ML20217K744

Rev 1 to QDC-0010-M-0396, Quad Cities Station Unit 2:ECCS Strainer Head Loss Estimates

ML20217K744
Person / Time
Site:	Quad Cities
Issue date:	04/23/1999
From:	Choromokos R COMMONWEALTH EDISON CO.
To:
Shared Package
ML20217K705	List: ML20217K701 ML20217K744 ML20217K764
References
QDC-0010-M-0396, QDC-0010-M-0396-R01, NUDOCS 9910260230
Download: ML20217K744 (58)

Text

{{#Wiki_filter:NEP-12-02 t Revision 7 CALCULATION TITLE PAGE I Calculation No: ODC-0010-M-0396 DESCRIPTION CODE:M03 DISCIPLINE CODE: M STATION / UNIT: Quad Cities Unit 2 SYSTEM CODE: RH. PC. CS TITLE: Ouad Cities Station Unit 2: ECCS Strainer Head Loss Estimates X Safety Related Augmented Quality Non-Safety Related REFERENCE NUMBERS Type Number Type Number PROJ VO1500 For additional References see Section 5. COMPONENT EPN: DOCUMENT NUMBERS: EPN Compt Type Joe Type / Subtype Document Number 2-1600-4 F10 Corr /NDIT D104-00014 2-1600-8 F10 Corr /NDIT ODC-97-052 2 1600-12 F10 Corr /NDLT ODC-97-034 2-1600-16 F10 REMARKS: i REV. RE>ISING APPROVED DATE ORGANIZATION PRINT / SIGN 0 DE&S (D104) Rob Choromokos / Signature on File 9/1/97 l 1 DE&S (D104) 7,b 4,y /[j_ [) ef/g3/ ply em l NN PPtREh CE wy L y 774 2gggg gg;g;g, P PDR

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION REVISION PAGE CALCULATION NO. QDC-0010-M-0396 PAGE NO.: 2 REVISION SUMMARIES REV:0 REVISION

SUMMARY

Initial Issue Electronic Calculation Data Files: (Program Name, Vmion, File name ext /siz:/date/hourh min) 9 Prepared by-Francisco J. Souto / Signature on File 8/28/97 Print / Sign Date Reviewed by: Jan Bostelman / Sienature on File 8/28/97 Print / Sign Date Type of Review [ x ) Detailed [ ] Altemate [ ] Test Supplemental Review Reauired I 1 Yes (NEP-12-05 documentation attached) I1No Supervisor / (Comed USE ONLY) DO ANY ASSUMPI1ONSIN THIS CALCULATION REQUIREIATER VERIHCATION [] YES [x]NO Tracked by: (NTS #, ER #, etc.) REV: 1 REVISION

SUMMARY

Calculation is reformatted per the requirements of NEP-12-02, Rev. 7. Sections 2,5,6 and 7 are modified to account for 2.0 and 2.5 mil stainless steel foils. Assumptions 3.7 and 3.8 are changed. Ref. 5.25 is added. No other debris types were evaluated in this revision Electronic Calculation Data Files: (Program Name, Version. File name ext / size /date/ hour /: min) P upared by: Gil Zieler / .u 4/23/99 ' Print / Sign / Date Reviewed by: Jan Bostelman tbab 4/23/99 P i'ni/ Sign Date f Type of Review /' [ x ] Detailed [ ] Altemate [ ] Test Supplemental Review Required I 1 Yes (NEP 12-05 documentation attached) I1No Supervisor / (Comed USE ONLY) DO ANY ASSUMPI1ONSIN THIS CALCULAT10N REQUIRE IATER('E_IGTAllON [] YES [x) NO Tracked by: (NTS #, ER #, etc.) .,J t ~

• w-=.

..w...,m .w._,,,

i l NEP-1242 j l Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION TABLE OF CONTENTS l CALCULATION NO. ODC-0010-M-0396 REV. NO.1 PAGE NO. 3 SECTION PAGE NO. SUB-PAGE NO. TITLE PAGE l REVISION

SUMMARY

2 i TABLE OF CONTENTS 3 l PURPOSE / OBJECTIVE 4 METHODOLOGY AND ACCEPTANCE CRJTERIA 4 ASSUMPTIONS 11 DESIGN INPUT 12 REFERENCES 12 CALCULATIONS 14

SUMMARY

AND CONCLUSIONS 28 5 l r [ - -... +

i l NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M 0396 PAGE NO. 4 l 1.0 PURPOSE / OBJECTIVE The purpose of this calculation note is to estimate the head loss across the debris bed formed on the propoJ,ed strainers at the suction of the emergency core cooling system (ECCS) pumps at the Quad Cities Station Unit 2 (QC2), due to accumulation of insulation debris (fibrous, calcium silicate and reflective metallic) and paniculate matter produced as a result of a loss of coolant accident (LOCA). The level of margm in the estimatica of the head loss is assessed. Additionally. a !!mited parametric analysis is performed on key variables affecting head loss estimates. The heal loss estimates reported harein do not include the head loss associated with the clean strainer, 2.0. METHODOLOGY AND ACCEPTANCE CRITERIA 2.1 Methodology J-The methods used for the estimation of the head loss across the proposed strainers at the suction of the ECCS of QC2 are based on the NUREGICR-6224, Parame:ric Study of the Potentialfor BWR ECCS Strainer Blockage due to LOCA Generated Debris (Ref. 5.24) 2.nd the SER for NEDO 32686 "URG for ECCS Suction Strainer Blockage," (Ref. 5.6). The NUREG/CR4224 fundamental models were implemented by the U.S. Nuclear Regulatory Commission in the BL,0CKAGE 2.5 computer code ' (Ref. 5.20 and 5.22). This section summarizes the methods used in this calculation note. Further details on the SER, the NUREG/CR-6224, and the BLOCKAGE 2.5 methods and models, as well as their corresponding ranges of applicability, are presented in the original references (Ref. 5.16), (Ref. 5.24), (Ref. 5.20 and 5.22), and (Ref. 5.23), respectively. In particular, the BLOCKAGE 2.5 computer code will be used to estimate the head loss due to fibrous and particulate matter debris, whereas the URG methodology used to estimate the head loss due to Ref!ective Metallic Inculation (RMI) debris is based ou the methodology proposed in the SER for the BWROG-URG. 2.1.1 BLOCKAGE 2.5 Computer Code The BLOCKAGE 2.5 code was developed by Science and Engineering Associates, Inc. for the U.S. Nuclear Regulatory Commission (NRC) as a toal to evaluate licensee compliance regarding the design of ECCS suction strainers as required by NRC Bulletin 96-03, Potential Plugging ofEmergency Core Cooling Suction Strainers by Debris in Dolling Water Reactors. As stated in its teference manual (Ref. 5.22), the BLOCKAGE code was developed to predict whether or not accumulation of debris on the ECCS su: tion strainers following a postulated LOCA would lead to loss of net positive suction head (NPSH) in a Boiling Water Reactor (BWR), BLOCKAGE 2.5 allows the user to simulate debris generation and subsequent transport of different types of debris including fibers, particles and metals, using user r.pecified debris generation and transport factors. The transport of debris from the drywell to the suppression pool can be simulcted as location-dependent and time-dependent. Alternatively, BLOCKAGE 2.5 allows the user to transport ins'.antaneously to the suppression pool a specified quantity of debris. In the pool, BLOCKAGE 2.5 allows the possibility of modeling the transport of REVISION NO. l 0 l CM D C Cpp r LW ~~ \\ e w.r

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0016 M-03% PAGE NO. 5 debris to the strainers by considering the specific sedimentation characteristics for each debris type and size group. Four user-optional head loss correlations, including the NUREG/CR-6224 head loss correlation for a mixture of fibrous and particulate matter debris, are available in BLOCKAGE 2.5 to estimate the pressure drop across the debris bed on the strainers. To estimate the quantity of particulate matter debris retained in the fibrous dabris bed, BLOCKAGE 2.5 allows the user to provide the filtration efficiency for each type of debris, 2.1.2 Had Less due to Fibrous Insulation Debris According to the URG (Ref. 5.6, p.14), the head loss from fibrous debris, in combination with suppression pool sludge, is expected to dominate the strainer head loss. Consequently, considerable attention was given to the available head loss correlations for fibrous insulation debris. BLOCKAGE 2.5 contains four optional correlations that may be selected by the user to model the head loss across a debris bed consisting of fibrous and particulate debris. The four correlations are the NUREG/CR-6224 correlation, the empirical correlation developed by the BWROG in 1994 (Ref. 5.5) (not to be confused with the URG head loss correlation for fibrous debris), and two generic correlations that can be used to implement an alternate user correlation. Based on the assessment ' documented in DE&S report TR-ECCS-GEN-03 (Ref. 5.23), it was determined that the fibrous debris head loss correlation based on the NUREG/CR4224 study was the most applicable and defensible for the QC2 analysis. Hence, it forms the basis for these calculations. The NUREG/CR-6224 head loss correlation, as descrii>ed in Appendix B to NUREG/CR-6224 (Ref. 5.24), is a semi-theoretical head loss model. It is based on theoretical and experimental research for the pressure drop across a variety of fibrous and porous media carried out since the 1940s. 'Ihis head loss model has been implemented in BLOCKAGE 2.5, as described in the corrasponding reference manual (Ref. 5.22). Its results have been extensively validated for debris beds composed of fiberglass and simulated suppression pool sludge (Ref. 5.21), as well as mineral wool fibrous materials (Ref. 5.15) for flat strainers. The NUREG/CR4224 head loss correlation was also shown to produce conservative results for the stacked disk strainers tested at EPRI when the amount of fibrous debris did not exceed 10 ft'(Ref ~.I 23). I i p,', ;. ~~l REVISION NO. l 0 l 1 l l 2_.,.-..-.....-.--..- %~

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-03% PAGE NO. 6 l 2.1.3 Head Loss Correlation due to Reflective Metallic Insulation Debris The type of foil of the originally installed Reflective Metallic Insulation (RMI) (Ref. 5.2) at the Quad Cities Station is 6 mil Aluminum. In the last few years, the foil type in replacement RMI cassettes has been either ) 2 mil or 2.5 mil stainless steel. In order to provide an estunate of the differences between two types of RMI, this calculation will present results for both 2/2.5 mil stainless stect and 6 mil aluminum foils. The BWROG study (Ref. 5.6) provides an empirical correlation to estimate the head loss due to different types of RMI debris for BWR ECCS suction stramers. However, while these efforts provided some valuable insights into differences between the different types of RMI, the NRC's SER (Ref. 5.16) l concluded that the resulting co: relation could not be d=anetrated to be conservative under all conditions. The NRC instead presented an alternate correlation which forms the basis for the results presented herein. i URG-SER Head Loss Correlationfor RMIDebris l The SER of the URG presents the following correlation (Equation K.Sa in the SER (Ref. 5.16)) that is stated to adequately bound the test data from the ARL (Ref. 5.25) and URG (Ref. 5.6) RMI tests: AH = 0.108U' A" (1) A,

where, M

is the head loss (ft-water), U is the approach velocity (ft/s) based on the available strainer area, l A is the RMI foil surface area (ft'), and y 1 2 A, is the available area of the strainer (ft ) which is taken as the circumscribed area of the outer cylindrical strainer shape. l l REVISION NO. l 0 l 1 l l - ~ ~ l

i I l NEP-12-02 ) Revision 7 l, l COMMONWEALTH EDISON COMPANY i CALCULATION NO. QDC-0010 M-0396 PAGE NO. 7 i I This equation is derived based on the head loss tests conducted by the NRC at the ARL test loop facility, using debris generated by the NRC RMI debris generation test (Ref. 5.25). The NRC debris generation RMI test was a steam test using a 2.5 mil Stainless Steel foil RMI Diamond Power cassette mounted on a circumferential weld break simulator. The SER also concluded that this correlation adequately predicted experimental data reported in the URG for gravity head loss tests using debris from the NRC RMI debris generation test, as well as tests conducted using 2.5 mil Stainless Steel debris manually generated by CDI. This correlation was also adopted to estimate head losses due to 2 mil Stainless Steel RMI debris. The % mil thickness difference between the two types of Stainless Steel RMI is not expected to cause measurabN differences in head loss. Both types of foil are expected to form very similar debris beds given the anticipated minimal variation in the strength of the crumbled debris pieces. This correlation is also assumed to bound head loss estimates if the RMI debris comes from 6 mil Aluminum instead of 2.5 mil Stainless Steel. The SER suggests that the smaller sized RMI debris would form beds with lower void fractions than larger sized RMI debris. The URG RMI debris generation tests showed that the 6 mil Aluminum RMI debris pieces were much larger than the debris pieces generated from the NRC 2.5 mil Stainless Steel. As such, a 6 mil Aluminum RMI debris bed will have larger void fractions than a 2/2.5 mil Stainless Steel RMI debris bed. Therefore, for the same foil area, the head losses of a 6 mil Aluminum RMI debris bed will be lower than a 2/2.5 mil i Stainless Steel RMI debris bed. The effect of larger pieces generating lower head losses than smaller i pieces in the flow velocity regime of the Quad Cities replacement strainers is clearly shown in the NRC sponsored RMI head loss tests [Ref. 5.25, Appendix D, Figure 3]. RMI Saturation Thickness Experimental evidence and theoretical reasoning suggest that RMI debris buildup on the strainer would l reach a saturation limit, beyond which local surface flow velocities would not induce sufficient drag to overcome forces imposed primarily by turbulence and gravity. 'Ihe URG experiments suggest that this i I limit is given when the local surface flow velocity is one half of the average terminal settling velocity of the RMI debris. l A spherictd RMI debris buildup model can be derived based on the simplified Figure 1 illustration. For a spherical RMI debris deposition on a stacked-disk strainer, the ratio of strainer approach velocity based on the circumscribed strainer area, U,, to the local flow velocity at the debris surface, U, may be approximated by 1 U, A 4x R' -Q (2) -=-= i U A, x L D, + x R,' + x (R,' - R,* ) l l l FOR REFERENCE REVISION NO. l 0 l 1 l f) M ?lir

1 NEP 12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 8 where (see Figure 1): A is the surface area of the RMI spheroid debris bed (ft'), - A,is the circumscribed area of the strainer (ft'), R is the radius of the RMI spheroid debris bed (ft), L is the strainer active length (ft), D, is the strainer outer diameter (ft), R,is the outlet pipe radius (ft), and Dis the area of spherical segment associated with the interference between the RMI debris bM and the outlet pipe (ft'). The radius of the RMI debris spheroid as a function of the average local flow velocity at the debris surface is then approximated by: R =)4 U* h D, + 2 R,2 -R,2 )+k (3) x i " ' r,? T. : 7. i' ::: ' } "~ REVISION NO. l 0 l 1 l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY ~ CALCULATION NO. QDC-0010-M-03% PAGE NO. 9 '~,s' s, R, N / N /n \\ / 's n ,I L '\\ l \\ f i n D. i Ri l I \\ / 1 \\ / \\ / 'N V / '\\, L E / N, ',~, ,'~~........... 'p* Figure 1. Schematics of a spheroid RMI debris bed on a strainer. Since the local flow velocity at saturation conditions is approximately % of the average settling velocity of the RMI debris, U,,,, the saturation bed U corresponding to a radius R can be approximated by: U (at R = R,) = U, = U"' (4) 2 Hence, the equivalent volume of RMI debris required to produce saturation conditions, %, may be estimated by: V, = x R,' - x R,* L - x R,' (R, Lfy (5) The corresponding RMI debris foil area, A, is then given by: V"^ Ag = K, -- (6) 1 ~ REVISION NO. l 0 l 1 l l ..f

l NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO.10 where K,(in ft) is the thickness constant for RMI debris. Based on experiments reported in the URG, K, is equal to 0.014 ft for 2.5 mil stainless steel debris whereas for 6 mil aluminum K, is equal { l 0.073 ft (Ref. 5.6). The K, value of 0.014 ft will also be used for the 2 mil stainless steel. The above methodology can be applied to Quad Cities Station as follows:

1. Determine the saturated bed thickness for a 6 mil aluminum RMI debris bed using equations 2 through 6.

2. Determine the head loss for a 6 mil aluminum saturated debris bed using equation 1.

3. Determine the saturated bed thickness for a 2/2.5 mil stainless steel RMI debris bed using equations 2 through 6.

4. Determine the head loss for a 2/2.5 mil stainless steel saturated debris bed using equation 1.

l 2.1.4 Head IAss due to Other Insulation Debris: Calcium Silicate Destruction of insulation based on calcium silicate (Cal-Sil) will produce particulate matter debris. Without enough fibrous materials to cover the strainer holes, calcium silicate debris (as is the case for any type of particulate matter) will not produce a significant head loss due to insufficient filtration. In the presence of enough fibrous debris to cover the strainer holes, the contribution of calcium silicate debris to the head loss will be modeled by BLOCKAGE 2.5 by including this debris as particulate matter in the NUREG/CR-6224 correlation. 2.1.5 BIDCKAGE 2.5 Verification and Validation BLOCKAGE 2.5 has been subjected to rigorous coding verification by its developers to ensure that the code performs as it was designed to perform, and extensive quality assurance (QA) was integrated into the development of the BLOCKAGE 2.5 code (Ref. 5.22). Based en this information, BLOCKAGE 2.5 is an approved code by DE&S (Ref. 5.8). 2.2 Acceptance Criteria l There are no acceptance criteria for this calculation. The results presented herein will provide input to a subsequent NPSH margin calculation. FOR REFERENCE REVISION NO. l 0 l 1 l (*) M i Vl L 1

h NEP-12-02 I Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-0396 PAGE NO.11 3.0 ASSUMPTIONS 3.1: Due to the common ring header, the ECCS flow is assumed to be equally distributed omong the four i strainers. 3.2 - The debris bed is formed and distributed uniformly over the surface of the strainer. This assumption es conservative, becaase if the debris bed were non-uniform, the debris bed thickness would be non-l uniform, allowing for the possibility of having relatively " clean" regions on the strainer, and thus reducing the hul loss. 3.3 ~ The densities and characteristic dimension of the miscellaneous fibrous debris are considered to be similar to those of NUKON 3.4 he densities and characteristic dimension of each drywell particulate material (i.e., equivalent diameter for calcium silicate debris, ditt/ dust and sludge, and thickness for paint / coating chips and rust flakes) will be assumed based on generic data. When large uncertainties exist in the characteristic size of particulate materials, such as in the case of unqualified paint chips or rust flakes, the smallest reasonable value will be used for conservatism. 3.5 - In the base case calculations, it will be considered that 33% of the quantity of large particles (paint chips and rust flakes) initially in the pool is actually deposited within the fibrous debris bed. This is based on assuming a sedimentation velocity 10 times lower than the terminal settling velocity for paint chips and rust flakes. No credit is given to the deposition of fibers, calcium silicate debris, sludge and dirt / dust particles on the floor (mainly by gravitational i.:dimentation) or on other structures in the suppression pool. ^ 3.6 In the base case calculations, it will be considered that 50% of the quantity of small particles (sludge and dirt / dust) initially in the pool is actually deposited vithin the debris bed. This is based on the assumption that the filtration efficiency model developed from the U.S Nuclear Regulatory Commission (NRC) experiments on head lo!.s (Ref. 5.21) is valid for sludge and dirt / dust particles. 'For paint chips and rust flakes, a filtration ef ficiency of 1.0 will be assumed for debris bcd thicknesses higher than 0 25 inches, and 0.50 for debris bed thicknesses lower than 0.25 inches. 3.7 - 'This calculation assumes that the NRC SER RMI head loss correlation is applicable to the Quad Cities strainer and all RMI debris types expected. The SER RMI head loss correlation adequately predicted experimental data for tests conducted using 2.5 mil Stainless Steel debris. Itis reasonable to asstime that the 2 mil Stainless Steel debris would be similar in shape and size to the 2.5 mil Stainless Steel debris tested. ' Hence, the thickness parameter, K, settling velocity, and head losses are expected to be the same. The co.Telation should also bound the head losses from 6 mil aluminum RMI. The URG RMI debris characterization information clearly shows larger debris pieces and lower packing density for the 6 mil aluminum as compared to the 2.5 mil Stainless Steel debris. This higher void fraction for the aluminum RMI debris .would result in a lower head loss for the same foil a.rea. ~. "l-L REVISION NO. -l .0 l 1 l e .wa,,--s. I

w L'. NEP-12-02 ~ Revision 7 I r. COMMONWEALTH EDISON COMPANY yALCULATION NO. QDC-00155-0396 l PAGE NO.12 3.8 - This calculation assumes that the head loss of a debris bed composed of a mixture of RMI debris and fibrous debris is bounded by the sum of the head loss contribution of each type of debris. 4.0 DESIGN INPUTS 'Ihe design h:put information for this calculation is obtained from Refs. 5.6, 5.16, 5.20, 5.21, 5.22 and 5.24. ] 5.0 RFIERENCES 5.1 Aldinger, T.L, R.A. White, and R.A. Manley, Performance of Containment Coatings During a less of Coolant Accident, Rechtel Power Corporation,. November 10,1994. (in Volume III of the Utility Resolution Guidance (URG) for ECG Suction Strainer Blockage, BWROG, Ncvembar 1996). 1 5.2 ' Ashbaugh, S. (a), Quad Orles Station Unit 2: Es:Imat!an ofinsulation Debris Sourx for ECCS Strainer Head Ioss Calculations, Innovative Technology Solutions Calculation Note No. QDC-0010-M-0394 (ITS/CN-97-20), Rev. O, May 1997. 5.3.. Ashbaugh, S. (b), Quad Cities Station - Unit 2: Estimation of Non-Insulation Drywell Debris Sources for ECCS Strainer IIcad loss Calculadons, innovative Technology Solutions Calculation Note No. QDC 0010-M-0395, Rev. O, May 1997. 5.4 Blomquist, M. and M. Dellby, Report From Tests Concerning the Efea of Steam Jet on Caposil Insulation at Karlshamn, SDC-93-1174, June 1993. 5.5 BWROG, Interim Report of the BWR Owners' Group ECG Suction Committece, Boiling Water Owners' Group, December 1994. 5.6 BWROG, Utility Resolution Guidance for ECG Suction Strainer Blockage, Boiling Water Owners' Group, NEDO-32686, Rev. O, November 1996. 5.7 Diertl, R., R., Louderback, A. Kauftum, and A Bilanin, Testing of Alternate Stralners With lasulation Fiber and Other Debris Revision 2, C.D.I. Report No. 95-09, October 1996. On Volume 1 of Utility Resolution Guidance (URG) for ECCS Spction Straigr__Blockare, BWROG, October 1996). ~ 5.8 DEAS, Computer Program Verfication and Validation, Procedure No. DPR-3.5, Rev. 4, Duke Engineering and Services, Jancary 10,1997. 5.9 Fuchs, N., 7he Mechanics ofAerosols, Pergamon Press Ltd., New York,1964. 5.10 Hart, G., Summary report on Performance of Performanct Contacting, Inc.'s Sure-flow" Suction . Strainer with Various Mixes of Simulated Post-LOCA Debris, Revision 0, Performance Contracting Inc., February 14, 1997 REVISION NO. l 0 l 1 l l h m

NEP-12-02 ~ Revision 7 a COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010M-0396 PAGE NO.13 5.11 Heames, T., Limits of Applicabilityfor the URG RMI Head Loss Q>rrelation, Report No. TR-ECCS-GEN-04 (ITS/1TS-97-04), Revision 0, Duke Engineering and Services, Inc., July 18,1997. 5.12 Munchausen, J.H., Air Jet impact Testing of Fibrous and Rejectiw Metallic Inswalon, Revision A, Final Draft, C.D.I. Report No. 96-06, Continuum Dynamics, Inc., September 199( 5.13' NDIT No. D104-Q0014, Quad Cities Station Unit 1 and 2 Design information for NPSH Calculations, Nuclear Design Information Transmittal (NDIT), Commonwealth Edison Co., November 27,1996. (Source ofInformation:. Tech Specs. 3.7/4.7 A.1.a/b). 5.14 NDIT No. 97-052, ECG Suction Strainerfow rates andpool temperaturesfor DBA LOCA, Nuclear Design Information Transmittal (NDITA Commonwealth Edison Co., April 25, 1997 (Sources of Information: 1) Quad Cities Calculation No. QDC-1000-M-0291, Rev. O,2) Quad Cities Calculation No. QDC-1000-M 0292, Rev. O,3) Quad Cities NTS No. 254525%DRE134,4) Dresden Calculation i No. DRE97-0012, Rev. O, 5) General Electric Report No. GENE 437-022-0293, 6) Facsimile from K. Ramsden to J. Garrity dated 12/30/96). 5.15 NEA, Knowledge Base for Emergency Core Cooling System Rediculation Reliability, NEAICSNIfR (95)11, Prepared by the U.S. Nuclear regulatory Commission for the Nuclear Energy Agency (NEA) of the OECD, February 1996. j 5.16. NRC, Safety Evaluationfor'NEDO-32686, Rev. O, ' Utility Resolution Guidance Documentfor ECG - Suction Strainer Blockage, U.S. Naclear Regulatory Commission, August 20,199E. 5.17 OG94461-161, Letter from T.A. Green (BWROG) to A. Serkiz (USNRC), "BWR Owners' Group ECCS Strainer Committee Suppression Pool Sludge Particle Size Distribution", Dated September 13, 1994.'(In Volume III of Ulility Resolution Gulduce (URG) for ECCS Suction Strainer Bloc);ggg, BWROG, November 1996). 5.18 PC1, Quad Cities Unit-2. Sure-Flow Strainer, Diagram QCO2-SUMP-8002-1100, Rev. 5, l'erformance Contracting, Inc., November 27,1996. 5.19 Reist, P.C., Introduction to Aerosol Science, Macmillan Publishing Company,1984. ) 5.20 Rao, D.V., W. Bernahl, J. Brideau, C. Shaffer and F. Souto, BLOCKAGE 2.5 User's Manual, . NUREG/CR-6370,' U.S. Nuclear ReFulatory Commission, l>ecember 19%. 1 5.21 Rao, D.V. and Souto,' F., Experimental Study of Head Loss and Filtration for LOCA Debris, , NUREG/CR-6367, U.S. Nuclear Regulatory Commission, February 1906. i 5.22

Shafier, C.J.,

W. Bernahl, J. Brideau and D.V. Rao, BLOCKAGE 2.5 Reference Manual, ' NUREG/CR-6371, U.S. Nuclear Regulatory Commission, December 1996. I I REVISION NOc l 0 l 1 1 M c5 n t r r b r r. p e r-e va ss i u,. s c a s u O N LY

g NEP-12-02 Revision 7 1 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC4010 M43% PAGE NO.14 5.23 Souto, F., Range ofApplicabilityfor GilG and NUREGICR-6224 Head Loss Correlations, Report No. - TR-ECCS-GEN-03 (ITS/lTS-97-03), Revision 0, Duke Engineering and Services, Inc., July 14, 1997. 5.24 Zigler, G., J. Brideau, D.V. Rao, C. Shaffer, F. Souto, and W. Thomas, Parametric Study of the - Potentialfor BWR ECCS Straker Blockage Due to LOCA Generated Debris, NUREGICR-6224, U.S. Nuclear Regulatory Commission,03ober 1995. ] 5.25 ' Zigler, G., J.' Brideau, and B. Ziglet, Experimental Investigation of Head Loss and Sedimentation Garacteristics of Reflectin Metallic Insulation Debris, SEA No. 95-970-01-a:2, Science a*1d Engineering Associates, May 1996. 6.0 CALCULATIONS 6.1 l Technical Input This section describes the infonnation used to develop the BLOCKAGE 2.5 specific input data file for - QC2. Basically, this 'information consists of plant specific parameters, quantities and physical } characteristics for each type of debris, as well as the modeling considerations with respect to debris j sedimentation into the suppression pool floor and filtration of paniculate matter within the debris bed. 6.1.1. Strainer Data Table 1 presents the dimensions of each of the four stacked 4isk strainers installed at QC2. Table 1. Quad Cities Station Unit 2: Strainer Dimensions Length 42 inches (Ref. 5.18) Maximum Outside Diameter 45 inches (Ref. 5.18) Inside Core Tuh,,e_ Diameter 20 iraches (Ref. 5.18) Gap Diameter 24.5 inches (Ref. 5.18) Gap Width 2 inches (Ref. 5.18) Disk Width 2 inches (Ref. 5.18) Number of Disks 11 (Ref. 5.18) 2 Total Surface Area 207 ft (Ref. 5.2) Circumscribed Area 61ft (See Section 6.1)) 2 Interstitial Volume 1 13 ft' (See Section 6.1) - - ~.., _. ,b WW'h REVISION NO. l 0 l l' l 'l - - - = =. - + - - +,, -

l i l NEP-12-02 Revision 7 J COMMONWEALTH EDISON COMPANY CALCULATION NO, QDC-0010 M-0396 PAGE NO.15 I d i 6.1. 2 Flow Conditions The flow rate and suppression pool water temperature as a function of time considered in these head loss estimates are presented in Table 2 (Ref. 5.14)'. 1 Table 2.' Ouad Cities Station Unit 2: Flow Conditions (Ref. 5.14) Time Pool Water Temperature Total ECCS Flow Rate (s) FF) (gpm) 16 106 33400 { 31 117 33400 1 ~ ~ 59 129 33400 ~ 337 144 33400 600 149 _ 33400_ { 601 149 29000 1 1000 1542 29000 j L 10000 1762 29000 l 6.1.3 Debris Quantities 6.1.3J Fibrous Insubtion Debris As estimated in calculation note QDC-0010-M-0394 (Ref. 5.2), the worr,t case break location in the QC2 drywell generates and transports 6.74 ft' of NUKON* fibrous debris to the suppression pool. All fibrous debris quantities include 2 ft' of miscellaneous fibrous debris materiais transported to the suppression pool (Ref. 5.2). In addition,0.78 ft) calcium silicate insulation debris is generated with this maximum quantity of fibrous debris. This break also tr:msports 11773 ft of RMI foil debris to the 2 pool. 6.1.3.2 Reflective Metallic Insulation Debris Calculation note QDC-0010-M4394 presents an estimate of 19405 ft' for the largest quantity of RMI foil debris transported to the suppression pool from the drywell of the QC2. Note that all RMI foil debris estimates are increased by 10% over the calculated values to account for damage to the RMI cassettes (Ref. 5.2). ' The sources of information for each NDIT appear in the list of References in Section 5.0 2 These values nre estimated based on a plot provided in (Ref. 5.14) ~~~~7 REVISION NO. l 0 l 1 l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COlWPANY ~" CALCULATION NO. QDC-0010-M-0396 g ,_PAGg g g pw n :. t.w ~~ 0 T'i L.\\ r

• t s

6.1.3.3 Calcium Silicate Insulation Debris - Calculation note QDC-0010-M-0394 presents a conservative estimate of 0.78 ft' for the quantity of calcium silicate insulation debris transported to the suppression pool from the drywell of the QC2 for every analyzed break.- 6.1.3.4 Particulate Debris Calculation note QDC-0010-M-0395 (Ref. 5.3) estimates conservative quantities for particulate debris, composed of sludge and drywell particulate matter, in the QC2 suppression pool. The values are presented in Table 3. Table 3. Quantity of Particulate Debris in the Quad Cities Station Unit 2 Suppression Pool Debris Type Mass (Ib) Dirt / Dust 150 Rust Flakes 50 Qualified Paint or Other Surface Coating 85 Unqualified Paint or Other Surface Coating 85 Total Drywell Particulate Debris 370 Suppression Pool Sludge 443 6.1.4 Debris Characteristics The NUREG/CR-6224 head loss correlation considers each type of debris by specifying the fiber diameter, the as-fabricated (or macroscopic) and the material (or microscopic) Lbrous material densities, and the characteristic sizes and average microscopic densities of suppressi'en pool sludge and drywell particulate matter. The following paragraphs present the proposed debris characteristics in this calculation. The material (or microscopic) density of NUKON fiberglass insulation is 175 lb/ft' (2800 kg/m') and the macroscopic pack density of this material is 2.4 lb/ft' (38 kg/m') (Ref. 5.24). The SEM analysis of NUKON fiberglass debris (Ref. 5.21) shows that the diameter of the fibers is fairly uniform and approximately equa! to 7.1 m. The microscopic density of sludge, which is basically iron oxide, is 324 lb/ft' (5190 kg/m') (Ref. 5.24). The mass median diameter of the sludge particle size distribution is estimated to be 2.5 pm (Ref. 5.17). This value represents the size dictribution of the sludge in the suppression pool. However, the size distribution of the sludge particles actually deposited on the fibers in the debris bed has a mass median diameter much larger than the corresponding mass median diameter of the sludge particles in the suppression pool, as suggested by the SEM photographs of typical debris beds (Ref. REVISION NO. l 0 l 1 l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-0396 PAGE NO.17 5.21), which show particle sizes in the order of 100 pm. Consequently, in these calculations an average debris bed sludge particle size of 10 pm will conservatively be used. The microscopic density of calcium silicate is reported to be 143 lb/ft' (2300 kg/m') in European studies (Ref. 5.4). These studies also indicate that the majority of the calcium silicate debris is greater than 0.85 mm in diameter, with less than 1% smaller than 20 pm. In this calculation, a characteristic size of about 40 m (1.2x10".ft) will conservatively be used for a typical calcium silicate debris particle. As indicated in Appendix E to " Testing of Alternate Strainers With Insulation Fiber and Other Debris" (Ref. 5.7), the macroscopic density of calcium silicate is 16.3 lb/ft'. In the absence of more detailed information, a microscopic density of dirt / dust of 156 lb/ft' (2500 . kg/m') (Ref. 5.24) will be used. An average equivalent diameter of 10 m, based on a typical diameter of dust particles (Ref. 5.19), will be used in this calculation. In general, the following types of coatings are found inside the primary containment of BWR nuclear plants: inorganic Zinc, epoxy, and alkyd. The microscopic densities of these materials (based on the specific gravity values reported (Ref. 5.1)) are: 90 lb/ft' (1430 kg/m') for epoxy, 94 lb/ft' (1500 kg/m') ' for alkyd, and 156 lb/ft' (2500 kg/m') for inorganic Zinc. In the absence of specific details about the paint / coatings chips in QC2, an average microscopic density of 124 lb/ft' will be used in these calculations (Ref. 5.24). The thickness of the paint chips will be a function of the coating thickness in ) the drywell. A typical lower bound for such coatings is 1 mil. To account for the uncertainty in this value, particularly in the case 'of unqualified coatings, a characteristic size of 0.69 mil will conservatively be used in these calculations. Rust flakes will be considered as iron oxides, with a microscopic density of 324 lb/ft' (5190 kg/m'). Since rust flakes appear to be visually similar to paint chips, an equivalent diameter of 0.69 mil (17 pm) will conservatively be used for the characteristic size. The debris characteristics used in this calculation are summarized in Table 4. Table 4. Quad Cities Station Unit 2 Debris Characteristics Debris Type Microscopic Density Characteristic Size Ob/ft') (ft) Ium) Fibers 175 2.3x10-' [7.11 Calcium Silicate 143 1.2x 10" [36.61 Sludge - 324 3.3x 10-'[10] Drvwell Particles Dirt / Dust 156 3.3x10'[10] Rust Flakes - 324 5.7x10 ' [17] Paint Chips 124 5.7x 10 ' [17] 1 ~w -- p.- - REVISION NO. l '0 l 1 l j-

3 { NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO.18 l 6.1.5 - Considerations in Estimating Head Imsses 6.1.5.1 Debris Sedimentation In the base case calculations, gravitational sedimentation of sludge, dirt / dust, calcium silicate particles and fibrous debris in the suppression pool will be conservatively neglected. Paint chips, however, have an average terminal settling velocity of 0.3 ft/s (9x10 m/s) (Ref. 5.24) which is about a factor of 4 30 greater than the average terminal settling velocity for sludge particles, i.e., 0.01 ft/s (3x10 m/s) 4 (Ref. 5.24). 'Ihus, neglecting gravitational. sedimentation of paint chips is considered to be overly conservative. In the NUREG/CR-6224 study, it was judged that, after cessation of the high-energy phase in a suppression pool during a LOCA, the settling rates will not be lower than 50% of those corresponding to the measured terminal settling velocity for quiescent pools [Ref. 5.24, p.B-30]. To have an upper bound, a settling velocity 10 times lower than settling velocity for quiescent pools, or 0.03 ft/s, will be assumed (see section 3.0). Based on these considerations, the removal fraction of 0.67 (67%) for paint chips and rust flakes, calculated in section 6.1, is used in these base calculations. No experimental data are available to estimate the sedimentation rate of rust flakes. However, with a characteristic particle size comparable to paint chips and density a factor between 2 and 3 higher, one - would expect a greater rate of sedimentation for rust. For conservatism, the same factor 0.67 (67%) proposed for the sedimentation of paint chips (67%), will be used in this analysis. 6.1.5.2 Debris Filtration Not all of the debris particles reaching the strainer would be trapped or filtered by the strainer to form a debris bed on the strainer surface. The fraction of the debris particles approaching the strainer that is deposited and contained in the fibrous debris bed is referred to as the filtration efficiency (Ref. 5.24). Several experiments were conducted by _the NRC to provide bounding estimates for the filtration efficiency of sludge particles (Ref. 5.21). Based on these experiments, a conservative upper-bound value of 0.50 was used for the sludge particle filtration efficiency for debris bed thicknesses higher than 0.25 inches in the NUREG/CR-6224 analysis; for thicknesses lower than 0.25 inches, 0.50 filtration efficiency was deemed overly conservative and a linear variation for the filtration efficiency from. O to 0.5 was used for theoretical thicknesses lower than 0.25 inches [Ref. 5.24, p. B-34]. Consequently, in this calculation the NUREG/CR-6224 filtration efficiency model for sludge particles is used. The characteristic sizes presented in Table 3 suggest that the dominant filtration mechanisms for dust and sludge particles are impaction and interception. For these mechanisms, the filtration efficiency is essentially the same for particles with diameters between 2 and 10 m (Ref. 5.9). On this basis, this i - calculation uses the same efficiency model for sludge and dirt / dust particles, i.e., an efficiency 0.5 for l theoretical debris bed thicknesses lower than 0.25 inches and a linear variation for the filtration efficiency from 0 to 0.5 for theoretical thicknesses lower than 0.25 inches. I-REVISION NO. l 0 l 1 l l u

NEP-12-02 Revision 7 i COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO.19 i To account for the fact that the calcium silicate debris, paint chips and rust flakes are larger than the sludge particles, a filtration efficiency of 1.0 is used for these drywell particles with debris bed thicknesses higher than 0.25 inches; for beds thicknesses lower than 0.25 inches, a filtration efficiency of 0.5 for these drywell particles will be conservatively used in this calculation. 6.1.6 Debris Quantities for the Base Case Analyses Table 5 summarizes the debris quantities for the base case analysis. The flow conditions specified in Table 2 are used. Table 5. Quad Cities Station Unit 2: Debris Quantities Fibrous Debris 6.74 ft' Calcium Silicate Debris 0.78x0.5 ft' (a) RMI Foil Debris 19405 ft 2 l Sludge 443x0.2 lb (a) Dirt / Dust 150x0.2 lb(a) Rust Flakes 50x0.33x0.5 lb (a,b) Total Paint Chips (Qualified + Unqualified) 170x0.33x0.5 lb (a,b) (a) Reduction fraction due to filtration (b) Reduction fraction due to sedimentation 6.2. Supporting Calculations 6.2.1 Strainer Circumscribed Area The strainer circumscribed area, A , is just the surface area of the cylinder, including the end plates, enveloping the strainer. The value in Table 1 is calculated as follows: j 2 A,,,, = fr D,L + 1 (2 D,* - D,*)~1 ft i g g2 l

where, l

D, = 45" is the maximum outside diameter, D, = 20" is the inside core tube diameter and L = 42" is the active length. Substituting numerical values, the circumscribed area is calculated to be 61 ft'. ) 1 FOR REFERENCE AB f t \\/ REVISION NO. l 0 l 1 l () kN L. t l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 20 6.2.2 Gap Volume The gap volume, V,, _ is the total volume within the gaps between the strainer disks. The value in Table 1 is calculated as follows: V' = hD*2 - D,2) d' (N - 1) 1728 / n' 4

where, D, = 24. 5" is the gap diameter, d, = 2" is the gap width and N = 11 is the number of disks.

Substituting numerical values, the gap volume is calculated to be 13 ft'. 6.2.3 Sedimentation Fraction To estimate the fraction of 0.33 (33%) for the quantity of paint chips deposited in the fibrous debris bed on the strainer, the method described in the following paragraphs is used. Consider a water volume in the suppression pool, V of 3177 m' (112200 ft') (Ref. 5.13), the g, maximum credible flow rate through the strainers, Or, of 2 m' /s (4 x 8350 gpm), the maximum water level in the suppression pool, hg, of 4.5 m (14.88 ft), and an average paint chip sedimentation velocity, va,10 times lower than the terminal settling velocity measured for quiescent pools,9 x 102 m/s (Ref. 5.24), i.e., ya = 9 x 10-' m/s. The characteristic sedimentation time, ta, is defined by: hp t,a = vsa Substituting numerical values, the characteristic sedimentation time is calculated to be: ta = 500 s The characteristic suppression pool turn-over time, tg, is defined by: V4 *" " Q7 1 7, g pr,. .~ REVISION NO. l 0 l 1 l l -l-k ..s,v-+--

NEP-12-02 ~ ~ Revision 7 l COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 21 Substituting numerical values, the characteristic suppression pool turn-over time is calculated to be: tg = 1600 s. Based on these considerations, the characteristic sedimentation time is approximately a factor of 3 lower than the characteristic suppression pool turn-over time. i The total quantity of paint chips debris in the pool, m, is given by: r 1 % = m + m, f where mf and m, are the quantities of paint chips deposited into the pool floor and approaching the strainer, respectively. Now, the quantity of paint chips debris that is deposited into the pool floor by sedimentation is inversely proportional to the sedimentation time, whereas the quantity of paint chips debris ~ approaching the strainer, is inversely proportional to the characteristic pool turn-over time (which was estimated to be 3 times lower than the sedimentation time). This results in: m = 3 m, f l Substituting this result in the equation for m gives: ) r ' + 2- = 1 m m r r => m, / m = 0.25, or m;/ m = 0.75 r r l This suggests that approximately 3/4 (75%) of the paint chips would settle to the suppression pool i floor. To have a conservative bound, this calculation considers that 67% of the paint chips settle to the suppression pool floor. 6.3 Base Case Analysis 1 Based on the maximum quantities of each type of insulation (fibrous, metallic and calcium silicate) l l described in Section 4.0, the case presented in Table 6 will be analyzed in this calculation note. All quoted head losses will be reported in ft-water at 60*F. ) 1 ) 1 i REVISION NO. l 0 l 1 l l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 22 l Table 6. Quad Cities Station Unit 2: Conditions for the Base Case Analysis l Fibrous Debris - 6.74 ft' RMI Foil Debris 19405 ft2 Calcium Silicate Deb:is 0.39 ft* Sludge 87lb Dirt / Dust 30 lb Rust Flakes 8 lb Paint Chips 28 lb 6.3.1 Maximum Fibrous Debris Case l As indicated in Table 6, the largest quantity of fibrous debris expected in the QC2 pool is 6.74 ft'. l The theoretical (or as-fabricated) bed thickness, d,, associated with this quantity of fibrous debris is calculated as follows: V _12 i n f i A L* = N, A, ft x

where, V = 6.74 ft'is the largest quantity of fibrous debris in the pool, f

2 A, = 207 ft is the surface area of the strainer and N, = 4 is the number of strainers in QC2. l Substituting numerical values, the theoretical debris bed thickness is calculated to be: 1 d, = 0.698" ] This theoretical thickness is less than 0.125" (1/8 of an inch), which is the minimum debris bed thickness required to bridge the holes in the strainer (1/8 of an inch). At the calculated theoretical bed thickness, a uniform debris bed can not be sustained on the surface of the strainer and some of the fibrous material will penetrate the holes in the strainer. Visual observations during tests on stacked disk strainers (Ref. 5.10) suggest that, at the beginning, the fibrous debris initially accumulates primarily within the gap volume between the disks, and later deposits on the remaining surface area of the strainer. As a result of this, the expected head losses with theoretical bed thickness less than j 0.125" are expected to be negligible. As a matter of fact, the EPRI test P4 (Ref. 5.7) shows that, under these circumstances (4, = 0.088"), a non-measurable head loss was obtained. In particular, the base case at QC2 considers 1.7 ft' of fibrous debris per strainer (6.74 ft' of fibrous i debris in the pool), compared with the 13 ft' of vclume available in the gaps between the disks. ' Consistent with the visual observations during the tests (Ref. 5.10), it is estimated that most of the 1.7 I - - - t, r e-REVISION NO. l 0 l 11 NK K C F C I5 Q N L F. O iAl l.,V,, ynw l

NEP-12-02 Rev% ion 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 23 ft' fibrous debris will be deposited within the 13 ft' of gap volume, leaving most of the strainer surface area without debris. Consequently, the expected head loss due to a combination of fibrous debris and particulate matter for the QC2 strainers is estimated to be negligible. For completeness, however, the head loss per strainer predicted by BLOCKAGE 2.5 for the combination of fibrous debris and particulate matter indicated in Table 6 is presented in Figure 1. 0.45 0.4-y 0.35 - l 0.3 - E. 0.25 - e 8 0.2 - a y 0.15 - Base Case

E 0.1 0.05 -f 0

0 5000 10000 15000 20000 25000 Time (s) Figure 2. Quad Cities Station Unit 2: BLOCKAGE 2.5 calculated head loss per strainer due to 6.74 ft' of fibrous debris,443 lb of sludge,150 lb of dirt / dust,50 lb of rust flakes and 170 lb of paint chips in the pool. As indicated in Figure 1, BLOCKAGE 2.5 predicts a maximum head loss of about 0.42 ft-water at approximately 6000 s into the accident. Since there is not enough fibrous debris to cover the strainer, the expected head loss due to fibrous debris is actually 0 ft-water'. The over-estimation in the BLOCKAGE 2.5 results for thin beds is basically a result of the assumption of considering a uniform and homogeneous distribution of debris on the surface of the strainer. In reality, with the open areas in the debris bed, the flow may pass through relatively clean regions in t' e strainer, thereby resulting in much lower head losses than those predicted by BLOCKAGE 2.5. ' Note that this estimate does not consider the contribution of RMI foil debris to the head loss; such contribution is estimated in Section 6.2.2 I REVISION NO. l 0 l 1 l l .~ 2

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010.M-03% PAGE NO. 24 i 6.3.2 Maximum RMI Foll Debris Case Saturated RMI Debris Bed Head lass at a Bow Rate of 8,350 gym Based on the methodology described in Section 2.1.3, the estimated volume (Vaw) and corresponding foil area (Ar a) to cause a saturated bed at a flow rate of 8,350 gpm on one strainer is depicted in Table 7 for the two types of RMI analyzed. 'Ibe settling velocity and the thickness constants used are also presented. Table 7 RMI Debris Saturation Bed Quantities - 8,350 rpm RMI Debris U. Vaw Au Type (ft/s) (ft) (ft') (ft ) 2 2/2.5 mil SS 0.39 0.014 50 3,590 6 mil Al 0.25 0.073 132 1,815 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.58 ft-water 6 mil aluminum 0.30 ft-water 'Ih highest RMI saturated bed head loss at a flow rate of 8,350 gpm is 0.58 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved during the first 10 minutes of ECCS operation. Saturated RMI Debris Bed Head Loss at a Row Rate of 7,250 gpm The estimated volume and corresponding foil area to cause a saturated bed on one strainer at a flow rate of 7,250 gpm is depicted in Table 8 for the two types of RMI analyzed. The settling velocity and the thickness constants used are also presented. Table 8 RMI Debris Saturation Bed Quantities - 7,250 gpm RMI Debris U, K Vmm Au Type (ft/s) (ft) (ft') (ft') 2/2.5 mil SS 0.39 0.014 34 2,410 6 mil Al 0.25 0.073 100 1,373 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 1 6 mil aluminum 0.17 ft-water FOR -{t,,p QC y(( 2/2.5 mil stainless steel 0.30 ft-water 0N LY REVISION NO. l 0 l 1 - l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 25 The highest RMI saturated bed head loss at a flow rate of 7,250 gpm is 0.30 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved auer the first 10 minutes of ECCS operation. 6.3.3 Maximum Calcium Silicate Debris Case Given the lack of sufficient fibrous debris to form a homogeneous bed, the calcium silicate debris will not contribute to a head loss across the strainer. 6.3.4 Maximum Particulate-to-Mber Mass Ratio As indicated in Section 6.2.1, even the largest quantity of fibrous debris expected in the QC2 pool will not be sufficient to cover the strainer surface. To consider a smaller quantity of fibrous debris, increasing therefore the particulate-to-fiber mass ratio with respect to that in the base case (estimated to be approximately 9), will result in even thinner debris bed thickness. To consider more fibrous debris will result in a lower particulate-to-fiber mass ratio than that considered in the base case. Consequently, this maximum particulate-to-fiber mass ratio case is not applicable to QC2. 6.4 Assessment of Margin As discussed in Section 6.2, ever. considering conservative assumptions the estimated head loss is no worse than maximum RMI head loss of 0.58 ft-water, a value which is already close to the minimum re.iding of the instruments typically used in head loss measurements. Relaxing some of these conservative assumptions, such as those related with sedimentation and filtration of debris, will result in even smaller head losses. Consequently, no assessment of margin will be conducted for QC2. 6.5 Parametric Analysis 6.5.1 Effect of Water Temperature According to the methods used in this calculation, BLOCKAGE 2.5 for head losses due to fibrous insulation and the URG based methodology for head losses due to RMI foil debris, the temperature of the water in the suppression pool only affects the head loss due to fibrous debris. As indicated in Section 6.2.1, in the worst case for fibrous debris in the QC2 pool the head loss is negligible. The temperature of the water only affects the water propenies used in the head loss model, i.e., the dynamic viscosity and density. A change in water temperature will not affect the theoretical debris bed thickness of 0,098" (which, as shown in Section 6.2.1, is less than the minimum required to produce a measurable head loss). Therefore, the predicted head loss due to fibrous debris will still be negligible. The SER based RMI head loss correlation is independent of the water temperature. Consequently, any change in the pool water temperature will result in the same maximum estimated head loss of 0.58 ft-water due to RMI foil debris. T REVISION NO. l 0 l 1 l l \\

NEP-12-02 ~ Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 26 6.5.2 Effect of Strainer Surface Area Reduction ne impact on the head loss of a reduction in the surface area of strainers by 2 ft for each strainer, 2 due to deposition of extraneous materials such as plastic bags, can be investigated by considering a strainer surface area of 205 ft and the debris quantities presented in Table 6. In this situation, the theoretical debris bed thickness is es,timated to be AL, = 0.01", which is still f below the mimmum theoretical thickness of 0.125" required to sustain a fibrous debris bed. Consequently, the head loss due to fibrous debris is estimated to be negligible. Saturated RMI Debris Bed Head Imss at a Flow Rate of 8,350 gym Based on the methodology described in Section 2.1.3, the estimated volume (Van.) and corresponding foil area (Ar a) to cause a saturated bed at a flow rate of 8,350 gpm on one reduced surface area strainer is depicted in Table 9 for the two types of RMI analyzed. The settling velocity and the thickness constants used are also presented. Table 9 RMI Debris Saturation Bed Quantities - 8,350 mm RMI Debris U. K Vans A Type (ft/s) (ft) (ft') (ft') 2/2.5 mil SS 0.39 0.014 54 3,907 6 mil Al 0.25 0.073 141 1,934 Based on the RMI foil quantitles of the above table and the SER RMI head loss correlation previously ~ described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.70 ft water 6 mil aluminum 0.35 ft-water The highest RMI saturated bed head loss on one reduced surface area strainer at a flow rate of 8,350 gpm is 0.70 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved during the first 10 minutes of ECCS operation. Saturated RMI Debris Bed Head Loss at a Row Rate of 7,250 gym The estimated volume and corresponding foil area to cause a saturated bed on one reduced surface area strainer at a flow rate of 7,250 gpm is depicted in Table 10 for the two types of RMI analyzed. The settling velocity and the thickness constants used are also presented. Table 10 RMI Debris Saturation Bed Quantities - 7,250 pm RMI Debris U, N Vans A Type (ft/s)- (ft) (ft*) (ft') 2/2.5 mil SS 0.39 0.014 37 2,666 6 mil Al 0.25 0.073 107 1,469 3 REVISION NO. l 0 l 1 l l. i

NEP-12 02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 27 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.36 ft-water 6 mil aluminum 0.20 ft-water The highest RMI saturated bed head loss on one reduced surface area strainer at a flow rate of 7,250 gpm is 0.36 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved after the first 10 minutes of ECCS operation. This represents an increase in the estimated head loss by approximately 20%. 6.5.3 Quantity of Mbrous Debris Nee==aary to Produce a Measurable Head 14ss As indicated in Section 6.2.1, the largest quantity of fibrous debris expected in the QC2 pool, estimated to be 6.74 ft', is not enough to produce a theoretical debris bed thickness of 0.125", which is the minimum required to result in a measurable head loss, as explained in Section 6.2.1 and demonstrated by the EPRI experiments (Ref. 5.7). It may be interesting, however, to estimate what is the quantity of fibrous debris in the pool required to produce this minimum theoretical debris bed thickness of 0.125". This quantity of fibrous debris, V,, is given by: V=(N,A,AL,,,,,,) f

where, N, = 4 is the number of strainers, A, = 207 ft'is the surface area of each strainer and A = 0.125" is the minimum theoretical thickness to produce a measurable head loss.

Substituting numerical values, this minimum quantity of fibrous debris in the QC2 pool required to result in a measurable head loss is estimated to be: V,. = 8.63 ft' Note, however, that this value represents just the quantity of fibrous debris in the pool that may result in a measurable (i.e.,

• O ft water) head loss. This quantity of fibrous debris in the pool, distributed on each of the strainers at QC2 resuhs in about 2.2 ft', a quantity that is expected to by primarily deposited in the gap volume of 13 ft' per strainer, leaving most of the remaining surface area of the strainer relatively clean of debris. Consequently, this estimated minimum quantity of fibrous debris in the pool to produce a measurable head loss is still extremely conservative.

i mn n ecr.o cNCE REVISION NO. l 0 l 11 I V T* ( " ' '..,, l V (N L. k L

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-03% PAGE NO. 28 7.0

SUMMARY

AND CONCLUSIONS In this calculation note, the following cases were analyzed: Base Case: 6.74 ft' of fibrous debris,19405 ft of RMI foil debris,0.78 ft' of calcium silicate debris, 2 443 lb of sludge,150 lb of dirt / dust, 50 lb of rust flakes and 170 lb of paint chips in the pool. Parametric Case 1: Same conditions as in the Base Case, but considering a reduction in the pool water temperature by 35'F. Parametric Case 2: Same conditions as in the Base Case, but considering a reduction in the strainer surface area by 2 ft: due to extraneous materials, i.e., plastic bags. Parametric Case 3: Estimation of the minimum quantity of fibrous debris in the pool required to produce a measurable head loss. The largest quantity of fibrous debris expected in the pool at QC2,6.74 ft', results in a theoretical bed thickness of 0.098", which is smaller than the minimum theoretical bed thickness required to produce a measurable head loss, i.e., 0.125". Reducing the quantity of fibers, to maximize the particulate-to-fiber mass ratio, would result in an even thinner bed thickness and, therefore, this case was not considered applicable to QC2. The conditions and corresponding head losses estimated for each of these cases are presented in Table 7. Table 7. Qaad Cities Station Unit 2: Summary of Head Loss Estimates Case Fibrous RMI Cal-Sil Sludge *) Dirt / Rust Paint Head Debris Debris " Debris *) (Ib) Dust *) Flakes ** Chips *d Loss (ft') (ft') (ft') Ob) Ob) Ob) (ft-water) Base 6.74 3,590 0.39 87 30 8 28 0.58 1 6.74 3,590 0.39 87 30 8 28 0.58 2 6.74 3,590 0.39 87 30 8 28 0.70 3 8.63 3,590 0.39 87 30 8 28 0.31* The highest head loss estimated for QC2, considering conservative assumptions, is 0.70 ft-water and is due solely to the contribution of RMI foil debris, i.e., the contribution of fibroue debris is negligible. 'Ihis highest head loss occurs during the first ten minutes of ECCS operation. Consequently, the impact of relaxing conservative assumptions in the estimation of the head loss due to fibrous debris was not further investigated in this calculation, a) This is the quantity of RMI debris required to produce the saturation bed thickness (i.e., maximum head loss due to RMI debria) for 2/2.5 mil stainless steel foils, b) Considering debris filtration. c) Considering debris sedimentation. d) Estimated qualitatively. REVISION NO. l 0 l 1 l l l .O

F-NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 29 ' The most relevant conclusions are summarized as follows: 1. The largest quantity of fibrous debris expected in the pool of QC2, 6.74 ft', is not enough to produce the minimum theoretical bed thickness required to result in a measurable head loss due to fibrous debris. 2. Metallic debris produces the worst head loss results. The highest head loss for RMI foil debris occurs at the saturation bed thickness. He quantity required to produce saturation bed thickness, estimated to be 3,590 ft' for QC2, results in a head loss no worse than 0.6 ft-water. 3. He limited parametric analysis conducted suggests that the impact on the head loss estimates due to a reduction in the surface area of each strainer by 2 ft', such as that originated by deposition of extraneous materials on the strainers, would increase the estimated head loss by approximately 20%. 4. The minimum amount of fibrous debris in the pool required to produce a measurable head loss is I estimated to be 8.63 ft', a quantity that is approximately 28% larger than the maximum quantity of fibrous debris expected to be transported to the QC2 suppression pool. This quantity is still very conservative, because it results in about 2 ft' of fibrous debris per strainer. Based on visual observations during tests conducted on stacked disk strainers, most of these debris are expected to be accumulated primarily within the 13 ft' of volume available within the gaps between the stacked disks, leaving a considerable area of the strainers relatively " clean" from debris, resulting thereby in very low head losses. t FINAL e REVISION NO. l 0 l 1 l l ~.

r NEP-12-02 Revision 7 CALCULATION TITLE.PAGE Calculation No: ODC-0010 M-0396 DESCRIPTION CODE:M03 DISCIPLINE CODE: M STATION / UNIT: Quad Cities Unit 2 SYSTEM CODE: RH. PC. CS TITLE: Ouad Cities Station Unit 2: ECCS Strainer Head Loss Estimates X Safety Related Augmented Ocality Non-Safety Related REFERENCE NUMBERS Type Number Type Number PROJ VO1500 For additional References see Section 5. COMPONENT EPN: DOCUMENT NUMBERS: EPN Compt Type Doc Type / Subtype Document Number 2-1600-4 F10 Corr /NDIT D104-00014 2-1600-8 F10 Corr /NDIT ODC-97-052 2-1600-12 F10 Corr /NDIT QDC-97-084 2-1600-16 F10 V REMARKS:

REV, REVISING APPROVED DATE ORGANIZATION PRINT / SIGN 0

DE&S (D104) Rob Choromokos / Signature on File 9/1/97 1 DE&S (D104) ~~p,b 4,y, [j_[) ef,/p3,/pp IOh r=L = nm _ __ j wN iMPtREh CE i

a NEP-12-02 Revision 7 l COMMONWEALTH EDISON COMPANY CALCULATION RFXISION PAGE j CALC'ULATION NO. QDC 0010-M-0396 PAGE NO.: 2 REVISION SUMMARIES REV:0 i REVISION

SUMMARY

) Initialissue 1 Electronic Calculation Data Files: i (Program Name. Version, File name exusue/date/ hour /: min) A ) !I I Prepared by: Francisco J. Souto / Sig iature on File 8/28/97 l Print / Sign Date Reviewed by: Jan Bostelman / Sienaturepn File 8/28/97 Print / Sign Date Type of Review [ x ] Detailed [ ] Altemate [ ] Test I Supplemental Review Reauired i 1 Yes (NEP-12-05 documentation attached) I1No Supem.or / (Comed USE ONLY) DO ANY ASSUMFDONS IN THIS CAlfULAllON REQUIRE LATER VERIHCATION [] YES [ x] NO Tracked by: (NTS #, ER #, etc.) REV: 1 REVISION

SUMMARY

Calculation is reformatted per the requirements of NEP-12-02, Rev,7. Sections 2,5,6 and 7 are modified to account for 2.0 and 2.5 mil stainless steel foils. Assumptions 3.7 and 3.8 are changed. Ref. 5.25 is added. No other debris types were evaluated in this revision Electronic Calculation Data Files: (Program Name, Version, File name extsimdate/ hour /; min) Prepared by:_pil Zieler / u .4/_2)_/99 ' Print / Sign / Date Reviewed by: Jan Bostelman L [kb 4/23/99 ~ Pfnt/ Sign Date Type of Review / [ x ] Detailed [ ] Altemate [ ] Test Supplemental Review Reauired i 1 Yes (NEF-12-05 documentation attached) [1No Supervisor / (Comed USE ONLY) DO ANY ASSUMFDONS IN11IIS CAlfUIAilON REQUIREIATER \\_'ERIRCAllON [ ] YES [x]NO Tracked by: (NTS #, ER #, etc.)

• WA ce = mew 4e eoe

--ee.,

t NEP-12-02 Revision 7 COMh10NWEALTH EDISON COMPANY CALCULATION TABLE OF CONTENTS CALCULATION NO. ODC-0010-M-03% REV. NO.1 PAGE NO. 3 SECTION PAGE NO. SUB-PAGE NO. -TITLE PAGE - 1 REVISION

SUMMARY

2 TABLE OF CONTENTS 3 PURPOSE / OBJECTIVE 4 METHODOLOGY AND ACCEPTANCE CRITERIA 4 ASSUMPTIONS 11 DESIGN INPUT 12 REFERENCES 12 CALCULATIONS ' 14 l

SUMMARY

AND CONCLUSIONS 28 l 4 I 1 1 i I I i J

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010.M.0396 PAGE NO. 4 1.0 PURPOSE / OBJECTIVE The purpose of this calculation note is to estimate the head loss across the debris bed formed on the proposed strainers at the suction of the emergency core cooling system (ECCS) pumps at the Quad Cities Station Unit 2 (QC2), due to accumulation of insulatica debris (fibrous, calcium silicate and reflective metallic) and paniculate matter produced as a result of a loss of coolant accident (LOCA). He level of margin in the estimation of the head loss is assecsed. Additionally, a limited parametric analysis is performed on key variables affecting head loss estimates, ne head loss estimates reported herein do not include the head loss associated with the clean strainer. 2.0 METHODOLOGY AND ACCEPTANCE CRITERIA 2.1 Methodology ne methods used for the estimation of the head loss across the proposed strainers at the suction of the ECCS of QC2 are based on the NUREGICR4224, Parametric Study of the Potentialfor BWR ECCS Strainer Blockage due to LOC 4 Generated Debris (Ref. 5.24) and the SER for NEDO-32686 "URG for ECCS Suction Strainer Blockage," (Ref. 5.6). He NUREG/CR4224 fundamental models were implemented by the U.S. Nuclear Regulatory Commission in the BLOCKAGE 2.5 computer code (Ref. 5.20 and 5.22). This section sununarizes the methods used in this calculation note. Further details on the SER, the NUREG/CR4224, and the BLOCKAGE 2.5 methods and models, as well as their corresponding ranges of applicability, are presented h the original references (Ref. 5.16), (Ref. 5.24), (Ref. 5.20 and 5.22), and (Ref. 5.23), respectively. In particular, the BLOCKAGE 2.5 computer code will be used to estimate the head loss due to fibrous and particulate matter debris, whereas the URG methodology used to estimate the head loss due to j Reflective Metallic Insulation (RMI) debris is based on the methodology proposed in the SER for the i BWROG-URG. I 2.1.1 BLOCKAGE 2.5 Computer Code i The BLOCKAGE 2 5 code was developed by Science and Engineering Associates, Inc. for the U.S. Nuclear Regulatory Commission (NRC) as a tool to evaluate licensee compliance regarding the design of ECCS suction strainers as required by NRC Bulletin 96-03, Potential Plugging ofEmergency Core Cooling Suction Strainers by Debris in Boiling Water Reactors. As - 'd in its reference manual (Ref. 5.22), the BLOCKAGE code was developed to predict whether or accumulation of debris on the 1 ECCS suction strainers following a postulated LOCA would lead to loss of net positive suction head j (NPSH) in a Boiling Water Reactor (BWR), BLOCKAGE 2.5 allows the user to simulate debris ) generation and subsequent transport of different types of debris including fibers, particles and metals, i using user specified debris generation and transport factors. De transpon of debris from the drywell to the srppression pool can be simulated as location-dependent and time-dependent. Alternatively, BLOCKAGE '2.5 allows the user to transpon instantaneously to the suppression pool a specified quantity of debris, in the pool, BLOCKAGE 2.5 allows the possibility of modeling the transpon of j REVISION NO. l 0 l cab D C C C D C M ("' C .was

J NEP-12-02 Revision 7 1 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-33% PAGE NO. 5 debris to the strainers by considering Ae specific sedimentation characteristics for each debris type and size group. Four user optional head loss correlations, including the NUREG/CR4224 head loss correlation for a mixture of fibrous and particulate matter debris, are available in BLOCKAGE 2.5 to estimate the pressure drop across the debris bed on the strainers. To estimate the quantity of particulate matter debris retained in the fibrous debris bed, BLOCKAGE 2.5 allows the user to provide the filtration efficiency for each type of debris. 2.1.2 Head less due to Fibrous Insulation Debris According to the URG (Ref. 5.6, p.14), the head loss from fibrous debris, in combination with suppression pool sludge, is expected to dominate the strainer head loss. Consequently, considerable attention was given to the available head loss correlations for fibrous insulation debris. BLOCKAGE 2.5 contains four optional correlations that may be selected by the user to model the head less across a debris bed consisting of fibrous and particulate debris. he four correlations are the NUREG/CR4224 correlation, the empirical correlation developed by the BWROG in 1994 (Ref. 5.5) (not to be confused with the URG head loss correlation for fibrous debris), and two generic correlations that can be used to implement an alternate user correlation. Based on the assessment documented in DE&S report TR-ECCS-GEN-03 (Ref. 5.23), it was determined that the fibrous debris head loss correlation based on the NUREG/CR4224 study was the most applicable and defensible for the QC2 analysis. Hence, it forms the basis for these calculations. He NUREG/CR4224 head loss correlation, as described in Appendix B to NUREO/CR4224 (Ref. 5.24), is a semi-theoretical head loss model. It is based on theoretical and experimental research for the pressure drop across a variety of fibrous and porous mcdia carried out since the 1940s. His head loss model has been implemented in BLOCKAGE 2.5, as described in the corresponding reference manual (Ref. 5.22). Its results have been extensively validated for debris beds composed of fiberglass and simulated suppression pool sludge (Ref. 5.21), as well as mineral wool fibrous materials (Ref. 5.15) for flat strainers. The NUREG/CR4224 head loss correlation was also shown to produce conservative results for the stacked disk strainers tested at EPRI when the amount of fibrous debris did not exceed 10 ft'(Ref. 5.23). I 1 l -i i f h., REVISION NO. l 0 l 1 l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 6 2.1.3 Head less Correlation due to Reflective Metallic Insulation Debris 'Ibe type of foil of the originally installed Reflective Metallic Insulation (RMI) (Ref. 5.2) at the Quad Cities i Station is 6 mil Aluminum. In the last few years, the foil type in replacement RMI cassettes has been either 2 mil or 2.5 mil stainless steel. In order to provide an estimate of the differences between two types of RMI, this calculation will present results for both 2/2.5 mil stainless steel and 6 mil aluminum foils. j The BWROG study (Ref. 5.6) provides an empirical correlation to estimate the head loss due to different i types of RMI debris for BWR ECCS suction stramers. However, while these efforts provided some valuable insights into dibuces between the different types of RMI, the NRC's SER (Ref. 5.16) concluded that the resulting correlation could not be d=anetrated to be conservative under all conditions. The NRC instead presented an alternate correlation which forms the basis for the results presented herein. URG-SER Head Loss Correlationfor RMI Debris The SER of the URG presents the following correlation (Equation K.Sa in the SER (Ref. 5.16)) that is stated to adequately bound the test data from the ARL (Ref. 5.25) and URG (Ref. 5.6) RMI tests: A 2 m AH = 0.108U (3) 4

where, AH is the head loss (ft-water),

U is the approach velocity (ft/s) based on the available strainer area, 2 A is the RMI foil surface area (ft ), and y A, is the available area of the strainer (ft ) which is taken as the circumscribed area of the 2 outer cylindrical strainer shape. REVISION NO. l 0 l-1 l l

NEP-12-02 Revision 7 j COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 7 This equation is derived based on the head loss tests conducted by the NRC at the ARL test loop facility, using debris generated by the NRC RMI debris generation test (Ref. 5.25). The NRC debris generation RMI test was a steam test using a 2.5 mil Stainless Steel foil RMI Diamond Power cassette mounted on a circumferential weld break simulator. The SER also concluded that this correlation adequately predicted experimental data reported in the 'URG for gravity head loss tests using debris from the NRC RMI debris generation test, as well as tests conducted using 2.5 mil Stainless Steel debris manually generated by CDI. This correlation was also adopted to estimate head losses due to 2 mil Stainless Steel RMI debris. The % mil thickness difference between the two types of Stainless Steel RMI is not expected ti cause measurable differences in head loss. Both types of foil are expected to form very similar debris beds given the anticipated minimal variation in the strength of the crumbled debris pieces. This correlation is also assumed to bound head loss estimates if the RMI debris comes from 6 mit Aluminum instead of 2.5 mil Stainless Steel. The SER suggests that the smaller sized RMI debris would form beds with lower void fractions than larger sized RMI debris. The URG RMI debris generation tests showed that the 6 mil Aluminum RMI debris pieces were much larger than the debris pieces generated from the NRC 2.5 mit Stainless Steel. As such, a 6 mil Aluminum RMI debris bed will have larger void fractions than a 2/2.5 mit Stain ess Steel RMI debris bed. Therefore, for the same foil area, the head losses of a 6 mil Aluminum RMI debris bed will be lower than a 2/2.5 mil Stainless Steel RMI debris bed. The effect of larger pieces generating lower head losses than smaller pieces in the flow velocity regime of the Quad Cities replacement strainers is clearly shown in the NRC sponsored RMI head loss tests [Ref. 5.25, Appendix D, Figure 3]. RMI Saturation Thickness Experimental evidence and theoretical reasoning suggest that RMI debris buildup on the strainer would reach a saturation limit, beyond which local surface flow velocities would not snic:e sufficient drag to i overcome forces imposed primarily by turbulence and gravity. The URG experiments suggest that this limit is given when the local surface flow velocity is one half of the average terminal settling velocity of the RMI debris. A spherical RMI debris buildup model can be derived based on the simplified Figure I illustration. For a spherical RMI debris deposition on a stacked-disk strainer, the ratio of strainer approach velocity based on the circumscribed strainer area, U,, to the local flow velocity at the debris surface, U, may be approximated by: U A 4#R -O 2 (2) U A, x L D, + x R,' + x (R,' - R,* ) j FOR REFERENG, REVISION NO. l 0 l 1 1 l Q M ;lv l r l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC.4010-M-0396 PAGE NO. 8 where (see Figure 1): . A is the surface area of the RMI spheroid debris bed (ft'), A,is the cNcumscribed area of the strainer (ft'), R is the radius of the RMI spheroid debris bed (ft), ~ L is the strainer active length (ft), D is the strainer outer diameter (ft), R is the oudet pipe radius (ft), and i D is the area of spherical segment associated with the interference between the RMI debris bed and the outlet pipe (ft'). The radius of the RMI debris spheroid as a function of the average local flow velocity at the debris surface is then approximated by: l l \\ \\ R = )1 h D, + 2 R,2 - R,2 )+ b (3) 4 U x t 3. 7. m., -. r - t *: 3,- ., J, REVISION NO. l 0 l 1 l l l

i NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO. 9 ',,,,....------...,,~~~ ', ~. R, N, s /n i, / ^ \\ / A \\. i l \\- fD. I ^ l \\ } \\ i / \\ / l \\v i '\\, 4 L >,/ s ~. d' ',~~.. -.............. ',,- Figure 1. Schematics of a spheroid RMI debris bed on a strainer. Since the local flo$v velocity at saturation conditions is approximately % of the average settling velocity of the RMI debris, U,, the saturation bed U corresponding to a radius R can be approximated by: U (at R = R,) = U, = U"' (4) 2 Hence, the equivalent volume of RMI debris required to produce saturation conditions, Vu, may be estimated by: V, = x R,' - x R

• L - x R,' (R, Ll)

(5) The corresponding RMI debris foil area, A, is then given by: i A,=y" (6) X, I l REVISION NO. l 0 l 1 l l

i C l 1 i y<, NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY l CALCULATION NO. QDC-0016.M.0396 PAGE NO.10 1 1 where K,(in ft)is the thickness constant for RMI debris. Based on experiments reported in the URG, K, is equal to 0.014 ft for 2.5 mil stainless steel debris whereas for 6 mil aluminum K, is equal 'O.073 ft (Ref. 5.6). The K, value of 0.014 ft will also be used for the 2 mil stainless steel. He above methodology can be applied to Quad Cities Station as follows: ' 1. Determine the saturated bed thickness for a 6 mil aluminum RMI debris bed using equations 2 through 6.

2. Detertaine the head loss for a 6 mil aluminum saturated debris bed using equation 1.

3. Determine the saturated bed thickness for a 2/2.5 mil stainless steel RMI debris bed using equations 2 through 6.

4. Determine the head loss for a 2/2.5 mil stainless steel saturated debris bed using equation 1.

2.1.4 Head Loss due to Other Insulation Debris: Calcium Silicate Destruction of insulation based on calcium silicate (Cal-Sil) will produce particulate matter debris. Without enough fibrous materials to cover the strainer holes, calcium silicate debris (as is the case for any type of particulate matter) will act produce a significant head loss due to insufficient filtration. In the presence of enough fibrous debris to cover the strainer holes, the contribution of calcium silicate debris to the head loss will be modeled by BLOCKAGE 2.5 by including this debris as paniculate untter in the NUREG/CR-6224 correlation. I 2.1.5 BLOCKAGE 2.5 Verification and Validation i BLOCKAGE 15 has been subjected to rigorous coding verification by its developers to casure that the cede performs as it was designed to perform, and extensive quality assurance (Q A) was integrated into ] the development of the BLOCKAGE 2.5 code (Ref. 5.22), Based on this information, BLOCKAGE 2.5 is an approved code by DE&S (Ref. 5.8). 2.2 Acceptance Criteria here are no acceptance. criteria for this calculation. The results presented herein will provide input to a subsequent NPSH margin calculation. i FOR REFERENCE i ~ REVISION NO. l 0 l 1 l ([1 id I Yl 1

n ( NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY l-CALCULATION NO. QDC-0010.M-0396 PAGE NO.11 1 3.0 ' ASSUMPTIONS 3.1 Due to the common ring header, the ECCS flow is assumM to be equally distributed among the four strainers. 3.2 he detris bed is formed and distributed uniformly over the surface of the strainer. This assumption is ; conservative, because if the debris bed were non uniform, the debris bed thickness would be non-uniform, allowing for the possibility of having relatively " clean" regions on the strainer, and thus reducing the head loss. 3.3 The densities and characteristic dimension of the miscellaneous fibrous debris are considered to be similar to those of NUKONS. De ' ensities and characteristic dimension of each drywell particulate material (i.e., equivalent 3.4 d diameter for calcium silicate debris, dirt! dust and sludge, and thickness for paint / coating chips and rust flakes) wdl be assumed based on generic data. When large uncertainties exist in the characteristic size of particulate materials, such as in the case of unqualified paint chips or rust flakes, the smallest reasonable value will be used for conservatism. i i 3.5, In the base case calculations, it will be considered that 33% of the quantity of large particles (paint { chips and rust flakes) initially in the pool is actually deposited within the fibrous debris bed. This is based on assuming a sedimentation velocity 10 times lower than the terminal settling velocity for paint chips and rust flakes. No credit la given to the deposition of fibers, calcium silicate debris, sludge and dirt / dust paalcles on the floor (mainly by gravitational sedimentation) or on other structures in the suppression pool. 3,6 In the base case calculations, it will be considered that 50% of the quantity of small particles (sludge and dirt / dust) initially in the pool is actually deposited within the debris i>ed. This is based on the assumption that the filtration efficiency model developed from the U.S Nuclear Regulatory . Commission (NRC) experiments on head loss (Ref. 5.21) is valid for sludge and dirt / dust particles. For paint chips and rust flakes, a filtration efficiency of 1.0 will be assumed for debris bed thicknesses higher than 0.25 inches, and 0.50 for debris bed thickasses lower than 0.25 inches. 3.7 This calculation assumes that the NRC SER RMI head loss correlation is applicable to the Quad Cities strainer and all RMI debris types expected. The SER RMI head loss correlation aAquately predicted experimental data for tests conducted using 2.5 mi! Stainless Steel debris. Itis reasonable to assume that the 2 mil Stainless Steel debris would be similar in shape and size to the 2.5 mil Stainless Steel debris tested. Hence, the thickness parameter, K,. settling velocity, and head losses are expected ta be the same. 'Ihe correlation should also bound the head losses from 6 mil aluminum RMI. The URG RMI debris characterization information clearly . shows larger debris pieces and lower packing density for the 6 mil aluminum as compared to n the 2.5 mit Stainless Steel debris. This higher void fraction for the aluminum RMI debris would result in a lower head loss for the same foil area REVI5 ION NO. l .0 [ 1; l l- - e ~ ~ %.m... e=---

t NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0310-M 03% PAGE NO.12 3.8 This calculation assumes that the head loss of a debris bed composed of a mixture of RMI debris and fibrous debris is bounded by the sum of the head loss contribution of each type of debris. 4.0 . DESIGN INPUTS 'Ibe design input information for this calculation is obtained from Refs. 5.6, 5.16, 5.20, 5.21, 5.22 and 5.24. 5.0 REFE. ENCES R 5.1 'Aldinger, T.I., R.A. White, and R.A. Manley, Performance of Contcinment Coatings During a Loss . of Coolant Accident, Bechtel Power Corporation, November 10,1994. (In Volume III of the Utilhy Resolution Guidance (URG)for ECG Suction Strainer Blockage, BWROG, November 1996). 5.2 Ashbaugh, S. (a), Quad Cities Station Unit 2: Estimation ofInsulation Debris Source for ECCS ) Strainer Head Loss Calculations, Innovative Technology Solutions Calculation Note No. QDC-0010-M-0394 (ITS/CN-97-20), Rev. O, May 1997. 5.3 Ashbaugh, S. (b), Quad Cities Station - Unit 2: Estimation of Non Insulation Drywell Debris Sources for ECG Strainer Healless Calculations, Innovative Technology Solutions Calculation Note No. QDC-0010-M-0395, Rev. O, May 1997 5.4 Blomquist, M. and M. Dellby, Report From Tests Concerning the Efect of Steam Jet on Caposil Insulation at Karlshamn, SDC-93-1174, June 1993. 5.5-BWROG, Interim Report of the BWR Owners' Group ECW Suction Committece, Bolling Water Owners' Group, December 1994. 5.6 BWROG, Utility Resolution fiuldancefor ECG Suction Strainer Blockage, Boiling Water Owners' Group, NEDO-37686, Rev. O, November 1996. 5.7 ' Dieril, R.;. R., Louderback, A. Kaufman, and A. Bilanin, Testing of Alternate Strainers With Insulation Fiber and Other Debris Revision 2, C.D.I. Report No. 95 09, October 1996. (in Volume I - of Litility Resolution Guidance (URG) fer ECCS Suction Strainer Blochee, BWROG, October 1996). I 5.8 ' 'DE&S, Computer Program Verification and Validation, Procedure No. DPR-3.5, Rev. 4, Duke ' Engineering and Services, January 10, 1997. 5.9 Fuchs, N., 7he Mechanics ofAerosols, Pergamon Press Ltd., New York,1964. 5.10 Hatt, G., Summary report on Performance of Performance Contacting, Inc.'s Sure-Flow" Suction St'ainer with Various Mixes of Simulated Post-LOCA Debris, Revision 0, Performance Contracting Inc., February 14,1997 i REVISION NO.- -j 0 l 1 l l ~

• W.

m

1 NEP-12 02 1 Revision 7 COMMONWEALTH EDISON COMPANY l CALCULATION NO. QDC-0010-M-03% PAGE NO.13 5.11 Henmes, T., Limits of Applicabilityfor the URG RM1 Read Loss Correlation, Report No. TR-ECCS-GEN-04 (ITS/ITS-97-04), Revision 0, Duke Enginaering and Services, Inc., July 18, 1997. 5.12 Munchausen, J.H., Air Jet hnpact Testing of Fibrous and Reflectiw Metallic insulation, Revision A, Final Draft, C.D.L' Report No. 96-06, Continuum Dynamics, Inc., September 1996. J 5.13 NDIT No. D104-Q0014,. Quad Cities Station Unit 1 and 2 Design Information for NPSH Calculations, Nuclear Design Information Transmittal (NDIT), Commonwealth Edison Co., November 27,1996. '(Source ofInformation: Tech Specs. 3.7/4.7 A.1.a/b). 5.14 NDIT No. 97-052, ECCS Suction Strainerflow rates arvipool temperaturesfor DBA LOC 4, Nuclear Design Information Transmittal (NDIT), Commonwealth Edison Co., April 25, 1997. (Sources of Information: 1) Quad Cities Calculation No. QDC-1000.M 0291, Rev. O,2) Quad Cities Calculation No. QDC-1000-M-0292, Rev. O, 3) Quad Cities NTS No. 254525%DRE134, 4) Dresden Calculation 'No. DRE97-0012, Rev. O, 5) General Electric Report No. GENE-637-022-0893, 6) Facsimile fram K. Ramsden to J. Garrity dated 12/30/96). ( 5.15 NEA, Knowledge Basefor Emergency Core Cooling System Recirculation Reliabillsy, NEAICSNllR (95)11, Prepared by the U.S. Nuclear regulatory Commission for the Nuclear Energy Agency (NEA) of the OECD, February 1996. 5.16 NRC, Safety Evaluationfor NEDO-32686, Rev. O, ' Utility Resolution Guidance Documentfor ECCS Suction Strainer Blockage, U.S. Nuclear Regulatory Commission, August 20.1998. 5.17 OG94-661-161, Letter from T.A. Green (BWROG) to A. Serkir (USNRC), "BWR Owners' Group ECCS Strainer Committee Suppression Pool Sludge Particle Size Distribution", Dated September 13, 1994. (In Volume III of Utihty Resolution Guidance (URG) for ECCS Suction Strainer Blockare, 8 BWROG, November 1996). I5.18 PCI, Gusd Cities Unit-2. Sure-Flow Stratact, Diagram QCU2-SUMP-8002-1100, Rev. 5, l Performance Contracting, Inc., November 27,1996. 5.19 Reist, P.C., Intraluction to Aerosol Science, Macmillan Publishing Company,1984. j 5.20- Rao, D.V., W. Bernahl, J. Brideau, C. Shaffer and F. Souto, BLOCKAGE 2.5 User's Manual, NUREO/CR4370, U.S. Nuclear Regulatory Commission, December 1995. ] 5,21 Rao, D.V. and Souto, F., Experimental Study of Head inss and Mitration for LOC 4 Debris, } NUREG/CR-6367, U.S. Nuclear Regulatory Commission, February 1996. i 5.22

Shaffer, C.J., W. Bernahl, J. Brideau and D.V. Rao, BLOCKAGE 2.5 Reference Manual, j

NUREG!CR 6371, U.S. Nuclear Regulatory Commission, December 1996, i i I l REVISION NO. l 0 l 1 1 t/*g r3 D [ r r b r-e. a e r j

• v aw c a s s, L.

j

i l NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO.14 5.23 Souto, F., Range ofApplicabilityfor URG and NUREG/CR-6224 Head Loss Correlations, Repost No. TR-ECCS-GEN-03 (ITS/ITS-97-03), Revision 0, Duke Engineering and Services, Inc., July 14, 1997. 5.24 Zigler, G., J. Brideau, D.V. Rao, C. Shaffer, F. Souto, and W. Thomas, Parametric Study of the Potentialfor BWR ECC5 Strainer Blockage Due to LOCA Generated Debris, NUREGICR-6224, U.S. Nuclear Regulatory Commission, October 1995. 5.25 Zigler, G., J. Brideau, and B. Zigler, Experimental inwstigation of Head Less and Sedimentation Garacteristics of Reflectiw Metallic insulation Debris, SEA No. *i5-970-01-a:2, Science and Engineering Associates, May 1996. 6.0 CALCULATIONS 6.1 TecimicalInput This section describes the information used to develop the BLOCKAGE 2.5 specific input data file for QC2. Basically, this information consists of plant specific parameters, quantitles and physical characteristics for each type of debris, as well as the modeling considerations with respect to debris sedimentation into the suppression pool floor and filtration of particulate matter within the debris bed. 6.1.1 Strainer Data Table 1 presents the dimensions of each of the four stacked-disk strainers installed at QC2. Table 1. Quad Cities Station Unit 2: Strainer Dimensions Length 42 inches (Ref. 5.18) Maximum Outside Diameter - 45 inches (Ref. 5.18) Inside Core Tube Diameter 20 inches (Ref. 5.18) Gap Diameter 24.5 inches (Ref. 5.18) Gap Width 2 inches (Ref. 5.18), Disk Width 2 inches (Ref. 5.18) Number of Disks 11 (Ref. 5.18) Total Surface Area 207 ft* (Ref. 5.2) I Circumvribed Area 61ft (See Section 6.1)) 2 Interstitial Volume 13 ft' (See Section 6.1) l REVISION NO. l 0 l l' l l g ~

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO.15 6.1.2 Flow Conditions The flow rate and suppression pool water temperature as a function of time considered in these head loss estimates are presented in Table 2 (Ref. 5.14)'. Table 2. Quad Cities Station Unit 2: Flow Conditions (Ref. 5.14L Time Pool Water Temperature Total ECCS Mow Rate (s) (T) (gpm) 16 106 33400 ~ 31 117 33400 59 129 33400 337 144 33400 600 149 33400 601 149 29000 1000 154 29000 2 10000 1762 29000 6.1.3 Debris Quantitles 6.1.3.1 librous Insulation Debris As estimated in calculation note QDC-0010 M-0394 (Ref. 5.2), the worst case break location in the QC.2 drywell generates and transports 6.74 ft' of NUKON fibrous debris to the suppression pool. All fibrous debris quantities include 2 ft' of miscellaneous fibrous debris materials transported to the suppression pool (Ref. 5.2). In addition,0.78 ft' calcium silicate insulation debris is generated with this maximum quantity of fibrous debris. This break also transports 11773 ft of RMI foil debris to the 2 pool. 6.1.3.2 Reflective Met.allic Insulation Debris Calculation note QDC-0010-M-0394 presents an estimate of 19405 ft' for the largest quantity of RMI foil debris transported to the suppression pool from the drywell of the QC2. Note that all RMI foil debris estimates are increased by 10% over the calculated values to account for damage to the RMI cassettes (Ref. 5.2). ' The sources of information for each NDIT appear la the list of References in Section 5.0 2 These values are estimated based on a plot provided in (Ref. 5.14) ~ ~ ' ~ ~ ' ~ REVISION NO. l 0 l 1 l l w-

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% 1 .PAGE M )6g Niih: L M 6 '*

• ON!.Y 6.1.3.3 Calcium Silicate Insulation Debris Calculation note QDC-0010-M 0394 presents a conservative estimate of 0.78 ft' for the quantity of calcium silicate insulation debris transported to the suppression pool from the drywell of the QC2 for every analyzed break.

6.1.3.4 Particulate Debris Calculation note QDC 0010-M-0395 (Ref. 5.3) estimates conservative quantiti:s for particulate debris, j composed of sludge and drywell particulate matter, in the QC2 suppression pool. The values are i presented in Table 3. Table 3. Quantity of Particulate Debris in the Quad Cities Station Unit 2 Suppression Pool Debris Type Maa (Ib) Dirt / Dust 150 Rust Flakes 50 Qualified Paint or Other Surface Coating 85 Unqualified Paint or Other Surface Coating 85 Total Drywell Particulate Debris 370 Suppression Pool Sludge 443 6.1.4 Debris Characteristics ne NUREG/CR-6224 head loss correlation considers each type of debris by specifying the fiber diameter, the as-fabri:.ated (or macroscopic) and the material (or microscopic) fibrous material densities, and the characteristic sizes and average microscopic densities of suppression pool sludge and drywell particulate matter. The following paragraphs present the proposed debris characteristics in this calculation. The material (or microscopic) density of NUKON fiberglass insulation is 175 lb/ft' (2800 kg/m') and the macroscopic pack density of this material is 2.4 lb/ft' (38 kg/m') (Ref. 5.24). The SEM analysis of NUKON" fiberglass debris (Ref. 5.21) shows that the diameter of the fibers is fairly uniform and .approximately equal to 7.1 m. The microscopic density of sludge, which is basically iron oxide, is 324 lb/ft' (5190 kg/m') (Ref. 5.24). He mass median diameter of the sludge particle size distribution is estimated to be 2.5 pm (Ref. 5.17). His value represents the size distribution of the sludge in the suppression pool. However, the size distribution of the sludge particles actually deposited on the fibers in the debris bed has a mass median diameter much larger than the corresponding mass median diameter of the sludge particles in the suppression pool, as suggested by the SEM photographs of typical debris beds (Ref. REVISION NO. l 0 l 1 l l

NEP-12-02 e Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-03% PAGE NO.17 5.21), which show particle sizes in the order of 100 pm. Consequently, in these calculations an average debris bed sludge particle size of 10 pm will conservatively be used. The microscopic density of calcium silicate is reported to be 143 lb/ft' (2300 kg/m') in European studies (Ref. 5.4). Rese studies also indicate that the majority of the calcium silicate debris is greater than 0.85 mm in diameter, with less than 1% smaller than 20 m. In this calculation, a characteristic size of about.40 pm (1.2x10" ft) will conservatively be used for a typical calcium silicate debris particle. As indicated in Appendix E to " Testing of Alternate Straine.rs With Insulation Fiber and Other Debris" (Ref. 5.7), the macroscopic density of calcium silicate is 16.3 lb/ft'. In the absence of more detailed information, a microscopic density of dirt / dust of 156 lb/ft' (2500 kg/m') (Ref. 5.24) will be used. An average equivalent diameter of 10 m, based on a typical diameter of dust particles (Ref. 5.19), will be used in this calculation. In general, the following types of coatings are found inside the primary containment of BWR nuclear plants: inorganic Zinc, epoxy, and alkyd. The microscopic densities of these materials (based on the specific gravity values reported (Ref. 5.1)) are: 90 lb/ft' (1430 kg/m') for epoxy, 94 lb/ft' (1500 kg/m') for alkyd, and 156 lb/ft' (2500 kg/m') for inorganic Zine. In the absence of specific details about the paint / coatings chips in QC2, an average microscopic density of 124 lb/ft' will be used in these calculations (Ref. 5.24). He thickness of the paint chips will be a function of the coating thickness in the drywell.- A typical lower bound for such coatings is 1 mil. To account for the uncertainty in this value, particularly in the case of unqualified coatings, a characteristic size of 0.69 mil will conservatively be used in these calculations. ' Rust flakes will be considered as iron oxides, with a microscopic density of 324 lb/ft' (5190 kg/m'). Since rust flakes appear to be visually similar to paint chips, an equivalent diameter of 0.69 mil (17 m) will conservatively be used for the characteristic size. The debris characteristics used in this calculation are summarized in Table 4. Table 4. Quad Cities Station Unit 2 Debris Characteristics Debris Type Microscopic Density Characteristic Size Ob/ft') (ft) lum] Fibers 175 2.3x10 [7.11 Calcium Silicate 143 1.2x104 [36.6] Sludge 324 3.3 x 10[10] Drywell Particles Dirt / Dust 156 3.3x10 5 [10] Rust Flakes 324 5.7x10-5 [37j Paint Chips 124 5.7x10-8 [17] ... ~. - _ - '.7;}7^T- ^, REVISION NO.- l 0 l 1 l ~ j- -~~

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO.18 l 6.1.5 Considerations in Estimating Head IAssis 6.1.5.1 Debris Sedimentation In the base case calculations, gravitational sedimentation of sludge, dirt / dust, calcium silicate particles j ~ and fibrous debris in the suppression pool will be conservatively neglected. Paint chips, however, l have an average terminal settling velocity of 0.3 ft/s (9x104 m/s) (Ref. 5.24) which is about a factor of 30 greater than the average terminal settling velocity for sludge particles, i.e., 0.01 ft/s (3x10-' m/s) (Ref. 5.24). Thus, neglecting gravitational sedimentation of paint chips is considered to be overly conservative. In the NUREG/CR-6224 study, it was judged that, after cessation of the high-energy phase in a' suppression pool during a LOCA, the settling rates will not be lower than 50% of those corresponding to the measured terminal settling velocity for quiescent pools [Ref. 5.24, p.B-30). To have an upper bound, a settling velocity 10 times lower than settling velocity for quiescent pools, or 0.03 ft/s, will be assumed (see section 3.0). Based on these considerations, the removal fraction of L 0.67 (67%) fmaint chips and rust flakes, calculated in section 6.1, is used in these base calculations. No experimental data are available to estimate the sedimentation rate of rust flakes. However, with a characteristic particle size comparable to paint chips and density a factor between 2 and 3 higher, one would expect a greater rate of sedimentation for rust. For conservatism, the same factor 0.67 (67%) proposed for the sedimentation of paint chips (67%), will be used in this analysis. 6.1.5.2 Debris Mitration l Not all of the debris particles reaching the strainer would be trapped or filtered by the strainer to form l a debris bed on the strainer surface. 'Ihe fraction of the debris particles approaching the strainer that is l . deposited and contained in the fibrous debris bed is referred to as the filtration efficiency (Ref. 5.24). Several experiments were conducted by the NRC to provide bounding estimates for the filtration 1 - efficiency of sludge particles (Ref. 5.21). Based on these experiments, a conservative upper-bound value of 0.50 was used for the sludge particle filtration efficiency for debris bed thicknesses higher than 0.25 inches in the NUREG/CR-6224 analysis; fot thicknesses lower than 0.25 inches, 0.50 filtration efficiency was deemed overly conservative and a linear variation for the filtration efficiency from 0 to 0.5 was used for theoretical thicknesses lower than 0.25 inches [Ref. 5.24, p. B 34]. Consequently, in this calculation the NUREGICR-6224 filtration efficiency model for sludge particles is used.

The characteristic sizes presented in Table 3 suggest that the dominant filtration mechanisms for dust and sludge particles are impaction and interception. For these mechanisms, the filtration efficiency is essentially the same for particles with diameters between 2 and 10 pm (Ref. 5.9). On this basis, this calculation uses the same efficiency model for sludge and dirt / dust particles, i.e., an efficiency 0.5 for theoretical debris bed thicknesses lower than 0.25 inches and a linear variation for the filtration efficiency from 0 to 0.5 for theoretical thicknesses lower than 0.25 inches.

REVISION NO. l 0 l 1 l l l

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010 M-03% PAGE NO.19 l To account for the fact that the calcium silicate debris, paint chips and rust flakes are larger than the sludge particles, a filtration efficiency of 1.0 is used for these drywell particles with debris bed thicknesses higher than 0.25 inches; for beds thicknesses lower than 0.25 inches, a filtration efficiency of 0.5 for these drywell particles will be conservatively used in this calculation. 6.1.6 Debris Quantities for the Base Case Analyses Table 5 summarizes the debris quantities for the base case analysis. The flow conditions specified in Table 2 are used, i Table 5. Quad Cities Station Unit 2: Debris Quantities Fibrous Debris 6.74 ft' Calcium Silicate Debris 0.78x0.5 ft' (a) RMI Foll Debris 19405 ft' Sludge 443x0.2 lb (a) Dirt / Dust 150x0.2 lb(a) Rust Flakes 50x0.33x0.5 lb (a,b) Total Paint Chips (Qualified + Unqualified) 170x0.33x0.5 lb (a,b) (a) Redus. ion fraction due to filtration (b) Reduction fraction due to sedimentation 6.2. Supporting Calculations 6.2.1 Strainer Circumscribed Area The strainer circumscribed area, A , is just the surface area of the cylinder, including the end plates, enveloping the strainer. The value in Table 1 is calculated as follows: A,,,, = fr D,L + 1 (2 D,2 _ p,2)'1/f'

where, D = 45" is the maximum outside diameter, D, = 20" is the inside core tube diameter and L = 42" is the active length.

Substituting numerical values, the circumscribed area is calculated to be 61 ft'. FOR REFERENCE -*,esr REVISION NO. l 0 l 1 l V kN L. t l l

[- ) i NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010.M-03% PAGE NO. 20 6.2.2 Gap Volume 'Ibe gap volume, V,, is the total volume within the gaps between the strainer disks. The value in l Table 1 is calculated as follows: l V = hD*2 - D,2) d (N - 1) 1728 in' 8 4 8

where, D, = 24. 5" is the gap diameter, l

d, = 2" is the gap width and N = 11 is the number of disks. Substituting numerical values, the gap volume is calculated to be 13 ft'. 6.2.3 Sedimentation Fraction i To estimate the fraction of 0.33 (33%) for the quantity of paint chips deposited in the fibrous debris i bed on the strainer, the method described in the following paragraphs is used. Consider a water volume in the suppression pool, V of 3177 m' (112200 ft') (Ref. 5.13), the g, maximum credible flow rate through the strainers, Or, of 2 m' /s (4 x 8350 gpm), the maximum water I level in the suppression pool, h,,,,, of 4.5 m (14.88 ft), and an average paint chip sedimentation velocity, y,,,,10 times lower than the terminal settling velocity measured for quiescent pools,9 x 10-2 m/s (Ref. 5.24), i.e., v,,, = 9 x 10 m/s. The characteristic sedimentation time, t,,,, is defined by: '~' t,,s = sed Substituting numerical values, the characteristic sedimentation time is calculated to be: t,,, = 500 s The characteristic suppression pool turn-over time, t,,,,, is defined by: ) t'"' ~ Gr I 1, ) g .,7 REVISION NO. l 0 l 1 i l --l- _.s .r. -

NEP-12-02 Revision 7 i COMMONWEALTH EDISON COMPANY i CALCULATION NO. QDC-0010-M-0396 PAGE NO. 21 l Substituting numerical values, the characteristic suppression pool turnsver time is calculated to be: tg = 1600 s. Based on these considerations, the characteristic sedimentation time is approximately a factor of 3 lower than the characteristic suppression pool turn-over time, l The total quantity of paint chips debris in the pool, m, is given by: r i m = m + m, 7 y where mf and m, are the quantities of paint chips deposited into the pool floor and approaching the strainer, respectively, Now, the quantity of paint chips debris that is deposited into the pool floor by sedimentation is inversely proportional to the sedimentation time, whereas the quantity of paint chips debris approaching the strainer, is inversely proportional to the characteristic pool turn-over time (which was estimated to be 3 times lower than the sedimentation time). ' Ibis results in: m = 3 m, f Substituting this result in the equation for m gives: r '+ 2-=1 m m 7 7 1 => m, / m = 0.25, or m,Im = 0. 75 r r This suggests that approximately 3/4 (75%) of the paint chips would settle to the suppression pool l floor To have a conservative bound, this calculation considers that 67% of the paint chips settle to the j suppression pool floor. 6.3 Base Case Analysis l Based on the maximum quantities of each type of insulation (fibrous, metallic and calcium silicate) described in Section 4.0, the case presented in Table 6 will be analyzed in this calculation note. All quoted head losses will be reported in ft-water at 60'F. h REVISION NO. l 0 l 1 l l L

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY l CALCULATION NO. QDC-0010.M-0396 PAGE NO. 22 Table 6. Quad Cities Station Unit 2: Conditions for the Base Case Analysis Fibrous Debris 6.74 ft' RMI Foil Debris 19405 ft' Calcium Silicate Debris 0.39 ft' Sludge 87lb Dirt / Dust 30 lb i Rust Flakes 8 lb Paint Chips 28 lb 6.3.1 Maximum Mbrous Debris Case As indicated in Table 6, the largest quantity of fibrous debris expected in the QC2 pool is 6.74 ft'. 'Ihe theoretical (or as-fabricated) bed thickness, d,, associated with this quantity of fibrous debris is i calculated as follows: V 12 in r A L* = N, A, x ft

where, l

V = 6.74 ft' is the largest quantity of fibious debris in the pool, f l A, = 207 ft is the surface area of the strainer and 2 N, = 4 is the number of strainers in QC2. Substituting numerical values, the theoretical debris bed thickness is calculated to be: E, = 0.098" This theoretical thickness is less than 0.125" (1/8 of an inch), which is the minimum debris bed thickness required to bridge the holes in the strainer (1/8 of an inch). At the calculated theoretical bed thickness, a uniform debris bed can not be sustained on the surface of the strainer and some of the fibrous material will penetrate the holes in the strainer. Visual observations during tests on stacked disk strainers (Ref. 5.10) suggest that, at the beginning, the fibrous debris initially accumulates primarily within the gap volume between the disks, and later deposits on the remaining surface area of the strainer. As a result of this, the expected head losses with theoretical bed thickness less than 0.125" are expected to be negligible. As a matter of fact, the EPRI test P4 (Ref. 5.7) shows that, under these circumstances (d. = 0.088"), a non-measurable head loss was obtained. In particular, the base case at QC2 considers 1.7 ft' of fibrous debris per strainer (6.74 ft' of fibrous debris in the pool), compared with the 13 ft' of volume available in the gaps between the disks. Consistent with the visual observations during the tests (Ref. 5.10), it is estimated that most of the 1.7 - - - m, f,- REVISION NO. l 0-l L fV (, K C F Cti C[N b L ON LY

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 23 ft' fibrous debris will be deposited within the 13 ft' of gap volume, leaving most of the strainer surface area without debris. Consequently, the expected head loss due to a combination of fibrous debris and l particulate maner for ce QC2 strainers is estimated to be negligible. For completeness, however, the head loss per strainer predicted by BLOCKAGE 2.5 for the combination of fibrous de'. 't and particulate matter indicated in Table 6 is presented in Figure 1. c 0.45 ~ 0.4-C 0.35 - 2 l 0.3 - E 0.25 - e80.2 I y 0.15 ' Base Case j

E 0.1 -

0.05 d 0 0 Sn00 10000 15000 20000 25000 Time (s) Mgure 1. Quad Cities Station Unit 2: BLOCKAGE 2.5 calculated head loss per strainer due to 6.74 ft' cf fibrous debris,443 lb of sludge,150 lb of dirt / dust,50 lb of rust flakes and 170 lb of paint chips J in the pool. As indicated in Figure 1, BLOCKAGE 2.5 predicts a maximum head loss of about 0.42 ft-water at approximately 6000 s into the accident. Since there is not enough fibrous debris to cover the strainer, the expected head loss due to fibrous debris is actually 0 't-water'. The over-estimation in the BLOCKAGE 2.5 results for thin beds is basically a resul: of the assumption of considering a uniform and homogeneous distribution of debris on the surface of the strainer. In reality, with the open areas in the debris bed, the flow may pass through relatively clean regions in the strainer, thereby resulting j in much lower head losses than those predicted by BLOCKAGE 2.5. l ' Ncte that this estimate does not consider the contribution of RMI foil debris to the he:d loss; such contribution is estimated in Section 6.2.2 I REVISION NO. l 0 l 1 l l -. 2

i NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 24 6.3.2 Maximum RMI Foil Detris Case Satwated RMI Debris Bed fiend 14ss at a Flow Rate of 8,350 gpm Basul on the methodology described in Section 2.1.3, the eatimated volume (Vu ) and corresponding ~ foil area (A,,a) to cause a saturated bed at a flow rate of 8,350 gpm on one strainer is dep.icted in Tatle 7 for the two types of RMI analyzed. De settling velocity and the thickness constants used are also presented. l - Table 7 RMI Debris Saturation Bed Quantities - 8,350 rpm RMI Debris U. Vans l A. Type (ft/s) (ft) (ft') (ft') 2/2.5 mil SS 0.39 0.014 50 3,W 6 mil Al 0.25 0.073 132 1,815 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.58 ft-water ti mil aluminum 0.30 ft-water The highest RMI saturated bed head loss at a flow rate of 8,350 gpm is 0.58 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved during the first 10 minutes of ECCS operation. Saturated RMI Debris Bed Head Inst at a Row Rate of 7,250 gpm The estimated volume and corresponding' foil area to cause a saturated bed on one strainer at a flow rate of 7,250 gpm is depicted in Table 8 for the two types of RMI analyzed. The settling velocity and the thickness constants used are also presented. Table 8 RMI Debris Saturation Bed Quantities - 7,250 gpm RMI Debris U. 5. Vana A Type (ft/s) (ft) (ft') (ft') 2/2.5 mit SS 0.39 0.014 34 2,410 6 mit Al 0.25 0.073 100 1,373 ' Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously ' described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: t: FOR RFt R NCE 2!2.5 mil stainless stee10.30 ft-wa:er P 6 mil aluminum 0.17 ft-water 0N LY REVISION NO. l 0 l 1 - I

NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M 0396 PAGE NO. 25 De highest RMI saturated bed head loss at a flow rate of 7,250 gpm is 0.30 ft-water due to 2/2.5 mil stainless steel RMI debris. His corresponds to the maximum head loss that could be achieved after the first 10 mir.utes of ECCS operation. 6.3.3 Maximum Calcium Silicate Debris Case - Given the lack of sufficient fibrous debris to form a homogeneous bed, the calcium silicate debris will not contribute to a head loss across the strainer. 6.3.4. Maximum ParticulatHo-Fiber Mass Ratio As indicated in Section 6.2.1, even the largest quantity of fibrous debris expected in the QC2 pool will not be sufficient to cover the strainer surface. To consider a smaller quantity of fibrous debris, increasing therefore the particulate-to-fiber mass ratio with respect to that in the base case (estimated to be approximately 9), will result in even thinner debris bed thickness. To consider more fibrous debris will result in a lower particulate-to-fiber mass ratio than that considered in the base case. Consequently, this maximum particulate-to-fiber mar,s ratio case is not applicable to QC2. 6.4 Assessment of Margin As discussed in Section 6.2, even considering conservative assumptions the estimated head loss is no worse than maximum RMI head loss of 0.58 ft-water, a value which is already close to the minimum reading of the instruments typically used la head loss measurements. Relaxing some of these conservative assumptions, such as those related with sedimentation and filtration of debris, will result in even smaller head losses. Consequently, no assessment of margin will be conducted for QC2. 6.5 Parametric Analysis 6.5.1 Effect of Water Temperature - According to the methods used in this calculation, BLOCKAGE 2.5 for head losses due to fibrous insulation and the URG based methodology for head losses due to RMI foil debris, the temperature of the water la the suppression pool only affects the head loss due to fibrous debris. As indicated in Section 6.2.1, in the worst case for fibrous debris in the QC2 pool the head loss is negligible. The j temperature of the water only affects the water properties used in the head loss model, i.e., the dyr.amic viscosity and density. A change in water temperature will not affect the theoretical debris bed i thickness of 0.098" (which, as shown in Section 6.2.1, is less than the minimum required to produce a measurable head loss). Therefore, the predicted head loss due to fibrous debris will still be negligible. The SER based RMI head loss correlation is independent of the water temperature. Consequently, any change in the pool water temperature will result in the same maximum estimated head loss of 0.58 ft-water due to RMI foil debris. ) i REVISION NO. l 0 l 1 i l l !g __ ___ __ _

i NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPAIW CALCULATION NO. QDC 0010 M 0396 PAGE NO. 26 6.5.2 Effect of Strainer Surface Area Reduction The impact on the head loss of a reduction in the surface area of strainera by 2 ft' for each strainer, due to deposition of extraneous materials such as plastic bags, can be investigated by considering a strainer surface area of 205 ft' and the debris quantities presented in Table 6. In this situation, the theoretical debris bed thickness is estimated to be AL, = 0.01", which is still below the minimum theoretical thickness of 0.125" required to sustain a fibrous debris bed. Consequently, the head loss due to fibrous debris is estimated to be negligible. Saturated RMI Debris Bed Head less at a Flow Rate of 8,350 gym Based on the methodology described in Section 2.1.3, the estimated volume (Vm) and corresponding foil area (Ar.a) to cause a saturated bed at a flow rate of 8,350 gpm on one reduced surface area strainer is depicted in Table 9 for the two types of RMI analyzed. 'Ihe settling velocity and the thickness constants used are also presented. Table 9 RMI Debris Saturation Bed Ouantities - 8,350 em RMI Debris U. Vm Am Type (ft/s) (ft) (ft') (ft') 2/2.5 mil SS 0.39 0.014 54 3,907 6 mil Al 0.25 0.073 141 1,934 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.70 ft-water 1 6 mil aluminum 0.35 ft-water The highest RMI saturated bed head loss on one reduced surface area strainer at a flow rate of 8,350 gpm is 0.70 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved during the first 10 minutes of ECCS operation. Saturated RMI DeMs Bed Head lass at a Row Rate of 7,250 gpm The estimated volume and corresponding foil area to cause a saturated bed on one reduced surface area strainer at a flow rate of 7,250 gpm is depicted in Table 10 for the two types of RMI analyzed. The settling velocity and the thickness constants used are also presented. Table 10 RMI Debris Saturation Bed Quantities - 7,250 mm RMI Debris U, N Vm Au Type (ft/s) (ft) (ft') (ft') 2/2.5 mil SS 0.39 0.014 37 2,666 6 mil Al 0.25 0.073 107 1,469 REVISION NO. l 0 l 1 l l

NEP-12-02 Revision 7 COMMONWEALTH EDISCN COMPANY CALCULATION NO. QDC-0010 M-0396 PAGE NO. 27 Based on the RMI foil quantities of the above table and the SER RMI head loss correlation previously described, the following are the saturation bed RMI head losses for the two types of RMI analyzed: 2/2.5 mil stainless stee10.36 ft-water 6 mil aluminum 0.20 ft-water De highest RMI saturated bed head loss on one reduced surface area strainer at a flow rate of 7,250 gpm is 0.36 ft-water due to 2/2.5 mil stainless steel RMI debris. This corresponds to the maximum head loss that could be achieved after the first 10 minutes of ECCS operation. This represents an increase in the estimated head loss by approximately 20%. 6.5.3 Quantity of Fibrous Debris N== mary to Produce a Measurable Head Imss As indicated in Section 6.2.1, the largest quantity of fibrous debris expected in the QC2 pool, estimated to be 6.74 ft', is not enough to produce a theoretical debris bed thickness of 0.125", which is the minimum required to result in a measurable head loss, as explained in Section 6.2.1 and demonstrated by the EPRI experiments (Ref. 5.7). It may be interesting, however, to estimate what is the quantity of fibrous debris in the pool required to produce this minimum theoretical debris bed thickness of 0.125". This quantity of fibrous debris, V, is given by: I ft V, = N, A, AL,,,) 12 :. n

where, N, = 4 is the number of strainers, 2

A, = 207 ft is the surface area of each strainer and A = 0.125" is the minimum theoretical thickness to produce a measurable head loss. Substituting numerical values, this minimum quantity of fibrous debris in the QC2 pool required to result in a measurable head loss is estimated to be: V, = 8.63 ft' Note, however, that this value represents just the quantity of fibrous debris in the pool that may result in a measurable (i.e.,

• O ft-water) head loss. Bis quantity of fibrcus debris in the pool, distributed on each of the strainers at QC2 results in about 2.2 ft', a quantity that is expected to by primarily deposited in the gap volume of 13 ft' per strainer, leaving.nost of the remaining surface area of the strainer relatively clean of debris. Consequently, this estimated minimum quantity of fibrous debris in the pool to produce a measurable head loss is still extremely conservative.

I rnn n c r.PD !: NCE REVISION NO. l 0 l 1 i V P O " ',; ;, l U N t. k L

NEP-12-02 Revisbn 7 COMMONWEALTH EDISON COMPANY CALCULATION NO ODC-0010-M-0396 l PAGE NO. 28 7.0

SUMMARY

AND CONCLUSIONS In this calculation note, the following cases were analyzed: Base Case: 6.74 ft' of fibro a debris,19405 ft' of RMI foil debris, 0.78 ft' of calcium silicate debris, 443 lb of sludge,150 lb of dirt / dust,50 lb of rust flakes and 170 lb of paint chips in the pool. l Parametric Case 1: Same conditions as in the Base Case, but considering a reduction in the pool water temperature by 35'F. Parametric Case 2: Same conditions as in the Base Case, but considering a reduction in the strainer surface area by 2 ft' due to extraneous materials, i.e., plastic bags. Parametric Case 3: Estimation of the minimum quantity of fibrous debris in the pool required to l produce a measurable head loss. The largest quantity of fibrous debris expected in the pool at QC2,6.74 ft', results in a theoretical bed thickness of 0.098", which is smaller than the minimum theoretical bed thickness required to produce a measurable head loss, i.e.,0.125". Reducing the quantity of fibers, to maximize the particulate 40-fiber mass ratio, would result in an even thinner bed thickness and, therefore, this case was not considered applicable to QC2. He conditions and corresponding head losses estimated for each of these cases are presented in Table 7. Table 7. Quad Cities Station Unit 2: Summary of Head Loss Estimates Case Fibrous RMI Cal Sil Sludge *) Dirt / Rust Paint Head Debris Debris ") Debris *) Ob) Dust *) Flakes *") Chips *') Loss (ft') (ft') (ft') Ob) Ob) Ob) (ft-water) Base 6.74 3,590 0.39 87 30 8 28 0.58 1 6.74 3,590 0.39 87 30 8 28 0.58 2 6.74 3,590 0.39 87 30 8 28 0.70 3 8.63 3,590 0.39 87 30 8 28 0.31* De highest head loss estimated for QC2, considering conservative assumptions, is 0.70 ft-water and is due solely to the contribution of RMI foil debris, i.e., the contribution of fibrous debris is negligible. This highest head loss occurs during the first ten minutes of ECCS operation. Consequently, the impact of relaxing conservative assumptions in the estimation of the head loss due to fibrous debris was not further investigated in this calculation. a) This is the quanuty of RMI debris required to produce the saturation bed duckness (i.e., maximum head loss due to RMI debris) for 2/2.5 mil stamless steel foils.- b) Considering debris fikrataon. c) Considering debris sedimentstaon, d) Estimated quahtatively. 7,, ~ m ~,7 y T z a l REVISION NO. l 0 l 1 l l i

h ] NEP-12-02 Revision 7 COMMONWEALTH EDISON COMPANY CALCULATION NO. QDC-0010-M-0396 PAGE NO. 29 he most relevant conclusions are summarized as follows: L 1. He largest quantity of fibrous debris expected in the pool of QC2, 6.74 ft', is not enough to produce the minimum theoretical bed thickness required to result in a measurable head loss due to fibrous debris. 2. ~ Metallic debris produces the worst head loss results. He highest head loss for RMI foil debris occurs at the saturation bed thickness. He quantity required to produce saturation bed thickness, j estimated to be 3,590 ft2 fer QC2, results in a head loss no worse than 0.6 ft-water. 3. The limited parametric analysis conducted suggests that the impact on the head loss estimates due to a reduction in the surface area of each strainer by 2 ft', such as that originated by deposition of i extraneous materials on the strainers, would increase the estimated head loss by approximately 20%. 4. He minimum amount of fibrous debris in the pool required to produce a measurable head loss is estrmated to be 8.63 ft', a quantity that is approximately 28% larger than the maximum quautity of I fibrous debris expected to be transported to the QC2 suppression pool. His quantity is still very conservative, because it results in about 2 ft' of fibrous debris per straintr. Based on visual observations during tests conducted on stacked disk strainers, most of these debris are expected to be accumulated primarily within the 13 ft' of volume available within the gaps between the stacked disks, leaving a considerable area of the strainers relatively " clean' from debris, resulting thereby in very low head losses. REVISION NO. l 0 l 1 _l l ~ ..}}