ML101300225

Draft Responses to NRC RAI in Preparation for Meeting on May 12 and 13, 2010 - GL 2004-02

ML101300225
Person / Time
Site:	Palisades
Issue date:	05/07/2010
From:	Mahesh Chawla Plant Licensing Branch III
To:
Chawla M, NRR/DORL, 415-8371
References
Download: ML101300225 (148)

Text

Palisades Draft RAI Responses for May 2010 Public Meeting 1 NRC Request1. Entergy Nuclear Operations, Inc. (the licensee) stated in its submittal datedJune 30, 2009, that the break selection process was re-performed afterfibrous debris zones of influence (ZOI) were reduced. However, the licenseefurther stated that after a significantly large fiberglass component wasidentified, deference was given to evaluating breaks for other debris sources.The Nuclear Regulatory Commission (NRC) staff cannot determine from thisinformation whether the break selection was conservative. Please verify andjustify that the break selection process identified the break that results in themaximum potential fibrous debris load and that this debris load wasconsidered in the remaining portions of the head loss evaluation.Entergy Nuclear Operations Response:The break selection process identified Break S5 as the maximum fibrous debrisload which was used in subsequent head loss evaluations and strainer testing.Use of Break S5 as the limiting break for fiber is detailed in several locationswithin Palisades 6/30/2009 submittal. The break selection process utilized theguidance of NEI 04-07 in deriving the maximum fibrous debris load. Thefollowing is an excerpt from Reference 1.1, Section 4.2 Break Locations:Identifying the break locations is the first step in determining the LOCAgenerated debris. The break selection process is described in the NEIMethodology in Section 3.3.4 (Reference 6.1.1). Per Section 1.3.2.3 of Reg.Guide 1.82 (Reference 6.1.6) at minimum, the following postulated breaklocations were considered: Breaks in the reactor coolant system (e.g., hot leg, crossover leg, cold leg,pressurizer surge line) and, depending on the plant licensing basis, mainsteam and main feedwater lines with the largest amount of potential debriswithin the postulated ZOI, References1.1 EA-MOD-2005-04-06 Revision 3 dated 2/9/2009, "Acceptance of DebrisGeneration Calculation 2005-01340 Rev 2" Palisades Draft RAI Responses for May 2010 Public Meeting 2 NRC RequestNote: Questions 2 through 9 below are being addressed in whole or in part bythe Pressurized Water Reactor Owners Group. The NRC staff expects thedegree to which the Owners Group is able to generically resolve these issues tobe clear by the time the licensee's responses to these requests for additionalinformation (RAIs) are due. As appropriate, the licensee may respond to theseRAIs with reference to NRC staff official correspondence on resolution of thestaff's questions of the Owners Group. In any event, some of the questions willneed a plant-specific response.2. The supplemental response dated June 30, 2009, credited a reduced ZOIfor low-density fiberglass. Please describe the jacketing/insulationsystems used in the plant for which the testing was conducted andcompare those systems to the jacketing/insulation systems tested.Demonstrate that the tested jacketing/insulation system adequatelyrepresented the plant jacketing/insulation system. The description shouldinclude differences in the jacketing and banding systems used for pipingand other components for which the test results are applied, potentiallyincluding steam generators, pressurizers, reactor coolant pumps, etc. At aminimum, the following areas should be addressed:a. How did the characteristic failure dimensions of the testedjacketing/insulation compare with the effective diameter of the jet at theaxial placement of the target? The characteristic failure dimensionsare based on the primary failure mechanisms of the jacketing system,e.g., for a stainless steel jacket held in place by three latches where allthree latches must fail for the jacket to fail, then all three latches mustbe effectively impacted by the pressure for which the ZOI is calculated.Applying test results to a ZOI based on a centerline pressure forrelatively low L/D nozzle to target spacing would be non-conservativewith respect to impacting the entire target with the calculated pressure.b. Was the insulation and jacketing system used in the testing of thesame general manufacture and manufacturing process as theinsulation used in the plant? If not, what steps were taken to ensurethat the general strength of the insulation system tested wasconservative with respect to the plant insulation? For example, it isknown that there were generally two very different processes used tomanufacture calcium silicate whereby one type readily dissolved inwater but the other type dissolves much more slowly. Suchmanufacturing differences could also become apparent in debrisgeneration testing, as well.c. The information provided should also include an evaluation of scalingthe strength of the jacketing or encapsulation systems to the tests. Forexample, a latching system on a 30 inch pipe within a ZOI could be Palisades Draft RAI Responses for May 2010 Public Meeting 3stressed much more than a latching system on a 10 inch pipe in ascaled ZOI test. If the latches used in the testing and the plants arethe same, the latches in the testing could be significantly under-stressed. If a prototypically sized target were impacted by anundersized jet it would similarly be under-stressed. Evaluations ofbanding, jacketing, rivets, screws, etc., should be made. For example,scaling the strength of the jacketing was discussed in the OntarioPower Generation (OPG) report on calcium silicate debris generationtesting.3. There are relatively large uncertainties associated with calculating jetstagnation pressures and ZOIs for both the test and the plant conditionsbased on the models used in the Westinghouse Commercial Atomic Power(WCAP) reports the licensee used to justify reduced ZOIs. What stepswere taken to ensure that the calculations resulted in conservativeestimates of these values? Please provide the inputs for thesecalculations and the sources of the inputs.4. Describe the procedure and assumptions for using the ANSI/ANS-58-2-1988 standard to calculate the test jet stagnation pressures at specificlocations downrange from the test nozzle.a. Was the analysis based on initial conditions (temperature) thatmatched the initial test temperature? If not, please provide anevaluation of the effects of any differences in the assumptions.b. Was the water subcooling used in the analysis that of the initial tanktemperature or was it the temperature of the water in the pipe next tothe rupture disk? Test data indicated that the water in the piping hadcooled below that of the test tank.c. The break mass flow rate is a key input to the ANSI/ANS-58-2-1988standard. How was the associated debris generation test mass flowrate determined? If the experimental volumetric flow was used, thenexplain how the mass flow was calculated from the volumetric flowgiven the considerations of potential two-phase flow and temperaturedependent water and vapor densities? If the mass flow wasanalytically determined, then describe the analytical method used tocalculate the mass flow rate.d. Noting the extremely rapid decrease in nozzle pressure and flow rateillustrated in the test plots in the first tenths of a second, how was thetransient behavior considered in the application of the ANSI/ANS-58-2-1988 standard? Specifically, did the inputs to the standard representthe initial conditions or the conditions after the first extremely rapidtransient, e.g., say at one tenth of a second?

Palisades Draft RAI Responses for May 2010 Public Meeting 4e. Given the extreme initial transient behavior of the jet, justify the use ofthe steady state ANSI/ANS-58-2-1988 standard jet expansion model todetermine the jet centerline stagnation pressures rather thanexperimentally measuring the pressures.5. Describe the procedure used to calculate the isobar volumes used indetermining the equivalent spherical ZOI radii using the ANSI/ANS-58-2-1988 standard.a. What were the assumed plant-specific reactor coolant system (RCS)temperatures and pressures and break sizes used in the calculation?Note that the isobar volumes would be different for a hot leg breakthan for a cold leg break since the degrees of subcooling is a directinput to the ANSI/ANS-58-2-1988 standard and which affects thediameter of the jet. Note that an under calculated isobar volume wouldresult in an under calculated ZOI radius.b. What was the calculational method used to estimate the plant-specificand break-specific mass flow rate for the postulated plant loss-of-coolant accident (LOCA), which was used as input to the standard forcalculating isobar volumes?c. Given that the degree of subcooling is an input parameter to theANSI/ANS-58-2-1988 standard and that this parameter affects thepressure isobar volumes, what steps were taken to ensure that theisobar volumes conservatively match the plant-specific postulatedLOCA degree of subcooling for the plant debris generation breakselections? Were multiple break conditions calculated to ensure aconservative specification of the ZOI radii?6. Provide a detailed description of the test apparatus specifically includingthe piping from the pressurized test tank to the exit nozzle including therupture disk system.a. Based on the temperature traces in the test reports it is apparent thatthe fluid near the nozzle was colder than the bulk test temperature.How was the fact that the fluid near the nozzle was colder than thebulk fluid accounted for in the evaluations?b. How was the hydraulic resistance of the test piping which affected thetest flow characteristics evaluated with respect to a postulated plantspecific LOCA break flow where such piping flow resistance would notbe present?

Palisades Draft RAI Responses for May 2010 Public Meeting 5c. What was the specified rupture differential pressure of the rupturedisks?7. If the application of the reduced ZOI is applied to components other thanpiping, please respond to this question. Please provide the basis forconcluding that a jet impact on piping insulation with a 45° seamorientation is a limiting condition for the destruction of insulation installedon steam generators, pressurizers, reactor coolant pumps, and other non-piping components in the containment. For instance, considering a breaknear the steam generator nozzle, once insulation panels on the steamgenerator directly adjacent to the break are destroyed, the LOCA jet couldimpact additional insulation panels on the generator from an exposed end,potentially causing damage at significantly larger distances than for theinsulation configuration on piping that was tested. Furthermore, it is notclear that the banding and latching mechanisms of the insulation panels ona steam generator or other RCS components provide the same measure ofprotection against a LOCA jet as those of the piping insulation that wastested. One WCAP reviewed asserts that a jet cannot directly impact thesteam generator, but will flow parallel to it. It seems that some damage tothe SG insulation could occur near the break, with the parallel flow thenjetting under the surviving insulation, perhaps to a much greater extentthan predicted by the testing. Similar damage could occur to othercomponent insulation. Please provide a technical basis to demonstratethat the test results for piping insulation are prototypical or conservative ofthe degree of damage that would occur to insulation on steam generatorsand other non-piping components in the containment.8. Some piping oriented axially with respect to the break location (includingthe ruptured pipe itself) could have insulation stripped off near the break.Once this insulation is stripped away, succeeding segments of insulationwill have one open end exposed directly to the LOCA jet, which appears tobe a more vulnerable configuration than the configuration tested byWestinghouse. As a result, damage would seemingly be capable ofpropagating along an axially oriented pipe significantly beyond thedistances calculated by Westinghouse. Please provide a technical basis todemonstrate that the reduced ZOIs calculated for the piping configurationtested are prototypical or conservative of the degree of damage that wouldoccur to insulation on piping lines oriented axially with respect to the breaklocation.9. At least one WCAP noted damage to the cloth blankets that cover thefiberglass insulation in some cases resulting in the release of fiberglass.The tears in the cloth covering were attributed to the steel jacket or the testfixture and not the steam jet. It seems that any damage that occurs to thetarget during the test would be likely to occur in the plant. Was thepotential for damage to plant insulation from similar conditions considered?

Palisades Draft RAI Responses for May 2010 Public Meeting 6For example, the test fixture could represent a piping component orsupport, or other nearby structural member. The insulation jacketing isobviously representative of itself. What provides the basis that damagesimilar to that which occurred to the end pieces is not expected to occur inthe plant? It is likely that a break in the plant will result in a much morechaotic condition than that which occurred in testing. Therefore, it wouldbe more likely for the insulation to be damaged by either the jacketing orother objects nearby.Entergy Nuclear Operations Response, RAI's 2-9:The PWR Owners Group (PWROG) provided a response to generic RAIsassociated with reduced ZOI testing on March 5, 2010, (Reference 2.1). Theopen RAI issues and responses in Reference 2.1 were grouped and wordeddifferently than Palisades RAIs 2 through 9, but a direct comparison is notrequired due to discussion that follows. The PWROG has not adequatelyresolved all open issues at this time. Additionally, smaller upstream dimensionsin the test rig than the nozzle diameter were identified during the process ofaddressing the generic RAIs. The NRC staff has provided their conclusions inNRC letter dated 3/31/2010, (Reference 2.2), regarding the current status ofcrediting reduced ZOI WCAP test reports. In summary, the NRC staff concludedthat the small diameter locations upstream of the test nozzle constitute significanttest design errors, and, absent substantial additional information, render allrecommended ZOIs in similar test reports invalid.As Palisades was explicitly crediting reduced ZOI test reports WCAP-16836-Pand WCAP-16710-P cited in Reference 2.2, Palisades will need to re-evaluatethe amount of debris generation without crediting the currently assumed reducedZOI for Nukon and Thermal Wrap. Use of associated ZOIs provided by NEI 04-07 SER is planned along with evaluating the possible exclusion of pressurizerinsulation above the support skirt for a break below the support skirt provided thesupport skirt would physically block any jet associated with the break, which is anallowance provided by References 2.1 and 2.2.

References2.1 PWROG Letter OG-10-84, "PWROG Response to Request for AdditionalInformation Regarding Pressurized Water Reactor Owners Group Basesfor Licensee Debris Generation Assumptions for GSI-191, (PA-SEE-0639Revision 1)," March 5, 2010.2.2 NRC Letter, Jonathan Rowley of NRR to Anthony Nowinowski of the PWROwners Group Program Management Office, "Nuclear RegulatoryCommission Conclusions Regarding Pressurized Water Reactor OwnersGroup Response To Request For Additional Information Dated January Palisades Draft RAI Responses for May 2010 Public Meeting 725, 2010 Regarding Licensee Debris Generation Assumptions For GSI-191," March 31, 2010. (ADAMS Accession Number: ML100570364)

Palisades Draft RAI Responses for May 2010 Public Meeting 8 NRC Request 10.In RAI 2 from the NRC's letter dated December 24, 2008 (ADAMSAccession No. ML083450689), the NRC staff requested informationconcerning debris characteristics for several debris sources listed in theFebruary 27, 2008, supplemental response to GL 2004-02. The licenseeresponded in the June 30, 2009, supplemental response. However, thestaff has the following questions remaining on this response.a. Please provide the basis for concluding that exposure of unjacketedfibrous material to containment spray will not result in the generation ofdebris.b. Please provide the basis for concluding that pieces of Marinite debriswill not transport to the strainers or erode in the post-LOCAcontainment pool. Although Section 4.2.2.2.5 of NEI 04-07 states thatMarinite can be assumed to be broken into large chunks, the staffcould not determine that NEI 04-07 or the accompanying safetyevaluation provides a basis for concluding that this Marinite is notsusceptible to transport or erosion. The staff has seen results fromtesting demonstrating that Marinite does erode when submerged andexposed to flow.Entergy Nuclear Operations Response to 10a:The NRC SER for NEI 04-07 Appendix VI at ml043280016 provides the belowexample material which appear to have been meant as tutorial material.Appendix VI, Detailed Blowdown/Washdown Transport Analysis forPressurized-Water Reactor Volunteer Plant,on page VI-29Table VI-4 summarizes the assumed fractions of fibrous debris that wereeroded. It was assumed that condensate drainage would not cause furthererosion of debris and that intact or covered debris would not erode further.Erosion does not apply to fine debris because that debris is already fine.About 1 percent of the small- and large-piece debris that the spraysdirectly impacted was considered to have eroded. This amount of erosionwas considered to be conservative because the DDTS concluded that theerosion was less than 1 percent. No erosion of the intact debris wasassumed because the canvas cover likely would protect the insulation.Table VI-4. Total Erosion Fractions for Fibrous DebrisExposure FinesSmallLargeIntact Condensate N/A 0 0 0 Sprays N/A 1%1%0 Palisades Draft RAI Responses for May 2010 Public Meeting 9From context the above table applies to material outside its jacket and exposedand shredded by break blowdown effects. Both smalls and larges are assigned1% erosion due to sprays.From the original submittal of June 30, 2009 in Table 3a2, "Summary of DebrisGenerated" it can be seen in the rows for unjacketed Nukon that a 17D ZOI wasassumed and for breaks S1 through S6 the amount of such insulation is 0.8,0.89, 0.8, 0, 0.8 and 1.79 cubic feet. From the repeated numbers it is assumedthat breaks were recounting the same insulation.The row for low density fiberglass unjacketed also used a ZOI of 17D and showsfor the same break set 0.59, 0.59, 0.59, 0.59, and 0.59 cubic feet. Again therepeat numbers are indicative of each break counting the same insulation.Since this set of breaks with a 17D ZOI specified encompasses very nearly theentire volume of containment containing insulated pipes, it is concluded that thevolume of unjacketed insulation is extremely small and approaches 2 cubic feet.It is further noted that during plant walkdowns for GSI-191 in 2004 attempts weremade to photograph all insulated piping. Unjacketed fiberglass would have beenconsidered anomalous and would have been specially photographed for thatreason. Review of the walkdown photographs agrees with the ZOI debrisgeneration calculation that such instances are rare and the associated volume isminimal. In most cases it was "stuffed" into pipe support gaps or was the resultof missing or pulled open jacketing.If the 2 cubic foot estimate is accepted, then 1% erosion is an insignificant (0.02cubic feet) contribution to the total fiber in the sump and is well within the errormargins of the debris generation calculation.The unjacketed quantity estimate is necessary because only piping within theZOI of the breaks was cataloged for GSI-191 and no specific record was made of"all insulation within containment". It is also noted that loose fiberglass without aconvection blocking cover is ineffective piping insulation and not sanctioned byplant insulation specifications.Entergy Nuclear Operations Response to 10b:Large chunks of Marinite are unlikely to transport due to their large size anddensity. This tendency not to transport is exacerbated by the fact that themajority of the "Marinite" insulation is in fact Transite insulation. Transiteinsulation is a similar material but has a significantly higher density (100 lb/ft 3[Ref. 10.2]) than Marinite-I and Marinite-M (46 lb/ft 3 for both [Refs. 10.3, 10.7]).Based on best available information, it appears that the "Marinite" at Palisades is19.6% Marinite I, 13.7% Marinite M, and 66.7% Transite [Ref. 10.1].

Palisades Draft RAI Responses for May 2010 Public Meeting 10NUREG/CR-6772 presented experimental data which showed the incipienttumbling velocity of Marinite to be 0.77 ft/s, with large pieces (4x4 in. flat pieces)experiencing no movement up to the test maximum of 0.99 ft/s flow velocity [Ref.10.6]. Therefore it is reasonable to assume that the large pieces of Marinitedebris in the pool will not transport.NUREG/CR-6772 states that, "Considering the amount of plastic deformationneeded to pull these small rubbery pieces apart, the disintegration of Mariniteinto smaller fragments as a result of turbulence was judged to be highly unlikely."[Ref. 10.6] This is restated as well in NUREG/CR-6808, where it is noted that,"-it was conjectured that the levels of agitation that might develop in acontainment pool would not cause Marinite material to disintegrate." [Ref. 10.5]Erosion testing of Marinite showed that when exposed to flow velocity of 0.4 ft/sfor 32 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br />, the Marinite samples experienced a maximum weight loss of 1.22%[Ref. 10.5]. However, this quantity was determined to be primarily the wearingaway of the rough edges initially found on the broken Marinite sample pieces.This erosion pattern was very similar to that observed on Cal-Sil samples testedconcurrently with the Marinite. Images of the Cal-Sil before and after erosiontesting are shown in Figure 10.1 and Figure 10.2, respectively.Figure 10.1: Cal-Sil Sample Before Erosion Testing Palisades Draft RAI Responses for May 2010 Public Meeting 11Figure 10.2: Cal-Sil Sample After Erosion TestingThus, although there was a 1.22% erosion of sample weight over 32 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br />, it isexpected by the erosion pattern that this would represent the majority of theerosion of the sample expected to take place over the 30 day mission time.Further, it should again be taken into account that the flow erosion testing wasperformed on Marinite 1 insulation material, where as the majority of theinsulation material designated as Marinite for the Palisades plant is actuallyTransite [Ref. 10.1], which in addition to being more dense is also significantlyharder. Marinite-I insulation has Brinell hardness number of 1.2 (45.5 kg load,19.05 mm [Ref. 10.3]) while the Transite insulation has a Brinell hardnessnumber of 17 (500 kg load, 6 mm diameter [Ref. 10.2]. Thus it is reasonable toassume that quantity of debris which erodes off the "Marinite"-type insulationdebris pieces in the pool is insignificant.Therefore it is appropriate to assume that the "Marinite" insulation debris in thePalisades plant will neither transport to the strainers nor erode in the pool.

Palisades Draft RAI Responses for May 2010 Public Meeting 12 References10.1 EA-MOD-2003-0021-01.10.2 BNZ Transite Specification Sheet.10.3 BNZ Marinite I Specification Sheet.10.4 ALION-REP-LAB-2352-218, "Marinite 1 Flow Erosion Testing Report", Rev. 0.10.5 UREG/CR-6808, "Knowledge Base for the Effect of Debris on PressurizedWater Reactor Emergency Core Cooling Sump Performance".10.6 NUREG/CR-6772, LA-UR-01-6882, "GSI-191: Separate EffectsCharacterization of Debris Transport in Water".10.7 BNZ Marinite M Specification Sheet.

Palisades Draft RAI Responses for May 2010 Public Meeting 13 NRC Request 11.In RAI 3 from the NRC's letter dated December 24, 2008, the NRC staff requestedthat the licensee discuss any changes that have been made to the licensee'sanalysis that are associated with debris characterization at a level of detailconsistent with the NRC supplemental response content guide. The licenseeresponded in the June 30, 2009, supplemental response by providing tables ofdebris quantities for each break. Although knowing the debris quantities is helpful,it is unwieldy to compute the percentages of debris in each size category for eachZOI subregion for each break. Since the percentages are the most importantaspect of determining the adequacy of treatment of debris characterization, pleaseprovide the percentages of debris assumed in each category within the ZOI or ZOIsubregion, as applicable.Entergy Nuclear Operations Response:Provided below are the supporting Size Distribution Tables for those giving CubicFeet and Pounds included original June 30, 2009 supplement. They appear tohave the information the NRC is seeking. These tables are provided byReference 11.1. Also see RAI 13 response which is associated with Table 3.1.

Palisades Draft RAI Responses for May 2010 Public Meeting 14 References11.1 ENERCON Report ENTP-003-PR-02, Revision 0, "GSI-191 Debris SizeDistribution and Debris Erosion Report for Palisades Nuclear Station" Reference 11.1, Table 3.1 Reference 6.8 is Design Input Record for EC7833, "Enercon Size Distribution and Erosion Report" Reference 11.1, Table 3.2 Reference 6.1 is ALION-REP-ALION-2806-01,Rev. 3, "Insulation Debris Size Distribution for use in GSI-191 Resolution" Palisades Draft RAI Responses for May 2010 Public Meeting 15 NRC Request 12.In RAI 4 from the NRC's letter dated December 24, 2008, the NRC staffrequested that the licensee provide debris characteristics information forAlpha Maritex insulation. The licensee responded that characterizationinformation for this debris was obtained from WCAP-16727. The licenseesummarized WCAP-16727 as stating that from 10-25 percent of thematerial would be characterized as small pieces and fines, and furtherstated that the debris would readily settle and would not transport in thecontainment pool based in part on a comparison to paint chip transporttest data and an experiment in the head loss test flume at Alden ResearchLaboratory. The NRC staff did not consider this response to be adequatefor the following reasons: (1) it was not clear that the debris generationtesting discussed in WCAP-16727 was adequately scaled to collect debrischaracterization information, (2) it was not clear that comparison of AlphaMaritex (a fibrous material) to paint chips is valid with respect to debristransport properties, (3) it was not clear that the transportability of AlphaMaritex fines was addressed by the licensee, and (4) as discussed in asubsequent RAI, it is not clear that the "averaged" flow conditionssimulated in the head loss testing for Palisades were prototypical orconservative with respect to the plant condition. Please address the aboveissues regarding the Alpha Maritex insulation.Entergy Nuclear Operations Response:Palisades utilized the WCAP-16727 stated Westinghouse opinion that the debrisfrom lead blankets would not transport. Squares pieces of Alpha Maritex of 1inch size were tested at Palisades's containment velocities as modeled in theAlden flume and did not transport which confirmed the Westinghouse statement.We understand that one or more other plants shredded whole sections of leadblankets and in those flumes some of the string like material from the shredderaction on Alpha Maritex cloth may have transported to the strainer at thevelocities utilized in the testing for those plants. It is noted that for the one testwe are aware which was also done at Alden, the velocity in the test flume wassignificantly higher than what Palisades used.If additional testing is done for Palisades a scaled amount of shredded cloth willbe included depending upon the state of knowledge as to the ZOI for leadblankets that exists at the time of the test.The blankets are constructed in long strips and are hung, as a general rule, fromone end. This facilitates the blankets blowing out and away from the break flowand to there by escape destruction. With the exception of 2 installations on thePressurizer Spray lines, all the blankets at Palisades are hung in this manner.The 2 spray line installations are draped across an open frame surrounding 3 Palisades Draft RAI Responses for May 2010 Public Meeting 16sides of the pipe. They are bolted at mid blanket to avoid falling off the pipeframe during plant operation due to vibration from the Primary Coolant Pumpsand unsteady flow in the Pressurized Spray lines. This would represent a smallamount of debris of a type not viewed as a strong participant in strainer pressure drop.Those blankets which are not within the ultimately acceptable ZOI can, ifrequired, be removed during plant operation and reinstalled during outages asthey had previously been. If they can ultimately be re-qualified for Post LOCAsurvival or by acceptable testing, then they can be left on during operation andthat will reduce the dose required to remove and reinstall them. This would be anegligible amount in proportion to that incurred for other GSI-191 activities suchas modification or removal of insulation. The blankets have no credited functionduring plant operation; they are only an outage ALARA issue.

Palisades Draft RAI Responses for May 2010 Public Meeting 17 NRC Request 13.The assumed debris size distribution of 60 percent small fines and40 percent large pieces for jacketed Nukon and Thermal Wrap within a 5DZOI is inconsistent with Figure II-2 of the NRC staff's SE dated December 6,2004 (Agencywide Documents Access and Management System (ADMAS)Accession No. ML0432800007), on NEI 04-07, which considers past air jettesting and indicates that the fraction of small fines should be assumed toreach 100 percent at jet pressures in the vicinity of 18-19 pounds persquare inch (psi). At 5D (5 times the pipe diameter), the jet pressure isclose to 30 psi, which significantly exceeds this threshold. Furthermore, thelicensee's assumption that the size distribution for debris in a range of 5D to7D is 100 percent intact blankets also appears to be inconsistent withexisting destruction testing data. These assumptions for the jacketedNukon and Thermal Wrap debris size distributions appear to be based onthe recent Westinghouse/Wyle ZOI testing discussed in WCAP-16710-P.However, that testing was not designed to provide size distributioninformation, and much of the target material was exposed to jet pressuresmuch lower than would be expected for a prototypically sized break.Furthermore, given the assumption that insulation between 5D and 7D is100 percent intact pieces that do not transport or erode, the licensee haseffectively assumed a 5D ZOI rather than a 7D ZOI for jacketed Nukon andThermal Wrap. In light of this information concerning previous testingexperience, please provide a basis for considering the assumed debris sizedistribution of 60 percent small fines and 40 percent large pieces within a5D ZOI to be conservative or prototypical, as well as the distribution of 100percent intact pieces in a 5D to 7D ZOI subregion.Entergy Nuclear Operations Response:The assumed size distribution for jacketed Nukon and Thermal Wrap was basedon Reference 13.1. Based on response to RAI's 2 through 9, attempting tosupport a size distribution for 5D and 7D is no longer necessary. For futurecalculations, Palisades will either evaluate use of the fibrous-debris sizedistribution provided in NEI 04-07 SER Appendix VI, Table VI-2 or will utilize thesize distribution approach noted in RAI 11 Response, Table 3.2.

References 13.1 Westinghouse Letter Report CPAL-09-3 dated January 30, 2009, "FibrousDebris Size Distribution for Palisades Nuclear Plant Based on JetImpingement Testing of Jacketed NUKON Insulation Pillows Reported inWCAP-16710-P" Palisades Draft RAI Responses for May 2010 Public Meeting 18 NRC Request 14.The June 30, 2009, supplemental response appeared to indicate that only50% of the calcium silicate within a 5.45D ZOI was assumed to bedestroyed into small fines. The other 50% of the calcium silicate within thisZOI was assumed to remain intact following the impact of the LOCA jet.The assumed size distribution appeared to be based on OPG testing forwhich less than 50% of the target calcium silicate was damaged by jetimpingement. Please address the following items concerning the assumedcalcium silicate debris distribution:

a.Significant portions of the target insulation in the OPG tests were notexposed to jet forces representative of the calculated ZOI. In otherwords, the insulation targets used for testing were so long (48 inches)that, due to the small size of the jet nozzle (2.86 inches), a significantportion of the insulation targets was not subjected to destructionpressures prototypical of a complete rupture of RCS piping. In addition,some OPG tests demonstrated the occurrence of insulation damage atdistances in excess of that corresponding to a spherical 5.45D.Although a 5.45D ZOI was accepted for calcium silicate based on theOPG testing, the NRC staff's safety evaluation on NEI 04-07conservatively recommended that 100% of the calcium silicate within aspherical 5.45D ZOI be assumed to be destroyed into small fines, whichcompensated for these test setup issues. Thus, please justify assuminga 5.45D ZOI with 50% intact pieces in light of the information presentedherein.Entergy Nuclear Operations Response to 14a:Based on the photographic evidence from the OPG tests used in determining thesize distribution of destroyed Cal-Sil debris (5D, 7D, 9D, 11D, 13D, and 20D testspresented in Figures 14.1 through 14.6) it is apparent that most significantdestruction of the target insulation was not directly in front of the jet nozzle butwas instead located almost randomly along the length of the target [Ref. 5]. The9D and 7D tests (OPG tests 8 and 12) show a very uniform destruction patternalong nearly the entire length of the target, with the far side of the targetinsulation removed and the nearside remaining intact under intact jacketing. The13D and 20D tests (OPG tests 14 and 15) show a destruction pattern centered atone end of the target, with the centerline of the target directly in front of thenozzle relatively intact.

Palisades Draft RAI Responses for May 2010 Public MeetingfThe approximations given were determined by identification of the jet impingementpressure experienced by the target for each test from Figure II-12 of the SER [Ref. 22].These jet impingement pressures were then converted to approximate equivalent ZOI's byinterpolation of the values in Table 3-1 of the SER [Ref. 22].

19It is important to note that the nozzle diameters (i.e. 5D, 7D, etc.) used in thenomenclature of the OPG tests are not representative of the spherical ZOI'swhich are used in debris generation calculations. These spherical ZOI aregenerated based on the jet impingement pressure experienced by the insulationtarget. An approximation f of the spherical ZOI's equivalent to each of the OPGtests pictured above can be determined by interpolation from the methodology fordetermining spherical ZOI from impingement pressures outlined in the SER.These approximate ZOI are shown in Table 14.1.Table 14.1 - Approximate Equivalent Spherical ZOI'sOPG NomenclatureApproximate Equivalent ZOI 5D 2.92D 7D 3.36D 9D 3.65D 11D 3.90D 13D 4.32D 20D 5.45DFigure 14.1 - OPG 5D Test Target Post-Destruction (Jacketing Removed)(Figure D4 in Ref. 5)

Palisades Draft RAI Responses for May 2010 Public Meeting 20Figure 14.2 - OPG 7D Test Target Post-Destruction(Figure D5 in Ref. 5)Figure 14.3 - OPG 9D Test Target Post-Destruction (Rear View)(Figure D7 in Ref. 5)

Palisades Draft RAI Responses for May 2010 Public Meeting 21Figure 14.4 - OPG 11D Test Target Post-Destruction (Rear View)(Figure D9 in Ref. 5)Figure 14.5 - OPG 13D Test Target Post-Destruction (Rear View)(Figure D11 in Ref. 5)

Palisades Draft RAI Responses for May 2010 Public Meeting 22Figure 14.6 - OPG 20D Test Target Post-Destruction (Rear View)(Figure D14 in Ref. 5)In addition to visual observation of the results of the tests, it is stated explicitly inthe test report that "For tests with target distances from 7 to 13 D, presented inTable 2, it was found that the zone of damage extended to one or both ends ofthe target [Ref. 5]."The OPG report also includes a discussion of the pressure profile on the target,based on correlations for pressure distribution detailed by Kastner [Ref. 20]. Thisdiscussion notes that assuming the jet is centered between two bands spaced8.25" apart, the maximum pressure between the bands at 3D would be 0.65MPa, while the average pressure would be 0.60 MPa. The discussion furthernotes that at a 5D distance the maximum pressure is 0.28 MPa and the averagepressure is 0.27 MPa. Finally at 10D the difference between the maximumpressure and the average pressure is negligible. These pressure profiles aredisplayed in Figure 14.7 [Refs. 5, 20].

Palisades Draft RAI Responses for May 2010 Public Meeting 23Figure 14.7 - Pressure Difference Between Bands as Given by Kastner(Figure 6 in Ref. 5)Comparison to Alternate Testing - General Electric Nuclear Energy (GE)conducted air jet destruction testing on aluminum-jacketed Calcium Silicateinsulation at the Colorado Engineering Experiment Station, Inc. (CEESI), asdocumented in NUREG/CR-6808 [Ref. 15]. The results of the CEESI air jetdestruction testing recommended a destruction pressure of 160 psig foraluminum-jacketed Calcium Silicate insulation. The SER recommends (Section3.4.2.2) reducing the destruction pressures determined in air jet destruction testsby 40% to account for two phase jet effects. Thus the adjusted destructionpressure determined in the CEESI testing is 96 psig. By comparison the OPGresults are conservative, documenting destruction of similar Aluminum-jacketedCalcium Silicate insulation at pressures as low as 24 psig, for an approximateequivalent ZOI of 5.45D. The adjusted destruction pressure of 96 psigrecommended by the CEESI air jet destruction testing would convert to anapproximate equivalent ZOI of 2.6D.Conservatisms -There are a number of conservatisms to be considered whichwere used in the process of converting the OPG test results into a destroyeddebris size distribution.First, the destroyed debris size distribution determined from the OPG testsutilized only those tests in which the longitudinal seam was placed at 45° fromthe jet. The 45° tests where shown overwhelmingly to have the greatestdestruction impact on the target. The tests performed with the longitudinal seamplaced at 0° from the jet produced destroyed debris out to a maximum of 7D, andfor the tests performed with the longitudinal seam placed at 180° no damage wasobserved even at a distance of 3D. This is in stark contrast to the damageobserved out to the 20D range for the 45° tests. As it is extremely unlikely that Palisades Draft RAI Responses for May 2010 Public Meeting 24the jet expanding from a pipe break would hit all possible insulation at 45° withrespect to their longitudinal seams, the use of only 45° tests in the determinationof the debris size distribution and ZOI represents a significant conservatism [Ref.

5].Further, the OPG tests were conducted solely for freely expanding jets. As notedin the report, it is reasonable to assume that within the congested areas commonto plant containments, blockage may occur and dissipate some of the energy ofthe jet [Ref. 5].Inherent conservatisms are also present in the size distribution used byPalisades in comparison to the results of the OPG 45° testing, as shown in Table 14.2.Table 14.2 - Cal-Sil Debris Size Distribution Compared to Average OPG SizeDistribution SizePalisades Size Distribution 0-2.7D ZOI Palisades SizeDistribution 2.7D-5.45D ZOIAverage SizeDistribution ofOPG Destroyed DebrisFines (Particulate) 50%25%20%Small Pieces (Under1" to Over 3")

50%16%13%Remains on Target 0%59%67%For the reasons stated, Palisades believes that the 5.45D ZOI paired with thesize distribution displayed in Table 14.2 is sufficiently conservative. However,based on the NRC's concerns voiced during the teleconference held April 26,2010, Palisades will utilize a 6.4D ZOI paired with the size distribution shown inTable 14.3. This ZOI and size distribution combination is consistent with ZOI andsize distribution outline for Cal-Sil in the ALION size distribution document [Ref.25]. The ZOI and size distribution outlined for Cal-Sil in the ALION sizedistribution document has been used for the debris generation of Cal-Sil in theIndian Point plant. In the Section 3.2 of the Indian Point Audit Report [Ref. 26]the size distribution outlined in the ALION size distribution document is acceptedas sufficiently conservative in part due to the conservative extension of the ZOI to6.4D (equivalent to a destruction pressure of 20 psig, reduced from the SE-accepted recommendation of 24 psig) and the conservative extension of the clothcovered Cal-Sil insulation ZOI to 28.6D. Palisades will utilize the sameconservative extension of the ZOI of jacketed Cal-Sil to 6.4D, and will continue toextend the cloth covered Cal-Sil to 28.6D.

Palisades Draft RAI Responses for May 2010 Public Meeting 25Table 14.3 - Cal-Sil Debris Size Distribution Compared to Average OPG SizeDistribution SizePalisades Size Distribution 0-2.7D ZOI Palisades SizeDistribution 2.7D-6.4D ZOIAverage SizeDistribution ofOPG Destroyed DebrisFines (Particulate) 50%23%20%Small Pieces (Under1" to Over 3")

50%15%13%Remains on Target 0%62%67%NRC Request b.Please identify the jacketing and banding, latching, etc., of the calciumsilicate insulation at Palisades and compare this insulation material tothe material that was tested by OPG to support the assumed debrischaracterization of 50% intact pieces based on the application of theOPG test results. Please also compare the manufacturing process forthe calcium silicate at Palisades with that used for the OPG testing(i.e., hydraulically pressed or molded - see Section 3.3.3 of IndianPoint Audit Report).Entergy Nuclear Operations Response to 14b:There are at least 2 kinds of Cal-Sil in Palisades Plant [Ref. 23]. The originalplant specified asbestos binder and aluminum "jacketing". By 1980 it was illegalto reinstall this material if it was removed from the piping. Also after 1975 it wasknown that there was a possible post LOCA hydrogen generation problem withaluminum so plant procedures were amended to require installation of Stainlessjacketing if the existing aluminum had to be replaced. Thus the majority of theCal-Sil insulation in the Palisades plant is assumed to be either asbestos-reinforced material jacketed with aluminum (referred to as Palisades Type 1) orasbestos-free Thermo-12Ž Gold Cal-Sil material jacketed with stainless steel(referred to as Palisades Type 2) [Ref. 23, 24].The relevant characteristics of the insulation material used in the OPG testing arecompared to those of the Cal-Sil insulation materials installed in the Palisadesplant in Table 14.4 [Ref. 23, 24].Table 14.4 - Cal-Sil Insulation Characteristic ComparisonCharacteristicOPG [Ref. 5]Palisades Type 1Palisades Type 2*Jacketing MaterialAluminum-AluminumStainless Steel Palisades Draft RAI Responses for May 2010 Public Meeting 26 1100Jacketing Thickness 0.016 in.0.016 in.minimum 0.016 in.minimumBand MaterialStainless SteelStainless Steel** Stainless Steel**Band Thickness 0.02 in.0.016-0.025 in.0.016-0.025 in.Band Width 0.5 in.0.5 in.0.5 in.Band Spacing (Centerline)8 in.12" maximum12" maximum12" maximum Cal-SilManufacturing ProcessNot provided Asbestosreinforced; post-autoclave processAsbestos-freeThermo-12ŽGold Cal-Sil; Pabco Process*Palisades Type 2 Cal-Sil is allowed by specification at the Palisades plant, but isbelieved to be installed in minimal quantities.**All bands are observed to be SS but specification does allow aluminum orgalvanized metal. [Ref. 23].Jacketing Material and Thickness - The thickness of both aluminum andstainless steel jacketing types is a minimum of 0.016 in [Ref. 7]. Thus it isreasonable to conclude that the aluminum jacketing is of equivalent or greaterstrength, as it is of equivalent or larger thickness to that used in the OPG tests.Further, the aluminum cladding used at the Palisades plant is ASTM B 209,Alloy: 5005 (Temper: H-14, Finish: mill). This aluminum is known to have amaterial tensile strength of 23 ksi [Ref. 21]. The aluminum jacketing used in theOPG testing was Aluminum 1100 [Ref. 5]. The tensile strength of Aluminum 1100is 13 ksi [Ref. 8]. As the aluminum jacket used in the Palisades plant is known tohave greater material tensile strength (23 ksi) than the aluminum jacketing usedin the OPG testing (13 ksi) it is reasonable to conclude that the jacketing in useat the Palisades plant is of equivalent or greater strength to that used in the OPGtesting. This is especially important as the failure mode observed in the OPGtesting was shearing of the cladding [Ref. 5].The stainless steel jacketing allowed by specification at the Palisades plant isbelieved to be present only in minimal quantities. The majority of the cladding isbelieved to be aluminum [Ref. 23].Banding Style and Spacing - The specification for the bands used to fasten thejacketing characterizes the bands as 0.016-0.02 in. thick for pipes of outerdiameter up to 12 in. and 0.02-0.025 in. thick for pipes of outer diameter greaterthan 12 in. The specification allows for aluminum bands as opposed to stainlesssteel bands for use with aluminum jacketing, however all bands are observed tobe stainless steel. Similarly galvanized metal straps are allowed by thespecification in place of stainless steel bands, but galvanized metal straps werenot found in the plant during walkdowns [Ref.23].

Palisades Draft RAI Responses for May 2010 Public Meeting 27It is thus concluded that the banding used to fasten the jacketing to the insulationin the Palisades plant is equivalent to that used in the OPG testing, comprisedprimarily of stainless steel bands of 0.02 in. thickness.The bands used in the Palisades plant to fasten the jacketing are spaced amaximum of 12 in. apart [Ref. 23]. This is marginally further apart than the 8in.spacing used in the OPG tests [Ref. 5]. However, the failure mode observed inthe testing was not band failure, but rather shearing of the cladding.Cal-Sil Manufacturing Process- In his document,Summary of Calcium SilicateInsulation Types Used inside Containment at US Nuclear Plants, [Ref. 6] GordonH. Hart, P.E., describes that over the past 50 years or so of nuclear power plantconstruction and operation, several different types of Cal-Sil pipe and blockinsulation have been used. The summary prepared by Hart was based uponinformation provided by Tom Whitaker of Technical Support at IndustrialInsulation Group, LLC, (11G), currently the only North American manufacturer ofCal-Sil.Hart categorizes Cal-Sil with asbestos fiber as Type I Cal-Sil created with a post-autoclave process that was discontinued in the early 1970's due to asbestos'carcinogenic attributes. Type II is free of asbestos fibers and is made by a filterpress pre-autoclave. Type III is also free of asbestos fiber and is made in a pourand mold process known as the Pabco Process, also a Post Autoclave process.Further, according to IIG, Type III Cal-Sil is more friable than Type II Cal-Sil.Type III Cal-Sil is softer and will most likely erode faster in a moving fluid thanType II Cal-Sil. Further according to IIG

,"To my (Whitaker's) knowledge, there isno published information related to erosion rates of any of these products. Thecomparison of the erosion rates is based on personal experience in the CalciumSilicate business since 1972."Also according to IIG

,Whitaker indicates that, "My opinion is that the asbestos-based Cal-Sil would have less erosion due to moving fluids than the non-asbestos based insulation products."In summary, based upon the above information, Type 1, asbestos fiber-reinforced, Post Autoclave process, Cal-Sil is theleast susceptible to erosion of the three Cal-Sil types described by Hart, Type II, non-asbestos, Johns-Manville Process Cal-Sil is less susceptibleto erosion than the Type III Cal-Sil, and Type III, Pabco/Post Autoclave Process Cal-Sil is the most susceptibleCal-Sil to erosion of the three Cal-Sil types described by Hart.

Palisades Draft RAI Responses for May 2010 Public Meeting 28As stated above the first type of Cal-Sil insulation installed in the Palisades plantis asbestos containing (Palisades Type 1) and is assumed to be equivalent toHart Type I. The second type of Cal-Sil insulation installed in the Palisades plantis asbestos-free (Palisades Type 2) formed using the Pabco process and isassumed to be equivalent to the Hart Type III. The Hart Type II Cal-Sil is notbelieved to be in use at Palisades because the Owens-Corning plant was in theeast coast area and also most of the new Cal-Sil was installed by PCI personnelfrom Kansas (PCI was the installation arm of Owens-Corning through the 1990's)who likely used one of the western (Colorado) Cal-Sil Plants some of which were built by Pabco.The asbestos reinforced Palisades Type I Cal-Sil insulation, as equivalent to theleast friable of all Hart-type Cal-Sil insulations, is assumed to be equal to orsuperior in destruction resistance to that used in the OPG tests.The asbestos free Palisades Type 2 Cal-Sil insulation is equivalent to the mostfriable of all Hart-type Cal-Sil insulations. However, this type of Cal-Sil insulationis present in much lower quantities within the plant than the more robust Type 1Cal-Sil insulation (equivalent to Hart-Type I). Further, dissolution testing wasperformed on Thermo-12Ž Gold Cal-Sil type Cal-Sil insulation (equivalent toPalisades Type 2 Cal-Sil insulation) at Alion test facilities. These tests showedthat large scale dissolution did not occur [Ref. 9]. As documented in Section5.2.5 of NUREG/CR-6808 [Ref. 15], at least one type of Cal-Sil has been shownto undergo as much as 76% weight loss under hot water and occasional stirring.Therefore, it has been shown that the Palisades Type 2 Cal-Sil insulation is notthe most friable of available Cal-Sil insulation.Based on discussion with the Staff during a 4/26/2010 Technical Call, it isunderstood that further support for the Palisades banding spacing differencesfrom that used in the OPG testing is needed. This may be in the form of acalculation to determine whether the additional strength of Palisades jacketingoffsets the added banding spacing. Installing additional banding could also be asolution. Additionally, Palisades needs to better address whether all bandinginstalled is in fact stainless steel. Both these items will be evaluated further.

NRC Request c.The 50% of the calcium silicate considered to be undamaged was notconsidered for potential erosion and transport to the strainer.However, for a number of the OPG tests, the insulation jacketing wasremoved, even when the base insulation material was not completelyremoved from the pipe. In light of the removal of the jacketing in anumber of tests, please justify that erosion of the calcium silicateremaining on pipes need not be considered.Entergy Nuclear Operations Response to 14c:

Palisades Draft RAI Responses for May 2010 Public Meeting 29While a sizable portion of jacketing was removed in a number of the OPG tests,the jacketing removed was overwhelmingly from areas in which the baseinsulation was removed (see Figures 14.8 through 14.10). For the insulationmaterial which remains on the target, the overwhelming majority of the surfacearea of the Calcium Silicate material remains protected by jacketing material.Only approximately 20% of the material remaining on the target is unprotected byjacketing material. The potential for erosion of the intact calcium silicateinsulation from the containment spray flow and the impact on the overall quantityof calcium silicate material reaching the containment sump strainer is considerednegligible for the following reasons: A significant portion of the Calcium Silicate material that is unprotected bythe jacketing material, but remains on the piping, would not becoincidentally in the path of the containment spray flow. The safety evaluation (SE) for NEI 04-07 [Ref. 22] concludes that thecontainment spray erosion rate for low density fiberglass insulation is lessthan one (1) percent. Although the SE does not provide a containmentspray erosion rate for calcium silicate insulation, the containment sprayerosion rate for calcium silicate insulation is expected to be extremely low,similar to low density fiberglass insulation. The recirculation pool erosionrate of 17% was measured for calcium silicate insulation, whereas arecirculation pool erosion rate of 10% was measured for low densityfiberglass insulation. The relationship between the Calcium Silicate andlow density fiberglass recirculation pool erosion rates is expected to besimilar to the relationship between the Calcium Silicate and low densityfiberglass containment spray erosion rates. Due to uncertainties in the Zone of Influence for unjacketed CalciumSilicate insulation, an extremely conservative ZOI of 28.6D (largest ZOI forall insulation provided in NEI 04-07) is used for unjacketed calcium silicateinsulation. Further, none of the unjacketed Calcium Silicate insulationwithin the ZOI is assumed to remain on the piping. All the calcium silicateinsulation reaches the containment recirculation pool. Therefore, the largepieces of calcium silicate insulation that are generated as part of the breakare subjected to erosion within the recirculation pool. As provided in Table 14.3, the size distribution to be used for Palisadesdebris generation calculation is extremely conservative when compared tothe size distribution observed in the OPG testing.Therefore, it is concluded that the erosion of the Calcium Silicate pieces thatremain on the piping after the break is negligible when compared to the otherconservatisms included in determination of the quantity of calcium silicateinsulation that reaches the Palisades containment sump strainer.

Palisades Draft RAI Responses for May 2010 Public Meeting 30Figure 14.8 - OPG 9D Test Target Post-Destruction (Front View)(Figure D8 in Ref. 5)Figure 14.9 OPG 11D Test Target Post-Destruction (Front View)(Figure D10 in Ref. 5)

Palisades Draft RAI Responses for May 2010 Public Meeting 31Figure 14.10 OPG 20D Test Target Post-Destruction (Front View)(Figure D13 in Ref. 5)

Palisades Draft RAI Responses for May 2010 Public Meeting 32 References1. EA-MOD-2003-0021-01.2. BNZ Transite Specification Sheet.3. BNZ Marinite I Specification Sheet.4. ALION-REP-LAB-2352-218, "Marinite 1 Flow Erosion Testing Report", Rev. 0.5. Ontario Power Generation, "Jet Impact Tests-Preliminary Results and TheirApplications," N-REP-34320-10000-R00. April 2001.6. Gordon H. Hart, P.E , "Summary of Calcium Silicate Insulation Types Usedinside Containment at US Nuclear Plant", December 31, 2007.7. CS85-P-HT1.8. Avallone, Eugene A. et. al.,"Marks' Standard Handbook for MechanicalEngineers", Ninth Edition.9. ALION-REP-PAL-7199-22, "Palisades Cal-Sil Flow Erosion and DissolutionTesting Report", Rev. 0.10. ALION-REP-ENT-7199-21, "Palisades Fiberglass Debris Flow ErosionTesting Report", Rev. 0.11. ALION-PLN-I006-02, "Erosion Testing of Small Low Density FiberglassDebris: Test Plan", Rev. 1.12. ALION-REP-LAB-2352-99, "Test Report: Erosion Testing of Mineral WoolInsulation", Rev. 0.13. ALION-REP-LAB-2352-77, "Test Report: Erosion Testing of Low DensityFiberglass Insulation", Rev. 3.14. ALION-REP-ENT-7637-02, "Palisades Erosion Assessment - AverageVelocity and Turbulence for Non-Transporting Debris", Rev. DRAFT.15. NUREG/CR-6808, "Knowledge Base for the Effect of Debris on PressurizedWater Reactor Emergency Core Cooling Sump Performance".16. NUREG/CR-6772, LA-UR-01-6882, "GSI-191: Separate EffectsCharacterization of Debris Transport in Water".17. BNZ Marinite M Specification Sheet.18. NUREG/CR-6224, "Parametric Study of the Potential for BWR ECCS StrainerBlockage Due to LOCA Generated Debris".19. ALION-REP-ALION-I006-04, "Erosion Testing of Small Pieces of Low DensityFiberglass Debris - Test Report", Rev. 0.20. Kastner, W. and Rippel, R., Jet Impingement Forces on Structures-Experiments andEmpirical Calculation Methods, Nuclear Science andEngineering, 105, pp269-284 (1988).21. Aluminum 5005-H14 Physical Data based upon Metals Handbook, Vol.2 -Properties and Selection: Nonferrous Alloys and Special-Purpose Materials,ASM International 10th Ed. 1990; Metals Handbook, Howard E. Boyer andTimothy L. Gall, Eds. American Society for Metals, Materials Park, OH, 1985;Structural Alloys Handbook, 1996 edition, John M. (Tim) Holt, Technical Ed;C. Y. Ho, Ed., CINDAS/PurdueUniversity, West Lafayette, IN, 1996; andAluminum Standards. (http://www.matweb.com

)

Palisades Draft RAI Responses for May 2010 Public Meeting 3322. NEI 04-07, Volume 2, "Safety Evaluation by the Office of Nuclear ReactorRegulation Related to NRC Generic Letter 2004-02," Rev. 0, December 6, 2004.23. Palisades Design Input for RAI Responses, DRAFT.24. Palisades Specification M-136.25. ALION-REP-ALION-2806-01, "Insulation Size Distribution for use inGSI-191Resolution", Rev. 3.26. Indian Point Audit Report.

Palisades Draft RAI Responses for May 2010 Public Meeting 34 NRC Request 15.The licensee's June 30, 2009, supplemental response stated that large pieces ofdebris were assumed to be retained on the 608'6" elevation (EL) rather thantransporting to the 590' EL. The licensee's response added that it was ultimatelyinconsequential whether the large pieces were retained on the 608'6" EL orwashed down to the 590' EL, since direct transport to the strainer would notoccur in either case, and an equivalent degree of erosion would occur regardlessof the elevation on which the debris was assumed to settle. It was not clear to theNRC staff that it was inconsequential whether large pieces remained on the608'6" EL or transported to and settled on the 590' EL. In particular, the NRCstaff expected that the flow conditions (e.g., velocity and turbulence) on these twoelevations could be significantly different, thereby leading to differentpercentages of eroded debris. In particular, the NRC staff expected that erosionwould be more severe on the 608'6" EL because (1) shallow pools exist, (2) thereis significant potential for break and/or spray drainage to create high-velocityand high-turbulence flow conditions, and (3) the potential exists for directexposure to break and/or spray drainage flows. It is not clear to the NRC staffthat flow conditions of this sort are bounded by industry erosion testing conductedwith significant submergence, no direct exposure to drainage flow, and relativelylow velocity and turbulence values, and how this potential discrepancy has beenaddressed in the licensee's strainer performance evaluation. Therefore, pleaseidentify the range of water velocity and turbulence levels expected on the 608'6"EL, the expected water depth, as well as the basis for concluding that anequivalent degree of erosion would occur for debris retained on the 608'6" EL aswould occur on the 590' EL considering the industry erosion test conditions andresults. As applicable, please also provide similar discussion for any otherlocations where significant debris holdup was credited for which erosion mayoccur under similar conditions to that applicable to the 608'6" EL (e.g., thestairwells leading from the 608'6" EL to the 590' EL).Entergy Nuclear Operations Response:For Palisades containment the limiting break from a debris generation and also adebris transport perspective occurs in the A-loop. The A loop geometry is shownin Figure 3 with run-off paths identified by letter code. Overall, 83 ft of 6" highcurbing surround the area marked in yellow. Only one trench-like opening existsin the curbing. This run-off path is marked K. Approximately half of the area isprotected from direct spray from above by the steam generator and otherconcrete closures in the upper elevations of containment. Given the spray andbreak-flow inputs that exist during an A-loop break, the floor of the area markedin yellow will flood to over-flow the curbing. To support the run-off of water, thewater level will rise to 7". Given the length of the curbing, an approach velocity tothe curb of 0.09 ft /sec can be calculated. This velocity is insufficient to allow thelarges to transport out of the curbed area since the lift-over-curb velocity forlarges is 0.34 ft/sec [Ref. 15.2 and 15.3]. While the areas of direct break flow Palisades Draft RAI Responses for May 2010 Public Meeting 35impact may be subjected to high velocities it is expected that a) the duration ofexposure of a large amount of insulation to such high velocities is very short asthe insulation is pushed away by sheeting water after the initial blast and blow-down b) the amount of insulation exposed to high velocities for an extendedperiod of time is very small. Even though a large amount of debris will be pushedto areas where break flow no longer provides significant velocity and spray flowdoes not exist (half of the area is covered by concrete slab or protected by theupper geometry of the steam generator) large debris is eroded to fines at a valueof 10%. Furthermore, note that the velocities approaching the curb will drop offdramatically to the point where larges will stop transporting and not block thecurbing.Figure 3. Immediate area of A-Loop SG compartment [Ref. 15.1]Another significant conservatism is provided by the treatment of smalls in theapproach to debris transport on the 608'6" EL. While the above calculations showthat the approach velocity to the curbing is limited to 0.09 ft/sec and a 6" curbexists for all but 9" (trench marked K) of the perimeter of the area, all smalls aretransported to the sump from the 608'6" EL. A significant portion of the smallstransported to the sump (37%) were calculated to be transportable to the strainer Palisades Draft RAI Responses for May 2010 Public Meeting 36based on a very conservative 0.06 ft/sec incipient tumbling velocity [Ref. 15.1].This fraction is consistent with ignored areas of possible smalls transport (seeRAI 18). In addition, the fraction of smalls not considered transportable to thestrainers was eroded at 10%. Therefore significant additional conservatism existsin the overall treatment of debris generated on the 608'6" EL which should betaken into account in evaluating the treatment of large fiber transport on 608'6" EL.Debris transport treatment of larges for B-loop breaks was similar. The geometryfor the B-loop is somewhat different from the A-loop due in part to the fact thatthe A-loop compartment includes the pressurizer. While less of the perimeter onthe B-loop is bounded by curbing, significantly less debris is generated in thecompartment. Significantly greater overall debris erosion of larges can besupported on B-Loop breaks before fines generation for these breaks begin toexceed A-Loop breaks fines generation. The erosion percentage necessarybefore B-loop breaks fines generation eclipses that of A-loop breaks is 33%. It istherefore clear that A-loop debris transport represents the limiting case for thePalisades nuclear plant.

References15.1 AREVA Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport15.2 NEI Guideline, "Pressurized Water Sump Performance EvaluationMethodology," December 2004.15.3 SER-GSI-191 SE, Revision 0, "Safety Evaluation of NEI Guidance onPWR Sump Performances", Office of Nuclear Reactor Regulation,December 2004 Palisades Draft RAI Responses for May 2010 Public Meeting 37NRC Request 16.Although the licensee's June 30, 2009, supplemental response quoted astatement in NUREG/CR-2982 indicating that mineral wool can remainafloat for extended periods of time, it did not appear that the licensee hadaccounted for the potential for floatation adequately in the following two respects:a. No basis was presented to justify the assumption that no large piecesof mineral wool would be capable of transporting over the 6-inch curb,in part due to floatation by virtue of trapped air.b. No basis was presented to justify that no large and small pieces ofmineral wool would be capable of transporting to the strainers, in partdue to floatation by virtue of trapped air. Although strainer testing wasperformed with mineral wool debris, based on the PCI test procedure,this material was thoroughly soaked with water, and thus the potentialfor transport by floatation was not examined.Please provide additional information to justify the assumption regardingflotation of mineral wool, considering the above.Entergy Nuclear Operations Response:The previous submittal to the NRC detailed the equipment contributing mineralwool debris ([Ref. 16.2], Table 3a1, p.17) in the considered LOCA debrisgeneration scenarios. Mineral wool debris is obtained from the pressurizer shell.The pressurizer is located in a corner of the A-loop steam generatorcompartment. The postulated breaks in the A-loop compartment would tend tokeep debris in the immediate equipment corner where the pressurizer is located.Note that the pressurizer is located greater than 7D away from the limiting debrisbreak ([Ref. 16.2], Table 5.2-6, p.35). Furthermore, the pressurizer floor elevationis below the main steam generator compartment and grating [Ref. 16.1] coversthe pressurizer floor area. The remainder of the steam generator compartmentdoes not exhibit grating except for in very small areas. The pressurizer area isnot open to any spray flow from above. Figure 16.46.1 is provided to illustrate thelayout of the area of the steam generator compartment under consideration here.Based on this information the following is expected with respect to the behaviorof mineral wool debris: The mineral wool debris originating from the pressurizer will largely remainnear the pressurizer. Transport of debris off the grating is not expected since water flow in thearea is small and no standing water exists to support floatation. Debris deposited onto the grating on the pressurizer grating will not besubject to increased erosion since spill flow in this area is very low (0.35 ft 3/sec)0 and the area does not see any spray flow.

Palisades Draft RAI Responses for May 2010 Public Meeting 38 While it is possible that some very small fraction of mineral wool smallsfrom the pressurizer could end up on the main steam generator floor, thisdebris will mix with Nukon fiber debris. Small mineral wool debris that remains afloat in the main steam generatorcompartment could flow over the existing curb and down into thecontainment pool. The test basis was developed by neglecting thispossible transport (floating), instead eroding all mineral wool debris,including the vast majority which is not subject to spray flow or othertransport mechanisms.Based on the above discussion debris transport of mineral wool larges byfloatation is not credible and only possible for small fractions of mineral woolsmalls. Although the bulk of the mineral wool debris will be located on largely drygrating, all mineral wool debris smalls and larges were eroded at 10% to finesleading to a very conservative testing basis for mineral wool debris.Figure 16.4. View of SG-A compartment near pressurizer References16.1 32-9099369-000 AREVA Calculation dated February 2009, PalisadesNuclear Power Station - Debris Transport Palisades Draft RAI Responses for May 2010 Public Meeting 3916.2 Palisades June 30, 2009 Submittal to NRC, Follow-up SupplementalResponse to NRC Generic Letter 2004-0216.3 NEI Guideline, "Pressurized Water Sump Performance EvaluationMethodology," December 2004.16.4 SER-GSI-191 SE, Revision 0, "Safety Evaluation of NEI Guidance onPWR Sump Performances", Office of Nuclear Reactor Regulation,December 2004.

Palisades Draft RAI Responses for May 2010 Public Meeting 40 NRC Request17. The June 30, 2009, supplemental response provided cumulative erosionpercentages for frangible debris in Table 3e4. Several values in this tablein the copy of the document available to the NRC staff in ADAMSAccession No. ML091820275 were illegible. Please provide another copyof Table 3e4 that can be clearly read. Please also provide a description ofany testing performed to support the assumed erosion percentages fordebris pieces in the containment pool or other areas of containment whereerosion was assumed to occur (e.g., 608'6" EL, stairwells, etc.). Pleasespecifically include the following information:Entergy Nuclear Operations Response:Below Table 17.1 provides the information that was in Table 3e4 of the 6/30/09supplemental response. Balance of RAI 17 response is provided below aftereach specific NRC request.Table 17.1 Tumbling velocities and erodible fractions Debris TypeErosion Factor Incipient Tumbling Velocity (ft/s)Nukon/Thermal Wrap Jacketed 10%0.06Calcium Silicate Metal Jacketed 17%0.25Transco RMI N/A0.20Low Density Fiberglass Jacketed 10%0.06Calcium Silicate Cloth Covered 17%0.25Low Density FiberglassUnjacketed 10%0.06Nukon Unjacketed 10%0.06Mineral Wool Jacketed 10%0.30Qualified Coatings N/A100% transport assumedUnqualified Coatings N/A100% transport assumedLatent Debris N/A100% transport assumed Palisades Draft RAI Responses for May 2010 Public Meeting 41 NRC Requesta. Please describe the test facility used for any erosion testing anddemonstrate the similarity of the flow conditions (velocity andturbulence), chemical conditions, and fibrous material present in theerosion tests to the analogous conditions applicable to the plantlocations where erosion could occur.Entergy Nuclear Operations Response to 17a:The original erosion tests were performed at the ALION Science & Technology'sHydraulics Laboratory in Warrenville, IL. Transport tests were performed todetermine tumbling velocities and were performed in the Transport Flume.Erosion tests were performed in the Transport Flume as well as in the enclosedVertical Test Loop [Refs.12, 13].New erosion tests were recently performed at ALION Science & Technology'sHydraulics Laboratory in Warrenville, IL to conservatively determine erosionfactors for Low Density Fiberglass insulation materials (LDFG) and to addressthe NRC's concerns with the original testing. The results of the new testsdesigned to address the concerns with the original testing have given results thatsupport the 10% erosion factor developed in the original tests [Ref. 19].

Palisades Draft RAI Responses for May 2010 Public Meeting 42Table 17.2 - Incipient Tumbling Velocities and Test Velocities Debris Type Erosion FactorIncipient TumblingVelocity (ft/s)Erosion TestingVelocity (ft/s)Nukon/ThermalWrap Jacketed 10%0.060.12Calcium SilicateMetal Jacketed 17%0.250.40Transco RMI N/A0.20 N/ALow DensityFiberglass Jacketed 10%0.060.12Calcium SilicateCloth Covered 17%0.250.40Low DensityFiberglassUnjacketed 10%0.060.12Nukon Unjacketed 10%0.060.12Mineral Wool Jacketed 10%0.300.12Qualified Coatings N/A100% transportassumed N/AUnqualifiedCoatings N/A100% transportassumed N/ALatent Debris N/A100% transportassumed N/ATable References 8, 10, 11, 19.LDFG - Flow Velocity and Kinetic Energy:After the bulk transport of thedebris, a total of 64.2% of the small fiberglass debris generated would be on thecontainment pool floor where it could be exposed to erosion. The average flowvelocity and turbulent kinetic energy (TKE) was calculated in six different zoneswithin the containment [Ref. 14]

Palisades Draft RAI Responses for May 2010 Public Meeting 43Figure17.1 - Containment Erosion ZonesTable 17.3 - Summary of Settled Fibrous Insulation DebrisLocationPercent of smallpieces settledAverage Velocity (ft/s)Average TKE (ft 2/s 2)Zone 120.4%0.00800.00010Zone 28.9%0.00810.000078Zone 30.4%0.0200.00039Zone 425.2%0.0130.000044Zone 52.6%0.00710.000028Zone 66.7%0.00590.000030Total/WeightedAverage64.2%0.00980.000067As presented in Table 17.3, the average velocity, weighted by quantity of settledfiber experiencing the flow in each location, is 0.0098 ft/s, with an average TKE,weighted similarly, of 0.000067 (ft 2/s 2) [Ref. 14].

Palisades Draft RAI Responses for May 2010 Public Meeting 44The new erosion tests were conducted utilizing a flow velocity of 0.12 ft/s,bounding not only the weighted average velocity, but the maximum average flowvelocity in any zone (zone 3) by a factor of 6.While it is not possible to determine the exact quantity of turbulent kinetic energyin the testing flume apparatus, it is possible to determine the total kinetic energyof 0.0072 ft 2/s 2 which compares favorably to the total kinetic energy of thehighest average velocity zone (zone 3), which has a turbulent kinetic energy of 0.00039 ft 2/s 2. Thus the water in the test flume is significantly more energeticand bounds the energy present within containment.LDFG - Target Size:The erosion fraction of 10% determined in the originaltests and confirmed in the new tests is applied to both small and large fiberpieces in the pool. The erosion fraction determined is assumed to be applicableto all affected sizes of LDFG due to the results of the original testing whichdetermined that proportional erosion of small pieces conservatively bounds theproportional erosion of larger pieces.During the original testing, small pieces (1"x1"x1") were tested under anapproach velocity of 0.12 ft/s, while large pieces (6"x3"x1") were tested under ahigher approach velocity of 0.37 ft/s [Ref. 13]. Even with the higher approachvelocity the erosion experienced by the large pieces was significantly less (1% to6%) than the erosion experienced by the small pieces (2% to 18%) [Refs.13, 19].The results clearly showed how the proportional erosion of small pieces boundsthe proportional erosion of larger pieces.LDFG - Chemical Conditions:NUKON and Mineral Wool "erosion" areessentially the departure of individual fibers from a piece of NUKON/MineralWool, not the wearing down of the fibers themselves, i.e., "erosion" is not a resultof fibers being "worn/eroded" down to a finer diameter by removal of smallamounts of the basic material. While insulation made of NUKON or Mineral Woolmaterial is a compressible, porous material without a solid binder holding ittogether, prior to the start to each erosion test, ALION boiled the NUKONsamples in tap water for over 10 minutes and used pre-baked Mineral Wool inorder to remove any oils that could be considered to act as a binder [Ref. 10].ALION is not aware of any experimental evidence that suggests that thedeparture of individual fibers from a piece of NUKON or Mineral Wool would bedifferent for either chemically non-neutral and/or elevated temperature solutionversus flow of equal velocity of tap water. In the absence of any experimentalevidence to the contrary, and the absence of a "binder" that could react tochemistry or temperature, there is no reason to believe that the "erosion" results(departure of individual fibers) for an elevated-temperature, chemically non-neutral solution would be significantly different from that obtained with room-temperature tap water.

Palisades Draft RAI Responses for May 2010 Public Meeting 45Both the Cal-Sil erosion tests (completed) and the LDFG erosion tests (ongoing)are conducted at room temperature, and this temperature is recorded with eachdata point. The increased post-LOCA water temperature at Palisades would havelittle effect on the flow erosion of fiberglass and Cal-Sil with respect to waterdensity and viscosity differences. The lack of containment recirculation chemicalsand neutral pH of the tap water would also have little effect on the flow erosionmechanism. Testing was mainly measuring the mechanical affect of circulatingtap water around stationary insulation samples to observe the amount offiberglass that is released into the flowing water [Refs. 8, 10].

Cal-Sil: The Cal-Sil erosion tests were performed under a flow velocity of 0.40ft/s, which significantly bounds both the tumbling velocity of the Cal-Sil piecesinvolved (0.25 ft/s) as well as the flow velocity conditions predicted to beexperienced in the containment (see Table 17.1). Like the on-going LDFGerosion testing, this significant margin in flow velocity also causes the Cal-Silerosion tests to be bounding from a kinetic energy standpoint, with a total kineticenergy of 0.08 ft2/s2 in the tests compared to a maximum turbulent kineticenergy of 0.00039 ft/s predicted to be experienced in the containment.Additional dissolution tests were conducted for the Cal-Sil debris at 190+/-5° F with3000 ppm boron (as the limiting boron concentration) under different chemistries.The pH range of dissolution testing was 4.8 to 8.75 which bounds the finalminimum and maximum pH at Palisades plant, 7.0 to 8.0. These dissolutiontests found that large scale dissolution of Cal-Sil did not occur due to chemicaldifferences in the water and are the basis for the application of the neutral pH tapwater erosion tests to the Palisades plant [Ref. 8].Mineral Wool: The mechanism for mineral wool and fiberglass was examined ina series of tests conducted by ALION Science & Technology [Ref. 10]. The sameerosion mechanism (attrition) that affected the NUKON LDFG testing [Ref. 10]was observed during the erosion of mineral wool. The mineral wool sampleswould release smaller fibers or loosely bound pieces of fiber; the sample did notactually erode into fines. This mechanism gives reason to an upper bound oferosion once the lose fibers detach from the sample. After such a washing, themineral wool sample would rest in the flowing water and cease to "erode" further.Table 17.4 - NUKON and Mineral Wool Characteristics [Ref. 10]Density(lbs/ft)Fiber Diameter(microns)Constitution(% substance)NUKON 2.4 7Fiber glass wool: 85-96Phenolic Formaldehyde Binder: 4-15Mineral Wool 85-7Mineral Fiber: 94-99Phenolic Formaldehyde Binder: 1-6 Palisades Draft RAI Responses for May 2010 Public Meeting 46Mineral wool has less binder and a higher density than NUKON. Mineral woolalso has a component called "shot" which is a particulate type matter that isfabricated into the mineral wool fibers. This shot increases the density of themineral wool, but the shot did not appear to erode during the flow erosion testingbeyond the initial washing of the outer most layers. If a great deal of the shot didindeed wash out during erosion, then a more substantial weight loss would beobserved for the mineral wool. The fact that the mineral wool behaves similarly tothe shot-free NUKON conveys the impression that the mineral wool shot is notsignificantly affected by the flow erosion. Although NUKON LDFG and mineralwool insulation are not the exact same material, the erosion properties of bothinsulations compare readily. It was observed during testing that both materialserode through the release of constituent fibers and not through the wearing awayof fibers or binder. Both material samples rested in the testing environmentsimilarly: neither material dissolved or disintegrated when placed into the flowingwater, and both materials retained their shapes when taken out of the water aftertesting. The test results for each duration of mineral wool test were also similar,specifically the pattern of decreasing erosion per hour over time. Table 17.5[Refs.12, 13] presents the results of the original mineral wool and NUKONerosion tests performed by ALION in 2007. The weight loss percentagepresented for NUKON after 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> is the average over the two data pointsavailable, the percentage presented after 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> represents the sole data pointdetermined in testing [Ref. 13]. Both mineral wool percentages represent thesole data points determined at the respective time during testing [Ref. 12].Table 17.5 - Fibrous Debris Material Erosion ResultsPercentage WeightLost after 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />sPercentage Weight Lostafter 16 hour1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br />sNUKON4.56%7.54%Mineral Wool2.44%6.55%Due to the similarities between LDFG and Mineral Wool in regards to the methodof erosion, the results of the LDFG erosion testing will be applied to Mineral Woolas well.NRC Requestb. Please identify the duration of the erosion tests and how the resultswere extrapolated to the sump mission time.Entergy Nuclear Operations Response to 17b:

Palisades Draft RAI Responses for May 2010 Public Meeting 47 Cal-Sil: Tests were conducted with a range of durations from 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> to 101hours. The data was then conservatively extrapolated to the 720 hour0.00833 days <br />0.2 hours <br />0.00119 weeks <br />2.7396e-4 months <br /> sumpmission time by the application of a linear fit curve to the entire set of data points(inclusive of all tests, regardless of duration). This extrapolation is conservativedue to the assumption of a constant rate of erosion, where it is more realistic toassume a reduction in the rate of erosion over time.LDFG: The new testing completed to address the concerns of the NRC dealtwith extrapolation by performing a 30 day test, matching the 720 hour0.00833 days <br />0.2 hours <br />0.00119 weeks <br />2.7396e-4 months <br /> sumpmission time and removing the necessity of any form of extrapolation.

NRC Requestc. Please provide a basis for adding debris attributed to erosionprocesses after the addition of small debris pieces. The addition offines after small pieces is not consistent with previous debrissequencing discussions for the Performance Contracting, Inc. (PCI)test protocol or the NRC staff's March 2008 head loss reviewguidance. In addition, some industry test results have suggested thaterosion occurs to a significant extent relatively soon after debris piecesare exposed to flow.Although the June 30, 2009, supplemental response provided discussionintending to demonstrate that sufficient margins existed in the licensee'sstrainer performance analysis to bound the quantity of fibrous fines thatwould be associated with explicitly analyzing the erosion of small piecesof fibrous debris settling in the test flume, the NRC staff did not considerthe margins discussed to be sufficient based on the following reasons:Although the analytical transport calculation did not credit the settlementand holdup of fine debris, the NRC staff considers that the head losstesting permitted substantial credit for fine settlement.Based on the flume velocities used for the licensee's head loss testing, theNRC staff expected that a significant majority of the small pieces of fiberadded to the test would not have reached the strainers.Based on the test scaling leading to denser debris piles in the test flumethan expected for prototypical plant conditions and the fact that the smallpieces of debris added to the test had been mechanically shaken torelease loosely attached fibers, the NRC staff expected that the actualerosion in the test flume would be very small and would not berepresentative of erosion under plant conditions.

Palisades Draft RAI Responses for May 2010 Public Meeting 48It was not clear to the NRC staff that significant credit could be taken forthe capture of fines in inactive volumes based on the accepted guidancepositions that a significant fraction of the fines would tend to be distributedto upper containment during blowdown, and that inactive volumes wouldbe filled early in the event, prior to the completion of the bulk of debris washdown.It was not clear to the NRC staff that significant retention credit could betaken for small pieces in upper containment, since small pieces of debrisare considered to be of a size small enough to pass through gratings.Based on NRC-sponsored washdown tests, it was recommended thatretention credit on gratings not be taken on debris smaller than the size ofthe grating opening. It is not clear to the NRC staff that sufficientretention credit is appropriate for other locations in upper containment.Entergy Nuclear Operations Response for 17c:With respect to the main 17c RAI, ENO believes that the sequencing of erodedfines after the introduction of smalls was appropriate and prototypical, referencefurther discussion provided in the response to RAI 26a. Timing for most erosionto reach the strainers would exceed the time for smalls to transport based onerosion trend data seen in Reference 19 versus tumbling velocities of the smalls.However, based on discussions to date with the NRC, certain elements of theexisting testing protocol will not be accepted by the NRC without refinements.For any future testing using a refined PCI flume test protocol, ENO would plan toadd the eroded fiber fines with the initial fiber fines to eliminate this potential NRC concern.The several paragraphs in RAI 17 that follow 17c seem to deal with a separateissue, namely fiber smalls added to the flume that did not transport should havebeen accounted for in the amount of eroded fines added. This issue is a genericPCI test protocol open item that may be discussed at a June 2, 2010 meetingbetween the PCI strainer testing team and the NRC. Palisades will follow theoutcome from the June 2, 2010 meeting in this area if covered. If the June 2,2010 meeting does not specifically cover this topic, then for any future testingusing a refined PCI flume test protocol, Palisades will add the appropriateamount of eroded fines to cover the amount of smalls entered into the flume,conservatively assuming none will transport.

Palisades Draft RAI Responses for May 2010 Public Meeting 49 References 1.EA-MOD-2003-0021-01.

2.BNZ Transite Specification Sheet.

3.BNZ Marinite I Specification Sheet.

4.ALION-REP-LAB-2352-218, "Marinite 1 Flow Erosion Testing Report", Rev. 0.5.Ontario Power Generation, "Jet Impact Tests-Preliminary Results andTheir Applications," N-REP-34320-10000-R00. April 2001.

6.Gordon H. Hart, P.E , "Summary of Calcium Silicate Insulation Types Usedinside Containment at US Nuclear Plant".

7.CS85-P-HT1.

8.Avallone, Eugene A. et. al.,"Marks' Standard Handbook for MechanicalEngineers", Ninth Edition.

9.ALION-REP-PAL-7199-22, "Palisades Cal-Sil Flow Erosion andDissolution Testing Report", Rev. 0.10. ALION-REP-ENT-7199-21, "Palisades Fiberglass Debris Flow ErosionTesting Report", Rev. 0.11. ALION-PLN-I006-02, "Erosion Testing of Small Low Density FiberglassDebris: Test Plan", Rev. 1.12. ALION-REP-LAB-2352-99, "Test Report: Erosion Testing of Mineral WoolInsulation", Rev. 0.13. ALION-REP-LAB-2352-77, "Test Report: Erosion Testing of Low DensityFiberglass Insulation", Rev. 0.14. ALION-REP-ENT-7199-21, "Palisades Erosion Assessment - AverageVelocity and Turbulence for Non-Transporting Debris", Rev. 0.NOTCURRENTLY ISSUED.15. NUREG/CR-6808, "Knowledge Base for the Effect of Debris onPressurized Water Reactor Emergency Core Cooling Sump Performance".16. NUREG/CR-6772, LA-UR-01-6882, "GSI-191: Separate EffectsCharacterization of Debris Transport in Water".17. BNZ Marinite M Specification Sheet.18. NUREG/CR-6224, "Parametric Study of the Potential for BWR ECCSStrainer Blockage Due to LOCA Generated Debris".19. ALION-REP-ALION-I006-04, "Erosion Testing of Small Pieces of LowDensity Fiberglass Debris - Test Report", Rev. 0.NOT CURRENTLYISSUED.

Palisades Draft RAI Responses for May 2010 Public Meeting 50 NRC Request 18.The June 30, 2009, supplemental response states that flow pathlines andvelocity isocontours were used to identify isolated eddies that hadvelocities higher than the incipient tumbling velocity but did not contributeto debris transport from the zone. A plot that appears to show minimaltrapping in eddies enclosed by or connected to transporting isocontoursfor a sample case was included in the supplemental response, but it wasnot clear whether this behavior is also characteristic of the limiting designcase. Therefore, for the limiting design basis conditions, please eitherdemonstrate that minimal credit for hold-up in such eddies was taken, orelse identify the types and quantities of debris assumed to be trapped ineddies of this sort, and provide the basis for not considering debrisassumed to be present in these areas at the switchover to recirculation astransporting to the strainers considering the following points:a. Even in steady-state flow problems, chaotic perturbations result invariance in the solution that will alter the flow pattern in isolated eddiesand allow fluid and debris elements in these eddies to escape as timeprogresses (or the number of computational iterations increases).b. Sophisticated turbulence models are expected to be necessary toaccurately predict the behavior of eddies if they are credited with theretention of debris. Please discuss the fidelity of the turbulence modelused in the computational fluid dynamics code and discuss whetherthe converged solution was run further and checked at variousintervals after convergence was reached to demonstrate evidence ofthe stability of any eddies credit debris hold up.c. Suspended phases and floor-transporting debris do not preciselyfollow streamlines of fluid flow. Phase slip can be particularlysignificant when the streamlines exhibit significant curvature, such asin an eddy.d. There are significant uncertainties associated with modelingblowdown, washdown, and pool fill transport mechanisms. The initialdebris distribution at switchover can vary significantly.Entergy Nuclear Operations Response:The debris transport quantity is determined by examining the iso-contour area ofthe tumbling velocity of the debris under consideration. Some iso-contour areaswere neglected in determining the transportable debris quantity. However thisprocedure was only applied to smalls debris. These iso-contour islands wereneglected because the iso-contour was not contiguous with the strainer. Ingeneral a significant distance exists between the accounted for iso-contour and Palisades Draft RAI Responses for May 2010 Public Meeting 51the neglected iso-contour islands. The minimum distance between any neglectediso-contour island and the closest accounted for iso-contour area in Zone 1(CWST) is greater than 5 ft. There is a very small neglected iso-contour island inZone 4 where the distance to the next connected contour is only 2 ft but theneglected area is only 2% of total iso-contour area in that zone and an evensmaller fraction of the overall iso-contour area.If the analysis would have accounted for all the neglected iso-contour islands, anincrease of 15% in the amount of small fiber would result. Coupled with this, adecrease of 1.5% in the fiber fines is also caused since the increased transport ofsmalls leaves less fiber behind that must be eroded. The neglect of 15% of thefiber smalls put an additional 1.5% of fiber fines into the test debris load. There istherefore a balance between conservatism in the amount of fines andconservatism in the amount of smalls when considering the neglected smallstransport. In addition, the isolated contours neglected are indeed unlikely totransport across the low velocity area that disconnects them from a transportableiso-contour.Furthermore, isolated iso-contour islands for the Palisades conditions do notconsist of turbulent eddies exhibiting enclosed recirculation. The isolated islandsignored are due to one of two conditions. The first condition (type 1) is due to anisolated drainage source that locally generates velocities in excess of thetumbling velocity but diffuses before getting close to a transporting iso-contourarea. The second condition occurs where a portion of the flow passes through arestriction which speeds up the flow locally but does not result in connection to atransporting iso-contour. Figure 5 shows the neglected iso-contour portionscolored in red. Adjacent to the iso-contour plot within Figure 5 is a pathline plotwhich illustrates the general flow field within the containment sump. Singlearrows in the iso-contour portion of the figure identify type 1 isolated contourareas (due to flow sources). Double arrows linking the two portions of the figureidentify type 2 isolated contour areas. The pathline plot underlines that none ofthe isolated contour areas correspond to recirculating eddies whose stability wasquestioned in the request for addition information.

Palisades Draft RAI Responses for May 2010 Public Meeting 52Flow sources fromaboveFigure 5. Origin of isolated iso-contour islands [Ref. 18.1]

References18.1 AREVA Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport Palisades Draft RAI Responses for May 2010 Public Meeting 53 NRC Request19. Please provide cumulative transport percentages for each type of debristhat are integrated over all regions of the containment pool for the limitingdesign basis case. Please specify for which break the results areapplicable.Entergy Nuclear Operations Response:Below is a copy of Table D-1 from the Palisades Transport analysis, [Ref. 19.1] done forthe November 2008 Flume test that gives percent of debris transported. It should be usedwith Table 3e7 on page 81 of the June 30, 2009 submittal for limiting break S5. It can beused with Table 3e8 for break S6 which was used for enveloping the CalSil insulation.

References19.1 AREVA Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport Palisades Draft RAI Responses for May 2010 Public Meeting 54 NRC Request 20.The June 30, 2009, and February 27, 2008, supplemental responsesindicate that a blowout panel has been installed to impede air flow into theair room while simultaneously preventing water hold up. The NRC staffnoted that the transport analyses for Palisades appeared to have beenperformed with the air room door closed (NRC staff could not determinewhether this referred to the blowout panel or another door). Please clarifywhether this assumption refers to the blowout panel or another door, andadditionally confirm that the assumption of the air room door being closedeither results in conservative debris transport results or is a validassumption.Entergy Nuclear Operations Response:The Air room doors and the Air Room Blowout Panels are separate anddifferently designed items. The assumption applies to a door, the blowout panelis always assumed to be open. The below provides the details reasons why.The air room extends from 590' Elevation upwards to approximately 625'Elevation and encompasses the containment personnel lock and the portion ofthe North Stairwell from the lock to the 590' Elevation floor which encompasses apart of the ECCS sump. Its function was initially to allow more rapid purging ofthat space with outside air and shorten the time for personnel to gain access tocontainment. Much of the Primary Coolant System instrumentation is mountedon the walls of this space and quick access was judged to be desirable. AfterTMI the ability to purge containment was severely restricted and this function wasno longer very useful. However, the air flow restriction was utilized in the plantfire analysis to help restrict the progress of a fire so the air room is beingmaintained for that purpose.On 590' Elevation there are 2 doors which swing inward so that water pressure ofaccumulating Post LOCA water will tend to force them into their jams (load not onthe latches) and hence would require that the doors buckle in order to open. Thedoors are standard 3 by 7 foot industrial strength steel doors set in steel frameswhich are anchored to reinforced concrete. One of the doors faces West and thispath is a long way from the strainers. The other faces East and separates thewater sources falling inside the air room from the AB bank of strainers thereforehaving a potential impact on the path of the water flow on 590' Elevation. and onthe path of the debris which is assumed to be falling with the water. Each ofthese doors is equipped with an anti-shine L shaped radiation barrier on theoutside of the air room. This tends to affect the flow streams and tends to reducethe effect of opening the door in the analysis.Directly to the north of the West Door are 2 Blowout Panels. [Ref. 20.1] Thesepanels are each 6 by 10 feet thus being about 3 times the size of a door (60 sq ft Palisades Draft RAI Responses for May 2010 Public Meeting 55vice 21 sq ft) and are set in the 2 foot thick concrete north shield wall and begin 6inches off the 590' Elevation floor. The blowout panels are only 2 feet apart andthe first one is only 1.5 feet north of the West Door [Ref. 20.2]. The blowoutpanels are half inch thick Marinite and the panels are cut and suspended tofacilitate easy blow out due to air pressure or breakage due to water pressurelow on the panel [Ref. 20.3]. The blowout panels have no radiation shine baffles.The blowout panels are assumed to be open in the analysis since they areengineered to be in that condition following a large LOCA due to break pressureeffects and due to water differential pressure for a small LOCA.Palisades has run numerous CFD cases while preparing for a "Test for Success"strategy during the November 2008 flume testing campaign. In general the airroom doors were assumed closed since that was known by past CFD analysis tousually be the worst case. It has been shown that the position of the West Doorhas almost no effect due to the 2 blowout panels being open and having openspace 4 times the door.The position of the East Door has relatively little effect on breaks on the "A"Steam Generator side (Pressurizer side) because the break flow and hence mostof the debris falls a long ways away. For "B" Steam Generator Side breaks it hasa major effect and for this reason was explicitly run in the transport analysis foreach situation, [Ref. 20.4]. Thus there was a three break set of "limiting breaks"always run. They were Hot leg A side, Hot leg B side, and Hot leg B with Eastdoor open. Thus actual amounts of debris on the screen were calculated for allthree cases and the output was used to judge the worst case for flume testing.For the November 2008 Tests the S5 Hot Leg A (East air room door closed)break was chosen as the worst break, however, the S6 Hot Leg B (East air roomdoor closed) had more CalSil so the CalSil from that break was used to"envelope" the situation and avoid having to make a judgment on the trade offquantity of fine fiber for fine particulate.

References20.1 Palisades Plant Drawing C-7320.2 Palisades Plant DrawingC-14420.3 Palisades Modification EAR-2003-0176, MOD-2003-002120.4AREVA Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport Palisades Draft RAI Responses for May 2010 Public Meeting 56 NRC Request 21.The NRC staff understood that the head loss testing conducted by PCImodeled flow conditions during the recirculation phase of a LOCA andmodeled all debris (other than a fraction of the latent debris added with therecirculation pump stopped) as entering the containment pool oneflume-length (nominally 30 feet) away from the containment sumpstrainers. Flow conditions during the pool-fill phase of the LOCA were notmodeled in the testing, nor was the potential for debris to enter thecontainment pool closer than one flume-length from the strainer due to theeffects of blowdown, washdown, and pool fill transport. The lack ofmodeling of these two transport aspects of the head loss testing appearedto result in a non-prototypical reduction in the quantity of debris reachingthe test strainer and, ultimately, non-conservative measured head lossvalues. This is a significant issue because of the large settlement creditallowed during the head loss test. Please provide the technical basis fornot explicitly modeling transport modes other than recirculation transport,considering the following points:a. As shown in Appendix III of he NRC staff's SE on NEI 04-07,containment pool velocity and turbulence values during fill up mayexceed those during recirculation, due to the shallowness of the pool.Some debris that would not transport during recirculation maytransport during the pool-fill phase. In addition, latent debris on thecontainment pool floor could be stirred into suspension by thesehigh-velocity, turbulent flows, unlike the latent debris added to thequiescent PCI flume.b. The pool fill phase will tend to move debris away from the locationswhere it washes down to the 590' EL, and a fraction of this debriswould be moved nearer to the recirculation sump strainers.c. Representatively modeling the washdown of some fraction of thedebris nearer the strainer than one flume-length away would beexpected to increase the quantity of debris transported to the strainerand the measured head loss, since a shorter flume would offer lessopportunity for debris to settle. This statement applies both to debristhat tends to settle in the head loss test flume, as well as debrisconsidered to settle analytically.Entergy Nuclear Operations Response:Figure 21.1 shows the modeled inlets for the Palisades containment sump [Ref.21.1]. Figure 21.1 shows that in the compartmentalized containment of Palisadesspray and break-flow run-offs combine and split into many different locations.Debris allocation to the sump is by proximity zone where the amount of debrisput into each proximity zone is proportional to the flow into it. The procedureemployed in the debris transport calculation follows recommended guidelines inNEI-04/07 [Ref. 21.3, 21.4]. There is also is a dependence on where the floworiginates and how much debris is at that location from debris generation. The Palisades Draft RAI Responses for May 2010 Public Meeting 57proximity zones are shown in Figure 21.2. Although pool fill-up is not explicitlyaccounted for by analysis, debris is distributed evenly in the proximity zoneseven though the debris enters the zones from particular areas. As an example,the area marked K&L on Figure 21.1 receives 30% of the overall recirculationflow. Conversely inlet Q receives only 1% of the flow. The pool fill phase isaccounted for by distributing the debris evenly in proximity zone 1 which largelymoves the debris closer to A & B strainer trains. As discussed further below, theopening between the containment wall and the CWST room wall near C & Dstrainer trains only carries 19% of the total CWST outflow and therefore does notrepresent a major path for debris to the C & D strainer trains. The distribution ofdebris evenly in proximity zone 1 therefore conservatively accounts for pool-filltransport of the debris.Figure 21.1 CFD domain flow inlets [Ref. 21.1]

Palisades Draft RAI Responses for May 2010 Public Meeting 58Figure 21.2. Proximity zones for Palisades containment sump [Ref. 21.1]Note that similar arguments and conservatisms are implied in assuming debris isdistributed evenly in the remainder of the proximity zones. Finally if the total flowinto Zones 4-6 is compared to the surface area of Zones 4-6, it is seen thatZones 4-6 receive 54% of the flow while occupying 60% of the total surface area.This implies that the general flow direction upon pool fill will be out of Zones 2and 3 which contain all four strainer trains towards Zones 4-6. The debristransport calculation conservatively does not account for this preferentialmovement of debris away from the strainers. Neglecting cross-zone debristransport during pool fill, as done for the debris transport calculation, isconservative relative to the plant condition.To further quantify the conservatism contained in adding debris at 30 ft from thestrainer, an analysis was conducted to conservatively estimate the distancedebris has to travel to reach each of the strainer trains. To arrive atrepresentative averages for the distance traveled, the overall flow paths and flowsplits were respected while drawing linear segments following these paths toconservatively estimate the total distance traveled by debris.The general flow-splits among the proximity zones is shown in Figure 21.. Thearrows indicate the flow direction across each of the boundaries. Based oncomparison of the calculated debris transport quantities, the bounding scenariofor debris transport is obtained when considering the east air room door closed Palisades Draft RAI Responses for May 2010 Public Meeting 59and therefore, there is no flow from zone 6 directly to zone 3. Note however alsothat only 25% of the flow comes from the intersection of zones 2 and 3.2.0 ft 3/s0.5 ft 3/s4.5 ft 3/s0.6 ft 3/s0 ft 3/s(door closed)1.1 ft 3/s1.7 ft 3/sFigure 21.3 Proximity zone flow splitNote that the only Zone exhibiting two exit flows is Zone 1. Zone 1 is thereforesubdivided by the relative flow fractions exiting from each side. Figure 21.therefore shows a dividing line which apportions 81% of the Zone 1 area to theexit near strainer banks A & B and 19% to the exit near strainer banks C & D.Pathlines indicate that none of the flow from the CWST room exit near strainerbanks C & D actually goes to strainer banks A & B which is expected based onthe overall layout of the strainers. Furthermore, the flow-source indicated as letterM in Figure 21.1 (located within proximity Zone 2) contains ample flow volume tomake up the cross-flow from Zone 2 to Zone 3. No flow from Zone 4-6 thereforeis considered to be required to make up the strainer train flow demand in Zone 3.With this information the distances traveled from each of the proximity zones tothe strainer banks can be appropriately weighted.To define the average origination point for the debris, the centroid of the relevantarea is employed. Fine debris is assumed to be uniformly distributed within eachproximity zone and so the centroid of the complete proximity zone is employedfor Zones 2-6 to define the fine material origination point. For Zone 1, multipleexit flow points exist and the zone is subdivided to apportion the zone by theCFD predicted flow split (81%/19%) as discussed above. Fine debris from Zone Palisades Draft RAI Responses for May 2010 Public Meeting 601 is therefore considered to originate from the centroid of each of the two subzones comprising Zone 1. For small debris, the velocity iso-contour centroidswithin each zone are employed to describe the origination point of the debris. ForZone 1, the iso-contours are split along the overall sub-zone boundary and thenanalyzed for centroids. The two separate iso-contour centroids for Zone 1 arethen used to describe the origination point for the debris within each of the sub-zones of Zone 1. The destination point for the strainer train is the centroid of thestrainer module train. In some cases the centroid corresponding to the originationpoint lies in a non-fluid zone (e.g. inside a CWST tank). In this case, theorigination point is moved on an arc to the nearest intersection between theboundary of the fluid zone and the arc. The center point of the arc is the previousdistance point drawn from the strainer origin. This method needed to be appliedonly to Zone 1A, both for smalls and fines. The best graphical example is shownin Figure 21.4 for the treatment of the centroid of all of Zone 1A (since fines are100% transportable in each zone and uniformly distributed upon generation).Debris quantities are derived from the debris transport calculation [Ref 21.1].Flow rates through each zone are derived from those indicated in Figure 21.3,which stem from the CFD calculation employed to perform the transportcalculation [Ref. 21.1]. Within each zone the debris and flow is divided among themodules according to the module flow rate. For example, in Zone 3, 0.6 ft 3/s offlow originates from Zone 2 going to Zone 3. Since both strainer trains A & Bdraw the same flow, half of this flow is apportioned to strainer train A and half tostrainer train B. For Zone 2, the situation is slightly different as strainer train Chas a higher flow capacity than strainer train D. But the debris and flow are againapportioned according to the relative flow between the two strainer trains.The following equation (1) calculates the average distance a given type of debristravels to a strainer train i.

(1)In the above equation is the flow mapped to strainer train i from Zone Z.

isthe debris quantity mapped to strainer train i from Zone Z. is the distancedebris has to travel from Zone Z to strainer train i. The summation uses Zone 1Afor strainer trains C & D and Zone 1B for strainer trains A &B. Following equation(1), an average distance can be calculated for all strainer trains. An overallaverage distance is then derived from these four distance measurements byaveraging all four distances weighting each of the distances by the flow to eachstrainer train. Should a given combination of transport and strainer train not bepossible due to the viable flow paths available, the entry is marked "NA" in thetables below. For example, since no flow from Zone 4 goes to strainer trains A &B, the table entries for transport from Zone 4 to strainer trains A&B are marked"NA".

Palisades Draft RAI Responses for May 2010 Public Meeting 61Further explanation is necessary with respect to the flow weighting employed inEquation (1). The flows used in the weighting are the flows transporting debris tothe strainer train in a given area. This means that for strainer train C, forexample, the flow rate weighting for debris in Zone 2 (the location of strainer trainC) is equal to the full strainer train flow rate since all of the flow going to thestrainer could transport debris. Conservatively, this approach weights the debrisdistances for debris located within the same zone as the strainer train quiteheavily. Note also for example that the flow weighting for debris originating fromZone 4 is relatively high since the flows from Zone 5 and Zone 6 run throughZone 4 and could therefore transport debris from Zone 4.The calculation also respects the applied limitation for small debris that thisdebris cannot bridge a separation between iso-contours of its tumbling velocity.This means that although strainer trains A and B receive flow from Zone 2, thedistance from the transportable iso-contour in Zone 2 to strainer trains A and B isnot used. The iso-contour within Zone 2 is not connected via a viable flow path tothe iso-contour in Zone 3. This treatment is only required for non-fine debris andfor this debris transport calculation and analysis only applied to fiberglass smalls.Figure 21.4 shows the drawing used to calculate the distances for fine fiberglassdebris from the various zones to each of the strainer trains. Table 21.1 shows theresults of the calculations using Equation (1) and then performing an overall flowweighted average to obtain a final representative measure of the distance thatdebris has to travel to the strainer. The table indicates that fine fiberglass debrisis required to move an average of 45 ft to reach the strainer trains. Thecalculation shown was repeated for the distribution of fine mineral wool debrisand a distance of 45 ft was again calculated. Results for the mineral wool finedebris are not summarized in a table. Considering the conservatism used indistributing debris for the pool fill phase, the distance clearly underlines theconservative test approach of adding debris at 28-30 ft from the strainer.

Palisades Draft RAI Responses for May 2010 Public Meeting 62Figure 21.4 Fines debris distancesTable 21.1 Averaging summary for fiberglass fines debrisFigure 21.5 shows the drawing used to calculate the distances for small fibrousdebris from the various zones to each of the strainer trains. Table 21.2 shows theresults of the calculations using Equation (1) and then performing an overall flow Palisades Draft RAI Responses for May 2010 Public Meeting 63weighted average to obtain a final representative measure of the distance thatdebris has to travel to the strainer. The table indicates that small fibrous debris isrequired to move an average of 31 ft to reach the strainer trains. This distanceexceeds the debris addition distance used of 30 ft. It should be noted that pool-filltransport processes will largely result with small debris on the floor and transportalong the containment floor by tumbling is significantly more difficult thantransport in the top of the water column where the fibrous debris is insertedduring testing. Adding debris at 30 ft is therefore a justifiable approach to use in testing.Following additional discussions with NRC staff (i.e., 4/26/2010 Technical Calland 4/28/2010 Meeting with PCI on Large Flume test Protocol), the aboveapproach employed to determine the average distance traveled by debris wasdeemed insufficient to justify adding debris at 30 ft from the strainer. In order toaccept the above analysis, information would be required to show that the likelyvarying transport fraction for debris outside and inside of 30 ft from the strainerswill result in a conservative amount of debris transported to the strainers. For anypotential tests considered going forward, debris will not be added at the end ofthe flume only. The methodology for determining where debris would be added inpossible future testing will be discussed with the staff in the framework of themeetings with PCI on the Large Flume Test Protocol, the next meeting which isscheduled to occur in June 2010.

Palisades Draft RAI Responses for May 2010 Public Meeting 64Figure 21.5 Fiber smalls debris distancesTable 21.2. Averaging summary for fiber small debris References21.1 Areva Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport Palisades Draft RAI Responses for May 2010 Public Meeting 6521.2 June 30 Submittal to NRC Palisades June 30, 2009 Submittal to NRC,Follow-up Supplemental Response to NRC Generic Letter 2004-0221.3 NEI 04-07, NEI Guideline, "Pressurized Water Sump PerformanceEvaluation Methodology," December 2004.21.4 SER-GSI-191 SE, Revision 0, "Safety Evaluation of NEI Guidance onPWR Sump Performances", Office of Nuclear Reactor Regulation,December 2004.

Palisades Draft RAI Responses for May 2010 Public Meeting 66 NRC Request 22.Please discuss any sources of drainage that enter the containment poolnear the containment sump strainers (i.e., within the range of distancesmodeled in the head loss test flume) and identify their locations and flowrates. Please identify whether the drainage would occur in a dispersedform (e.g., droplets) or a concentrated form (e.g., streams of water runningoff of surfaces). Based on the June 30, 2009, supplemental response, itappears that sources of drainage were not modeled in the head loss testflume. The NRC staff expects that the lack of modeling of these drainagesources led to non-prototypically low transport results in the test flume.Therefore, please provide contour plots of the calculated turbulence(which include a numerical scale with units) for the CFD calculation for thetest flume and compare it to that for the full-containment plant CFDcalculation. The NRC staff does not consider the licensee's theoreticalarguments supporting a higher level of turbulence in the linear flume ascompared to the plant to be convincing in light of CFD comparisons forother plant conditions to flume flows at corresponding velocities that havetypically shown the plant condition to experience significantly higherturbulence.Entergy Nuclear Operations Response:Figure 21.1 shows the spray and drainage sources into the containment sump.Note that the areas designated as AA represent direct containment spray and aretherefore by definition inlets in the form of droplets whose near surfaceturbulence impact has negligible effect on debris transport. Three other areaswith flow input, marked X, M and P have boundaries within 30 ft of the strainertrains A-D. Of these, X is extremely weak at less than 0.1 ft 3/s. P represents aminor run-off flow of about 0.35 ft 3/s, originating from grating under thepressurizer and is therefore justifiably modeled as dispersed input flow withlimited turbulence generation potential [Ref. 21.5]. This leaves area M (flow at 0.76 ft 3/s) to be considered which represents flow from the steam generatorcompartment. More than half of area M is covered above with grating. Althoughsome of the water could remain concentrated, this is improbable given theunstable nature of a water sheet and the near proximity of the broken up waterdroplets from the flow that has passed through the grating. Note also that theflow not affected by grating only amounts to 5% of the total recirculation flowfurther reducing the potential impact on debris transport. Finally, the edge ofspray / drainage area M is more than 25 ft away from the nearest strainer bankand therefore barely within the modeled length of the flume. Thecompartmentalized nature of the Palisades containment distributes the break-flow energy effectively throughout containment and results in relatively quiescentconditions in the containment sump.

Palisades Draft RAI Responses for May 2010 Public Meeting 67The quiescent conditions of the containment sump are illustrated by theturbulence contours shown in Figure 22.1, Figure 22.2 and Figure 22.4. Thefigures show turbulence at two different depths to illustrate the diminished roleturbulence is expected to play in debris transport, as explained below. The lowerlimit of the scale in Figures 22.2 and 22.3 corresponds to the expected midwater-column turbulence level in the flume based on initial calculations. Thisturbulence level is not corrected for the difference between containment andflume temperatures. Note that the maximum turbulence level outside of thestrainers themselves is approximately 0.01 ft 2/sec 2 which is quite low from anabsolute standpoint. A contour plot illustrating the full range of turbulence incontainment is shown in Figure 22.1. Note that the 0.01 ft 2/sec 2 level is onlyattained in a very small area of containment. A more focused range in thecontour plot is shown in Figure 22.2, illustrating that the bulk of Palisadescontainment is at turbulence levels far below even 0.001 ft 2/sec 2. The previoussupplemental response [Ref. 21.2] indicated that the turbulence levels betweenflume and containment would be expected to be similar, if not higher in the flumebecause of the presence of shear producing walls. This statement is correct forcases where there is no significant source of turbulence near the strainers thatwould allow turbulence to be transported to the near field of the strainers. Due tothe compartmentalized nature of the Palisades containment, there are no strongsources of turbulence due to break or spray flow in containment. For the A&Bstrainer trains, the strongest source of turbulence occurs due the 90 degree turnthat the flow has to execute to leave the CWST room and make it past thecontainment NaTB baskets to the strainers. For the C&D strainer trains, thestrongest source of turbulence is from the restriction of flow originating from thewest side of containment and then turning toward the strainers.

Palisades Draft RAI Responses for May 2010 Public Meeting 68Figure 22.1 Turbulent kinetic energy contours (ft 2/sec 2) at half-foot water depth Palisades Draft RAI Responses for May 2010 Public Meeting 69Figure 22.2 Turbulent kinetic energy contours (ft 2/sec 2) at 1in water depthThe dominant turbulence producing features are thus bluff bodies located (e.g.NaTB baskets, columns etc) that lie within a concentration of flow. Theturbulence is not associated with concentrated sources of water entering thecontainment sump or break flow energy. The fact that the source of turbulencecan be traced back to bluff-body type separations is important since it implies theexistence of an associated low velocity region behind the bluff body. Theunsteadiness in the flow moving around the bluff bodies would tend to over timeaccumulate debris in the low velocity region. This drop-out and low velocityregion is ignored in the conducted transport analysis and associated testing.Once the turbulence is generated, it is carried toward the strainer. However, incontrast to the test flume, as Figure 22.2 and Figure 22.3 imply, the turbulencenear the floor becomes significantly lower than further up in the water column. Inthe test flume the flow-guiding walls and floor provide a concentration ofturbulence near the floor likely reaching to approximately 5e-4 ft 2/sec 2 (green inFigure 22.2 and Figure 22.3). Since debris transport is predominately expected tooccur in the lower parts of the water column, it is turbulent kinetic energy in theseregions that is expected to be the best measure of the likelihood that debrissettling will be prevented (or debris suspension enabled) by turbulent kineticenergy. Flume turbulence levels are higher near the floor relative to the mid Palisades Draft RAI Responses for May 2010 Public Meeting 70water column in the flume whereas containment turbulence levels near the floorare lower near the floor relative to the mid water column.Furthermore, areas of relatively high turbulence are small on a recirculation flowbasis when considering that a large portion of the overall recirculation flow isdrawn in by strainer banks C & D (65%). Areas bounded by the expected flumeturbulence level are even larger for these two strainer trains.Figure 22.3 Turbulent kinetic energy contours (ft 2/sec 2) at half-foot water depthIn summary therefore, it can be expected that effective turbulence levels(accounting for the difference in temperature and associated material settlingrates between conducted testing and expected containment conditions) of near 1e-3 ft 2/sec 2 will be seen in a comparison between flume turbulence levels andcontainment turbulence levels. The 1e-3 ft 2/sec 2 turbulence level bounds adominant portion of the approach areas employed by the flow on its way to thestrainers. While somewhat larger turbulence levels exist in small areas ofcontainment on selected approaches, the 1e-3 ft 2/sec 2 turbulence level is aprototypical turbulence level for flow approaching the Palisades strainers. In viewof this, Palisades proposes to do a CFD of the Palisades flume test setup whichcan be compared with the above plots. It is anticipated that such an analysis willshow the flume and the containment are comparable.

Palisades Draft RAI Responses for May 2010 Public Meeting 71Transport conditions for several breaks were considered before determining thatbreak S5 with East Air Room Door closed was limiting. Based on discussionswith NRC staff (i.e., 4/26/2010 Technical Call and 4/28/2010 Meeting with PCI onLarge Flume test Protocol), additional justification must be provided for thediffuse flow boundary conditions employed in the near field of the strainers. It isanticipated that a) additional information with respect to containment geometryand b) considerations of water sheet atomization regardless of structure impactwill allow justification of the boundary conditions employed and/or provide clearprototypical treatment of all run-off flows. Considerations of b) will be discussedwith the staff going forward in the framework of the meetings with PCI on theLarge Flume Test Protocol, the next meeting which is scheduled to occur in June 2010.References22.1 AREVA Calculation 32-9099369-000 dated February 2009, PalisadesNuclear Power Station - Debris Transport22.1 Palisades June 30, 2009 Submittal to NRC, Follow-up SupplementalResponse to NRC Generic Letter 2004-0222.3 NEI Guideline, "Pressurized Water Sump Performance EvaluationMethodology," December 2004.22.4 SER-GSI-191 SE, Revision 0, "Safety Evaluation of NEI Guidance onPWR Sump Performances", Office of Nuclear Reactor Regulation,December 2004.22.5 Hur, Hong B., "Modeling a rain-induced mixed layer", Thesis, NavalPostgraduate School, Monterey, CA, June 1990.

Palisades Draft RAI Responses for May 2010 Public Meeting 72 NRC Request 23.The June 30, 2009, supplemental response described the licensee'smethodology of averaging velocities along different approaches to thestrainer modules in order to determine the flow conditions for the headloss test flume for the Palisades strainer testing. During the chemicaleffects audit for Palisades, the NRC staff also considered thismethodology based on the more detailed descriptions and results from theCFD modeling report. Based on this information, the NRC staffconsidered the licensee's methodology for determining the head loss testflume flow conditions as lacking adequate justification and appearing non-conservative. In particular, the average velocities calculated for a numberof the cases and configurations included approaches to the strainer thatexperienced relatively little flow of water and debris. The velocitiesassociated with these relatively stagnant approaches that appeared tohave little impact on debris transport to the strainer were arbitrarilyweighted equally with higher velocity pathways by which the majority ofthe water flow and generated debris appeared to reach the strainers. Thispractice appeared to result in non-prototypically low flume velocities forstrainer testing, leading to increased debris settling and lower head lossesthan expected for the plant condition. Therefore, please provide thefollowing information to justify the velocities chosen for the head loss test flume:a. Velocity contour plots for the containment pool, including close-upplots in the vicinity of the strainers, as well as a table of the velocitiesused in the head loss test flume for comparison.b. Justification for weighting stagnant approaches to the strainers, alongwhich little debris transport occurs, equally with high-velocityapproaches by which the majority of debris transports to the strainers.Entergy Nuclear Operations Response:The requested flume approach velocity profile as a function of distance from thetest strainer is given in Table 23.1. The velocities given in the table are a result ofapplying the methodology described previously [23.1]. Additional details andjustification for the approach definitions employed is illustrated and explainedbelow. The approach to applying the methodology to the parallel strainer trains A,B, C and D was to consider A and B strainer trains as one ensemble and C andD, each separately. The discussion below will detail how an approach velocityprofile was derived for each of these cases. To obtain the overall averageapproach velocity given in Table 23.1, a strainer module flow based average wasemployed where the weighting of a given ensemble's (A&B, C or D) averageapproach velocity was equal to the number of modules contained in the Palisades Draft RAI Responses for May 2010 Public Meeting 73ensemble. Since each module draws the same amount of flow, the resultingaverage is a flow-weighted average.Table 23.1 Flume approach velocityFigure 23.1 shows a velocity contour plot near strainer trains A & B. The contourplot is taken at an elevation of 591 ft, 1 ft off the containment sump floor. The fourapproaches are indicated by sets of arrows on the figure. The arrows show thatApproach 1 and Approach 3 are relatively short approaches whereas the majorapproaches are Approach 2 and Approach 4. For areas where only twoapproaches exist (beyond 7 ft from the strainer) the higher velocity approach(Approach 4) is weighted double. Note that Approach 2, although exhibiting a lowvelocity contains more than a quarter of the amount of flow relative to the overallflow going to strainers A & B. By double weighting Approach 4, the fastestapproach is appropriately emphasized relative to Approach 2. Note also that theapproaches for strainers A & B only make up 35% of the overall recirculationflow. Approach 2 shows arrows running across a cylindrical tank. Some of thevelocity stream does divert under the tank to approach the strainers and this pathis represented by the area weighted average velocity of this approach.

Palisades Draft RAI Responses for May 2010 Public Meeting 74Approach 1Approach 3Approach 4Approach 2Figure 23.1 Strainer trains A&B approaches (Velocity - ft/s)Approach 1Approach 2Approach 3Approach 4Figure 23.2 Approach descriptions for strainer bank C (Velocity - ft/s)

Palisades Draft RAI Responses for May 2010 Public Meeting 75Figure 23.2 shows the approaches used to describe the approach velocity tostrainer train C. Note that Approach 1 and Approach 2 are short and theapproach velocity is dominated by Approach 3 and Approach 4. The fastest ofthese (Approach 4, generally) was weighted double in the determination of theaverage approach velocity for strainer train C.Approach 1Approach 2Figure 23.3 Approach descriptions for strainer bank D (Velocity - ft/s)Figure 23.3 shows the approaches used to describe the approach velocity tostrainer train D. Note that Approach 1 is fairly short and the approach velocity isdominated by Approach 2. Approach 2 is also the faster approach and istherefore double-weighted for the cases where two approaches are observed.Reviewing Figure 23.1 to Figure 23.3, the contour plots show that while someareas exist with velocities greater than those tested in the flume, these areas arerelatively small and the flume velocities employed are prototypical of those incontainment approaching the strainers.

References23.1 Palisades June 30, 2009 Submittal to NRC, Follow-up SupplementalResponse to NRC Generic Letter 2004-02.

Palisades Draft RAI Responses for May 2010 Public Meeting 76 NRC Request 24.In RAI 13 from the NRC's letter dated December 24, 2008, the NRC staffrequested information regarding the postulated single failure of a low-pressure safety injection (LPSI) pump to automatically trip followingreceipt of a recirculation actuation signal. After reviewing the licensee'sresponse in the June 30, 2009, supplemental response, the NRC staff didnot consider the issue fully addressed with respect to debris transport forthe following reasons: (1) modeling the failure of a LPSI pump to tripautomatically would likely lead to additional debris transport in the headloss test flume, as well as in the analysis, (2) the NRC staff expects thatthe transport of additional fines and/or small pieces would have increasedthe head loss for the Palisades design basis test, and (3) it is not likelythat the LPSI pump flow would be terminated in time such that thetransport effects from this flow can be completely neglected. Thesupplemental response argues that 15 minutes is an appropriate time forcrediting remote operator actions to ensure that the LPSI pump flow isterminated. However, because the failure mechanism and necessaryrecovery action (e.g., tripping the pump breaker or remotely tripping thepump) would not be known ahead of time, based on the informationprovided by the licensee, it was not clear to the NRC staff that the remoteoperator actions could reasonably be accomplished in 15 minutes. It isnot clear, for example, that the first remote operator action would besuccessful for all potential failure modes, or that applicable proceduresinclude sufficiently detailed guidance for the operators. Please provideadditional information to address the NRC staff's concerns and state howmuch time is required for one turnover of the containment pool volume inthe case of a LPSI pump failure to trip.Entergy Nuclear Operations Response:Palisades approach for addressing the LPSI failure to trip scenario was analytical(Reference 24.1) with consideration of strainer test results based on designmaximum flow with no LPSI pumps running. Main reason for this approach isthere was no way to adequately model such a scenario using the existing testprotocol that requires introduction of all fine particulate before fine fiber additionfollowed by the small debris. Use of the LPSI failure to trip flow rate wouldrequire introduction of all non-chemical debris before the flow rate could bereduced, simulating the subsequent termination of the LPSI pump running,before the chemical debris is added. Such a test would not be representative ofwhat would occur in the plant with respect to debris transport and would beextremely conservative. Given the low probability for a LPSI failure to trip tooccur (estimated ~ 0.001), Palisades performed strainer testing using flow ratesand velocities that were representative of the maximum design flow rateassuming with no LPSI flow.

Palisades Draft RAI Responses for May 2010 Public Meeting 77The strainers have a 2.6 ft water head loss allocated to address both cleanstrainer head loss and debris head loss. The clean strainer head loss, (1.026 ft head loss at 212 oF), can be readily calculated for the increased flow for a LPSIfailure to trip scenario. The assumed debris head loss for the LPSI took intoconsideration the measured head loss from the design basis strainer test toestimate what the corresponding head loss might be for the starting point of theLPSI failure to trip evaluation.The November 2008 tests for the design basis debris loading resulted in 0.44feet of head loss with all but the chemical precipitate debris in the flume(Reference 24.2). The design basis debris load used in the November 2008testing is the full 30 day loading.The EOP's direct the operators to manually trip the LPSI pumps if they continueto operate post RAS. The trip verification happens very soon after RAS andsimulator experience shows that a running LPSI pump could be tripped or itssupply bus could be tripped within 15 minutes of RAS. Note that this 15 minutetime frame is not a defined time critical operator action and specific timevalidation has not yet occurred. The current EOP direction is detailed for eachaction to be taken and the component the action is to be taken on. Verifying thatthe LPSI pumps have tripped post RAS is the first step in EOP Supplement 42.The first attempt for tripping the operating LPSI pump is using the control switchin the control room. This would be performed and if successful, the LPSI pumpwould be expected to be running less than 5 minutes post-RAS. If the controlroom control switch is unsuccessful, the next step in the procedure woulddispatch an operator to trip the pump at the LPSI pump bus breaker. If thesecond effort is also unsuccessful, then the associated safety bus is tripped inthe control room. This last step is also relatively quick. If Palisades were to godirectly to tripping the safety bus in the control room if the pump control switchfailed, the entire effort is then estimated at less than 10 minutes. This couldeliminate any uncertainty in timing associated with tripping the pump at the LPSIpump bus breaker. Palisades proposes to time validate the current EOP actionsand if difficulties exist in completing actions within 15 minutes, then considerationwill be given to going directly to tripping the safety bus in the control room if thepump control switch failed.The sump contains approximately 30,000 cu ft of water equal to approximately225,000 gallons. At approximately 7,000 gpm flow with the LPSI pump running,one turnover is approximately 30 minutes. Therefore 15 minutes isapproximately one half turnover.It is known that the following portions would not be on the strainer at 15 minuteseven with the additional flow:1. Chemical Precipitates Palisades Draft RAI Responses for May 2010 Public Meeting 782. Eroded fibers (approximately 1/3 or more of the total fine fibers, Reference RAIResponse 26, Table26a-1)3. Failed paint particulate and chips (do not fail instantly as shown by test & mostof the volume is unqualified paint not in the ZOI of the break)4. A significant fraction of the Calcium Silicate particulate (fines form by erosion)The combination of less than 2/3 of the sensitive fiber, less than half of the paintrelated particulate and significantly reduced Cal-Sil fine particulate, whencombined with one half turnover of the sump volume suggests that assumingonly half of the measured 30 day sump debris pressure drop would beconservative. Therefore a debris related head loss of 0.22 ft at 3591 gpm wasjustified as an appropriate starting point of the LPSI failure to trip evaluation.However, a higher value of 0.4 ft head loss was assumed for the staring point at3591 gpm. When clean strainer head loss and debris head loss was corrected tothe LPSI failure to trip flow conditions, the results was 4.82 ft head loss,(reference Table 3g1 on page 135 of the 6/30/2009 submittal) which wasacceptable.The CFD and Debris Transport analysis supporting the November 2008 strainertest used the lower non-LPSI failure flow and that output was used to computedropout fractions. However the fine fiber component was assumed to transport100% independent of CFD velocity and increased velocity would not haveincreased the quantity of fines used in the test flume. The amount of fiber finesthat may have transported in test flume in November 2008 would have beengreater if the test was run at the higher LPSI failure to trip flow conditions for the30 day debris load. However, for the arguments offered in the precedingparagraphs, the starting point for the LPSI failure to trip evaluation used aconservative value to address added transport. With respect to fiber smalls, theamount of smalls used in the November 2008 strainer test was conservativelyhigh as demonstrated by the fact that much did not transport due to theconservatively low 0.06 ft/s tumbling velocity assumed in the CFD and debristransport evaluation. If a more representative tumbling velocity of 0.12 ft/s hadbeen used, it would have provided at least some offset for any increase intransport of smalls under LPSI failure to trip flow conditions. With respect to ahigher flow in the test flume under LPSI failure to trip flow conditions, additionalfiber smalls might have transported but those have smaller effect on pressuredrop and a case can be made for additional smalls reducing head loss byinterfering with development of a thin bed.The above discussion provides justification for the analytically derived head lossfor the short period of time that the LPSI pump is running. Once the runningLPSI pump is terminated, the amount of debris transported to the strainers couldthen be higher once the remaining 30 day design basis debris amount transports.The November 2008 strainer test results (<0.75 ft with chemical) did have marginto the 1.568 ft head loss available for debris (i.e., 2.6 ft available for design minusclean strainer and associated piping head loss). Given the LPSI failure to trip is Palisades Draft RAI Responses for May 2010 Public Meeting 79a low probability event combined with the worse case debris case based on alarge break double guillotine being a very low probability event, the combinationof these two events probably could be supported as not required to be analyzed,if necessary, through an appropriate License Amendment Request.

References24.1 EA-MOD-2005-004-03 Rev 3 dated 4/6/2010, "ESS Flow Rates andNPSH during Recirc Mode with CSS Thottling"24.2 AREVA Document 66-9097941-000, "Palisades Test Report for ECCSStrainer Performance November 2008 Testing", 2/16/2009.

Palisades Draft RAI Responses for May 2010 Public Meeting 80 NRC Request25. In RAI 13 from the NRC's letter dated December 24, 2008, the staff askedfor information regarding the single failure of a LPSI pump to trip at thetime of switchover to recirculation, or to be restarted during the event.This issue was not fully addressed in the supplemental response withrespect to the affect on head loss and vortexing. The licensee stated thatthe emergency operating procedures have been revised to remove thesteps that directed restarting the LPSI pump, and the licensee evaluatedthe overall head loss associated with a running LPSI pump. Theevaluation was acceptable except that it did not address the potential foradditional transport to the strainer (as discussed above) if head losstesting had been conducted with the higher flow rate.Please provide the assumptions of the analytical evaluation and its results,with adequate bases for the assumptions, and verify that the results do notaffect the head loss evaluated both for the short-term LPSI pump runduration and for the mission time of the strainer. In addition, please verifythat the higher LPSI flow rates would not result in air entrainment(vortexing or deaeration) either at the strainer or the ECCS pump suctionpipe in the sump due to the potentially higher head losses if the levelbehind the strainer were drawn down into the sump. Also, please verifythat the strainer testing conservatively represented the flow velocities nearthe strainer under the LPSI failure-to-trip scenario or provide informationthat shows that testing contained adequate representation of the flowsunder this condition.Entergy Nuclear Operations Response:The first identified item, assumptions of the analytical evaluation and its results,is addressed in RAI 24 response. The last identified item, strainer testing flowvelocities, is addressed in RAI 24. The response to the balance of the RAI 25request follows.Vortex testing was performed on the Palisades strainer design at a submergenceof 2" over the top of strainer consistent with the minimum containment waterelevation of 593'-4" (Reference EC12249, Rev. 0) and flow rates associated witha large break LOCA (LBLOCA). Testing performed on the strainer determinedthat the strainer did not exhibit any characteristics associated with a vortex orvortex formation. Based on these test results, Entergy concluded that an air corevortex that could draw air into the Palisades strainer modules is not expected forthe LBLOCA conditions.The low pressure safety injection (LPSI) failure to trip flow rate of 7,148 gpm(Reference EA-MOD-2005-004-03, Rev. 3) is approximately 2 times the 3,591gpm LBLOCA flow rate (Reference EC12249, Rev. 0). Although the strainer Palisades Draft RAI Responses for May 2010 Public Meeting 81vortex testing was not performed at the higher LPSI failure to trip flow rate, it isnot expected that air would be drawn into the ECCS pump suction lines at thesehigher flow rates for the following reasons:i. Air core vortices were not observed during the testing at the minimumLBLOCA water level that fully submerges the strainer modules, 593'-4"(Reference EC12249, Rev. 0), consistent with the minimum water levelduring a LPSI failure to trip event. The strainer perforated plate providessome flow straightening that minimizes the potential for air core vortices.Air core vortices were not observed in the testing until the water levelwater was decreased to the small break LOCA (SBLOCA) elevation of592.34 feet (Reference EA-SDW-97-003, Rev. 3). For a SBLOCA, thewater level is below the top of the strainer modules (strainer is partiallysubmerged). The LPSI failure to trip condition would not occur when thestrainer is partially submerged at the SBLOCA water level since the LPSIpumps could not inject into the Primary Coolant System for breaks thatgenerate the SBLOCA water level.ii. If an air core vortex was drawn into the strainer during a LPSI failure to tripevent at the LBLOCA minimum water level of 593'-4" (ReferenceEC12249, Rev. 0), any air that could be ingested into the containmentsump would be released through the containment sump vents since thestrainer is vented. This would eliminate the possibility of any potential airingestion into the Emergency Core Cooling System (ECCS) pump suctionlines.iii. The containment water level would submerge the containment sump ventsif the containment water level reached the 595 feet elevation (ReferenceDrawings M-74, Sheet 1, Rev. 13 and VEN-M802, Sheet 1, Rev. 0). Atthis water level, the strainer is submerged over 20 inches (10 times thesubmergence at the tested minimum LBLOCA water level). Using theFroude number as defined in Appendix A of the NRC Regulatory Guide1.82, Revision 3, dated November 2003; the following can be concluded:The Froude number (Fr) is a non-dimensional parameter used to describea fluid flow field and consequently evaluate hydraulic performance. TheFroude Number (Fr) is the ratio of inertial forces to gravitational forces thathas been used to evaluate the susceptibility to air core vortices. TheFroude number as defined in Appendix A of the NRC Regulatory Guide1.82 is: gs U erFroudeNumb Palisades Draft RAI Responses for May 2010 Public Meeting 82Where:g = acceleration due to gravitys = minimum submergenceU = velocitySince vortices were not evident at the tested LBLOCA flow rate and theminimum strainer submergence condition, the Froude number for thetested condition can be compared to the Froude Number for the LPSIfailure to trip condition to determine the susceptibility to air core vortices.The Froude Number for the tested condition can be defined as:

1 1 1 gs U erFroudeNumb,and the Froude Number for the LPSI fail to trip condition will be defined as: 2 2 2 gs U erFroudeNumb.Equating the Froude Numbers for both flow conditions yields the followingequation: 2 2 1 1 gs U gs USolving the equations for the ratio of s 2/s 1, a comparison can be made tothe ratio of the velocities using the following equation:

2 1 2 1 2U U S SThe strainer surface area is constant for the strainer tested condition at3,591 gpm and the LPSI failure to trip condition (7,148 gpm). Although theflow distribution between the strainer modules will be slightly different forthe LPSI failure to trip condition due to the strainer internal losses, anassessment of the susceptibility of the strainer modules to an air corevortex can be performed by assuming that the strainer approach velocityis proportional to the strainer flow rate. Solving the above equation with aflow ratio of 2 (ratio of flow rate for the LPSI failure to trip condition to thetested condition) gives a required submergence ratio of 4 for the higherflow rate for the LPSI failure to trip condition.In summary, the LPSI failure to trip condition with an associated waterlevel of 595 feet results in a required submergence of approximately 4times the submergence for the tested conditions. Since the strainer Palisades Draft RAI Responses for May 2010 Public Meeting 83submergence at the water elevation of 595 feet is 10 times greater thanthe tested condition; ingestion of air core vortices is not expected atcontainment water levels that submerge the containment vents.

Palisades Draft RAI Responses for May 2010 Public Meeting 84 NRC Request 26.Some subparts of the licensee's response to RAI 14 from the NRC's letterdated December 24, 2008, were not responded to satisfactorily, as follows.a. 14.e: Debris preparation and introduction methods. The licensee provided the debris preparation and introductionmethods. The staff observed issues with the debris preparation andintroduction during the review of test video. The Palisadessupplemental response stated that some of the fines were removedfrom the small fibrous debris prior to addition to the test flume. ThePCI method of removing these fines from the smalls is to shake thesmalls on a shaker table. The NRC staff does not believe that this isconservative or prototypical of plant debris as some fines are likelycontained in small fibrous pieces. The removal of the fines from smallpieces for testing may be non-conservative when compared to theplant condition. The licensee should provide justification that theremoval of the fines from the small fibrous debris prior to theintroduction of smalls into the flume is prototypical or conservative withrespect to debris that would be generated in the plant. Additionally, thelicensee should provide information that justifies that debris used in thetesting was prototypically sized. Please provide information thatjustifies that the debris introduction sequence did not non-conservatively affect transport during testing.Entergy Nuclear Operations Response 26a:Utilizing the information provided in Table 26a-1 (below), each of the Staff's three(3) specific concerns/issues for RAI 26a can be addressed. The three (3) Staffconcerns/issues are as follows:1. The Palisades supplemental response stated that some of the fines wereremoved from the small fibrous debris prior to addition to the test flume.The PCI method of removing these fines from the smalls is to shake thesmalls on a shaker table. The staff does not believe that this isconservative or prototypical of plant debris as some fines are likelycontained in small fibrous pieces. The removal of the fines from smallpieces for testing may be non-conservative when compared to the plantcondition. The licensee should provide justification that the removal of thefines from the small fibrous debris prior to the introduction of smalls intothe flume is prototypical or conservative with respect to debris that wouldbe generated in the plant.

Palisades Draft RAI Responses for May 2010 Public Meeting 852. Additionally, the licensee should provide information that justifies thatdebris used in the testing was prototypically sized.3. Please provide information that justifies that the debris introductionsequence did not non-conservatively affect transport during testing.Even though the three (3) Staff concerns/issues are related, they are individuallyaddressed.The following chronological and historical summary of issues directly related tothe three (3) Staff concerns/issues is provided in order to obtain a betterperspective of the background associated with the Staff's issues: separation offines from small fines, prototypically sized debris, and debris sequencing. NEI 04-07 (May 2004) does not provide specific guidance regarding anintegrated head loss test protocol. SE (December 2004) for NEI 04-07 also does not provide specificguidance regarding an integrated head loss test protocol. PCI/AREVA/Alden and Licensees (i.e., the Team) utilized available andrelevant NUREG/CRs such as 2982 (December 1983) and 6773(November 2002) to develop the Large Flume Test Protocol sinceguidance was not available in either NEI 04-07 or the SE for same. PCI/AREVA/Alden and Licensees (i.e., the Team) met with and discussedthe Large Flume Test Protocol on more than nine (9) occasions (February2007 (ML072530885) - February 2008 (ML080370262)). The Staff in various public meetings with the NEI/PWROG/Licensees andPublic praised the fact that the Team had engaged the Staff regarding theproposed test protocol when other strainer vendors had not includingsome that had actually completed testing without ever discussing theprotocol with the Staff in advance of testing. The 'draft' of theNRC Staff Review Guidance Regarding Generic Letter2004-02 Closure in the Area of Strainer Head Loss and Vortexing(ML072600348) (i.e., so-called March Guidance Document) issued forcomment in approximately October 2007 did not provide anyguidance/objections/clarifications to integrated head loss testing as well asthe preparation and classification of fine fibrous debris. In addition theterms prototypical and conservative were not defined individually or whatwas meant by the term 'prototypical or conservative'. During the October 24, 2007 public meeting between the Staff andrepresentatives from NEI, the PWORG, Licensees, and the public, theStaff responded to a specific question regarding wording in the 'draft'March Guidance Document that 'fine fibrous debris is individually orreadily suspendable fibers'. The Staff response at the direction andquestioning of the director was'fines' were not single fibers, but couldbe 'clumps' or 'bunches' of fibers.

Palisades Draft RAI Responses for May 2010 Public Meeting 86 The Team and Licensees initiated Large Flume Test activities in January2008. Members of the Staff and a Staff outside contractor witnessed thefirst test. The Staff provided feedback to the Team which the Teamincorporated as well as 'lessons learned' into a revision to the LargeFlume Test Protocol. The Staff documented the positive observations of the initial Licensee test(ML081840095). In the 'final' issued version of the so-called March Guidance Document,there is much discussion by the Staff that 'fine' fibrous debris beindividually or readily suspendable fibers (Page 4 as well as others).The discussion in the subject document is in direct contradiction with theStaff SER for NEI 04-07 and the Staff response to a specific question fromNEI, the PWROG representative, and licensees made in the October 24,2007 public meeting, that'fines' were not single fibers, but could be'clumps' or 'bunches' of fibers. The terms prototypical and conservativewere still not defined individually or what was meant by the term'prototypical or conservative' even though they are utilized extensively inthe document. Subsequent Licensee tests were performed and successfully completedutilizing the revised Large Flume Test Protocol. The Staff on a number ofoccasions observed and witnessed the subject tests. The Staffdocumented their observations (ML083590250). There is no mention thatthe revised Large Test Flume Protocol that was utilized was deficient orthat there were any issues regarding debris preparation, debris settlement,or agglomeration of fibrous debris. The Staff was fully aware of the fact that fine fibrous debris had beenremoved from small fines for a Licensee test, since the Staff had beenpresent at the Licensee test and had also received a copy of the debrisallocation sheets which specifically indicated that fines were removed fromsmall fines (ML083590250). All Licensee Large Flume Tests utilizing the revised Large Flume TestProtocol were completed on November 11, 2008. The Staff stated in a Licensee's July 9, 2009 public meeting regardingopen RAIs that fiber fines are expected to conform to Classifications 1, 2 &3 of NUREG/CR-6808 as shown in Table 3-2Size Classification Schemefor Fibrous Debris. This is a new (i.e., July 2009) 'definition' for finefibrous debris as stated by the Staff to the industry which is different fromthat provided in the SE for NEI 04-07 and the March 2008 Staff issued,NRC Staff Review Guidance Regarding Generic Letter 2004-02 Closure inthe Area of Strainer Head Loss and Vortexing which also occurred after allof the Licensee Large Flume Testing was completed in November 2008. Staff issues, (DRAFT) NRC Staff Synopsis of Issues and StatusRegarding Credit for Debris Settlement on February 25, 2010 - more thansixteen (16) months after all Licensee testing had been completed.

Palisades Draft RAI Responses for May 2010 Public Meeting 87Response to RAI26a Issue 1, Removal of fines from smalls:PCI separated the fines from smalls in order to test the fibrous debris form knowas small fines in a more conservative manner. The reason to remove fines fromsmall fines was in response to the staff's concerns regarding debris introductionsequence for the Large Flume Test Protocol. It was agreed by the staff that theProtocol would adhere to the basic criterion of introducing debris from the mosttransportable to the least transportable. This is not prototypical, but it is veryconservative. Since fine fibrous debris based on the staff's opinion will morereadily transport than small fibrous debris, the fines were separated from thesmall fines.Regarding the staff statement, -The staff does not believe that this isconservative or prototypical of plant debris as some fines are likely contained insmall fibrous piece.PCI agrees this is not prototypical; however, it is veryconservative to separate and segregate the fines from the classification of smallfines, and introduce the separated fines before the remaining small fibrous debrisin our test protocol. The Large Flume Test Protocol was specifically developedto be more conservative than the prototypical plant condition.By removing the fines from the small fines, and introducing them to the test flumebefore the remaining small fibrous debris, a greater percentage of separated finefibrous debris was introduced for the PNP test. This is clearly more conservativethan introducing fines as a sub-set of small fines.If the staff is suggesting that by removing fines from the debris classificationsmall fines is not conservative, there is no basis for the staff's suggestion. ThePNP Design Basis debris types and quantities were fully met during the PNPhead loss test at ARL. It would have been overly conservative since the fineswithin the small fines are either contained within the quantity of small fines orthey are separated from the small fines as fine fibrous debris. Fine fibrous debriscannot exist in both fibrous debris classifications, that is small fines and fines. Inaddition, there is no regulatory requirement or guidance document that supportssuch a staff suggestion.Table 26a-1 provides a summary of the fibrous debris types, quantities, andclassifications utilized for the PNP Design Basis Test performed in November 2008.Table 26a-1 PNP Fibrous Debris Types, Quantities & ClassificationsFibrous Debris Type 3Unit DebrisAllocation ScaledQuantity 1 DebrisRoundedAllocationQuantity 1MeasuredQuantity -

Used inPNP Test 2 Palisades Draft RAI Responses for May 2010 Public Meeting 88NUKON Fineslbm 8.991 9.0 9.05NUKON Smalls (W/FinesRemoved)lbm 7.642 7.7 7.75NUKON Larges - Eroded(Treated as Fines)lbm 3.327 3.4 3.45Mineral Wool Fineslbm 6.909 7.0 7.05Mineral Wool Larges - Eroded(Treated as Fines)lbm 3.720 3.8 3.85Latent Fiber (Fines)lbm 1.342 1.4 1.45Total Fibrous Debris Quantity lbm 31.931 32.3 32.6% Increase from Scaled Qty %

NA 1.2 2.1% Fines including Latent

%76.1 76.2 76.2% Smalls%23.9 23.8 23.8Fines & Small Fines DebrisSummary for Testing% Fines including Latent

%76.1 76.2 76.2% Smalls W/Fines Removed

%23.9 23.8 23.8% Additional Fines Contained(i.e., 16%) in Smalls W/FinesRemoved 4%3.82 (16.0% x23.9%)3.81 (16.0% x23.8%)3.81 (16.0% x 23.8%)% Total Smalls - Tested

%20.08 19.99 19.99% Total Fines - Tested

%79.92 80.01 80.01NOTE:(1) Values based on PCI calculation TDI-6031-02(2) Values based on AREVA Test Plan 63-9095797-001 and Test Report66-9097941-000(3) Debris descriptions and quantities from Areva Calculation 32-9099369-000(4) Initial percentage of fines contained in small fines smalls after initialfines removed (i.e., 25%) is an additional 16% fines based on PCIdocument Performance Contracting, Inc., SSFS-TD-2007-004,Supplement 1, Rev. 1,Sure-Flow Suction Strainer - Testing DebrisPreparation & Surrogates (ML092430056 & ML092580203)

Palisades Draft RAI Responses for May 2010 Public Meeting 89As can be seen from the results of the subject table, PNP classified more than80% of their entire debris quantity as fine fibrous debris which was also utilized inthe PNP head lost test (Test 4: Design Basis Test).Finally, it should be noted that the staff was fully aware of the fact that finefibrous debris had been removed from small fines for a previous licensee test,since the staff had been present at the licensee test and had also received acopy of the debris allocation sheets which specifically indicated that fines wereremoved from small fines.In addition, the staff observed and documented as part of the same licensee test(ADAMS ML083590250), the following:The PCI/Areva debris preparationmethodology has been revised to remove any fine fibrous debris from the smalldebris category by subjecting "smalls" processed through a wood chipper to ashaker table with coarse screen. The old methodology left any fine debris thatwas created in the process of making the small debris in the mixture; whereasthe new process allows loose "fines" to be removed from the "smalls". Thisreduces the total fine debris available for transport to the strainer to be morerepresentative of the design basis specified by the client.It is evident that the staff was fully aware of the change in debris preparation toremove fines from small fines for licensee testing. It should be further noted thatthe total quantity of fine fibrous debris utilized in both the PNP and licensee testswere specified by the licensee and representative of the expected quantities forthe licensee's plant. Simply stated the removal of fines from small fines fibrousdebris did not have any affect on nor reduced the total specified Design Basisquantity of fine fibrous debris utilized for a licensee's test, and in all cases the'recommended' percentage of fine fibrous debris as discussed in the SE for NEI04-07 was significantly exceeded.Response to RAI26a Issue 2, testingdebris prototypically sized

The design input for strainer testing defined the debris size distribution for PNPstrainer test in terms of fines and smalls (no larges were determined to transportto the strainers). The debris types, scaled quantities, and classifications used forthe PNP strainer qualification test were documented in the AREVA NP Test Plan(Test 4: Design Basis Test). Please refer to Table 26a-1 for additionalinformation regarding the various fibrous debris classifications and quantities.The fibrous debris utilized for the PNP test was prototypically sized.The Staff has raised a number of issues and questions regarding fibrous debrissizing with respect to the fibrous debris classifications of both fine and small finesfibrous debris. The following discussion provides a historical perspective of theissue of fibrous debris sizing and fibrous debris size distribution as it relates tothe Large Flume Test Protocol as utilized for the PNP test. The discussionperspective also provides clarification of the various NUREG/CRs and their Palisades Draft RAI Responses for May 2010 Public Meeting 90definition of and classification of fibrous debris sizing. The subject referencedNUREGs and other documents were used to develop the subject Protocol.Generally speaking, there is no regulatory and/or industry definition of fine fibrousdebris generated as the result of a Design Basis LOCA. This conclusion isbased on an extensive review of regulatory and industry documents. Therefore,the term 'prototypically sized debris' is subjective and has no known objectivecriteria with respect to size and/or classification. However, PCI has processedand prepared dry fibrous debris into classifications that are prototypical of post-LOCA conditions based on various NUREG/CRs related to the issue.That being said, in Section VI.3Methodology of the Safety Evaluation (SE) forNEI 04-07, the BWR drywell debris transport study (DDTS) is discussed. Thestudy is documented in NUREG/CR-6369-1, -6369-2, and -6369-3. It should benoted that the configuration of a BWR Mark I drywell (containment) on which thesubject DDTS evaluations were performed is very small, very different, moreconfined, and has considerably more platforms, gratings, walkways, and otherstructures that would significantly 'shred' fibrous debris during the initial post-LOCA blow down period than would be reasonably expected for a typical, largePWR containment. Therefore use of the DDTS results, conclusions, and theirapplication to a large, dry PWR containment and the associated post-LOCAgenerated fibrous debris would be extremely conservative, but not prototypical.The DDTS (page VI-9) concluded that based on CEESI air blast tests of fibrousinsulation, that when an LDFG blanket was completely destroyed, 15 to 25percent of the insulation was in the form of very fine debris (i.e., debris too fine tocollect readily by hand). It should be noted that very fine debris as defined in thesubject NUREG/CR does not mean nor infer individual fibers of fibrous insulation.On page VI-14 of the SE for NEI 04-07, it states in part -The analysis of theAJIT testing performed at CEESI to support the DDTS determined that wheneverentire blankets were completely destroyed, 15 to 25 percent of the insulation wastoo fine to collect by hand. In this case, complete destruction means nearly all ofthe insulation was either fine or small pieces. In any case, 15 to 25 percent ofthe blanket (an average of 20 percent) can be considered fine debris for thepurposes of this analysis.It can be concluded thatcomplete destruction meanseither fine or small pieces(of fibrous debris), and not individual fibers. It isrecognized that individual fibers constitute an unknown portion of the fine orsmall fibrous debris pieces, and most likely not the majority of the debris pieces.It should be further noted that, NUREG/CR-6369 Volume 1, specifically Section 2.2Debris Size Methodology states in part -the fact that debris sizedistribution that is universally applicable to plant conditions and accidentscenarios is not available. On the other hand, there is consensus that the debriscan be broadly divided into three size classes: small 9, large, and large-canvassed, according to their relative size and pathways available for theirtransport.

Palisades Draft RAI Responses for May 2010 Public Meeting 91Footnote 9 identified with the class of smalls from the previous paragraph statesas follows, 9It should be noted that, initially, efforts were made to further classifysmall debris into sub-groups fines, smalls, and medium consistent withNUREG/CR-6224 study [Ref. 2.1]. However, a decision was made to collapsethem into a single group called small because: a) fines, smalls and mediumpieces have very similar transport pathways and b) existing debris sizedistribution data does not differentiate these three size groups [Ref. 2.3].Even though NUREG/CR-6369 states that -the debris can be broadly dividedinto three size classes, NUREG/CR-6369 is actually based on NUREG/CR-6224which classified fibrous debris notby size, but actually by shape (refer tosubsequent discussion below).Finally, Table 2-5Debris classification from NUREG/CR-6369 Volume 1 providesthe following information.In addition to the information provided in Table 2-5, NUREG/CR-6369 Volume 1also provides photographs of the various fibrous debris classifications includingsmalls. Figure 2-5 depicting the fibrous debris classification smalls is shownbelow. Note the physical size of the small classification fibrous debris.

Palisades Draft RAI Responses for May 2010 Public Meeting 92As stated in Table 2-5 from NUREG/CR-6369 Volume 1, smalls are classified asfibrous debris passing through a standard floor grating of less than6" x 4". NEI04-07 specified that fibrous debris classified as small fines was fibrous debristhat would pass through a standard4" x 4" floor grating. PCI document, Sure-Flow Suction Strainer - Testing Debris Preparation & Surrogates, TechnicalDocument No. SFSS-TD-2007-004 very conservatively defines the small finesclassification as fibrous debris that would pass through a standard4" x 1" floorgrating. Therefore, the PCI classification of small fines which was used for alllicensee testing including PNP is significantly more conservative than that statedin either NUREG/CR-6369 Volume 1 or NEI 04-07.Table 2-5 also discusses the transport characteristics of the fibrous debrisclassified as smalls. In part, the subject Table states that:Gravitation settling isnegligible andPool turbulence can keep them in suspension. These statementsin the subject NUREG/CR are in relation to the stated definition for small, whichis any fibrous debris that will pass through a 6" x 4" grating. This would includefine fibrous debris, but it also would include significantly larger pieces of fibrousdebris. It could therefore be concluded that NUREG/CR-6369 Volume 1supports the fact that fibrous debris passing through a 6" x 4" grating will not beaffected by gravitational settling and that fluid turbulence will keep the subjectfibrous debris in suspension.As part of Table 2-5, reference is made to NUREG/CR-6224 terminology withregard to fibrous debris classification. Section B.4.1Classification of Fibers and Palisades Draft RAI Responses for May 2010 Public Meeting 93Table B-3Fibrous Debris Classification by Shape found in NUREG/CR-6224discuss fibrous debris classification. Table B-3 is provided below.

`Section B.4.1Classification of Fibers discusses the methodology that wasutilized in classifying fibrous debris and that is presented in Table B-3. SectionB.4.1 states in part -The debris classes of Table B-3 can best be described asshape classes since their classification is based solely on their shape. Implicitly,however, each shape is associated with a narrow range of sizes and thus anarrow range of settling velocities. The Section further states -However,owing to their ill-defined shapes, it is difficult to further classify these debris bytheir shape classes and to develop appropriate size distribution curves (i.e., it isdifficult to determine what fraction of the residual debris belongs to each shape class). Finally the Section states -Usage of settling groups instead of theshape classes described above provides for finer classification of debris. Section Palisades Draft RAI Responses for May 2010 Public Meeting 94B.5 presents further discussion on the settling groups used for classifying theNUKON material and their relationship to shape classes in Table B-3.It should be noted that Section B.4.1Classification of Fibers and Table B-3Fibrous Debris Classification by Shape found in NUREG/CR-6224 does notprovide any specific information with regard to fibrous debris size and/or if ascreening method such as a 4" x 4" or a 6" x 4" standard floor grating openingwas utilized. Therefore the relative sizes of the subject fibrous debrisclassifications found in Table B-3 are unknown and could be significantly large aslong as the classification shape of the fibrous debris was such that it settled at acertain rate.This is a very important distinction since NUREG/CR-6224 concluded that size isnot really important with regard to fibrous debris transport and settling, but in factthe shape of the fibrous debris controls the transport and settling. Therefore itcan be inferred that the actual size of fibrous debris per NUREG/CR-6224 is notimportant.In addition, the NUREG/CR-6224 settling rates provided in Table B-3 are notstated in relation to a water temperature or temperature range. It should benoted that water temperature has a significant effect on fibrous debris settlementrates even more so than the shape and/or configuration of the fibrous debris.This conclusion is documented in both NUREG/CR-2982,Buoyancy, Transport,and Head Loss of Fibrous Reactor Insulation and NUREG/CR-6772, GSI-191:Separate-Effects Characterization of Debris Transport in Water, and to a lesserextent in NUREG/CR-6773,GSI-191: Integrated Debris-Transport Tests in WaterUsing Simulated Containment Floor Geometries (Please refer to followingdiscussion).Following the first Licensee Large Flume Test, the Staff in March 2008 issued thedocument,NRC Staff Review Guidance Regarding Generic Letter 2004-02Closure in the Area of Strainer Head Loss and Vortexing (also known as theMarch Guidance Document). It should be noted that no reference is made toNUREG/CR-6224, - 6369, or -6808 (discussed below) in the March GuidanceDocument with regard to the specific subject of fibrous debris sizingclassifications or Table 2-5, Table B-3, or Table 3-2. However, the various termssuch as fines, fine fiber, finer fiber, etc. are utilized in the subject document, butare never objectively defined within the document and/or reference made to anyNUREG/CR or other guidance documents.In the March Guidance Document, there is much discussion by the Staff that'fine' fibrous debrisbe individually or readily suspendable fibers (Page 4 aswell as others). The discussion in the subject document is in direct contradictionwith the Staff SER for NEI 04-07 and the Staff response to a specific questionfrom NEI, the PWROG representative, and licensees made in the October 24, Palisades Draft RAI Responses for May 2010 Public Meeting 952007 public meeting, that'fines' were not single fibers, but could be 'clumps'or 'bunches' of fibers

.In the March Guidance Document, there is also much discussion by the Staff thatfine fibrous debris should be mixed so that agglomeration does not occur thatwould not be prototypical. However, no guidance or description is provided bythe Staff of what is meant by agglomeration as has been previously discussed. Itwould appear that 'clumps' and 'bunches' based on previous Staff discussion(i.e., October 24, 2007 public meting) would be prototypical and conservativewith regard to defining agglomeration. It appears that the Staff's primaryconcern with regard to fine fibrous debris agglomeration is that the subject debriswill settle and not transport as readily as individual fibers. However, there is nobasis provided by the Staff as to why they believe that fine fibrous debris wouldnot agglomerate in the post-LOCA containment environment.As a matter of fact it would seem that it would be more prototypical and likely thatthe fine fibrous debris would actually agglomerate, would not readily transportand would readily sink as individual fibers in the quiescent post-LOCAcontainment 212 oF plus fluid prior to the initiation of ECCS/CSS recirculation. Areview of various regulatory documents such as NUREG/CRs provides nosupport that fine fibrous debris will exist as individual fibers and not agglomeratein the post-LOCA containment. However, on the other hand there are regulatorydocuments that support the opposite position, that is, fine fibrous debris includingindividual fibers will settle and not readily transport.NUREG/CR-2982Buoyancy, Transport, and Head Loss of Fibrous ReactorInsulation on Page 19, Section 3.1 Buoyancy Tests, specifically paragraphs a),c), and e) state in part:a) In general, the time needed for insulation to sink wasfound to be lessat higher water temperatures. (emphasis added)c) The fiberglass (Filomat) readily absorbs water,particularly hot water,and sinks rapidly (from 20 to 30 seconds in 120 oF water).This isalso true for individual fibers. (emphasis added)e) - Damaged fiberglass insulation pillowswill sink before activationof the recirculation system - (emphasis added)Since the initial post-LOCA water temperature in all cases for all Licensees willeasily exceed 120 oF, it can be concluded based on the subject NUREG/CRreport that fiberglass insulation and more importantly individual fiberglass fiberswill sink before the initiation (i.e., approximately 15 - 40 minutes post-LOCA) ofthe ECCS/CSS in the recirculation mode. It is interesting to note thatNUREG/CR-6808, -6224, or -6369 nor the March Guidance Document makesany mention of NUREG/CR-2982, NUREG/CR-6772, or NUREG/CR-6773, and Palisades Draft RAI Responses for May 2010 Public Meeting 96specifically that fiberglass individual fibers will sink in 20 to 30 seconds in 120 o Fwater. It should be noted that NUREG/CR-2982 predates all of the subjectNUREG/CRs by more than 13 years and NUREG/CR-6772 and NUREG/CR-6773 were specifically funded to investigate GSI-191 issues. In addition duringLicensee fiber by-pass testing at the Alden Research Laboratory it was alsonoted that significant quantities of fibrous debris including fine fibrous debrissettled, and did not reach the strainer in the Licensee prototypical flow streamsand at water temperatures of approximately 120 oF. It should be noted that thesubject fibrous debris in accordance with the Large Flume Test Protocol was notallowed to 'sit' in a quiescent flume, but was instead added to a moving flumeflow stream, and still significant settling of the fibrous debris occurred. Therecent Licensee fiber by-pass testing supports the conclusion reached inNUREG/CR-2982 and NUREG/CR-6772 that individual fibers will rapidly sink inwater greater than 120 o F.The extremely important conclusion supported by testing associated with anddocumented in NUREG/CR-2982, that is - fiberglass (Filomat) readily absorbswater,particularly hot water, and sinks rapidly (from 20 to 30 seconds in 120 o F water).This is also true for individual fibers. - has some bearing on theinformation contained in Table B-3 of NUREG/CR-6224. Note that as waspreviously discussed NUREG/CR-6224 is the 'basis' for NUREG/CR-6808 and -6369 regarding debris size classifications.The buoyancy test facility at ARL which served as the primary research facility forNUREG/CR-2982 was 5.0' in depth. The subject NUREG/CR documents thatindividual fiberglass fibers will sink in 20 to 30 seconds results in settlement ratesof 0.167 fps to 0.25 fps. It could therefore be concluded that the settling testsperformed and documented in NUREG/CR-6224 were performed at a muchcolder water temperature (i.e., < 120 oF) which is both conservative and non-prototypical of post-LOCA conditions. This conclusion is reached based on thestated settlement rates for fibrous debris classes 1 - 3 summarized in Table B-3of NUREG/CR-6224. The subject table lists settlement rates of 0.003 - 0.011 ft/sfor class 1 and 0.04 - 0.06 ft/s for class 3. These settling rates are significantlydifferent than those stated in NUREG/CR-2982 as well as recent testingperformed at ARL.In addition to the conclusions reached in NUREG/CR-2982 regarding rapidsinking of fibrous debris in heated water prototypical of containment post-LOCAfluid conditions, NUREG/CR-6772,GSI-191: Separate-Effects Characterizationof Debris Transport in Water,provides further evidence and support that fibrousdebris will rapidly sink in heated water. Section 2.2.1.1Effect of Temperaturestates in part -Post-LOCA water temperature is approximately 80 oC, which issignificantly different for the ambient temperature proposed for use in the testprogram. It can be postulated that water temperature could affect debris settlingcharacteristics because density and viscosity are temperature-dependent.Further the saturation rates of debris may be temperature-dependent because Palisades Draft RAI Responses for May 2010 Public Meeting 97surface tension varies with temperature. This set of experiments provided datato quantify these effects, as described below.Saturation of Debris in Hot Water. When fibrous debris was introduced to waterat ambient temperature, it was observed to float on the surface for more than 24h. Even when shredded fiber fragments were forcibly immersed in ambient-temperature water for 24 h, they subsequently would rise up to the surface whenreleased. However, if the fiber shreds were immersed in hot water (80 oC) for aslittle as 2 min, they readily sank and remained submerged. In the aftermath of aLOCA, the temperature of water in a PWR recirculation sump is likely to becloser to 80 o C than to (~ 20 o C).It should be noted that 80 o C is 176 o F, and 20 o C is 68 oF. Therefore debrissettlement and transport tests in ambient (i.e., cold) water are very conservativebut are significantly non-prototypical of the post-LOCA containment fluid conditions.The difference in settlement rates between NUREG/CR-6224 and NUREG/CRs -2982 and -6772 can be easily explained. Simply stated, NUREG/CR-6224testing was performed at non-prototypical relatively cold water temperatures,while NUREG/CR-2982 and -6772 utilized prototypical relatively hot watertemperatures. This conclusion is supported by Section 4.1Buoyancy Tests ofNUREG/CR-2982 which indicates that fiberglass in 50 oF water took 20 to 60minutes to sink and only took 20 - 30 seconds to sink in 120 oF water, and Section 2.2.1.1Effect of Temperatureof NUREG/CR-6772 which indicates thatfibrous debris introduced to water at ambient temperature (i.e., 68 oF) wasobserved to float on the surface for more than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Even when shreddedfiber fragments were forcibly immersed in ambient-temperature water for 24hours, they subsequently would rise up to the surface when released. However,if the fiber shreds were immersed in hot water (i.e., 176 oF) for as little as 2 min,they readily sank and remained submerged. In the aftermath of a LOCA, thetemperature of water in a PWR recirculation sump is likely to be closer to 176 o Fthan to 68 oF . Accordingly, the settling rates established in NUREG/CR-6224and which subsequently were utilized to establish fibrous debrisshapes/sizes/configurations in Table 2-5 of NUREG/CR-6369, and eventuallyTable 3-2 of NUREG/CR-6808, and all of which form the basis for the MarchGuidance Document appear to be based on potentially erroneous data based onnon-prototypical test water temperatures that resulted in unsupported staffpositions regarding non-prototypical fibrous debris size classifications, fibrousdebris settling, and fibrous debris transport.The Staff stated in a Licensee's July 9, 2009 public meeting regarding open RAIsthat fiber fines are expected to conform to Classifications 1, 2 & 3 of NUREG/CR-6808 as shown in Table 3-2Size Classification Scheme for Fibrous Debris (seetable below).This is a new (i.e., August 2009) 'definition' for fine fibrous debrisas stated by the Staff to the industry which is different from that provided in the Palisades Draft RAI Responses for May 2010 Public Meeting 98SE for NEI 04-07. It should also be noted that Class 3 fibrous debris is definedand pictured as 'clumps' measuring up to ~ 1" in accordance with NUREG/CR-6808, Figure 3-3 (see photo (Figure 3-3) following Table 3-2 below).It should be noted that the subject 'new' fibrous debris classification criteriabased on NUREG/CR-6808, as well as similar classification schemes are all based on dry fibrous debris. There are no known available objective criteria orguidance to assess or evaluate processed wet fibrous debris with regard to sizeclassification criteria, the mixing/preparation of same in support of integratedtesting, the addition of same to a test flume for integrated testing, and finally theability to assess, evaluate, and/or determine prototypical agglomeration andconcentration of same.It is also noted that the 'sketches' (Table 3-2 from NUREG/CR-6808) of thedebris classifications do not agree with the actual photographs of debrisclassifications (refer to the photographs (Figure 3-3) below for debris classes 3and 5). It should be further noted, that NUREG/CR-6808 does not contain anyphotographs of the very important debris 1 and 2 classes which the staff isstating is their basis for acceptable small fine debris. Furthermore, there are no Palisades Draft RAI Responses for May 2010 Public Meeting 99'hard' physical descriptions of the debris classifications with respect tosize/dimensions, volume, weight, 'screening' characterization (i.e., passingthrough a grating opening), etc. This makes it extremely difficult to objectivelyassess the exact nature of each debris classification as discussed inNUREG/CR-6808.Finally, the most important aspect of the subject debris classifications as theissue relates to NUREG/CR - 6224, - 6369, and -6808, and the SE for NEI 04-07is the fact that all of the subject NUREG/CRs and the supporting tables,sketches, and descriptions contradict each other.Simply stated there is nodefinitive mechanism to classify fibrous debris in an objective manner, nor isthere an objective mechanism to distribute the fibrous debris by classification.NUREG/CR-6369 concluded that small (not fines) classified fibrous debris wasnot affected by gravitational settling and that pool turbulence would keep thedebris in suspension. Contrary to the statements made in the SE for NEI 04-07regarding debris sizing and classifications, the Staff stated during a Licensee'smeeting to discuss RAIs that they considered the fibrous debris classifications 1- 3 found in NUREG/CR-6808 as meeting their definition of fine fibrous debris. Itshould be noted that the 'screening' criteria utilized to 'define' small fines for thesubject SE for NEI 04-07 and the subject NUREG/CRs (-6369 & -6808) was a 4"x 4" and a 6" x 4" standard floor grating opening, respectively. However, PCIutilized a 4" x 1" standard floor grating opening which is therefore veryconservative when compared to either the SE or NUREG/CR screening criterion.Simply stated, PCI's processing methods result in a significant order ofmagnitude smaller fibrous debris classified as small fines than if the guidancefound in the SE for NEI 04-07 or the subject NUREG/CRs were utilized. The useof PCI's classification of fibrous debris resulted in much smaller fibrous debristhat would also be more readily transportable during Licensee Large FlumeTesting than if SE or NUREG/CR screening criterion were utilized for the fibrousdebris. The subject PCI processed small fines fibrous debris was utilized in thesubsequent Large Test Flume for Licensees.

Palisades Draft RAI Responses for May 2010 Public Meeting 100Based on the discussion provided, it can be concluded that the processing of dryfibrous debris used for PNP testing resulted in very conservative and prototypicalfibrous debris size classifications that are readily transportable.In summary, PNP fibrous debris has been processed and prepared in accordance with thePCI 'white paper' SSFS-TD-2007-004Sure-Flow Suction Strainer - TestingDebris Preparation & Surrogates, the PCI 'white paper' SSFS-TD-2007-004,Supplement 1, Rev. 1,Sure-Flow Suction Strainer - Testing DebrisPreparation & Surrogates (ML092430056 & ML092580203), and the LargeFlume Test Protocol which have been provided to and discussed with the Staff. PCI has processed raw fibrous debris materials into 'fines' representative ofboth eroded or latent fibrous debris and 'fines/smalls' by recognizedmechanical process devices (i.e., chipper (smalls) & Munson machine(fines)). (ML092580203) The PCI definition of small fines has resulted in even more conservative andsignificantly smaller fibrous debris for ARL testing than if either theNUREG/CR-6369 Volume 1 or NEI 04-07 definition of small debris wereutilized. PCI utilizes a 4" x 1" standard grating opening for 'sizing' processedfibrous debris as opposed to the 6' x 4" or 4" x 4' standard grating openingfound to be acceptable by both the Staff's SE for NEI 04-07 and NUREG/CR-6369 Volume 1. All of PCI's processed small fines fibrous debris meets and/or exceeds thedry fibrous debris classes of 1, 2 & 3 as defined by NUREG/CR-6808, thesmalls classification as defined by NUREG/CR-6369 Volume 1, and classes 2- 6, inclusive for NUREG/CR-6224. The SE for NEI 04-07 found no issue with the NEI 04-07 definition of smallfines which are both larger in size classification and less conservative withrespect to fibrous debris transport issues than the PCI definition of small fines. Samples of dry latent, fines/smalls, and larges' were provided to the Staffbefore any Large Flume Testing was initiated and were found to be'representative of what the Staff had expected

'. Dry fibrous debris meeting the aforementioned fibrous debris classificationshas been deemed to readily transport based on the testing performed insupport of the subject NUREG/CRs.

Palisades Draft RAI Responses for May 2010 Public Meeting 101 The percentage of fine fibrous debris 'contained' in the PCI classification ofsmalls was 16% that required additional processing to free the fine fibrous debris. Observations and comments by the Staff and lessons learned byPCI/AREVA/Alden during the initial Large Flume Test for a Licensee wereincorporated into all subsequent tests and the Large Flume Test Protocol. Actual preparation (i.e., mixing of debris) of fibrous debris and introduction ofsame for Licensee Large Flume Tests has consistently been performed bymany of the same Alden personnelAs previously discussed, the Large Flume Test fibrous debris preparation,sequencing, concentration, and potential agglomeration of debris following apost-LOCA event is both prototypical and conservative for PNP.In addition, the following should be noted. There is no objective regulatory and/or industry guidance or criteria thataddresses the issues ofwet fibrous debris size classification, debrissequencing, concentration of debris slurries prepared for testing, thepotential for debris to agglomerate during preparation and addition ofsame to the test flume in support of a prototypical integrated test. The fibrous debris preparation and introduction of same to the Large TestFlume does represent a very conservative quantity of small fines includingfines and small fiber clumps which are transported in the representativeLicensee flow streams. Since the fibrous debris is transportable, it willcollect on the Licensee test strainer based on the expected Licensee post-LOCA conditions.To re-state - the debris preparation, sequencing of debris, concentration of debrisslurries prepared for testing, the potential for debris to agglomerate duringpreparation and addition of same was prototypical of the expected Licensee post-LOCA conditions, and very conservatively represented the actual and expectedLicensee post-LOCA conditions within the containment.In conclusion, the preparation, concentration, and introduction sequencing offibrous debris did not promote the agglomeration of the debris and did not inhibitthe transport of same other than what would have naturally occurred in an open,free flowing water stream such as what would occur in a Licensee's post-LOCAcontainment following initiation of ECCS/CSS recirculation. Therefore, thesubject fibrous debris was properly prepared and did not agglomerate oradversely affect the debris transport to the strainer. The PNP Large Flume Testwas conservative and prototypical of the actual Licensee post-LOCA containmentconditions and LOCA scenario.

Palisades Draft RAI Responses for May 2010 Public Meeting 102Response to RAI26a Issue 2,debris introduction sequence not non-conservatively affect transport

The PNP Design Basis test (Test 4) debris addition sequence was conservativeand prototypical for an integrated debris transport and head loss test.Accordingly, the PNP ARL Large Flume Test debris addition sequence did notadversely affect the ability of transportable debris to reach the test strainer. Thesequencing and concentration of debris slurries prepared for testing, the potentialfor debris to agglomerate during preparation and addition of same wasprototypical and also very conservatively, since the actual and expected PNPpost-LOCA conditions within the containment were represented. Please refer tothe responses to RAI 17e and 28 for additional information and discussion. Thesubject RAI response addresses and provides additional and related information for RAI 26a.Debris Introduction & SequencingThe PNP bounding debris head loss tests performed in November 2008 asdocumented in the test report met the intent of the discussions between the Staffand PNP/PCI/AREVA/ARL (ADAMS Accession No. ML080310263), and utilizedthe following sequence in order of debris addition: Batch 1 (25% of latent fiber debris quantity) Batch 2 (Cal-Sil) Batch 3 (acrylic powder - coating surrogate) Batch 4 (dirt & dust)Batch 5 (Sil-Col-Sil) Batch 6 (tin powder - IOZ surrogate) Batch 7 - 11 (NUKON & mineral wool - fine fibrous debris)Batch 12 (1/32" acrylic paint chips - coating surrogate)Batch 13(NUKON - small fibrous debris)Batch 14(Cal-Sil - eroded)Batch 15 & 16(NUKON & mineral wool - eroded)Batch 17 - 55 (aluminum oxyhydroxide - PNP design basis chemicals)Batch 56 - 117 (aluminum oxyhydroxide - PNP beyond design basischemicals)Each debris type in a 'batch' was mixed in an individual container and separatelyadded to the test flume. In addition, a minimum of two (2) pool turnovers tookplace between the additions of subsequent debris types.Also, the debris types were sequenced from the most transportable to the leasttransportable with the exception of batch 15 and 16. These batches representedthe potentially eroded NUKON and mineral wool fibrous debris that was Palisades Draft RAI Responses for May 2010 Public Meeting 103determined not to transport. The sequencing was based on the prototypical post-LOCA conditions that fibrous debris erosion would not immediately occur, butwould take place at some later time post-LOCA when the combination of wash-down, spray, and recirculation were present.Therefore, it can be concluded that the sequencing of fibrous, particulate, andchemical precipitate debris did not affect the transport of the subsequent debrisbeing added to the test flume nor the PNP bounding debris head loss test results.This conclusion is supported based on the following facts: There were two (2) or more flume turnovers before the addition of the nextdebris type to ensure that the debris type had adequate time to transportwithin the flume. The delay in the addition of subsequent debris wouldensure that the previous debris was not prohibited from potential transportby the subsequent debris. One (1) flume turnover for PNP takes approximately 14.15 minutes basedon the PNP Design Basis scaled ECCS/CSS flow rates and ARL CFDmodel. At the slowest PNP scaled flume flow velocity of 0.0682 ft/s,debris added to the test flume could travel a distance of approximately 58ft during the approximately 14.15 minute duration required for one (1)flume turnover. The PNP test flume length from the point of debrisintroduction to the PNP strainer module is approximately 30 ft. Therefore,the added debris could travel a potential distance ofalmost 4 times thelength of the PNP test flume before the subsequent addition of debristakes place based on thelowest PNP scaled flume flow velocity and two(2) flume turnovers. Subsequent added debris would not be prohibitedfrom being transported by previously added debris. Floating and/or settled debris would not preclude the transport ofsubsequent debris since it is located at the outside 'extremes' of the flumeflow boundaries, away from the subject debris.In conclusion, the introduction sequencing of PNP debris did not promote theagglomeration of the debris and did not inhibit the transport of same other thanwhat would have naturally occurred in an open, free flowing water stream suchas what would occur in the PNP post-LOCA containment following initiation ofECCS/CSS recirculation.

NRC Request b.14.q: The February 27, 2008, supplemental response states thatcontainment accident pressure was not credited in evaluating flashingacross the strainer. However, the submergence of the strainer is smallwhen compared to the strainer head loss. In the supplementalresponse the licensee's discussion of air ingestion into the strainer is Palisades Draft RAI Responses for May 2010 Public Meeting 104not well supported. The licensee stated on Page 41 that theNUREG/CR-6224 correlation indicates 0.0% void fraction downstreamof the screen. This statement does not appear correct, particularly inlight of the statement on Page 46 of the supplemental response that nocontainment accident pressure was credited in the flashing calculation.The bases for the conclusion that flashing will not occur should beprovided.The licensee provided updated information on the potential for flashingacross the debris bed. Upon further review of the strainer design, theNRC staff believes that, since the strainer is vented to the atmosphere,flashing will not occur across the debris bed. The exception to this is ifthe water height in containment exceeds the elevation of the strainervents. This could occur because the vents are relatively low inelevation. The vents are about 2.2 ft above the top of the strainer.Therefore if debris head loss is maintained less than 2.2 ft, flashingshould not occur. The strainer design limits debris head loss to 1.57 ft.If this design parameter is not changed flashing will not occur. Pleaseprovide the maximum LOCA flood level and verify that the vent is notcovered or evaluate the potential effects on deaeration and flashing.Entergy Nuclear Operations Response 26b:The staff inquired about whether flashing would occur across the strainer debrisbed. The updated submittal provided in June 2009 summarized the conditionsthat impact flashing across the PLP strainer debris bed. The updated submittalconcluded that there is only marginal strainer submergence for a large breakLOCA to preclude flashing and the containment sump strainer is partiallyuncovered for a small break LOCA. ENO evaluated the containmentoverpressure for the short time period after the switchover to recirculation modeand before there is significant subcooling in the sump. The updated submittalprovided in June 2009 concluded that the available containment overpressure ismore than adequate to preclude the sump inventory from flashing when passingthrough the debris bed.Under the conservative assumption of zero containment overpressure which wasshown to be overly conservative based on the June 2009 submittal, there is apotential concern with flashing for the LBLOCA scenario with the flow passingthrough the strainer debris bed. Flashing will occur across the debris bed if thestatic head of water above the strainer module is less than the maximum strainerdebris head loss limit of 1.57 feet (Reference EC12249, Rev. 0). Therefore, thepotential exists for flashing across the strainer debris bed with the conservativeassumptions that (a) the containment pressure equals the vapor pressure at thecontainment sump inventory temperature and (b) the head loss across thestrainer debris bed is at its maximum design limit. However, the following shouldbe noted:

Palisades Draft RAI Responses for May 2010 Public Meeting 105i. Flashing across the debris bed is not considered a significant concernsince the flow must descend to the level of the core tube, and below, toreach the sump; providing additional static head to collapse any voidscreated associated with flashing.ii. Long term in the event, when the containment sump inventory cools, thecontainment pressure will be greater than the vapor pressure of the sumpinventory providing significant pressure margin to prevent flashing. Theminimum containment pressure long term in the event is bounded by theminimum containment pressure prior to the accident.iii. At containment water levels that do not submerge the containment sumpvents, where the submergence above the strainer modules is less than1.82 feet (595' - 593.18') (Reference Drawing M-74, Sheet 1, Rev. 13;VEN-M802, Sheet 1, Rev. 0 and VEN-M802, Sheet 2, Rev. 1), thecontainment sump vents will be remain open to atmosphere allowing theescape of any possible entrained vapor.Under the conservative assumption of zero containment over-pressure for theSBLOCA condition, the sump inventory may experience flashing when passingthrough the debris bed. The condition does not represent a concern since thepotential for vapor phase transport through the strainer is minimal, given theinterior of the strainer is in direct communication with the containmentatmosphere.This RAI inquires specifically about the impact of higher containment water levelsthat submerge the containment sump vents on flashing across the debris bed. Ata containment water level of 595 feet, when the water level begins to exceed thecontainment sump vent elevation, there would be 1.82 feet (595' - 593.18')(Reference Drawing M-74, Sheet 1, Rev. 13; VEN-M802, Sheet 1, Rev. 0 andVEN-M802, Sheet 2, Rev. 1) of static head above the top of strainer modules.Since this static head exceeds the debris head loss design limit of 1.57 feet(Reference EC12249, Rev. 0), there is adequate static head of water above thestrainer modules to prevent flashing assuming no containment overpressure. Asdescribed above, ENO concludes that the available containment overpressure ismore than adequate to preclude the sump inventory from flashing when passingthrough the debris bed. Therefore, ENO concludes that the increasedcontainment water level that submerges the containment sump vents will providean increased static head of water above the strainer modules that will furthersuppress the potential for the sump inventory from flashing when passingthrough the strainer debris bed.

Palisades Draft RAI Responses for May 2010 Public Meeting 106 NRC Request 27.(Audit Open Item 6.1) The licensee should provide a justification that Test4 resulted in a realistic or conservative head loss for the strainer.Specifically, the licensee should provide additional information thatjustifies that a change in strainer hole size from 0.045 inches (Test 2, highhead loss) to 0.095 inches (Test 4, low head loss) would result in achange in head loss of greater than an order of magnitude. The issuesidentified briefly below and discussed in detail in the NRC staff ChemicalAudit Report of Palisades (ADAMS Accession No. ML091070664) shouldbe considered in the development of the response to this open item. Theaudit report should be referenced and the issues discussed in the reportshould be fully addressed. The licensee provided some informationregarding this issue and its sub-parts. However, additional information isrequired as described below.a. Please describe the testing methodology for the referenced PCI testingfor foreign plants that shows that strainer hole size results in significantchanges in head loss across the strainer. Provide information thatvalidates that the testing was conducted similarly to the Palisadestesting or that the results of the testing can be applied to the Palisadesstrainer.Entergy Nuclear Operations Response 27a:The foreign utility strainer head loss testing and test results described in PCIproprietary document (ML090050043) as submitted by PNP to the Staffdiscusses the test methodology. Specifically, on Attachment 1, page 1 of 2 thesubject document it states in part, -All tests were in the Alden small flume to atesting protocol similar to that used by PCI for small flume testing in the U.S..The Staff is very familiar with the Small Flume Test Protocol utilized for USLicensee head loss testing at ARL. The Staff has previously witnessedapplication of the Small Flume Test Protocol for US Licensee plants on a numberof occasions, which were subsequently documented by the Staff (i.e.,ML052230269 (03/17/05) and ML060750340 (01/18/06)).As stated, the Small Flume Test Protocol was utilized for the foreign utilitystrainer head loss test program. On the other hand, PNP utilized the LargeFlume Test Protocol which is similar, but is based on a number of differentcriteria and objectives. However in both cases, the PCI Sure-Flow Strainertechnology was utilized in the strainer design and therefore a comparison of thetest results can be made, specifically the effect of strainer perforated plate holesize and head loss.The document (ML090050043) previously provided to the Staff presented acomparative summary that concluded that a smaller strainer perforated plate hole Palisades Draft RAI Responses for May 2010 Public Meeting 107size opening will result in an increased strainer head loss when the design basisparameters are essentially the same. This conclusion was true for both theforeign utility and PNP when the strainer head loss test was performed to thesame test protocol and similar design basis parameters with the only variablebeing the strainer perforated plate hole size in both cases.In addition to the conclusions reached in the document (ML090050043), thereare other technical references that support the findings previously offered in theRAI response by PNP. Some of them are discussed as follows.The technical paper,An Investigation of Flow Through Screens by W.D. Barnesand E.G. Peterson that was contributed by the Hydraulic Division and presentedat the Annual Meeting of the American Society of Mechanical Engineers(November 26 - December 1, 1950) further supports the test results thatconcluded that a smaller perforated plate hole size will have a higher head lossthan a larger hole size. The paper concluded that the greater the solidity ratio,that is the fractional degree to which the screen or perforated plate obstructs theflow, the greater the head loss. In other words, the smaller the hole size, thegreater the head loss. The paper states in part, -If a considerable dissipationof energy , i.e., reduction of pressure, is required, this may be obtained with asingle screen of high solidity ratio - but at the expense of evenness in thevelocity distribution

.Another technical paper,Discharge Coefficients Through Perforated Plates byP.A. Kolodzie, Jr. and Matthew Van Winkle printed in the AIChE Journal, Volume3, No. 3, 1957 further supports the test results and concluded that -Thevariables which affect the orifice coefficients were found to be the hole diameter,hole pitch, plate thickness, fraction of the plate covered by the perforated platearea, and a Reynolds number based on the hole diameter.In the text,Handbook of Hydraulic Resistance (3 rd Edition) by I.E. Idelchik,specifically Chapter 8,Resistance to Flow through Barriers Uniformly DistributedOver the Channel Cross Section: Resistance Coefficients of Grids, Screens,Porous Layers, and Packings there is additional supporting discussion that holesize affects the resistance coefficient as it relates to head loss.It can therefore be reasonably concluded, that a smaller perforated plate holesize will have a higher head loss than a larger hole size.

NRC Request b.Please provide an evaluation of the information provided by PCI to theNRC staff in ADAMS Accession No. ML090050043 (proprietary) thatstates that hole size has not been observed to directly impact headloss performance and provide information as to why the Palisadesstrainer would not behave similarly to the description in that document.

Palisades Draft RAI Responses for May 2010 Public Meeting 108Entergy Nuclear Operations Response 27b:As discussed in detail for the RAI 27a response, it was concluded that a smallerperforated plate hole size will have a higher head loss than a larger hole size.Therefore, it does not appear that the Staff's RAI statement,Please provide anevaluation of the information provided by PCI to the staff in ML090050043(proprietary)that states that hole size has not been observed to directlyimpact head loss performance and provide information as to why the Palisadesstrainer would not behave similarly to the description in that document.

, iscorrect, since there is no supporting basis for the subject Staff statement.As a matter of fact, the PCI document (ML090050043) supports, states, andconcluded that perforated plate hole size does make a difference. Specifically, asmaller perforated plate hole size based on the same design basis parameterswill have a higher head loss than a larger perforated plate hole size.It appears that the Staff's statement may be incorrect. If the Staff can providefurther clarification regarding RAI 27b, PNP will respond accordingly. However,based on the statements and information provide by the Staff, PNP cannotprovide further expalantion at this time.

NRC Request c.Please provide an evaluation of why the addition of eroded finesresulted in a significant head loss increase indicating that additionalfibrous debris transport affected the ability of the bed to create headloss. This evaluation should concentrate on justification that fibrousdebris transport issues during the test did not result in the largedifference in head loss between the two tests.Entergy Nuclear Operations Response 27c:The AREVA Test Plan (63-9095797-001) and the Test Report (66-9097941-000)were reviewed with regard to both Tests 2 and 4, respectively. This was donesince the Staff did not state which test was the basis for their first statementregarding the addition of eroded fibers and increased head loss. The review ofthe both AREVA documents could not support the Staff's statement regarding theaddition of eroded fines. PNP is not sure as to what documents the Staff hasreviewed that supports their statement.Furthermore, in the case of Test 2, eroded fines were never added to the testflume due to the documented high head loss. Test 2 was terminated before theeroded fines were added to the flume. A further review of Test 4 also did notsupport the Staff's statement.

Palisades Draft RAI Responses for May 2010 Public Meeting 109It appears that the Staff's statement may be incorrect or based on otherinformation that PNP has not seen. If the Staff can provide further clarificationregarding RAI 27c, PNP will respond accordingly. However, based on thestatements and information provide by the Staff, PNP cannot respond at thistime.NRC Request d.Please provide an evaluation of the information supplied by PCI thatspeculates that the effect of hole size diminishes above 0.045 inchesand the how this relates to the smaller Palisades hole size of 0.045inches.Entergy Nuclear Operations Response 27d:As discussed in detail for the RAI 27a response, it was concluded that a smallerperforated plate hole size results in a larger head loss. As stated in the PCIdocument (ML090050043), PCI's response and statements are based on thelimited test data from foreign utility strainer testing and that of PNP. Therefore,based on the limited test data, PCI speculated that the difference in head lossvalues was due to the hole size differences. The foreign utility strainer head losstesting and test results described in PCI proprietary document (ML090050043)as submitted by PNP to the Staff discusses and compares the head loss testingfor that of the foreign utility and PNP. Specifically, on Attachment 1, page 2 of 2the subject document it states in part, -PCI speculates the affect of hole sizediminishes above the 0.045" size hole; and becomes increasingly detrimental tohead losses below 0.045" since PCI has had some success in testing with 0.045"holes-PCI's statements were based on the limited testing data available. In addition,the response to RAI 27a provides further support for the conclusion that smallerperforated plate hole sizes will have a higher head loss than a larger perforatedplate hole size.

NRC Request e.Please provide an evaluation of the physical phenomenon caused bythe difference in strainer hole size that resulted in a debris bedmorphology difference significant enough to result in the largedifference in head loss values between the two tests.Entergy Nuclear Operations Response 27e:As discussed as part of the response to RAI 27a and to a lesser extent in RAI27d, it was concluded that a smaller perforated plate hole size will have a higherhead loss than a larger perforated plate hole size. PNP did not make any Palisades Draft RAI Responses for May 2010 Public Meeting 110statements nor reach conclusions with regard to small or large perforated platehole size and their potential to initiate physical phenomenon and resultant debrisbed morphology changes as they relate to head loss.Any analysis and/or evaluation would be purely speculative regarding peroratedplate hole size and debris bed morphology as the result of physical phenomena.As far as PNP knows there is no know publicly available test reports that addressthe issue of perforated plate hole size and post-LOCA debris bed head loss. Inlight of the unavailability of public test reports, the PCI test data for a foreignutility and that for PNP with regard to perforated plate hole size as it relates tostrainer debris laden head loss may be the only test data available. Thereference documents discussed in RAI response 27a do support the effect ofperforated plate hole size and head loss, but they are all based on testsperformed with clean fluids and no debris of any kind.

NRC Request f.Please provide an evaluation of how the debris addition sequencedifferences between the two tests affected the results. Specifically,address why the separate addition of Nukon and mineral wool in Test 2provided similar transport of the fiber when compared to transport whenthe fiber addition consisted of a mixture of Nukon and mineral wool.Provide justification that excessive interaction of the fiber types did notoccur and that transport was not affected non-conservatively. Considerthat the concentration of the fibrous debris in the test flume is likelymuch higher than would be expected in the plant. Provide informationfrom the testing that validates the evaluation. Alternately provideinformation that shows that the two fibrous debris types would both begenerated by all break scenarios and arrive at the strainer near fieldsimultaneously, and that the concentration in the test flume had nosignificant effect.Entergy Nuclear Operations Response 27f:The June 30, 2009 submittal provided information related to RAI 27f. In general,the mixing was discussed as prototypical and similar transport was expected.Based on discussions to date with the NRC, certain elements of the existingtesting protocol will not be accepted by the NRC without refinements. Therefore,to eliminate any concern over whether the mixing of Nukon and mineral wool hadany unjustified advantageous effect, for any future testing using a refined PCIflume test protocol, ENO would plan to add the Nukon and mineral woolseparately to eliminate this potential NRC concern.

Palisades Draft RAI Responses for May 2010 Public Meeting 111 NRC Request 28.(Audit Open Item 6.2, related to 14.e above) The licensee should provideinformation that the test methodology resulted in realistic or conservativestrainer head loss testing results. In particular, debris preparation andintroduction methods used during testing should be justified asprototypical or conservative. The issues identified briefly below anddiscussed in detail in the NRC staff Chemical Audit Report of Palisades(ADAMS Accession No. ML091070664) should be considered in thedevelopment of the response to this open item. Items 1, 2, and 3discussed in Section 5.4.2 of the audit report should be considered whendeveloping the response to this open item. The licensee providedresponses to the issues listed below, but additional information is requiredas described below:a. Observation of test video documenting the addition of fibrous debrisindicated that the debris may not have been prepared as finely as NRCstaff guidance would suggest or may have agglomerated during thedebris introduction process. There are several examples on the videothat indicate that fiber preparation and/or introduction may not havebeen controlled to the degree prescribed in NRC staff guidance. Thelicensee stated that the debris distribution used during the testing wasrepresentative of the post-LOCA debris distribution expected for theplant. The licensee further stated that the protocol implemented wasconsistent with other tests run by PCI that the NRC staff had observed.In addition, it was noted that some variation in debris form is expecteddepending on the debris preparation and introduction methods. TheNRC staff's position is that these processes should be controlled suchthat consistent debris characteristics are maintained. The NRC staff'sobservations of previous testing for various plants resulted in tripreports that documented that fibrous debris preparation andintroduction needed to be more closely controlled to ensure that finefibers were properly prepared and introduced into the test flume. TheNRC staff has noted inconsistency with these processes during PCItest observations as documented in trip reports. The agglomerationnoted on the video of the Palisades test was excessive for fine fibers.The licensee further noted that any clumped fibrous debris would breakup as it entered the flume and that the video alone is not a valid basisfor judging the final form of the fiber once it enters the flow stream.The NRC staff agrees that some break down of clumps will likely occurwhen the fiber enters the flow stream. However, the degree of thiseffect is unknown. In addition, the NRC staff could not determine thatthe debris added to the flume was actually fine debris and may havebeen larger pieces. In the past, when the NRC staff noted excessiveagglomeration, PCI has adjusted its practices to ensure that the debriswas separated adequately that it could be seen that the debris was fine Palisades Draft RAI Responses for May 2010 Public Meeting 112and not agglomerated. The licensee's response does not adequatelyaddress this issue. Please provide information that justifies that debrisintroduction and preparation provided prototypical or conservative test results.Entergy Nuclear Operations Response 28a:The PNP Design Basis test (Test 4) debris preparation and addition wasconservative and prototypical for an integrated debris transport and head losstest. Accordingly, the PNP ARL Large Flume Test debris addition did notadversely affect the ability of transportable debris to reach the test strainer. Theconcentration of debris slurries prepared for testing, the potential for debris toagglomerate during preparation and addition of same was prototypical and alsovery conservatively, since the actual and expected PNP post-LOCA conditionswithin the containment were represented. Please refer to the responses to RAI17c and 26a for additional information and discussion. The subject RAIresponses address and provide additional and related information for RAI #28a.The concentration of debris slurries prepared for testing, the potential for debristo agglomerate during preparation and addition of same was prototypical of theexpected PNP post-LOCA conditions, and very conservatively represented theactual and expected PNP post-LOCA conditions within the containment.The agglomeration of debris did not occur due to prototypical debrisconcentrations during the preparation, during the addition of same, or within theARL test flume for PNP. During debris addition slurries prepared for testing, thepotential for debris to agglomerate during preparation and addition of same wasprototypical of the expected PNP post-LOCA conditions, and very conservativelyrepresented the actual and expected PNP post-LOCA conditions within thecontainment.It should be noted that there are no guidance documents, such as NUREG/CRsor similar type documents that define excessive agglomeration, prototypicaldebris concentrations as they relate to agglomeration during debris preparationor addition, or higher than expected debris concentrations during debris additionas related to testing. Simply put, there is absolutely no available definitive orsubjective documents and/or information that discuss the term prototypicalagglomeration, the aforementioned issues or their potential impact on head loss testing.Since there are no definitive or subjective criteria and/or guidance available withregard to the subject issues, the agglomeration of fibrous debris and moreimportantly, the homogeneous 'mixing' of both fibrous and particulate debris inthe period prior to and after ECC/SCSS pump recirculation initiation would berealistically expected and prototypical. Numerous NUREG/CRs including 6773as well as others support the agglomeration of debris, the settlement of same, Palisades Draft RAI Responses for May 2010 Public Meeting 113the concentration of same, as well as the homogeneous 'mixing' of fibrous andparticulate debris prior to and after the initiation of ECCS/CSS pumprecirculation. Accordingly, the PNP test at ARL utilized the limited guidancefound in NUREG/CR-6773.PCI also utilized the guidance provided in NEI 04-07 and the Staff's SER for NEI04-07 for the initial preparation of the Large Flume Test Protocol. The subjectProtocol was discussed face-to-face and in numerous telephone conversations(ML072530885 & ML080370262) with the Staff prior to the first Licensee test inearly 2008. Based on observation by and comments from the Staff(ML081830645) during the first Licensee test as well as 'lessons learned' fromthe subject test, the Large Flume Test Protocol was further revised to addressboth the Staff's comments and the 'lessons learned'.The Staff stated in a Licensee's July 9, 2009 public meeting to discuss applicableRAIs, that the fiber classes 1 - 3 per NUREG/CR-6808, Table 3-2 are acceptablewith regard to 'defining' fine fiber. It should be noted that the subject NUREG/CRand Table are specifically associated withSection 3.1.2.1 Size Classification ofFibrous Debris of the subject NUREG/CR. The subject Section is based on air-blast testing experiments of fibrous debris. In other words, the debris sizingclassification in Table 3-2 is based ondry simulated post-LOCA destroyedfibrous debris. The subject NUREG/CR does not address the aspects of testingwith the subject dry fibrous debris, the sizing of same, agglomeration of same, orthe affects of agglomeration on testing. It should be noted that no reference ismade to NUREG/CR-6808 in either NEI 04-07 or the Staff's SER with regard tothe specific subject of fibrous debris sizing classifications or Table 3-2. Mostimportantly, it should be noted that the Staff's 'new' position with regard to fibrousdebris classification for testing was made public approximately eight (8) monthsafter the PNP testing was completed in November 2008.Prior to the initiation of any Licensee testing in early 2008 and the completion ofthe Large Flume Test Protocol, PCI presented various samples of processed dryfibrous debris to the Staff in late 2007 during a GSI-191 public meeting (the Staffwas given the samples and is believed to still have the subject samples in theirpossession). The samples were presented to the Staff in order to solicitcomments or recommendations regarding the processing and size classificationof the processed dry fibrous debris with regard to the proposed Large Flume TestProtocol. The subject samples consisted of the three (3) classifications of fibrousdebris: latent/fines, small fines, and larges as defined in NEI 04-07 and thesubsequent Staff SER.The Staff indicated that the subject samples were representative of what theyexpected for each of the three (3) subject classifications of dry fibrous debris. Itshould be noted that PCI and the Licensees utilize a more conservative definitionof small fines than that of the guidance documents (i.e., NEI 04-07 and the Staff'sSER for NEI 04-07). PCI utilizes a 1" x 4" grating in lieu of the recommended 4" Palisades Draft RAI Responses for May 2010 Public Meeting 114x 4" grating to separate fines/smalls from larges. Therefore, the PCI definition ofsmall fines results in significantly smaller-sized fibrous debris than, if theguidance recommendation were followed. Again, it should be noted that NEI 04-07, the Staff's SER for same, or the March Guidance Document specificallyaddressed, provided guidance, or discussed the size classification of small finesand larges.Following the first Licensee Large Flume Test in February 2008, the Staff inMarch 2008 issued the document,NRC Staff Review Guidance RegardingGeneric Letter 2004-02 Closure in the Area of Strainer Head Loss and Vortexing(also known as the March Guidance Document) (ML072600348). It should benoted that no reference is made to NUREG/CR-6808 in the March GuidanceDocument with regard to the specific subject of fibrous debris sizingclassifications or Table 3-2. However, the various terms fines, fine fiber, finerfiber, etc. are utilized in the subject document, but are never defined within thedocument and/or reference made to NUREG/CR-6808, specifically Table 3-2.During the public GSI-191 meeting of October 24, 2007, held prior to the initialLicensee Large Flume Test, the issue of fibrous debris 'fines' was specificallydiscussed and questions were raised by the Licensee, PWROG, and NEIrepresentatives. In this meeting, the Staff agreed and stated in the meeting that'fines' were not single fibers, but could be 'clumps' or 'bunches' of fibers

.No further description or definition of what the Staff meant with regard to 'clumps'or 'bunches' was provided, in no case was NUREG/CR-6808 discussed in thisregard, and most importantly, the term prototypical agglomeration was neverdiscussed. Attendees at the subject meeting left with the understanding thatfibrous debris 'fines' were not individual fibers.In the final version of the March Guidance Document (ML072600348), there ismuch discussion by the Staff that 'fine' fibrous debrisbe individually or readilysuspendable fibers (Page 4 as well as others). The discussion in the subjectdocument is in direct contradiction with the Staff SER for NEI 04-07 and the Staffstatements made in the October 24, 2007 public meeting, that'fines' were notsingle fibers, but could be 'clumps' or 'bunches' of fibers

.In the March Guidance Document, there is also much discussion by the Staff thatfine fibrous debris should be mixed so that agglomeration does not occur thatwould not be prototypical. However, no guidance or description is provided bythe Staff of what is meant by agglomeration or 'agglomeration does not occurthat would not be prototypical' as has been previously discussed. It wouldappear that 'clumps' and 'bunches' based on previous Staff discussion (i.e.,October 24, 2007 GSI-191 public meeting) would be appropriate with regard todefining agglomeration.It appears that the Staff's primary concern with regard to fine fibrous debrisagglomeration is that the subject debris will settle and not transport as readily as Palisades Draft RAI Responses for May 2010 Public Meeting 115individual fibers. However, there is no basis provided by the Staff as to why theybelieve that fine fibrous debris would not agglomerate in the post-LOCAcontainment environment. As a matter of fact it would seem that it would bemore likely that the fine fibrous debris would actually agglomerate, would notreadily transport, and instead would readily sink as individual fibers in thequiescent post-LOCA containment 212 oF plus fluid prior to the initiation of ECCSrecirculation. A review of various regulatory documents such as NUREG/CRsprovides no support that fine fibrous debris will exist as individual fibers and notagglomerate in the post-LOCA containment. However, on the other hand thereare regulatory documents that support the opposite position, that is, fine fibrousdebris including individual fibers will settle and not readily transport.NUREG/CR-2982Buoyancy, Transport, and Head Loss of Fibrous ReactorInsulation on Page 19, Section 3.1 Buoyancy Tests, specifically paragraphs a),c), and e) state in part:b) In general, the time needed for insulation to sink wasfound to be lessat higher water temperatures. (emphasis added)c) The fiberglass (Filomat) readily absorbs water,particularly hot water,and sinks rapidly (from 20 to 30 seconds in 120 oF water).This isalso true for individual fibers. (emphasis added)e) - Damaged fiberglass insulation pillowswill sink before activationof the recirculation system - (emphasis added)In addition to the conclusions reached in NUREG/CR-2982 regarding rapidsinking of fibrous debris in heated water prototypical of containment post-LOCAfluid conditions, NUREG/CR-6772,GSI-191: Separate-Effects Characterizationof Debris Transport in Water,provides further evidence and support that fibrousdebris will rapidly sink in heated water. Section 2.2.1.1Effect of Temperaturestates in part -Post-LOCA water temperature is approximately 80 oC, which issignificantly different for the ambient temperature proposed for use in the testprogram. It can be postulated that water temperature could affect debris settlingcharacteristics because density and viscosity are temperature-dependent.Further the saturation rates of debris may be temperature-dependent becausesurface tension varies with temperature. This set of experiments provided datato quantify these effects, as described below.Saturation of Debris in Hot Water. When fibrous debris was introduced to waterat ambient temperature, it was observed to float on the surface for more than 24h. Even when shredded fiber fragments were forcibly immersed in ambient-temperature water for 24 h, they subsequently would rise up to the surface whenreleased. However, if the fiber shreds were immersed in hot water (80 oC) for aslittle as 2 min, they readily sank and remained submerged. In the aftermath of a Palisades Draft RAI Responses for May 2010 Public Meeting 116LOCA, the temperature of water in a PWR recirculation sump is likely to becloser to 80 o C than to (~ 20 o C).It should be noted that 80 o C is 176 o F, and 20 o C is 68 oF. Therefore debrissettlement and transport tests in ambient (i.e., cold) water are very conservativebut are significantly non-prototypical of the post-LOCA containment fluidconditions. In addition, the tests and subsequent results would provide non-prototypical guidance with regard to fibrous debris transport and settlementissues and the evaluation of same. Simply stated, the fibrous debris includingindividual fibers would settle at a much greater rate and would not transport asreadily.Since the post-LOCA water in all cases for all Licensees will initially easilyexceed 120 oF, it can be concluded based on the subject NUREG/CR report thatfiberglass insulation and more importantly individual fiberglass fibers will sinkbefore the initiation (i.e., approximately 15 - 40 minutes post-LOCA) of the ECCSin the recirculation mode. It is interesting to note that neither NUREG/CR-6808nor the March Guidance Document (ML072600348) makes any mention ofNUREG/CR-2982 or NUREG/CR-6772, and specifically that fiberglass individualfibers will sink in 20 to 30 seconds in 120 oF water. In addition during Licenseefiber by-pass testing at the Alden Research Laboratory it was also noted thatsignificant quantities of fibrous debris including fine fibrous debris settled and didnot reach the strainer in the prototypical flow streams and at water temperaturesof approximately 120 oF. It should be noted that the subject fibrous debris wasnot allowed to 'sit' in a quiescent flume, but was instead very conservativelyadded to a moving flume flow stream, and still significant settling of the fibrousdebris occurred.NUREG/CR-6224 may be the only document that indirectly discusses non-agglomeration of debris (i.e., fibrous and particulate). However, this NUREG/CRis neither prototypical nor realistic of post-LOCA containment conditions withregard to debris size, configuration, agglomeration, mixing, etc. Instead, thesubject NUREG/CR discusses closed vertical pipe loop (CVPL) testing of variousdebris types under theoretical extremely conservative, non-prototypical, and non-realistic conditions (i.e., addition of debris types in non-prototypical configurationsand quantities, non-prototypical sequencing of debris types andsizes/configurations, non-prototypical flow velocities, non-prototypical applicationof gravity, non-prototypical post-LOCA conditions, etc.). The subject NUREG/CRalso applies pipe flow parameters and conditions to the post-LOCA containment,when in reality the post-LOCA containment is open-channel flow. In addition, theanalysis of the two (2) flow types is considerably different and not related. ThePNP Large Flume Test performed at ARL is an integrated test that incorporatesboth debris transport and debris head loss testing. Therefore, most if not all ofthe content of NUREG/CR-6224 does not apply to the subject PNP test.However, it appears that the Staff is applying portions of NUREG/CR-6224 to the Palisades Draft RAI Responses for May 2010 Public Meeting 117PNP Large Flume Test, specifically debris sequencing, fibrous debrisclassification, and fibrous debris transport/settlement (i.e., agglomeration).The application of NUREG/CR-6224 test conditions and criteria with regard to thereview of the PNP Large Flume Test at ARL which is an integrated test (i.e.,debris transport and head loss) is therefore neither appropriate nor relevant withrespect to the realistic and prototypical post-LOCA conditions expected in thePNP containment, and which were utilized to the extent possible in the PNPLarge Flume Test at ARL.As previously discussed, the PNP concentration and potential agglomeration offibrous debris following a post-LOCA event is prototypical and do not reflect thenon-prototypical but extremely conservative positions of NUREG/CR-6224. TheStaff's real concern is therefore interpreted to mean - did the fibrous debrisutilized in the PNP test have adequate separation (fibers) to facilitate fibrousdebris transport and allow collection on the test strainer in a representative orbounding manner based on the PNP expected post-LOCA conditions? Theaffect on head loss caused by fibrous debris agglomerating in the test flume isnot relevant so long as the fibrous debris transports in the test flume and thecollection of fibrous debris on the test strainer is representative and bounding tothe specific PNP plant conditions.The responses to RAIs 17e and 26a, inclusive form a comprehensive answer tothis question. To re-state the response to this RAI - The concentration of fibrousdebris slurries prepared for testing, the potential for fibrous debris to agglomerateduring preparation, and addition of same was prototypical of the expected PNPpost-LOCA conditions, and very conservatively represented the actual andexpected PNP post-LOCA conditions within the containment.In addition, the following should be noted. The fibrous debris preparation and introduction of same to the ARL test flumedoes represent a very conservative quantity of small fines including fines andsmall fiber clumps which are transported in the representative PNP flowstreams. Since the fibrous debris is transportable, it will collect on the PNPtest strainer based on the expected PNP post-LOCA fluid flow conditions. PCI has processed raw fibrous debris materials into fine fibrous debrisrepresentative of either eroded or latent fibrous debris and small fines byrecognized mechanical process devices (i.e., chipper (smalls) & Munsonmachine (fines)). PCI has separated (i.e., size distribution) the processed fibrous debris utilizinga 1' x 4" grating opening which ismore conservative than the 4" x 4" gratingopening identified in NEI 04-07 and the Staff's SER for the same.

Palisades Draft RAI Responses for May 2010 Public Meeting 118 Samples of latent, fines, small fines, and larges were provided to the Staffbefore any Large Flume Testing was initiated and were found to berepresentative of what the Staff had expected

. Fibrous debris is processed and prepared in accordance with the PCI 'whitepaper' SSFS-TD-2007-004Sure-Flow Suction Strainer - Testing DebrisPreparation & Surrogates, the PCI 'white paper' SSFS-TD-2007-004,Supplement 1, Rev. 1,Sure-Flow Suction Strainer - Testing DebrisPreparation & Surrogates (ML092430056 & ML092580203), and the LargeFlume Test Protocol which have been provided to and discussed with the Staff. Documented positive observations and comments by the Staff(ML081840095) and lessons learned by PCI/AREVA/Alden during the initialLarge Flume Test for a Licensee were incorporated into all subsequent tests. Actual processing, preparation (i.e., mixing of debris for testing), andintroduction of fibrous debris for ARL Large Flume Testing has consistentlybeen performed by the same PCI and Alden personnel, respectively.To summarize, all of the fibrous debris (i.e., latent, fines, small fines, and smalls)for PNP as well as that for all of the other Licensees has been processed,prepared, and introduced to the test flume in accordance with the PCI 'whitepaper'Sure-Flow Suction Strainer - Testing Debris Preparation & Surrogates , thePCI/AREVA/Alden Large Flume Test Protocol, and most importantly by the samePCI and Alden personnel (in most cases). Observations and comments by theStaff and lessons learned by PCI/AREVA/Alden during the initial Large FlumeTest for a Licensee were incorporated into all subsequent tests. Simply put therehas been a significant level of consistency in the processing, preparation, andintroduction of latent, fines/smalls, and large fibrous debris into the Large TestFlume. It should be further noted that samples of processed fibrous debris (drymaterial) as latent, fines/smalls, and larges were provided to the Staff and thedetermination was made that 'the samples were representative of what the Staffhad expected.'In conclusion, the preparation, concentration, and introduction of fibrous debrisdid not promote the agglomeration of the debris and did not inhibit the transportof same other than what would have naturally occurred in an open, free flowingwater stream such as what would occur in the PNP post-LOCA containmentfollowing initiation of ECCS recirculation.

NRC Request b.The debris introduction sequence for the testing did not appear to beperformed consistent with the procedure previously discussed betweenPCI/AREVA and the NRC staff. Some more easily transportable Palisades Draft RAI Responses for May 2010 Public Meeting 119debris was added after less transportable debris. For example, debrisadded as eroded fibrous material was added after larger fibrouspieces. This is a potential non-conservative practice because in thetest a large debris pile may form in the test flume. This pile may act asan impediment to the transport of debris that may otherwise transport ifthe pile was not present. In the plant such a debris pile is less likely toform because the concentration of debris is much lower than in thetest. The debris captured in the flume overflow filters was also addedat the original drop zone which is behind the debris pile. A portion ofthe latent fiber was added to the test flume prior to starting therecirculation pump. This may be non-conservative from a transportperspective because washdown and pool fill up transport is notmodeled. It has been noted that the velocity of the flume is increasedif a debris pile is present. While the debris pile will increase flumevelocity to some extent, a porous debris pile on the flume bottom couldcapture debris such that the affect of higher flume velocity is negated.There are many variables that affect debris transport. The NRC staffcould not determine that an adequate evaluation of these variables andtheir uncertainties was attained prior to the determination of debrisintroduction sequencing. The licensee stated that the addition of theeroded debris near the end of the non-chemical debris introduction isrepresentative of the expected plant response. With the exception ofthe debris pile in the flume that would not be present in the plant, this istrue. However, there are competing effects such that the debris pile inthe flume may adversely affect the transport of the eroded fines. TheNRC staff expects that the transport of fibrous debris in the flume overthe debris pile may be dependent on the flume flow velocity withtransport being more likely in higher flow streams. Because finefibrous debris is known to be problematic for head loss, the licenseeshould employ test practices that ensure that the fibrous debris has anopportunity to transport similarly to expected plant transport.Alternately, the licensee could provide information that shows that thefine debris transports. During observations of a different test thatadded eroded fibers to the flume after other debris, the NRC staff wasable to verify that the debris transported (based on changes in thehead loss trend). A review of the head loss trend supplied in thelicensee's supplemental response did not provide definitive informationfor this issue. Please provide justification that the debris introductionsequence did not affect head loss test results non-conservatively.Entergy Nuclear Operations Response 28b:It should be noted that RAI 28b is almost identical in content to both RAI 17c andStaff Concern/Issue 3 from RAI 26a with regard to the issue of fibrous debrissequencing. Please refer to both RAI 17c and Staff Concern/Issue 3 from RAI26a for additional discussion and explanation with regard to this issue.

Palisades Draft RAI Responses for May 2010 Public Meeting 120The specific issue of eroded fibrous debris being added after other lesstransportable fibrous debris is in the test flume is addressed by the response toRAI 17c.The new issue identified in RAI 28b, A portion of the latent fiber was added tothe test flume prior to starting the recirculation pump.,was not addressed bythe responses to RAIs 17c and Staff Concern/Issue 3 from 26a.Based on the AREVA NP Test Plan for PNP, 25% of the latent fibrous debris wasintroduced along the length of the test flume prior to starting the recirculationpump. NUKON fine fiber was used as the debris constituent for latent fibers. Atotal of 0.40 lbm of NUKON fine fiber was weighed and introduced along thelength of the test flume.The location of post-LOCA debris before ECCS pump recirculation initiation isbased on prototypical post-LOCA containment conditions indicating that thedebris including latent fibrous debris would be scattered to varying degreesthroughout the containment floor areas with heavier debris concentrationimmediately adjacent to the actual LOCA pipe break location. This prototypicalcondition is consistent with and based on the methodology and objectivesdocumented in NUREG/CR-6773 regarding debris-transport tests for PWRs.Therefore, it is certainly reasonable to expect some latent fiber to be located nearthe strainer and on the containment floor prior to the initiation of ECCSrecirculation. Please note the following:1. Numerous NUREG/CRs including 2982, 6772, and 6773 confirm that finefibrous debris settles quickly (i.e., one (1) minute or less) in watertemperatures as low as 120 ºF which is considerably less than the actualinitial post-LOCA expected water temperature of greater than 212 ºF.2. There is approximately 15 - 45 minutes following the post-LOCA eventprior to the initiation of ECCS recirculation. This would allow any fibrousdebris including latent fiber to be subjected to both a significant timeperiod and the greater than 212 ºF temperature water which would resultin ideal conditions for settlement of the fibrous debris in the post-LOCAcontainment fluid as well as in the near field of the strainers. It should alsobe noted that the strainer/sump location is considered to be in a quietzone when post-LOCA containment fill is complete. Regardless, 'some'fibrous debris is expected to be on the containment floor at the time ofECCS pump recirculation.Therefore, it can be reasonably concluded that it is prototypical andrepresentative that 'some' fibrous debris is located on the containment floor nearthe strainer/sump location when recirculation begins. The absence of 'some'fibrous debris during an initial Licensee's qualification test at ARL near the test Palisades Draft RAI Responses for May 2010 Public Meeting 121strainer was, in fact, observed to be a weakness in the test protocol by the Staffwho witnessed the test. In response to the comment, PCI revised the Test Planand Protocol to begin introduction of either 0.5 lbs or 25% of the design basislatent fibrous quantity into the ARL test flume between the debris introductiondrop zone and test strainer five (5) minutes prior to pump start to address thisconcern. All subsequent tests have followed this refinement to the Test Protocolwhich was discussed with the Staff approximately six (6) months prior to the PNP test.To further address the Staff's concern, PCI/ARL implemented a number ofspecial tests (August - September 2009) to observe in clean water (i.e., noparticulate or chemical debris) the transportability of latent fibrous debrisintroduced five (5) minutes prior to pump start-up (i.e., simulated initiation of post-LOCA ECCS/CSS recirculation). The use of clean water without particulatedebris is very conservative, since the particulate debris would have 'mixed' withthe fibrous debris and resulted in significant trapping' or settlement of the 'mixed'fibrous and particulate debris. Implementation of the test without particulatedebris is not prototypical of the expected Licensee post-LOCA conditions, but isvery conservative with regard to the Test Protocol and test results. The testutilized various flow velocities that were bounding for various Licensee facilitiesand as were implemented in the ARL large flume test. ARL observed that thelatent debris placed on the small flume floor with the pump off resulted in thesubject fibrous debris being transported at various Licensee flow velocities, butvery little of the debris reached the strainer. The flow velocity required to initiatethe fibrous debris movement was usually greater than the actual Licensee flowrates and associated velocities. However, not much of the fibrous debris wastransported to the strainer. In order to actually transport the fibrous debris to thestrainer, flow velocities of more than 300% of the actual Licensee flow velocitieswere required.Since the fibrous debris was observed to transport from its initial resting positionin the subject small flume special tests, the introduction of 'some' latent fibrousdebris prior to pump start was concluded to be realistic, representative, andprototypical of the actual PNP post-LOCA containment conditions and LOCAscenario.NRC Request c.In some photos (especially the fiber only test photos), some finefibrous debris appeared to be clumped into balls. The NRC staff hasobserved other tests during which shredded fiber has clumped intoballs if not properly blended. The observed fibrous debris did notappear to exhibit properties that would be expected to result from jetimpingement. The licensee stated that the debris was preparedprototypically. However, this question was specifically related to thepea-sized balls of fiber observed in the fiber-only test. The licensee Palisades Draft RAI Responses for May 2010 Public Meeting 122should provide additional information that justifies that the fine fiber inthe plant would exist, at least partially, in the form of small tightspheres, or provide an alternate explanation of the appearance of thefinely prepared fiber.Entergy Nuclear Operations Response 28c:A review of the videos was performed that documented the PNP Large FlumeTest performed in November 2008. Specifically chapters 8, 9, 11, 13, 19, and 21of the subject video were reviewed since they dealt with the issues regardingfibrous debris mixing and addition of the mixed fibrous debris to the test flume.The review of the subject video chapters did not produce specific observationand/or recognition of the Staff's statedpea-sized balls of fiber orin the form ofsmall tight spheres. It was observed that some of the fibrous debris within thecontainers as it was added to the test flume was somewhat 'lumpy' indicating thatsome of the fibrous debris was loosely connected, but was also indicative ofClass 1, 2 & 3 fibrous debris as documented in NUREG/CR-6808, Table 3-2, thatthe Staff has previously indicated was an acceptable fibrous debris form for finefibrous debris. However, there was as previously stated no specific observationthat the subject fibrous debris was in the form of eitherpea-sized balls orin theform of small tight spheres

.The addition of the mixed fibrous debris to the test flume utilized 5-gallonbuckets. As the fibrous debris is added to the test flume, the liquid portion of themixed fibrous debris will 'separate' from the heavier fibrous debris, leaving thefibrous debris in a somewhat 'lumpy' state, but not in the form of eitherpea-sized balls orin the form of small tight spheres. As the remaining portions of mixedfibrous debris are introduced to the test flume, small quantities of fibrous debrismay remain either in the bucket and/or on the test flume addition chute. Asshown in the subject PNP videos a significant quantity of water is utilized towash-down and 'mix' the remaining fibrous debris to ensure that it all enters thetest flume in a diluted manner.The effect of utilizing the buckets and the test flume chute result in theappearance of the 'lumpy' mixed fibrous debris. However, the observed 'lumpy'mixed fibrous debris is but a small quantity of the total fibrous debris introducedto the test flume.It should also be noted that Class 3 fibrous debris is defined and pictured as'clumps' measuring up to ~ 2" in accordance with NUREG/CR-6808, Figure 3-3(see photo (Figure 3-3).

Palisades Draft RAI Responses for May 2010 Public Meeting 123In addition, the observed 'lumpy' fibrous debris form is in agreement with theStaff's response to a specific question from NEI, the PWROG representative, andlicensees made in the October 24, 2007 public meeting, that'fines' were notsingle fibers, but could be 'clumps' or 'bunches' of fibers

.Recently (August 2009 - September 2009), Licensees implemented a number oftests to observe in clean water (i.e., no particulate or chemical debris) thetransportability and potential release/separation of fine fibrous debris and finefibrous debris from NUKON small fibrous debris. The use of clean water withoutparticulate debris was utilized for the subject tests, since it is very conservative,as the particulate debris would have 'mixed' with the fibrous debris and resultedin significant trapping and/or settlement of the 'mixed' fibrous and particulatedebris. Implementation of the test without particulate debris is not prototypical ofthe expected Licensee post-LOCA conditions, but is very conservative withregard to the Test Protocol and test results. The test utilized Licensee specificflow velocities that were bounding and as were implemented in the ARL LargeFlume Test.ARL observed and documented, that -The Nukon small fiber is added in thesame manner as the fiber was added in the large flume testing. - A debriscloud of lighter debris can be observed breaking away during the introduction. -The lighter smaller fiber is transporting to the strainer at this time, while theheavier fiber material is starting to accumulate on the floor. The fiber on the flooris transporting across the floor however.ARL further documents, that

- Thefiber that has been introduced into the flume has all either settled or transportedacross the bed of fiber on the flume floor. The fiber is eroding from the leadingedge of the debris pile. - The erosion of the fiber pile has slowed. Some smallfiber pieces continue to break away at this time, but the pile is negligibly affected.ARL goes on to further document, that -shows a picture of the strainer nearthe end of testing. It is clear from this picture that a significant amount of fines Palisades Draft RAI Responses for May 2010 Public Meeting 124were released from the smalls and traveled to the strainer. - shows the pictureof 0.25 lb of latent fine fiber on the screen, it is possible to conservativelyconclude that at least twice this amount separated from the introduced small fiber(2.24 lb) and made it to the strainer here -.The responses to RAI #28 as well as RAI #13, and others all form acomprehensive answer to this RAI. To re-state the response to this RAI - Thepreparation of fibrous debris slurries, concentration of debris slurries prepared fortesting, the potential for debris to agglomerate during preparation and addition ofsame was prototypical of the expected PNP post-LOCA conditions, and veryconservatively represented the actual and expected PNP post-LOCA conditionswithin the containment.In addition, the following should be noted. There is no regulatory and/or industry guidance or criteria thataddresses the issues of fibrous debris preparation, concentration ofdebris slurries prepared for testing, the potential for debris toagglomerate during preparation and addition of same to the test flumein support of an integrated test. The fibrous debris preparation and introduction of same to the ARL testflume does represent a very conservative quantity of small finesincluding fines and small fiber clumps which are transported in therepresentative PNP flow streams. Since the fibrous debris istransportable, it will collect on the PNP test strainer based on theexpected PNP post-LOCA conditions. PCI has processed raw fibrous debris materials into 'fines'representative of either eroded or latent fibrous debris and small finesby recognized mechanical process devices (i.e., chipper (smalls) &Munson machine (fines)). The SE for NEI 04-07 found no issue with the NEI 04-07 definition ofsmall fines. PCI has separated (i.e., size distribution) the processed fibrous debrisutilizing a 1' x 4" grating opening to establish the category of smallfines. The PCI definition of small fines resulted in even moreconservative and significantly smaller fibrous debris for ARL testingthan if either the NUREG/CR-6369 Volume 1 or NEI 04-07 definition ofsmall debris were utilized - PCI utilizes a 1" x 4" standard gratingopening for fibrous debris separation as opposed to the 6' x 4" or 4" x4' standard grating opening found to be acceptable by both the SE forNEI 04-07 and NUREG/CR-6369.

Palisades Draft RAI Responses for May 2010 Public Meeting 125 Samples ofdry latent, fines/smalls, and larges were provided to theStaff before any Large Flume Testing was initiated and were found tobe 'representative of what the Staff had expected'. Fibrous debris has been processed, prepared, and introduced inaccordance with the PCI document,Sure-Flow Suction Strainer -Testing Debris Preparation & Surrogates, Technical Document No.SFSS-TD-2007-004 and the PCI/AREVA/Alden Large Flume TestProtocol which have both been provided to and discussed with the Staff. Observations and comments by the Staff and lessons learned byPCI/AREVA/Alden during the initial Large Flume Test wereincorporated into all subsequent tests. Actual preparation (i.e., mixing of debris) of fibrous debris andintroduction of same for the ARL Large Flume Testing has consistentlybeen performed by the same Alden personnel During various Licensee tests at ARL (August 2009 - September2009) latent fibrous debris introduced upstream of the test strainer andwith the test pump operating at a range of scaled Licensee DesignBasis flow rates and flow velocities, resulted in a 'cloud' of fine fiber(i.e., latent fibrous debris) that transported to the strainer which wasobserved and documented by ARL. In addition, the pump scaled flowrate required to start movement of the latent fibrous debris was in allcases greater than the various Licensee Design Basis flow rates. Aflow rate and associated velocity greater than the various LicenseeDesign Basis conditions is required to transport latent fibrous debris.To summarize, all of the fibrous debris (i.e., latent, fines, small fines, smalls, andlarges), particulate, chemical, and miscellaneous debris for PNP as well as all ofthe other Licensees has been processed, prepared, and introduced to the testflume in accordance with the PCI 'white paper'Sure-Flow Suction Strainer -Testing Debris Preparation & Surrogates, PCI Debris Preparation SSFS-TD-2007-004 Supplement 1, the PCI/AREVA/Alden Large Flume Test Protocol, andmost importantly by the same Alden personnel (in most cases). Observationsand comments by the Staff and lessons learned by PCI/AREVA/Alden during theinitial Large Flume Test for a Licensee were incorporated into all subsequenttests. Simply put there has been a significant level of consistency in theprocessing, preparation, and introduction of latent, fines/smalls, and large fibrousdebris into the Large Test Flume. It should be further noted that samples ofprocessed fibrous debris (dry material) as latent, fines, small fines, smalls, andlarges were provided to the Staff and the determination was made thatthe dryfibrous debris samples were representative of what the Staff had expected

.

Palisades Draft RAI Responses for May 2010 Public Meeting 126In conclusion, the debris preparation, debris classification (sizing), amount ofeach debris classification (size) for each debris classification (size) category, andthe basis for the debris classification (size) distribution chosen for the debris anddebris surrogates including debris concentration and introduction sequencing offibrous debris did not promote the agglomeration of the debris and did not inhibitthe transport of same other than what would have naturally occurred in an open,free flowing water stream such as that which would occur in the PNP post-LOCAcontainment following initiation of ECCS/CSS recirculation.

NRC Requestd. Some debris may enter the containment pool closer than 30-40 ft fromstrainers during the blowdown, washdown, and pool fill-up phases ofthe LOCA. This debris would be more likely to transport to the strainerand less likely to contribute to the debris pile in the test flume. The testprocedure did not attempt to model this aspect of the postulated event.This potential issue would likely have more influence as flume flowvelocities decrease because settling would tend to occur over a shorterdistance in a low velocity flow stream. Palisades' velocities arerelatively low. The licensee provided information regarding theprobable distribution of debris in the containment at the start ofrecirculation. The information appears to be reasonable. However,the test input parameters were based on debris amounts predicted bythe transport evaluation to be at the strainer, not 30 ft away. It isapparent that not all of the debris reached the strainer during the test.The NRC staff believes that some debris would be closer than 30 ftand some further at the onset of recirculation. Due to variables in postaccident debris distribution and transport phenomenon, the NRC staffcannot conclude whether the practice of placing most of the debris 30ft from the strainer is conservative or prototypical. The level ofconfidence with this practice resulting in a realistic or conservativehead loss is dependent upon individual plant parameters and testimplementation. Because of the level of uncertainty in some aspectsof the head loss evaluation, the licensee should ensure that sufficientconservatism is included in the testing to assure an overallconservative or prototypical result. In this area conservatism has notbeen demonstrated. Please discuss the acceptability of testingprocedures given the above.Entergy Nuclear Operations Response 28d:See detailed discussion provided in response to RAI 21.

NRC Request Palisades Draft RAI Responses for May 2010 Public Meeting 127 e.The relatively low flume volume has an effect on the concentration ofparticulate and fine debris suspended in the flume. The volume of theflume affects the scaling between the strainer surface and the poolvolume. Having a flume with a larger volume could avoid some of theconcerns with over-concentration of debris in the flume and mayreduce agglomeration. Flume debris concentration is significantlyhigher than the plant condition. The licensee did not provide aseparate response to this issue, but referenced the response to theissues discussed above. Please provide justification that theconcentration of debris in the test flume did not affect transport duringhead loss testing.Entergy Nuclear Operations Response 28e:The staff's understanding of the test flume scaling methodology is not correct.There are actually two (2) scaling methodologies that are employed, but they areseparate and mutually exclusive of each other. Neither scaling methodologydetermines the scaling factor between the strainer module surface area and thepost-LOCA containment fluid volume. This has not been a requirement and/orissue for any Licensee to date, nor are there any requirements and/or guidanceto make such a scaling determination. Simply put, it has not been required ofany Licensee, strainer vendor, or test facility to make this scaling determination,and apply it to an actual strainer head loss test.The first scaling methodology is associated with and is utilized to determine thedebris quantities that will be used for the PNP test. PNP has twenty-three (23)strainer modules. One (1) strainer module was used for the PNP Design BasisTest (Test 4). Therefore, the ratio (i.e., 1/23) is multiplied times the PNP totalstrainer surface area represented by the entire twenty-three (23) modules minus 100 ft 2 of sacrificial area (PNP value) to address miscellaneous tags, labels,stickers, etc. that could potentially cover and block some of the strainer totalsurface area. The result is the scaling factor, which for PNP was 4.475%.The PNP scaling factor, 4.475% is used to determine all debris type (i.e., fibrous,particulate, miscellaneous, and chemical precipitate) quantities and the test flumeflow rate (i.e., gpm). The scaling factor is multiplied by each debris type totalquantity to establish the debris quantity of each debris type that will be utilized forthe PNP test. Likewise, the total PNP ECCS/CSS flow rate is multiplied by thescaling factor to establish the test flume flow rate (gpm).The second scaling methodology is associated with the PNP test flume. Aldenperforms a CFD analysis and generates a fluid flow model of the PNP strainerarrangement and configuration within the PNP containment. The PNP CFDmodel is then utilized to develop the PNP test flume configuration based on thescaled fluid flow rate that was previously discussed. The result is a PNP specificscaled test flume that is prototypical and representative of the PNP post-LOCA Palisades Draft RAI Responses for May 2010 Public Meeting 128fluid flow within containment. The basis for utilizing CFD modeling and themethodology utilized in developing the CFD PNP plant and test flume models isaddressed in NEI 04-07, the SE for NEI 04-07, NUREG/CR-6773,GSI-191:Integrated Debris-Transport Tests in Water Using Simulated Containment FloorGeometries,NRC Staff Review Guidance Regarding Generic Letter 2004-02Closure in the Area of Strainer Head Loss and Vortexing (ML072600348) (i.e.,so-called March Guidance Document), and PCI/AREVA/Alden/Licenseemeetings and telephone discussions that discussed the Large Flume TestProtocol including CFD application and methodology (i.e., More than nine (9)occasions (February 2007 (ML072530885) - February 2008 (ML080370262)).It should be noted, that at no time did the Staff bring up or discuss the issue ofthe scaling factor between the strainer module surface area and the post-LOCAcontainment fluid volume with regard to the test flume. As previously stated thisissue as can best be determined has not been brought up by the Staff, been aStaff concern, or has been documented in a Licensee RAI for any Licensee,strainer vendor, or test facility activities.The test flume size (i.e., water volume) is directly related to and is a function ofthe first scaling methodology (i.e., scaling factor) and the second scalingmethodology (i.e., CFD model). The test flume cannot be changed for anyreason since it is an accurate representation of the flow velocities along the fluidflow path to the strainer module. Any changes such as making the test flumelarger based on the scaling factor between the strainer module surface area andthe post-LOCA containment fluid volume would have a negative and direct affecton the approach velocity within the test flume. Simply put, the test flume flowvelocity would not be prototypical, conservative, or representative of the PNPpost-LOCA containment fluid flow velocities associated with the strainer moduleconfiguration. Any such changes would also be in direct conflict with the LargeFlume Test Protocol that was discussed with the Staff (ML072530885 &ML080370262), as well as the other previously identified reference documentsthat address the use and application of CFD methodology.Please refer to the responses for RAIs 17c, 26a, and portions of the responses to28a-d for additional information and discussion. The subject RAI responsesaddress and provide additional and related information for RAI 28e.Based on the discussion provided in the subject responses, it can be concludedthat the quantity of fine fibrous debris and the fibrous debris classification size ofsame utilized for the PNP Large Flume Test was realistic, representative, andprototypical of the actual PNP post-LOCA containment conditions and LOCAscenario.The PNP Large Flume Test debris addition activities did not adversely affect theability of transportable debris to reach the strainer module. The sequencing andconcentration of debris slurries prepared for testing, the potential for debris to Palisades Draft RAI Responses for May 2010 Public Meeting 129agglomerate during preparation and addition of same was prototypical of theexpected PNP post-LOCA conditions, and very conservatively represented theactual and expected PNP post-LOCA conditions within the containment.The Staff has used the all encompassing term 'agglomeration' to describe thesequencing of debris and the related issue of debris concentration as it affectsdebris transport, settling, and strainer head loss testing. For PNP, theagglomeration of debris did not occur due to prototypical debris sequencing,concentrations during the preparation, during the addition of same to the testflume, or within the ARL test flume for PNP.It should be noted that there are no documents, such as NUREG/CRs or similartype documents that provide guidance and/or define excessive agglomeration,prototypical debris concentrations as they relate to agglomeration during debrispreparation or addition, or higher than expected debris concentrations duringdebris addition as related to integrated (i.e., transport, settling, and head loss)strainer testing. Simply put, there are absolutely no available definitive orsubjective documents and/or information that discuss the term prototypicalagglomeration, addresses the aforementioned issues, or their potential impact onintegrated strainer testing.Since there are no definitive or subjective criteria and/or guidance available withregard to the subject issues, the agglomeration of fibrous debris and moreimportantly, the homogeneous 'mixing' of both fibrous and particulate debris inthe period prior to and after ECCS/CSS pump recirculation initiation would berealistically expected and prototypical. Numerous NUREG/CRs including 6772and 6773 as well as others support the agglomeration of debris during the post-LOCA period, the settlement of same, the concentration of same, as well as thehomogeneous 'mixing' of fibrous and particulate debris prior to and after theinitiation of ECCS/CSS pump recirculation. Accordingly, the PNP test at ARLutilized the limited guidance found in NUREG/CR-6773. It should be noted thatnone of the NUREG/CRs or other documents discuss, address, or support theidea of debris sequencing for transport and/or integrated testing (i.e., debristransport, settling, and head loss).NUREG/CR-6224 may be the only document that indirectly discussessequencing and non-agglomeration of debris (i.e., fibrous and particulate).However, this NUREG/CR had a very different objective than integrated testing,which was to determine the most conservative sequencing that would achievethe highest possible head loss. In other words, the subject NUREG/CR had theprimary objective of determining the worse-case head loss regardless if the testprotocol was prototypical and realistic of the post-LOCA conditions withincontainment. Accordingly, this NUREG/CR is neither prototypical nor realistic ofpost-LOCA conditions with regard to debris transport, settling, size, sequencing,configuration, agglomeration, mixing, etc. Instead, the subject NUREG/CRutilizes a closed vertical pipe loop (CVPL) test apparatus to 'matrix' test various Palisades Draft RAI Responses for May 2010 Public Meeting 130debris types under theoretical extremely conservative, non-prototypical, and non-realistic conditions (i.e., addition of debris types in non-prototypical sequences,sizes, configurations and quantities, non-prototypical flow velocities, non-prototypical application of gravity, non-prototypical post-LOCA conditions, etc.).The subject NUREG/CR also applies pipe flow parameters and conditions to thepost-LOCA containment and debris, when in reality the post-LOCA containmentis open-channel flow which would result in transported debris being directlyaffected by gravity. In addition, the analysis of the two (2) flow types is verydifferent and not related. The PNP Large Flume Test performed at ARL is anintegrated test that incorporates debris transport, debris settling, and debris headloss testing. Therefore, most if not all of the content and conclusions ofNUREG/CR-6224 does not apply to the subject PNP test.The application of NUREG/CR-6224 test methodology, protocol, conditions andcriteria, specifically the sequencing of debris by size and/or transportability withregard to the review of the PNP Large Flume Test at ARL is therefore neitherappropriate nor relevant with respect to the realistic and prototypical post-LOCAconditions expected in the PNP containment, and which were utilized in the PNPLarge Flume Test at ARL.As previously discussed, the PNP ARL Large Flume Test sequencing,concentration, and potential agglomeration of debris following a post-LOCA event is prototypical. The Staff's real concern is therefore interpreted to mean - did thefibrous debris utilized in the PNP test have adequate separation (fibers) tofacilitate fibrous debris transport and allow collection on the test strainer in arepresentative or bounding manner based on the PNP expected post-LOCAconditions? The affect on head loss caused by fibrous debris agglomerating inthe test flume is not relevant so long as the fibrous debris transports in the testflume and the collection of fibrous debris on the test strainer is representativeand bounding to the specific PNP plant conditions.The responses to RAIs 17c, 26a, as well as portions of 28a-d, and others all forma comprehensive answer to this question. To re-state the response to this RAI -The sequencing of debris, concentration of debris slurries prepared for testing,the potential for debris to agglomerate during preparation and addition of samewas prototypical of the expected PNP post-LOCA conditions, and veryconservatively represented the actual and expected PNP post-LOCA conditionswithin the containment. Most importantly, the PNP test protocol utilized theguidance (i.e., NEI 04-07; the SE for NEI 04-07; NUREG/CRs - 2982, -6772 & -6773; and the March Guidance Document (ML072600348)) available at the timeof the test.In addition, the following should be noted.

Palisades Draft RAI Responses for May 2010 Public Meeting 131 There is no regulatory and/or industry guidance or criteria that addressesthe issues of debris sequencing, concentration of debris slurries preparedfor testing, the potential for debris to agglomerate during preparation andaddition of same to the test flume in a detailed manner that would supportan integrated test. The fibrous debris preparation and introduction of same to the ARL testflume does represent a very conservative quantity of small fines includingfines and small fiber clumps which are transported in the representativePNPP flow streams. Since the fibrous debris is transportable, it will collecton the PNP test strainer based on the expected PNP post-LOCA fluid flow conditions. PCI has processed raw fibrous debris materials into 'fines' representativeof either eroded or latent fibrous debris and small fines by recognizedmechanical process devices (i.e., chipper (smalls) & Munson machine(fines)). PCI has separated (i.e., size distribution) the processed fibrous debrisutilizing a 1' x 4" grating opening which is significantly more conservativethan the 4" x 4" standard grating opening identified in NEI 04-07 and theStaff's SER for the same. Samples of latent, fines/smalls, and larges were provided to the Staffbefore any Large Flume Testing was initiated and were found to be'representative of what the Staff had expected'. Fibrous debris has been processed, prepared, and introduced inaccordance with the PCI document,Sure-Flow Suction Strainer - TestingDebris Preparation & Surrogates, Technical Document No. SFSS-TD-2007-004 and the PCI/AREVA/Alden Large Flume Test Protocol whichhave both been provided to and discussed with the Staff. Observations and comments by the Staff and lessons learned byPCI/AREVA/Alden during the initial Large Flume Test were incorporatedinto all subsequent tests. Actual preparation (i.e., mixing of debris) of fibrous debris and introductionof same for the ARL Large Flume Testing has consistently beenperformed by the same Alden personnel During various Licensee tests at ARL (August 2009 - September 2009)latent fibrous debris introduced upstream of the test strainer and with thetest pump operating at a range of scaled Licensee Design Basis flow ratesand flow velocities, resulted in a 'cloud' of fine fiber (i.e., latent fibrousdebris) that transported to the strainer which was observed and Palisades Draft RAI Responses for May 2010 Public Meeting 132documented by ARL. In addition, the pump scaled flow rate required tostart movement of the latent fibrous debris was in all cases greater thanthe various Licensee Design Basis flow rates. A flow rate and associatedvelocity greater than the various Licensee Design Basis conditions isrequired to transport latent fibrous debris.To summarize, all of the fibrous debris (i.e., latent, fines/smalls, and larges),particulate, chemical, and miscellaneous debris for PNP as well as all of theother Licensees has been processed, prepared, and introduced to the test flumein accordance with the PCI 'white paper'Sure-Flow Suction Strainer - TestingDebris Preparation & Surrogates, PCI Debris Preparation SSFS-TD-2007-004Supplement 1, the PCI/AREVA/Alden Large Flume Test Protocol, and mostimportantly by the same Alden personnel (in most cases). Observations andcomments by the Staff and lessons learned by PCI/AREVA/Alden during theinitial Large Flume Test for a Licensee were incorporated into all subsequenttests. Simply put there has been a significant level of consistency in theprocessing, preparation, and introduction of latent, fines/smalls, and large fibrousdebris into the Large Test Flume. It should be further noted that samples ofprocessed fibrous debris (dry material) as latent, fines, small fines, smalls, andlarges were provided to the Staff and the determination was made thatthe dryfibrous debris samples were representative of what the Staff had expected

.In conclusion, the debris preparation, debris classification (sizing), amount ofeach debris classification (size) for each debris classification (size) category, andthe basis for the debris classification (size) distribution chosen for the debris anddebris surrogates including debris concentration and introduction sequencing offibrous debris did not promote the agglomeration of the debris and did not inhibitthe transport of same other than what would have naturally occurred in an open,free flowing water stream such as that which would occur in the PNP post-LOCAcontainment following initiation of ECCS/CSS recirculation.

Palisades Draft RAI Responses for May 2010 Public Meeting 133 NRC Request 29.The June 30, 2009, supplemental response indicates that throttling thecontainment spray flow is credited to ensure adequate NPSH margin.Please provide the basis for determining the reduced spray flow requiredduring recirculation mode.Entergy Nuclear Operations Response:Modification EC8350 "Replace Containment Spray Isolation Valves per GSI-191Resolution" [Ref. 29.1] was performed in order to maintain NPSH margin for theContainment Spray Pumps and allow installation of the new PCI designed sumpstrainers. An analysis EA-MOD-2005-005-003 Rev 1 was performed todetermine the maximum flow under strainer design conditions in EC496 [Ref.29.2] that would provide adequate NPSH to the pumps.The 819 gpm valve throttled flow rate requirement (Ref. Specification M0255 andEA-MOD-2005-004-03 R1, ESS Flow Rates & Pump NPSH during RecirculationMode with CSS Throttling (Case 1AAA D)) is to provide adequate NPSHavailable to the Containment Spray pumps during a design basis event where aLeft Channel Power Failure (Loss of the Emergency Diesel Generator 1-1)occurs resulting in the loss of two of the three Containment Spray Pumps.In this limiting case the amount of flow which passes through the containmentspray header valve CV-3002 valve must be limited to the 819 gpm. However, asthe LOCA event progresses the NPSH available increases due to containmentsump temperature continuing to decrease below saturation to where throttling theContainment Spray flow would, though not credited, no longer be critical.The design specification for the Strainers in EC496 was taken from SpecificationM0802 which was 2.6 foot water head loss due to debris at 1849 gpm flowthrough the strainers. This value reserved some capability for uncertainty andmargins in the pump flows and NPSH calculations.With this value established, a calculation EA-Gothic-04-08 Rev 2 of post LOCAcontainment conditions was done with the un-throttled flow in the injection modetime frame and the throttled flow in the Post RAS time frame. The resultsshowed that the FSAR maximum temperature and pressure conditions weremaintained.EA-MOD-2005-004-03 determined minimum and maximum system flow rates,pump flow rates, and the minimum pump NPSH margins. The bounding caseused for establishing the analytical limiting replacement valve throttling flow of803 gpm (valve min throttling flow is 819 gpm (Ref. DIT CCI03)) in case 1AAAwhere after RAS initiates a Left Channel Failure, the failure of EDG 1-1 occurs.This alignment involves the CSS Pump P-54A supplying the HPSI Pump P-66A, Palisades Draft RAI Responses for May 2010 Public Meeting 134CSS Header A, and the four PCS cold legs following the isolation of CSS HeaderB. This alignment would continue from shortly after RAS initiation until HPSIthrottling or hot leg injection is initiated. This alignment can also occur followinga failure of Sump Suction Valve CV-3030 to open.

References29.1 Modification EC8350 "Replace Containment Spray Isolation Valves perGSI-191 Resolution"29.2 Modification EC496, "Replace Containment Sump Screens Per GSI-191Resolution (Passive Strainer)"

Palisades Draft RAI Responses for May 2010 Public Meeting 135 NRC Request30. The June 30, 2009, supplemental response described components forwhich friction losses were considered. As requested in the content guidefor GL 2004-02 responses, please also describe the methodology and/orreferences used to derive the formulas for calculating flow losses for thesecomponents in the NPSH margin calculation.Entergy Nuclear Operations Response:The analysis to evaluate containment spray pump NPSH post-RAS wasperformed using the Palisades ESS integrated hydraulic model developed withPipe-Flo version 4.11 software. The software allows the user to specify a losscoefficient (K), enter a head loss vs. flow rate curve, or use standard industryformulas to calculate the loss coefficient for a piping system component. Withrespect to the strainer piping components, debris head loss, and clean strainerhead loss, these values were all manually entered as either a function of headloss or a specific loss coefficient.The clean strainer head loss was modeled in the software by first calculating theclean strainer head loss at flow rates of 1405 gpm and 3521 gpm as described insection 6.2.1 of EA-MOD-2005-004-03 Rev. 3. The regression function forcalculating clean strainer head loss, as well as the additional head loss due tocore tube length, was based on PCI Technical Document SFSS-TD-2007-002submitted to the NRC March 25, 2009. The clean strainer and core tube headlosses were combined to develop a total head loss data point for each strainerassembly at each of the two flow rates.Using this approach, three data points were developed for each of the strainerassemblies to create four functions of head loss vs. flow rate which were enteredinto the hydraulic model (the first data point being 0 head loss at 0 flow rate) assystem components. Pipe-Flo then uses this data to calculate an exponentialequation of the form:

dPcomp = CW n dPcomp= pressure drop across the component W = Flow RateC and n = Values determined using geometric regressionThe approach to modeling head loss for the piping and components downstreamof the strainer assemblies is described in section 6.2.2 of EA-MOD-2005-004-03Rev. 3. Values for piping bends, orifices, enlargements, and exit losses from thedowncomers to the sump were manually developed and entered into the modelas specified loss coefficients (K) on a given pipeline. Friction loss for the pipingis calculated by the Pipe-Flo software as a function of piping length, material anddiameter, as well as water properties which were entered into the model.

Palisades Draft RAI Responses for May 2010 Public Meeting 136The software uses the Darcy-Weisbach method for calculating piping frictionlosses. From the Pipe-Flo User's Guide:The Darcy-Weisbach method takes into account fluid viscosity and piperoughness, providing valid results for incompressible Newtonian fluids flowing inany round fully charged pipe. This formula can also be extended to compressiblefluids with some restrictions.The Darcy-Weisbach equation is as follows:

dP = f(L/D)v 2/2gdP = pressure drop = fluid densityf = Darcy friction factorL = length of pipeD = pipe diameterv = mean fluid velocityg = gravitational constantOften, the Darcy-Weisbach is expressed in the following way:

dP = Kv 2/2gwhere K = f(L/D)The K in the above equation is the total resistance coefficient for the pipeline.This "total K" is a combination of the K value for the pipe and the K value for thevalves and fittings in the pipeline. Therefore, the pipeline pressure dropcalculated is a combination of the pressure drop due to the pipe and valves.Separate components to model head loss through the debris bed on the strainerwere developed for the model using a function of head loss vs. flow rate, as wasdone for the clean strainer head loss. The head loss across the clean strainers,core tubes, and piping to the containment sump, was calculated for each strainerassembly at the maximum design total flow rate (3591 gpm). These head losseswere subtracted from the maximum design fouled strainer head loss of 2.6 ft thusgenerating values of maximum additional head loss due to strainer fouling vs.flow rate. This data was extrapolated to a higher flow rate to provide a third datapoint (the first being 0 flow at 0 head) and entered into the Pipe-Flo model asdebris components for each strainer module. When modeling scenarios thatconsider fouled strainers, the components were inserted into the model upstreamof the clean strainer components. Further details may be found section 6.2.4 ofEA-MOD-2005-004-03 Rev. 3.

References Palisades Draft RAI Responses for May 2010 Public Meeting 13730.1 EA-MOD-2005-004-03 Rev. 3 dated 4/6/2009, "ESS Flow Rates andNPSH During Recirc Mode with CSS Throttling"30.2 PCI Technical Document SFSS-TD-2007-002, Rev 1 dated December 11,2008, "Sure-Flow Suction Strainer - Suction Flow Control Device(SFCD) Principles abd Clean Stariner Head Loss Design Procedures"30.3 Pipe-Flo User's Guide Palisades Draft RAI Responses for May 2010 Public Meeting 138 NRC Request31. A number of objects were cited as displacing water in the post-LOCAcontainment sump pool, including the reactor vessel insulation,pressurizer heater transformers (both of which objects appeared to bepotentially non-leak tight and to have some hollow internal volume) andthe containment buffer, which would presumably dissolve in thecontainment pool water. Please clarify whether other dissolvable or hollowand non-Ieaktight objects are credited with displacing volume in the post-LOCA containment pool, estimate the total displaced volume credited, andprovide justification for crediting such objects with displacing water.Response:The minimum post-LOCA containment water level is determined in calculationEA-SDW-97-003, Revision 3. EA-SDW-97-003 utilizes an equation that relatescontainment water volume (V [ft 3]) to containment water level (h [ft]), referred tobelow as V(h). EA-SDW-97-003 determines the level corresponding to a givenwater volume by iteratively solving V(h), which is tantamount to finding theinverse of the equation.The V(h) equation is developed from a similar equation derived in calculation EA-C-PAL-94-0016A-01, Revision 1. The equation in EA-C-PAL-94-0016A-01 isused to determine the maximum containment water level. Therefore, adjustmentsto this equation are made to develop an appropriate equation for determining theminimum containment water level.The V(h)equation in EA-C-PAL-94-0016A-01 (the maximum containment waterlevel calculation) treats the following objects as displacing water: Reactor Vessel Reactor Vessel Insulation Bioshield Clean Waste Receiver Tanks Concrete Structures Pressurizer Heater Transformers Miscellaneous Equipment Containment BufferThe displacement volume calculations in EA-C-PAL-94-0016A-01 are assessedfor conservatism and biases that are not applicable to a minimum water levelcalculation as follows.Reactor VesselNo biases towards maximum containment water level are noted in thedevelopment of the reactor vessel displacement volume calculation. Therefore, Palisades Draft RAI Responses for May 2010 Public Meeting 139no adjustments are made to the V(h) equation for this displacement volume inthe minimum water level calculation.Reactor Vessel (Cavity) InsulationNo significant biases towards maximum containment water level are noted in thedevelopment of the reactor vessel insulation displacement volume calculation.Note that insulation referred to in the calculation as reactor vessel insulation isactually reactor vessel cavity insulation. Reactor vessel insulation located on thereactor vessel itself is not included. Reactor vessel cavity insulation (present onthe reactor cavity wall and floor) is not assumed to be water tight. Reactor vesselcavity insulation is assumed to displace 1/3 of the equivalent water volume forthe displacement volume calculation. Prior to the reduction described below, thetotal volume ranges from 102 ft 3 to 117 ft 3, for the range of calculated minimumwater levels (7.563*h + 84.474; h = 2.34 ft, 4.26 ft) out of total water volumes of 18,772 ft 3 and 34,385 ft 3, respectively. The range of minimum water level heightsand volumes correspond to a 4-inch small break LOCA with left channel failure,SIRWT level at the technical specification minimum, adjusted to 212°F sumptemperature; and a 42-inch hot leg large break LOCA with left channel failure,SIRWT level at the Admin limit, adjusted to 212°F sump temperature,respectively. For conservatism, the V(h) equation was adjusted by reducing thedisplacement volume for reactor vessel cavity insulation by an additional 25% inthe minimum water level calculation.BioshieldNo significant biases towards maximum containment water level are noted in thedevelopment of the bioshield displacement volume calculation. The two 1" drainlines from the reactor cavity to the containment sump are considered insignificantand are not removed from the displacement volume calculation. The total volumeof these lines is less than 0.02 ft 3 (2*1.5'**(0.5/12)^2, M-74, Sheet 2); however,the lines are filled with corium plugs and the actual non-displacement volume isconsiderably less. Therefore, no adjustments are made to the V(h) equation forthis displacement volume in the minimum water level calculation.Clean Waste Receiver TanksThe bottom hemisphere of the clean waste receiver tanks (596' elevation) isabove the calculated minimum flood levels. The clean waste receiver tanks donot displace any water for the minimum water level calculation. Therefore, theV(h) equation was adjusted by removing the displacement of the clean wastereceiver tanks in the minimum water level calculation.Concrete StructuresNo biases towards maximum containment water level are noted in thedevelopment of the concrete structures displacement volume calculation.Therefore, no adjustments are made to the V(h) equation for this displacementvolume in the minimum water level calculation.

Palisades Draft RAI Responses for May 2010 Public Meeting 140Pressurizer Heater TransformersNo significant net biases towards maximum containment water level are noted inthe development of the pressurizer heater transformers displacement volumecalculation. Transformer cabinets are assumed to displace 100% of theequivalent water and the switchgear cabinets are assumed to displace 25% ofthe equivalent water volume. The total volume of these components ranges from 146 ft 3 to 266 ft 3, for the range of calculated minimum water levels (62.326*h, h =2.34 ft, 4.26 ft) out of total water volumes of 18,772 ft 3 and 34,385 ft 3 ,respectively. The assumed displaced volume amounts to less than 0.8% of thewater volume. In addition, the switchgear are assumed to be 75% free space.Therefore, no adjustments are made to the V(h) equation for this displacementvolume in the minimum water level calculation.Miscellaneous EquipmentThe development of the miscellaneous equipment displacement volumecalculation is biased towards maximum containment water level. An additional15% was added to the displacement volume equation to account formiscellaneous small equipment that may not have been identified explicitly in thewalkdown. Prior to the reduction described below, the total volume of theadditional miscellaneous small equipment ranges from 260 ft 3 to 265 ft 3, for therange of calculated minimum water levels (16.910*h + 1692.128 = 2.34 ft, 4.26 ft)out of total water volumes of 18,772 ft 3 and 34,385 ft 3, respectively. Therefore,the V(h) equation was adjusted by reducing the displacement volume foradditional miscellaneous small equipment by 25% in the minimum water levelcalculation.Containment BufferNo significant biases towards maximum containment water level are noted in thedevelopment of the containment buffer displacement volume calculation. Thetotal displacement volume assumed for the containment buffer is 200 ft

3. The"displacement" of water by the dissolved buffer depends on the partial molarvolumes of the solvent (water) and the solute (buffer) in the solution and on thenumber of moles of each in the solution. Partial molar volumes are difficult topredict and may be larger or smaller than the pure molar volumes. That is, thevolume of the solution may be different than the sum of the volumes of thesolvent and solute, depending on the specific solvent and solute underconsideration.However, the mass of the solution is preserved such that the important quantityfor NPSH calculations (static head) of the solution increases in proportion to thefractional mass increase. The buffer "displacement' calculation resulted in 200 ft 3of displacement volume, resulting in an absolute increase in static head (waterlevel) of 0.3", which represents a fractional static head increased of about 0.1%for the range of calculated minimum water levels (2.34 ft to 4.26 ft, with pumpsuction at the 573' elevation). The mass of buffer in solution is required to begreater than 8,186 lbm of sodium tetraborate decahydrate equivalent (LCO Palisades Draft RAI Responses for May 2010 Public Meeting 1413.5.5), which represents a fractional mass increase and corresponding fractionalstatic head increase of 0.4% to 0.7%. The additional water level predicted istherefore not significant and is compensated for by the minimum required buffermass. Therefore, no adjustments are made to the V(h) equation for thisdisplacement volume in the minimum water level calculation.In addition, the following items were noted with respect to the development of theV(h) equation for the maximum water level calculation. An approximation is usedto determine the free space in the tapered containment wall. The calculation ofthe additional volume of the sloped floor region of containment is biased low.Several short 4" diameter drains are ignored.Tapered Wall Volume ApproximationThe containment cylinder wall tapers to a smaller diameter as the 590' elevationis approached. No biases towards maximum containment water level are notedin the development of the tapered wall volume calculation. The approximation forthe tapered wall volume is slightly conservative for the minimum water levelcalculation. Therefore, no adjustments are made to the V(h) equation for thisvolume.Sloped Floor VolumeThere is a bias towards maximum containment water level noted in thedevelopment of the sloped floor volume calculation. The development of thesloped floor volume contains conservatisms that lead to a smaller sloped floorvolume. The total sloped floor volume calculated for the maximum water levelcalculation is 391.107 ft

3. Therefore, the V(h) equation was adjusted byincreasing the volume for the sloped floor by 25% in the minimum water levelcalculation.590' Floor DrainsSeven 4" diameter floor drains were conservatively ignored in the development ofthe sump volume calculation. The total drain volume is estimated to be less than 10 ft 3. Therefore, no adjustments are made to the V(h) equation for this volume.Summary:The credit taken for displacement volumes has been addressed. If no credit istaken for any displacement volume for the reactor vessel cavity insulation,pressurizer heater transformers, containment buffer, or miscellaneous additionalequipment, the minimum water level is reduced by 0.08 ft to 0.09 ft for the rangeof minimum water levels calculated (the range 2.34 ft to 4.26 ft changes to 2.26 ft- 4.17 ft).While justification for crediting displacement volumes has been discussed above,the cumulative impact of not crediting these volumes is on the order of a 1"reduction in minimum containment water level. Given the limited margin, theminimum water level calculation will be revised to provide a more rigorous Palisades Draft RAI Responses for May 2010 Public Meeting 142accounting of both free space and displacement volume and biased consistentlyfor minimum water level determinations.

References31.1 EA-C-PAL-94-0016A-01, Revision 1, Containment Flood Analysis,December 1994.31.2 EA-SDW-97-003, Revision 3, Minimum Post-LOCA Containment WaterLevel Determination, February 2009.31.3 M-74, Sheet 2, Revision 0, Reactor Cavity Drain Plug, February 1997.

Palisades Draft RAI Responses for May 2010 Public Meeting 143 NRC Request32. Please clarify the assumption concerning the pump curves applied in thedesign analysis including a 7 percent allowance for flow degradation in thecontainment spray pumps, and an 8 percent allowance for flowdegradation in the high-pressure safety injection pumps as it relatesdetermination of conservative pump flow rates for the I\IPSH analysis.While not fully understanding the licensee's methodology, the NRC staffgenerally believes that assuming pump degradation when determiningminimum NPSH margins for pumps could be non-conservative, since therequired NPSH would seemingly be lower for a degraded pump withreduced flow. Please clarify whether the pumps are assumed to operateabove or below their certified head-flow performance curves indetermining NPSH margin and provide a basis for considering thecalculated NPSH margins to be limiting in light of this flow assumption.Entergy Nuclear Operations Response:The HPSI and containment spray degraded pump curves were developed byapplying a uniform percentage flow reduction at all pump heads. As pumpinternal leakage is proportional to the square root of the pump head, thisapproach is conservative because no credit was taken for the smaller flowreduction that would occur at lower pumping head. The percentage of flowreduction was determined graphically by uniformly reducing the flow at all pumpheads until the curve intersected with lower limit for inservice testing acceptancecriteria. For the containment spray pumps this value was 7% and for the HPSIpumps 8%. The degraded pump curves also accounted for the minimum alloweddiesel generator frequency by application of pump affinity laws to determine theflow rate and pump head at reduced pump rotation speed. Finally, instrumentuncertainty was applied to further reduce pump flow rate and discharge head ateach data point.To develop the strong pump curves, pump affinity laws were used to calculatethe pump head and flow rate at each data point due to the increased rotationspeed at the maximum diesel generator frequency, then instrument uncertaintywas applied in a positive fashion to increase further increase flow rate and pumphead. Using this approach, the degraded pumps operate below their nominalperformance and the strong pumps operate above their nominal performance curve.The methodology for applying strong or degraded pumps in the hydraulic modelis briefly described in section 3.g of the June 30, 2009 GL 2004-02 supplementalresponse on pages 138 - 140. However, it should be clarified that whendegraded pump curves were used for purposes of calculating containment spraypump NPSHa, the degraded curves were applied on the opposite train of thepump being evaluated (e.g. a degraded HPSI and containment spray pump Palisades Draft RAI Responses for May 2010 Public Meeting 144tandem on the first diesel train and a strong pump tandem on the second train).As the pumps are operating in parallel, the stronger pump train overpowers thedegraded pump train increasing its flow rate and NPSHr. Therefore, theapproach used was conservative.

Palisades Draft RAI Responses for May 2010 Public Meeting 145 NRC Request 33.In RAI 18 from the NRC's letter dated December 24, 2008, the NRC staffasked about the typical amounts of algae and/or slime that are removedfrom the sump and why this amount of biological material does not need tobe considered as an additional debris source after a postulated LOCA.The licensee provided the requested information which identified anamount of oil sludge, estimated at 27.5 gallons, which is contained in acavity which connects, by means of a pipe, to the sump area downstreamof the strainer. The licensee stated that due to the large amount (250,000gallons) of what it considered "hot soapy sump water," the 27.5 gallons ofeither oily emulsion or algae created biological material would bedissolved. In addition, the licensee stated that the extreme agitation as ittransits through the sump area will ensure good mixing takes place andwill enhance the process of dissolution of solubles or suspension of smallparticles. The NRC staff believes that it is unlikely that the oil-based slimeidentified by the licensee will dilute in the sump pool as stated in theresponse, and that there is a potential for this material to be transported tothe reactor core and spray nozzles. The bases for the NRC staff's concernis that the sludge has been proven difficult to remove, as indicated in page168 of the June 30, 2009 RAI response, which states, "...the high viscosityof the material made a bubble form in the small screen squares and itresisted removal by a stiff wire brush that rode over the high points on thescreens. Adding soapy cleaning solution did nothing to help thisphenomenon." The physical and chemical properties of the oil sludge maynot allow it to be easily diluted. Please provide further information on yourplans to address the oil sludge during future plant operation. If it were totransport downstream of the strainer, has the amount of sludge beenevaluated as part of the downstream effects analysis? Were any otherapproaches (e.g. hard piping the sump where the sludge is located fromthe strainer) considered?Response:In RAI 18 of December 24, 2008, the NRC staff asked about the typical amountsof algae and/or slime that are removed from the sump. Details of the amount ofmaterial removed in days gone by during cleaning are usually not kept in anyformally retrievable manner. It was known that some informal records wereavailable from past System Engineers notes and also from Radiation WorkPermit records but none of these were subject to independent validation andsome of the descriptions were authored by cleaning crews who had very littlesystem knowledge and were phrased in casual worker jargon.To answer the request for information what records existed were used regardlessof the pedigree. The original draft of the response to RAI 18 of 2/27/08 Palisades Draft RAI Responses for May 2010 Public Meeting 146attempted to retain as much as possible of the original wording from severalsources with several authors.The present RAI appears to be based on a miss-understanding of the answer tothe original RAI 18 from the NRC's letter dated December 24, 2008.In the statement "The licensee provided the requested information whichidentified an amount of oil sludge, estimated at 27.5 gallons, which is containedin a cavity which connects, by means of a pipe, to the sump area downstream ofthe strainer" the 27.5 gallons was in a waste drum which was a part of the sumpcleaning equipment that contained the results of vacuuming out the fluid that wasleft in the bottom of the sump. It is not a part of plant equipment and it isremoved from containment after the cleaning work order is completed and beforeplant startup. The drum contents gravity separate in time and the 27.5 gallons isthe result of the record having stated that the disposal drum contained about halfwater and half "slime". This item was quoted in the RAI response to fulfill therequirement of "typical amounts" requested by the RAI. It was one of the fewrecords that were in any way quantified. Most estimated it by talking aboutfractional inches of water slime mix observed in the 22 foot diameter sump "tank".Because the bottom of the "tank" is uneven, the depth depends upon where themeasurement is made. Furthermore the typical record made no attempt to statethe how much of the residual fluid was water.In the statement "The licensee stated that due to the large amount (250,000gallons) of what it considered "hot soapy sump water," the 27.5 gallons of eitheroily emulsion or algae created biological material would be dissolved. In addition,the licensee stated that the extreme agitation as it transits through the sump areawill ensure good mixing takes place and will enhance the process of dissolutionof solubles or suspension of small particles." there are several miss-statementsresulting from removal from the original context. First it must be recognized thatthe area under discussion is a part of the plant that is before the ECCS pumpsand is behind the current sump strainers. Thus the water from the "sump tank"must be taken into the pump suction pipes which are about 4 inches above thesump floor, transit down approximately 75 feet of 24" piping, enter the pump andbe violently agitated by its impeller, exit the pump and transport through a heatexchanger, be returned to containment via either the HPSI injection valves whichseverely throttle flow or via containment spray which involves a throttled (multi-passage Drag Valve Style) valve and a spray nozzle. From there it must traversethe PCS and come out the break or fall as spray on to the floor. The runoff fromboth of these processes must transit containment and get into the 590 elevation"basement" floor and move in the 3 foot deep water containing Sodium Tetra-borate (aka the soapy component) where the ECCS suction strainers arelocated. This is the first contact that the "slime" could have with the strainers andwould be the first opportunity to participate in the filter formation/blockingprocess. At this point it would be one part in 10,000 parts of water there havingbeen multiple intense mixing steps prior to the contact point. Note that Sodium Palisades Draft RAI Responses for May 2010 Public Meeting 147Tetra-Borate is sold commercially as 20 Mule Team Borax specifically for thepurpose of breaking down the surface tension of oily material embedded inclothing to facilitate dissolution in a washing machine and also to be abactericide.In the statement "The bases for the NRC staff's concern is that the sludge hasbeen proven difficult to remove, as indicated in page 168 of the June 30, 2009RAI response, which states, "...the high viscosity of the material made a bubbleform in the small screen squares and it resisted removal by a stiff wire brush thatrode over the high points on the screens. Adding soapy cleaning solution didnothing to help this phenomenon" again the problem is removal of the statementfrom context. This was a part of a 2 page discussion which was given for thesake of satisfying the "typical" part of the RAI and was a short discussion ofcleaning efforts that had been employed in the past which were not continueddue to poor results. The cleaning under discussion is done by hand under verytrying circumstances.To quote a corrective action document CR-PLP-2007-05055:The sump cleaning and inspections were performed under the followingconditions:1" to 1/4" of mud on sump bottomDark - Poor lightingWearing face shields and multiple layers of protective clothing and rubber gloves3.5' tall ceilingWarm and humidHigh contamination potentialThe sump screen is welded in place so that only the outside can be accessed.To further complicate the cleaning the screens are very close to the open mouthof the 24" recirculation suction pipes and extraneous fluids and possibly removedmaterial easily enters the pipe and can not be easily removed. For containmentisolation reasons these pipes are encased into a secondary isolation boundaryfrom the containment wall to the Recirculation isolation valve. Thus there are noflush points or piping drain lines to remove material so it must be flushed throughthe ECCS System.Cleaning tools typically used were stiff wire brushes, spray bottles, and moppingrags or absorbent paper materials supplied by Radioactive Material Control(RMC) personnel who are professional decontaminators and who actually do themajority of the cleaning with the help and direction of the System Engineer.In the 2003 to 2006 time frame the methods had progressed to use ofpressurized sprayers (like lawn hand pumped sprayers) using hot water andRMC supplied detergent. The heated spray when delivered to the screens in afine spray would allow the material to run down the screens and be removed by Palisades Draft RAI Responses for May 2010 Public Meeting 148wiping. It was this process which was meant to be referenced as evidence thathot soapy water worked to break up the slime. It was the prior to that cold spray,wire brush, vacuum suction hoses and hand towels that did not work well andcaused severe ALARA concerns during some refueling outages if fuel failureshad occurred.The strainers in the immediately above discussion were cut out of the sump in2007 when the 3,500 sq. ft. of PCI strainers were installed to meet GSI-191criteria. Since the area is behind these strainers the procedure of clean thesump tank every Refueling Outage was retained.The final NRC concern was driven by the miss-interpretation of the previous RAIand is resolved by the above discussion. In essence what is hard for a singleRMC employee crawling around on his hands and knees in radioactive waterwith a small vacuum hose, a spray bottle, cold water and a brush being carefulnot to put foreign material into the ECCS/Shut Down Cooling system is notcomparable to 400 horsepower of pumps moving essentially boiling water andintensely mixing it under pressure up to 1200 psi.Furthermore more recent experience has been that less slime is seen than theabove noted 27.5 gallons. This is believed to be due to better control of PrimaryCoolant Pump oil control, far less Containment Air cooler leakage due to coilreplacement, finer screens installed in the 16" downcomer openings and closefitting 590 Elevation floor drain screens installed for GSI-191.It is suggested that this material is similar to "sludge" seen in Boiling WaterReactor suppression pools. This is handled in PWR by the 200# of Latent Debrismocked up in testing as fine fiber and sand like material. Palisades has enoughmargin between the measured amount and the default 200# used to cover thedried "solid" portion of the slime.An order of magnitude estimate would be 1000 sq ft of 3 mil thick paint or 0.25cuft. This assumes the surface of the sump top, bottom, and sides were paintedwith an opaque layer of dried slime conservatively assumed to be 3 mils thick (infact the pictures indicate that it is semi-transparent). [(2 x 22 x 22 x pi / 4 + 22 x3.5 x pi) x 0.003 /12 = 0.2505 cu ft]To consider it liquid slime, in Palisades testing the amount of extra chemicalprecipitate material was also more than adequate to cover the material in thesump. Chemical precipitate used in testing is slime like material and could beconsidered a surrogate approximately representative of the variable andunknown constituents of the extraneous dump material. Sufficient materialbeyond the design calculated amount was used to cover the 27 gallonapproximation to sump condition (27/23 ~1 gallon when scaled to the test).

Palisades Draft RAI Responses for May 2010 Public Meeting 149CR-PLP-2007-5055 Photo of 1 of 2 Sump Exits with Old Screen RemovedWhite material is dried Boric Acid