfor Investrgating the Effet ofCoatyngs Debris on ECCS SrainerPerfomance for Vewrnont Yankem Doaiment No:2.; OM-PA Revfrlon 2 Executive Summary Two series of tests were conducted at the Alden Research Laboratories (ARL) to investigate the effect of paint chips and fibrous debris on the performance of the new ECCS suction stacked-disks strainers at the Vermont Yankee plant. The two series of tests were: 1) tests to determine head losses due to a mixture of different types of fibrous debris and paint chips, hereafter denoted asl-Series because they were conducted in the same closed-loop facility used for earlier head loss tests sponsored by the USNRC [Zigler, et at, 19951 and 2) tests conducted to investigate the effzcts of suppression pool tuibulence on the debris bed formation, hereafter designated as the C-Series because they were conducted in the saee suppression pool segment chugging facility previously used for the debris settling tests, sponsored by the USNRC [S auto and Rao, 19961.

LSeies Tests: The specific objective of the L-series tests (loop test facility) was to estimate the head losses due to a mixture of fibrous insulation debris (0.35 lb in mass), composed of 75% (by mass) of N(JXON"W, 20%0 of Fibera and 5% of Temp-MatV, and simulated paint chips (cured epoxy) of given shape (flat pieces) and size distribution, ie., 50% (by mass) of small (118"xl/8" to V2x V?"), 25%/6 of medium (I2xlt2" to 1"xI"), and 25% of large (1"xl' to 28x2") paint chips. Additionally, some tests included 168 g of simulated suppression pool sludge.

The quantities of these materials and their radios to each other were representative of the worst-case debris loading (on a per strainer surface area basis) predicted to result from a hypothetical DBA LOCA at Vermont Yankee.

The rage of approach velocities tested (0.01 to 0.061Il/s),

encompassed the expected approach velocities for the Vermont Yankee strainers (0.02 to 0.04 ills). Significant findings of this series of tests are summarized as follows:

1. Debris was deposited non-uniformly over the strainer, with the bed thickness being higher in the center than at the edges of the strainer. Non-uniform distribution of debris on the strair indicates that deposition of debris on the strainer surface was significantly impacted by gravitational settling and wall eects. This situation (i.e., gravity and wall effects dominating debris deposition) is not expected for the Vermont Yankee stacked-disks strainers. In this
conte, note that these approach velocities were significantly lower than those approach velocities previously tested in this ficility iLe, 0.15 to 1.5 ft/s) [Zigler, et at, 1995] and, therefore, these gravity related effects were never observed before.
2. The head losses measured were significantly lower than those calculated before the tests using the EL.OSS 1.0 computer code. This overestimation in the computer code calculations is main attributed to the intrinsic assumption of uniform debris distribution on the strainer made in the pro-test anasis.
3. One test was conducted with a representative quantity of shredded Annaflex-insulation material, also included as part of the debris mixture. This sample of shredded Anuaflex remained floating on the water surface, without reaching the strainer and, therefore, was not fiurher tested. This behavior suggests that Arraflex inulation debris would not contribute to thehead loss due to postulated post-LOCAdebris collecting ontthe Vermont Yankee stacked-disics strainers.

3 of 32

Anabsir of Testsforinvestlgafng the Effect ofCoaings Debris onECCSSrralner~erformancefor Verront Ynke Dnt.mt~o.ll T-980--NP. Revslon 2 C-Series Tests: The specific objective of the Cseries tests (chugging facility tests) was to investigate the effect of varying levels of turbulence on debris deposition and retention on a cylindrical strainer at the relatively low approach velocities typical of the Vermont Yankee stacked-disks straies. To this effec; two bounding strainer approach velocities, ite, 0.06 and 0.12 ft/s, and four levels of turbulence, ie., higA mediuni low and no simulated chugging, were investigated. Note that the high level of turbulence was estimated to be representative of suppression pool tuibuence occurring late in time during a mediumLOCA blowdown. The same fibrous debris composition and pait chips used for the L-series were investigated in the C-series tests. Significant findings of this series of tests are summarized as follows:

1. The C1 tests, conducted with fibrous debris only, showed that at an approach velocity of 0.12 ft/s, strainer suction forces were dominant over gravity or turbulence related forces, for all of the turbulence levels tested. Note that this approach velocity, 0.12 ils, is at least 3 times hiher than the highest approach velocity expected at the Vermont Yankee stacked-disks strainers (about 0.04 ils). For an approach velocity of 0.06 fls, the fibrous debris remained suspended in the water at the high and medium turbulence lcvels, indicating than turbulence-related forces were dominant over strainer suction and gravity related forces.

.2. The C2 tests, conducted with fibrous debris and paint chips, showed that both types of debris (ier fibers and paint chips) can only be deposited on the strainer at an approach velocity of 0.12 fits and the high level of turbulence required to filly suspend all debris materials. For the mediun and low levels of chugging-induced turbulence tested, very few paint chips were deposited on the strainer, suggesting that gravity related forces (i.e., sedimentation onto te pool floor) doninated the behavior of paint chips. At the approach velocity of 0.06 flls, these tests clearly indicated that neit fiber or paint chips can be suctioned by the strainer at the high and mediumr chugging-induced turbulence levels. After the simulated chugging VWas stopped, only a small quantity of fibrous debris was deposited on the strainer.

3. Overall, the C-series tests indicate that paint chips do not contribute to the head loss due to post-LOCA debris for the srainer approach velocities and suppression pool turbulence conditions calculated forVermont Yankee.
4. As in the mase of the IL-series tests, the head loss measurements for the C-series tests were lower than those head losses calculated with the 3LOSS 1.0 computer code before the tests.

Again, this over-estimation in the pre-test head loss calculations is partially related to the non-uniform distribution of debris on the cylindrical strainer that was observed during the C-senies tests.

S. An indirect result from this series of tests was the estimation of the density and characteristic fiber diameter of the fibrous debris mnxture for use in the Vermont Yankee ECCS strainer per~frmance analysis. Based on a characteristic debris bed thickness estimated from video tapes of the C-series tests, a density of 2.1 lb/f1 is proposed for the mixture of fibrous debris types poftulated to occr during a DBA.

Based on the bead loss measurements, a characteristic fiber diameter of 8.3 um is proposed for the Vermont Yankee ECCS stacked-disks stainer performance analysis under DBA conditions.

4 of 32

  • Anarysts ofTesfsforlnveffigating the Effectof CoatfngsDerisonECCSSralnerPeforrm=cefor Vermont Yankee DocumentNoaIIfTJ-98-NP, Revision2
O INtroduction During a loss-of-coolant accident (LOCA) in a Boiling Water Reactor (BWR) nuclear power plant, pipe insulation in the drywell may be dislodged by the force of the LOCA jet and be transported to the suppression pool. Ths ination debris, along with corrosion products and other miscellaneous debris can block the strams on the suction lines supplying the emergency core cooling system (ECCS) pumps Consequently, this could result in a sufficiently large pressure drop across the strain= surfice that would degrade pump performance.

For the Vermont Yankee (VY) nuclear station, stners have been. designed to accommodate worst-case debris loading such that the resulting head loss across the debris bed does not degrade pump perfornance. This design effort involved predicting the debris head loss at the strainer surface obr a specified set of assumptions, which included:

  • fibrous insulation debris tansported to the strainers,
  • corrosionproducts and othermiscelaneous debrisinthe suppressionpool,
  • pump flow rate, and
  • paintchip debrisinthe suppressionpool.
  • For this design effort to be Vahid, it is necessary to have confidence that the correlation used to

'predict head loss is applicable and conservative for the insulation type and other conditions relevant to Vermont Yankee.

The methodology used to calculate fibrous debris head loss as part of the Vermont Yankee stainer seing activity is based on that developed by the Nuclear Regulatory Commission (NRC) as documented in NAUEG/CR-6224 [Zigler, et. aL, 1995]. This methodology is implemented in theRLOSS code developed by Innovative Technology Solutions (ITS) Corporation JLMast P. Ki and Souto, F. T., I997]. The validation eort conducted for the ELOSS code demonstrated the applicability of the code for calculating debris head loss across NUKON fibrous debris, with and

-without the presence of corrosion products. The primary fibrous insulation found at Vermont Yankee, however, is a mixture offibers rather than just NUKON. While all the fibrous insulation ypes are blankets of fibers, they have somewhat different densities, and the size of the individual fibers is also different. Because the NUREG/RC-6224 head loss correlation is based on the actual physical parameters of the fibrous debris material (le., fiber diameter, densit, and porosity), it is expected to be valid for fiber mitures as welt However, limited data on mixed fiber bed head loss is available to confirm this expctation.

Tests to evaluate generic strainer performance under the specified Vermont Yankee conditions were performed at Alden Research Labs (ARL) in Holden, Massachusetts [lobrson, 1998]. The purpose of the testing was to investigate the effect of paint chips and fiber debris on the performance of new ECCS suction sainers to be installed at the Vermont Yankee plant. Two

  • separate sets of tests designated as L tests and C tests were planned and executed ICopus, January 15, 1998]. These tests are summarized as follows:

S af 32

Analys of Trestsforluvesdgating the Eed of CoatingsDebris on ECCSStrainerPerfonnancefor Vermont Yankee DvumentNo. /PY-98-01.NF, ILvision 2

  • LISeries Tests: These tests were designed to quantify the effect of known quantities of paint chips (wvith known size distnbution) in a frous debris bed on strainer head loss at low approach velocities prototypic to the Vermont Yankee ECCS flow under accident conditions.
  • C-Series Tests: These tests were designed to investigate the effect of varying levels of pool turbulence on paint chip deposition on the strainer surface at low approach velocities prototypic to the Vermont Yankee ECCS flow under accident conditions.

Both test series represented previously untested conditions unique to the Vermont Yankee ECCS performance expectations under accident conditions (Betti, 1997]. Previous head loss' and turbulence2 testing has been performed, but not at the debris loading conditions specfied for Vermont Yankee Design Basis Accident (DBA) and btemeiteeBreak Accident (IBA) scenarios and not at the low approach velocities that would be prototypic for the new ECCS strainers.

This report summarizes the results of the ARL tests performed for Vermont Yankee, along with analysis of the tests using HLOSS 1.0 and BLOCKAGE 2.5. The results can be used to apply the HLO SS code to Vermont Yankee strainer designlevaluation efforts.

1.

Hcad &m Ted mcks iatabhzu-Tef atd wr ocast.15 - L5WAol & fbtsh;dg- (nwO loads 1sce2, -UREwe

24., A.

F.,Y

-24. Tobl EIS 9IMgcetsI. 399531 2..

Qh:6n Teatm

  • st rs

-Tm zdsdcb letli 10 J: 15 n2zf-rA c

1 (NUKOId3s4 dsk) as1I u hu oxip~dcks.

eN G-F B.5 Fy, 36-29 [Zlglet. d al, 19953 6d32

Analysis of Tests forInvestigatng the Effect of CoaingsDebris on ECCSStraterPerjormancefor Vermont Yankee Document Yo.:d.-98-01.-NP, RiP's5on 2 2.0 -Series Tests These tests were performed in the 'Tead Loss Loop Pacility', which consists of a 100 gallon flow loop with a flat disk strainer of approximately 0.8 ft2 in area Flow through the loop can be closely controlled via variable speed pumps so that values of head loss versus (steady-state) velocity can be generated for a known debris bed on the strainer. Values of head loss at representative VY flow velocities were generated for a range of debris quantities/composition.

The important parameters that were varied aere:

  • Fier qatiy. A miiture of 75% (by mass) NIXON, 20% Flbermat, and 5% Tempmat representative ofthe insation mix in the Vermont Yankee VY containment was used for 6 of the 7 tests. Two quantities of shredded fiber were used:
  • 0.35 lb., which is representative of the fiber loading per urnit area expected from a D}3A, was used intests L-1, 2, 3, L-4, and L-5,
  • 0.05 lb., which is representative of the fiber loading per unit area expected from an IBA, was used intest L-14.
  • SMudge: All of the tests were run inital without sludge. Pour tests included 168 g of (BWROG simulant, NUREGaCR-6224-iron oxide mixture) sludge at the end of the test to quantify the impact of combi&ed sludge and paint chip particulate in the debris bed.
  • Paint Chip Size: The shape and size distribution of unqualified coatings debris is highly uncertain. Thus, a rdix of VY specified sizes [Betti, 19973 was explored as shown in the following table. In most cases, a thickness of 7.5 mlls was used, with one comparison test (J-5) at a thickness of 15 mils.

Table 1. VYPaint Chip Size Distribution Size l

Size Ringe Small 118" x 1/8, to V" x 1/2" Medium V2n 1/2to 1" x l" Large I nxl 'to 2" x 2 Mx 50°h sma%, 25% medium, 25% large

  • Pai Chip Quanti The quantity of paint coatings that would be expected to be destroyed during a LOCA is also highly uncertai Thus, a range of values for the quantity of coatings debris was explored. Note that the largest quantity specified (20 ft?)

represents slightly more than 20 fi? of coatings debris per square foot of strainer surface area

  • Strainer approach veocity: The Vermont Yankee strainer is expected to operate tnder ECCS flow conditions of 7400 - 14200 Spm for the 800 2 MM system strainer and 4000

- 4600 gpm for the 430 ft2 CS system strainer. The range of approach velocities (0.01 -

0.06 fWls) which encompasses these conditions was applied for each of the debris loading cases.

7 of 32

Analysis of TestsforInve~srgatg the Effect of CoatingsDebris on ECCSStrainerPerformancefrr Venront Yankee DOcumentNo.:Z1S1VY-98-01-NRP, Revision 2 Table 2 summarizes the 1L series matrix of tests, varying the relative quantifies of each of the above debris types. These tests were perfnoed in January 1998.

Table 2. L Series Test Matrix Test Fiber Type Fiber Sludge Paint Size Paint Flow Velocity Mass Mass Amot

(

1 (thickness - mi)

(e (Ils)

II1 Mix 035 0.0 0.01-0.06 L-2 Mix 035 0.4 0.0 0.01-0.06 13 MI 0.35 Mix (7.5)

S.0 0.01-0.06 LA MIx 035 o.4 Mim (7.5) 5.0 0.01-0.06 L-5 035 0.4 mA (15) 5.0 0.01-0.06

-l1 Mix (7.5) 20.0 0.01-0.06 L-14 Mm 0.05 0.4 Mix (7.5) 20.0 0.01-0.06 2.1 Qualitative RIsults from the L Tests The L tests effectively met tlheir designed goal of covering the parametric space, which included low flow conditions and the individual and combined debris constituents of paint, sludge, and fiber. Tests I-1 and L-3 provided a baseline head loss comparison between fibrous debris with and without paint chips at the relatively low flow velocities representative of VY strainer conditions. Results ofthese two tests indicated that head losses were less than pretest predictions

[Copus, January 19,1998] using theoretical input values or the extrapolated results from the high flow velocity database as indicated by Appendix E of NUREG/CR-6224. Tests L-2 and [A repeated the comparison between fiber and paint debris beds when sludge is also present These tests indicated that sludge was a significant contributor to total head loss and that the total head loss was still less than pre-test predictions based on the high velocity database. Tests L4 and L-5 compared 15 mil thick paint debris to 7.5 ml paint debris. These tests indicated that paint of either thickless produced similar result, both having head losses less than pretest prediction values. As an option, a small amount of Armnaflex insulation was to be added to the test loop dutng Test L-3. The purpose for this additional point for Test 13 was to determine the transportability of the Arnaflex insulation type. Due to its closed-cell construction, Armaflex was found to be highly buoyant, was not trAportable, and could not be added to the debris bed. This test confirmed that Armaflex debris would not contribute to head losses across the strainer surfice.

All of the above tests were for a fiber loading representative of DBA conditions. Tests L-1l and I-14 repeated the ivestigation of low flow conditions for paint alone and a fiber loading representative of IBA conditions. The I I test innestigated the separate effect of a large paint loading (20 fe vs. a strainer area of less than 1 fi) without a Aber or sludge component. The test indicated that paint alone could not completely block flow and that head losses would be relatively low at low flow velocities.. The L-14. test combined the effects of a small.fiber bed, a 8 of32

Analysis of TezsforInvestigating the Effect of Coatgs Debris on ECCS Strainer Performuncefor Vermont Yankee DacwmerztXo.*7SY-98-O1-NP, Rvsion 2 large paint loading, and sludge. This test indicated that the addition of fiber and sludge would increase the head loss due to large paint loadings alone.

The general data trend is that head losses across a generic strainer for the.VY loading conditions and flow velocities have been lower than predicted. Also, fiber bed densities have been much less than the theoretical values used for input to the pretest calculations.

2.2 L Series Tests - Quantitative Results Seven tests (L1, L2, 13, I4, L5, L11, & L14) were performed, which resulted in ten separate data sets and 69 separate data points at conditions that are applicable to VY specifications. A summaryofthis data is given in Table 3.

i 0.01-0.06 f/s flow range 2.2.1 L Series Data Arange of pressure drop values in terms of inches of water was neasured for each type of debris over the expected range of VY ECCS flow velocities. The results for two of the flow velocities are reported below in Table 4 by debris type.

9 of32

. Analysis of Tessforlnvestgating the Effectof CoatlngsDebris on ECCSStrainerPerformancefor Yermont Yankee DownentAro.U7S'VT-9&O-N. Revision 2 Table 4. L Series Head Loss results PROPETI~rARYINFORM&ATIONREMOVED>

2.22 Density Esimate for fiber bed in the L Tests All but one of the tests formed thick 0.35 lb fiber beds on the 0.8 ft2 flat plate strainer. Visual observations indicated non-uniform loading densities due to large fiber fiagments and a dominant settling velocity (vs. flow velocity) force with debris thicknesses estimated at 2-6 inches and slightly different behavior fr6m test to test These thicknesses can be estimated/veified perhaps to within 0.5 inch using the filmed results of the tests [ARL, 1998]. The debris was thickest in the center of the strainer and thinnest at the edge position, with the difference being about a factor of two. Turning off the flow resulted in ecpansion of the fiber debris bed by as much as 1 or 2 inches. The thickness was measured during the IA test at 3-3.5 inches at the edge and 5-5.5 inches in the center of the strainer. All of the debris was estimated to be on the strainer at that time, and using a density of 1.5 lblft would yield an average thickness on the strainer of about 3.7 inches. The approach velocity was 0.02 W/s for these initial bed thickness measurements. For 100% deposition of the 0.35 lb fiber mass at a density of 2.77 lb/Ct3, the fiber would fbrm an average fiber bed of -2 inches. Accordinly the apparent density of the fiber debris bed could range from 1.8 lbWI 3 to 0.9 lb/ft3 depending on the flow velocity, location on the straer, and time of debris bed measurement. Based on an average fiber debris depth of 3.7 inches, a nominal L test density for head loss estimates would be -1.5 lb'f with an uncertainty factor of 30%.

2.2.3 Fiber Diameter Estimate for the L tests Pretest predictions were performed using a density of 2.77 Ib/fe and an effective fiber diameter of 7.1 m based on a wei pernt a of the am1facturcr as-fabricated data. These values produced estimated head losses of approximately 6 inches of water (vs. 1.5 inches acWua measured) at 65F and an approach velocity of 0.04 s. Both the average fiber diameter (which is used to estimate the average surface-to-volume ratio for the entire debris bed) and the debris bed average density are dependent on specific debris loading conditibns, and different input values for both-values are indicated in order to analytically match the observed results. FirsM different densities were clearly observed.' Based on an observed average fiber debris depth of 3.7 indhes, the L test density for head loss estimates would more appropriately be 1.5 lbN. Second, very non-uniform debris loading was seen in the L tests, indicating non-uniform flow through the strainer. Read loss measurements from the L tests can be conservatively matched using a density of 1.5 lbIf and an effective fiber diameter of 10.7 pm. Table 5 summarizes this information.

.10of32

  • Analysis of TestxforJlvestfgating Ow EffectofCoaffngsDebris on CCSSaronererrfomoancefor Vermont Yankee DocumentfNT 7-9rf $9-0 P, Reisioon 2 Table 5. Hoss vs. L Series Test Values r

<PROPIBTARY INFORATION REMOV.ED>

i 11 of 32 t

Analysis of estrforInvestigahing the Effect of CoatingsDebris onECCSStrainerPefoinnancefor Vermont Yankee Doarnent No. T 01-P, Revsfon 2 3.0 C-Series Tests These tests [Ripp, January 1998] were performed in the (modified) "Chugging Facility." This facility, which is a scale model of a section oftheBDWRMarkI torus, uses pistonsto simulatethe turbulence induced by the lowncomers. Previously this facility had been used to investigate debris sedimentation under various levels of turbulence [Souto and Rao, 1995]. A modification to this facility added a small (2.8 f11) cylindrical strainer in the pool along with the associated piping and a variable speed pump. This fiity was then used to investigate the effect of varying degrees of pool turbulence on the rate and amount of fiber and paint chip debris buildup on the strainer (as inferred from the time-dependent head loss across the strainer). No sludge debris was used In these tests. This was done to observe the fiber and paint chip debris behavior in the suppression pool for the relatively low strainer approach velocities (0.02 - 0.04 fiWs) representative of the Vermont Yankee strainers.

The fiber and paint chip debris quantities used in these tests were scaled (on a per square foot of strainer surface area basis) to those specified for an IBA and DBA at Vermont Yankee. The cylindrical strainer mockup used in these tests had approximately 1/400 times the surface area of the stacked disk strainers in the VY facility.

Four levels of pool turbulence were investigated in these tests:

El

- Stroke of 2 ft and a frequency of 57 strokes per minute for four downcomer tubes each with an area of 0.5 ft2. This was estimated to be representative of suppression pool turbulence energy level occurring late in time during a Medium LOCA blowdown which relates roughly to a hypothetical 900 s VY IBA in a Mark I torus. A discussion of the chugging phenomena expected during a LOCA, the scaling analogy used to design the ARL test facility, and some previous results for fiber and sludge materials may be found in the NLMEG/CR-6368 report [Souto and Rao, 1995]. The EI turbulence level used in the C test series resulted in a turbulence level that suspended both paint and fiber debris.

MD - Stroke of 2 ft and a frequency of 27 strokes per minte. This resulted in a turbulence energy level at 22% of the HEtubulence leveL This level begins to suspend paint debris in addition to fiber debris.

LoQ

- Stroke of 2 ft and a frequency of 13 strokes per minute. This resulted in a turence energy level at 5% of the EI turbence level. This level was sufficient to keep fiber debris suspended off floor, but could not suspend any paint.

Zero - No stroke, recrculation flow only at 75 and 150 gpm through a 1 ft2 splash plate at the pool surface. This results in a turbulence energy level of less than 1% of the EH turbulence level. This level was sufcient to partially circulate fiber and extend its settling time (relative to the quiescent pool settling time), but had no effect on paint settling times.

Two strainer approach velocities were investigated in these tests:

12 of32

An=4ys of Testrforl IlestigatdngitheEfofeCoarngsDebrfs on ECCSStrainerPerfonranc Jfor Vermont Yoakee onentSJ.MYZY-9801, Revosfon 2 V- 0.0611Ws

- nominal approach velocity (based on total. surface area) for VY V = 0.12 W/s

- high approach velocity (based on circumscribed surface area) for VY strainers The fiber and the paint debris used for pool loading was the same as that used in the L test series.

For fber debris, the mixture of NUKONTenmpmatfFbennat described in Section 2 was used.

Tbis material is representative of the materials in Vermont Yankee and also compares reasonably to the generic reactor maerials used to dvl the s

paint chip debris, a cured epoxy paint 7

in the sizes and size distribution shown m Section 2 -was supplied by Keeler and which was similar to epoxy paint found in the VY plant. Tbis paint debris had an approximate thickness of 7.5 ils and the weight per square foot was about 312 g/tf. This material is reasonably representative of reactor paint materials in terms of size, density, and settling velocity based on test results documented in the BWROG URG.

Although the actual coatings in the VY drywell vary in thickness (7.5 mil to 15 mil range considered in the L-seies tests), the 7.5 mi paint debris was used for these tests, since the thinnr debris would have a lower setting velocity and thus would be more likely to transport to the strainer. The use of the thinner paint debris is considered conservative for that reason.

The parameters varied in the test matix for the C series tesits, shown in Table 6, are. the pool turbulence level, the strainer approach velocity, and the debris loading in the pooL Of these prameters, both the material parameters and the tuibulece parameters were intended to be relatively generic due to their high degree of uncertainty, with the key parameter being the relatively well defined Vermont Yankee specific approach velocity (0.02 fits on average for the VY CS system and 0.04 fWls on average for the VY RHR system) to the BCCS stainers This approach velocity would be representative ofthe flid velocity at the debris surface prior to hilling of the gaps betwcen the strainer disks. After the gaps are filled with debris, the fluid velocity at the debris sface would be determined by considering the total straner flow through the ciurimscibed area of the strainer. This velocity ranges from 0.06 fils for the CS system to 0.11 W/s for the PER system.' As can be seen in Table 6, the strainer approach velocities actually tested ranged from 0.06 ft/s to 0.12 ft/s. Thus, this is somewhat higher than actually expected, for conservatism (highes surface velocities would tend to favor debris deposition on the strar).

This 'sa tvwoaraghical error.

Ph alnt teste tALwsB 45 s

cdocumented in the Enclosure A purchase documentation, the Paint ordered by DE&SW was K&t 7475. This is the epoxy topcoat oriainallv used in the VY drywell. K&L sup ied)

E-1I-7475 as the equivalent paint for/ARL testina.

13 f32

-Analysis of Testsfor nvestigating thre Effect CoalingsDehri onECCSStrainer.Perfonnancefor Vernmnt a

Yankee DocumentNo.jTVY-98SD0IMPF Revision 2 Table6.CScriesTestMatrLx Test StrainerApproach Velocity Pool Turbulence Pool Debris Load co 0.06 -.12 fts a, MED, LO & ZERO None-clean plate Cla-0.12 0.,12 rs LO Fiber only Clc-0.06-1irz 0.06 NO D

(1 lb mibred)

ClC-O.06-OHz 0.06 flIs ZERO Cld-0.12 0.12 fs LO Clb-0.06 0.06 ft/s NME Clb-0.12r 0.12 fls

_M_

C2a-0.06 0.06 I/s LO DBA load C2b-0.06 0.06 flIs MED (1 lb fiber, C2b-0.12 0.12 f/s MED 170 ft2 paint)

C2c-0.12 0.12 fr/s

-E C2-0.06-l1z 0.06 W/s Ii C2-0.06-OHz 0.06fl/s ZERO__

For a given test nm (C-2, for examnple), a comparison of the time dependent head loss at the four pool turbulence levels was designed to demonstrate how pool turbulence and debris settling affect deposition of (especially paint chip) debris on the straier surface. If turbulent forces are dominant, then head losses would remain low during the turbulent period and a large fraction of debris would still be suspended in the pool whea the turbulence ends. Head losses would then be expected to remain low if settling forces are dominant late in time, as a larger fraction of the paint chip debris would then be sulject to settling to the floor of the suppression pool. If strain suction forces are dominant, then head losses could increase steadily throughout both the turbulence and the settling phases in the pool as debris continues to collect on the strainer.

Prior to the C tests, three possible scenarios were postulated:

1) Turbulcnce keeps all debris suspended in the pool, but does not impact the rate of debxis buildup on the strainer. In this case, one would expect a combination of fiber and paint debris to build up on the strainer during the blowdown phase of the accident. Following the end of blowdown, relatively rapid sedimentation of paint debris wouod occur such that the debris buildup onthe strainerwouldbe pimalyfiber. (thisisthe dcf;ult set ofassumptions used to develop prior estimates of debris buildup on the strainers.)
2) Turbulence not only keeps the debris suspended in the pool during the blawdown phase of the accident, but also impedes the deposition Qf (especially paint) debris on the strainer. In that case, there is no period of time during which significant paint debris is deposited, and the.

overall debris deposition on the strainer is very similr to what would be expected for fibrous debris only.

14 of 32

  • Analsits o TestrforImestgatfng the Effect of CoatingsDebris on ECCSStrainerPefrmaonefor Vennont Yankee Docuent No.XSW-98-01-NP, Revision 2
3) The level of turbulence in the pool following the blowdown phase is sufficiently high that there is a long-term impact on both paiit chip and fiber deposition on the strainer surface. In that case, the ultimate head loss is reduced because of the reduction in both paint chip and fibrous debris deposition.

A comparison of results at different pool turbulence levels and a comparison between the results of the C-1 series (fiber only) and the C-2 series (fiber with paint) determined which of the above scenarios is a closer description of reality.

The C tests were riu during the week of Februaty 16-20, 1998. The measured parameter was the head loss (i inches of water) as a flmcidon of time, approach velocity, and turbulence (chug) level. Key results from these tests are sumrarized in Table 7. The enainment values are estimates of the debris fraction initially suspended or entrained in the pool based on observation.

Table 7. C Series TcstResults Designator HeadLoss@tinie C.uglevel Flowvellevel Fiber Paint

(-_.

entrained entrained co AU_

An Clb-O.12r

_Med

i All U Cla-0.12 LO Hi AU Cld-0. 12 Cl-0.06-I i

a All Clb-0.06 Med LO AU Clc-0.06-O1z

_Zero*

Lo AlL C2-0.12 Hi-H An 8l/l C2b-0.12 Med I3 AU 10%/o C2c-0.06-lHz Hi LO AU 80%

C2b-0.06 Med to All

_10%

C2a-0.06 LO LO AU 1%

C2o-0.06-o1Z_

Zcro*

LO All 80%

  • The zero chugging-induced turbulence follows an initial period of bigh turbulence that was required to initially suspend fte debris.

3.1 QualitativeAnalysis of theCTestS Test C0 was a baseline test series vfithout ay debris in the pooL Data vas taken at three chugging levels and two flow velocities. These tests demonstrated that the chuggin turbulwce levels had a negligible impact on head loss across a clean strainer.

The C1 test series was performed with only fibrous debris in the pool. At the bigher strainer approach velocity of 0.12 Wlls, complete deposition on the strainer of all fibrous debris initially in the pool was observed at both the medium and low pool turbulence level. As would be expected in that case, the final measured head loss was approximately the same in both cases (independent i of 32

Analysis of restjbr.Invseigaftng Tihe ffea of CoatlngsDebris on ECCSStrainerPerfobmancfor Vermont DocmentNo.:1TS-9"1

-NP, Rvasion 2 of pool turbulence level). This indicates that for these flow versus pool turbulence conditions, the strainer suction forces were dominant relative to the pool turbulence forces for fibrous debris.

At the lower strainer approach velocity, both the medium and high pool turbulence levels were sufficient to keep all fibrous debris in suspension and prevent its deposition on the strainer. This indicates that for these flow versus pool turbulence conditions the pool turbulence forces dominated the strainer suction forces for fibrous debris, In the final fiber-only test run at this lower flow rate, debris was initially suspended in the pool through induced chugging tirbulence.

The piston chugging was then teriniated such that the only source of pool tulence was the recirculation of the flow through the strainer. At this very low turbulence level, fibrous debris was collected on the straimer. However, significant fiber sedimentation was also observed, thereby limiting the total quantity of debris coilected on the strainer and hence the fnal head loss value.

The C2 test series was perftbned wih both fiber and paint debris in the pool. At the higher strainer approach velocity, the degree of paint debris deposition on the strainer was a strong function of pool turbulence. The high pool turbulence level (Test CMc-.12) was sufficient to keep most of the paint debris suspended in the pool rather than settling to the pool foor, and this suspended paint debris was then readily deposited on the strainer along with the fibrous debris.

Thus, the strainer suction forces dominated the pool turbulence forces for both paint and fibrous debris under these conditions. At the medium pool turbulence level (Test O2bO.12), most of the paint debris settled to the pool floor, and little remained suspened where it could be ultimately depsoited on the strainer surface. Hence, the measured head loss in this case was only slightly higher than the corresponding result fom the fiber-only tests. This indicates the dominance of the settling forces for paint debris even at the medium pool turbulence level.

At -the lower strainer approach velocity, one observed similar results with respect to paint debris sedimentation. Thus, at -the high pool turbulence level (Test C2c-0.06-Mz), most of the paint debris did not settle to the pool floor, whereas at the medium pool turbulence level (Test C2b-0.06), signifcant sedimentation was observed. However, at this lower approach velocity, the pool turbulence in both cases was also sufficently high to keep all debris (both paint chip and fiber) in suspension and prevent its deposition on the strainer. Thus, as was the case in the fiber only tests conducted at this flow rate, no measurable head loss was observed. This indicates that for these conditions, the pool turbulence faces dominate the strainer suction fbrces for both debris types. At the low pool turbulenoe level (rest CMa-O.06), deposition of fiber on the strainer was observed. However, paint debris sedimentation was suffcienly complete such that no more than a negligible quantity of such debris was deposited on the strainer. This same effect was observed in Test C2c0.06-OI, wherein the debris was initially filay suspended through high chuggng turbulence, with the piston chuggg then terminated such that the cnly source of pool turbulence was the recirculation of the flow through the strainer. As in Test C2a-O.06, little or no paint debris deposition on the strainer was observed. Also as in the fiber-only test conducted in this manner, significant settling of even fibrous debris was obsserved. Thus, at the lower strainer approach velocity of 0.06 Ws, significant paint debris deposition on the strainer was not observed 16 of32

  • Analysis of Testfror Investigating the Effect of Coatngs Debris on ECCS Sbrainer Performwicefor Vermont Yarkee DoatwnentNo.fl 7T.98OI-NP, Revsion 2 under any pool turbulence conditions. Figures IA and 1B illustrate the critical debris deposition results from the C series tests.

Based on the results of these tests as summarized above, one can draw several qulitative conclusions relative to the expected behavior of the actual Vermont Yankee strainers during a postulated LOCA. Duing the initial stages ofa LOCA, either DBA or IBA, little debris will have built up on the strainr and the gaps between the stainer stacked disks will be open (not filled with debris). During that Initial time period, the flow velocity at the strainerfdibris surface is determined by the total strainer surface area and ranges from 0.02 W1s to 0.04 fWs as previously stated. This surface velocity is the important parameter for characterizing the relative importance of the strainer suction force relative to the force of the random bulk turbulence in the pool Visual interpretation of the test video obtained during the C-series testing demonstrated that the key factor in determining whether debriswas deposited on the straier was not whether the debris impacted the strainer surface, but whether the flow rate was sufficient to keep that. debris on the surface. At both approach velocities, debris was observed to continually impact the stainer surface. However, for certain combinations of approach velocity and pool turbulence, the random bulk fluid velocity was sufficiently high to reentrain the debris into the pool. Thus, the expected behavior of the actual Vermont Yarkee strainers prior to gap closure is best represented by the resuts of the testing done at the lower strainer approach velocity of 0.06 Ms. For that approach velocity, it was demonstrated that no significant deposition of paint debris on the strainer could occur. The paint debris either rapidly settled to the pool floor at low turbulence, or remained in suspension at medium to high turbulence In fact, at this approach velocity, even fibrous debris deposition on the strainer was inhibited by the turbulence. Thus, during the early stages of a DBA LOCA and during all phases of an IBA LOCA (during which insufficient debris is generated to fRl the gaps), no paint debris deposition on the straie is expected.

During the later stages of aDB3ALOCA, afterthe Saps have ess all filled with debris, the flow velocity at the debris surface is determined by the ruscribed straini surface area and ranges from 0.06 fEs to 0.11 Pls as previously stated. Thus, fvr the CS stainer (0.06 fWts velocity), the conclusions drawn from the low velocity tests sunmnarized above are still valid, even after gap closure For the PER strainer (0.11 fils velocity), the testing done at 0.12 fWts is most directly relevant Those results would suggest that paint debris suspended in the pool at the time of gap closure could subsequently be deposited on the strainer surface, causing a significant bead loss increase. However, the timing of debris. deposition on the strainers is such that pool turbulence is significantly dimi ed by the tine gap closure occurs.- The results of the C-2 tests done at

rnedmnto low pool Utrbulence demonstrate that rapid settling of the paint debris is then expected to occur. Thus, a negigible quantity of paint debris remains suspended in the pool by the time gap closure occurs, and negligible paint debris deposition on the strainer is expected.

These qualitative arguments on paint debris behavior suggest that paint debris does not impact Vermont Yankee straier head loss under any relevant DBA or IBA LOCA conditions. These arguments are based on best-estimate anticipated debris quantities, pump flows, and strainer sizes.

However, these arguments should be revisited as part of the final Vermorn Yankee strainer head loss performance assessment to confirm that the prelimzinary conclusions reached herein are valid.

17 of 32

Analysis of TesiforInvesdgating thd Effect of Congs Debris on ECCSStrainer Performancefor Vermont Bandwe DocntNot.:

Rvf-98-01-.P, Revsion 2 C2a - lo chug lo v gTCRErRMaMraCM i

Figure lA. Head loss results for test Ca-0.06

<PROPRIETARY INFOPOMATION REMOVED>

C2b - med chug lo v id l OsRe ToRYReRsUtCNb OVED0 Figure 1 B. Head loss results for test CMb0.06 18of32

II AnatySls of Tes:forInvestigatlng he Effect of Coatinp Debrls on ECCSSraoner.PeofonnmnceJfor Vennont Yankee DoanentNo.dlTSr7 01-NP Revision 2 3.2 Quantitative Analysis for the C Tests 3.2.1 C tests - Density Estimate for fiber bed in the C Tests

  • PROPRIETARY INFOIMATIONREMOVED>

i 19 of 32

  • 8 Analysis of Testsfornvesffgaulngihi Effectof CoatingsDebrison ECCSSraner PerformanceforVennonr Yan~cee floament~o.:IT&9$W-OM-P, Bevislon 2 i

TROPMTARYYINFORMATION REMOVED>

2Oct32

Analysis of Testiforlnvestigafing thi Effect of Coatfn Deb&i on ECCS Strainer Performance for Verm ont rankee DocumentMO.:lT-98-01-NP, Revislon 2 3.2.2 FiberDiameter Estimate for the C tests I

<PROPRIETARY lNORMATION KEMOVED>

Table 8. Comparison oftest results and ELOSS calculions for different fiber diameters (The debris denrt used is 2.1 WbYf)

<PROPR1ETARY INFORMATION REMOVED>

3.2.3 Debris Deposition Analysis using the ARL C tests

<PROPRIETARY 1NFORMATION REMBVED>

3.2.3.1 Test C2a-.06 (C2a)- Low flow and the lowest chugging energy. Fiber plus paint, but very litle paint on strainer.

<PROPRIETARY lNFORMATION REMOVED>

3.2.3.2 Test C1c-.06-0fz (Cic) -Low flow and zero chugging. Fiber debris.

<PROPRIETARY INFORMATION REMOVED>

21 of32 I

  • Analsis of2'estsforlm'stigatng td Effect ofCoahngs Debris on ECCS Strainer Performancefor Vermont Yankee Docnmnt)o.aS'VY-9&01-NP, Revsion 2 3.2.3.3 Test C2c-.06-O"Uz (C2f)-Low flow and zero chugging. Fiber debris plus paint but very little paint on strainer.

I

  • EQROPRIETARY INFORMATION REMOVED>

22af32

Analysis of Tests forInvesfigating the Effet of Coatings Debris on ECCSStraner Perfobmwre for Vennont Ycmree DocmetNo.:;/Vt-9

-NP. RevWon 2 3.2.4 Comparison to Blockage Calculations I

<PROPRIETARY lNFORM:AkIONREMfOVBD>

i Table 9. Blockage Ieslts vs. C Series Test Values

<:PROPERITAPY IFORMATIONEREMOVED>

23 d32

Analysis of Tesfrforlfigih eat of CoatingsDebrfr on ECCS Straln rPerfomocs for Vermont Analyio Sestsforlavesfiga( IkeEffect of Coatrags~ebnr on ECC4StrafnerPerferscwjor Vermont Yankee DocwnentNo..IT T-98-01-NP, RevIsion 2

<PROPRETARY INFORMATION REMOVED>

24 of 32

Analysis of Testsforlnvestigactng Ox Effect of CoatingsDebris on ECCS StrainerPerformnance for Vermont Yankee DocurmentrA.l11A-98-C1-NP, Revision 2 4.0 Summary and Conclusions Tests to evaluate generic strainer performance under the specified Vermont Yankee conditions were performed at Alden Research labs in Holden, Massachusetts [ohnson, 1998]. The purpose of the testing was to investigate the effect of paint chips and fiber debris on the performance of new ECCS suction strainers to be installed at the Vermont Yankee plant. Two separate sets of tests described as L (Poop facility) tests and C (Chugging facility o: pool facility) tests were planned and exected. Both test series represented previously untested conditions unique to the Vermont Yankee ECCS performance expectations under accident conditions. The results of these tests can be used to apply the ELOSS code to Vermont Yankee strainer designtevaluation efforts.

The L tests were performed in the "Head Loss Loop Facility", wimch consists of a 100 gallon flow loop with a dat disk strainer of approidmately 0.8 f in area. The purpose of the testing was to investigate the effect of VY specific debris on a generic strainer in a loop geometry. Plow through the loop can be closely controlled via variable speed pumps so that values of head loss versus (steady-state) velocity can-be generated for a known debris bed on the strainer. Values of head loss at representative VY flaw velocities were generated for a range of debris quantities/composition.

All but one of the L tests formed thick 035 lb. fiber beds on the 0.8 fe flat plate strainer. Visual observations indicated non-unfofrm loading across the strainer due to wal effects and a dominant settling velocity (vs. flow velocity) force with debris thicknesses estimated at 2-6 inches and slightly different loading behavior from test to test. These thicknesses can be estimated/verified perhaps to within 0.5 inch using the filmed results of the tests [ARL, 1998]. The debris was thickest in the center of the strainer and thinnest at the edge position with the difference being about a factor of two. Turning off the flow resulted in expansionfof ote fiber debris bed by as much as 1 or 2 inches. The thickness was measured during the IA test at 3-3.5 inches at the edge and 5-5.5 inches in the center of the strainer. All of the debris was estimated to be on the strainer at that time and using a density of 1.5 lbt would yield an average tickness on the strainer of about 3.7 inches. The approach velocity was 0.02 Nt/s for these initial bed thickness measurements. For 100e deposition ofthe 0.35 lb. fiber mass at a reference density of 2.77 lb/l 3, the fiber would form. an average fiber bed of 2 inch. Accordingy the apparent density could range from 1.8 Ib/ft3 to 0.9 lb/fIV depending on the flow velocity, location on the strainer, and time of debris bed measurement. Based on an average fiber debris depth of 3.7 inches, a nominal L test density for head loss estimates would be 1.5 l/

-with an uncertainty factor of 30%.

Pretest predictions for the L tests were performed using a reference density of 2.77 lb/ft3 and an effective fiber density of 7.1 micron based on a weight percent averaging of the manufacturer as-fabricated data. These values produced estimated head losses of approximately 6 inches (vs. 1.5 irces actually measured) at 65F and a flow velocity of 0.04 ftIs. Both the effective fiber diameter (which is used to estimate the average surface-to-volume ratio for the entire debris bed) and the debris bed average density are dependent on specific debris loading conditions and dierent input values for both values are indicated in order to analytically match the observed results. Based on an observed average fiber debris depth of 3.1 inches, the L test density for head loss estimates 25 of 32

  • Analysis ofTesssforlnvestianing th Effectof Coaoings~ebr&s on ECCS Sbriner Perfoamne for Vermont Yan2ee DocwnentNo..:4Y-98-O1 -AP, Revision 2 would more appropriately be 1.5 lb/ft. Head loss measurements from the L tests can then be conservatively matched using a density of 1.5 WbeI3 and an effective fiber diameter of 10.7 pm.

The C tests were performed in the (modified) "Chugging Facility." This facility, which is a scale model of a section of the B)WR MIrk I torus, uses pistons to simulate the turbulence induced by the dowxnconer Previously, tHis facility had been used to investigate debris sedimentation under various levels of turbulence. A modification to this facility added a small (2.8 f) cylindrical strainer in the pool along with the associated piping and a variable speed pump. This facility was then used to investigate the effect of varying degrees of pool turbulence on the rate of debris buildup on the strainer (as ificed from the timedependent head loss across the strainer). This was done to simulate debris removal behavior from a pool for the very low strainer approach velocities (0.02 fXts on average for the CS system and 0:04 fiWs on average for the RHR systern) representative of the Vermont Yankee strainers.

The CI test series was performed with fiber debris in the pool. Fiber debris was suctioned from the pool completely in tests Cla-0.12, Cld-0.12, and Clb-0.12r. over a period of thirty minutes and deposited on the strainer this indicated that the strainer suction forces were dominant at the higher approach velocity of 0.12 *ts regardless of turbulence level Fiber debris remained suspended in the pool indefinitely in tests Clc-0.06-lHz and Clb-0.06, which were performed at the expected VYDBA flowrate. These tests showed that turbulent forces dominated the strainer suction force at the HI and MED turbulence levels for the lower approach velocity of 0.06 fls Fiber debris was partially suctioned from the pool in the Clc-0.06-OUz test over a 60 minute period and deposited on the strainer, indicating that settling forces were a factor which limit fiber debris deposition on the strainerunderthe condition of recirculation only.

The C2 test series was performed with both Sb and paint debris in the pool. When paint was added to the debris nfucture, both paint and fiber could orgy be auctioned from the pool and deposited on the strainer under the higher approach velocity and the HI turbulence level as indicated by test C2c-0.12. Fiber with small amounts of paint were suctioned from the pool and deposited an the strainer under MED turbulence levels in the C2b-0.12 test performed at the higher flow velocity which was very sinilar to the fiber only result seen in Clb-0.12r. For the postulated VY DBA approach velocity of 0.06 NIo, no fiber or paint was collected on the strainer as indicated by the C2b-0.06 and C2c-0.06-11z tests. Under post DBAJIBA conditions of recirculation flow, only fiber was collected as indicated by the C2c-0.06-Oz test. Comparison of the data taken at the 0.06 fs approach velocity (C2a-0.06, C2b0.06 and C2c-0.06 tests) clearly indicate that neither fiber or paint can be suctioned firo the pool and deposited on the strainer during for tubulent conditions in excess of tulence driven by recirculation flow alone at the initial Vermont Yankee approach velocities of 0.02 Rls and 0.04 fIs. For conditions where pool turbulence is driven by recirculation flow, only fiber can be suctioned from the pool and deposited on the strainer and only a fraction of the fiber is deposited with the remainder settling to the bottom of the pool.

The qualitative results from the C-series tests concerning paint debris deposition on the strainer suggest that paint debris does not impact Vermont Yankee strainer head loss under any relevant DBA orIBALOCA conditions. At pool turbulence levels that are sufficient to keep paint debris suspended in the pool (rather than settling to the pool floor), the turbulence is also sufficient to 26cf32

Anay5ris of ZestsforInvestigating the Effec of CoatingsDebris on ECCS StrcnerPerformancefor Vermont Yankae DocumentNo.JMSIT-98-01-NP, Revision 2 prevent deposition of the paint debris on the strainers. Once turbulence levels are reduced so as to allow the deposition of debris on the strainers, the tuVhulence is no longer sumicient to prevent rapid settling of the paint debris. In no case was it possible to observe a measurable impact on strainer head loss due to coatings debris.

For the C tests, depending on the flow velocity and time of debris bed measurement, the density of the debris bed could range from 1.4 lb/A3 to 4 lb/e. Based on an observed average fiber debris thickness of 2 inches from test Cla at the 0.12 and 0.06 fils approach velocities, a -nominal C test density for head loss estimates would be 2.1 lb/f 3 with an uncertainty :factor of 50°/0 due to non-uniform deposition and the uncertainty il thefickness measurement Head loss measurements from the C tests which suctioned fiber from the pool and deposited it on the strainer were lowerthan would be predicted using ertrapolated values from the high velocity database found in NUREG/CR-6224. Pretest predictions for the C tests were performed using a density of 1.5 l/t and an effective fiber diameter of 10.7 pm based on results of the L series tests. These values produced estimated head losses of approximately 5 inches (vs. 10 inches actually measured) for 10OW1, debris deposition. Both the effective fiber diameter (which is used to estimate the average surface-to-volume ratio for the entire debris bed) and the debris bed average density are dependent on specific debns loading conditions and different input values for both values am indicated in order to analytically match the observed results. Head loss measurements from the C tests can be reasonably matched using the observed average density of 2.1 lb/f and an effective fiber diameter of 8.3 pm.

Comparison of the BLOCKAGE calculations to the settling results seen in the C tests 156 W pool, 75 gpm flow, 035 ft3 fiber] indicates agreement somewhere between the tau--1 case where quiescent pool debris settling velocities are used and the tau=0.5 case where the settling velocity for the debris is one half the quiescent pool settling velocity. The tau-case indicates approximately 75% fiber deposition on the strainer in 60 minutes and the tau=0.5 case indicates appro ately 80% deposition in 60 mirnutes Vs. a 7M% removial in 50-60 minutes for the C tests.

The C tests indicated near complete settling in 40-50 nimites where BLOCKAGE indicates that about 1l0% of the fiber would still be in the pool at that time. This seems to indicate that the settling times are slightly faster and the setting fraction is sligly higher in the C cases than what would be predicted using the defaIt (IJR-EG/CI-6224) settling velocities found in BLOCKAGE Comparison of the results of the L series tests and the C series tests indicates that the average debris density was lower and the effective surface to volume ratio was higher in the L tests which were conducted on a fiat plate and ina flow loop system vs.. the C tests which were conducted using a clindrical strainer in a pool system Tests under both conditions produced head loss values lower than those, which would be extrapolated from the database, found in NUREG/CR-6224.

i 27 of 32

Analyvis of Testsfor.nvestigafing OVEffect c/oatof g, D6ebris on ECCS SfraWzeF'er/ormancefor mernont Ykee DomDum n:o.1 7 Y 01-NP, Revsion 2 5.0 References ARL: "Video tapes of the L and C Tests Performed at Alden Research Laboratory, Inc.",

January-Febnrary 1998.

Bettd EJ., Specificarfon for RM1 and CS Suction Strainers, VYS-049: Rev. 2, Change 1, Vermont YanlkeeElectric Company, November 12,1997.

BWROG, Utlity Resolution Guldance for ECCS Suction Strainer Blockage, Boiling Water Owners' Group, NBDO-32686, Rev. 0, November 1996.

Carbonmdum, Company, Fiberfiz Blanker and Mat Products - Product Specificatfons, The Carborundum Company, Fibers Division, Marcd, 1990.

Copus, E. R., "Test Plan for Tests For Investigating The Effect Of Coatings Debris On ECCS StrainerPerformance For Vermont Yankee", ITSC LetterReport, January 15, 1998.

Copus, E. R "VYI-SeriesPretest Calculations using LOSS", ITSC Letter Report, January 19, 1998.

Copus, E. R., VY C-Series Pretest Calculations using HLOSS", ITSC Letter Report, February 5, 1998.

Johnson, A. B. and Padmanaban, M., "Head Loss of Fibrous Insulafon Debris, Paint Chips, Sludge nd Turbulence Effectsfor Vermoni Ykve BMR Suction Strainers, AIL 65-98/M208F, April 1998 Mast, PK and Souto, F.J., LOSS 1.0: A Code for the Prediction of ECCSStainer Head Loss, Innovative Technology Solutions Corporation, }IS/ITS-97-01, Rev. 0, My 15, 1997.

Shaffer, C.J., W. Bernabl, J. Brideau and D.V. Rao, BLOCKAGE 2.5 Reference Ma l, NUREGICR-6371, U.S. NudlearRegulatory Commission, December 1996.

Pipp, R., Test Plan for the Vermont Ymankee ECCS Suction Strainer Design, Chugging Tests, C senes, Duke Engineering & Services, TR-A34600-05, January 1998.

Ripp, R, Test Proceedure for the Vermont Yankee ECCS Suction Strainer Design, Chugging ests, C series, Duk Eingineering & Services, TR-A34600-07, February 1998.

Souto, F. J. & Rao, D. V., Exerimentza Investigation of Sedimenwtaton of LOCA-Generated Fibrous Debris and Sudge in BW? Suppression Pools, NURECICR-6368, US. Nuclear Regilatory Commission, December 1995.

28cf32

Ana4zsls of TestsforInvesfigating the Effect of Coatings Debris on ECCSSbrranerPerformancefor Vermont rne Document No.dl=S 0J-NP, Revision 2 Zigler, G.,. Brideau, D.V. Rao, C. Shaffer, F. Souto, and W. Thomas, Parametric Study of the Potentalfor BWR ECCS Strainer Blokage Due to LOCA Generated Debris; NUREG/CR-6224, U.S. NuclearRegulatory Commission, October 1995.

29of32 i

Analysis of Teslfor Invesfigating the.Fffect of Coatings Debris on ECCS StrainerPefonmanceforYermont Yankee DocumentNo..:l SIVY-98O1-NP, Revson 2 Attachment A The following is an ELOSS Output File used in the analyses to estimate the head loss across the cylindzical strainer under the conditions ofthe Cla-.12 test.

30 f32

  • . Analysis of Testsfor nvestgating thzeEffectofCoatingsDebrisonECCSStraInerPerfonnanceefor Vermont Yankee DocumentNo.:fSV7-98.01-NP, Revision 2 I

<PROPRIETARY FO1PMEJSION1.BMOVB>

i 31of32

l.

Analyss of 7estsfor !vesrlgatlng thfffectof rOagsDebrUs onECCS&PoJnerPeiformancefor Vennont Y07*ee DocumentNo.:lWTA-9801-XNP, Revislon 2

<PROPRIETARY NFORMATION REMOVED>

I I

I I

I I

32 Of32.

Retrieved from "https://kanterella.com/w/index.php?title=ML050750145&oldid=5386915"